Atomic-level direct imaging for Cu(Ⅰ) multiple occupations and migration in 2D ferroelectric CuInP2S6

Atomic-level direct imaging for Cu(Ⅰ) multiple occupations and migration in 2D ferroelectric CuInP2S6 | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Atomic-level direct imaging for Cu(Ⅰ) multiple occupations and migration in 2D ferroelectric CuInP 2 S 6 Wanbiao Hu, Changjin Guo, Xiali Liang, Jiyang Xie, Caifu Wen, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4487714/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 22 Nov, 2024 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract CuInP 2 S 6 (CIPS) is an emerging 2D ferroelectric material known for disrupting spatial inversion symmetry due to Cu(I) position switching. Its ferroelectricity strongly relies on the Cu(I) atom/ion occupation ordering and dynamics. Nevertheless, the accurate Cu(I) occupations and correlated migration dynamics under the electric field, which are key to unlocking ferroelectric properties, remain controversial and unresolved. Herein, an atomic-level direct imaging through aberration-corrected scanning transmission electron microscopy is performed to precisely trace the Cu(I) dynamic behaviours under electron-beam irradiation along (100)-CIPS. It clearly demonstrates that Cu(I) possesses multiple occupations, and Cu(I) could migrate to the lattice, vacancy and interstitial sites between the InS 6 octahedral skeletons of CIPS to form local Cu x InP 2 S 6 ( x = 2–3) structure. Cu(I) multi-occupations induced lattice stress results in a layer sliding along the b -axis direction with generating a sliding size of 1/6 b axis. The Cu x InP 2 S 6 ( x = 2–3) exists in a type of dynamic structure, only metastable with electron dose over 50 e − Å −2 , thus generating a dynamic process of Cu x InP 2 S 6 (x=2-3) ⇌ CuInP 2 S 6 , a completely new phenomenon. These findings shed light on the novel mechanism underlying the Cu(I) migration in CIPS, providing crucial insights into the fundamental processes governing its ferroelectric properties. Physical sciences/Materials science/Condensed-matter physics/Ferroelectrics and multiferroics Physical sciences/Materials science/Nanoscale materials/Two-dimensional materials ion migration CuInP2S6 ferroelectric materials crystal structure Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction 2D Van der Waals ferroelectric materials are emerging with intriguing physical and quantum phenomena for a broad span of applications in e.g. field-effect transistors (FETs), non-volatile memory, ferroelectric tunnel junctions (FTJs), and beyond. A prominent representative one is 2D monoclinic CuInP 2 S 6 (CIPS). It manifests room-temperature and tunable quadruple-well ferroelectricity, 1–4 and therefore, many electronic devices based on CIPS have been rapidly developed. 5–9 With regard to the ferroelectric polarizations, great efforts have been devoted, with having been uncovering the dominated role from the lattice Cu(I) site occupations. To clearly illustrate it, Fig. 1 demonstrates the Cu(Ⅰ) occupation and polarization characteristics. Cu(Ⅰ) ions are usually located between two InS 6 octahedral skeletons (positions I or III in Fig. 1 a). The InS 6 skeletons with Cu(Ⅰ) locked inside combine with P-P ion pairs to form a continuous layer. The layers are connected by Van der Waals forces stacking in the ABAB stack, allowing it to reach the low dimensional limit by separating layers from each other experimentally. The primitive ferroelectricity of 2D CIPS originates from the spatial inversion symmetry breaking with the Cu(Ⅰ) ions occupying either I or III position (Fig. 1 a) where the feasible mobility of Cu(Ⅰ) ions within the in-plane (IP) and out-of-plane (OOP) orientations in CIPS crystal structure also happens. But more crucially, when an external field ( e.g. electric field, stress, thermal, etc .) is applied, the Cu(I) occupations may switch to align the polarization direction with the electric field direction ( e.g. all the Cu(I) ions switch to position III to make the polarization is up when an upward electric field applied in Fig. 1 a). This is related to the Cu(I) migration and similar phenomena have been well explored in CIPS. However, several fundamental questions raise, e.g. how many possible positions that could accommodate Cu(I) switching or additional migration? Are there any other Cu(I) occupations except the I and III positions (Fig. 1 a)? Whether Cu(I) possibly migrates to occupy the interstitial site (position II in Fig. 1 a) or interlayer site (position IV in Fig. 1 a) exists or not? 10,11 Definitely, these basically structural and physical pictures are fundamentally important to understanding the ferroelectric nature and polarization switching of CIPS. There are many studies on Cu(I) migration and associated mechanisms, mainly including two categories: ( 1 ) Cu(I) ions migration and ferroelectric polarization switching under the external electric field. As shown in Fig. 1 b, both OOP and IP modes ( i.e. applying external electric field styles) have been performed. Under OOP migration path (Fig. 1 b-i), some Cu(I) ions could be occasionally trapped in the vdW gap and even extracted out of the lattice because of the coexistence of polarization switching and ionic conduction. 12 Under IP migration path (Fig. 1 b-ii), Cu(I) ions are proven to move along the electric field by in-situ scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) measurements. 13 The Cu(I) ions migration toward the negative electrode leads to a typical diode-like resistive switching behavior in the homojunction device. 14 Further, the robust threshold-switching behavior assisted by Cu(I) migration is also found in CIPS heterostructure even if using asymmetrical electrodes ( e.g. graphene, Au vs. Cu). 15 To conclude, the Cu(I) ions migration under the external electric field is a universal behavior unaffected by the electrodes. ( 2 ) Ferroelectric domain distributions and Cu(I) ions location under the flexoelectric effect. Polarization switching based on the flexoelectric effect can effectively avoid triggering longer-range ionic redistribution (Cu(I) ions just move their positions confined between InS 6 octahedral skeletons). As depicted in Fig. 1 c, when stress is applied on OOP mode, the free energies for occupying position up and position down are not equal, and Cu(I) ions tend to move and occupy positions in the direction gradient, thus exhibiting polarization ( P up and P down in Fig. 1 c). The synergistic effect of Cu ion migration and interfacial Schottky barrier controls the current and device performance. 16–21 These above studies combine with the existing cognition inferred from multiple characterizations have played a great role in promoting the understanding of the Cu(I) migration ferroelectric polarization. However, the basically physical picture with showing accurate Cu(I) lattice-site occupations and Cu(I) configurations to the surrounding arrangement in CIPS is far beyond established. In particular, in addition to the primitive Cu(I) lattice positions I and III (Fig. 1 a), whether the positions II and IV (Fig. 1 a) that are only predicted theoretically really occur and how they contribute to the observed overall ferroelectricity are also the fundamental questions to be resolved. Herein, an atomic-level direct imaging to precisely trace the Cu(I) migration and its local dynamics in CIPS is performed through aberration-corrected scanning transmission electron microscopy (AC-STEM). The cleavage plane (100) based direction ( i.e. a -axis direction) is chosen for the imaging view because the out-of-(100) plane best reflects the atom occupying situation as well as the clear Cu(I) migration along c direction (Fig. 1 d). Accordingly, the configuration and moving dynamics (speed, energy, and structural stability) of Cu(I) ions in the CIPS under an electric field ( i.e. electron beam irradiation) are well monitored, with finding the presence of Cu(I) interstitial occupations i.e. positions II presented in Fig. 1 a. The Cu(I) outside the field of view converges towards the irradiation location and crowds in the possible location of Cu(I), eventually forming an unstable Cu x InP 2 S 6 ( x = 2–3) local structure. 2. Results Atomic-level imaging observation along (100) plane The pivotal criterion to get accurate and sufficient atomic-level structural information on Cu(I) occupations is the exposed plane aligning flawlessly with a -axis, i.e. (100) cleavage plane, as proposed in Fig. 1 d. For this, the as-grown CIPS single crystal is precisely cut into nanosheet along a -axis by focused-ion-beam (FIB) technology (Figure S1 ). The low-dose integrated differential phase-contrast coupled scanning transmission electron microscopy (iDPC-STEM) is performed to achieve atomic-level imaging. Figure 2 a shows a dark field (DF) mode image where all the Cu(I)/In/P/S atoms and their positions are clearly seen. Strikingly, the atoms between the InS 6 octahedral skeletons are not solitary, but instead there occur obvious multiple lattice sites for Cu(I) occupations, e.g. Cu(I) ions occupy both position I and III (Fig. 2 b-i); Cu(I) ions occupy simultaneously position I, II and III (Fig. 2 b-i i ). For convenience, we here refer to the local crystal configuration situation of Region I and Region II by Cu 2 InP 2 S 6 and Cu 3 InP 2 S 6 , respectively. It is worth noting that the local structure Cu x InP 2 S 6 ( x = 2, 3) only occurs in a tiny region but does not change the overall chemical formula of the CuInP 2 S 6 single-crystal. The structural configurations with forming local Cu x InP 2 S 6 ( x = 2, 3) can be universally observed from different areas or regions in CIPS (Figure S2), which is definitely attributed to the Cu(I) migration and multiple occupations under the electric field. To gain the insights on the microscopical results, standard atomic structure and lattice occupations are compared. The theoretical angular/distance relationship between adjacent In ions, as the (001) plane demonstrated in Fig. 2 c, is ∠ ABC = 59.969 º (~ 60 º) and BC = 0.6099 nm, respectively. Experimentally, the projected distance between the two nearest In ions in a layer from the (100) plane is 0.5500 nm (highlighted by red line in Fig. 2 a), which yields an actual distance of 0.55/cos(59.969/2) = 0.6350 nm according to the geometric relationship. This value is 4.12% larger than the theoretical In-In distance ( i.e. BC = 0.6099 nm in Fig. 2 c), which suggests the existence of the multiple Cu(I) occupations between the InS 6 octahedral skeletons that thus leads to lattice expansion in the direction. The occurrence of the atomic multi-occupation and lattice constant increase is to release the local stress