You are currently viewing a new version of our website. To view the old version click .
MDPIMDPI
Search all articles
by
  • Yuanda Li1,
  • Jinshun Bi1,* and
  • Xuefei Liu1,*
  • et al.

Reviewer 1: Maurizio Millesimo Reviewer 2: Anonymous Reviewer 3: Khurram Karim Qureshi

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors present a TCAD-based investigation of how the self-heating effect influences single-event transient responses in triple-layer stacked nanosheet GAA-FETs. The study explores many aspects within a single framework, which makes the work worthy of publication.

Here are a few comments/suggestions:

  1. Some temperature-assisted SET effects have already been reported in literature for such technology. This reviewer suggests the author to extend the references and to highlight in a more explicit manner what aspects of the present work are fundamentally new compared with the existing literature.
  2. Can the discussion of proposed physical mechanisms be improved by extending the discussion with TCAD simulated electric field?
  3. The chosen heating/cooling intervals appear arbitrary. How do these time constants relate to realistic operational conditions? Are they chosen just for the analysis to better investigate the phenomena?

Author Response

We thank the referee for his/her comments on this paper. They are addressed in the following. The changes in the text are highlighted in red.

Comment 1: Some temperature-assisted SET effects have already been reported in literature for such technology. This reviewer suggests the author to extend the references and to highlight in a more explicit manner what aspects of the present work are fundamentally new compared with the existing literature.

Response 1: We sincerely thank the reviewer for the valuable feedback. We fully agree with your perspective and have further expanded the relevant literature citations in the revised manuscript. While existing studies on temperature-assisted single-event transients primarily focus on the influence of uniform ambient temperature on device behavior, our work distinguishes itself by examining the effect of localized non-uniform temperature distribution—induced by self-heating during device operation—on charge collection and transport mechanisms in a gate-all-around (GAA) device. To the best of our knowledge, this phenomenon has been reported in planar LDMOS power devices but has not yet been investigated in three-dimensional GAA architectures. Furthermore, our study employs simulations at a finer temporal scale to capture the device’s internal temperature dynamics of the device, providing deeper insight into the transient process.

These points have been clarified and elaborated in Lines 63–79 of the revised manuscript. We genuinely appreciate your insightful comments, which have helped us better highlight the novelty and significance of our work. Should you have any further suggestions, we are very willing to refine our discussion accordingly.

Comment 2: Can the discussion of proposed physical mechanisms be improved by extending the discussion with TCAD simulated electric field?

Response 2: We fully agree that incorporating the electric field analysis strengthens the discussion of the physical mechanisms.

In the revised manuscript, we have added Figure 13, which compares the internal electric field distributions before, during, and after heavy-ion incidence for devices with and without SHE. Based on this simulation, we extended the discussion in Section 3.3. The evolution of electric field reveals that the elevated lattice temperature induced by SHE degrades carrier mobility. This degradation slows down carrier transport, prolongs the residence time of excess carriers, and consequently delays the recovery of the electric field. This mechanism provides a solid physical explanation for the increased charge collection observed in the SHE case.

Comments 3: The chosen heating/cooling intervals appear arbitrary. How do these time constants relate to realistic operational conditions? Are they chosen just for the analysis to better investigate the phenomena?

Response 3: You are correct that the time scale selected in our work is on the order of picoseconds, which is indeed shorter than that of actual operational conditions. This choice was made in line with similar simulation-based studies in the field, particularly as referenced in References [25] and [26], where ps-scale analysis is employed to capture fast transient phenomena. Our primary objective was to investigate the device’s internal temperature evolution of the device in response to varying gate voltages and over time. A finer temporal resolution allows us to observe the dynamic thermal behavior with greater clarity, which is essential for understanding the underlying mechanisms of self-heating and its interaction with single-event effects.

