106 Journal of the Tunisian Chemical Society, 2016, 18, 106-113 * Corresponding author, e-mail address : hajerdouahem@gmail.com - Tel: +216.79.32.54.70; Fax: +216.79.32.53.14 INTRODUCTION Phosphogypsum is an industrial waste produced by phosphate industry [1]. The manufacturing 1 ton of phosphoric acid produces 5 tons of Phosphogypsum as a by-product. Accordingly, the annual word production of phosphogypsum exceeds 200 million tons [2]. This waste causes many environmental problems since it contains metals in high concentration [3]. Despite these drawbacks, it is accumulated in large stockpiles and occupies huge land areas. The phosphogypsum causes a very serious problem, especially when it is considered over production, for the damage that it causes to the environment due to its negative impact on water, air and soil [1]. Several efforts of valorization have been conducted by researchers in order to solve this environmental problem. Recently, many research works have studied the possibility of valorizing this waste and several commercial applications have been found [4]; its use as a fertilizer in agriculture, a building material, an amendment for soil stabilization and a coating pigment in the paper industry. Modeling and optimization of phosphogypsum transformation into calcium fluoride using experimental design methodology Hajer Douahem * ; Halim Hammi; Ahmed Hichem Hamzaoui; Adel M’nif Useful Materials Valorization Laboratory National Centre of Research in Materials Science; Technological Park of Borj Cedria B.P.73–8027.Soliman. TUNISIE (Received: 11 January 2016, accepted: 24 February 2016) Abstract: In this work, the experimental design methodology has been used for the modeling and optimization of phosphogypsum transformation into calcium fluoride and sodium sulfate. For this reason a four factors box behnken design has been chosen to be used. The studied factors were reaction time (X 1 ), NaF/phosphogypsum ratio (X 2 ), mass of phosphogypsum (X 3 ) and stirring speed (X 4 ) while Y 1 (the calcium yield) and Y 2 (the purity of the formed precipitate) are selected as the responses variables. A compromise between the studied responses has been determined and the optimum conditions for this reaction have been fixed at a reaction time of 97.80 min, a NaF/phosphogypsum ratio equal to 2.06, with 8.13 g of phosphogypsum mass and a stirring speed of 506.39 rpm. At these experimental conditions the calcium yield is 95.61% and the purity of the formed precipitate is 98.86%. Keywords: phosphogypsum; box behnken design; Response surface methodology; optimization Today’s priority is to try to find new applications of phosphogypsum in order to minimize its disposal and stockage [5] but there is a big problem in phosphogypsum because it contains many impurities [6]. Singh et all declared that they can solve this problem by washing phosphogypsum [7] or by treatment with ammonium hydroxide [8] or with aqueous ammonium sulfate [9] and a mixture of sulfuric acid and silica [10] before use. Despite the purification of phosphogypsum and the previously- mentioned applications, problems caused by the stockage of phosphogypsum have not been completely solved. Therefore this work aim to study the transformation of phosphogypsum into CaF 2 and Na 2 SO 4 by adopting the response surface methodology which is an efficient statistical method that has been successfully used in testing process parameters and their interactive effect [11-14]. Firstly, a box behnken experimental design was used to determine the most significant factors and the mathematical model that predict the response