Assessing the Stress Induced by Novel Packaging in GaN HEMT Devices via Raman Spectroscopy

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have fabricated the packaged Device 112 (PD) AlGaN/GaN HEMT prepared by the MOCVD.  With details, investigations of the method and fabrication of the material were adopted for the AlGaN/GaN HEMT device. The device was studied by Micro-Raman. However, some deviancies need to be addressed before the article can published.

1.    Why the authors don’t implement any additional techniques to properly check the materials before packaging?

2.    It is suggested to check the electrical results of the fabricated packaging device.

3.    The authors have to check the Raman of the connected packing devices other than separate ones.

4.    It is suggested to perform the SEM to see the packaging side and the quality of the fabricated device and materials.

 

Please remove the words like successfully and write the names of the symbols they used in the article for different parameters.

 

Some typos need to be properly checked. 

 

 

Comments on the Quality of English Language

No

Author Response

Comments:

The authors have fabricated the packaged Device 112 (PD) AlGaN/GaN HEMT prepared by the MOCVD.  With details, investigations of the method and fabrication of the material were adopted for the AlGaN/GaN HEMT device. The device was studied by Micro-Raman. However, some deviancies need to be addressed before the article can published.

1. Why the authors don’t implement any additional techniques to properly check the materials before packaging?

Answer: We thank the reviewer for his/her comment, which helped us understand that in the experimental description, it was not made sufficiently clear that the investigation was conducted on a commercial device produced by ST Microelectronics. Given that it is a commercial device, the quality and chemical characteristics of the materials have already been studied and are guaranteed by the production process. To clarify the aim of the work we added in the introduction the following sentence: In this work, we studied a commercial electronic device produced by ST Microelectronics equipped with a wire-bonding-free packaging designed to be integrated around the chip to save space and optimize thermal dissipation. The investigation proposed in the manuscript emphasizes the importance of choosing and designing the appropriate packaging to ensure performance and lifetime suitable for the device's mission profile.

2. It is suggested to check the electrical results of the fabricated packaging device.

Answer: We agree with the reviewer point of view, but as the device is a commercial one, hence the electrical performances have been properly studied by the manufacturer. On the other side it is not possible to compare the bare and packaged devices because the electrical connections are integrated in the package itself, meaning that it is almost impossible to perform electrical measurement on the bare device.

3. The authors have to check the Raman of the connected packing devices other than separate ones.

Answer: We're sorry, but we're having trouble understanding the meaning of the suggestion. The work compares the Raman behavior of a device without packaging and one with packaging, highlighting that the intrinsic tensile stress present in the device without packaging is compensated by the presence of the package, which induces compressive stress on the layers.

4. It is suggested to perform the SEM to see the packaging side and the quality of the fabricated device and materials.

Answer: As required by the referee, we added the SEM image of the cross-sectional view of the device HEMB-PD (i.e. the entire device including the package) as obtained after ion milling. All the layer composing the device have been identified.   

Please remove the words like successfully and write the names of the symbols they used in the article for different parameters.

Some typos need to be properly checked.

Answer: As requested, we have removed the words identified by the reviewer, we have checked and listed in a proper table (Table 1.) all the names of the symbols used in the manuscript. Moreover, we solved some typos.

 

Reviewer 2 Report

Comments and Suggestions for Authors

This work focuses on an interesting field relating to the semiconductors industry, where new chips packaging strategy for high power integrated circuit, becomes everyday more important.

In fact, Raman spectroscopy is a useful tool to evaluate the stress / strain in (particularly) low dimensional materials, although can be used in many other materials / situations. The vibration signature, on a molecular scale, can be easily measured and compared in different situations in which the material is exposed. The authors attempt to follow a similar strategy to evaluate the induced stress in GaN HEMT packaging devices. The concept is not new, and, in my opinion, the authors didn’t show a work with significant prime novelty, although they had the chance to try. In fact, and without doubt, the topic underlying this work, although framed from a scientific point of view, is clearly of technological interest. And, under this concept, I think that the work needs a major revision / new experimental data before can be considered acceptable for publication. I suggest the authors to consider the following aspects:

1) A more detailed state of the art regarding the packaging / strain / stress / temperature for electronic devices needs to be done, in particular (including bibliographic references) in the paragraph in the second page, in lines 59 – 73 (even discarding vague and unimportant inscriptions in other parts of the introduction). A real prime novelty needs to be shown.

