Reduction in Dark Current in Photodiodes: A Review

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript reviews dark current mechanisms and reduction strategies in photodiodes, covering diffusion, SRH generation, trap-assisted tunneling, band-to-band tunneling, surface leakage, thermal management, passivation, guard-ring engineering, gettering, and architecture optimization. The topic is relevant and potentially suitable for Micromachines. The manuscript is generally readable and broad in scope. However, in its present form, I do not think it is yet ready for publication. My main concerns relate to the completeness of the recent literature coverage, the rigor of the comparative framework, and several reference-level and presentation-level inconsistencies. I therefore recommend major revision.

1. The review would benefit from a more rigorous comparative framework. At present, the manuscript is mostly organized as a sequential narrative by mechanism and technique, but it does not always help the reader understand which reduction strategy is most effective under which material system, bias regime, temperature range, and device architecture. I strongly encourage the authors to add at least one summary table or figure that compares, across different photodiode platforms, the dominant dark current source, the primary mitigation strategy, the reported reduction magnitude, and the practical trade-offs such as temperature requirement, process complexity, compatibility with CMOS, and scalability to arrays.

2. Table 1 is useful in intent, but in its current form it mixes heterogeneous metrics and outcomes, such as dark current reduction ratio, breakdown voltage increase, quantum efficiency, photon detection efficiency, and even qualitative statements, within the same comparison framework. This makes cross-platform comparison difficult and, in some places, potentially misleading. The table should be restructured so that the meaning of each metric is standardized or clearly separated by category. Otherwise, the table risks functioning more as a literature inventory than as an analytical synthesis.

3. The manuscript needs stronger critical analysis rather than only listing representative studies. For example, in the sections on passivation, guard rings, and gettering, the authors summarize many studies, but the discussion of limitations remains relatively brief. A high-quality review should not only state what worked, but also clarify when a given strategy fails, what the scaling bottlenecks are, and whether the reported improvements are limited by material quality, surface chemistry, electric-field crowding, process budget, or measurement conditions. This is especially important for emerging materials and cryogenic detectors.

4. The material-specific discussion is uneven. Silicon CMOS and black silicon are discussed in relatively good detail, but the treatment of GeSn, and narrow-bandgap infrared platforms is still too light considering the topic of dark current reduction. Since the manuscript explicitly aims to cover silicon, germanium, III-V, and emerging materials, the authors should rebalance the discussion and clarify the criteria used to select representative literature.

5. The manuscript claims to be a comprehensive review, but the coverage of recent progress is still incomplete in several important areas, especially for very-long-wavelength infrared blocked-impurity-band photodetectors and focal plane arrays. These devices are highly relevant to the theme of dark current suppression because interface control, low-temperature operation, and system-level detectivity are central to their performance. The authors should discuss and cite the following recent works: DOI: 10.1109/LED.2025.3558427 and DOI: 10.1063/5.0278909. The first paper is directly relevant to interface-controlled performance in VLWIR blocked impurity band photodetectors, and the second reports a broadband phosphorus-doped silicon blocked impurity band VLWIR focal plane array. Both are highly pertinent to the review’s discussion of dark current control under cryogenic operation and should be incorporated into the sections on material-specific considerations, device architecture optimization, and future directions.

6. Some statements require clarification or correction of physical meaning. For example, the sentence stating that P-Well and P+ implant geometry optimization around the STI edge can “reduce dark current by over 40 mV/s in terms of voltage ramp rate” is confusing because mV/s is not a direct dark-current unit. The authors should carefully check whether this refers to a dark signal slope, dark current equivalent, or another proxy metric, and revise the wording accordingly.

7. The reference list requires careful proofreading. I noticed duplicated or apparently duplicated references, for example Refs. 104 and 105 appear to refer to the same GeSn paper with the same DOI, and Refs. 97 and 98 also appear to duplicate the same ion implantation gettering paper. These issues should be corrected throughout the manuscript, and the full bibliography should be checked for consistency in author names, titles, page ranges, and DOI formatting.

