Ideas for an overview manuscript

[ondrejkrejci]

• [x] Journal Ideas
• [ ] List of tasks to be completed and to be mentioned in the manuscript:
  • [x] working on PiP (#55) and all platforms (#27 ):
  • [x] working damping (#48) and parameters for Pauli (full density)
  • [x] KPFM
  • [ ] GUI (#42)
  • [x] GPU version (#35)

To be closed after the paper has been accepted.

[NikoOinonen]

For the GUI, it's working pretty well right now and not all of the issues mentioned in #42 are critical, but there are some obvious ways to make it crash still, and I would at least like to fix those before the publication.

[NikoOinonen]

We should also make a DOI for the repo to cite in the paper. It's more permanent than an URL. I also want to cite it in my thesis.

[ondrejkrejci]

I think that we agreed on Computer Physics Communications and the overleaf link is here.

[ProkopHapala]

Hi, I sketched some initial (introduction and discussin of different simulations modesl, GPU and GUI), also I added added some relevant citations. I will work on it more later. Please have a look if you want.

[ProkopHapala]

I think we should think about figures now. I'm now thinking about these 2-3

1. Comparison of image contrast over characteristic molecules with all the different methods. Methods should be like:

   1. Lenard-Jones + Point Charges,
   2. Lenard-Jones + Quadrupole
   3. Lenard-Jones + CO-desnity derived electrostatics
   4. FDBM = Pauli form CO density + D3 dispersion atractve vdW + CO denisity derived electrostatics/

   The moleucles could be e.g.

   1. pyridine
   2. PTCDA
   3. BPBA
   4. phtalocyanine with hydrogens in the center
   5. fluor-chlor-brom-benzene
   6. FFPB
   7. CaF2 surface

2. Comparison of image contrast obtained for different parameters woth FDBM (@NikoOinonen)

3. Screenshot of the GUI

@ProkopHapala, @ondrejkrejci, (we should find if we have still some data )
@yakutovicha - do you have data for the old paper where you did the fitting of probe particle parameters? I think the set of molecules you ued was quite representative/ relevant for this

[NikoOinonen]

Hi, I sketched some initial (introduction and discussin of different simulations modesl, GPU and GUI), also I added added some relevant citations. I will work on it more later. Please have a look if you want.

Good work! I fixed some typos and left some comments there.

3. Screenshot of the GUI

We can probably use the one that is on the wiki, unless we want some specific molecule in the image: https://github.com/Probe-Particle/ppafm/wiki/PPAFM-GUI

[ondrejkrejci]

Hei everyone, I am done for this week.

@yakutovicha - could you have a look on some parts like the complex-tip notes and technical details

@mondracek and @aureliojgc - could you write something on the KPFM.

@ProkopHapala and @NikoOinonen - could you work on that image for different theory stages?

And @ProkopHapala - really thank you for the work on the text. I've put some of my visions into it. Have a look if that is OK with you.

[ProkopHapala]

@ondrejkrejci

I was quickly looking into you changes. few comments:

1. ad Figure 1 - Do we really want to introduce double-particle model there? I think the image gives it too much attention (half of the first figure), while in the paper it will be just side-note - or you have different opinion and want to expand some section about double-probe particle ?
2. ad Figure 1 - I don't like visualy to font "verbatin", it is not good on figures which has to be re-scaled, can we use different font (e.g. verdana bold) ?
3. ad Eq.2 - I think it should be written using vectors rather than per-component. We can also write just potential energy (as it is done later)
4. Not sure we have to write explicitly $F = - \nabla E$, everybody knows that

In general:

• Should we continue writing notes Blue directly into the text, or using the comment system of overleaf?
• feel free to add any citations you can think of

[ondrejkrejci]

Ok, 1 -> new image at the end of the paper 2 -> adjusted, please check 3 -> I want to clearly state how the force is applied on the tip. As we have 1 radial and 2 lateral springs, Feel free to change or suggest a better version. 3b -> As for the energies versus forces. Whatever. At the moment energies are used just for the PP-sample interactions. and it is said in the beggining of paragraph 2.2. 4 -> removed. I have adjusted the text once more. And especially the images. I have uploaded svgs so you can adjust the images as well if you want. In blue there is only what images are missing. I have started to use comments, instead of blue. There is quite a lot of them. For example we have at the moment somehow good description of the model, but we are missing some discussion of normally used parameters e.g. for CO tip.

