Skip to main content
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Jun 18;38(42):e73776. doi: 10.1002/adma.73776

Water‐Triggered Domino‐Like Phase Transition in a Molecular Ferroelectric

Chen‐Kai Yang 1,✉, Yong Ai 1, Xian‐Jiang Song 1, Huan‐Huan Chen 1, Xiao‐Gang Chen 1, Yuan‐Yuan Tang 1, Yao Zhang 1, Juanxia Wu 2,3, Bing Han 4, Liming Xie 2,3,✉, Ren‐Gen Xiong 1,✉
PMCID: PMC13410439  PMID: 42316911

ABSTRACT

Domino‐like phase transitions offer a powerful route to propagate structural changes through local stimuli, yet their coupling with ferroelectricity in molecular materials has remained elusive. Here we show that water enables a domino‐like phase transition in the hybrid copper halide Gua3Cu2I5 (Gua+ = CN3H6 +), providing non‐thermal access to a polar molecular ferroelectric. We find that the polar α phase cannot be reached through conventional thermal pathways but instead emerges through a hierarchical transition mechanism. In this mechanism, water first induces a surface solid–liquid–solid conversion via localized dissolution and reprecipitation, which subsequently triggers an interface‐propagated domino‐like transformation from the non‐polar β phase into the ferroelectric α phase in the bulk. This process can be directly visualized through real‐time luminescence contrast and is accompanied by robust, switchable ferroelectric polarization at room temperature. By identifying water as an active mediator rather than a passive environment, our results establish a new pathway for accessing polar phases beyond thermal equilibrium and provide a conceptual framework for coupling domino‐like phase transitions with ferroelectric functionality in molecular materials.

Keywords: domino‐like phase transition, ferroelectricity, molecular ferroelectric, water


Domino‐like phase transitions can amplify local stimuli but rarely couple with ferroelectricity. We show that water triggers a non‐thermal domino‐like transition in Gua3Cu2I5, enabling access to a polar ferroelectric phase inaccessible thermally. A water‐induced surface reconstruction initiates interface‐propagated transformation, yielding robust room‐temperature switchable ferroelectricity.

graphic file with name ADMA-38-e73776-g003.jpg

1. Introduction

Solid‐state structure phase transfer materials (PTMs) exhibit dramatic switching of physical properties in response to various external stimuli such as heat, mechanical force, light, and electricity [1, 2, 3, 4, 5]. These materials are regarded as promising candidates for next‐generation intelligent applications in data storage, sensors, energy conversion, and smart actuators [5]. Among them, ferroelectric materials play an irreplaceable role in non‐volatile memories, piezoelectric sensors, and electrocaloric cooling due to their unique spontaneous polarization and its reversibility under an external electric field [6, 7, 8, 9]. Compared to traditional inorganic ferroelectrics, organic–inorganic hybrid molecular ferroelectrics offer distinct advantages, including structural flexibility, tunability via molecular design, light weight, solution processability, and good biocompatibility [8, 10, 11, 12, 13, 14]. More importantly, their rich dynamic behaviors enable many molecular crystals to respond to external stimuli such as temperature and pressure through structural phase transitions, often accompanied by switching between ferroelectric and paraelectric states [15, 16]. This provides a pathway for manipulating functional properties via external fields. However, achieving controllable modulation of ferroelectricity through novel phase transition mechanisms remains a significant challenge.

Recently, various external fields beyond electric fields have been explored to modulate ferroelectricity [17, 18, 19]. For example, photoswitchable ferroelectrics based on azobenzene or Schiff base derivatives enable optical control of polarization through reversible photoisomerization [20]. However, such photomodulation is typically transient, with the system reverting to its initial state upon light removal, which limits its utility for permanent information storage. In contrast, a unique phenomenon known as the “domino phase transition” has attracted considerable research interest. This phenomenon describes a chain reaction where a local mechanical stimulus or contact with a seed crystal triggers a phase transition front that propagates spontaneously throughout the entire crystal [21, 22, 23, 24, 25]. For instance, Ito et al. pioneered the report of mechanical‐ or seeding‐triggered single‐crystal‐to‐single‐crystal domino transformations in gold(I) isocyanide complexes, accompanied by changes in photoluminescence [26]. More recently, Wu et al. observed a similar crystalline‐phase‐recognition‐induced domino phase transition in hybrid copper halides α/β‐Gua3Cu2I5, alongside luminescence and nonlinear optical (NLO) switching behaviors [27]. While these seminal works demonstrate the potential of domino transitions for optical switching, a critical question remains unexplored, whether this fascinating phenomenon can be coupled with technologically vital ferroelectricity, thereby offering a new paradigm for controllable ferroelectricity.

This gap reflects a more fundamental conceptual challenge. In most ferroelectric materials, polar phases are accessed through thermally driven symmetry breaking under equilibrium conditions [16]. By contrast, domino‐like phase transitions often involve reconstructive processes that propagate far from thermal equilibrium, navigating complex energy landscapes that include metastable intermediates [26]. Molecular materials, in particular, are characterized by rich structural flexibility and competing local minima, making them fertile ground for such non‐thermal pathways. Yet, the mechanisms that enable these pathways to be activated and stabilized remain poorly understood. Notably, ubiquitous species such as water are commonly treated as passive environmental factors, despite their well‐known ability to mediate dissolution, reconstruction, and interfacial processes [28, 29]. Whether such overlooked components can actively enable non‐thermal, propagating phase transitions that access polar states represents an open and largely unexplored question.

