I think you have two options:
1. Use "merge" but then you need the full syntax for "merge" which expects an update clause and an insert clause. You may be able to use a "where" clause in the update so it doesn't actually do anything.
2. Use a PL\SQL procedure with a cursor loop to fetch each row, then check if it already exists, or not, in the destination table. If yes, do nothing, if not, do an insert. (This will be faster than trying an insert in all cases, but catching duplicates via a unique or primary key violation exception.)