[ProkopHapala]

1 -> new image at the end of the paper

Good, maybe we make the Fig. 1 smaller, or athther something because to my taste it takes much space relative to amount of information.

3 -> I want to clearly state how the force is applied on the tip. As we have 1 radial and 2 lateral springs, Feel free to change or suggest a better version.

I changed it, since how was it written was not really how is it calculated (I think). Originally I did not want to go into these technical details (they are also described in previous paper), but if you want to go there we should do it completely. Not sure if $\odot$ as element-wise product of two vector is something everybody understand.

3b -> As for the energies versus forces. Whatever. At the moment energies are used just for the PP-sample interactions. and it is said in the beggining of paragraph 2.2.

Yes, true, for calculation we actually need forces rather than energies. But in the paper it is easier to write equations for energies. (I think it does not matter much, assuming we did the derivative right :) )

[yakutovicha]

Outline:

1. Intro 1.1. Why SPM (nc-AFM, kpafm)? 1.2. Why ppafm? 1.3. Versatility of ppafm (LJ, electrostatics, KPAFM) 1.4. What to use PPAFM for? 1.5. Brief description of the following topics of the paper (?)
2. NC-AFM models 2.1. What nc-AFM can teach us? 2.2. Description of the tip. 2.3. Very basic LJ model. Characteristics of the model: what does it do and doesn't do? 2.4. + point charges. -/- 2.5. + dft based electrostatics. -/- 2.6. full density-based model 2.7. Show a model system with different sets of features enabled and discuss how they improve the model performance.
3. Other models 3.1. What can be done on top of nc-AFM? 3.2. KPAFM + examples 3.3. STM + examples 3.4. EITS?
4. How to use ppafm? 4.1. Installation (conda, pip, Docker) 4.2. User interfaces (GUI, cmd, python-API)
5. Implementation. 4.1. Numerical methods (here we discuss main numerical tools used in chapter 2) 4.2. Python/C++/openCL
6. Conclusion

[NikoOinonen]

We agreed on some division of responsibilities for each chapter in the last meeting last week. But should some one person first take care of reorganizing the current text into the new chapter structure, or should we all just move the relevant pieces of text to our respective chapters (which could become messy)?

[ProkopHapala]

We agreed on some division of responsibilities for each chapter in the last meeting last week. But should some one person first take care of reorganizing the current text into the new chapter structure, or should we all just move the relevant pieces of text to our respective chapters (which could become messy)?

I think it is better that everybody work on the paper when he has time, and not wait for anybody else. Therefore I think it is better if everybody move his part where he finds it fits. Also the "responsibility" is relative, definitely it is appreciated to improve any other part even if it is not your part.

[ondrejkrejci]

Hi everyone, I have added the chapters and rearranged accordingly. But now, some parts of the text does not make sense (at all). I will look at the introduction @ProkopHapala - could you have a deeper look on the nc-AFM simulations (section 2) now. Especially the section 2.1. - what we can learng from AFM @yakutovicha and @NikoOinonen - could you please look on the images for the AFM methods comparison @aureliojgc and @mondracek - could you have a look on 3.2 KPFM ? @yakutovicha - please have a look on the chapter 4 - How to use PPAFM @NikoOinonen - please have a look on the chapter 5 - implementation Thank you!

[ProkopHapala]

@ondrejkrejci : I went today through the manuscript from 1.Introduction to end of 2. nc-AFM models. I improved what I could and I'm quite satisifed with that parts. (maybe tomorrow I can see the rest)

@NikoOinonen : I think the most crucial part is to have the images comparing various models. Do you have some idea when you can do it? I think large part of the paper revolves around it, and once this is done, the rest of the paper is just polishing details. .... Do you need some help?

