Elsevier

Organic Electronics

Volume 78, March 2020, 105598
Organic Electronics

Letter
Tailored interfacial crystal facets for efficient CH3NH3PbI3 perovskite solar cells

https://doi.org/10.1016/j.orgel.2019.105598Get rights and content

Highlights

  • A SSG technology is demonstrated to modulate interfacial grain facets in CH3NH3PbI3 PSC.

  • Interfacial grain facets of CH3NH3PbI3 film are transformed from (100)/(112) to (110)/(002).

  • Tailored interfacial grain facets lead to improved average PCE from 16.51 ± 0.64% to 18.40 ± 0.67%.

Abstract

Interface engineering is generally requisite for highly efficient perovskite solar cells (PSCs). However, the current interface engineering methods inevitably introduce extra modifier layers into PSCs, which not only complex the configurations and fabrication procedures, but also increase the production cost of PSCs. Herein, we propose an interface engineering strategy for PSCs by controlling the nature of Lead halide perovskite films, and specifically their interfacial grain facets. In detail, a solution-mediated secondary growth (SSG) technology is demonstrated to tailor interfacial grain facets in CH3NH3PbI3 PSC. The precise tailoring ability of interfacial grain facets is achieved by controlling SSG temperature. When it is optimized to 60 °C, interfacial grains of CH3NH3PbI3 film can be fully transform from dodecahedral-shaped ones enclosed by (100) and (112) facets to the cubic-shaped ones enclosed by (110) and (002) facets, while maintaining the film's crystalline phase and composition. More importantly, such transitions are accompanied by significantly improved average PCE from 16.51 ± 0.64% to 18.40 ± 0.67% for the optimized CH3NH3PbI3 PSCs, benefiting from the greatly suppressed recombination and enhanced extraction of carriers.

Graphical abstract

Transition of interfacial CH3NH3PbI3 grain facets from (100)/(112) to (110)/(002) triggers much improved interfacial carrier dynamics and hence cell efficiency.

Image 1
  1. Download : Download high-res image (212KB)
  2. Download : Download full-size image

Introduction

Lead halide perovskite materials such as CH3NH3PbI3 are endowed with a variety of attractive photo-physical properties coupled with low-cost, low-temperature, and high-throughput solution processibility of films for optoelectronic applications [[1], [2], [3]], in particular of perovskite solar cells (PSCs) [[4], [5], [6], [7], [8], [9], [10], [11]]. Over the past decade, PSCs have achieved a rapid improvement in power conversion efficiency (PCE) from 3.8% to the certified value of 25.2% [4,12], which has been comparable to CIGS and Si solar cells. In general, interface properties determine to a large extent the PCE and stability of a PSC [13,14], as well as its unique current–voltage hysteresis (J-V) phenomenon [15,16]. Thereby, extensive research efforts have been devoted to interface engineering of PSCs for passivating interfacial defects and promoting band alignment, both of which could contribute to favorable interfacial carrier dynamics and thus improved performance of PSCs [[17], [18], [19], [20], [21], [22]]. Yet, the current interface engineering methods surely introduce extra modifier layers into PSCs, which not only complex the configurations and fabrication procedures, but also increase the production cost of PSCs.

More recently, several independent works reveal that photoluminescence (PL) intensities and carrier lifetimes vary remarkably among different grain facets of polycrystalline perovskite film [[23], [24], [25], [26], [27], [28], [29], [30], [31]]. Such spatial heterogeneity is later observed in resulting PSC in terms of open-circuit voltage (Voc) and short-circuit current density (Jsc), because of facet-dependent fluctuations in each individual crystal grains [23,24]. This unique phenomenon is finally correlated with anisotropic distributed trap densities on different grain facets of polycrystalline perovskite film [23,24]. On the other hand, the studies focused on perovskite single crystals similarly figure out that their trap densities, ion migration behavior, and photoelectric properties are facet-dependent [[32], [33], [34], [35], [36], [37], [38]]. For example, W. C. H. Choy et al. [39] recently demonstrated that the CH3NH3PbI3 single crystals with exposed {110} and {001} facets have better responsivity and faster response speed in photodetection than the ones with exposed {100} and {112} facets. These results make clear that the nature of lead halide perovskite materials, specifically their trap densities and ion migration behavior, along with photoelectric features are influenced strongly by grain facets. In fact, having facet-dependent properties is a general phenomenon for perovskite-type oxides, especially when they are served as catalysts [33,40]. Thus, it seems quite possible to modulate interfacial properties of a PSC by controlling interfacial grain facets of perovskite film. Yet, a significant challenge hindering the popularity of facet-related interface engineering in PSCs lies in the effective preparation of polycrystalline perovskite films with tunable and perfect interfacial grain facets.