Here we address this challenge by demonstrating, in the hybrid copper halide Gua3Cu2I5, that a non‐thermal, water‐triggered pathway can couple domino‐like phase transitions with ferroelectric functionality. We show that water acts as an active mediator, initiating a surface reconstruction process that subsequently triggers an interface‐propagated, domino‐like phase transformation in the bulk. This hierarchical transition pathway provides access to a polar ferroelectric phase that is inaccessible under thermal equilibrium, enabling irreversible and spatially propagating ferroelectric order. By revealing how cooperative phase‐transition dynamics and ferroelectric functionality can be intrinsically linked through a non‐thermal route, our work establishes a new framework for phase‐transition control in molecular materials.

2. Results and Discussion

2.1. Crystal Structures and Phase Transition Behaviors

Single crystals of the β phase of Gua3Cu2I5 were grown by a solvent evaporation method following reported procedures [27]. The phase purity of the as‐prepared samples was confirmed by the excellent agreement between the experimental powder x‐ray diffraction (PXRD) pattern and the simulated pattern at 293 K (Figure S1A). Variable‐temperature single‐crystal x‐ray diffraction measurements reveal that Gua3Cu2I5 is a polymorphic material. Upon heating, two successive phase transitions occur, yielding the γ phase at T c = 362 K and the δ phase at T c = 400 K, respectively (Figure 1F). In contrast, the α phase cannot be directly obtained by thermal cycling. Instead, it can be rapidly induced from β‐phase single crystals through contact‐triggered phase transition, or slowly formed upon prolonged exposure to ambient atmosphere. Powder x‐ray diffraction measurements confirm the phase purity of the β‐to‐α transformation (Figure S1B).

FIGURE 1.

FIGURE 1

Crystal structure of Gua3Cu2I5. (A–D), Packing diagram of Gua3Cu2I5 in α phase (A), γ phase (B), δ phase(C), and β phase (D). The hydrogen atoms of the guanidinium organic cations are omitted for clarity. (E), Reconstructive transformation between zero‐dimensional and one‐dimensional frameworks in Gua3Cu2I5. Polyhedral representations of zero‐dimensional and one‐dimensional frameworks are distinguished and represented in blue and yellow, respectively. (F), The triggering method for transitions between different phases.

At room temperature, the α phase crystallizes in the polar space group Fdd2. The inorganic [Cu2I5]3− units consist of two face‐sharing CuI4 tetrahedra, forming distorted trigonal bipyramids that are spatially isolated by ordered guanidinium cations, giving rise to a zero‐dimensional framework (Figure 1A). Notably, the terminal iodine ions in each trigonal bipyramid are collectively displaced toward the same direction along the c axis, which serves as the structural origin of the spontaneous polarization in the α phase. Upon heating, the packing of the guanidinium cations becomes progressively disordered, driving two consecutive structural phase transitions into the γ phase (Figure 1B) and the δ phase (Figure 1C), respectively. Upon cooling from the δ phase back to room temperature, the cations recover an ordered arrangement; however, the system does not revert to the original α phase. Instead, a distinct β phase is obtained (Figure 1D), indicating an irreversible thermodynamic phase transition.

The irreversibility of this phase evolution originates from a reconstructive transformation of the inorganic [Cu2I5]3− framework. Specifically, the zero‐dimensional configuration in the α phase transforms into a one‐dimensional chain‐like framework in the γ phase, which is retained upon further cooling (Figure 1E,F). This reconstruction involves the rotation of half of the trigonal bipyramids and extensive breaking and reformation of Cu─I bonds, ultimately yielding one‐dimensional chains composed of CuI4 tetrahedra connected through both edge‐ and corner‐sharing motifs (Figure 1E; Figure S2). The formation of the one‐dimensional framework leads to pronounced changes in interchain spacing and lattice parameters, as summarized in Figure S3 and Table S1. Despite this thermodynamic irreversibility, the β phase can be converted back to the α phase through direct contact with an α‐phase crystal, a process reminiscent of a domino‐like phase transition (Figure 1F) [27].

