Photonic lattices achieved with high-power femtosecond laser … · 2016-11-21 · Photonic lattices achieved with high-power femtosecond laser microexplosion in transparent solid

Photonic lattices achieved with high-power femtosecond lasermicroexplosion in transparent solid materials

Hong-Bo Sun* Ying Xub, Saulius Juodkazisa, Kai Sunb, Junji Nishiic,

Yoshihisa Suzukid, Shigeki Matsuob and Hiroaki Misawa* b

aSatellite Venture Business Laboratory, The University of Tokushima,

2-1 Minamijyosanjima, Tokushima 770-8506, Japan

bDepment of Ecosystem Engineering, Graduate School of Engineering,The University of Tokushima, 2-1 Minamijyosanjima, Tokushima 770-8506, Japan

COptical Materials Division, Osaka National Research Institute,

1-8-31 Midorigaoka, Iketa. Osaka 563-8577,Japan

doptoTechnology Laboratory, The Furukawa Electronic Co. Ltd.,

6 Yawata-Kaigandori, Ichihama, Chiba 290-8555,Japan

ABSTRACT

We propose and utilize ultrashort laser pulses to tailor three-dimensional microstructures and their optical properties.

When an intense femtosecond pulse was tightly focused into some transparent materials, a laser-induced microexplosion

occurred, generating void holes inside the medium. When the thus-fabricated holes or cylinders were regularly organized, a

microstructure with a periodic refractive index distribution was accomplished, which was liable to act as a photonic crystal

structure. One-, two, and three-dimensional photonic lattices have been acquired by using this technique. Significant

photonic band gap effects were confirmed by transmission measurements. The unique feature of the ultrashort laser

micromachining of photonic crystal structures was the availability of arbitrary spatial geometry.

Keywords: Transparent media, Vitreous silica, Femtosecond laser. Ultrashort laser pulse, Laser-induced microexplosion,

Photonic band gap, Photonic crystal, Photonic lattice, Laser mcirofabrication

1. INTRODUCTION

Since the establishment of the first ruby laser in 1960, light-matter interactions have been a hot topic for 40 years. In

the mean time, laser fabrication has grown as a powerful tool for industrial applications'2 such as laser annealing and circuit

drawing of semiconductor materials, laser drilling, welding, and cutting, and laser prototyping. More recently the

development of high-power laser systems with lower pulse energy but untrashort pulse duration has created an opportunity

to acquire fine optical and photonic microstructures. When a femtosecond (fs) laser pulse of proper energy is tightly focused

into some transparent media, very localized structural changes, which result in a strongly modified index of refraction, are

*Correspondence: [email protected] .ac.jp, or hbsun @ieee.org

In High-Power Lasers in Manufacturing, Xiangli Chen, Tomoo Fujioka, Akira Matsunawa, Editors,Proceedings of SPIE Vol. 3888 (2000) • 0277-786X/OO/$1 5.00 131

Downloaded from SPIE Digital Library on 01 Mar 2012 to 136.186.72.84. Terms of Use: http://spiedl.org/terms

132

produced3. Three dimensional (3D) optical memory bits4' and total internal reflection fiber waveguides6 have been written

in silica glass. In this paper, we report the fabrication of photonic crystal (PhC) structures by periodically arraying the laser-

damaged spots in three spatial dimensions. The validity of this technique has been approved by pronounced photonic band

gap effects.

The laser microfabrication methodology has at least three unique merits comparing to the conventional

photorefraction technology7 (e.g. the photogenerated change of the refractive index) and semiconductor large-scale

integrated circuit (LSI) techniques such as dry and wet etching8. (i) Penetrativitv. Light, different from particle beams, can

penetrate into transparent materials to tailor a desired structure from inside materials. The final structure, in principle,

doesn't depend on the order of processing. For particle (electron or ion) beam fabrication, the undercutting is a touchy task

due to a shadowing effect from the outer part of media. (ii) High spatial resolutions. Since the bandgap width of general

transparent dielectrics is larger than the photon energy of the visible and the infrared (IR) laser, electron excitation in

materials is from a multiphoton absorption (MPA) process. The probability of MPA depends exponentially on the laser

intensity, so under a tightly focusing condition, the absorption is confined at the focal point to a volume of the cubic laser

wavelengths. The response of materials to excitation such as the fluorescence or a photoinduced chemical reaction are also

localized in this small volume9. The photogenerated spot size3, in a range of several hundreds of nanometers. is much

smaller than the optical diffraction limit of the fabricating laser. (iii) Extensitv. No photosensitive materials are required and

a variety of transparent media such as glasses, crystals, and plastics are suitable as matrix materials, while the induced

refractive index contrast is much larger than that acquired by the photorefraction.

Picosecond and longer laser pulses don't serve such a purpose from two aspects: (i) Material response to the longer

pulses is a thermal process (see later in §2.2), the local material modification doesn't give a strong enough contrast of the

refractive index. In addition, with longer pulses, e.g., hundreds of picosecond duration, the resulting structures are

irregularly shaped', and cracks appear in the bulk of medium even at energies only slightly above the laser-induced

damaging threshold (LIDT). (ii) Self-focusing effect becomes pronounced with an increased pulse width". A filament-like

damaging trace causes a severe inter-contamination between the nearest-neighbored spots. On the other hand, recent

experiment observations in the short-pulse regime (<ps) suggest that the group-velocity dispersion (GVD) becomes

increasingly important and that, therefore, ultrashort pulses resists the self focusing'2. Numerical studies'4 of the self-

focusing of femtosecond laser pulses in a normally dispersive medium show the self-focusing threshold was increased much

higher than that for damaging. The laser microfabrication was managed at an energy level of several times of LIDT, free of

self-focusing.

2. INTENSE FEMTOSECOND LASER-INDUCED DAMAGES

IN TRANSPARENT DIELECTRICS

2.1 Optical Excitation and Carrier Relaxation Process

Generally speaking, MPA processes are much less likely to occur than the single photon absorption (SPA) but their

probability increases with an increasing of the laser intensity. For an insulator dielectric (as commonly delined, Eg> 3 eV)


7cN(a;) (b) (c) (d)

Fig. I Several processes for electrons excitation (a, b) and relaxation (c, d), including free-carrier absorption (a), impact ionization (b),carrier-carrier scattering (c), and inelastic scattering of carriers by phonons (d). When transparent dielectrics (commonly with a widebandgap) were irradiated by the infrared or the visible laser, the electron excitation from the valence band to the conduction band wasfrom a multiphoton absorption process.

irradiated by the visible (Eu: I .5—3.OeV) or the infrared (IR) laser, free carrier pairs are only created by an MPA process.

Si02 has a band gap of approximately 8.9 eV, and 400 nm (3. 10 eV) and 800 nm (I .55 eV) excitation correspond to 3- and

6- photon absorption. The carriers in the conduction band can be further excited by absorbing photons [Fig. 1(a)]. No new

carriers are generated in such a process. However, the carrier energy is indeed increased. If some of the free carriers have

sufficient access energy above the conduction band minimum, impact ionization will occur. The free carrier transfers its

excess energy to a valence electron, creating a new electron hole pair. This process is shown in Fig. I (b).

After the electrons are excited from the valence band to the conduction band, they relax quickly through a number of

processes'3. Firstly a thermalization of the nonequilibrium distribution created by the excitation at a fixed wavelength takes

place by carrier-carrier scattering [Fig. 1(c)] on a 10 fs time scale. This process changes neither the total energy in the

carriers nor the carrier density but redistributes carriers over the band. The highly excited free carriers can emit phonons

[Fig. 1 (d)], having the lattice heated. Lattice thermalization, a slow relaxation process. generally requires picosecond (ps)

timescale, especially for the excitation with large excess energy.

