Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors brazed Al₂O₃ ceramic and Kovar alloy by AgCu-2Ti filler. The microstructure of the brazed joint was investigated, and the effects of the brazing parameters on the microstructure and shear strength of the joints were studied. To improve the quality of the work, the following comments should be considered.

 

1. In the statement “A belt-like TiFe2+TiNi3 IMC could be found in the filler layer”, what do the numbers (2 and 3) in the chemical formulas represent? Do they need to be subscript? (others e.g. line 168, line 169 and Table 3, carefully review the entire manuscript)

 

2. Pay attention to the long sentences which can be more clarity. (e.g. lines 46-49 is actually composed of two sentences, the same to lines 128-130)

 

3. Since the cooling rate is also an important parameter for the performance of the brazed joints, it should be provided clearly in the Section 2.

 

4. Are Figures 3a-3c secondary electron images or backscattering electron images? The contrast of the images needs to increase to clearly present different phase distributions. Same suggestion to Figure 7,

 

5. Why the (Fe,Co)0.64Ni0.36 peak is the most obvious in the XRD results (Figure 4)? How to prepare the corresponding test sample?

 

6. What is the selection principle of the brazing temperature?

 

7. Some notes in the figures are not clear enough (indistinct or too small). (e.g. Figures 5c-5i, Figures 6c-6i, Figures 8e-8h, Figures 9e-9h, Figure 10, Figure 11a and Figure 13a)

 

8. The vertical coordinate scales in Figure 10 are suggested to remain consistent to make an easier comparison.

 

9. Why the shear strength value at 950 ^oC under the heating rate of 2.5 ^oC/min has a relatively large error (170 ± 61 MPa)?

Author Response

Comments 1: In the statement “A belt-like TiFe2+TiNi3 IMC could be found in the filler layer”, what do the numbers (2 and 3) in the chemical formulas represent? Do they need to be subscript? (others e.g. line 168, line 169 and Table 3, carefully review the entire manuscript)

Response 1: Thank you for pointing this out. We agree with this comment. The numbers represent the atom numbers in a single molecule, which should be subscript. Therefore, we have checked the entire manuscript and revised the manuscript as you advised in line24, line 26, line 169, line 170 and Table3.

Comments 2: Pay attention to the long sentences which can be more clarity. (e.g. lines 46-49 is actually composed of two sentences, the same to lines 128-130)

Response 2: Thank you for pointing this out. We agree with this comment. Therefore, we have changed the structure of the sentences in lines 47-50, and lines 129-132.

Comments 3: Since the cooling rate is also an important parameter for the performance of the brazed joints, it should be provided clearly in the Section 2.

Response 3: Thank you for pointing this out. We agree with this comment. Based on the obtained cooling curve of the furnace temperature, the data was well fitted by a single-exponential decay function and we have changed the description of cooling process in lines 135-137.

Comments 4: Are Figures 3a-3c secondary electron images or backscattering electron images? The contrast of the images needs to increase to clearly present different phase distributions. Same suggestion to Figure 7

Response 4: Thank you for pointing this out. We agree with this comment. The SEM images in Figure 3 are SE2 images, so we changed the captions for Figure 3. Meanwhile, we increase the contrast of the images in Figure 3 and Figure 7.

Comments 5: Why the (Fe,Co)0.64Ni0.36 peak is the most obvious in the XRD results (Figure 4)? How to prepare the corresponding test sample?

Response 5: Thank you for pointing this out. The test samples prepared for XRD stay the same with the samples for the SEM, cutting the brazing joint transversely at their center for cross-sectional analysis. So the (Fe,Co)0.64Ni0.36 phase which we believe consistent with Kovar is the most obvious.

Comments 6: What is the selection principle of the brazing temperature?

Response 6: Thank you for pointing this out. Before carrying out brazing experiments, we thoroughly investigated the brazing research and found the propriate range for brazing Al2O3 and Kovar. The study of the microstructure and mechanical properties of joint around 900 °C have been fully studied, so we were trying to fulfill the research gap of brazing under super high temperature and low heating rate.

Comments 7: Some notes in the figures are not clear enough (indistinct or too small). (e.g. Figures 5c-5i, Figures 6c-6i, Figures 8e-8h, Figures 9e-9h, Figure 10, Figure 11a and Figure 13a)

Response 7: Thank you for pointing this out. We agree with this comment. Therefore, we checked all the figures in the manuscript and changed the size and color of notes in all the pictures to make the figures more readable.

Comments 8: The vertical coordinate scales in Figure 10 are suggested to remain consistent to make an easier comparison.

Response 8: Thank you for pointing this out. We agree with this comment. Therefore, we changed the vertical coordinate scales in Figure 10.

