Perspectives in Liquid-Crystal-Aided Nanotechnology and Nanoscience

Featured Application

Nanomaterial-doped liquid crystals.

Abstract

The research field of liquid crystals and their applications is recently changing from being largely focused on display applications and optical shutter elements in various fields, to quite novel and diverse applications in the area of nanotechnology and nanoscience. Functional nanoparticles have recently been used to a significant extent to modify the physical properties of liquid crystals by the addition of ferroelectric and magnetic particles of different shapes, such as arbitrary and spherical, rods, wires and discs. Also, particles influencing optical properties are increasingly popular, such as quantum dots, plasmonic, semiconductors and metamaterials. The self-organization of liquid crystals is exploited to order templates and orient nanoparticles. Similarly, nanoparticles such as rods, nanotubes and graphene oxide are shown to form lyotropic liquid crystal phases in the presence of isotropic host solvents. These effects lead to a wealth of novel applications, many of which will be reviewed in this publication.

1. Introduction

For many materials, the transition between the liquid and the solid phase is not a single-step process, but a range of various mesophases, which are called liquid crystals (LCs). LCs are self-organized anisotropic fluids that are thermodynamically located between the isotropic liquid and the crystalline phase, exhibiting the fluidity of liquids as well as the long-range lattice order that can only be found in crystalline solids [1,2,3]. Generally, LCs are composed of anisotropic building blocks (usually of rod or disc shape), which are spontaneously oriented along a specific direction, called the director n [4]. Without an external alignment force, the director of a nematic LC, the simplest phase of LC, whose molecules are only orientationally ordered, is usually spatially changed continuously but randomly over large spatial extensions (except for defects, where the director may vary suddenly and drastically) [5,6].

Normally, one can split LCs into two typical categories, i.e., thermotropic LCs and lyotropic LCs [7,8,9,10]. Thermotropic LCs are usually further distinguished according to the shape of their constituent molecules, being called calamitic for rod-like, discotic for disk-like and sanidic for brick- or lath-like molecules (Figure 1a) [11]. A typical calamitic mesogen is generally composed of a rigid core, often incorporating phenyl and biphenyl groups, and two flexible endgroups, often alkyl or alkoxy chains. A common structure of discotic mesogens is a rigid, disk-like core to which six flexible endgroups are attached. Apart from these conventional mesogens, research attention has been recently focused on the so-called non-conventional LCs [12], which are neither rod- nor disk-shaped. Among them, bent-core LCs [13], in particular, have received attention due to their unique effects of the observation of chirality from achiral molecules, resulting from sterically induced packing of the bent-core mesogens [10], such as ferroelectricity or the formation of helical superstructures in the B7 phase [14]. Thermotropic LCs are commonly constituted by single organic entities or mixtures thereof, which exhibit various mesophases at different temperatures or pressures [15], illustrated in Figure 1b. As the temperature rises, a typical thermotropic LC passes through higher ordered phases, also called soft crystals, the hexatic smectic phases with positional order as well as bond orientational order, through the fluid smectic phases (SmC and SmA), which exhibit both positional and orientational order, and finally to the nematic phase (N) with purely orientational order, into the isotropic phase. The number of different phases observed depends on the chemical composition, symmetry and order of the LC molecules. About 25 different thermotropic phases are known to date, and they are still increasing in number.

In contrast, lyotropic LCs consist of at least two different kinds of components: a collection of anisodiametric molecules and particles dispersed in a suitable solvent (Figure 2). These systems will thus always be mixtures, with the main control variable being the concentration. Unlike thermotropics, the phase transitions of lyotropic LCs are not merely dependent on the temperature of the system, but also, mainly, on the relative concentration of each component [15]. Being mixtures, there will always be two-phase regions at the phase boundaries between two different lyotropic phases. One can distinguish lyotropic LCs into three different kinds: (i) amphiphilic lyotropics, (ii) colloidal lyotropics, and (iii) chromonics, where the constituent molecules are dye molecules in a suitable solvent. As the name indicates, amphiphilic lyotropic LCs are usually composed of amphiphilic molecules upon addition of a solvent [16], often water. As Figure 2 shows, with increasing amphiphile concentration, due to the segregation of hydrophobic and hydrophilic regions, these molecules self-assemble into spherical or rod-like micelles, leading to the formation of the hexagonal phase, cubic phase or lamellar phase. Nevertheless, the solute component of a lyotropic LC need not always be molecular in nature, but may also consist of much larger (solid) particles with anisotropic shapes. It can also be of colloidal size [7], as we will discuss further in Section 3. Such materials would then be known as inorganic liquid crystals, LC clays, but also nanotubes, graphene oxide or biological structures such as viruses (Figure 3).

The fact that the director of LCs, which is often equivalent to the optic axis of the material, is easily influenced by a variety of external stimuli such as mechanical, magnetic, electric, or optic fields, as well as temperature, makes liquid crystals attractive to both the industry and academia [6,17,18,19,20,21,22,23,24,25,26], as exemplified by the $100 billion industry built around displays and large-screen, flat-panel consumer electronics. Due to the great achievement in displays since the 1970s, LCs are one of the most popular materials around the world [27]. However, with the rapid development and impressive advantages of organic light-emitting displays (OLEDs), there is some competition emerging. Over the last decade, more and more scientists have moved their attention away from display materials to a diversity of new fields—for instance, new optical devices, telecommunication, information storage, energy conservation, elastomer robots, sensors, biotechnologies, nano-/micromanipulation, just to name a few [15,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45]. These new fields can be quite different from displays and more stimulating due to their novelty, leading to a new era for LCs, materials design and technology.

At the same time, revolutionary developments have been made in the fields of nanotechnology and nanoscience, leading to the birth of a series of novel nanomaterials [10,47,48,49,50,51,52,53,54,55]. These nanostructured materials, whose size in at least one dimension is in the range from 1 nm to 100 nm, have gained a wealth of academic and industrial attention due to their size-dependent electronic, optical, magnetic, and chemical properties, which are significantly different from those of bulk materials as well as from individual atoms or molecules [55,56,57,58,59,60,61,62]. These nanostructured materials have been extensively applied in nearly every field of science from energy, optics, computing to catalysis, biosciences and medical sciences [48,63,64,65]. Undoubtedly, when these novel nanometre-scale structures encounter liquid crystals, a highly interesting and unique synergy will be observed, leading to an abundance of entirely new and potential applications [2,8,42,66,67,68,69,70].

The addition of nanomaterials to a LC material produces a composite or colloidal dispersion [71,72]. The new materials are expected to behave differently from their individual components (nanomaterials and LCs) both on the microscopic as well as the macroscopic scale [63,64,65,66,67,68,69,70,71,72,73,74,75,76,77]. There are three basic objectives to making a LC–nanomaterial composite system: (i) to modify the primary physical or chemical properties of the pristine LCs [78,79,80]; (ii) manipulating and ordering nanomaterials in LCs to modify the properties of nanomaterials [81]; and (iii) to obtain additional functionalities that are available from neither the LCs nor the nanomaterials in their intrinsic states [82].

A fundamental investigation of the nanophysics of LCs was published by Brochard and de Gennes, who discussed a theory of magnetic nanoparticles suspended in a nematic LCs in 1970 [83]. Later in the same year, Rault et al. reported the first suspension composed of nematic LCs and small magnetic particles [84]. The main idea of these pioneering studies was to control the director of the LC hosts by a coupling between LC molecules and elongated magnetic particles, producing what has been called magneto-nematics, i.e., nematic liquid crystals with ferromagnetic properties. In fact, it should be mentioned that this conclusion is wrong, as the produced system is not ferromagnetic. As the magnetic field is removed, the magnetization does in fact decrease to zero, while it should retain a finite value for ferromagnetics. Nevertheless, one could call the produced system an anisotropic ferrofluid. Still, this idea led the trend of studies on LC–nanomaterial suspensions for the next few decades. With the rapid development of composite LC materials, it was recognized that long-range orientation and interactions in mesophase can lead to a strong impact of nanomaterials on the properties of the LC hosts [85,86,87] and vice versa; the LC matrix can rearrange the orientational and positional order of nanomaterials [88,89]. The coupling of long-range ordering in LCs and the unique properties of dopant nanomaterials allows us to change or even impose unique physical properties on the LC–nanomaterial composites by doping different kinds of nanomaterials in LCs [70].

