In this study, xylitol production of Candida tropicalis M2 was optimized in chestnut shell
hemicellulosic hydrolysate. Initial xylose concentration, initial concentration of ammonium
chloride which was chosen as a nitrogen source and shaking rate were selected as the independent
variables. Modelling the effects of the independent variables on maximum xylitol concentration
(Y1), xylitol yield (Y2), maximum xylitol production rate (Y3), percentage of substrate conversion
(Y4) and specific product formation rate (Y5) were performed by response surface methodology
using central composite design. Variations of the responses with the independent variables were
defined by a quadratic model except the Y4. According to ANOVA results, initial concentration
of ammonium chloride was found to be the most important parameter. Linear effect of the
initial xylose concentration was found as significant for only Y1. The effect of shaking rate was
very significant for the Y1 and Y4. By numeric optimization conducted; the conditions with
80 g/L initial xylose concentration, 3.761 g/L initial ammonium chloride concentration and
80 strokes/min of shaking rate were predicted as optimum conditions. In optimum conditions; Y1,
Y2, Y3, Y4 and Y5 were obtained as; 24.50 g/L, 31.517%, 0.497 g/L.h, 97.35% and 0.06 g/L.h.N,
respectively.