Research articleHydrophobic gold nanoparticles via silane conjugation: chemically and thermally robust nanoparticles as dopants for nematic liquid crystals
Abstract
We examine for the first time how chemically and thermally stable gold nanoparticles (NPs), prepared by a silane conjugation approach, affect both the thermal and the electro-optical properties of a nematic liquid crystal (LC), when doped at concentrations ranging from 0.25 to 7.5 wt%. We find that the octadecylsilane-conjugated gold NPs stabilize both the enantiotropic nematic and the monotropic smectic-A phases of the LC host with a maximum stabilization of 2°C for the nematic and 3.5°C for the smectic-A phases for the mixture containing 1 wt% of the silanized particles. The same mixture shows the lowest values for the Fréedericksz transition threshold voltage and the highest value for the dielectric anisotropy. Generally, all NP-containing mixtures, except mixtures with NP concentrations exceeding 5 wt%, reduce the threshold voltage, increase the dielectric anisotropy and reduce both rise and decay time; the latter particularly at temperatures at least 10°C below the isotropic–nematic phase transition on cooling.
1. Introduction
Doping (i.e. the intentional incorporation of small quantities of an additive into a liquid crystal (LC)) is one of the most prominent processes to tune optical and electro-optical properties of anisotropic ordered fluids. Traditionally, dyes [1,2] and chiral additives [3–7] were most commonly pursued to induce properties such as photoswitching, lasing [8–10], helical twisting or spontaneous polarization.
Many of these induced properties were tuned and optimized for applications in LC displays using chiral smectic-C and nematic (or chiral nematic) phases. The response of a nematic LC to an applied electric field is of great importance for such display applications. While the key display modes, namely twisted nematic, vertical alignment and in-plane switching, use different nematic LC materials and alignment approaches, the ON and OFF state of each pixel always relies on an electric field being applied through the LC material to induce a bulk change of the LC orientation [11,12]. In general, it is desirable that the LC material or the mixture of interest for an LC display has a minimal threshold voltage (V th), maximal dielectric anisotropy Δε (positive or negative), low viscosity (γ), high birefringence (Δn) and short switching times (i.e. rise and fall times, τ).
Minimizing V th is important for maximizing energy efficiency of an LC display [13]. V th is typically measured as the voltage at which a 5 or 10 per cent change in capacitance (or 10% in transmittance) occurs. The dielectric anisotropy and the elastic constants of the nematic LC are governed by the slope of the C–V plot. The splay elastic constant (K11) is particularly important for the cell geometry used to study these systems (Fréedericksz transition in a planar cell), because the splay deformation (reorientation from a planar to a vertical configuration and vice versa) is the deformation that occurs when an electric field is applied to a rubbed polyimide-coated planar cell.
The splay elastic constant (K11) is derived from
Over the past decade, doping LCs with nanoparticles (NPs) has provided an intriguing and viable way of manipulating the properties of LCs [14–17]. If stable dispersions can eventually be achieved, this pathway may well provide the means to be of significant importance for LC-based device and display applications. Tremendous global research efforts demonstrated that doping LCs with NPs can introduce original effects related to NP pattern formation, NP-guided LC alignment and alteration of optical or electro-optic characteristics in both nematic and smectic LC phases [18]. For example, several reports demonstrated lower values for V th, induced homeotropic alignment or faster switching in nematics, and higher values for the spontaneous polarization in ferroelectric SmC* materials [18].
Using larger, micrometre-sized particulate systems as dopants to alter the properties of nematic LCs is an area of research that has been ongoing for quite some time. Early studies focused on defect formation induced by aerosil particles [19–34]. Commonly, these particles produced a large number of defects resulting in intense light scattering of the mixture in the field-OFF state, and becoming transparent when driven with an electric field [35–37]. As more nanomaterials varying in shape, size and composition became available, research focused on incorporating nanomaterials.
Previous work in our laboratory has shown that alkylthiolate-capped gold NPs (Au NPs) and silver NPs (Ag NPs) can modulate the optical and the electro-optic properties in dispersion with nematic LC hosts in a unique fashion [16,38–40]. For example, we have demonstrated that alkylthiol-capped Au NPs induce multiple alignment modes in nematic LCs (e.g. a thermal history-dependent dual alignment and electro-optic switching mode) in addition to minimized threshold voltages and altered elastic constants (particularly, the splay elastic constant K11) and dielectric properties (ε∥,ε⊥ and Δε) [38,40].
