Is there an explanation for high CPU spikes when reclaiming space from recyclebin?

We had a situation on a production system running on HP-UX 11.2.0.1.0 Enterprise edition.  Soon after midnight on a Saturday, the CPU spiked to 100% on the system.  Looking at the v$session table we had most of our sessions blocked with TX row lock contention, mutex, and library cache locks.  Our system switched over to the secondary node (causing the database to restart).  After the switchover the problems were not evident and the system operated as expected.  After running an AWR for the half hour right before the switchover, we saw three SQL statements in the CPU section of the report;