Numerical Investigation of the Effects of the Beam Scanning Pattern and Overlap on the Temperature Distribution during the Laser Dopant Activation Anneal Process
Round 1
Reviewer 1 Report
In this work, the authors present a numerical investigation related to the effects of the beam scanning pattern and overlap on the temperature distribution during the laser dopant activation anneal process. The research appears to be efficiently done and appropriately reported; however, the standard of English must be improved. Nevertheless, there are some questions and corrections that must be answered to improve and complete the document.
Lines 6 and 8. Please, correct “1Departament…” and “2Departament…”
Lines 15 and 20. Please correct “pro-cess…” and “how-ever …” to “process…” and “however…”
Introduction section: In this section, the authors don’t indicate the novelty of their work. what is the innovation of your work when compared with the other researchers? The "Knowledge gap to be filled"? In this introduction, the authors must describe or indicate the work that will be done to test their "hypothesis". On the other hand, the most of references are not updated.
Line 114: The authors claim that they implemented a #numerical simulation. However, this sentence is very incomplete. Did you use commercial numerical simulations software, or did you develop the numerical codes on your own?
If you used commercial software, which one? If you developed your own code, which platform did you use?
Lines 185, 299, 309. The authors must delete the “tab” before the sentence.
Figure 3. I think Figures 3a and 3b represented zig-zag and 3c and 3d are bidirectional. However, the schematic, presented on the top right, is the same (zig-zag)
Lines 288 until 311. I think that this information is not about the “Results and discussion”. In my opinion, this part of the text must be written in the previous section (Development of a numerical model).
This work is a very interesting numerical approach; however, I would like to know if you have done any experimental validation or if you are thinking to implement it in future works. Have you any idea about your experimental approach? Which experimental techniques are more adequate for your validations? What measurement methods are you planning to use? I think that you must discuss these points in your conclusions.
References: Most references are not very up to date. It would be interesting if you could include a larger number of more up-to-date references.
Author Response
To editor and reviewers :
We really appreciate your valuable comments for this paper. I have finished revision and also made responses for corresponding comments from reviewers. The detail contents of the revision and response are described below.
Reviewer 1
In this work, the authors present a numerical investigation related to the effects of the beam scanning pattern and overlap on the temperature distribution during the laser dopant activation anneal process. The research appears to be efficiently done and appropriately reported; however, the standard of English must be improved. Nevertheless, there are some questions and corrections that must be answered to improve and complete the document.
1) Lines 6 and 8. Please, correct “1Departament…” and “2Departament…”
Ans) “1” & “2” are removed from lines 6 and 8
2) Lines 15 and 20. Please correct “pro-cess…” and “how-ever …” to “process…” and “however…”
Ans) These words are changed as commented above
3) Introduction section: In this section, the authors don’t indicate the novelty of their work. what is the innovation of your work when compared with the other researchers? The "Knowledge gap to be filled"? In this introduction, the authors must describe or indicate the work that will be done to test their "hypothesis". On the other hand, the most of references are not updated.
Ans) The novelty of this study is described in the last paragraph of the introduction as shown below.
“However, in most studies, simulations of the LTA process were carried out under very limited condition (a fixed beam position or a single pulse), with the results not providing enough information about the phenomena occurring during the actual annealing process. In actual applications of the LTA process, the scanning pattern of the laser beam and the related beam overlap are very important to achieve a uniform temperature and proper dopant activation throughout the wafer area. The purpose of this study is to investigate the temperature distributions created by different laser scan paths and beam overlaps.”
4) Line 114: The authors claim that they implemented a #numerical simulation. However, this sentence is very incomplete. Did you use commercial numerical simulations software, or did you develop the numerical codes on your own?
Ans) We used commercial CAE software, called COMSOL multiphysics, for simulation as described in the last paragraph of the introduction part. We revised the sentence (lines 108 & 114) above to clearly give information.
5) If you used commercial software, which one? If you developed your own code, which platform did you use?
Ans) Please see the answer of comment No. 4)
6) Lines 185, 299, 309. The authors must delete the “tab” before the sentence.
