Pedestal High-Contrast Gratings for Biosensing

Round 1

Reviewer 1 Report

This manuscript describes a type of high contrast grating that may be employed for biosensing. The pedestal high contrast grating is a rib structure that is undercut to produce a structure that is elevated above the substrate by a pedestal. The undercutting of the amorphous silicon structure provides a high refractive index contrast on nearly all four sides of the structure and allows for a sensitive response to the adsorption of target analytes.

The paper needs some additional information before being in a form that is suitable for publication. It is somewhat incomplete in its current form.

• Figure 1 shows a schematic of the grating structures and Figure 2 shows SEM images; however, it would be useful for the reader to see a schematic of the experimental configuration. Show the relation of the beam and detector so it is clear how the measurement is made. An accompanying paragraph is also necessary.
• Line 35: I suggest changing “expose” to “exhibit”. The former does not make sense.
• Line 48: “does” should be “do”
• Line 59: BRIS is a terrible acronym given the actual definition of the word. It would be sufficient to refer to it as bulk in subscripts and the like. I suggest changing this. Furthermore, the manuscript is loaded with acronyms that seem unnecessary. I recommend reducing the number of acronyms used.
• Line 59: An “e” was added to “glycerol”. Correct this misspelling.
• With regard to the SEM of the pedestal grating shown in Figure 2, how much was the device undercut? How deep is the gap under the pedestal? When immersed in the liquid, are you sure that the solution fills the gap and does not produce an air bubble?
• At several points in the document, the authors refer to “data not shown” (lines 103 and 114). Could this material be added to a supplemental document for the reader?
• Comment: Figure 4 shows large shifts which correspond to large refractive index changes. To be honest, the reality is when detecting trace analytes such refractive index changes (0.1) would not be seen. The effective index of additions of small quantities of biomaterial would not produce this kind of change and when they get dense enough to do so then there is no need for a sensitive detector.
• With respect to the data shown in Figure 5, the authors note a larger response with the pedestal structure and argue that it is due to an increase in the exposed area. I’m skeptical. Is it proportional to the observed area increase? Is it simply because of the increased refractive index contrast? Were experiments performed with different degrees of undercutting to understand the area effect?
• Reference 1: this is the template reference. Remove it and correct all the references thereafter.

Overall, the manuscript has potential, but it is lacking in details in its present form. The authors should clarify some points by including additional figures and text. Once revised the manuscript could be considered again for publication.

Author Response

Overall, the manuscript has potential, but it is lacking in details in its present form. The authors should clarify some points by including additional figures and text. Once revised the manuscript could be considered again for publication.

We would like to thank the reviewer for the overall positive evaluation of our manuscript and valuable suggestions on how to improve it. Below are our responses to reviewer’s comments, which are also highlighted in the modified manuscript. Furthermore, we went carefully through the entire manuscript text, striving to improve the English throughout. For convenience or reading, these changes were not highlighted.

