Toward the Targeted Design of Molecular Ferroelectrics: Modifying

2 days ago - ... which are unavailable in their inorganic counterparts, make molecular systems a good platform for manipulating ferroelectricity. Rema...
0 downloads 0 Views 7MB Size
Article Cite This: Acc. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/accounts

Toward the Targeted Design of Molecular Ferroelectrics: Modifying Molecular Symmetries and Homochirality Han-Yue Zhang,† Yuan-Yuan Tang,‡ Ping-Ping Shi,† and Ren-Gen Xiong*,†,‡ †

Jiangsu Key Laboratory for Science and Applications of Molecular Ferroelectrics, Southeast University, Nanjing 211189, People’s Republic of China ‡ Ordered Matter Science Research Center, Nanchang University, Nanchang 330031, People’s Republic of China

Acc. Chem. Res. Downloaded from pubs.acs.org by IDAHO STATE UNIV on 04/15/19. For personal use only.

S Supporting Information *

CONSPECTUS: Although the first ferroelectric discovered in 1920 is Rochelle salt, a typical molecular ferroelectric, the front-runners that have been extensively studied and widely used in diverse applications, such as memory elements, capacitors, sensors, and actuators, are inorganic ferroelectrics with excellent electrical, mechanical, and optical properties. With the increased concerns about the environment, energy, and cost, molecular ferroelectrics are becoming promising supplements for inorganic ferroelectrics. The unique advantages of high structural tunability and homochirality, which are unavailable in their inorganic counterparts, make molecular systems a good platform for manipulating ferroelectricity. Remarkably, based on the Neumann’s principle and the Curie symmetry principle defining the group-tosubgroup relationship, we have found some outstanding high-temperature molecular ferroelectrics, like diisopropylammonium bromide (DIPAB) with a large spontaneous polarization up to 23 μC/cm2 (Fu, D. W.; et al. Science 2013, 339, 425). However, their application potential is severely limited by the uniaxial nature, leading to major issues in finding proper substrates for thinfilm growth and achieving high thin-film performance. Inspired by the commercialized inorganic ferroelectrics like Pb(Zr, Ti)O3 (PZT), where the multiaxial nature contributes greatly to the optimized ferroelectric and piezoelectric performance, developing high-temperature multiaxial molecular ferroelectrics is an imminent task. In this Account, we review our recent research progress on the targeted design of multiaxial molecular ferroelectrics. We first propose the “quasi-spherical theory”, a phenomenological theory based on the Curie symmetry principle, to modify the spherical cations to a low-symmetric quasi-spherical geometry for acquiring the highly symmetric paraelectric phase and the polar ferroelectric phase of multiaxial ferroelectrics simultaneously. Besides the sizes and weights of the cation and anion, the intermolecular interactions are particularly crucial for decelerating the molecular rotation at low temperature to reasonably induce ferroelectricity. It means that the momentums of the cation and anion should be matched, so we describe the “momentum matching theory”. In particular, introducing homochirality, a superiority of molecular materials over the inorganic ones, was demonstrated as an effective approach to increase the incidence of ferroelectric crystal structures. Thanks to the striking chemical variability and structure−property flexibility of molecular materials, our research efforts outlined in this Account have led to and will further motivate the richness and the application exploration of high-temperature, highperformance multiaxial molecular ferroelectrics, along with the implementation and perfection of the targeted design strategies. flexibility, low cost, easy processing, and environmental friendliness. Nevertheless, there is still a long way for them to go from curiosity-driven discoveries to real-life applications. Compared with their inorganic counterparts, whose basic properties are easily optimized by using dopants or modifiers, the advance of molecular ferroelectrics relies on exploring new families. Unfortunately, this seems like looking for a needle in a haystack, due to the lack of effective theories to control and optimize ferroelectricity. After years of being focused on this field, we developed the strategies of distinguishing potential

1. INTRODUCTION Since the first discovery of ferroelectricity in a molecular ferroelectric Rochelle salt in 1920, ferroelectrics have gained widespread attention for their great academic value and practical significance.1,2 However, the ones that shine light in applications like memory elements, capacitors, sensors, and actuators are inorganic ferroelectrics such as BaTiO3 and Pb(Zr, Ti)O3 (PZT), which have exceptional electrical, mechanical, and optical properties.3−5 To address the growing issues in environment, energy, and cost, faced by these inorganic ferroelectrics, the past decade has witnessed the renaissance of molecular ferroelectrics as promising supplements with intrinsic advantages of light weight, mechanical © XXXX American Chemical Society

