Transient Light-Induced Refractive Index Change Made by LaserMicrofabrication in Nitroaniline-Doped PMMA Film

Kazuhiko YAMASAKI*, Saulius JUODKAZIS**, Mitsuru WATANABE*, Shigeki MATSUO*, Kenji*** *** *KAMADA , Koji OHTA , and Hiroaki MISAWA

*Graduate School of Engineering, The University of Tokushima, 2-1 Minamijyosanjima, Tokushima770-8506, Japan

E-mail: misawaeco.tokushima-u.ac.jp (Corresponding author: H iroaki M isawa)**SvBL The University oflokushima, 2-1 Minamijyosanjima, Tokushima 770-8506, Japan

***Osaka National Research Institute, I -8-3 1 Midorigaoka, Ikeda, Osaka 563-8577,Japan

Keywords: PMMA, nonlinear properties, optical memory, laser microfabrication, light-induceddamage threshold

We report the observation of high light-induced change in refractive index (recognizable by observation in conventionalmicroscope) in PMMA film doped with an optically non-linear dye 2-nitroaniline (N02(C6H4)NH2 abbreviated as 2NA). Theoptically altered micrometer-sized regions were fabricated by single-shot irradiation of I 20 fs laser pulses into doped PMMAfilm using high numerical aperture I .3 and high magnification xlOO objective lens. The doping of films can be achieved in awide range of 2NA concentrations (up to 40 wt%) without precipitation. This allows to control a storage time of an opticallyaltered region up to one month by the adjusting the energy of the femtosecond (fs) recording pulse at 800 nm. Typicalrecording energy was I 0-80 nJ/pulse at the point of irradiation. Total recovery of transmission of the PMMAI2NA film wasconfirmed by optical transmission measurements in microscope. The light induced damage threshold (LIDT) (for permanentdamage) was increased more than by 4 times (up to 40 nJ/pulse) when 2NA doping were ca. 13 wt%. While the LIDT fortransient damage was decreased by I .5-2 times. Total optical recovery was observed single exponential with decay time of ca.0.5-1 mm for moderate irradiation intensities (O.IxLIDT of permanent damage). The damage induced with at the higherintensities lasts up to month, but the recovery was not total (residual transmission changes were observable).

The phenomenon can be applied for the optical memory, photonic crystal, and micro-mechanical applications. Theunderlying mechanism of the phenomenon is discussed in terms of anelastic a- and /3-relaxation (polymer backbone and sidechain relaxation, respectively).

1. Introduction

Polymethylmethacrylate (PMMA) is gaining an interest inthe field of optical applications due to its high opticaltransmission in visible, easy dye-doping, preparation andprocessing including the laser ablation and controllablephoto-modification. A three-dimensional (3-D) opticalmemory' and an optical waveguide2 have been demonstratedrecently. On the part of optical 3-D memory variousapproaches have been proposed: a bacteriorhodopsinprotein-based two-photon absorption (TPA) inducedwrite/read memory,3 femtosecond (fs) pulse induced micro-explosion recording,4 photorefractive,5 photochromic6 andphotopolymer7 materials have been used as recording media.

Here we demonstrate a novel mechanism of 3-Drecording, which is transient with controllable duration ofrecording (I O-l O s) made by a single pulse fs-irradiation

of PMMA film doped with 2-nitroaniline, N02(C6H4)NH2.The dye was responsible for TPA of 795 nm illuminationand lowered the light induced damage threshold (LIDT) ofPMMA. The recovery of optical transmission was explainedin terms of viscoelastic a- and /3-relaxations of PMMA.