(generally I think it would be feasible to finish the paper this year or early January - which I would very much appreciated, in order to mention the paper in yearly report to my grant)

[NikoOinonen]

I think the most crucial part is to have the images comparing various models. Do you have some idea when you can do it? I think large part of the paper revolves around it, and once this is done, the rest of the paper is just polishing details. .... Do you need some help?

@ProkopHapala Yeah, sorry for being slow. I'll try to get to it next week. If you have time, you could maybe look at the data and think if we want all of the different examples, or only some of them.

[ProkopHapala]

I finished reading and editing till the end. I even sketched some summary (feel free to modify it). I think we are not so far from finishing it.

The main things where we should focus now is to add simulated images images for each technique mentioned.

@NikoOinonen

Beside making the figures for 2.2.5. Comparison of AFM simulation models please look here:

• 4.3. Integration with other software - please add something about how you integrate it with ASD-ML (e.g. mentioning the relevatn classes in python). you know more about this than I do.

• Also you may look on 5. Implementation Details to check what I wrote there, and see if you have some ideas what to add

@ondrejkrejci

• Please add some example of simulated image with PP-STM. I think it would be better to take the simulation schematics from Fig.2c and put it together with some example of simulated image into new figure.

• Not sure what to do with Fig.2a (schematics of double probe particle). Consider with @yakutovicha if you want to put 3.4. Double probe-particle model. If yes, then it would be nice to have some example of simulated image, and put it next to the schematics (Fig.2a)

@yakutovicha

I a bit rewrited 4.1. Installation, please check if it is corred (can we install on Windoes and MacOS using pip ?), eventually polish it.

@aureliojgc and @mondracek

• Please add some figure showing the simulation results with KPFM module.
• also look on the section 3.1. Kelvin probe force microscopy after my modifications.
  • I removed some text where you discuss $V{LCPD}$, because I think if we discuss it, we should provide some background formula (kelvin parabola) where $V{LCPD}$ is mentioned. Feel free to do that.

@ProkopHapala

I have to look what is the problem with IETS and eventually also make some example simulation image

[NikoOinonen]

@ProkopHapala I added some text on the Python API into 4.4. I chose not to mention there any specific class names, because these are likely subject to change in the future. Also rewrote the installation in 4.1 a bit.

[yakutovicha]

@yakutovicha

I a bit rewrited 4.1. Installation, please check if it is corred (can we install on Windoes and MacOS using pip ?), eventually polish it.

For your information. I am leaving for the holidays this week, then I will have a tutorial to conduct. I can only contribute during the week of 18-22nd December.

[ondrejkrejci]

@NikoOinonen please could you ask Adam, from which project you were paid, we should add it there! Thanks!

[NikoOinonen]

@ondrejkrejci I added the projects to the acknowledgements.

[ProkopHapala]

Hi all, I added preliminary comparison for AFM simulation models from @NikoOinonen into the paper, and wrote some discussion of the observed contrast. With that I consider the paper almost done. Please have a look on my changes.

• In the figure the pentacene is still asymetric for some reason (due to error in hartree potential) and there is missing column with molecular geometry, but that can be esily supplemented. @NikoOinonen could you please also add all simulation details (parameters) to the text?
• We are still missing some example image of KPFM @mondracek , @aureliojgc . Since Pavel Jelinek does not want to put there any details, it makes it even easier. Let's just put there whatever you already have, just tho show some example how it looks, it does not matter much.
• @ondrejkrejci could you please finish the things related to STM - especially to change the image.
• @ondrejkrejci - I forgot about IETS part. You perhaps also did not do any progress in searching for source of the bug?

EVERYBODY: Please have a look on the part relevant for you. I would be really happy if we can finish and submit it before chrismass or early in january.

[NikoOinonen]

@ProkopHapala I updated the figure with the geometries and added the simulation parameters to the text.

[NikoOinonen]

@ProkopHapala I updated the figure again, now with the pentacene fixed, so you could add discussion on that example as well.