Herein, we propose a strategy to modulate interfacial grain facets in CH3NH3PbI3 PSC via a facile solution-mediated secondary growth (SSG) technique. The precise tailoring ability of interfacial grain facets is achieved by controlling SSG temperature. When it is optimized to 60 °C, interfacial grains of CH3NH3PbI3 film can be transform from dodecahedral-shaped ones enclosed by (100) and (112) facets to the cubic-shaped ones enclosed by (110) and (002) facets, while maintaining the film's crystalline phase and composition. More importantly, such transitions are accompanied by improved average PCE from 16.51 ± 0.64% to 18.40 ± 0.67% for the CH3NH3PbI3 PSCs.

The proposed SSG technique is mainly consisted of two steps: the first is the preparation of CH3NH3PbI3 film via a solvent engineering assisted one-step spin-coating method reported by Park et al. [41]; and the second is the simple dip of obtained CH3NH3PbI3 film into 10 mg mL−1 CH3NH3I isopropanol solution under controlled temperature of 40 or 60 °C for 1 h for secondary growth. Detailed preparation procedures are described in experimental section of Supporting Information. Hereinafter, the CH3NH3PbI3 films along with the PSCs formed by SSG technique under 40 and 60 °C are labeled as “SSG@40 °C″ and “SSG@60 °C″ for simplicity, while the CH3NH3PbI3 film and PSC prepared without SSG is marked as “Pristine”. Fig. 1(a) shows surficial and cross-sectional scanning electron microscope (SEM) images of pristine CH3NH3PbI3 film. One can see that the film is composed of closely-packed crystal grains with dodecahedral shape, which is frequently observed in previous works [24,32,41]. The investigations on CH3NH3PbI3 single crystals indicate that the dodecahedral-shaped crystals are generally enclosed by (100) and (112) ending facets [33,35,37]. At such, we infer that the pristine CH3NH3PbI3 film exhibits (100) and (112) exposed facets. After SSG at 40 °C, the film has also full surface coverage, as indicated by Fig. 1(b). However, the shape of its crystal grains evolves from dodecahedral to truncated cubic. Since cubic-shaped CH3NH3PbI3 single crystals usually possess (110) and (002) exposed facets [33,35,37], we deduce that the (110) and (002) facets are dominant in the SSG@40 °C CH3NH3PbI3 film. When SSG temperature is elevated to 60 °C, the resulting CH3NH3PbI3 film has still desirable surface coverage in spite of significant reconstruction of crystal grains, as shown in Fig. 1(c). Meanwhile, the original dodecahedral-shaped crystal grains disappear completely in it, instead, they are replaced by perfect cubic crystal grains enclosed by (110) and (002) facets. That is to say, the exposed facets of SSG@60 °C CH3NH3PbI3 film are fully converted into (110) and (002) from the original (100) and (112). At this point, one can draw a conclusion that SSG method can effectively modulate the exposed facets of CH3NH3PbI3 polycrystalline films and even tailor the ratio of different exposed facets. As far as we know, this is the first report of successful regulation of exposed facets of polycrystalline perovskite film in a controlled manner [[23], [24], [25], [26], [27], [28], [29], [30], [31]].

