Novel Fabry-Pérot Filter Structures for High-Performance Multispectral Imaging with a Broadband from the Visible to the Near-Infrared

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript considers the design of multispectral filter arrays (MSFA) for complementary metal-oxide-semiconductor (CMOS) image sensors based on Fabry-Perot cavity. According to the authors, the main proposals of the work are a novel dielectric-metal mirror and a novel filter structure with a dielectric thin film inside the cavity, which together provide stable maximum transmission for single-peak filtering across a broad spectral range of 400-1100 nm. The design of the filter was mainly done using the finite-difference time-domain (FDTD) calculations to obtain spectral transmission curves for the specified thicknesses of metal and dielectric layers. As an experimental evaluation of the proposed filter structure, eight large-scale samples were manufactured and characterized by scanning electron microscopy (SEM) and transmission measurements showing lower performance compared to the design values, which is presumably explained by fabrication inaccuracies. Additionally, the manuscript demonstrates an example of the practical use of the manufactured filters as the eight-band sequential imaging of plant leaves and a color chart with subsequent classification.

The manuscript has major issues that should be addressed before it can be considered for publication.

First, the claimed novelty of the proposed dielectric-metal mirror and filter structure should be justified. The dielectric coatings complementing the metal thin films are widely used for anti-reflective purposes providing increased or decreased reflectance in a certain spectral range. A particular example of their application in the Fabry-Perot resonator can be found in [1] (see Additional references at the end of the review). The optimization of the thicknesses in such dielectric-metal reflectors can be done using different methods, e.g., limited-memory Broyden–Fletcher–Goldfarb–Shanno with bound constraints (L-BFGS-B) algorithm, see its application for a quad-layer structure with TiO2-Ag layers in [2] or a penta-layer structure with TiO2, SiO2 and Ag in [3]. Moreover, similar structures with TiO2, SiO2 and Ag layers were presented in [4]. The optimization for selectively suppressing higher-order modes in tri-layered structures was demonstrated in [5]. Therefore, it is highly recommended to clearly formulate the optimization task, the optimization technique and the novelty for the designed TiO2-Ag-SiO2-Ag-TiO2 resonator in the manuscript. The same can be said about the proposed filter structure with a dielectric thin film inside the cavity: a similar structure with a thin metal film in the middle of the SiO2 cavity was utilized in [6] to retain the second-order resonance peak while the first-order one was almost completely suppressed. The authors should provide differences between their manuscripts and the other works, especially [6], and describe the benefits of their approach. Otherwise, the manuscript will most likely be rejected due to lack of novelty.

Next, the manuscript contains many unnecessary repetitions and rather general phrases, but it lacks important details necessary to establish the reproducibility of the results and a clear understanding of the physical effects underlying the design. The list of these issues along with some typos and errors is given below.

1. Abstract, line 26. Probably, “does not degrade”.

2. Page 3, lines 94-103 are an example of the aforementioned unnecessary repetition.

3. Page 3, Eq. 1 and 2. The notation ‘m’ – the order of resonance is absent. It is not clear whether Eq. 2 is for peak transmission or its spectral dependence. The phase difference ‘delta’ is out of context and can be misleading. Moreover, Ref. 25 does not contain these equations. It is highly recommended to use a reference to a fundamental book like Born and Wolf or similar works like [7] instead and add proper notation and explanations.

4. “Peter H. Berning and colleagues introduced the concept of induced transmission [29]” (page 3, line 128). “Jingyuan Zhu and others also have applied a similar principle to reduce the FWHM of Fabry-Pérot resonators [30]” (page 3, line 130-131). These references do not correspond to the list of references at the end of the manuscript. Some references are also mixed up. Ref. 28 looks strange. Check the list carefully.

5. Page 4, Figure 1. It should be stated whether the results were obtained using Eqs. 3-5 or the FDTD simulation. Additionally, the difference between the simplified model (eqs. 3-5) and the FDTD simulation should be clarified from the point of view of considered physical effects and assumptions. If Eqs. 3-5 were not used, they should not be placed in the manuscript. Which software was used for the FDTD simulations?

6. Figure 1 and 2. Providing real values instead of ‘min’ and ‘max’ will be beneficial for the readers.

7. “Figures 2c and 2d show the reflectance and transmittance“ (Page 5, line 172), but “(c) transmittance and (d) absorbance between” (caption of Figure 2).

8. Page 6, Figure 3. It is recommended to add a plot of peak transmission and FWHM vs wavelength for both resonators for clear comparison. Similarly, a plot of peak wavelength and FWHM vs cavity thickness can be added. Also, the description “These secondary peaks, indicated by dashed lines in Figure 3(e),” (page 7, line 208) is quite misleading. Do dashed lines indicate plots with double peaks or the secondary peaks only? Which orders of resonance were used?

9. “was validated through simulations” (Page 6, line 188). Which simulations and which software were used?

10. “Previously, we proposed a design” (page 7, lines 215-218). A reference should be given.

11. Page 7, Eq. 6. The equation is out of context since the phase ‘phi’ was not used elsewhere. A complete picture including also Eqs. 1,2 should look like Eqs.1-3 from [7]. Similar to Eqs. 3-5, it should be stated whether the results in Figure 4 were obtained using Eq. 6 or some other simulation.

12. Figure 4 demonstrate the results for s- and p-polarized light separately, but the effect of polarization is not discussed in the other parts of the manuscript. Is the transmission of the designed filter polarization-dependent?

13. “single-medium resonator with the same optical thickness” (page 9, line 265). The optical thickness depends on wavelength. For what wavelength is the equality valid?

14. “The transmission spectra for 16 (4×4) different central wavelengths are shown in Figure 6(b). The results confirm that” (page 9, line 288). The results of simulation? Which techniques were used to optimize the design?

