Article pubs.acs.org/Langmuir
Cite This: Langmuir XXXX, XXX, XXX−XXX
Anchoring Energy Measurements at the Aqueous Phase/Liquid Crystal Interface with Cationic Surfactants Using Magnetic Fréedericksz Transition Fatma Yesil,* Masayori Suwa, and Satoshi Tsukahara Department of Chemistry, Graduate School of Science, Osaka University, 1-1 Machikaneyama, Toyonaka, Osaka 560-0043, Japan S Supporting Information *
ABSTRACT: We constructed the apparatus to observe the Fréedericksz transition of liquid crystal in contact with water. The Fréedericksz transition is a distortion of nematic liquid crystals (LCs) induced by external fields. In the present system, sweeping homogeneous magnetic field was applied to the sample, and the distortion of the LC was visualized with a polarized light microscope with the crossed Nichols configuration. The anchoring energy (WAQ/LC) at the aqueous phase/LC interface was measured in the presence of surfactant from the threshold magnetic field of the Fréedericksz transition. We studied two cationic surfactants: dodecyltrimethylammonium bromide and tetradecyltrimethylammonium bromide. A nematic LC, 4cyano-4′-pentylbiphenyl (5CB), was examined, which was confined in a copper grid on an octadecyltrichlorosilane-treated microscope glass plate. Measured WAQ/LC were reproducible and showed consistence with the reported region for the water/LC interface. Interfacial excess of surfactants was also measured by the pendant drop method, and the relationship between the obtained WAQ/LC and the interfacial excess was investigated. Experiments showed that an increase in the anchoring energy depends on the surfactant and its interfacial excess. The region of the interfacial coverage, at which WAQ/LC increases, varied with the chain length of the surfactant. The measurement of the anchoring energy will provide new fundamental information on aqueous phase/LC interface.
■
INTRODUCTION
σ = σ0 +
In 1888, Friedrich Reintzer, an Austrian botanist, discovered a new phase of matters that possessed physical properties between those of liquids and solids; today, they are known as liquid crystal (LC). LC molecules are composed of rod-like or disk-like rigid cores and a long flexible tail, and they are aligned approximately parallel to one another. The long axis of LC molecules is called the director. After his discovery, LCs were widely studied,1 and C. Mauguin adapted LCs to solid interfaces in 1911.2 His study lead to studies of the interaction between LC molecules and solid surfaces modified with some chemicals. To evaluate the interaction, the anchoring energy, W, is frequently used. In 1969, Rapini and Papoular29 introduced a simple presentation of W as a component of the interfacial tension, σ, as eq 1, where σ0 is the interfacial tension without an external force, θ is a titled angle of LC molecules, and θ0 is that without an external force. It should be noted that the range of the interfacial tension is ∼10−3 J m−2, but the range of the anchoring energy is about 10−7 to 10−5 J m−2. W is a parameter representing how strong the orientation of LC molecules is fixed at the interface. Several methods to measure W are available for solid/LC interface. One of the methods is microscopic technique by using mechanical torque,4 and most of others use high electric fields. © XXXX American Chemical Society
1 W sin 2(θ − θ0) 2
(1)
An applied external electric or magnetic field can override the default director orientation determined by the boundary conditions. In 1934, Fréedericksz and Tsvetkov in a Russian group showed the first magnetic field effects;5 since then, a variety of new magnetic phenomena have been discovered.3 The Fréedericksz transition is a distortion of nematic LCs induced by an external field; it is an orientational change occurring in LCs under a sufficiently high external field. Electric field is easy to use and has spectacular and important applications in liquid crystal displays (LCDs), but the action of LCs by electric field is influenced by impurity carriers and electrochemical phenomena. In the case of magnetic field, LC molecules align with their easy axis magnetization parallel to the applied magnetic field. Recently, LC is widely used in LCDs, which were invented with twisted LC/solid interfaces and Fréedericksz transition by an electric field or a light source. LCDs had a great impact on optoelectronics because the devices work with very low voltage Received: August 31, 2017 Revised: November 26, 2017 Published: December 4, 2017 A