Up to now, there have been four primary groups of nanomaterials used as dopants in LCs, i.e., metal nanoparticles, ferroelectric nanoparticles, semi-conductor nanoparticles, and carbon nanoparticles. Among these dispersions, improved physical characteristics and novel functionalities can be obtained depending on the physical and chemical characteristics of the dopants as well as the interaction between LCs and nanomaterials [90,91,92,93,94,95]. For example, there are abundant reports declaring that doping metallic, semiconducting, oxide and ferroelectric nanoparticles into LCs can efficiently modify the electrical and mechanical (viscoelastic) properties of LC host, leading to large dielectric and optical anisotropy, low threshold voltage, and an improved electro-optical response [87,92,94,96,97]. In addition, ferroelectric nanoparticles can also increase the order parameter as well as the clearing point of the LC host due to the interaction between the elastic forces of LC molecules and the spontaneous polarization of ferroelectric nanoparticles [86,98,99,100,101]. Doping nematic LCs with ferromagnetic particles can effectively reorient the director of LCs by magnetic fields induced by the coupling of magnetic particles with LC molecules [83,102,103,104,105,106]. At the same time, there have been reports claiming the exact opposite behaviour: an increase in threshold voltage, a slower optical response, and a decrease in order and in transition temperatures. This indicates that the properties of the composite materials are very much dependent on the dopant size, materials employed, preparation conditions and time. It further indicates that a lot more research is needed to fully understand these systems. With largely varying results, even on a qualitative basis, the field of liquid crystal nanoparticle composites is only in its infancy.

Metal and semiconducting nanoparticles have been used to stabilise the mesophase known as the blue phase, effectively broadening the temperature range of this frustrated phase from 1 K to up to 20 K [107,108]. Silica, ferroelectric and metal nanoparticles can effectively influence the electro-optical response of nematic LCs and induce a memory effect, i.e., a residual transmittance can be maintained without external electric fields being applied [109,110,111,112]. Moreover, a frequency-dependent electro-optical response can be obtained in suspensions of metal nanoparticles and nematic LCs due to the coupling of the dielectric properties of the nanoparticles and the LCs [78,113]. It was also reported that carbon nanotubes dispersed in nematic LCs would align along the direction of the LC director, therefore largely changing the conductivity depending on the orientation of the tubes [114]. In fact, it could be shown that the nanotubes can be reoriented elastically by reorienting the director through applied electric or magnetic fields. Fullerenes could effectively improve the switching speed of a nematic host LC [95], and improved electro-optic responses could be induced by doping graphene oxides in LCs [115].

In addition to changing the intrinsic physical properties of the pristine LC materials mentioned above, intriguing photonic functionalities can also be introduced into the composites by combining nanomaterials with LCs, which is considered a promising method for the building of novel metamaterials [116]. Metamaterials are artificial engineered bulks with regular nanostructures that show exotic optical properties [117]. Combining the emerging field of optical metamaterials with LCs provides an extremely attractive quality, i.e., tunability, which is of the utmost importance in applications such as optically addressed spatial light modulators, tunable photonic materials and dynamic holography [118,119,120]. When integrating LCs with metamaterials, one can modify the director alignment of LCs by applying external stimuli and therefore manipulating the overall optical characteristics of the composite [119]. For instance, upon periodically embedding particular metal nanomaterials, such as gold nanoparticles, into LC matrixes, a localized surface plasmon resonance (LSPR) will be obtained, which can be tuned by adjusting the birefringence of the surrounding LC matrix [77,81,121,122,123,124,125]. Apart from the abovementioned technologies, novel photonic properties and applications can be further exploited for nanodopants including semiconducting nanomaterials, dyes, oxides, quantum dots (QDs), etc. [126,127,128,129]. Instead of the SPR effect in the metallic nanomaterial–LC colloidal systems, the main property of interest in these cases is photoluminescence. The optical excitation and emission of these nanomaterials could be tuned effectively through the interaction between nanomaterials and the long-range ordered LC molecules, leading to a series of potential applications such as information storage, displays, LC lasers, etc. [130,131,132,133,134].

At present, one of the central challenges faced by the development of novel LC–nanomaterial composites is reliable assembly of nanoscale building blocks into functional bulk materials by using distorted LCs. It has already been demonstrated that the localisation of a nanoparticle in a distorted LC region allows the decrease of the free energy, and the directional motion of nanomaterials in LCs can be driven by the collective long-range interactions mediated by the LC director field. As a result, the energetic cost of LC defects enables them to entrap and reorient nanomaterials in a reversible manner and with well-manipulated position and orientation [135,136,137,138]. In addition, nanomaterials that are suspended in LCs and have a given anchoring energy at the surface will induce a deformation of the orientation in the surrounding LCs, producing topological defects in the vicinity of the nanoparticle. The elastic distortion exerts a force on neighbouring nanomaterials at a range of up to a few micrometres, resulting in defect lines or points around nanomaterials [139,140,141,142,143,144]. This LC-mediated interaction can be either repulsive or attractive and, as a consequence, can be used as an effective method to manipulate the spatial distance between nanoparticles [88,145,146,147,148]. It is thus anticipated that, by coupling these two LC-mediated effects will result in the assembly of nanomaterials with one-, two- or even three-dimensional (3D) periodic superstructures [143,149,150,151,152,153,154]. Due to the great potential in both fundamental science and applications, this top-down assembly approach, where spontaneous or artificially generated LC textures are utilized as templates for organizing nanoparticles on controlled lattices or confining them in designed defects, has been increasingly coming into the spotlight.

Another interesting research field of LC–nanomaterial technology that is worthwhile to discuss is the self-assembly of lyotropic LCs composed of 1- or 2D anisotropic nanomaterials [3,7]. This topic has been investigated from time to time as people have realized that suspensions of anisodiametric nanomaterials can form LC phases, even at a very low concentration, due to Bernal’s seminal work on suspensions of tobacco mosaic virus [155] and Onsager’s theory of the excluded volume mechanism [156]. However, with the boom in the development of nanotechnology and nanoscience in the last decade, various types of novel nanomaterials have emerged. The size dispersion, shape anisotropy, and surface morphology of the nanomaterials are now well controlled. All of these developments will definitely lead to a dramatic increase in the use of nanoparticle-based LCs. Nowadays, it is well known that suspensions of anisotropic nanorods or nanoplates, viruses, nano-cellulose, carbon nanotubes and graphene oxide can form lyotropic LC phases, which are usually accompanied by surprising physical properties [157,158,159,160,161,162,163,164,165].

In this review, we introduce recent studies on the interactions between LCs and nanomaterials leading to improved or novel physical properties for new applications in various research fields. We first discuss some results on improved electro-optical and other physical properties of LCs by doping with various nanoparticles. This is followed by an outline of the additional exotic photonic functionalities of nanoparticle-doped LCs. In the next section we emphasize the use of LC matrixes for nanoparticle assembly as well as the self-assembly of anisotropic nanoparticles into lyotropic LC phases. In the final part of this review our attention is focused on biological applications of LC–nanoparticle composites, i.e., biosensor and drug delivery systems. We anticipate reaching a wider readership, beyond the liquid crystal community, by presenting some recent examples that are illustrative and representative. The attention of this review is mainly focused on the functionalities of nanomaterial-doped LCs, but the rich LC world also provides further nanotechnology applications such as organic semiconductors, nanoporous materials, ion conducting materials, etc. Readers who are interested in these areas can find more information in a recently published review by Kato et al. [166] and the book by Li [167].

2. Modification of Physical Properties of LC Materials by Nanodopants

LCs have been well identified with the applications in commercial displays due to their outstanding electro-optic performance, which is a consequence of their anisotropy of the dielectric constant and the optical refractive index in combination with their elasticity and ready response to external stimuli. However, with the rapid development of technology, liquid crystal displays (LCDs) can no longer meet the increasingly demanding visual requirements with traditional LC mixtures. In the search for LCDs with lower driving voltage, higher colour contrast ratio and shorter response times, novel materials need to be employed to meet increasing demands. At the same time, for progress in other research and technology areas, such as novel optical devices, advanced materials, nanotechnologies and biosciences, LCs with improved physical properties are needed. Currently, one of the effective primary solutions is introducing various functional nanodopants into existing high-performance LC materials. As reported by many studies, nanodopants, such as ferromagnetic, ferroelectric, semiconducting particles of various shapes, as well as carbon nanoparticles, can effectively modify electro-optic responses as well as other physical characteristics of LCs.

