1. Introduction

Due to their diverse areas of research and possible applications, colloidal systems in anisotropic systems have attracted a growing interest in the past years. Understanding the nature of interparticle interactions in nematic liquid crystals [1–3] provides a rich variety of new physical phenomena and represents a step forward in the fundamental physics of atomic and molecular systems as well as in applied science. Anisotropic molecular interactions lead to self-assembly of particles into linear chains [3–5] and more complex structures, such as colloidal crystals [6–8] that are useful in the important application of photonic crystals. Recetntly, the value of liquid crystal colloids is becoming increasingly successful as they are upgraded with optical manipulation [1,2,9–12]. Control over positioning, exerted force and torque on spherical objects in an anisotropic media present a step forward in development of tools for microscopic characterization and imaging.

Qualitatively new behavior of colloidal systems is observed when spherical symmetry of a colloidal particle is broken by its internal structure or by surface chemical composition. Extended studies of cholesteric and smectic droplets in aqueous media have therefore been performed to investigate the nature of interaction between polarized light and the liquid crystal. Since it is possible to transfer angular momentum from light to birefringent media [13], it has been shown that the dipole nature of interaction between light and liquid crystal can induce both, alignment of the internal droplets' structure [14] as well as their rotation [15]. There have been also done several studies of 'Janus' particles in aqueous suspension [16–18]. The idea of the Janus motif divides the surface of a spherical particle into two hemispheres of different chemical composition. Here, it is the surface assymetry of Janus particles that leads to rotational motion of such particles [18–20] in the presence of a laser field. Rotation is based on absorption of polarized light that transfers spin or angular momentum of the beam by spherical objects [13,19,20]. Such optical micro-rotators have potential applications in the field of microfluidics and sensing.

We report on experimental study of the behavior of Janus colloidal particles in the anisotropic media, nematic liquid crystal 5CB. The assymetry of colloidal particles is produced by “capping” one half of silica micro-spheres with a thin layer of gold to induce planar alignment of 5CB. The uncovered hemisphere is treated with a homeotropic DMOAP surfactant. As a result of detailed preparation chemistry we distinguish between DMOPA/Au and Au/DMOAP Janus particles showing different equilibrium orientations in a planar nematic cell. Using polarizing optical microscopy, we discuss the observed nematic profiles. Finally, we use optical tweezers to control particle's orientations and transformations in the presence of a laser field.

2. Sample preparation and experimental set-up

Experiments were performed with 4.3 µm silica particles (Bangs Laboratories Inc.). Sample preparation starts with the production of a monolayer of silica colloids on a glass substrate (Fig. 1 left). First, the surface of silica particles is methoxylated by treating it with methanol. The particles are then dispersed in methanol and centrifuged. After repeating this procedure several times, we add the prepared suspension onto the surface of water, and wait for the colloidal particles to suspend at the air-liquid interface. A pure glass substrate is then mounted in the water containing column, in the direction perpendicular to the water-air interface. Finally, as the water slowly poures out of the column, the silica colloidal particles are deposited onto the glass substrate in a form of a colloidal monolayer.

 

Fig. 1 Left: SEM image of 2D array of 4.3 µm silica particles on a glass substrate. Right: SEM image of 4.3 µm silica particles coated with 3 nm of Cr and 50 nm of Au.

Download Full Size | PDF

Two different procedures have been used to produce a combination of planar – homeotropic alignment, at the particle surfaces. Following the first method of surface functionalization, silica microspheres were first capped with a gold layer. The monolayer sample was transferred into an evaporation chamber, where a 3 nm adhesive layer of chromium and 50 nm layer of gold was sputtered onto the spheres. Afterwards, the particles were detached from glass substrate into water with ultrasound. SEM image (see Fig. 1) of silica capped colloids clearly shows that one hemisphere of colloidal particle is capped with gold. One can also see, that the line, dividing gold-coated and non-coated hemispheres is not straight, but “wavy”, which is due to not optimized conditions during the sputtering process. In the next step, a N,N-dimethyl-N-octadecyl-3-aminopropyl- trimethoxysilyl chloride (DMOAP) monolayer was chemically attached to the non-capped hemisphere. As the DMOAP silane does not bind to the Au surface, we have obtained capped colloidal particles, where the surface of DMOAP-covered hemisphere ensured strong homeotropic surface anchoring of a nematic liquid crystal, whereas gold-covered hemisphere induced planar surface anchoring conditions. According to the preparation procedure, we shall refer to those samples as Au/DMOAP.

