Correlation of Crystal Defects with Device Performance of AlGaN/GaN High-Electron-Mobility Transistors Fabricated on Silicon and Sapphire Substrates
, Tae-Hoon Jang 2, Kyu-Hwan Shim 1,2,* and Chel-Jong Choi 1,*
Round 1
Reviewer 1 Report
Dear editorial board of MDPI electronics,
The authors presented two AlGaN/GaN HEMT epi on Si and Al2O3 with different quality level, and compared their characteristic with and without passivation. The result itself isn’t far from expectation. However, from the point of scientific publication, the quality is fundamentally flawed and make the reviewer questioning the credibility of content. Here are some critical problems:
1. From line 34 to 38, the author stated that there’s no commercially available HEMT epi on market and sapphire is a more suitable substrate for nitride growth than SiC. Both are not true.
2. Line 55 and 56, “…The AlGaN/GaN HEMT has positive surface donor-like states, which have been reported to be the actual origin of two-dimensional electron gas (2DEG)” This is also not true. If this is true, surface passivation of these donor state shall terminate the 2DEG, then the whole study become invalid.
3. Line 66 to 97, “…Despite the advantages of SiO2 passivation, there are only a
few reports available on the SiO2 passivation mechanism” Surface passivation of HEMT is an intensively-studied topic, the authors should not create the novelty by being blind to other people’s work.
4. In figure 1, both sapphire and Si substrate used GaN buffer for growth. However, it’s well known that GaN layer has a melt-back etching effect on Si in MOCVD, so its’ always AlN buffer directly on Si. Did the authors have some fundamental breakthrough on MOCVD growth technology?
5. Line 113 to 116 repeated the step from 104 to 110.
6. Line 139 to 140 said the sapphire has a higher thermal conductivity, but line 148 said the results implied Si has a higher thermal conductivity. This is an unforgivable self-inconsistency. Thermal conductivity of materials is a nature constant that doesn’t need to be implied.
7. Line 161, “…the effectiveness of the SiO2 passivation current”, what does that even mean?
8. Line 162, “…SiO2 passivated Si substrate”. No, you don’t grow HEMT on SiO2.
9. Line 164, “… SiO2 passivation in reducing the electric field between the gate and the drain” No, SiO2 passivation might reduce the leakage current, but SiO2 won’t reduce the electric field. SiO2 is not metal.
… and the reviewer gave up reading. In general, the authors might need to strengthen their basic knowledge in relevant physic and technology, and then re-examine their work thoroughly before submitting again. Or the reviewing process will be a time-wasting and painful process.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The proposed manuscript presents a theoretical study on the new stacked gate oxide L-shaped tunnel field effect transistor (LTFET) while the stacked gate oxide structure consists of high-k and SiO2 dielectrics. Before the manuscript is accepted for publication, there are some comments that the authors are required to address them.
Comment 1: There is a contradictory statement in the abstract and Results sections.
In the abstract, the authors claim that the performance of the device is improved by passivation, but in the results section, it has been said against it.
Comment 2: In Fig.2, the schematic of the fabricated device is illustrated. It is very useful if the experimental images of the fabricated devices are added.
Comment 3: In Fig.2, the I-V characteristics of devices are drawn, are there for On-state or Off-state?
The authors state that the performance of the devices will be improved while the Drain current is increased, but in the Off-state the less current will cause better performance.
Comment 4: The authors, state that the increasing drain current could be due to the thermal conductivity of the substrate, it is useful to explain it from the physical point of view.
Comment 5: In Fig. 2(b), it seems that the current-voltage curve for zero gate voltage, while silicon is used as a substrate is omitted.
Comment 6: The definition of VBR is not correct, "VBR is defined as VDS at IDS = 1 mA at a gate bias VGS = –10 V", in Fig. 3, the gate current versus drain voltage for a gate voltage of -10 V is illustrated.
Comment 7: According to the data reported in the manuscript, the breakdown voltage does not change after passivation in the devices with Sapphire substrate compared to with Si substrate, what is the physics behind it?
Comment8: The caption of Fig.4, must also include the trans-conductance.
Comment9: Based on the manuscript the authors have stated that:
"staIt can also 179 be observed that the gm of the AlGaN/GaN HEMT fabricated on the Si substrate is lower 180 than that of the HEMT fabricated on the sapphire substrate at a particular VGS because of 181 the higher self-heating effect in Si."
· The gm of the sapphire substrate is not higher than that of the Si substrate for all VGS, for a wide range of gate voltage the performance of such devices is better in the case of the Si substrate.
· In the special VGS, the gm curves of Si and sapphire substrate devices cross each other, while the behavior of the device is changed at that point, it will be very valuable if the authors explain this according to the physical point of view.
· What is the particular VGS?
