Microfabrication and Surface Functionalization of Soda Lime Glass through Direct Laser Interference Patterning

Round 1

Reviewer 1 Report

The manuscript "Interference-based laser direct nano/microstructuring  of soda lime glass via non-linear absorption" describes a technique to fabricate (sub)micronanostructures on soda lime glasses surfaces. It also analyzses some optical and chemical chaceretistics (optical grating dispersion and wetting angle measurements).

The manuscript is well written but I think it overstretches itself in the scope as it discusses too many things. This generates too many different discussions that, while they are correctly stablished, they do not developed its fully scientific potential as there is no space in them for the paper. This damage the readibility of the paper and blurs the most interesting conlusion:

 

1. The title itself it is not entirely accurate:"Interference-based laser direct nano/microstructuring  of soda lime glass via non-linear absorption". It should read "Optical gratings and water wetting analysis of interference-based laser direct nano/microstructures  fabricated on soda lime glass" or something  close to this. The characterization of the applications takes significant space in the manuscript. If the authors would like to stick to the title they should get rid of the optical/chdemical characterization  of the fabricated structures.
2.  The most interesting part is the fabrication results. LAser interference has traditionally been applied to high absorption materials such as metals or photoresists. The application to transparent materials is something new and the results are very interesting. The authors decreased the pulse time down to the tens of ps to active the nonlinear regime.  But the number of results are limited and I would like to see here a larger set of experimental conditions, and a broader discussion on the origins of the "narrow window" to pattern with DLIP. Why is it so narrow? Absorption limited? Themal diffusion limited? Both?
3. It is surprising that the authors do not discuss the onset and the quantitative effects of the non-linear absoprtion that enables the (sub)micrometric estructures and that should have a lead role in understanting the generated patterns. Can you provide a (rough) model? Cab you show what happens to the non-linear effect with the 2 pulse times  (12ps or 70ps ) used in the papers?
4. The discussion on optical properties, is too descriptive. This may need some more insight and it propably can make a paper on ts own. Because of the surface  (as shown in figure 3) not being as clearly defined as standard gratings made by etching or lithography, the application does not make technical sense. I agree that there can be some room for scientific discussion of the properties. But this is also too short as there is no model (RCWA or FDTD), no clear quantitative estimation of the scattering lossess of the structure, no direct quantitative stimation of the effect of the non constant depth of the etching... There might be some change of refractive index below the etched pattern which may have an effect on the difraction. Also, can the authors try to smooth the surface structures  with some chemicals? HF maybe?
5. Figure 6 and figure 7 have the efficincy scales in different directions (probably figure 7 is upside down).
6. The wetting characteristics are also presented. This is unrelated to the optical properties and somehow adds additional noise to the paper. The results here are intriguing as the wetting angle is independent of the roughness factor (there seems to be a dependence on the spatial period). This however cannot be explained by the current data and, hence, any conclusion about the origin of the wetting properties remains open. The only valid conclusion is that the wetting angle is dramatically reduced by the interference process. This could be good enough given the experimental approach of the manuscript but it does not add significant novelty.

 

Author Response

We would like to thank the Reviewer for her/his detailed feedback on our manuscript that allowed us to enhance the quality of the text.

 

Query 1: The title itself it is not entirely accurate: "Interference-based laser direct nano/microstructuring  of soda lime glass via non-linear absorption". It should read "Optical gratings and water wetting analysis of interference-based laser direct nano/microstructures  fabricated on soda lime glass" or something  close to this. The characterization of the applications takes significant space in the manuscript. If the authors would like to stick to the title they should get rid of the optical/chdemical characterization  of the fabricated structures.

