If You want to obtain thicknesses for layers thinner than 4nm I recommend you any Spectroscopic Ellipsometer operating at Vacuum or Nitrogen Pure System (i.e. operating at the wavelengths below 190 nm /at photon energy greater than 6.5 eV/). Currently, there are three company which produce such ellipsometers: J.A. Woollam, Horiba Jobin Yvon, Sopra.
Is it possible to use Scanning Electron Microscopy to measure the thickness? I meant, we make the position of thin film in perpendicular with respect to beam..
Yes, as long as your SEM machine 1) allows a vertical wall (cross-section imaging) sample mount (e.g: Prod #16354 of http://goo.gl/UdsOH) and 2) can tilt to align your film for cross-sectional imaging (as just simply mounting your sample sideways will not be perfectly straight due to various reasons: substrate, film, adhesive that is used to hold the sample to the holder, etc)
Not to mention your film thickness should be sufficiently thicker than SEM resolution limit (~1nm ish). Generally if your film is thicker than ~20nm or so, you should be able to see it via SEM. If not, you can technically do TEM (where the resolution limit is up to 1 Angstrom. An example of TEM image of a film on the substrate can be seen in this article: http://goo.gl/miq9zs) if you want to obtain an image of your film.
Thank you for your answers. I tried to measure in SEM but I did not get good results, it is complicated. As said Jea Cho, is necessary some minimum characteristics. I will read about TEM fos this use.
Ask to Laurent Broch or Aotmane Ennaciri on research gate, they develop their own ellisometers have buyed several ones and can give you advice on the best ellipsometer for your use.