High-performance hysteresis-free perovskite transistors through anion engineering

Despite the impressive development of metal halide perovskites in diverse optoelectronics, progress on high-performance transistors employing state-of-the-art perovskite channels has been limited due to ion migration and large organic spacer isolation. Herein, we report high-performance hysteresis-free p-channel perovskite thin-film transistors (TFTs) based on methylammonium tin iodide (MASnI3) and rationalise the effects of halide (I/Br/Cl) anion engineering on film quality improvement and tin/iodine vacancy suppression, realising high hole mobilities of 20 cm2 V−1 s−1, current on/off ratios exceeding 107, and threshold voltages of 0 V along with high operational stabilities and reproducibilities. We reveal ion migration has a negligible contribution to the hysteresis of Sn-based perovskite TFTs; instead, minority carrier trapping is the primary cause. Finally, we integrate the perovskite TFTs with commercialised n-channel indium gallium zinc oxide TFTs on a single chip to construct high-gain complementary inverters, facilitating the development of halide perovskite semiconductors for printable electronics and circuits.


REVIEWER COMMENTS
Reviewer #1 (Remarks to the Author): I read carefully the response letter based on previous comments, I think the authors have almost addressed my concerns. There is still one question: For the R1Q3 about the XPS results, the authors argued that the perovskite is inevitably exposed to the air, I suggest the authors to coat some inert materials such as PMMA on perovskite and then carried out XPS measurement, after etching the PMMA in XPS instrument, you can get more real result about the perovskite.
Reviewer #2 (Remarks to the Author): The authors have provided a clear and satisfactory response to the issues raised in my report. However, one issue requires further clarification. The authors now argue that the concentration of Sn4+ defects might be due to accidental oxygen exposure of the films and they have moved the XPS results to the SI. I think this is not very satisfactory given that the question of Sn4+ defects is an important and widely appreciated concern.
I would encourage the authors to perform their XPS experiments under conditions where the effect of oxygen exposure can be eliminated by an inert atmosphere transfer into the vacuum chamber (or at least minimized by investigating how the Sn4+ concentration varies when the films are exposed to air for different length of times before the transfer into UHV.) Reviewer #3 (Remarks to the Author): The authors, have taken the corerctions very seriously and have submitted a very good revised manuscript that in my opinion should be publish without any further need of corrections. 1

Reviewer #1 (Remarks to the Author):
I read carefully the response letter based on previous comments, I think the authors have almost addressed my concerns. There is still one question: For the R1Q3 about the XPS results, the authors argued that the perovskite is inevitably exposed to the air, I suggest the authors to coat some inert materials such as PMMA on perovskite and then carried out XPS measurement, after etching the PMMA in XPS instrument, you can get more real result about the perovskite.
Reply: We would like to thank the reviewer for reviewing our manuscript and for positive comments.
Following the reviewer's suggestion, we re-conducted the XPS measurement under a minimized oxygen exposure condition using high-vacuum metal/rubber/metal containers (which can be high-vacuumed under an inert atmosphere) instead of the previous low-vacuum package ( Figure R1) for the sample transfer process. In this way, we obtained more reliable data.
We updated the data in Figure S7 and Figure 3b, also shown below. The new data on Sn 3d5/2 XPS analysis shows the same relative Sn 4+ content trend (I > I/Br > I/Cl > I/Br/Cl) with the original ones but lower absolute value for each film, supporting the previous discussion in the main text well. The relative Sn 2+ /Sn 4+ ratios are found reasonable compared to previous reports on Sn-based perovskites. [1][2][3][4] We also thank the reviewer for suggesting the PMMA coating/etching method. The high-power ion etching can cause the degradation of the local bonding environment considering the vulnerable property of halide perovskites, which may influence the measurement reliability. Therefore, we chose the method above, and it turned out effective. Figure R1. Previous low-vacuum package with sample tray and the newly-developed 2 high-vacuum metal/rubber/metal container. The new container can be first high-vacuumed under an inert atmosphere using the anti-chamber in an N2-filled glove box, then vacuumed with a plastic bag for sample transfer. Supplementary Fig. 7. Analyses of the Sn 3d5/2 core level XPS spectra with fitting results for Sn δ<2+ , Sn 2+ , and Sn 4+ .

Reviewer #2 (Remarks to the Author):
The authors have provided a clear and satisfactory response to the issues raised in my report. However, one issue requires further clarification. The authors now argue that the concentration of Sn 4+ defects might be due to accidental oxygen exposure of the films, and they have moved the XPS results to the SI. I think this is not very satisfactory given that the question of Sn 4+ defects is an important and widely appreciated concern.
I would encourage the authors to perform their XPS experiments under conditions where the effect of oxygen exposure can be eliminated by an inert atmosphere transfer into the vacuum chamber (or at least minimized by investigating how the Sn 4+ concentration varies when the films are exposed to air for different length of times before the transfer into UHV.) Reply: We thank the reviewer for reviewing our work again and for the positive comments on our Response Letter. We also thank the reviewer's helpful suggestions on XPS experiment.
Inspired by the reviewer's good suggestion, we re-conducted the XPS measurement under a minimized oxygen exposure condition using high-vacuum metal/rubber/metal containers (which can be vacuumed under an inert atmosphere) instead of the previous low-vacuum package ( Figure R1) for sample transfer. In this way, we obtained more reliable data.
We updated the data in Figure S7 and Figure 3b, also shown below. The new data on Sn 3d5/2 XPS analysis shows the same relative Sn 4+ content trend (I > I/Br > I/Cl > I/Br/Cl) with the original ones but lower absolute value for each film, supporting the previous discussion in the main text well. The relative Sn 2+ /Sn 4+ ratios are found reasonable compared to previous reports on Sn-based perovskites. 1-4 Figure R1. Previous low-vacuum package with sample tray and the newly-developed high-vacuum metal/rubber/metal container. The new container can be first high-vacuumed under an inert atmosphere using the anti-chamber in an N2-filled glove box, then vacuumed 4 with a plastic bag before transfer. Supplementary Fig. 7. Analyses of the Sn 3d5/2 core level XPS spectra with fitting results for Sn δ<2+ , Sn 2+ , and Sn 4+ .