Intermediate-phase-assisted low-temperature formation of γ-CsPbI3 films for high-efficiency deep-red light-emitting devices

Black phase CsPbI3 is attractive for optoelectronic devices, while usually it has a high formation energy and requires an annealing temperature of above 300 °C. The formation energy can be significantly reduced by adding HI in the precursor. However, the resulting films are not suitable for light-emitting applications due to the high trap densities and low photoluminescence quantum efficiencies, and the low temperature formation mechanism is not well understood yet. Here, we demonstrate a general approach for deposition of γ-CsPbI3 films at 100 °C with high photoluminescence quantum efficiencies by adding organic ammonium cations, and the resulting light-emitting diode exhibits an external quantum efficiency of 10.4% with suppressed efficiency roll-off. We reveal that the low-temperature crystallization process is due to the formation of low-dimensional intermediate states, and followed by interionic exchange. This work provides perspectives to tune phase transition pathway at low temperature for CsPbI3 device applications.

2) The authors mentioned that other large ammonium salts worked as well. How crucial is the identity of this ammonium salt for general device performance?
3) The imidazole has helped in the formation of CsPbI3 at low temperature. Could the authors comment on whether it is helpful for keeping CsPbI3 in its black phase over the long term? How does it compare to CsPbI3 formed at high temperature without the help of imidazole?
Reviewer #3 (Remarks to the Author): The authors has developed a low-temperature fabrication method of γ-CsPbI3, where organic ammonium cations were introduced (through imidazolium iodide (IZI)) to form an intermediate phase to mediate the formation process of the perovskite. Through XPS and ATR-FTIR characterizations, the authors pointed out that the deprotonation of the IZ+ cations by ZnO promotes interionic exchange between IZ+ and Cs+, and therefore facilitates the formation of the γ-CsPbI3. The work also demonstrated a record EQE of 10.4% in 3D CsPbI3-based red PeLEDs. The findings reported in this manuscript are interesting and may have important impact on low-temperature fabrication of thin film all-inorganic perovskites. In general, this is a high-quality work. I would recommend acceptance of this manuscript after the following issues are properly addressed.
1. Figure 2b shows different N 1s peaks of IZI on ZnO and on ITO. It is suggested that ZnO can deprotonate the IZ+ cation of IZI through formation of Zn-O-H bonds. In principle, ITO should also allow for a similar deprotonation process through formation of Sn-O-H bonding. Why does the latter process not occur? Can the authors please clarify what properties are required for the metal oxide to initiate the deprotonation process? 2. Regarding the mechanism illustrated in Figure 2f: if the interionic exchange is enabled by ZnO, which contacts the bottom surface of the perovskite, could the exchange happen efficiently at the top surface of the perovskite? Does the thickness of the perovskite film impose a limit to the effectiveness of this method?
3. It might be helpful if the authors could provide some morphological characterizations to assist understanding of the perovskite formation process. 4. For the stage 2 of chemical reaction, Zn2+ ions are formed. Will these ions induce trap states? 5. Ammonium salts are reported to passivate defects in perovskite films and at interfaces in PeLEDs, leading to promotion of device performance and lifetime. In this regard, could the addition of IZI also lead to improvement in film quality or interfaces and therefore better LED performances? Also, how about stability?

Point-by-Point Response to Referees
Reviewer #1 (Remarks to the Author): Comment #1: J. Wang et al report the fabrication of black phase CsPbI 3 at low annealing temperature of 100˚C with high photoluminescent quantum efficiency (PLQY) for light-emitting diodes (LEDs) application. The use of organic ammonium imidazolium iodide (IZI) precursor and zinc oxide (ZnO) electron transport layer enables to control the phase transition. The IZI-CsPbI 3 film shows PLQY of 38% and negligible crystal degradation even after exposing 36 days in the air. The LED based on IZI-CsPbI 3 film shows a peak external quantum efficiency (EQE) of 10.4% with good reproducibility. The addition of IZI into perovskite played a role for stable crystal structure and LED efficiency. However, the discussions and results regarding the optoelectrical properties such as optical stability, PL decay time, and energy levels (valence/conduction band) are insufficient in this manuscript. In addition, a lot of papers regarding the high efficiency red perovskite LEDs with EQE of over 10% have already been reported. Therefore, this manuscript is not suitable for publication in this stage.