1. Figure 1 shows a schematic of the grating structures and Figure 2 shows SEM images; however, it would be useful for the reader to see a schematic of the experimental configuration. Show the relation of the beam and detector so it is clear how the measurement is made. An accompanying paragraph is also necessary.
   A schematic representation of the optical setup was added to Figure 2, panel (d), together with appropriate descriptions of the figure.
2. Line 35: I suggest changing “expose” to “exhibit”. The former does not make sense.
   Corrected.
3. Line 48: “does” should be “do”
   Corrected.
4. Line 59: BRIS is a terrible acronym given the actual definition of the word. It would be sufficient to refer to it as bulk in subscripts and the like. I suggest changing this. Furthermore, the manuscript is loaded with acronyms that seem unnecessary. I recommend reducing the number of acronyms used.
   We made the corresponding changes in the text (blue-highlighted), as well as in Figure 4b.
5. Line 59: An “e” was added to “glycerol”. Correct this misspelling.
   Corrected.
6. With regard to the SEM of the pedestal grating shown in Figure 2, how much was the device undercut? How deep is the gap under the pedestal? When immersed in the liquid, are you sure that the solution fills the gap and does not produce an air bubble?
   The grating has been undercut approximately at 120 nm, providing an increase in surface area of ​​about 15%. This value coincides approximately with the data on the improvement in surface and bulk sensitivity. The gap under the grating bar is also 120 nm if counted from the edge of this grating, since the “supporting pillar” is rounded due to the specificity of the etch process. It is difficult to judge as accurately as possible whether the formation of air bubbles occurs or not. We assume that since the data on the increase in effective area have fairly similar values ​​for both theoretically calculated and experimentally measured values ​​for volume and surface sensitivities, then the creation of air bubbles is minimal.
7. At several points in the document, the authors refer to “data not shown” (lines 103 and 114). Could this material be added to a supplemental document for the reader?
   Thank you for pointing this out. Since this data was not the core part of our work and was based on previous results, we chose not to include it in the manuscript. Instead, we now reference the original works that our experiments were based on, and we edited the text to highlight this. For the former line 104, we now cite the M. Sc. thesis Hussein El Dib, S.E.; Nanostructures for Bio-Chemo Sensing, and for the former line 114, we now cite  https://0-doi-org.brum.beds.ac.uk/10.1002/sia.5311, from which our protocol was adapted.
8. Comment: Figure 4 shows large shifts which correspond to large refractive index changes. To be honest, the reality is when detecting trace analytes such refractive index changes (0.1) would not be seen. The effective index of additions of small quantities of biomaterial would not produce this kind of change and when they get dense enough to do so then there is no need for a sensitive detector.
   It is absolutely true that when it comes to the real analyte, the detectable shifts will be much smaller than shown in Figure 4. For example, this can be seen in Figure 5c, where the detected shifts are below 1 nm, while in Figure 4 they are in the order of tens of nm. However, these measurements were taken to evaluate the volumetric (bulk) sensitivity of our device , for which we used water-glycerol solutions with different refractive indices 1.33-1.47 to mimic a dielectric environment relevant to label-free assays. A large number of points are presented in order to try to estimate this in linear scale as accurately as possible.
9. With respect to the data shown in Figure 5, the authors note a larger response with the pedestal structure and argue that it is due to an increase in the exposed area. I’m skeptical. Is it proportional to the observed area increase? Is it simply because of the increased refractive index contrast? Were experiments performed with different degrees of undercutting to understand the area effect?
   Experiments with varying degrees of undercutting have not been carried out, although this is an interesting point and may require further research in more detail. Validating the leading mechanism of improved response for the pedestal HCGs is beyond the scope of this paper, as this has been previously demonstrated. Articles https://0-doi-org.brum.beds.ac.uk/10.1039/C3LC50177A and https://0-doi-org.brum.beds.ac.uk/10.1021/acsami.7b15396 ) describe that the main mechanisms for improving the sensitivity of pedestal structures are associated with the increase in surface area, which allows more portions of analyte to interact with the electric fields supported by the structure, and with narrower resonances, (higher Q-factors), appeared due to the reduced influence of the substrate. To differentiate the individual contributions of each of these reasons is currently hardly feasible. In the revised manuscript, we added these two references in support of our statement.
10. Reference 1: this is the template reference. Remove it and correct all the references thereafter.
    Corrected.

Author Response File: Author Response.pdf

Reviewer 2 Report

In this paper, Leonid Yu. Beliaev et al. reported pedestal high-contrast gratings (HCGs) for better sensing performance than conventional HCGs. Their results showed pedestal HCGs have enhanced sensitivity, FOM, LoD, and LoQ, compared to conventional ones. I think their results are interesting, but their discussion is not convincing enough, and much key information is missing in their paper, thus I cannot recommend publishing their results in the current form. My comments are listed below:

(1) The authors didn’t discuss why they choose pedestal HCGs, i.e., why the pedestal design is better than the conventional design.

(2) The authors should add simulations to show the field profiles of both conventional and pedestal HCGs, which can help to explain why pedestal ones have enhanced performances. The simulations also illustrate the field distribution, which helps to design better sensors.