 

Answer 1: We thank the Reviewer for his/her hint regarding the title. We agree with the comment and therefore we changed the title to “Microfabrication and surface functionalization of soda lime glass through Direct Laser Interference Patterning”

 

Query 2: The most interesting part is the fabrication results. LAser interference has traditionally been applied to high absorption materials such as metals or photoresists. The application to transparent materials is something new and the results are very interesting. The authors decreased the pulse time down to the tens of ps to active the nonlinear regime.  But the number of results are limited and I would like to see here a larger set of experimental conditions, and a broader discussion on the origins of the "narrow window" to pattern with DLIP. Why is it so narrow? Absorption limited? Themal diffusion limited? Both?

 

Answer 2: We thank the Reviewer for the comment regarding the processing conditions. Indeed this is an important point and, as the Reviewer suggested, needed to be emphasized in the manuscript. Nevertheless, we strongly believe that the amount of experimental work regarding the microfabrication is sufficient for the reader to understand the behavior of glass under ultrashort pulsed interference patterning and, eventually, to reproduce these results in another lab. On the other hand, the limited processing window is strictly related to the multi-photon nature of the ablation induced through our method. Following the Reviewer’s suggestion, we updated the introductory part as follows, addressing the points raised in this query:

 

“This topic has received considerable attention in recent years and the main processes can be summarized as follows:

• Multiphoton ionization: an electron in the valence band can absorb several visible or near-infrared photons and gain enough energy to cross the band gap [47].
• Tunneling photoionization: the strong electric field suppresses the Coulomb barrier and allows the electron to tunnel through. The free electrons created by nonlinear photoionization can then absorb more energy from the laser pulse by inverse bremsstrahlung. If the energy of the free carriers becomes high enough, they can also promote an electron from the valence to the conduction band by impact ionization leading to an avalanche process [48 ].

One of the most distinguishing features of two-photon absorption is that the amount of absorbed power in a thin layer of the medium is proportional to the square of the light intensity (or power), while in one-photon absorption the ratio of absorption depends linearly with respect to intensity [49]. Therefore, considering a Gaussian beam, the absorption rate drops quadratically moving from the center towards the periphery of the beam and any linear variation in the power would produce a quadratic variation of the absorption coefficient and, therefore, of the ablation depth [50]. As a result, the process window for the two-photon based microfabrication in non-absorbing materials is very narrow, compared with linear absorption.”

 

Also section 3.1 has been updated with the following sentence:

“As mentioned in Section 1, since the onset of non-linear absorption mechanisms strictly depends on the laser fluence dose, slight variations in pulse energy or pulse overlap may lead to large differences in the texturing results.”

 

Query 3: It is surprising that the authors do not discuss the onset and the quantitative effects of the non-linear absoprtion that enables the (sub)micrometric estructures and that should have a lead role in understanting the generated patterns. Can you provide a (rough) model? Cab you show what happens to the non-linear effect with the 2 pulse times  (12ps or 70ps ) used in the papers?

 

Answer 3: Non-linear absorption was identified as the optical mechanism, by which technical glasses samples absorb the laser energy allowing the fabrication of the surface micropatterns in agreement with many previous reports (Niino, H. et al In Proceedings of the Photon Processing in Microelectronics and Photonics III; International Society for Optics and Photonics, 2004; Vol. 5339, pp. 112–117; Herman, P.R. et al Appl. Surf. Sci. 2000, 154–155, 577–586; Schwerter, M. et al. J. Microelectromech. S. 2016, 25, 517–523; Ihlemann, J. et al. Appl. Surf. Sci. 1996, 106, 282–286; Stoian, R. et al. Appl. Phys. Lett. 2002, 80, 353–355; Du, D. et al. Appl. Phys. Lett. 1994, 64, 3071–3073.). However, it is not the scope of this work to dig into the physical processes that lead to the modelling and quantification of the intertwined microscopic processes typical of non-linear absorption. Although DLIP ablation of transparent materials has not been extensively investigated so far, several works have modeled and discussed the multi-photon absorption processes in transparent materials assuming a single incident beam. Particularly, Sun and coworkers (Sun, M. et al. Laser Ablation Mechanism of Transparent Dielectrics with Picosecond Laser Pulses. In Proceedings of the Laser-Induced Damage in Optical Materials: 2012; International Society for Optics and Photonics, 2012; Vol. 8530, p. 853007) have modeled the ablation in soda lime glass considering the ultrafast dynamics of free-electrons using a rate equation for free-electron density including multi-photon ionization, avalanche ionization and loss terms. With this model they found a relationship between the onset of ablation with the laser fluence and pulse duration (Fig. 3b of their article). Interestingly, that model predicts that using the pulse durations (12 and 70 ps) and laser fluences (1.7-5 J/cm²) employed in our work the process lies in the ablation regime. For the same pulse durations, the model predicts a fluence threshold in the range 05-1 J/cm², which agrees also with our preliminary results.