2) In the experimental part, the authors need to explain why the samples preparation does not introduce extra effects in them. I think that it is not really clear.

3) A more scientific discussion about the overall data will be welcome as for that point, the authors don’t present absolutely no new information.

4) The authors need to make some work under with samples subjected to different temperatures. I think that is obvious (including the effects they will have on the Raman signature). And it also seems that it won't be difficult with the employed experimental setup.

5) The authors need to elaborate (as it is not clear in the manuscript) the effective correlation of the current work with the impact in the chip devices expected performance.

6) Finally, where the current work can be an enhanced technique compared to the current ones in the same technological field?

Comments on the Quality of English Language

Please look carefully at the English as there are some sentences that could be improved for clarity.

Author Response

Comments:

This work focuses on an interesting field relating to the semiconductors industry, where new chips packaging strategy for high power integrated circuit, becomes everyday more important. In fact, Raman spectroscopy is a useful tool to evaluate the stress / strain in (particularly) low dimensional materials, although can be used in many other materials / situations. The vibration signature, on a molecular scale, can be easily measured and compared in different situations in which the material is exposed. The authors attempt to follow a similar strategy to evaluate the induced stress in GaN HEMT packaging devices. The concept is not new, and, in my opinion, the authors didn’t show a work with significant prime novelty, although they had the chance to try. In fact, and without doubt, the topic underlying this work, although framed from a scientific point of view, is clearly of technological interest. And, under this concept, I think that the work needs a major revision / new experimental data before can be considered acceptable for publication. I suggest the authors to consider the following aspects:

• A more detailed state of the art regarding the packaging / strain / stress / temperature for electronic devices needs to be done, in particular (including bibliographic references) in the paragraph in the second page, in lines 59 – 73 (even discarding vague and unimportant inscriptions in other parts of the introduction). A real prime novelty needs to be shown.

Answer: We agree with the referee point of view: may be that the novelty of work is not clear enough in the manuscript. We added a more detailed state of the art (lines 73–80 and 98–104) and pointed out with the following sentence that at the best of our knowledge the impact of the packaging on the stress of a GaN based HEMT has never been studied:

“Moreover, the impact of the packaging on the stress of a GaN based HEMT has never been investigated, but those phenomena have to be taken into account to design the packaging of high-performance devices.”

Furthermore, we stressed that a key point consists in the package structure. In our case the package is very complex as it includes the electrical connection.   

• In the experimental part, the authors need to explain why the samples preparation does not introduce extra effects in them. I think that it is not really clear.

Answer: This point is, in our opinion, much important. To study the stress among the layers by means of Raman spectroscopy it needs to expose the active part of the device. A simple cutting is not suitable as, by this procedure, the obtained roughness surface should be very high, and an apparent additional stress should be added to the layers. To avoid the above, we used an ion milling procedure that has been already suggested by other groups. In this case, of course, the first layers of the semiconductor will undergo amorphization. Hence, the Raman spectrum might be affected by the treatment; however, considering the volume of Raman scattering, taking into account the depth of penetration of the optical signal into the material (about 100nm) and the focal depth of the objective, we can conclude that the majority of the Raman signal comes from the deeper layers of the semiconductors. The surface amorphization may, at most, result in a slight broadening of the peak without significantly affecting its position. Additionally, it should be noted that the analysis was conducted by comparing two samples (with and without packaging) that were treated in the same manner. Therefore, even if there were an influence on the stress due to ion milling treatment, this influence would be the same in both samples. Moreover, we performed a SEM analysis on the milled HEMT-PD to attest that no (visible) damages have been induced.