Since Reznikov et al. first reported that dispersing low concentrations of submicron ferroelectric particles (Sn₂P₂S₆) in commercial nematic LCs (ZLI-4801) could enhance the dielectric response, lower the operating voltage and induce a linear response to the applied electric field [168], ferroelectric nanoparticle dispersions have been studied extensively as dopants for LC modifications. It was argued that the dispersion of ferroelectric nanoparticles can induce an enhancement of the orientational order of LCs because of the coupling between the spontaneous polarization of the ferroelectric nanoparticles and the liquid crystal via an elastic field, and hence effectively improve the electro-optical properties of LCs [86,169,170] through increased molecular order. Recently, Singh et al. systemically investigated the physical characteristics, including dielectric constants, threshold voltage and elastic constants, of the composite consisting of the nematic LC 6CHBT and ferroelectric nanoparticles of barium titanate (BaTiO₃) [94]. The ferroelectric nanoparticle–LC composite in their system showed an improved electro-optical characteristic, accompanied by an increase in the dielectric anisotropy as well as a decrease in the threshold voltage. Mishra et al. found that the physical properties of suspensions of ferroelectric nanoparticles of barium titanate dispersed in the nematic liquid crystal 5CB are largely dependent on the concentration of nanoparticles [171]. For samples with a low concentration (<1 wt %), the threshold voltage, elastic constants, dielectric coefficients and ionic conductivity could be effectively changed. For those with high concentrations (>1 wt %), the changes were relatively small. Although there have been numerous papers proposing that doping ferroelectric nanoparticles will improve the physical properties of the dispersion, the experimental results still appear sketchy, sometimes even contradictory, and improvements are not straightforward to achieve. As pointed out above, the effect of ferroelectric particle doping strongly depends on the preparation conditions, time, size of the dopant particles, and various other uncontrollable parameters; for example, barium titanate exhibits a particle size-dependent transition between ferroelectric and paraelectric behaviour, whose transition, in addition, depends on the actual milling process employed to produce the nanoparticles. Glushchenko et al. found that doping ferroelectric nanoparticles (BaTiO₃) did not change the threshold voltage of nematic LCs (5CB) [169], while Klein et al. even presented an increase in the Fréedericksz threshold voltage and no change in the dielectric anisotropy [172]. The reasons for the different behaviour observed may be the uncontrollable ionic contamination, broad size distribution and irregular shapes of particles, poor dispersibility, etc. For example, it has been reported that large ferroelectric particles (>100 nm) can form a polydomain structure but very small particles (<10 nm) may lose their ferroelectricity [98].

A recent study by Zangana et al. systemically investigated the dependence of the electric-optic and dielectric properties of the LC (nematic 5CB and ferroelectric SmC* Felix M4851/050), suspended [173] with barium titanate nanoparticles of different size and concentration (Figure 4). They demonstrated that the ferroelectricity of nanoparticles was highly dependent on the particles’ sizes, and the particles did not influence the nematic LC properties unless the concentration was high enough (>0.5 vol %), where a reduction of electro-optic response time was observed, which was due to the enhancement of dielectric anisotropy. Dubey et al. did a low-frequency dielectric investigation of a Schiff-based LC compound dispersed with BaTiO₃ nanoparticles [174]. They found that the doped LC–ferroelectric nanoparticle suspension exhibited a temperature- and frequency-dependent dielectric permittivity. In addition, due to the spontaneous dipole moment, a huge number of free ions in LCs were captured by the ferroelectric nanodopants, leading to a diminishing of the Maxwell–Wagner–Sillars effect. This will in fact have a pronounced effect on the threshold voltage, and ion capture would be expected to reduce the threshold voltage. Alternatively, this implies that when using a pure, commercially cleaned liquid crystal, the ion concentration is already low, such that only minute effects on the threshold voltage would be expected, explaining some of the discrepancies reported in different publications. Furthermore, the authors found in general that the conductivity, dielectric anisotropy and relaxation frequency of the compound decreased due to the strong interaction between BaTiO₃ particles and LC molecules.

It seems that ferroelectric nanoparticles do modify the physical properties of LC matrixes, sometimes leading to an improvement in the electro-optic performance; however, due to the complicated size-dependent ferroelectricity and aggregation behaviour of ferroelectric nanoparticles in LCs, additional research is necessary.

Except for ferroelectric nanoparticles, inorganic nanoparticles, such as metallic nanoparticles, quantum dots (QDs) and oxides, can also effectively modify the physical parameters of LC materials, sometimes leading to an improvement in the electrical and electro-optic performance.

Just like ferroelectric nanoparticles, metallic nanoparticles, such as gold nanoparticles, are expected to potentially improve the physical properties of liquid crystal materials due to the anticipation that the electrostatic interactions between LC molecules and metal nanoparticles can significantly enhance the orientational order of LCs [175,176,177,178,179,180,181]. The argument of performance enhancement through an increase in orientational order has been used in many publications to explain the observed results. Nevertheless, to our knowledge such an increase has not been demonstrated quantitatively by a strongly reliable experimental method, such as X-ray diffraction at equal reduced temperatures, for example, with values for the neat and the doped system being well outside the limits of error. Recently, Elkhalgi et al. investigated the thermodynamics, electro-optic response and dielectric permittivity of a nematic LC (6CHBT)-gold nanoparticle composite [182] (Figure 5). It was found that the threshold voltage required for switching LC molecules from homogeneous to homeotropic alignment was decreased, while the first phase transition temperature (isotropic to nematic) as well as the dielectric permittivity for the composite, was raised due to the increased nematic ordering of the host LC medium. Apart from that, the presence of gold nanoparticles also promoted the reorientational motion of LC molecules by applying an electric field. Furthermore, an increase of dielectric permittivity and ionic conductivity was found in the gold nanoparticle-columnar discotic LC (HAT4) suspensions [183]. Silver nanoparticles are also quite popular among the metal nanoparticles and have been widely used as nanodopants for the modification of LC physical properties [179] due to their high electric and thermal conductivity, nano-scale calorimetric effect, excellent chemical stability, and various medical and catalysts applications [184,185,186]. The thermodynamic, dielectric and electro-optic characteristics of a nematic LC (6CHBT) mixed with silver nanoparticles with different sizes have been studied by Tripathi et al. They found that the first-order phase transition temperature of the mixture was marginally increased, while the dielectric permittivity, and elastic constants decreased compared to the pure LC [187].

QDs are semiconductor nanoparticles in the range from two to 10 nanometres in diameter. They are usually composed of thousands of atoms of group II and VI elements (e.g., CdSe, CdTe, ZnO), group III and V elements (e.g., InP, InAs), and group IV–VI elements (e.g., PbS) [188]. As a semiconductor material, the electrons of QDs will be promoted from the valence band to the conduction band by the irradiation of photons with energies larger than the band gap. Since these electrons are not stable in the conduction band, they will transit back to the valence band and recombine with the holes. As a result, the extra energy will be emitted as photons with a wavelength related to the bandgap. According to the developed bulk Wannier-Hamiltonian model of Luis Brus [189,190], this bandgap is dependent on the size of the employed QDs. Their unique size- and shape-dependent electro-optic characteristics made QDs the focal point of research in various fields including photonics, sensors, medicine, information storage, etc.

The special physical properties have made QDs a promising nanodopant for modifying physical properties and improving the electro-optic responses of LC materials [191,192,193,194]. Singh et al. recently found that the inclusion of QDs (CdSe-QDs) would effectively influence phase transition temperatures and enhance the dielectric anisotropy of nematic LCs (4PP4OB). Moreover, it increased the transition voltage for switching LC molecules from homogeneous alignment to homeotropic alignment, decreased the splay elastic constant, and increased the relaxation frequency of the LC matrix [195]. Roy et al. reported that nematic LC 1832A dispersed with QDs (InP/ZnS) exhibited a promoted molecule alignment, leading to an enhanced non-linear optical anisotropy. They also found that the physical characteristics of the composite, including elastic constants, switching time, and viscosity, were distinctly decreased, while faster electro-optical response, enhanced birefringence and larger dielectric anisotropy were observed [196].

Oxides such as zinc and titanium oxide nanoparticles have also regularly been used as dopants in LC materials due to their exotic electrical and optical properties. Numerous works focusing on the physical properties of LCs with oxide entities demonstrated new or improved electro-optic performance, including altered dielectric anisotropy, lower threshold voltage or shorter response and relaxation time [197,198,199]. In a recent publication, Oh et al. demonstrated the effects on the physical characteristics of nematic LCs (ZSM-50087XX, JNC) by doping with titanium silicon oxide (TiSiO₄) nanoparticles. The authors found that the electro-optic characteristics of the composite were apparently improved. By increasing the doping concentration, the electro-optic properties can be further improved until the doping concentration was too large (0.2 wt %), which would influence the LC behaviour and deteriorate the electro-optic behaviour. In addition, the authors also demonstrated that the dispersed TiSiO₄ nanoparticles could effectively improve the thermal stability of the pristine LC system, capture impurity ions and thus lead to hysteresis-free voltage-transmittance characteristics [200].