Au NPs, as the most likely frequent choice to dope LCs, display unusual and unique size-dependent chemical and physical properties [41]; their surface can be passivated in situ or via place exchange reactions with molecules ranging from simple alkanethiols, phosphines and amines to more complex molecular structures, including dendrimers, polymers and LC moieties. Drawbacks of these Au NP dopants have always been the fairly large size distribution and the difficulty to obtain series of these particles just differing in size using the one- [42,43] or two-phase Brust–Schiffrin methods [44]. While the ease as well as speed in preparing and modifying these Au NPs has to be considered an advantage, their stability under various conditions is not. Factors to consider arise from the chemical and thermal (in)stability of metal NPs as well as from the mobility of surface-bound capping agents (usually thiols). The chemical stability or the instability of thiol-capped metal NPs towards oxidation (i.e. oxidation of surface-bound thiols to disulfides in air or in the presence of other oxidants) [45], towards halides [46] and towards alkaline metal ions has been studied by a number of groups [47] highlighting the importance of determining NP purity. The thermal stability of Au NPs is particularly important considering the use in composites with LCs, where sample preparation protocols often call for annealing and slow cooling from isotropic liquid phases. Both thermogravimetric analysis data [48] and temperature-dependent X-ray photoelectron spectroscopy (XPS) studies frequently reveal that thiols with an alkyl chain length longer than C6 start to or completely desorb (forming dithiols) from the Au NP surface at temperatures above 160°C; thiols with shorter chain length desorb at even lower temperatures (between 90°C and 140°C for C3–C5) [49]. Sintering of Au NPs is often observed for particle sizes below 4 nm as shown by Brust and co-workers [48], rapidly (i.e. 10 min) at elevated temperatures (e.g. 250°C), but also appreciably at room temperature, an effect known as coarsening. For example, alkylthiol-capped Au NPs with an average core diameter of 2 nm are easily converted into Au NPs with a core diameter larger than 5 nm simply by refluxing them in toluene [40]. Related to this coarsening process, migration of thiols on the NP surface is relevant. Using electron paramagnetic resonance spectroscopy on Au NPs passivated with spin labels, such as TEMPO, Chechik and others demonstrated that thiol ligands on Au NP surfaces undergo slow lateral diffusion (even at room temperature) that may take several hours at temperatures around 90°C [50–52]. While slow, such migration might lead to segregation of thiols on the surface of mixed monolayer-capped Au NPs (different ligands simultaneously cap the surface of the NP) [53,54].
2. Results and discussion
To enhance both thermal and chemical stability of Au NPs used as additives in LCs and circumvent thiol desorption and migration issues that could potentially falsely contribute as free thiol ligands to some of the observed effects, we here present a new robust type of hydrophobic Au NPs prepared using a simple, versatile silane conjugation approach, first reported by Jana et al. [55] for hydrophilic thiol-capped Au NPs. In this two-step silanization, the first step delivers 3-mercaptopropyltrimethoxysilane (MPS)-capped Au NPs. These NPs, after isolation and purification, then undergo a silane conjugation step (i.e. hydrolysis and dehydration condensation under basic conditions) with a second, long chain alkytrimethoxysilane (here H37C18Si(OCH3)3), which results in Au NPs that are thermally and chemically robust, survive extensive sonication without decomposition or thiol desorption, and are well dispersible in both organic solvents and nematic LCs featuring aliphatic side chains such as LC1 used in this study (figure 1). We systematically characterized the NPs synthesized using the above-described silane conjugation approach by 1H nuclear magnetic resonance (NMR) spectroscopy (see the electronic supplementary material, figure S1), UV–visible spectrophotometry, XPS and transmission electron microscopy (TEM). Batches of the same NPs kept at elevated temperatures and under prolonged sonication gave identical 1H NMR spectra, visible absorption spectra and TEM images, which confirms their exceptional stability. A visible absorption spectrum in hexane and a representative high-resolution TEM image are shown in figure 2. TEM image analysis provided an average NP size of 4.1±0.7 nm, and XPS confirmed the formation of a silica shell around the Au NP core (see the electronic supplementary material, table S1). More detailed information about the synthesis and characterization of these NPs is given in §4 and in the electronic supplementary material.