Received: December 31, 2018

A

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

chemistry”. Considering the structural flexibility and tunability of molecular materials, it leaves room for those nonpolar crystals to be endowed with ferroelectricity under precise molecular modifications or molecular tailoring. Even slight modification of molecular structures may change the crystal symmetry. When we introduced specific chemical groups into the spherical cations, the lowered molecular symmetry and the resultant molecular dipole moment are promising for arousing low-symmetric polar crystal structures and ferroelectric behaviors. As the temperature rises, the modified quasispherical cations still preserve high-speed, high-energy molecular rotations and appear to be spherical in a dynamical disordered, isotropic state, corresponding to the highsymmetric paraelectric phase. Hence, the “quasi-spherical theory” is actually using a chemical perspective to understand the Landau theory of phase transitions and the Curie symmetry principle. Beyond that, another key factor to modulate ferroelectricity should be the matching degree between cations and anions, involving their sizes, weights, and particularly intermolecular interactions, so the “momentum matching theory” is put forward reasonably (see the Supporting Information for details). Specifically, the intermolecular interactions between the cations and the anions are like ropes to decelerate the molecular rotations during the phase transition. From the crystallography, it generally causes a lower symmetry at room temperature, and the consequent striking symmetry breaking is favorable for multiaxial ferroelectricity. These ideas have enabled us to readily construct a series of molecular perovskite ferroelectrics: [Me3NCH2X]MnX3 (X = Cl−, Br−),15,16 [Me3NCH2Cl]CdX3 (X = Cl−, Br−),15,17 [Me3NCH2I]PbI3,18 [Me3NOH]2[KFe(CN)6],19 [Mdabco]RbI3 (Mdabco = N-methyl-N′-1,4diazoniabicyclo[2.2.2]octonium),20 [Mdabco]NH4X3 (X = Br−, I−),21 and [Odabco]NH4X3 (X = Cl−, Br−) (Odabco = N-hydroxy-N′-1,4-diazoniabicyclo[2.2.2]octonium).21 Similar to the structural tunability, homochirality, which has been largely ignored within the scope of ferroelectrics, is also a superiority of molecular systems over the inorganic ones that disallow homochiral centers. Homochiral molecules can form enantiomorphic crystals of the corresponding handedness, whose crystal structures must belong to the 11 chiral point groups: 1 (C1), 2 (C2), 222 (D2), 4 (C4), 422 (D4), 3 (C3), 32 (D3), 6 (C6), 622 (D6), 23 (T), and 432 (O), whereas racemic mixtures that contain equal amounts of molecules of each homochirality may crystallize in nonchiral or even centrosymmetric point groups. Five of the ten polar point groups are chiral (C1, C2, C4, C3, C6), constituting as many as 22 species of chiral-to-chiral ferroelectric phase transitions (Table 1). Apparently, in contrast with the achiral or racemic compounds, it will be easier for homochiral compounds to crystallize in the five chiral-polar point groups, where the probability increases

ferroelectric candidates from the Cambridge Crystallographic Data Centre (CCDC) and designing new systems by taking the Curie symmetry principle and the Neumann principle as the beacon, as fully discussed in our previous work.6 The paraelectric phase falls into the 32 point groups, but the ferroelectric phase must adopt one of the 10 polar point groups, i.e., C1 (1), Cs (m), C2 (2), C2v (mm2), C3 (3), C3v (3m), C4 (4), C4v (4mm), C6 (6), and C6v (6mm). Between them, a group-to-subgroup relationship is defined by the Curie symmetry principle; Hence, Aizu derived 88 species of potential ferroelectric phase transitions (Table S1, Supporting Information), enabling one to initially identify the ferroelectricity.7 Through continuous efforts, we obtained some hightemperature molecular ferroelectrics at a glacial pace, including diisopropylammonium bromide (DIPAB),8 tetraethylammonium perchlorate, 9 (3-pyrrolidinium)MnCl 3 , 10 and so forth.11−14 Notably, those discovered before 2016 are mostly uniaxial, where the ferroelectric space group is a maximal nonisomorphic subgroup of the paraelectric one along the direction of spontaneous polarization (Ps), following a “parent−child” relationship (see the Supporting Information for details). In contrast, the multiaxial ferroelectrics follow a “grandparent−grandchild” relationship, and the greater the symmetry change between paraelectric and ferroelectric phases is, the more equivalent polarization directions will appear. An ideal case of the most polarization directions is the cubic-totriclinic phase transition with an Aizu notation of m3mF1. (3Pyrrolidinium)MnCl3 with the mmmFmm2 phase transition is a typical uniaxial ferroelectric triggered by the face-to-back vibration of the N atom in the pyrrolidinium plane at a low amplitude.10 Not only is there the difficulty of finding suitable substrates for growing high-quality single-crystalline films with preferred orientations, such uniaxial systems also suffer from limited ferroelectric performance of thin films as well as low piezoelectricity. The usability of molecular ferroelectrics can improve, only if they share the benefit of conventional inorganic systems like PZT, whose superior performance and broad applications are inseparable from the multiaxial feature. Consequently, targeted design strategies for developing new multiaxial molecular ferroelectrics are urgently required. To induce the multiaxial ferroelectric phase transition accompanying significant symmetry change, first, a highsymmetric paraelectric crystal structure is particularly desirable. In this way, the spherical cations, such as [Me4N]+ and 1,4diazabicyclo [2.2.2]octonium (dabco), become the preferred building blocks, which can easily undergo dynamic rotation or orientational disorder to facilitate the formation of a highsymmetric, high-temperature paraelectric phase. When the temperature decreases, however, their high symmetry and nonpolar features still favor centrosymmetric crystal structures incompatible with the symmetry requirement of ferroelectric phase. Although symmetry breaking might also occur, the molecular dipole moment still does not appear, which makes it difficult to induce the emergence of spontaneous polarization. Accordingly, how to resolve this contradiction and to acquire the highly symmetric paraelectric phase and the polar ferroelectric phase simultaneously is the crucial problem in the field of multiaxial molecular ferroelectrics. To place our empirical knowledge in a broader setting, herein we first propose the “quasi-spherical theory”, a phenomenological theory aiming at the chemical design of the cations as well as a specific application of the group-tosubgroup relationship in the context of “ferroelectric

Table 1. Twenty-Two Species of Chiral-to-Chiral Ferroelectric Phase Transitions

B

crystal system

Aizu notation

monoclinic orthorhombic tetragonal trigonal hexagonal cubic

2F1 222F1; 222F2 4F1; 422F1; 422F2 (s); 422F4 3F1; 32F1; 32F2; 32F3 6F1; 622F1; 622F2 (s); 622F6 23F1; 23F2; 23F3; 432F1; 432F2 (s); 432F4; 432F3 DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 1. Targeted Design Strategy of 1 and 2a

a

(a) Modification of molecular symmetry; (b) initial structural phase transition; and (c) modified ferroelectric phase transition.

2.1. Design Ferroelectrics by Modifying [Me4N]+

from 10/32 to 5/11, pointing out a rational way to the targeted design of molecular ferroelectrics. As expected, this idea has succeeded in several optically active ferroelectrics using homochiral templating cations: (R)-(−)-3-hydroxlyquinuclidinium, R- or S-3-ammoniopyrrolidinium, and R- or S-3ammonioquinuclidinium. Modifying molecular structure, tuning intermolecular interactions, and constructing the homochiral center enable the targeted design of molecular ferroelectrics as well as effective manipulation of their physical properties. In this Account, we present the design strategies taking “quasispherical theory” and “momentum matching theory” as the main core, for the sake of foreseeing more outstanding candidates and then breathing new life into the century-old but once-fading field of molecular ferroelectrics.