We acknowledge that this simplification differs from real-word operation, and we have explicitly addressed this limitation in the Conclusion section to ensure transparency. We sincerely thank you for highlighting this aspect, as it encourages us to contextualize our methodology and findings better. Your feedback has helped us further clarify the scope and assumptions of our study in the revised manuscript.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This work uses TCAD simulations to study self-heating and single-event transient effects in triple-layer stacked nanosheets, showing that self-heating both degrades electrical performance and significantly increases susceptibility to heavy-ion strikes This causes amplifying transient currents and collected charge, especially with longer heating times and partially mitigated by longer cooling times. Although the paper provides some useful information, it lacks many essential details. Therefore, a major revision is required before the manuscript can be considered for further processing. My main comments are as follows:

  1. The abstract should be rewritten for better clarity. It must be concise, start with a clear problem statement, and accurately reflect the content of the paper. Symbols such as Ion (and similar quantities) should be written with proper subscripts. The abstract should avoid listing too many detailed quantitative results.
  2. The reference list needs to be updated. In the current version, only one reference from 2024 is cited and there are no references from 2025, which suggests that recent developments in the field have not been adequately considered.
  3. The authors state that the device was “fabricated in this work,” whereas the study is purely based on simulations. This wording must be corrected throughout the manuscript.
  4. The device structure reported in Ref. [16], which is used for model calibration, should be clearly highlighted. More details about this reference device (geometry, doping, bias conditions, etc.) must be provided.
  5. The data in Table 1 require more discussion. The authors should explain their physical significance.
  6. The simulation methodology needs to be described in more detail. In particular, the temperature-dependent models and equations employed in the simulations should be explicitly given and properly referenced.
  7. The sentence “During carrier equation calculations, the keyword ‘Temperature’ was coupled into the carrier continuity equation to compute lattice temperature” is unclear and should be rewritten.
  8. In Figure 3, the bias is labeled as Vg = Vd = 0.6 V; this should be checked and corrected. Moreover, the authors report that Ion is degraded by 2.11%, while the on/off ratio is degraded by only 1.89%. This implies that the off current decreases with increasing temperature, which contradicts the typical behavior where off current usually increases with temperature. The authors must comment on this critical issue and verify whether SRH recombination and other leakage mechanisms have been modeled correctly.
  9. Figure 12 appears to be plotted incorrectly, and the actual intended Fig. 12 seems to be missing. This should be fixed.
  10. The SET simulations are performed using only one LET value (10 MeV·cm2/mg), one strike radius (10 nm), and normal incidence. This very limited parameter space restricts the generality of the conclusions, and the authors should at least discuss this limitation.
  11. Many key parameters, such as thermal contact resistance, oxide and substrate thermal conductivities, carrier lifetimes, and the heavy-ion track profile, are assumed but not varied. As a result, there is no assessment of parameter uncertainties or process variations and their impact on the results.
  12. The conclusion should be strengthened by clearly summarizing the most important quantitative findings of the work. The limitations of the work should also be addressed.
Comments on the Quality of English Language

The manuscript contains numerous grammatical and typographical errors and should be thoroughly revised for language quality.

Author Response

We thank the referee for his/her comments to improve this paper. They are addressed in the following. The changes in the text are highlighted in blue.

Comments 1: The abstract should be rewritten for better clarity. It must be concise, start with a clear problem statement, and accurately reflect the content of the paper. Symbols such as Ion (and similar quantities) should be written with proper subscripts. The abstract should avoid listing too many detailed quantitative results.

Response 1: Thank you for your valuable suggestion. We have carefully incorporated the recommended changes into the revised manuscript, and the modifications have been highlighted in blue for your convenience. These revisions have strengthened the relevant section and addressed your comment appropriately.

Comment 2: The reference list needs to be updated. In the current version, only one reference from 2024 is cited and there are no references from 2025, which suggests that recent developments in the field have not been adequately considered.

Response 2: We have carefully updated the reference list to reflect the state of the art. We have added several recent references regarding thermal effects on semiconductor devices [14,15]. It is worth noting that most of these recent studies focus on uniform ambient-temperature environments. Our work differs significantly by focusing on the internal self-heating effects and local temperature variations within the device at a much finer time scale. We have clarified this distinction in the Introduction section to better position our contribution within the context of recent developments.

 

Comment 3: The authors state that the device was “fabricated in this work,” whereas the study is purely based on simulations. This wording must be corrected throughout the manuscript.

Response 3: We have revised the text accordingly. The changes have been highlighted in blue for your convenience.

 

Comment 4: The device structure reported in Ref. [16], which is used for model calibration, should be clearly highlighted. More details about this reference device (geometry, doping, bias conditions, etc.) must be provided.

Response 4: We have revised the text accordingly. The changes have been highlighted in blue for your convenience with reference numbers updated to [17,19].