The text was modified as follows:

 

“Although DLIP ablation of transparent materials has not been extensively investigated so far, several works have modeled and discussed the multi-photon absorption processes in transparent materials assuming a single incident beam. Particularly, Sun et al. have modeled ablation in soda lime glass considering the ultrafast dynamics of free-electrons using a rate equation for free-electron density including multi-photon ionization, avalanche ionization and loss terms [72]. Interestingly, that model predicts that using pulse durations in the picosecond regime and laser fluences in the 1.7 - 5 J/cm2 range, as employed in our work, the process lies in the multi-photon ablation regime, in accordance with the results presented in this section.”

 

 

Query 4: The discussion on optical properties, is too descriptive. This may need some more insight and it propably can make a paper on ts own. Because of the surface  (as shown in figure 3) not being as clearly defined as standard gratings made by etching or lithography, the application does not make technical sense. I agree that there can be some room for scientific discussion of the properties. But this is also too short as there is no model (RCWA or FDTD), no clear quantitative estimation of the scattering lossess of the structure, no direct quantitative stimation of the effect of the non constant depth of the etching... There might be some change of refractive index below the etched pattern which may have an effect on the difraction. Also, can the authors try to smooth the surface structures  with some chemicals? HF maybe?

 

Answer 4: We would like to thank the Reviewer for suggesting new insights to further discuss the optical properties of our samples. We would like to stress that the main advantage of DLIP is that it is a single-step process that does not require chemicals, masks or stringent ambient conditions as in a clean room. Thus, we believe that our work can have a technological relevance if the main features of DLIP process can be exploited and consequently, we ruled out further post-processing steps such as chemical etching or ultra-sound cleaning. We agree that modelling and further optimizing the optical properties could deserve a paper on its own, but here we wanted to focus on the feasibility of using DLIP to structure and functionalize conventional glasses. Although relatively complex numerical models based on FEM, RCWA or FDTD were not used here, we compared the diffraction efficiency of the fabricated quasi-sinusoidal gratings with that of perfectly sinusoidal thin phase gratings according to the simpler Fraunhofer diffraction theory. We assume that the difference of the diffraction efficiencies of the fabricated and theoretical gratings comes mainly from the lack of uniformity of the fabricated patterns. Further process optimization procedures should then aim to enhance the homogeneity of the gratings to increase the optical performance. On another note, we cannot discard a change in the refractive index of glass, as suggested by the Reviewer and observed already in previous reports. However, to detect such small changes (in the order of ~10^-3) on the rough glass surfaces, specific equipment must be used, which is not currently available at our labs. Although this point can also be interesting to further understand the microscopic physical processes at the surface of glass upon irradiation with ultra-short laser pulses, it lies beyond the scope of the present study.

 

Query 5: Figure 6 and figure 7 have the efficincy scales in different directions (probably figure 7 is upside down).

 

Answer 5: We thank the Reviewer for spotting the mistake. We corrected the colorbar of Figure 6 in the new version of the manuscript to be consistent with the colorbars of the other plots.

 

Query 6: The wetting characteristics are also presented. This is unrelated to the optical properties and somehow adds additional noise to the paper. The results here are intriguing as the wetting angle is independent of the roughness factor (there seems to be a dependence on the spatial period). This however cannot be explained by the current data and, hence, any conclusion about the origin of the wetting properties remains open. The only valid conclusion is that the wetting angle is dramatically reduced by the interference process. This could be good enough given the experimental approach of the manuscript but it does not add significant novelty.

 

Answer 6:   We appreciate the comment of the Reviewer and the critics about the wetting results and novelty of the work. Nevertheless, we disagree with the fact that the wettability analysis significantly disturbs the manuscript’s uniformity. We tried to summarize our points in the following reasons:

1. The description of the wettability results takes place in a separated section from the rest; therefore, we see no particular interruption in the topics written in the manuscript.
2. Following the previous answers to the Reviewer’s query 4, we would like to remark that the wettability analyses are part of the characterization of the fabricated surfaces and do bring an added value to the research. Although not carried out with fully analyses on the surface tension, the contact angle measurements that we proposed are routine measurements carried out in most of labs dealing with surface engineering and microfabrication and help future researchers to have a reference on interference texturing of glass.
3. As the Reviewer states, all data-points are consistent with each other and show that microtexturing significantly lowers the surface tension of the glass surface. The origin of this behavior can be described with the Wenzel wetting model, as reported in the manuscript. In particular, the glass surface, already slightly hydrophilic, experiences a much lower contact angle when microstructures are fabricated on it. However, the data points in Figure 5 do not follow the trend of the Wenzel equation and this can be ascribed to the fact that, as it can be observed in the SEM images, a nanoroughness is created simultaneously to the interference process which, however, is below the resolution of the confocal microscope employed for retrieving the Wenzel roughness factor. Therefore, it can be assumed that the real roughness of the samples is higher than the one measured and that the contact angle measured can be ascribed to the interaction of a water droplet with a hierarchically rough surface. We believe that this has been already explained in the last paragraph of section 3.2.
4. The characterization of the surface functionalities, optical and wettability properties in this case, was devised as a mean to test the feasibility of DLIP to produce multi-functional structured surfaces on glass. We believe that the ability to fabricate such surfaces have a technical relevance because, as stated in the introduction, glass substrates featuring superhydrophilicity and light management ability, for instance by diffraction, can be used in different application fields, for example for self-cleaning solar modules with increased efficiency.

Reviewer 2 Report

This article reads well, which means presentation of the results is good. Even though this article appears to focus more on the laser processing (or modification) of glass rather than laser-glass interactions, the contents described in it would deserve publication in the Journal.    

Author Response

We thank the Reviewer for her/his positive feedback and appreciate the time she/he spent on our manuscript.

Reviewer 3 Report

This paper presents results on the structuring of glass surfaces using interference-based laser writing exploiting non-linear absorption. This is an interesting topic, however, some information is missing in order to fully access the importance of the manuscript.

The paper state that lithography is the most widespread method for making microstructures on glass, stating that laser based structuring has the advantage of fewer fabrication steps and no use of hazardous chemicals, however, a very critical issue  is fabrication speed and though the paper does give some information on fluences, I am missing, for example in table 1 clear statements of patterning speed: how many square centimeters can be patterend pr. second. for the different pattern structures.

In the introduction I am missing a discussion of the fact that Hydrophilic surfaces will be oliophobic under water.  Micron-scale structuring of optically transparent materials with an aspect ratio that maintains at least partially the transparency is interesting for making under-water oliophobic windows with superior under-water self-cleaning surfaces. This has been demonstrated recently on sapphire and should be discussed in the introduction (Nanoletters, Pillars or Pancakes, N. Akthar et al, 2018).

I also miss in the introduction a clear discussion of the application of these windows. It would be interesting to measure the 0-order transmisison of a normal incidence plane wave of light, to access how suitable they would be as "normal windows".

I am also missing in the introduction a discussion of whether this fabrication principle can be easily extended to other transparant substrates such as SiC, Quartz and Sapphire.