2.1.1. [Me3NCH2Cl]CdX3 (X = Cl− (1) and Br− (2)). While searching the CCDC, we noted that [Me4N]CdCl3 (ref code in CCDC: TMACDC) with BaNiO3-like perovskite structure crystallizes in the centrosymmetric space groups P63/m and P63/mmc at room temperature and 410 K, respectively, which denotes a common structural phase transition.22,23 Because of the lower potential energy barrier of the tumbling motion, the spherical [Me4N]+ cation adopting the point group Td exhibits high-speed rotations in the high-temperature phase (HTP), giving rise to a highly disordered, isotropic state (Scheme 1). Following the temperature down, it turns into an ordered state with the freezing of molecular rotations, while the high molecular symmetry and the absence of molecular dipole moment are responsible for the failure to form polar crystal structure in the room-temperature phase (RTP). The “quasispherical theory” provide us the bright promise of endowing this prototypical system with ferroelectricity. As a result, we replaced one methyl of the [Me4N]+ cation by a chloromethyl, and then assembled the modified quasi-spherical cation with the same one-dimensional (1D) {CdCl3}− chain. As shown in Scheme 1, in analogy to [Me4N]+, the [Me3NCH2Cl]+ cation also rotates at a high speed and manifests an isotropic spherical geometry, so that in the HTP [Me3NCH2Cl]CdCl3 (1) maintains the same high-symmetric space group P63/mmc as that of [Me4N]CdCl3. And the exciting thing is that, as the introduction of electronegative Cl atom facilitates that the molecular symmetry of [Me3NCH2Cl]+ cation is smoothly lowered to C3v and equipped with a molecular dipole moment, the RTP symmetry group gets validly lowered to the polar Cc. With an Aizu notation of 6/mmmFm, 1 is really a multiaxial ferroelectric with 12 equivalent polarization directions.15,24 However, not all modifications of cation and anion lead to multiaxial ferroelectrics. As the “momentum matching theory” says, the matching degree between cation and anion plays a crucial role in modulating multiaxial ferroelectricity. For instance, when one methyl was substituted with an ethyl, [Me3NEt]CdCl3 crystallizes in a centrosymmetric space group Pbca in the RTP and still fails to be a ferroelectric, despite that the molecular symmetry is successfully lowered to C3v. The

2. MODIFICATION OF MOLECULAR STRUCTURES Increasing the incidence of polar crystal structures is the basic approach to realize the symmetry-dependent ferroelectricity, which requires specific molecular organization into welldefined architectures. To move molecular ferroelectrics from rough and casual construction in most of the past works to targeted design, more dedicated research efforts should be first devoted to the choice of molecular building blocks. The spherical cations with large conformational freedom, such as [Me4N]+, dabco, and quinuclidine, have been frequently used as active components in structural phase transitions. In the light of the “quasi-spherical theory”, they were modified into quasi-spherical geometry that is desirable for the multiaxial ferroelectric phase transitions. Then considering that not only the individual molecules and their organization in the crystal lattice but also the intermolecular interactions exert critical impact on the ferroelectric response, the “momentum matching theory” is derived for the balance of cationic and anionic parts. Several examples presented below, encompassing the “chemical chameleons”, the perovskite architecture, highlight the correlations between the ferroelectricity and the molecular structures in terms of shape, size, and intermolecular interactions. C

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 2. Targeted Design Strategy of 3 and 4a

a

(a) Modification of molecular symmetry; (b) initial structural phase transition; and (c) modified ferroelectric phase transition.

Scheme 3. Targeted Design Strategy of 5a

a

(a) Modification of molecular symmetry; (b) initial structural phase transition; and (c) modified ferroelectric phase transition.

mismatch of [Me3NEt]+ cation and {CdCl3}− ligand, between which only weak van der Waals forces exist, eliminates the possibility of regulating the lattice symmetry through intermolecular interactions. In contrast, for compound 1, the [Me3NCH2Cl]···Cl halogen bonding between the cationic and anionic moieties is effective enough to cause the deformation of anionic chain and further contribute to the ferroelectric polarization. Next, we substituted {CdBr3}− ligand for {CdCl3}−, and another molecular ferroelectric isostructural with 1, [Me3NCH2Cl]CdBr3 (2), was achieved.17 Similar to 1, 2 also crystallizes in the hexagonal space group P63/mmc in the HTP, while the [Me3NCH2Cl]···Br halogen bonding between the modified cation and the anionic chain remains a critical point for the lattice deformation. Nevertheless, with increasing the size of anion, the initial balance of cation and anion is gradually broken. The larger {CdBr3}− anion gives rise to the relatively mismatched rotary movement in 2, and therefore, in the RTP, the crystal symmetry gets only lowered to the polar

hexagonal space group P63mc. With an Aizu notation of 6/ mmmF6mm, 2 is classified as a uniaxial ferroelectric, but not a multiaxial one like 1. The situation in the test case of [Me3NCH2Cl]CdI3 is even worse, where the extremely large anion and the stronger [Me3NCH2Cl]···I halogen bonding make it a nonferroelectric material with a centrosymmetric space group Pnma in the RTP. Remarkably, in addition to the reservation of ferroelectricity, the moderate halogen bonding interactions further act as an essential driving force in the attractive piezoelectric performance of 1 and 2, whose piezoelectric constants (d33) are as large as 220 and 139 pC/ N, respectively. These examples illustrate the significance of combining the “quasi-spherical theory” and the “momentum matching theory” in the targeted design of molecular ferroelectrics. 2.1.2. [Me3NCH2X]MnX3 (X = Cl− (3) and Br− (4)). [Me4N]MnCl3, an analogue to [Me4N]CdCl3, has been reported to crystallize in the centrosymmetric space group P63/m (ref code in CCDC: TMAMMN).25 Enlightened by the D

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 4. Targeted Design Strategy of 6a

a

(a) Modification of molecular symmetry; (b) initial structural phase transition; and (c) modified ferroelectric phase transition.

