Observation of a photo wetting effect on anisotropic ... - OSA Publishing

Observation of a photo wetting effect on anisotropic liquid-solid interfaces Guillaume Goubert and Tigran Galstian* Center for Optics, Photonics and Laser, Department of Physics, Engineering Physics and Optics Université Laval, Pavillon d’Optique-Photonique, 2375 Rue de la Terrasse, Quebec, Canada G1V 0A6 * Corresponding author: [email protected]

Abstract: This paper reports on the experimental observation of an anisotropic photo controlled wetting effect. We discuss how capillary propagation of nematic liquid crystals in a “sandwich” cell is blocked in areas exposed to light. We postulate that the underlying mechanism is related to the optical reorientation of molecules on the surface of the surrounding solid, which in turn, forces a corresponding reorientation of the liquid crystal molecules. The corresponding orientation deformation energy of the liquid then changes the balance of forces and the corresponding capillary action. © 2009 Optical Society of America OCIS codes: (160.5470) Polymers; (230.3720) Liquid-crystal devices; (230.5440) Polarization sensitive devices; (310.6860) Thin films, optical properties; (350.6670) Surface photochemistry.

References and links 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

12. 13. 14.

15. 16.

M. K. Chaudhury and G. M. Whitesides,“How to make water run Uphill,” Science 256, 1539-1541 (1992). R. Fürstner, W. Barthlott, C. Neinhuis, P. Walzel,“Wetting and Self-Cleaning Properties of Artificial Superhydrophobic Surfaces,” Langmuir 21, 956-961 (2005). A. M. Cazabat, F. Heslot, S. M. Troian, P. Carles,“Fingering instability of thin spreading films driven by temperature gradients,” Nature 346, 824-826 (1990). M. Schneemilch, W. J. J. Welters, R. A. Hayes, J. Ralston,“Electrically Induced Changes in Dynamic WettabilityElectrically Induced Changes in Dynamic Wettability,” Langmuir 16, 2924-2927 (2000). T. D. Blake, A. Clarke, E. H. Stattersfield,“An Investigation of Electrostatic Assist in Dynamic Wetting,” Langmuir 16, 2928-2935 (2000). S. Kataoka and M. A. Anderson,“Capillary rise between two TiO2 thin-films: evaluating photo-activated wetting,” Thin Solid Films 446, 232-237 (2004). M. Miyauchi, N. Kieda, S. Hishita, T. Mitsuhashi, A. Nakajima, T. Watanabe, K. Hashimoto,“Reversible wettability control of TiO2 surface by light irradiation,” Surface Science 511, 401-407 (2002). R. Wang, K. Hashimoto, A. Fujishima, M. Chikuni, E. Kojima, A. Kitamura, M. Shimohigoshi, T. Watanabe,“Light-induced amphilic surfaces,” Nature 388, 431-443 (1997). J. Y. Shin and N. L. Abbott,“Using Light to Control Dynamic Surface Tensions of Aqueous Solutions of Water Soluble Surfactants,” Langmuir 15, 4404-4410 (1999). K. Ichimura and S-K. Oh, M. Nakagawa,“Light-Driven Motion of Liquids on a Photoresponsive Surface,” Science 288, 1624-1626 (2000). Z. Sekkat, M. Biichel, H. Orendi, H. Knobloch, T. Seki, S. Ito, J. Koberstein, W. Knoll,“Anisotropic alignment of a nematic liquid crystal controlled by a polarization sensitive Langmuir-Blodgett command layer,” Optics Communications 111, 324-330 (1994). D. Dumont, T. Galstian, S. Senkow, A. M. Ritcey,“Liquid Crystal Photoalignment using New Photoisomerisable Langmuir-Blodgett Films,” Mol. Cryst. Liq. Cryst. 375, 341-352 (2002). V. Chigrinov, V. Kozenkov, H-S. Kwok, Photoalignment of Liquid Crystalline Materials, Wiley, 2008. E. Belamie, G. Mosser, F. Gobeaux, M. M. Giraud-Guille,”Possible transient liquid crystal phase during the laying out of connective tissues: alpha-chitin and collagen as models,” Journal of Physics: Condens. Matter 18, S115-S129 (2006). V. P. Pham, T. V. Galstyan, A. Granger, R. Lessard,“Novel azo dye-doped poly (methyl methacrylate) films as optical data storage media,” Jpn. Journal of Applied Physics 36, special issue, Part 1, No. 1B, 429-438 (1997). A. W. Adamson and A. P. Gast, Physical Chemistry of Surfaces, Wiley, New York, 1997.

#107595 - $15.00 USD Received 19 Feb 2009; revised 11 Apr 2009; accepted 19 Apr 2009; published 26 May 2009

(C) 2009 OSA

8 June 2009 / Vol. 17, No. 12 / OPTICS EXPRESS 9637

17. A. Yavrian, T. V. Galstian, M. Piché,“Circularly polarized light-induced rearrangement of optical axis in photoanisotropic recording media,” Optical Engineering 41(04), 852-855 (2002). 18. X. Tong, G. Wang, Y. Zhao,“Photochemical Phase Transition versus Photochemical Phase Separation,” J. Am. Chem. Soc. 128, 8746-8747 (2006). 19. E. Grelet and H. Bock,“Control of the orientation of thin open supported columnar liquid crystal films by the kinetics of growth,” Europhys. Lett. 73 (5), 712–718 (2006). 20. T. T.-Katona and A. Buka,“Nematic-liquid-crystal–air interface in a radial Hele-Shaw cell: Electric field effects,” Physical Review E 67, 04171(1-7), (2003). 21. P. G. de Gennes, J. Prost, The Physics of Liquid Crystals, Second Edition (Oxford University Press, 1995). 22. L. M. Blinov and V. G. Chigrinov, Electrooptic Effects in Liquid Crystal Materials, (Springer, 1994).

1. Introduction Interaction mechanisms on the interface between liquids and solids are of paramount importance in our day to day life. A simple example of the important consequences of such an interaction is the capacity of liquids to wet surfaces. Our level of knowledge and capacity to control these interactions open up exciting avenues for various important applications ranging from industrial (e.g., painting) to medical (e.g., drug delivery) areas. It is well known that the final geometrical form and capillary behavior of a liquid is defined by the balance of the attractive forces between the surface and internal molecules of the liquid, the resistance of the liquid to compression and forces produced by the surface molecules of solid substrate on the liquid. Since antiquity, there have been many attempts to modify the wetting process by changing the properties of liquids and/or solid surfaces. These modifications may have had a permanent chemical [1] or architectural character (such as the lotus flower [2]) or a dynamic (reversible) character. Known dynamic control mechanisms are the use of temperature [3], electric field [4,5] and light (see below). In fact, one of the most promising methods of dynamic controlling the wetting effect is the use of light, particularly in the ultra-violet (UV) and visible (VIS) spectra. Different mechanisms for this kind of control have been previously reported. Photo activated (via the photocatalytic process) oxygen (bridging oxygen) defects (holes) have been used to increase the wetting of water to the surface of TiO2 [6], for example. A similar reversible effect was reported in Ref. [7] based on the UV induced increase in dissociated adsorbed water on the surface and its return to the hydrophobic state by VIS light induced elimination of surface OH (by the thermal process). In fact, these effects are so strong that the TiO2-coated glass may become reversibly anti-fogging and self-cleaning [8]. A different class of reversible photo wetting phenomena was reported in material systems containing azobenzene molecules, either doped into the wetting liquid or introduced onto the solid surface. Thus, a dynamic UV control of the surface tension of liquid solutions was demonstrated, using light to change the extent of aggregation of surfactants within the bulk solution (by using water soluble surfactants that contain azobenzene [9]). In contrast, for a solid surface, which contains azobenzene molecules, a net mass transport was demonstrated by asymmetrical photo irradiation, which caused a gradient in surface free energy due to the photo isomerization of surface azobenzenes, leading to the directional motion of the droplet [10]. The above mentioned experiments have investigated azo dye interactions with isotropic liquids only. The study of interaction of azobenzene containing surfaces with anisotropic liquids, such as liquid crystals (LC), was reported in several works [11,12]; (see also the review in Ref. [13]). However, to the best of our knowledge, all previous investigations

focused on the optical control of the orientation of the director n (average direction of long molecular axes) of the LC and no observations were reported concerning the photo induced wetting effects due to the anisotropic character of liquid – surface interactions. In this present work we are reporting on the observation and preliminary study of light induced reorientation of molecules of an azo dye containing surface and the resulting un-


(C) 2009 OSA


wetting of nematic LCs. The fundamental peculiarity of this material system is the presence of an additional source of free energy, which is the energy of orientational deformation of the director of the LC, which affects the wetting conditions. One of the reasons of our interest in such systems is the abundant presence of anisotropic liquid structures (cell membranes, fibrils, DNA, collagen, etc.) and their crucial role in living organisms [14]. 2.