[ProkopHapala]

Hi @ondrejkrejci I updated the Fig.1 (schematics of PP force), maybe you can look on it if you agree ?

Also slightly modified the Fig.3 (PPSTM). I put there as placeholder image of HOMO from our old article. Please replace it (it was monolayer, not single molecule, also there is this stupid red line). And check if what I wrote there is correct? Is it HOMO, is it with $p_z$ orbital on the tip?

Now it remains to put there the image of IETS. Should I make such image or will you do it? The thing is I don't remember very well the last simulations we did in Bruno's paper (because you were doing it). It would be good to make it somewhat consistent. But in principle I can do it as well.

[yakutovicha]

An idea for the title:

Advancing Scanning Probe Microscopy Simulations: A Decade of Development in Probe-Particle Models

[NikoOinonen]

So I will try to soon add an example that plots the main simulations figure: #246

[ProkopHapala]

An idea for the title: Advancing Scanning Probe Microscopy Simulations: A Decade of Development in Probe-Particle Models

I quite like it, I only worry that it would look like historical summary without any new features. But we have a lot o new features.

[NikoOinonen]

An idea for the title: Advancing Scanning Probe Microscopy Simulations: A Decade of Development in Probe-Particle Models

I quite like it, I only worry that it would look like historical summary without any new features. But we have a lot o new features.

I think it's good. I would say the word Advancing is enough for suggesting that there are new developments.

[ProkopHapala]

Hi, I was just looking on the state of the article. To me the text looks fine. Does anybody have any serious objections or suggestions with respect to the text?

We are still missing the the figures:

• @ondrejkrejci - please update the STM figure (I put there just as placeholder screenshor from original article (https://journals.aps.org/prb/abstract/10.1103/PhysRevB.95.045407). I think you can still take data from this article just re-plot the so that it is not just screenshot.
• @mondracek - The KPFM of fluorinated molecule you send by email looks perfect to me. Can you incorporate that figure into the manuscript?

Then I think we just have to check format of references, remove the blue comments, and we are ready to go. Or do you have other ideas?

[yakutovicha]

To me the text looks fine. Does anybody have any serious objections or suggestions with respect to the text?

Tonight I will go through the text and fix issues that I find. I will make sure to have it done by tomorrow's meeting.

[NikoOinonen]

I went through the text just now. Mostly I think it's complete. A couple of things I noticed:

• @ProkopHapala your affiliation is not set.
• The first figure uses some symbols that don't match the text. @ProkopHapala I think you made this figure?
• In the KPFM figure, if possible it would be better to put the units next to the colorbars (@mondracek).
• In the STM section there is a reference to a figure that does not exist (@ondrejkrejci).
• The IETS figure is missing a proper caption (@ProkopHapala?)
• Some missing project number in the acknowledgements (@ondrejkrejci)

[ondrejkrejci]

Thanks @NikoOinonen!

IETS is my problem. @mondracek - add your projects into acknowledgments, if you want. @mondracek and @ProkopHapala - add acknowledgements for the supercomputers.

BTW, I will work on abstract and cover letter, hopefully over the weekend.

[ProkopHapala]

OK, I did now:

• I wrote IETS figure caption and connedted it to the text
• wrote proper affiliation for @ProkopHapala and @mondracek, I also assigned imdea to @aureliojgc (it was assigned to @mondracek for some reason)
• I added acknowledgment of metacetrum for @ProkopHapala and @mondracek
• I modified the variable symbols in the text to match the figure (hope I did not overlooked something - if sombody check it it would be good)

[ondrejkrejci]

@yakutovicha - overleaf line 480, could you please provide a XC that you have used for your DFT? Thank you!

[ondrejkrejci]

@ProkopHapala - I have checked those symbols - it seems to be fine and consistent as far I can tell.

Please have a quick look on:

• STM part
• 1st sentence of concluasions (adjusted slighly)
• I do not like that the IETS image caption is so long, but I am out of ideas now. In the worst case we can work on it, if the referee will complain about that one. Have a deeper look at:
  • abstract
  • data availabilty section. I do not know, if we should put it in a new section. E.g. in here they have is in the conclusions.