Fig. 1(d) presents X-ray diffraction (XRD) patterns of pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films. The entire XRD peaks except the ones from substrates can be indexed to perovskite CH3NH3PbI3 with tetragonal phase [5,6,26]. Such results suggest that SSG process does not alter the crystal phase and component of original CH3NH3PbI3 films, even though it triggers significant changes of their exposed grain facets. In addition, the XRD peaks of (110) and (200) planes for SSG@40 °C and SSG@60 °C CH3NH3PbI3 films are slightly higher than those of the pristine one, which are indicative of fewer structural defects in the former [42,43]. Fig. 1(e) shows UV–vis absorption spectra of the investigated CH3NH3PbI3 films. All the films exhibit similar absorption onset at ~785 nm, in consistent with the optical bandgap of tetragonal-phase perovskite CH3NH3PbI3 polycrystalline materials [1,5,6]. By contrast, there are some increases in absorption intensities for CH3NH3PbI3 films after SSG process, in particular for SSG@60 °C film. Such phenomenon could be assigned to their uneven surficial microstructures with strong light scattering effect [44]. Fig. 1(f) gives steady-state PL spectra of the samples. Obvious PL peaks at ~778 nm can be detected from all the CH3NH3PbI3 films, which are generated by band-band recombination of photo-generated carriers in them [30,31]. Meanwhile, one can see that the PL intensity increases in turn for pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films. A higher PL intensity generally means a weaker nonradiative recombination of photo-generated carriers, as a result of fewer defects in CH3NH3PbI3 film [30,31,42]. Thus, we infer that defect density decreases in sequence for them with the increased SSG temperature, which is consistent with the XRD results. Overall, the above characterizations convey that the SSG CH3NH3PbI3 films possess similar crystal phase and component with the pristine one, while they exhibit much reduced defects with elevated SSG temperature. This desired feature can be attributed to the increased portion of (110) and (002) exposed interfacial facets of SSG CH3NH3PbI3 films, since ever-increasing studies focused on CH3NH3PbI3 single crystals have revealed that the one with exposed (110) and (002) facets has reduced trap densities [33,37,39].

To investigate the influence of SSG on interfacial dynamics of carriers in.

Resulting PSCs, we record time-resolved photoluminescence (TRPL) spectra of obtained CH3NH3PbI3 films deposited on insulating glass substrates and sandwiched within spiro-MeOTAD hole transporting layer (HTL) and compact TiO2 (c-TiO2) electron transporting layer (ETL), respectively. Fig. 2(a) shows TRPL results of CH3NH3PbI3 films deposited on insulating glass substrates. By using an exponential decay function of time, the PL lifetimes are fitted to 44.8, 53.5, and 128.6 ns for pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films, respectively. As depicted in insert of Fig. 2(a), the detected PL signals mainly come from radiative recombination of carriers in this case [3,21,32]. Therefore, the longer PL lifetimes for SSG films are indicative of their smaller defects densities, agreeing with the XRD and steady-state PL results above. When are sandwiched within HTL and ETL, the PL decay is accelerated obviously for all the CH3NH3PbI3 films, as presented in Fig. 2(b). This is due to the extraction of carriers by HTL and ETL [21,30,32], as illustrated in insert of Fig. 2(b). Accordingly, the PL lifetimes are calculated to 2.3, 1.3, and 0.4 ns for pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films, respectively. Thus, one can derive that the SSG CH3NH3PbI3 films, in particular of SSG@60 °C, can enable more effective extraction of carriers from them to the carrier transporting layers.

The effects of exposed interfacial facets of CH3NH3PbI3 film on PSC's performance are examined based on the common planar-heterojunction architecture of FTO/c-TiO2/CH3NH3PbI3/spiro-MeOTAD/Ag, as illustrated in Fig. 3(a). Statistical PCEs of 20-independent PSCs based on as-synthesized CH3NH3PbI3 films are provided in Fig. 3(b), which are measured under simulated AM 1.5G illumination (100 mW cm−2). From them, the average PCEs are calculated to be 16.51 ± 0.64%, 17.18 ± 0.77%, and 18.40 ± 0.67% for PSCs fabricated with pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films, respectively. These results tell that the SSG CH3NH3PbI3 films such as SSG@60 °C can promote much enhanced performance of ultimate PSCs.

