Frequently Asked Questions (FAQ)
- What wavelength range does your ellipsometer system cover?
- We cover the range from 190 nm to 30 µm.
- Is one measurement for whole range?
No. There are several detectors and different light sources. Therefore, it needs at least two different measurements.
- What are the incident angles for ellipsometry measurement?
They depend on your substrate and desired precision. The incident angles could vary from 30 to 89 degree automatically or manually.
- Do you use multiple angle measurement?
Yes or no. It depends on your specific samples.
- How long will it take for one measurement?
Measurement time varies from sample to sample and also model to model. For example, one full measurement at one incident angle or one mapping site only takes 1 second with our array based ellipsometer tool. While one measurement may take several minutes or even hours with scanned single element detector systems.
- How big samples can you handle?
TFProbe SE under standard configuration, can handle samples up to 300mm in diameter.
- How small samples you can take?
We can take samples at a minimum size of 1mm in diameter.
- How about the sampling area on sample?
- The parallel beam size is adjustable and it is between 1 to 5mm in diameter. For small samples, we use focused beam that has a beam size about 100µm only.
- Can you check uniformity information over a wafer?
Yes, we can run as many point measurements as you want over a maximum 300mm wafer.
- Will your tools come with turn-key operation?
Yes. There is full recipe setup sp user can click button once to get final results.
- Can you measure Ge concentration in SiGe alloy?
Yes, we can measure Ge concentration in both strained and relaxed SiGe films as the same principle for Al concentration in AlGaN films.
- Can you determine optical band gap for GaN, AlGaN, or SiC, etc.?
Yes, we do run band gap calculations based on the obtained optical constants from the ellipsometry modeling.
- Can you measure multiple layer stack?
Yes. A spectroscopic ellipsometry run at variable angles will help to get accurate results for each layer.
- Can you measure embedded ITO layer conductivity with ellipsometry technique?
Yes, with near infrared or middle infrared ellipsometer, Drude dispersion could be used to obtain electrical properties of conducting films.
- What materials ellipsometer can work with?
Almost all materials can be characterized with ellipsometry, such as Metals, polymers, ceramics, glasses, semiconductors and their compounds, composites etc.
- How about application fields for ellipsometry?
- Here is a partial list of application fields for ellipsometry:
- Semiconductor Industry:
- Gate dielectrics
- Semiconductors and their alloys or compounds such as, SiGe, InGaAs
- Optical coatings
- Semiconductor compounds
- Functional films in Optical MEMS
- Data Storage
- Diamond-like carbon (DLC)
- Magnetic films
- Flat Panel Display (FPD)
- Thin film transistors (TFT) stack
- Conductive oxide: Indium Tin Oxide (ITO)
- Any Advanced Applications for ellipsometry?
- Yes. Here are some examples for advanced applications for ellipsometry.
- Ultra-thin gate dielectric films
- Porous structure
- Deep trench profile
- Graded structure
- Corrosion & protection of metals and alloys
- Real time process monitoring for deposition rates, etching rates, etc.
- Ultra-thin gate dielectric films
- Semiconductor Industry:
- Other questions?
Please contact us.
- What is the accuracy of the system?
In general, there two different cases. One is direct technique and the other one is indirect. For reflection and transmission measurement, they are direct measurement although they need background dark measurement and reference measurement. These measurements basically are light level measurement. Therefore the accuracy for these measurement purely depends on hardware setup.
For film thickness measurement, it is indirect measurement comparing transmission measurement. A model is always needed to get thickness information. therefore, thickness accuracy not only relies on hardware but also data's mathematically handling which is the second step involved in thickness measurement. From theory section, it has been learnt that thickness is obtained from phase thickness in the layer which is related to wavelength, physical thickness and its optical properties such as refractive index and extinction coefficient. Therefore, thickness accuracy also relies on accuracy of optical properties used in the layer stack.
- What if I can not find optical constants in the database for our films?
Optical constants are not constants. They vary at different wavelength and by different processing technique. Processing techniques affect packing density, materials crystalinity etc. therefore affect optical properties. It is recommended that user should evaluate what optical constants sets are proper for his films. If user can not find optical constants in database, we can provide service to determine optical constants precisely with spectroscopic ellipsometry technique and then provide to user for use.
- What if the film is not homogeneous?
User can use average optical properties for that film.
- Why measured thickness depends on initial input or start values?
To speed up calculation, program always tries to start from the input thickness to minimize the sigma between measured and calculated spectra. Depends on the range, it is possible to miss true minima (global) and then will give out thickness obtained at local minima. There is trade off here. Wide range provides high possibility to find global minima but take more time to do the job.
- What if the detector is saturated?
You need to start over all measurement again including measuring dark background after you reduced integrations time.
- When should I use small pixel group?
Pixel group function provides average signal over boxed pixels. If you measure very thick films which will display high dense fringes in the reflection spectrum, it is recommended that use small pixel grouping.
- What assumptions used in film thickness measurement?
- Like all other optical model based technique, there are following assumptions in film thickness measurement:
- Layer stack generally describes the real structure of measuring sample
- Film and substrate are homogeneous
- Over sampling beam area, film properties are uniform
- Sample surface are flat and interface between film and substrate are sharp
- Sample is flat
- For reflection measurement, sample surface and reference surface are adjusted to the same level (height)
- Nothing changed or moved such as fiber bending or power fluctuation
- Light source is stable
- How to improve measurement accuracy?
Any improvement in above question will lead to a better result.
- What is the difference between reflectometry and ellipsometry?
Both reflectometry and ellipsometry are optical, non-contact and non-destructive methods. In general, reflectometer is used to acquire reflection spectrum over a wavelength range. If there is film or coating on some kind of substrate, the film or coating thickness can also be figured out from the measured reflection spectrum. Ellipsometry measurement is performed at non normal incident angles. The two ellipsometry parameters, Psi and Del, give more information than reflectance itself. Therefore, more information can be accessible through ellipsometry technique, such as multiple layer analysis, dielectric constants calculation, surface or interface roughness, inhomogeneity behavior etc. Of course, the information is always derived from an optical model, which leads to some difficulties to use this technique although it is much more powerful than reflectometry technique.
- Which one should I choose for thickness measurement then ?
For thickness measurement, both techniques rely on modeling. In general, ellipsometry gives better accuracy than reflectometry in thin thickness range such as below micron level. For a typical ellipsometer configuration, the maximum thickness measurable is below 10 microns. However, reflectometer can measure up to hundreds microns of non absorbing thick coatings.
- How thick and how thin I can measure for thickness?
These optical methods need to have light to penetrate through film and reflect back from film/substrate interface. Without meeting such condition, the film thickness can not be figured out because of lacking necessary phase information. One example is a sample with thick metal films on it. Because metal has high absorption in visible and Near infrared range, light only can penetrate metal film with a depth of less than 1000 Angstroms. For such film with a thickness above 1000 Angstroms, it is impossible to measure its thickness by reflectometry or ellipsometry although reflection spectrum can be obtained with reflectometer and optical properties for metal film can be obtained with ellipsometry. Roughly, the measurable thickness for various films can be estimated from penetration depth if knowing its extinction coefficient or absorption coefficients.
- Can you give some application examples with spectroscopic ellipsometry measurement?
- Optical constants (refractive index n and extinction coefficient k) for thin films, coatings and bulk substrate
- Accurate nondestructive thickness determination for thin films
- Alloy concentration determination for various thin films such as Ge in SiGe alloy, Al in AlGaN films
- Band gap determination for GaN, SiC, AlN, AlGaN, etc.
- Porosity measurement in low-K films
- Relative volume fraction determination for each component in nano-composite
- Physical thickness and optical properties for each layer in a multiple layer stack or periodic structure such as quantum well structure
- Thickness and optical properties uniformity information
- Inhomogeneous film analysis in physical density or alloy concentration
- Optical properties for high-k films
- Nondestructive measurement for electrical conductivity of metal films, metallic compounds (such as WN, TiN, TaN, etc.), doped semiconductor epi layers (thickness can be also determined at the same time), other compound oxides such as ITO films
- Nondestructive measurement for doping concentration in doped semiconductors (active dopant! not total concentration as given by destructive SIMS analysis)