Open Access
Add to BibSonomyAbstract
An ISO certified laser-induced damage threshold testing method was applied to characterize photopolymers widely used in 3D laser micro/nano-lithography. For the first time, commercial as well as custom made materials, including epoxy based photoresist (SU-8), hybrid organic-inorganic polymers (OrmoComp and SZ2080), thermopolymer (PDMS) and pure acrylate (PMMA), are investigated and directly compared. The presence of photoinitiator molecules within host matrix clearly indicating the relation between damage threshold and absorption of light is revealed. To simulate single- and multiphoton absorption processes optical resistance measurements were carried out at both fundamental (1064 and 1030 nm) and second harmonic (532 and 515 nm) wavelengths with laser pulse duration’s representing nanosecond and femtosecond regimes. Damage morphology differences from post mortal microscopic analysis were used to enrich the discussion about the possible breakdown mechanisms. The obtained characteristic values of damage threshold reveal potential of photopolymers and their possible applications in high power lasers.
© 2014 Optical Society of America
1. Introduction
Advantages of the integrated photonic circuits over the traditional electronic devices stimulate rapid progress in both optics manufacturing technology and material science. Accordingly, new ways of high power micro- and nanooptics component integration are of major importance. As an alternative to traditional lithographic methods, direct laser writing (DLW) based on multi-photon polymerization was proposed for the production of 3D micro/nanostructures in 1997 [1]. Among other technologies such as focused ion beam milling (FIB) based on material removal, electron beam lithography (e-beam) and UV lithography incorporating multi step mask processes, DLW is distinguished for its unique properties to prototype true 3D objects, fast and straightforward process, high structuring accuracy and no need to use masks, molds or vacuum [2]. Typically, the liquid solution of monomers (pre-polymers) is exposed with the laser radiation according to the predefined geometrical model. Tightly focused laser beam (in the range of TW/cm2 intensity at the focal spot) induces cross-linking of the monomers into long polymer chains and the material is selectively solidified. Afterwards, the unexposed material is washed-out during wet chemical development procedure (in the case of negative pre-polymer) and free-standing solid objects are obtained. Moreover, most of the photopolymers used in DLW inherit desired optical properties such as high transmission in the visible (VIS) and near-infrared (NIR) spectral ranges, chemical and mechanical stability. This was exploited for the successful fabrication of the complex-shaped and integrated microoptical components (MOC) via single DLW process [3–5]. Up to known, numerous state of the art examples of refractive and diffractive MOCs were reported [6–13]. Despite of successful proof of principle demonstrations the technology was challenged by several limitations. Much effort has been devoted to overcome low fabrication throughput, for instance by introducing multiple beam continuous flow lithography, and to achieve appropriate surface roughness as the later one plays a critical role in optical performance [14]. Outer shell fabrication [15–17] and diverse voxel’s overlap [18,19] approaches were employed to speed up the fabrication process up to ∼400×. Better than λ/20 surface roughness in the visible spectral range of the MOC’s was achieved after implementation of the various scanning strategies [6, 16, 18, 20]. In addition, self-smoothing [19] and non-local polymerization [21] effects were found to reduce surface roughness. High repeatability, negligible practical deviations from designed shape and 100% fill factor of MOCs was further step leading to practical applications of the 3D microoptics. However, the implementation of the polymeric solid state MOCs into optical circuits is still problematic in terms of used optical power.
At the early stage of the DLW conventional lithographic UV photopolymers were widely exploited: acrylate based optical glue [1], epoxy based photoresin (SU-8) [22] and hybrid organic-inorganic materials (ORMOCERs) [15]. Despite of many attractive features of photopolymers such as high third order nonlinear susceptibility [23], low cost, light weight, ease of processing and simplicity to tailor their properties (e.g. doping), they were found to have weak optical resistance in comparison to traditional glasses and dielectric coatings. As the technology evolved, novel photopolymers were also developed to meet the requirements of DLW technique [24]. However, laser-induced damage threshold (LIDT) performance of the solidified pre-polymers is still unknown and it can be the limiting factor of the polymerized micro/nanodevices in many future high power applications. Thus, DLW processable polymers with higher optical resistance could stimulate new breakthrough in pertinent optics technologies.
Mostly, testing of the MOCs involve investigations of light propagation and diffraction patterns which are obtained under illumination with low power sources such as continuous wave He-Ne or semiconductor lasers [8, 10]. Pulsed lasers were used as well [25]: the combination of 15 ps pulse duration, 50 kHz repetition rate and 532 nm wavelength resulted in rather weak peak intensity ∼0.8 W/cm2 and the LIDT was not reached. On the other hand destructive laser treatment of conventional polymers such as polyimide (PoI), polyethylene terephthalate (PET), polycarbonate (PC) and polymethylmethacrylate (PMMA) were also studied in some ways during the last twenty years [26]. Several wavelength and pulse duration dependent mechanisms were assumed as a possible reasons of polymer destruction [27]. In the case of UV excimer laser ablation it is considered as a purely photochemical process as high energy photons (∼5 – 6.4 eV) directly brakes C – C and C – H bonds without signature of melting [28]. Meanwhile, optical damage at the visible spectrum is most likely caused by the excitation of vibrational modes and is attributed to photothermal process [29]. The same mechanism of laser ablation was also assumed at longer wavelengths (10.6 μm) using CO2 lasers [30]. Multiphoton absorption [31] and incubation [32] effects were observed using ultrashort laser pulses. Nevertheless, there are no clear benchmark on main DLW materials performed under identical experimental conditions and also simulating different regimes of laser induced destruction that could allow direct comparison of them.
Here, we seek to characterize LIDT performance of the solidified polymers thin films which are used as preprocessed materials for the manufacturing of MOCs, preferable for high light intensity applications. We used PMMA as a reference material to relate LIDTs of the new materials in absolute scale and make our data comparable to other publications as it is the mostly investigated polymer for the such kind of studies [27,28,32]. The only difference was in preparation method: PMMA is used as solid bulk material in previous cases while in DLW it is obligatory to use PMMA in its liquid form. The effect of preparation method on LIDT is considered as negligible. Thermo-polymer polydimethylsiloxane (PDMS) was investigated as well as it is widely used for the mass manufacturing of the 3D microstructures using soft-lithography replication technique [33] and as low refractive index substrate [34]. Moreover, recently the structuring at high throughput (∼720 μm3/s) of the PDMS doped with various photoinitiators was demonstrated [35]. A special attention is given to the novel hybrid photopolymer SZ2080 which is specially synthesized for the production of 3D micro/nano-structures by DLW [24]. Lately, it was demonstrated the fabrication of the microstructures out of non-photosensitized photopolymer at tight focusing (∼TW/cm2) and at (pre)breakdown conditions [36]. Here we also considered the possibility to study pure and photosensitezed photopolymers in order to reveal the effect of the photosensitization on the LIDT. Damage morphology differences from post mortal microscopic analysis were used as to enrich the discussion about the possible breakdown mechanisms. Thus, performed research may pave the way for new avenue of practical applications of the photopolymers into 3D optical circuitries, field of nonlinear optics and manufacturing of miniature polymer lasers.
