Charge Transport Mechanism in the Forming-Free Memristor Based on PECVD Silicon Oxynitride

Round 1

Reviewer 1 Report

1.       A few quantitative results should be added to the abstract. An abstract should also include the motivation, a brief of the methodology, and the results. Therefore, the abstract should be re-written.

2.       All abbreviations and acronyms should be clearly defined in both abstract and body of the paper.

3.       There are few grammar and punctuation mistakes.

4.       Since it is non-stoichiometric, is there no other information about the possible values of mole fractions x and y in SiNxOy?, and how do these quantities affect its suitability for memristor.

5.       With the sccm quantities of constituents SiH4, N2 and O2 included for SiNxOy:H films production, for how long their respective valves were opened? are there no changes in the pressure and temperature of the chamber throughout the process? Please include time-profile diagram.

6.       A concise description of substrate preparation should be included. Washing with acetone, DI water and drying with N2 gun, baking etc. What type of photoresist was used?

7.       Replot Figs. 4(a&b) and scale the current axes at least to the min. and max. data values.

8.      Finally please bear in mind that conclusions should synthesize not summarize paper.

9

Author Response

First of all, the authors would like to thank the distinguished referee for valuable comments. We hope that taking into account these comments and corrections will improve the article.

 

Reviewer 1 - Comments and Suggestions for Authors

1. A few quantitative results should be added to the abstract. An abstract should also include the motivation, a brief of the methodology, and the results. Therefore, the abstract should be re-written.

Answer: thank you for the comments. The abstract was re-written:

Abstract: A memristor is a new generation memory that merges dynamic random access memory and flash properties. In addition it can be used in neuromorphic electronics. The advantage of silicon oxynitride, as an active memristor layer, over other dielectrics it is compatibility with the silicon technology. It is expected that SiN_xO_y-based memritors will combine the advantages of memristors based on nonstoichiometric silicon oxides and silicon nitrides. In the present work, the plasma-enhanced chemical vapor deposition (PECVD) method was used to fabricate a silicon oxynitride-based memristor. The memristor leakage currents determine it is power consumption. To minimize the power consumption, it is required to study the charge transport mechanism in the memristor in the high-resistance state and low-resistance state. The charge transport mechanism in the PECVD silicon oxynitride-based memristor in high and low resistance states cannot be described by the Schottky effect, thermally assisted tunneling model, Frenkel effect model of Coulomb isolated trap ionization, Hill-Adachi model of overlapping Coulomb potentials, Makram-Ebeid and Lannoo model of multiphonon isolated trap ionization, Nasyrov-Gritsenko model of phonon-assisted tunneling between traps and the Shklovskii-Efros percolation model. The charge transport in the forming-free PECVD SiO_0.9N_0.6-based memristor in high and low resistance states is described by the space charge limited current model. The trap parameters responsible for the charge transport in various memristor states are determined: for the high resistance state trap ionization energy W is 0.35 eV, the trap concentration N_d is 1.7×10¹⁹ cm^-3; for the low resistance state, the trap ionization energy W is 0.01 eV and the trap concentration N_t is 4.6×10¹⁷ cm^-3.

2. All abbreviations and acronyms should be clearly defined in both abstract and body of the paper.

Answer: thank you for the comments. All abbreviations and acronyms were checked and defined in both the abstract and body of the paper.

3. There are few grammar and punctuation mistakes.

Answer: We fixed some grammar and punctuation mistakes.

4. Since it is non-stoichiometric, is there no other information about the possible values of mole fractions x and y in SiNxOy?, and how do these quantities affect its suitability for memristor.

Answer: The film composition was studied using the method of X-ray photoelectron spectroscopy (XPS). Based on the intensity of the peaks of silicon, oxygen, and nitrogen in the XPS spectra, the composition was determined as SiO_0.9N_0.6. The analysis of the composition of silicon oxynitride films from the XPS spectra will be the subject of a separate article; these data have not yet been published. Therefore, this article only briefly mentions the composition of the silicon oxynitride film, minor changes were made to the text of the article and are highlighted in yellow.

5. With the sccm quantities of constituents SiH4, N2 and O2 included for SiNxOy:H films production, for how long their respective valves were opened? are there no changes in the pressure and temperature of the chamber throughout the process? Please include time-profile diagram.

Answer: thank you for the comments. The working gases were supplied to the discharge chamber of the PECVD facility through separate channels, each of which had an integrated electronic gas flow controller.Plasma was ignited after the stabilization of the established gas flow values.The film growth rate was about 7Å/sec, and the growth time did not exceed 45 sec.During this time, the regulators showed stable values.

The substrate temperature was set at 200 °C and maintained during the deposition process automatically by a heater.

6. A concise description of substrate preparation should be included. Washing with acetone, DI water and drying with N2 gun, baking etc. What type of photoresist was used?

Answer: thank you for the comments. The following explanations have been added to the text of the article (marked in yellow).

SiN_xO_y:H films were obtained by PECVD with a gas mixture of SiH₄, N₂ and O₂ on the p⁺⁺-type Si (100) wafers after the RCA-1 cleaning procedure followed by the oxide removal in diluted hydrofluoric acid, final rinsing in deionized water and drying in a nitrogen stream. However, after all the procedures, the loading substrate into the PECVD chamber was carried out through air; accordingly, the silicon was covered with a fresh natural oxide layer

As for the photoresist, we did not use photolithography processes in this work. The respected reviewer, perhaps, did not pay attention to the fact that the upper electrode of the MIS structure was deposited by the magnetron sputtering through a mask with rectangular holes 0.7x0.7 mm.

7. Replot Figs. 4(a&b) and scale the current axes at least to the min. and max. data values.

Answer: thank you for the comments. To get similar max and min values we deleted the current value at 0V. It is not a physical current value and it is rather for the devise sensibility. The 0V point value is also not included in the double-logarithmic figure for the SCLC model. Figures 4, 5, and 6 were replotted.

8. Finally please bear in mind that conclusions should synthesize not summarize paper.

Answer: thank you for the comments. We improved the conclusions:

The IV characteristics of the obtained PECVD SiO_0.9N_0.6-based memristor structures and their switching revealed that the PECVD SiO_0.9N_0.6-based memristor is forming-free. To clarify the charge transport mechanism in the PECVD SiO_0.9N_0.6-based forming-free memristor, the current at different temperature values in the high resistance state and low resistance states was simulated by the Schottky effect, thermally assisted tunneling model, Frenkel effect model of Coulomb isolated trap ionization, Hill-Adachi model of overlapping Coulomb potentials, Makram-Ebeid and Lannoo multiphonon isolated trap ionization model, Nasyrov-Gritsenko model of phonon-assisted tunneling between traps, Shklovskii-Efros percolation model and the space-charge limited current model. After filtering out the models by various fitted parameters, the charge transport mechanism in the forming-free PECVD SiO_xN_y-based memristor in the high and low resistance states can be described by the space-charge-limited current model. The trap ionization energy W_t = 0.35 eV and the trap concentration N_d = 1.7×10¹⁹ cm^-3 were obtained from the simulation by the SCLC model in the high resistance state. For the low resistance state, the trap ionization energy is W_t = 0.01 eV and the trap concentration is N_t = 4.6×10¹⁷ cm^-3.

Reviewer 2 Report

 

This is an interesting paper that systematically explores various current conduction models for charge transport through SiOxNy. The authors also calculate relevant trap parameters for different transport models explored and find space charge limited current transport to best explain the experimental results. This manuscript makes a significant contribution by numerically modelling experimental results. By successfully synthesizing experimental results with theoretical analysis, this study advances the understanding on current transport through SiOxNy for memristor applications. Further, the manuscript is well-structured, has clear figures and sufficiently referenced.

Here are some points which the authors should address to enhance clarity and make this manuscript a valuable contribution to the field.

 

 

1Major Comments:

 

1.    Lines 47-52: The claims related to defect diffusion, their cause and their rates should be referenced from research article that shows preferably experimental proof of the same.

2.    Lines 76-77: The authors should mention which type of cleaning treatment, RCA 1&2 cleans, or ozone clean or something else was used. Also, the type of surface is critical and hence it should be indicated if the Si surface is expected to be H-terminated or of it has OH bonds or native oxide right before the deposition.