The thermodynamic phase transitions among the α, γ, δ, and β phases were further verified by differential scanning calorimetry (DSC). During the heating process, the phase transitions from α to γ and γ to δ occur at 362 and 400 K, respectively. Upon cooling, the transitions from δ to γ and from γ to β take place at 391 and 307 K, respectively (Figure 2A). The notably large endothermic peak associated with the α‐to‐γ transition reflects the occurrence of a substantial structural reorganization. These phase transitions are further corroborated by temperature‐dependent changes in the real part of the dielectric constant (Figure 2B). In addition, variable‐temperature second‐harmonic generation (SHG) measurements confirm the symmetry change from the polar α phase to the centrosymmetric γ phase (Figure 2C). The relatively dispersed packing and weak interactions between the cations and the discrete trigonal bipyramids in the polar α phase of Gua3Cu2I5 collectively underpin its propensity for drastic reconstructive phase transitions and create favorable conditions for electric‐field‐induced polarization switching. Consequently, the ferroelectricity of the polar α phase is unambiguously confirmed by a well‐saturated polarization‐voltage (P–V) hysteresis loop (Figure 2D), yielding a remanent polarization of approximately 1.2 µC/cm2, which is comparable to those of some typical representative molecular ferroelectrics, such as ammonium metal formates [NH4][M(HCOO)3] (M = Zn, Mn, Co, Fe, Ni; 0.97–2.2 µC/cm2) [30, 31] and layered lead iodide perovskite ferroelectric, (4,4‐DFHHA)2PbI4 (1.1 µC/cm2) [32]. This value is higher than those of classical Rochelle salt (0.25 µC/cm2) [33], Cu‐based hybrid molecular ferroelectric [DFCBA]2CuCl4 (0.29 µC/cm2) [34] crown‐ether based ferroelectric Ca(NO3)2(15‐crown‐5) (0.55 µC/cm2) [35], and the recently developed 2D van der Waals molecular ferroelectric sodium 2,2,3,3‐tetrafluoropropionate (2,2,3,3‐TFPS; 0.5 µC/cm2) [36], but lower than that of the organic–inorganic perovskite TMCM‐MnCl3 (4.0 µC/cm2) [37] and organic ferroelectric crystal HOCH2(CF2)3CH2OH (6.22 µC/cm2) [8]. Figure S4 exhibits the P–V hysteresis loops under different testing frequencies of 0.03, 0.02, and 0.01 Hz, respectively, and the coercive voltage values exhibit a decreasing trend, and the remanent polarization values increase slightly as frequency descends, which is consistent with the general expectation for a ferroelectric that a longer testing period allows more sufficient time for the polarization reversal process. Noteworthy, to avoid large leakage current due to the hygroscopicity and semiconducting feature of Gua3Cu2I5, the P–V hysteresis loops were measured in the glove box filled with nitrogen atmosphere, and the temperature was controlled to be 233 K. Under such conditions, the leakage current of the Gua3Cu2I5 film sample has been recorded as shown in Figure S5. The I‐V curve is almost linear, and the electrical resistivity (ρ) of Gua3Cu2I5 could thus be roughly estimated to be 3.73 × 1010 Ω·cm, close to the insulator range. Thus, the leakage current is not that large, satisfying the requirement of ferroelectric characterization. Furthermore, according to the Aizu's rule [38], the phase transition between β and γ phases belongs to the mmmF2/m‐type ferroelastic phase transition, as evidenced by the observed alternating bright and dark stripe ferroelastic domains under a polarizing microscope (Figure S6). The interesting thing is that through a domino‐like phase transition, the transition from ferroelasticity to ferroelectricity of the Gua3Cu2I5 crystals can be further achieved (Figure S7), with the clear ferroelastic domains vanishing after the phase transition. Unlike conventional inorganic ferroelectrics such as BaTiO3 and PZT, which possess rigid lattices that preclude any domino‐like mechanism, molecular‐based ferroelectrics feature flexible lattices that enable diverse phase transition mechanisms, including order‐disorder [10] and reconstructive transitions [39]. Recently, although several reports have documented domino‐like phase transitions in molecular or hybrid materials, none of them have investigated the associated ferroelectricity or ferroelasticity, let alone demonstrated a process that converts a ferroelastic state into a ferroelectric state via a domino‐like transition [22, 24, 26, 27]. To the best of our knowledge, this transition from ferroelasticity to ferroelectricity via a domino‐like phase transition observed in Gua3Cu2I5 crystals represents an unprecedented case in the field of ferroelectric materials.

FIGURE 2.

FIGURE 2

Phase transitions and ferroelectricity of Gua3Cu2I5. (A) DSC curves upon heating and cooling runs. (B) Real part (ɛ') of the dielectric permittivity at 1 MHz for Gua3Cu2I5 as a function of temperature recorded in a glovebox. (C) Temperature‐dependent SHG response for Gua3Cu2I5. (D) Polarization‐voltage hysteresis loop of Gua3Cu2I5 measured by the double‐wave method.

2.2. Gua3Cu2I5 Ferroelectrics

We further utilized piezoresponse force microscopy (PFM) to comprehensively characterize the ferroelectric domains, domain switching behavior, and piezoelectric properties of the α‐Gua3Cu2I5 film. One of the defining characteristics of ferroelectric materials is the existence of domain structures. PFM allows for non‐destructive visualization and manipulation of ferroelectric domains, as well as direct measurement of local physical properties. In Figure 3A–C, we present the PFM amplitude and phase images of the α‐Gua3Cu2I5 thin film in vertical mode, along with the topography image associated with the PFM measurements. In the phase image, contrasting colors indicate distinct polarization orientations, thereby delineating different domains. The amplitude signal reflects the magnitude of the local piezoelectric response. Since the polarization directions on both sides of a domain wall are opposite, the piezoelectric response at the domain wall is minimal, appearing as red lines in the amplitude images. There is no correlation between the domain structure and the surface topography, which rules out topographic crosstalk. PFM can also apply a conductive probe to apply a local electric field to the ferroelectric material for manipulating its polarization state, thereby verifying its polarization switching. We first performed Switching Spectroscopy PFM (SS‐PFM) measurements on the thin film sample. As shown in Figure 3D, the characteristic butterfly‐shaped amplitude loops and well‐defined square‐shaped phase loops provide clear evidence of reversible polarization switching.

FIGURE 3.

FIGURE 3

PFM measurement of α‐Gua3Cu2I5. (A–C) Topology image of Gua3Cu2I5 film (A) and the corresponding vertical PFM amplitude (B) and phase images (C). (D) Phase‐voltage hysteresis loop and amplitude‐voltage butterfly loop. (E, F) PFM phase image of Gua3Cu2I5 film showing a box‐in‐box amplitude (E) and domain pattern (F), which was obtained by successively applying a tip voltage of +100 V and −100 V to a region with an initial single‐domain state. Scale bars: A–C, 100 nm; E, F, 1 µm.