Another kind of nanomaterial that has been widely used as a dopant for influencing the physical properties of LC materials is carbon nanomaterials. Carbon nanomaterials, including fullerene, carbon nanotube, graphene and graphene oxides, have attracted great attention due to their exotic physical and chemical properties.

Fullerenes are hollow spherical molecules of nano-sized diameter, generally composed of 60–70 carbon atoms (C₆₀, C₇₀). Among the various fullerenes, C₆₀ has been studied particularly extensively. The molecule of C₆₀ is a strong acceptor capable of accepting from one to six electrons to form the corresponding anions [201]. In order to improve the electrical and electro-optic properties of LC materials, great efforts have been made to attach C₆₀ covalently to mesogens as well as to dope C₆₀ into LC matrices [95,202,203,204]. The recent, significant developments in fullerene-containing LCs have been reviewed by Zhang et al. [205].

In 1991, the discovery of quasi-1D carbon nanotubes (CNTs) was reported by Iijima [206], which ignited scientific interest in 1D nanostructures. CNTs are crystalline allotropes of carbon with atoms hexagonally arranged in curved lattices, forming hollow cylinders with diameters on the nanometre scale and aspect ratios up to 10⁸. The CNTs composed of a single layer are called Single-Walled CNTs (SWCNTs), which can be either metallic or semi-conductive, depending on the diameter and rolling angle, while the ones formed by several concentric cylinders are called Multi-Walled CNTs (MWCNTs). Due to their exceptional anisotropic thermal, electrical, optical and mechanical properties, CNTs have been found to have great potential as additives to LCs, which will result in the improvement of electro-optic properties. A few papers have demonstrated the beneficial effects of CNTs doping on the physical properties of LC materials, such as reduced amplitude of residual DC, altered dielectric anisotropy, lower threshold voltage and shorter response time [207,208,209,210,211,212,213]. This improved electro-optical performance can be attributed to the higher dielectric anisotropy of the suspension, due to the extremely high permittivity and anisotropy of CNT inclusions as well as the ion-trapping effect of CNTs reducing the free ionic concentration of the solvent, which leads to an increase in the effective applied cell voltage (a reduction of electric field shielding). The inclusion of CNTs in LCs can also lead to other modifications of physical properties, such as viscosity, elastic constants and phase transition temperature, to name just a few. Very recently, Singh et al. reported the enhancement of orientational order, increase of dielectric anisotropy and decrement of threshold voltage in the composite of SWCNTs and nematic LCs (3017) [214,215] (Figure 6). However, just like ferroelectric nanoparticles, the experimental results of the CNT-LC switching behaviour are not consistent. Some authors claimed neither a decrease in threshold voltage nor an increase in dielectric anisotropy was found in respective composites. On the contrary, a higher concentration of CNTs could even increase the switching voltage [216,217].

Another allotrope of carbon, graphene, which is a monolayer of sp2-hybridized carbon atoms arranged into a 2D honeycomb lattice, has been one of the most attractive nanomaterials since its discovery in 2004 [218]. Due to the excellent mechanical, electrical, optical, thermal and electronic properties, such as high thermal conductivity, charge mobility, mechanical strength, optical transmittance and chemical stability, graphene has been widely applied in a variety of applications such as electronics, energy storage, biomedical engineering, new composite materials, or waste water management. However, graphene is known for its notoriously poor dispersibility and solubility in solvents, organic liquids, including LCs, which greatly inhibits their use in LC composites so far, despite large efforts having been made to increase dispersibility via surface treatment and surface decoration with covalently bound mesogens. In contrast to graphene, graphene oxide (GO), obtained by treating graphite with strong oxiders, is readily dispersible in solvents, including water, due to its hydrophilic nature. In fact, in certain concentration ranges, GO actually forms a lyotropic liquid crystal in many such solvents, as well as in water. Although GO is electrically insulating, it can be converted into conducting graphene (rGO) by chemical or thermal reduction methods, at least partially regaining the conductive properties of neat graphene.

So far, a series of studies have shown some improvement in the electro-optic performance of thermotropic LCs doped with graphene oxide (GO) [219,220,221,222,223,224]. The reason for this improvement may be the coupling of electrical dipoles on the surface of the GO with the ones of mesogens and the trapping of free ionic contamination by GO. Zangana et al. have dispersed GO of different sizes in a nematic LC (5CB) with a wide range of concentrations (Figure 7). They found that the LC molecules would be strongly anchored on the surface of GO flakes. As a result, the threshold voltage and elastic constants were strongly increased, while the elastically driven electro-optic off-response time decreased when increasing the concentration of GO flakes [225,226]. Similar results in GO-doped nematic LC (E_5CN7) were reported by Dalir et al. They found that the capacity of charge storage was expanded, while the mobility of free ions and the diffusion coefficient were significantly decreased due to the strong interaction between LC molecules and GO flakes [227]. In another publication, they also reported that GO doping would lead to an increase in the nematic-isotropic phase transition temperature, which could be shifted by changing the GO concentration [224]. Moreover, Lapanik et al. reported a 30–50% reduction of the threshold voltage and switching time in partially reduced graphene oxide (PRGO)-doped nematic LCs and a 20–25% increment of the spontaneous polarization in PRGO-doped ferroelectric LCs [228]. Ozgan et al. found that the dielectric anisotropy of nematic LC (6CB) would be increased with an increasing concentration of GO [115]. In a very recent investigation, Mrukiewicz et al. systematically investigated the physical properties of nematic LC (5CB) dispersed with GO at a wide concentration range from 0.05 to 0.3 wt %. They observed a decrease in the threshold voltage in electro-optic and dielectric spectroscopy measurements. This was attributed to the disrupted homogeneous alignment induced by the strong π–π stacking effect between the graphene oxide and the LC molecules’ benzene rings. In addition, they presented the dependence of the first-order phase transition temperature, elastic constants, dielectric permittivity and switching duration on the GO doping concentration [229]. Additionally, except for fullerenes, CNTs and GOs, modified physical properties of LC hosts can also arise from other carbon nanoparticles including carbon dots [230], detonation nanodiamonds [231] and graphene quantum dots [232].

A small number of added nanomaterials in LCs can produce composite materials with modified physical properties and improved electro-optical performance—for example, a shorter response time, a higher birefringence, a lower threshold voltage, a larger dielectric anisotropy, better contrast, enhanced nonlinear-optical properties, etc., which has great potential for the design of next-generation LC equipment including LC displays, tunable LC lasers and filters, electro-optical switchers and shutters, nonlinear-optical valves for photonic information processing systems, telecommunication and many other aspects of modern technology. At present, though, experimental results are still inconclusive, lacking reproducibility. Thus, further research is needed in the near future, both experimental as well as theoretical, to identify the main underlying aspects that lead to the modification of liquid-crystal-based composite materials’ properties, as well as the causes for the range of behaviours observed. Developing a good understanding of the latter will clear the way for an optimized development of novel materials for improved future technologies.

3. Photonic Applications

Surface plasmon waves, which are the collective oscillation in electron density at the interface of a dielectric and a metallic material, have received great attention since the 1960s [233]. They offer the opportunity to merge electronics and photonics at nanoscale dimensions due to their particular capability to guide and localize electromagnetic waves in sub-wavelength metallic structures, and have been widely exploited in various research fields including transforming photo-conversion processes [234,235], improving weak fluorescence signals for biological imaging [236,237], enabling nano-lasers or spontaneous amplified irradiation [238,239], realizing photo-thermal conversion [240] and so on. Moreover, a series of plasmonic devices, such as light sources, waveguides, lenses, filters, or dynamic colour displays, have already been demonstrated [241,242,243,244].

In the past few decades, metal nanoparticles have already been widely utilized as functional building blocks for obtaining new generations of nanomaterials with tunable surface plasmonic resonance (SPR). Metal nanoparticles show SPR when stimulated by electromagnetic radiation of an appropriate wavelength due to the quantum confinement effect of free electrons. In particular, gold nanoparticles have received great attention and been found to be pivotal for SPR applications such as scanning near-field optical microscopy [245] and biosensors [246] for several reasons. First, the plasmonic resonance of most metals occurs in the ultraviolet (UV) region of the spectrum, while gold shows this resonance in the visible region. Secondly, gold is chemically inert and does not oxidize, which is a very important property in optical applications since oxide layers will drastically complicate the calculation of optical properties. Thirdly, gold nanoparticles are compatible with biological systems, which will be quite important for investigations and in vivo analysis. Last but not least, there are now many new physical and chemical methods for the synthesis of gold nanoparticles with a wide range of diameters and shapes [247].