Ans) Tab before the sentence (lines 185, 299 and 309) are all deleted.
7) Figure 3. I think Figures 3a and 3b represented zig-zag and 3c and 3d are bidirectional. However, the schematic, presented on the top right, is the same (zig-zag)
Ans) Yes, we made a mistake for schematic presentation. We have changed it correctly.
8) Lines 288 until 311. I think that this information is not about the “Results and discussion”. In my opinion, this part of the text must be written in the previous section (Development of a numerical model).
Ans) We have moved the text from line 288 to line 311 to “development of a numerical model” section as suggested above. For this, we add a new subsection (section 2.6.), and we make minor changes (please see corrected parts in the section 2.6 of the revised manuscript) in the text (to avoid awkward insertion of the text)
9) This work is a very interesting numerical approach; however, I would like to know if you have done any experimental validation or if you are thinking to implement it in future works. Have you any idea about your experimental approach? Which experimental techniques are more adequate for your validations? What measurement methods are you planning to use? I think that you must discuss these points in your conclusions.
Ans) We have a plan to do experimental study for this area in the near future. We are going to use experimentally measured dopant concentration distribution data (obtained by SIMS analysis) for validation of the simulation results. (As we mentioned in our paper, direct measurement of the temperature distribution is very difficult due to the small beam spot and high speed of the process). We newly add the discussion for this part to conclusions.
10) References: Most references are not very up to date. It would be interesting if you could include a larger number of more up-to-date references.
Ans) We add 5 up-to-date references as suggested above (No. of newly updated references.: 3, 5, 7, 13 and 14).
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Interesting work. The modelling work is rather well implemented and is of interest for the community. Major revisions are in order for the authors to address the comments detailed as seen below:
Language needs to be revised by a native speaker. Several minor errors found. Please revise.
“On the other hand, the heat in LTA is only confined to local and shallow surface regions of the wafer; hence, there is no warpage of the wafer when using this process.”: this is related to the characteristics of the laser radiation as it occurs in order laser-based processes. See for example 10.1016/j.scriptamat.2021.114219 and 10.1016/j.matdes.2021.109533 and further complement the introduction.
“e LSA process produced better electrical performance 68 (i.e., a larger drain current and lower sheet resistance) of a CMOS device compared to 69 RTP”: why? Clarify.
“measurements of the temperature history in the LTA process are usually very difficult 80 due to the short anneal time and small laser spot.”: There are thermal models that can applied for such goals.
“computational domain is 7.5 mm × 2.7 mm × 0.75 mm (length”: why these dimensions? Justify.
“It has been reported that a ??2 laser with a long wavelength (l = 10.6 mm) can reduce 170 the pattern effect and produce a relatively uniform temperature distribution within a chip”: why? Justify and compare to fiber laser for example.
“quickly stabilizes in a few milliseconds, and it mostly does not change until the first scan- 204 ning track of the laser beam ends”: was this expected? I would be expected an increase along the laser travel because of conduction of heat.
Fig 5 the color scale axis is correct? Is the minimum temperature 40 ºC? Why not room temperature?
“Severe diffusion in the junction area can induce the detrimental short chan”: how much is severe?
Do the authors have any experimental data that can validate the modelling work? The modelling works is very nice but some validation would interesting to be presented.
Author Response
To editor and reviewers :
We really appreciate your valuable comments for this paper. I have finished revision and also made responses for corresponding comments from reviewers. The detail contents of the revision and response are described below.
Reviewer 2
Interesting work. The modelling work is rather well implemented and is of interest for the community. Major revisions are in order for the authors to address the comments detailed as seen below:
1) Language needs to be revised by a native speaker. Several minor errors found. Please revise.
Ans) Language revision was already made by a native speaker before submission. We carefully check a whole manuscript again, and we find several errors and revise them (i.e., misused hyphens, spacing and indentation for the 1st line of the paragraph).
2) “On the other hand, the heat in LTA is only confined to local and shallow surface regions of the wafer; hence, there is no warpage of the wafer when using this process.”: this is related to the characteristics of the laser radiation as it occurs in order laser-based processes. See for example 10.1016/j.scriptamat.2021.114219 and 10.1016/j.matdes.2021.109533 and further complement the introduction.
