Method for determining layer thickness ranges
First Claim
1. A method for determining layer thickness ranges of a plurality of layers (Si) of a specimen, the method comprises the steps of:
- generating for each of the layers (Si), the wavelength-dependent refractive index and absorption functions ni(λ
) and ki(λ
), as well as starting values (Timin,0, Ti0, Timax,0) for a minimum, nominal, and maximum layer thickness;
measuring a reflection spectrum of the specimen in a wavelength range (λ
min, λ
max) smoothing the reflection spectrum by diminishing noise caused predominantly by external influences and thereby generating a smoothed reflection spectrum;
determining the number of extremes in the smoothed reflection spectrum, wherein two adjacent extremes, of which the one must be a minimum and the other a maximum, being evaluated only if they differ at least by a specified contrast criterion;
determining the layer thickness ranges based on the number of extremes, such that for each layer (Si) the thickness (Ti) is varied between (Timin,0) and (Timax,0) in steps having a predetermined increment (dTi), a reflection spectrum being modeled within the wavelength range (λ
min, λ
max), and the number of extremes is determined and stored and compared to the number of extremes in the smoothed reflection spectrum, specifying the wavelength range (λ
min, λ
max) and the increments (dTi) in self-consistent fashion by means of a sensitivity criterion, such that for each layer Si for a starting increment (dTi0) and in a wavelength range (λ
0min, λ
0max), a reflection modulation spectrum is modeled, the absolute value thereof is determined, and the resulting spectrum is then smoothed;
limiting the wavelength range to a continuous range (λ
imin, λ
imax), wherein (λ
imin) is the smallest and (λ
imax) is largest wavelength, for which the modeled spectrum still exhibits a value above or equal to a specified threshold value and the modeled spectrum substantially exhibits, for values between those wavelengths, values above the threshold value;
selecting (λ
min) as the minimum of all (λ
imin), and selecting (λ
max) as the maximum of all (λ
imax);
calculating the increments (dTi) for the wavelength range (λ
min, λ
max) and are in turn used to model a reflection modulation spectrum for each layer (Si); and
repeating until the wavelengths (λ
min) and (λ
max) change, between two repetitions, only by less than a specified self-consistency criterion, in which case the wavelengths (λ
min), (λ
max) and increments (dTi) are used for the further method steps.
1 Assignment
0 Petitions
Accused Products
Abstract
The invention concerns a method for the determination of layer thickness ranges of layers of a specimen, in which the reflection spectrum of the specimen is measured in a specified wavelength range and then smoothed, the number of extremes in the smoothed reflection spectrum is determined, and the determination of the layer thickness ranges is accomplished by comparison with the number of extremes in the modeled reflection spectra, such that for each layer the thickness in that layer is varied in steps having a predetermined increment, and a reflection spectrum is modeled. In a method of this kind, the wavelength range and the increments are specified in self-consistent fashion using a sensitivity criterion.
9 Citations
2 Claims
-
1. A method for determining layer thickness ranges of a plurality of layers (Si) of a specimen, the method comprises the steps of:
-
generating for each of the layers (Si), the wavelength-dependent refractive index and absorption functions ni(λ
) and ki(λ
), as well as starting values (Timin,0, Ti0, Timax,0) for a minimum, nominal, and maximum layer thickness;
measuring a reflection spectrum of the specimen in a wavelength range (λ
min, λ
max)smoothing the reflection spectrum by diminishing noise caused predominantly by external influences and thereby generating a smoothed reflection spectrum;
determining the number of extremes in the smoothed reflection spectrum, wherein two adjacent extremes, of which the one must be a minimum and the other a maximum, being evaluated only if they differ at least by a specified contrast criterion;
determining the layer thickness ranges based on the number of extremes, such that for each layer (Si) the thickness (Ti) is varied between (Timin,0) and (Timax,0) in steps having a predetermined increment (dTi), a reflection spectrum being modeled within the wavelength range (λ
min, λ
max), and the number of extremes is determined and stored and compared to the number of extremes in the smoothed reflection spectrum,specifying the wavelength range (λ
min, λ
max) and the increments (dTi) in self-consistent fashion by means of a sensitivity criterion, such that for each layer Si for a starting increment (dTi0) and in a wavelength range (λ
0min, λ
0max), a reflection modulation spectrum is modeled, the absolute value thereof is determined, and the resulting spectrum is then smoothed;
limiting the wavelength range to a continuous range (λ
imin, λ
imax), wherein (λ
imin) is the smallest and (λ
imax) is largest wavelength, for which the modeled spectrum still exhibits a value above or equal to a specified threshold value and the modeled spectrum substantially exhibits, for values between those wavelengths, values above the threshold value;
selecting (λ
min) as the minimum of all (λ
imin), and selecting (λ
max) as the maximum of all (λ
imax);
calculating the increments (dTi) for the wavelength range (λ
min, λ
max) and are in turn used to model a reflection modulation spectrum for each layer (Si); and
repeating until the wavelengths (λ
min) and (λ
max) change, between two repetitions, only by less than a specified self-consistency criterion, in which case the wavelengths (λ
min), (λ
max) and increments (dTi) are used for the further method steps. - View Dependent Claims (2)
-
Specification