DOI: 10.1021/acs.langmuir.7b03005 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir and power.3 A better understanding of the Fréedericksz transition may lead to more performing devices. Abbott and colleagues6 reported that the anchoring of the aqueous phase/LC interface is controllable with an addition of specific anions,7 surfactants,8 phospholipids,9 proteins,9 or synthetic polymers 9 to the aqueous phase at certain concentrations. These compounds can trigger rapid orientational ordering from tilted to homeotropic alignments of waterimmiscible thermotropic LCs.10 Aqueous phase/LC interfaces have their anchoring energy, W, like solid/LC interfaces, but there are no reports on direct measurement of W at aqueous phase/LC interfaces by using magnetic field to the best of our knowledge. As mentioned above, electric field is commonly used for the measurement of W at solid/LC interfaces, but it would be hard to apply it to the W measurements at aqueous phase/LC interfaces for the following reasons. In the case when electrodes are parallel to the interface, the distribution of electric field depends on the dielectric constant and depth of the mediums. In the case when the electrodes are normal to the interface, they cause electrolysis in the aqueous phase. As for magnetic field, no direct contact of magnets to medium is required. The relative magnetic permeabilities of glass, aqueous solutions, and LC are almost unity, and thus it is easy to calculate the strength of magnetic field. Therefore, we decided to employ the Fréedericksz transition induced by a magnetic field to measure W values at aqueous phase/LC interfaces. The present study is important for the knowledge of the chemical properties at the aqueous phase/LC interface as well as for the development of new devices using LC at the water interface. We measured W values at aqueous phase/LC interfaces in the presence of surfactant with a home-built apparatus and studied the relationship between W and chemical properties of the aqueous phase/LC interfaces. The LC used in this study was 4-cyano-4′-pentylbiphenyl (5CB). It is one of the common LCs, and its physical properties are well known. For example, 5CB is nematic at ambient temperature, between 22.5 and 35 °C.5 The aqueous phase/LC interface was modified by two cationic surfactants, dodecyltrimethylammonium bromide (C 12 TAB)11 and tetradecyltrimethylammonium bromide (C14TAB). C12TAB and C14TAB have the same headgroup and different alkyl chain length. The experimental methods and the theoretical approaches will be shown, and the W dependency on the surfactant and its concentration will be discussed.
■
solution was removed, they were washed with dichloromethane. The resultant glass plates were dried in vacuum for 2 h and stored in a vacuum desiccator. Preparations of an Optical Cell. The preparation of 5CB thin layer is a very important process in the present study. It has to be prepared in a clean environment because impurities disturb the orientation of 5CB and form some topological defects called disclination lines.12 We prepared two types of LC sample: One was an LC layer between two OTS glass plates and the other was an LC layer between an aqueous phase and an OTS glass plate. The former cell was made as follows: A copper specimen grid for transmission electron microscopy was placed on an OTS glass plate; then, 0.25 μL (1 L = 1 dm3) of 5CB was put into the grid. The grid and the LC layer were covered with another OTS glass plate. The latter cell was prepared according to the protocol reported by Abbott et al.8 The copper specimen grid was placed onto an OTS glass plate; then, 0.25 μL of 5CB was put into the grid. The excess 5CB was removed with a capillary tube. The prepared cell was heated for 3 min at 50 °C to flatter the surface of the confined 5CB. This 5CB optical cell was set into a cylindrical tube made from polychlorotrifluoroethylene (PCTFE), as shown in Figure 1. An O-ring was inserted
Figure 1. Schematic illustration of the cell assembly. between the PCTFE tube and the glass plate to prevent a leakage of the aqueous solution. The O-ring and the glass plate were fixed to the PCTFE tube by a screw with a hole. The prepared sample was placed between a pair of electromagnets; then, 300 μL of surfactant solution was added to the tube. All surfactant solutions contained 0.1 M lithium sulfate and 0.01 M sulfuric acid to adjust at pH 2 to prevent electrostatic repulsion between head groups of surfactant at the interface and to dominate the hydrophobic force.13,14 The salt concentration and the pH were fixed because they can affect the W value.7 Observation of the Fréedericksz Transition. Figure 2 shows the schematic representation of the experimental setup for the observation of the Fréedericksz transition of the 5CB layer. 5CB confined in the grid was observed with a polarized light inverted microscope (IX71, Olympus) and an objective lens of 4× magnification under crossed Nichols configuration, which allowed us to detect the birefringence change of 5CB accompanied by the
EXPERIMENTAL SECTION
