Investigation of the Interfacial Reactions between the CoCuFeNi High Entropy Alloy and Sn Solder
, Yong-Chi Chang 1, Chao-Hong Wang 2, Andromeda Dwi Laksono 1 and Yee-Wen Yen 1,*
Round 1
Reviewer 1 Report
Metals review
Investigation of the Interfacial Reactions between the 2
CoCuFeNi High Entropy Alloy and Sn Solder
Dear authors,
Thank you for your article. I guess, it needs some corrections.
1) Is the format of chemical formulae correct? FeSn2 x FeSn2 . Lower exponent is more appropriate, I guess.
2) Line 28 – If Heas consist of 5 elements, then your alloy with 4 elements is not a HEA, probably MEA, Medium Entropy Alloy
3) Line 32 – can you give an example of limitations of conventional lead-frame materials?
4) Line 42 – Is this alloy, or similar more conductive than pure Cu?
5) Line 50 – What were the conditions of the Li et al. study?
6) Line 58 – what is new in your study compared to referred articles?
7) Line 59 – If you use the word „identified “, I would expect XRD or any other diffraction technique, which is not your case.
8) Line 69 – Can you give also dimensions of the ingots?
9) Line 80 – How did you choose the temperature and time of the homogenisation treatment?
10) Line 80 – The term “aging” is usually connected with precipitation, which is not this case
11) Line 83 – Metallurgical of metallographic?
12) Line 85 – How was the Sn and HEA specimens arranged in the capsule to be in contact?
13) Line 93 – Probably metallographic, rather than metallurgical
14) Line 95 – What was inspected? Surface of metallographic cuts?
15) Line 99 – probably particles, not elements
16) Line 99 – Why did you use ImageJ? Was it properly calibrated?
17) Line 100 – What magnification was used for the measurement?
18) Line 104 – Results and discussion
This is probably allowed, but in this case, I would like to suggest you to give Results separately and Discussion separately. Results could by together for all 3 temperatures.
Raw results of the layer thickness shall be given, e.g., in tables.
The discussion sections in current paragraphing are repeated in each section. So once would be enough for all experimental conditions. More general and wider discussion will be useful.
19) Line 105 – How did you choose experimental conditions? Time and temperatures of annealing? Are they different to referred experiments?
20) Line 110 – The chemical composition is from EDS? Do you have any spectra, image?
21) Figure 1, 2, 3 – Pictures are too small, with too small magnification, contrast is low too.
For image quality see e.g. DOI 10.1007/s10854-006-9031-5
22) Line 116-143 – this is discussion
23) Line 159 – Any comment of the different morphologies of the 2 precipitated phases?
24) Line 164 – Are other elements diffusion rate also higher with temperature, as in the case of Co?
25) Is there any Cu diffusion in Sn? E.g., in solid solution?
26) Are the observed IMCs thermally and electrically conductive?
27) Hume Rothery rules – AFAIK thus rules are intended for metals, not for intermetallic compounds. Of course, some elements are interchangeable in IMCs but the explanation could be probably different.
28) Line 204 – This problem of diffusion growth of IMC layer between 2 solids is probably well known. What is the physical meaning of K constant? Any connection with any diffusion coefficients? E.g., Fe in Sn?
e.g. https://www.sciencedirect.com/science/article/pii/S2095034920300076
29) Line 218 – Any connection of Q with diffusion coefficients, or something similar? Or different interpretation?
30) Fig. 4 – It is not sufficient to give the data just in this form. This is data interpretation. Data tables needed. Also, image on linear time is needed.
31) Figure 4 – Why are error bars only for 1 curve? These bars are calculated from 3 measurements only?
32) Discussion – Is presence of intermetallic layer acceptable or not? How is the situation of pure elements from the HEA with Sn, or binary or ternary alloys?
Author Response
Answers to Reviewer 1's comments:
A-1: We are very grateful for Reviewer 1’s valuable advice. A lower exponent is more appropriate. We have checked all chemical formulas in the manuscript, and they are all the lower exponent.
A-2: We are very grateful for Reviewer 1’s valuable advice. The definition of high entropy alloy usually requires the inclusion of at least five different significant elements, but this is not a strict rule, and some scholars may refer to alloys composed of four elements as high entropy alloys. We also have revised the word in the manuscript at line 29. For related literature see e.g., DOI 10.1016/j.ssc.2022.114980.