Free carrier recombination'3 also plays an important role in the relaxation process. In a radiative recombination

process, an electron and a hole of the identical wave vector recombine emitting a photon with their energy difference. This

process lowers both energy and density of free carriers, taking place in a timescale of nanosecond. The non-radiative

recombination is an inverse of the impact ionization process: an electron and a hole recombine transferring momentum and

energy to a third carrier. This three-body process is important mainly at the high carrier density, N, and its probability is

proportional to N .

2.2 Laser.Induced Microexplosion

In § 2.1. we have given some general descriptions on the optical excitation and relaxation of carriers. For

microfabrication, we are concerned much more on the properties and applications of the damage induced by irradiation.

Since the discovery in the 1970's that semiconductor can be annealed by laser irradiation, two models" have been

proposed to interpret the structural changes resulting from the laser excitation. One, known as thermal model", describes the

133


structural changes as a thermal melting process. It assumes that the hot electrons rapidly equilibrate with the lattice through

the phonon emission. With this assumption, the laser energy deposited in materials can be treated as though it is instantly

converted to heat. If the excitation is strong enough, the irradiated region of the sample will be heated up to melting

temperature and undergo a phase transition to the liquid phase. For nanosecond and picosecond pulse irradiation, the

processes including heating of solid, thermal diffusion, melting, motion of the interface boundary and recrystallization has

been well confirmed. Another model, known as plasma model'4, ascribes to a slow rate of phonon emission by the excited

electronic system compared to the energy deposition time, i.e., the laser pulse duration. According to this model, the

10000

300 400 500 600 700.-; Wavelength (nm)c ..•: 1000 ..Cl)

100 . .

1 10Laser Pulse Energy QiJ)

Fig.2 Linear dependence power law of the PL intensity on the fabricating laser pulse energy. Before irradiation, no any PL features weredetected in vitreous silica samples. Three PL bands at 1 .9 (s). 2.7 () and4.4 eV (a). respectively, emerged after irradiation, which wererelated with different species of the laser-induced point defects. The inset is the PL spectrum of the irradiated sample excited by 5.0 eVphotons. Since the excitation was from a Xenon lamp. the PL intensity depended linearly on the defect density (band a was taken as theindex in this figure). For proving the linear absorption of photon by excited electrons as discussed in the text, it is necessary to assumethat the photogenerated defect density was proportional to the photon number absorbed by the material,

structural change is driven directly by the excited electronic system. If high enough fraction of the valence electrons are

excited from the valence band to the conduction band, the crystal becomes unstable, and a structural phase transition occurs.

Such a model has been accepted for the femtosecond laser excitation through numerous experimental and simulation.

With the fs laser excitation, the initial electrons are excited across the bandgap through MPA, then they acquire excess

kinetics through a sequential linear absorption [Fig. I (a)] and then scatter off other electrons exciting them across the

bandgap by impact ionization [Fig. 1(b)]. The newly generated electrons can then undergo the similar linear absorption. and

this process continues as an avalanche. For SiO,, the difference of bandgap width and photon energy demands a MPA

process for the excitation, and the ensuing linear absorption is critical for the launch of an avalanche, which can be

134


approved by monitoring the intensity of photoluminescence (PL) emitted by defects in damaged materials (Fig. 2), provided

that the defect density is proportional to the irradiated laser intensity.

The electrons are ionized and heated faster than they transfer their excess energy to the lattice. Therefore hot plasma

is formed before the lattice is significantly heated. The highly excited and tightly localized plasma will expand in an

explosive way, generating a hole surrounded with a more densified phase. Fig. 3 shows optical microscopic images of spots

irradiated in v-silica and PMMA.

(a)

(b)

Fig.3 Optical microscopic images of spots induced by single-shot 800 nm and 150 fs laser pulses inside the bulk of vitreous silica(a) and PMMA (b). The laser was focused 50 mm below the surface by a objective lens of IOOX, and 1.35 ofNA. The scale baris 10 jim.

