Wafer bin map based root cause analysis

1. A method comprising:

obtaining a first wafer bin map for a first semiconductor wafer manufactured in a current production run;

identifying a product flow for the first semiconductor wafer;

retrieving a plurality of prior wafer bin maps corresponding to prior production runs for semiconductor wafers manufactured using the identified product flow, and loading the plurality of prior wafer bin maps into a user interface;

setting boundary conditions for a plurality of selected wafer parameters;

comparing a first pattern on the first wafer bin map to a plurality of prior patterns respectively associated with the plurality of prior wafer bin maps in order to identify a subset of the plurality of prior wafer bin maps that have patterns similar to the first pattern of the first wafer bin map on the basis of the boundary conditions for the plurality of selected wafer parameters;

determining a probability that a root cause associated with at least some of the subset of the plurality of prior wafer bin maps is a primary root cause for a defect associated with the first pattern on the first wafer bin map; and

when the probability exceeds a threshold, assigning a label to the first wafer bin map indicating that the first root cause is the primary root cause.

2. The method of claim 1 , further comprising:

manually setting the boundary conditions for the selected wafer parameters via the user interface.

3. The method of claim 1 , further comprising:

automatically setting the boundary conditions for the selected wafer parameters.

4. The method of claim 3 , further comprising:

tuning the set of boundary conditions in order to identify a first number of neighboring wafers as the subset of prior wafer bin maps.

5. The method of claim 1 , further comprising:

determining a first statistical distance dc between the first semiconductor wafer and a plurality of neighboring wafers that is considered close enough to associate a root cause of the neighboring wafers with the first semiconductor wafer.

6. The method of claim 5 , further comprising:

determining a second statistical distance df between the first semiconductor wafer and the plurality of neighboring wafers that is considered too far away to associate the root cause.

7. The method of claim 1 , the comparing step further comprising:

determining a first statistical distance dc between the first semiconductor wafer and a plurality of neighboring wafers that is considered close enough to associate a root cause of the neighboring wafers with the first semiconductor wafer;

determining a second statistical distance df between the first semiconductor wafer and the plurality of neighboring wafers that is considered too far away to associate the root cause of the neighboring wafers with the first semiconductor wafer;

weighting a first number k of the plurality of neighboring wafers equally when all k neighboring wafers are within the distance dc;

assigning the root cause of the k neighboring wafers to the first semiconductor wafer when all k neighboring wafers have the same root cause label; and

assigning a plurality of root causes of the k neighboring wafers to the first semiconductor wafer in proportion to a frequency of occurrence when all k neighboring wafers do not have the same root cause label.

8. The method of claim 1 , further comprising:

retrieving a second plurality of prior wafer bin maps corresponding to prior production runs for semiconductor wafers manufactured using a similar product flow to the identified product flow; and

scaling the retrieved wafer bin maps as necessary.

9. The method of claim 1 , further comprising:

displaying the first wafer bin map and at least some of the subset of prior wafer bin maps having patterns similar to the first wafer bin map on the user interface; and

providing a first set of user controls to enable drilling down into additional wafer data for the first semiconductor wafer and the subset of semiconductor wafers manufactured in prior production runs.

10. The method of claim 9 , further comprising:

providing a second set of user controls to override and update a dominant root cause or secondary root causes associated with the first semiconductor wafer.

11. The method of claim 1 , the comparing step further comprising:

evaluating possible matches for the first wafer bin map based on clustering of the prior patterns of the plurality of prior wafer bin maps;

generating a statistical distribution for a plurality of calculated distances between the first semiconductor wafer and a plurality of neighboring wafers based on the clustering;

determining a first statistical distance dc between the first semiconductor wafer and a plurality of neighboring wafers that is considered close enough to associate a root cause of the neighboring wafers with the first semiconductor wafer; and

determining a second statistical distance df between the first semiconductor wafer and the plurality of neighboring wafers that is considered too far away to associate the root cause of the neighboring wafers with the first semiconductor wafer.

12. The method of claim 1 , the comparing step further comprising:

calculating zonal bin deltas for the prior patterns of the plurality of prior wafer bin maps;

generating a statistical distribution for a plurality of calculated distances between the first semiconductor wafer and a plurality of neighboring wafers based on the zonal bin deltas;

determining a first statistical distance dc between the first semiconductor wafer and a plurality of neighboring wafers that is considered close enough to associate a root cause of the neighboring wafers with the first semiconductor wafer; and

determining a second statistical distance df between the first semiconductor wafer and the plurality of neighboring wafers that is considered too far away to associate the root cause of the neighboring wafers with the first semiconductor wafer.

13. The method of claim 1 , the comparing step further comprising:

convolving the prior patterns of the plurality of prior wafer bin maps;

generating a statistical distribution for a plurality of calculated distances between the first semiconductor wafer and a plurality of neighboring wafers based on the convolved patterns;

determining a first statistical distance dc between the first semiconductor wafer and a plurality of neighboring wafers that is considered close enough to associate a root cause of the neighboring wafers with the first semiconductor wafer; and

determining a second statistical distance df between the first semiconductor wafer and the plurality of neighboring wafers that is considered too far away to associate the root cause of the neighboring wafers with the first semiconductor wafer.

14. A method comprising:

receiving a first wafer bin map for a first semiconductor wafer manufactured in a current production run;

retrieving a plurality of prior wafer bin maps corresponding to prior production runs for semiconductor wafers;

computing a plurality of statistical distances between respective ones of the prior wafer bin maps and the first wafer bin map;

assigning a single root cause to the first wafer bin map when each one of a subset of the plurality of prior wafer bin maps has a computed statistical distance less than a threshold and each one of the subset has the single root cause; and

assigning a plurality of root causes to the first wafer bin map when each one of the subset has a computed statistical distance less than the threshold but each one of the subset does not have the first root cause, each of the plurality of root causes being assigned as a function of the computed statistical distance between respective ones of the prior wafer bin maps and the first wafer bin map.

15. The method of claim 14 , further comprising, prior to computing the statistical distances, clustering the prior wafer bin maps.

16. The method of claim 14 , further comprising, prior to computing the statistical distances, calculating zonal bin deltas for the prior wafer bin maps.

17. The method of claim 14 , further comprising, prior to computing the statistical distances, convolving the prior wafer bin maps.

18. The method of claim 14 , further comprising:

identifying a first product flow for the first semiconductor wafer; and

retrieving the plurality of prior wafer bin maps corresponding to prior production runs having the first product flow.

19. The method of claim 14 , further comprising:

identifying a first product flow for the first semiconductor wafer; and

retrieving the plurality of prior wafer bin maps corresponding to prior production runs having a product flow similar to the first product flow.

20. The method of claim 14 , further comprising:

setting wafer limits for a plurality of selected wafer parameters;

comparing the first wafer bin map to the plurality of prior wafer bin maps in order to identify a first group of the plurality of prior wafer bin maps that are similar to the first wafer bin map on the basis of the wafer limits; and

computing the plurality of statistical distances between the first group of the prior wafer bin maps and the first wafer bin map.