Design of experiments

The tests were aimed to assess the performance of the PBBR under a wide range of dilution rate and by feeding the reactor with stream bearing a spectrum of substrate. The D was quasi-steadily increased in each experimental set: the D was increased at the new value, close to the previous one, and it was kept constant until a steady-state condition established.

The set of experiments was carried out by feeding the PBBR, with a glucose-based medium (glucose concentration set at 50 g L⁻¹) and the dilution rate was set between 0.5 and 2.4 h⁻¹.

The second set of experiments was aimed to adapt the cells to a xylose-based medium, the main pentose sugar present in a lignocellulosic hydrolysate. The feeding was a solution of glucose and xylose at percentage of xylose progressively increased from 0 up to 100%. The total sugar concentration (glucose + xylose) in the feeding was set at 50 g L⁻¹. The dilution rate was set at 1.24 h⁻¹.

After evaluating how the PBBR performances changed increasing the xylose concentration in the media, the effect of the dilution rate on the succinic acid production by using a xylose-based medium was investigated. For this set of experiments, the xylose concentration was set to 40 g L⁻¹ and the dilution rate was ranged between 0.5 and 1.44 h⁻¹.

The fourth set of experiments was carried out by feeding the PBBR with a synthetic medium that mime the composition of a lignocellulosic hydrolysate (inhibitor-free). The feeding was a solution of glucose, arabinose and xylose (GAX): the total sugar concentration was set at 80 g L⁻¹ and the mass ratio between the sugars was set to at 55:15:30 [24]. The dilution rate was set between 0.7 and 1.44 h⁻¹.

A D-jumping strategy was adopted to assess the repeatability of the biofilm reactor performance with respect to the dilution rate. Provided that the reactor steady-state at the prefixed value of D = D^*, the feeding stream rate was changed to operate the reactor at a D equal to a fraction of the previous one (say D⁺) and close to a D value already investigated. The performances of the reactor were measured until the new steady-state condition established and they were compared with those measured under the previous D⁺. The comparison of the performances of the biofilm reactor assessed by changing D according to the two strategies (quasi-steadily increase vs. D-jumping) pointed out that the performances of the biofilm reactor depended only of the D and not on the D-tuning strategy.

The last set of experiments was carried out to investigate the effects of the two principal inhibitors: furfural and HMF [25], found in lignocellulosic hydrolysate on the fermentation process. The concentration of the inhibitors in the feeding was set at 1 and 0.28 g L⁻¹ [26], for furfural and HMF, respectively. The dilution rate was set at 0.75 and 1.00 h⁻¹.