Open Access
Add to BibSonomy
Abstract
This paper considers the nanostructuring of the surface of dielectrics under the effect of two successive femtosecond laser pulses, one of the fundamental frequency (FF) and the other of the second harmonic (SH) of a Ti:sapphire laser, through a layer of polystyrene microspheres 1 µm in diameter, which act as microlenses. Polymers with strong (PMMA) and weak (TOPAS) absorption at the frequency of the third harmonic of a Ti:sapphire laser (sum frequency FF + SH) were used as targets. Laser irradiation led to the removal of microspheres and the formation of ablation craters with characteristic dimensions of about 100 nm. Depending on the variable delay time between pulses, the resulting structures differed in their geometric parameters and shape. Based on statistical processing of the depths of the obtained craters, the optimal delay times for the most effective structuring of the surfaces of these polymers was found.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
When high-power femtosecond laser pulses irradiate dielectrics in the transparency range, a plasma is formed in the material due to multiphoton or tunneling ionization. Electrons in the conduction band absorb the energy of laser radiation when they collide with the lattice. If the energy of such an electron is higher than the bandgap energy, then impact ionization that promotes another electron into the conduction band is possible. This can lead to an avalanche-like increase in the number of free electrons and enhance the absorption of light energy. The release of energy within the irradiated region can result in irreversible changes in the material, laser breakdown, and laser ablation. If the generation of seed electrons in the conduction band occurs due to multiphoton ionization, then the efficiency of this process strongly depends on the photon energy, i.e., on the number of photons necessary to overcome the bandgap energy. This leads to the natural idea of using two-color pulses. In this case, part of the fundamental frequency energy that is typically in the infrared range is converted into the second or higher harmonic [1,2]. The second harmonic generates free electrons much more efficiently by multiphoton absorption than the fundamental frequency, the energy of which, in turn, is well absorbed by free electrons. Thus, the conversion of part of the energy into the second harmonic reduces the threshold for laser material alteration.
Another idea is that the second harmonic beam can be better focused; thus, when using a two-color pulse, the actual transverse size of the alteration region, for example, the diameter of the ablation crater, will be determined by the second harmonic, while the energy required for laser ablation will be supplied by a pulse of a more powerful fundamental frequency [3,4]. This is especially important when a laser pulse irradiates the surface of solid dielectrics through a layer of colloidal microparticles, which serve as microlenses here [4]. These microlenses strongly focus laser radiation [5,6], acting as near-field lenses in some cases.
From this point of view, it is interesting to consider polymer materials whose modification and, in particular, ablation thresholds are much lower than the corresponding threshold of inorganic dielectrics. This allows one to create structures over a large area by using a single millijoule laser pulse [4,7]. This method can be an alternative to the interference method for creating structures, since femtosecond laser pulses interfere poorly due to low coherence. Using two-color laser pulses (800 and 400 nm), structures with a density of about 109 cm-2 can be obtained using polystyrene microspheres with a diameter of about 0.5 µm [7]. These spheres do not focus the fundamental frequency radiation at all, but focus the second harmonic radiation. In this work, we study the features of the effect of two-color laser pulses on polymer materials using a layer of polystyrene microspheres deposited in an ordered manner (hexagonal packing) in one layer on the surface of the samples.
Controlled nano- or micro-patterning of polymer surfaces provided by this technique is promising for various biological applications [8], the manufacture of photovoltaics, micro-optics, lab-on- a- chip devices [9], and the development of functional materials [10].
Another aspect of dielectric surface irradiation through a layer of identical close-packed microspheres is the possibility of using them to study the interaction of laser radiation with matter. Indeed, when exposed to a pair of laser pulses, many similar nanostructures form under the microspheres array in a single run. Statistical processing of the obtained structures helps to reliably identify their characteristic parameters for various irradiation conditions.
