Microorganism: Zymomonas mobilis NRRL-B14023 was used in this study. Stock culture was incubated in YM liquid medium (Difco laboratories, Michigan, USA) for 18 h at 30°C and stored for up to two weeks at 4°C in YM liquid medium.

Growth medium: The growth medium consisted of 100 g.1-1 glucose, 10 g/*1 yeast extract, 1 g. H KH2PO4, 1 g. tl (NH^SC^, and 0.5 %.tx MgS04.?H20. Composition of fresh medium feed was the same as the medium employed as growth medium. The substrate medium had the same composition as the growth medium.

Inoculation and Preculture: Z mobilis was first propagated at 30°C for 18 h without agitation by transferring 0.5 ml of culture from the stock culture to

In Fuels and Chemicals from Biomass; Saha, B., et al.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1997.

10 ml of YM liquid medium. The starter culture was then transferred to 40 ml of seed culture medium at pH 5.5 and incubated for 18 hours at 30°C. This seed culture was transferred into 450 ml of growth medium in the fermentor to make a final volume of 500 ml.

Continuous culture: The schematic diagram of the experimental set-up is shown in Figure 1. Continuous culture was performed in а 1-/jar fermentor with a working volume of 500 ml. The composition of the medium and culture conditions were the same as used for the batch and continuous culture experiments. Fresh substrate of the same composition and same initial pH of 5.5 as that in the fermentor was fed intermittently into the fermentor. The amount of fresh substrate fed into the fermentor depended on the amount of alkali (0.5 M NaOH) pumped into the fermentor by the pH control unit (Biott Co. Ltd.,Току о Japan, model KVS-15-B) to neutralize the acid produced by the metabolizing cells, and to maintain the pH of the fermenting broth at the set level of 5.5. pH was monitored using a glass Pt/KCl electrode (Toko Chemical Laboratories Co., Ltd. Tokyo Japan). Fresh medium and 0.5 M NaOH solution were contained in 1 / and 100 ml graduated cylinders, respectively. The fresh substrate and alkali reservoirs were placed on sensitive electronic balances (Shinko Denshi Co. Ltd. Tokyo Japan, models SK-6000H and SK-600H) with weight limits of 6000 g and 600 g, respectively. The weight signals (voltage) from the feed substrate and alkali balances were inputted into a computer (NEC, PC 9801 VM) through an analog-to-digital (A/D) converter. The digital signal was converted to the voltage signal by means of a digital-to-analog (D/A) converter. Using software written in C++ computer language loaded on the computer hard disk, any decrease in weight detected by the alkali reservoir balance, resulting from the pumping of alkali into the fermentor to neutralize any acid produced, led to the switching on of a peristaltic pump (Tokyo Rikikai Co. Ltd., Tokyo Japan, EYELA Microtube pump MP-3) connected to the D/A converter. The pump then delivered specific weight of fresh substrate into the fermentor. This was achieved by monitoring the decrease in weight recorded by the fresh medium pump when the pump was activated. The pump was switched off when the required amount of fresh substrate was delivered. The same pump was employed to draw off exactly the same volume of culture broth equal to the volume of fresh substrate fed into the fermentor. This was to ensure that culture volume was maintained at constant level in the fermentor. Each experiment was carried out until steady state was achieved, after about 30 to 50 hours.