Domain Decomposition Methods in Optimal Control of Partial Differential Equations