To clarify the issue the referee raised, we modified as follows some sentences in the manuscript, and added some references:

• In subsection 2.1: “However, a simple cut and subsequent polishing were not the suitable experimental strategies, because the obtained roughness surface should be very high, and an apparent additional stress would be added to the layers. For this reason, ion beam milling technique was employed on both the samples with the same milling conditions. The Raman spectrum might be affected by the treatment, however it will result only in a slight broadening of the peak without significantly affecting its position, since the volume of Raman scattering and the focal depth of the objective causes the majority of the Raman signal coming from the deeper layers of the semiconductors [1109/77.621012 , 10.1063/1.3658866 , 10.1364/OME.6.002052]. To attest the quality of the growth and the absence of defect or phase transformations induced by ion milling, a SEM analysis was performed on a milled HEMT-PD device with a beam energy of 5 keV.”
• In section 3: “It should be noted that, even if there were an influence on the stress due to ion milling treatment, this influence would be the same in both samples since the milling conditions were the same. Therefore, Δω wouldn’t be affected by the samples’ preparation.”
• A more scientific discussion about the overall data will be welcome as for that point, the authors don’t present absolutely no new information.
• Answer: We have added several pieces of information to the text also to address the requests of the other referees, including some SEM images and a discussion explaining the logic behind the sample preparation and the reasons why the preparation only marginally influences stress measurement. The main purpose of the work is to demonstrate that the used package mitigates the stress induced by lattice mismatch. It has been highlighted that, for GaN devices, this type of study, to our knowledge, has never been published. We believe that in this way, the main information of the work has been sufficiently clarified.The authors need to make some work under with samples subjected to different temperatures. I think that is obvious (including the effects they will have on the Raman signature). And it also seems that it won't be difficult with the employed experimental setup.

 

Answer: We agree with the reviewer: it is crucial to explore the effects of different temperatures on the Raman signature and investigating temperature effects would undoubtedly enhance the comprehensiveness of our study. Unfortunately, our current experimental setup lacks the necessary sample holder to precisely control and measure temperature changes. Addressing this aspect would require acquiring additional equipment and modifying our setup, which may extend the scope of our research beyond its current capabilities.

 

• The authors need to elaborate (as it is not clear in the manuscript) the effective correlation of the current work with the impact in the chip devices expected performance.

 Answer: It is well known that local stress limits the dynamic performance of power electronic devices. The problem primarily lies in the different coefficients of thermal expansion of the layers composing the device, which can lead to phenomena such as delamination or cracking between layers, ultimately leading to premature device failure or drastically reducing its lifetime. The following sentence, and references, have been added: The study of the aging effect and failure mechanism caused by thermal and thermomechanical phenomena on electronic devices is currently an active research field [10.1109/ISPSD.2016.7520794 , 10.1109/PESC.2008.4592088], hence the evaluation of the built-in stress and strain is mandatory to correctly estimate the device lifetime.

In our case, to verify that the stress reduction due to the package effect positively influences the quality and reliability of the device, it would be necessary to operate both the bare device and the packaged one. Unfortunately, this is almost impossible due to the very special design of the device that integrate the electrical connections into the package. We added two references in the manuscript in which the dependance of AlGaN/GaN HEMT performance and reliability on mechanical stress and strain is investigated [10.1109/TSM.2018.2865106 , 10.1063/1.4826524].

 

6) Finally, where the current work can be an enhanced technique compared to the current ones in the same technological field?

Answer: Raman spectroscopy has been employed several times to study stress between the layers of different semiconductors and structures, so from this perspective, we cannot highlight any scientific improving. However, to the best of our knowledge, the study of stress induced by the package on GaN HEMT devices has never been addressed. The technologically innovative aspect lies in the particular design of the package, which, by compensating for the residual stress between the layers induced by lattice mismatch, should improve the device's performance.