Comment 5：The data in Table 1 require more discussion. The authors should explain their physical significance.

Response 5: A deeper discussion regarding the physical significance of the data in Table 1 was necessary. The relevant discussion has been added to the revised manuscript, which is highlighted in blue.

“The geometric parameters are calibrated according to the IRDS roadmap, dnsuring the simulation reflects realistic aggressively scaled devices.”

“The channel doping is kept low to minimize impurity scattering and enhance carrier mobility, while the source/drain regions are heavily doped to reduce parasitic contact resistance.”

“A thermal boundary resistance of 2×10⁻6 cm²K/W was applied to the contact electrodes, and a value of 2×10-4cm2K/W was specified for the interface between the silicon and SiO2 layers [17. The high resistance at the oxide/semiconductor interface represents the critical heat dissipation bottleneck.”

Comment 6：The simulation methodology needs to be described in more detail. In particular, the temperature-dependent models and equations employed in the simulations should be explicitly given and properly referenced.

Response 6: In the revised manuscript, we have significantly expanded the description of the simulation methodology and workflow. As requested, we have explicitly provided the governing equations and the temperature-dependent models used in the study, along with the corresponding references.

Furthermore, to better illustrate the simulation process, we have optimized Figures 1 and 2. These changes can be found in the Section 2 (Lines 107-130), and the modifications are marked in blue.

 

Comment 7：The sentence “During carrier equation calculations, the keyword ‘Temperature’ was coupled into the carrier continuity equation to compute lattice temperature” is unclear and should be rewritten.

Response 7: We have significantly elaborated the descriptions in Section 2 by incorporating relevant mathematical formulas and references to substantiate our approach.

Comment 8：In Figure 3, the bias is labeled as Vg = Vd = 0.6 V; this should be checked and corrected. Moreover, the authors report that Ion is degraded by 2.11%, while the on/off ratio is degraded by only 1.89%. This implies that the off current decreases with increasing temperature, which contradicts the typical behavior where off current usually increases with temperature. The authors must comment on this critical issue and verify whether SRH recombination and other leakage mechanisms have been modeled correctly.

Response 8: We sincerely apologize for the oversight regarding the bias labels in Figure 3 and, more importantly, for the inconsistency in our previous electrical data. We fully agree that the earlier results, implying a decrease in off-current with temperature, contradicted fundamental physical principles. Taking this comment very seriously, we have corrected the bias labels in Figure 3(Vg=Vd=0.65V) and performed a rigorous re-simulation of the device characteristics to ensure that all leakage mechanisms, including SRH recombination/generation, are correctly modeled. The updated simulation results have resolved the contradiction. As shown in the revised Figure 3(a), the Self-Heating Effect (SHE) degrades the on-state current (Ion) by 1.81% (from 114.383μA to 112.303μA). Crucially, the on/off current ratio degrades by approximately 1.82% (a larger degradation than Ion). Mathematically and physically, this confirms that the off-state current (Ioff) indeed increases under self-heating conditions. Our specific data shows that Ioff rises from 4386.95 pA to 4386.98 pA. Although the absolute increase is slight, the trend is now fully consistent with the mechanism of thermally induced carrier generation. We have added a detailed explanation of this phenomenon based on the SRH generation rates (Figure 3b) in the revised manuscript (please see Lines 149-159).

Comment 9：Figure 12 appears to be plotted incorrectly, and the actual intended Fig. 12 seems to be missing. This should be fixed.

Response 9: We would like to apologize for the omission. Due to an error on our part, the figure was not uploaded previously. We have now corrected this and inserted it in the appropriate section.

Comment 10：The SET simulations are performed using only one LET value (10 MeV·cm2/mg), one strike radius (10 nm), and normal incidence. This very limited parameter space restricts the generality of the conclusions, and the authors should at least discuss this limitation.

Response 10: We acknowledge that the current simulations are performed under specific parameters settings. The primary goal of this work was to provide a proof-of-concept for the proposed design. The selected parameters (LET=10 MeV·cm²/mg, normal incidence) serve as a representative standard to demonstrate the device's capabilities. Exploring the whole parameter space would require significantly more computational resources and is planned as a dedicated follow-up study.

Comment 11：Many key parameters, such as thermal contact resistance, oxide and substrate thermal conductivities, carrier lifetimes, and the heavy-ion track profile, are assumed but not varied. As a result, there is no assessment of parameter uncertainties or process variations and their impact on the results.