Figure 1. Silane conjugation synthesis and schematic of the silanized, hydrophobic Au NPs, and structure as well as phase transition temperatures of the Felix-2900–03 nematic liquid crystal host (LC1). (Online version in colour.) Figure 2. Visible absorption spectrum of Au NPSil in hexane (a) as well as TEM (b) and high-resolution TEM images (c) of Au NPSil. (Online version in colour.)
Based on protocols for investigating NP-doped nematic LCs established in our laboratory [38], we tested seven different concentrations of the silanized Au NPs in Felix-2900–03 (LC1) from 0.25 to 7.5 wt% covering an extended range from low to high NP loadings. Our group has used the same nematic LC for several studies in the past [38,56–58], which will allow us to compare current with prior test results later. Following the sample preparation protocol outlined in §4 provided the first visual clues that these silanized NPs are able to form stable suspensions in nematic solvents such as LC1. After complete solvent evaporation with the vial placed on a hotplate adjusted to a temperature below the isotropic-to-nematic phase transition (TNI), all seven mixtures of Au NPSil in LC1, even up to 7.5 wt%, appeared homogeneous, showed no signs of NP settling, and displayed a continuous colour intensity increase based on the surface plasmon resonance (SPR) frequency of the NPs in LC1 (figure 3).
Figure 3. Photograph showing the quality of dispersion of Au NPSil in LC1 in solid glass V-vials after complete solvent evaporation; vials are placed on a hotplate with the NP/LC1 dispersions in the nematic phase at TNI−T=2°C (left to right: concentration ranging from 0.25 to 7.5 wt%). One can clearly see the colour of the SPR of the Au NPSil increasing in intensity as the concentration is raised from 0.25 to 7.5 wt%. No settling or macroscopic aggregation is visible. (Online version in colour.)
Thin films of the LC colloids between plain, untreated glass slides were first investigated by polarized optical microscopy (POM) to test alignment and thermal properties. Predominantly at NP loadings ranging from 0.25 to 2.5 wt% in LC1, large domains of NP-induced homeotropic alignment due to partial segregation of NPs to both glass substrates are observed. These homeotropic domains surround birefringent stripe features (figure 4a–c) frequently observed for many other types of Au NPs and quantum dots [59,60]. We recently investigated these textural features in more detail using fluorescence confocal polarizing microscopy (FCPM) and established both the director field within the stripes and the segregation of NPs causing the induced homeotropic alignment [59]. Two additional striking features were observed by POM both in untreated glass slides and rubbed polyimide-coated indium tin oxide (ITO) glass cells (cell gap: 4 μm). First, at NP concentrations exceeding 2.5 wt% an increasing number of isotropic domains persist well below TNI of the mixture on cooling. Coexistence ranges observed for all mixtures by POM on cooling are plotted in figure 5. A very similar trend was also observed on heating. Second, and noticeable from the temperature plot in figure 5, mixtures in the mid-concentration range (from 0.5 to 2.5 wt%) show an apparent stabilization of the nematic phase on heating from the solid state and on cooling from the isotropic liquid. To confirm these optical microscopy results, we performed differential scanning calorimetry (DSC) studies on all mixtures. DSC does not register the biphasic behaviour (coexistence of nematic and isotropic phases), but measures the bulk thermal behaviour of these samples. Nevertheless, the DSC data show the same phase transition temperature trends as noticed by POM (figure 6). The bar diagrams show the measured phase transition temperatures on heating (figure 6a) and cooling (figure 6b) taken at a heating/cooling rate of 10 K min−1. Practically identical DSC traces were observed during the first and the second heating/cooling run. Notably, not only the nematic, but also the monotropic smectic-A phase is stabilized (higher TN-SmA) for all and especially in the lower to mid-concentration range mixtures (0.25–2.5 wt%), which we were able to confirm by POM. The full series of DSC traces can be found in the electronic supplementary material, figure S2. This thermal phase stabilization by about 1.5–2.0°C for the nematic and about 3.5°C for the smectic-A phases (at 0.5–1.0 wt% of the NPs; far above the experimental error) is a very surprising and stimulating result. We have not observed such stabilizing effect for any of the other NPs we tested in LC1 until that time, including similar-sized Au NPs or CdSe quantum dots capped with alkylthiols, amines or carboxylic acids featuring similar hydrocarbon chain lengths [56]. Any other NP doped into a nematic or a smectic mixture showed either no or a slight phase destabilizing effect. While some of the silanized Au NPSil appear to segregate to the interfaces inducing homeotropic alignment between plain glass slides, these NPs appear to enhance the overall order parameter of the nematic and the smectic-A phases. Similar drastic stabilizing effects have so far only been described for anisometric NP inclusions such as carbon nanotubes [61] or polar ferroelectric NPs [62]. In general terms, a mesophase stabilizing effect by NP inclusions is believed to be caused by weak interactions between NP and LC host or by NPs that are well compatible with (well-dispersed in) a given nematic LC host [63].
Figure 4. POM photomicrographs (crossed polarizers) of LC1 doped with Au NPSil at TNI−T=5°C between plain glass slides: (a) 0.25 wt%, (b) 0.5 wt%, (c) 1.0 wt%, (d) 2.5 wt%, (e) 5.0 wt% and (f) 7.5 wt%. (Online version in colour.) Figure 5. Plot of the phase transition temperatures (TNI and biphasic N/Iso regions) of all mixtures of LC1 doped with Au NPSil observed by POM on cooling at a rate of 1 K min−1 (plain glass and LC test cells). Red colour denotes nematic–isotropic phases; blue colour, nematic phase. (Online version in colour.) Figure 6. Plots of the phase transition temperatures measured by DSC at a rate of 10 K min−1: (a) on heating highlighting the stabilization of the nematic phase (blue, nematic; black, crystalline), and (b) on cooling demonstrating the stabilization of both the enantiotropic nematic (blue) and the monotropic smectic-A (green) phases. (Online version in colour.)
The next step was to test the alignment in rubbed polyimide-coated ITO glass cells. Such cells were independently prepared in both participating laboratories, and the images in figures 7 and 8 show the POM results with respect to alignment and biphasic behaviour. All mixtures in the cells align perfectly planar without the formation of birefringent stripe defects or homeotropic domains as observed for other Au NPs in prior work. The cells only show the aforementioned isotropic domains, even ‘deep’ into the nematic phase on cooling (figures 7e,f and 8f,g). FCPM tests using a dichroic dye (see §4) with a much better solubility in the isotropic liquid phase of LC1 confirm the isotropic or biphasic nature of these domains (figure 9). In addition, between uncrossed, parallel polarizers the locally elevated concentration of the Au NPs is clearly visible (dark spot-like regions in figure 9c). This is not surprising for two reasons. First, NPs are in general more soluble in the isotropic liquid phase of an LC host, and second, the concentration greater than or equal to 5 wt% in these mixtures is far above the typical solubility limit of NPs in nematic phases; only a limited number of reports claims higher well-dispersed NP loads [64].
Figure 7. POM photomicrographs (crossed polarizers) of LC1 doped with Au NPSil at TNI−T=5°C in planar cells (cell gap: 4 μm)—a selected, representative section is shown: (a) 0.25 wt%, (b) 0.5 wt%, (c) 1.0 wt%, (d) 2.5 wt%, (e) 5.0 wt% and (f) 7.5 wt% (arrows in bottom right indicate rubbing direction; small dots in these images are the spacers of the 4 μm cells). (Online version in colour.) Figure 8. POM photomicrographs (crossed polarizers) of LC1 doped with Au NPSil at TNI−T=5–7°C in planar cells (cell gap: 4 μm)—entire cell is shown: (a) 0.25 wt%, (b) 0.5 wt%, (c) 1.0 wt%, (d) 2.5 wt%, (e) 5.0 wt% and (f) 7.5 wt% (tilting of the sample indicates the isotropic nature of the dark spots—coexistence of nematic and isotropic even 7°C below TNI). (Online version in colour.) Figure 9. Photomicrographs showing the 5 wt% Au NPSil in LC1 mixture at TNI−T=5°C in a planar cell: (a) POM image (crossed polarizers) (figure 7e), (b) combined POM/FCPM image (both polarized transmission, red channel and fluorescence confocal, green channel) showing the larger concentration of the dye in the isotropic domains, and (c) POM image under identical conditions with parallel, uncrossed polarizers showing the larger concentration of NPs in these isotropic regions. (Online version in colour.)