Author Response

Reviewer_1_Comment_1: The review would benefit from a more rigorous comparative framework. At present, the manuscript is mostly organized as a sequential narrative by mechanism and technique, but it does not always help the reader understand which reduction strategy is most effective under which material system, bias regime, temperature range, and device architecture. I strongly encourage the authors to add at least one summary table or figure that compares, across different photodiode platforms, the dominant dark current source, the primary mitigation strategy, the reported reduction magnitude, and the practical trade-offs such as temperature requirement, process complexity, compatibility with CMOS, and scalability to arrays.

Response to Reviewer_1_Comment_1: We thank the reviewer for this constructive suggestion. We agree that a cross-platform comparative framework adds significant value to the review by helping readers navigate the landscape of dark current reduction strategies across different material systems and device architectures. We have added a new section (Section 11, "Cross-Platform Comparative Analysis") containing a comprehensive landscape tables that compares nine major photodiode platforms—Si PIN/PPD, black silicon, Si SPAD, Ge-on-Si, InGaAs/InP APD, InGaAs/InP SPAD, GeSn, Si:P BIB, and 2D van der Waals heterostructures—across the following dimensions: spectral range, dominant dark current source, primary mitigation strategy, best reported reduction magnitude, operating temperature, CMOS compatibility, process complexity, array scalability, and key practical trade-off. The table is accompanied by an analytical discussion highlighting three cross-cutting observations: (1) the systematic shift in dominant dark current mechanism from interface SRH generation in wide-bandgap silicon to tunneling in narrow-gap materials and hopping conduction in VLWIR BIB detectors; (2) the strong platform-specificity of optimal mitigation strategies; and (3) the role of practical deployment constraints in governing the viable strategy space. This new section is positioned between the technique-level literature classification and the Summary & Conclusions, providing the reader with a platform-level synthesis before the concluding discussion.
—-
Reviewer_1_Comment_2: Table 1 is useful in intent, but in its current form it mixes heterogeneous metrics and outcomes, such as dark current reduction ratio, breakdown voltage increase, quantum efficiency, photon detection efficiency, and even qualitative statements, within the same comparison framework. This makes cross-platform comparison difficult and, in some places, potentially misleading. The table should be restructured so that the meaning of each metric is standardized or clearly separated by category. Otherwise, the table risks functioning more as a literature inventory than as an analytical synthesis.

Response to Reviewer_1_Comment_2: We thank the reviewer for this important observation. We agree that the original Table 1 mixed heterogeneous metrics—dark current reduction ratios, absolute current values, breakdown voltage gains, quantum efficiency, photon detection probability, and qualitative statements—within a single undifferentiated "Reduction" column, making cross-platform comparison difficult and potentially misleading. We have restructured Table 1 into 4 tables with adding metrics (1) "Outcome Metric," which uses a standardized taxonomy to label the type of quantity reported (e.g.,  `Id red.` for dark current reduction factor, 'Id abs.' for absolute dark current, DCR for dark count rate, 'Vbr' for breakdown voltage, QE for quantum efficiency, PDP/PDE for photon detection probability/efficiency, DSR for dark signal rate, 'Diag.' for diagnostic studies, and 'Qual.' for qualitative improvements), and (2) "Reported Value," which gives the actual numerical result. This taxonomy is fully defined in the table caption. All 82 entries across 12 technique categories have been individually reclassified under this framework. This restructuring, combined with the new cross-platform comparative Tables, ensures that the reader can distinguish between fundamentally different figures of merit and make meaningful comparisons both within and across device platforms.
—-
Reviewer_1_Comment_3: The manuscript needs stronger critical analysis rather than only listing representative studies. For example, in the sections on passivation, guard rings, and gettering, the authors summarize many studies, but the discussion of limitations remains relatively brief. A high-quality review should not only state what worked, but also clarify when a given strategy fails, what the scaling bottlenecks are, and whether the reported improvements are limited by material quality, surface chemistry, electric-field crowding, process budget, or measurement conditions. This is especially important for emerging materials and cryogenic detectors.