2. Experimental

2-nitroaniline (2NA) was introduced into chloroformsolution of PMMA ( I 5 wt.%) then casted over a microscopecover glass to form a film, which was used as recordingmedia. The 2NA doping up to 40 wt.% was precipitationfree. Samples were dried in a desiccator at a vacuum of '—10mm Hg for 24 h. Eventual thickness of the film was ca. 150-300 .tm. It was crack free and stable on the yearly scale evenafter the optical recording. 3-D recording was made by laser

First International Symposium on Laser Precision Microfabrication, Isamu Miyamoto, Koji Sugioka,Thomas W. Sigmon, Editors, Proceedings of SPIE Vol. 4088 (2000) 2000 SPIE . 0277-786X/OO/$1 5.00 51

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

irradiation through an oil-immersion objective (x100magnification and numerical aperture. IsA l.3)through thecover glass in order to avoid any oil contact with PMMAfilm. An irradiation spot (lateral size) can be evaluated asdiffraction-limited spot size (diameter of Airy disk) dI .22)JNA 746 nm and the corresponding axial size was:2nA/N42 = 701 nm. where A 795 nm is the wavelength offabrication laser pulse, n is the refractive index at focal point

(flPMMA 1.49 at 600-800 nm). The sample was translated inthe focal plane during irradiation. Every single bit seen as anoptically-altered transmission region was recorded in asingle shot of 120 fs duration.

3. Results

Permanent optical damage (bit) can be recorded in PMMAfilm as we reported earlier,' when the pulse energy densityis higher than LIDT of material. For undoped PMMA, wefound it ca. 2.3 J!cm (10 nJ) for 795 nm, 120 l's pulses. Thisdetermination was made from an optical transmission, whenthe bit is recognizable by observation in 100-times

magnification. NA I .35 objective. The recovery of thetransmission of irradiated spot was observed (Fig. I). whenthe pulse energy was smaller that IxLIDT and PMMA filmwas doped with 2NA. The sequence of snap shots in Fig. Iallow to trace the recovery of transmission of selected bitsA. B and C on a time scale on tens-of-seconds as depicted inFig. 2. The transmission at the center of the bits showedsingle-exponential recovery with time constant TR = 10-30 s(Fig. 2(a)). More complicated recovery can be observedwhen the average transmission on a square of lxi jim2 wasplotted over time (Fig. 2(b)). It can be seen from Fig. I thatthe bit was enlarging while its transmission was decaying.The increase in average transmission up to 10-15 % wasobserved at irradiated spot immediately after fabrication inPMMA2NA-32 wt.% film, when the pulse energy was 12

Fig. 1. Time series (a-c) of optical transmission recovery afterirradiation of 15 ni (3.4 i/cm) pulses (l2Ofs at 795nm). Thefilm was PMMA2NA-32wt.°/o (307 jim-thick). Scale bar 7jim.

Tan(s)Fig. 2. Time decay of maximum (background subtracted (a))and average (b) bit transmission. (a) the hits A, B, and Ccorresponds to those in Fig. I. (b) averaging was made on 40-

by-4() pixels. BKG marks the background transmission.

nJ (in terms of' energy density 2.7 i/cm2, the upper limit ofintensity can be evaluated using 120 fs pulse duration,2.3xlO'3 W/cm2. since the effect of pulse spreading mainlyin the objective lens is unknown).

The optical absorption spectrum was found withoutapparent modifications after the recovery (Fig. 3). Thestrong absorption band at Ca. 400 nm is due to 2NAabsorption. The recovery became faster with increasingconcentration of 2NA in the PMMA film (Fig. 4). Thethreshold pulse energy was found slightly decreasing withthe increasing density of 2NA. The recovery time at theLIDT was Ca. Is. what allowed to repeat writing/re-writingcycles at 1 Hz pulse repetition rate and no films alterationwas observed after all 3 h procedure. This implied that theexplanation of the results lies in the light-induced chargeand dielectric relaxation.

1.0

500 1000 1500 2000 2500 3000

Wavelength (nm)

Fig. 3. Absorbtion spectra of fresh PMMAI2NA-3Owt.° film(dashed line) and during the recovery (r5 1-3 h) of opticaldamage (shifted upwards by 0.25). Thickness of film was157 jim.

(I)

C

-D

C0(I)U)

EU)C

I—

200

100

10

1

1.1

1.0

20 25 30

4A ':

35 40 45

—.-— 2

BtNb.: (1))-. A- 1

4A

0.5

a)0Cci:i

0(I,.0

during ri

smai1erY1I1PMMAI2NA0f3OwI/

I---,

52 Proc. SPIE Vol. 4088

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

Fig. 4. Summary of the recovery time vs. 2NA concentrationand the recording pulse energy. Simulated Y Const+aXdependencies depict observed recovery time. Grey markersand dashed line depicts partial recovery of transmission: solidmarkers — permanent optical damage.

4. Discussion

Optical alteration of materials is usually considered in termsof refractive index changes. directly related to the localmass-density modifications or in terms of an absorption ofoptically induced defects, which are, in turn, causing thechanges of refractive index via Kramers-Kronig relation.5Astonishing high refractive index changes (> 5xl0) werereported made by microexplosion in silica and sapphire.Indeed, if one is measuring the diffraction efficiency of agrating written by microexplosions and if the absorption isconsidered negligible the effect of diffraction on thinsinusoidal grating gives overestimated values of refractiveindex changes. The transmission contrast such as in Fig. 2(a)gives the value of ca. 0.02 for the grating written in silica(similar value would be found in PMMA). We shouldproperly address the refraction, absorption and scattering toget insight into transmission image formation in order tounderstand the images such as in Fig. I (identical imagescan be observed in silica, sapphire, TiO-rutile).

Obviously, when the optically altered volume is of thesize comparable with the wavelength of the image forminglight (550 nm mid-wavelength of condensor illumination ina microscope) the light scattering is one of the mostimportant factors. To look at a near-field distribution of thelight passing through a small sphere of decreased andincreased refractive index as compared with the index ofsphere's ambience, we carried out numerical simulationsaccording the code developed by Barber and Hill.' Theprogram allows to calculate the light intensity inside andoutside sphere, when incident plane wave has intensity I.The central part of the bit was considered of decreasedrefractive index and the diameter is given by fabrication spotradius of 373 nm. This can be described as a light scatteringsphere, which relative refractive index is defined by rnfllfl() < I, where n is the refractive index of the void-likesphere and n0 is that of sphere's surrounding. Another

Fig. 5. Near-field intensity distribution of scattering made byspheres of given refractive index and size. Equivalentpresentations of densified shell of a bit (a) and void-like center(b). (c,d) internal light fields inside spheres. The intensity ofincident plane wave is I (polarization and direction ofincidence arc given by arrow in (c.d)).

sphere considered for calculations was of increased index, m> 1, and the diameter was twice larger than that of the void-like sphere's. This simulates the real structure of the hit.which is void-like inside and surrounded by densified shell.In the following part we will show how the employed valuesof rn are justified (m 0.99 for the void-like sphere and iv

1.002 for the higher density shell). The results ofsimulations in Fig. 5 shows, that the center of the spherewith m -> I appears bright centered in transmission (Fig.5(a)), while the void-like sphere should give an oppositecontrast. Also, the larger sphere of m 1 caused largerintensity contrast as compared with that of sphere with ivI. This qualitatively explains the observed transmission of3-D memory written in PMMA, although the real hitstructure was treated simplified as two separate spheres.

To answer the question, what is the refractive index ofdensified and dilated regions of' bit, let us compare ourobserved recovery time of transmission of 1-100 s withcreep compliance data in the experiments of dielectric

Proc. SPIE Vol. 4088 53

-)

0r(I)

>.,c,)ci)

w

Intensity: 0.92-1.06 0.95-1.01

0 10 20 30Concentration of 2NA in PMMA (wt.%)

(a)

:ia)

U)

-0.5 0.0

Distance (p.m)

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

cooling and strain release). Are the amorphous polymersresponding as a glass or as a rubber is the matter of time and

3 The temperature control over the processesconsidered can be expected to follow general dependence ofglass-rubber transition, i.e. the time and temperature areacting equivalently on the strain relaxation in polymers. Fig.2(b) can be explained by glass-rubber transition as well. Theplateau region in the transition (Fig. 6) is whereentanglement ofpolymer coils is relaxing. This process doesnot change the average refractive index as strongly as it doesat the faster $-relaxation.

5. Conclusions

We have demonstrated for the first time the 3-D transientoptical memory in PMMAJ2NA dye doped films. Therecovery time is widely varied by the doping concentrationand fs-laser pulse energy. The mechanism of opticalrecovery, most probably, is caused by an anelastic dielectricrelaxation, mainly /3-relaxation. The phenomenon can findapplications in the micro-systems as fluid pumps (driven bystrain), optically controlled capacitance (driven by dielectricconstant), and temperature sensors (temperature control overtransmission). The phenomenon can be expected to be foundin other polymers as well.

References

I . K. Yamasaki, S. Juodkazis, M. Watanabe, H. Sun, S. Matsuo,and H. Misawa, Appi. Phys. Lett. 76, (2000) 1000.

2. C. Wochnowski, S. Metev, G. Sepold, App!. Surf Sci. 154-155,(2000) 706.

3. Z. Chen, D. Govender, R. Gross, R. Brige, Biosensors &Bioelectronics I I , xv (1 996); BioSystems 35, (1 995) 145.

4. M. Watanabe, S. Juodkazis, H. Sun, S. Matsuo, H. Misawa, M.Miwa, R. Kaneko, App!. Phys. Let!. 74, (1999) 3957.

5. Y. Kawata, H. Ishitobi, S. Kawata, Opt. Lett. 23, (1 998) 756.6. D. A. Parthenopoulos, P. M. Rentzepis, Science 245, (1989)

843.7. J. H. Strickler, W. W. Webb, Opt. Let!. 16, (1 991) 1780.8. P. R. Herman, R. S. Marjoribanks, A. Oct11, K. Chen, I.

Konovalov, S. Ness, App!. Surf Sci. 154-155, (2000) 577.

(2) 9. E. N. Glezer and E. Mazur, App!. Phys. Lett. 71, (1997) 882.10. P. W. Barber and S. C. Hill, Light scattering by particles:

computational methods (World Scientific, Singapore, 1990).11. N. G. McCrum, B. E. Read, and G. Williams, Anelastic and

dielectric effects in polymeric solids (Dover Publications, Inc.,New York, 1991) ch.8.

12. B. E. Yoldas, App!. Optics 19, (1980) 1425.13. G. Strobl, The physics ofpolymers (Springer, Berlin, 1997).

(1)

PMMA Terminal

: 1.5 flow

:1:

Ig(Time) [Ig(s)]

Fig. 6. Strain relaxation in PMMA (reconstructed from ref.11). 106 is ca. lOdays, l0.— 100 days.

relaxation. In Fig. 6 the master creep curve is plotted (basedon ref. I I ). As it can be seen the recovery of I -1 00 s(plateau region in Fig. 6) corresponds to the 0.7-0.8% strain.Such a strain is already larger than typical limit of elasticityof ca. 0.2% even in ductile polymers. The strain, as ameasure of tV/V, there V is the volume, is directly related tothe refractive index. Yoldas,'2 derived the relation betweenporosity and refractive index, which we can adopt for thedensified/dilated material:

i±5V purou.s— = Strain 1 — _______V n,nce 1

The relation does not depend on the elastic response ofmaterial and can be applied to microexplosion alteredmaterial. By taking Strain 0.8% (plateau in Fig. 6) we findcorresponding refraction index changes of 3.3xl 0'. Whenunperturbed index of film was taken as that of PMMA,namely I .490. The relative refraction index values of mI .002 and 0.998, for the densified and dilated regions can befound. These are the values employed in the modelization oflight scattering (Fig. 5). If the absorption needs to be takeninto account, the numerical simulations are done withcomplex refractive index m (the calculations in Fig. 5 wereobtained without consideration of absorption). Typical resultof absorption on the transmission intensity is the loweredcontrast. The direct evaluation of transmission withoutconsideration of scattering is straight forward I I( I -R)e",where R is reflectance, which for the right angle incidence

is given by: R = (nI)2 ÷ 2 where k = aAJ4ir(n+l)2 +k2

Here, the absorption coefficient can be consideredproportional to the strain, since more material is encounteredin the light pass in denser region and vice versa for dilatedone. When pure relaxation mechanism of transmissionrecovery is considered, the absorption coefficient is a 0cm' . Even then a realistic transmission image can bereconstructed by taking into account light scattering (Fig. 5).

Glass-rubber transition is mainly due to /3-relaxation(also known as Johari-Goldstein), i.e. a side chain kinematicmovement. At high power of irradiation the temperature canbe higher than glass transition of PMMA at Tg 105°C atirradiation spot (then a-relaxation is involved during

54 Proc. SPIE Vol. 4088

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