The comment above ties in with my major objection to the paper. It comes across essentially as a "trial and error" report, there is no support of the results by theoretical modelling, predicting the expected features, would it for example be possible to surpress the nanofeatures in order to get higher 0-order transmission? Without a theoretical framework the impact of the paper is limited.

Further comments:

In section 2.2. on the setup. It should be clarified what laser spotsizes are used.

Problem with reference on line 132

Table 1 Please include a pattern writing speed in the form of cm2/s, also desired aspect ratio parameter not included here.

I would like an additional table with some numbers stating how well the production parameters in table 1 were reflected in the real results. What were the optimum fluence for producting specific features. It is discussedto some extend with reference to the SEM images, and depth diagrams in figure 4 but should be quantified also in numbers.

l. 152 - were there no issues with charging when imaging the glass structures, how was this done? In figure 3 specify electron current and what detector was used for imaging (secondary emission or backscattered)

l.157. If available state what the humidity was in the lab.

l. 256. Was the flat reference glass subject to any cleaning before the contact angle was measured? Surface contaminants  in the form of physisorbed hydrocarbons can have major influence on the contact angle - this was demonstrated for sapphire a few years back (Underwater superoleophobic Sapphire(0001) surfaces, Akhtar et al, J. Phys. Chem C. 2015) and is a major topic of debate for example in the graphene community. See for example: Effect of airborne contaminants on the wettability of supported graphene and graphite, Li et al,  Nature Materials, 2013. In the extreme case the observed hydrophilicity may be due, not to the structuring, but to the removing of hydrocarbon contaminants from the glas surface through the laser processing. Did the authors check that the properties of the structured surfaces remained constant over time? This should be clarified

figure 5 - inset - based on how many measurements?

l. 251, please explain in more details how the roughness factor was measured. As pointed out by the authors themselves there are indications that it was wrongly measured.

 

 

 

 

Author Response

We would like to thank the Reviewer for her/his time invested in revising our manuscript and appreciate the comments that allowed us to improve the text.

 

Query 1: The paper state that lithography is the most widespread method for making microstructures on glass, stating that laser based structuring has the advantage of fewer fabrication steps and no use of hazardous chemicals, however, a very critical issue  is fabrication speed and though the paper does give some information on fluences, I am missing, for example in table 1 clear statements of patterning speed: how many square centimeters can be patterend pr. second. for the different pattern structures.

 

Answer 1: We agree with the Reviewer, since throughput is a key parameter for every industrial process. In the experiments carried out in our work, we have used lab-scale laser systems based on laser sources with modest power outputs of 2.7 W and 61.4 W, for the four-beams and two-beams configuration, respectively. In turn, the structuring strategy was limited by the use of linear stages with a maximum speed of 300 mm/s. The employed setups thus yielded a maximum throughput of 3.3 cm²/min for patterning glass samples with 2.3 µm line-like features (two-beam configuration) and 2.4 cm²/min for patterning hole-like textures (four-beam setup). Although these throughputs are relatively low to be attractive for industrial applications, the DLIP method can be scaled up to mass production by using laser sources with higher power output that allows larger spot sizes. In addition, optimizing the DLIP optics and employing faster beam guidance systems, such as galvoscanners or polygon scanners, can further increase the patterning speed. For instance, Lang and coworkers (Lang et al. World record in high speed laser surface microstructuring of polymer and steel using direct laser

interference patterning, Proc. SPIE 9736, Laser-based Micro- and Nanoprocessing X, 97360Z, 2016) reported the world record in DLIP processing by using a ns-laser source with a power of 180 W at 10 kHz, coupled to an in-house developed DLIP optics providing a rectangular spot (15 mm x 50 µm). Polycarbonate and stainless steel samples, that were mounted in linear axes with maximum speeds of 1 m/s, were processed at throughputs of 0.9 m²/min and 0.36 m²/min, demonstrating that this technology can be compatible with industrial requirements. Further optimizing the DLIP setup and process, it is also expected that glass structuring can reach throughputs in the same order of magnitude.

The section 2.3 was modified as follows:

“With the used DLIP setups the maximum patterning throughputs for the two-beam and four-beam configurations were 3.3 cm2/min and 2.4 cm2/min, respectively.”

The conclusions section was modified as follows:

“Although the throughputs achieved in this study (~3 cm²/min) are relatively low to be attractive for industrial applications, the DLIP method can be scaled up to mass production, e.g. throughputs ~1 m²/min, by using laser sources with higher power output that allows larger spot sizes, by optimizing the DLIP optics and employing faster beam guidance systems, such as galvoscanners or polygon scanners.”

 

Query 2: In the introduction I am missing a discussion of the fact that Hydrophilic surfaces will be oliophobic under water.  Micron-scale structuring of optically transparent materials with an aspect ratio that maintains at least partially the transparency is interesting for making under-water oliophobic windows with superior under-water self-cleaning surfaces. This has been demonstrated recently on sapphire and should be discussed in the introduction (Nanoletters, Pillars or Pancakes, N. Akthar et al, 2018).

 

Answer 2: We consider that the point that the Reviewer is mentioning is indeed is very interesting and could enhance the functionalities of nano/microstructured glass. Despite the exploitation of this aspect is beyond the scope of our study, we modified the Introduction section to provide the readers a brief insight into this application. The text then was modified as follows:

“Superhydrophilic surfaces immersed in water can repel oil contaminants due to the differences in the surface tensions of the liquids paving the way for novel oil/water separation applications [14,15]. In addition, such underwater self-cleaning surfaces can also prevent biofouling, which could be attractive for the food processing sector, biomedical devices and marine industry [16].”

 

Query 3: I also miss in the introduction a clear discussion of the application of these windows. It would be interesting to measure the 0-order transmisison of a normal incidence plane wave of light, to access how suitable they would be as "normal windows".

 

Answer 3: In the Introduction we have mentioned several applications where structured glasses can offer new or enhanced functionalities, for example in microfluidics chips that do not require pumps, for anti-fog and self-cleaning applications, and for optoelectronics devices with improved efficiencies as in thin-film Si solar cells. The zero-order transmission was indeed measured for all the patterned samples with line-like features and shown exemplarily in Fig. 6 for three selected samples (vertical stripe at an angle of 0°). Although the samples have a high global transmittance, they are not suitable for use as “normal windows” as the transmitted image appears distorted due to the rough surface. However, it is still possible to tune the filling factor of the structured area relative to the untreated glass surface in order to reduce the transmitted distortion at the expense of the efficacy of additional functionalities (light diffraction or hydrophilicity, for instance).

 

Query 4: I am also missing in the introduction a discussion of whether this fabrication principle can be easily extended to other transparant substrates such as SiC, Quartz and Sapphire.

 

Answer 4: We appreciate the Reviewer’s question about the possibility to texture other high energy-gap dielectrics and the answer is yes. On this topic some early works have been published on SiO2 and diamond using femtosecond interfering pulses and we already mentioned them in the introduction (References 62-64). Moreover, we also recently published about the texturing of Sapphire by means of interference texturing employing a sub-ps infrared laser (Reference 54).

To make this point even clearer we modified the following sentence in the Introduction:

“Although a few pioneering works on direct structuring by interfering ultra-short laser pulses on transparent ceramics, such as amorphous SiO2, sapphire, SiC, TiO2, diamond, among others, have been reported [62-64], no systematic study has been done yet to optimize the laser processing parameters and link them with controlled surface properties.”

 

Query 5: The comment above ties in with my major objection to the paper. It comes across essentially as a "trial and error" report, there is no support of the results by theoretical modelling, predicting the expected features, would it for example be possible to surpress the nanofeatures in order to get higher 0-order transmission? Without a theoretical framework the impact of the paper is limited.