Ans) Additional description for the characteristics of laser radiation has been inserted below the sentence mentioned above. And the references suggested above are added to the reference list.
3) “e LSA process produced better electrical performance 68 (i.e., a larger drain current and lower sheet resistance) of a CMOS device compared to 69 RTP”: why? Clarify.
Ans) LSA produced better electrical performance than RTP owing to the higher heating temperature and shorter anneal time. The reason for better electrical performance in the case of the LSA is additionally described in the revised manuscript.
4) “measurements of the temperature history in the LTA process are usually very difficult 80 due to the short anneal time and small laser spot.”: There are thermal models that can applied for such goals.
Ans) Some pyrometer tools using high response speed diode could be used for the measurement of temperature at the interaction zone between a laser beam and workpiece. However, this only give average temperature of the interaction spot and still can not show the data for detail temperature distribution within or around the spot. Simulation based on relevant thermal model can give estimated temperature distribution as you commented.
5) “computational domain is 7.5 mm × 2.7 mm × 0.75 mm (length”: why these dimensions? Justify.
Ans) Through preliminary numerical tests, we found that the temperature change at the boundary of the computational domain (7.5 mm (x-dir) × 2.7 mm (y-dir)) was negligible. 0.75 mm depth corresponds to the thickness of real commercial wafer. We imitated this. Additional explanation for the computational domain is inserted in the revised manuscript.
6) “It has been reported that a ??2 laser with a long wavelength (l = 10.6 mm) can reduce 170 the pattern effect and produce a relatively uniform temperature distribution within a chip”: why? Justify and compare to fiber laser for example.
Ans) Since the photon energy of the long wavelength beam is usually much lower than the band gap energy of patterned thin layers (i.e., multilayer structure composed of Si, dielectric layers and etc.), the absorption of the CO2 laser beam mostly occurs in the Si substrate below the patterned device. And heat is transferred to the device (of patterned multilayer structure) area through heat diffusion (or conduction). This mechanism improves the temperature uniformity within a chip. However, if the photon energy of a laser beam (laser beam of a shorter wavelength, such as a fiber (1.07 um wavelength), green or UV laser) is higher than the band gap energy of thin layers, there is more possibility of the direct absorption of the laser beam energy by these layers. Under this situation, the degree of direct absorption mainly depends on the band gap energy and thickness of layers. This induces a non-uniformity of the temperature within a chip (The degree of the non-uniformity usually increases as the wavelength of the beam decreases). Additional explanation for reducing the pattern effect in the case of the CO2 laser beam is inserted in the revised manuscript.
7) “quickly stabilizes in a few milliseconds, and it mostly does not change until the first scan- 204 ning track of the laser beam ends”: was this expected? I would be expected an increase along the laser travel because of conduction of heat.
Ans) Stabilization of the temperature is achieved through the thermal equilibrium between laser energy input to wafer and energy loss (convective and radiative heat transfer loss) from the wafer. If the temperature keeps increasing along the beam scan path, then uniform temperature over the whole wafer can not be obtained (this means that laser annealing process can not be used for the real mass production, however LTA process is being used by chip makers).
8) Fig 5 the color scale axis is correct? Is the minimum temperature 40 ºC? Why not room temperature?
Ans) Yes, it is correct. As described in section 2.3., wafer preheating of was assumed in the simulation. Absorption rate of a laser beam in material (Si) can be highly increased by increasing the temperature of the material (efficient use of the laser energy is possible). Owing to this, in the real semiconductor industry, the wafer is usually preheated with a heater (or a chuck) placed below the wafer during the laser annealing process. In this study, we imitated this wafer preheating condition. The reason for preheating is explained in the revised manuscript.
9) “Severe diffusion in the junction area can induce the detrimental short chan”: how much is severe?
Ans) Here, the sever diffusion of dopant means a certain level of diffusion that electrically connects a source with a drain in a transistor even under the off state (no bias violated is applied to the gate of the transistor) of the transistor (this condition usually induces increase of the leakage current of the transistor)
10) Do the authors have any experimental data that can validate the modelling work? The modelling works is very nice but some validation would interesting to be presented.
