1 Nanostr uctural polymorphism in the low -birefringence chiral phase of an achiral bent - shaped dimer

2025-04-30 0 0 3.26MB 25 页 10玖币

侵权投诉

Nanostructural polymorphism in the low-birefringence chiral phase of an achiral bent-

shaped dimer

Khoa V. Le, 1,2,* Michael R. Tuchband, 3 Hiroshi Iwayama,4 Yoichi Takanishi,5 Noel A. Clark,

3 and Fumito Araoka1,*

1 RIKEN Center for Emergent Matter Science (CEMS), 2-1 Hirosawa, Wako, Saitama 351-

0198, Japan

2 Department of Chemistry, Faculty of Science, Tokyo University of Science, 1-3 Kagurazaka,

Shinjuku-ku, Tokyo 162-8601, Japan

3 Department of Physics and Soft Materials Research Center, University of Colorado,

Boulder, Colorado 80309-0390, USA

4 UVSOR Synchrotron Facility, Institute for Molecular Science, Okazaki 444-8585, Japan

5 Department of Physics, Kyoto University, Kitashirakawaoiwake-cho, Sakyo-ku, Kyoto, 606-

8502, Japan.

Correspondence and requests for materials should be addressed to F.A. (email:

fumito.araoka@riken.jp) or to K.V.L. (email: khoa@rs.tus.ac.jp).

Abstract

Polymorphism, the phenomenon that a species can exist in many discrete forms, is common

in nature, such as hair colors in an animal species, flower colors in a tree species, and blood

types in humans, etc. In materials science, it refers to a solid that can exist in multiple forms

with different crystalline structures. In the liquid crystals field, however, polymorphism is

hard to find because a discontinuous structural variation is basically impossible because of

their fluid or partially fluid nature. Herein we show that the B4 and DC phases that for many

years have been classified as distinctive phases are connected, in terms of their nano-

architectures, based on the study of a single compound, a flexible bent-shaped dimer. The

surrounding solvent is the key to assisting the dimeric molecules in morphing and adopting

different supramolecular structures at the mesoscale. Furthermore, we accidentally find a

novel nanotube-like structure that has not yet been reported in view of the B4/DC phases.

Together with the known sponge (DC) and the helical filament (B4) structures, they are just

some of the manifestations of the polymorphism in a class of low-birefringence, chiral phase

from achiral liquid crystals.

The processes underlying the self-assembly of constituent building blocks into complex,

hierarchical superstructures remain among the most intriguing mysteries in materials science

today. This research field has been actively pursued and widely exploited to develop novel

functional materials, including both nature-inspired biomimetic materials and metamaterials

[1] [2]. Controlling the shape, size, orientation, and arrangement of constituents at the

molecular level, giving rise to desirable emergent properties at macroscopic scales is,

however, a significant challenge. Organization can arise spontaneously through a variety of

non-covalent intermolecular forces [3] or can be induced by external factors such as electric

and magnetic fields, temperature, light, pH, or interactions with foreign molecules through

the so-called solvent or dopant effects, examples of which are aqueous systems of peptide- or

glycolipid amphiphiles [4] [5] [6]. At first glance, one might imagine that the constituent

building blocks of a material made up of chiral superstructures must be chiral, as a

manifestation of chiral amplification. In this context, the self-assembly of achiral bent-core

molecules is unique and interesting, since they can geometrically pack to form chiral and

polar symmetry-breaking phases with large conglomerate domains over micron dimensions

or larger [7] [8] [9] [10]. Hough et al. reported that in some bent-core molecules which have

intralayer semi-crystalline order, a twisted helical nanofilament (HNF), called the B4 phase,

or lamellar sponge (SP) structure, called the dark conglomerate (DC) phase, can form [11]

[12]. The B4 phase is particularly intriguing because it has strong optical activity [13] [14],

large effective nonlinear optical coefficients [15] [16], strong gelation ability [17] [18],

enhanced hydrophobicity [19], anisotropic charge transport properties in photovoltaic devices

