連續波及飛秒脈衝雷射應用於微米至奈米大小球狀物質的光學捕捉

(1)

i

國立交通大學

應用化學系

碩士論文

連續波及飛秒脈衝雷射應用於微米至奈米大小球狀

物質的光學捕捉

Optical Trapping Dynamics of Micron- to Nanometer-Sized Spherical

Target Materials by Continuous Wave- and Femtosecond Pulse-Mode

Laser Beams

研究生：江威逸

指導教授：增原宏教授

(2)

ii

連續波及飛秒脈衝雷射應用於微米至奈米大小球狀物質的光學捕

捉

Optical Trapping Dynamics of Micron- to Nanometer-Sized Spherical

Target Materials by Continuous Wave- and Femtosecond Pulse-Mode

Laser Beams

研究生：江威逸 Student: Wei-Yi CHIANG

指導教授：增原宏教授 Advisor: Prof. Dr. Hiroshi MASUHARA

國立交通大學

應用化學系

碩士論文

A Thesis

Submitted to M. S. Program Department of Applied Chemistry

College of Science National Chiao Tung University in Partial Fulfillment of the Requirements

for the Degree of Master

in

Applied Chemistry July 2012

Hsinchu, Taiwan, Republic of China

(3)

iii

Published paper

Anwar Usman, Wei-Yi Chiang, Hiroshi Masuhara, Optical trapping and polarization-controlled scattering of dielectric spherical nanoparticles by femtosecond laser pulses, J. Photochem. Photobiol. A: Chemistry 234 (2012) 83-90.

Published proceeding

Anwar Usman, Wei-Yi Chiang, Hiroshi Masuhara, Laser Trapping-Induced Reconfiguration of Individual Smectic Liquid Crystal Micro-Droplet Showing Size-Dependent Dynamics, Proc. SPIE 8274 (2012) 82740L1-8.

Conference

1. “Orientation Modulation in Individually Trapped Liquid Crystal Droplets by Tightly Focused Laser Beam”,

Wei-Yi Chiang, Anwar Usman, Takayuki Uwada, Hiroshi Masuhara,

2011 Spring Symposium of Photochemistry Association in Taiwan, 15 - 16 February, 2011, Hsinchu, Taiwan

2. “Orientation Modulation in Individually Trapped Liquid Crystal Droplets by Tightly Focused Laser Beam”,

Wei-Yi Chiang, Anwar Usman, Takayuki Uwada, Hiroshi Masuhara,

The 3rd Workshop on Laser Bio/Nano Science, 4th March, 2011, Hsinchu, Taiwan 3. “Size-dependent phase modulation of individually trapped liquid crystal micro-droplets”,

Wei-Yi Chiang, Anwar Usman, Hiroshi Masuhara,

2011 年光化学討論会, 6-8 September, 2011, Miyasaki, Japan

4. “Size-dependent phase modulation of individually trapped liquid crystal micro-droplets”, Wei-Yi Chiang, Anwar Usman, Hiroshi Masuhara,

Japan-Taiwan joint workshop: Future Perspective on NanoBio Science Pioneered by Light, 4th Octorber, 2011, Hsinchu, Taiwan

5. “Femtosecond Laser Pulse-Induced Optical Trapping and Polarization-Controlled Off-Axis Scattering of Dielectric Spherical Nanoparticles”,

Wei-Yi Chiang, Anwar Usman, Hiroshi Masuhara,

2012 Spring Symposium of Photochemistry Association in Taiwan, 9-10 February, 2012, Kaohsiung, Taiwan

6. “Concentration Dependence of Polarization-Controlled Scattering of Polystyrene Nanoparticles by Femtosecond Laser Pulses”,

(4)

iv

応用物理学関係連合講演会, 15-18 March, 2012, Tokyo, Japan

7. “Axial Field-Controlled Direction of Scattered Dielectric Spherical Nanoparticles in Optical Trapping”,

Anwar Usman, Wei-Yi Chiang, Hiroshi Masuhara,

2012 Spring Symposium of Photochemistry Association in Taiwan, 9-10 February, 2012, Kaohsiung, Taiwan

8. “Dynamic Optical Force Exerting on Dielectric Nanoparticles by Femtosecond Laser Pulses, Leading to Polarization-Controlled Off-Axis Scattering”,

応用物理学関係連合講演会, 15-18 March, 2012, Tokyo, Japan

9. “Femtosecond Laser-Induced Trapping and Off-Axis Ejection of Dielectric Spherical Nanoparticles”,

The 1st Workshop on Laser Trapping with Laser BioNano Science, 4-6 April, 2012, Hsinchu, Taiwan

10. “Laser Trapping-Induced Reconfiguration of Individual Smectic Liquid Crystal Micro-Droplet Showing Size-Dependent Dynamics”,

Anwar Usman, Wei-Yi Chiang, Takayuki Uwada, Hiroshi Masuhara, SPIE photonic west, 22-28 January, 2012, San Francisco, USA

(5)

v

連續波及飛秒脈衝雷射應用於微米至奈米大小球狀物質的光學捕捉

研究生：江威逸指導教授：增原宏教授

國立交通大學應用化學系碩士班

摘要

微奈米結構有機物或聚合物的光學捕捉提供光學調控分子排列以及聚焦雷射產生動態變化的資訊。這篇論文含蓋了兩個方向，分別是藉由聚焦連續波雷射光學捕捉引發微米大小液晶液珠內的分子排列以及飛秒雷射照射在 50 奈米大小的聚苯乙烯球所產生的動態變化。我們探討在聚焦連續波雷射照射下液晶液珠內的分子重排而重排發生自內部並遍及整個液珠，且相轉變的門檻能量隨液珠大小而改變。我們認為當焦點體積內的光學重組克服液珠的介面錨定效應便會產生重排。藉由探討電場對液晶液珠的作用能以及引入液珠介面錨定效應的概念，我們成功建立了可解釋隨液珠大小而產生不同變化的模型。飛秒脈衝雷射照射在 50 奈米大小聚苯乙烯球的動態變化呈現出極化方向控制的兩個相反方向並交互變換的散射且與雷射極化方向相互垂直。由於光電場和極化介電球誘導偶極矩之間的相互作用，我們重新討論作用在介電奈米球體上的勞倫茲力用以解釋新現象的發生。我們考慮高數值孔徑物鏡產生的軸向及橫向電場導出軸向電場是散射力及時間變化力橫向組成的原因並且控制雷利粒子散射的方向。透過這些發現，我們提供了聚焦超快雷射脈衝雷利粒子動態光學捕捉很重要的資訊。

(6)

vi

Optical Trapping Dynamics of Micron- to Nanometer-Sized Spherical Target

Materials by Continuous Wave- and Femtosecond Pulse-Mode Laser Beams

Student: Wei-Yi CHIANG Advisor: Prof. Dr. Hiroshi MASUHARA

M. S. Program, Department of Applied Chemistry

National Chiao Tung University

Abstract

Optical trapping of organic or polymeric micro- and nano-structures provides

information on how their molecular configuration is optically manipulated or how their

dynamical motions are controlled by the focused laser beam. This thesis covers both ideas of

optical trapping-induced molecular configuration in micro-sized individual

4’-n-pentyl-4-cyanobiphenyl liquid crystalline droplets by tightly focused continues wave

laser beam and dynamical motions of 50-nm-sized polystyrene beads by femtosecond pulsed

laser beam.

We explore the reconfiguration in liquid crystal droplet under the irradiation of

tightly focused cw laser beam. We show that reconfiguration took place throughout the inside

of droplet, and the threshold power of the phase transition depends on droplet size. We

propose that the droplet reconfiguration takes place when optical reorientation at the focal

volume overcomes the droplet-liquid interfacial anchoring effect. By considering interaction

energy of liquid crystal droplet by electric field and introducing liquid crystal droplet

(7)

vii dynamic change depending on size.

The dynamic motions of the 50 nm-sized polystyrene beads impinged by

femtosecond pulsed laser beams show the polarization-controlled scattering in two opposite

directions, in alternating manner, perpendicular to the laser polarization. To understand this

new phenomenon we reconsider Lorentz force acting on a dielectric spherical nanoparticle

due to interactions between light electric field and induced dipole moment of the polarizable

dielectric sphere. We take axial and lateral electric fields into account produced by high

numerical aperture objective lens. The axial electric field is responsible for lateral components

of the scattering and temporal forces, and hence, controls the scattering directions of the

Rayleigh particles. These findings provide important information about the dynamic optical

(8)

viii

Acknowledgment

Two years ago when I joined this laboratory I was worry with the new environment

and unfamiliar people. Specially, all of the professors in this laboratory come from other

countries, so I need to communicate in English, which is not familiar language for me, with

them. Fortunately, even though my poor English, all of the professors have nice patience to

listen and teach me.

Now, upon completing this thesis, I would like to express my gratitude to whom

that have helped me in this two years. Particularly, I am deeply indebted to my supervisors

Prof. Hiroshi Masuhara and Dr. Anwar Usman who supervise, advice, and encourage me to

perform my research works and to write this thesis.

I also want to express my appreciation to Prof. Atsushi Miura, Prof. Takayuki

Uwada, and Dr. Kenichi Yuyama who gave me helps, supports, and suggestions during my

research works.

Additionally, I am obliged to acknowledge to my reviewers Prof. Hiroshi Masuhara,

Prof. Atsushi Miura, Prof. Yen-Ju Cheng and Prof. Yoshiaki Teranishi for their kind evaluation

on my thesis and permission to get the Master degree.

In this occasion, I also want to thank my seniors, Ping-Yu Hee, Jing-Ru Tu,

Chung-Wei Huang and Tsung-Han Liu; classmates, Shun-Fa Wang, Ching-Hsu Tseng,

(9)

ix

encouraging, and help during these two years of laboratory life.

Besides, I want to thank the Department of Applied Chemistry of National Chiao

(10)

x

List of Figures

Figure 1.1 The optical forces acting on particle performed by momentum transfer of photons. 3 Figure 2.1 A schematic diagram of the experimental setup ... 13 Figure 2.2 Sequences of time evolution of POM images of an individual smectic 5CB droplet

optically trapped by a linearly polarized trapping beam at; (a) 130 mW (or 60 MW/cm2); (b) 850 mW (or 360 MW/cm2), and by a circularly polarized trapping beam at; (c) 130 mW (or 60 MW/cm2), (d) 460 mW (or 190 MW/cm2), and (e) 940 mW (or 390 MW/cm2) ... 17 Figure 2.3 Plot of threshold laser power density to induce the phase transition as a function of

droplet radius ... 19 Figure 2.4 (a) The total energy as a function of the radius of the droplet at different level of

laser power. (b) Plot of the predicted dependence of threshold laser power density to induce a negative local minimum, which is related to the reconfiguration, as a function of droplet radius relative to the beam waist, based on Eqn. (2.2) (see text for details). ... 28 Figure 3.1 Schematic diagram of the experimental set-up ... 37 Figure 3.2 A combination of two halves of two different image frames showing optical

trapping and scattered polystyrene nanoparticles by femtosecond-pulse modes and showing their optical trapping by cw Ti:sapphire laser beams (λ= 800 nm) ... 41 Figure 3.3 (a) Time-dependent line profile intensity under femtosecond-pulse (black line) and

cw mode (gray line). (b) The typical time-dependent event when bright locus of scattered polymer beads from the focus spot is along the two alternating left and right directions within the observation window of 120 s. Experimental information for data shown in the figures (a) and (b) are related to those in Figure 3.2a... 43 Figure 3.4 Plots of calculated gradient, scattering, and temporal forces acting on a

polystyrene nanoparticle locating in the focal plane (z = 0) as a function of either the normalized lateral position of x ω⁄ vor normalized time t τ₀ ⁄ (a) Lateral component of the gradient force, Fgrad, parallel to Ex, at the peak of a pulse duration (t τ = 0⁄ ) (b) Axial (black) and lateral (red) components of the scattering force, Fscatt, at the peak of a pulse duration (t τ = 0⁄ ) (c) Axial (black) and lateral (red) components of the temporal force, Ftemp. (d) A schematic illustration of the gradient, scattering, and temporal forces ... 56

(13)

1

CHAPTER 1

(14)

2

The interactions of light with matter result in many fascinating phenomena. Over a

long history of investigation of such interactions, the researches nowadays have arrived at

nanometer scale of target material, a dimension much smaller than the wavelength of utilized

light. One of the fascinating ideas in optics is the use of light to trap atoms and small particles

[1,2]. This so called optical tweezers or optical trapping technique has already proven to be an

essential tool in physics, biology, and chemistry [3,4]. Among the interesting applications of

this technique in fabrication and characterization at the nanoscale [5], in such a thriving field

of research over the past 40 years the laser trapping-induced crystallization of amino acids has

generated attentions as a model for three-dimensional molecular rearrangement by focused

laser beam [6-10]. The basic concept of the optical trapping is the generation of gradient force,

as illustrated in Figure 1.1 that the incident light is refracted through the particle, leading to

the change of momentum of the photons. According to the conservation of momentum, a

(15)

3

Figure 1.1 The optical forces acting on particle performed by momentum transfer of photons.

This thesis attempts to understand molecular reconfiguration inside microparticles

trapped by focused laser beam as well as to explore the novel trapping phenomena of

nanoparticles by ultrafast laser pulses. For the earlier attempt, liquid crystal droplets with soft

and reconfigurable molecular organization properties were used as the target material

optically trapped by continuous wave-mode laser beam. Though the researches on the liquid

crystal droplets by focused beam have been started a few years ago, most researches

employed the laser power and optical condition below those necessary to induce a laser

trapping crystallization similarly to the amino acids [12-15]. Thus, in this work, the trapping

behavior was studied in a wide range of laser power tightly focused by a high numerical

(16)

4

the target material, and femtosecond pulse-mode laser beam was used as the trapping beam

substituting the traditional continuous wave laser. We note that in the beginning of ultrashort

pulsed laser trapping was started, the laser pulse trapping was believed to be as just effective

as cw laser trapping of microparticles in particular [16]. Indeed, when picosecond pulse-mode

laser beam was finally shown to be able to trap less-than-5-nm sized CdTe quantum dots in

2007, optical trapping experiments with laser pulses started to attract more and more attention

with many questions still to explore. More recently, research works on the ultrashort laser

pulse-induced trapping behavior of nanoparticles provide several excellent phenomena, for

instance, the control of crystalline properties and dimensions of nanodeposits of 25-nm-sized

CdS particles [17] and the split of trapping site of gold NPs into two equivalent positions

shifted from the focal center [18]. These findings highlight the advantages of pulse- over

cw-mode laser beam in trapping ability and control on Rayleigh particles.

However, such fundamentally different behaviors of the three-dimensional optical

trapping with the laser pulses to the cw laser also require theoretical elucidations, because the

new phenomena are beyond the concept of conventional optical trapping. Thus, for a

quantitative explanation we should rely upon the scattering theory of electromagnetic

radiation, reexamining gradient, scattering, and temporal pulse-radiation forces, which are

respectively related to spatial Lorentz force [19,20], momentum transfer from light to the

(17)

5

cross section value of the nanoparticles was obtained based on Mie theory [21].

In Chapter 2 we describe the polarization- and droplet size-dependence of laser

trapping-induced reconfiguration in individual radial symmetric liquid crystalline

micro-droplets suspended in water. We found that the optical trapping of the droplet is indeed

followed by a molecular reconfiguration, which takes place only when laser power density is

above a definite threshold level. The relevant qualitative theoretical approximation is

described based upon a likely mechanism of the molecular reconfiguration inside the droplet

that involves local optical reorientation at focal volume when the interaction energy of liquid

crystals with dielectric anisotropy aligned by the electric field of the focused laser beam

exceeds Frank distortional energy. With this proposed mechanism, we show that the predicted

dependence of the threshold power on the droplet size is in qualitative agreement with the

experimental observation.

Chapter 3 discusses femtosecond laser pulse trapping behavior of polystyrene

nanoparticles. The main focus of this chapter is laser pulse trapping behavior in comparison

with the cw mode. We found that the laser pulses can trap a larger number of nanoparticles. In

addition, the laser pulses induce nanoparticle flows out of the focal spot in two opposite

directions, in an alternating manner, controlled by the laser polarization. We evaluate both

gradient, scattering, and temporal forces by adopting Lorentz force of fundamental Gaussian

(18)

6

nanoparticles obtained based on Mie theory. We demonstrate that the lateral and axial electric

fields produced by the high numerical aperture objective lens are responsible for the present

novel phenomenon.

In Chapter 4 we conclude some important findings of the research works in this

thesis, and provide the future outlooks of this research works, which already start to open new

vistas for controlling the molecular reordering and reconfiguration of micro-droplets by the

cw laser trapping, controlling dynamical motion of nanoparticle assembly by ultrafast laser

pulses. The latter finding, in particular, is considered to be applicable for separation and

sorting of nanoparticles with either different polarizabilities or scattering cross sections. Thus,

(19)

7

References

[1] A. Ashkin, Phys. Rev. Lett. 24 (1970) 156..

[2] A. Ashkin, Proc. Natl. Acad. Sci. USA 94 (1997) 4853.

[3] K.C. Neuman, S.M. Block, Rev. Sci. Instrum. 75 (2004) 2787. [4] K. Dholakia, P. Reece, M. Gu., Chem. Soc. Rev. 37 (2008) 42. [5] E. Ozbay., Science 311 (2006), 189.

[6] T.Sugiyama, T. Adachi, H. Masuhara, Chem. Lett. 36 (2007) 1480. [7] Y. Tsuboi, T. Shoji, N. Kitamura, J. Phys. Chem. C 114 (2010) 5589. [8] T. Sugiyama, T. Adachi, H. Masuhara, Chem. Lett. 38 (2009) 482.

[9] T. Rungsimanon, K. Yuyama, T. Sugiyama, H. Masuhara, N. Tohnai, M. Miyata, J. Phys. Chem. Lett. 1 (2010) 599.

[10] T. Rungsimanon, K. Yuyama, T. Sugiyama, H. Masuhara, Cryst. Growth Des. 10 (2010) 4686.

[11] J.E. Molloy, M.J. Padgett, Contemp. Phys. 43 4 (2002) 241

[12] E. Brasselet, S. J. Joudkazis, Nonlinear Opt. Phys. Mater. 18 (2009) 167. [13] N. Murazawa, S. Joudkazis, H. Misawa, Eur. Phys. J. E 20 (2006) 435. [14] N. Murazawa, S. Joudkazis, S. Matsuo, H. Misawa, Small 1 (2005) 656. [15] N. Murazawa, S. Joudkazis, H. Misawa, Opt. Express 14 (2006) 2481.

[16] B. Agate, C.T.A. Brown, W. Sibbett, K. Dholakia, Opt. Express 12 (2004) 3011. [17] M. Sanz, R. de Nalda, J.F. Marco, J.G. Izquierdo, L. Banares, M. Castillejo, J. Phys.

Chem. C 114 (2010) 4864

[18] Y. Jiang, T. Narushima, H. Okamoto, Nat. Phys. 6 (2010) 1005 [19] P. Bartlett, S. Henderson, J. Phys.: Condens. Matter 14 (2002) 7757 [20] Y. Harada, T. Asakura, Opt. Commun. 124 (1996) 529

[21] C.F. Bohren, D.R. Huffman, Absorption and Scattering of Light by Small Particles, Wiley, New York, 1983.

(20)

8

CHAPTER 2 Polarization and Droplet Size Dependences of Laser

Trapping-Induced Reconfiguration in Individual Radial

(21)

9

1. Introduction

Optical trapping technique utilizing a tightly focused continuous-wave (cw) laser

beam allows one to manipulate micron- or a few tens of nm-sized objects with a high

precision [1,2]. Therefore, this technique becomes indispensable and has been widely

employed to control the deposition of dielectric particles [3,4], as well as to induce optical

manipulation on single cells [5], polymers [6,7], micelles [8], J-aggregates [9], and amino

acids [10,11]. In particular, the optical trapping of glycine, the simplest amino acid, dissolved

in water leads to nucleation and growth of its single crystal [10,12] with its polymorphs being

controllable by the laser power [13,14]. Such a three-dimensional molecular alignment

accurately driven by a focused light is of interest for both fundamental sciences and

applications. This finding also implies migration of molecules or clusters from the

surrounding area to the focal spot as well as cooperative molecular realignment, and such

effects have been demonstrated for gold nanoparticles and liquid crystal (LC) thin film by the

focused laser beam [15-17].

The laser trapping-induced optical realignments have also been revealed in

spherical LC droplets with various kinds of self-organized configurations depending on

droplet-liquid boundary conditions in the medium they were dispersed [18-24]. Typically, the

laser trapping beam can induce reconfiguration in radial nematic LC droplets, leading to

(22)

10

transfer [20,23].Similar phenomena were also observed for optically trapped bipolar nematic

[18,19] or cholesteric LC droplets [21,22]. In the latter case, in particular, the rotation is

accompanied by a distinct vertical motion of the droplet in the optical trap [21,22]. By

contrast, at the similar level of laser power, molecular reconfiguration inside smectic LC

droplets has never been observed, although the laser trapping beams can induce axial and

lateral repositioning leading to wobbing rotation of the trapped droplets when the laser beam

is circularly polarized [25]. Lamellar organization of LC molecules inside the droplets has

been considered to induce large elastic rigidity, preventing the light-induced molecular

reconfiguration inside the droplets [24,26]. Measuring rotation frequency of the optically

trapped droplet and polarization change of the laser beam passed through the micro-sphere

allow one to determine the viscosity of surrounding media precisely [27,28].

Recently, we have reported laser trapping behavior of radial symmetric

4'-pentyl-4-cyanobiphenyl (5CB) liquid crystalline micro-droplets dispersed in heavy water

(D2O) under highly focused near-infrared cw laser beam [29]. In our experiments, we found

that the optical trapping of the radial symmetric droplet is indeed followed by a molecular

reconfiguration, which takes place only when laser power density is above a definite threshold

level. Here, we report in some detail a qualitative understanding of the experimentally

observed reconfiguration by considering that the threshold and the dynamics of the

(23)

11

likely mechanism of the molecular reconfiguration inside the smectic LC droplet involves

local optical reorientation at focal volume when the interaction energy of LCs with dielectric

anisotropy aligned by the electric field of the focused laser beam exceeds Frank distortional

energy, followed by reconfiguration inside the confined droplet. With this mechanism, we

show that the predicted dependence of the threshold power on the droplet size is in qualitative

(24)

12

2. Experimental

2.1 Optical setup

The experimental setup is shown in Figure 2.1. A continuous wave (cw) 1064 nm

Nd:YVO4 laser beam (Spectra Physics; J20I-BL-106C) was focused into the sample cell, held

on the stage of an inverted microscope (Olympus IX71). The beam was focused by an

UPlanApo oil immersing objective lens (magnification 100 times and numerical aperture

(N.A.) 1.35). After passing through the power adjustment, the laser beam was expanded and

collimated to about 7 mm in diameter and the polarization state was controlled by a half-wave

plate. After the objective lens, the laser power was operated in the range of 0.1-1.0 W, and the beam waist for 1064 nm was calculated to be 0.39 μm. The liquid crystal droplet reoriented by

the highly focused laser beam was monitored conventionally by polarization optical

microscopy (POM), which is composed by two orthogonal polarizers. From the top of the

microscopy, visible probe light from a halogen lamp (λ=400-750 nm) was passed through a

pair of polarizers and condenser lens, N.A. is 0.55, sandwiching the sample. The coupled

charge-coupled device (CCD) camera (JAI; CV-A55IR E), running at 30 interlaced frames per

second, was used to detect the transmittance probe light. The scattering light from near

(25)

13

Figure 2.1 A schematic diagram of the experimental setup; BS = beam splitter, λ/2=half-wave plate, LPF = low pass filter with transmission edge at 900 nm. Inset: the cell containing an individually trapped smectic LC droplet by focused laser beam.

2.2 Sample preparation

The sample cell was prepared by two cover glasses (Matsunami; thickness

0.12-0.17 mm), and a chamber between the two cover glasses was formed by strips of

parafilms along the glass edges with height 15-25 μm. The chamber was filled by 5CB

(26)

14

and extraordinary refractive indices,

n

_o and

n

_e, are 1.54 and 1.74, respectively) suspended

in D2O without any additional surfactants. Comparing with H2O, in this kind of condition, we

can avoid temperature elevation induced by 1064 nm laser beam that has been reported to be

2.6 ºC/W and 24 ºC/W in H2O [30,31]. Due to the interfacial tension, 5CB droplet always

(27)

15

3 Results

3.1 Reconfiguration of the LC droplet

The diameter of the micro-droplets was within submicron to several microns. Due

to their high refractive index, the droplets are highly polarizable and easily trapped by a steep

optical-field gradient. In Figure 2.2 (a-b) are shown POM images for a ~2.5-μm-sized smectic

droplet upon irradiation by linearly polarized laser beam with different laser power densities.

We found that at low laser powers the droplet is optically trapped, so that its center is located

at the focal area, its POM images show no time dependence, and the intrinsic lamellar

configuration inside the droplet apparently remains intact. When the laser power was

increased to few hundreds of MW/cm2, we observed that the POM images of the optically

trapped droplet are time-dependent. The images showed clearly that the radially symmetric

configuration of the smectic droplet disappears, followed by the appearances of ring patterns

of transmitted probe light passing through the trapped droplet. Typically, the ring patterns

consisting of a small ring near the center and one or two larger concentric rings showed up

clearly on time scales of seconds to a few tens of seconds depending on the laser power. Such

time evolution of POM images indicate unambiguously that the radial symmetric droplet is

reconfigured under the high power of laser trapping beam. In addition, we observed

(28)

16

transmission images, though it is not so clearly observed for droplets with diameter less than 2 μm.

Figure 2.2 (c-e) shows the POM images for a ~3.5-μm-sized droplet trapped by

circularly polarized laser beam with different laser power densities. Notably, for the small

droplet, the circularly polarized laser trapping beam did not induce wobbing rotation. This

suggests that optical trapping of the small droplet is on-center and it is consistent with the

earlier report, in which off-center optical trapping is observed for smectic droplets larger than 4 μm in diameter independent on the laser power [25]. When we increased the laser power,

the POM images show that the droplet was also trapped steadily at the early time of

irradiation, and such a stable optical trap was followed by a wobbing rotation of the droplet

around the trapping axis. The transition from the static to rotation states takes place in a few

tens of milliseconds. In this case, once the droplet started to undergo a wobbing rotation, the

rotation continues without stopping within our observation window of 90 s. Further increasing

the laser power led a wobbing rotation to undergo temporarily. When such rotation stopped,

the appearances of ring patterns similarly to those for a droplet trapped by the linearly

(29)

17

Figure 2.2 Sequences of time evolution of POM images of an individual smectic 5CB droplet optically trapped by a linearly polarized trapping beam at; (a) 130 mW (or 60 MW/cm2); (b) 850 mW (or 360 MW/cm2), and by a circularly polarized trapping beam at; (c) 130 mW (or 60 MW/cm2), (d) 460 mW (or 190 MW/cm2), and (e) 940 mW (or 390 MW/cm2). The trapping time in second is indicated in each snapshot; while the most left and most right snapshot in each sequence is the image just before laser trapping beam is switched on and just after the beam is switched off, respectively. The red circles denote the focal spot area. The scale bar of 3 μm is applied for all images.

In addition to such polarization-dependent dynamics of the reconfiguration,

increasing the laser power above the threshold led all the dynamics to be faster. When the

laser beam was switched off, the transient ring patterns vanish immediately on the time scale

of tens of milliseconds, restoring the initial radially symmetric pattern of a radial symmetric

0.0 5 10 12 15 20 45 0.0 5 10 12 15 20 45 80 80 (a) 0.03 3 6 9.9 10 1 20 30 0.03 1 1.1 1.7 2 2. 2.3 34 0.03 5 8 10 15 2 25 30 (b) (c) (d) (e)

(30)

18

droplet. Such an immediate restoration of the initial pattern, however, is independent on the

laser power, and it is evidence for the existence of droplet-liquid interfacial anchoring effect

as well as for the absences of optical memory, hysteresis, or storage effects.

3.2 Droplet size-dependence of threshold laser power

By varying the laser power to trap an individual LC droplet, we found that there is a

clear threshold power, above which the optical trapping is followed by molecular

reconfiguration throughout the inside of the droplet [29]. To evaluate the dependence of the

threshold power on the droplet size and polarization state, we have performed and repeated

the above process for different sizes of LC droplets for both linearly and circularly polarized

beams. We found that the threshold is higher for a larger droplet. As far as the optical

reconfiguration throughout the inside of the droplet as indicated by the formation of ring

patterns is considered, we found that the threshold for circularly polarized laser beam tends to

be slightly larger than that for linearly polarized beam. The droplet size and polarization

dependences of the threshold trapping laser power to induce reconfiguration are shown in

(31)

19

Figure 2.3 Plot of threshold laser power density to induce the phase transition as a function of droplet radius. The calculated beam waist is denoted by an asterisk. Data points denoted with filled rectangular and open circular are for linearly and circularly polarized beams, respectively.

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

300

400

500

600 T

hr

es

hol

d P

ow

er

[

M

W

/c

m

2

]

Droplet Radius [µm]

*

(32)

20

4. Discussion

4.1 Laser trapping-induced reconfiguration

We should note that our observation on laser trapping properties of the radial

symmetric 5CB droplet at low laser powers are in accordance with those recently reported and

well documented in works by Murazawa et al. [25] and Brasselet et al. [24] Our important

experimental finding is that, actually, optically induced reconfiguration does exist in the

droplet under high laser powers with a sharp threshold, similarly to the cases of other LC

droplets with radial configurations [23,32,33].

To interpret the optically-induced reconfiguration in the radial nematic LC droplet,

we first consider optical reorientation of LCs, analogous to optical Fréederickscz transition

(OFT) in LC thin films. The optical reorientation should take place when the free-energy by

the light field exceeds Frank distortional energy. In a corresponding LC thin slab analog, in

which reorientation driven by optical nonlinearities has been accurately described, for

conventional light irradiations, the threshold of light density to induce molecular reorientation,

Ith, is given by [34,35] ) ( π I ₂ ₂ ₂ 2 2 th o e o e n n n d cKn − = (2.1)

(33)

21

extraordinary and ordinary refractive index, and K is the average Frank elastic constant. If

one considers that the diameter of the droplet is approximately equal with the thickness of the

LC thin slab analog, with K ≅10 pN, Eqn. (1) gives the threshold light density for a droplet

of 2.5-μm-diameter to be approximately 2 MW/cm2, corresponding to electric field strength of 2 V/μm. This calculated threshold power, however, is about two orders of magnitude lower

than the actual power to induce the reconfiguration throughout the inside of the droplet (the

threshold power density is ~360 MW/cm2). The striking difference between the calculation

and experimental value is due to that laser intensity of a focused beam is steeply distributed

around the diffraction limited size, not all inside the droplet volume. Nevertheless, with this

calculation, we proposed that the focused beam with laser powers slightly above OFT

threshold should not only generate gradient force, which confines stably the birefringent

droplet, but such a polarized laser beam propagating in the confined droplet can also induce

optical reorientation locally within the focal spot at the droplet center, although such local

reorientation is too small inside the droplet to be detected by the POM imaging.

As a self-organized structure, droplet-liquid interface energetic will also control the

molecular orientation inside the droplet through the interfacial anchoring effect. In this sense,

the interface can be considered to act as the anchoring layer. Under the optical trapping at

laser powers between the OFT and reconfiguration threshold, the reorientation propagates

(34)

22

thus, the droplet may adopt a kind of intermediate configuration. In this situation, the local

reorientations inside the droplet will induce non-symmetrical orientations, and such symmetry

breaking inside the droplet leads to spatially unequal torque. This is clearly indicated by

rotation of the droplet under circularly polarized beam. At higher powers, when optical

reorientation overcomes the anchoring effect, rotation of the droplet is coupled with its

reconfiguration. Thus, the rotation is induced as the result of local optical reorientations and it

can be considered as an early process or precursor of the reconfiguration, which most

probably is an equilibrium reorientation throughout the inside of the droplet. This means that,

when such an equilibrium configuration is formed, the local birefringence ultimately becomes

negligible, the torque is spatially balanced, and the rotation stops as observed with the

appearances of ring patterns.

The ring patterns of the optical transmission can be usually related to a far-field

intensity distribution of the transmitted light due to its spatial self-phase modulation and the

wave-front curvature, when the light passes through the nematic LC thin film [36,37]. In this

droplet case, the ring patterns may indicate that, in the equilibrium state, the LC molecules

inside the droplet is reoriented, but since both linearly and circularly polarized beams result in

the same POM images of the reconfigured droplet, we could draw a conclusion that their LC

orientations are not preferred along the laser polarization. This is an indication that

(35)

23

cooperative effects throughout the inside of the droplet. Though the LC molecular

configuration inside the droplet is still an open question, we qualitatively interpreted that one

of possible structures is a kind of twisted configuration [29]. However, as the initial pattern is

immediately restored when the laser beam is switched off, we could consider that the

molecular alignment at the interfacial layer which acts as an anchoring surface should always

remain intact.

Upon the reconfiguration, the local refractive index inside the droplet should jump

between the extraordinary and ordinary refractive indices. As a consequence, the optical

forces exerting on the droplet should be also modified, resulting in a relocation of the droplet

slightly from the initial trapping center. This could be attributable to the change in droplet

diameter in the POM image. In comparison, the similar effect has been observed in

photo-induced molecular reorganization inside optically trapped cholesteric LC droplets

[21,22].

4.2 A proposed model mechanism for droplet size dependence of threshold

power

Considering that the reconfiguration is initially induced and started by the optical

(36)

24

be the key factor in determining the threshold power. For droplets with radius comparable to

the beam waist,

a

≅

r

₀, in which all LC molecular dipoles in the droplets interact with a

tightly focused laser beam and are directly reoriented by the light, the reconfiguration

undergoes at the lowest threshold power. For larger droplets with

a

>

r

₀, the relevant scale

for the interactions is limited in the small volume of focal spot. Thus, the LC molecules at the

positions

r

>

r

₀ inside the droplet are reoriented either by the propagation of optical

reorientation at the focal spot due to cooperative effects or by the small light field of

transverse Gaussian intensity, leading to a higher threshold power for a larger droplet. This

interpretation is clearly demonstrated by the tendency of the threshold power in the

experiment (Figure 2.3). However, this finding contradicts essentially to the prediction of the

OFT threshold by conventional light irradiations (Eqn. 2.1), by which the threshold is

inversely proportional to square of droplet diameter. To clarify this issue, here, we evaluate

the optical reorientation and reconfiguration by a tightly focused laser beam as follows.

Firstly, we consider the interaction energy of LCs aligned throughout the inside of

the droplet by the electric field of the focused beam can be expressed as

∫

−

∝

αQI(

,

)dV

)

,

U(

r

z

r

z

, where α is polarizability, I(r,z) is the light field intensity as a

function of the lateral (r) and axial (z) distances from the center of the focal spot, V is volume

of the droplet, and Q is dimensionless factor for the target materials in the geometrical optics

(37)

25

Gaussian beam around the focus area is approximately expressed as,

( )

r,z I₀exp

(

2kr2 (kr₀2 4z2)

)

I = − + , where

I

₀ is the light power intensity at the center of

focal spot is k wavenumber in the surrounding medium. Secondly, since the light-induced

reconfiguration always competes with LC droplet-liquid interfacial anchoring effect [39], just

like a system consisting of two-coexisting processes, namely, an optical controlled-lattice site

surrounded by the interface-controlled self-aligned geometric configuration, we evaluate

qualitatively the threshold of light-induced reconfiguration based on the mean-field theory

[40]. In this case, we take into account the surface free energy of the droplet-liquid interfacial

anchoring effect, which is defined as the multiplication of surface free energy per unit area (β)

and the phase boundary area [26]. Considering that the depth of focus of the near-infrared

laser beam is much smaller in the front of the droplet size, the total energy can be

approximately given by,

2 0 2 0 2 2 0 _exp(-2 ₎ β αQI a dr r r r U a total ∝−

∫

+ (2.2)

This equation indicates that the total energy is a function of the radius of the droplet

and laser power in terms of

a

and αQI0/β. We show in Figure 2.4(a) that the total energy shows a negative local minimum when αQI0/β is larger than 7.5, indicating clearly that when

the laser powers is above such a threshold condition, the interaction energy of the LCs

(38)

26

interface, and thus, molecular reconfiguration can undergo throughout the inside of the

droplet [29]. To place the droplet-size dependence in context, with the criterion that the total

energy reaches a negative local minimum, in Figure 2.4(b) we show the

polarization-insensitive plot of αQI0/β as a function of

a

/ r

0, indicating two distinct regions

in the plane of parameters; where in the region above the graph, the optical trapping is

followed by molecular reconfiguration throughout the inside of the droplet. The slightly

higher threshold power for the circular polarized beam can be understood as a cause of

angular momentum transfer of the light inducing the rotation of the trapped droplet prior to

the reconfiguration. From the plot, it now became clear that a larger threshold power is

required for a larger droplet. The qualitative similarity between the tendency predicted by the

calculation and that in the experimental observation supports the relevance of the above

mentioned approximation.

Finally, for droplets with diameter much larger than the beam waist, in this case

>

a

1 μm, the predicted tendency shows a larger increase than that of the experimental

observation. This fairly indicates that some parameters in Eqn. (2) may change drastically for

the droplets with

a

>>

r

₀, resulting in such deviation from the experimental data. The

physical reasons are yet unclear, though one may consider that Q is also a function of

droplet-size [38]. The other likely reason for the deviation is that optical reorientation of LCs

(39)

27

through dipolar interaction, which may reduce the threshold power. We recall that such

cooperative effects of smectic layers in the thin slab film results in a significant reduction of

the predicted critical electric field at which the OFT effect takes place [41]. Hence, further

(40)

28 (a)

(b)

Figure 2.4 (a) The total energy as a function of the radius of the droplet at different level of laser power. (b) Plot of the predicted dependence of threshold laser power density to induce a negative local minimum, which is related to the reconfiguration, as a function of droplet radius relative to the beam waist, based on Eqn. (2.2) (see text for details).

0.0

0.2

0.4

0.6

0.8

1.0

1.2 -0.05

0.00

0.05

0.10 T

ot

al

E

ner

gy

Relative Droplet Radius (a/r

0

)

aQI

₀

/β=8

aQI

₀

/β=7

aQI

₀

/β=6

aQI

₀

/β=0

1

2

3

4

5

0

30

60

90

120

150 aQI

0

/β

(

a.

u.

)

(41)

29

5. Summary

We have presented laser-induced reconfiguration of micrometer-sized radial

symmetric droplets of liquid crystal 5CB dispersed in D2O. The definite threshold laser power

to induce such reconfiguration depends on the droplet size and polarization state. The

tendency of the droplet size dependence can be qualitatively explained by considering that the

direct optical reorientation by laser beam is mostly distributed around the focal spot and that

the LC droplet is a self-organized structure with droplet-liquid interfacial layer controlling the

molecular orientation inside the droplet, rather than by conventional non-focused light

irradiations. Thus, the likely mechanism involves the optical reorientation (OFT effect) at the

focal volume and propagating out of the focal spot by cooperative effect. When the interaction

energy of the LCs realignment by the electric field overcomes the surface free energy of the

droplet-liquid interface, the radially symmetry of the radial symmetric structure is completely

broken, and ultimately leading to a reconfiguration throughout the inside of the confined

droplet. With this proposed mechanism, we show qualitatively that the calculated tendency of

(42)

30

6. References

[1] A. Ashkin, Phys. Rev. Lett. 24 (1970) 156.

[2] A. Ashkin, Proc. Natl. Acad. Sci. USA 94 (1997) 4853.

[3] A. Ashkin, J. M. Dziedzic, J. E. Bjorkholm, S. Chu, Opt. Lett. 11 (1986) 288.

[4] K. Sasaki, M. Koshioka, H. Misawa, N. Kitamura, H. Masuhara, Jpn. J. Appl. Phys. 30 (1991) L907.

[5] A. Ashkin, J. M. Dziedzic, T. Yamane, Nature 330 (1987) 769.

[6] P. Borowicz, J. Hotta, K. Sasaki, H. Masuhara, J. Phys. Chem. B 101 (1997) 5900. [7] J. Hofkens, J. Hotta, K. Sasaki, H. Masuhara, T. Taniguchi, T. Miyashita, J. Am. Chem.

Soc. 119 (1997) 2741.

[8] J. Hotta, K. Sasaki, H. Masuhara, J. Am. Chem. Soc. 118 (1996) 11968. [9] Y. Tanaka, H. Yoshikawa, H. Masuhara, J. Phys. Chem. B 110 (2006) 17906. [10] T. Sugiyama, T. Adachi, H. Masuhara, Chem. Lett. 36 (2007) 1480.

[11] Y. Tsuboi, T. Shoji, N. Kitamura, J. Phys. Chem. C 114 (2010) 5589. [12] T. Sugiyama, T. Adachi, H. Masuhara, Chem. Lett. 38 (2009) 482.

[13] T. Rungsimanon, K. Yuyama, T. Sugiyama, H. Masuhara, N. Tohnai, M. Miyata, J. Phys. Chem. Lett. 1 (2010) 599.

[14] T. Rungsimanon, K. Yuyama, T. Sugiyama, H. Masuhara, Cryst. Growth Des. 10 (2010) 4686.

[15] T. Uwada, T.Sugiyama, H. Masuhara, J. Photochem. Photobiol. A: Chemistry 221 (2011) 187.

[16] X. Sun, B. A. Garetz, M. F. Moreira, P. Palffy-Muhoray, Phys. Rev. E 79 (2009) 021701. [17] A. Usman, T. Uwada, H. Masuhara, J. Phys. Chem. C 115 (2011) 11906.

[18] S. Joudkazis, M. Shikata, T. Takahashi, S. Matsuo, H. Misawa, Appl. Phys. Lett. 74 (1999) 3627.

[19] S. Joudkazis, S. Matsuo, N. Murazawa, I. Hasegawa, H. Misawa, Appl. Phys. Lett. 82 (2003) 4657.

[20] T. A. Wood, H. F. Gleeson, M. R. Dickinson, A. J. Wright, Appl. Phys. Lett. 84 (2004) 4292.

[21] H. F. Gleeson, T. A. Wood, M. Dickinson, Phil. Trans. Royal Soc. A 364 (2006) 2789. [22] Y. Yang, P. D. Brimicombe, N. W. Roberts, M. R. Dickinson, M. Osipov, H. F. Gleeson,

Opt. Express 16 (2008) 6877.

[23] E. Brasselet, N. Murazawa, S. Joudkazis, H. Misawa, Phys. Rev. E 77 (2008) 041704. [24] E. Brasselet, S. J. Joudkazis, Nonlinear Opt. Phys. Mater. 18 (2009) 167.

[25] N. Murazawa, S. Joudkazis, H. Misawa, Eur. Phys. J. E 20 (2006) 435.

[26] P. G. de Gennes, J. Prost, The Physics of Liquid Crystals, 2nd ed.; Clarendon Press: Oxford, 1993.

(43)

31

[27] A. I. Bishop, T. A. Nieminen, N. R.; Heckenberg, Rubinsztein-Dunlop, H. Phys. Rev. Lett. 92 (2004) 198104.

[28] N. Murazawa, S. Joudkazis, V. Jarutis, Y. Tanamura, H. Misawa, Eur. Phys. Lett. 2006, 73, 800.

[29] A; Usman, W.-Y; Chiang, T; Uwada, H. Masuhara, Proc. SPIE 8274 (2012) 82740L1-8. [30] S. Joudkazis, N. Mukai, R. Wakaki, A. Yamaguchi, S. Matsuo, H. Misawa, Nature 408

(2000) 178.

[31]S. Ito, T. Sugiyama, N. Toitani, G. Katayama, H. Miyasaka, J. Phys. Chem. B, 111 (2007) 2365-2371.

[32] N. Murazawa, S. Joudkazis, S. Matsuo, H. Misawa, Small 1 (2005) 656. [33] N. Murazawa, S. Joudkazis, H. Misawa, Opt. Express 14 (2006) 2481.

[34] P. Palffy-Muhoray, The Nonlinear Optical Response of Liquid Crystals. In Liquid Crystals: Applications and Uses; Bahadur B., Ed., World Scientiﬁc, New Jersey, 1990. [35] I.-C. Khoo, S. T. Wu, Optics and Nonlinear Optics of Liquid Crystals; World Scientific:

Singapore, 1993.

[36] I.-C. Khoo, R. R. Michael, P. Y. Yan, IEEE J. Quantum Electron QE-23 (1987) 267. [37] E. Santamato, Y. R. Shen, Opt. Lett. 9 (1984) 564.

[38] A. Ashkin, Biophys. J. 61 (1992) 569.

[39] I. I. Smalyukh, S. Chernyshuk, B. I. Lev, A. B. Nych, U. Ognysta, V. G. Nazarenko, O. D. Lavrentovich, Phys. Rev. Lett. 93 (2004) 117801.

[40] S. A. Safran, Statistical Thermodynamics of Surfaces, Interfaces, and Membranes, Westview: Colorado, USA, 2003 (2003).

(44)

32

CHAPTER 3 Optical Trapping and Polarization-Controlled Scattering

of Dielectric Spherical Nanoparticles by Femtosecond

(45)

33

1. Introduction

One of the successful applications of mode-locked lasers is the ultrafast

time-resolved spectroscopies, which provide the absorption, vibrational, or emission spectra

of atoms or molecules on extremely short time scales after their excitation with ultrashort

laser pulses. The research group of Prof. M. Martin is one of pioneers who have utilized

transient absorption spectroscopy to decipher the dynamics and mechanisms of fundamental

photo-induced processes [1]. Their reports on the insights of driving forces and primary

occurring events in the photo-induced dynamics of various chromophores, photoactive

proteins, or biomimetics are important advances in our understanding of the photo-processes,

particularly the functionality of the biomaterials in relation with their electronic structures [2–

7].

Another important laser application is optical trapping (also called optical tweezers),

exploiting the optical gradient force, which can confine micrometer to a few tens nm-sized

objects in the focal spot [8,9]. In this phenomenon, a high numerical aperture lens is

necessarily required to focus tightly the continuous-wave (cw) laser beams into a

diffraction-limited spot size [10,11]. With its potential ability of non-destructive tool to

immobilize, reorient, and transfer the dielectric or metallic particles, this technique has been

widely applied in various fields of sciences with target materials ranging from small particles

(46)

34

become indispensable in single-molecule measurements [21,22].

Recently, the optical trapping technique is further developed by utilizing ultrashort

laser pulses. By the femtosecond laser pulses, optical trapping of micrometer-sized silica

spheres was found to be as just effective as cw optical tweezers, and trap stiffness was related

to average power of the laser pulses [23]. With the ultrashort laser pulses, however, several

phenomena have been revealed, including optical trapping of as small as a few nm-sized

CdTe quantum dots or the depositions of CdS nanoparticles with grain size down to 25 nm

[24,25]. For the trapping of gold nanoparticles by laser pulses, the trapping site splits up into

two equivalent positions around the focal center, demonstrating that high nonlinear optical

susceptibility of the target materials can modify the shapes of gradient force and trapping

potential [26]. More recently, the femtosecond laser pulses with the power less than 200 mW

has been successfully applied to confine an individual polystyrene bead with a diameter of a

few tens of microns (the particle sizes within the framework of geometrical optics regime),

but the microparticle was pushed away from the trapping site when the focal position was

shifted to its downstream surface due to secondary convergence of the laser pulses that

reduces water breakdown threshold [27].

In this article, we report on an experimental study exploring the trapping behavior

of the dielectric spherical nanoparticles (50-nm-diameter polystyrene beads), suspended in

(47)

35

lens. We show that as compared with the cw mode, the laser pulses can trap a larger number

of nanoparticles. In addition, the laser pulses induce nanoparticle flows out of the focal spot in

two opposite directions, in an alternating manner, controlled by the laser polarization. To

understand this phenomenon, we evaluate both radiation (gradient and scattering) and

temporal forces (the latter is also called pulse radiation force) by adopting Lorentz force of

fundamental Gaussian beam exerted on Rayleigh particles [10,28], and by applying scattering

cross section value of the nanoparticles obtained based on Mie theory [29]. We demonstrate

that the axial electric field produced by the high numerical aperture objective lens is

(48)

36

2. Experimental

2.1 Optical setup

To experimentally exemplify the trapping behavior of the nanoparticles by

femtosecond laser pulses, we developed an experimental setup based on an inverted

microscope (Olympus IX71), as shown in upper Figure 3.1. We used a 800-nm fundamental

mode of Ti:sapphire (Tsunami; Spectra Physics) laser beam, which can be operated in cw or

femtosecond-pulse mode, acting as the trapping beam. When it was operated in the pulse

mode, the pulse duration was compressed by a pair of prisms to be about 90 fs, and the

repetition rate was 80 MHz. The polarization direction of the laser beam was controlled by a

half-wave plate before the beam was collimated and expanded to ~5 mm in diameter by a pair

of positive lenses with focal length being 100 and 200 mm, respectively. The beam then was

focused through an objective lens (60×, NA = 0.90) at normal incidence into a sample cell,

which was placed on the sample stage of the microscope. The light power after the objective

lens was controlled in the range of 0.10–0.35 W. The beam waist, ω0, at the focal spot was

calculated to be 460 nm, equivalent to the calculated radius when the beam intensity of its

(49)

37

Figure 3.1 Schematic diagram of the experimental set-up; λ/2 is halfwave plate, BS is beam splitter, and LPF is low pass filter. Inset is a schematic illustration of the sample cell containing the colloidal solution of nanoparticles.

2.2 Sample cell and detection system

The sample cell consisted of a silicon chamber (1 mm thickness) sandwiched

between two cover-glass plates (Matsunami). The inner well of the chamber (10 mm in

diameter) was filled with colloidal solution containing spherical polystyrene beads

(PolyScience; radius = 25 nm, density = 1.06 g/cm3) suspended in distilled water (refractive

index = 1.33 at room temperature). Concentration of the polystyrene beads was 3.79 × 1014

(50)

38

calculated by Cauchy dispersion relation to be 1.59 [30].

The sample cell containing the dielectric spherical nanoparticles was illuminated by

white light (λ = 400–750 nm) from a halogen lamp passing through a cardioid immersion dark

field condenser lens (Olympus; U-DCW NA = 1.4–1.2). The elastic light scattering originated

from the laser trapping beam was completely cut by a shortwave-pass filter with transmission

at 380–720 nm (Semrock; Brightline 750/SP) in front of charge-coupled device (CCD)

camera (JAI; CV-A551R E). With such a setup only the scattering light from halogen lamp by

the nanoparticles was collected by the objective lens, and was detected by using the CCD

camera running at 30 interlaced frames per second. Thus, the detected light intensity can be

related mainly to the scattering light intensity, although there is possibly a very minor

contribution of three-photon excited fluorescence of the bare polystyrene beads due to

nonlinear optical effects if such the fluorescence wavelength is longer than 380 nm to pass the

shortwave-pass filter. The positions of the nanoparticles were associated with the image of the

scattering light detected by the CCD camera. The resolution of the image in the lateral

direction was 94 nm per pixel, and our observation layer was limited within the axial

(51)

39

3. Results

3.1 Polarization-controlled off-axis scattering of nanoparticles

With the laser trapping beam operated in the femtosecond pulse mode at the average

power of 350 mW, we observed a brighter scattering light at the focal spot compared with the

surrounding area. We should note that such bright scattering light was never observed in a

neat solvent. In addition to scattering light at the focal spot, bright locus of scattered polymer

beads, just like multiple shooting stars, from the focus spot to the surrounding area was also

observed. The bright locus was shaped like a partially opened folding fan along two opposite

directions, in an alternating manner, perpendicular to the polarization direction. Such the

event occurred randomly, and simultaneous bright locus along the two opposite directions like

a pair of two partially opened folding fans was never observed. Thus, we show the bright

locus along the two opposite directions by combining two halves of different video frames in

Figure 3.2(a)–(c). Since the scattering lights detected in the video image represent the

positions of the nanoparticles, we therefore extracted the profiles of scattering light intensity

passing through the focal center as shown in each panel. Such line profile parallel to the

polarization direction shows clearly a single sharp peak with an approximately 1.6 μm full

width at half maximum at the focal spot, whereas that perpendicular to the polarization

(52)

40

opposite directions in an alternating manner. The intensity of the bright locus is comparable to

each other. When the laser beam was operated in the cw mode at the same laser power, only a

tiny scattering light at the focal spot was observed under the similar experimental conditions,

but there were no any observable bright locus of scattered polymer beads from the beam

center to the surrounding area. An image frame under the cw-mode laser irradiation is shown

in Figure 3.2(d). The line profiles passing through the focal center parallel and perpendicular

to the polarization direction reveal that the scattering light at the laser focal spot of the cw

mode is very low.

By varying the laser power, the threshold of the femtosecond laser pulses to induce

observable scattering light at the focal point and bright locus along the two opposite directions

was observed at 264 mW for the highly concentrated nanoparticle solution. We also found

that the concentration of the nanoparticles was a crucial parameter to observe the bright locus.

Under the same optical conditions, the laser pulses at the average power of 350 mW, the

bright locus was not observed when the solution was diluted by a factor of 4, equivalent to the

concentration of 0.95×1014 particles/mL. The fourfold dilution led to severe reductions in the

trapping rate, size of trapped assembly, and number of scattered nanoparticles, giving no

(53)

41 (a) (b)

(b) (d)

Figure 3.2 A combination of two halves of two different image frames showing optical trapping and scattered polystyrene nanoparticles by femtosecond-pulse modes and showing their optical trapping by cw Ti:sapphire laser beams (λ= 800 nm). (a)-(c) A sharp scattering light at the focal spot and bright locus of scattered polymer beads from the focus spot to the surrounding area towards two opposite directions in an alternating manner perpendicular to the laser polarization direction of femtosecond laser pulses, indicating the optical trapping of polystyrene nanoparticles at the focal spot and nanoparticle flows along the two opposite directions. The laser power for each case is 350 mW after the objective lens. Arrow in each panel indicates polarization direction of the laser beam. The line profiles in each panel were taken from one cursor passing through the focal center to the opposite cursor, parallel and perpendicular to the polarization direction. The two line profiles (solid and dotted lines) perpendicular to the polarization direction are related to the two alternating directions of the scattering light. (d) An unstable and low scattering light intensity from the focal spot of cw laser beam at the same laser power. Arrow in the panel indicates polarization direction, and the line profiles were taken from parallel and perpendicular to the polarization direction.

(54)

42

3.2 Temporal evolutions of the bright spot at the focal spot

The temporal evolutions of scattering light intensity at the focal spot when the laser

beam was operated in femtosecond-pulse or cw mode are shown in Figure 3.3(a). In contrast

to high intensity of scattering light when the laser beam was operated in the pulse mode, low

intensity, unstable, and fluctuated scattering light intensity was observed under the cw mode.

In comparison, the scattering light intensity is about one order higher when the laser beam

was operated in the pulse mode than that under cw mode.

Further, in Figure 3.3(b) we show the plot of the temporal random distribution of

the event, where the bright locus was observed along one of the two alternating directions

perpendicular to the polarization of the femtosecond laser pulses, for the observation window

about 120 s. The bright locus emerged in one direction on the timescale of seconds before

they changed into the opposite direction, and they continued in the same way. The probability

and total duration of the bright locus along the two alternating directions almost balanced

(55)

43 (a)

(b)

Figure 3.3 (a) Time-dependent line profile intensity under femtosecond-pulse (black line) and cw mode (gray line). (b) The typical time-dependent event when bright locus of scattered polymer beads from the focus spot is along the two alternating left and right directions within the observation window of 120 s. Experimental information for data shown in the figures (a) and (b) are related to those in Figure 3.2a.

0 20 40 60 80 100 120 0 5 10 15 20 150 200 250 L ig h t S c a tt e ri n g I n te n s it y / a .u . Trapping Time/ s 0 40 80 120 T im e / s

Scattering Direction Probability 1 0

1

(56)

44

4. Discussion

4.1 Optical trapping and nanoparticle flows

Extracting the intensity of scattering light by line profile is one of useful practical

ways to identify the spatial position of nanoparticles in colloidal solution. Typically, a sharp

light intensity at the focal spot in the line profiles can be attributed to nanoparticles

accumulated by laser trapping, and the number of trapped nanoparticles is associated with the

light intensity [31]. Similarly, the existence of such single and sharp peak intensity from the

spherical polystyrene beads under the femtosecond laser pulses, as shown in Figure 3.2,

indicates the possible of a single trap site at the focal spot. Such a single trap site is commonly

observed in conventional optical trapping experiments [8] and [32], but it is in contrast to the

trap split of 60-nm sized gold nanoparticles by the femtosecond laser pulses on the same level

of average laser power [26]. It is noteworthy that since the third-order nonlinear optical

susceptibility is responsible and sensitive to split the trap site into two equivalent positions

shifted from the beam axis, the lower third-order susceptibility of polystyrene nanoparticles (0.8×10−8 esu) [33] as compared with that of gold nanoparticles (5×10−8 esu) [34] is the

reason for the polymer nanoparticles having no clear observable nonlinear optical effect on

the trap site.

Here, we interpret our experimental results as follows. When a nanoparticle enters

(57)

45

intensity of the highly focused ultrashort laser pulses is exerted on the nanoparticle, and the

force drags the nanoparticle towards the single trap site at the focal center. A stable trapping

can only be achieved when the gradient force overcomes scattering force, and the

characteristic stiffness of optical trap should be proportional to the laser intensity, as it has

been reported for the trapping laser in cw mode [21] and [35]. For the ultrashort optical pulses,

in particular, in addition to the gradient and scattering forces, we also have to consider

temporal force within the pulse duration, which is defined as instantaneous Lorentz force at

the time over the entire duration of the pulse envelope [36], [37] and [38]. In this case, the

gradient force of the laser pulses should overcomes the scattering and their temporal forces to

achieve a stable trapping. Considering the sizes of the nanoparticles and beam waist, multiple

nanoparticles can be trapped in the single potential minimum at the focal spot, and the

scattering light intensity is proportionally enhanced with the number of trapped nanoparticles

at the trap site. Thus, monitoring such the scattering light intensity, similarly to monitoring a

stepwise behavior of fluorescence intensity increase in optical trapping of 100- or

200-nm-sized dye-doped polystyrene nanoparticles by cw laser beam [39], would provide the

information on the number of nanoparticles entering the trap site. However, the video frames

in our experiment are saturated within the first integrating time, hindering a precise

observation on the exact number of the optically trapped nanoparticles. Nevertheless, based

連續波及飛秒脈衝雷射應用於微米至奈米大小球狀物質的光學捕捉

國立交通大學

應用化學系

碩 士 論 文

連續波及飛秒脈衝雷射應用於微米至奈米大小球狀

物質的光學捕捉

Optical Trapping Dynamics of Micron- to Nanometer-Sized Spherical

Target Materials by Continuous Wave- and Femtosecond Pulse-Mode

Laser Beams

研究生：江威逸

指導教授：增原 宏 教授

連續波及飛秒脈衝雷射應用於微米至奈米大小球狀物質的光學捕

捉

Optical Trapping Dynamics of Micron- to Nanometer-Sized Spherical

Target Materials by Continuous Wave- and Femtosecond Pulse-Mode

Laser Beams

研究生：江威逸 Student: Wei-Yi CHIANG

指導教授：增原 宏 教授 Advisor: Prof. Dr. Hiroshi MASUHARA

國立交通大學

應用化學系

碩士論文

Published paper

Published proceeding

Conference

連續波及飛秒脈衝雷射應用於微米至奈米大小球狀物質的光學捕捉

研究生：江威逸 指導教授：增原宏 教授

國立交通大學 應用化學系碩士班

摘要

Optical Trapping Dynamics of Micron- to Nanometer-Sized Spherical Target

Materials by Continuous Wave- and Femtosecond Pulse-Mode Laser Beams

Student: Wei-Yi CHIANG Advisor: Prof. Dr. Hiroshi MASUHARA

M. S. Program, Department of Applied Chemistry

National Chiao Tung University

Abstract

Acknowledgment

Contents

List of Figures

CHAPTER 1

References

CHAPTER 2

Polarization and Droplet Size Dependences of Laser

Trapping-Induced Reconfiguration in Individual Radial

1. Introduction

2. Experimental

2.1 Optical setup

2.2 Sample preparation

n

n

3 Results

3.1 Reconfiguration of the LC droplet

3.2 Droplet size-dependence of threshold laser power

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

300

400

500

600

T

hr

es

hol

d P

ow

er

[

M

W

/c

m

]

Droplet Radius [µm]

*

4. Discussion

4.1 Laser trapping-induced reconfiguration

4.2 A proposed model mechanism for droplet size dependence of threshold

power

a

≅

r

a

>

碩士論文

指導教授：增原宏教授

指導教授：增原宏教授 Advisor: Prof. Dr. Hiroshi MASUHARA

研究生：江威逸指導教授：增原宏教授

國立交通大學應用化學系碩士班