3.    In Section 2: Materials and Methods, no information is provided on the software package used for simulations.

4.    Section 2: SiOxNy is a non-standard material, and its properties are a strong function of x and y. The authors should mention the expected value of x and y in the experimental film. Further, what are the material parameters other than the ones that are fitted) used in the simulations of SiOxNy should also be mentioned.

5.    In Section 3.2: Enough information is not provided on how the current conduction is modelled for HRS. There is no explanation given to why should there be any conduction in the HRS state. There appears to be significant trap density in the bulk of dielectric. Hence, it should be mentioned if trap assisted tunnelling or something else is enabled which accounts for the leakage current.

6.    The authors should provide an explanation to why the SCLC characteristics are significantly different from those of the experimentally observed ones. The experimental results show an increasing IV curve which would yield a negative second derivative of current with voltage while the simulated SCLC ones would give a positive one.  

7.    There is no clear demarcation of the parameters that are fitted/varied and the ones that are given as constant inputs to the simulator. The authors should include a table that provides this information as Appendix section. This would greatly improve the clarity of results and discussions section.

8.    The authors should present results (trap ionization energy, effective masses, trap densities etc.) of section 3.1 and 3.2 in tabular format for both HRS and LRS. This will enable a useful comparison of the appropriateness of the models explored for the study. 

Minor Comments:

1.    Line 32: Instead of writing very long, the authors should quote the order of timescales for which the state needs to be held and reference the information accordingly.

2.    Line 41: Authors should mention in which of the parameter is the performance of SiOx based memristor better than those of SiNx based ones. Is the parameter leakage current or something else, the parameter should be explicitly mentioned.

3.    Line 64-65: There appears to be a type as ‘forming-free PECVD SiOx’ is repeated twice. As indicated in the references, the ‘forming-free PECVD SiOx’ should be modified to ‘forming-free PECVD SiNx’

4.    Line 67: Grammar: The authors should modify ‘Not much paper study in depth charge transport mechanism in SiNxOy films.’ To ‘Not many papers study the charge transport mechanism in depth’.

5.    Line 68: Typo: Change ‘SiNxOy film obtained’ to ‘SiNxOy film is obtained’.

6.    Line 71: Grammar: Change ‘but the simulation experimental data’ to ‘but the modelling of experimental data’.

7.    Line 97: Typo: The used instrument is spelled incorrectly as ‘Keythley’. It is correctly spelled as ‘Keithley’.

 

8.    Line 206 and Line 226: The authors repeatedly use ‘simulating experimental data’. This can be more clearly written as ‘modelling the experimental data’ as it makes a better demarcation between experimental and simulation data. 

 

 

Author Response

First of all, the authors would like to thank the distinguished referee for valuable comments. Also, the authors are very grateful for the huge work of the reviewer on correcting errors and typos. We hope that taking into account these comments and corrections will improve the article.

 

Reviewer 2 - Comments and Suggestions for Authors

This is an interesting paper that systematically explores various current conduction models for charge transport through SiOxNy. The authors also calculate relevant trap parameters for different transport models explored and find space charge limited current transport to best explain the experimental results. This manuscript makes a significant contribution by numerically modelling experimental results. By successfully synthesizing experimental results with theoretical analysis, this study advances the understanding on current transport through SiOxNy for memristor applications. Further, the manuscript is well-structured, has clear figures and sufficiently referenced.

Here are some points which the authors should address to enhance clarity and make this manuscript a valuable contribution to the field.

 

1Major Comments:

1. Lines 47-52: The claims related to defect diffusion, their cause and their rates should be referenced from research article that shows preferably experimental proof of the same.

Answer: thank You for the comments. We found the paper with the defect diffusion topic, and with the information, we re-wrote the introduction as follows:

This hypothesis is based on the fact that the information storage time is determined by the defect diffusion involved in the conductive filament forming (filament consists of nitrogen vacancies for SiN_x-based memristors and of oxygen vacancies for SiO_x-based memristors), and the nitrogen self-diffusion coefficient value in SiN_x (0.212 cm^-2/sec at 25 °C [11]) is lower than that of oxygen in SiO_x (0.232 cm^-2/sec [11]). The diffusion coefficient value of nitrogen in SiN_x [12] and diffusion coefficient value of oxygen in SiO_x [13] are moredifficult to compare due to their very strong diffusion coefficient dependence on temperature and the layer obtaining technique [13].