3). Among the family of [Me3NCH2X]+ (X = F, Cl, Br and I), [Me3NCH2I]+ has the largest volume of 182.01 Å3 and contributes to the strongest [Me3NCH2I]···I halogen bonding (2.89 × 10−3 hartree).18 The resultant ordered [Me3NCH2I]+ cation array in [Me3NCH2I]PbI3 (5) is the main reason for its low-symmetric, polar monoclinic space group C2 in the RTP. At 318 K, 5 undergoes a transition to the paraelectric phase adopting the space group C2/m, while the Aizu notation 2/ mF2 means that it is a uniaxial ferroelectric. These systematic works point out a useful avenue for designing high-temperature molecular ferroelectrics, where the phase transition temperature and can be efficiently tuning by the modulation of halogen bonding. 2.1.4. [Me3NOH]2[KFe(CN)6] (6). Beyond halogen atoms permitting halogen bonding, there are many other chemical groups that are capable of inducing molecular dipole and asymmetry as well as bringing desirable intermolecular interactions. Under the guidance of the “quasi-spherical theory” and the “momentum matching theory”, on [Me4N]2[KFe(CN)6], which undergoes the structural phase transition from centrosymmetric tetragonal I4/m to centrosymmetric cubic Fm3̅m, a hydroxyl replaced one methyl of the high-symmetric [Me4N]+ cation (Scheme 4).27 That way, the [Me3NOH]+ cation not only has a lower symmetry but also facilitates the increased intermolecular interactions, such as hydrogen bonding and coordination interactions, so as to achieve the polar monoclinic space group Cc in the RTP.19 For [Me3NOH]2[KFe(CN)6] (6), due to the high-speed rotation of the [Me3NOH]+ cation that makes the hydroxyl from methyl difficult to distinguish, along with the weakened hydrogen bonding and coordination effect, the HTP maintains Fm3̅m. With the Aizu notation of m3̅mFm, 6 is characterized as a multiaxial ferroelectric with 24 equivalent polarization directions.

aforementioned works on [Me4N]CdCl3, we inferred that taking halogen bonding into consideration is anticipated to trigger ferroelectricity in this system. As a consequence, two isostructural perovskite structures, [Me3NCH2X]MnX3 (X = Cl− (3) and Br− (4)) were designed.15,16 At room temperature, they both crystallize in the polar monoclinic space group Cc, and belong to the centrosymmetric hexagonal space group P63/mmc in the HTP (Scheme 2). With the Aizu notation of 6/mmmFm, both 3 and 4 possess 12 crystallographically equivalent polarization directions, corresponding to 6 ferroelectric axes. The targeted design strategies of multiaxial molecular ferroelectrics succeeded again in 3 and 4, confirming the rationality of “quasi-spherical theory” and “momentum matching theory”. 2.1.3. [Me3NCH2I]PbI3 (5). Lead halide perovskites like [MeNH3]PbI3 show great potential for next-generation solar cell devices. Although debate on the ferroelectric nature of [MeNH3]PbI3 and whether it is related to the ultrahigh photovoltaic performance is ongoing, the exploration of lead iodide perovskite ferroelectrics has attracted much attention. As a matter of course, [Me4N]PbI3 (ref code in CCDC: CERBUW) that belongs to the centrosymmetric hexagonal space group P63/m at room temperature becomes a valuable test subject.26 Taking advantage of the “quasi-spherical theory”, halogen atoms from F to Br were introduced to the spherical [Me4N]+ cation sequentially to increase the potential energy barrier of the tumbling motion, so the phase transition temperature raises from 184 K in [Me4N]PbI3 to 189, 269, and 291 K in [Me 3 NCH 2 F]PbI 3 , [Me 3 NCH 2 Cl]PbI 3 and [Me3NCH2Br]PbI3, respectively.18 Unfortunately, the F, Cl, and Br atoms are not large or heavy enough for making the tumbling motion totally freezing, so the modified cations preserve severe disorder similar to the [Me4N]+ cation. It is the canceled molecular dipole moments in the disordered state that render the three compounds centrosymmetric, with hexagonal space groups P63/m, P63/mmc, and P63/mmc, respectively.18 Seeing that these cations mismatch with the {PbI3}− anion and are unable to arouse ferroelectricity, the “momentum matching theory” leads us to seek the ideal one with suitable size and weight, that is, [Me3NCH2I]+ (Scheme

2.2. Design Ferroelectrics by Modifying dabco

2.2.1. [Mdabco]RbI3 (7). A series of simple organic salts, [dabco]X (X = ClO4−, BF4−, and ReO4−), have been reported as high-temperature hydrogen-bonded multiaxial molecular ferroelectrics.28−31 This inspired us to attempt to assemble the intensively studied dabco in the fascinating perovskite E

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 5. Targeted Design Strategy of 7a

a

(a) Modification of molecular symmetry; (b) initial structural phase transition; (c) and modified ferroelectric phase transition.

Scheme 6. Targeted Design Strategy of 7a

a

(a) Modification of molecular symmetry; (b) initial structural phase transition; and (c) modified ferroelectric phase transition.