Response11: We agree that a comprehensive sensitivity analysis would add value to the study. However, the primary focus of this work is to elucidate the fundamental physical mechanism underlying SHE’s impact on SET in the proposed structure. The parameters used in our simulations were selected based on references reported in standard literature to ensure the baseline is representative. Due to the high computational cost of 3D-TCAD simulations, performing a full statistical variation analysis is beyond the scope of the current manuscript. Therefore, we plan to conduct a systematic study on parameter sensitivity and process variations in our future work.

Comment 12：The conclusion should be strengthened by clearly summarizing the most important quantitative findings of the work. The limitations of the work should also be addressed.

Response 12: We fully agree that discussing the practical limitations of the proposed design in the conclusion section is highly important. In response, we have added a paragraph in the revised conclusion to address the limitations of this study explicitly.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This paper investigates the interaction between self-heating effects (SHE) and single-event transient (SET) responses in a triple-layer stacked nanosheet GAA-FETs using extensive TCAD simulations. The authors study the DC, AC, and radiation-driven transient behavior and find that self-heating greatly increases the device’s lattice temperature, affects the SET peak current, and boosts the collected charge during heavy-ion events. I have the following comments that should be incorporated by the authors’.

  1. The authors should cite the most recent papers in the area and also highlight their main contributions in this research.
  2. The manuscript relies entirely on TCAD simulations, yet, no experimental validation is provided for, self-heating temperature rise (>170 K), AC-mode heating dynamics, SET current/charge magnitudes to name a few.
  3. The reported lattice temperature rises (e.g., 472–485 K in Fig. 4 & Fig. 9), which appear physically questionable for nanosheet transistors, the authors should comment on this and provide justification.
  4. The study examines only one specific stacked nanosheet geometry (5 nm thickness, 15 nm width, 12 nm gate length); the results cannot be generalized without exploring structural variations.
  5. The AC heating was performed with, 2 ps rise time, which is not realistic in actual systems.
  6. It is seen that some of the writing is repetitive, especially when explaining heating and cooling effects.
  7. The authors should comment on the practical limitations of their proposed design in the conclusion section.
  8. There are a few typos in the paper to be fixed during the revision.
Comments on the Quality of English Language

There are a few typos in the paper to be fixed during the revision.

Author Response

We thank the referee for his/her comments to improve this paper. They are addressed in the following. The changes in the text are highlighted in purple.

 

Comment 1: The authors should cite the most recent papers in the area and also highlight their main contributions in this research.

Response 1：Thank you for the comment. We have updated the reference list to include the most recent studies in this field. In the revised manuscript, we have detailed how these references provide the theoretical foundation and data verification for our approach. Furthermore, as suggested, we have explicitly highlighted the main contributions of the work in the Section 3 to clearly distinguish our research from the cited literature.

Comment 2: The manuscript relies entirely on TCAD simulations, yet, no experimental validation is provided for, self-heating temperature rise (>170 K), AC-mode heating dynamics, SET current/charge magnitudes to name a few.

Response 2：We acknowledge that the lack of experimental validation is a limitation of the current study. Due to the limitations of our current experimental facilities, we were unable to perform the specific measurements for self-heating and SET dynamics presented in the manuscript.

We have addressed the limitations of our study in the Conclusion section.

Comment 3: The reported lattice temperature rises (e.g., 472–485 K in Fig. 4 & Fig. 9), which appear physically questionable for nanosheet transistors, the authors should comment on this and provide justification.

Response 3：We have verified our results against existing studies to ensure their validity. The observed lattice temperatures (472–485 K) are in good agreement with similar devices reported in the literature. Specifically, Ref. [27] demonstrates a temperature rise up to 485 K, and Ref. [28] reports 484 K under similar operating conditions. This confirms that our simulation data falls within a feasible and expected range for this technology node. We have cited these references in the revised manuscript (line 219) to support our findings.

Comment 4: The study examines only one specific stacked nanosheet geometry (5 nm thickness, 15 nm width, 12 nm gate length); the results cannot be generalized without exploring structural variations.