I will provide @NikoOinonen my data at latest tomorrow, so he can provide the citations/url there.

@mondracek - did Pavel checked your section? Don't you want acknowledge him in the acknowledgments?

@yakutovicha - what about your supervisor.

I will work on the cover letter at latest tomorrow.

[NikoOinonen]

@mondracek If you could also send me the KPFM-related files, then I can add them to the dataset at the same time.

[ProkopHapala]

@ondrejkrejci did we already settle the journal? We have a rule that before submission we have to report to head of department where we want to submit, and he must agree (main purpose is avoid MDPI or predatory journals)

[ondrejkrejci]

Sure, we go now for: https://www.sciencedirect.com/journal/computer-physics-communications

[ProkopHapala]

@ondrejkrejci

• read the abstract and made some style modifications
• read STM and made some modification and the end (mostly style), added some reference.
• read KPFM and made minor modification at the end. Added reference to Nikolaj Molls paper about FFPB molecule.
• ad "IETS image caption is so long" - if you have idea, shorten it, if not, I don't think it is a problem. I don't think referee would complain about shuch things.
• please share the Cover Letter, best in Google Doc.

[ondrejkrejci]

Done from my side. We are almost finished, we are discussing some final details about the cover letter. Please have a look there, see todays email. Does anyone wants to have some final look? Or could we go through?

[ondrejkrejci]

@NikoOinonen - could you check, if you can change the description of the dataset on Zenodo? Like that we just add there the connection to ArXiv, once it is uploaded? Thank you! @yakutovicha - Would it be possible to do a new release, so the Zenodo database is directly connected with the latest version? Thank you!

[NikoOinonen]

@ondrejkrejci Only changing the uploaded files requires a new version. All the other information can be edited at any time.

[ProkopHapala]

@ondrejkrejci @NikoOinonen

Hi, there are data for the IETS, the zip has 430MB, I did not want to spend time contemplating what to include, co basically I just zipped whole simulation folder (including scripts, inputs, but also output FF, vibration eigenvectros, and PP positions, and AFM, all in both .npz and .png). If you think we should sort it out and add there just the necessary, feel free to repack it.

All data (.zip, 430MB) https://mega.nz/file/bOxSGDAY#aeT5N4oyurJlzi7FDFeRs46p6uEJW7HQ6Qr6dg6yHX0

Or if you want just inputs, there it is, it has 5kB IETS_FePc_Au-just-inputs.zip

[NikoOinonen]

@ProkopHapala If the size of the input files is only 5kB, I think we could just upload it as an example to the repo. I'm thinking we only use Zenodo for the large files.

[ProkopHapala]

@ProkopHapala If the size of the input files is only 5kB, I think we could just upload it as an example to the repo. I'm thinking we only use Zenodo for the large files.

yes, why not. We can put it as example.

But I was thinking you wanted also to have somewhere the results because (e.g. to monitor eventual changes in the code, as @mondracek was emphasizing)

[NikoOinonen]

to monitor eventual changes in the code

Well, we have versioning for that reason. I would say most of the time people are looking how to run the latest version instead of a historic one. And the general goal is to maintain the existing examples anyway.

If we really want to make sure that there is an archive version, I think Zenodo also supports archiving a Github repo and giving a DOI to it.

[ondrejkrejci]

@ProkopHapala

• yeah let's upload that test to the examples and have let's just use the Zenodo description to it.
• I have made one last adaptation to the abstract - I am using just probe-particle model as standing for anything inside and I am very consistent with it. @mondracek
  • please send the KPFM data (just the input data) to Niko

We are ready to go tomorrow.

[NikoOinonen]

I got all the files now, and uploaded them to Zenodo: https://doi.org/10.5281/zenodo.10563098.

I added a basic description of the files. Feel free to suggest a more detailed description and I can add it.

[ondrejkrejci]

Hi Everyone:

• submitted to Computer Physics Communications

• arXiv: identifier 2401.14564

  Thank you all and fingers crossed!