Furthermore, we investigated domain switching in the Gua3Cu2I5 film under an electric field. Figure S8A–C shows the initial PFM images along with the corresponding topography image of a 30 µm × 30 µm region. There is a multi‐domain structure in this area. We then selected the green rectangular area in the lower‐left corner and applied a +100 V voltage to the center of the magnified region. Subsequent PFM imaging results are shown in Figure S8D–F. Upon application of the tip bias, the new domain first nucleated at the grain boundary and then propagated toward the center of the blue domain, where the electric field was applied. Following this, an opposite tip voltage of −100 V was applied to the newly formed red domains, resulting in reversible switching back to the original state (Figure 3E,F; Figure S8G–I). In this case, it can be clearly observed that the domain switching is achieved via the movement of the existing domain walls. Furthermore, the switched domains exhibit superior ferroelectric retention as no degradation after 28 h (Figure S8J–L). These PFM measurements clearly demonstrate that the polarization in α‐Gua3Cu2I5 can be repeatedly switched by an external electric field, confirming the ferroelectric nature of α‐Gua3Cu2I5.

2.3. Domino Phase Transition

The phase transition between the α and β phases is not a simple contact‐induced transformation. Instead, water plays a decisive role in triggering and sustaining the phase transition between the two phases. A series of control experiments and in situ observations (Figure S9) reveal that β‐phase samples undergo phase transition only under ambient air conditions, whereas the transformation is strongly suppressed in an oxygen‐ and moisture‐free glovebox. When bulk samples are transferred into the glovebox environment, the phase transition gradually slows down and eventually terminates, indicating that the presence of water is essential for the phase evolution. Direct spectroscopic evidence for the involvement of water is provided by in situ Raman measurements (Figure S10). For initially dry samples, a pronounced Raman band centered at ∼3400 cm−1, corresponding to adsorbed water [40, 41, 42], gradually emerges upon exposure to ambient atmosphere. Conversely, when a vacuum environment is introduced using a pumping system, this water‐related Raman feature disappears, indicating the reversible adsorption and desorption of water molecules at the sample surface. These observations further corroborate the critical role of water in initiating and mediating the phase transition process.

To directly elucidate the role of water at the phase boundary, the initially induced β phase was investigated by in situ environmental scanning electron microscopy (ESEM), which enables precisely controlled humidity under an oxygen‐free atmosphere (Figure 4B; Figures S11–S16). As the relative humidity (RH) increases, the interface between the α and β phases exhibits a pronounced dissolution‐reprecipitation behavior (Figures S11A–D and S15). Specifically, the β phase absorbs water and undergoes partial dissolution at relative humidities between 20% and 25% RH, no obvious interfacial evolution at 10–20% RH (Figure S15), with the dissolution rate accelerating at higher humidity levels. This humidity‐dependent behavior is further supported by quantitative grain‐boundary migration analysis, where the migration rate increases markedly from 0.23 µm/min at ∼40% RH to 8.04 µm/min at ∼60% RH (Figure S16B,C). Notably, a persistent groove with a characteristic width of approximately ∼1.3 µm is consistently observed between the α and β thin films (Figure S17), indicating that the two phases are not continuously connected in thin‐film samples. Instead, the β phase progressively dissolves while the α phase precipitates, corresponding to a water‐driven solid–liquid–solid phase transition, as schematically illustrated in Figure 4A. These observations provide direct evidence that adsorbed water plays a decisive role in activating and sustaining the interfacial solid–liquid–solid phase‐transition pathway.

FIGURE 4.

FIGURE 4

Water‐driven phase transition with coupled surface solid–liquid–solid conversion and bulk domino‐like propagation. (A) Schematic illustration of the humidity‐driven phase transition mechanism. (B) In situ environmental scanning electron microscopy (ESEM) images were recorded under controlled humidity. (C) Time‐resolved optical imaging of an α/β bulk junction. (D) In situ piezoresponse force microscopy (PFM) mapping across the phase boundary. (E) 2D plot of the position‐dependent Raman spectrum of an α/β bulk junction, confirming the spatially resolved phase evolution. (F) Total energies of the α and β phases obtained from first‐principles calculations, with the relative total energies ▵E = E α— E β = −63.3 meV/f.u. (G) Water‐coordinated initial configuration of the β‐Gua3Cu2I5 framework in the bc plane, atomic configuration showing a water molecule anchored in the vicinity of the Cu1 site, where H2O provides a persistent coordination environment to the Cu center. The local coordination of Cu1, Cu2, Cu3, and the bridging I1 site is highlighted. The dashed arrow indicates the constrained reaction coordinate defined by the Cu1‐I1 distance (d Cu1‐I1), which is initially set to the same value as in the dry and I1‐hydrated configurations. (H) Reaction energy profiles as a function of the Cu‐I distance along the constrained phase‐transition pathway, comparing the dry system (black), H2O coordinated to the Cu1 site (red), and H2O interacting with the migrating iodine atom I1 (blue). The energy reference is set to the initial configuration at d = 4.627 Å. The presence of water markedly lowers the energetic cost of Cu─I bond rearrangement, providing a microscopic origin for the experimentally observed water‐assisted phase transition.

The phase identity and ferroelectric nature of the transformation were further confirmed by in situ piezoresponse force microscopy (PFM) measurements (Figure 4D; Figure S18). The left region of the thin film exhibits nearly zero piezoresponse amplitude, characteristic of the centrosymmetric β phase, whereas the right region shows a pronounced piezoresponse signal, corresponding to the ferroelectric α phase. The discontinuous interface between the two regions further supports a phase transition mediated by dissolution and reprecipitation. This process can therefore be regarded as a “ferroelectric writing” process enabled by water‐assisted phase conversion.

First‐principles calculations provide additional insight into the thermodynamic driving force of the transition. The calculated static total energies indicate that the α phase is energetically more stable than the β phase by approximately 63.5 meV per formula unit at 0 K (Figure 4G). This thermodynamic preference rationalizes why the α phase preferentially precipitates during the solid–liquid–solid transition, even when both phases coexist at the interface.

Distinct from the solid–liquid–solid transition observed in thin films, the phase transition behavior in bulk crystals is more complex. As shown in Figure 4C and Figure S19, in situ observations under ambient conditions reveal a directional, domino‐like phase transition propagating across the bulk sample. To systematically investigate this contact‐induced transformation from the β phase to the α phase, we studied the dynamics of the domino‐type transition using a self‐built fluorescence imaging system. Under 355 nm laser excitation, the two phases can be readily distinguished by their characteristic luminescence, with the α phase exhibiting blue emission and the β phase showing yellow‐green emission. Time‐resolved imaging shows that the phase boundary propagates unidirectionally from the α/β interface into the β phase along specific crystallographic orientations, forming a characteristic angle of approximately 124° (Figure 4C; Figure S20). This angle likely corresponds to the (011) direction of the α phase or the (001) direction of the β phase, indicating a strong crystallographic orientation dependence of the domino‐like phase transition. Notably, an estimation of the propagation velocity reveals that the phase transition in bulk crystals proceeds at a rate of approximately 150.50 µm/min, which is more than two orders of magnitude faster than that observed for the solid–liquid–solid transition in thin films, where the velocity is only about 0.23 µm/min (Figure S16A,B). This pronounced difference in kinetics indicates that, in bulk crystals, the phase transition is dominated by a rapid domino‐like propagation process, while the solid–liquid–solid transition primarily serves as a local trigger at the surface. Together, these observations point to a cooperative transition mechanism in which a water‐assisted solid–liquid–solid process initiates the transformation, followed by a crystallographically guided domino‐like phase transition that governs the bulk propagation.

Consistent with the thin‐film behavior, in situ ESEM observations of single‐crystal surfaces under controlled humidity also reveal a clear solid–liquid–solid transition at the surface. As shown in Figure S21, cross‐sectional SEM of the α/β boundary reveals a surface groove while the interior remains continuous, supporting the coexistence of surface dissolution–reprecipitation and bulk domino‐like propagation. These results indicate a hierarchical phase transition mechanism in bulk materials, while the surface region is dominated by water‐driven solid–liquid–solid conversion, the interior undergoes an interface‐propagated domino‐like phase transition, as summarized in Figure 4A and Figures S21 and S22. To gain deeper insight into the structural evolution during the phase transition, we performed in situ Raman spectroscopy measurements (Figure 4E; Figure S19). In the low‐wavenumber region (25–150 cm−1), associated with Cu‐I vibrations, the emergence of new peaks and pronounced intensity changes during the transition confirm a substantial reorganization of the Cu─I bonding environment at the phase boundary. In contrast, the torsional vibration mode of NH2 at ∼515 cm−1 remains unchanged, indicating that the organic cations retain their structural integrity throughout the process [43]. Combined with crystallographic analysis, these results demonstrate that the phase transition involves the reconstruction of the inorganic [Cu2I5]3− framework from one‐dimensional chains with corner‐ and edge‐sharing connectivity to isolated zero‐dimensional bioctahedral units with face‐sharing connectivity. The breaking of Cu─I bonds at corner‐sharing sites and the migration of Cu ions through I‐I‐I triangular channels are identified as the key microscopic steps underlying this structural reorganization.

To further elucidate the correlation between the domino‐like phase transition and the involvement of water molecules, we constructed the model shown in Figure 4G to explicitly evaluate the role of H2O during the transition process. Constrained reaction‐pathway calculations were carried out within a first‐principles density functional theory framework to probe the microscopic mechanism of Cu─I bond rearrangement and to quantify the effect of water on the associated energy barrier (Figure 4G; Figure S23). In the dry system, the energy increases steeply as the Cu─I bond rearrangement proceeds, indicating a substantial kinetic barrier. In contrast, the presence of water significantly lowers the energy profile along the entire pathway, with the most pronounced reduction observed when H2O interacts directly with the Cu1 atom (Figure 4H). These results demonstrate that water does not merely act as a passive environment but actively participates in the transition by reshaping the local energy landscape. The calculated barrier reduction provides a microscopic explanation for the experimentally observed domino‐like phase transition behavior and further establishes the presence of water as a necessary factor for enabling and sustaining the transition process.

3. Conclusions

In summary, we demonstrate that a domino‐like phase transition can be intrinsically coupled with ferroelectric functionality in the hybrid copper halide Gua3Cu2I5 through a non‐thermal, water‐enabled pathway. The polar α phase exhibits robust room‐temperature ferroelectricity, while the system hosts a rich polymorphic landscape inaccessible through thermal equilibrium alone. Crucially, water acts as an active mediator that initiates a surface solid–liquid–solid reconstruction, which subsequently triggers an interface‐propagated, chain‐reaction transformation from the non‐ferroelectric β phase to the ferroelectric α phase in the bulk. This process enables a directional and non‐volatile propagation of ferroelectric order, accompanied by luminescence change and structural reconstruction of [Cu2I5]3− units, effectively realizing a contact‐driven ferroelectric writing mechanism. By identifying water as a key control parameter for activating domino‐like ferroelectric phase transitions, our work establishes a general framework for accessing polar phases beyond thermal equilibrium and broadens the strategies for phase‐transition control in molecular materials.

