| No model to simulate or fit optical data is defined immediately after newAD2 starts; it must be loaded using the command:
newAD2> model modelfile
where modelfile is a text file containing the model definition.
Only the first line is obligatory for newAD2 and it has the following syntax:
newAD2: model sid
where sid is the name of one of the implemented model versions. At the moment, models for optics of
thin films are implemented in versions OTF2.1 and OTF2.2, so sid can have the value
OTF2.1 or OTF2.2 (for backward compatibility the program still understands the older, version-less
spelling OTF, and treats it as OTF2.1).
Command model reads the first line of modelfile, calls slaveAD2-sid on the master computer and passes it information about experimental data as well as the rest of the preprocessed modelfile.
Preprocessing
Preprocessing is governed by two special characters:
- !
Denotes a comment. Every character from the first occurrence of "!" until the end of the line will be ignored.
- #
When the multisample mode is used, lines containing the character "#" are cloned and expanded for each sample number.
OTF modelfile
Format of modelfile for optics of thin films is as follows:
newAD2: model OTF2.1 structural parameters: ... media: ... boundary systems: ... sample measurements: ... functions: ... modeled functions: ...
where "..." are placeholders for real definitions.
The order of sections is not obligatory, however every object must be defined before its first use.
The only exception is the definition of the range of a structural parameter.
structural parameters:
In this section a list of structural parameters used for sample description is introduced.
Format:
id [= value] : nice_id : range [unit]
More information can be found on the Structural parameters page.
media:
This section is used to define the media (dielectric tensors) occurring in the systems used for modeling. Dispersion parameters are automatically generated.
Format:
medium_name = dispersion_model_used
For details of the implemented models see the Dispersion models page.
boundary systems:
This section defines Yeh matrices representing boundary systems that can be used to construct the structural models using the Yeh matrix formalism. Boundary and transfer matrices are available, as well as more complex constructs that can be used to define complicated systems.
Format:
boundary_name = definition
For a description of the implemented complex 4×4 Yeh matrices, see the Yeh matrices page.
sample measurements
In this section the Mueller matrices representing different optical systems are defined.
Format:
measurement_name = definition
For a description of the implemented real 4×4 Mueller matrices, see the Mueller matrices page.
functions
This section is dedicated to defining the functions of optical quantities, which the program can either fit or calculate based on the selected model. All experimental functions must be listed here. The format of the experimental function identifier is fixed. Individual experimental function identifiers are separated by commas, followed by a hyphen separating the measurement identifier, and then the sample number in square brackets. Format
function_id = definition
Example
Is,Ic,In-UVV[1] = IsIcIn(EUVV)
The program only checks whether the number of return values matches the dimension of the vector (number of identifiers) on the left side. It does not check the identifier names. The list of implemented functions can be found on the Functions page.
modeled functions
This section defines the ranges of variables for which the functions defined earlier in the functions section will be calculated. The calculation is performed by invoking the calculation command, or automatically during fitting whenever the algorithm finds new model parameters. The format is as follows:
function_id(var1=min[~max:number_of_points[:distribution]])
where the function identifier is identical to one of the identifiers listed in the functions section. The parentheses contain a comma-separated
list of variable identifiers; any variable that is not listed is assumed to have its default value.
Variable values are given either as a numeric constant, or as an interval between min and max together with the number of points in that interval. The distribution value can be either lin or exp, representing a linear or exponential distribution of the variable
(the value lin does not need to be stated).
Example modelfile
Here we present an example model for fitting an unknown thin film described using the universal dispersion model (UDM) on silicon. The optical constants of silicon are likewise described using UDM, with predefined parameters that the program finds in the parfile c-Si in the standard directory.
newAD2: model OTF2.1 structural parameters: ds = 0.38 : d<sub>s</sub> : [0.1,1] mm ! substrate thickness df = 100 : d<sub>f</sub> : [10,1000] nm ! film thickness media: a = Vacuum ! dispersion model of the ambient f = Universal:ex=3:he ! dispersion model of the film s = Universal:c-Si ! dispersion model of the substrate boundary systems: Mf = B(a,f)*T(f,df)*B(f,s) ! description of the front side of the system Mb = B(a,s) ! description of the back side of the system sample measurements: EUVV = Slab(R,Mf,Mb,s,ds,UVISEL) ! description of the measurement by a Mueller matrix: ! reflectance slab measurement, with the front boundary defined by the Mf Yeh matrix, ! the back boundary defined by the Mb Yeh matrix, on substrate s with thickness ds and ! measured with a UVISEL ellipsometer. functions: Is,Ic,In-UVV = IsIcIn(EUVV) ! description of the measured ellipsometric quantities n,k-f = OC(f) ! description of the modeled optical constants of the film n,k-s = OC(s) ! description of the modeled optical constants of the substrate modeled functions: n,k-f(E=0.005~50:4001:exp) ! calculation of the modeled optical constants of the film