2. Materials and methods
2.1. Preparation of the polymers thin films
One inch diameter and 1 mm thickness soda-lime glass was used as a substrate for the spin-coating of photopolymers. Standart cleaning procedure was applied to them before covering with materials. Firstly, substrates were wiped with acetone to remove impurities and afterwards immersed in a 4:1:2 solution of H2O (water), NH4OH (ammonium hydroxide) and H2O2 (hydrogen peroxide) at 80 °C for 20 min to remove contaminants. In a second step, substrates were immersed into isopropanol and brought to ultrasonic bath for the 5 min. In the last step, the residuals of the isopropanol were blown with compressed air flow.
Five different commercial and laboratory synthesized materials were used for the LIDT measurements: epoxy based photoresin SU-8 2015 (MicroChem Corp.), hybrid photopolymers OrmoComp (Microresist Technology GmbH) and SZ2080 (FORTH, Greece), thermo polymer PDMS (Sylgard 184, Down Corning) and pure acrylate PMMA (Microresist Technology GmbH). 2-Benzyl-2-(dimethylamino)-1-[4-(4-morpholinyl) phenyl]-1-butanone (Sigma Aldrich) at 2 wt. % concentration was used as a photoinitiator (PI) for the SZ2080. Doping of the polymers with PI is commonly used in order to increase photosensitivity and efficiency of the polymerization. Summarized data of the photopolymers preparation for the LIDT measurements are presented in Table 1.
Table 1. Preparation of the photopolymers thin films (v – spin-coating velocity, t – time, T – temperature, λ – wavelength).
SU-8 is an epoxy based based photoresist which is dissolved in a organic solvent (cyclopentanone) with photoacid generator (triarylsulfonium salt). A single molecule of SU-8 consists of eight reactive epoxy groups with the back-bone formed from benzene rings. Cationic photopolymerization of SU-8 is followed by a strong acid generation during UV light exposure and subsequent cross-linking at post-bake procedure.
OrmoComp is a hybrid organic-inorganic polymer synthesized during sol-gel process and depends to ORMOCER material class. The back-bone of it is formed from inorganic silicate network ≡ Si – O – Si ≡ with the single bonds between atoms. Organo(alkoxy)silanes are used as an organic crosslinking units in order to functionalize material with polymerizable groups, namely methacryloxy group. Also, OrmoComp is additionally photosensitized for the UV light exposure by the manufacturer.
SZ2080 is a hybrid organic-inorganic polymer which chemical structure is similar to Ormo-Comp, but depends to ORMOSILs materials class. The difference of these polymers lies in the structure of inorganic network, i.e. some Si atoms are replaced with Zr heteroatoms for the SZ2080. It this case, the back-bone is formed from ≡ Si – O – Zr ≡ inorganic network. In our study we used a non-photosensitized SZ2080 and doped with the photoinitiator (SZ2080 + PI) in order to reveal the influence of the photosensitization.
Polydimethylsiloxane (PDMS) is a silicon based organic thermo-polymer with chemical formula CH3 [Si(CH3)2O]nSi(CH3)3 consisting of single bonds between atoms. A curing agent and PDMS prepolymer were mixed at 1:10 ratio and later mechanically stirred for 10 min. To remove air bubbles occurring during mixing, the prepolymer mixture was degassed in a desiccator with a vacuum pump. Cross-linking of the polymer begins since the mixing of liquid prepolymer and curing agent and is stepped-up during heating.
Poly(methyl methacrylate) (PMMA) is a positive tone organic polymer which chemical formula is (C5O2H8)n. Atoms along back-bone of the polymer are connected with single bonds. Upon exposure, long polymer chains of PMMA are decomposed via chemical bond breaking to lower molecular weight fragments which are later dissolved in organic solvent.
High spin-coating velocities varying from 3000 to 6000 rpm were applied to ensure uniformity of the polymers thickness according to the recommendations of the suppliers. After the spin-coating, all samples except thermopolymer PDMS were cured under UV lamp (NU-4 KL, Benda Laborgeraete) with an emission centered around 365 nm while PDMS was cured on a hot plate for 1 hour at 100 °C. 254 nm wavelength and 18× longer time was used to cure pure SZ2080 in comparison to photosensitized SZ2080. To verify the quality of photopolymerized films, the samples were immersed into the appropriate developer bath. No delamination of the thin films was observed and it confirmed that they are completely polymerized.
2.2. Characterization of the polymers thin films
The spectral distribution of transmittance T(λ) and reflectance R(λ) was measured at close to a normal incidence in the wavelength range of 190–1100 nm with a 2 nm spectral resolution using double beam spectrophotometer (Specord® Plus 250, Analytik Jena AG). 24×24 mm2 UV fused silica instead of soda lime glasses were used as a substrates for the measurements as it does not cut off the transparency window of the polymers.
Physical thickness (l) of the photopolymers thin films was determined using optical profilometer (PLμ 2300, Sensofar). Either the height of the photopolymerized waveguides fabricated via DLW technique (for SZ2080+PI, OrmoComp and SU-8), scratch made with diamond wedge scriber (for SZ2080 and PDMS) or laser ablated region (for PMMA) was measured in a five different places to determine the height’s uniformity of the thin films.
Characterization of the optical resistance by means of LIDT has been carried out at Laser Research Center of Vilnius University by two automated in-house build test stations [37]. Different (IR and VIS) wavelengths as well as pulse durations (nanosecond (ns) to femtosecond (fs)) were used for testing. In case of ns pulses, LIDT test bench is based on a single longitudinal mode injection seeded Nd:YAG laser (“SpitLight Hybrid”, Innolas Laser GmbH) delivering linearly polarized pulses of 11 ns duration (FWHM). Its frequency can be doubled by built in harmonic generator. The repetition rate of optical pulses was set to 50 Hz. In case of fs pulses, a diode pumped Yb:KGW oscillator-amplifier system (“Pharos”, Light Conversion Ltd) has been used with similar LIDT test bench. The repetition rate of optical pulses can be tuned by built-in Pockels cell in range of 1 – 200 kHz. Measurements have been carried out under repetition rate of 50 kHz as the the effect of repetition rate was not observed in the range from 50 Hz to 50 kHz with fs pulses. LIDT testing procedures were performed according to ISO 21254 standard [39]. Fluence was adjusted with a motorized attenuator consisting of half-wave plate and polarizer. The damage threshold has been estimated for the single pulse per site (1-on-1 regime) and a series of pulses with constant energy per site (S-on-1 regime, S = 1000). The on-line damage detection system was based on optical scattering from irradiated sample surface. A photo diode sensor was used to track laser-induced surface changes. Damage detecting optical scattering signal was recorder for every pulse. The off-line inspection of irradiated sites was performed by Nomarski microscopy (BX51, Olympus) after irradiation exposure. Here, criterion of damage is any visible modifications that can be seen with it. LIDT have been calculated using non-linear fitting procedure based on maximum-likelihood approach described elsewhere in details [38]. Estimated error bars correspond to a confidence level of 95 %. Distance between neighboring exposition points was fixed to 500 μm in order to avoid re-deposition of the damaged material on the surrounding surface, in advance. In all cases, damage threshold of the substrate was higher than the tested polymers and it did not influenced the results. Testing conditions are summarized in Table 2.
Table 2. LIDT testing conditions (τ – pulse duration, λ – laser wavelength, f – repetition rate and ω – beam diameter measured at 1/e2 fluence level at target plane).
Single-shot (1-on-1) testing regime represents more fundamental material properties responsible for initiation laser-induced damage while multi-shot (1000-on-1) testing is used to evaluate fatigue effects attributed to defect state generation (bond braking) and thermal energy accumulation. As in practical applications fatigue effects are important issue, multi-shot LIDT is used for the benchmarking of different polymers. Single shot LIDT values will be considered discussing optical breakdown phenomena in the polymers.
3. Experiment and results
3.1. Spectrophotometric results
Figure 1 shows the results of the measured transmittance spectra of the investigated polymers thin films. Preparation of the samples was identical as and for samples used in LIDT measurements except that the UV fused silica (UVFS) was used as a substrate instead of soda-lime glass. All measurements were performed at the same experimental conditions. As one can observe, all photopolymers inherits high transmittance in the VIS and NIR spectral ranges while in UV transmission differs significantly. The highest transmittance in UV spectral range is characteristic for the PDMS and PMMA polymers (they start to absorb in UV-C) while the lowest values are for the SU-8 and SZ2080 + PI (at the edge of the VIS and UV-A). Transmittance of the OrmoComp and pure SZ2080 lies in between. Photosensitization of the SZ2080 photopolymer with the photoinitiator molecules leads to the increased absorption in the near UV spectral range in comparison to the pure SZ2080 as the absorption edge shifts to the longer wavelength side of the spectrum.
Fig. 1 Measured transmittance of the photopolymers thin films (solid line shows the transmittance of the soda-lime glass substrate used for the LIDT measurements whereas shaded area shows the transmittance of the UV FS substrate used for the spectrophotometric measurements).
Polymers thin films optical band-gap energy (Eg) was characterized using classical methods even though their are amorphous molecular solids which do not have long-range order. Eg was found from the absorption coefficient (α) dependence on the photon energy (hν) according to the Cody [40] proposed empirical model for amorphous materials. To determine Eg, we have plotted (α/hν)1/2 as a function of hν and linearly extrapolated to zero excluding the points in the low energy tail. Absorption coefficients of the photopolymers were derived from the transmittance and reflectance measurements using fundamental expression:
where l is a physical thickness of the photopolymer’s thin film. The obtained values of the thickness (l) and optical energy band-gap according to Cody (Eg) model are given in Table 3. Comparison of the Eg values of the various polymers presented in Table 3 indicates, that the highest Eg are characteristic for PDMS and PMMA polymers while the lowest – for SU-8. Similar optical band-gap values are obtained for pure SZ2080 and OrmoComp as one would expect since both materials are hybrid photopolymers and optical properties of them differs slightly. The presence of the PI molecules within the host the SZ2080 photopolymer lowers Eg in comparison to non-photosynthesized SZ2080. Therefore, it is expected that the LIDT of the photosensitized SZ2080 should be smaller than pure SZ2080. Summarized data of the photopolymers physical and optical are presented in Table 3.Table 3. Physical and optical properties of the polymers thin films: l – thickness, Tg – glass-liquid transition temperature, Tm – melting temperature, λT – wavelength at which transmittance is over 85 %, n – refractive index at 632.8 nm wavelength (data taken from polymers manufactures data sheets), α – absorption coefficient at 532 nm wavelength and Eg – optical band-gap energy.
3.2. Damage performance of the polymers at nanosecond pulse duration regime
Experimental values of the LIDT for the various photopolymers, irradiated with nanosecond pulse duration laser radiation, are shown in Fig. 2. For the direct comparison of the wavelengths and number of pulses influence to laser-induced damage, results are plotted on the same graph for the fundamental (1064 nm) and second (532 nm) harmonics and at single (1-on-1) and multi-shot (S-on-1) regimes. Obtained LIDT values of polymers ranges from ∼8 J/cm2 to ∼25 J/cm2 at single-shot regime and from ∼0.3 J/cm2 to ∼15 J/cm2 at multi-shot regime. Specific values of the damage threshold are listed in the Table 4.
Fig. 2 A histogram plot showing the LIDT of the photopolymers thin films at ns regime (AOI = 0 deg, f = 50 Hz, τ = 11 ns and 6.2 ns, ω = 250.2 ± 10.0 μm and 133.5 ± 4.6 μm respectively for the first (1064 nm) and second (532 nm) harmonics).
Table 4. LIDT values (J/cm2) of the investigated polymers.
It is evident that non-photosensitized hybrid photopolymer SZ2080 exhibits the highest damage threshold at both wavelength and multi-shot regime in comparison to all other. However, we have found that doping of the SZ2080 with the photoinitiator molecules leads to the decreased damage threshold by almost one order of magnitude from 13.67 J/cm2 to 1.44 J/cm2 for the second harmonics. Interestingly, the effect of the PI is much smaller at the fundamental harmonic. LIDT values of the pure and photosensitized SZ2080 are equal to 14.73 J/cm2 and 10.63 J/cm2, respectively. Notwithstanding different threshold values, damage morphology after single laser pulse shown in Fig. 3 (green framed boxes) reveals that damage mechanism is similar in both cases. The material is completely removed and clear boundaries having regular circular geometrical shape are apparent. Such features are typical for delamination which is caused due to high tensile strengths originating from melting by intense plasma during laser exposure.