 

Answer 5: In this case we do not fully agree with the Reviewer’s comment. The resulting functionalities of the textured samples were modeled with well-known models and their respective equations. For instance, Wenzel’s model was used for understanding the wettability behavior, whereas the diffraction grating equation allowed us to predict the diffraction angles as function of wavelength and diffraction order. The applicability and limitations of the models to explain the observed behaviors were also pointed out and discussed in section 3.2 and 3.3. Moreover, the diffraction efficiency of the produced diffraction gratings with a near sinusoidal shape was contrasted with that of a perfectly sinusoidal thin grating according to the Fraunhofer diffraction theory. As stated in the manuscript, the difference between the measured and calculated diffraction efficiency comes mainly from the lack of spatial uniformity of the fabricated gratings due to the chipped material, variable structure height and LIPSS. Although the presence of LIPSS is inherent to the laser process employing ultra-short laser process, further process optimization might lead to more homogeneous surface patterns and therefore higher diffraction efficiencies.

 

Query 6: In section 2.2. on the setup. It should be clarified what laser spotsizes are used.

 

Answer 6: We mentioned all the experimental details regarding the interference setup in the section 2.3 (Table 1) and we believe that this position better suits the manuscript structure.

 

Query 7: Problem with reference on line 132

 

Answer 7: We corrected the mistake.

  

Query 8: Table 1 Please include a pattern writing speed in the form of cm²/s, also desired aspect ratio parameter not included here.

 

Answer 8: As stated in our reply to query 1 of the same Reviewer, we include now the maximum throughputs achievable with the two- and four-beams DLIP systems in section 2.3, which are 3.3 and 2.4 cm²/min, respectively. The aspect ratio cannot be estimated a priori, as the ablation rate depends heavily on the material properties as well as on the process parameters, such as accumulated fluence, spatial period, repetition rate and laser wavelength.

 

Query 9: I would like an additional table with some numbers stating how well the production parameters in table 1 were reflected in the real results. What were the optimum fluence for producting specific features. It is discussedto some extend with reference to the SEM images, and depth diagrams in figure 4 but should be quantified also in numbers.

 

Answer 9: In the revised manuscript, we have included a new table as an appendix (Appendix B) where the samples with spatial periods of 2.3 µm and 3.9 µm showing the highest diffraction efficiencies together with their process parameters are listed. In addition, we have included the sample with a spatial period of 4.9 µm with the deepest microstructures and good uniformity.

  

Query 10: l. 152 - were there no issues with charging when imaging the glass structures, how was this done? In figure 3 specify electron current and what detector was used for imaging (secondary emission or backscattered)

 

Answer 10: We thank the Reviewer for point out this topic. In order to avoid charging issues during the SEM analysis, the samples where previously coated with a 30 nm thick gold layer in a sputtering chamber. Moreover, the electrons detected where secondary electrons. We updated the manuscript with these information in the materials and methods section and in Figure 3, respectively.

 

Query 11: l.157. If available state what the humidity was in the lab.

 

Answer 11: The humidity was 35%. This value is now included in the manuscript (section 2.2).

 

Query 12: l. 256. Was the flat reference glass subject to any cleaning before the contact angle was measured? Surface contaminants  in the form of physisorbed hydrocarbons can have major influence on the contact angle - this was demonstrated for sapphire a few years back (Underwater superoleophobic Sapphire(0001) surfaces, Akhtar et al, J. Phys. Chem C. 2015) and is a major topic of debate for example in the graphene community. See for example: Effect of airborne contaminants on the wettability of supported graphene and graphite, Li et al,  Nature Materials, 2013. In the extreme case the observed hydrophilicity may be due, not to the structuring, but to the removing of hydrocarbon contaminants from the glas surface through the laser processing. Did the authors check that the properties of the structured surfaces remained constant over time? This should be clarified

 