Next, we performed electro-optic measurements on all mixtures. Earlier studies on alkylthiol-capped metal NPs (dodecanethiol-capped Au NPs with a core diameter of 5.4 nm [38] as well as hexanethiol-capped Ag NPs with a core diameter of 4.2 nm [40]) showed a tremendous reduction of V th despite much reduced values for the dielectric anisotropy (Δε). In contrast to the Au NPSil described here, these NPs induced homeotropic alignment in both untreated glass slides as well as rubbed polyimide-coated planar ITO glass cells. For these NPs, we always observed the dual and reverse electro-optic mode (switching from homeotropic to planar for a nematic LC with positive Δε, particularly at low field frequencies) [38,57] that was not observed in the current mixtures. The robustness of the current Au NPSil, allowing for the preparation of much-improved homogeneous and stable dispersion in LC1, appears to be responsible for the diminished NP segregation to interfaces, and hence the lack of homeotropic alignment in planar cells.
V th was determined from capacitance and transmittance measurements. However, transmittance measurements are sometimes more stable than capacitance measurements, as the capacitance plateau in the latter can occasionally exhibit significant deviations. In addition, transmission measurements can selectively choose small regions from the cell for data acquisition, whereas the capacitance is always measured over the complete electrode area. In previous experiments using for example quantum dots, we were able to show that different areas within one cell can show different behaviour: when a slight surface coverage of particles occurs owing to NP segregation, the surface energy is changed and the threshold gets lost. For the calculation of threshold, the transmission data are transformed into phase retardation data, using the equation
Plots of V th versus temperature and versus reduced temperature T−TNI are collected in figure 10. In general, a slight decrease in threshold with increasing particle concentration was found. Neglecting the mixtures with coexistence of nematic and isotropic domains (5 and 7.5 wt%), the mixtures with a concentration of 0.5, 1.0 and 2.5 wt% of Au NPSil in LC1 show the lowest values for V th in the V th versus reduced temperature plots (which are corrected for the difference in clearing point between mixtures). Maybe not surprisingly, the mixtures showing the most significant stabilization of the nematic and the monotropic smectic-A phases (0.5 and 1.0 wt%) are the mixtures with the overall lowest V th (figure 10b). The slope in each set of plots remains approximately identical, also in comparison with pure LC1, hence we conclude that the temperature dependence of V th is not affected by the presence of suspended NPs. Accordingly, the same mixtures of 0.5 and 1.0 wt% Au NPSil in LC1 also show among the highest values for Δε in the investigated series (figure 11). For both of these values, V th and Δε, the mixture containing 0.25 wt% of the silanized NPs appears to show almost identical values to pure LC1 within experimental error. It appears that this concentration is the lower threshold for electro-optic alterations in the current series, while concentrations around 5 wt% and higher represent the upper useful limit.
Figure 10. Plots of the threshold voltage (V th): (a) versus temperature and (b) versus reduced temperature T−TNI. (Online version in colour.) Figure 11. Plots of the dielectric anisotropy (Δε): (a) versus temperature and (b) versus reduced temperature T−TNI. The plots for the mixtures containing 5 and 7.5 wt% of Au NPSil (AuSi_5.0 and AuSi_7.5) show lower values for Δε because of defects originating due to isotropic domains. (Online version in colour.)
