Methodology for extracting effective lens aberrations using a neural network

A method of extracting effective imaging system aberrations from test data collected from test structures constructed from a lithography system having an imaging system associated therewith, comprising the steps of:

• inputting experimental critical dimension data corresponding to fabricated features on a substrate to an intelligent system;
• inputting nominal critical dimension data corresponding to the fabricated features on the substrate to the intelligent system;
• and
• determining the effective aberrations of the imaging system associated with the lithography system used to fabricate the features using a calibrated neural network, and wherein calibrating the neural network comprises the steps of:
• (a) inputting an imaging system aberration condition to a lithography simulator;
• (b) performing a lithography simulation for a plurality of input features representing a set of nominal critical dimensions at a fixed set of lithography process conditions and at the imaging system aberration condition:
• (c) saving an output of the lithography simulation of step (b) in a memory, wherein the output comprises a plurality of simulated features corresponding to the input features having critical dimensions that differ from the nominal critical dimensions due to the aberration condition;
• (d) inputting another imaging system aberration condition to the lithography simulator;
• and
• (e) repeating steps (b), (c) and (d) a predetermined number of times, thereby generating and saving in the memory a training set of data representing critical dimensions of simulated features corresponding to the input features at a plurality of imaging system aberration conditions;
• (f) inputting the simulated features to the neural network;
• (g) inputting the input features to the neural network;
• (h) comparing an output of the neural network, which represents an imaging system aberration condition, with an expected imaging system aberration condition using the training set;
• (i) adjusting one or more interconnections of nodes within the neural network in response to the comparison;
• and
• (j) repeating steps (h) and (i) until the output is within a predetermined amount of the expected imaging system aberration condition.

A system for characterizing an imaging system for a projection-type photolithography system, comprising:

• a lithography simulator, wherein the lithography simulator is operable to simulate a lithography process step for one or more nominal features based on a set of lithography process conditions and one or more imaging system aberration conditions, and wherein the lithography simulator outputs one or more simulated features which vary with respect to the nominal features due to the imaging system aberration conditions;
• a first memory operatively coupled to the lithography simulator, wherein the memory stores input data and output data associated with one or more lithography simulations performed by the lithography simulator;
• a second memory which stores experimental test feature critical dimension data for one or more experimental test features which were formed using the imaging system being characterized;
• and
• an intelligent system operatively coupled to the first memory and the second memory, wherein the intelligent system receives the input data and the output data associated with the one or more lithography simulations and the experimental test feature critical dimension data and determines an effective imaging system aberration condition at one or more points in response thereto.

A system for characterizing an imaging system for a projection-type photolithography system, comprising:

• a lithography simulator, wherein the lithography simulator is operable to simulate a lithography process step for one or more nominal features based on a set of lithography process conditions and one or more imaging system aberration conditions, and wherein the lithography simulator outputs one or more simulated features which vary with respect to the nominal features due to the imaging system aberration conditions;
• a first memory operatively coupled to the lithography simulator, wherein the memory stores input data and output data associated with one or more lithography simulations performed by the lithography simulator;
• a second memory which stores experimental test feature critical dimension data for one or more experimental test features which were formed using the imaging system being characterized;
• and
• an expert system operatively coupled to the first memory and the second memory, wherein the intelligent system receives the input data and the output data associated with the one or more lithography simulations and the experimental test feature critical dimension data and determines an effective imaging system aberration condition at one or more points in response thereto, wherein the expert system contains a plurality of expert rules which are applied to the input data, output data and the experimental test feature critical dimension data to determine the effective imaging system aberration condition at the one or more points.

A system for characterizing an imaging system for a projection-type photolithography system, comprising:

• a lithography simulator, wherein the lithography simulator is operable to simulate a lithography process step for one or more nominal features based on a set of lithography process conditions and one or more imaging system aberration conditions, and wherein the lithography simulator outputs one or more simulated features which vary with respect to the nominal features due to the imaging system aberration conditions;
• a first memory operatively coupled to the lithography simulator, wherein the memory stores input data and output data associated with one or more lithography simulations performed by the lithography simulator;
• a second memory which stores experimental test feature critical dimension data for one or more experimental test features which were formed using the imaging system being characterized;
• and
• a neural network operatively coupled to the first memory and the second memory, wherein the intelligent system receives the input data and the output data associated with the one or more lithography simulations and the experimental test feature critical dimension data and determines an effective imaging system aberration condition at one or more points in response thereto, and wherein the intelligent system comprises a neural network, wherein the neural network uses the input data and the output data associated with the one or more lithography simulations as a training set to calibrate the neural network and thereby establish a transfer function which maps a deviation between one or more nominal features and the one or more experimental test features to an imaging system aberration condition which could have produced such deviation.