4. Experimental Section

4.1. Crystal Growth of β‐Gua3Cu2I5

The single crystals were prepared by the solvent evaporation method. Gua(H2CO3)0.5, CuI, and HI were mixed at a molar ratio of 2:1:3 and dissolved in acetonitrile to form a clear and transparent precursor solution under continuous stirring. The mixture then slowly evaporated at room temperature. Finally, the colorless block crystals with yellow luminescence were obtained.

4.2. Crystal Growth of α‐Gua3Cu2I5

Single crystals of α‐Gua3Cu2I5 were obtained from β‐Gua3Cu2I5 through two complementary routes. Upon exposure of β‐Gua3Cu2I5 crystals to ambient air with moderate humidity for several days, the crystals gradually transformed into the α phase, accompanied by the emergence of blue photoluminescence. Alternatively, α‐Gua3Cu2I5 crystals could be rapidly generated via solid–solid contact between α‐ and β‐phase crystals, where contact with an α‐phase seed triggered a domino‐like phase transition in the β‐phase crystal.

4.3. Thin‐Film Preparation of β‐Gua3Cu2I5 and α‐Gua3Cu2I5

The thin films were prepared through a drop‐casting method. 10 mg of the β‐Gua3Cu2I5 crystals was dissolved in 200 mL acetonitrile to prepare a precursor solution, 20 µL of which was then spread on a UV‐ozone‐treated indium tin oxide (ITO) substrate (1 × 1 cm2). After being left at room temperature for 1 h, a polycrystalline β‐phase film with yellow fluorescence was obtained. The α‐phase film with blue fluorescence was obtained by inducing crystallization through direct contact with an α‐phase seed crystal placed on a β‐phase film.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: adma73776‐sup‐0001‐SuppMat.docx.

ADMA-38-e73776-s001.docx (22.9MB, docx)

Acknowledgements

We gratefully acknowledge the support from the National Natural Science Foundation of China (22488101, 22271131, 22575110, 22201120, U24A20504, and 22475054) and the Beijing Gaoju Project (202504841015).

Contributor Information

Chen‐Kai Yang, Email: yangck@ncu.edu.cn.

Liming Xie, Email: xielm@nanoctr.cn.

Ren‐Gen Xiong, Email: xiongrg@seu.edu.cn.

Data Availability Statement

Research data are not shared.

References

  • 1. Bermúdez‐García J. M., Sánchez‐Andújar M., Castro‐García S., López‐Beceiro J., Artiaga R., and Señarís‐Rodríguez M. A., “Giant Barocaloric Effect in the Ferroic Organic‐inorganic Hybrid [TPrA][Mn(dca)3] Perovskite under Easily Accessible Pressures,” Nature Communications 8, no. 1 (2017): 15715, 10.1038/ncomms15715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Kou Y., Sun K., Luo J., et al., “An Intrinsically Flexible Phase Change Film for Wearable Thermal Managements,” Energy Storage Materials 34 (2021): 508–514, 10.1016/j.ensm.2020.10.014. [DOI] [Google Scholar]
  • 3. Li B., Kawakita Y., Ohira‐Kawamura S., et al., “Colossal Barocaloric Effects in Plastic Crystals,” Nature 567, no. 7749 (2019): 506–510, 10.1038/s41586-019-1042-5. [DOI] [PubMed] [Google Scholar]
  • 4. Tokoro H., Yoshikiyo M., Imoto K., et al., “External Stimulation‐controllable Heat‐storage Ceramics,” Nature Communications 6, no. 1 (2015): 7037, 10.1038/ncomms8037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Usman A., Xiong F., Aftab W., Qin M., and Zou R., “Emerging Solid‐to‐Solid Phase‐Change Materials for Thermal‐Energy Harvesting, Storage, and Utilization,” Advanced Materials 34, no. 41 (2022): 2202457. [DOI] [PubMed] [Google Scholar]
  • 6. Asadi K., De Leeuw D. M., De Boer B., and Blom P. W., “Organic Non‐volatile Memories from Ferroelectric Phase‐separated Blends,” Nature Materials 7, no. 7 (2008): 547–550, 10.1038/nmat2207. [DOI] [PubMed] [Google Scholar]
  • 7. Ma Y.‐B., Novak N., Koruza J., Yang T., Albe K., and Xu B.‐X., “Enhanced Electrocaloric Cooling in Ferroelectric Single Crystals by Electric Field Reversal,” Physical Review B 94, no. 10 (2016): 100104, 10.1103/PhysRevB.94.100104. [DOI] [Google Scholar]
  • 8. Zhang H.‐Y., Tang Y.‐Y., Gu Z.‐X., et al., “Biodegradable Ferroelectric Molecular Crystal with Large Piezoelectric Response,” Science 383, no. 6690 (2024): 1492–1498. [DOI] [PubMed] [Google Scholar]
  • 9. Zhang Y., Jie W., Chen P., Liu W., and Hao J., “Ferroelectric and Piezoelectric Effects on the Optical Process in Advanced Materials and Devices,” Advanced Materials 30, no. 34 (2018): 1707007, 10.1002/adma.201707007. [DOI] [PubMed] [Google Scholar]
  • 10. Pan Q., Gu Z.‐X., Zhou R.‐J., et al., “The Past 10 Years of Molecular Ferroelectrics: Structures, Design, and Properties,” Chemical Society Reviews 53, no. 11 (2024): 5781–5861. [DOI] [PubMed] [Google Scholar]
  • 11. Xu Z. K., Song X. J., Qin Y., et al., “Room‐Temperature Metal‐like Ferroelectric Molecular Crystal with Malleability and Ductility,” Advanced Materials 38, no. 2 (2026): 10218, 10.1002/adma.202510218. [DOI] [PubMed] [Google Scholar]
  • 12. Chen H. H., Chen X. G., Xu Z. K., et al., “A 3D Hybrid Perovskite Ferroelastic with Triclinic‐to‐Cubic Phase Transition Boosts Temperature/Pressure Dual On/Off Switchable Birefringence,” Angewandte Chemie International Edition 64, no. 24 (2025): 202503681, 10.1002/anie.202503681. [DOI] [PubMed] [Google Scholar]
  • 13. Liao W.‐Q., Zhao D., Tang Y.‐Y., et al., “A Molecular Perovskite Solid Solution with Piezoelectricity Stronger than Lead Zirconate Titanate,” Science 363, no. 6432 (2019): 1206–1210, 10.1126/science.aav3057. [DOI] [PubMed] [Google Scholar]
  • 14. Jiang H.‐H., Song X.‐J., Lv H.‐P., Chen X.‐G., Xiong R.‐G., and Zhang H.‐Y., “Observation of Ferroelectric Lithography on Biodegradable PLA Films,” Advanced Materials 36, no. 8 (2024): 2307936. [DOI] [PubMed] [Google Scholar]
  • 15. Sato O., “Dynamic Molecular Crystals with Switchable Physical Properties,” Nature Chemistry 8, no. 7 (2016): 644–656, 10.1038/nchem.2547. [DOI] [PubMed] [Google Scholar]
  • 16. Shi P.‐P., Tang Y.‐Y., Li P.‐F., et al., “Symmetry Breaking in Molecular Ferroelectrics,” Chemical Society Reviews 45, no. 14 (2016): 3811–3827, 10.1039/C5CS00308C. [DOI] [PubMed] [Google Scholar]
  • 17. Zhang Y., Du S., Liu X., et al., “Low‐force Pulse Switching of Ferroelectric Polarization Enabled by Imprint Field,” Nature Communications 16, no. 1 (2025): 5271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Wang B. Y., He X., Luo J. J., et al., “Ultralow‐pressure Mechanical‐motion Switching of Ferroelectric Polarization,” Science Advances 11, no. 18 (2025): adr5337, 10.1126/sciadv.adr5337. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Chen H. H., Peng H., Xu Z. K., et al., “Spatial Symmetry Operation Breaking Sulfur‐Chiral Photoswitchable Ferroelectrics,” Journal of the American Chemical Society 147, no. 12 (2025): 10715–10723. [DOI] [PubMed] [Google Scholar]
  • 20. Tang Y.‐Y., Liu J.‐C., Zeng Y.‐L., et al., “Optical Control of Polarization Switching in a Single‐Component Organic Ferroelectric Crystal,” Journal of the American Chemical Society 143, no. 34 (2021): 13816–13823, 10.1021/jacs.1c06108. [DOI] [PubMed] [Google Scholar]
  • 21. Ogawa T., “Domino Mechanisms in Photoinduced Phase Transitions,” Phase Transitions 74, no. 1‐2 (2001): 93–128, 10.1080/01411590108224566. [DOI] [Google Scholar]
  • 22. Seki T., Sakurada K., and Ito H., “Controlling Mechano‐ and Seeding‐Triggered Single‐Crystal‐to‐Single‐Crystal Phase Transition: Molecular Domino with a Disconnection of Aurophilic Bonds,” Angewandte Chemie International Edition 52, no. 49 (2013): 12828–12832, 10.1002/anie.201307672. [DOI] [PubMed] [Google Scholar]