and maintain the structural stability when the electron-beam irradiation continues. Atomic distance relationships for P-P in one layer and In-In between adjacent two layers are further analyzed. As shown in Fig. 2 d, the theoretical P-P distance is 0.2218 nm, almost analogous to the experimental value of 0.2256 nm (highlighted by blue line in Fig. 2 a). The theoretical In-In distance between the nearest layers in Fig. 2 e is 0.6812 nm, while the experimentally measured distance is 0.6812 nm (highlighted by yellow line in Fig. 2 a). It is worth noting that the lattice constant c (= 1.3524 nm) experimentally determined also matches well the theoretical one (1.3623 nm). These percentage differences, i.e. 1.7% in P-P, 0.0058% in In-In in layers, and − 0.73% in lattice constant c , respectively, are small enough to be considered no variations in the c -axis direction. To conclude, the Cu(I) atomic multi-occupations do not cause the out-of- ab -plane structural fluctuations but apparent in- ab -plane structural expansion. Then, what is the structural reason? This can be attributed to the sliding mis-arrangements for the CIPS layers. As indicated in Fig. 2 f, the theoretical In ions between the CIPS layers should be arranged in a straight line from the perspective of the (100) plane, but the microscopical observation clearly exhibits that the In ions in CIPS layers are not situated in a straight line along the c -axis direction. The array paths of In ions are shown in Fig. 2 b- iii dotted orange lines, which is a broken line, containing lines segment parallel to the c -axis. Accordingly, the sliding boundary position can be deduced, which is rightly the interface of the two adjacent CIPS layers, as indicated by the green line in Fig. 2 a. This suggests that the smallest unit in which the glide occurs is a single cell, without compromising its double-layer structure. The sliding distance between the layers is carefully confirmed, and its magnitude is 1/6 lattice constant b . This further confirms that in order to keep the stability of the crystal structure induced by the Cu(I) multiple occupations, the interlayer sliding releases the lattice stress but does not cause changes in lattice constant c . In addition to the atom positions or occupations, more atomic-level information e.g. atomic distribution/intensity/shift etc. are further extracted. To this, precise analysis and transformation are performed on the iDPC-STEM images (Figs. 3 a, S3) by optimized Multiple-Ellipse Fitting method with the detailed calculations are seen in Supporting information section (Figures S3-7). 22 The noise reduction image in Figure S3a is obtained by BM3D method and the position of the atoms are determined by threshold method. 23 Fig. 3 a and Fig. 3 b show respectively the pseudo-color display raw image and fitting image that exhibit quite high matchiness, thus giving a very small difference in intensity, morphology and strength errors (Figs. 3 c, S3) with all the fitting parameters can be obtained in Table S1 ). Accordingly, the 2D intensity distributions of Cu, In, P, and S based on the elemental mapping and statistical analysis (corresponding to the highlighted area in Fig. 3 d, Figure S8) are deduced, as shown in Fig. 3 e, and meanwhile, the distribution of the atom shift is also obtained (Fig. 3 f). With regard to the Cu(I) occupation, Cu(I) ions are present between the InS 6 octahedral skeletons in the form of Cu x InP 2 S 6 ( x = 2–3) local structure. As a matter of fact, the STEM image contrast predominantly depends on the atomic number (Z) and the atom count. 24,25 The intensities of Cu(I) are primarily distributed in 5–7×10 6 and 9–16×10 6 for positions II and I & III, respectively in the intensity distribution histogram of Fig. 3 e. These results show that Cu(I) in CIPS have their own filling rules: Under electron irradiation, Cu(I) will preferentially occupy positions I and III to form Cu 2 InP 2 S 6 local structure. When positions 1 and 3 are filled, Cu(I) then will occupy position II to further form Cu 3 InP 2 S 6 local structure. These results are very significant for the understanding of the kinetic process of copper ion occupation in CIPS. The Cu(I) preferential occupation might induce the atom/ion shift/displacement in CIPS crystal structure, shown in Fig. 3 f, which favors the understanding of lattice distortion and polarization. There occur two types of the atom shift (the shift vectors are highlighted by yellow arrows). One shifts along the directions ([010] or [0–10]) distinguished from the middle position, which contributes to the lattice distortion by atomic multiple occupancies between the InS 6 octahedral skeletons. The other shifts along direction, which is due to the intrinsic out-of-plane (OOP) polarization in the lattice. Because the atomic shift vector in direction is overall upward, it does not affect the lattice constant of the c -axis. The fitting result corresponding to the iDPC-STEM image indicates the dependability of the structural evolution process of Cu(I) multiple occupancies between the InS 6 octahedral skeletons of CIPS under the electron-beam irradiation along direction. Dynamic and stability of Cu x InP 2 S 6 ( x = 2–3) local structure The above precise iDPC-STEM imaging and analysis prove the Cu(I) multiple occupations and the structural changes of CIPS under electron irradiation, but the dynamic process and stability for the local structural configurations on Cu x InP 2 S 6 ( x = 2–3) remain to be investigated clearly. Herein, a facile strategy is devised to observe the Cu x InP 2 S 6 ( x = 2–3) configurations. Specifically, a large current (50 pA) is used to irradiate partial region (red box in Fig. 4 a) for 30 seconds, and EDS mapping is then carried out over a larger area with a small current (33 pA). The Cu x InP 2 S 6 ( x = 2–3) configurations can be judged by the fluctuations in elemental contents in the red box. The resulting EDS mapping (Fig. 4 b) shows that there is no clear distinction of element contrast. Another similar experiment by line scanning (Fig. 4 c, d) for element mapping also gives the same conclusion, i.e. the electron-irradiated area exhibits the uniform elemental distribution. As a matter of fact, the Cu(I) contents (26.01% for Fig. 4 a and 29.52% for Fig. 4 c) are much higher than the theoretical value from the standard structure (Table S2 and Table S3), which means the occurrence of the Cu(I) multi-occupancy with forming Cu x InP 2 S 6 ( x = 2–3). However, the Cu x InP 2 S 6 ( x = 2–3) is quite unstable because of undergoing the extremely severe lattice distortion, unless continuous injection energy (electron-beam irradiation) maintains Cu x InP 2 S 6 ( x = 2–3) configurations. Otherwise, when the energy (electron-beam irradiation) is removed, the Cu( 1 ) would move back to its original position to retain the standard CuInP 2 S 6 structure (Figure S9). This dynamic process that clearly uncovers locally structural configuration fluctuations i.e. is obviously a result of the Cu(I) migration under the electric field behavior. The Cu(I) migration dynamics is further investigated by monitoring the elemental fluctuations during the continuous electron-beam irradiations. To figure out the energy dependence or threshold for the Cu(I) migration, the varied current dose rates are performed. Here Fig. 4 e-h only show the cases of two typical current rates, i.e. 30 pA and 72 pA. Note that too small dose ( e.g. 2 pA) can not drive the Cu(I) migration (Figure S9) while too large dose makes Cu(I) migration too quick for inaccurate detection. It can be seen that the detected elemental amounts exhibit different changes with increasing the irradiation time. Clearly from both the EDS imaging (Fig. 4 e) and quantitive analysis (Fig. 4 g), the Cu(I) amount gradually increases with the irradiation time at a low current (30 pA), from initial 25.9% to final 27.09% (irradiation for 30 mins) and keeping unchanged upon longer irradiation in atomic proportion. While, high current irradiation e.g. 72 pA (Figs. 4 f and 4 h), might be due to the quite fast Cu(I) migration, does not bring about any apparent fluctuations in elemental distributions that nearly undergo unchanged at about 28%. The energy injected for Cu(I) migration induced crystal structure distortion is further quantified, which is referred to the current dose (D) that can be obtained by Eq. ( 1 ) is 50 e − Å −2 and 120 e − Å −2 for 30 pA (Fig. 4 e) and 72 pA (Fig. 4 f), respectively. Although the energy dose increases doubly, the Cu(I) content is not increased all the way, but finally fixed at about 28%. This is, again, indicative of the Cu(I) migration but the available positions in the lattice sites ( i.e. positions I, II, III, IV illustrated in Fig. 1 a) for Cu(I) occupation are in a certain level. Even if forming the Cu 3 InP 2 S 6 structural configuration, Cu(I) has an atomic percentage content of 25%, still less than the value obtained from experimental determinations. One possibility is Cu(I) may occupy the interlayer position IV (Fig. 1 a) that has been predicted by the theoretical calculation, but here from the atomic-level microscopy experiment, no direct evidence that Cu(I) could occupy the interlayer position. 