The third idea in studying the features of the effect of two-color pulses on dielectric surfaces is that if the second harmonic is a source of seed electrons, and the fundamental frequency is used to “heat” these electrons, then for the optimal effect it is necessary that the second harmonic is ahead of the fundamental frequency [2,3,11, 12 ]. In this case, the impact on the electronic system will be optimal. This phenomenon was experimentally considered when two-color pulses of a Ti:sapphire laser were applied to fused silica, that is if the sum frequency (fundamental + second harmonic) falls into the transparency band of the medium. In this work, we study the effects of femtosecond pulses of the fundamental frequency and the second harmonic of a Ti:sapphire laser, shifted relative to each other in the case where the frequency of the third harmonic falls into the region of strong absorption, and in the case where the frequency of the third harmonic falls into the region of weak absorption, the tail of the absorption band. We show that in these cases there are deviations from the above statement.
2. Experimental
2.1 Sample preparation
In the experiment, polymer samples were irradiated with two-color femtosecond pulses with wavelengths of 800 nm (fundamental frequency, FF, of a Ti:sapphire laser and 400 nm, second harmonic, SH, of this laser with an adjustable delay between them. The samples were films of polymethylmethacrylate (PMMA) and TOPAS copolymer [13] with layers of polystyrene microspheres 1 µm in diameter deposited on the sample surface. Figure 1 shows the absorption spectra of the corresponding polymer films and the position of the wavelength of the SH and a third harmonic of a Ti:sapphire laser, corresponding to the sum frequency, FF + SH.
Fig. 1. Absorption spectrum of TOPAS and PMMA samples (a), film thickness 0.1 mm. It is shown that the sum frequency (FF + SH) falls into the region of weak linear absorption of the TOPAS sample and that the sum frequency (FF + SH) falls into the region of strong linear absorption of the PMMA sample. The figure also shows the evolution of absorption associated with the solvent due to annealing of the TOPAS samples during fabrication. (b) Energy diagrams for two-photon absorption.
Polymer samples of PMMA were plates 2.5 × 7.5 cm in size, cut by a laser from a sheet of commercial plexiglass (TOSP-1, OAO Dzerzhinsky plexiglass). The sample surface was freed from the protective film and cleaned with an alcohol solution. AFM scanning of such a sample showed a roughness of 2 nm in a scan field of 10 µm. TOPAS (8007 × 10 olefin copolymer, TOPAS advanced polymers, Germany) samples were obtained by casting a polymer layer about 100 µm thick on a fused silica substrate from a solution of TOPAS granules in toluene. However, it was found that after the deposited polymer dries, the surface becomes very rough (with irregularities at a level of up to 20-30 nm), and when determining the transmission spectra of the polymer film on fused silica substrate, characteristic absorption peaks of the solvent were found. Vacuuming the samples for a day did not give satisfactory removal of solvent traces. For the TOPAS polymer, it was found that at temperatures above 130°C, a strong softening of the polymer occurs, which is sufficient for the action of surface tension forces and improves the characteristics of the samples. The problem of uneven surface during solvent removal and surface leveling was solved by placing the samples for a short time (10-80 min) in a furnace at a temperature of 130-135°C. An AFM scan of such a sample showed a roughness of less than 1 nm in a scan field of 10 µm. The transmission spectra of the polymer itself did not change in the UV part of the spectrum, which indicated the absence of polymer degradation at temperatures above the glass transition temperature. Micron polystyrene spheres (polystyrene microspheres: Sigma Aldrich (Merck) analytical standard 1 µm) with a diameter of 1 µm were cast by dropping an aqueous colloidal solution of microspheres onto polymer films. In this case, the boundaries of the droplets were fixed on scratches made in advance in certain places, and the water was evaporated at 60°C on an inclined plane.
Figure 2 shows photographs of samples with polystyrene microspheres deposited on them.
Fig. 2. PMMA (a) and TOPAS (b) film samples coated with polystyrene microspheres 1 µm in diameter. The photograph was taken with a Nikon Eclipse Ci-S optical microscope (Japan).
2.2 Exposure techniques
Figure 3 shows the optical scheme of sample irradiation.
Fig. 3. General optical scheme of sample irradiation by two femtosecond pulses with a variable delay time. HR is a mirror with a high reflection coefficient, BS is a beam splitter for pulse energy control, DM is a dichroic mirror that separates radiation into FF and SH, TF is a filter turret, DL is a delay line with micrometer screws and an electronic micrometer for position determination, L is a converging lens, S is the sample, MP is a mechanical three-coordinate platform, F is the lens focus point, Mi is a digital microscope with up to x200 magnification, PC are computers for data acquisition, Sp is the spectrometer, SF is the optical input of the spectrometer, CI is the BBO crystal for generating the second harmonic according to the type ooe synchronism, CII is the BBO crystal for generating the third harmonic during the circuit timing control. Line color corresponds to the beam wavelength: red – 800 nm, green – 532 nm, blue – 400 nm, and violet – 266 nm.
When exposed to a pair of laser pulses, the microspheres fly off, leaving a large field of nanostructures in the irradiated area suitable for investigation using probe microscopy. Due to the presence of a large number of structures at once, it is possible to single out only those of them that are not on the defective areas of the surface. Statistical processing of the obtained structures helps to reliably identify their characteristic parameters for different delays between the FF and SH pulses.
To obtain a stable result of cleaning the surface from microspheres, several different predetermined values of the intensities of the FF and SH laser pulses were used, which was achieved by shifting the sample by a few millimeters along the propagation axis (closer or farther from the focus). For each type of samples, more than 100 successful experiments were performed with high-quality cleaning of the impact area from microspheres. Under laser exposure, the FF pulse energy was determined using a control photodiode. If the pulse energy significantly deviated from the average (more than 2% at the fundamental frequency), the irradiated region was omitted from consideration.
When irradiating PMMA samples, applied laser pulse energy for FF was 110 µJ, pulse duration was 60 fs, peak intensity was about 1.5 TW/cm2, fluence was 90 mJ/cm2. Applied laser pulse energy for SH was 5 µJ, pulse duration was 80 fs, peak intensity was about 90 GW/cm2, fluence was 7 mJ/cm2.
When irradiating TOPAS samples, applied laser pulse energy for FF was 150 µJ, pulse duration was 60 fs, peak intensity was about 3 TW/cm2, fluence was 180 mJ/cm2. Applied laser pulse energy for SH was 9 µJ, pulse duration was 80 fs, peak intensity was about 300 GW/cm2, fluence was 25 mJ/cm2.
Since the FF and SH beams have different diameters in the interaction region, the idea of beams centers shifted relative to each other was implemented (see Fig. 4), so that the region under study was at the FF beam boundary, where the microsphere removal threshold was not yet reached during the effect of the FF pulse alone. Under the employed conditions of laser exposure, we ensured that the threshold was also not reached when using SH pulses alone and even when using a pair of FF and SH pulses with large relative delays (more than 4 ps). At the same time, we check that in the area of intersection of the beams, the microsphere removal threshold was reached under the synchronous effect of the FF and SH pulses (zero delay). This arrangement of pulses simplifies the detection of areas available for studies with an atomic force microscope.
Fig. 4. Schematic of the two-beams irradiated zone and the microphotograph of the areas where the spheres have been removed. The centers of the beams are shifted in such a way that the SH beam irradiates the area near the microsphere removal threshold by the FF beam alone. The AFM data acquisition point occurs out of the area where the microspheres have been removed by the FF beam alone.
The disadvantage of such an approach is the intensity gradient of the FF beam. However, the size of the region scanned by the AFM does not exceed 50 µm whereas the FF beam diameter was about 500 µm and the above gradient affects the parameters of the structures less significantly than the effect of the relative delay between laser pulses. This was verified in separate experiments.