Reagents. Heptane, sulfuric acid, hydrogen peroxide, methanol, dichloromethane, and 5CB were obtained from Wako. C14TAB was obtained from Sigma-Aldrich. Lithium sulfate and C12TAB were obtained from Nacalai tesque and octadecyltrichlorosilane (OTS) was obtained from Shin-Etsu Chemical. All reagents were used without further purification, excepting heptane. Heptane was dried with molecular sieves. Water was purified with a direct-Q UV system (Millipore). Microscope Glass Plate Bound with OTS. The Fisherfinest Premium plain microscope slides were obtained from Fisher Scientific. Round-shaped glass of 1 cm in diameter was cut out from the slide. The round glass plates were first washed with piranha solution (sulfuric acid/hydrogen peroxide 7:3 in volume) for 1 h at 80 °C in a water bath. After this immersion, they were rinsed with water, with ethanol, followed by washing with methanol. Then, they were put in a vacuum oven for 2 h at 110 °C. They were immersed in the 0.5 mM (1 M = 1 mol dm−3) OTS heptane solution for 30 min. After the OTS
Figure 2. Schematic illustration of the polarized light microscope for the observation of the Fréedericksz transition. B
DOI: 10.1021/acs.langmuir.7b03005 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 3. Schematic illustration of the deformation of the LC layer under a magnetic field. Red arrows indicate the direction of the magnetic field. Fréedericksz transition. Polarized monochromatic light of 480 ± 5 nm in wavelength obtained with an optical filter was irradiated to the sample. The optical axis of the analyzer was set to be perpendicular to that of the polarizer. The microscope images of the optical cell were obtained with a CCD camera (Neptune 100, Watec) and stored in a personal computer through a video converter (DFG/USB2pro, The Imaging Source). A pair of electromagnets (TMS202−1504530, Gigateco) generated a homogeneous magnetic field at the sample position. The electromagnets were connected to the programmable bipolar power supply (BP4620, NF Corp.), and the strength of the magnetic field was monitored with a Gauss meter (HGM-300P, Magnix). The direction of polarization of light was adjusted to be 45° against the magnetic field. The magnetic field was increased from 0 to 0.43 T at a set sweep rate and then was decreased from 0.43 to 0 T in the same sweep rate. The sweep rate was changed from 0.72 to 14 mT s−1. We built a program to control the magnetic field and to accumulate polarized microscope images of the 5CB layer simultaneously with LabVIEW. To avoid an increase in temperature of the sample induced by Joule heat of the electromagnets, the observation system was equipped in a thermostatic holder, which was a copper cylinder wounded by a thermal exchange tube. The sample was kept at 25 °C by circulating thermostated water in the tube. The threshold magnetic field for the Fréedericksz transition, Bc, was determined from the change of the intensity of light passing through each compartment. The region of interest was set at the middle of each compartment to avoid the edge effect of the Cu grid.15 Measurements of the Thickness of the LC Layer. The thickness of the 5CB layer, d, was determined from an interference appearing in reflection spectrum. The details of this method are explained in the SI, part III. In brief, light of a spectral range from 575 to 740 nm from a xenon arc lamp was irradiated coaxially from the cell bottom; then, the interference of the reflected light between the upper (aqueous phase/LC) and the lower (LC/OTS) interfaces was observed. The Fourier transform of the measured spectrum yielded d value. Our setup allowed us to measure the d of the 5CB layer in each compartment of the grid. Estimation of the Interfacial Excess of Surfactant at the Aqueous Phase/LC interface. For each surfactant, the interfacial tensions between its aqueous phase and 5CB were measured as a function of the surfactant concentration by the pendant drop method16 to determine the interfacial excess of surfactant, Γ, from the Gibbs isotherm.17 By a nonlinear least-squares fitting, the variables of the Szyszkowski equation (see SI, part I) were determined, which provided the relationship between the bulk concentration and Γ of surfactant.
phase/LC interface and a strongly anchored LC/OTS interface. The anchoring energy at the LC/OTS interface is relatively high (8 × 10−5 J m−2).18 Under these conditions, W at the aqueous phase/LC interface (WAQ/LC) is related to Bc with eq 2 (see SI, part II)11 Bc =
π d + (K3/WAQ/LC)
K3μ0 Δχ
(2)
Here K3 is the elastic constant of LC for bent deformation, μ0 is the vacuum magnetic permeability, and Δχ is the magnetic anisotropy of LC. The K3 and Δχ values of 5CB were reported to be 8.6 × 10−12 N19 and 1.56 × 10−6 20 at 25 °C, respectively. The substitution of obtained Bc and d in eq 2 yields W value. Fréedericksz Transition of the 5CB Layer between Two OTS-Treated Glass Plates. Figure 4 shows examples of
Figure 4. Changes of polarized light microscope images with the increase in the magnetic field.