 

Response to Reviewer_1_Comment_3: We thank the reviewer for this important feedback. We agree that a high-quality review should not only catalogue what has been achieved but also critically assess the conditions under which strategies fail, the scaling bottlenecks, and the gaps between laboratory results and practical deployment.
We have added five new critical-analysis subsections and paragraphs throughout the manuscript:
(1) Section 4.5 ("Limitations and Failure Modes of Passivation Approaches") discusses thermal budget incompatibilities, parasitic inversion risks from Al₂O₃ on n-type surfaces, long-term UV stability under space radiation, the fundamental instability of GeO₂ above 450 °C, the absence of a viable native oxide for GeSn, and the general gap between optimized test-structure results and production-line reliability.
(2) Section 5.4 ("Scaling Bottlenecks and Design Trade-offs in Guard Rings") addresses the fill factor vs. guard ring area trade-off in high-density FPAs, the sensitivity of optima to process variations, the limited transferability of discrete-device results to array-scale fabrication, Zn diffusion control challenges in InGaAs/InP, and the vulnerability of floating guard rings to surface charge drift.
(3) Section 6.4 ("Limitations of Gettering and Defect Engineering") examines the ineffectiveness of gettering against slow-diffusing contaminants, the distinction between bulk impurity removal and interface/structural defect mitigation, the placement dilemma in thinned BSI wafers, residual hot-pixel tails, and the material-quality floor imposed by residual threading dislocations in Ge-on-Si.
(4) A critical paragraph in Section 10 (Emerging Materials) addresses the intrinsic tension between Sn fraction and defect density in GeSn, the non-scalability of exfoliated-flake 2D device results, and the undemonstrated CMOS integration path for 2D heterostructures.
(5) A limitations paragraph in Section 7 (BIB Photodetectors) discusses the mandatory deep cryogenic cooling, the sensitivity of dark current to interface abruptness during large-area growth, the gap between the demonstrated 16×16 format and megapixel scales, and uncharacterized doping uniformity for astronomical applications.
These additions ensure that each major technique category now includes an explicit discussion of when the strategy fails, what limits further improvement, and what gaps remain between reported results and practical scalability.
—-
Reviewer_1_Comment_4: The material-specific discussion is uneven. Silicon CMOS and black silicon are discussed in relatively good detail, but the treatment of GeSn, and narrow-bandgap infrared platforms is still too light considering the topic of dark current reduction. Since the manuscript explicitly aims to cover silicon, germanium, III-V, and emerging materials, the authors should rebalance the discussion and clarify the criteria used to select representative literature.

Response to Reviewer_1_Comment_4: We thank the reviewer for identifying this imbalance. We have substantially expanded the treatment of narrow-bandgap and emerging material platforms throughout the revised manuscript.
(1) GeSn alloy photodiodes now receive a dedicated subsection (10.4) with significantly expanded discussion including: the composition-dependent crossover from SRH-dominated to diffusion-dominated dark current; quantitative threading dislocation density scaling with Sn fraction; the severe thermal budget constraint imposed by Sn segregation above 350–400 °C; the Si overgrowth passivation mechanism and its limitations at high Sn fractions; and a quantitative comparison of GeSn MWIR dark current densities against competing technologies.
(2) A new subsection on narrow-bandgap infrared platforms (10.5) has been added covering HgCdTe and type-II superlattice detectors, providing essential benchmarking context for emerging alternatives. This includes the Auger-limited dark current floor in HgCdTe, the SRH-limited performance of T2SL devices, and photon-trapping thin-absorber results.
(3) The InGaAs/InP section (10.3) has been significantly expanded to address the fundamental narrow-bandgap dark current physics: the delicate SRH-vs-tunneling balance, the critical multiplication layer thickness window, the distinct surface chemistry challenges of InGaAs compared to silicon, planar vs. mesa geometry trade-offs, and the strong temperature sensitivity that mandates TEC cooling.
(4) 2D van der Waals heterostructures have been separated into their own subsection (10.6) for clearer organization.
(5) The Introduction now explicitly states the criteria used to select representative literature, including: direct relevance to dark current mechanisms, quantitative reporting of dark current or related metrics, balanced coverage of mature and emerging platforms, emphasis on recent (2020–2025) work alongside foundational references, and preference for the most complete characterizations where studies overlap.
—-
Reviewer_1_Comment_5: The manuscript claims to be a comprehensive review, but the coverage of recent progress is still incomplete in several important areas, especially for very-long-wavelength infrared blocked-impurity-band photodetectors and focal plane arrays. These devices are highly relevant to the theme of dark current suppression because interface control, low-temperature operation, and system-level detectivity are central to their performance. The authors should discuss and cite the following recent works: DOI: 10.1109/LED.2025.3558427 and DOI: 10.1063/5.0278909. The first paper is directly relevant to interface-controlled performance in VLWIR blocked impurity band photodetectors, and the second reports a broadband phosphorus-doped silicon blocked impurity band VLWIR focal plane array. Both are highly pertinent to the review’s discussion of dark current control under cryogenic operation and should be incorporated into the sections on material-specific considerations, device architecture optimization, and future directions.