[20], and structural colors [21] [22] [23]. One can envision a wide range of applications for

this material, such as tunable optical rotators, piezoelectric elements, chiral detectors,

asymmetric chemical syntheses, ultra-dry surfaces, solar cells, color reflectors, etc., among

other possibilities [10] [24] [25] [26]. The formation of the dark conglomerate SP and HNF

structures appear to be driven by the same mechanism, i.e., a mismatch of the two half-layers

formed by the molecular arms of the compound in a smectic layer induces saddle-splay

curvature to relieve the elastic strain caused by their mismatch. This leads to the spontaneous

formations of left- and right-handed HNFs or dark conglomerate SP domains at the

macroscopic scale with a huge optical activity, on the order of 1 deg/m. However, this still

leaves open the following question: what causes a particular material to adopt either the SP or

the HNF structure?

In this work, and for the first time to our knowledge, we report that an achiral bent dimer

molecule which exhibits an SP morphology as the ground state structure can be continuously

transformed to an HNF structure and an unusual hollow nanotube (tubular) structure simply

by making mixtures with either mesogenic or organic isotropic solvents and tuning the

concentration. We characterize this behavior with polarizing optical microscopy (POM),

scanning electron microscopy (SEM), freeze-fracture transmission electron microscopy

(FFTEM), atomic force microscopy (AFM) and synchrotron X-ray diffraction (XRD). The

thorough understanding of the mechanism of this polymorphism may provide not only

insight into the nature of the formation of the B4 “banana” phase and other bent-core liquid

crystal phases but may also shed light on the materials science pursuit of bottom-up

supramolecular design approaches which exhibit multiple tiers of self-assembly.

Results

Neat material 12OAz5AzO12. In this work, we study a bent-shaped symmetric dimer

compound which we call 12OAz5AzO12. It has a pentamethylene spacer connecting two

mesogenic wings, namely, 4-n-alkoxyazobenzene-4'-carbonyloxy-n-dodecanes (Fig. 1a) [27].

The odd number of methylene units in the linkage makes the molecule adopt a bent shape in

the all-trans conformation. Its phase sequence is Iso (108°C) SmCA (94.1°C) SmX, where the

SmX phase is a low birefringence phase with chiral conglomerate domains and is stable

down to room temperature. This compound has been utilized in earlier works to investigate

chirality control in the DC/B4 phase [28], twisted nematic director fields [29], and

biomolecular adsorbates [30]. Upon cooling to the SmX phase (Fig. 1b,c), POM reveals a

very dark texture with some birefringent inclusions which are most probably thin layers of

SmCA-like structures pinned on the surfaces remaining from the higher temperature SmCA

phase, as their retardation does not change with cell thickness. By slightly decrossing the

polarizers, the spontaneous chiral resolution can be easily checked (Fig. 1d,e). When the

SmX phase forms, cracks in the material appear due to volume shrinkage, which is common

in the B4 phase (Fig. 1d,e). For many years, this SmX phase was identified as the ‘B4 phase’

since it is a solid, low-birefringence phase that appears below a smectic phase and shows no

response to electric field, all characteristics of a B4 phase. However, for the first time, our

SEM and FFTEM observations show that neat 12OAz5AzO12 has a sponge structure, rather

closer in morphology to the DC phase (Fig. 3a,e). We observe only disordered focal conic

domains with clear saddle-splay curvature throughout the bulk sample. In particular, when

the sample is fractured near the glass interface, the layers stand on the surface with the layer

normals parallel to the glass surface [Fig. S1], as reported in ref. [12]. These observations

correspond well with those reported by Hough et al. for a DC phase nanostructure [12].

Observations using AFM also support the suggestion of a sponge structure [Fig. S2].

