In this paper, the new results obtained with the development of a novel thermodynamic equilibrium-inspired optimization algorithm are presented. This technique was developed in order to solve nonlinear optimization problems, with continuous domains. In our proposal, each variable is considered as the most volatile chemical component of a saturated binary liquid mixture, at a determined pressure and temperature. In the search process, the new value of each decision variable is obtained at some temperature of bubble or dew of the binary system. The search includes the random change of the chemical species and their compositions. The algorithm has being tested by using well-known mathematical functions as benchmark functions and has given competitive results in comparison with other metaheuristics.