Response to Reviewer_1_Comment_5: We thank the reviewer for highlighting the important topic of VLWIR blocked impurity band (BIB) photodetectors and focal plane arrays. We agree that these devices are highly relevant to the review's central theme of dark current suppression, particularly given the critical role of interface control and cryogenic operation in their performance.
We have made the following additions to the revised manuscript:
(1) A new subsection entitled "Blocked Impurity Band (BIB) Photodetectors for VLWIR" has been added within Section 5 (Device Design and Architecture Optimization). This subsection introduces the BIB operating principle, discusses the recent interface model for Si:P BIB detectors showing that a sharp epitaxial interface suppresses dark current by five orders of magnitude (maintaining <1 pA up to 2.7 V forward bias with D* > 10¹² cm Hz¹/² W⁻¹ at 28.3 µm) [DOI: 10.1109/LED.2025.3558427], and reports the 16×16 Si:P BIB FPA results demonstrating broadband VLWIR detection to 32.2 µm with sub-40 fA dark current at 4 K [DOI: 10.1063/5.0278909].
(2) The classification tables have been extended with a new "BIB Photodetectors" category containing both cited works.
(3) The Summary and Conclusions section has been updated to discuss how the interface-controlled dark current principles examined throughout the review extend naturally to impurity-band architectures under deep cryogenic conditions, and to identify scaling of BIB FPAs as an open challenge for next-generation VLWIR imagers.
Both recommended references have been cited and incorporated into the bibliography.
—-
Reviewer_1_Comment_6: Some statements require clarification or correction of physical meaning. For example, the sentence stating that P-Well and P+ implant geometry optimization around the STI edge can “reduce dark current by over 40 mV/s in terms of voltage ramp rate” is confusing because mV/s is not a direct dark-current unit. The authors should carefully check whether this refers to a dark signal slope, dark current equivalent, or another proxy metric, and revise the wording accordingly.

Response to Reviewer_1_Comment_6: We thank the reviewer for catching this ambiguity. The reviewer is correct that mV/s is not a direct dark current unit. In CMOS image sensor characterization, the dark signal rate (expressed in mV/s) refers to the voltage accumulated on the pixel sense node per unit integration time and is a widely used proxy metric for dark current at the pixel level. The original wording failed to make this distinction clear. We have revised the sentence in Section 8.1 to explicitly define the dark signal rate and clarify that it is a standard CIS characterization metric, not a direct current measurement. The corresponding entry in the classification table (Table 1) has also been updated for consistency.
—-
Reviewer_1_Comment_7: The reference list requires careful proofreading. I noticed duplicated or apparently duplicated references, for example Refs. 104 and 105 appear to refer to the same GeSn paper with the same DOI, and Refs. 97 and 98 also appear to duplicate the same ion implantation gettering paper. These issues should be corrected throughout the manuscript, and the full bibliography should be checked for consistency in author names, titles, page ranges, and DOI formatting.

Response to Reviewer_1_Comment_7: We thank the reviewer for this important observation. We have conducted a thorough audit of the entire bibliography and identified 10 pairs of duplicate references where different citation keys pointed to the same publication (identical DOI). These include the GeSn passivation paper (formerly Refs. 104/105), the ion implantation gettering paper (formerly Refs. 97/98), and eight additional duplicated pairs. All duplicates have been consolidated to a single entry, and every citation throughout the manuscript has been updated accordingly. In addition, we corrected several entry-type misclassifications (conference proceedings that were erroneously tagged as journal articles, and books tagged as articles), and verified that all 138 cited keys resolve correctly to entries in the bibliography. The full reference list has been proofread for consistency in author names, titles, page ranges, and DOI formatting.

With Best Regards,

Alper Ulku

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript presents a review of dark current mechanisms in photodiodes and summarizes techniques used to reduce dark current across a wide range of material systems and device architectures. This comprehensive literature review examines 145 publications covering the physical mechanisms underlying dark current generation and diverse techniques employed for its reduction.

The topic is relevant for the photodetector and imaging sensor communities, with a large number of references retrieved from the literature.

The paper is generally well structured, and the coverage of mitigation strategies such as surface passivation, guard rings, device architecture optimization, and temperature control is useful. In particular, the effort to summarize results from many publications in Table 1 is valuable.

However, in its current form the manuscript is largely descriptive rather than analytical, and several aspects require improvement before the paper is suitable for publication as a review article. In particular, the review manuscript would benefit from stronger critical analysis of the literature, clearer quantitative comparisons, and improved presentation of the summarized data.

1)Limited critical analysis of literature. Although the manuscript cites a large number of studies, most sections primarily summarize results without critically evaluating them. I invite the authors to:

a)Compare the effectiveness of different dark-current reduction techniques.

b)Identify the advantages and limitations of each approach.

c) Discuss trade-offs between dark current reduction and other device parameters e.g., responsivity, noise, bandwidth.

2) Lack of quantitative aspects. The manuscript reports many dark current values from different publications, but these results are not presented in a consistent or easily comparable format. For instance, important factors such as operating temperature, bias conditions or device areas vary significantly between studies and are not reported.

3) Table 1 needs restructuring. It reports results from numerous publications but it is very large and difficult to interpret, mixing different device types, materials, and experimental conditions.

To improve clarity, the authors should consider for instance splitting the table into multiple smaller tables. This would significantly improve the usefulness of the table for readers.

4)Lack of figures or conceptual diagrams. The manuscript contains almost no figures explaining key physical concepts. I would suggest including schematic diagrams illustrating dark current mechanisms would greatly improve readability.

5)Some parts of the manuscript need a more formal scientific tone. Minor grammatical issues appear throughout the manuscript and should be corrected during revision.

Comments on the Quality of English Language

Moderate revision.

Author Response

Reviewer_2_Comment_1: Limited critical analysis of literature. Although the manuscript cites a large number of studies, most sections primarily summarize results without critically evaluating them. I invite the authors to:
a) Compare the effectiveness of different dark-current reduction techniques.
b) Identify the advantages and limitations of each approach.
c) Discuss trade-offs between dark current reduction and other device parameters e.g., responsivity, noise, bandwidth.