Material system

To build our photo controllable surfaces, we used a commercially available reactive mesogene (RM) RMS03-001 (from Merck). We used the solvent PGMEA (from Merck) for spin coating the RM. The concentration of the RM in the PGMEA was 30 wt%. This RM was used for several reasons: firstly, this material is in the liquid crystalline mesophase at room temperature, which greatly facilitates experimental work (the Nematic-Isotropic phase transition is 75°C and the solution must be kept above 4°C to avoid crystallization). Also, we used this RM since it takes on a planar alignment when spin coated on a rubbed “planar” polyimide (PI) surface. Before being spin coated on the PI, the mixture of RMS03-001 / PGMEA was doped (1wt%) by the azo dye (further called AZD2 [15]) as shown below:

The mixture was homogenized using magnetic mixers and an ultrasound bath for 20 minutes. 1wt% of PI (PI150), dissolved in the solvent S21 (both from Nissan Chemicals, LTD), was used to generate planar alignment (see later). The LC material used was TL216 (from Merck). It has a positive optical anisotropy. 3.

Cell fabrication

Commercially available (from TFD) indium tin oxide (ITO) coated Float glass substrates were used to build cell “sandwiches”. Those substrates (with sizes 10 mm x 10 mm x 0.7 mm) were first cleaned and then covered by a thin film of PI by spin coating. The spin coating conditions were: 500 rpm for 5s, then 3000 rpm for 20s. The PI coated ITO-glass substrates were then heated in an oven (80°C for 10 minutes, then 280°C for 45 minutes) to remove the rest of the solvent and to bake the PI prior to mechanical rubbing. The thickness of PI layer obtained was between 60-80 nm (as measured by a Dektak profilemeter and a Horiba Jobin Yvon ellipsometer). The final step of the substrate preparation was to spin coat a thin layer of the AZD2 doped RMS03-001 / PGMEA solution described above. The spin coating conditions were: 5000 rpm for 25 sec. The film was then prebaked at 55°C for 1 minute. At room temperature, the RM layer was in the LC phase and aligned in the direction of the PI rubbing. This was verified using a standard polarimetric set-up. These films had to be polymerized to avoid their dissolving into the LC (see below). The glass substrate coated with ITO/PI/RM films was then inserted into a chamber with optical windows. The chamber was filled with Nitrogen (to minimize the access of Oxygen) and the RM layer was photopolymerized for 30 minutes by means of a spatially uniform non polarized UV light at normal incidence (UV lamp intensity was ≈ 5 mW.cm-2 and its spectra was centered at 375 nm; measured with a Gentec UV power meter). The oriented and solid RM layers, thus obtained, had thicknesses ranging from 600-800 nm (measured by


(C) 2009 OSA


profilemeter and ellipsometer). Two such substrates were then used (with the RM layers facing each other) to build a LC cell of thickness L=45µm (using cylindrical spacers mixed into a UV curable glue, which was dispensed on the periphery of the cell). In contrast to the above mentioned procedure, another set of substrates were UV polymerized while being simultaneously exposed locally by a circularly polarized CW Ar+ laser beam (operating at 514 nm, near the absorption peak of the AZD2). The laser beam was incident from air (at normal incidence) onto the RM layer to avoid change of its polarization if it would be propagated in the RM layer (incidence from glass side). This was done to reorient the azo dye and RM molecules in the direction normal to the surface (see later). The diameter of the laser beam was 2 mm ± 0.2 mm and its power was varied (for different experiments) between 50-100 mW. Special care was made to assemble the cell in such a way to have the laser exposed areas of two substrates facing each other. The LC was introduced into the cell by capillary action at room temperature. These cells had well aligned LC orientation in the areas which were not exposed to the laser beam during the UV photopolymerization (verified by the microscope and polarimetric set-up). 4.