the changes of polarized light microscope images of the 5CB thin layer between two OTS glass plates with the increase in the magnetic field. Under lower magnetic fields, no light passing through the 5CB layer was detected and there were no changes (a,b). 5CB stayed in the homeotropic alignment under the low magnetic fields, and therefore the optical anisotropy was not observed. Transmitted light through the 5CB layer was observed (c). Because the easy axis of magnetic anisotropy of 5CB was along its director, an application of a higher magnetic field than Bc induced a reorientation of 5CB molecules. This lead to the difference in the refractive indices between parallel to and perpendicular to the magnetic field. Therefore, linearly polarized light passing through the tilted 5CB molecules became elliptically polarized light, which was partially transmissive through the analyzer. Under higher magnetic fields, a blinking of the transmitted light (d,e) was observed with the increase in the magnetic field. This blinking was due to an
■
RESULTS AND DISCUSSION Principle of Measurements of the Anchoring Energy from Fréedericksz Transition. Fréedericksz transition is the deformation of a uniformly oriented LC layer by an external field. The present study employed a magnetic field as the external field. The schematic illustration of the deformation is presented in Figure 3. In our study, the LC layer was placed between two different interfaces, a weakly anchored aqueous C
DOI: 10.1021/acs.langmuir.7b03005 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 5. Observed light intensities passing through the 5CB layers against the magnetic field (a) with a sweep rate of 14 mT s−1 and (b) with a sweep rate of 0.72 mT s−1.
Figure 6. Variation of light intensities under a magnetic field that is linearly increasing (a) and decreasing (b). The magnetic field was plotted against the retardation, Δr (c).
magnetic field, at which the edge was observed, was not reproducible, and (2) the difference exists between the rising edge in the increasing process and the falling edge in the decreasing process. The problem (1) is attributed to the difficulty in adjusting the intensity of the incident light. With low intensity light, the transmitted light intensity just after the Fréedericksz transition in the increasing process was too weak to be detected, so the threshold magnetic field was overestimated. With high intensity light, the scattering of light induced glare in the image, and thus Bc was underestimated. It was difficult to find the appropriate light intensity for deciding the position of the correct edge of the curve. To avoid these problems, we utilized the oscillation of the transmission intensity after the Fréedericksz transition in Figure 5a,b because it was found that the magnetic fields at which the peaks and valleys appeared were invariant with the intensity of the incident light. Because this oscillation arose from the change of the birefringence in the 5CB layer, its retardation, Δr, can be determined from the number of the peaks and the valleys. When Δr is equal to (m − 1/2)λ or mλ,
increase in birefringence. When the retardation was an odd multiple of the half wavelength, the direction of polarization became parallel to the axis of the analyzer and the maximum light intensities were observed. Inversely, when the retardation equaled to an integral multiple of the wavelength, the polarization state turned back to the original linearly polarized light perpendicular to the analyzer and the minima were observed. Figure 5a shows typical changes of the intensity of the transmitted light passing through the 5CB layer against the magnetic field that was swept at 14 mT s−1. The red and blue lines indicate the light intensity measured during the increasing and decreasing magnetic fields, respectively. A significant difference was observed; the 5CB deformation by the magnetic field exhibited a hysteresis. Under a lower sweep rate (0.72 mT s−1), the difference reduced, as shown in Figure 5b. Therefore, this difference was attributed to a slow response of the 5CB deformation induced by the magnetic field. It seemed to be easy to determine the threshold magnetic field, Bc, from the edge of the curve, but there were two problems: (1) The D