Response 4：The dimensions selected for this study were specifically chosen to represent the aggressive scaling limits of current state-of-the-art GAA technology. The primary objective of this work is to elucidate the underlying coupling mechanism between the self-heating effect (SHE) and single-event transient (SET) in these aggressively scaled devices. By fixing the geometry to this benchmark, we establish a controlled baseline to isolate and analyze the impact of SHE on SET, avoiding the confounding factors introduced by geometric variations. While we agree that exploring structural variations is valuable, the fundamental mechanism revealed in this work—specifically the impact of thermal confinement on SET—is intrinsic to the Gate-All-Around (GAA) architecture. The systematic study on geometric scaling is planned for our future work.

Comment 5: The AC heating was performed with, 2 ps rise time, which is not realistic in actual systems.

Response 5: We chose the 2 ps rise time to simulate the theoretical limit of the device's switching speed and to evaluate the thermal impact under extreme transient stress.

This setting is supported by existing literature, such as Ref. [31] (0.25 ps) and Ref. [18] (5 ns). By selecting 2 ps, we aimed to provide an assessment that covers the fast-transient regime, which is critical for understanding reliability risks.

Comment 6: It is seen that some of the writing is repetitive, especially when explaining heating and cooling effects.

Response 6: To address this, we have carefully revised the relevant sections by removing redundant statements and consolidating the discussion to improve clarity and conciseness.

Comment 7: The authors should comment on the practical limitations of their proposed design in the conclusion section.

Response 7：In response, we have added a paragraph in the revised conclusion to explicitly address the limitations of this study. These primarily include the gap between the idealized assumptions of the simulation models and real-world conditions, as well as practical experimental constraints such as process variations and the metrological bounds of instrumentation accuracy.

Comment 8: There are a few typos in the paper to be fixed during the revision.

Response 8: We had carefully checked through whole manuscript and corrected all typos and grammatical errors.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have addressed most of the issues raised from the previous review. However, some points need to be addressed before accepting this paper:

  1. More quantitative results should be included in the conclusion.
  2. Most of figures need to be modified for better clarity. For instance:

Figure 2(b): all symbols should be written in a scientific way (e.g. Id should be I_d (as subscript), Vd, Vth, Vg, .... etc.). the same holds for Figure 3(a).

Figure 6: fonts can be enlarged for better reading values. The same holds for Figure 7, 9 10(b), 11, 12, and 15.

3. grammar should be checked

 

 

 

 

Comments on the Quality of English Language

The manuscript contains numerous grammatical and typographical errors and should be thoroughly revised for language quality.

Author Response

We thank the referee for his/her comments on this paper. They are addressed in the following. The changes in the text are highlighted in red.

Comment 1: More quantitative results should be included in the conclusion.

Response 1: We sincerely thank the reviewer for the valuable feedback. In the revised Conclusion section, we have incorporated additional quantitative results to support our findings. Specifically, we state that the device reaches thermal equilibrium at approximately 150 ps. The maximum lattice temperatures are observed to be 472.35 K in DC mode and 484.56 K in AC mode. Furthermore, compared to the scenario without self-heating effects (SHE), the presence of SHE increases the transient current by approximately 7.9% and the collected charge by about 11.9%.

Comment 2: Most of figures need to be modified for better clarity. For instance:

Figure 2(b): all symbols should be written in a scientific way (e.g. Id should be I_d (as subscript), Vd, Vth, Vg, .... etc.). the same holds for Figure 3(a).

Figure 6: fonts can be enlarged for better reading values. The same holds for Figure 7, 9 10(b), 11, 12, and 15.

Response 2: We have corrected the notation symbols in Figures 2(b) and 3(a) to use proper subscripts. Furthermore, we have increased the font size in Figures 6, 7, 9, 10(b), 11, 12, and 15 (and others where applicable) to improve the clarity of the values.

Comments 3: grammar should be checked

Response 3: We have conducted a thorough grammatical check of the whole manuscript to corrected errors and typos. All changes are marked in red in the revised manuscript.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

I am satisfied with the revised manuscript. 

Author Response

We thank the referee for his/her comments on this paper. They are addressed in the following. The changes in the text are highlighted in red.

Thank you very much for your valuable comments. We have carefully proofread the manuscript to correct grammatical errors and improved the quality of the figures throughout the text. Additionally, we have incorporated quantitative results into the Conclusion section (Lines 362 and 366) to better support our findings. All changes are marked in red in the revised manuscript.

Author Response File: Author Response.pdf