Comments 9: Why the shear strength value at 950 ^oC under the heating rate of 2.5 ^oC/min has a relatively large error (170 ± 61 MPa)?

Response 9: Thank you for pointing this out. 3 samples were tested for shear strength, and the maximum shear strength obtained in the experiment reached 224 MPa. Nevertheless, a subset of samples exhibited a significantly lower strength of 84 MPa, which is attributed to the inherent brittleness of the ceramic material and suboptimal control of experimental conditions. In the interest of presenting the authentic experimental data, all results have been included. Future work should focus on increasing the number of tests to enhance the statistical reliability of the data and to improve the consistency of experimental conditions.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Refer to the report

Comments for author File: Comments.pdf

Author Response

Comments 1: Novelty

Please clarify what is new compared to previous studies (e.g., Cao et al., Wang et al.). Ti–O and IMC layer formation in this system is already reported. Is the novelty the heating-rate effect, microstructural quantification, or mechanical correlation?

Response 1: Thank you for pointing this out. We agree with this comment that numerous studies have confirmed the presence of a reaction layer on the ceramic side in the brazing of alumina ceramics to Kovar alloys. However, this work is the first research to propose that this ceramic-side reaction layer can be specifically subdivided into two distinct sub-layers as Ti-O rich layer and IMC rich layer. A novel approach, by manipulating the heating rate, was employed to elucidate the formation mechanisms of these two sub-layers. Furthermore, this study provides a thorough report on establishing the crucial "process-microstructure-property" relationship by analyzing the shift in fracture location from the ceramic-side reaction layer to the interior of the filler metal upon changing the brazing temperature and heating rate. More importantly, we significantly enhanced the shear strength of the joints (achieving a maximum exceeding 200 MPa) through optimized heating rates, thereby exploring the potential of process parameters. But we do need to emphasize the novelty of this work, therefor we made adjustment in Abstract an Introduction in lines 17-20, lines105-108.

Comments 2: Brazing Parameters

• Why were only 2.5 °C/min and 10 °C/min chosen? Would intermediate rates provide more useful trends?
• At 1100 °C, the seam is almost entirely IMC. Is this condition practically relevant in real brazing applications?

Response 2: Thank you for pointing this out. We agree with this comment. A heating rate of 10°C/min, commonly employed in brazing due to its balance between joint quality and process efficiency, served as the baseline (control group) in this study. The rate of 2.5°C/min was designated as the experimental group to investigate the effect of a slower heating rate on the joint's microstructure and properties. Based on our findings, the Ti-O layer formed during the 2.5°C/min heating stage is sufficient to inhibit the formation of the IMC rich layer. This insight suggests that future work could explore higher heating rates to identify the critical threshold for this transition. Subsequent research should focus on the interplay between brazing temperature and heating rate by systematically studying different parameter combinations. The ultimate goal is to identify the optimal process that yields the highest joint strength while minimizing overall process time. We thank you for this valuable suggestion. This avenue of research is indeed significant, and we have incorporated the points you raised into the research prospects in Conclusion sections of our manuscript in lines 494-495.

Thank you for pointing this out. Although the joint brazed at 1100°C exhibits substandard mechanical properties with no practical application, it was investigated solely for fundamental research purposes. Studying this parameter allows us to explore further microstructural evolution and compare it directly with the microstructure and properties of the joint brazed at 1000°C, thereby helping to define the viable process window.

Comments 3: Mechanical Properties

• The reported scatter in shear strength (±61 MPa) is quite high. How many replicates were tested per condition?
• Please include error bars and perform statistical analysis (e.g., ANOVA or t-test) to confirm differences are significant.
• Why was only shear strength measured? Have you considered tensile strength or fracture toughness?

Response 3：Thank you for pointing this out. 3 samples were tested per condition in the shear test.

Thank you for pointing this out. We agree with this comment. We performed ANOVA and t-test using SPSS to get a better statistical analysis of the shear test results. The results show that reducing the heating rate significantly enhances the shear strength of the joints and no statistically significant difference was found in shear strength between the high brazing temperature (1000°C, 1100°C) and low brazing temperature (900°C, 940°C) groups due to substantial within-group variation and the limited sample size. Therefore, we changed the Result part in lines 292-327.

Thank you for pointing this out. In brazing research, the shear test is commonly adopted as the standard method for evaluating mechanical properties. Therefore, to facilitate comparison with and reference to other studies, we selected the shear test for this work. Future research could involve testing additional mechanical properties, such as toughness and tensile strength, to gain a more comprehensive understanding of the brazed joint's performance. We consider your suggestion highly valuable as it outlines a clear direction for future work. Accordingly, we have incorporated this perspective into the Conclusion section of our manuscript as part of the research outlook in lines 498-500.