The optical characteristics of gold nanoparticles in visible and near-infrared (NIR) wavelength ranges are mainly determined by the collective response of electrons. The strong interactions between the incident electro-magnetic field and free electron clouds at the gold nanoparticle’s surface will displace the electron clouds from their equilibrium position, thus inducing surface polarization charges that act as a restoring force on the electron clouds. This collective oscillation of free electrons will enhance the electric field that is produced at the interfaces of nanoparticles, and also confine it to a scale shorter than the wavelength of the incident electromagnetic field. At a macroscopic scale, such resonant excitation will lead to both strong scattering and absorption of light [247,248,249].

As a result, SPR is highly dependent on the diameter, shape, and dielectric constant of gold nanoparticles as well as the surrounding medium [248,250,251,252,253], implying that the SPR of gold nanoparticles can be tuned by changing these parameters. Although recent progress in the fabrication of nanomaterials has enabled us to produce metal nanoparticles with arbitrary size and shape, it is still time-consuming and costly to design plasmonic architectures with complicated structures for different optical applications. New methods for externally tuning SPR are still in demand. Among the various methods, one feasible approach is combining plasmonic units with host matrixes whose dielectric characteristics can be manipulated through external stimuli [254]. To this end, LCs have come into the spotlight. LCs hold great potential for enabling effective manipulation over the optic characteristics of plasmonic architectures due to their large refractive index anisotropy, real-time response to the external stimuli and the unique ability to act as dynamic templates for the self-assembly of nanomaterials. As a result, LCs have been widely used to manipulate SPR effects, not only in nanoparticles, but also in binary gratings, antenna arrays, patterned conductive surfaces and metamaterials.

To date, a wealth of gold nanoparticle–LC composites with tunable SPR have been reported [81,82,124,125,181,254,255,256,257,258,259,260,261,262,263]. In recent years, among the different kinds of gold nanoparticles, due to the anisotropic mechanical and optical properties, gold nanorods have especially attracted attention [263,264]. Very recently, a LC modulated tunable filter based on SPR was designed and discussed by Li et al. [265]. The filter consisted of Au nanorod arrays and an Au film separated by a dielectric ITO glass layer. The numerical simulation demonstrated that the filter could achieve double absorption peaks at NIR wavelengths, which were generated by the strong plasmonic coupling between the plasmon-induced transparency mode of the Au nanorod dimer and the cavity mode of the Au cavity structure. A tunable band range of 160 nm was confirmed. On the other hand, apart from the size and shape of single nanoparticles, the SPR also depends on the interparticle spacing and spatial structure of nanoparticles. Rozic et al. demonstrated a self-organization of gold nanorods into end-to-end chains by using oriented linear arrays of smetic A defects (see Figure 8). These nano-chains led to strongly anisotropic SPR absorption, which varied from 530 nm to 920 nm when end-to-end coupled gold nanorods were employed [266].

Another possible SPR effect, which is attractive but as yet largely unexploited, arises from the design and realization of tunable emissive materials, in which the SPR-related losses of light can be compensated for by the corresponding enhancement effect. For instance, in [238], nanoparticles composed of gold cores and dye-doped silica shells could efficiently offset the loss of SPR through the gain of dyes, thus enabling surface plasmon amplification by stimulated emission of radiation. Based on this idea, the authors of [267] successfully realized the control of both the intensity and decay rates of fluorescence by combining the absorption of SPR and the fluorescence anisotropy of dyes (Figure 9). In this system, the LC medium acts as an active template, imposing orientational ordering on the nanorods and mediating synchronous switching of the anisotropic particle orientation. Although gold nanoparticles have been widely used as plasmonic building blocks for various plasmonic optical applications, they are rarely applied in applications such as optical shutters, smart windows or spatial light modulators. The main reason is due to the confinement of the size and shape. The SPR peaks of these gold nanoparticles are normally limited to within the near-infrared (NIR) spectral range. One effective solution to this problem was put forward by Sheetah et al. [268]. They developed a guest–host mesostructured composite in which dye molecules and plasmonic nanorods spontaneously aligned either parallel or orthogonally to the director of the LC (5CB and AMLC-0010) host. Such a system allows the feasible tunability of the SPR peaks from the NIR to the visible spectral range. By changing the geometry of nanoparticles and different kinds of LC materials (nematic and cholesteric LCs), the authors also showed that the switching time of the composite could be tuned from milliseconds to seconds and even the polarization of the transmitted light could be manipulated according to the application demands. Another solution, which was proposed by the same group, was co-dispersing different anisotropic nanoparticles in LCs [269]. The authors used gold nanorods with varying aspect ratios in a nematic LC host (5CB). By applying appropriate surface treatments to the particles, the nanorods oriented either parallel or orthogonal to the LC director. With the application of an electric field, the nanorods would rotate their short or long axes due to elastic coupling between particles and the LC director, thus leading to a tunable SPR spectral range as well as polarization-dependent optical characteristics.

Recent progress has demonstrated that SPR also exists in monolayers of graphene [270], graphene ribbons [271] and other 2D materials [272,273]. Compared to the SPR in gold nanoparticles, the SPR in graphene possesses lower propagation loss and stronger mode confinement in the mid-IR spectra [218,274,275]. Reshetnyak et al. recently discussed the theoretical possibility of manipulating the SPR effect of graphene monolayers and ribbons by modulating the dielectric properties of the neighbouring LC media [276]. According to their modelling, the graphene monolayers/ribbons were coated with a LC layer and, by altering the applied electric field or the surface anchoring, the LC director would be reoriented and thus vary the SPR spectra due to changes in the dielectric characteristics of the LC layer on the graphene monolayers/ribbons and the polarization of the incident light.

On the other hand, due to the enhanced light absorption of SPR, gold nanoparticles and graphene can also act as heat nano-sources when excited by a specific wavelength of light, i.e., act as photothermal agents. Over the past decade, the photothermal effect has been a fertile ground for fundamentally new scientific research. Among the various inorganic (e.g., metal and carbon nanomaterials) and organic (e.g., indocyanine green and polyaniline) photothermal agents, gold nanoparticles and graphene have attracted particular attention due to their high photothermal conversion efficiency as well as their NIR-absorbing spectra, which make them extraordinarily popular in biological applications [277,278,279,280]. Remote manipulation of LCs through the photothermal effects induced by particular nanomaterials is an emerging field of promising materials [281,282,283,284,285]. The functional nanoparticle–LC composites are driven by light when the photothermal nanomaterial absorbs the light energy and converts it into heat, leading to a local increase in temperature. Consequently, the remotely triggered local temperature increase induces morphological and physical changes [23]. Recently, Palermo et al. presented an optical manipulation of the plasmonic thermal effect induced by randomly distributed gold nanoparticles coupled with a nematic LCs director field (E7) as an active medium. The authors reoriented the LC director by employing a photoalignment material, which produced changes in both the thermal conductivity and the refractive index of the LCs surrounding the nanoparticles, thus enabling optical control of their SPR and of the resulting heating [286]. Moreover, a tunable helical pitch and a reversible handedness inversion of cholesteric LCs (CLCs), enabled by the photothermal effect, were demonstrated by Wang et al. The authors dispersed mesogen-functionalized gold nanorods into CLCs. Upon irradiation with NIR light, the photothermal effect of the dispersed gold nanorods commenced, triggering the change of the helical pitch and the handedness inversion of the helical superstructure from left-handedness to right-handedness through an untwisted transient state (see Figure 10). The effect is based on the temperature dependence of the pitch of the helical superstructure of the cholesteric (chiral nematic) liquid crystal. To achieve an inversion of the helical handedness, special materials or mixtures are needed, which exhibit a temperature-induced twist inversion. Together with the optical properties of the cholesteric phase, known as selective reflection, the structure reflects left- or right-handed circular polarized light, depending on the handedness of the CLC helix, while the wavelength of the reflected circular polarized light depends on the pitch and average refractive index of the helix. By removing the NIR irradiation, the reverse process would spontaneously occur [287]. In other recent work, the authors designed an adaptive smart window whose optical transmittance could be switched between transparent and opaque states due to the phase transition between the SmA* and the N* phase induced by the photothermal effect of dispersed graphene (Figure 10c) [288].

Light-driven phase transitions in LCs are a fascinating topic from both the scientific and the technological point of view. Apart from the photothermal effect [289], such phase transition can also be obtained by the photoisomerization ability of azo benzene dyes [290,291,292,293,294], which, at the same time, is also usually used for fabricating LC lasers [32,295,296,297,298]. However, because these dyes are not nanoparticles, they will not be discussed within the framework of this review. However, since we have mentioned LC lasers on several occasions, it is sensible to go into somewhat more detail by discussing quantum dots (QDs) in this respect.

