Non-Contact Optical Control of Colloidal Architectures Using Molecular Monolayers

by user

Category: Documents





Non-Contact Optical Control of Colloidal Architectures Using Molecular Monolayers
Non-Contact Optical Control of Colloidal
Architectures Using Molecular Monolayers
Angel Martinez
Many efforts have been devoted to achieve optical non-contact manipulation of
colloidal particles and architectures at the micro- and nano-scales. In this paper, one
more approach is introduced using a monolayer of light-sensitive DMR molecules
bonded to glass substrates. By controlling tightly-focused laser beams, at relatively low
powers, one can induce cis-trans photoisomerization to manipulate the orientation of the
DMR molecules and, through this, locally manipulate the liquid crystal director. It is
demonstrated here that this kind of manipulation allows one to exert forces and torques
on micro- and nano-scale colloidal particles and architectures consisting of a wide variety
of materials to control translational and orientational position.
Manipulation of colloidal particles has been achieved using many different
techniques. One common and widely known method is the use of optical tweezers1-2.
However, there are several important limitations. First, optical tweezers typically require
high power outputs. If trapping several particles is required, the power needed can
quickly increase. Second, because the method is based on refraction, the particle being
trapped must be transparent to the laser’s wavelength in addition to having a greater
index of refraction than its surrounding medium. Third, the trapping volume of the
focused laser is typically subject to the diffraction limit. Fourth, manipulation of metal
partricles using this approach is typically impossible.
These limitations have been overcome with other approaches. The need for
trapping large amounts of small scale particles has led to research in using multiple
beams to create interference patterns.3 But this method would only provide means to
achieve highly ordered patterns, leaving little possibility for more irregularly shaped
structures if they are so desired. Control of different kinds of materials has also been
achieved with the method of optoelectronic tweezers.4 However, the means to do so
involves a process of doping silicon to create photoconductive materials as well as having
to carefully construct multi-layered structures using this material. The limitation of the
diffraction limit has been overcome using metals nanostructured on substrates.5 But
creating such nanostructures can be difficult and expensive. In this work, I describe
another method with the potential to overcome all of these difficulties by utilizing surface
monolayers of light-sensitive azobenzene molecules, which differs from research already
done where azobenzene molecules are already applied to the liquid crystal (LC) systems.6
In my research project, I achieve optical manipulation of nano- and micro-sized
colloidal particles and architectures through the use of molecular monolayers of an
azobenzene variety called methyl red derivative (DMR) bonded to glass substrates to
control the director field of a liquid crystal medium. The chemical diagram of the DMR
molecule is shown in Figure 1a. The azobenzene molecule within DMR transforms
between the cis and trans states upon being irradiated with an appropriate wavelength of
light. In this case, around 488 nm does this effectively. The geometry of this light
sensitive cluster in the two states is depicted in Figure 1b. The DMR monolayer becomes
oriented in plane at 90 degrees to the polarization direction of the incident light. In turn,
the liquid crystal molecules (5CB shown in Figure 1) align themselves along with the
DMR molecules in the monolayer. This manipulation of the monolayer can be done
using powers on the order of 10 to 100 times less than that used by methods such as
optical tweezers.
Having multiple alignment domains in the LC, we can use the barriers between
them to exert forces on colloidal architectures to achieve control of translational position.
In addition, by manipulating the orientation of the director field within the domains, we
can exert torques on these architectures to achieve orientational control. So, in essence,
by controlling the position and polarization of tightly focused laser beams, we can gain
an indirect translational and orientational control of nano- and micro-sized colloidal
Figure 1: a) Chemical diagram of a DMR molecule. The cluster of atoms at the bottom bond to the
glass surface. b) The light sensitive azobenzene cluster of the DMR molecule. Exposure to specific
visible light causes cis-trans photoisomerization. c) Chemical diagram of 5 CB LC.
Equipment and Experimental Procedures
My work over this summer was based around microscopy techniques provided by
a fluorescent confocal microscope from Olympus. The microscope was equipped with a
variety of objectives that provided a variety of magnifications: 2x, 10x, 60x, and 100x.
My project used mainly the 100x objective. The microscope also provided a variety of
methods of observing a sample. I was able to observe the sample by choosing from a
couple of transmission modes and a fluorescent mode. In addition, the microscope setup
included a variety of lasers, each with a different wavelength. The wavelengths I was
able to use were 458 nm, 488 nm, 515 nm, 543 nm, and 633 nm.
The preparation of the substrates involved two main phases: glass cleaning and
monolayer application. The glass-cleaning phase begins with scrubbing the glass surface
using a soft brush and detergent to wash away oils and remove any large particulate
matter that may be stuck on the surface. The glass is then rinsed and dried using
compressed nitrogen gas to blow across the surface. Once dry, the glass is then sonicated
in Acetone, Isopropanol and DI water, one by one and in that order for 5 minutes each.
The sonication helps remove any microscopic junk attached to the glass surface. The
glass is dried again with compressed nitrogen gas in preparation for the plasma cleaner.
The glass is placed between electrodes and under a bell jar. The bell jar is then put under
a vacuum and a high voltage is put across the electrodes to create a plasma environment
for the final step of the cleaning phase. Once the glass substrates are cleaned, they are
submerged in DMR solution at 45 degrees Celsius for about 90 minutes to allow the
DMR molecules to bond to the glass surface. Then, they are rinsed with toluene to wash
away any DMR molecules that have not bonded to the glass surface. Finally, they are
heated in air to about 115 degrees Celsius for about 2 hours to cure.
Once the monolayer has been applied to the substrates, the LC sample cells can be
constructed. Some cells were constructed with UV curable glue, which contained
spacers, sandwiched in between the glass substrates at several points throughout the cell
area. Other cells were constructed simply with epoxy (also sandwiched between)
containing no spacers. An example of this construction is shown in Figure 2. The
spacers helped keep the gap between the substrates even and at a desired thickness. In
my particular case, the cells with spacers contained 30 m spacers. The cells without
spacers (with only epoxy) had gaps with higher variations. Measurements of these gaps
showed that the spacing ranged from 3 m to about 20 m. The results in this paper
were constructed with epoxy, so the cell thickness was usually very thin. After the cell
has been constructed and the glue has set, the gap between the substrates is filled with the
LC sample by touching a drop to an open edge and allowing it to be pushed into the
empty gap by capillary action, as shown in Figure 2. When the cell has been sufficiently
filled, the edges are sealed with epoxy to physically isolate the inside of the cell from the
rest of the world and to keep the LC from escaping.
LC Droplet
Glass Substrate
DMR Monolayer
Figure 2: A small section of a constructed cell. The LC droplet is touched to the gap as shown and is
pushed between the substrates by capillary action.
The constructed LC sample cell is mounted on the microscope stage for
observation. The laser scanning transmission mode was most useful in my research. It
allowed me to observe the sample in real time while simultaneously scanning an area of
the sample with the desired laser. I used two of the available lasers to conduct my
research. The 488 nm laser proved to be the most efficient of the available lasers for
realigning the DMR molecules, while the 633 nm laser did not noticeably affect it. So,
when realigning was desired, I turned on the 488 nm laser, but when I wished to observe
the sample without affecting the monolayers, I switched to the 633 nm laser. Areas
scanned with the 488 nm undergo a visible change in real-time as the alignment of the
liquid crystal is altered. The square areas of different colors in Figures 3-7 are areas
scanned with this laser.
The micro-particle structures shown in Figures 3-7 were constructed manually
(with the exception of some of the chains, which can usually be found naturally
assembling due to elasticity-mediated forces throughout the sample) using a laser
trapping system. Once assembled, the laser trap was turned off and the rest of the particle
and architecture manipulation was done using the laser scanning mode (LSM) provided
by the microscope.
The initial motivation for this research was an investigation of the elastic forces at
the barriers between different alignment domains. Domains can be created by scanning a
particular area of the substrates with the 488 nm laser with a particular polarization, upon
which the DMR molecules and LC molecules would align themselves perpendicularly.
The laser intensities used in this research were kept at about 3-5% of the maximum level,
which translates to an order of about 100 W at the sample.
3 μm MR bead
5 μm
Figure 3: Micron-sized particles can be pushed by advancing barriers between alignment domains as
shown in the sequence a-e above. Nano-scale rods, however, are hardly pushed, but do respond to their
changing alignment environment by aligning themselves accordingly. The bright area is aligned at
about 45 degrees as is revealed by the rod’s alignment.
By advancing the scanning area, the barrier advances with it. This advancing
barrier is seen as the leading edge of the bright rectangular area in Figure 3, which is a
domain aligned at about 45 degrees to the surrounding dark purple area. Also visible
nearby is a nano-rod, about 3-4 m long and 200 nm in diameter, and a 3 m melamine
resin (MR) sphere. As the barrier advances toward the colloids, the MR sphere is pushed
and stays ahead of the barrier due to the elastic forces. The nano-rod, however, is hardly
pushed at all (Figure 3). Presumably, this is due to its small thickness. However, its long
aspect ratio makes it sensitive to changes in the director, which exerts elastic torques on
the rod. This is evident by the rotation of the rod as it is enveloped by the advancing
domain. Figure 4C demonstrates this more clearly for a lone rod. All images shown in
Figures 3-7 were taken under illumination by a bright lamp. After a few seconds of
exposure the lamp begins to noticeably change the director in the domain. As a result,
the orientations of the nano-rods and particle structures have slightly deviated. This is
also the reason why the rectangular area in Figure 3 has a fade effect trailing behind it as
it is advanced toward the particle.
When MR resin spheres cluster together, they can form architectures without the
spherical symmetry that are subject to elastic torques upon the reorientation of the LC
director. The simplest kind of structures involving more than one colloidal sphere is a
chain of spheres. Because of the potential for a long axis along one direction (a long
lever arm), chains can be highly responsive to changes in the director. The LC molecules
like to align parallel to the MR sphere surface, which induces the director distortion of a
quadrupolar character. As a result, the chains align themselves at about 30 degrees to the
director field as is depicted in Figure 4A. In Figure 4B, you can see an actual micronsized chain of MR spheres surrounded by a domain with alignment shown. The chain’s
30 degree tilt off the director is seen. As the director field in the domain is reoriented (by
adjusting the laser polarization to reorient the DMR monolayer), and elastic torque is
exerted and the chain rotates.
Other structures were investigated. Figures 5-7 show effects on diamond and
triangular clusters.
It has been demonstrated that using molecular monolayers on substrates allows
for precise control of translational and orientational positioning of colloidal architectures.
In addition, this approach provides the ability to overcome certain limitations introduced
by other methods of micro- and nano-scale particle manipulation, such as those faced by
optical tweezers. Because the method of manipulating monolayers using polarized light
does not depend on the refractive index of any particle or structure being manipulated, it
has the potential to provide control of a wide variety of materials and also of complex
composite materials. Furthermore, the powers necessary to reorient the DMR molecules
within the monolayer are minute compared to the powers of other methods and the
manipulation of multiple particles and structures would not require any additional
amounts of laser power. Also, since the monolayers are composed of individual
molecules, there is also the potential to overcome the diffraction limit.
At this point, only some of the qualitative aspects have been investigated using
molecular monolayers on substrates. All of the results in this paper only show effects of
manipulating the monolayers on both substrates in the same way to create a uniform
alignment along the z-direction. Manipulating each of the monolayers on the two
substrates can prove more interesting and is a natural step in gaining a more complete
understanding of this approach. Obtaining the best results for a separate manipulation of
the monolayers on each substrate would require larger gaps between the substrates to
help reduce the effects on one surface while focusing the laser on the other. So, using the
30 m spacers for this case would be ideal. Moreover, efforts in making quantitative
measurements of the forces and torques exerted will be made. Also, in addition to
manipulating monolayers on the surfaces of the substrates, investigations of manipulating
monolayers on the surfaces of the micro-particles themselves will be performed.
Figure 4A: To the right you see a MR bead chain
and a nano-rod as they should appear within a
nematic LC with director alignment as shown. The
LC molecules align themselves parallel to the surface
of the MR beads which results in chains tilting 30
degrees off of the director as shown. The nano-rods
align along the director.
LC director direction
Nano-rod alignment
Figure 4B: Chain of 5 3m MR beads. a) The LC in a square surrounding the chain is aligned along the
direction shown. b) The same square area has been realigned to about 90 degrees with respect to direction
in image “a” along the direction shown. You can see the chains aligning at about 30 degrees to the director
as is expected.
Figure 4C: Silver nano-rod. a) The LC in a square surrounding the nano-rod is aligned along the direction
shown. b) The same square area has been realigned to about 90 degrees with respect to direction in image
“a” along the direction shown. The rods deviate slightly off of the direction due to the incident light from
the lamp when taking a picture. The light begins to alter the director in the domain after several seconds of
Figure 5: Diamond cluster. a) The LC in a square surrounding the diamond is aligned along the direction
shown. b) The same square has been realigned to about 90 degrees of direction shown in image “a”. It
appears that at least two of the sides like to align themselves at about 30 degrees to the director as a chain
Figure 6: Triangular cluster. a) The LC in a square surrounding the triangle is aligned along the direction
shown. b) The same square has been realigned to about 90 degrees of direction shown in image “a”.
Similar to the diamond, some sides align themselves along a direction about 30 degrees to the director.
Figure 7: Simultaneous rotation of a chain and a diamond. a) The LC in a square surrounding the chain
and diamond cluster is aligned along the direction shown. b) The same square has been realigned to about
90 degrees of direction shown in image “a”. It seems in this case, that the large diamond has found an
equilibrium point between those of the individual sides. None of the sides align quite like the lone chain.
1. Ashkin, A., Dziedzic, J. M., Bjorkholm, J. E., and Chu, S., “Observation of a
single-beam gradient force optical trap for dielectric particles,” Optics Letters
11(5) 288-290 (1986)
2. Ashkin, A., “Optical trapping and manipulation of neutral particles using lasers,”
Proceedings of the National Academy of Sciences of the United States of America
94(10) 4853-4860 (1997)
3. Rohner, J., “Light structuring for massively parallel optical trapping,” PhD
dissertation, EPFL, Lausanne, 2007.
4. Chiou, P., “Massively parallel optical manipulation of single cells, mico- and
nano-particles on optoelectronic devices,” PhD dissertation, University of
California, Berkeley, 2005.
5. Grigorenko, A. N., Roberts, N. W., Dickinson, M. R., Zhang, Y., “Nanometric
optical tweezers based on nanostructured substrates,” Nature Photonics 2 365-370
6. Yamamoto, T., Tabe, Y., Yokoyama, H., “Photochemical transformation of
topological defects formed around colloidal droplets dispersed in azobenzenecontaining liquid crystals,” Colloids and Surfaces A: Physicochemical and
Engineering Aspects 334 155-159 (2009)
Fly UP