Ans) We have a plan to do experimental study for this area in the near future. We are going to use experimentally measured dopant concentration distribution data (obtained by SIMS analysis) for validation of the simulation results. (As we mentioned in our paper, direct measurement of the temperature distribution is very difficult due to the small beam spot and high speed of the process). We newly add the discussion for this part to conclusions.
Author Response File: Author Response.pdf
Reviewer 3 Report
Dear authors,
when I received your article for review, I was looking forward to assessing the effect of the laser beam on the anneal process and dopant activation based on the scanning pattern and overlap size. Unfortunately a reality was diferent.
You state right at the beginningIn of introduction „In a shallow junction, an abrupt doping profile which can provide low resistance is crucial to improve the electrical performance of the device. In general, activation of the dopant increases with the temperature and time of the activation anneal process. On the other hand, this enhances the diffusion of the dopant, which in turn can restrict the formation of the abrupt doping profile and induce a detrimental short-channel effect in the device.” I would therefore expect that you will optimize (with help of numerical simulations) an anneal process and for the final solution you will verify the process. However, I was disappointed that you only used informations from published articles to define the numerical simulation input data, but your personaly contribution in this area is highly debatable.
Below are a few points where you did not explain why the values given by you were used in the work, or what is the benefit, or how it is possible to use the knowledge gained from the simulation.
1) You used simplistic boundary conditions, it can be understood. But nowhere you explained, for example, why you used preheating 400°C or what would happen without the using of preheating, or at a different value of preheating.
2) You performed a simulation of laser annealing with a zig-zag pattern and a bidirectional pattern with 0% and 50% overlap, but nowhere did you state which process is more advantageous from the point of view of the abrupt doping profile.
3) From the point of view of possible verification of controlled temperature profiles, you state that „The temporal and spatial profiles of the temperature at the annealed area during the LTA process are the key factors determining the degrees of dopant activation and diffusion. However, precise measurements of the temperature history in the LTA process are usually very difficult due to the short anneal time and small laser spot.“ Whether the temperature profiles correspond with reality can be verified indirectly. In your case can be used EDX analysis (SEM microscope) of the diffusion of elements from the dopant.
4) In Chapter 3 you summarized the surface temperature distribution, but I lack a deeper discussion. Similarly, in Chapter 3.3 you describe the dopant concentration distribution, but the result is basically the general statement that with higher temperature, more intense diffusion occurs. However, this is already clear from general equation 10.
5) You take all physical and material quantities used in the simulation from the literature. In your case, the value of activation energy can be determined experimentally. If you will use activation energy from literature, what influence it will have on simulation accuracy for yours boundary conditions and what influence it will have on change in the dopant concentration? You did not discussed it.
6) Conclusion is just a summary of the results, but what knowledge should the reader take away? So what exactly should be done if the LTA process should be used to improve the properties?
For the above reasons, I cannot recommend an article for publication in the Applied sciences journal.
Author Response
To editor and reviewers :
We really appreciate your valuable comments for this paper. I have finished revision and also made responses for corresponding comments from reviewers. The detail contents of the revision and response are described below.
Reviewer 3
Dear authors,
when I received your article for review, I was looking forward to assessing the effect of the laser beam on the anneal process and dopant activation based on the scanning pattern and overlap size. Unfortunately a reality was diferent.
You state right at the beginningIn of introduction „In a shallow junction, an abrupt doping profile which can provide low resistance is crucial to improve the electrical performance of the device. In general, activation of the dopant increases with the temperature and time of the activation anneal process. On the other hand, this enhances the diffusion of the dopant, which in turn can restrict the formation of the abrupt doping profile and induce a detrimental short-channel effect in the device.” I would therefore expect that you will optimize (with help of numerical simulations) an anneal process and for the final solution you will verify the process. However, I was disappointed that you only used informations from published articles to define the numerical simulation input data, but your personaly contribution in this area is highly debatable.
Below are a few points where you did not explain why the values given by you were used in the work, or what is the benefit, or how it is possible to use the knowledge gained from the simulation.