Response:
We thank the reviewer for the constructive comments and useful suggestions, which have helped us to improve the manuscript.
In this work, we have successfully fabricated high-efficiency red perovskite LEDs based on in situ prepared CsPbI 3 films at low temperature. More importantly, we have clearly demonstrated the mechanism behind and shown how to form the optically-active γ-CsPbI 3 though tuning the phase transition pathway and overcoming the energy barrier at low temperature (~ 100 o C). These in situ solution-processed γ-CsPbI 3 film-based LEDs exhibits an EQE of 10.4 % with suppressed efficiency roll-off and color stability under a high current density of 100 mA cm −2 . We believe that our work provides new perspectives to tune phase transition pathway at low temperature for CsPbI 3 based applications.
Regarding the extra measurement required, we have included additional data in the revised manuscript, such as optical stability, PL decay time, and energy levels, which can be found in detail below.

Response:
We agree that the ZnO/PEI electron transport layer has been intoduced to the perovskite LED community from the beginning to reduce the defect states and enhance EQE. As we mentioned in the Response to the Comment #1, the conventional CsPbI 3 films are fabricated from colloidal nanocrystals, which need extra synthesis with much higher temperature and suffering from strong Auger recombinations. The low-temperature in situ solution-processed CsPbI 3 film are still rarely reported at present due to their high formation energy and thermodynamic instability. Therefore, the 38% PLQE from low-temperature in-situ formation pervskite with supressed Auger effect represent an important advance in red pervskite emitters.

Comment #3:
In order to confirm the effect of ZnO in the interionic exchange process, the authors should be added the reference sample without ZnO layer.

Response:
The crystal phase evolution of the reference sample without ZnO layer as shown in Supplementary Fig. 6. The low temperature phase transformation cannot be observed without ZnO layer, where both the 1D phase and CsI remain unchanged even after 10 min annealing. UV-vis absorption spectra measurement result is consistent with the above XRD result (Supplementary Fig.   7). This result suggests that during the formation of γ-CsPbI 3 from the intermediate phase IZPbI

Response:
We compared the crystal phase evolution of IZI-CsPbI 3 film on ZnO layer with and without the underneath ultrathin PEIE layer, we find that there is no difference between on ZnO/PEIE and on ZnO for phase transformation (Fig. 2b, c and Supplementary Fig. 5). The absorption shoulder of γ-CsPbI 3 phase at around 687 nm are observed for IZI-CsPbI 3 film prepared on ZnO/PEIE substrate (Line 121 to 123, Page 5, highlighted, Supplementary Fig. 7 blue line). The XPS result indicates that this deprotonation of IZ + can be also observed in the IZI-CsPbI 3 film with ZnO and ZnO/PEIE substrate, respectively (Line 136 to 138, Page 5, highlighted, Fig.2b and Supplementary Fig. 8). The ATR-FTIR spectroscopy measurement results are consistent with the XPS results (Supplementary Fig. 9). These results suggest that the ultrathin PEIE layer dose not completely prevent the direct contact and interactions between ZnO and perovskite precursor film.  respectively. Note that the underneath ultrathin PEIE does not prevent the deprotonation of the IZ + cation by ZnO. Supplementary Fig. 9a ATR-FTIR spectra of IZI-CsPbI 3 films on ZnO and ZnO/PEIE substrate, respectively. The signature peak of imidazole (IZ) at about 3345 (ν N-H ), 3130 (ν C-H ), 1543 (δ NH ), 1328 (δ CH ), 1263 (ring breathing), 1055 (δ CH ), 841 (ring bend), 757 (γ CH ) and 658 (torsion) cm -1 are clearly observed on ZnO/PEIE substrate, indicating the IZ + cation could also be deprotonated by ZnO/PEIE to form IZ.

