Hi Sourav mandal,

In my opinion the best possible way to see smoothness of any surface is by taking SEM and AFM images of both substrate and deposited thin film. You can easily compare the surface roughness of substrate and deposited film by AFM data by checking Rq parameter.

I did same procedure in my previous work for 3 nm gold films on different substrates. 

One question for you to answer first.  What is the diameter of a hydrogen atom?

Perform an AFM scan - in tapping-mode preferably to not potentially scratch your surface - and calculate Ra and Rq RMS values from the scans. Compare before and after. I'd say a 500x500 nm scan size would be good proof.

Hi Sourav Mandal,

Take AFM images of both substrate and deposited thin film. You can easily compare the surface roughness by checking RMS parameter.

measure roughness before and after deposition. it should count the same roughness if it was uniform. Specially be careful of Rmax!

BTW, it absolutely will not be the same!!!

The question was about uniformity of the coverage. It isn`t so easy as it might seem. Surely roughness (at this place one can specify what is meant by "roughness", but let`s skip that part) can be the same for substrate and deposited layer but it doesn't give the information about uniformity of the deposited layer. In this case the best way, to my mind, is doing SEM of the FIB crossection of the deposited sample in order to resolve vertical structure of the layers. Since the layer thicknes as well as materials had not been specified by TS it is hard to predict whether it would be possible to visualize layers by SEM easily. This may be a little tricky.

Best

Sergey, I agree. Uniformity is hard to assess, except by doing scans over large areas or many smaller areas. In any case, both SEM and AFM might work but will be tedious and time-consuming.

I totally agree with you, that roughness in an average value over a certain scan area and may not reflect the uniformity well. I believe a more statistical survey covering several areas of the sample is needed. 

Sourav, what is the area over which you expect uniformity?

1cm2 area and the deposited film thickness was 2nm. The film thickness is main problem for cress-sectional FE-SEM.

how about using AFM for this cross-sectional measurement? 

1 cm2 is quite large for AFM. 

Have you determined the roughness of the surface before adding the film layer yet?

You can get a number to quantify the surface roughness. Just take the average of the Height measured over an unit lenght. or alternatively you can integrate the height measured at different points along X - axis (H)over unit length (L). Roughness parameter = integral of H(X)/L; where L = 0 to L = L. or just take a summation and divide it by umber of points Roughness = Summation of H(X)/N; N of points measured.

We use this technique to measure the surface roughness of the crystals....again the roughness depends on the benchmark material that you would like to compare....(Roughness of sample)/(Roughness of benchmark or standard smooth surface)

Kelvin Probe Force Microscopy can be useful, though it depends on the deposited and the substrate material. However it can hardly be used to characterize the full 1cm^2 area.

You can try scanning Auger microscopy. With this method you can also scan the whole 1cm^2 area by stitching several element mappings together.

It works well on hot-dip galvanized Zn coatings on steel in order to see the homogeneity of the Al-Oxide layer on tip of the Zn.