Answer 12: We thank the Reviewer for point out also this topic. As the Reviewer mentions, any surface gets contaminated by a nanometer-thick layer of organic molecules as soon as this is in contact with the atmosphere, even within some minutes. Nevertheless, we disagree with the Reviewer when he/she writes that “In the extreme case the observed hydrophilicity may be due, not to the structuring, but to the removing of hydrocarbon contaminants from the glas surface through the laser processing”.  Due to the fact that these physisorbed molecules are hydrocarbons, the organic layer is apolar which, in turn, makes the surface more hydrophobic. In fact, the mentioned paper of Li et al. (Nat. Mater. 2013, 12, 925–931) is supporting exactly this assumption and demonstrating it with their data.

However, this gave us the hint that this marginal contamination contributed to slightly increase the real contact angle of the measurements reported in Figure 5 and that, therefore, most of the values do not reach contact angles lower than 5 degrees. Following this, we updated the manuscript in Section 3.2 as follows:

“On the other hand, most of the measured values do not reach contact angles lower than 5° although the samples show high roughness values. This behavior can be ascribed to the presence of atmospheric organic contaminants on the samples surface, which deposit on the surface through physisorption (Van der Waals bonds) [87]. In fact, a common source of such accidental contamination are hydrocarbons present in ambient air and this has been demonstrated to be present even in high-end nanofabrication clean rooms [88,89]. This contamination, very common in metal surfaces, has been confirmed also on inert surfaces as SiO2 by means of spectroscopic investigations [90]. As commonly known, this thin organic layer leads to a decrease of the surface energy of the substrate and increases its hydrophobicity (i.e. an increase of the CA) [87], which may prevent the contact angles reported in this work to reach a value close to zero.”

 

Query 13: figure 5 - inset - based on how many measurements?

 

Answer 13: The inset in Figure 5 shows a box chart with the statistical distribution of the contact angle measured along and across the direction of the grooves of the 108 structured samples with line-like features. To make this point clearer we added the number of samples in section 3.2.

 

Query 14: l. 251, please explain in more details how the roughness factor was measured. As pointed out by the authors themselves there are indications that it was wrongly measured.

 

Answer 14: As defined in the text, the adimensional roughness factor r is the ratio of the real area of the sample divided by the projected area in the horizontal plane (r =1 for a flat surface, r > 1 for a rough surface). We added the following reference in the text (next to the definition of the roughness factor), so that the reader can find related information on the definition of the roughness factor and its relation with Wenzel’s model: C. Ran et al., Langmuir 2008, 24, 18, 9952–9955. In our study, the difficulty in estimating the roughness factor lies in the limited lateral resolution of the used confocal microscope, which is approximately 140 nm. Therefore, we have assumed that we have underestimated the real roughness factor considering the complex topography of the nano-scales ripples and other nano-features. We thus suggest that the mismatch between the measured contact angles and the contact angles predicted by Wenzel’s model (Fig. 5) might come from the underestimation of the roughness, i.e. the real roughness factors of the textured samples might lie above 1.6.

Reviewer 4 Report

overall, I think this paper is acceptable and have no big problem.

topic is not controversial and reasonable also experimental procedure is also general method. 

but I want to know why did you use "non-linear absorption" in title.

of course, there are no wrong information because your pattern was generated from non-linear absorption. but it seems like not main topic in your paper.

you verified non-linear absorption but I think it is just a kind of demonstration, no theoretical discussion

how you think of just using "LIPSS" instead of non-linear absorption?

 

thank you

Author Response

We thank the Reviewer for his/her comment about the title. Although not strictly proven, our intention was to make clear to the reader that this texturing approach differentiates from common interference texturing due to the fact that it can only take place under high laser energy and ultra-short pulse durations, which trigger non-linear absorption mechanisms. Following the suggestion from another Reviewer, we changed the title to: “Microfabrication and surface functionalization of soda lime glass through Direct Laser Interference Patterning”