There is provided a learning method. The method includes performing preprocessing on light emission data in a chamber of a plasma processing apparatus, setting a constraint for generating a regression equation representing a relationship between an etching rate of the plasma processing apparatus and the light emission data, selecting a learning target wavelength from the light emission data subjected to the preprocessing, and receiving selection of other sensor data different from the light emission data. The method further includes generating a regression equation based on the set constraint while using, as learning data, the selected wavelength, the received other sensor data, and the etching rate, and outputting the generated regression equation.

There is provided a learning method. The method includes performing preprocessing on light emission data in a chamber of a plasma processing apparatus, setting a constraint for generating a regression equation representing a relationship between an etching rate of the plasma processing apparatus and the light emission data, selecting a learning target wavelength from the light emission data subjected to the preprocessing, and receiving selection of other sensor data different from the light emission data. The method further includes generating a regression equation based on the set constraint while using, as learning data, the selected wavelength, the received other sensor data, and the etching rate, and outputting the generated regression equation.

An optical manufacturing process sensing and status indication system is taught that is able to utilize optical emissions from a manufacturing process to infer the state of the process. In one case, it is able to use these optical emissions to distinguish thermal phenomena on two timescales and to perform feature extraction and classification so that nominal process conditions may be uniquely distinguished from off-nominal process conditions at a given instant in time or over a sequential series of instants in time occurring over the duration of the manufacturing process. In other case, it is able to utilize these optical emissions to derive corresponding spectra and identify features within those spectra so that nominal process conditions may be uniquely distinguished from off-nominal process conditions at a given instant in time or over a sequential series of instants in time occurring over the duration of the manufacturing process.

An optical manufacturing process sensing and status indication system is taught that is able to utilize optical emissions from a manufacturing process to infer the state of the process. In one case, it is able to use these optical emissions to distinguish thermal phenomena on two timescales and to perform feature extraction and classification so that nominal process conditions may be uniquely distinguished from off-nominal process conditions at a given instant in time or over a sequential series of instants in time occurring over the duration of the manufacturing process. In other case, it is able to utilize these optical emissions to derive corresponding spectra and identify features within those spectra so that nominal process conditions may be uniquely distinguished from off-nominal process conditions at a given instant in time or over a sequential series of instants in time occurring over the duration of the manufacturing process.

The present invention is enclosed in the area of machine learning, in particular machine learning for the analysis of High or Super-resolution spectroscopic data, which typically comprises analysis of highly complex samples/mixtures of substances and/or data with low resolution, for instance Laser-Induced Breakdown Spectroscopy (LIBS). It is an object of the present invention a method of computational self-learning for characterization of one or more constituents in a sample, from electromagnetic spectral information of such sample, which changes the paradigm associated with prior art methods, by using only sub-optical spectral information, i.e., obtaining the resolution of the spectral information and thereby be able to extract spectral lines—thus determining a spectral line position—from such spectral information, hence avoiding all the uncertainty associated with pixel based methods. It is also an object of the present invention a computational apparatus configured to implement such method.

The present invention is enclosed in the area of machine learning, in particular machine learning for the analysis of High or Super-resolution spectroscopic data, which typically comprises analysis of highly complex samples/mixtures of substances and/or data with low resolution, for instance Laser-Induced Breakdown Spectroscopy (LIBS). It is an object of the present invention a method of computational self-learning for characterization of one or more constituents in a sample, from electromagnetic spectral information of such sample, which changes the paradigm associated with prior art methods, by using only sub-optical spectral information, i.e., obtaining the resolution of the spectral information and thereby be able to extract spectral lines—thus determining a spectral line position—from such spectral information, hence avoiding all the uncertainty associated with pixel based methods. It is also an object of the present invention a computational apparatus configured to implement such method.

Provided is an inductively coupled plasma spectrometric system for measuring an emission state of plasma into which a measurement target sample is fed, the inductively coupled plasma spectrometric system including: a spectrometer configured to resolve light emitted in a measurement region set in the plasma into a plurality of wavelength components; a detection device configured to detect a spatial distribution of the resolved light; and a measuring device configured to measure the detected spatial distribution at every measurement unit time, the measurement unit time being at least shorter than time required for the sample to pass through the measurement region.

Provided is an inductively coupled plasma spectrometric system for measuring an emission state of plasma into which a measurement target sample is fed, the inductively coupled plasma spectrometric system including: a spectrometer configured to resolve light emitted in a measurement region set in the plasma into a plurality of wavelength components; a detection device configured to detect a spatial distribution of the resolved light; and a measuring device configured to measure the detected spatial distribution at every measurement unit time, the measurement unit time being at least shorter than time required for the sample to pass through the measurement region.

Disclosed are embodiments of an improved apparatus and system, and associated methods for optically diagnosing a semiconductor manufacturing process. A hyperspectral imaging system is used to acquire spectrally-resolved images of emissions from the plasma, in a plasma processing system. Acquired hyperspectral images may be used to determine the chemical composition of the plasma and the plasma process endpoint. Alternatively, a hyperspectral imaging system is used to acquire spectrally-resolved images of a substrate before, during, or after processing, to determine properties of the substrate or layers and features formed on the substrate, including whether a process endpoint has been reached; or before or after processing, for inspecting the substrate condition.

An apparatus to determine occurrence of an anomalous plasma event occurring at or near a process station of a multi-station integrated circuit fabrication chamber is disclosed. In particular embodiments, optical emissions generated responsive to the anomalous plasma event may be detected by at least one photosensor of a plurality of photosensors. A processor may cooperate with the plurality of photosensors to determine that the anomalous plasma event has occurred at or near by a particular process station of the multi-station integrated circuit fabrication chamber.