木糖醇為天然糖醇的一種，可用於食品、醫學以及保健食品，為用途廣泛的天然物。本研 究以frap/mfo BCRC21436做為實驗菌株，進行生產木糖醇最適化之探討。先以一次 一因子實驗探討C. frop/cafo發酵生產木糖醇實驗最適條件，結果發現C. frop/cato最適條件為 125 rpm、40°C、pH 6，木糖醇產量為25.03 g/L。以C. frap_ca/_s進行回應曲面法找尋最適生產 木糖醇之條件，將一次一因子所得條件做為中心點進行23因子設計實驗，經一階回應曲面實驗 後發現125 rpm、40C、pH 6仍為C. fropCafo生產木糖醇之最佳條件，以此點為原點進行陡升 路徑實驗，結果發現最高木糖醇產量點仍為原點，將此點做為中心混成實驗之原點進行尋找極 值點之實驗。結果發現二階回應曲面實驗設計之迴歸方程式無法適切表達此模式在實驗數據上 之適用性，因此本研究發現一次一因子實驗所得之最適條件，亦是一階回應曲面時所得之最佳 條件，此為目前以C. fropCato生產木糖醇之最適條件，此條件應可作為後續以生物法進行工 業化大量生產木糖醇之參考依據。

In this study, Candida tropicalis BCRC 21436 was used as an experimental strain to be optimized for the production of xylitol. The optimal conditions for the production of xylitol from C. tropicalis were studied using the one-factor-at-a-time method. The optimal conditions for C. tropicalis were found to be an agitation speed of 125 rpm, a temperature of 40°C, and a pH of 6. These conditions resulted in xylitol production of 25.03 g/L. In addition to the one-factor-at-a-time method, the response surface method (RSM) was used to optimize the production of xylitol by this strain. In the 23 factorial design experiment, the linear temperature term was demonstrated to have a significant effect on xylitol production, Furthermore, the first-order with interaction model proved adequate (R2 = 0.997). The optimal condition was then assessed using the steepest ascent method and central composite design. However, the regression equation of the second-order response surface experiment could not fit the applicability of this model for the experimental data. The one-factor-at-a-time method and the RSM method both indicated that 125 rpm, 40°C, and pH 6 were the optimal conditions for xylitol production by C. tropicalis. This is by far the most suitable set of conditions for the production of xylitol from C. tropicalis. These conditions should be used as a reference point for subsequent industrial production of xylitol through biological methods.