7 with the Aizu notation 432F3 is definitely a multiaxial ferroelectric with eight equivalent polarization directions. 2.2.2. [Mdabco]NH4X3 (X = Br− (8) and I− (9)). The unique and adaptable crystal structure of perovskites leaves plenty of room for us to target design ferroelectrics by modifying either the templating cation or the anionic framework. A scarce metal-free perovskite, [dabco]NH4Br3, undergoing a ferroelastic phase transition from P432 to the trigonal P3121, deserved particular attention.21 Upon applying the “quasi-spherical theory” to tailor the organic cation and applying the “momentum matching theory” to select the matched anionic framework satisfying the Goldschmidt tolerance factor, a similar method to that of 7 enabled the discovery of a family of 3D molecular metal-free perovskite ferroelectrics. Striking examples are [Mdabco]NH4X3 (X = Br− (8) and I− (9)), which adopt the polar trigonal space group R3 and cubic P432 in the RTP and HTP, respectively, with an Aizu notation of 432F3 and having eight possible polarization directions (Scheme 6). Significantly, 9 has a high Curie temperature of 448 K and a large Ps of 22 μC/cm2, comparable to that of BaTiO3. The hydrogen bonding interactions between

architecture. [dabco]RbCl 3 comes into our view, but unfortunately, it adopts a nonpolar space group P3121.20,32 Following the “quasi-spherical theory”, a methyl is added to break the σh mirror plane of the dabco, so that the molecular symmetry of [Mdabco] is smoothly lowered from D3h to C3v, accompanied by the equipment of a dipole moment. In the HTP, the quasi-spherical [Mdabco] cation will still exhibit isotropic rotation at a high speed, in line with the necessary high-symmetric paraelectric phase of multiaxial ferroelectrics (Scheme 5). Then the question arises: the large size of the [Mdabco] cation is no longer suitable for the stabilization of the three-dimensional (3D) {RbCl3}− framework, which should obey size restrictions quantified by the Goldschmidt tolerance factor. For this, we finally chose the {RbI3}− anionic framework and targeted [Mdabco]RbI3 (7).19 At room temperature, 7 does crystallize in polar trigonal space group R3, and in the HTP it adopts the cubic space group P432, when the severely disordered cations indicate nearly free rotation. Owing to the order−disorder transition of the [Mdabco] cation and the distortion of the anionic framework, F

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 7. Targeted Design Strategy of 10 and 11a

a

(a) Modification of molecular symmetry; (b) initial structural phase transition; and (c) modified ferroelectric phase transition.

Scheme 8a

a

Targeted design strategy of 12−14: (a) introduction of homochirality; (b) initial structural phase transition (for details, see Table S2 and Figures S1−S8 in the Supporting Information); and (c) modified ferroelectric phase transition.

the [NH4]+, [Mdabco] cations and the I− contribute greatly to the stable perovskite structure and the ferroelectric behavior. 2.2.3. [Odabco]NH4X3 (X = Cl− (10) and Br− (11)). Similar to the case of [Mdabco], attaching a hydroxyl group on the dabco can also lower the molecular symmetry down to C3v. The resultant [Odabco] cation with a molecular dipole accords well with the “quasi-spherical theory”. With our expectations, [Odabco]NH4Cl3 (10) and [Odabco]NH4Br3 (11) belonging to the family of 3D molecular metal-free perovskite ferroelectrics were well designed.21 In the RTP, they crystallize in the polar orthorhombic space group Pca21 and the polar trigonal R3, respectively, and the corresponding HTP transform into the cubic space groups P4̅3m and P432, respectively, with Aizu notation of 4̅3mFmm2 and 432F3. The strong NOdabco−H···Cl hydrogen bonding interactions and the flexible hydroxyl make it easier to reorient the dipole moment and further induce a low-symmetric ferroelectric phase. The high symmetry of the HTP is ascribed to the high-speed, isotropic rotation of the [Odabco] cations (Scheme 7).

3. ASSEMBLY OF HOMOCHIRALITY As a well-established tool to induce noncentrosymmetric lattice structure, homochirality drives optimal molecular orientations in the crystals. It even has had particular relevance to the development of ferroelectrics, as Rochelle salt ([KNaC4H4O6]·4H2O) is just a typical homochiral compound. Homochirality is an important superiority of molecular crystals over the inorganic ones, which will greatly increase the possibility of forming the five chiral-polar point groups and promote the success rate of designing new molecular ferroelectrics. However, the use of homochiral building blocks for ferroelectrics has long been overlooked. Hence, combining with the precious experiences on the “quasi-spherical theory” and the “momentum matching theory” discussed above, nitrogen-containing heterocycles including quinuclidine, pyrrolidinium, and so on are considered as template cations to conduct subtle molecular modifications and the transcription of homochirality. G

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 9. Targeted Design Strategy of 15 and 16a

a

(a) Introduction of homochirality and (b) ferroelectric phase transition.

Scheme 10. Targeted Design Strategy of 17 and 18a

a

(a) Introduction of homochirality and (b) ferroelectric phase transition.

3.1. R- or S-3-Quinuclidinol Halides

(PFM) images were obtained, explicitly signifying the existence of non-180° domains which is unique for the multiaxial ferroelectrics. Hence, the Aizu notation of 12−14 should be assigned as 432F4, because the 422F4 represents a uniaxial ferroelectric. As shown in Scheme 8, in the high-symmetric HTP, the quasi-spherical 3-quinuclidinol cation rotates at a high speed and exhibits isotropic disorder. In the RTP, with the freeze of the molecular rotation, the lowered symmetry and equipped dipole moment of the 3-quinuclidinol cation start to work in the anisotropic ordered state, leading to ferroelectricity. This case is a perfect combination of the “quasi-spherical theory” and the idea of introducing homochirality.