Fig. 3 Nomarski microscope images of the laser-induced damage morphology at single-shot regime (a) and damage profile of the PDMS at fs pulse regime and 515 nm wavelength along dashed line in asterisk symbol marked picture (b). Colored and numbered framed boxes highlight different types of morphology: green (1) – catastrophic damage: delamination and coating removal, blue (2) – absorption centers induced breaks, red (3) – heat affected damage: increase of refractive index or thin film height and black (4) – surface swelling and evaporation. Note the difference of scales at ns and fs pulse regimes. Numbered areas in (b) picture shows non-affected zone of the polymer after laser irradiation (1), region of surface swelling (2) and concave pit formed during evaporation of few hundreds nanometers of top polymer layer (3).
One would expect that LIDT values of the OrmoComp should be similar to SZ2080 as both materials depends to the organic-inorganic photopolymers class. However, even in comparison to SZ2080 + PI, damage threshold of OrmoComp decreases by approximately 3 and 8 times upon 532 nm and 1064 nm wavelengths and multi-shot regime. Damage morphology of the OrmoComp changes either. At 532 nm wavelength several isolated damage spots can be observed which result from the decreased optical band-gap energy due to transitional states between valence and conduction bands introduced by local absorption centres (blue framed box in Fig. 3). At 1064 nm we have noticed the change in damage morphology as well. Instead of multiple damage points, the circular rainbow pattern in laser exposed region is clearly apparent (red framed box in Fig. 3). Discoloration effect is caused due to alteration either of refractive index or physical thickness of the film since absorption centers are inactive at this wavelength. We assume that it is a consequence of the light absorption throughout entire laser exposed area which indicates heat affected zone (HAZ).
Furthermore, it is evident that LIDT value of the epoxy based photoresin SU-8 is one of the lowest at multi-shot regime. The lowest optical band-gap energy, thus the highest absorption coefficient, can cause such results. Besides, damage morphology of the SU-8 at the first harmonics (red colored frame box in Fig. 3) is similar to the one of OrmoComp at the same wavelength. As it was explained previously, it can be attributed to the HAZ.
Qualitative analysis of the PMMA morphology indicates that damage is related to localized inclusions – absorption centers embedded in spin-coated polymer film as and for OrmoComp at 532 nm wavelength. During rapid heating with the laser beam at long pulses, inclusions tend to explode due to higher absorption than the surrounding polymer. Nevertheless, LIDT values of the PMMA are still relatively high in comparison to OrmoComp and SU-8. Similar LIDT values are noticeable for the thermo-polymer PDMS at multi-shot regime as well, accordingly – 3.51 J/cm2 and 5.50 J/cm2. Likely, it is the results of the highest optical band-gap energy.
3.3. Damage performance of the polymers at femtosecond pulse duration regime
Measurements of the damage threshold fluence of polymers at fs pulse duration laser radiation as a function of wavelengths and number of pulses are shown in Fig. 4 and summarized data are provided in the Table 4. The LIDT value decrease by one order of magnitude in comparison to results at ns pulses is evident for all materials. Also, negligible scattering of the LIDT values of the polymers at multi-shot regime and 515 nm wavelength is noticeable except for pure and photosensitized SZ2080. In comparison, LIDT of OrmoComp, SU-8, PDMS and PMMA is equal to 0.1 J/cm2 while for SZ2080 and SZ2080 + PI – 0.13 J/cm2 and 0.06 J/cm2. The photosensitization effect of the SZ2080 is clearly pronounced at the second harmonics while it has as no considerable influence at the fundamental wavelength. In the case of 515 nm wavelength, damage threshold of SZ2080 + PI decreases by a factor of 2. Obtained results of the doping SZ2080 photopolymer with photoinitiator prove the same tendency noticed at ns regime. Consequently, LIDT measurements of the photosensitized and pure polymer indicate that photoinitiator within polymer matrix lowers optical resistance to the laser radiation and limits their applicabilities in high power optical and laser systems.
Fig. 4 A histogram plot showing the LIDT values of the photopolymers thin films at fs regime (AOI = 0 deg, f = 50 kHz, τ = 343 fs, ω = 65.0 ± 0.2 μm and 46.5 ± 0.2 μm respectively for the first (1030 nm) and second (515 nm) harmonics).
Slightly broader scattering of the LIDT fluencies from 0.17 J/cm2 to 0.57 J/cm2 are apparent at 1030 nm wavelength. SZ2080, PDMS and PMMA have the highest LIDT (0.57 J/cm2, 0.55 J/cm2 and 0.57 J/cm2, respectively), SU-8 has the lowest (0.17 J/cm2), whereas damage threshold of SZ2080 + PI and OrmoComp lies in between (0.49 J/cm2 and 0.33 J/cm2, respectively). Damage morphology of the polymers thin films is distinctive as well. Elliptical laser radiation affected zones shown in Fig. 3 (black framed boxes) are observed for the OrmoComp, SU-8 and PDMS at both wavelengths. Geometrical shape of the damage profile is caused due to slightly elliptical laser beam. However, it does not influenced LIDT value. In all cases, the dip and blurred rim around it can be seen. Likely, the rim is formed due to surface swelling and the dip is formed after laser ablation. Material is evaporated in the central part and melted around it after single laser pulse. During post-exposure, melted polymer re-solidifies and a rim with a smooth surface is formed. Although, no debris around laser exposed region can be detected. For the sake of clarity, crater profile of the one typical morphology (PDMS at 515 nm wavelength) was measured with the optical profilometer and it is shown in Fig. 3(b). Three zones can be distinguished: (1) non-affected area upon laser radiation, (2) region of slight surface swelling and (3) few hundreds nanometers evaporation of the polymers upper layer. Obtained results confirm that optical damage of the OrmoComp, SU-8 and PDMS is associated with strong surface absorption which leads to the surface ablation. In the case of SZ2080 and SZ2080 + PI, tiny spots with a diameter of few micrometers are apparent. Such type of damage can be attributed for optical breakdown due to photoionization. Slightly larger damage spots and chromatic pattern is observed of the PMMA morphology. It suggests, that probably “femtosecond melting” is a mechanism responsible of the optical breakdown of the PMMA.
