Temperature-Controlled CO₂ Laser Polishing of Fused Silica Microlens Arrays

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presents a temperature-controlled CO₂ laser polishing approach for improving the surface quality of fused silica microlens arrays (MLAs) fabricated via direct laser writing. By integrating an infrared temperature sensing system with a PID-based feedback loop, the authors aim to maintain the polishing process within a controlled molten regime. A coupled thermal–fluid finite element model is developed and compared with experimental data. The results show a significant reduction in surface roughness (from 173 nm to 25 nm) and improved optical transmittance exceeding 90% in the visible range, while largely preserving the microlens geometry. Despite its merits, the manuscript requires substantial improvement before it can meet the standards of publication.

1. While the manuscript claims novelty in “temperature-controlled polishing”, closed-loop thermal control in laser polishing has already been explored in multiple recent works. For instance, studies such as Ref. [5] and Ref. [20] already address temperature window control and process stability. The manuscript does not clearly articulate what is fundamentally new beyond incremental system integration. The authors must explicitly differentiate their contribution (e.g., MLA-specific control challenges, spatial resolution of control, or advantages in dynamic response). This should be clarified explicitly.
2. The manuscript lacks direct benchmarking against alternative polishing approaches (e.g., magnetorheological polishing, ion beam figuring, or uncontrolled CO₂ laser polishing). Without comparative data, it is difficult to assess whether the reported 25 nm roughness is competitive. Current state-of-the-art for fused silica polishing can achieve sub-10 nm Sa; thus, the reported results may not represent a significant advancement.
3. Several simplifying assumptions in Section 2.3 are not adequately justified. For instance, (1) Treating a pulsed laser as a continuous heat source based on thermal diffusion time is oversimplified and may neglect transient peak effects. (2) Ignoring the temperature dependence of properties other than viscosity (e.g., thermal conductivity, surface tension) may significantly affect melt flow predictions. (3) The 2D model cannot capture the full 3D curvature effects of microlenses. These assumptions limit the model's predictive validity, yet the authors claim strong agreement with experiments based solely on temperature validation.
4. The manuscript would benefit from a stronger connection to recent advances in optical field engineering and surface–light interaction enabled by high-quality micro/nano-structured optics. In particular, the authors are encouraged to cite and discuss Chen et al. (doi: https://doi.org/10.1515/nanoph-2021-0787) on Broadband generation of accelerating polygon beams with large curvature ratio and small focused spot using all-dielectric metasurfaces, Bai et al. (doi: https://doi.org/10.1364/OE.398115) on Shift of the surface plasmon polariton interference pattern in symmetrical arc slit structures and its application to Rayleigh metallic particle trapping, and Jing et al. (doi: https://doi.org/10.3390/nano13030508) on Wavelength-Independent Excitation Bessel Beams for High-Resolution and Deep Focus Imaging, as these works highlight the critical role of surface quality and profile fidelity in determining optical performance and would better position the present study within the broader photonics context.
5. Validation is limited to temperature and power evolution at a single point. There is no validation of the Melt pool geometry, Surface profile evolution, and Flow dynamics. The reported temperature error (0.11%) is unusually low and suggests either overfitting or insufficient measurement uncertainty analysis. A rigorous validation must include spatial and morphological comparisons.
6. The manuscript presents single-value roughness (Sa = 173 nm → 25 nm) and transmittance results without standard deviation, number of samples, and repeatability analysis; this is insufficient. Variability across the MLA array and between samples must be reported.
7. The claim of shape preservation is weakly supported. Although a small reduction in shape error is reported (from 5.88 μm to 5.18 μm), the manuscript does not provide a clear definition of the error metric, nor does it include spatial error distributions or comprehensive profile comparisons. Given that laser-induced melting can easily distort microstructures, stronger evidence is required.
8. The optical characterization is limited to bulk transmittance measurements. For microlens arrays, functional optical performance, such as focusing quality, spot size, or imaging capability, is critical. The absence of such characterization significantly limits the impact of the work.
9. The PID control strategy is insufficiently described. The selection and tuning of the control parameters are not justified, and no analysis of system stability or sensitivity is provided. This weakens the reproducibility and technical depth of the control aspect.
10. The manuscript claims to provide a “mechanistic analysis model”, but the analysis remains largely qualitative. The relationship between temperature gradients, Marangoni flow, and final surface smoothing is not quantified.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This work proposes PID temperature‑controlled CO₂ laser polishing to smooth stepped textures on DLW‑fabricated fused silica microlens arrays (MLAs). Roughness is reduced by 85.5% and transmittance exceeds 90%. The closed‑loop control idea is not without merit, and the simulation‑experiment temperature agreement is acceptable. However, the manuscript suffers from fatal flaws: extremely low processing efficiency, negligible form error correction, an unacceptably large residual form error (~5.2 µm), an oversimplified 2D model, lack of statistical rigor, and no meaningful comparison with alternative methods. These issues severely undermine the practical and scientific value. Major revision is recommended.

1. Processing efficiency is critically low. Scanning speed is 0.2 mm/s with 25 µm spacing. Polishing a single 1 mm² microlens takes ~3.3 minutes; a 10×10 array (1 cm²) would require ~5.5 hours. The authors do not discuss this limitation or propose any acceleration strategy, making the method impractical even for laboratory use.
2. The method barely corrects form error. Shape error improves from ≤5.88 µm to ≤5.18 µm, a reduction of only 0.7 µm (~12%). This shows that polishing only removes superficial steps and cannot rectify the macroscopic profile inherited from DLW. For any precision optics (imaging, beam shaping, single‑mode fiber coupling), this lack of figure correction is unacceptable.
3. The residual form error of 5.18 µm is far too large. With a lens sagitta of ~42 µm, this error represents 12% of the sagitta, translating to tens of wavelengths of wavefront aberration in the visible range. Such errors would severely degrade optical performance. The authors provide no error budget or intended application to justify this value.
4. The 2D FE model is a serious oversimplification. It cannot account for 3D heat diffusion, varying laser‑surface incidence angle, thermal crosstalk between lenslets, or edge effects. The pulsed laser is simplified to a continuous source without justification. No sensitivity analysis or experimental validation on curved surfaces is provided, undermining simulation credibility.
5. Experimental validation lacks statistical rigor. Roughness and form error are reported as single values or ranges without standard deviations, confidence intervals, or repetition counts. Measurement positions (apex, edge, or average) are not specified. This lack of reproducibility data makes the method’s reliability impossible to assess.
6. No meaningful comparison with alternative methods. Established techniques (ion beam figuring, plasma polishing, optimized CO₂ laser scanning) can achieve sub‑nm roughness while preserving form accuracy. The authors provide no quantitative comparison, so the claimed advantages of their temperature‑controlled approach remain unclear.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript presents a temperature‑controlled CO₂ laser polishing method for fused silica microlens arrays prepared by layer‑by‑layer direct laser writing. The research topic is practically significant for the manufacturing of high‑performance micro‑optical components. The work includes a reasonable experimental design, a closed‑loop temperature control strategy, finite element simulation, and sufficient performance characterization. The results are reliable and show that surface roughness is significantly reduced and optical transmittance is effectively improved. Overall, the manuscript is well‑organized and scientifically sound. Only minor revisions are required before publication:

1、Supplement the average value and standard deviation of multi‑point surface roughness measurements to improve the statistical analysis and reliability of the results.

2、Describe in detail the alignment and calibration procedures between the infrared thermometer and the laser processing spot to ensure the reproducibility of temperature control accuracy.

3、Add an analysis of the rationality and error associated with simplifying the pulsed laser as a continuous heat source in the simulation.

4、Although the literature review is relatively comprehensive and covers related studies from the past five years, it lacks a dedicated overview of real‑time temperature‑controlled laser polishing for microstructures of hard and brittle materials.

5、Add comparisons with results from similar studies on CO₂ laser polishing of fused silica in the discussion section to highlight the advantages of the present work.

6、Supplement characterization data of the heat‑affected zone (HAZ) to improve the evaluation system of processing quality.

7、Deepen the quantitative mechanistic discussion on the correlation between the Marangoni effect, surface tension‑driven flow, and the temperature field.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Accept

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript has been revised accordingly with appropriate clarifications, added discussions, and updated data. The responses are generally reasonable and the revisions have improved the clarity and credibility of the paper.I recommend accepting the manuscript.