Methods and mechanisms for process recipe optimization

1 . A method, comprising:

performing, by manufacturing equipment, a first process on a first substrate according to a process recipe, wherein the process recipe comprises a plurality of setting parameters, wherein the first process generates a plurality of features on a surface of the first substrate;

generating metrology data associated with the plurality of features;

inputting the metrology data and at least one setting parameter of the plurality of setting parameters into one or more Bayesian probabilistic models;

receiving an output from the one or more Bayesian probabilistic models, wherein the output comprises output data indicative of correlations between the at least one setting parameter and the metrology data;

obtaining at least one target parameter comprising a physical attribute of a feature of the plurality of features;

generating, based on the output data and the at least one target parameter, a Pareto frontier including at least one set of Pareto efficient manufacturing settings;

selecting a value associated with the Pareto frontier which is closest to a curve indicative of a setting of target parameters;

updating the process recipe by modifying the at least one setting parameter of the plurality of setting parameters using the selected value; and

performing, by the manufacturing equipment, a second process on a second substrate according to the updated process recipe.

2 . The method of claim 1 , wherein the features are etch features that comprise a set of mandrels.

3 . The method of claim 1 , further comprising:

updating the one or more Bayesian probabilistic models based on at least one of the updated process recipe or metrology data associated with the second process.

4 . The method of claim 1 , further comprising:

generating the one or more of the Bayesian probabilistic models using a design of experiment (DoE) technique.

5 . The method of claim 1 , further comprising:

performing a dimensionality reduction operation on the metrology data.

6 . An electronic device manufacturing system, comprising:

a memory device; and

a processing device, operatively coupled to the memory device, to perform operations comprising:

causing manufacturing equipment to perform a first process on a first substrate according to a process recipe, wherein the process recipe comprises a plurality of setting parameters, wherein the first process generates a plurality of features on a surface of the first substrate;

generating metrology data associated with the plurality of features;

inputting the metrology data and at least one setting parameter of the plurality of setting parameters into one or more Bayesian probabilistic models;

receiving an output from the one or more Bayesian probabilistic models, wherein the output comprises output data indicative of correlations between the at least one setting parameter and the metrology data;

obtaining at least one target parameter comprising a physical attribute of a feature of the plurality of features;

generating, based on the output data and the at least one target parameter, a Pareto frontier including at least one set of Pareto efficient manufacturing settings;

selecting a value associated with the Pareto frontier which is closest to a curve indicative of a setting of target parameters;

updating the process recipe by modifying the at least one setting parameter of the plurality of setting parameters using the selected value; and

causing the manufacturing equipment to perform a second process on a second substrate according to the updated process recipe.

7 . The electronic device manufacturing system of claim 6 , wherein the features are etch features that comprise a set of mandrels.

8 . The electronic device manufacturing system of claim 6 , wherein the processing device is further to perform operations comprising:

updating the one or more Bayesian probabilistic models based on at least one of the updated process recipe or metrology data associated with the second process.

9 . The electronic device manufacturing system of claim 6 , wherein the processing device is further to perform operations comprising:

generating the one or more of the Bayesian probabilistic models using a design of experiment (DoE) technique.

10 . The electronic device manufacturing system of claim 6 , wherein the processing device is further to perform operations comprising:

performing a dimensionality reduction operation on the metrology data.

11 . A non-transitory computer-readable storage medium comprising instructions that, when executed by a processing device operatively coupled to a memory, performs operations comprising:

causing manufacturing equipment to perform a first process on a first substrate according to a process recipe, wherein the process recipe comprises a plurality of setting parameters, wherein the first process generates a plurality of features on a surface of the first substrate;

generating metrology data associated with the plurality of features;

inputting the metrology data and at least one setting parameter of the plurality of setting parameters into one or more Bayesian probabilistic models;

receiving an output from the one or more Bayesian probabilistic models, wherein the output comprises output data indicative of correlations between the at least one setting parameter and the metrology data;

obtaining at least one target parameter comprising a physical attribute of a feature of the plurality of features;

generating, based on the output data and the at least one target parameter, a Pareto frontier including at least one set of Pareto efficient manufacturing settings;

selecting a value associated with the Pareto frontier which is closest to a curve indicative of a setting of target parameters;

updating the process recipe by modifying the at least one setting parameter of the plurality of setting parameters using the selected value; and

causing the manufacturing equipment to perform a second process on a second substrate according to the updated process recipe.

12 . The non-transitory computer-readable storage medium of claim 11 , wherein the features are etch features that comprise a set of mandrels.

13 . The non-transitory computer-readable storage medium of claim 11 , wherein the processing device is further to perform operations comprising:

updating the one or more Bayesian probabilistic models based on at least one of the updated process recipe or metrology data associated with the second process.

14 . The non-transitory computer-readable storage medium of claim 11 , wherein the processing device is further to perform operations comprising:

generating the one or more of the Bayesian probabilistic models using a design of experiment (DoE) technique.