I am trying to analyze using several softwares, how much of the oxide area has been deposited by my technique. The layer formed is not homogeneous and I need a way of calculating how much material has been deposited. I am currently using WSxM software with the flooding option which estimates this same thing, but I am noticing it is very arbitrary as it all depends the height you want to choose as your threshold. For example some times the substrate can be considered 0nm so anything above 0.1nm can be considered material deposited, but other times, the substrate seems to be starting at 0.5nm so if you use the previous value of 0.1nm you wouldn't get anything flooded. Hence I need to use 0.6nm and this is quite arbitrary and depends completely of what I want.
If anyone can tell me a better method for calculating the percentage coverage, taking in mind the material is very poorly adhered to the surface (so any contact or intermitent contact between tip and sample will destroy both, hence techniques such as adhesion comparison cannot be used as a contact is needed).
And a second question: Is there any way to do adhesion mapping in tapping mode in no contact regime completely?
Hi Guillermo, I am just wondering aloud here -- correct me if I am wrong. Generally my samples are well structured nano materials with perfect pitch and diameter -- so its easy for me to do this. Assuming your sample as random shapes of patches of different dimensions, here are few things one can try (I suppose). Also this is assuming thresholding the image isn't working for you (did you try imageJ as well?).
1. If you have good contrast in your image (or you can enhance them) and patches are pretty large, flatten / make grey scale image and import it to MATLAB, apply Canny's edge detection followed by area measurements -- you should be able to download these (if you do not have already) from Mathworks, there are several codes available. MATLAB struggles with 1024 x 1024 images, but 512 x 512 should work very well.
2. If the patches are very small and random, you can use FFT feature in WSxM software, get a plot between PSD and k, get the k value from the intersection of plateau and decay region, use this intersection k value as correlation length, (measure from multiple frames, get an error bar). Now is the tricky part -- you need to have an approximate idea of the average size of the patches -- approximate should be good enough. if thresholding isn't working, just use profile measurement options and get some decent number of patches measured manually. More the better -- if I were you, I would target somewhere between 15 - 20 frames of soft tapping images (using same tip, different parts of the sample). Now you have area of the patches plus an error bar. Number of patches ~1/L^2 /unit area where L is the correlation length determined from the FFT / PSD process I mentioned. From these two you should be able to get the % coverage -- accuracy would depend on number of frames you analyze.
3. As for the last question regarding adhesion mapping -- phase image is one way -- better is doing force modulation microscopy (if you are using Veeco / Bruker systems, others may call it different). Else you can functionalize your tip with suitable molecules and use that to create a force map / image. Either one generally produces decent results given appropriate tip - sample interaction.
I understand there are not a lot of concrete ideas here -- but hope this helps,
cheers,
Palash
I have tried it but I cannot find any kind of utility to do what I need. At least WSxM software has the flooding but nanoscope doesn't look like it has one. Do you know a way with nanoscope software? Thanks by the way.
Excuse me, telling me to use a program doesn't really help me, as I know all of these programs already. I need to know HOW to do it not in which program. Thanks anyway.
you can use phase image instead of the topography. Phase images give an idea about the different interactions between the tip and the different components of the surface, thus it will be possible to discriminate your deposition and to calculate the coverage percentage.
Hi Guillermo, I am just wondering aloud here -- correct me if I am wrong. Generally my samples are well structured nano materials with perfect pitch and diameter -- so its easy for me to do this. Assuming your sample as random shapes of patches of different dimensions, here are few things one can try (I suppose). Also this is assuming thresholding the image isn't working for you (did you try imageJ as well?).
1. If you have good contrast in your image (or you can enhance them) and patches are pretty large, flatten / make grey scale image and import it to MATLAB, apply Canny's edge detection followed by area measurements -- you should be able to download these (if you do not have already) from Mathworks, there are several codes available. MATLAB struggles with 1024 x 1024 images, but 512 x 512 should work very well.
2. If the patches are very small and random, you can use FFT feature in WSxM software, get a plot between PSD and k, get the k value from the intersection of plateau and decay region, use this intersection k value as correlation length, (measure from multiple frames, get an error bar). Now is the tricky part -- you need to have an approximate idea of the average size of the patches -- approximate should be good enough. if thresholding isn't working, just use profile measurement options and get some decent number of patches measured manually. More the better -- if I were you, I would target somewhere between 15 - 20 frames of soft tapping images (using same tip, different parts of the sample). Now you have area of the patches plus an error bar. Number of patches ~1/L^2 /unit area where L is the correlation length determined from the FFT / PSD process I mentioned. From these two you should be able to get the % coverage -- accuracy would depend on number of frames you analyze.
3. As for the last question regarding adhesion mapping -- phase image is one way -- better is doing force modulation microscopy (if you are using Veeco / Bruker systems, others may call it different). Else you can functionalize your tip with suitable molecules and use that to create a force map / image. Either one generally produces decent results given appropriate tip - sample interaction.
I understand there are not a lot of concrete ideas here -- but hope this helps,
cheers,
Palash
Thank you very much Palash. I will try what you say but it looks good to work with for now. This is what I meant by a "new idea". I thank you very much for your help and for taking time helping with this issue.
Guillermo,
Image tool, the Image J software is what I use as well. That would be my starting point. PicoScan, an Agilent software, also has a feature similar to this that allows you to determine this as well. It is called slices.
I will try it too. Every little help is appreciated as maybe this time one thing works but for future experiments it might not so every answer is helpful to me. Thank you very much
I have tried also Phase imaging, but it doesn't work as expected because I am scanning in a non-contact regime hence phase diagram is not too clear. It would be different if I was doing an intermittent contact or complete contact, but the problem is my material would get stuck on the tip changing morphology and other properties which I am interested in analyzing.
And by the way, anyone know where to get ImageJ from? The webpage is down and the installation folder I found isn't for windows 64bits...
http://rsbweb.nih.gov/ij/download.html
You can download 64 bit software from there -- the site is up, I checked just now. Plugins such as Fractal Surface Measurement and surfaceJ may be of interest for you.
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructures
- Thin Films