As we mentioned above, QDs are semiconductor nanoparticles with unique physical characteristics such as high quantum yields, large absorption cross-section, tunability of fluorescence emission, excellent colour purity, broad excitation spectra, good stability and long lifetimes. As a result, QDs have already been widely used as a nanodopant in LC materials for introducing additional photonic functionalities to LCs including LC lasers. LC lasers are usually made by doping laser dyes (e.g., DCM) into cholesteric LCs (CLCs), which, due to the selective reflection, are photonic band gap materials. A CLC has a 1D modulation of the refractive index due to its unique self-organized helical superstructures, in which rod-like LC molecules self-assemble into monomolecular layers, with their long axes continuously rotating along the helical axis, perpendicular to the director. The helical superstructure is described by its pitch length P and the helical handedness, where P is the distance along the helix axis over which the director rotates 360°, while the helix handedness is the direction of rotation. According to Bragg’s law, such helical superstructures can selectively reflect circularly polarized light with the same handedness as the helix and the reflection wavelength is determined by its pitch λ = <n>P, where <n> is the average refractive index. Thus, CLCs are usually considered a 1D photonic crystal with an adjustable photonic bandgap (PBGs) via the temperature dependence of the pitch P. After doping the gain medium (e.g., dyes and QDs), the CLC helix acts as a resonator necessary for laser emission, which can be tuned by various external stimuli. Compared to conventional LC lasers, which are fabricated by dispersing laser dyes into CLCs, LC lasers based on QDs can lase emissions with improved features such as narrower line width, lower lasing threshold and low intensity to noise ratios, as well as high stability due to the 3D quantum confinement effect of QDs [131,299,300,301] (Figure 11). On the other hand, amplified spontaneous emission (ASE) from colloids of QDs dispersed in LCs has also been investigated widely [301]. Cao et al. reported an ASE effect of QDs doped into polymer-dispersed LCs (PDLCs). They found that, due to the multiple scattering in the LC–polymer matrix, the dwell time and path distance of light in the active media was drastically increased, thus leading to enhanced ASE and s decreased threshold. Such a QD-PDLC system is promising for applications such as random fibre lasers and laser amplifiers [302]. Recently, tunable ASE in graphene QD-doped CLCs have also been realized by the authors [303].

Nanomaterials with SPR effects, fluorescence emission and other particular optical properties can introduce unique photonic functionalities to LC materials. Vice versa, LCs with large refractive index and long-range elastic anisotropies can function as dynamic medium for externally tuning the optical properties of nanomaterials. Such multifunctional nanomaterial–LC composites pave the way for the realization of various novel optic devices.

4. LC-Aided-Assembly and Self-Assembly of Nanomaterials

As we have seen, nanomaterials show unique physical and chemical properties that are significantly different from bulk materials as well as from individual atoms and molecules. However, single nanomaterial entities cannot be directly applied in functional devices. The generation of macroscopic assemblies of nanomaterials with controlled localization and orientation is necessary for their practical application. A common method is to use a “bottom-up” approach whereby functional nanomaterials are designed and directed through self-organization into a particular structure, which requires the targeted ability to manipulate interactions between nanomaterials. To this end, the properties of self-organization of LCs may be exploited. The unique ability of LC materials to self-assemble into long-range ordered superstructures has been widely used as a potential method for synthesizing various functional materials with complicated spatial structures. By virtue of the chemical reactions performed in aqueous lyotropic LCs, different mesoporous materials have been successfully obtained [304]. In addition, thermotropic LCs can also act as an appropriate medium to design “top-down” approaches. LCs composed of anisotropic molecules possess a long-range orientational or even positional order and a high susceptibility to various external stimuli. Thanks to these properties, anisotropic torques and forces will be imposed on the micro-additives when they are dispersed in a LC matrix, which will also direct them on a macroscopic scale or even arrange their position on specific micro-scale locations. In the past several decades, a large number of studies have demonstrated the interactions between microparticles in LCs [145,147,148,305,306] as well as between microparticles and LC defects [137,307,308]. Some of them even reported the successful fabrication of complex periodic superstructures consisting of microparticles [146,149,150,154,308,309]. However, when it comes to the nanoscale, the problem becomes much more complicated because of the uncontrolled aggregation as well as the limited number of effective characterization methods.

Nanoparticles localized in distorted LC regions can effectively decrease the free energy of the system due to the release of the elastic energy generated by the volume occupied by the nanoparticles. In the presence of a distortion gradient, the nanoparticles tend to be automatically dragged into the centre of the distortion [310,311,312,313]. As a result, topological defects, which are the cores surrounded by strongly elastically distorted regions, are usually used as templates for the bottom-up assembly of nanoparticles [314]. Moreover, in the case of anisotropic nanoparticles, the surface anchoring energy plays a vital role in the orientation of nanoparticles. With tangential anchoring, the particles tend to align along the nematic director, while, with radial anchoring, the alignment of the particles tends to be perpendicular to the director. Furthermore, nanoparticles suspended in LCs with a given anchoring energy at the surface will induce a local deformation of LC director field. According to the studies of Koenig et al. [142], the elastic distortion exerts a weak force (<5 k_BT) on neighbouring nanoparticles at a range of up to a few micrometres. Coupling the LC-mediated elastic interaction between nanoparticles with the long-range drag force exerted by the distortion gradient of director field, nanoparticle assemblies with ordered mesostructures in LCs are expected to be obtained, which will potentially be useful in applications connected with the fields of photonics and electronics [128,254,261,263,264,315,316,317,318,319].

It has been reported that gold nanoparticles can be aligned along linear disclinations [316,320] and trapped in the centre of toric focal conic defects of smectic phases [321]. Moreover, they can be captured by the topological defects induced by the inclusion of microparticles in nematic phases [311] (Figure 12). Such periodically organized gold nanoparticles are especially interesting for optical applications due to their intriguing SPR effects. Recently, Rozic et al. demonstrated that the orientation of gold nanorods with a large range of diameters could be oriented along the direction of the linear arrays of smectic A defects, so-called smectic oily streaks. The gold nanorods self-organized into chains, increasing the density of nanorods. The gap between nanorods could be controlled through the balance between the steric repulsion and the van der Waals attraction. Furthermore, it was found that these nanorods were more likely to be trapped by the smectic dislocations with respect to ribbon-like defects. Such special microstructures of gold nanorods could lead to an electromagnetic coupling controlled by light polarization [266].

Additionally, as mentioned above, CNTs are famous for their remarkable anisotropic thermal, electrical, optical and mechanical characteristics as well as their exceptional thermal and chemical stability and have attracted interest in relation to various applications such as sensors, electronic devices, conductive and transparent films, as well as strong, light and conductive fibres. The realization of macroscopic ordered assemblies of CNTs is significant for the development of electronics, new materials and energy storage devices. In the past few years, some studies have reported on the ordered alignment and assemblies of CNTs in lyotropic LCs [322,323,324]. CNTs dispersed in films of polymerizable lyotropic LC surfactants can be aligned in the film thickness direction by the alignment of a suitable host mesophase under an external magnetic field [325], and filaments of uniformly aligned CNTs can be extracted from CNT-dispersed lyotropic LCs [326]. In a recent investigation, Kasprzak et al. reported that, by using photo-polymerization, a large group of SWCNT-based polymer nanocomposites was obtained from polymerizable quaternary ammonium surfactants. Efficient dispersion of CNTs within cylindrical micelles was observed due to the physical adsorption of the nonpolar tails of the surfactants. After photo-polymerization, the composites with SWCNTs are organized into ordered nanostructures with both lamellar and gyroid nanostructures [327]. Apart from this, the alignment and assembly of CNTs in traditional thermotropic LCs [2,328,329,330,331] and macromolecular LCs [332,333] have also been reported.