Mixtures of 12OAz5AzO12 and a rod-like nematic liquid crystal (NLC) ZLI-2293. We

made mixtures of different concentrations of 12OAz5AzO12 with a rod-like nematic material

called ZLI-2293 (Merck) to observe the various effects on the nanoscale morphology of the

phase. ZLI-2293 was chosen because it has a wide nematic temperature range (91°C to below

room temperature). The phase diagram of the mixtures in Fig 2a is based on POM

observations. The SmCA phase is completely suppressed at 20 wt% ZLI-2293, with this same

behavior reported in other systems of bent-core LCs with rod-like LC solvents [31] [32] [33].

At higher concentrations of ZLI-2293 (>20 wt%), all the mixtures exhibit two distinct phase

transitions: the Iso–N transition and a transition from the homogenous N phase to the

composite SmX phase, composed of nano/micro-phase segregated 12OAz5AzO12 and

nematic ZLI-2293 [14] [32]. Fig. 2b contains a POM image of a contact cell at room

temperature demonstrating continuous changes in texture along the concentration gradient of

ZLI-2293. On the right-hand side of the image, the neat 12OAz5AzO12 exhibits its

characteristic dark texture with some birefringent inclusions (region A) similar to those in Fig.

1c. On increasing the ZLI-2293 concentration, the texture becomes darker and smoother with

no more birefringent inclusions (region B). Further increasing the ZLI-2293 concentration,

we observe many dendrites growing and separating from the nematic background (region C).

Finally, in region D, the ZLI-2293-rich nematic phase exhibits a Schlieren texture.

Decrossing the polarizers reveals that the conglomerate domains in region A persist into

region B and even C [see more in Fig. S3]. But how could such a sponge structure nucleate,

develop, and phase separate from the nematic background? To our surprise, SEM and

FFTEM observations reveal that the nano-phase segregated 12OAz5AzO12 in the mixtures

(observed after ZLI-2293 is removed, see Methods) no longer has the SP structure, but

transforms into an HNF morphology with increasing concentration of ZLI-2293 (up to ~60

wt%) (Fig. 3c). Even more surprisingly, at a very high concentration of ZLI-2293 (≈ 80 wt%

or above) a hollow nanotube structure of diameter ≈ 100 nm, slightly larger than that of the

HNFs (70 – 80 nm), predominantly forms (Fig. 3d) [see also Fig. S4].

As the SEM resolution is generally quite limited for organic materials, FFTEM

observations were also performed on these mixtures to visualize the layering, which should

be approximately 4 – 5 nm. For the mixture with 60 wt% ZLI-2293, in which the HNFs are

dominant, the HNF layer stacks bending and twisting with saddle-splay curvature is clearly

observed (Fig. 3g). We also observe several HNFs branching out from a single HNF, a

process that enables the chirality preserving growth which leads to the macroscale chiral

conglomerate domains in the B4 phase [34]. We also observe several examples of coherently

twisting HNFs, with the layers in adjacent filaments face-to-face [Fig. S5]. A larger scale

view of the HNF morphology [Fig. S5] demonstrates that the filaments can grow

independently or collectively in the bulk and form a nanoporous structure, as expected for

HNFs [11]. These structures are undoubtedly HNFs, but are remarkably different from those

formed by traditional bent-core molecules [11] [34] [35], as analyzed from FFTEM data: (i)

the cross-section of the HNFs formed of 12OAz5AzO12 is rectangular with the dimension

along the layer normal (≈ 80 nm) about 4 times longer than the width (≈ 20 nm), as opposed

to a nearly square cross-section in other bent-core LCs, and (ii) the layers (each layer ≈ 5 nm)

stack such that a layer ‘shelf’ is evident every 3 – 4 layers, with the layer shelves stack on

one another while slightly shifting towards to the edges of the filament (the direction normal

to the filament axis). On the other hand, when the concentration of ZLI-2293 is between 20 –

40 wt%, neither a clear HNF or SP structure is discernable; rather, an intermediate

morphology between SP and HNF observed (Fig. 3b,f). The presence of this transitional

regime demonstrates that the composite SmX phase of chiral conglomerate domains is a

spectrum of morphologies that exist between the SP and HNF structures. All these

morphologies have a saddle-splay curvature (negative Gaussian curvature) and are identical

from a topological viewpoint, i.e., catenoid and helicoids (see Supporting Information of ref.

[35]), respectively, and blends of the two in the intermediate regime. This shows that the

interfacial tension between these two immiscible components plays an essential role in

determining the morphology. Finally, in good accordance with SEM, when the concentration

of ZLI-2293 is high (80 wt% or above), almost solely nanotubes are observed (Fig 3h). Their

surfaces are very smooth smectic layers with zero Gaussian curvature, as also confirmed by

AFM [Fig. S6]. Some of the nanotubes are even connected with the twisted HNF structures,

indicating that the mechanisms of the formation of the nanotubes and HNFs are closely

related. Here also, we believe it is the interfacial tension that transforms the HNF into the

nanotube structure.

Our experiments with other mesogenic solvents (5CB, CB15) and non-mesogenic solvents

(n-dodecane, chloroform) also show such polymorphic behavior, with lower concentrations

of the non-mesogenic guest material necessary to form the nanotube structures [Fig. S7]. In

particular, we also find that nanotube structures even form at extremely low concentrations of

12OAz5AzO12, e.g., 98.7 wt% n-dodecane [Fig. S8]. We stress that if the starting material

has the HNF structure, no matter how much solvents are mixed, the HNF structure always

remains, either in the form of helical twists or helical ribbons [17] [18].

X-ray diffraction and resonant carbon K-edge X-ray scattering.

To determine the nature of the molecular packing within these various nanostructures, we

performed synchrotron X-ray diffraction experiments. Fig. 4a shows the data taken on three

samples with the distinct morphologies after the solvent was removed. Curiously, they show

essentially the same behavior, except for some variations in the peak widths of particular

features. At least 8 different harmonics of the first-order diffraction attributed to the layer

periodicity (≈ 5 nm) appear at small angles, signifying the robust lamellar ordering. The full

width at half maximum of these peaks reveals the finite size of the lamellar correlations to be

is about 20 layers [11] [18], which agrees well with the results obtained from the electron

microscopy observations. At wide angles, the numerous peaks point to long-range crystalline

or semi-crystalline order within the layers (Fig. 3) and that 12OAz5AzO12 maintains this

rigid ordering when nano-/micro-phase segregating from the guest solvent, regardless of the

concentration.

The resonant soft X-ray scattering (RSoXS) at the carbon K-edge (284.5 eV), which is a

powerful technique to probe the periodic modulation of the orientation of carbon bonds [36]

[37], shows a clear distinction between the three morphologies (Fig. 4b). HNFs, appearing in

the middle doping range, show relatively sharp scattering peaks at the scattering angle

corresponding to the half-pitch (≈150 nm) periodicity of the helical filaments, which is

consistent with that observed by FFTEM (Fig. 3g) and analogous to that in another dimer

with similar structure [26]. On the other hand, the SP structures show broad peaks at the

slightly smaller scattering angle region, which could be corresponding to the average

distances between the lamellar inter-connections. Interestingly, the nanotubes show just

scarce signals in this region, meaning almost no nanoscopic structural periodicity.

Discussion

A study by Lin et al. showed that by mixing a bent-core molecule with a poor solvent

(THF/water solution) they could force their assembly into different superstructural

organizations, including flat, elongated lamellar crystals, helical ribbons, and tubules [38].

They reported that the helical morphology is driven by conformational chirality of the bent-

core molecules which results in the twisting and bending of lamellar crystals into the various

complex shapes [39] [40]. More recently, Cano et al. showed that some ionic bent-core

dendrimer molecules can aggregate in water to form a variety of assemblies, including rods,

spheres, fibers, helical ribbons, or tubules [41]. In our case, mesogenic and non-mesogenic

organic guest solvents are used, and 12OAz5AzO12 always completely phase separates from

the guest due to the strong crystalline ordering it possesses, as gleaned from the X-ray data

(Fig. 4).