A-3: We are very grateful for Reviewer 1’s valuable advice. Conventional lead-frame materials, such as Cu and C alloys, have some limitations that can affect their performance and suitability, such as limited mechanical strength, which can limit their ability to withstand stress and strain, especially in high-temperature and high-vibration environments. In addition, we have added an example of the limitations of conventional lead-frame materials with a red mark in lines 34-35.
A-4: We are very grateful for Reviewer 1’s evaluation of our manuscript. However, we didn't measure some conductivity in this study. On your valuable suggestion, we will do the conductive tests in the next research. In addition, High entropy alloy is generally lower than that of pure copper. Cu has the highest electrical conductivity of all pure metals at room temperature. In contrast, the electrical conductivity of HEAs is typically lower than that of pure metals due to the presence of multiple elements in their structure, which can increase the number of scattering sites for electrons. That being said, the electrical conductivity of HEAs can still be relatively high and comparable to that of some copper alloys or other conductive materials, depending on their composition and processing conditions. HEAs can also offer other advantages such as high strength and resistance to corrosion and wear, which make them attractive for use in applications where conductivity is not the only consideration.
A-5: We are very grateful for Reviewer 1’s evaluation of our manuscript. We have updated the conditions of Li et al. study with a red mark in lines 63-64.
A-6: We are very grateful for Reviewer 1’s valuable advice. Our study investigates raising the temperature to 300, 375, and 450 °C, and using a simpler and common quaternary composition as the substrate to help tevaluate the feasibility of using HEA as a desirable lead-frame material in electronic packaging.
A-7: We are very grateful for Reviewer 1’s valuable advice. We have added an XRD result in new figure (Figure 4) of the XRD diffraction patterns of the Sn/CoCuFeNi couple reacted at 450°C.
A-8: We are very grateful for Reviewer 1’s valuable advice. We have added the ingot dimension with a red mark in lines 91-92.
A-9: We are very grateful for Reviewer 1’s valuable advice. We chose this temperature and time for homogenization treatment based on the experience and the studies of other researchers, and we checked the homogeneity of the alloy through the SEM and EDS. For related studies see e.g., DOI 10.3390/ma15186324.
A-10: We are very grateful for Reviewer 1’s valuable advice. We have correct it to “homogenized” with a red mark in line 94.
A-11: We are very grateful for Reviewer 1’s valuable advice. We have corrected it to “metallographic” with a red mark in line 97.
A-12: We are very grateful for Reviewer 1’s valuable advice. The Sn and HEA encapsulated in a quartz tube were placed by first placing the high-entropy alloy into the tube with pliers, and then placing the Sn behind the high-entropy alloy to enclose it. We have added how the Sn and HEA specimens are arranged in the capsule to be in contact with a red mark in lines 104-106.
A-13: We are very grateful for Reviewer 1’s valuable advice. We have corrected it to “metallographic” with a red mark in line 109.
A-14: We are very grateful for Reviewer 1’s valuable advice. We inspected the surface of metallographic cuts using a scanning electron microscope (SEM; Hitachi, TM-3000; Tokyo, Japan), and we added it with a red mark in line 111.
A-15: We are very grateful for Reviewer 1’s valuable advice. We have corrected it to “particles” with a red mark in line 115.
A-16: We are very grateful for Reviewer 1’s valuable advice. ImageJ is commonly used for measuring the thickness of intermetallic compounds (IMCs) and grain size, and it was calibrated by using the scale bar in the original high-resolution SEM image. In addition, we measure multiple times to ensure accuracy, so we use ImageJ.
A-17: We are very grateful for Reviewer 1’s valuable advice. We used the magnification of 3000x for the measurement.
A-18: We are very grateful for Reviewer 1’s valuable advice. We have revised to give Results separately and Discussion separately in the new manuscript.
A-19: We are very grateful for Reviewer 1’s valuable advice. We chose the temperature and the time for annealing based on the experiment results and the studies of other researchers because we want to show the IMC changes and states completely in the annealing time and temperature. In addition, our study is different from referred experiments (ref. [9][10]), and we investigate raising the temperature to 300, 375, and 450 degrees, and using a simpler and common quaternary composition as the substrate to help to evaluate the feasibility of using HEA as a desirable lead-frame material in the electronic packaging.
A-20: We are very grateful for Reviewer 1’s valuable advice. Our EDS results were analyzed at multiple points and many times within the scope of the IMCs area, and then averaged to ensure accuracy. Therefore, no relevant spectra were attached.
A-21: We are very grateful for Reviewer 1’s valuable advice. We have updated Figures 1, 2, and 3 according to DOI 10.1007/s10854-006-9031-5.