Experimental observations

The set of control cells, which were fabricated without the laser exposed substrates, were filled in the standard way. Uniform fill and orientation of the LC were obtained. In contrast to these, cells which were build by using laser exposed substrates, showed an extraordinary behavior. Namely, capillary action moved the LC up to the laser exposed area as usual. However, this LC movement stopped near to the border of the area exposed to laser light. The LC completed the fill process everywhere else, while an air bubble was formed in the area (Fig. 1) which was exposed to the laser beam. We wish to emphasize that, in contrast to


(C) 2009 OSA


Fig. 1. The photography of the LC cell (placed between two polarizers) after the capillary fill process. The white arrow shows the laser exposed area. Colors of the photo were changed to better reveal the air bubble in the laser exposed area.

occasional air bubbles that may time-to-time appear in capillary fill experiments, the air bubbles obtained in our case are always in the same position (in the laser exposed area), they have the same form, they do not move either during or after the capillary fill and they are extremely stable. Also, the diameter of the air bubble (2.3 mm ± 0.2 mm) is approximately equal to the diameter of the laser beam. Closer analyses (done by using a polarizing microscope) of the “wall”, which separates the air bubble from the LC, shows complex multiple zone structure (Fig. 2), which may be roughly regrouped into three zones noted as external (on the right side of the border 3), internal (on the left side of the border 1) and mid (noted as border 2) “zones”, the terms


(C) 2009 OSA


1 2 3 Air bubble

Liquid crystal

Fig. 2. The micro photography of the complex wall of the air bubble obtained by using a polarizing microscope. The director of the LC is aligned in a way to optimize the contrast of different zones. The dotted arrow shows the direction of movement of “walls”.

“external” and “internal” being applied to the air bubble. The application of an electric field (using the two ITOs of opposed substrates) caused the well known dielectric-torque induced reorientation of LC molecules, aligning them perpendicular to the substrate surfaces (the LC has positive dielectric anisotropy at the driving frequency). This resulted in the modification of the above mentioned zones at different degrees. Thus, the Fig. 3 a), b) and c) represents the changes of the positions of the borders 3, 2 and 1, respectively. For a reference, the initial

18 16

Distance (µm)

14 12 10 8 6 4 2 0 -2 0

5

10

15

20

RMS voltage (V) a)


(C) 2009 OSA


4

Distance (µm)

3

2

1

0

0

5

10

15

20

RMS voltage (V) b) 3.5 3.0

Distance (µm)

2.5 2.0 1.5 1.0 0.5 0.0 0

5

10

15

20

RMS voltage (V) c) Fig. 3. Electrically induced movement of the complex wall structure for a) external zone 3, b) mid zone 2 and c) internal zone 1.

distance between the “walls” 1 and 2 is approximately 35 µm (it shrinks upon the application of an electric field, see later). The electric field induced movement of those zones showed some initial “fluctuation” of their position and further overlap of some of “walls” (particularly in the zone between ‘walls” 2 and 3), but was then directed towards to the center of the air bubble (thus slightly shrinking the bubble). Note that the dielectric torque induced movement of the “wall” shows the importance of its anisotropic character (see below).


(C) 2009 OSA


5.

Discussion

We believe that the key factor, in the origin of the observed phenomenon, is surface interaction, since the thickness of our LC layer is 45µm, while the diameter of the “bubble” is approximately 2.3 mm (in addition, the LC propagation is stopped once it “meets” the first border of the area exposed to laser light). It seems that the laser exposition creates an energetic barrier for the propagation of the LC by capillarity near the periphery of the laser exposed area and reduces the surface wetting by the LC. In contrast to “traditional” mechanisms, this exposure did not change the chemical composition of the surface. In traditional systems, the contact angle θ is described by the Young’s equation for flat and solid surfaces [16]: γLV cos θ = γSV − γSL

where γSL , γSV and γLV are surface tensions (interfacial free energy per unit area) of the solid-liquid, solid-gas (vapor) and liquid-gas interfaces, respectively. Note that the surface analysis (made by means of a Quesant atomic force microscope, AFM) did not reveal any noticeable relief modulation in the area exposed by the laser. Also, both AFM and microscope analyses do not show any sign of phase separation in that area. This is why we think that the formation of the air bubble during the capillary fill is primarily related to the reorientation of surface molecules (both azo and RM). In fact, in our material system, the laser exposition may generate two phenomena: trans to cis photo isomerization of the azo dye and reorientation of the trans molecules (as a result of cis to trans relaxation) into directions, where the dipole moments of those molecules have lower projections on the polarization of the excitation laser [15]. Thus, in the case of a circularly polarized and normally incident laser beam, this is the direction of the normal of the surface [17]. Given the liquid crystalline character of the host matrix (the RM), into which the azo dye molecules are doped, the reorientation of azo dyes should force the “collective” reorientation of the molecules of the RM too [18]. For the same reason (the liquid crystalline character of the RM) the molecules of the LC are forced to follow the same reorientation. In fact, initially the RM aligns naturally in the rubbing direction of the PI. The same happens to the LC molecules when they are put in direct contact with the same PI surface. So it is quite natural that the LC is aligned in the planar direction on non exposed (to the circularly polarized laser) surfaces after the capillary fill. We believe that it is also natural that the orientation of the LC must undergo a strong deformation when approaching the laser exposed area, since in this area, the surface molecules are reoriented to be perpendicular to the surface (as a result of the circularly polarized light exposure and UV polymerization). The schematic representation of such capillary fill may be found in the Fig. 4 (the cross section of the cell). The arrow (from left to right) shows the propagation direction of the LC. Fig.4a shows the LC propagation process in the non exposed (by laser) area in the case where the air-LC interface is homeotropic (LC molecules are aligned perpendicular to the air-LC surface). Since the alignment of surface molecules (both azo and RM) is planar, then the capillary action should create two disclination lines near to the top and bottom corners of the propagation front (shown by two dots). Still in the case of homeotropic air-LC interface, the situation changes when the LC approaches the laser exposed areas, as demonstrated by the


(C) 2009 OSA


a)

b)

c)

d)

Fig. 4. Schematic (cross section) representation of the capillary fill process of the liquid crystal in the cases of a)&b) homeotropic alignment of molecules at air-LC interface and c)&d) planar alignment of molecules at air-LC interface. a)&c) show the capillary propagation in uniform areas, while the b)&d) show the capillary propagation near to the laser exposed areas. The dashed lines inside of the two parallel (horizontal) solid lines represent the LC molecules. The dashed lines outside of the two parallel (horizontal) solid lines represent the molecules of reactive mesogene and azo dye. The arrow shows the LC flow direction (from left to right). The area on the left side of the curved surface is the air bubble. θ is the wetting angle. Dots represent the disclination lines. Laser exposed area is on the right side of the vertical line 3.

vertical dotted line in the Fig. 4b. As demonstrated schematically, in this area, the perpendicular orientation of surface molecules should reduce the elastic deformation energy, helping to keep the wetting angle low and promoting good propagation due to capillary. In contrast, in the case of planar air-LC interface (LC molecules are parallel to the air-LC surface), the initial orientational deformation energy (with only one disclination line when propagating in non exposed areas, Fig. 4c) increases when the LC approaches the laser exposed area (Fig. 4d). This laser exposed area is shown extending from the line 3 and to the right of the line 1. One can see schematically that the further propagation of LC in those areas is possible “at the price” of the formation of two additional disclination lines (shown by two dots). Thus, those areas create an energetic barrier for the capillary propagation. We must emphasize that the air-LC interface is believed to provide homeotropic alignment, which does not corresponds to the model described above. The previously discussed zones (1-3) are schematically positioned (in a very approximate way) in the Fig. 4d by vertical dotted lines 1-3. The area on the left side of the vertical line 3 is the uniform (planar) oriented LC. The area between lines 3 and 2 is the orientational transition zone. The area on the right side of the line 1 is the air bubble. The “movements” of the three “walls” under the action of the electric field are thus related to the fact that the liquid crystal has a positive dielectric anisotropy and is forced to align parallel with the applied #107595 - $15.00 USD Received 19 Feb 2009; revised 11 Apr 2009; accepted 19 Apr 2009; published 26 May 2009

(C) 2009 OSA


electric field. However, it is well known [19,20] that the air-LC interface has a very strong energy and may be considered as strongly oriented layer. That is why there is a narrow zone (delimited by the line 2) near the air-LC interface which is affected much less. Similarly, the extreme limit of that zone (delimited by the line 1) is affected very weakly. Thus, the movements of the lines 1&2 are related to the change of molecular orientation of the near surface (air-LC) zone and of the LC wetting angle. To recap, we obtain a solid surface, the chemical composition of which is uniform and which must be wet to allow the LC propagation. However, the molecular orientation of this surface is not uniform. When approaching the laser exposed zone, the LC molecules tend to be oriented in the direction of RM/AZD2 molecules (along the normal of the surface) given their liquid crystalline character. This creates additional free energy (due to the additional orientational deformation of LC, see below) and prohibits the capillary fill. The application of an electric field causes the reorientation of LC molecules and moves border 3 towards the center of the air bubble. This reorientation is significantly less for border 2 since there is already a bend reorientation (towards the normal of the surface) of LC molecules and the border is too close to the air-LC interface. The movement of border 1 is rather small too and is related to the movement of the internal wall of LC and its wetting angle. If we suppose that, in the static regime, there is, for example, a bend deformation of the director n , then the corresponding (to that deformation) additional volume free energy 1 density Fd of the nematic LC may be expressed by the equation Fd = K 3 (n × rot n)2 , where 2 K 3 is the corresponding elastic constant (typically at the order of 10-6 erg/cm) [21,22]. For a deformation scale of l , and in the so called one-constant elasticity approximation (K1=K2=K3≡ K), the very rough estimation of the energy Fd may be evaluated by the K [21,22]. Thus, for a typical value of l ≈ 0.2 x 10-3 cm (the l2 typical size of the “movement” of the line 2, at the order of 2 µm) we obtain Fd ≈ 25 erg/cm3 = 2.5 N/m2. At the same time the surface tension coefficient A (which is at the order of UNS/a2; where UNS is the anisotropic part of the energy of Nematic-Substrate interaction and a is the average molecular length) may be estimated as A ≈ 0.5 x K 3 L2q3, where q=2π/ l )

following expression Fd ≈

[21,22]. We obtain (by using the values L ≈ 45 x 10-4 cm and l ≈ 0.2 x 10-3 cm; and thus q=31.4 x 103 cm-1) the order of magnitude of A ≈ 0.5 x 10-6 erg x cm-1 x 2025 x 10-8 cm2 x 30959.14 x 109 cm-3= 313.46 x 10-3 N/m, which is a rather significant amount. In fact, more accurate estimations should take into account the energy of orientation disclinations, which is a matter for separate study. Just for reference, let us note however, that in the situation of almost complete wetting (with wetting angle θ →0) the typical interfacial free energy per unit area, e.g., for the water wetting on the TiO2 surface, is γLV ≈ 73 x 10-3 N/m, while we have the value γLV ≈ 485 x 10-3 N/m for the wetting angle θ between 122-140° (which is almost an unwetting state) of the mercury droplets on the surface of TiO2 [6]. Thus, in our case, the additional energy generated by the anisotropic character of the wetting liquid as well as of the surface (to be wetted) is quite significant, which may explain the un-wetting of the LC in the laser exposed areas. We must realize that those estimations are based on the relatively “soft” deformations and more detailed theoretical analyses must be conducted to take into account the energy of disclinations. Intensive experimental and theoretical work is under way to better understand and describe this phenomenon. 6.

Conclusion

Preliminary study shows that the propagation (due to capillarity) and wetting condition are dramatically changed as a result of the internal surfaces of substrates, composing the LC cell, #107595 - $15.00 USD Received 19 Feb 2009; revised 11 Apr 2009; accepted 19 Apr 2009; published 26 May 2009

(C) 2009 OSA


being exposed to circularly polarized light. The mechanism of this change is not yet clear and the proposed model has significant limitations. However, we believe that this phenomenon should be related to the orientation of molecules of the solid surface, which in turn, forces the corresponding reorientation of the LC director. The additional energy, related to the director’s deformation, changes the balance of forces and the corresponding capillary action. We believe thus that this additional source of energy must be taken into account to correctly describe the wetting process and that this phenomenon may have significant fundamental and practical importance. Acknowledgments

We would like to thank Dr. E. Brasselet for very valuable discussions and comments. We would like also to acknowledge the financial support of NSERC Canada.


(C) 2009 OSA