(3) For HCGs, the thickness of the high-index layer is critical, so the authors should explain why the Si layer with a thickness of 500nm was chosen. For instance, the authors should illustrate whether the thickness of 500nm gives the best performance compared to other thicknesses, at least through simulations.

(4) Some experimental details should be added. For instance, the authors should mention how did they deposit thick aluminum, hafnium, and titanium oxides.

(5) Authors should calculate the Q factors of their conventional HCGs and Pedestal HCGs.

(6) In figure 4, the reflectance dips of both HCG and Pedestal HCG are in the spectral range from 1450 nm to 1550 nm, while in figure 5, the reflectance dips are in the spectral range from 1286 nm to 1302 nm. Can authors explain why they choose those different spectral ranges and what are the differences between those conventional HCG and Pedestal HCG samples?

(7) In figure 5c, the data point around 10ng/mL significantly deviates the fitting curve, can the authors explain why?

Author Response

We would like to thank the reviewer for taking the time to review our manuscript. We would like to point out that many of the experiments and questions posed by the reviewer are discussed in our previous paper, G. Finco et al., Guided-mode resonance on pedestal and half-buried high-contrast gratings for biosensing applications. Nanophotonics 2021, 10, 4289–4296 https://0-doi-org.brum.beds.ac.uk/10.1515/nanoph-2021-0347 (Ref. 30).. Below are our point by point answers to each comment.

1. The authors didn’t discuss why they choose pedestal HCGs, i.e., why the pedestal design is better than the conventional design.
   Our reason for choosing pedestal HCG is given in the introduction: "The pedestal gratings offer a larger surface area for the analyte to interact with electric fields of GMRs supported in the grating [37– 39 ], which enables larger resonance shift in the presence of the analyte. Another advantage of the pedestal Si gratings lies in their even narrower resonance compared to conventional HCGs, which is advantageous for refractometric sensing. These two properties contribute to an overall enhanced performance of PHCGs over conventional gratings [30]."
   We blue-highlighted these sentences in the main text.
2. The authors should add simulations to show the field profiles of both conventional and pedestal HCGs, which can help to explain why pedestal ones have enhanced performances. The simulations also illustrate the field distribution, which helps to design better sensors.
   The simulation results and the field profiles of pedestal structures are already presented in our previous article, which is why we instead of providing them, refer to these results using Ref. [30].
3. For HCGs, the thickness of the high-index layer is critical, so the authors should explain why the Si layer with a thickness of 500nm was chosen. For instance, the authors should illustrate whether the thickness of 500nm gives the best performance compared to other thicknesses, at least through simulations.
   It is definitely important to optimize the thickness of the amorphous silicon layer in order to obtain the best performance. Our choice of thickness is based on the results  of optimization reported in our theoretical article (Ref. [30]). 
4. Some experimental details should be added. For instance, the authors should mention how did they deposit thick aluminum, hafnium, and titanium oxides.
   These details were already described in the Fabrication section of our manuscript: "For the surface sensitivity measurements, the structures were coated with 2-5 nm thick layers of Al2O3, HfO2 and TiO2 using ALD. The ALD of oxides is based on two types of precursors – H2Oas the oxidation agent and trimethylaluminum (TMA), titanium tetrachloride (TiCl4) or tetrakis (ethylmethylamido) hafnium (TEMAHf) as sources of aluminum, titanium and hafnium, respectively. The deposition process was carried out at 200 ◦C, 150 ◦C and 350 ◦C and the deposition rates were 0.097 nm/cycle, 0.0385 nm/cycle and 0.090 nm/cycle, correspondingly"
   We blue-highlighted these sentences in the main text.
5. Authors should calculate the Q factors of their conventional HCGs and Pedestal HCGs.
   Q-factor, as well as other parameters, were calculated in our previous article (Ref. [30]). In general, the main trend that the Q-factor is better for the pedestal structure remains unchanged. However, the theoretical data values are higher than the experimental ones. Possible explanations may be in the imperfection of the nanofabrication process, as well as in the limitation of the resolution of the receiving device for optical measurements.
6. In figure 4, the reflectance dips of both HCG and Pedestal HCG are in the spectral range from 1450 nm to 1550 nm, while in figure 5, the reflectance dips are in the spectral range from 1286 nm to 1302 nm. Can authors explain why they choose those different spectral ranges and what are the differences between those conventional HCG and Pedestal HCG samples?
   The main difference is that the data shown in Figure 4 with the position of the resonance in the region of 1450-1550 nm, was obtained when our devices were immersed in a liquid, whereas Figure 5 shows data measured in air. The air background is the reason for the different position of the resonance dip. Our devices were designed to give a reflection dip in the region of 1450-1550 nm in aqueous solutions.
7. In figure 5c, the data point around 10ng/mL significantly deviates the fitting curve, can the authors explain why?
   We cannot explain why this data point deviates so significantly from the curve for the conventional HCG and we do not wish to speculate on this topic. However, we would like to point out that the data for the pedestal-HCG does not exhibit any similar anomalies.

Author Response File: Author Response.pdf

Reviewer 3 Report

In this study, authors present the experimental study of the biosensing platform based on designed pedestal high-contrast gratings. Previously authors investigated these structures numerically and this is a very interesting result that the enhancement factor was the same in theory and experiment. The results may be very useful for the refractive-based biosensing.
However, since this study is about the practical platform for biosensing it is important to show the errors of all measurements. The authors presented errors in figure 5c, but it would be nice to do that for figures 4b and 5a, also for all values in Table 1, and for values of improvement.  

Author Response

In this study, authors present the experimental study of the biosensing platform based on designed pedestal high-contrast gratings. Previously authors investigated these structures numerically and this is a very interesting result that the enhancement factor was the same in theory and experiment. The results may be very useful for the refractive-based biosensing.

However, since this study is about the practical platform for biosensing it is important to show the errors of all measurements. The authors presented errors in figure 5c, but it would be nice to do that for figures 4b and 5a, also for all values in Table 1, and for values of improvement.

We would like to thank the reviewer for the positive evaluation of our manuscript. In the revised manuscript, we added the error values as suggested (blue-highlighted).

Author Response File: Author Response.pdf

Reviewer 4 Report

In this paper the authors describe an experimental comparison of two different types of nanostructures for biosensing applications: high contrast gratings (HCG) and pedestal high contrast gratings (PHCG). To my knowledge this is a novel study and the results should be of interest to readers of Nanomaterials. The nanofabrication is generally well-described, and the experimental methods are also clearly presented. However, the authors need to provide a better explanation of the data analysis, since some aspects are unclear. In terms of detailed comments:

1. In section 1 it would be helpful if the authors would provide the reference to their own numerical analysis of PHCG structures (ref 31) in this section and describe the numerically predicted sensitivity enhancement.
2. In section 2 (fabrication) how were the dimensions for the gratings selected?
3. Can you please provide more details on the HF etch process which forms the pedestals.
4. In section 3, figure 4; since the authors have a numerical model for the refractometric response of the structure, it would be helpful to compare the measured results with numerical predictions.
5. Equations 5 and 6: I suggest it would be better stylistically to present a single equation with parameters and then give the values for the parameters in the text.
6. Figure 5(c) : I was curious about the equations and so I plotted the figure myself. However, I did not obtain the same results. When I use the data provided I get graphs where the HCG gives a greater response than the PHCG (see attached file). Also, the intercept is below zero, which does not seem to make physical sense. Can the authors check this please?
7. The authors report much improved LOD and LOQ for the PHCG structure. Since this increase cannot be explained from the improvement in sensitivity alone, it must come from the reduction in standard deviation. Can the authors report on the standard deviations of the measurements so that we can verify.