Comments 4: Interfacial Microstructure

• The Ti–O layer thickness remains almost constant across conditions. Please explain why temperature increase does not affect growth—could oxygen diffusion be rate-limiting?
• Provide evidence distinguishing between TiO, Ti₂O, or amorphous Ti–O in the XRD results.

Response 4：

Thank you for pointing this out. The primary reason for the constant thickness of the Ti-O layer is that the formation of a dense Ti-O layer itself hinders the further diffusion of oxygen, making it a self-limiting reaction process. Although the thickness of the Ti-O layer undergoes changes during the reaction, thereby influencing the formation of the IMC layer, it ultimately reaches a stable equilibrium thickness by the end of the process.

Thank you for pointing this out. The XRD results constitute a key finding of this study. A comparison between the detected diffraction patterns and the standard peaks of potential compounds such as TiO₂ reveals that some peak positions can be matched (32.6° and 82.5°), albeit with low intensity due to their relatively low content. Considering the elemental system of the material, it is reasonable to assign these features to the presence of TiO₂ compounds because there is no peak at these positions for those phases which have been detected like Al₂O₃ and (Fe, Co)_0.64Ni_0.36.

Comments 5: Fracture Analysis

• The shift in fracture mode (reaction layer → filler layer) is central to the paper. Can you quantify fracture locations (e.g., % fracture area in IMC vs. filler vs. ceramic)?
• Did you consider residual stresses in the analysis? Finite element modeling or stress discussion would strengthen this claim.

Response 5: Thank you for pointing this out. We agree with this comment. Therefore, low-magnification BSE images of the fracture surfaces are presented below. It can be observed that for all joints, except for the one brazed at the extremely high temperature of 1100°C, the fractures predominantly occurred on the ceramic side. Furthermore, a statistical analysis was performed on the BSE images using the ImageJ software. The ratio of the fracture area on the ceramic side to the fracture area within the filler metal and interfacial region was quantified. However, this ratio did not exhibit a clear pattern and appeared largely random, with greater variation within groups than between groups. Therefore, it is considered that these results do not significantly contribute to supporting the main conclusions of this study. This may be attributed to the considerable scatter in shear strength measurements, but this represents a direction for further investigation. Consequently, this aspect has been included in the prospects for future work in lines 500-502.

Thank you for pointing this out. We agree with this comment. We are currently developing a finite element analysis (FEA) model, and the integration of simulation and experimental results will be a key focus of our subsequent publication. The present study, however, was deliberately designed to investigate the influence of heating rate and brazing temperature on microstructural evolution and the resulting mechanical properties of the joints. Residual stresses primarily develop during the cooling stage and since the primary process variations in this work (heating rate and brazing temperature) would not significantly alter the cooling dynamics, the residual stress distribution was not the central theme of our investigation. Nevertheless, we acknowledge the constructive nature of your suggestion and have incorporated it into the future perspectives in Conclusions section of the manuscript in lines 495-498.

Comments 6: Figures and Presentation

• Micrographs in Figs. 7 to 9 are difficult to interpret; please annotate IMC boundaries more clearly.
• In conclusion, “10 °C/s” appears—should this be “10 °C/min”? Please check units.
• Language needs polishing in some places for clarity.

Response 6: Thank you for pointing this out. We agree with this comment. The Figures 7 to 9 have been reproduced with modifications to the annotation size and color for better clarity. The intermetallic compound (IMC) layers have been re-annotated accordingly.

Thank you for pointing this out. We agree with this comment. Therefore we rechecked the units in the whole paper and change the unit in lines 485-486.

Thank you for pointing this out. We agree with this comment. Therefore we used professional paper polishing service to improve the English in lines 14-16, lines 31-33, lines 41-43, lines 56-57, lines 74-76, lines 79-81, lines 92-95, lines 108-109, lines 114-117, lines 132-135, lines 159-162, lines 174-176, lines 182-184, lines 186-187, lines 195-197, lines 212-214, lines 220-225, lines 227-229, lines 245-246, lines 268-270, lines 285-286, lines 333-334, lines 340-343, lines 345-349, lines 352-353, lines 363-367, lines 376-378, lines 388-390, lines 405-408, lines 448-456.

 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This work is very interesting and data-rich, holding significance for both scientific theoretical research and practical applications. The characterizaion is detailed and the shear strength of the joint is optimum. Minor revision is required before it can be accepted. 