3. Discussion The ferroelectricity of CuInP 2 S 6 (CIPS) strongly relies on the Cu(I) atom/ion occupation ordering and dynamics. When Cu(I) only occupies either the position I or position III (Fig. 1 a) orderly, the spatial inversion symmetry is broken, thus exhibiting apparent ferroelectricity. Such a state is structurally and thermally stable at room temperature that has been proven by various experimental and theoretical studies. Up to date, there is no doubt on the Cu(I) migration, but the direct evidences especially the atomic-level imaging that demonstrate the specific occupation sites as well as the locally structural configurations under the electric field are not explored previously. Therese questions are fundamental for understanding ferroelectric mechanism and exploiting ferroelectric-based applications. This room-temperature ferroelectricity must allow the occurrence of the intrinsic and ordered Cu(I) vacancy (conversely, position III/I vs . occupied I/III), and meanwhile, endow the adjacent Cu(I) a chance for migrating to occupy the Cu(I) vacancy, which is structurally responsible of the Cu(I) migration. When the Cu(I) has migrated to occupy the vacancy, new locally structural configurations would form, with the theoretically possible configurations could be Cu x InP 2 S 6 ( x = 2–4). Here, we noted that no any traces from our results indicate the Cu(I) could migrate to the interlayer site (Position IV in Fig. 1 a), despite of this possibility has been predicted by John A. Brehm et al. that Cu(I) will migrates to the interlayer site (position IV in Fig. 1 a) and bond with the S atoms in the next layers under the external field to reveal tunable quadruple-well ferroelectric by DFT calculation, 4 so it only leaves the possibility of the Cu x InP 2 S 6 ( x = 2–3) that have been observed in our experiment. When the three positions I/II/III (Fig. 1 a) are simultaneously occupied (Fig. 2 b- ii ) with forming local Cu 3 InP 2 S 6 , or both positions I and III (Fig. 1 a) are occupied (Fig. 2 b-i, Cu 2 InP 2 S 6 ), these formed locally structural configurations should not exhibit the ferroelectricity because the occupied Cu(I) causes the structural symmetry. The case of the Cu(I) occupies the interstitial site (position II) and one of lattice sites (positions I/III) but not simultaneously occupies vacancy site should also generate/maintain the ferroelectricity because the spatial inversion symmetry is still broken, but unfortunately, it is not observed by current atomic imaging that may be due to the Cu(I) fast migration effect. Anyway, these locally structural configurations exist only at a very tiny scale instead of generating long-range migration, as indicated from the relatively small change in Cu(I) atom proportion (Fig. 4 ), which should not influence the ferroelectricity to a large extent. In summary, through iDPC-STEM direct imaging on (100)-CIPS, we obtained the atomic-level structure clearly. We confirm that Cu(I) possesses multiple occupations (lattice, vacancy and interstitial sites) and Cu(I) will migrate to positions I, III, and II between the InS 6 octahedral skeletons of CIPS to form local Cu x InP 2 S 6 ( x = 2–3) structure when electron-beam irradiation along direction. As a result of the Cu(I) multiple-occupation, lattice distortion occurs for stress release to induce layer sliding along the b -axis direction with generating a sliding size of 1/6 b axis. The dynamic process of the locally formed Cu x InP 2 S 6 ( x = 2–3) configurations monitored as a function of the irradiation times and currents indicates that Cu x InP 2 S 6 ( x = 2–3) is an unstable structure, which just exists by external energy injection ( i.e. electron-beam irradiation), but it will return back to standard CuInP 2 S 6 immediately when the external energy vanishes. With new findings of Cu(I) multiple occupations and migration dynamics, this work provides the deep insights on the Cu( 1 ) migration origin and ferroelectric mechanism in CIPS, which are significant for guiding device design and performance improvement based on CIPS. 4. Materials and methods Synthesis of Single Crystal CuInP 2 S 6 The raw materials and chemicals used in this study were sourced from reputable suppliers. Copper powder (99.9%), Indium powder (99.99%), and Sulfur powder (99.999%) were purchased from Zhongnuo New Material Co., LTD. Phosphorus particle (99.999%) was obtained from Sichuan Khaqikun Technology Co., LTD. I (99.8%) was procured from Macklin Co., LTD. The CuInP 2 S 6 crystals were synthesized by a chemical vapor transport method. The raw materials Cu, In, P, and S were ground into powder with a mortar for homogenization purposes according to the stoichiometric ratio and iodine was used as the transport agent (50 mg/25 mL). Then the mixture into the bottom of a quartz test tube (inner diameter, 16 mm; length, 30 cm) subjected to evacuation via a mechanical pump and sealed via a Partulab MRVS-1002 system. The temperature for the evaporation and crystallization zones was initially set at 650 ℃ and 750 ℃, respectively, and the chemical reaction was allowed to proceed for 2 days. Following this, the temperature in the evaporation and crystallization zones was adjusted to 700°C and 665°C, respectively. Then the process was continued for 5 days. Finally, cool down to room temperature at a cooling rate of 10 °/h. Crystal structure characterizations The CIPS crystal was cut along the (100) plane by Focus Ion Beam (FIB) double-beam electron microscopy (Thermofisher Helios 5 CX) for electron microscopy measurement. A thickness of 5 µm platinum is deposited on the surface of CIPS as a protective layer, leaving 1 µm remaining after cutting. iDPC-STEM imaging were performed on Thermofisher Spectra 300, equipped with Gatan 1069 camera for collecting EDS mapping data. Plan-view STEM Imaging The integrated differential phase-contrast scanning transmission electron microscopy (iDPC-STEM) imaging was conducted using Cs-corrected (S) TEM (Thermofisher Spectra 300), operated at 300 kV. The collection semi-angles for the high-angle annular dark field (HAADF) STEM images were 51–200 mrad and the accumulated electron dose D in iDPC-STEM imaging was calculated by $$\text{D}=\frac{I\times t}{A}$$ 1 Where I was the electron-beam current, t was the irradiation time, and A was the area of irradiation. The electron-beam current was 30 pA and 72 pA for Fig. 4 e and Fig. 4 f, respectively. The dwell time is 20 µs. The pixel size of Fig. 4 e, f was 8.67 Å, and the electron dose D is about 50 e − Å −2 and 120 e − Å −2 , respectively. Analysis of atom shift in CIPS iDPC-STEM images of CIPS were analyzed by CalAtom Software to extract the atomic position by multiple-ellipse fitting. Atomic position and species are determined by fitting the intensity of the iDPC image. The atomic shift is determined by comparing the fitting results with the standard CIPS structure. References Maisonneuve V, Evain M, Payen C, Cajipe VB, MoliniC P (1995) Room-temperature crystal structure of the layered phase Cu I In III P 2 S 6 . J Alloys Compd 218:157–164 Belianinov A et al (2015) CuInP 2 S 6 Room Temperature Layered Ferroelectric. Nano Lett 15:3808–3814. 10.1021/acs.nanolett.5b00491 Liu F et al (2016) Room-temperature ferroelectricity in CuInP 2 S 6 ultrathin flakes. Nat Commun 7. 10.1038/ncomms12357 Brehm JA et al (2019) Tunable quadruple-well ferroelectric van der Waals crystals. Nat Mater 19:43–48. 10.1038/s41563-019-0532-z Neumayer SM et al (2020) Alignment of Polarization against an Electric Field in van der Waals Ferroelectrics. Phys Rev Appl 13. 10.1103/PhysRevApplied.13.064063 Li P et al (2022) Electrostatic Coupling in MoS 2 /CuInP 2 S 6 Ferroelectric vdW Heterostructures. Adv Funct Mater 32. 10.1002/adfm.202201359 Taylor PD, Tawfik SA, Spencer MJS (2022) Ferroelectric van der Waals heterostructures of CuInP 2 S 6 for non-volatile memory device applications. Nanotechnology 34. 10.1088/1361-6528/aca0a5 Li W et al (2022) A Gate Programmable van der Waals Metal-Ferroelectric‐Semiconductor Vertical Heterojunction Memory. Adv Mater 35. 10.1002/adma.202208266 Wu J et al (2020) High tunnelling electroresistance in a ferroelectric van der Waals heterojunction via giant barrier height modulation. Nat Electron 3:466–472. 10.1038/s41928-020-0441-9 Maisonneuve V, Cajipe VB (1997) Ferrielectric ordering in lamellar CuInP 2 S 6 . Phys Rev B 56:10860–10868 Zhang D et al (2021) Anisotropic Ion Migration and Electronic Conduction in van der Waals Ferroelectric CuInP 2 S 6 . Nano Lett 21:995–1002. 10.1021/acs.nanolett.0c04023 Zhou Z et al (2023) Unconventional polarization fatigue in van der Waals layered ferroelectric ionic conductor CuInP 2 S 6 . Nat Commun 14:8254. 10.1038/s41467-023-44132-y Sun Y et al (2023) Internal ion transport in ionic 2D CuInP 2 S 6 enabling multi-state neuromorphic computing with low operation current. Mater Today 66:9–16. 10.1016/j.mattod.2023.04.013 Zhu H et al (2023) Highly Tunable Lateral Homojunction Formed in Two-Dimensional Layered CuInP 2 S 6 via In-Plane Ionic Migration. ACS Nano 17:1239–1246. 10.1021/acsnano.2c09280 Zhong Z et al (2023) Robust Threshold-Switching Behavior Assisted by Cu Migration in a Ferroionic CuInP 2 S 6 Heterostructure. ACS Nano 17:12563–12572. 10.1021/acsnano.3c02406 Chen C et al (2022) Large-Scale Domain Engineering in Two-Dimensional Ferroelectric CuInP 2 S 6 via Giant Flexoelectric Effect. Nano Lett 22:3275–3282. 10.1021/acs.nanolett.2c00130 Huang Y et al (2023) Cu + Migration and Resultant Tunable Rectification in CuInP 2 S 6 . ACS Appl Electron Mater 5:5625–5632. 10.1021/acsaelm.3c00973 Jiang X et al (2023) Strong Piezoelectricity and Improved Rectifier Properties in Mono- and Multilayered CuInP 2 S 6 . Adv Funct Mater 33. 10.1002/adfm.202213561 Jiang X et al (2022) Manipulation of current rectification in van der Waals ferroionic CuInP 2 S 6 . Nat Commun 13. 10.1038/s41467-022-28235-6 Liu Y et al (2023) Stress and Curvature Effects in Layered 2D Ferroelectric CuInP 2 S 6 . ACS Nano 17:22004–22014. 10.1021/acsnano.3c08603 Ming W et al (2022) Flexoelectric engineering of van der Waals ferroelectric CuInP 2 S 6 . Sci Adv 8. 10.1126/sciadv.abq1232 Zhang Q, Zhang LY, Jin CH, Wang YM, Lin F, CalAtom (2019) A software for quantitatively analysing atomic columns in a transmission electron microscope image. Ultramicroscopy 202:114–120. 10.1016/j.ultramic.2019.04.007 Dabov K, Foi A, Katkovnik V, Egiazarian K (2007) Image denoising by sparse 3-D transform-domain collaborative filtering. IEEE Trans Image Process 16:2080–2095. 10.1109/tip.2007.901238 Shibata N et al (2017) Electric field imaging of single atoms. Nat Commun 8. 10.1038/ncomms15631 Seki T, Ikuhara Y, Shibata N (2021) Toward quantitative electromagnetic field imaging by differential-phase-contrast scanning transmission electron microscopy. Microscopy (Oxf) 70:148–160. 