Comments for author File: Comments.pdf

Author Response

In this paper the authors describe an experimental comparison of two different types of nanostructures for biosensing applications: high contrast gratings (HCG) and pedestal high contrast gratings (PHCG). To my knowledge this is a novel study and the results should be of interest to readers of Nanomaterials. The nanofabrication is generally well-described, and the experimental methods are also clearly presented. However, the authors need to provide a better explanation of the data analysis, since some aspects are unclear. 

We would like to thank the reviewer for overall positive evaluation and valuable comments on how to improve the quality of our manuscript. Our point-by-point responses to the reviewer’s comments are given below.. 

1. In section 1 it would be helpful if the authors would provide the reference to their own numerical analysis of PHCG structures (ref 31) in this section and describe the numerically predicted sensitivity enhancement.
   In accordance with the reviewer's note, we added the following sentence to the first section (blue-highlighted):
   “According to the numerical data in [30], the PHCGs showed a 12% improvement in the bulk refractive index sensitivity.”
   
2. In section 2 (fabrication) how were the dimensions for the gratings selected?
   The choice of the grating dimensions was determined by the numerical data described in our previous article (Ref. [30]), as well as the parameters of the nanofabrication process itself. In the theoretical article, we found out the optimum parameters for SOI wafer which are: aSi thickness 500 nm and 1100 nm of  SiO₂ layer. This information is now included in the manuscript. The width of the bars fabricated experimentally was however reduced compared to the optimal one theoretically predicted, from 390 to 340 nm, in order to increase the reproducibility of the fabrication for the pedestal and conventional high contrast grating structures. 
3. Can you please provide more details on the HF etch process which forms the pedestals.
   The following sentences were added in the main text and blue-highlighted: “The process includes a pressure and temperature stabilization in the presence of N₂ (1425 sccm) and (EtOH) 210 sccm gasses; etch step, when HF 190 sccm flow is introduced into the chamber and the pumping. The etch time was found to be 600 seconds.” 
4. In section 3, figure 4; since the authors have a numerical model for the refractometric response of the structure, it would be helpful to compare the measured results with numerical predictions.
   Since the structures presented in this article are not identical to those from the numerical calculations article, we chose not to include a detailed comparison. Instead, we simply say that “These results are in full agreement with modeling of similar structures, where a 12% improvement in surface sensitivity was shown [30]."
5. Equations 5 and 6: I suggest it would be better stylistically to present a single equation with parameters and then give the values for the parameters in the text.
   Changed according to comments. The following part were added in the main text and blue-highlighted: “where C is the concentration of the analyte, A₁, A₂, A₃ and A₄  – fitted constants that were equal to 0.58919, 0.63468, 15.87538, 0.51237 and 0.58548, 0.6989, 9.90319, 0.41903 for the PHCG and HCG, respectively. The values R² for PHCG and HCG were 0.98656 and 0.9487, respectively.”
6. Figure 5(c) : I was curious about the equations and so I plotted the figure myself. However, I did not obtain the same results. When I use the data provided I get graphs where the HCG gives a greater response than the PHCG (see attached file). Also, the intercept is below zero, which does not seem to make physical sense. Can the authors check this please?
   The reviewer's comment is absolutely correct. Figure 5c is presented correctly, as well as the parameter R², but we accidentally mislabelled the coefficients in the text (coefficients for HCG should have been for PCHG and vice versa). We corrected this in the revised manuscript. The fact that the intercept is below zero for the conventional HCG does indeed have no physical meaning. This is a result of averaging, and is most likely due to the imperfections of the experiments.
7. The authors report much improved LOD and LOQ for the PHCG structure. Since this increase cannot be explained from the improvement in sensitivity alone, it must come from the reduction in standard deviation. Can the authors report on the standard deviations of the measurements so that we can verify.
   Certainly. The standard deviation for the measurements taken for the pedestal structure at the lowest concentration was ±0.04 nm, while for the conventional one it was ±0.05 nm. If plotted in the correct order, then the same LoD and LoQ values for the pedestal and regular structures should be obtained.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors have satisfactorily addressed all my comments. While I still suggest the authors add some simulated field distributions in the new manuscript and experimental obtained Q-factors for readers' convenience.