Comment 10: All the data and also the image related to the experimental results are valuable and can improve the quality of the manuscript, it is recommended that all this information be included in the manuscript.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors studied the impact of substrate nature on physical properties of AlGaN/GaN structure which affect directly the HEMT performances. This work has some merit; however it should be thoroughly improved following these comments:
1/ The introduction should be enhanced; the authors should be careful to repeat the same reference as references 11 and 25 are the same and the references 1 and 37 too.
2/ Only the reference 4 reported on device and the others references 1, 2 and 3 reported on growth conditions and their impact on the GaN quality. This is not as mentioned by the authors in the beginning of the introduction, so they need to change them by others. For example the authors can add the reference on the blue LEDs (Effect of SiN Treatment on Optical Properties of InxGa1−xN/GaN MQW Blue LEDs. DOI: 10.1007/s11664-017-5383-2).
3/ In line 43 and 43, the authors talk about the optimisation of the growth conditions, and they forget to put some references on the used process to reduce the dislocation density of GaN layer using thin buffer layer grown at low temperature or use patterned substrate. This type of substrate can be obtained by ELOG or SiN treatment of sapphire. They can use this reference (Correlation between morphological, electrical and optical properties of GaN at all stages of MOVPE Si/N treatment growth. DOI:10.1016/j.spmi.2006.09.018)
In addition, they need to mention that the growth conditions affect the dislocations density too such as the effect of nature of the carrier gas. They can add this reference (Comparative study on the nanomechanical behavior and physical properties influenced by the epitaxial growth mechanisms of GaN thin films. DOI: 10.1016/j.apsusc.2021.152188).
Moreover, they need to add some references on the optimisation of AlGaN layer growth. There are some parameters that affect the quality of AlGaN layer in particular the percentage of aluminium. They can add this reference (Characterization of low Al content AlxGa1-xN epitaxial films grown by atmospheric-pressure MOVPE. DOI: 10.1002/pssa.201127529). On the other hand, they need to add reference for the thickness effect on the morphological and electrical properties of AlGaN layer due to the change of the dislocation density. They can use this new reference (Thickness-dependent physical and nanomechanical properties of AlxGa1−xN thin films. DOI: 10.1016/j.mssp.2022.107023).
4/ In the end of the introduction, the authors should put the aim and the novelty of this work without filling this paragraph with results.
5/ The experimental section need to be improved by adding the used growth process of these structures (if the growth was by MOCVD or MBE…), the pressure, the precursors and gases, the growth temperature of each layer (especially to understand how to distinct between the elaboration of 2 micro-meters of GaN and 300 nano-meters of GaN)…. Because the growth conditions are fundamental. The authors should also specify the nature of the used substrates; for Si : is it n or p type and for sapphire: is it orientated along a or c axis?.
6/ The authors should specify the choice of 25% of aluminium (is it optimised after a systematic study?). In addition, the authors should explain the thickness choice of 300nm of GaN channel and 23nm of AlGaN layer. How to measure these thicknesses?
7/ The authors should give the electrical properties of each layer (at least their electron mobility and density).
8/ In line 217, the authors claimed that “Fig. 6(a) that the GaN epilayer grown on the Si substrate exhibits a large number of surface pits”, but in the figure caption they put “Figure 6. Optical microscope images of AlGaN/GaN surface grown on (a) Si and (b) sapphire” in line 222. It is not clear what kind of layer they characterize. The same problem is found with table caption in line 218 (Table 1. Summary the parameters of AlGaN/GaN grown on different substrates) but the authors put in the table the FWHM values of GaN and not for AlGaN/GaN structure! The authors should clarify this issue specially they are already put this information in the abstract.
9/ The authors should add the HR-XRD figure (2θ-ω scan mode) and AFM images (in tapping mode to show clearly the pits) of AlGaN/GaN structure.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
The authors investigated the performance of AlGaN/GaN high-electron-mobility transistor (HEMT) devices fabricated on Si and sapphire substrates in this MS. The effect of SiO2 surface passivation on the electrical characteristics of AlGaN/GAN HEMTs grown on Si and sapphire substrates was explored and compared with those of unpassivated HEMT. The research is design appropriate and the structure is relative complete. It should be accepted after a minor revise.
1. The high-resolution X-ray diffraction patterns should be presented to visualize the quality of the crystal structure.
2. Relevant literature for the most recent year needs to be presented.
3. Figure 2(b), the output (IDS-VDS) characteristics of AlGaN/GaN HEMTs grown on Si substrate at VGS = 0 V was absent.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The author made an extensive revision based on the previous critic. The reviwer strongly suggest the authors not to take a shortcut while considering scientific publication in the future.
Reviewer 2 Report
The manuscript has been improved in the new version and can be published in the journal.
Reviewer 3 Report
The authors corrected and improved the manuscript as recommended by the reviewer. This work could be considered for publication in Electronics.