A-22: We are very grateful for Reviewer 1’s valuable advice. We have revised to give Results separately and Discussion separately in the new manuscript.
A-23: We are very grateful for Reviewer 1’s valuable advice. The IMC at 375°C was similar to that formed in the Sn/CoCuFeNi couple reacted at 300°C. Still, it became thicker as the reaction temperature increased, and it was also more obvious between Sn solder and HEA alloy. We have added a new comment with a red mark in lines 169-170.
A-24: We are very grateful for Reviewer 1’s valuable advice. Besides cobalt, nickel also accelerates its diffusion rate to the solder as the temperature increased because the precipitated phase of (Co, Ni)Sn2 could be observed at the shorter reaction time with higher temperature, as shown in Fig. 1(f), Fig. 2(d)-(f), and Fig. 3(e)-(f).
A-25: We are very grateful for Reviewer 1’s valuable advice. In the Sn/CoCuFeNi system, we did not observe any Cu diffusion in Sn solder by SEM and EDS. However, we will add more copper to our HEA, such as CoFeNiCu2 HEA to confirm the diffusion behavior of copper in the next research.
A-26: We are very grateful for Reviewer 1’s valuable advice. However, we did not measure some thermally and electrically conductive tests in this study. On your valuable suggestion, we will do the tests in the next research.
A-27: We are very grateful for Reviewer 1’s valuable advice. This study employs high-entropy alloys, and since the difference in atomic radii between Fe and Co is extremely small, the substitutional reaction could be explained by the Hume-Rothery rule occurs. In the case of IMCs, according to XRD phase structure analysis at the Sn/CoCuFeNi couple, FeSn2 was the base phase in the couple. However, due to the diffusion of elements in the high-entropy alloy to the solder side, the substitutional reaction described by the Hume-Rothery rule occurs, so it confirmed of the IMC phases was (Fe, Co)Sn2 by EDS.
A-28: We are very grateful for Reviewer 1’s valuable advice. The growth of a compound or IMC layer can be represented by the x = (k×t)1/2 empirical power law, where k indicates the growth rate constant. k is the temperature-dependent rate constant, and its physical meaning is to determine the rate of IMC growth in the specific temperature. In addition, the growth rate constant represents the overall growth rate of the IMC, while the diffusion coefficients represent the diffusion rate of atoms. Each type of atom has a different diffusion rate, but the growth of the IMC is also affected by the diffusion rate of specific atoms. Therefore, there is a positive relationship between the growth rate constant and the diffusion coefficients of specific atoms, such as Fe in Sn.
A-29: We are very grateful for Reviewer 1’s valuable advice. Q is the apparent activation energy of IMC growth, and it could be obtained by the Arrhenius equation k = k0[exp(-Q/RT)], which is the variation in the growth rate constant (k) with temperature. Therefore, there is a relationship between the growth rate constant and the Q because of the Arrhenius equation.
A-30: We are very grateful for Reviewer 1’s valuable advice. We have added an IMC thickness data interpretation in the new table (Table 2) of the thickness of IMC reacted at times and temperatures in the Sn/CoCuFeNi couples and linear time in the new figure (Figure 5) of the thickness of IMC versus reacted at times and temperatures in the Sn/CoCuFeNi couples.
A-31: We are very grateful for Reviewer 1’s valuable advice. The error bars are for three curves, but the slope and the scale in the three temperature was a huge difference, so they are not obvious in the figure.
A-32: We are very grateful for Reviewer 1’s valuable advice. According to other studies (ref. [9][10]) and the Fe-Sn phase diagram, it is reasonable and acceptable for the (Fe, Co)Sn2 phase to be generated in this system. Additionally, according to the Fe-Sn phase diagram, FeSn2 phase is expected to form between 300°C to 450°C. Therefore, it is expected that the reaction between pure Fe and Sn will also form the FeSn2 phase. In addition, we will add more Cu to our HEA, such as CoFeNiCu2 HEA to confirm whether the Cu6Sn5 or other phases including Cu formed or not in the next research.
Reviewer 2 Report
The article presents the results of studies of various properties of the high-entropy CoCuFeNi alloy, as well as their characterization and phase transformations under thermal heating conditions. In general, this line of research is very promising, and the methods used by the authors make it possible to estimate the established dependences with high accuracy. The article corresponds to the subject of the declared journal and can be accepted for publication after the authors answer a number of questions that the reviewer has when analyzing it.
1. First, the authors should provide more details about the experimental procedure, as well as the reasons for the choice of exposure temperatures.
