Commercial P-Channel Power VDMOSFET as X-ray Dosimeterâ€
, Stefan D. Ilić 1,2, Sandra Veljković 1, Aleksandar S. Jevtić 1, Strahinja Dimitrijević 1, Alberto J. Palma 3, Srboljub Stanković 4 and Marko S. Andjelković 5Round 1
Reviewer 1 Report
The authors investigated the possibility of using the commercial p-channel power vertical double-diffused metal-oxide-semiconductor field-effect transistors (VDMOSFETs) as X-ray dosimeters, but there is no comparison to commercially available RADFETs.
In the abstract (lines 15-16) it is written that "there is very little research on the effects of X-rays on semiconductor electronic components". This is a very strong statement and nowhere in the article this is justified.
There are 8 transistors used in the experiment but in the discussion of the results there is nothing explained if the results are from all the transistors or from one of them. This is connected with the other issue. There are no errors for the experimental values obtained. This is especially crucial when displaying the golden ratio results (figure 11) and in figure 8.
English should be corrected. Especially the use of "the" and "a". Among others there are clear errors:
- sentence in the lines 30-31
- the use of word different instead of various (for example in line 32)
- discrepancy was appeared (in lines 22-23)
- is well -> is good (in lines 141 and 245)
- radiation -> irradiation (in lines 169 and 211)
Author Response
Response to Reviewer 1 Comments
The authors investigated the possibility of using the commercial p-channel power vertical double-diffused metal-oxide-semiconductor field-effect transistors (VDMOSFETs) as X-ray dosimeters, but there is no comparison to commercially available RADFETs.
Response: We have not found adequate research, but we plan to do so.
In the abstract (lines 15-16) it is written that "there is very little research on the effects of X-rays on semiconductor electronic components". This is a very strong statement and nowhere in the article this is justified.
Response: We agreed, and deleted it.
There are 8 transistors used in the experiment but in the discussion of the results there is nothing explained if the results are from all the transistors or from one of them. This is connected with the other issue. There are no errors for the experimental values obtained. This is especially crucial when displaying the golden ratio results (figure 11) and in figure 8.
Response: We used one transistor for each polarization (a total of 8 transistors) and therefore errors could not be given. However, the transistors are very well paired and their electrical characteristics matched perfectly before the radiation. The experimental apparatus did not allow us to irradiate more than 8 transistors at the same time, which would be the only thing relevant for comparison.
English should be corrected. Especially the use of "the" and "a". Among others there are clear errors:
- sentence in the lines 30-31
- the use of word different instead of various (for example in line 32)
- discrepancy was appeared (in lines 22-23)
- is well -> is good (in lines 141 and 245)
- radiation -> irradiation (in lines 169 and 211)
Response: We agreed, and corrected it.
Reviewer 2 Report
This manuscript studies the possibility of commercial p-channel power VDMOSFETs as an x-ray sensor. Firstly, the equations of threshold voltage shift ΔVT, area density ΔNft and ΔNst are given. Then, the dependence between sensitivity S and gate voltage VG is verified by experiments for three x-ray beams, and a good fitting effect is obtained. The sensitivity increases first and then decreases with the increase of gate voltage. RQR8 beam obtained the highest sensitivity and ΔNft is more dependent on VG than ΔNst. In addition, the author uses silicon instead of silicon dioxide to verify the relationship between absorbed dose D and energy. There is an error between the experimental results and the theory. Finally, the golden ratio GR is used as a new dose parameter to compare the effects of different transistors or operating conditions. This manuscript still has the following problems, which need to be further improved, and it is suggested to review after revision.
- In this manuscript, there are too many details about the experimental results in the abstract. It is suggested to make appropriate modifications and add concluding sentences.
- In the introduction, the author should give the background knowledge about X-ray and MOSFET, and sort out the content and order of the whole manuscript for better understanding.
- The linear regression r2correlation coefficient mentioned in the third part is greater than 0.99, please give the corresponding experimental proof or explanation.
- Why choose Kair= 50 Gy to study the relationship between ΔNft and gate voltage?
- In Figure 8, it is suggested to supplement the dotted line diagram when VG= 0 to make the verification part more complete.
- Is it reasonable to use silicon instead of silicon dioxide to verify the relationship between absorbed dose Dand X-ray energy in this manuscript?
- The conclusion of the manuscript should be summarized in the order discussed above, so as to make the manuscript more organized so that readers can see the conclusion more intuitively.
Author Response
Response to Reviewer 2 Comments
1. In this manuscript, there are too many details about the experimental results in the abstract. It is suggested to make appropriate modifications and add concluding sentences.
Response: We made it.
2. In the introduction, the author should give the background knowledge about X-ray and MOSFET, and sort out the content and order of the whole manuscript for better understanding.
Response: We went through the whole text.
3. The linear regression r2correlation coefficient mentioned in the third part is greater than 0.99, please give the corresponding experimental proof or explanation.
Response: We did not show the VT = f(Kair) dependence for all used gate polarizations, but only for Vg = 21 V in Figure 2 and for Vg = 0 V in reference [26], because we would have to add more figures, which would be too much for paper.
4. Why choose Kair= 50 Gy to study the relationship between ΔNft and gate voltage?
Response: This was limited by the X-ray machine we used.
5. In Figure 8, it is suggested to supplement the dotted line diagram when VG= 0 to make the verification part more complete.
