**Optimization of Baker's Yeast Production on Date Extract Using Response Surface Methodology (RSM)**

**Mounira Kara Ali 1,\*, Nawel Outili <sup>2</sup> , Asma Ait Kaki <sup>3</sup> , Radia Cherfia <sup>1</sup> , Sara Benhassine <sup>1</sup> , Akila Benaissa <sup>2</sup> and Noreddine Kacem Chaouche <sup>1</sup>**


Received: 16 July 2017; Accepted: 3 August 2017; Published: 7 August 2017

**Abstract:** This work aims to study the production of the biomass of *S. cerevisiae* on an optimized medium using date extract as the only carbon source in order to obtain a good yield of the biomass. The biomass production was carried out according to the central composite experimental design (CCD) as a response surface methodology using Minitab 16 software. Indeed, under optimal biomass production conditions, temperature (32.9 ◦C), pH (5.35) and the total reducing sugar extracted from dates (70.93 g/L), *S. cerevisiae* produced 40 g/L of their biomass in an Erlenmeyer after only 16 h of fermentation. The kinetic performance of the *S. cerevisiae* strain was investigated with three unstructured models i.e., Monod, Verhulst, and Tessier. The conformity of the experimental data fitted showed a good consistency with Monod and Tessier models with *R <sup>2</sup>* = 0.945 and 0.979, respectively. An excellent adequacy was noted in the case of the Verhulst model (*R <sup>2</sup>* = 0.981). The values of kinetic parameters (*Ks*, *Xm*, *µm*, *p* and *q*) calculated by the Excel software, confirmed that Monod and Verhulst were suitable models, in contrast, the Tessier model was inappropriately fitted with the experimental data due to the illogical value of *Ks* (−9.434). The profiles prediction of the biomass production with the Verhulst model, and that of the substrate consumption using Leudeking Piret model over time, demonstrated a good agreement between the simulation models and the experimental data.

**Keywords:** *Saccharomyces cerevisiae*; biomass; date extract; optimization; response surface methodology; kinetic models