10.1093/jmicro/dfaa065 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryD.docx Cite Share Download PDF Status: Published Journal Publication published 22 Nov, 2024 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4487714","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":308548810,"identity":"405e648f-b4bc-4fa7-8c47-1a24b3d6579e","order_by":0,"name":"Wanbiao Hu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2klEQVRIiWNgGAWjYBACAxDB2CDBwMDewABhE6+F5wBpWoCERAKRWswlcgw/F+6wkOeXfGP2mIfBRnbDAeZnD/BpsZyRYyw984yE4czZOebGPAxpxhsOsJkb4HXYjdwN0rxtEowbbueYSfMwHE7ccICHTYKAls2/gVrsN9w8A9Lynygt20C2JG64wQPScoAILWfef7PmPSORPLMnrUxyjkGy8czDbGb4tRxPS77Nu6POtp/98DaJNxV2sn3Hm5/h1YIEOAwg0cRMpHogYH9AvNpRMApGwSgYUQAAgPpF+25KWWYAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0003-0180-9648","institution":"Yunnan University","correspondingAuthor":true,"prefix":"","firstName":"Wanbiao","middleName":"","lastName":"Hu","suffix":""},{"id":308548811,"identity":"562f9a90-e1a6-4325-b234-74c8cab0c89e","order_by":1,"name":"Changjin Guo","email":"","orcid":"","institution":"Yunnan University","correspondingAuthor":false,"prefix":"","firstName":"Changjin","middleName":"","lastName":"Guo","suffix":""},{"id":308548812,"identity":"aff06433-b754-487e-a627-16528b96ae7c","order_by":2,"name":"Xiali Liang","email":"","orcid":"","institution":"Yunnan University","correspondingAuthor":false,"prefix":"","firstName":"Xiali","middleName":"","lastName":"Liang","suffix":""},{"id":308548813,"identity":"5a2fe9dc-34a4-43e3-abf3-49070ce439cd","order_by":3,"name":"Jiyang Xie","email":"","orcid":"","institution":"Yunnan University","correspondingAuthor":false,"prefix":"","firstName":"Jiyang","middleName":"","lastName":"Xie","suffix":""},{"id":308548814,"identity":"96c8ebe9-8881-491e-9372-1eb486ac5bbf","order_by":4,"name":"Caifu Wen","email":"","orcid":"","institution":"Yunnan University","correspondingAuthor":false,"prefix":"","firstName":"Caifu","middleName":"","lastName":"Wen","suffix":""},{"id":308548815,"identity":"88f58e62-0edd-44d8-906c-98112943446a","order_by":5,"name":"Chengding Gu","email":"","orcid":"","institution":"Yunnan University","correspondingAuthor":false,"prefix":"","firstName":"Chengding","middleName":"","lastName":"Gu","suffix":""}],"badges":[],"createdAt":"2024-05-28 03:00:07","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4487714/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4487714/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-024-54229-7","type":"published","date":"2024-11-22T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":57518282,"identity":"cda730ea-16e3-4b39-a16a-91ecea81cd32","added_by":"auto","created_at":"2024-05-31 20:30:37","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":308613,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSchematic illustration of field-correlated structural origin and methodologies of 2D ferroelectric CuInP\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003eS\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e6 \u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e(CIPS). \u003c/strong\u003e(a) Four possible types of Cu(I) ion occupations, \u003cem\u003ei.e.\u003c/em\u003e lattice sites (I, III), interstitial site (II) and interlayer site (IV). (b) Two modes (OOP and IP) for Cu(I) ion migration under electric fields. The purple-dotted arrows represent the possible Cu(I)-migration paths, while the red arrows denote the polarization. (c) Flexoelectricity induced ferroelectric polarizations (\u003cstrong\u003eP\u003c/strong\u003e\u003csub\u003eup\u003c/sub\u003e and \u003cstrong\u003eP\u003c/strong\u003e\u003csub\u003edown\u003c/sub\u003e) that are correlated with Cu(I) ion deflection. (d) Proposed (100)-plane view of atomic-level observation for Cu(I) migration dynamics under the electron-beam irradiation, which allows for monitoring the Cu(I) migration with out-of-(100) plane, \u003cem\u003ee.g. \u003c/em\u003ealong \u003cem\u003e\u003cstrong\u003ec\u003c/strong\u003e\u003c/em\u003e direction.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4487714/v1/75ed8e5bc92afded2fcccd8b.png"},{"id":57518281,"identity":"836b43b1-f773-4040-8871-9cad3ba300df","added_by":"auto","created_at":"2024-05-31 20:30:37","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1165565,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAtomic-level imaging and structural correlation of CIPS.\u003c/strong\u003e (a) iDPC-STEM taken along \u003cem\u003e\u003cstrong\u003ea\u003c/strong\u003e\u003c/em\u003e-axis direction. The green line distinguishes the different layers for sliding boundaries. The marked yellow, red, and blue line are the distances of In-In between layers, In-In in layers, and P-P in layers, respectively. Scale bar in (a): 1 nm. (b) Detailed display of Regions \u003cem\u003ei-iii\u003c/em\u003e in Figure 2a. Figure 2b-\u003cem\u003ei\u003c/em\u003e and Figure 2b-\u003cem\u003eii\u003c/em\u003e show two type configurations (Cu\u003csub\u003e2\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e and Cu\u003csub\u003e3\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e) diagram of the CIPS local structure. Figure 2b-\u003cem\u003eiii\u003c/em\u003e highlight the relative position of In ions between layers by orange dotted line. (c-f) Atomic structure of CIPS viewing from different directions.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4487714/v1/02ccf016965e74175c1f3ee9.png"},{"id":57518280,"identity":"2ab95a82-0ac4-4176-a3ee-89b9dcd5729f","added_by":"auto","created_at":"2024-05-31 20:30:37","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1178330,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAtomic distribution/intensity/shift fitting from STEM image\u003c/strong\u003e. (a-c) Result of Atoms by Optimized Multiple-Ellipse Fitting. Scale bars: 1 nm. (a) Raw image. (b) Fitting image. (c) Intensity difference between the fitting image and raw image. (d-e) Elemental Mapping and Statistical Distribution of Elements. (f) Atom shift fitting diagram. The red circles and red dots are on behalf of atomic standard and experimental occupancy (determined by threshold and Merge method), respectively. The center positions of red circles are obtained by abstracting the CIPS of the standard structure (the probability of Cu(I) occupying positions I and III is 50% each) into pixels after two step Fourier transforms. The arrows from the centers of the circles to the dots respond to relative atom shift, and the length of the arrow represents the relative atomic displacement.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4487714/v1/50bee1cda1b4d6d5ab02f509.png"},{"id":57518283,"identity":"2688e681-f956-4dd2-9613-2f04a3e7471a","added_by":"auto","created_at":"2024-05-31 20:30:38","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":731280,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDynamic process and stability for the local structural Cu\u003c/strong\u003e\u003csub\u003e\u003cem\u003e\u003cstrong\u003ex\u003c/strong\u003e\u003c/em\u003e\u003c/sub\u003e\u003cstrong\u003eInP\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003eS\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e6\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e (\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003ex\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e=2-3) configurations during electron-beam irradiation. (a-d) Distribution of elements after electron irradiation. \u003c/strong\u003e(a) Electron irradiates the red box area by a current of 50 pA for 30 seconds, then a full field of view area HAADF is obtained by a current of 33 pA. (b) Its corresponding EDS element mapping, scale bars: 50 nm. (c) HAADF image. Electron irradiation along the arrow direction under the red box region by a current of 50 pA and the trace was made by electron irradiation (d) Its corresponding EDS mapping, scale bars: 500 nm. \u003cstrong\u003e(e-h)\u003c/strong\u003e \u003cstrong\u003eElemental content with time under different currents. \u003c/strong\u003e(e-f) EDS mapping of Cu, In, P, and S with times increase under 30 pA and 72 pA, respectively. (g-h) The percentage content of the four elements in Figure 4a-b with times. The red arrows represent the trend of Cu(I) in the two processes. Scale bars: 200 nm.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4487714/v1/450d69970e096c28abcb1e49.png"},{"id":69684825,"identity":"817fd329-1bee-40d7-8ae7-f10024742a04","added_by":"auto","created_at":"2024-11-23 08:07:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4021751,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4487714/v1/81ac741f-f9a1-4e94-a19b-6ead5c4b107c.pdf"},{"id":57518279,"identity":"3a7d8e58-562a-41a2-a9e0-ba8c508e93f1","added_by":"auto","created_at":"2024-05-31 20:30:37","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":36140189,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryD.docx","url":"https://assets-eu.researchsquare.com/files/rs-4487714/v1/425792e88dc6b5e851043fe1.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Atomic-level direct imaging for Cu(Ⅰ) multiple occupations and migration in 2D ferroelectric CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003e2D Van der Waals ferroelectric materials are emerging with intriguing physical and quantum phenomena for a broad span of applications in \u003cem\u003ee.g.\u003c/em\u003e field-effect transistors (FETs), non-volatile memory, ferroelectric tunnel junctions (FTJs), and beyond. A prominent representative one is 2D monoclinic CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (CIPS). It manifests room-temperature and tunable quadruple-well ferroelectricity,\u003csup\u003e1\u0026ndash;4\u003c/sup\u003e and therefore, many electronic devices based on CIPS have been rapidly developed.\u003csup\u003e5\u0026ndash;9\u003c/sup\u003e With regard to the ferroelectric polarizations, great efforts have been devoted, with having been uncovering the dominated role from the lattice Cu(I) site occupations. To clearly illustrate it, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e demonstrates the Cu(Ⅰ) occupation and polarization characteristics. Cu(Ⅰ) ions are usually located between two InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons (positions I or III in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). The InS\u003csub\u003e6\u003c/sub\u003e skeletons with Cu(Ⅰ) locked inside combine with P-P ion pairs to form a continuous layer. The layers are connected by Van der Waals forces stacking in the ABAB stack, allowing it to reach the low dimensional limit by separating layers from each other experimentally.\u003c/p\u003e \u003cp\u003eThe primitive ferroelectricity of 2D CIPS originates from the spatial inversion symmetry breaking with the Cu(Ⅰ) ions occupying either I or III position (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) where the feasible mobility of Cu(Ⅰ) ions within the in-plane (IP) and out-of-plane (OOP) orientations in CIPS crystal structure also happens. But more crucially, when an external field (\u003cem\u003ee.g.\u003c/em\u003e electric field, stress, thermal, \u003cem\u003eetc\u003c/em\u003e.) is applied, the Cu(I) occupations may switch to align the polarization direction with the electric field direction (\u003cem\u003ee.g.\u003c/em\u003e all the Cu(I) ions switch to position III to make the polarization is up when an upward electric field applied in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). This is related to the Cu(I) migration and similar phenomena have been well explored in CIPS. However, several fundamental questions raise, \u003cem\u003ee.g.\u003c/em\u003e how many possible positions that could accommodate Cu(I) switching or additional migration? Are there any other Cu(I) occupations except the I and III positions (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea)? Whether Cu(I) possibly migrates to occupy the interstitial site (position II in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) or interlayer site (position IV in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) exists or not?\u003csup\u003e10,11\u003c/sup\u003e Definitely, these basically structural and physical pictures are fundamentally important to understanding the ferroelectric nature and polarization switching of CIPS.\u003c/p\u003e \u003cp\u003eThere are many studies on Cu(I) migration and associated mechanisms, mainly including two categories: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) \u003cb\u003eCu(I) ions migration and ferroelectric polarization switching under the external electric field.\u003c/b\u003e As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, both OOP and IP modes (\u003cem\u003ei.e.\u003c/em\u003e applying external electric field styles) have been performed. Under OOP migration path (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb-i), some Cu(I) ions could be occasionally trapped in the vdW gap and even extracted out of the lattice because of the coexistence of polarization switching and ionic conduction.\u003csup\u003e12\u003c/sup\u003e Under IP migration path (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb-ii), Cu(I) ions are proven to move along the electric field by in-situ scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) measurements.\u003csup\u003e13\u003c/sup\u003e The Cu(I) ions migration toward the negative electrode leads to a typical diode-like resistive switching behavior in the homojunction device.\u003csup\u003e14\u003c/sup\u003e Further, the robust threshold-switching behavior assisted by Cu(I) migration is also found in CIPS heterostructure even if using asymmetrical electrodes (\u003cem\u003ee.g.\u003c/em\u003e graphene, Au \u003cem\u003evs.\u003c/em\u003e Cu).\u003csup\u003e15\u003c/sup\u003e To conclude, the Cu(I) ions migration under the external electric field is a universal behavior unaffected by the electrodes. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) \u003cb\u003eFerroelectric domain distributions and Cu(I) ions location under the flexoelectric effect.\u003c/b\u003e Polarization switching based on the flexoelectric effect can effectively avoid triggering longer-range ionic redistribution (Cu(I) ions just move their positions confined between InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons). As depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec, when stress is applied on OOP mode, the free energies for occupying position up and position down are not equal, and Cu(I) ions tend to move and occupy positions in the direction gradient, thus exhibiting polarization (\u003cb\u003eP\u003c/b\u003e\u003csub\u003eup\u003c/sub\u003e and \u003cb\u003eP\u003c/b\u003e\u003csub\u003edown\u003c/sub\u003e in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). The synergistic effect of Cu ion migration and interfacial Schottky barrier controls the current and device performance.\u003csup\u003e16\u0026ndash;21\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eThese above studies combine with the existing cognition inferred from multiple characterizations have played a great role in promoting the understanding of the Cu(I) migration ferroelectric polarization. However, the basically physical picture with showing accurate Cu(I) lattice-site occupations and Cu(I) configurations to the surrounding arrangement in CIPS is far beyond established. In particular, in addition to the primitive Cu(I) lattice positions I and III (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea), whether the positions II and IV (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) that are only predicted theoretically really occur and how they contribute to the observed overall ferroelectricity are also the fundamental questions to be resolved.\u003c/p\u003e \u003cp\u003eHerein, an atomic-level direct imaging to precisely trace the Cu(I) migration and its local dynamics in CIPS is performed through aberration-corrected scanning transmission electron microscopy (AC-STEM). The cleavage plane (100) based direction (\u003cem\u003ei.e.\u003c/em\u003e \u003cb\u003ea\u003c/b\u003e-axis direction) is chosen for the imaging view because the out-of-(100) plane best reflects the atom occupying situation as well as the clear Cu(I) migration along \u003cb\u003ec\u003c/b\u003e direction (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). Accordingly, the configuration and moving dynamics (speed, energy, and structural stability) of Cu(I) ions in the CIPS under an electric field (\u003cem\u003ei.e.\u003c/em\u003e electron beam irradiation) are well monitored, with finding the presence of Cu(I) interstitial occupations \u003cem\u003ei.e.\u003c/em\u003e positions II presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea. The Cu(I) outside the field of view converges towards the irradiation location and crowds in the possible location of Cu(I), eventually forming an unstable Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) local structure.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"2. Results","content":"\u003cp\u003e \u003cb\u003eAtomic-level imaging observation along (100) plane\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe pivotal criterion to get accurate and sufficient atomic-level structural information on Cu(I) occupations is the exposed plane aligning flawlessly with \u003cb\u003ea\u003c/b\u003e-axis, \u003cem\u003ei.e.\u003c/em\u003e (100) cleavage plane, as proposed in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed. For this, the as-grown CIPS single crystal is precisely cut into nanosheet along \u003cb\u003ea\u003c/b\u003e-axis by focused-ion-beam (FIB) technology (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The low-dose integrated differential phase-contrast coupled scanning transmission electron microscopy (iDPC-STEM) is performed to achieve atomic-level imaging. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea shows a dark field (DF) mode image where all the Cu(I)/In/P/S atoms and their positions are clearly seen. Strikingly, the atoms between the InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons are not solitary, but instead there occur obvious multiple lattice sites for Cu(I) occupations, \u003cem\u003ee.g.\u003c/em\u003e Cu(I) ions occupy both position I and III (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-i); Cu(I) ions occupy simultaneously position I, II and III (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-i\u003cem\u003ei\u003c/em\u003e). For convenience, we here refer to the local crystal configuration situation of Region I and Region II by Cu\u003csub\u003e2\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e and Cu\u003csub\u003e3\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e, respectively. It is worth noting that the local structure Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2, 3) only occurs in a tiny region but does not change the overall chemical formula of the CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e single-crystal. The structural configurations with forming local Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2, 3) can be universally observed from different areas or regions in CIPS (Figure S2), which is definitely attributed to the Cu(I) migration and multiple occupations under the electric field.\u003c/p\u003e \u003cp\u003eTo gain the insights on the microscopical results, standard atomic structure and lattice occupations are compared. The theoretical angular/distance relationship between adjacent In ions, as the (001) plane demonstrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec, is \u0026ang;\u003cem\u003eABC\u003c/em\u003e\u0026thinsp;=\u0026thinsp;59.969 \u0026ordm; (~\u0026thinsp;60 \u0026ordm;) and \u003cem\u003eBC\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.6099 nm, respectively. Experimentally, the projected distance between the two nearest In ions in a layer from the (100) plane is 0.5500 nm (highlighted by red line in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea), which yields an actual distance of 0.55/cos(59.969/2)\u0026thinsp;=\u0026thinsp;0.6350 nm according to the geometric relationship. This value is 4.12% larger than the theoretical In-In distance (\u003cem\u003ei.e. BC\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.6099 nm in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec), which suggests the existence of the multiple Cu(I) occupations between the InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons that thus leads to lattice expansion in the \u0026lt;\u0026thinsp;010\u0026thinsp;\u0026gt;\u0026thinsp;direction. The occurrence of the atomic multi-occupation and lattice constant increase is to release the local stress and maintain the structural stability when the electron-beam irradiation continues.\u003c/p\u003e \u003cp\u003eAtomic distance relationships for P-P in one layer and In-In between adjacent two layers are further analyzed. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed, the theoretical P-P distance is 0.2218 nm, almost analogous to the experimental value of 0.2256 nm (highlighted by blue line in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). The theoretical In-In distance between the nearest layers in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee is 0.6812 nm, while the experimentally measured distance is 0.6812 nm (highlighted by yellow line in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). It is worth noting that the lattice constant \u003cb\u003ec\u003c/b\u003e (=\u0026thinsp;1.3524 nm) experimentally determined also matches well the theoretical one (1.3623 nm). These percentage differences, \u003cem\u003ei.e.\u003c/em\u003e 1.7% in P-P, 0.0058% in In-In in layers, and \u0026minus;\u0026thinsp;0.73% in lattice constant \u003cb\u003ec\u003c/b\u003e, respectively, are small enough to be considered no variations in the \u003cb\u003ec\u003c/b\u003e-axis direction. To conclude, the Cu(I) atomic multi-occupations do not cause the out-of-\u003cb\u003eab\u003c/b\u003e-plane structural fluctuations but apparent in-\u003cb\u003eab\u003c/b\u003e-plane structural expansion. Then, what is the structural reason?