2. The authors state that the growth mechanism of the (Fe, Co)Sn2 phase is controlled by diffusion in Sn/CoCuFeNi vapors. At the same time, the authors should give more details about the growth mechanisms themselves, as well as the influence of external conditions.
3. On the presented SEM images, it is necessary to highlight the thickness of the forming solder and the new phase.
4. These changes in thickness have large measurement errors, the authors should explain why such large errors are associated?
5. The authors should also cite the results of X-ray phase analysis of the structures under study, as well as the dynamics of phase changes.
6. In conclusion, the authors should provide data on further plans in this area of research.
7. When interpreting the data, the authors should pay attention to the comparison of the obtained results with the results of previous studies.
Author Response
Answers to Reviewer 2's comments:
A-1: We are very grateful for Reviewer 2’s valuable advice. We chose the temperature and time for homogenization treatment based on the experience and the studies of other researchers, and we checked the homogeneity of the alloy through the SEM and EDS. In addition, we chose the temperature and the time for annealing based on the experiment results and the studies of other researchers because we want to show the IMC changes and states completely in the annealing time and temperature. In addition, our study is different from referred experiments (ref. [9][10]), and we investigate raising the temperature to 300, 375, and 450 degrees, and using a simpler and common quaternary composition as the substrate to help to evaluate the feasibility of using HEA as a desirable lead-frame material in the electronic packaging.
A-2: We are very grateful for Reviewer 2’s valuable advice. We have added more details about the growth mechanisms with a red mark on lines 261-265.
A-3: We are very grateful for Reviewer 2’s valuable advice. We have highlighted the IMC phase, substrate, and solder by the arrow in Figures 1 to 3, and also added an IMC thickness data interpretation in the new table (Table 2).
A-4: We are very grateful for Reviewer 2’s valuable advice. We have corrected our error bar calculation and updated the new figure (Figure 5). The error bar in the new figure is much more reasonable.
A-5: We are very grateful for Reviewer 2’s valuable advice. We have added an XRD result in the new figure (Figure 4) of the XRD diffraction patterns of the Sn/CoCuFeNi couple reacted at 450°C to prove the phase we found.
A-6: We are very grateful for Reviewer 2’s valuable advice. We have added further plans in the conclusion.
A-7: We are very grateful for Reviewer 2’s valuable advice. We have added a comparison of the obtained results with the results of previous studies in the discussion part.
Reviewer 3 Report
1、This paper introduces the interfacial reaction between CoCuFeNi HEA and Sn and it's very traditional. What are the new findings from the reference [9]?
2、The experiment results of EDS should be given, including specific values for the content of each element. Besides, the detection position should be indicated on the figure.
3、To investigate the growth mechanism of IMC, the growth kinetics curve(log curve of thickness and time) and the value of growth kinetic index should be given.
Author Response
Answers to Reviewer 3's comments:
A-1: We are very grateful for Reviewer 3’s valuable advice. Our study is different from referred experiments (ref. [9][10]), and we investigate raising the temperature to 300, 375, and 450 degrees, and using a simpler and common quaternary composition as the substrate to help evaluate the feasibility of using HEA as a desirable lead-frame material in the electronic packaging. Therefore, we have new findings, such as another precipitated phase (Co, Ni)Sn2 was formed on the solder, FeSn2 phase formed at 300 to 450°C, and the Iron and Cobalt had substitutional reaction occurred, etc. In the future, we will also do thermally and electrically conductive tests to study IMC thermally and electrically conductive.
A-2: We are very grateful for Reviewer 3’s valuable advice. We have added EDS results in a new table (Table 1). In addition, our EDS results were analyzed at multiple different points and many times within the scope of IMCs area, and then averaged to ensure accuracy. Therefore, there was not have specific detection position in the results.
A-3: We are very grateful for Reviewer 3’s valuable advice. We have added an IMC thickness data interpretation in the new table (Table 2) of the thickness of IMC reacted at times and temperatures in the Sn/CoCuFeNi couples and a growth curve in the new figure (Figure 5) of the thickness of IMC versus reacted at times and temperatures in the Sn/CoCuFeNi couples.
Round 2
Reviewer 1 Report
Thank you for modyfying the article. I think Table 2 shall be moved to Results section. Otherwise it is good.
Good luck
Author Response
We are very grateful for Reviewer 1’s valuable advice. We remove Table 2 in the revised manuscript.
Reviewer 3 Report
The author addressed my concerns about this article and this article can be published in this journal
Author Response
Thanks for Reviewer 3's comment.
