This work is focused on the optimization of a reverse osmosis network proposed for the ultrapurification of chemicals, from technical grade to semiconductor grades demanded in electronic applications that require the use of high-purity wet chemicals in the semiconductor manufacturing, as it is the case of hydrogen peroxide that is commonly used in the wafer cleaning and surface conditioning processes. An industrial installation able to produce simultaneously the five different Semiconductor Equipment and Materials International Grades of hydrogen peroxide is represented by a simplified superstructure configuration formulated as a nonlinear programming problem. The network integrates reverse osmosis membrane modules, mixers, and split functions, defined by the equations of mass balances and the KedemKatchalsky transport model for the description of the permeate flux and the metal rejection coefficients at each membrane stage. The objective of the design is to maximize the daily profit obtained from the sale of the electronic grades of hydrogen peroxide produced by the ultrapurification system. (C) 2012 American Institute of Chemical Engineers AIChE J, 2012