Hi Xavier,

We need more information to give a good answer:

How thick is the film?  (And how accurately do you know the thickness?)
Is your film freestanding or coated on a substrate?  If a substrate, what
material?
How much accuracy do you need in your measurement?
Do you know for certain that your film is fully transparent at 1.55 µm?

Depending on the setup, an ellipsometer can be used to extract the refractive
index, as can a spectrophotometer (R, T measurements at normal incidence and
known polarization).  Most importantly, the instrument needs to include 1.55 µm
in its spectrum - you can't assume that the index in the NIR is the same as
you'd measure with a visible-spectrum tool.  A variable-wavelength ellipsometer
with model fitting software will certainly do the trick.  You do need to select
an appropriate model, but that doesn't necessarily require you to know the
optical properties in advance.  For a well-behaved transparent film, you can
probably get away with assuming a simple Cauchy-style dispersion in the
transparent region.  This does require fitting software which can adjust the
model parameters to fit the measurements.  Many ellipsometers have that; some
don't.

You can even extract n from a single-wavelength ellipsometer (though only at the
measured wavelength!) if you know the thickness with reasonable accuracy.
Ellipsometric measurements (psi, delta) at a single wavelength are periodic in
film thickness with a period of (n*lambda), so as long as you know which period
you're in, you can extract n.  At a single wavelength, you can also use variable
angle ellipsometry to disambiguate your result if your ellipsometer supports
adjusting the arms' angle.

If your film is coated on a substrate, you should ideally take a set of
measurements of the substrate before coating, since its optical properties
affect measurements made by ellipsometry.  Don't assume that "default" values
for a particular material are a good match.  Si in particular is annoying in
ellipsometry because its optical properties vary with doping, crystal
orientation, native oxide thickness, and phase of the moon - always best to
extract an (n,k) curve for your specific substrate before coating.

For a more "lab-bench" approach, if you have a correctly polarized light source
at 1.55 µm and a good NIR photodetector, you could also just try measuring the
Brewster angle, which is independent of film thickness and might get you one or
two significant figures of accuracy in the extracted index.  (For a film in air,
just use n = tan(theta_B) ).

-Mike

On Apr 6, 2011, at 12:45, Xiaohui Lin wrote:

> We have got a plastic transparent film and we would like to measure its
> optical properties.
>
> Could anyone suggest a way or equipment that can be used to determine its
> refractive index for 1.55um light?
>
> Will ellipsometer do? My understanding is that you have to know the material
> and choose model when dealing with ellipsometer.
>
> Any suggestion is appreciated.