Comment #5:
In this manuscript, various experimental data were provided regarding the crystal structure characterization of IZI-CsPbI 3 film. However, there is no information about the PL decay profile, which is very important data to discuss an optical property of IZI-CsPbI 3 film.

Response:
We thank the reviewer for the constructive comments. We have performed the PL lifetime measurements of the perovskite films with different contents of IZI in precursor solutions (IZI /PbI 2 molar ratio is x, x= 1, 2, 3, 4) (Supplementary Fig. 1a). The results show that the PL lifetime increases with the increasing amount of IZI, suggesting that the non-radiative recombination of the γ-CsPbI 3 is be suppressed with increasing IZI, which is consistent with the PLQE results ( Supplementary Fig. 1b). We have added this in the revised manuscript (Line 84 to 88, Page 3, highlighted, Supplementary Fig. 1

Response:
We thank the reviewer for this useful comment. The valence bands of IZI-CsPbI 3 film was obtained via ultraviolet photoelectron spectroscopy (UPS) measurement (Supplementary Fig.   14), and the conduction band was estimated using the band gap derived from the absorption band edge. we have added the energy levels of each layer for the device in the revised manuscript (Fig.   4a). We have added this in the revised manuscript (Line 206 to 209, Page 8, highlighted, Supplementary Fig. 14). Supplementary Fig. 14  substrate.

Fig. 4a
Flat-band energy level diagram of LED structure. the valence bands of ZnO/PEIE substrate and IZI-CsPbI 3 film was obtained via ultraviolet photoelectron spectroscopy (UPS) measurement, and the conduction band was estimated using the band gap derived from the absorption band edge.

Comment #7:
The authors showed a crystal stability test by XRD measurement. The optical stability test of IZI-CsPbI 3 film is also important.

Response:
We thank the referee for the helpful suggestion. We have added the evolution of the normalized PL intensity of the CsPbI 3 films with different contents of IZI in precursor solutions (IZI/PbI 2 molar ratio is x, x= 1, 2, 3, 4) on ZnO/PEIE substrate in ambient air. The PL intensity of Reviewer #2 (Remarks to the Author):

Comment #1:
The authors report efficient CsPbI 3 light-emitting diodes by employing an imidazolium additive. The authors have also performed thorough experimental studies to determine the mechanism of low temperature CsPbI 3 perovskite formation through the use of imidazolium.
Given that cesium-based perovskites show good promise in long-term device stability, this work is important and could be published in nature communications upon minor revision.

Response:
We thank the reviewer for recognising the importance of our work. Guided by these constructive comments, we have made improvements throughout the paper.

Comment #2:
Could the authors comment if it is experimentally confirmed that the imidazole remains in the perovskite film/layer? Presumably this could be determined by some spectroscopic techniques.

Response:
We appreciate this suggestion. In this work, we detect the chemical structure of the perovskite film by the XPS measurement. two peaks assigned to 401.7 eV (N-3) and 400.2 eV (N-1) of imidazole are observed, confirming the existence of imidazole in the IZI-CsPbI 3 films (Fig. 2).
ATR-FTIR spectroscopy measurement result can further confirm this result. As shown in out-of-plane bend) and 658 (torsion) cm-1 are clearly observed.

Comment #2:
The authors mentioned that other large ammonium salts worked as well. How crucial is the identity of this ammonium salt for general device performance?

Response:
We appreciate the reviewer for raising this interesting point. In this work, we have revealed that the CsPbI 3 black phase could also be fabricated at low temperature through synergistic effect of ZnO and other large organic ammonium salt, such as BAI, HAI, PEAI and NMAI, demonstrating the generality of the approach. The IZI has been successfully used to improve the performance of CsPbI 3 LED by composite engineering and device optimization. We have not made similar effort with other ammonium salts, but we believe it shall be possible to achieve high performance devices with them. The key point we think is to achieve high quality crystals with low defect densities, presumbly the ammonium can effectively passiviate defects, and to achive desirable morphology to maximize the light outcoupling efficiency.

