Specific Process Knowledge/Characterization/Measurement of film thickness and optical constants
Feedback to this page: click here
Unless otherwise stated, this page is written by DTU Nanolab internal
Measurement of Film Thickness and Optical Constants
Thickness measurement of thin films can be done by optical measurement methods. We have various techniques for this: Ellipsometry (with the Ellipsometer VASE) and reflectivity measurements (with the FilmTek or optical profilers) or X-ray diffractometer.
If the thin film is not transparent to light or there is insufficient reflectance at the interfaces in your sample, you must find another way to measure the thickness. You can for instance etch a pattern down to the next layer or substrate and then measure the etch step by profilometry. Or if you deposit a layer by PVD you may be able to physically mask part of your sample or an adjacent dummy sample to create a sharp step that can be measured by profilometry (etching will usually create a sharper step, but a physical mask can give you a quick-and-dirty estimate).
Measurement if the optical constants of a thin film is measured together with the thickness of the film either by ellipsometry or by refraction (using the FilmTek). Some of these methods also allow you to indirectly determine, e.g., the surface roughness of the layers in your sample, or, by X-ray reflectivity, the density of the layers.
Comparison of the different methods
FilmTek 4000 | Ellipsometer | Optical Profiler | XRD SmartLab | Etch/mask + stylus method | |
---|---|---|---|---|---|
Generel description | Thin films up to 250 µm, Especially good for thick thin films and for wafer mapping | Good for very thin films down to a few Å | Thin films up to 20µm, especially good for measuring in a small point (down to 4µm) | Good for very thin films down to a few Å | Technique for thickness measurement for non-transparent films |
Method | Reflection | Ellipsometry | Reflection | X-ray reflection | Etch or physical mask combined with stylus profiling |
Film thickness range | <250 µm (for silicon oxides > ~75nm, thinner layer can be measured using the UV light source) | 20 Å to ~10 µm (for silicon oxide) | ~30 nm-20 µm (down to 10nm when using the base without the microscope) | few nm to approx. 100 nm (somewhat thicker for low Z materials) | ~100 nm to 1 mm |
Film thickness accuracy | Very dependent of how good the model fits (if the fit is good it could be within 1 % for a single layer) | Very dependent of how good the model fits. | Very dependent of how good the model fits.(Not so good fitting possibility) | Very good, depends on the model fit | Depends on end-stop accuracy (of the etch) or step sharpness (if a physical deposition mask is used) |
Index range | not any limits | not any limits | not any limit | No index measurement | No index measurement |
Index accuracy | not known | not known | not know | not relevant | not relevant |
Wavelength range | 400-1000 nm (with UV source down to 250nm) | 210-1690 nm | 350-950 nm | Cu Kα radiation (1.5406 Å) | Not relevant |
What kind of thin films can be measured | Any film that is transparent to the light in the given wavelength range
ex:
|
Any film that is transparent to the light in the given wavelength range
ex:
|
In principle any film that is transparent to the light in the given wavelength range. It is limited by the refractive index files in the software and only one model (Cauchy)
ex:
|
Any film that is transparent to the X-rays
including thin layers of metals directly on Si wafers |
|
Substrate size |
|
|
|
|
Limited by the etching method and the stylus profiling instrument. |
Allowed materials |
|
|
|
|
|
Comments |
If no model exists for your layer structure, developing a model may require some work. The instrument has many models available already. |
Measurement and analysis quite time consuming. Can determine layer roughness and density including density gradients |
Direct measurement. |