You can use scanning probe microscope such as tapping mode AFM or SICM.

Hi, you can use Raman spectroscopy or PL and compare their wavelength to find out the thickness of your desired layer. Besides you can use atomic force microscopy to understand it although AFM may generate the error of one or two layers. Moreover, the high laser power may damage your substrate if it is sensitive.

The other way is the optical path length.

Dear Laura,

I usually measured thin films (deposited by sputtering) thickness by using a marker to cover a significative surface of substrate. Then, I put the sample (after film deposition) into acetone to remove the marker ink. This simple process generally enableS to get a well-defined substrate-to-film step, which can be easily measured by Atomic Force Microscopy (AFM).

You could try to deposite one bilayer on substrate, and use the method I described to calibrate the bilayer thickness. I usually mark the 4 corners and center of a squared- substrate to have considerable data for reproducibilty studies. However, if you want to minimize the damage on your sample, you can mark a little zone in the substrate corner.

Kind regards,

Simón Roa

I understand you look for a fast technique, but TEM is the best (and surprise that you say it is not reliable!).

Anyways, I can give you a general comment in this case, I suggest you try X-ray reflectivity XRR measurements, if you access to XRD equipment. This is a popular common technique used for determining thickness of many thin films, in particular, very thin layers.

The measurement is definitely non-destructive and quick, but you may need spend some time to fitting the results to get the thickness. To have the easiest and quickest way,

1. grow one layer of SiO2 and one layer of Si3N4 (two different samples + preferably thin layers).

2. do XRR to get thickness of each later.

3. then you will have deposition rate for each layer (already know deposition time for each and now you have thickness from XRR; so, deposition rate is film thickness/deposition time), with a high accuracy not like SEM which has high error bars.

4. finally, grow the multilayers with a high, accurate control over each bilayer thickness.

If you also access RBS, you can get combine that with XRR. General speaking, XRR gives thickness, roughness, and density. So, you may get density easily and use that in RBS simulations to get thickness from RBS measurements. Both are non-destructive thin-film methods with high accuracy.