1) You used simplistic boundary conditions, it can be understood. But nowhere you explained, for example, why you used preheating 400°C or what would happen without the using of preheating, or at a different value of preheating.
Ans) As described in section 2.3., wafer preheating of was assumed in the simulation. Absorption rate of a laser beam in material (Si) can be highly increased by increasing the temperature of the material (efficient use of the laser energy is possible). Owing to this, in the real semiconductor industry, the wafer is usually preheated (up to 400 , for stable use of an AlN heater) with a heater (or a chuck) placed below the wafer during the laser annealing process. In this study, we imitated this wafer preheating condition. The reason for preheating is explained in the revised manuscript.
2) You performed a simulation of laser annealing with a zig-zag pattern and a bidirectional pattern with 0% and 50% overlap, but nowhere did you state which process is more advantageous from the point of view of the abrupt doping profile.
Ans) The main objective of this study was to investigate the temperature distribution created by different scan paths and beam overlaps (which combination is more advantageous to produce uniform temperature distribution in annealed region) as described in the last paragraph of the introduction. We included the contents of the shallow junction and abrupt doping profile in introduction part to point out the reason why LTA was introduced for fabrication of the scaled device.
3) From the point of view of possible verification of controlled temperature profiles, you state that „The temporal and spatial profiles of the temperature at the annealed area during the LTA process are the key factors determining the degrees of dopant activation and diffusion. However, precise measurements of the temperature history in the LTA process are usually very difficult due to the short anneal time and small laser spot.“ Whether the temperature profiles correspond with reality can be verified indirectly. In your case can be used EDX analysis (SEM microscope) of the diffusion of elements from the dopant.
Ans) We have a plan to do experimental study for this area in the near future. We are going to use experimentally measured dopant concentration distribution data (obtained by SIMS analysis) for validation of the simulation results. We newly add the discussion for this part to conclusions.
4) In Chapter 3 you summarized the surface temperature distribution, but I lack a deeper discussion. Similarly, in Chapter 3.3 you describe the dopant concentration distribution, but the result is basically the general statement that with higher temperature, more intense diffusion occurs. However, this is already clear from general equation 10.
Ans) I agree with your comment above, but in this study, we want to show the concentration distributions obtained by using the specific temperatures (numerically calucated) at three distinctive spots (normal beam path part, corner part, overlapped part) in multipath laser scan which corresponds to actual laser annealing process. From this study, we could expect how much temperature difference and concentration difference can be induced in the multipath laser annealing.
5) You take all physical and material quantities used in the simulation from the literature. In your case, the value of activation energy can be determined experimentally. If you will use activation energy from literature, what influence it will have on simulation accuracy for yours boundary conditions and what influence it will have on change in the dopant concentration? You did not discussed it.
Ans) In most simulation study, material properties are generally adopted from the literature (they were the ones either experimentally obtained or theoretically obtained). We also adopted them from the literature (academic journals). As you mentioned, the results of simulation highly depend on material properties. For example, if we used a higher thermal conductivity in the simulation, we would get low temperature value (or low dopant diffusion, accordingly the concentration distribution more similar to that of as-implanted case) since the heat can be quickly dissipated through conduction heat transfer.
6) Conclusion is just a summary of the results, but what knowledge should the reader take away? So what exactly should be done if the LTA process should be used to improve the properties?
Ans) The multipath scanning method (scan pattern and beam overlap condition) to produce uniform temperature distribution (within the scanned area) is essential for actual application of the LTA. In conclusion section, we mentioned that the most uniform heating is possible when the zig-zag pattern and the 50% overlap condition are used.
For the above reasons, I cannot recommend an article for publication in the Applied sciences journal.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The second version of manuscript improved significantly when compared with first version. So, in my opinion the manuscript can be accepted for publication.
Reviewer 2 Report
The authors are revised the manuscript properly and acceptance is recommended.
Nice work.
Reviewer 3 Report
Dear authors,
I have checked yours explanation to my questions and remarks. Some of them are valuable and logical. I have to also state that simulation without verification are very weak. At the end I have decide to agree with publishing.