  • 23. Wu H. L., Wang J. B., Zhao C. B., et al., “Domino‐like Magnetic Phase Transition Induced by a Bias Voltage in a FeRh Thin Film,” Physical Review B 111, no. 21 (2025): 214414. [Google Scholar]
  • 24. Zhang Z. Z., Lin Y. P., Jin J. C., et al., “Crystalline‐Phase‐Recognition‐Induced Domino Phase Transition and Luminescence Switching for Advanced Information Encryption,” Angewandte Chemie International Edition 60, no. 43 (2021): 23373–23379, 10.1002/anie.202110088. [DOI] [PubMed] [Google Scholar]
  • 25. Liu J., Wu H. C., Chen B., et al., “Lighting up the Non‐Conjugated Polymers in Full Color via Domino‐Effect Triggered Emission,” Advanced Materials 38, no. 6 (2025): 10125, 10.1002/adma.202510125. [DOI] [PubMed] [Google Scholar]
  • 26. Ito H., Muromoto M., Kurenuma S., et al., “Mechanical Stimulation and Solid Seeding Trigger Single‐crystal‐to‐single‐crystal Molecular Domino Transformations,” Nature Communications 4, no. 1 (2013): 2009. [DOI] [PubMed] [Google Scholar]
  • 27. Wu J., Guo Y., Qi J. L., et al., “Multi‐Stimuli Responsive Luminescence and Domino Phase Transition of Hybrid Copper Halides with Nonlinear Optical Switching Behavior,” Angewandte Chemie 135, no. 18 (2023): 202301937. [DOI] [PubMed] [Google Scholar]
  • 28. Wang Y.‐H., Zheng S., Yang W.‐M., et al., “In Situ Raman Spectroscopy Reveals the Structure and Dissociation of Interfacial Water,” Nature 600, no. 7887 (2021): 81–85. [DOI] [PubMed] [Google Scholar]
  • 29. Wu L. Q., Huang W. X., Li D. Y., et al., “Unveiling the Structure and Dissociation of Interfacial Water on RuO2 for Efficient Acidic Oxygen Evolution Reaction,” Angewandte Chemie 137, no. 1 (2025): 202413334. [DOI] [PubMed] [Google Scholar]
  • 30. Xu G.‐C., Ma X.‐M., Zhang L., Wang Z.‐M., and Gao S., “Disorder−Order Ferroelectric Transition in the Metal Formate Framework of [NH4][Zn(HCOO)3],” Journal of the American Chemical Society 132, no. 28 (2010): 9588–9590, 10.1021/ja104263m. [DOI] [PubMed] [Google Scholar]
  • 31. Xu G.‐C., Zhang W., Ma X.‐M., et al., “Coexistence of Magnetic and Electric Orderings in the Metal–Formate Frameworks of [NH4][M(HCOO)3],” Journal of the American Chemical Society 133, no. 38 (2011): 14948–14951, 10.1021/ja206891q. [DOI] [PubMed] [Google Scholar]
  • 32. Chen X.‐G., Song X.‐J., Zhang Z.‐X., et al., “Confinement‐Driven Ferroelectricity in a Two‐Dimensional Hybrid Lead Iodide Perovskite,” Journal of the American Chemical Society 142, no. 22 (2020): 10212–10218, 10.1021/jacs.0c03710. [DOI] [PubMed] [Google Scholar]
  • 33. Valasek J., “Piezo‐Electric and Allied Phenomena in Rochelle Salt,” Physical Review 17, no. 4 (1921): 475–481, 10.1103/PhysRev.17.475. [DOI] [Google Scholar]
  • 34. Huang C.‐R., Luo X., Chen X.‐G., Song X.‐J., Zhang Z.‐X., and Xiong R.‐G., “A Multiaxial Lead‐free Two‐dimensional Organic‐inorganic Perovskite Ferroelectric,” National Science Review 8, no. 5 (2021): nwaa232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Ye H. Y., Zhang Y., Fu D. W., and Xiong R. G., “A Displacive‐Type Metal Crown Ether Ferroelectric Compound: Ca(NO3)2 (15‐crown‐5),” Angewandte Chemie International Edition 53, no. 26 (2014): 6724–6729, 10.1002/anie.201402339. [DOI] [PubMed] [Google Scholar]
  • 36. Chen X. G., Qin Y., Lv H. P., et al., “A 2D Van Der Waals Molecular Ferroelectric,” Advanced Materials 38, no. 4 (2026): 07450, 10.1002/adma.202507450. [DOI] [PubMed] [Google Scholar]
  • 37. You Y.‐M., Liao W.‐Q., Zhao D., et al., “An Organic‐inorganic Perovskite Ferroelectric with Large Piezoelectric Response,” Science 357, no. 6348 (2017): 306–309. [DOI] [PubMed] [Google Scholar]
  • 38. Aizu K., “Possible Species of “Ferroelastic” Crystals and of Simultaneously Ferroelectric and Ferroelastic Crystals,” Journal of the Physical Society of Japan 27, no. 2 (1969): 387–396, 10.1143/JPSJ.27.387. [DOI] [Google Scholar]
  • 39. Chen H. H., Qin Y., Zhang Y., et al., “3D Chiral Perovskitoid‐to‐Perovskite Reconstructive Phase Transition Designed Molecular Ferroelectric Thin‐Film with Out‐of‐Plane Polarization,” Advanced Materials 38, no. 9 (2026): 20443, 10.1002/adma.202520443. [DOI] [PubMed] [Google Scholar]
  • 40. Socrates G., Infrared and Raman Characteristic Group Frequencies: Tables and Charts (John Wiley & Sons, 2004). [Google Scholar]
  • 41. Kolesov B., “Raman Investigation of H2O Molecule and Hydroxyl Groups in the Channels of Hemimorphite,” American Mineralogist 91, no. 8‐9 (2006): 1355–1362, 10.2138/am.2006.2179. [DOI] [Google Scholar]
  • 42. Frost R. L., Reddy B. J., and Keeffe E. C., “Structure of Selected Basic Copper(II) Sulphate Minerals Based Upon Spectroscopy—Implications for Hydrogen Bonding,” Journal of Molecular Structure 977, no. 1‐3 (2010): 90–99, 10.1016/j.molstruc.2010.05.019. [DOI] [Google Scholar]
  • 43. Drozd M., “Molecular Structure and Infrared Spectra of Guanidinium Cation,” Materials Science and Engineering: B 136, no. 1 (2007): 20–28, 10.1016/j.mseb.2006.08.063. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting File: adma73776‐sup‐0001‐SuppMat.docx.

ADMA-38-e73776-s001.docx (22.9MB, docx)

Data Availability Statement

Research data are not shared.


Articles from Advanced Materials (Deerfield Beach, Fla.) are provided here courtesy of Wiley

RESOURCES