Trap-Related Reliability Problems of Dielectrics in Memory Cells

Round 1

Reviewer 1 Report

This is a very nicely written paper. A good literature review is given which is important for further studies in the same area. 

I have a few comments:

1. Please give full forms of the abbreviation used in the paper, whenever you are using something for the first time. For e.g. TEM on line 87.
2. Page 13 and Page 16, please correct the format.
3. Please explain the x and y-axis in your figures, even if you cite some paper, make sure it is understandable from your paper, what the x and y-axis means.

Author Response

Dear Reviewer 1

Thank you very much for your evaluation of our works and kind suggestions to refine the manuscript. Please see the attachment about our reply. 

Best regards, 

Authors

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript reports experimental results and theoretical investigation about the effect of traps on the current in dielectric stacks. The analysis is interesting, but the writing must be significantly improved since in the present for it is very hard to understand the procedures followed to interpret the experiments. Beside having the manuscript proof-read (not for the grammar, but because a lot of sentence have no logical sense to me), that authors should address the following major comments

 

1. I would not call the trap “noise sources”. I understand that the authors also analyze noise, but in general traps are source of fast as well as slow fluctuations, whereas noise usually applies to fast fluctuations around a “stable” bias point. I would call them “Perturbations”
2. About Figure6. It is not obvious that the current increases because we have tunneling across the percolation path produced by two traps: the second trap may just be in more favorable position to enhance the current more than the first one, but the two traps may still be at two different position and involved in single trap tunneling phenomena
3. Why the Fourier transform cannot be applied to the noisy waveform? After removal of the “slow” time evolution one can perform FFT and find the time constants of the RTN.
4. What does it mean “apply the Fourier transformation to the fitting formulae”? This would just transform the transient described by 7 and 8 and not the fluctuation related o 1/f noise, that take place on shorter times…I guess that the fitting formulas have been removed from the measured IV so to leave the short time fluctuations (see my point above) that are then Fourier transformed to get power-spectral-densities (PSD). From the slope of the PSD, alpha in Fig.17 is obtained. Am I correct? What procedure has been used for the green line in Fig.17?
5. It is hard to find the link between Eqs.9, 10 and the modeling in Figs.18 and 19: in my view the process is random, I cannot see how to model the exact timing at which a current transition takes place. At least one may model the PSD of the process and generate simulated sample waveforms with similar features as the experiments, but not the exact timing.

 

Other “minor” adjustments are:

1. Try to keep the symbols in (2) and (4) as close as possible not to confuse the reader. For example “Ei” and “Ef“ in (4) are equivalent to “E” and “E+qVg” in (2). Eq.(49 uses the Dirac’s delta, whereas (2) has already integrated it to get a density of states.
2. Something wrong in the calculation “CR = 0.52⁄0.4 ≅565”. It is true that 0.52/0.4=1.3. Maybe the authors meant: 0.52/(0.4+0.52)=0.565
3. 6, “D(t)” should read “D(tau)”
4. Again on Eq.6: what is the D(tau) giving alpha=0? This should be reported to help the reader connect the 1/f noise to the distribution of time constants
5. Keep the caption of Fig.7 in the same page of the figure
6. 11: what is the applied voltage?

 

About the sentences (or portions) that I found hard to understand, I report below just a few:

1. “..are somewhat to record changes in..”
2. “…as emerging memories to draw extra values which cannot….”
3. “The major innovations that the human being has made are two….. “ over which time period? Last century? Since the appearance of mankind?
4. “During the period that we treat with”
5. “If it is as is, then the improvement of information communication efficiency may cancel that of the energy conversion efficiency by a considerable amount”. Better to say that the improvement in communication efficiency may surpass the one associated with conversion efficiency? Any number to support that?
6. “..then Equation 2 can be validated” should it be something like “the Eq.2 is valid”?
7. “Electrons coming from the diffusion layers G to the inversion layer can likely emit to the ..” I would say “.. can be injected into the…””
8. “In other words, the emission current reduces over time due to the emission itself. While the emission dominates the DT”. Please specify in the two sentences from where the emission takes place. Otherwise the two sentences appear to contradict one another.

Author Response

Dear Rev-2

Thank you very much for your valuable comments and suggestions. Everything you gave us is really helpful to refine the manuscript. 

Please see the attachment on our reply. 

The other revisions to reply other reviewers and further minor revisions we added are underlined. 

Best regards, 

Authors

Author Response File: Author Response.pdf

Reviewer 3 Report

See attached

Comments for author File: Comments.pdf

Author Response

Dear Rev-3

Thank you very much for your interest on this research review and your effort and  valuable comments to refine the manuscript.

The other revisions to reply to other reviewers and further minor revisions we added are underlined. 

Best regards, 

Authors

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The revised version shows significant improvements with respect to the original one. I have a few additional suggestions.

1. I do not really like the “trap free” expression (e.g. in Fig.4 and Fig.7 and in part of the text). In the text I would write “tunneling not assisted by traps”, in the sketches I would write “Direct Tunneling”
2. 15 and 16: please make explicit in the caption that the plot b is the application of Eq.6b to the functions in plot a one the model (solid liens) is removed from the experiments (symbols). Am I correct?
3. Regarding Figs.18, 19 and associated text, I would not say “discrete perturbations and phonon scattering are synchronized.”. As I see it, the model reproduces the “jumps” seen in the experiments and those jump are thus artificially lined with the ones in the experiments. Am I correct?

Author Response

Dear Rev-2;

Thank you very much for your further careful review and helpful suggestions. According to your help, we have successfully refined the manuscript. 

Please see the attached document in more detail. 

Best regards, 

Authors

 

Author Response File: Author Response.pdf