A split lot sends some wafers of one lot through a test condition and the rest through the baseline, for example half the slots with the new etch recipe and half with the process of record. Because all wafers share one start, one history and one test setup, lot-to-lot noise drops out of the comparison and small effects become visible. Splits are the fab's everyday designed experiment.
Split lot came back clean: odd slots on the new liner, even slots on POR, no difference in via resistance.
Interleave odd and even slots. A top-half versus bottom-half split confounds the treatment with slot position in batch tools.
Also heard as
- split
- wafer split
Related terms
Block
Group of experimental runs made under similar conditions, used to remove a known nuisance source such as day, batch or tool.
t-test
Test that compares a mean against a target, or two means against each other, using the t distribution.
Confounding
Mixing of two effects so that the data cannot separate them, whether by design or through an uncontrolled variable.
DOE
Design of experiments: planning a set of runs that vary several factors at once so their effects can be separated.