\u003c/p\u003e \u003cp\u003eThis can be attributed to the sliding mis-arrangements for the CIPS layers. As indicated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, the theoretical In ions between the CIPS layers should be arranged in a straight line from the perspective of the (100) plane, but the microscopical observation clearly exhibits that the In ions in CIPS layers are not situated in a straight line along the \u003cb\u003ec\u003c/b\u003e-axis direction. The array paths of In ions are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-\u003cem\u003eiii\u003c/em\u003e dotted orange lines, which is a broken line, containing lines segment parallel to the \u003cb\u003ec\u003c/b\u003e-axis. Accordingly, the sliding boundary position can be deduced, which is rightly the interface of the two adjacent CIPS layers, as indicated by the green line in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea. This suggests that the smallest unit in which the glide occurs is a single cell, without compromising its double-layer structure. The sliding distance between the layers is carefully confirmed, and its magnitude is 1/6 lattice constant \u003cb\u003eb\u003c/b\u003e. This further confirms that in order to keep the stability of the crystal structure induced by the Cu(I) multiple occupations, the interlayer sliding releases the lattice stress but does not cause changes in lattice constant \u003cb\u003ec\u003c/b\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn addition to the atom positions or occupations, more atomic-level information \u003cem\u003ee.g.\u003c/em\u003e atomic distribution/intensity/shift \u003cem\u003eetc.\u003c/em\u003e are further extracted. To this, precise analysis and transformation are performed on the iDPC-STEM images (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, S3) by optimized Multiple-Ellipse Fitting method with the detailed calculations are seen in Supporting information section (Figures S3-7).\u003csup\u003e22\u003c/sup\u003e The noise reduction image in Figure S3a is obtained by BM3D method and the position of the atoms are determined by threshold method.\u003csup\u003e23\u003c/sup\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea and Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb show respectively the pseudo-color display raw image and fitting image that exhibit quite high matchiness, thus giving a very small difference in intensity, morphology and strength errors (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec, S3) with all the fitting parameters can be obtained in Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Accordingly, the 2D intensity distributions of Cu, In, P, and S based on the elemental mapping and statistical analysis (corresponding to the highlighted area in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed, Figure S8) are deduced, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee, and meanwhile, the distribution of the atom shift is also obtained (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef).\u003c/p\u003e \u003cp\u003eWith regard to the Cu(I) occupation, Cu(I) ions are present between the InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons in the form of Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) local structure. As a matter of fact, the STEM image contrast predominantly depends on the atomic number (Z) and the atom count.\u003csup\u003e24,25\u003c/sup\u003e The intensities of Cu(I) are primarily distributed in 5\u0026ndash;7\u0026times;10\u003csup\u003e6\u003c/sup\u003e and 9\u0026ndash;16\u0026times;10\u003csup\u003e6\u003c/sup\u003e for positions II and I \u003cem\u003e\u0026amp;\u003c/em\u003e III, respectively in the intensity distribution histogram of Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee. These results show that Cu(I) in CIPS have their own filling rules: Under electron irradiation, Cu(I) will preferentially occupy positions I and III to form Cu\u003csub\u003e2\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e local structure. When positions 1 and 3 are filled, Cu(I) then will occupy position II to further form Cu\u003csub\u003e3\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e local structure. These results are very significant for the understanding of the kinetic process of copper ion occupation in CIPS.\u003c/p\u003e \u003cp\u003eThe Cu(I) preferential occupation might induce the atom/ion shift/displacement in CIPS crystal structure, shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef, which favors the understanding of lattice distortion and polarization. There occur two types of the atom shift (the shift vectors are highlighted by yellow arrows). One shifts along the \u0026lt;\u0026thinsp;010\u0026thinsp;\u0026gt;\u0026thinsp;directions ([010] or [0\u0026ndash;10]) distinguished from the middle position, which contributes to the lattice distortion by atomic multiple occupancies between the InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons. The other shifts along \u0026lt;\u0026thinsp;001\u0026thinsp;\u0026gt;\u0026thinsp;direction, which is due to the intrinsic out-of-plane (OOP) polarization in the lattice. Because the atomic shift vector in \u0026lt;\u0026thinsp;001\u0026thinsp;\u0026gt;\u0026thinsp;direction is overall upward, it does not affect the lattice constant of the \u003cb\u003ec\u003c/b\u003e-axis. The fitting result corresponding to the iDPC-STEM image indicates the dependability of the structural evolution process of Cu(I) multiple occupancies between the InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons of CIPS under the electron-beam irradiation along \u0026lt;\u0026thinsp;100\u0026thinsp;\u0026gt;\u0026thinsp;direction.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eDynamic and stability of Cu\u003c/b\u003e \u003csub\u003e \u003cb\u003ex\u003c/b\u003e \u003c/sub\u003e \u003cb\u003eInP\u003c/b\u003e \u003csub\u003e \u003cb\u003e2\u003c/b\u003e \u003c/sub\u003e \u003cb\u003eS\u003c/b\u003e \u003csub\u003e \u003cb\u003e6\u003c/b\u003e \u003c/sub\u003e \u003cb\u003e(\u003c/b\u003e\u003cb\u003ex\u003c/b\u003e\u0026thinsp;\u003cb\u003e=\u0026thinsp;2\u0026ndash;3) local structure\u003c/b\u003e\u003c/p\u003e \u003cp\u003eThe above precise iDPC-STEM imaging and analysis prove the Cu(I) multiple occupations and the structural changes of CIPS under electron irradiation, but the dynamic process and stability for the local structural configurations on Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) remain to be investigated clearly. Herein, a facile strategy is devised to observe the Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) configurations. Specifically, a large current (50 pA) is used to irradiate partial region (red box in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea) for 30 seconds, and EDS mapping is then carried out over a larger area with a small current (33 pA). The Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) configurations can be judged by the fluctuations in elemental contents in the red box. The resulting EDS mapping (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb) shows that there is no clear distinction of element contrast. Another similar experiment by line scanning (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec, d) for element mapping also gives the same conclusion, \u003cem\u003ei.e.\u003c/em\u003e the electron-irradiated area exhibits the uniform elemental distribution. As a matter of fact, the Cu(I) contents (26.01% for Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea and 29.52% for Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec) are much higher than the theoretical value from the standard structure (Table S2 and Table S3), which means the occurrence of the Cu(I) multi-occupancy with forming Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3). However, the Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) is quite unstable because of undergoing the extremely severe lattice distortion, unless continuous injection energy (electron-beam irradiation) maintains Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) configurations. Otherwise, when the energy (electron-beam irradiation) is removed, the Cu(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) would move back to its original position to retain the standard CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e structure (Figure S9). This dynamic process that clearly uncovers locally structural configuration fluctuations \u003cem\u003ei.e.\u003c/em\u003e\u003cimg src=\"data:image/png;base64,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\" height=\"26\" width=\"263\"\u003e is obviously a result of the Cu(I) migration under the electric field behavior.\u003c/p\u003e \u003cp\u003eThe Cu(I) migration dynamics is further investigated by monitoring the elemental fluctuations during the continuous electron-beam irradiations. To figure out the energy dependence or threshold for the Cu(I) migration, the varied current dose rates are performed. Here Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee-h only show the cases of two typical current rates, \u003cem\u003ei.e.\u003c/em\u003e 30 pA and 72 pA. Note that too small dose (\u003cem\u003ee.g.\u003c/em\u003e 2 pA) can not drive the Cu(I) migration (Figure S9) while too large dose makes Cu(I) migration too quick for inaccurate detection. It can be seen that the detected elemental amounts exhibit different changes with increasing the irradiation time. Clearly from both the EDS imaging (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee) and quantitive analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eg), the Cu(I) amount gradually increases with the irradiation time at a low current (30 pA), from initial 25.9% to final 27.09% (irradiation for 30 mins) and keeping unchanged upon longer irradiation in atomic proportion. While, high current irradiation \u003cem\u003ee.g.\u003c/em\u003e 72 pA (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh), might be due to the quite fast Cu(I) migration, does not bring about any apparent fluctuations in elemental distributions that nearly undergo unchanged at about 28%.\u003c/p\u003e \u003cp\u003eThe energy injected for Cu(I) migration induced crystal structure distortion is further quantified, which is referred to the current dose (D) that can be obtained by Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) is 50 e\u003csup\u003e\u0026minus;\u003c/sup\u003e \u0026Aring;\u003csup\u003e\u0026minus;2\u003c/sup\u003e and 120 e\u003csup\u003e\u0026minus;\u003c/sup\u003e \u0026Aring;\u003csup\u003e\u0026minus;2\u003c/sup\u003e for 30 pA (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee) and 72 pA (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef), respectively. Although the energy dose increases doubly, the Cu(I) content is not increased all the way, but finally fixed at about 28%. This is, again, indicative of the Cu(I) migration but the available positions in the lattice sites (\u003cem\u003ei.e.\u003c/em\u003e positions I, II, III, IV illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) for Cu(I) occupation are in a certain level. Even if forming the Cu\u003csub\u003e3\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e structural configuration, Cu(I) has an atomic percentage content of 25%, still less than the value obtained from experimental determinations. One possibility is Cu(I) may occupy the interlayer position IV (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) that has been predicted by the theoretical calculation, but here from the atomic-level microscopy experiment, no direct evidence that Cu(I) could occupy the interlayer position.