11. Winn, E.B. The Temperature Dependence of the Self-Diffusion Coefficients of Argon, Neon, Nitrogen, Oxygen, Carbon Dioxide, and Methane. Physical Review, 1950, 80(6), 1024–1027.
12. Kijima, K.; Shirasaki, S.. Nitrogen self‐diffusion in silicon nitride. The Journal of Chemical Physics, 1976, 65(7), 2668–2671. 
13. Roussel, M.; Talbot, E.; Pareige, P.; Gourbilleau, F. Influence of the supersaturation on Si diffusion and growth of Si nanoparticles in silicon-rich silica. Journal of Applied Physics, 2013, 113(6), 063519.
14. Lines 76-77: The authors should mention which type of cleaning treatment, RCA 1&2 cleans, or ozone clean or something else was used. Also, the type of surface is critical and hence it should be indicated if the Si surface is expected to be H-terminated or of it has OH bonds or native oxide right before the deposition.

Answer: thank You for the comments. The following explanations have been added to the text of the article (marked by yellow).

SiN_xO_y:H films were obtained by PECVD with a gas mixture of SiH₄, N₂ and O₂ on the p⁺⁺-type Si (100) wafers after the RCA-1 cleaning procedure followed by the oxide removal in diluted hydrofluoric acid, final rinsing in deionized water and drying in a nitrogen stream. However, after all the procedures, the loading substrate into the PECVD chamber was carried out through air; accordingly, the silicon was covered with a fresh natural oxide layer.

3. In Section 2: Materials and Methods, no information is provided on the software package used for simulations.

Answer: We added the information about the software package we used and the information about a more accurate calculation:

To model the experimental data, we used only the analytical calculation in the Mathematica software. To make a more accurate simulation, we need to account for the local field distribution in the sample with Poisson equation and Shockley-Read-Hall statistics for the trap occupation. However, that required more specialized and complex software for numerical calculation.

4. Section 2: SiOxNy is a non-standard material, and its properties are a strong function of x and y. The authors should mention the expected value of x and y in the experimental film. Further, what are the material parameters other than the ones that are fitted) used in the simulations of SiOxNy should also be mentioned.

Answer: The film composition was studied using the method of X-ray photoelectron spectroscopy (XPS). Based on the intensity of the peaks of silicon, oxygen, and nitrogen in the XPS spectra, the composition was determined as SiO_0.9N_0.6. The analysis of the composition of silicon oxynitride films from the XPS spectra will be the subject of a separate article; these data have not yet been published. Therefore, this article only briefly mentions the silicon oxynitride film composition. Minor changes were made to the text of the article and are highlighted in yellow.  

5. In Section 3.2: Enough information is not provided on how the current conduction is modelled for HRS. There is no explanation given to why should there be any conduction in the HRS state. There appears to be significant trap density in the bulk of dielectric. Hence, it should be mentioned if trap assisted tunnelling or something else is enabled which accounts for the leakage current.

Answer: We added the text in Section 3.2 explaining the current conduction in the HRS.

In the memristor, the HRS conduction filament is partially or fully disrupted. In that case, the conduction in the HRS is similar to the dielectric one. The thin dielectric film conductivity can be divided into two processes: the charge-carriers injection from the contact into the dielectric and the charge transport in the dielectric bulk. Because the current depends on temperature, we exclude Fowler-Nordheim model [40]. If there is a high trap concentration in the dielectric, then the space charge on them limits the injection from the contact, and the leakage current is limited by the trap ionization rate. The trap ionization can be described by six analytical bulk-limited current models.

40. Fowler, R. H.; Nordheim, L. Electron emission in intense electric fields, Proceedings of the Royal Society of London A., 1928, 119, 173-181.
41. The authors should provide an explanation to why the SCLC characteristics are significantly different from those of the experimentally observed ones. The experimental results show an increasing IV curve which would yield a negative second derivative of current with voltage while the simulated SCLC ones would give a positive one.  

Answer: We added text in Section 3.2 explaining the SCLC characteristics difference from of the experimentally observed ones:

The SCLC characteristics are significantly different from those of the experimentally observed ones in the HRS at 400 K from the 3 V to the 4 V curve part. This difference is due to the pre-switching part of the IV curve and it is similar to the IV curve in Figure 3 from 6 V to 8 V. At a higher temperature, the switching begins at a low voltage as we can see it in Figure 6f. We simulated the IV curves only at a fixed resistance. But at this pre-switching part the resistance begins to change. So we did not account it in the simulation.

7. There is no clear demarcation of the parameters that are fitted/varied and the ones that are given as constant inputs to the simulator. The authors should include a table that provides this information as Appendix section. This would greatly improve the clarity of results and discussions section.

Answer: We added Table A1 and Table A2 in the Appendix section with all the parameters used in the modeling.

8. The authors should present results (trap ionization energy, effective masses, trap densities etc.) of section 3.1 and 3.2 in tabular format for both HRS and LRS. This will enable a useful comparison of the appropriateness of the models explored for the study. 

Answer: We added Table A1 and Table A2 in the Appendix section with all the parameters used in the modeling.

Minor Comments:

1. Line 32: Instead of writing very long, the authors should quote the order of timescales for which the state needs to be held and reference the information accordingly.

Answer: We changed the text accordingly to the comment:

The memristor is an element that can switch between high resistance state (HRS) and low resistance state (LRS) and save those states for 10 years at 85 °C [3, 4]

3. Lee, MJ.; Lee, C.; Lee, D; Chang, M.; Hur, J. H.; Kim, Y.-B.; Kim, C.-J.; Seo, D. H.; Seo, S.; Chung, U-I.; Yoo I.-K.; Kim, K. A fast, high-endurance and scalable non-volatile memory device made from asymmetric Ta₂O_5−x/TaO_2−xbilayer structures. Nature Mater, 2011, 10, 625–630.
4. Prakash, A.; Jana, D.; Maikap, S. TaOx-based resistive switching memories: prospective and challenges. Nanoscale Research Letters, 2013, 8, 418.
5. Line 41: Authors should mention in which of the parameter is the performance of SiOx based memristor better than those of SiNx based ones. Is the parameter leakage current or something else, the parameter should be explicitly mentioned.

Answer: We changed the text accordingly to the comment:

According to our previous research data, the performance of SiO_x-based memristors is higher than that of SiN_x-based memristors [9, 10]: the switching voltage in the SiO_x-based memristors (+10V and-10V) is lower than in the SiN_x-based memristors (+10V and-16V) with the similar leakage current in the LRS and HRS. Therefore, the power consumption in the SiO_x-based memristors is less than in the SiN_x-based memristors.

3. Line 64-65: There appears to be a type as ‘forming-free PECVD SiOx’ is repeated twice. As indicated in the references, the ‘forming-free PECVD SiOx’ should be modified to ‘forming-free PECVD SiNx’

Answer: Thank you for this comment. We fixed our mistake.

4. Line 67: Grammar: The authors should modify ‘Not much paper study in depth charge transport mechanism in SiNxOy films.’ To ‘Not many papers study the charge transport mechanism in depth’.

Answer: Thank you for this comment. We have slightly modified the text you suggested.

The charge transport mechanism is studied in depth in not so many papers.

5. Line 68: Typo: Change ‘SiNxOy film obtained’ to ‘SiNxOy film is obtained’.

Answer: Thank you for this comment. We fixed our mistake.

6. Line 71: Grammar: Change ‘but the simulation experimental data’ to ‘but the modelling of experimental data’.

Answer: Thank you for this comment. We used your text modification.

7. Line 97: Typo: The used instrument is spelled incorrectly as ‘Keythley’. It is correctly spelled as ‘Keithley’.

Answer: Thank you for this comment. We fixed our mistake.

8. Line 206 and Line 226: The authors repeatedly use ‘simulating experimental data’. This can be more clearly written as ‘modelling the experimental data’ as it makes a better demarcation between experimental and simulation data. 

Answer: Thank you for this comment. We used your text modification.

Author Response File: Author Response.docx