Polishing semiconductor wafers using causal models
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for optimizing a process of polishing semiconductor wafers. In one aspect, the method comprises repeatedly performing the following: i) selecting a configuration of input settings for polishing a semiconductor wafer, based on a causal model that measures current causal relationships between input settings and a quality of semiconductor wafers; ii) receiving a measure of the quality of the semiconductor wafer polished with the configuration of input settings; and iii) adjusting, based on the measure of the quality of the semiconductor wafer polished with the configuration of input settings, the causal model.
1 . A method comprising:
repeatedly performing the following:
selecting a configuration of input settings for polishing a semiconductor wafer, based on a causal model mapped to a distribution over possible control settings such that the causal model measures current causal relationships between input settings and a quality of semiconductor wafers, wherein control settings are assigned to different procedural instances that allow blocked groups to later be identified in order to compute impact measurements between blocked groups, wherein probability matching is utilized to map the impact measurements and confidence intervals in the causal model to probabilities and the control settings are assigned based on these probabilities;
receiving a measure of the quality of the semiconductor wafer polished with the configuration of input settings;
adjusting, based on the measure of the quality of the semiconductor wafer polished with the configuration of input settings, the causal model; and
re-computing the causal model by computing overall impact measurements based on confidence intervals around the overall impact measurements and means of d-scores.
2 . The method of claim 1 , wherein:
selecting a configuration of input settings comprises selecting the configuration of input settings based on the causal model and a set of internal control parameters, and
the method further comprises adjusting the internal control parameters based on the measure of the quality of the semiconductor wafer polished with the configuration of input settings.
3 . The method of claim 1 , wherein the measure of quality of the semiconductor wafer comprises one or more of:
one or more measures related to removing material from the semiconductor wafer;
a final flatness of the semiconductor wafer;
a planarization efficiency of the semiconductor wafer;
a uniformity of features of the semiconductor wafer;
a measure of variations within the semiconductor wafer;
a measure of variations across multiple semiconductor wafers in a batch of semiconductor wafers;
a measure of defects of the semiconductor wafer;
one or more measures related to production yields; or
one or more measures related to production costs.
4 . The method of claim 3 , wherein the measures related to removing material from the semiconductor wafer comprise one or more of:
an average rate of removal of the material;
a standard deviation of the rate of removal of the material;
a spatial uniformity of the removal of the material; or
a removal rate profile of the semiconductor wafer.
5 . The method of claim 3 , wherein the measures related to production yields comprise one or more of:
a total yield;
a time to complete one or more predefined steps; or
a frequency of part rejection.
6 . The method of claim 3 , wherein the measures related to production costs comprise one or more of:
a production time;
a burn rate of consumables; or
a measure of inventory management.
7 . The method of claim 1 , wherein the input settings comprise one or more of:
one or more settings related to a polishing pad;
one or more settings related to the semiconductor wafer;
one or more settings related to a conditioner; or
one or more settings related to a slurry.
8 . The method of claim 7 , wherein the settings related to a polishing pad comprise one or more of:
a speed of the polishing pad;
a decision regarding when to change the polishing pad; or
a type of the polishing pad.
9 . The method of claim 7 , wherein the settings related to the semiconductor wafer comprise one or more of:
a speed of the semiconductor wafer; or
a pressure for one or more zones of the semiconductor wafer; or
one or more wafer sweep parameters.
10 . The method of claim 7 , wherein the settings related to a conditioner comprise one or more of:
a speed of the conditioner; or
one or more conditioner sweep parameters.
11 . The method of claim 7 , wherein the settings related to a slurry comprise one or more of:
a slurry flow rate; or
a type of slurry used.
12 . The method of claim 1 , wherein:
selecting the configuration of input settings comprises selecting the configuration based on the casual model and respective measures of a predetermined set of external variables; and
adjusting the internal control parameters comprises adjusting internal control parameters that parameterize an impact of the predetermined set of external variables on the selecting of the configuration.
13 . The method of claim 12 , wherein the predetermined set of external variables comprises one or more of:
properties of one or more materials used during polishing;
one or more variables related to a film;
one or more variables related to a slurry;
one or more variables related to a polishing pad;
one or more variables related to a conditioner;
an amount of debris generated or accumulated;
one or more variables related to consumables or other components; or
one or more variables related to a surrounding environment during polishing.
14 . The method of claim 13 , wherein the variables related to a film comprise one or more of:
a composition of the film;
a thickness of the film; or
a topography of the film.
15 . The method of claim 13 , wherein the variables related to a slurry comprise one or more of:
a brand of the slurry;
variables related to a batch of the slurry;
a type of the slurry;
a composition of the slurry; or
a dilution level of the slurry.
16 . The method of claim 13 , wherein the variables related to a polishing pad comprise one or more of:
a brand of the polishing pad;
variables related to a batch of the polishing pad;
a diameter of the polishing pad;
a construction of the polishing pad;
a groove pattern of the polishing pad;
a material of the polishing pad; or
a stiffness of the polishing pad.
17 . The method of claim 13 , wherein the variables related to a conditioner comprise one or more of:
a brand of the conditioner;
variables related to a batch of the conditioner;
a diameter of the conditioner;
a construction of the conditioner;
a grit of the conditioner;
a material of the conditioner; or
a stiffness of the conditioner.
18 . The method of claim 13 , wherein the variables related to a surrounding environment comprise one or more of:
a temperature during polishing;
a humidity during polishing; or
a level of lighting during polishing.
19 . A method comprising:
repeatedly performing the following:
obtaining a configuration of input settings for polishing a semiconductor wafer;
determining a measure of a quality of the semiconductor wafer polished with the configuration of input settings; and
providing the measure of the quality of the semiconductor wafer polished with the configuration of input settings to a system that updates a causal model mapped to a distribution over possible control settings such that the causal model that measures current causal relationships between the input settings and the quality of semiconductor wafers, wherein control settings are assigned to different procedural instances that allow blocked groups to later be identified in order to compute impact measurements between blocked groups, wherein probability matching is utilized to map the impact measurements and confidence intervals in the causal model to probabilities and the control settings are assigned based on these probabilities; and
re-computing the causal model by computing overall impact measurements based on confidence intervals around the overall impact measurements and means of d-scores.
20 . A system comprising one or more computers and one or more storage devices storing instructions that when executed by the one or more computers cause the one or more computers to repeatedly perform the following operations:
select a configuration of input settings for polishing a semiconductor wafer, based on a causal model mapped to a distribution over possible control settings such that the causal model measures current causal relationships between input settings and a quality of semiconductor wafers, wherein control settings are assigned to different procedural instances that allow blocked groups to later be identified in order to compute impact measurements between blocked groups, wherein probability matching is utilized to map the impact measurements and confidence intervals in the causal model to probabilities and the control settings are assigned based on these probabilities;
receive a measure of the quality of the semiconductor wafer polished with the configuration of input settings;
adjust, based on the measure of the quality of the semiconductor wafer polished with the configuration of input settings, the causal model; and
re-compute the causal model by computing computing overall impact measurements based on confidence intervals around the overall impact measurements and means of d-scores.