I modified my nanofibers by chemical and plasma etching and SEM shows simply that it`s surface roughness is increased is this sufficient to approve surface roughness increment or AFM is necessary?
AFM is much more appropriate for quantitative analysis for surface roughness, although it may be more expensive. The non-contact mode AFM is easier to work with for nanofibers. The contact mode is OK for regular fibers
"SEM shows simply that it`s surface roughness" Are You sure that SEM make 3D diagram, which may let to measure roughness? I dont think so. SEM works because of electron sputtering or/and absorbing. If some part of visible surface is dark and another looks white - it does not mean that dark is deep and white is peek. It means, some parts of structure has differ possibility to absorb/sputter electrons. For example, picture of polymer + ceramics + carbon wire composite will give you light ceramics particles, darker polymer matrix and dark carbon wires, which are all on the same height level. As Guillaume Vidal propose: use 3D measurement tool to get quantitative roughness measurements. In case of higher roughness, may be some problems with AFM (Z scale is low usually - less than 20um). But of course, AFM will give You roughness information of small regions (depends of Z), such as 10x10um. Another disadvantage of AFM is time of measurement, but it worth to try.
For sure AFM is the better choice with possibility of 3D reconstruction and qualitative surface roughness analysis,,,whereas the SEM in general provide you a quantitative results however you can have attachment of some digital camera plus some software program for detailed surface study. Preferably you may need both SEM and AFM to strengthen your findings
I agree with Guillaume Vidal, plus you could also use a confocal microscope to find values for Ra. As Janusz Rebis mentioned, AFM could have some trouble in a large Z scale, it usually works best for smaller scales. I hope this helps.
AFM is much more appropriate for quantitative analysis for surface roughness, although it may be more expensive. The non-contact mode AFM is easier to work with for nanofibers. The contact mode is OK for regular fibers
SEM can provide you with a QUALITATIVE indication of a change in roughness - if the change is very prominent, you might be satisfied with it. However, SEM is not QUANTITATIVE, for a quantitative evaluation, you will need to use AFM. If the change is roughness is not dramatic, it is more convincing if you can show a quantitative assessment of the change. Furthermore, if the change in roughness is a central part of your story, you want to be able to quantify it.
If you want to analyse the roughness quantitatively, AFM provides RMS which is an objective roughness measurement. This bocomes more important specially if you want to differentiate the roughness of your samples quantititatively.
AFM can quantify roughness, but keep in mind the tapered shape of an AFM probe will convolve the surface roughness - and if the texture is deep, narrow or steep, the AFM probe will not provide a realistic view. Thus AFM can be useful, especially for comparing different treatments, but it is imperfect and can be very unrealistic with deeply textured materials. SEM can also quantify, but for this you must do 3D stereo pair imaging. There are a couple of software programs that make this easier than doing the trig yourself - like Alicona's MeX. But you should at least look at some SEM stereo pairs - this will give you a surprising view and may let you know if the AFM images and measures are realistic. Accurate 3D SEM can be done.
In many cases SEM observation is good enough if you have just two groups of specimens. If you see difference, then you know that roughness was changed. If you have multiple groups then you need to quantify roughness. For this you need special software for SEM or AFM.
To all who replied that AFM is definitely needed: it is not so.
I would say SEM can not give you a solid evidence, AFM is good idea, however you can use nano indentation which is similar to AFM, using profilometer is another option. I think the easiest way to quantify surface roughness is by measuring surface energy or hydrophilicity of the surface.