Lastly, we also measured the switching times of all mixtures; both rise time (γ) and decay time (τoff). For all tested cells, we found an increase in τoff with increasing temperature. Obvious concentration dependencies were hard to find. Nevertheless, almost all NP-doped samples show lower decay times, increasingly so as we depart from TNI on cooling at T−TNI<−10°C. Interestingly, the mixture doped with only 0.25 wt% shows the lowest values for τoff in this series at T−TNI<−5°C (figure 12). Similar trends were also observed for the rise time. The NP-doped mixtures show lower rise times far from the isotropic–nematic phase transition on cooling (T−TNI<−10°C) and slower rise times closer to the clearing point of the mixture. Overall, however, an increase in γ with increasing NP concentration can be seen in figure 13.
Figure 12. Plots of the decay time (τoff): (a) versus temperature and (b) versus reduced temperature T−TNI. At temperatures closer to TNI, values of τoff significantly scatter, making a clear interpretation of trends with respect to NP concentration at higher temperatures somewhat difficult. (Online version in colour.) Figure 13. Plots of the rise time (γ): (a) versus temperature and (b) versus reduced temperature T−TNI. (Online version in colour.)
3. Conclusions
For the first time, we have been able to prepare and test truly robust Au NPs as additives in a nematic LC host. The silane conjugation chemistry used to synthesize these NPs provides access to chemically and thermally stable Au NPs without the risk of thiol desorption from or thiol migration on the Au NP surface. In addition, NP ripening or sintering at elevated temperatures in the LC solvent during sample preparation and testing is also prevented by the stable siloxane corona around the NP. In this way, exceedingly reliable optical, thermal as well as electro-optic data of nematic mixtures containing these Au NPs could be collected. The nematic LC host is structurally characterized by a 2-phenylpyrimidine rigid core flanked by two terminal hydrocarbon chains, which, based on past and current experiments, enhance the dispersibility of alkyl monolayer-capped Au NPs. The Au NPSil particles used in this study, however, are much easier to disperse in the same nematic host used in numerous prior studies (LC1). Considering calculated molecular parameters, this enhanced dispersibility is most likely due to the larger, fixed distance between neighbouring alkyl chains provided by the siloxane framework around the NP. Considering a homeotropic anchoring of the LC molecules to the NP surface, such larger distance may well provide better access for the LC molecules of the host, but could also mean weaker anchoring. In any event, the robustness imparted by the siloxane network allows for extended sonication times that help ensure well-dispersed LC colloidal samples.
Thermal characterization revealed that these silanzied Au NPs induce an NP concentration-dependent LC phase stabilizing effect of up to 2.0°C to the enantiotropic nematic phase and up to 3.5°C to the monotropic smectic-A phase of the LC host, which one would only expect in this magnitude from anisometric, magnetic or ferroelectric NP inclusions. The exact nature of this stabilizing effect is somewhat unclear at the moment, but additional experiments with modified alkyl chain lengths of the alkyltrimethylsilane used to conjugate the MPS-coated Au NPs are being pursued to shine more light onto this phenomenon. The mixture containing exactly 1 wt% of the suspended NPs shows the largest thermal stabilization effect. The same mixture also shows the lowest values for the threshold voltage and the highest values for the dielectric anisotropy, which indicates that thermal and electro-optic effects are to some extent connected. Finally, all NP-doped nematic samples show faster optical switching (both rise and decay time) particularly at lower temperatures away from the isotropic–nematic phase transition on cooling.
In conclusion, the chemically and thermally robust, silanized Au NPs presented here are ideal candidates for studies of NPs in LCs and applications of such composites in devices based on the manipulation of optical and electro-optic properties. With the option to translate this approach to other NP platforms, silanization appears as ‘the way to go’.
4. Experimental details
(a) Materials
Hydrogen tetrachloroaurate(III) trihydrate was purchased from Alfa Aesar, didodecyldimethyl-ammonium bromide (DDAB), tetrabutylammonium borohydride (TBAB), 3-mercaptopropyl-trimethoxysilane and trimethoxy(octadecyl)silane (TMODS) were purchased from Sigma-Aldrich. All chemicals were used without further purification. Pure LC1 (Felix 2900–03) was purchased from Synthon Chemicals. All solvents used for the synthesis and purification were Aldrich purification grade purified via a PureSolv solvent system (Innovative Technology). Millipore water (18 MΩ) was used to rinse glassware and vials used for sample preparation. In addition, all vials used to store, handle and mix NPs with the LC host were cleaned with aqua regia.