The unusual phase sequence that we find in 12OAz5AzO12 (SmCA–DC) has also been

observed recently by Chen et al., where a bent-core compound (W624 in ref. [42]) was

reported to exhibit a stable DC phase below the B2 phase. This is intriguing because,

according to other previous reports, while the B4 phase is rather solid (semi-crystalline) and

normally appears below the B2 and/or the B3 phases, the DC phase is more liquid-like and

appears below an isotropic liquid and not the ordered SmCA or B2 phase [12] [11]. The DC

phase can in principle be transformed into the ordered B2 phase on application of an electric

field [12] [43] or could be induced from the columnar B1rev phase [44]. Like in W624, we

observe re-crystallization [Fig. S9] when the DC phase is heated back up to the SmCA phase.

Therefore, the DC phase reported for 12OAz5AzO12 here can be considered to be in a meta-

stable state. As such, interfacial tension from the guest solvent can drive it to the HNF

structure which would make it the more stable configuration. On the other hand, by

continuing to increase the solvent content to some extent, the HNF is then transformed to the

nanotube structure. Recently, the observation of the nanotube structures similar to ours has

also been reported in mixtures of an acute-angle (ca. 45°) bent-core compound with

nematogenic or smectogenic additives over a narrow concentration range (Fig. 3h and Fig.

S6) [45]. The acute-angle bent-core compound also exhibits re-crystallization on heating,

which implies that molecules exhibiting metastable DC/B4 phases may be suitable candidates

to express a polymorphic behavior when mixed with a solvent or influenced by other external

factors. Because the nanotube structure does not have a negative Gaussian curvature like the

SP or the HNF structures, it appears that the interfacial energy is strong enough to suppress

saddle-splay structure and completely change the expected morphology. For instance, when

we consider the molecules at the outmost layers of each nanotube, whose exposed chemical

groups are the liquid-like dodecanoxy groups (-OC12H25), the system can gain high entropy

even if the dimer molecules strongly aggregated because the outer surface is more soluble

with the alkylated portions of the solvent molecules [46]. This also explains why we require

less n-dodecane (C12H26) to form the nanotubes.

From the symmetry viewpoint, Fig. 5 shows how a change in local in-plane hexatic lattice

symmetry, from a square lattice in the layer mid plane to a rectangular lattice in the layer mid

plane, mediates a change in growth morphology from twisted ribbon B4 to cylindrical ribbon

nanotube. In the typical B4 twisted ribbon case the upper and lower half layers are

structurally identical, related by a 2-fold rotation around P. The upper and lower half layers

relate identically to the overall ribbon geometry. In the nanotube (cylindrical ribbon) case, the

lower symmetry implies a structural difference between the upper and lower half layers, e.g..

a larger tilt in the upper half than that of the lower. In this case the orange lattice must be

rectangular and the upper and lower half layers relate differently to the overall structure with

one along the ribbon and the other normal to it. In both cases the filament edges are a low

Miller index face of the 2D lattice, 11 in the case of the ribbon and 10 in the case of the

nanotube.

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

1Nanostructuralpolymorphisminthelow-birefringencechiralphaseofanachiralbent-shapeddimerKhoaV.Le,1,2,*MichaelR.Tuchband,3HiroshiIwayama,4YoichiTakanishi,5NoelA.Clark,3andFumitoAraoka1,*1RIKENCenterforEmergentMatterScience(CEMS),2-1Hirosawa,Wako,Saitama351-0198,Japan2DepartmentofChemistry,FacultyofSci...

展开>> 收起<<

1 Nanostr uctural polymorphism in the low -birefringence chiral phase of an achiral bent - shaped dimer.pdf

共25页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

1 Nanostr uctural polymorphism in the low -birefringence chiral phase of an achiral bent - shaped dimer

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: