Underwater Laser Treatment of PET: Effect of Processing Parameters on Surface Morphology and Chemistry

: Rapid development of nanotechnology in processes of metal nanoparticle immobilization on solid surfaces, especially polymeric ones, requires the study of particular issues within these complex approaches. Numerous studies have been published on laser light mediated manipulation with single metal nanoparticles in water environment and even laser assisted immobilization of such particles on polymeric substrate, however, not much has been reported on fundamentals of underwater laser processing of polymer itself, especially regarding to resulting surface morphology and chemistry. In this work, we study surface morphology (atomic force microscopy (AFM)) and chemistry (angle-resolved X-ray photoelectron spectroscopy (ARXPS) and inductively coupled plasma-mass spectroscopy (ICP-MS)) of polyethylene terephthalate (PET) after underwater laser treatment in broad scale of applied laser ﬂuencies and operating voltages. Due to typical dependence of laser efﬁciency on operating voltage, induced nanostructures on PET exhibited a noticeable symmetry spread out around the maxima of laser efﬁciency for low laser ﬂuencies. The study of surface chemistry revealed that at high laser ﬂuencies, photochemical decomposition of macromolecular polymer structure took place, resulting in rapid material ablation and in balanced chemical composition of the surface throughout the studied proﬁle. Enrichment of the water bath by the low-molecular polymer degradation products proves that ablation mechanism is the governing process of surface nanostructure formation in underwater laser processing.


Introduction
With the development of new approaches of nanoparticle immobilization on polymer surfaces, especially those driven by forward-directed scattering forces in colloid solutions [1], the need to thoroughly investigate laser to polymer interaction in a water environment becomes of fundamental importance. Many theoretical models have been suggested to describe the processes at the water-polymer interface during laser treatment, however, experimental studies, addressing this issue in detail, are either missing or date back to early 1990s, which hinders comparison of obtained data to modern analytical outputs of present science. Several models of laser heating of polymers in air have been suggested so far. These include photothermal models [2,3], especially for higher wavelength radiation (such as for infrared lasers). Contrary to that, in case of UV lasers, models based on both photothermal and photodissociative or photochemical breakdown have been reported [4,5]. The photochemical model fundamentals lie in the chemical kinetics of bond dissociation and are coupled with the processes of photothermal decomposition of material. Some of these models also take into account dynamic attenuation of radiation by the expanding plasma plume [6]. Besides these, molecular dynamics is also frequently used to simulate laser ablation of polymers [7]. Although the fundamentals of the laser-matter interaction remain the same as in air even in the presence of water, such environment can have a significant effect especially to the heating of the polymer and in turn may result to the development of surface structures, which morphology is far away from those, typically observed in case of laser processed materials in air environment (LIPSS) [8,9]. In particular, the phenomena of plasma formation, its expansion, and subsequent radiation attenuation effects, are greatly influenced by the specific type of confining media [10].
In this work, we report on investigation of surface morphology and chemistry of polyethylene terephthalate (PET) after underwater treatment with excimer KrF laser in broad scale of applied laser fluencies (4-20 mJ·cm −2 ) and operating voltages (19-29 kV). Due to typical parabolic dependence of laser efficiency on operating voltage of main laser capacitor, induced structures exhibit a noticeable symmetry spread out around the maxima of laser efficiency at lower examined laser fluencies. Development of typical surface nanostructure takes place in a narrow interval of applied laser fluencies and is accompanied by considerable polymer ablation as the laser fluence increases.

Materials, Apparatus and Procedures
Polyethylene terephthalate foil (PET, thickness 50 µm, density 1.36 g·cm −3 ) supplied by Goodfellow Ltd., Cambridge, UK, was exposed to a 248-nm KrF excimer laser (COMPexPro 50F, Coherent, Inc., California, USA) with pulse duration of 20 to 40 ns, frequency of 10 Hz, 6000 pulses. A laser beam was linearly polarized by fused quartz prism (model PBSO 248-100) and directed through the aperture of 5 × 10 mm 2 to the polymer. Strip of PET foil (30 × 8 mm 2 ) was placed vertically into high precision spectro-optics cell (HellmaAnalitics GmbH, Müllheim, Germany, type No. 100-QS, light path of 10 mm), so that the polymer was centered in the middle of the cuvette. Afterwards, 3 mL of water (water for ion chromatography, Sigma-Aldrich Co., Missouri, USA) was added by using automatic pipette. Scheme of the process in depicted in Figure 1. In our experiments, we applied fluencies ranging from 4 to 20 mJ·cm −2 , however, worm-like morphology, typical for underwater polymer processing with linearly polarized laser light, occurred in much narrower interval of 8 to 18 mJ·cm −2 . All PET samples were modified under static conditions with 6000 pulses, perpendicularly to polymer surface. For the setting of required laser fluence (4-20 mJ·cm −2 ) at different operating voltage (19-29 kV, charging voltage applied on main storage capacitor), external laser beam attenuator (Attenuator module No. 2910216, Coherent, Inc., California, USA) was used. Attenuator was placed immediately behind the mouth of the laser tube. with the processes of photothermal decomposition of material. Some of these models also take into account dynamic attenuation of radiation by the expanding plasma plume [6]. Besides these, molecular dynamics is also frequently used to simulate laser ablation of polymers [7]. Although the fundamentals of the laser-matter interaction remain the same as in air even in the presence of water, such environment can have a significant effect especially to the heating of the polymer and in turn may result to the development of surface structures, which morphology is far away from those, typically observed in case of laser processed materials in air environment (LIPSS) [8,9]. In particular, the phenomena of plasma formation, its expansion, and subsequent radiation attenuation effects, are greatly influenced by the specific type of confining media [10].
In this work, we report on investigation of surface morphology and chemistry of polyethylene terephthalate (PET) after underwater treatment with excimer KrF laser in broad scale of applied laser fluencies (4-20 mJ· cm −2 ) and operating voltages (19-29 kV). Due to typical parabolic dependence of laser efficiency on operating voltage of main laser capacitor, induced structures exhibit a noticeable symmetry spread out around the maxima of laser efficiency at lower examined laser fluencies. Development of typical surface nanostructure takes place in a narrow interval of applied laser fluencies and is accompanied by considerable polymer ablation as the laser fluence increases.

Materials, Apparatus and Procedures
Polyethylene terephthalate foil (PET, thickness 50 µ m, density 1.36 g· cm −3 ) supplied by Goodfellow Ltd., Cambridge, UK, was exposed to a 248-nm KrF excimer laser (COMPexPro 50F, Coherent, Inc., California, USA) with pulse duration of 20 to 40 ns, frequency of 10 Hz, 6000 pulses. A laser beam was linearly polarized by fused quartz prism (model PBSO 248-100) and directed through the aperture of 5 × 10 mm 2 to the polymer. Strip of PET foil (30 × 8 mm 2 ) was placed vertically into high precision spectro-optics cell (HellmaAnalitics GmbH, Müllheim, Germany, type No. 100-QS, light path of 10 mm), so that the polymer was centered in the middle of the cuvette. Afterwards, 3 mL of water (water for ion chromatography, Sigma-Aldrich Co., Missouri, USA) was added by using automatic pipette. Scheme of the process in depicted in Figure 1. In our experiments, we applied fluencies ranging from 4 to 20 mJ· cm −2 , however, worm-like morphology, typical for underwater polymer processing with linearly polarized laser light, occurred in much narrower interval of 8 to 18 mJ· cm −2 . All PET samples were modified under static conditions with 6000 pulses, perpendicularly to polymer surface. For the setting of required laser fluence (4-20 mJ· cm −2 ) at different operating voltage (19-29 kV, charging voltage applied on main storage capacitor), external laser beam attenuator (Attenuator module No. 2910216, Coherent, Inc., California, USA) was used. Attenuator was placed immediately behind the mouth of the laser tube.

Analytical Methods
Laser fluence, applied on PET during underwater treatment, was measured by FieldMaxII-TOP Laser Power Meter (Coherent, Inc., California, USA), so that it was placed just behind the cuvettefilled water without the polymer. Owing to only negligible absorption of purified water at 248 nm

Analytical Methods
Laser fluence, applied on PET during underwater treatment, was measured by FieldMaxII-TOP Laser Power Meter (Coherent, Inc., California, USA), so that it was placed just behind the cuvette-filled water without the polymer. Owing to only negligible absorption of purified water at 248 nm (absorption coefficient ε H2O < 0.04 [11]), measured laser fluence in that way corresponds well to that applied to polymer during the experiments.
The atomic concentrations of elements, carbon C(1s) and oxygen O(1s), were studied by Angle Resolved X-ray Photoelectron Spectroscopy (ARXPS) on Omicron Nanotechnology ESCAProbeP spectrometer, Taunusstein-Neuhof, Germany. The X-ray source was monochromated at 1486.7 eV. The evaluation of spectra was performed by CasaXPS software, version 2.3.19. The samples were studied under the electron take-off angles of 0 and 81 • .
Surface morphology and roughness were examined by atomic force microscope (AFM) VEECO CP II device in a tapping mode. A Veeco oxide-sharpened P-doped silicon probe RTESPA-CP was used. Surface roughness, characterized by the mean roughness value (R a ), represents the arithmetic average of the deviation from the center plane of a sample. Variation of R a parameter did not exceed 5%. All samples were measured at 7 different points within modified area in the arrangement 2, 3, 2 along the longer side of the modified region (see Figure 2). In all cases the measurement provided similar surface morphologies within one sample, pointing to homogeneous distribution of induced nanostructure by laser treatment. Presented AFM images were carefully chosen to represent typical morphologies of laser processed samples under the specific modification conditions. (absorption coefficient εH2O < 0.04 [11]), measured laser fluence in that way corresponds well to that applied to polymer during the experiments. The atomic concentrations of elements, carbon C(1s) and oxygen O(1s), were studied by Angle Resolved X-ray Photoelectron Spectroscopy (ARXPS) on Omicron Nanotechnology ESCAProbeP spectrometer, Taunusstein-Neuhof, Germany. The X-ray source was monochromated at 1486.7 eV. The evaluation of spectra was performed by CasaXPS software, version 2.3.19. The samples were studied under the electron take-off angles of 0 and 81°.
Surface morphology and roughness were examined by atomic force microscope (AFM) VEECO CP II device in a tapping mode. A Veeco oxide-sharpened P-doped silicon probe RTESPA-CP was used. Surface roughness, characterized by the mean roughness value (Ra), represents the arithmetic average of the deviation from the center plane of a sample. Variation of Ra parameter did not exceed 5%. All samples were measured at 7 different points within modified area in the arrangement 2, 3, 2 along the longer side of the modified region (see Figure 2). In all cases the measurement provided similar surface morphologies within one sample, pointing to homogeneous distribution of induced nanostructure by laser treatment. Presented AFM images were carefully chosen to represent typical morphologies of laser processed samples under the specific modification conditions. Inductively coupled plasma with mass spectroscopy detector (ICP-MS) was used to determine total carbon content (TCC) in water from the cuvette after laser treatment of polymer. Changes in the TCC correspond to the release of low molecular weight polymer ablation product to water environment due to laser modification. The trace element analysis of TCC was conducted by using Agilent 8800 triple-quadrupole spectrometer (Agilent Technologies, California, USA) connected to an auto-sampler according to measurement protocol taken from [12]. Sample nebulization was performed using a MicroMist device equipped with a tempered cell. To minimize the interference of an analyte with multiatomic adducts, we used a collision cell (He collision gas) operating in a highenergy mode with double quadrupole filtration.

Surface Morphology
Development of typical surface morphology during underwater laser treatment of PET was observed in narrow interval of laser fluencies from 8 to 18 mJ· cm −2 , regardless of operating voltage applied on laser main storage capacitor. Operating voltage, however, had distinctive impact on the fine microstructure superimposed on the on PET surface, in particular at low applied fluencies. This interesting phenomenon most likely originates from the symmetry of laser efficiency on operating voltage (see Figure 3), which reaches a maximum at about 25 kV (ca 1%) and then falls evenly to both border values (ca 0.75 %).
Lower laser efficiency at low and/or high voltages is related to increased instability in laser beam profile at both border values [13], which obviously determines the resulting surface morphology, especially at lower laser fluencies. For clear illustration of this phenomenon, surface morphology in Inductively coupled plasma with mass spectroscopy detector (ICP-MS) was used to determine total carbon content (TCC) in water from the cuvette after laser treatment of polymer. Changes in the TCC correspond to the release of low molecular weight polymer ablation product to water environment due to laser modification. The trace element analysis of TCC was conducted by using Agilent 8800 triple-quadrupole spectrometer (Agilent Technologies, California, USA) connected to an auto-sampler according to measurement protocol taken from [12]. Sample nebulization was performed using a MicroMist device equipped with a tempered cell. To minimize the interference of an analyte with multiatomic adducts, we used a collision cell (He collision gas) operating in a high-energy mode with double quadrupole filtration.  One can clearly see that at lower laser fluencies and boarder values of applied voltages (Figure 4), the developed surface morphology is reminiscent to that of treated in air [14]. Even though the LIPSS are not precisely developed, typical prolongation of worm-like structures at specific operating voltage into linear ones is obvious. This may originate from the confined plasma plume in water, which expands at a much lower velocity than in air (order of magnitude less). This implicates that spatially confined plasma induces increase of surface temperature in water environment much intensively than in air [10], which may significantly contributes to the development of similar surface structure as in air when processing at sufficiently lower laser fluencies in water [15]. It is also obvious from the Figures 4-6, that the higher the laser fluence applied the more uniform developed nanostructure is and the possibility of modulation of fine microstructure by variation of applied voltage diminishes. This finding may be quantified by corresponding values of surface roughness, which are given in Table 1. Concurrently to AFM images, Ra values suggest that at higher applied fluencies the effect of voltage on surface roughness is negligible and surface roughness stabilizes, regardless of specific voltage value. At low fluence, when worm-like nanostructure is not fully developed (see Figure 4: 19 and 29 kV), however, the variation of Ra value corresponds to the possibility to vary surface morphology just by proper setting of applied voltage. Lower laser efficiency at low and/or high voltages is related to increased instability in laser beam profile at both border values [13], which obviously determines the resulting surface morphology, especially at lower laser fluencies. For clear illustration of this phenomenon, surface morphology in dependence of operating voltages (19-29 kV) is shown in Figures 4-6 for applied laser fluencies of 8, 12, and 16 mJ·cm −2 , respectively. We have chosen these fluencies, since they provide sufficient insight into the nature of morphology changes of underwater laser processed polymer. With increasing flow, massive ablation of the polymer surface took place, until the typical worm-like nanostructure disappeared completely (applied laser fluence > 18 mJ·cm −2 ).

Surface Morphology
One can clearly see that at lower laser fluencies and boarder values of applied voltages (Figure 4), the developed surface morphology is reminiscent to that of treated in air [14]. Even though the LIPSS are not precisely developed, typical prolongation of worm-like structures at specific operating voltage into linear ones is obvious. This may originate from the confined plasma plume in water, which expands at a much lower velocity than in air (order of magnitude less). This implicates that spatially confined plasma induces increase of surface temperature in water environment much intensively than in air [10], which may significantly contributes to the development of similar surface structure as in air when processing at sufficiently lower laser fluencies in water [15]. It is also obvious from the Figures 4-6, that the higher the laser fluence applied the more uniform developed nanostructure is and the possibility of modulation of fine microstructure by variation of applied voltage diminishes. This finding may be quantified by corresponding values of surface roughness, which are given in Table 1. Concurrently to AFM images, R a values suggest that at higher applied fluencies the effect of voltage on surface roughness is negligible and surface roughness stabilizes, regardless of specific voltage value. At low fluence, when worm-like nanostructure is not fully developed (see Figure 4: 19 and 29 kV), however, the variation of R a value corresponds to the possibility to vary surface morphology just by proper setting of applied voltage.

Chemical Composition of the Surface and Water Bath
Changes in chemical composition of polymer surface after underwater laser treatment were recorded by ARXPS and are given in Table 2. Atomic concentrations of O (35.2 at. %) and C (64.8 at. %) in pristine PET corresponds well to polymer stoichiometry with theoretical concentrations of O (33.0 at. %) and C (67.0 at. %), which are often shifted in favor of carbon due to hydrocarbon impurities being adsorbed on the surfaces of the polar polymeric materials from air [16]. Advantage of angle-resolved XPS analysis lies in the possibility to sensitively distinguish between the element concentration in thick surface polymer layer (typically 10 atomic layers, electron collection angle of 0 • ) and ultrathin surface shell (typically 1 to 2 atomic layers, electron collection angle of 81 • ). Thus, samples of pristine PET typically exhibit increase of carbon concentration when analyzed under 81 • , especially due to absorbed hydrocarbons in mono-atomic layer (see Table 2). Apparently, when increasing applied laser fluence the discrepancy between "bulk" and "very surface" concentration of C and O gradually disappears. This phenomenon most likely originates from (i) massive polymer ablation (absorbed hydrocarbons are removed), (ii) dramatic photochemical decomposition of macromolecular structure which unifies the chemical composition of the surface in a deeper profile [4]. Both of these contributions being affected by considerable changes in surface morphology, especially regarding data obtained under 81 • , where the analytical information comes predominantly from the ridges of induced nanostructure. Besides this, these changes are obviously accompanied by considerable graphitization of polymer surface [1] which contributes to increasing concentration of carbon with increasing applied fluence. This finding is similar to that of Elaboudi et al. [15], who identified the hydrolysis reaction responsible for decrease of laser threshold during excimer laser processing of polyesters which increases the efficiency at low applied fluencies. Their results showed that ablation products are rich in carboxylic groups and carbon, but did not correspond directly to hydrolysis products of polymer. It is also evident, that variation of voltage has no direct impact on chemical composition, which corresponds well with the fact that absorbed energy is invariant towards this parameter. Remarkable difference in concentrations of C and O in dependence of detection angle can only be observed at treatment fluence of 8 mJ·cm −2 , similarly as in case of pristine PET. This phenomenon most likely originates from considerably low absorbed energy when the ablation and photothermal cleavage of macromolecules is low.
To reveal the mechanism of surface structure development on laser treated polymer, identification of the rate of material ablation was accomplished. To address this issue, ICP-MS analysis of the water bath used in underwater laser treatment, aiming to total carbon content (TCC), was accomplished. Those results are summarized in Figure 7a,b. Changes in the TCC correspond to the release of low molecular weight polymer ablation product to water environment due to laser modification. Figure 7a shows dependence of TCC on applied voltage at fixed laser fluence of 8 mJ·cm −2 . Compared to measured TCC background in fresh water (blue line, 1.2 ± 0.3 mg·L −1 ), TCC in the bath increased approximately one order of magnitude after laser processing. It is clearly visible that applied voltage has no influence on TCC, which is in good concordance with results obtained from AR-XPS analysis, suggesting limited material removal from the PET surface at low fluencies. changes in surface morphology, especially regarding data obtained under 81°, where the analytical information comes predominantly from the ridges of induced nanostructure. Besides this, these changes are obviously accompanied by considerable graphitization of polymer surface [1] which contributes to increasing concentration of carbon with increasing applied fluence. This finding is similar to that of Elaboudi et al. [15], who identified the hydrolysis reaction responsible for decrease of laser threshold during excimer laser processing of polyesters which increases the efficiency at low applied fluencies. Their results showed that ablation products are rich in carboxylic groups and carbon, but did not correspond directly to hydrolysis products of polymer. It is also evident, that variation of voltage has no direct impact on chemical composition, which corresponds well with the fact that absorbed energy is invariant towards this parameter. Remarkable difference in concentrations of C and O in dependence of detection angle can only be observed at treatment fluence of 8 mJ· cm −2 , similarly as in case of pristine PET. This phenomenon most likely originates from considerably low absorbed energy when the ablation and photothermal cleavage of macromolecules is low.
To reveal the mechanism of surface structure development on laser treated polymer, identification of the rate of material ablation was accomplished. To address this issue, ICP-MS analysis of the water bath used in underwater laser treatment, aiming to total carbon content (TCC), was accomplished. Those results are summarized in Figure 7a,b. Changes in the TCC correspond to the release of low molecular weight polymer ablation product to water environment due to laser modification. Figure 7a shows dependence of TCC on applied voltage at fixed laser fluence of 8 mJ· cm −2 . Compared to measured TCC background in fresh water (blue line, 1.2 ± 0.3 mg· L −1 ), TCC in the bath increased approximately one order of magnitude after laser processing. It is clearly visible that applied voltage has no influence on TCC, which is in good concordance with results obtained from AR-XPS analysis, suggesting limited material removal from the PET surface at low fluencies. Additionally to that, Figure 7b summarizes the values of TCC measured for different laser fluence applied at fixed voltage of 25 kV. From the presented data, one can see exponential dependence of the amount of ablated material on applied fluence. As the absorbed energy grows, increase of TCC is detected, pointing to massive polymer ablation. Data presented in Figures 5 and 6  Additionally to that, Figure 7b summarizes the values of TCC measured for different laser fluence applied at fixed voltage of 25 kV. From the presented data, one can see exponential dependence of the amount of ablated material on applied fluence. As the absorbed energy grows, increase of TCC is detected, pointing to massive polymer ablation. Data presented in Figures 5 and 6 unambiguously demonstrate the enrichment of the water bath by the low-molecular polymer degradation products, which prove that ablation mechanism is the governing process of surface nanostructure formation. Besides others, this finding contributed to reveal one of the challenges of present material research, namely the mechanism of LIPPS formation. Great efforts have been made to address this issue so far, considering various approaches involving UV-Vis spectrometry [17] or gravimetry [18], which outputs were often loaded with considerable statistic error. Right the use of water bath and thus capturing of low molecular weight product of ablation process into the media allowed to determine TCC by ICP-MS method.

Conclusions
In this work we studied surface morphology and chemistry of polyethylene terephthalate after underwater laser processing in broad scale of applied laser fluencies and operating voltages. We determined the interval of laser fluencies causing development of typical "worm-like" structure on PET. While the value of laser fluence was crucial with respect to the development of specific nanostructure, the effect of operating voltage lied especially in gentle modulation of primary "worm-like" nanostructure at considerably low laser fluencies. Most reproducible structures were prepared at operating voltages closed to the maxima of laser efficiency, where the beam profile exhibited considerable stability. Increase of laser fluence significantly supported photochemical decomposition of polymer macromolecules, resulting in massive material removal which unified the chemical composition of the surface in entire accessible depth of angle resolved XPS analysis. Data from ICP-MS unambiguously proved that ablation mechanism is the governing process of surface nanostructure formation in the process of underwater laser processing of PET.