The quinuclidine molecule crystallizes in a centrosymmetric cubic space group Fm3̅m at room temperature (ref code in CCDC: QUINCL), and then a hydroxyl is introduced in order to lower the molecular and crystal symmetry. The obtained rac-3-quinuclidinol disappointedly belongs to the centrosymmetric space group P21/n, failing to exhibit ferroelectricity and optical activity. But as we expected, the homochiral molecules (R)-(−)-3-quinuclidinol and (S)-(+)-3-quinuclidinol crystallize in the chiral-polar space group P61 in the RTP and undergo ferroelectric phase transitions with Aizu notation of 622F6.33 Not only as ferroelectric unimolecules, a wider world of homochiral molecules assemble with diverse components addressing the occurrence of ferroelectricity. Not surprisingly, the rac-3-quinuclidinol hydrochloride belongs to the centrosymmetric space group Pnma, while the series of isostructural R- or S-3-quinuclidinol halides (chloride (12), bromide (13), and iodide (14)) are characterized in chiral-ferroelectric space group P41.34 According to the Aizu rule (see Table 1), there are only two possible paraelectric point groups: 422 and 432. From the Kleinman symmetry, χ(2) vanishes in the 11 centrosymmetric point groups and the 3 chiral point groups 422 (D4), 622 (D6), and 432 (O). The phenomenon that 12− 14 undergo a clear transition from a SHG-active state to the SHG-inactive state at around the Curie temperature supports this deduction.34 Then, piezoresponse force microscopy

3.2. R- or S-3-AQ-NH4X3 (X = Cl− (15) and Br− (16))

On the basis of the “quasi-spherical theory”, the 3ammonioquinuclidinium (3-AQ) cation is also a competitive candidate for assembling ferroelectrics, in which the addition of an amino to the quinuclidinium lowers the molecular symmetry and imports a molecular dipole. As depicted in Scheme 9, the quasi-spherical 3-AQ cation, which experiences high-speed rotation in the HTP, appears to be an isotropic structure resembling the quinuclidinium. In the RTP, within the matched 3D framework, the movement of the 3-AQ cation will be confined to be anisotropic to some degree, which may facilitate the generation of ferroelectricity. To further guarantee the formation of a polar crystal structure, R- and S-3-AQ were used to assemble with the aforementioned 3D metal-free H

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research perovskite framework. Just as expected, the isostructural R- or S-3-AQ-NH4X3 (X = Cl− (15) and Br− (16)) were obtained as molecular ferroelectrics with the chiral-polar space group P21.21 In view of the Curie symmetry principle, the paraelectric point group should fall into the parent groups of P21: 2/m, 222, 422, 4̅2m, 32, 23, 622, and 432. As PFM reveals the existence of non-180° domains and further confirms the multiaxial nature, the point groups 2/m and 222 can be excluded. Then, from the temperature-dependent SHG effect, which is inactive in the paraelectric phase, the SHG-active point groups 42m, 32, and 23 are also not available. Finally, in terms of the crystallographic parameters of the HTP, the paraelectric space group of 15 and 16 is concluded to be P432. It is worth mentioning that the confinement effect of the 3D framework also plays a critical role in inducing spontaneous resolution. The rac-3-AQ-NH4X3 (X = Cl− and Br−) can be spontaneously resolved during the process of crystallization, and thus crystals, containing each homochiral molecule can be separated spontaneously. With these homochiral ferroelectric crystals obtained naturally from the raceme, their usability will improve while the manufacturing cost decreases.

Undoubtedly, the representative examples explained in the text will stimulate more creative works. Learning from the [Me3NCH2X]+ (X = Cl−, Br−, I−), two similar transformations from [Me3NCH2OH]+ and [(CH2OH)4N]+ to the respective [Me3NCH2NH2]+ and [(CH2OH)3CNH3]+ are predicted for the targeted design of molecular ferroelectrics. Also, considering the modification of dabco, the replacement of N atoms by the P element may make sense in the same way. Surprisingly, though introduction of the F atom lowers the molecular symmetry, the space group will not change. This idea can be applied in many existing ferroelectric systems to further find a similar but new candidate for enriching the molecular ferroelectric family. Additionally, the F atom will also increase the plasticity and lower the crystallinity of molecular crystals, bringing in excellent flexibility beneficial for film preparation. For example, taking the (3-pyrrolidinium)CdCl3 as a prototype, through slight structural substitution by the fluorine atom, the resultant (3-fluoropyrrolidine)CdCl3 might raise the Tc in one-dimensional metal halides. Overall, the coupling “quasi-spherical theory”, the “momentum matching theory”, and the homochirality idea constitute the whole targeted design strategy concerning “ferroelectric chemistry”. With more and more molecular ferroelectric materials in hand, their commercial applications in molecular recognition, domain engineering, molecular catalysis, and biocompatible electric devices can be targeted in the near future.

3.3. R- or S-3-AP-NH4X3 (X = Cl− (17) and Br− (18))

3-Ammoniopyrrolidinium (3-AP) is a cation similar to 3-AQ, modified by using the “quasi-spherical theory”. The addition of amino not only lowers the symmetry of the pyrrolidinium but also introduces a chiral center. The smaller size of cations allows a high-speed rotation in rac-3-AP-NH4X3 (X = Cl− and Br−), where the confinement effect is weakened and hence hinders the spontaneous resolution (Scheme 10). Because of the asymmetric cation, rac-3-AP-NH4X3 (X = Cl− and Br−) crystallizes in the noncentrosymmetric space group Ia.21 To change this, homochiral R- and S-3-AP cations were used to induce the chiral-polar space group P21 in the 3D molecular metal-free perovskite ferroelectrics: R- or S-3-AP-NH4X3 (X = Cl− (17) and Br− (18)).21 Their paraelectric space group of 17 and 18 can be reasonably inferred in the way similar to the aforementioned cases.



ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.accounts.8b00677. Discussion of “parent−child relationship” and “grandparent−grandchild relationship” and “momentum matching theory”; species of potential paraelectric-toferroelectric phase transitions; lattice parameters of quinuclidine and quinuclidinum halides; crystal structures of quinuclidinum, quinuclidinum hydrochloride, quinuclidinum hydrobromide, and quinuclidinum hydroiodine at different temperatures (PDF)