DOI: 10.1021/acs.langmuir.7b03005 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir where m is a positive integer and λ is the wavelength, maximum and minimum intensities are observed, respectively, as shown in Figure 6a,b. The magnetic fields at which the light intensity exhibited the maxima or minima were plotted against the determined retardation (Figure 6c) in both the increasing and decreasing processes. Linear relationships were obtained; the intercepts of these lines correspond to the apparent threshold magnetic fields, Bc1 and Bc2, for increasing and decreasing processes, respectively. The problem (2) may be avoided by lowering the sweep rate. However, the elongation of the measurement time causes the temperature rise of the electromagnets by Joule heat, so higher sweep rate is preferable for this experiment. The dependency of Bc1 and Bc2 on sweep rate is shown in Figure 7. The gap
pure water or surfactants of an insufficient concentration, the tilted orientation of 5CB was observed as reported by Brake et al.13 With an increase in the surfactant concentration, the homeotropic alignment was induced by adsorbed surfactants at the aqueous phase/LC interface and by the octadecyl chains at the LC/OTS interface. So it was necessary to determine the minimum concentration of the surfactant required for the homeotropic alignment. Conversely, a higher concentration of surfactant caused a decrease in the accuracy in the reading of a reflection spectrum for the thickness measurements. This may be attributed to mechanical instability of the interface due to the decrease in the surface tension. The results are summarized in Table 1. The lower limit measured for C12TAB is in Table 1. Concentration Range of the Surfactants in Which the Anchoring Energy Was Measurable
C12TAB C14TAB
lower limit/ 10−6 M
upper limit/ 10−6 M
CMC/ 10−3 Ma
CMC/ 10−3 Mb
700 25
1000 70
∼15 ∼4
∼6 ∼ 0.8
a
Values in water from refs 22 and 23. bMeasured in the present study (0.1 M Li2SO4 and 0.01 M H2SO4).
agreement with that reported by Brake et al. 13 The concentration range became higher with the chain length. This would be due to the increase in the interfacial excess of the surfactant. It was reported that thermotropic LCs are soluble in micelles in aqueous solutions.21 Therefore, it is important that the surfactant concentration should not be higher than the critical micellar concentration (CMC). Table 1 also includes CMC values of the used surfactants in water and in the present system. Anchoring Energy of Aqueous Phase/LC Interface. WAQ/LC was measured within the limit concentrations of the surfactants shown in Table 1. Two alkyltrimethylammonium bromides, C12TAB and C14TAB, were examined to investigate the effect of chain length on the anchoring energy. As shown in Figure 8, the WAQ/LC increased with the surfactant concen-
Figure 7. Changes of Bc1 and Bc2 against the sweep rate of the magnetic field. Bc1 (red), Bc2 (blue), and their average, Bcm (yellow), are plotted.
between Bc1 and Bc2 increases with the increase in the sweep rate. The extrapolations of Bc1 and Bc2 to the sweep rate of 0 T s−1 fall down to almost the same point, which corresponds to the correct Bc. Figure 7 also shows that the averages of Bc1 and Bc2 were invariant with the sweep rate. Hence, the average Bcm can be used as the estimation of Bc. The sweep rate was set to 14 mT s−1 in the following experiments. The errors in Bcm and d were evaluated by repeating the measurements of the 5CB layer in a compartment, and their standard deviations (SDs) were obtained to be 0.7 mT and 0.09 μm, respectively. In the present case, the top and bottom of the 5CB layer were anchored by the OTS glass plates. Because the LC/OTS interface has a strong anchoring energy, Bc can be expressed with eq 3 (see SI, part II) Bc =
π d
K3μ0 Δχ
(3)
The Bc·d value should be constant and was calculated to be 8.27 × 10−6 T m−1 from literature values of K3 and Δχ with eq 3. On the contrary, Bcm and d were measured in nine different compartments of the same grid to calculate an average and a SD of Bcm·d, and these measurements were repeated for three different optical cell. The obtained Bcm·d values were (8.30 ± 0.09) × 10−6, (8.32 ± 0.09) × 10−6, and (8.26 ± 0.13) × 10−6 T m−1, which agreed well with the calculated value. Therefore, it was elucidated that the measurements and OTS coating were reliable. Fréedericksz Transition of the 5CB Layer between an Aqueous Phase and One OTS-Treated Glass Plate. In this technique, the homogeneous homeotropic alignment of the 5CB layer is required in the absence of magnetic field. With
Figure 8. Plots of the anchoring energy (WAQ/LC) against the surfactant concentration in the aqueous phase. C12TAB in green triangle and C14TAB in red square.
tration in the aqueous solution. Near the lower limit of the surfactant concentration, WAQ/LC was