4. Discussion
4.1. Laser-induced damage mechanism at ns pulse duration regime
Polymers of this study are much more complex organic molecular structures in comparison to inorganic materials such as fused silica, laser crystals and dielectric coatings. Therefore, the discussion of the LIDT mechanism of the polymers at ns regime requires reconsideration of their optical properties. Besides to the electronic states, vibrational and rotational states exists next to them for the organic compounds due to atoms possibility to vibrate and rotate around their chemical bounds in a molecule [41]. Rotation of the atom in a solid state is almost impossible and its influence to the optical transitions can be neglected. In a simplified model of the electronic-vibrational transitions in the polymers, the electrons are promoted to the excited vibrational state from the electronic ground state after the absorption of visible or ultraviolet light photon. Electron in the excited vibrational state inherits excess energy which is released during non-radiative relaxation to the excited electronic state via collisions with neighboring molecules [41]. Released energy is a source of the internal temperature rise and as a result, the polymer is heated. Subsequently, it causes melting, formation of mechanical stress and evaporation of the polymer. Thereby, laser induced damage of the polymers at nanosecond regime is thermal in nature. The latter reasonings of the polymers degradation concepts at ns pulse duration regime are supported by the inorganic solids LIDT examination on the various pulse lengths. It was demonstrated that at long pulse regime (τ > 20 ps: ns regime) LIDT dependence on the pulse duration scales as [42]. It is a consequence of the electron kinetic energy transfer to the lattice and heat diffusion during the laser pulse. Moreover, obtained damage morphology of the photopolymers shown in Fig. 3 clearly indicates the traces of the heat affected zones. However, what determines the distinct values of the LIDT of the investigated polymers?
One can assume direct dependence of the LIDT on their thermal properties (glass-liquid transition (Tg) and/or melting (Tm) temperature) as the mechanical and optical properties of them changes at these points. However, LIDT can not be linked to the thermal properties of the examined photopolymers as the Tg and Tm values are applicable only for the SU-8 and PMMA (see Table 3). Even more, half of the investigated materials are duromers – polymers which do not have glass-liquid transition and melting temperatures.
For a further understanding of the polymers degradation lets consider their chemical structure. Generally, polymers can be classified into two main groups: saturated and conjugated compounds [41]. The difference of these two groups lies in the chemical bond type between carbon atoms along polymer backbone, i.e. whether it is single (saturated) or alternating single-double (conjugated). In the first case, the valence electrons are firmly localized next to adjacent atoms and optical transitions of such compounds are possible under irradiation with UV light. Therefore, saturated polymers are colorless, they do not absorb in the visible spectrum and their optical properties are not significantly different to those of optical glasses. At the second case, the valence electrons are delocalized, they start to absorb in the visible spectrum and their binding energy is lower in comparison to saturated polymers [41].
In our case, PMMA and PDMS can be assigned to the saturated polymers class as the chemical bond type between carbon atoms in a backbone is a single whereas SU-8 belongs to conjugated polymer class due to electron delocalization in benzene rings. Attribution of the SZ2080, SZ2080 + PI and OrmoComp for the any one class is more complicated. It is likely that non-photosensitized SZ2080 belongs to the saturated polymers class [24]. However, doping of the polymers with the photoinitiators, which are typically conjugated compounds due to presence of the benzene rings in their chemical structure [43], changes their optical properties significantly. Therefore, we can assume that SZ2080 + PI and OrmoComp is a mixture of the saturated and conjugated compounds. Even though, the concentration of the photoinitiator is relatively low (commonly 0.2 – 2 % [2]), it is sufficient to alter polymers respond to the laser radiation.
Optical transmittance and band-gap energy of the polymers correlate with the properties of the conjugated and saturated polymers, i.e. the highest Eg is characteristic for PDMS and PMMA and the lowest Eg – for the SU-8 and SZ2080 + PI. Laser induced damage measurements at single-shot regime and 532 nm indicate that the highest LIDT is characteristic for saturated polymers (PDMS, PMMA and SZ2080) and the lowest – for conjugated polymers (SZ2080 + PI, OrmoComp and SU-8). The number of alternating single-double bonds is the highest for the SU-8 in comparison to OrmoComp and SZ2080 + PI and it explains the lowest damage threshold of SU-8. Such LIDT correlation on the polymers type can be explained by the fact that electron binding energy is higher for the saturated compounds than for the conjugated. It means that lower laser fluence is necessary in order to break chemical bond of the conjugated polymers in comparison to saturated.
4.2. Laser-induced damage mechanism at fs pulse duration regime
In the last decade it was shown that simultaneous absorption of the n photons with the sum energy equal to Eg or higher is one of the leading mechanisms of optical breakdown at ultra-short pulses of dielectric layers [44, 45]. This idea is supported by the evidence of the empirical data of LIDT cascade dependence on the optical band-gap energy. Abrupt increase of the damage threshold was observed at the point where n + 1 instead of n photons were required for the absorption. In order to verify this hypothesis for the polymers we have plotted LIDT as a function of Eg. Results at 515 nm and 1030 nm wavelength both at single and multi-shot regimes are presented in Fig. 5(a) and Fig. 5(b), respectively. Possible transitions from n to n + 1 were calculated according to n = [(Ege)/(ωh̄)] + 1 and are indicated in different colors in Fig. 5. An increase of the damage threshold value from ∼0.65 J/cm2 to ∼0.9 J/cm2 can be observed between n = 2 and n = 3 cases at single-shot regime in Fig. 5(a). This reveals that multiphoton absorption is probably a dominant process of the optical breakdown as the higher intensity is necessary for the absorption of 3 photons instead of 2. Meanwhile optical damage at multishot regime is influenced by either thermal accumulation or defect incubation effects as LIDT values decreases up to ∼ 0.1 J/cm2 and no cascade behavior can be observed [46].
Fig. 5 LIDT dependence on polymers optical band-gap energy at 515 nm (a) and 1030 nm (b) wavelength after fs pulse duration laser irradiation. Dashed blue line in the graph (a) is a guideline for the eye which indicates the increase of the LIDT at the point where n + 1 instead of n photons are required for the absorption. Numbers in brackets depict the number of photopolymer: 1 – SZ2080, 2 – SZ2080 + PI, 3 – OrmoComp, 4 – SU-8, 5 – PDMS and 6 – PMMA.
In the case of measurements at 1030 nm wavelength we did not noticed any correlation between LIDT and optical band-gap energy of the polymers in any cases (Fig. 5(b)). The probability of n photons absorption scales as P ∝ Inσnτ, where I is intensity of the light, σn – absorption cross-section of n photons and τ – pulse duration [47]. It becomes negligible small at longer wavelength as higher number of photons are necessary for the transition. Together with the absence of LIDT dependence on Eg, it implies that multiphoton absorption is no longer predominant mechanism of optical damage at 1030 nm wavelength and likely that other phenomenon have bigger influence. Generally, it is assumed that optical breakdown of the transparent dielectrics is caused due to tunneling or avalanche photoionization besides to multiphoton absorption as electron density reaches the plasma critical density (ncr ∼ 1021 cm−3) [48]. The same theory as a possible damage mechanism of the polymers at ultrashort laser pulse laser radiation was considered as well [49]. In our case, tunneling ionization can be neglected as it is expected to dominate when Keldysh parameter γ < 1.5, whereas for all investigated polymers it is equal to ∼3.5 – 5.6 and ∼11.4 – 13.1 for the first and second harmonics, respectively. However, further estimation of the optical breakdown of distinct photopolymers requires a more detailed study of the material properties (e.g. absorption cross-section, molar mass, etc.) which is already out of scope of this research.