In the second method, silica microspheres were first completely covered with DMOAP, followed by a monolayer formation described above. Monolayer was then capped with a gold layer and detached from the glass substrate into water. We can conjecture, that the planar alignment on sputtered Au is stronger compared to the homeotropic alignment on DMOAP, which was formed before the sputtering process. We shall therefore refer to DMOAP/Au samples in this case.

Finally, capped colloids were dried out of water solution and dispersed in the nematic liquid crystal 4-n-penthyl-4-cyanobiphenil (5CB). The colloidal dispersion was introduced into a 10 µm thick planar cell and analyzed under the polarizing microscope. After filling the cell with colloidal dispersion, we obtain different orientations and director patterns around microspheres, mainly depending on the method of surface preparation.

For optical manipulation we have used a laser tweezers setup built around an inverted microscope (Nikon Eclipse TE2000-U) with a fast video camera and a green laser source (Coherent Innova 90C) at wavelength 514 nm. We have used a computer-controlled acousto-optic deflector for trap manipulation. The incident laser light was linearly polarized and focused onto the sample by the water immersion 60x objective (Nikon, working distance of 2.8 mm).

We should stress, that the golden-capped colloidal particles had to be manipulated with lots of care and preferrably at very low laser power level. Optimal tweezing power in our experiments was 10-20 mW. Due to the strong absorption of light at the golden hemisphere, the long term trapping at higher laser powers could result in heating of the liquid crystal media around the colloidal particle into the isotropic phase. Once the liquid crystal was overheated, the colloidal particles stuck on to the glass wall of the liquid crystal cell after switching off the laser. The surrounding liquid crystal however cooled back into the nematic phase.

3. Results and discussion

3.1 Au/DMOAP Janus particles

Immediatelly after filling the LC cell with a nematic Au/DMOAP colloidal dispersion, we observed a dipolar director configuration around the capped particle (Fig. 2 ) in more than 95% of observations. The observed elastic dipole of a capped particle is almost indistinguishable from an elastic dipole formed around a non-capped, only silanated particle. Small difference is observed only in the position and local details of the hyperbolic point defect.

 

Fig. 2 Microscope images of the three possible configurations of 4.3 µm Au/DMOAP capped colloids in a 10 µm thick planar cell of 5CB. When the dipolar capped colloidal particle (middle) is disturbed with a weak laser filed, we observe an irreversible transition into two possible stable configurations: Saturn ring (left) and novel, boojum-ring (right) configuration. The arrows in the dipolar capped colloidal particle images indicate rubbing direction of the planar cell. In the upper images no polarizers are used, the lower images show capped colloidal particles as observed under crossed polarizers. Schematics of the director fields are sketched above the corresponding configurations.

Download Full Size | PDF

Such dipolar configuration of the capped particle is however unexpected, as the golden hemisphere, which is expected to induce planar alignment, acts as effectively homeotropic. It is the preparation procedure of the particles' surfaces that is most probably responsible for the effective change in the surface properties. This implies that the golden hemisphere determines the overall orientation of the particles through the energy balance between the two hemispheres with homeotropic anchoring of different strength. An alternative explanation, such as charging of the conductive golden surface layer is however also possible. Particles being effectively dipolar, screen to most extent particles' Janus nature.

When Au/DMOAP colloidal particles in the nematic liquid crystal were manipulated with a laser field, irreversible reorientations into two final stable states were observed (Fig. 2). The first configuration (Fig. 2, left) is a capped colloid with a Saturn ring, positioned roughly at the border line between the homeotropic and planar hemisphere. The second stable state (Fig. 2, right) represents a newly observed configuration with a surface boojum at the planar site (golden hemisphere) and a Saturn ring analog, positioned at the border line between the homeotropic and planar hemisphere. This configuration, called a boojum-ring, is discussed in details in Ref. 21.

We should stress, that in both stable states, the border line between the two hemispheres remains positioned perpendicular to the rubbing direction. In addition, the original dipolar structure could not be reestablished by any possible means of external field. This implies, that the original dipolar colloidal state was a metastable state, most probably induced by a liquid crystal flow during filling the cell with dispersion.

3.2 DMOAP/Au Janus particles

When the planar LC cell is filled with a DMOAP/Au colloidal dispersion, the capped particles orient themselves with a borderline between the Au and DMOAP hemisphere nearly horizonthal or inclined at an angle up to ~10⁰ with respect to the overall orientation of the LC director, n₀ (Fig. 3 ). Microscope images of the particle reveal two characteristic spots located symmetrically on each side of the equator of the particle. The spots are dark under crossed polarizers (Fig. 3a) and faintly dark with no polarizers (Fig. 3b), which is a characteristics of point defects, such as boojum defects [3].