3. FABRICATION AND CHARACTERIZATION OF PHOTONIC CRYSTAL

STRUCTURES IN VITREOUS SILICA

3.1 Principle of the laser microfabrication for photonic crystal structures

The utilization of vitreous silica (v-silica) as a matrix material in our current research is a natural choice since the idea

matured on a 3D optical memory in silica4 ' Actually there are much more candidates for this application only if the

material properties meet some fundamental requirements as follows. (i) Transparency. The materials should be transparent

at the fabrication wavelength (e.g. 400 nm or 800nm in this work) and the detection wavelength (up to 3—5 j.tm in this work,

dependent on the photonic lattice constant). The latter means a low-loss condition at the wavelength where the forbidden

band gap is lying. This is critical for the applications of waveguide structures. (ii) High refractive index. Numerical study

and experiments show that a ratio of refractive index between dielectric pairs in a 3D PhC should be not less than 2.0 for

opening a full bandgap'6. A refractive index of I .0 of the vacuum requires that of the transparent solid be larger than 2.0.

Tab. I lists the properties of several transparent solids. (iii) Optical isotropy. For objective lens of a large magnification and

high NA, the laser was focused from a cone shape with a large solid angle. In a birefringent crystal, the focusing will be

spread and deformed. Therefore optically isotropic crystals, amorphous materials, and polymers are much favorable. (iv)

Wide bandgap. This is commonly guaranteed by transparency condition as specified in (i), however, we want to emphasize

that the MPA process is an effective way to confine the modified region, as discussed in §2.2. An negative example is

As2S3 (Eu:— 2.0 eV) irradiated by a single shot laser pulse at 400 nm (3.1 eV), where a strong surface absorption even

135


prohibits any focusing into the bulk.

In addition to the choice of materials, technology parameters such as far-field pattern, pulse-to-pulse energy stability

of fabricating laser output, NA and magnification of objective lens, accuracy of movement of piezoelectric translator (PZT)

are critical for a successful fabrication. These affects will be mentioned in the following paragraphs.

Tab. I * Optical properties of several selected transparent materials: transparency windows. refractiveindexes, and bandgaps. Some of them are considered as promising for the fabrication of PhCs.

Materials Refractive indexTransparencyWavelength

(tm)

Bandgap (eV) States

Silica glass(S102)

1.4-1.5 0.19 - 3.0 8.9 Amorphous

Quartz(Si02)

n0= 1 .5280, 'e .5365(X==1.53 tm)

O.193—3.6 8.9Positive uniaxialCrystalline

Sapphire(Al203)

n0= 1 .7466

(?= 1 .53 rim)0.18—5.1

(cx= 1 cm1 level)Negative uniaxialCrystalline

LiNO3n0=2.2083, fle21356(X=: 1 .64 urn)

0.4-5.5(0 transmittance) 3.9

Negative uniaxialCrystalline.

Diamond(C)

n=2.4024 (X=O.76 rim)n=2.379 (A=2.5 Mm)

O.24-2.7(cx=1 cm1 level)

5 2 IsotropicCrystalline

Rutile(Ti02)

n0=2.4413, r1e26902(X= 1 .60 tim)

O.435.3(a=: 1 cm1 level) Crystalline

f3—ZnSn0=2.2716(1 .53 tim)

0.6-13(a= 1 cm1 level)

3.54 IsotropicCrystalline.

ZnSe n=2.491 (X=1.O lim)n=2.448 (?=2.O lim)

O.5118(a=1 cm1 level)

2.58 IsotropicCrystalline

CdS n0=2.296, ne=2.312(X= 1 .5 jim)

O.6514.8(a=1 cm1 level) 2.42

Positive uniaxialCrystalline

GaPn=3.0509(1 .6 jim)

0.56-1.4 (1 cm1 )0.52—7.2 (10cm1 )

2.25 IsotropicCrystalline

PMMAn=1.4815(A=: 1 .00 rim)

0.35-1.6(0 transmittance) 3.5 Polymer

*Data were mainly from "Handbook of Optics," Vol. 1 and II. 2nd edition, eds. M. Bass. E. W. VanStryland. D. R. Williams and W. L. Wolfe. McGRAW-Hill. mc, New York, 1995.

3.2 A Characterizationof Primitive Elements of Crystals

When a single-shot laser pulse with appropriate energy was focused into transparent solids, a nearly empty void was

generated. The refractive index contrast between the core region and matrix offered a modulation of light field. Therefore

the occurrence of the void and its size was critical for any applications. A clear evidence for the existence of the voids was

acquired by an atomic force microscopic (AFM) measurement. The vitreous silica sample was polished until the surface

level reached the fabricated layer then scanned by an AFM. Shown in Fig. 4 (a) is a 2.5X 2.5 im2 AFM image of spots that

irradiated by single pulse shots.

Formation of a void must be accompanied by material compaction at the surrounding region, which would be

presented as a dense crust. The refractive index would be increased in such a compacted layer since the mass-density,

therefore the dielectric constant, would be increased. A simplified model as shown in Figs. 4 (b) and (c) schematically

illustrate a distribution of the refractive index around the damaged sites. Instead of a square well presentation, the refractive

136


Fig.4 Voxels induced by single—shot laser pulses inside the v-silica. (a) Measured AFM images, (b) a suggested model of voxels after themicroexplosion, and (c) a possible refractive index distribution, where region A is the silica matrix, region B is the compressed crust, andC. the inside of void.

index profile was considered as gradually changing, implying

that the void may be not completely empty, but consisted of

less dense material except the central region.

3.3 Inlayed-Atom-Like Photonic Crystal Structures

The generic crystal is thought of as composed of unit

cells, which describes the smallest atomic building block

within ordered crystalline lattice with translational symmetry

in all directions. These unit cells are packed in a close-fitting,

three-dimensional array that results in long-range periodicity.

In the case of PhC'7, a similar lattice can be constructed by

packing the microexploded holes as "atoms". This was

fulfilled by accurately positioning and translating the focal

point of the laser with a computer controlled 3D PZT [Fig.5

(a)]. Each lattice point was written by a single-shot laser pulse.

By using a large NA objective lens, near-spherical voids can

be obtained and the high reproducibility of their shape was

favored by the stability (±3%) of the laser pulse output.

Intuitively the holes resembled the atoms in Thomson atom

model, and therefore we call these structures inlayed- "atom "-

like PhCs.

Since a laser can penetrate the medium without

damaging surface region and the microexplosion occurs only

Fig. 5 (a) Schematic laser microfabrication system withlight and electronic controlling units, where theabbreviations denote BS: beam splitter; RM: reflectionmirror; PD: photo diode. (b) Pattern of the Chinesecharacter of "light" as a fabrication example..

137

Laser RM

PD


in the focal point due to MPA, the voids indeed can be spatially arranged freely. Shown in Fig. 5 (b) is a pattern of Chinese

character of "light".

Different photonic lattices were realized just by varying computer-aided design (CAD) program. As an example, one

( 1 1 1 ) plane of fcc-lattice PhC is shown in Fig. 6 (a). The entire crystal was created through layer-by-layer stacking ( I 1 1)

planes.

Fig. 6 Fcc PhC structures. (a) Optical microscopic image of a (1 1 1 ) plane in the lattice, and (b) simulated (dashed line) and measured(solid) transmission spectra

To testify photonic bandgap effects, a transmission spectrum was measured with a Fourier transform infrared (FTIR)

spectrometer as presented by the solid line in Fig. 6 (b). The minimum of transmittance occurred at 3490 cm'. Changing the

lattice constant caused a synchronous variation of the wavelength of the transmission valley as predicted by Bragg law,

showing the valley is indeed from the photonic band gap effect. A transmission spectrum was calculated by a full-vector

calculation to reproduce the experimental valley wavelength of 2.87 im (3490 cm'), with the void radius r and the

difference in the refractive index, M, as fitting parameters. As a result, d 250 nm and An 0.45 gave a good agreement

between the measured and calculated dips. The refractive index difference Ltn 0.45 is much larger than those achieved by

using photorefraction (commonly less than 102 in order) techniques. The simulated spectrum is shown by dashed line in Fig.

6(b). What is noteworthy is the value of d measured by AFM is larger than 250 nm. The AFM-measured lateral size

depended on the polishing level and some deviation was possibly induced in polishing (widening) and from AFM lateral

measurement, that is why we haven't fixed d as a constant in the simulation. However, the average difference of refractive

index between silica matrix and inside the void, zn, would be a little smaller than 0.45 if the actual diameter of a void was

larger than 250 nm. This is consistent with the model in Figs. 4 (b) and (c).

3.4 Photonic Crystal Structures Consisting of Cylinder

100•

80

60>(1)CU)

C

U)N

E0z

40

20(b)

3200 3400 3600

Wavenumber (cm1)

38004000

138


An alternative way to create microstructures is (a)scanning the focal spot linearly so that a cylinder was

Polymeruacket Grating Pairintroduced. If a laser of high repetition rate was , ,,,•

. .selected, the nominal spot spacing can be very small,

I

for example, I 0 nm for 1 kHz laser and 1 0 jim/s stage Light

velocity. Therefore interwalls between bits neighbored Core High n layerin the same line were partly or completely crushed and

ejected by the ensuing pulse shocks. Shown in Fig. 7 is

a grating pair structure, i.e., a quasi-ID PhC,fabricated in a multi-mode optical fiber.

With the presence of PBG, the group velocity

thtildk decreases near the bandedge, where is the

angular frequency and k is the wavenumber. The phase

velocity w/k can be controlled by controlling the band

structures. Control of the group and phase velocity of

light has several advantages. For example, photon-

electron interactions such as nonlinear optical

susceptibilities per unit volume can be usefully Fig. 7 Optical fiber grating. a quasi-iD PhC structure. (a)Structural design of the orating pair PhC, and (b) image ofincreased if GVD is reduced. As regards second of the two gratings.

harmonic generation (SHG) a new phase matching can

be achieved by controlling the phase velocity of the fundamental and second harmonics. The wavevector in such a quasi-

one dimensional PhC can be calculated as follows'8:

k(w)n (co)aw ,n(w)bw n +n., . n,aw nbw— —arccos[cos(

')COS ) — sin(—)sin( )Id C c 2n,n2 c c

As shown in Fig. 7(b), the grating pitch is 0.8 jim, which corresponds to a transmission wavelength of approximately

2.5 .tm. Even the smallest lattice constant we achieved up to now, i.e., 0.6 .tm, put its transmission center to about I .8 tim,

which excludes the possibility to characterize this grating structure using general techniques applied to long wavelength JR

LED or LD such as InGaAsP quantum-well lasers. Further configuration of measurement devices is still undergoing.

By stacking the grating planes as shown in Fig. 7 (b). a variation of 2D and 3D photonic lattices consisting of

cylinders can be acquired. A direct application to 3D is so-called layer-by-layer'7 structure by in-plane turning every other

layer grating plane by an fixed angle, e.g. 900 then offset the nearest layers with the same orientation by half a period. As

an fabrication example, we present here the results from a 2D triangular lattice as shown in Fig. 8(a), which had been

theoretically and experimentally shown to be the best geometry for achieving a full in-plane band gap for both E- (electric

field parallel to the cylinder axis) and H- (magnetic field parallel to the cylinder axis) polarizations.

The cylinders were confirmed to be hollow by an AFM measurement. The fabricated structure was observed using

optical microscope from both the top and the side. For the side viewing, the glass plate was broken and end facet was a little

139


140

polished. Figure 8 (b) shows typical microscope images of cross-section of a triangular PhC structure. The lattice constant

of the triangular array was I .2 rim. Limited by the PZT working distance the entire area of PhC in x-yplane was 40x40 im2.

The almost round cross section ofthe air rod was favored by the utilization oflarge NA ofobjective lens (NA 1.35).

Typical transmission characteristics of the PhC structures with both polarizations are shown in Fig. 8 (c). To recognize

the bandgap feature, the spectra were normalized by the transmission of randomly irradiated samples. An approximately

10% transmission dip for H-polarization occurred at a wavenumber of 4100 cm', while that for E-polarization was much

less pronounced at 4000 cm'. A perfect air rod

structure with refractive index difference of I .45

permits almost a full bandgap, i.e., light propagation is

forbidden in all direction (in-plane for the 2D periodic

lattice) if a sufficient filling ratio was offered. However,

the diameter of the cylinder was smaller than necessary,

while reducing the line spacing or increasing the rod

diameter (by increasing the laser power) would lead to

a distortion of cylinders.(b)

Series of angle-dependent transmission spectra

will give much more information on the photonic band

structures. A rough measurement indicates that (i) the

first Brillouin zone band of r-x is broader for H- than

that for E-polarization, while the opposite is true for r-

J gaps. (ii) T-J gaps were at shorter wavelength than

corresponding T-X gaps. What is noteworthy is that the

band gap effects for H-polarization was always

observed to be stronger than E-polarization. This

showed that H-polarization (or TE modes) was favored

in a lattice of isolated low-n regions, consistent with

the theoretical expectation19. c0The existence of an uncoupled mode2° which

cannot be excited by an external plane in the cases of Eboth 2D and 3D PhC should not be neglected. It means

that the transmittance does not necessarily properly

reflect the photonic state density. The measured opaque

region can be either from a band gap or an uncoupled 3750 4000 4250 4500

mode. In both cases the transmission attenuation is Wavenumber (cm1 )

closely relevant to the periodicity of the structures.Fig.8 2D triangular PhC lattice. (a) Designed structure. (b) cross-

Finally, several problems for the laser fabrication sectional image and (c) transmission spectra

3500


should be mentioned. The spectra of fs laser pulse, due to uncertainty principle, are commonly broad, for example, I 0 nm of

FWHM in our case. A dispersion of the refractive index will degrade the focusing. An objective lens designed to

compensate the aberration for a specified wavelength would solve this problem. In addition, random light scattering in

fabricated PhC structures attenuates the propagating light power very much. Although it can be partly solved by high-

temperature annealing of samples. further measures are evidently necessary for the application of the PhC, for example, as

integrated-micro-optical waveguide.

ACKNOWLEDGEMENT

This work was supported in part by a Grant-in-Aid for Scientific Research (A)(2) from Ministry of Education,

Science, Sports and Culture (09355008), the Marubun Research Promotion Foundation and Satellite Venture Business

Laboratory of the University of Tokushima.

REFERENCES

1 . A. Blatter, M. V. Ailmen, "Laser-beam interactions with materials: Physical principles and applications." (Spnnger Series in Materials

Science. V. 2), Springer Verlag, 2nd edition, Berlin, 1995.

2. I. Nakagawa, Y. Marutani, "Layered manufacturing system: A new development of three-dimensional copying technology," Kogyo

Chosakai Publishing Co, Ltd. Tokyo, 1996.

3. E. N. Glezer and E. Mazur, "Ultrashort-laser driven microexplosion in transparent materials," Appl. Phys. Lett. 71, pp. 882-884, 1997.

4. M. Watanabe, H. -B. Sun, S. Juodkazis, T. Takahashi, S. Matsuo, Y. Suzuki, J. Nishii and H. Misawa, "Three-dimensional optical data

storage in vitreous silica," Jpn. J. Appl. Phys. 37, pp. L1527-L1530, 1998.

5. E. N. Glezer, M. Milosavljevic, L. Huang, R. J. Finalay, T.-H. Her, J. P. Callan and E. Mazur, "Three-dimensional optical storage inside

transparent materials," Opt. Lett. 21, pp. 2023-2025, 1996.

6. K. Miura, J, Qiu, H. Inouye, T. Mitsuyu and K. Hirao. "Photowritten optical waveguide in various glasses with ultrashort pulse

laser," Appl. Phys. Lett. 71, pp. 3329-3331, 1997.

7. H. Ueki, Y. Kawata and S. Kawata, "Three-dimensional optical bit-memory recording and reading with a photorefractive crystal:

analysis and experiment," AppI. Opt. 35, pp. 2457-2465, 1996.

8. Electron-beam etching is now a widely used technique for the fabrication of PhCs, see, for example, extended abstracts of the 601h

autumn meeting. The Japan Society of Applied Physics, Konan, pp. 894-901 , 1999.

9. H.-B. Sun, S. Juodkazis, M. Watanabe, S. Matsuo, H. Misawa and J. Nishii, "Generation and recombination of defects in vitreous silica

induced by irradiation with a near-infrared femtosecond" to be published.

10. B. C. Stuart, M. D. Feit, A. M. Rubenchik, B. W. Shore and M. D. Perry, "Laser-induced damages in dielectrics with nanosecond to

subpicosecond pulses," Phys. Rev. Lett. 74, pp. 2248-2252, 1995.

11. See Articles in "Energy-beam-solid interactions and transient thermal processing," Vol. 35, Pittsburgh: Materials Research Society,

ed. by D. K. Biegelsen, G. A. Rozgonyi, C. V. Shank, 1985.

141


c661 'ZOO6-668 dd 'q AOJ S/1j spWI ruoioqd jo /LAI1DOUOJ E1q pu 'PO)ES N O c661 'fN UOOZUUd 'SSaId UOPDUUd zuoioi,j,, UUIM I °PLAI .1 sojnodouuiof f 61

t661 'cP d j, IdO pOAJ f UOpMO AJ J Ut U[jMOU d f 81

sa1n1Jo111 pirn 'uipoooid sup ui 'sioiu pijos ut uoisoidxoioiw iosij puozsonuoj JoMod-qiq qiM pAoL1p soDiw

DLUOOq,, 'is . pu onsi,\ '1MU/\J 1A '11X •A UflS •H- H '/flUEUAUO3 SO1U 'SJq JO douo3 qi jo uonduDsop i io LI

9661 'ILI3OIPJOU "°''1N 'clfl°A 1 -'S ISV OIVN IflO)POS j q P 's1zwuI dvX puvq 31110101'd ut swuouw di puiq 3LUOwq,, '°H IA! N pu SiIflO)IflOS IA! J 'S1i5 IA! 'uqj j J 'SMSt H 91

6661 '6c6— L6 dd 'l'L 19 Scq IddV 'US SflOOJflA UI JOWOW 10S P-P JO S1UI

DuO3souiwnjooqd pu uoEssLwsUIJj,, 'OjOU) j puI 1Mij,\J IAJ 'MSAj H OflSPAJ S 'US H-H 'SiZpOflf S 'IA! 'ci

D'86I

'3U1 SSOJd 'otj u 'ouJy 'j ' cq po 'doosoipods iosij ioqsi.njn cq poqoid JOPflpUOOWO5,, UI 'UOqDoA v I 'Vt

cL6 I '1°A 1\0N OUL 'suournoqqn JOAOQ siopnpuooiuos in sossoooid 'oAo)1ud '1 '1 ' I

'9661 'OOipU5 '3iWopIoV uoUOlUOUOqd osjnd iosj ioqsiJç,, in 'qdopn 'i 'PA pui spi(j ' 'f i

ZfL


Documents

Photonic lattices achieved with high-power femtosecond laser … · 2016-11-21 · Photonic lattices achieved with high-power femtosecond laser microexplosion in transparent solid