Not all the areas were selected for further AFM scanning, but only those along the perimeter of which remnants of the lying ordered microspheres packed in a hexagonal domain structure were clearly observed in the optical microscope. This proved that indeed we are studying the action of laser radiation through a monolayer. When scanning with an atomic force microscope, first a survey scan was used with a scan area size of 20-50 µm, and then a detailed scan of selected segments of the area was used approximately at the center (possibly closer to the center) of the second-harmonic beam spot. In this case, the size of the scanned square regions was 5–10 µm, which was quite sufficient for obtaining profiles of the structures that appeared on the surface. In the detailed scans we obtained, we examined only craters that were not located at the edge of the domain, i.e., were in the inner position within the hexagonal structure, taking into account the edge effect studied in [14]. Next, to obtain the characteristics of each crater separately, the substrate level was first identified, which is the reference plane for further measurements and statistical analysis. The level of the surface far from the structures was taken as the substrate level, which was determined from the average of the height data outside the dust grains, defects, or debris. In total, from 5 to 80 profiles of single craters could be obtained from detailed scans, depending on the quality and location of the crater. During the statistical processing of the data, some outliers were identified, which were excluded from the results due to possible inhomogeneity of the substrate material.
3. Results and discussion
Figures 5 and 6 show examples of AFM images of irradiated samples.
Fig. 5. Images obtained by AFM methods during PMMA surface structuring during the implementation of the leading SH pulse (with a delay of 200 fs) (a, b) and the synchronous effect of the FF and SH laser pulses (c, d), zero delay.
Fig. 6. Images obtained by AFM methods during structuring of the TOPAS surface during the implementation of the leading SH pulse (a, b) and the leading FF pulse (c, d) (with a relative delay of 133 fs in each case), as well as under the synchronous action of both pulses (e, f). In the latter case, there is a significant scatter in the depths of ablation craters and different morphology of the structures.
Figure 7 shows statistical data on the distribution of depths of ablation craters in the case of PMMA and TOPAS depending on the relative delay between pulses.
Fig. 7. Depths of nanostructures from the level of the substrate surface for the materials under study depending on the delay time between FF and SH pulses, (a)-PMMA, (b) -TOPAS. The positive delay corresponds to the FF pulse preceding the SH one. In the case of PMMA, there is a single maximum corresponding to the zero delay time. In the case of TOPAS, there are two maxima, one at negative delay time and the other at zero delay time.
The real ionization potential of used polymers is about 10-12 eV. Thus, talking about the femtosecond “two-photon ionization”, we imply the model considered in [1]. It assumes that the two-photon electron transition to the strong absorption band is followed by a cascade of saturated single-photon transitions. That is, the electron promoted to the strong absorption band will end up in the continuum spectrum (conduction band) with the probability close to unity. It means that the first two-photon transition through the bandgap defines the overall rate of the ionization process. This model allows to apply the concepts developed previously for femtosecond laser interaction with dielectrics such as fused silica to polymers considered here. Small FF + SH absorption in the case of TOPAS means that it is an order of magnitude smaller than in its strong absorption band (see Fig. 1).
When two-color femtosecond pulses are applied to dielectrics with a wide bandgap, a high-frequency pulse (in our case, a SH one) plays the role of a generator of seed electrons into the conduction band due to the multiphoton absorption of only SH photons, which in this case is most effective, since it requires the absorption of the smallest number of photons for the transition of an electron from the valence band to the conduction band. The fundamental frequency pulse is used to increase the energy of electrons in the conduction band, including an increase in the density of such electrons due to impact ionization. In this case, the most optimal configuration is achieved when the high-frequency (SH) pulse precedes the low-frequency (FF), since in this case the FF pulse will affect all the electrons generated by the SH pulse. The presence of an efficiency maximum at a finite detuning of the FF pulse from the SH pulse is related to the relaxation time of electrons from the conduction band. The maximum at a negative delay of the SH pulse from the FF pulse in Fig. 7(b) is related precisely to this circumstance, since the doubled frequency of the SH falls within the fundamental absorption band of the TOPAS sample (see Fig. 1(a)), while the sum frequency of the FF + SH lies in the bandgap. If the sum frequency of FF + SH falls within the region of strong absorption of the material, as in the case of PMMA, then the synchronous action of pulses leads to the highest