Response to Reviewer_2_Comment_1: We thank the reviewer for this constructive observation. We agree that a review of this scope should move beyond summarizing individual results and provide a critical comparative perspective. We have made several substantive additions to address all three sub-points:
(a) Comparing effectiveness of different techniques: A new cross-platform comparative table (Table 2, Section 11) has been added that maps each major photodiode material system to its dominant dark current mechanism, the most effective mitigation strategy, the reported reduction magnitude, operating temperature, CMOS compatibility, process complexity, array scalability, and key trade-offs. An accompanying analytical paragraph (Section 11.1) synthesizes cross-cutting patterns—including the systematic shift of dominant mechanisms with bandgap and the platform-specificity of optimal strategies (e.g., ALD Al₂O₃ is transformative on silicon but plays only a supplementary role on InGaAs).
(b) Advantages and limitations of each approach: Dedicated critical analysis subsections have been added within the technique sections themselves: Section 4.5 (passivation limitations and failure modes), Section 5.4 (guard ring scaling bottlenecks and design trade-offs), Section 6.4 (gettering limitations), and critical paragraphs within the GeSn, HgCdTe/T2SL, 2D materials, and BIB subsections. These explicitly discuss thermal budget constraints, area penalties, material-specific failure modes, and the gap between laboratory results and production-scale performance.
(c) Trade-offs between dark current reduction and responsivity, noise, and bandwidth: A new subsection (Section 11.2, "Trade-offs Between Dark Current Reduction and Device Performance") has been added that systematically examines how dark current mitigation strategies interact with other device parameters. Specific trade-offs discussed include: guard ring area versus fill factor/responsivity; thin absorber designs versus single-pass absorption efficiency; depletion width optimization for dark current versus transit-time-limited bandwidth; multiplication layer thickness in APDs balancing gain, excess noise factor, and tunneling dark current; afterpulsing suppression deadtime versus maximum photon count rate in SPADs; and forming gas anneal reducing interface generation while potentially increasing 1/f noise. The subsection concludes that the optimal strategy is application-dependent, determined by the joint requirements on noise, responsivity, and bandwidth rather than by the maximum achievable dark current suppression alone.
—-

Reviewer_2_Comment_2: Lack of quantitative aspects. The manuscript reports many dark current values from different publications, but these results are not presented in a consistent or easily comparable format. For instance, important factors such as operating temperature, bias conditions or device areas vary significantly between studies and are not reported.

Response to Reviewer_2_Comment_2: We appreciate this important observation. We agree that the heterogeneity of reporting formats across the literature makes direct quantitative comparison difficult, and that our original manuscript did not sufficiently address this issue. We have made the following additions:
A new Table 3 ("Quantitative dark current benchmarks under reported measurement conditions") has been added to Section 11. This table presents representative dark current values from 16 key studies spanning all major platforms reviewed (Si PIN, Si PPD/CIS, black silicon, Ge-on-Si, InGaAs/InP APD and SPAD, Si SPAD, GeSn, HgCdTe, Si:P BIB, and 2D heterostructures). For each entry, we report: (1) the dark current as originally stated in the source publication, (2) the area-normalized dark current density in A/cm² where computable, (3) the operating temperature, (4) the reverse bias voltage, (5) the device active area, and (6) the key experimental condition or technique applied.
An accompanying methodological paragraph explains the normalization approach and explicitly notes the limitations of cross-platform comparison even after area normalization—in particular, that the relative contributions of bulk generation, surface generation, and peripheral leakage scale differently with device geometry, making direct comparison between large-area discrete photodiodes and small-pixel imaging arrays inherently approximate. Where the original publication did not report the device area or other conditions, this is marked in the table to maintain transparency about the available data.
Together with the previously existing classification table (Table 1) and the cross-platform comparison table (Table 2), this new quantitative benchmarking table provides the reader with three complementary levels of analysis: technique-level classification, platform-level strategy comparison, and device-level quantitative benchmarking under stated conditions.
—-

Reviewer_2_Comment_3: Table 1 needs restructuring. It reports results from numerous publications but it is very large and difficult to interpret, mixing different device types, materials, and experimental conditions. To improve clarity, the authors should consider for instance splitting the table into multiple smaller tables. This would significantly improve the usefulness of the table for readers.

Response to Reviewer_2_Comment_3: We agree that the original single classification table was unwieldy and made it difficult to compare results within specific technique categories. We have restructured the table as follows:
The original monolithic Table 1 has been split into four thematic tables, each organized around a coherent group of techniques: (1) thermal management and surface passivation (Tables 1), (2) guard ring structures and gettering/defect engineering (Table 2), (3) avalanche photodiodes and single-photon avalanche diodes (Table 3), and (4) CMOS image sensors, black silicon, device architecture, bias management, and BIB photodetectors (Table 4). Each table retains the standardized six-column format with the "Outcome Metric" taxonomy introduced. Additionally, a new quantitative benchmarking table (Table 7) has been added that presents representative dark current values alongside their measurement conditions (operating temperature, bias voltage, and device area), directly addressing the reviewer's concern about inconsistent reporting formats. Together with the cross-platform comparison table (Table 5) the manuscript now provides six complementary tabular views of the literature that the reader can navigate by theme, platform, or quantitative benchmark.
—-

Reviewer_2_Comment_4: Lack of figures or conceptual diagrams. The manuscript contains almost no figures explaining key physical concepts. I would suggest including schematic diagrams illustrating dark current mechanisms would greatly improve readability.

Response to Reviewer_2_Comment_4: We agree that schematic diagrams are essential for conveying the physical concepts discussed in this review. Three new figures have been added:
Figure 1 presents an energy band diagram of a reverse-biased p-n junction with all five dark current generation mechanisms annotated: (1) diffusion of thermally generated minority carriers in the quasi-neutral regions, (2) SRH generation through midgap trap states in the depletion region, (3) trap-assisted tunneling via intermediate traps under high field, (4) direct band-to-band tunneling at very high fields, and (5) surface/interface generation through dangling-bond states. The figure uses distinct visual encodings (carrier symbols, trap markers, tunneling arrows) to distinguish the mechanisms.
Figure 2 provides a schematic cross-section of a generic photodiode that maps each dark current mechanism to its spatial origin within the device structure, and annotates the corresponding reduction strategy (passivation, guard rings, gettering, doping/bias optimization) with cross-references to the relevant manuscript sections. This figure serves as a visual roadmap connecting the mechanism physics of Section 2 to the reduction techniques discussed in Sections 4–9.
Figure 3 summarizes the approximate dark current reduction magnitude achievable by each major technique category on a logarithmic scale, visually illustrating the wide dynamic range of effectiveness (from ~2× for basic gettering to ~10⁵× for BIB interface engineering) and the general trend that the most effective approaches impose the most demanding system-level constraints.
—

Reviewer_2_Comment_5: Some parts of the manuscript need a more formal scientific tone. Minor grammatical issues appear throughout the manuscript and should be corrected during revision.

Response to Reviewer_2_Comment_5: We thank the reviewer for this observation. We have performed a thorough revision of the manuscript to improve the formal scientific tone and correct grammatical issues. The entire manuscript has been reviewed from Introduction through Conclusions to ensure a consistent formal tone. Specific changes include: Informal and colloquial expressions have been replaced throughout with formal scientific language. Examples include: "it is worth taking stock of" → "it is instructive to examine"; "carries a distinctive fingerprint" → "exhibits a characteristic signature"; "No semiconductor crystal is perfect" → "All semiconductor crystals contain imperfections"; "workhorse detectors" → "predominant detectors"; "finding their footing" → "at an early stage of development"; and numerous similar replacements across all sections. Imprecise qualifiers such as "roughly" have been systematically replaced with "approximately" for consistent formal register. Colloquial phrasing ("tells a clear story," "hard ceiling," "pushes past," "sidesteps") has been replaced with standard scientific terminology ("provides a diagnostic indicator," "fundamental limit," "exceeds," "circumvents"). Minor grammatical corrections have been applied throughout to improve sentence structure and clarity.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have satisfactorily addressed the major comments raised in the previous review. The revised manuscript has been substantially improved in terms of literature coverage, comparative discussion, and reference accuracy. In my opinion, the manuscript is now acceptable for publication in Micromachines.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have adequately addressed all Reviewers remarks and the improved manuscript can be now accepted in the present form.