 

Fig. 3 Microscope images of a 4.3 µm DMOAP/Au capped microsphere in a 10 µm thick planar cell of 5CB. The arrows indicate rubbing direction of the planar cell. (a) The capped colloidal particle between crossed polarizers. (b) No polarizers are used. (c) Schematics of the director field.

Download Full Size | PDF

Light microscopy observations therefore suggest the DMOAP covered hemisphere acts as a surface with planar anchoring that is even stronger than on the gold-coated side despite being covered with a homeotropic surfactant. Again, this is a result of the detailed chemistry of the Janus particle's preparation process, where the Au vapor may have diffused during the sputtering to the other side of the microspheres. Such an equilibrium configuration was observed in more than 90% of observations. In addition, this configuration is very stable and colud not be permanently changed by external action, such as laser tweezers field.

However, when manipulating DMOAP/Au colloids with a laser field, we do observe a well defined, optically induced rotation of the trapped particle. After setlling itself in a stable trapping position close to the boojum defect, the colloid starts to rotate around the axis parallel to the borderline between the two hemispheres, that is, around an axis perpendicular to the laser beam axis (Fig. 4 ). Here, the light pressure propels the motion of colloids due to their assymetry which is introduced by capping them with a layer of gold [18,20]. The stability of optical trapping and rotation arises from the transfer of optical momentum to the particle, depending on the balance between the gradient force and the scattering force. When the gradient force overcomes the gravitational force and if the scattering force induces a torque, we can achieve both, trapping and rotation around the axis perpendicular to the beam axis [20].

 

Fig. 4 Left: Rotation of a 4.3 µm DMOAP/Au capped colloid in a 10 µm thick cell when trapped with a weak laser field of 10 mW. The colloid rotates around the axis parallel to the rubbing direction indicated by the arrow. The sequence of non-polarized images (a) - (h) depicts one cycle of 360°. Positions (a) and (e) are the two stable positions of a DMOAP/Au colloid in the planar nematic cell, positions (c) and (g) denote the orientation of a golden cap above and below the silanated hemisphere. The red cross marks the position of a laser trap. Right: Measured rotation frequency vs. incident laser power.

Download Full Size | PDF

The sequence of images in Fig. 4 depicts trapping and rotation of a capped colloid in the presence of a weak laser field of 10 mW. In one cycle, the colloid rotates monotonously from the starting position, which is observed as half bright, half dark (Fig. 4 a), through the position where the golden cap (i.e. non-transparent dark side) in Fig. 4 c is oriented above or below the silantated hemisphere into the opposite stable position in Fig. 4 e. The cycle is finished with further rotation through the state where the golden cap is oriented below or above the silantated hemisphere (Fig. 4 g) into the starting position (Fig. 4 h). If the trap is switched off during the rotation, the colloid always returns into one of the bistable equilibrium positions (Fig. 4a or 4e) with the border line between the silanated hemisphere and golden cap being aligned with the rubbing direction. One ahould note, that during the rotation, the charcteristic boojum defects remain located on each side of the particle's equator. Only small shifts of the boojums [21] in the z-direction towards the hemisphere with weaker anchorig are observed throguh the slight change of focus. However, the elastic deformation of the nematic induced by the particle's rotation remains negligible, which is reflected in monotonous rotation within a single cycle.

The frequency of rotation was measured by using the time sequence of recorded video images. The graph in Fig. 4 shows that the rotation frequency is approximately proportional to the incident laser power used for trapping. As the rotation within a single cycle appears constant, we conclude that the monotonous increase of rotation rate with increasing laser intensity is mainly a result of the balance between the externally applied optical torque and the torque arising form the viscous drag of the liquid crystal. The effects of elastic torque are unmeasurable since the expected non-monotonous rotation within each cycle is not observed.

4. Conclusions

We have presented a structural and light driven analysis of Janus nematic colloids, where one hemisphere was treated with DMOAP homeotropic surfactant, whereas the other hemisphere was gold-capped to assure planar alignment of a nematic liquid crystal 5CB. We have shown that orientations of Janus capped colloids as well as director field configurations around them strongly depend on the technique, used for preparation of surfaces. Au/DMOAP capped colloids, where Au layer was deposited before silanization with DMOAP, show a dipolar-like configuration. Using optical tweezers, we were able to irreversibly change the dipole either into a Saturn-ring or newly observed boojum-ring configuration. On the contrary, DMOAP/Au Janus colloids showed only one stable, boojum-like configuration. Using laser light, we were able to induce a rotation of the DMOAP/Au particle around the axis perpendicular to the laser beam as a consequence of optical momentum transfer. The rate of rotation resulted from the competition between the externally applied torque and the torque arising from the viscous drag of the surrounding liquid crystal.