efficiency, since here the most efficient channel for generating electrons in the conduction band is the two-photon transition of FF + SH with the participation of a powerful pulse of the fundamental frequency. This can be seen in Fig. 7(a), where the maximum efficiency of laser radiation exposure corresponds to zero delay between pulses. Note, however, that for PMMA there is an asymmetry in the efficiency curve in Fig. 7(a). Indeed, for negative delays (SH ahead), this curve is higher than for a positive delays (FF ahead). This is due to the fact that when the pulses are not synchronized, the two-photon FF + SH transition loses its efficiency and the weaker two-photon SH + SH and three-photon FF + FF + FF transitions come to the fore, which affects the depth of the ablation craters. In the case of negative delays, electrons are “heated” by the FF pulse generated by two-photon transitions SH + SH and FF + FF + FF. In the case of positive delays, there is heating of electrons generated only by the FF + FF + FF transition and generation of electrons by the SH + SH transition without heating by FF. In the first case, the absorbed energy will obviously be greater. The effect of asymmetry will actually be greater than it follows from the shape of the curves in Fig. 7. As can be seen in Figs. 5 and 6, ablation craters in the case where the FF pulse comes first have large rims. These rims indicate that in this case craters are formed not only due to ablation, but also due to the hydrodynamic removal of matter from the laser-heated area of the material, for example, due to the Marangoni effect by which a force arises at the surface along the surface tension gradient It is known that the surface tension decreases under heating. Another reason can be vapor pressure-based liquid expulsion [15]. This suggests that under the dominant influence of only the FF pulse, the shape of the crater changes, the crater becomes wider and a rim appears, i.e., the quality of the structure deteriorates. Finally, a very important circumstance is the appearance of an additional maximum in Fig. 6(b) (TOPAS) at a zero delay between pulses. This is due to the fact that some absorption is observed in the bandgap of the polymer, which can be associated with the tail of the absorption band. In this case, due to the high intensity of the FF pulse under the synchronous action of both pulses, the generation of electrons during the FF + SH transition will prevail even with weak resonant absorption. When two pulses are desynchronized, this transition becomes insignificant, and the ablation efficiency curve as a function of delays corresponds to the case of a wide bandgap, where the two-photon SH + SH transition discussed above is the predominant mechanism for generating seed electrons. It is interesting to note that, due to some “randomness” of weak absorption at the sum frequency FF + SH, which can be understood as the region of absorption by ‘defects,’ the variations in craters depths for zero delay is significantly larger than for the negative delays, as can be seen in Fig. 6(b). This suggests that in the case of weak absorption at the sum frequency (FF + SH), in order to obtain the optimal quality of structures, it is necessary to work with a negative delay between the pulses, where the SH + SH transition to the strong absorption band is used to generate seed electrons.
4 Conclusions
The effect of a pair of femtosecond pulses, one of the fundamental frequency and the other of the second harmonic of a Ti:sapphire laser, on the of polymer surfaces coated with a layer of colloidal polystyrene microparticles 1 µm in diameter, has been studied. PMMA and TOPAS copolymer were considered. In the case of PMMA, the sum frequency (FF + SH) falls into the region of strong absorption; in the case of TOPAS, the sum frequency is in the bandgap. AFM was used to study the shape and depth of ablation craters formed after the departure of polystyrene microspheres as a result of a single impact of a pair of pulses. In the experiment, the energies and intensities of the pulses were fixed, and the time delay between the pulses was varied. Statistical analysis of the ablated craters depth revealed, that the structures obtained at a zero relative delay between the FF and SH pulses have the maximum depth from the substrate level for PMMA samples. For TOPAS samples, the maximum depth of the ablation craters is achieved when the SH pulse comes 133–200 fs before the FF (with a 66-fs delay step). When the FF pulse arrives first, a qualitatively different shape of the created structures with significant rims is observed. Also, for zero delay between the FF and SH pulses in the case of the TOPAS polymer, the second maximum of the crater depth is observed with a large data scatter, which can be explained by the presence of inclusions or impurities in a polymer with absorption levels in the bandgap.
Funding
Russian Science Foundation (22-19-00322).
Acknowledgments
Scanning probe microscopy studies were performed using the equipment of the Center for Collective Use of the Research and Educational Center “Physics of Solid-State Nanostructures” at the N.I. Lobachevsky State University of Nizhny Novgorod.
Disclosures
The authors declare no conflicts of interest.
Data availability
No additional data were generated or analyzed in the presented research.