Reviewer 3 Report

Comments and Suggestions for Authors

In this manuscript titled “Assessing the stress induced by novel packaging in GaN HEMT devices via Raman Spectroscopy”, Zainab et al micro-Raman spectroscopy was successfully employed to assess the stress/deformation induced by packaging in AlGaN/GaN high-electron-mobility transistors (HEMTs) supplied by STMicroelectronics S.r.l.. The paper is well written and well-summarized. I recommend this paper to be published in the Applied Sciences journal. However, I have a few questions.

 

1.       What was the energy of the Nd: YAG laser device operating at 532 nm used in the study?

2.       Which spectrometer was employed for the measurement of the Raman spectra?

3.        Why the 128 spectra along the red line crossing the figure orthogonally were chosen? Any particular reason for that?

4.       In Fig 4b, the Raman spectra were captured at fixed laser parameters.

5.       Why does the Raman shift take place as a function of thickness? Please explain

6.        Identify the peaks of Raman spectra in Fig 5 and label it in Fig. 5. Corresponding elements etc.

 

 

Comments on the Quality of English Language

Minor editing of English language required

Author Response

Comments:

In this manuscript titled “Assessing the stress induced by novel packaging in GaN HEMT devices via Raman Spectroscopy”, Zainab et al micro-Raman spectroscopy was successfully employed to assess the stress/deformation induced by packaging in AlGaN/GaN high-electron-mobility transistors (HEMTs) supplied by STMicroelectronics S.r.l.. The paper is well written and well-summarized. I recommend this paper to be published in the Applied Sciences journal. However, I have a few questions.

 

1. What was the energy of the Nd: YAG laser device operating at 532 nm used in the study?

Answer: The energy of photons at 532nm is 2.33eV. The laser power used to perform the experiment was 30 mW, but a neutral filter was introduced in the laser path to limit the maximum power impinging the sample surface low to ~ 400 µW. This precaution needs to prevent any displacement of the Stokes peak induced by temperature increasing. This info has been added to the text.

2. Which spectrometer was employed for the measurement of the Raman spectra?

Answer: The experiment was conducted using a confocal Raman spectrometer based on a Sol MS3504i spectrometer equipped with a 2400 lines/mm^-1 grating blazed at 400 nm. This info has been added to the manuscript.  

3. Why the 128 spectra along the red line crossing the figure orthogonally were chosen? Any particular reason for that?

Answer: The spectra were collected along a vertical line crossing the layers because this allows us to observe how the Raman peaks shift from one layer to another and thus highlight the presence of stress at the interfaces. Since we had to integrate the signal for very long times, we chose to move along a line instead of performing a real Raman map, which would have required excessively long measurement times. We are grateful to the reviewer for this comment and added the following sentence to the manuscript to clarify this point: “In particular, the 128 spectra were acquired along a vertical line crossing the layers (the red line crossing the Figure 4a). This strategy allows us to observe how the Raman peaks shift from one layer to another and thus highlights the presence of stress at the interfaces. Since obtaining a sufficient signal-to-noise ratio required integrating the signal for very long periods, we chose to proceed along a line instead of performing a full Raman mapping. Indeed, the latter would have required extremely long measurement times, potentially leading to results affected by thermal drift and unpredictable variations in experimental conditions. Moreover, it would not have provided additional useful information to the measurement.”.

4. In Fig 4b, the Raman spectra were captured at fixed laser parameters.

Answer: Yes, all the Raman spectra were captured at fixed laser and experimental conditions.

5. Why does the Raman shift take place as a function of thickness? Please explain.

Answer: Growing materials with different lattice spacings on top of each other, lead at the interface a mismatch appears, and the lattice is deformed and therefore stressed. For this reason, the Raman peaks shift at higher or lower frequencies depending on whether the stress is tensile or compressive. Moving away from the interface, the lattice tends to release the stress by returning to its natural structure, and if the material thickness is sufficient, the stress completely disappears.

6. Identify the peaks of Raman spectra in Fig 5 and label it in Fig. 5. Corresponding elements etc.

Answer: We added a label in Figure 5b indicating the Raman Peaks referring to the GaN. In Figure 5a the spectra refer to GaN only because they were acquired far away from the interface. To avoid any confusion, we added a label indicating that the material was GaN.