Apart from the relatively simple one- (organized by nematics, smectics and cholesterics) and two- (columnar phases) dimensional nanoparticle structures, the structurally rich world of LCs also provides phases with complicated spatial modulation of density and molecular orientation, periodic in three (blue and cubic phases) dimensions [334]. As a distinct class of frustrated phases, Blue Phases (BPs) are composed of double-twist cylinders, energetically favoured at high enough molecular chirality, but always leading to geometric frustration since it is impossible to tile the whole 3D space with double-twist cylinders without introducing disclination lines. These disclination lines either self-assemble into a 3D periodic regular lattice (BPI and BPII) or remain amorphously distorted (BPIII), which make BPs potential candidates for the 3D template of trapping sites for nanoparticles [335]. According to the computer simulations of Stratford et al. [318], nanoparticle–BPLC composites can give rise to regular crystals, glasses, percolating gels, isolated clusters, twisted rings and undulating colloidal ropes (Figure 13a). The wide variety of structures can be tuned via the concentration and anchoring condition of particles as well as external stimuli. Such a new class of composites will be promising for switchable, multistable devices for optical technologies such as smart glass and e-paper. Furthermore, particle trapping can also substantially increase the thermal stability range of the blue phase, by a factor of two or more [107,108,336] (Figure 13b). Apart from the disclinations in BPs, other defects such as focal conic domains in smectic LCs [321], fingerprint textures in CLCs [337,338,339], or even the LC–water interface [340], as well as the topological defects induced by microparticles [311], can also be used to trap nanoparticles and produce periodic arrays.

On the other hand, it is well known that 1D or 2D anisotropic nanoparticles, such as CNTs and GOs, can act as LC building blocks in isotropic solvents and will entropically form lyotropic liquid crystalline phases above a critical volume fraction due to the excluded volume interactions, according to Onsager’s theory. According to the simple steric model, the molecular configurations are dominated by steric effects, which are dependent on the molecule concentration. In a dilute solution, the molecular configuration behaves isotropically due to the large free volume. In a condensed solution, the free volume is inhibited by the concentrated inclusions and, as a consequence, the suspension becomes liquid crystalline. In such suspensions, the nanoparticles exhibit the long-range orientational order with the thermodynamic stability [341]. Such nanoparticle LCs are particularly interesting and hold great potential in industrial applications such as thin films, fibres and various multifunctional composites. A variety of different materials have recently been studied to an increasing extent, such as carbon nanotubes, graphene oxide, nanorods and nanowires, viruses, clays and nanocrystalline cellulose.

Because of their anisotropic shape, CNTs may form lyotropic nematic LC phases when they are suspended in an isotropic liquid medium at a sufficiently high concentration [342,343,344]. At the same time, such carbon nanotube liquid crystals (CNTLCs) can also be formed with the assistance of external forces such as magnetic field [345], electric field [331], and especially shear [346]. Nevertheless, suspending CNTs at a high concentration in a solvent is not always straightforward. CNTs are notoriously difficult to disperse in any medium and the long-term stability of the dispersion is poor because of the strong attraction from van der Waals forces between CNTs. To obtain a homogeneous suspension of CNTs, various chemical approaches have been employed [347,348,349,350,351]. In addition, CNTLCs usually show numerous topological defects that do not coarsen with time due to their length and waviness [352,353], as well as the high viscosity (Figure 14a–d). This is unfavourable for the optimum performance of CNTs and thus seriously hinders their applications. Methods such as mechanical shearing [354], as well as changing the morphology of CNTs [355], have been applied to obtain homogeneously aligned CNTLCs. CNTs are excellent conductors and have exceptional conductivity anisotropy. The electrical conductivity along the main axis of the tube can typically be several orders of magnitude larger than that in the tube’s radial direction. As a consequence, with the assistance of self-assembly CNTLCs, CNT films with specific structures, vertically aligned CNT arrays or horizontally aligned CNT arrays, can be obtained, which will have considerable potential to be used as transparent electrodes [356,357,358,359]. Compared to the traditional ITO electrodes, which are suffering from difficulties such as the brittleness of ITO, the rising cost of indium, and the high vacuum and temperature processing during the production, CNT electrodes have great advantages in terms of their flexibility, low cost, safe fabrication process and special conductivity anisotropy, which make them a good substitute for ITO in the LCD industries [360,361]. CNT films can also act as alignment layers to orient LC molecules via the CNT substrates due to the orientational structure and the π–π interaction between them [212,362,363,364] (Figure 14e,f). Apart from that, CNTs also show anisotropic optical responses due to the anisotropy and the graphite π electrons. When the polarization of the incident light is perpendicular to the tube axis, there will be no attenuation; in contrast, if they are parallel to each other, there will be strong absorption. Such anisotropic absorption not only exists in the visible region, but also in ultraviolet and infrared regions. This optical property produces a highly aligned CNT film to be used in various optical applications [365,366,367] (Figure 14i,j). Last but not least, CNTLCs can also be used to fabricate continuous and multifilament fibres. These lightweight fibres show a uniform morphology with few defects and a high degree of CNT alignment (Figure 14g,h). They have excellent mechanical, electrical and thermal properties and can be applied in the field of smart and strong materials including smart textiles, conductive composites and electric wires [368,369,370].

Furthermore, since GO has an extremely large aspect ratio, it is expected that aqueous GO dispersions can also self-assemble into lyotropic LCs according to Onsager’s theory (Figure 15a,b). Although there have been a fair number of studies focused on the liquid crystallinity of aqueous GO dispersions, it should be noticed that the first relevant study for graphene LCs was that of Behabtu et al., who dispersed graphene in chlorosulfonic acid [163]. One should recall that graphene is hardly soluble in water and organic solvents, which severely limits its application in LCs. Instead, GO has long been known to disperse well in water due to its tunable amphiphilic properties. In 2011, Kim et al. first reported the liquid crystallinity of GO [164]. The liquid crystalline properties of GO dispersions have since been studied intensively. Early investigations found that the liquid crystalline properties of GO dispersions are highly dependent on the concentration [165,371] and aspect ratio of GO flakes [372]. A high concentration of GO can lead to a “pseudo” lamellar phase [165] and a tunable structural colour can be obtained by varying the concentration [371], while a large aspect ratio of GO flakes can efficiently decrease the critical concentration for the spontaneous formation of the nematic phase [373]. Furthermore, Jalili et al. found that the ability of GO to form lyotropic LCs is largely due to the polarity and the capacity of the solvents to form hydrogen bonds with GO flakes [374], while the groups of Xu et al. [375] and Tkacz et al. [376] discovered that the LC behaviour of GO dispersions was also related to the salt concentration and pH value, respectively. Kim et al. [164] and Shen et al. [377] showed that the macroscopic alignment of GO LC could be induced by magnetic and/or electric fields. Very recently, a systematic analysis of the electro-optical response of GO LCs under alternating fields was presented by Guerrero et al. [378] and a shear-induced stripe assembly was reported by Hong et al. [379]. Due to these intriguing physical properties, various potential applications based on GO LCs (Figure 15c–e), such as GO LC gels with “shape memory” effect [380], photoluminescence imaging [381], nanocomposites [382,383,384], sensitive strain sensors [385], transparent electrodes [386], conductive fibres [387,388,389,390], multifunctional yarns [391] and membranes [392], flame-retardant nanocoatings [393], optical devices [394] and biomedical applications [395], have been anticipated and some of them have been realized. More relevant information can be found in a recent review published by Sasikala et al. [396].

Before the conclusion of this section, it is necessary to introduce another nanomaterial that can self-assemble into cholesteric lyotropic LCs, i.e., cellulose nanocrystals (Figure 16). Cellulose nanocrystals are rod-like nanoparticles with a relatively high aspect ratio (from 10 to over 100) that can be extracted from an extensive range of materials including wood, bacteria, algae and tunicate [397,398,399]. Cellulose LCs have attracted much interest due to their mechanical and optical properties. The individual cellulose nanocrystal (CNC) rods are transparent across the visible spectrum and have low density and excellent mechanical strength. It is well known that suspensions of cellulose with an appropriate concentration can spontaneously form a CLC phase [159,400,401]. However, it is still not completely understood how the transfer of chirality from asymmetric carbons in the molecular structure of CNCs to the macroscopic scale of the LC phases occurs. A recent study by Usov et al. confirmed that the mechanism of chirality transfer involves multiple steps, including a geometric effect whereby the CNCs acquire a uniformly right-handed twisted morphology and transfer this to the helical modulation of the macroscopic phase [402]. The induced helical pitch of the cholesteric CNC suspension is also intricate. Although many studies have demonstrated a decrease of pitch with increasing CNC concentration [403,404,405], the CNCs in all these cases are electrostatically charged. It is thus unlikely that the decrease of the pitch is solely due to the concentration increase. It may also be related to electrostatic screening effects [406,407,408]. So far, the manipulation of the pitch has been successfully realized by controlling various parameters including the concentration of ions [409], different kinds of solvents [410], ultrasound treatment [411], etc. However, the methods for controlling the orientation of the helix axis are still limited. Recently, a cholesteric CNC film with controllable helix was achieved by Petesic et al. The authors successfully controlled the cholesteric domains by using small commercial magnets. By manipulating the direction of the magnetic field, the orientation of the helix can be adjusted from 0° to 17°, leading to interesting optical properties [412]. The geometrical characteristics of CNCs [413], as well as the nature of the solvents [414], appear to play an important role in the behaviour of pitch. A most desirable feature of the cholesteric CNC suspensions is the drying to solid films. The helical superstructure will not vanish due to kinetic arrest; in contrast, the pitch of the helix will be decreased to the range of a visible selective reflection, making the films like photonic crystal paper [415] that can be used as a colorimetric humidity sensor [416] or temperature and PH sensor [409], for optical encryption [417,418], in chiral plasmonics [419], as light scattering shutters [420], as nano-templates [421], and even in mirrorless lasing [160]. Recently, Rofouie et al. reported a cholesteric CNC photonic semi-spherical film with a polydomain morphology where the thickness, structure and optical properties of the film changed along the surface, leading to a broadband reflection covering the entire spectral range [422].

