Distributed Reinforcement Learning for Molecular Design: Antioxidant case