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"3. Discussion","content":"\u003cp\u003eThe ferroelectricity of CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (CIPS) strongly relies on the Cu(I) atom/ion occupation ordering and dynamics. When Cu(I) only occupies either the position I or position III (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) orderly, the spatial inversion symmetry is broken, thus exhibiting apparent ferroelectricity. Such a state is structurally and thermally stable at room temperature that has been proven by various experimental and theoretical studies. Up to date, there is no doubt on the Cu(I) migration, but the direct evidences especially the atomic-level imaging that demonstrate the specific occupation sites as well as the locally structural configurations under the electric field are not explored previously. Therese questions are fundamental for understanding ferroelectric mechanism and exploiting ferroelectric-based applications.\u003c/p\u003e \u003cp\u003eThis room-temperature ferroelectricity must allow the occurrence of the intrinsic and ordered Cu(I) vacancy (conversely, position III/I \u003cem\u003evs\u003c/em\u003e. occupied I/III), and meanwhile, endow the adjacent Cu(I) a chance for migrating to occupy the Cu(I) vacancy, which is structurally responsible of the Cu(I) migration. When the Cu(I) has migrated to occupy the vacancy, new locally structural configurations would form, with the theoretically possible configurations could be Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;4). Here, we noted that no any traces from our results indicate the Cu(I) could migrate to the interlayer site (Position IV in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea), despite of this possibility has been predicted by John A. Brehm \u003cem\u003eet al.\u003c/em\u003e that Cu(I) will migrates to the interlayer site (position IV in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) and bond with the S atoms in the next layers under the external field to reveal tunable quadruple-well ferroelectric by DFT calculation,\u003csup\u003e4\u003c/sup\u003e so it only leaves the possibility of the Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) that have been observed in our experiment. When the three positions I/II/III (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) are simultaneously occupied (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-\u003cem\u003eii\u003c/em\u003e) with forming local Cu\u003csub\u003e3\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e, or both positions I and III (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) are occupied (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-i, Cu\u003csub\u003e2\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e), these formed locally structural configurations should not exhibit the ferroelectricity because the occupied Cu(I) causes the structural symmetry. The case of the Cu(I) occupies the interstitial site (position II) and one of lattice sites (positions I/III) but not simultaneously occupies vacancy site should also generate/maintain the ferroelectricity because the spatial inversion symmetry is still broken, but unfortunately, it is not observed by current atomic imaging that may be due to the Cu(I) fast migration effect. Anyway, these locally structural configurations exist only at a very tiny scale instead of generating long-range migration, as indicated from the relatively small change in Cu(I) atom proportion (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), which should not influence the ferroelectricity to a large extent.\u003c/p\u003e \u003cp\u003eIn summary, through iDPC-STEM direct imaging on (100)-CIPS, we obtained the atomic-level structure clearly. We confirm that Cu(I) possesses multiple occupations (lattice, vacancy and interstitial sites) and Cu(I) will migrate to positions I, III, and II between the InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons of CIPS to form local Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) structure when electron-beam irradiation along \u0026lt;\u0026thinsp;100\u0026thinsp;\u0026gt;\u0026thinsp;direction. As a result of the Cu(I) multiple-occupation, lattice distortion occurs for stress release to induce layer sliding along the \u003cb\u003eb\u003c/b\u003e-axis direction with generating a sliding size of 1/6 \u003cb\u003eb\u003c/b\u003e axis. The dynamic process of the locally formed Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) configurations monitored as a function of the irradiation times and currents indicates that Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2\u0026ndash;3) is an unstable structure, which just exists by external energy injection (\u003cem\u003ei.e.\u003c/em\u003e electron-beam irradiation), but it will return back to standard CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e immediately when the external energy vanishes. With new findings of Cu(I) multiple occupations and migration dynamics, this work provides the deep insights on the Cu(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) migration origin and ferroelectric mechanism in CIPS, which are significant for guiding device design and performance improvement based on CIPS.\u003c/p\u003e"},{"header":"4. Materials and methods","content":"\u003cp\u003e \u003cstrong\u003eSynthesis of Single Crystal CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e\u003c/strong\u003e \u003cp\u003eThe raw materials and chemicals used in this study were sourced from reputable suppliers. Copper powder (99.9%), Indium powder (99.99%), and Sulfur powder (99.999%) were purchased from Zhongnuo New Material Co., LTD. Phosphorus particle (99.999%) was obtained from Sichuan Khaqikun Technology Co., LTD. I (99.8%) was procured from Macklin Co., LTD.\u003c/p\u003e \u003c/p\u003e \u003cp\u003eThe CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e crystals were synthesized by a chemical vapor transport method. The raw materials Cu, In, P, and S were ground into powder with a mortar for homogenization purposes according to the stoichiometric ratio and iodine was used as the transport agent (50 mg/25 mL). Then the mixture into the bottom of a quartz test tube (inner diameter, 16 mm; length, 30 cm) subjected to evacuation via a mechanical pump and sealed via a Partulab MRVS-1002 system. The temperature for the evaporation and crystallization zones was initially set at 650 ℃ and 750 ℃, respectively, and the chemical reaction was allowed to proceed for 2 days. Following this, the temperature in the evaporation and crystallization zones was adjusted to 700\u0026deg;C and 665\u0026deg;C, respectively. Then the process was continued for 5 days. Finally, cool down to room temperature at a cooling rate of 10 \u0026deg;/h.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eCrystal structure characterizations\u003c/strong\u003e \u003cp\u003eThe CIPS crystal was cut along the (100) plane by Focus Ion Beam (FIB) double-beam electron microscopy (Thermofisher Helios 5 CX) for electron microscopy measurement. A thickness of 5 \u0026micro;m platinum is deposited on the surface of CIPS as a protective layer, leaving 1 \u0026micro;m remaining after cutting. iDPC-STEM imaging were performed on Thermofisher Spectra 300, equipped with Gatan 1069 camera for collecting EDS mapping data.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePlan-view STEM Imaging\u003c/strong\u003e \u003cp\u003eThe integrated differential phase-contrast scanning transmission electron microscopy (iDPC-STEM) imaging was conducted using Cs-corrected (S) TEM (Thermofisher Spectra 300), operated at 300 kV. The collection semi-angles for the high-angle annular dark field (HAADF) STEM images were 51\u0026ndash;200 mrad and the accumulated electron dose D in iDPC-STEM imaging was calculated by\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv id=\"Equ1\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\text{D}=\\frac{I\\times t}{A}$$\u003c/div\u003e \u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eWhere I was the electron-beam current, t was the irradiation time, and A was the area of irradiation. The electron-beam current was 30 pA and 72 pA for Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee and Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef, respectively. The dwell time is 20 \u0026micro;s. The pixel size of Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee, f was 8.67 \u0026Aring;, and the electron dose D is about 50 e\u003csup\u003e\u0026minus;\u003c/sup\u003e \u0026Aring;\u003csup\u003e\u0026minus;2\u003c/sup\u003e and 120 e\u003csup\u003e\u0026minus;\u003c/sup\u003e \u0026Aring;\u003csup\u003e\u0026minus;2\u003c/sup\u003e, respectively.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eAnalysis of atom shift in CIPS\u003c/strong\u003e \u003cp\u003eiDPC-STEM images of CIPS were analyzed by CalAtom Software to extract the atomic position by multiple-ellipse fitting. Atomic position and species are determined by fitting the intensity of the iDPC image. The atomic shift is determined by comparing the fitting results with the standard CIPS structure.\u003c/p\u003e \u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMaisonneuve V, Evain M, Payen C, Cajipe VB, MoliniC P (1995) Room-temperature crystal structure of the layered phase Cu\u003csup\u003eI\u003c/sup\u003eIn\u003csup\u003eIII\u003c/sup\u003eP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. J Alloys Compd 218:157\u0026ndash;164\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBelianinov A et al (2015) CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e Room Temperature Layered Ferroelectric. Nano Lett 15:3808\u0026ndash;3814. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acs.nanolett.5b00491\u003c/span\u003e\u003cspan address=\"10.1021/acs.nanolett.5b00491\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu F et al (2016) Room-temperature ferroelectricity in CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e ultrathin flakes. Nat Commun 7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/ncomms12357\u003c/span\u003e\u003cspan address=\"10.1038/ncomms12357\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrehm JA et al (2019) Tunable quadruple-well ferroelectric van der Waals crystals. Nat Mater 19:43\u0026ndash;48. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41563-019-0532-z\u003c/span\u003e\u003cspan address=\"10.1038/s41563-019-0532-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeumayer SM et al (2020) Alignment of Polarization against an Electric Field in van der Waals Ferroelectrics. Phys Rev Appl 13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1103/PhysRevApplied.13.064063\u003c/span\u003e\u003cspan address=\"10.1103/PhysRevApplied.13.064063\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi P et al (2022) Electrostatic Coupling in MoS\u003csub\u003e2\u003c/sub\u003e/CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e Ferroelectric vdW Heterostructures. Adv Funct Mater 32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/adfm.202201359\u003c/span\u003e\u003cspan address=\"10.1002/adfm.202201359\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaylor PD, Tawfik SA, Spencer MJS (2022) Ferroelectric van der Waals heterostructures of CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e for non-volatile memory device applications. Nanotechnology 34. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1088/1361-6528/aca0a5\u003c/span\u003e\u003cspan address=\"10.1088/1361-6528/aca0a5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi W et al (2022) A Gate Programmable van der Waals Metal-Ferroelectric‐Semiconductor Vertical Heterojunction Memory. Adv Mater 35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/adma.202208266\u003c/span\u003e\u003cspan address=\"10.1002/adma.202208266\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu J et al (2020) High tunnelling electroresistance in a ferroelectric van der Waals heterojunction via giant barrier height modulation. Nat Electron 3:466\u0026ndash;472. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41928-020-0441-9\u003c/span\u003e\u003cspan address=\"10.1038/s41928-020-0441-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaisonneuve V, Cajipe VB (1997) Ferrielectric ordering in lamellar CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. Phys Rev B 56:10860\u0026ndash;10868\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang D et al (2021) Anisotropic Ion Migration and Electronic Conduction in van der Waals Ferroelectric CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. Nano Lett 21:995\u0026ndash;1002. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acs.nanolett.0c04023\u003c/span\u003e\u003cspan address=\"10.1021/acs.nanolett.0c04023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou Z et al (2023) Unconventional polarization fatigue in van der Waals layered ferroelectric ionic conductor CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. Nat Commun 14:8254. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-023-44132-y\u003c/span\u003e\u003cspan address=\"10.1038/s41467-023-44132-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun Y et al (2023) Internal ion transport in ionic 2D CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e enabling multi-state neuromorphic computing with low operation current. Mater Today 66:9\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.mattod.2023.04.013\u003c/span\u003e\u003cspan address=\"10.1016/j.mattod.2023.04.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu H et al (2023) Highly Tunable Lateral Homojunction Formed in Two-Dimensional Layered CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e via In-Plane Ionic Migration. ACS Nano 17:1239\u0026ndash;1246. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acsnano.2c09280\u003c/span\u003e\u003cspan address=\"10.1021/acsnano.2c09280\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhong Z et al (2023) Robust Threshold-Switching Behavior Assisted by Cu Migration in a Ferroionic CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e Heterostructure. ACS Nano 17:12563\u0026ndash;12572. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acsnano.3c02406\u003c/span\u003e\u003cspan address=\"10.1021/acsnano.3c02406\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen C et al (2022) Large-Scale Domain Engineering in Two-Dimensional Ferroelectric CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e via Giant Flexoelectric Effect. Nano Lett 22:3275\u0026ndash;3282. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acs.nanolett.2c00130\u003c/span\u003e\u003cspan address=\"10.1021/acs.nanolett.2c00130\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang Y et al (2023) Cu\u003csup\u003e+\u003c/sup\u003e Migration and Resultant Tunable Rectification in CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. ACS Appl Electron Mater 5:5625\u0026ndash;5632. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acsaelm.3c00973\u003c/span\u003e\u003cspan address=\"10.1021/acsaelm.3c00973\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang X et al (2023) Strong Piezoelectricity and Improved Rectifier Properties in Mono- and Multilayered CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. Adv Funct Mater 33. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/adfm.202213561\u003c/span\u003e\u003cspan address=\"10.1002/adfm.202213561\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang X et al (2022) Manipulation of current rectification in van der Waals ferroionic CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. Nat Commun 13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-022-28235-6\u003c/span\u003e\u003cspan address=\"10.1038/s41467-022-28235-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Y et al (2023) Stress and Curvature Effects in Layered 2D Ferroelectric CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. ACS Nano 17:22004\u0026ndash;22014. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acsnano.3c08603\u003c/span\u003e\u003cspan address=\"10.1021/acsnano.3c08603\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMing W et al (2022) Flexoelectric engineering of van der Waals ferroelectric CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e. Sci Adv 8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/sciadv.abq1232\u003c/span\u003e\u003cspan address=\"10.1126/sciadv.abq1232\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Q, Zhang LY, Jin CH, Wang YM, Lin F, CalAtom (2019) A software for quantitatively analysing atomic columns in a transmission electron microscope image. Ultramicroscopy 202:114\u0026ndash;120. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ultramic.2019.04.007\u003c/span\u003e\u003cspan address=\"10.1016/j.ultramic.2019.04.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDabov K, Foi A, Katkovnik V, Egiazarian K (2007) Image denoising by sparse 3-D transform-domain collaborative filtering. IEEE Trans Image Process 16:2080\u0026ndash;2095. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/tip.2007.901238\u003c/span\u003e\u003cspan address=\"10.1109/tip.2007.901238\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShibata N et al (2017) Electric field imaging of single atoms. Nat Commun 8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/ncomms15631\u003c/span\u003e\u003cspan address=\"10.1038/ncomms15631\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeki T, Ikuhara Y, Shibata N (2021) Toward quantitative electromagnetic field imaging by differential-phase-contrast scanning transmission electron microscopy. Microscopy (Oxf) 70:148\u0026ndash;160. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/jmicro/dfaa065\u003c/span\u003e\u003cspan address=\"10.1093/jmicro/dfaa065\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"ion migration, CuInP2S6, ferroelectric materials, crystal structure","lastPublishedDoi":"10.21203/rs.3.rs-4487714/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4487714/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (CIPS) is an emerging 2D ferroelectric material known for disrupting spatial inversion symmetry due to Cu(I) position switching. Its ferroelectricity strongly relies on the Cu(I) atom/ion occupation ordering and dynamics. Nevertheless, the accurate Cu(I) occupations and correlated migration dynamics under the electric field, which are key to unlocking ferroelectric properties, remain controversial and unresolved. Herein, an atomic-level direct imaging through aberration-corrected scanning transmission electron microscopy is performed to precisely trace the Cu(I) dynamic behaviours under electron-beam irradiation along (100)-CIPS. It clearly demonstrates that Cu(I) possesses multiple occupations, and Cu(I) could migrate to the lattice, vacancy and interstitial sites between the InS\u003csub\u003e6\u003c/sub\u003e octahedral skeletons of CIPS to form local Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e = 2–3) structure. Cu(I) multi-occupations induced lattice stress results in a layer sliding along the \u003cstrong\u003eb\u003c/strong\u003e-axis direction with generating a sliding size of 1/6 \u003cstrong\u003eb\u003c/strong\u003e axis. The Cu\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (\u003cem\u003ex\u003c/em\u003e = 2–3) exists in a type of dynamic structure, only metastable with electron dose over 50 e\u003csup\u003e−\u003c/sup\u003e Å\u003csup\u003e−2\u003c/sup\u003e, thus generating a dynamic process of Cu\u003csub\u003ex\u003c/sub\u003eInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e (x=2-3) ⇌ CuInP\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e6\u003c/sub\u003e, a completely new phenomenon. These findings shed light on the novel mechanism underlying the Cu(I) migration in CIPS, providing crucial insights into the fundamental processes governing its ferroelectric properties.\u003c/p\u003e","manuscriptTitle":"Atomic-level direct imaging for Cu(Ⅰ) multiple occupations and migration in 2D ferroelectric CuInP2S6","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-31 20:30:31","doi":"10.21203/rs.3.rs-4487714/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"25f93a1c-2293-49d1-987b-cbd5d658c74b","owner":[],"postedDate":"May 31st, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":32594633,"name":"Physical sciences/Materials science/Condensed-matter physics/Ferroelectrics and multiferroics"},{"id":32594634,"name":"Physical sciences/Materials science/Nanoscale materials/Two-dimensional materials"}],"tags":[],"updatedAt":"2024-11-23T08:07:01+00:00","versionOfRecord":{"articleIdentity":"rs-4487714","link":"https://doi.org/10.1038/s41467-024-54229-7","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2024-11-22 05:00:00","publishedOnDateReadable":"November 22nd, 2024"},"versionCreatedAt":"2024-05-31 20:30:31","video":"","vorDoi":"10.1038/s41467-024-54229-7","vorDoiUrl":"https://doi.org/10.1038/s41467-024-54229-7","workflowStages":[]},"version":"v1","identity":"rs-4487714","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4487714","identity":"rs-4487714","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}