Can PostgreSQL do on-demand synchronization during replication?

Question

PostgreSQL

Essence in the following:
There is a table in Postgre of 130 million records which is updated from the first and up to the last record in one iteration of a cycle. The idea is to delete all indexes before each iteration of the loop to increase the collection speed and add them again at the end of the iteration (experimentally, I found out that when collecting immediately with indexes, their size was 2 times larger than after reindex). But, if you delete the indexes on the master, then they will be deleted on the slaves as well. Is it possible to make it so that synchronization takes place at the end of each iteration on the master? Synchronize on demand, so to speak. Or maybe there is some flag on the master that will prevent the slaves from synchronizing? What are the solutions for such a task?