15. Section 3. Specify, which equipment (SEM, spectrometer, CMOS sensor, etc.) was used and provide their key specifications (e.g., spectral resolution for the spectrometer).

 

Additional references

1. Robert Jan Van Wijk, Fabry-perot with coated mirrors, Patent EP0571022A1 (1993).

2. Incheol Jung et al., Tailoring resonant modes in dual cavities for transmissive structural colors with high brightness and high purity, Opt. Express 32, 26136-26146 (2024).

3. Incheol Jung et al., Flexible transmissive colors with enhanced purity and brightness through overlapping multi-cavity resonances, Opt. Lett. 49, 7214-7217 (2024).

4. Guan Wang et al., Fabry–Perot multilayer for spectral imaging applications, J. Opt. Soc. Am. B 41, D75-D80 (2024)

5. Dohyun Kim et al., Optical interference control for selectively suppressing higher-order modes in all-dielectric tri-layered structures for reflective RGB colors, Opt. Lett. 50, 1779-1782 (2025).

6. Youfen Yuan et al., Narrow-band filter of Fabry-Pérot cavity over a wide spectrum range based on second-order resonant peak, Opt. Express 33, 28081-28096 (2025).

7. Roger Arsenault and Germain Boivin, Fabrication of metal-dielectric interference filters: a simple method, Appl. Opt. 16, 1890-1892 (1977).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

I have read your manuscript with great care and must admit that a text this clear and well-prepared is a rare occurrence. The article has been crafted with meticulous attention to detail; both its editorial form and its substance leave little to be desired. I am particularly impressed with the manner of presentation, the literature citation, and the coherence of the main narrative.

The manuscript demonstrates a high degree of innovation, which the authors clearly emphasize by referencing the existing state of knowledge. The main research achievement described in the article—the construction of a Fabry-Pérot resonator that achieves an excellent transmittance exceeding 50% and a full width at half maximum (FWHM) of less than 25 nm within the same spectral range—is well-highlighted.

Both the method of citing literature and the presentation of mathematical formulas are entirely correct and meet the standards of reputable scientific publications. I am pleased that the cited references come from various international sources, which proves that the work is solidly grounded in the realities of contemporary science.

As a specialist from a slightly different field, I can confirm the high methodological consistency of the reviewed article. The data visualization section has also been executed correctly, which, given the general nonchalance in the cartographic representation of research results, deserves full praise.

My minor editorial comments are as follows:

• a lack of spacing between the axis titles and the units in Figures 2-9, which should be corrected,

• an unnecessary red underline for the word "thickness" in Figure 5.

Beyond these points, I have no other significant concerns. I believe that after one more mandatory review of the entire text and the correction of these minor oversights, the manuscript can proceed toward publication. I see no need for a second review of the article. I congratulate the authors and wish them continued success in their research!

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Issue 1: The manuscript states that “surface roughness of the thin film causes lower transmittance than the theoretical value and broader FWHM,” but only qualitatively demonstrates roughness in Fig. 8(c). No quantitative data (e.g., surface roughness parameters Ra/RMS measured by AFM) are provided, nor is the quantitative impact of roughness on optical performance explained (e.g., how much transmittance decreases per 1 nm increase in Ra).
Suggestion 1: Provide AFM measurements of the film surface roughness and establish a correlation curve linking “roughness – transmittance – FWHM.” Additional comparative experiments could be added (e.g., reducing roughness via annealing and testing performance changes) to further validate the mechanism of roughness influence.

Issue 2: The multi-channel filter is designed with 16 channels (4×4), but only 8 samples were fabricated and tested, leaving half of the channels unverified. Moreover, no data are provided on long-term stability (e.g., performance under high temperature and humidity), which limits the argument for industrial applicability.

Suggestion 2: Include fabrication and testing data for all 16 channels to demonstrate generalizability of the multi-channel design. Add environmental stability tests (e.g., 85 °C / 85 % RH aging tests) and report the changes in transmittance and FWHM to strengthen the practical applicability argument.

Issue 3: The conclusion section only highlights advantages such as wide bandwidth and high transmittance but does not mention study limitations (e.g., the impact of surface roughness under the current process, limited number of fabricated samples, or unexplained transmittance fluctuations in the near-infrared long-wavelength range of 1000–1100 nm).

Suggestion 3: Add a discussion of the limitations in the conclusion or “Discussion” section.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have made impressive work and have significantly improved the manuscript. However, some minor improvements can still be recommended.

First of all, the authors provided a good description of novelty in the response to reviewer but added only two sentences to the text of the manuscript. I suggest further adding some parts of the response to the main text to better inform readers about the proposals of the manuscript compared to the previous works. For example, “In our design, a thin film inside the cavity is used to increase the separation between resonant frequencies rather than to eliminate odd-order resonances as in [12].” can replace the sentence in page 2, line 87. Similarly, main differences can be added to the description in Section 2.1 and 2.2.

The authors further mentioned in their response to reviewer that they used Lumerical software. Please, add this to the main text and specify the version of the software.

What is ‘d’ in Eq. 2?

What does a dashed white line in Figs. 3e,f indicate? Please, add this information to the caption of the figure. Also, it is recommended to indicate the order of resonance near the color lines in these figures.

Similarly to the caption of Figure 9, please, specify the spectrometer used for Figure 7b. E.g., in page 10, line 312.

The authors state that the proposed design overcomes the tradeoff between the working range and FWHM of the filters. At the same time, FWHM of the designed filters can only be roughly estimated by the readers from the plots like Fig. 7b. Therefore, I recommend indicate both design and experimental values of FWHM for several filters in the text.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf