Suppression of Cohesive Cracking Mode Based on Anisotropic Porosity in Sintered Silver Die Attach Encapsulated by Epoxy Molding Compounds

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presents a well-executed experimental and numerical study that investigates the suppression of cohesive cracking in sintered silver die attach layers through anisotropic porosity control using local press bonding (LP) versus full press bonding (FP). In the reviewer’s opinion, it is technically relevant, well-structured, and contributes meaningfully to the studied area. This said, the following aspects could be further strengthened:

(1) The manuscript contains several awkward sentences. It is recommended for a throughout professional English editing. For example, “By stark, LP showed no degradation…” in Line 17.

(2) It looks like the present discussions mainly focus on qualitative comparison. The authors are encouraged to include more quantitative comparison of porosity levels, crack dimensions, etc.

(3) The limitations and assumptions of using 2D FEA to simulate a 3D phenomenon should be discussed, especially near interfaces.

(4) The conclusion section discussed the directions of future work. It is recommended to more explicitly discuss the limitations in this study and connect them with the future work directions.

Author Response

Thank you so much for giving a review report on our manuscript. Please confirm the response letter to Reviewer 1 as attached PDF file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript proposes a method to suppress cohesive cracking in sintered silver (s-Ag) die-attach structures by introducing anisotropic porosity through two types of press sintering: full press (FP) and local press (LP). The concept is innovative and the combination of experimental observation and finite element modeling provides an interesting perspective. However, several critical issues must be addressed before the manuscript is suitable for publication.

1，The finite element analysis (FEA) relies solely on qualitative SEM observations for validation. There is no quantitative comparison (e.g., strain mapping, crack propagation metrics) to assess the fidelity of the APS distribution predicted. Additionally, the modeling setup lacks information about mesh sensitivity, element selection, and numerical convergence.

2,The manuscript lacks critical details needed for replicating the FEA results. Specifically, the mechanical property models for high-porosity (hp) and low-porosity (lp) s-Ag are not fully defined. No equations, fitting parameters, or porosity-property mapping curves are provided. A clearer methodology is required to enable reproducibility.

3,Although LP improves fatigue resistance, the practical feasibility of localized pressing is not discussed. How this technique would scale to actual manufacturing lines, including challenges in die alignment, uniformity, and process throughput, should be addressed to support its application potential.

4，The suppression of cohesive cracking is attributed to anisotropic APS redistribution via porosity gradients. However, the physical basis for this effect should be better supported. The role of stiffness contrast, local plasticity, and energy dissipation mechanisms across the hp/lp regions can be more deeply explored.

5，While the references cover key foundational studies, the authors are encouraged to cite more recent research on sintered Ag failure suppression strategies, porosity gradient design, and SiC die-attach reliability from journals such as IEEE T-CPMT, Microelectronics Reliability, or EFA. This will contextualize the study within current research efforts.

Author Response

Thank you so much for giving a review report. Please see the attachment as a response letter.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript addresses the suppression of cohesive cracking in sintered silver (s-Ag) die attach layers by introducing anisotropic porosity, with the encapsulant provided by epoxy molding compounds (EMC). While the work is relevant to packaging reliability in power electronics and presents a promising direction, several issues in the methodology, analysis, and explanation reduce its clarity and impact.

• The manuscript refers to “anisotropic porosity” but does not sufficiently explain how this porosity is generated during the sintering process. It is unclear whether this anisotropy arises due to pressure gradients, particle alignment, or temperature variation. The suggestion is to provide micrographs or schematic illustrations that clearly show how the anisotropy in porosity is introduced and controlled during processing.
• Although differences in cracking behavior are reported, the evidence remains largely qualitative. There is a lack of quantitative metrics to convincingly support the claim that anisotropic porosity suppresses cohesive cracking. The suggestion is to quantify the extent of cracking through metrics such as total crack area, average crack length, or fracture energy and include this data in the comparison of samples.
• The finite element modeling section lacks sufficient detail regarding the implementation of porosity. It is not clear whether porosity is explicitly modeled using microstructural representations or whether it is approximated through changes in material constants like the elastic modulus. The suggestion is to clarify how porosity anisotropy is introduced in the model, and to provide specifics on material models, boundary conditions, and validation steps.
• The manuscript asserts that anisotropic porosity reduces lateral stress transmission, thereby mitigating crack propagation. However, this claim is presented without experimental validation. The suggestion is to strengthen this claim using experimental stress field mapping techniques such as digital image correlation (DIC) or Raman stress spectroscopy, or at least to provide supporting simulation visuals showing lateral vs vertical stress transfer differences.
• The statistical robustness of the experimental results is not clear. The manuscript does not specify how many samples were tested under each condition or whether the results shown are representative. The suggestion is to include statistical information such as standard deviation, sample size, or box plots to show the repeatability of the observed crack suppression.
• There is insufficient visual comparison of the crack morphologies in samples with different porosity configurations. The few SEM or CT images provided do not fully demonstrate the differences in crack initiation and propagation. The suggestion is to include side-by-side high-resolution images showing cohesive vs interfacial cracking in isotropic and anisotropic porosity samples.
• The discussion does not address the long-term reliability of the anisotropic porosity approach. It is not clear whether the microstructure remains stable under thermal cycling, or whether pore coalescence or densification occurs. The suggestion is to include discussion or preliminary data on the aging behavior of the pores under accelerated thermal or power cycling tests.
• The abstract and introduction contain broad statements about the importance of reliability without anchoring these claims in specific industry metrics or challenges. The suggestion is to revise these sections to clearly identify the problem being addressed, the novelty of the approach, and the measurable improvement achieved in reliability or crack suppression.
• The manuscript lacks clarity regarding the scalability and reproducibility of the process for inducing anisotropic porosity in real manufacturing settings. There is no discussion of the process windows or sensitivity of the porosity profile to variations in sintering parameters. The suggestion is to include a discussion on process control, potential implementation challenges in a production environment, and alternative routes (e.g., templated sintering or 3D printing) that might provide more tunable porosity.
• Finally, while the concept is interesting, there is no discussion on potential trade-offs introduced by anisotropic porosity, such as reduced thermal conductivity or mechanical strength in certain directions. The suggestion is to include a discussion on whether vertical porosity, while effective in reducing cohesive cracking, could create other vulnerabilities under mechanical or thermal stresses.
• This manuscript could be significantly improved by clarifying the mechanisms behind porosity formation, quantifying cracking behavior, detailing modeling assumptions, and expanding the discussion on reliability and scalability.

 

Author Response

Thank you so much for giving the review report. Please see the attachment as a response letter.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have made substantial revisions to address the major concerns raised in the first review. The finite element analysis section now includes mesh strategy, element selection, numerical convergence details, and material fitting for high-porosity and low-porosity s-Ag. Additional SEM images and manufacturing feasibility discussion for local press (LP) have been added, along with more recent literature citations. These changes improve transparency and strengthen the link between APS/MPS distribution and cohesive cracking suppression. Remaining issues are relatively minor and can be addressed without fundamental changes.

1.The study still does not provide the number of tested samples and basic statistical metrics (such as mean, standard deviation, or error bars) for each experimental condition, which limits the repeatability and robustness of the results.

2.The APS threshold used to explain the suppression of CCM is still taken from existing literature, lacking direct quantitative validation within this study, for example through correlation analysis between crack density and APS level.

3.Some FEA contour plots still lack numeric scale bars, the SEM and SAT images have insufficient contrast, and the figure captions are not fully self-contained, making it difficult to interpret the figures without referring to the main text.

4.The manufacturability discussion of the LP process remains largely qualitative, and it is recommended to add a comparison table summarizing the differences between FP and LP in terms of pressing area control, achievable porosity gradient, processing time and complexity, alignment tolerance, and thermal resistance impact.

5.The expanded reference list addresses the previous gap, with more works from IEEE T-CPMT and Microelectronics Reliability. This strengthens the positioning. One minor suggestion is to ensure that the most relevant recent works on anisotropic porosity and graded sintering are cited not only in the introduction but also in the discussion when interpreting results.

 

Author Response

Dear Reviewer 2
Thank you so much for providing a review report again. We would like to answer your points as attached file.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

I have carefully examined the revised version of the manuscript. The authors have satisfactorily addressed all of my previous comments and questions. The revisions have significantly improved the clarity, depth, and scientific quality of the work.

The methodology and analysis are now presented more clearly, the discussion is better supported by relevant literature, and the figures/tables have been improved for better readability. I find that the manuscript now meets the standards for publication in electronics.

I therefore recommend acceptance in its current form.

Author Response

Dear Reviewer 3
Thank you so much for giving us about your feed back comments on our revised manuscript. Thanks to your comprehensive review report in the 1st round, we could improve the quality of our paper. We really appreciate your reviews. Best regards,