4. SUMMARY AND OUTLOOK This Account summarizes the targeted design strategies of multiaxial molecular ferroelectrics based on molecular modifications and homochiral assemblies. To address the extremely tough task to search for a one-in-a-million molecular ferroelectric among the countless compounds, we first proposed two chemistry-inspired practical approaches, named “quasi-spherical theory” and “momentum matching theory”. The first one focuses on high-symmetry or even spherical organic cations, for which the structural phase transitions are easy to realize, but the ferroelectric phase transitions are achieved by chance. When introducing electronegative species like halogen atoms or other chemical groups into such cations, it is anticipated that one would obtain polar crystal structure and ferroelectricity through the lowered molecular symmetry and the resultant molecular dipole. The “momentum matching theory” points out that the momentum of the cation and anion should also be matched, in a valid way, preserving easy molecular reorientation and further contributing to the ferroelectric polarization. In addition, the introduction of homochiral molecules is capable of greatly increasing the possibility of crystallizing in the five chiral-polar point groups, so as to facilitate the developments of new molecular ferroelectric systems.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Han-Yue Zhang: 0000-0001-6718-0665 Yuan-Yuan Tang: 0000-0002-8369-572X Ren-Gen Xiong: 0000-0003-2364-0193 Notes

The authors declare no competing financial interest. Biographies Han-Yue Zhang is a first-year Ph.D. student in Material Physics and Chemistry from Southeast University (Prof. Ren-Gen Xiong). She is working on the targeted design of multiaxial molecular ferroelectrics with homochirality. Yuan−Yuan Tang received his Ph.D. in 2019 from Southeast University in Material Physics and Chemistry. After graduation, he is a professor in the laboratory of Ren-Gen Xiong, working on electric I

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

ferroelectric thin-film with duodecuple-rotational polarization-directions. Nat. Commun. 2017, 8, 14934. (14) Ye, H. Y.; Liao, W. Q.; Hu, C. L.; Zhang, Y.; You, Y. M.; Mao, J. G.; Li, P. F.; Xiong, R. G. Bandgap Engineering of Lead-Halide Perovskite-Type Ferroelectrics. Adv. Mater. 2016, 28, 2579−2586. (15) You, Y.-M.; Liao, W.-Q.; Zhao, D.; Ye, H.-Y.; Zhang, Y.; Zhou, Q.; Niu, X.; Wang, J.; Li, P.-F.; Fu, D.-W.; Wang, Z.; Gao, S.; Yang, K.; Liu, J.-M.; Li, J.; Yan, Y.; Xiong, R.-G. An organic-inorganic perovskite ferroelectric with large piezoelectric response. Science 2017, 357, 306−309. (16) Liao, W. Q.; Tang, Y. Y.; Li, P. F.; You, Y. M.; Xiong, R. G. Large Piezoelectric Effect in a Lead-Free Molecular Ferroelectric Thin Film. J. Am. Chem. Soc. 2017, 139, 18071−18077. (17) Liao, W. Q.; Tang, Y. Y.; Li, P. F.; You, Y. M.; Xiong, R. G. Competitive Halogen Bond in the Molecular Ferroelectric with Large Piezoelectric Response. J. Am. Chem. Soc. 2018, 140, 3975−3980. (18) Hua, X.-N.; Liao, W.-Q.; Tang, Y.-Y.; Li, P.-F.; Shi, P.-P.; Zhao, D.; Xiong, R.-G. A Room-Temperature Hybrid Lead Iodide Perovskite Ferroelectric. J. Am. Chem. Soc. 2018, 140, 12296−12302. (19) Xu, W.-J.; Li, P.-F.; Tang, Y.-Y.; Zhang, W.-X.; Xiong, R.-G.; Chen, X.-M. A Molecular Perovskite with Switchable Coordination Bonds for High-Temperature Multiaxial Ferroelectrics. J. Am. Chem. Soc. 2017, 139, 6369−6375. (20) Zhang, W.-Y.; Tang, Y.-Y.; Li, P.-F.; Shi, P.-P.; Liao, W.-Q.; Fu, D.-W.; Ye, H.-Y.; Zhang, Y.; Xiong, R.-G. Precise Molecular Design of High-T c 3D Organic−Inorganic Perovskite Ferroelectric: [MeHdabco]RbI3 (MeHdabco = N-Methyl-1,4diazoniabicyclo[2.2.2]octane). J. Am. Chem. Soc. 2017, 139, 10897− 10902. (21) Ye, H.-Y.; Tang, Y.-Y.; Li, P.-F.; Liao, W.-Q.; Gao, J.-X.; Hua, X.-N.; Cai, H.; Shi, P.-P.; You, Y.-M.; Xiong, R.-G. Metal-free threedimensional perovskite ferroelectrics. Science 2018, 361, 151−155. (22) Morosin, B. Crystal structure of tetramethylammonium cadmium chloride. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1972, 28, 2303−2305. (23) Paściak, M.; Wołcyrz, M.; Pietraszko, A. Structural origin of the x-ray diffuse scattering in (CH3)4NCdCl3 and related compounds. Phys. Rev. B: Condens. Matter Mater. Phys. 2008, 78, No. 024114. (24) Liao, W.-Q.; Zhao, D.; Tang, Y.-Y.; Zhang, Y.; Li, P.-F.; Shi, P.P.; Chen, X.-G.; You, Y.-M.; Xiong, R.-G. A molecular perovskite solid solution with piezoelectricity stronger than lead zirconate titanate. Science 2019, 363, 1206−1210. (25) Morosin, B.; Graber, E. J. Crystal structure of tetramethylammonium manganese(II) chloride. Acta Crystallogr. 1967, 23, 766− 770. (26) Contreras, J. G.; Seguel, G. V.; Ungerer, B.; Maier, W. F.; Hollander, F. J. Triiodolead(II) complexes. structure and Raman spectra. J. Mol. Struct. 1983, 102, 295−304. (27) Xu, W.-J.; Chen, S.-L.; Hu, Z.-T.; Lin, R.-B.; Su, Y.-J.; Zhang, W.-X.; Chen, X.-M. The cation-dependent structural phase transition and dielectric response in a family of cyano-bridged perovskite-like coordination polymers. Dalton Trans. 2016, 45, 4224−4229. (28) Katrusiak, A.; Szafrański, M. Ferroelectricity in NH··· N hydrogen bonded crystals. Phys. Rev. Lett. 1999, 82, 576. (29) Szafrański, M.; Katrusiak, A.; McIntyre, G. Ferroelectric Order of Parallel Bistable Hydrogen Bonds. Phys. Rev. Lett. 2002, 89, 215507. (30) Tang, Y.-Y.; Zhang, W.-Y.; Li, P.-F.; Ye, H.-Y.; You, Y.-M.; Xiong, R.-G. Ultrafast Polarization Switching in a Biaxial Molecular Ferroelectric Thin Film: [Hdabco]ClO4. J. Am. Chem. Soc. 2016, 138, 15784−15789. (31) Tang, Y.-Y.; Li, P.-F.; Zhang, W.-Y.; Ye, H.-Y.; You, Y.-M.; Xiong, R.-G. A Multiaxial Molecular Ferroelectric with Highest Curie Temperature and Fastest Polarization Switching. J. Am. Chem. Soc. 2017, 139, 13903−13908. (32) Paton, L. A.; Harrison, W. T. Structural diversity in non-layered hybrid perovskites of the RMCl3 family. Angew. Chem., Int. Ed. 2010, 49, 7684−7687.

domains under the technique of piezoelectric force microscope (PFM). Ping−Ping Shi received her Ph.D. in Applied Chemistry from Southeast University under the supervision of Ren-Gen Xiong in 2016. Thereafter, she joined the group of Ren-Gen Xiong as an assistant professor. She is working on the targeted design of molecular ferroelectrics with excellent electrical, mechanical, and optical properties. Ren-Gen Xiong is the head of Jiangsu Key Laboratory for Science and Applications of Molecular Ferroelectrics in Southeast University. Since the mid-1990s, he has been working on the study of noncentrosymmetric coordination compounds. Now his research interests mainly cover the exploration of molecule-based ferroelectrics. In terms of the Curie symmetry principle and Neumann’s principle, the targeted design strategies on multiaxial molecular ferroelectrics are put forward in predicting potential ferroelectrics. Meanwhile, he and his team have successfully discovered a series of high-temperature and multiaxial molecule-based ferroelectrics from practice. This study can significantly contribute to the purposeful search for ferroelectrics.



ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (21427801, 21831004).



REFERENCES

(1) Valasek, J. Piezo-electric and allied phenomena in Rochelle salt. Phys. Rev. 1921, 17, 475. (2) Scott, J. F. Applications of modern ferroelectrics. Science 2007, 315, 954−959. (3) Lines, M. E.; Glass, A. M. Principles and applications of ferroelectrics and related materials; Clarendon Press: Oxford, UK, 1977. (4) Scott, J. F.; Araujo, C. P. d. Ferroelectric memories. Science 1989, 246, 1400−1405. (5) Goossens, D. J. Local Ordering in Lead-Based Relaxor Ferroelectrics. Acc. Chem. Res. 2013, 46, 2597−2606. (6) Shi, P.-P.; Tang, Y.-Y.; Li, P.-F.; Liao, W.-Q.; Wang, Z.-X.; Ye, Q.; Xiong, R.-G. Symmetry breaking in molecular ferroelectrics. Chem. Soc. Rev. 2016, 45, 3811−3827. (7) Aizu, K. Possible species of “ferroelastic” crystals and of simultaneously ferroelectric and ferroelastic crystals. J. Phys. Soc. Jpn. 1969, 27, 387−396. (8) Fu, D. W.; Cai, H. L.; Liu, Y.; Ye, Q.; Zhang, W.; Zhang, Y.; Chen, X. Y.; Giovannetti, G.; Capone, M.; Li, J.; Xiong, R. G. Diisopropylammonium bromide is a high-temperature molecular ferroelectric crystal. Science 2013, 339, 425−428. (9) Ye, H.-Y.; Ge, J.-Z.; Tang, Y.-Y.; Li, P.-F.; Zhang, Y.; You, Y.-M.; Xiong, R.-G. Molecular Ferroelectric with Most Equivalent Polarization Directions Induced by the Plastic Phase Transition. J. Am. Chem. Soc. 2016, 138, 13175−13178. (10) Ye, H. Y.; Zhou, Q.; Niu, X.; Liao, W. Q.; Fu, D. W.; Zhang, Y.; You, Y. M.; Wang, J.; Chen, Z. N.; Xiong, R. G. High-Temperature Ferroelectricity and Photoluminescence in a Hybrid OrganicInorganic Compound: (3-Pyrrolinium)MnCl3. J. Am. Chem. Soc. 2015, 137, 13148−13154. (11) Liao, W. Q.; Zhang, Y.; Hu, C. L.; Mao, J. G.; Ye, H. Y.; Li, P. F.; Huang, S. D.; Xiong, R. G. A lead-halide perovskite molecular ferroelectric semiconductor. Nat. Commun. 2015, 6, 7338. (12) Fu, D. W.; Zhang, W.; Cai, H. L.; Ge, J. Z.; Zhang, Y.; Xiong, R. G. Diisopropylammonium chloride: a ferroelectric organic salt with a high phase transition temperature and practical utilization level of spontaneous polarization. Adv. Mater. 2011, 23, 5658−5662. (13) You, Y. M.; Tang, Y. Y.; Li, P. F.; Zhang, H. Y.; Zhang, W. Y.; Zhang, Y.; Ye, H. Y.; Nakamura, T.; Xiong, R. G. Quinuclidinium salt J

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research (33) Li, P.-F.; Liao, W.-Q.; Tang, Y.-Y.; Qiao, W.; Zhao, D.; Ai, Y.; Yao, Y.-F.; Xiong, R.-G. Organic enantiomeric high-Tc ferroelectrics. Proc. Natl. Acad. Sci. U. S. A. 2019, 116, 5878−5885. (34) Li, P. F.; Tang, Y. Y.; Wang, Z. X.; Ye, H. Y.; You, Y. M.; Xiong, R. G. Anomalously rotary polarization discovered in homochiral organic ferroelectrics. Nat. Commun. 2016, 7, 13635.

K

DOI: 10.1021/acs.accounts.8b00677 Acc. Chem. Res. XXXX, XXX, XXX−XXX