An implicate is the dual of an implicant. A sum term S is an implicate of F if S = 0 forces F = 0. In other words, every row where S is 0 is a 0 of F. (Equivalently, F implies S: whenever F = 1, S = 1.)
A prime implicate is an implicate that stops being one if you remove any literal. Removing a literal from a sum doubles the rows where it is 0, and for a prime implicate that always takes in a 1 of F.
They play exactly the role for POS that prime implicants play for SOP:
- Every factor of a minimal pos is a prime implicate.
- You find them by combining adjacent maxterms, the 0s of F.
- You then choose a cheapest set that covers every 0.
So on a Karnaugh map, grouping the 0s finds prime implicates, just as grouping the 1s finds prime implicants. Many courses simply say "group the 0s" and skip the word, but you will meet it in textbooks.