(b) Methods
UV–visible spectra were recorded using a Varian Cary 5000 UV-vis-NIR spectrophotometer. TEM imaging for AuNPSil was performed using a JEOL JEM-2100F TEM instrument at an accelerating voltage of 120 kV or a Hitachi H-7000 electron microscope. 1H NMR spectra were acquired using a Bruker Avance 300 MHz spectrometer, and XPS data were collected using a Kratos Axis Ultra X-ray photoelectron spectrometer.
POM images were taken using an Olympus BX51-P microscope in conjunction with a Linkam LS350 heating/cooling stage, a Leica SD6 Macroscope, or an Ortholux II Pol-BK microscope (Leica). In the latter case, for illumination, a halogen lamp (XENOPHOT, Osram) with interference filter for 542 nm was used. The LC test cells used were planar cells (cell gap: 4.0 μm) with antiparallel polyimide alignment layers and 1°–3° pre-tilt (LC Vision). For all POM imaging, the LC mixtures were heated above the isotropic–nematic phase transition temperature (TNI) and cooled at a rate of 1°C min−1 until the desired temperature below TNI was reached. FCPM images were obtained using a Nikon LV 100D-U upright microscope coupled with a Nikon Eclipse C1 Plus scanner/controller using a 488 nm laser to excite the dichroic dye N,N′-bis(2,5-di-tert-butylphenyl)-3,4,9,10-perylenedicarboximide (obtained from Sigma-Aldrich) used at 0.002 wt%.
Electro-optic parameters were measured using two different set-ups. One set-up included an LCAS I automated liquid crystal analyser (LC Vision, LLC), a second one is described later. All test cells were positioned under an angle of 45° between crossed polarizers, and the light intensity passing through the cell was measured using a photomultiplier tube (Oriel). A Linkam LTS350 heating stage with TMS94 controller was used for temperature control. Before measurement, all samples were heated up to the isotropic phase and then cooled down to 58°C with a cooling rate of 1°C min−1. Capacitance measurements were performed simultaneously to transmission measurements on an HP 4284 A LCR bridge, varying the sine test signal level from 0.05 to 19.95 Vr.m.s. to reorient the LC. The test signal frequency was 1 kHz. Analysis or interpretation of the data was performed based on the single-cell method described by Wu et al. [65].
The mixtures of the AuNPSil particles in LC1 were prepared by weighing accurate amounts of the solid LC using a microbalance and dissolving it in a known amount of toluene. The as-prepared solutions of the respective NP were then combined in a V-vial with the LC solution to produce the exact concentration of the NPs in LC1. The solvent was then allowed to evaporate under a steady stream of dry N2 at approximately 70°C for about 24–48 h followed by sonication.
(c) Silanized gold nanoparticles
The synthesis of the silanized NPs was carried out using a procedure adopted from [55]. Briefly, a 10 ml solution of HAuCL4 in toluene (0.01 M) was prepared in the presence of an equimolar amount of DDAB using 5 min sonication. Next, 200 μl of a toluene solution of MPS (0.1 M) was added under stirring followed by addition of 1 ml toluene solution of TBAB and DDAB (30 mg of TBAB and 27 mg of DDAB). The reaction solution turned immediately black and changed then slowly to pink. After 1 h, 100 μl of TMODS and 230 μl of NaOH (1.0 M) were added to the reaction mixture and stirring was continued for 15 h. Finally, the solvent was evaporated, and the collected particles were washed several times with chloroform and isopropanol.
Acknowledgements
This work was financially supported by the Natural Science and Engineering Research Council (NSERC) of Canada, the Canada Foundation for Innovation (CFI), the Manitoba Research and Innovation Fund (MRIF), the Deutsche Forschungsgemeinschaft (DFG, KI 411), and the European Science Foundation (ESF-EUROCORES, SONS II program, LCNANOP project). We are grateful to K. McEleney for his valuable assistance with and M. S. Freund for access to XPS measurements. We also thank A. Dufresne for help and assistance with TEM imaging. T.H. thanks the Government of Ohio's Third Frontier Program (Ohio Research Scholar Program).
