Mass integration is an efficient tool to prevent pollution derived from chemical processes. In recent years, several advances have been reported in methodologies for the design of mass exchange networks (MENs), using approaches as the pinch point and mathematical programming. Recently, simultaneous design and optimization methodologies have been proposed. The design methods based on simultaneous optimization offer the possibility of synthetizing MENs in a single step, formulating the synthesis problem as a mixed-integer nonlinear programming problem (MINLP). In this work, a MINLP model for the synthesis of mass exchange networks is presented. The model is based on a superstructure represented with disjunctions. The proposed superstructure is obtained from a state-task-network representation and is applied to a process for copper removal from an etching process. The obtained mass exchange network showed a lower total annual cost than the networks previously reported for the same process. (c) 2018 Elsevier Ltd. All rights reserved.