• Please, can you describe more precisely how did you choose experimental. Is it in vacuum or in other protection?
• Please, can you describe how many repetitions in each experimental trial of shear strength you conducted. 
• There are two Table 1 in the paper. At the same time, the header of the table is also incorrect. Please double check.(Line 299)
• In Table 1 (Line149), "rest" is not commonly used for showing the chemical composition.
• The Kovar 4J34 alloy mentioned in the abstract are not the same as those mentioned in the conclusion.
• The conclusions are well-articulated and summarize the main findings of the study. However, the authors may consider discussing the broader implications of their work and potential directions for future research.
• Currently, the conclusion is too long. 

 

 

Author Response

Comments 1: Please, can you describe more precisely how did you choose experimental. Is it in vacuum or in other protection?

Response 1: Thank you for pointing this out. We agree with this comment. The brazing process was carried out in a vacuum with a pressure of 5 × 10⁻³ Pa. The Materials and Methods section has been revised to include a more detailed description of the experimental setup in lines 128-129.

Comments 2: Please, can you describe how many repetitions in each experimental trial of shear strength you conducted.

Response 2: Thank you for pointing this out. We agree with this comment. We conducted 3 samples for each process parameters and calculate the average shear strength for that condition. The Materials and Methods section has been revised in lines 145-148.

Comments 3: There are two Table 1 in the paper. At the same time, the header of the table is also incorrect. Please double check. (Line 299)

Response 3: Thank you for pointing this out. We agree with this comment. Therefore, we changed the Table number and the header of the Table 4.

Comments 4: In Table 1 (Line149), "rest" is not commonly used for showing the chemical composition.

The Kovar 4J34 alloy mentioned in the abstract are not the same as those mentioned in the conclusion.

Response 4: Thank you for pointing this out. We agree with this comment. Therefore, we changed the description “rest” to “Bal.” to in Table 1. Meantime, the Kovar we used in the experiment was accurately 4J34, so we changed the description in the conclusion in line 463.

Comments 5: The conclusions are well-articulated and summarize the main findings of the study. However, the authors may consider discussing the broader implications of their work and potential directions for future research.

Response 5: Thank you for pointing this out. We agree with this comment. Therefore, we add some frontiers for future research, in lines 494-502.

Comments 6: Currently, the conclusion is too long.

Response 6: Thank you for pointing this out. We agree with this comment. Therefore, we summarize the conclusion part of shear strength again and use professional paper polishing service to improve the English. The changed Conclusion part can be found in lines 485-489.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

1. Although the cooling rate is an important parameter for the brazing process, it does not mean that the authors have to fit the cooling curve into a formula. Instead, the authors should provide the information whether the brazed samples were furnace cooled down to room temperature or cooled at a special cooling rate (set in the thermal cycle program).

2. Replace the SE model images in Figures 3 and 7 with BSE images. The current images are not suitable for publication.

Author Response

Comments 1: Although the cooling rate is an important parameter for the brazing process, it does not mean that the authors have to fit the cooling curve into a formula. Instead, the authors should provide the information whether the brazed samples were furnace cooled down to room temperature or cooled at a special cooling rate (set in the thermal cycle program).

Response 1: Thank you for pointing this out. We agree with this comment. Therefore, we have changed the description of cooling process and deleted the fitted formula in lines 135 and 136.

Comments 2: Replace the SE model images in Figures 3 and 7 with BSE images. The current images are not suitable for publication.

Response 2: Thank you for pointing this out. We agree with this comment. Compared to SE2, BSE is more suitable for revealing the phase distribution in the microstructure. Therefore, we reacquired BSE results of the corresponding joints and compared them with the SE2 images presented in the paper, as presented in the figures below. It was showed that the BSE results did not differ significantly from the SE2 results and provided little additional information of the microstructures obtained under various process parameters. In addition, to maintain the correspondence between the EDS results and the SEM images, it should be better to retain the present SE2 images in the manuscript. Moreover, the clarity and contrast of some SE2 images (Figure. 7c, d, e, h) were adjusted this time to better illustrate the distribution of different phases in the joints.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

My concerns have been addressed. Thanks and good job!

Author Response

Thank you very much for your initial feedback. Your suggestions have been extremely helpful in enhancing the quality of our manuscript.

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

Only one chance to revise the Figures before the manuscript is accepted. Since the BSE images have been obtained, why didn't the authors carried out the corresponding element mapping scanning test? The authors can request several more days to revise the figures, but not conceal the flaws.

Author Response

Comments 1: Only one chance to revise the Figures before the manuscript is accepted. Since the BSE images have been obtained, why didn't the authors carried out the corresponding element mapping scanning test? The authors can request several more days to revise the figures, but not conceal the flaws.

Response 1: Thank you for pointing this out. We agree with this comment. We have re-acquired the corresponding BSE images and EDS mappings in Figure 3 and Figure 7, which have been updated in the manuscript accordingly. Additionally, we have re-collected the point scan data presented in Table 3.

Author Response File: Author Response.pdf