Fig. 3(c) gives reverse-scan light current density versus voltage (J-V) curves of champion PSCs based on pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films, respectively. From them, the detailed photovoltaic parameters are extracted and summarized in Table S1. The PSC with pristine CH3NH3PbI3 film yields a Jsc of 23.56 mA cm−2, a Voc of 1.103 V, and a fill factor (FF) of 0.68, leading to an inferior PCE of 17.67%. In terms of PSCs with SSG@40 °C and SSG@60 °C CH3NH3PbI3 films, their Jsc values are similar with that of pristine one, while there are obvious increases in Voc and FF values that render much enhanced PCEs of them. In particular, the PSC with SSG@60 °C exhibits the outstanding PCE of 19.66% with Jsc of 23.53 mA cm−2, Voc of 1.129 V, and FF of 0.74, which is ranking in the highest PCE range of pure CH3NH3PbI3-based PSCs reported previously [45].

Meanwhile, steady-state outputs of the champion PSCs under their respective maximum power points are recorded and given in Fig. S1. From them, the steady-state PCEs are estimated to 16.57%, 17.67%, and 19.17% for the champion PSCs with pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films, respectively. These results indicate that the J-V hysteresis happens in all the PSCs [16,19,42], which can be also observed from Fig. S2. Even so, the better PCEs and weaker J-V hysteresis for PSCs with SSG@40 °C and particularly SSG@60 °C CH3NH3PbI3 films are identified as before. Meanwhile, external quantum efficiency (EQE) spectra of the champion PSCs are collected, as given in Fig. 3(d). The slightly higher EQEs in the region from 650 to 760 nm can be identified for PSCs with SSG CH3NH3PbI3 films in particular of SSG@60 °C, further verifying their excellent photoelectric conversion features. Fig. S3 gives the integrated Jsc values form the EQE spectra. All of the values are slightly lower than those of measured from J-V curves. Such discrepancy is mainly caused by the absence of white light bias and/or the spectral mismatch between the EQE source and the solar simulator [42,43]. Together, the above investigations well validate the promising feasibility of SSG method in realizing high-efficiency PSCs by effectively modulating interfacial grains facets of CH3NH3PbI3 films.

Finally, we compare the interfacial carrier dynamics of PSCs with pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films by means of transient photovoltage (TPV), transient photocurrent (TPC), and electrochemical impedance spectroscopy (EIS) technologies, respectively. The results in Fig. 4(a–b) disclose that the PSCs with SSG CH3NH3PbI3 films yield the much faster TPC and slower TPV features, in contrast with the one based on pristine film. And, these phenomena are more obvious for the PSC based on SSG@60 °C film. That is to say, carrier extraction is more effective and carrier recombination is weaker in PSCs based on SSG CH3NH3PbI3 films, such as SSG@60 °C [24,43]. Fig. 4(c) gives Nyquist plots extracted from EIS spectra measured under bias voltage of 0.9 V and dark condition. Clear semicircles can be identified in intermediate frequency region for all the PSCs, which are related to interfacial carrier recombination process featured by recombination resistance (Rrec) [19,24,43]. One can see clearly that Rrec values increase in turns for PSCs with pristine, SSG@40 °C, and SSG@60 °C CH3NH3PbI3 films. The increased Rrec values suggest that the suppressed carrier recombination, which is well in accordance the TPV results. Therefore, as summarized in Fig. 4(d), one can realize that the improved interfacial carrier dynamics, as results of suppressed recombination and enhanced extraction, give rise to the improved performance of PSCs with SSG temperature, which are originally correlated with the increased portion of (110) and (002) exposed interfacial facets of CH3NH3PbI3 films correspondingly.

In summary, the SSG method is highly effective to tailor interfacial grains facets of CH3NH3PbI3 PSC. By adjusting the SSG temperature, interfacial grains of CH3NH3PbI3 film can be gradually transformed from dodecahedral-shaped ones enclosed by (100) and (112) facets to the cubic-shaped ones enclosed by (110) and (002) facets. Meanwhile, the film exhibits much reduced defects with increased SSG temperature, while it still can maintain its original crystal phase and component. Such an innovative interfacial facets engineering induces the improved average PCE from 16.51 ± 0.64% to 18.40 ± 0.67% for the optimized PSCs due to significantly suppressed interfacial carrier recombination and enhanced interfacial carrier extraction.

Section snippets

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

All the authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (61804113 and 61874083), Initiative Postdocs Supporting Program (BX20190261), National Natural Science Foundation of Shaanxi Province (2018ZDCXL-GY-08-02-02 and 2017JM6049).

References (45)

  • J. Ding et al.

    Mater. Chem. Phys.

    (2018)
  • T.C. Sum et al.

    Energy Environ. Sci.

    (2014)
  • Y. Zhao et al.

    Chem. Soc. Rev.

    (2016)
  • K. Chen et al.

    Chem. Soc. Rev.

    (2018)
  • A. Kojima et al.

    J. Am. Chem. Soc.

    (2009)
  • H.-S. Kim et al.

    Chem. Commun. (J. Chem. Soc. Sect. D)

    (2019)
  • M. Saliba et al.

    Angew. Chem. Int. Ed.

    (2018)
  • M.M. Tavakoli et al.

    Adv. Energy Mater.

    (2019)
  • H. Zhang et al.

    Adv. Mater.

    (2019)
  • H. Zhang et al.

    J. Mater. Chem.

    (2017)
  • H. Zhang et al.

    Adv. Mater.

    (2017)
  • F. Li et al.

    J. Mater. Chem.

    (2017)
  • Cell efficiency chart

  • P. Schulz

    ACS Energy Lett

    (2018)
  • H. Wang et al.

    Energy Environ. Sci.

    (2019)
  • S.A. Weber et al.

    Energy Environ. Sci.

    (2018)
  • P. Liu et al.

    Adv. Energy Mater.

    (2019)
  • A.N. Cho et al.

    ChemSusChem

    (2017)
  • A. Rajagopal et al.

    Adv. Mater.

    (2018)
  • M.M. Tavakoli et al.

    Adv. Energy Mater.

    (2018)
  • F. Meng et al.

    J. Mater. Chem.

    (2019)
  • M. Stolterfoht et al.

    Energy Environ. Sci.

    (2019)
  • Cited by (5)

    • Carbon-based, all-inorganic, lead-free Ag<inf>2</inf>BiI<inf>5</inf> rudorffite solar cells with high photovoltages

      2021, Solid-State Electronics
      Citation Excerpt :

      Fig. 4(c-d) present TPC/TPV curves for as-fabricated solar cells. Faster TPC and slower TPV decays are identified for the one prepared with air blowing modification, further consolidating improved carrier extraction and suppressed carrier recombination in it [17,27,28]. Overall, one can make sense that Ag2BiI5 solar cells could be attained based on a simplified carbon-based, all-inorganic configuration.

    • Effect of PbI<inf>2</inf> surface treatment with DMSO vapor on the properties and photodetector characteristics of CH<inf>3</inf>NH<inf>3</inf>PbI<inf>3–x</inf>Cl<inf>x</inf> perovskite films synthesized by a PbS-to-PbI<inf>2</inf>-to-perovskite sequence

      2020, Organic Electronics
      Citation Excerpt :

      In this way, the maximum of emission was identified at 1.61 eV for all the perovskite samples, as indicated in the same figure. This value agrees with the photoluminescence-determined energy bandgap reported by Yamada et al. [22] and with that previously reported by us for the perovskite synthesized by the solution/vapor/vapor three-step approach [8]; this value is similar to those obtained by other characterization techniques [23–25]. Fig. 6 shows the SEM micrographs of the surface of the untreated and the four DMSO-treated PbI2 samples, in which it is observed that the untreated sample is composed of PbI2 nanoplates (see the high-resolution SEM image of PbI2 included in our previous paper [8]).

    1

    These authors contributed equally to the work.

    View full text