Response: We added it.
6. Is it reasonable to use silicon instead of silicon dioxide to verify the relationship between absorbed dose Dand X-ray energy in this manuscript?
Response: There are expected to be some differences, but we were forced to do so.
7. The conclusion of the manuscript should be summarized in the order discussed above, so as to make the manuscript more organized so that readers can see the conclusion more intuitively.
Response: We went through the whole text.
Round 2
Reviewer 1 Report
The authors corrected the article in good direction but still the whole article is one case study and this should be corrected before publishing.
The answer to my objections was:
Response: We used one transistor for each polarization (a total of 8 transistors) and therefore errors could not be given. However, the transistors are very well paired and their electrical characteristics matched perfectly before the radiation. The experimental apparatus did not allow us to irradiate more than 8 transistors at the same time, which would be the only thing relevant for comparison.
I can not agree that nothing could be done to estimate errors. Because the authors use commercial transistors they could for example test at least another 8 transistors in the same configuration. The optimal solution would be to test few sets and estimate errors. Otherwise the whole article is one case study.
English also could be improved further.
There is one clear error in the line 210 where irradiatio should be replaced with irradiation.
Author Response
Author's Reply to the Review Report (Reviewer 1)
The authors corrected the article in good direction but still the whole article is one case study and this should be corrected before publishing.
The answer to my objections was:
Response: We used one transistor for each polarization (a total of 8 transistors) and therefore errors could not be given. However, the transistors are very well paired and their electrical characteristics matched perfectly before the radiation. The experimental apparatus did not allow us to irradiate more than 8 transistors at the same time, which would be the only thing relevant for comparison.
I can not agree that nothing could be done to estimate errors. Because the authors use commercial transistors they could for example test at least another 8 transistors in the same configuration. The optimal solution would be to test few sets and estimate errors. Otherwise the whole article is one case study.
Answer: Unfortunately, additional experiments are not possible.
English also could be improved further.
Answer: We went through the whole text.
There is one clear error in the line 210 where irradiatio should be replaced with irradiation.
Answer: We corrected it.
Reviewer 2 Report
This manuscript studies the possibility of commercial p-channel power VDMOSFETs as an x-ray sensor. Firstly, the equations of threshold voltage shift ΔVT, area density ΔNft and ΔNst are given. Then, the dependence between sensitivity S and gate voltage VG is verified by experiments for three x-ray beams, and a good fitting effect is obtained. The sensitivity increases first and then decreases with the increase of gate voltage. RQR8 beam obtained the highest sensitivity and ΔNft is more dependent on VG than ΔNst. In addition, the author uses silicon instead of silicon dioxide to verify the relationship between absorbed dose D and energy. There is an error between the experimental results and the theory. Finally, the golden ratio GR is used as a new dose parameter to compare the effects of different transistors or operating conditions. This manuscript still has the following problems, which need to be further improved, and it is suggested to review after revision.
- In the introduction, the author should give the background knowledge about X-ray and MOSFET for better understanding.
- The linear regression r2 correlation coefficient mentioned in the third part is greater than 0.99, please give the corresponding formula proof or experimental process.
- Figure and Fig in the manuscript should be in a unified format.
- In Figure 11, what is the reason for the obvious upward and downward trend of RQR8?
- The conclusion of the manuscript should be summarized in the order discussed above, so as to make the manuscript more organized so that readers can see the conclusion more intuitively.
Author Response
Author's Reply to the Review Report (Reviewer 2)
This manuscript studies the possibility of commercial p-channel power VDMOSFETs as an x-ray sensor. Firstly, the equations of threshold voltage shift ΔVT, area density ΔNft and ΔNst are given. Then, the dependence between sensitivity S and gate voltage VG is verified by experiments for three x-ray beams, and a good fitting effect is obtained. The sensitivity increases first and then decreases with the increase of gate voltage. RQR8 beam obtained the highest sensitivity and ΔNft is more dependent on VG than ΔNst. In addition, the author uses silicon instead of silicon dioxide to verify the relationship between absorbed dose D and energy. There is an error between the experimental results and the theory. Finally, the golden ratio GR is used as a new dose parameter to compare the effects of different transistors or operating conditions. This manuscript still has the following problems, which need to be further improved, and it is suggested to review after revision.
1. In the introduction, the author should give the background knowledge about X-ray and MOSFET for better understanding.
Answer: We added it.
2. The linear regression r2 correlation coefficient mentioned in the third part is greater than 0.99, please give the corresponding formula proof or experimental process.
Answer: We fitted the dependence of DVT on Kair (DVT = f(Kair) dependence) for all used gate polarizations and obtained that the r-square (r2) correlation coefficients of the linear regression () are higher than 0.99 for all cases.
3. Figure and Fig in the manuscript should be in a unified format.
Answer: We put Figure at the beginning of the sentences and Fig. inside the sentences.
4. In Figure 11, what is the reason for the obvious upward and downward trend of RQR8?
Answer: We have no explanation for that.
5. The conclusion of the manuscript should be summarized in the order discussed above, so as to make the manuscript more organized so that readers can see the conclusion more intuitively.
Answer: We are not sure that understand it correctly.
Round 3
Reviewer 1 Report
Author Response
Response: Thank you. We have further improved the English language.
Reviewer 2 Report
Accept
Author Response
Response: Thank you. We have further improved the English language.