In comparison to the previous studies of the PMMA ablation threshold determination, we have obtained similar results. S. Baudach et al. reported ablation threshold values equal to 2.6 J/cm2 and 0.6 J/cm2 at single and multi-shot (N = 100) regimes using 800 nm wavelength and 150 fs pulse duration laser radiation [32]. In the case of shorter pulse durations (30 fs), it was found that ablation threshold is equal to 0.9 J/cm2 at 1-on-1 regime and it drops-off by one order of magnitude at 1000-on-1 regime [51].
4.3. Comparison of the exposure parameters for the DLW and LIDT
Let us consider the typical exposure parameters of a DLW process and LIDT of the same materials in a sense of the employable intensities. For a direct comparison, fs pulse regime and SZ2080 + PI polymer was chosen since it is widely used in the DLW technology and exposure parameters can be found in the literature at the same experimental conditions as and performed LIDT characterization [36, 52]. Firstly, let us discuss the case of the fundamental harmonic (1030 nm). It was shown that polymerization threshold of the SZ2080 + PI is equal to 4.5 TW/cm2 [36]. This value corresponds to the intensity at the focal spot, but the transmission of the objective lens was not encompassed and one should multiply it by a factor of 0.1 which corresponds to the ∼10 % typical lens transmission at the 1030 nm wavelength. One will get 0.45 TW/cm2. Moreover, the repetition rate and scanning velocity used for the DLW experiments was f = 1 kHz and v = 100 μm/s, respectively [36]. It yields Np = (ωf/v) ∼ 10 pulses per spot (here ω = (1.22λ)/NA is a beam waist at the focus point). Hence, polymerization threshold should be compared to the LIDT at 10-on-1 regime as well. It is not provided in the Table 4, but it was measured during LIDT characterization. LIDT is 6.5 times larger than the polymerization threshold and equal to 2.95 TW/cm2. In the case of second harmonic (515 nm), the polymerization threshold is equal to 0.39 TW/cm2 (transmission is included) and the pulse number in the focal spot is ∼100 (f = 200 kHz and v = 1 mm/s) [52]. Corresponding damage threshold at 100-on-1 regime is equal to 0.47 TW/cm2. It can be observed than the ratio between LIDT and DLW threshold decreases considerably at the 515 nm wavelength.
It should be noted that provided comparison is a rough estimation as the different focusing conditions are used for the DLW and LIDT experiments. In the case of the DLW, laser beam is tightly focused into the droplet of the polymer using high numerical aperture objective lens (NA = 1.4) whereas for the LIDT experiments, the light was focused to the target plane with an ordinary lens (NA = 0.25). Moreover, the repetition rate is not the same in all cases: LIDT was performed with 50 kHz repetition rate, DLW at 1030 nm – with 1 kHz [36] and DLW at 515 nm – with 200 kHz [52]. It means that pulse energy was different in all cases. As a result it can influence the threshold values since it causes the energy deposition into the material per pulse which, in turns, leads to the heat accumulation effects.
5. Conclusion
Laser-induced damage experiments of six different polymers were conducted with Nd:YAG and Yb:KGW laser systems representing nanosecond and femtosecond regimes. Obtained results at ns pulse duration regime indicate that LIDT value correlates with the chemical structure of the photopolymers: it was found that LIDT of saturated polymers is higher than that of conjugates polymers at 532 nm wavelength. Moreover, non-photosensitized hybrid organic-inorganic photopolymer SZ2080 distinguished for the highest optical resistance at both wavelength and multi-shot regime in comparison to other materials. The obtained value of the SZ2080 damage threshold (∼14 J/cm2) possesses sufficiently high optical resistance in comparison to conventional coatings of optical glasses and nonlinear crystals. This reveals high potential of this material to be used for the manufacturing of microoptical elements and their applications at moderate intensities.
Study of the LIDT at fs pulse and multi-shot regime revealed that optical resistance of the polymers is lower by one order of magnitude in comparison to results at ns regime: ∼ 0.1 J/cm2 and ∼ 0.2 – 0.6 J/cm2 at 515 nm and 1030 nm, respectively. Also, neither of the materials can be distinguished as the most appropriate candidate for the fabrication of the microoptical elements, especially at 515 nm wavelength. The optical band-gap energy step-like dependence of the LIDT at 515 nm wavelength and single-shot regime suggests that optical breakdown of the photopolymers is caused primary due to multiphoton absorption. Absence of the such dependence at 1030 nm wavelength propose that other ionization paths (e.g. avalanche ionization) could be the dominant physical process leading to the damage. Furthermore, we have noticed that LIDT values of all polymers are decreasing while the number of the incident pulses increases at both wavelength. Such behavior can be attributed to either thermal accumulation or incubation effects related to the intermolecular bond breaking. Also, it explains damage threshold independence on the optical band-gap energy at multi-shot regime and both wavelengths.
Finally, the influence of the photosensitization was revealed both at ns and fs regimes. It was demonstrated that it reduces damage threshold significantly at second harmonics (ten and two folds at 532 nm and 515 nm, respectively) while at the first harmonics it has no significant influence. Hence, employment of the non-photosensitized polymers for the manufacturing of 3D micro/nano-structures should increase their applicability at green and shorter wavelengths.
Acknowledgments
The authors acknowledge the research support by a grant No. MIP-12241 (”Mikrošviesa”) from the Research Council of Lithuania.
