FTIR or spectroscopic ellipsometer will work, but only if the wavelength is long
enough.  In my experience, most spectroscopic ellipsometers do not suffice,
because their wavelength range is 190-1000 nm.

I have used a two-sided laser profilometer, with precision alignment, to measure
membrane thicknesses in this range to better than 0.1 um.  Cyber-Scan used to
make a product, but I'm not sure if they're in business anymore.  A single-sided
profilometer has the problems outlined by others, in terms of error due to
unknown reference distance or length.  A micro-caliber could work, but I am not
familiar with it.

Al




On Tue, Jan 5, 2010 at 1:38 PM, Andrea Mazzolari wrote:

> Hi All,
>
> i need to realize silicon membranes of thickness 14 and 29 um in (100)
> silicon.
>
> I need to measure membrane thickness with a tollerance of 0.5um.
>
> Any suggestion about a way to measure membrane thickness with a so good
> resolution ?
>
> Best regards,
> Andrea