Virtual metrology for wafer result prediction

A method for wafer result prediction and root cause analysis, comprising:

• before a failure occurs, executing a plasma etching process;
• collecting first data of the plasma etching process:
• determining predictor parameters of the plasma etching process using domain knowledge including knowledge of the plasma etching process, a plasma etching tool associated with the plasma etching process, a metrology tool, and/or the wafer;
• removing collinearity among the predictor parameters to obtain key predictor parameters;
• collecting metrology data of the wafer using the metrology tool, the metrology data related to the plasma etching process;
• selecting a subset of the key predictor parameters based on the metrology data of the wafer;
• building a virtual metrology (VM) model on the subset of the key predictor parameters;
• and
• predicting wafer results of the plasma etching process using the VM model,
• wherein determining the predictor parameters comprises determining a first subgroup of predictor parameters using domain knowledge, the first subgroup of predictor parameters expected to affect wafer results;
• and determining a second subgroup of predictor parameters based on Design of Experiments (DOE) and domain knowledge, the second subgroup of predictor parameters being a subset of the first subgroup of predictor parameters and known to affect wafer results,
• after the failure occurs, executing the plasma etching process;
• and
• collecting second data of the plasma etching process;
• identifying a parameter associated with a root cause of the failure using the VM model, the first data and the second data;
• fixing the failure by adjusting the parameter;
• and
• executing the plasma etching process using the parameter adjusted.