References and links
1. S. Maruo, O. Nakamura, and S. Kawata, “Three-dimensional microfabrication with two-photon-absorbed photopolymerization,” Opt. Lett. 22, 132–134 (1997). [CrossRef] [PubMed]
2. M. Malinauskas, M. Farsari, A. Piskarskas, and S. Juodkazis, “Ultrafast laser nanostructuring of photopolymers: a decade of advances,” Phys. Rep. 533, 1–33 (2013). [CrossRef]
3. G. Cojoc, C. Liberale, P. Candeloro, F. Gentile, G. Das, F. de Angelis, and E. Di Fabrizio, “Optical microstructures fabricated on top of optical fibers by means of two-photon photopolymerization,” Microelectron. Eng. 87, 876–879 (2010). [CrossRef]
4. J.H. Atwater, P. Spinelli, E. Kosten, J. Parsons, C. Van Lare, J. Van de Groep, J. Garcia de Abajo, A. Polman, and H.A. Atwater, “Microphotonic parabolic light directors fabricated by two-photon lithography,” Appl. Phys. Lett. 99, 1–3 (2011). [CrossRef]
5. A. Žukauskas, M. Malinauskas, and E. Brasselet, “Monolithic generators of pseudo-nondiffracting optical vortex beams at the microscale,” Appl. Phys. Lett. 103, 181122 (2013). [CrossRef]
6. R. Guo, S. Xiao, X. Zhai, J. Li, A. Xia, and W. Huang, “Micro lens fabrication by means of femtosecond two photon photopolymerization,” Opt. Express 14, 810–816 (2006). [CrossRef] [PubMed]
7. Q.-D. Chen, X.-F. Lin, L.-G. Niu, D. Wu, W.-Q. Wang, and H.-B. Sun, “Dammann gratings as integratable micro-optical elements created by laser micronanofabrication via two-photon photopolymerization,” Opt. Lett. 33, 2559–2561 (2008). [CrossRef] [PubMed]
8. X.-F. Lin, Q.-D. Chen, L.-G. Niu, T. Jiang, W.-Q. Wang, and H.-B. Sun, “Mask-free production of integratable monolithic micro logarithmic axicon lenses,” J. Lightwave Technol. 28, 1256–1260 (2010). [CrossRef]
9. Y. Li, Y. Yu, L. Guo, S. Wu, C. Chen, L. Niu, A. Li, and H. Yang, “High efficiency multilevel phase-type Fresnel zone plates produced by two-photon polymerization of SU-8,” J. Opt. 12, 1–4 (2010).
10. E. Brasselet, M. Malinauskas, A. Žukauskas, and S. Juodkazis, “Photopolymerized microscopic vortex beams generators: precise delivery of optical orbital angular momentum,” Appl. Phys. Lett. 97, 1–3 (2010). [CrossRef]
11. V. Osipov, V. Pavelyev, D. Kachalov, A. Žukauskas, and B. Chichkov, “Realization of binary radial diffractive optical elements by two-photon polymerization technique,” Opt. Express 18, 25808–25814 (2010). [CrossRef] [PubMed]
12. M.D. Turner, M. Saba, Q. Zhang, B.P. Cumming, G.E. Schröder-Turk, and M. Gu, “Miniature chiral beamsplitter based on gyroid photonic crystals,” Nat. Photonics 7, 801–805 (2013). [CrossRef]
13. Z.N. Tian, L.J. Wang, Q.D. Chen, T. Jiang, L. Qin, L.J. Wang, and H.B. Sun, “Beam shaping of edge-emitting diode lasers using a single double-axial hyperboloidal micro-lens,” Opt. Lett. 38(24), 5414–5417 (2013). [CrossRef] [PubMed]
14. S.C. Laza, M. Polo, A.A.R. Neves, R. Cingolani, A. Camposeo, and D. Pisignano, “Two-photon continuous flow lithography,” Adv. Mater. 24, 1304–1308 (2012). [CrossRef] [PubMed]
15. J. Serbin, A. Egbert, A. Ostendorf, B.N. Chichkov, R. Houbertz, G. Domann, J. Schulz, C. Croanauer, L. Fröhlich, and M. Popall, “Femtosecond laser-induced two-photon polymerization of inorganic-organic hybrid materials for applications in photonics,” Opt. Lett. 28(5), 301–303 (2003). [CrossRef] [PubMed]
16. D. Wu, S.-Z. Wu, L.-G. Niu, Q.-D. Chen, R. Wang, J.-F. Song, H.-H. Fang, and H.-B. Sun, “High numerical aperture microlens arrays of close packing,” Appl. Phys. Lett. 97, 1–3 (2010). [CrossRef]
17. M. Malinauskas, A. Žukauskas, V. Purlys, K. Belazaras, A. Momot, D. Paipulas, R. Gadonas, A. Piskarskas, H. Gilbergs, A. Gaidukevičiūtė, I. Sakellari, M. Farsari, and S. Juodkazis, “Femtosecond laser polymerization of hybrid/integrated micro-optical elements and their characterization,” J. Opt. 12, 1–8 (2010).
18. A. Žukauskas, M. Malinauskas, C. Reinhardt, B.N. Chichkov, and R. Gadonas, “Closely packed hexagonal conical microlens array fabricated by direct laser photopolymerization,” Appl. Opt. 51, 4995–5003 (2012). [CrossRef] [PubMed]
19. K. Takada, H.-B. Sun, and S. Kawata, “Improved spatial resolution and surface roughness in photopolymerization-based laser nanowriting,” Appl. Phys. Lett. 86, 071122 (2005). [CrossRef]
20. S.H. Park, S.H. Lee, D.-Y. Yang, H.J. Kong, and K.-S. Lee, “Subregional slicing method to increase three-dimensional nanofabrication efficiency in two-photon polymerization,” Appl. Phys. Lett. 87, 1–3 (2005). [CrossRef]
21. M. Malinauskas, G. Bičkauskaitė, M. Rutkauskas, D. Paipulas, V. Purlys, and R. Gadonas, “Self-polymerization of nano-fibres and nano-membranes induced by two-photon absorption,” Lith. J. Phys. 50, 135–140 (2010). [CrossRef]
22. G. Witzgall, R. Vrijen, E. Yablonovitch, V. Doan, and B.J. Schwartz, “Single-shot two-photon exposure of commercial photoresist for the production of three-dimensional structures,” Opt. Lett. 23(22), 1745–1747 (1998). [CrossRef]
23. Y. Enami, C.T. Derose, D. Mathine, C. Loychik, C. Greenlee, R.A. Norwood, T.D. Kim, J. Luo, Y. Tian, A.K.-Y. Jen, and N. Peyghambarian, “Hybrid polymer/sol-gel waveguide modulators with exceptionally large electro-optics coefficients,” Nature 1, 180–185 (2007).