Comment #3:
The imidazole has helped in the formation of CsPbI 3 at low temperature. Could the authors comment on whether it is helpful for keeping CsPbI 3 in its black phase over the long term?
How does it compare to CsPbI 3 formed at high temperature without the help of imidazole?

Response:
We have performed the phase stability measurements of the regular γ-CsPbI 3 obtained from the high-temperature annealing process. The γ-CsPbI 3 black phase can only retain for 4 hours in the ambient air ( Supplementary Fig. 2). In contrast, the IZI-CsPbI 3 film exhibits negligible degradation after exposing for 36 days in the same environment (Fig. 1a), suggesting significantly improved phase stability compared to previously reported results. So we believe that the IZI is helpful for keeping CsPbI 3 in black phase for long term. We have added this in the revised manuscript (Line 88 to 91, Page 3, highlighted, Supplementary Fig. 2). Supplementary Fig. 2 XRD pattern of CsPbI 3 film without IZI after exposed in the ambient air for various durations. The film is fabricated on ZnO/PEIE substrate at annealing temperature of 230 o C. Fig. 1a XRD pattern of IZI-CsPbI 3 film on ZnO/PEIE substrate after exposed in the air with 80 % relative humidity for various durations.
Reviewer #3 (Remarks to the Author): Comment #1: The authors has developed a low-temperature fabrication method of γ-CsPbI 3 , where organic ammonium cations were introduced (through imidazolium iodide (IZI)) to form an intermediate phase to mediate the formation process of the perovskite. Through XPS and ATR-FTIR characterizations, the authors pointed out that the deprotonation of the IZ + cations by ZnO promotes interionic exchange between IZ + and Cs + , and therefore facilitates the formation of the γ-CsPbI 3 .
The work also demonstrated a record EQE of 10.4% in 3D CsPbI 3 -based red PeLEDs. The findings reported in this manuscript are interesting and may have important impact on low-temperature fabrication of thin film all-inorganic perovskites. In general, this is a high-quality work. I would recommend acceptance of this manuscript after the following issues are properly addressed.

Response:
We thank the reviewer for well appreciating the importance of our work. We have revised our manuscript carefully according to the reviewer's suggestion.
Comment #2: Figure 2b shows different N 1s peaks of IZI on ZnO and on ITO. It is suggested that ZnO can deprotonate the IZ + cation of IZI through formation of Zn-O-H bonds. In principle, ITO should also allow for a similar deprotonation process through formation of Sn-O-H bonding. Why does the latter process not occur? Can the authors please clarify what properties are required for the metal oxide to initiate the deprotonation process?
Response: We think that the chemical acidity of the metal oxide plays a key role in the deprotonation process. The ZnO with a higher isoelectric point (IEP) value of 8.7-10.3 is likely to induce an easier deprotonation of the organic cation. However, the SnO 2 with IEP value of 6.6-9.5 present neutral and difficult react with the organic cation. So we believe that the basic metal oxide substrate could initiate the deprotonation process of organic ammonium cations easily.

Comment #3:
Regarding the mechanism illustrated in Figure 2f: if the interionic exchange is enabled by ZnO, which contacts the bottom surface of the perovskite, could the exchange happen efficiently at the top surface of the perovskite? Does the thickness of the perovskite film impose a limit to the effectiveness of this method?

Response:
We thank the referee for the insightful comment. Since the ZnO induced deprotonation process likely mainly occurs at the ZnO interface, it is interesting to investigate how thick the perovskite can be formed by this approach. Supplementary Fig. 11 shows the absorbance of the films with various thickness fabricated from different concentration of precursor solutions. It shows the absorbance at 687 nm from the black phase CsPbI 3 increases linearly with the film thickness less than 200 nm. Above 200 nm, the absorbance saturates and declines. This result suggests that the ZnO