5. Biological Applications: Biosensors and Drug Delivery

5.1. Biosensors

With increasingly improved medical techniques, it is expected that people will live longer. However, the current circumstances are not that promising. Due to the emerging problem of a “graying society” as well as the rapid spread of “illnesses of affluence,” chronic conditions such as overweight, hypertension, heart disease, diabetes, etc. are still severely threatening people’s health. Therefore, the advancement of more compelling techniques for immune detection of different kinds of disease biomarkers is necessary. LCs, which are susceptible to weak external stimuli, such as chemistry, solvent vapour, humidity or surface forces, have received increased attention for use in biological sensing since Abbott and his co-workers initiated the research field of using LC materials as sensing elements for detecting biomolecules [426]. The key idea of this study is using a nematic LC film with a specific alignment, either homogeneous or homeotropic, at an interface with an aqueous phase. Under polarizing optical microscopy, the LCs will show a particular texture. By introducing some “impurity” molecules into the aqueous phase, the alignment of the LC molecules at the interface will be altered due to the binding between the biomolecules and the doped impurity, thus leading to a texture transition (Figure 17). This transition is easily seen by the naked eye.

Such LC-based biosensors can localize biomolecules to the micrometre scale and the procedure can be executed under ambient light even without an external electric power source. All these advantages lend LC-based biosensors great potential for the next generation of high-sensitivity, low-cost and label-free bioassays. Up to now, various investigations of LC-based biosensors have been reported [427,428,429,430,431,432]. Very recently, Lee et al. reported a label-free protein (bovine serum albumin, BSA) quantitative detection by a dual-frequency LC (DFLC)-based biosensor [433]. The spectra of the dielectric properties of DFLC in the presence of BSA at varied concentrations were measured and analysed. They demonstrated that the difference in dielectric properties between the low- and high-frequency regimes was related to the concentration of BSA, which permitted a quantitative measurement of the BSA concentration.

The simplicity and versatility of LC-based biosensors and their sensitivity is restricted by the size and amount of biomolecules, which may restrict their applications in the bioassay range of low concentrations and trace molecules. It is thus significant to seek signal-enhancement strategies for LC-based sensors to circumvent the problem. Several studies have demonstrated that the disruption of the LC orientation can be enhanced by introducing nanoparticles, thus leading to an amplified optical signal [434,435,436,437,438,439,440] (Figure 18a,b). A recent study by Wei et al. demonstrated a nickel nanoparticle-assisted LC cell system that can be used to visualized the enzymatic activities between cholylglycine hydrolase (CGH) and cholylglycine (CG). The trace doping of nickel nanoparticles induced a uniform homeotropic LC alignment that could be disturbed by introducing CG to the binding-immobilized CGH surface. Such a system had a detection limit at the “10 pM” level, and could quantitatively define a working range from 0.1 nM to 1 μM [439]. The investigation by Kim et al. reported a different biosensor based on self-reporting and self-regulating LCs (Figure 18c,d). In this interesting work, the authors showed that a range of stimuli, including temperature, mechanical shear as well as cationic amphiphiles, can induce a continuous or transient release of microcargo (micro-particles or -droplets) that initially is trapped within the LC matrix. By pre-programming through an interplay of elasticity, electrical double layers, buoyancy and shear forces in diverse geometries, the LC matrixes could self-report and self-regulate their chemical response to targeted chemical, physical and biological events in particular ways. These LC materials can detect the weak mechanical shear stress generated by motile bacteria and then respond in a self-regulated manner via a feedback loop in order to release the minimum amount of biocidal agent required to cause bacterial cell death [441].

5.2. Drug Delivery

As one of the key features providing biological systems with unique physical and chemical properties, the self-assembly of biological materials, such as amphiphilic lipids, has been the focus of an increasing number of studies. Lyotropic LC droplets with diameters in the range of 100 nm, usually referred to as cubosomes and hexosomes, are potential candidates for constructing novel matrices mimicking biological systems in the formation of new nanoparticle carriers for delivering drugs and nutrition. These self-assembled nanostructures are spontaneously formed by biocompatible amphiphilic lipids, such as phospholipids, monoglycerides, glycolipids and urea-based lipids, mainly in water, and contained in a thin shell of amphiphilic block co-polymer (Figure 19).

Unlike traditional liposomes, which can only load a hydrophilic drug cargo and release it to the target immediately once the lipid membrane is broken, these LC nanoparticles have cores with equal amounts of polar and non-polar regimes, and thus can carry either hydrophilic or hydrophobic cargo, or even both simultaneously. At the same time, the interior drugs are protected from chemical and enzymatic degradation during the passage through the digestive system to the site of release, leading to a sustained release over time [15]. Due to such a slow release, the drugs are inclined to accumulate as much as possible in sickness-related tissues or cells to generate a maximal effect and reduce the negative effects on human bodies such as the toxicity of the drugs, making these LC colloidal nanoparticles particularly interesting in medical treatments [443,444]. Solid lipid nanoparticles (SLNs) have also attracted great attention as cosmetic and pharmaceutical formulations [445]. Due to their relatively small particle size but large surface area, SLNs possess very strong adhesive characteristics. This can induce film formation on the human skin, which may be helpful to keep the skin moist and even restore the previously damaged protective lipid film on the body through occlusion. Incorporation of active ingredients into a solid lipid matrix can provide protection against hydrolytic degradation, allowing drug release at a controlled rate. This implies either a transient or a continuous drug release, depending on the particle size and polymorphic transitions of the lipid matrix [446]. Continuous release is critical for active ingredients that may irritate the body if the concentration is too high or when a prolonged period of drug transport time is desired, whereas transient release will be helpful to improve the penetration of drugs. In addition to LC phases and solid lipids, a large variety of materials can be fabricated into drug-carrier nanoparticles. For instance, both biodegradable and non-biodegradable polymers have been used as drug carrier materials [447].

6. Summary

In this paper, we summarized the influence of nanoparticle doping on the electro-optical and other physical properties, of liquid crystals, including molecule alignment, viscosity, clearing point, or elastic constants, just to name a few of the properties introduced. Various kinds of nanodopants, such as ferroelectric nanoparticles, noble metallic nanoparticles, semiconductor nanoparticles and carbon nanoparticles, have been presented to induce modifications of the physical properties of LCs. However, these studies create an inconsistent picture of the beneficial effects of nanoparticles on the physical properties of LC hosts. This may be attributed to the effect of different nanoparticles’ size, shape, dispersibility and functionalization. To thoroughly understand the influence of nanoparticle doping on the physical properties of LCs, further systematic investigations are required. We further demonstrated the exciting photonic functionalities, including SPR effect, photothermal effect and lasing, of LC–nanoparticle composites. The intrinsic optical properties of nanoparticles can be effectively modified by a LC medium, leading to a variety of multifunctional photonic applications.

Various nanoparticle mesostructures, obtained either within distorted areas of LCs or topological defects, were discussed. The self-organization-assisted nanoparticle assembly within the disclination lines will effectively eliminate high-energy volume areas and reduce the elastic-free energy density of the LC system. Such an effect can even lead to a stabilization of the 3D blue phase defect lattice.

Additionally, we demonstrated the self-assembly of nanoparticles into lyotropic LC phases. Colloidal suspensions of anisotropic particles, including rods, tubes, disks, flexible chains and wires, can self-organize from the isotropic disordered to the nematic ordered phase as the concentration exceeds a critical value according to Onsager’s model. These lyotropic LC materials have intriguing physical properties and can be applied to various functional devices.

Finally, we briefly introduce the practical applications of LC–nanoparticle composites in biological systems. The aspects of biological sensors and drug delivery systems are of particular interest, and it is anticipated that this area of novel liquid crystal research will attract much interest in the liquid crystal community and soft matter community in particular.