24. A. Ovsianikov, J. Viertl, B. Chichkov, M. Oubaha, B. MacCraith, I. Sakellari, A. Giakoumaki, D. Gray, M. Vamvakaki, M. Farsari, and C. Fotakis, “Ultra-low shrinkage hybrid photosensitive material for two-photon polymerization microfabrication,” ACS Nano 2(11), 2257–2262 (2008). [CrossRef]
25. J.-F. Ku, Q.-D. Chen, R. Zhang, and H.-B. Sun, “Whispering-gallery-mode micro-disk lasers produced by femtosecond laser direct writing,” Opt. Lett. 15, 2871–2873 (2011). [CrossRef]
26. P.E. Dryer, “Excimer laser polymer ablation: twenty years on,” Appl. Phys. A 77, 167–173 (2003).
27. A.A. Serafetinides, C.D. Skordoulis, M.I. Makropoulu, and A.K. Kar, “Picosecond and subpicosend visible laser ablation of optically transparent polymers,” Appl. Surf. Sci. 135, 276–284 (1998). [CrossRef]
28. R. Srinivasan, B. Braren, D.E. Seeger, and R.W. Dreyfys, “Photochemical cleavage of a polymeric solid: details of the ultraviolet laser ablation of poly(methyl methacrylate) at 193 and 248 nm,” Macromolecules 19(3), 916–921 (1986). [CrossRef]
29. B.J. Garrison and R. Srinivasan, “Laser ablation or organic polymers: microscopic models for photochemical and thermal processes,” J. Appl. Phys. 57(8), 2909–2914 (1985). [CrossRef]
30. P.E. Dyer, G.A. Oldershaw, and J. Sidhu, “CO2 laser ablative etching of polyethylene terephthalate,” Appl. Phys. B 48, 489–493 (1989). [CrossRef]
31. H. Kumagai, K. Midorikawa, K. Toyoda, S. Nakamura, T. Okamoto, and M. Obara, “Ablation of polymer films by a femtosecond high-peak-power Ti:sapphire laser at 798 nm,” Appl. Phys. Lett. 65(14), 1850–1852 (1994). [CrossRef]
32. S. Baudach, J. Bonse, J. Krüger, and W. Kautek, “Ultrashort pulse laser ablation of polycarbonate and poly-methylmethacrylate,” Appl. Surf. Sci. 154–155, 555–560 (2000). [CrossRef]
33. C.N. LaFratta, L. Li, and J.T. Fourkas, “Soft-lithographic replication of 3D microstructures with closed loops,” P. Natl. Acad. Sci. USA 103(23), 8589–8594 (2006). [CrossRef]
34. L. Li, E. Gershgoren, G. Kumi, W.Y. Chen, P.T. Ho, W.N. Herman, and J.T. Fourkas, “High-performance micror-ing resonators fabricated with multiphoton absorption polymerization,” Adv. Mater. 20, 3668–3671 (2008). [CrossRef]
35. S. Rėkštytė, M Malinauskas, and S. Juodkazis, “Three-dimensional laser micro-sculpturing of silicone: towards bio-compatible scaffolds,” Opt. Express 21, 17028–17041 (2013). [CrossRef]
36. M. Malinauskas, A. Žukauskas, G. Bičkauskaitė, and S. Juodkazis, “Mechanisms of three-dimensional structuring of photo-polymers by tightly focussed femtosecond laser pulses,” Opt. Express 18, 10209–10221 (2010). [CrossRef] [PubMed]
37. A. Melninkaitis, M. Ščiuka, G. Batavičiūtė, J. Mirauskas, S. Bucka, and V. Sirutkaitis, “Automated test station for characterization of optical resistance with ultrashort pulses at multi kilohertz repetition rates,” Proc. SPIE 8530, 85301M (2012). [CrossRef]
38. G. Batavičiūtė, P. Grigas, L. Smalakys, and A. Melninkaitis, “Revision of laser-induced damage threshold evaluation from damage probability data,” Rev. Sci. Instrum. 84, 045108 (2013). [CrossRef]
39. ISO 21254-2:2011, “Test methods for laser-induced damahe threshold-Part 2: Threshold determination,” Lasers and laser-related equipment, 2011.
40. G.D. Cody, T. Tiedje, B. Brooks, and Y. Goldstein, “Disorder and the optical-absorption edge of hydrogenated amorphous silicon,” Phys. Rev. Lett. 47, 1480 (1981). [CrossRef]
41. M. Fox, Optical Properties of Solids (Oxford University, 2001).
42. B.C. Stuart, M.D. Feit, A.M. Rubenchik, B.W. Shore, and M.D. Perry, “Laser-induced damage in dielectrics with nanosecond to subpicosecond pulses,” Phys. Rev. Lett. 74(12), 2248–2251 (1995). [CrossRef] [PubMed]
43. K.J. Schafer, J.M. Hales, M. Balu, K.D. Belfield, E.W. Van Stryland, and D.J. Hagan, “Two-photon absorption cross-sections of common photoinitiators,” J. Photoch. Photobio. A 162, 497–502 (2004). [CrossRef]
44. C.W. Carr, H.B. Radousky, and S.G. Demos, “Wavelength dependence of laser-induced damage: determining the damage initiation mechanisms,” Phys. Rev. Lett. 91(12), 127402 (2003). [CrossRef] [PubMed]
45. M. Jupe, L. Jensen, K. Starke, D. Ristau, A. Melninkaitis, R. Grigonis, and V. Sirutkaitis, “Quantized behavior of LIDT in the fs-range in dielectric layers,” Proc. SPIE5647(2005). [CrossRef]
46. A. Chmel, “Fatigue laser-induced damage in transparent materials,” Mater. Sci. Eng. 49, 175–190 (1997). [CrossRef]
47. R. Srinivasan, E. Sutcliffe, and B. Braren, “Ablation and etching of polymethylmethacrylate by very short (160 fs) ultraviolet (308 nm) laser pulses,” Appl. Phys. Lett. 51, 1285 (1987). [CrossRef]
48. B. Rethfeld, “Unified model for the free-electron avalanche in laser-irradiated dielectrics,” Phys. Rev. Lett. 92(18), 187401 (2004). [CrossRef] [PubMed]
49. J. Krüger and W. Kautek, “Femtosecond-pulse visible laser processing of transparent materials,” Appl. Surf. Sci. 96–98, 430–438 (1996). [CrossRef]
50. S. Juodkazis, A.V. Rode, E.G. Gamaly, S. Matsuo, and H. Misawa, “Recording and reading of three-dimensional optical memory in glasses,” Appl. Phys. B. 77, 361–368 (2003). [CrossRef]
51. J. Krüger, S. Martin, H. Mädebach, L. Urech, T. Lippert, A. Wokaun, and W. Kautek, “Femto- and nanosecond laser treatment of doped polymethylmethacrylate,” Appl. Surf. Sci. 247, 406–411 (2005). [CrossRef]
52. A. ŽukauskasM. MalinauskasA. KadysG. GervinskasG. SeniutinasS. KandasamyS. Juodkazis, “Black silicon: substrate for laser 3D micro/nano-polymerization,” Opt. Express21(6), 6901–6909 (2013). [CrossRef] [PubMed]