References
1. N. Bityurin and A. Kuznetsov, “Use of harmonics for femtosecond micromashining in pure dielectrics,” J. Appl. Phys. (Melville, NY, U. S.) 93(3), 1567–1576 (2003). [CrossRef]
2. D. Grojo, M. Gertsvolf, S. Lei, T. Barillot, D. M. Rayner, and P. B. Corkum, “Exciton-seeded multiphoton ionization in bulk SiO2,” Phys. Rev. B 81(21), 212301 (2010). [CrossRef]
3. X. Yu, Q. Bian, B. Zhao, Z. Chang, P. B. Corkum, and S. Lei, “Near-infrared femtosecond laser machining initiated by ultraviolet multiphoton ionization,” Appl. Phys. Lett. 102(10), 101111 (2013). [CrossRef]
4. N. Bityurin, A. Afanasiev, V. Bredikhin, A. Alexandrov, N. Agareva, A. Pikulin, I. Ilyakov, B. Shishkin, and R. Akhmedzhanov, “Colloidal particle lens arrays-assisted nano-patterning by harmonics of a femtosecond laser,” Opt. Express 21(18), 21485 (2013). [CrossRef]
5. J. Kofler and N. Arnold, “Axially symmetric focusing as a cuspoid diffraction catastrophe: Scalar and vector cases and comparison with the theory of Mie,” Phys. Rev. B 73(23), 235401 (2006). [CrossRef]
6. B. S. Luk’yanchuk, R. Paniagua-Domínguez, I. Minin, O. Minin, and Z. Wang, “Refractive index less than two: photonic nanojets yesterday, today and tomorrow,” Opt. Mater. Express 7(6), 1820 (2017). [CrossRef]
7. A. Afanasiev, V. Bredikhin, A. Pikulin, I. Ilyakov, B. Shishkin, R. Akhmedzhanov, and N. Bityurin, “Two-color beam improvement of colloidal particle lens array assisted surface nanostructuring,” Appl. Phys. Lett. 106(18), 183102 (2015). [CrossRef]
8. C. Fernandez-Rodriguez, J. Paredes, J. Etxarri, E. Tellechea, M. L. Martinez-Chantar, L. A. Martinez-Cruz, and I. Quintana, “Laser ablation combined with Nanoimprint Lithography technology as new surface engineering approach to produce novel polymer-based heteronucleants for recalcitrant protein crystallization,” Appl. Surf. Sci. 607, 155029 (2023). [CrossRef]
9. R. van Dommelen, P. Fanzio, and L. Sasso, “Surface self-assembly of colloidal crystals formicro- and nano-patterning,” Adv. Colloid Interface Sci. 251, 97–114 (2018). [CrossRef]
10. R. Wang, Ph. Zimmermann, D. Schletz, M. Hoffmann, P. Probst, A. Fery, J. Nagel, and Ch. Rossner, “Nano meets macro: Furnishing the surface of polymer molds with gold-nanoparticle arrays,” Nano Sel. 3(11), 1502–1508 (2022). [CrossRef]
11. J. Peng, D. Grojo, D. M. Rayner, and P. B. Corkum, “Control of energy deposition in femtosecond laser dielectric interactions,” Appl. Phys. Lett. 102(16), 161105 (2013). [CrossRef]
12. F. Potemkin, E. Mareev, Y. Bezsudnova, V. Platonenko, B. Bravy, and V. Gordienko, “Controlled energy deposition and void-like modification inside transparent solids by two-color tightly focused femtosecond laser pulses,” Appl. Phys. Lett. 110(16), 163903 (2017). [CrossRef]
13. TOPAS COC Polymers, Advanced Resins for Your Leading Products, https://topas.com/products/topas-coc-polymers.
14. A. Afanasiev, A. Pikulin, I. Ilyakov, B. Shishkin, and N. Bityurin, “Edge effect at the microsphere colloidal array in near-field particle lithography on polymer surfaces,” Surf. Interfaces 29, 101735 (2022). [CrossRef]
15. D. Bäuerle, Laser processing and chemistry (Springer2011).
