March 15, 2014 (Vol. 34, No. 6)

Studying the Impact of Industrial-Scale Inhomogeneities on Microbial Cultures

Hamilton’s VisiFerm Arc 120 dissolved oxygen and Polilyte Plus Arc 120 pH sensors are valuable tools when used in a plug flow reactor (PFR) to measure the metabolic response of Bacillus subtilis to oxygen availability. The PFR is connected to a common stirred tank reactor (STR) to simulate conditions in large-scale industrial processes. The impact of inhomogeneities of oxygen, and potentially pH, on Bacillus subtilis cultivations were observed. This study leads to the determination that such inhomogeneities will also exist, and can be measured and corrected, in industrial processes.

To achieve conditions that mimic those in large scale, determination of the hydrodynamic behavior by ascertaining the residence time distribution is of basic interest. Therefore, a reliable measurement of the pH along the PFR is necessary.

During cultivation, the distribution of oxygen and pH along the PFR allows an exact description of the process conditions and enables the user to adopt them to large-scale conditions.

The PFR is characterized by high inhomogeneities in the axial direction with respect to oxygen concentration. Since oxygen limitation in industrial fed-batch processes occurs near the feed line, the feed is introduced at the entrance of the PFR. Unlike the first lab-scale two-compartment reactors, the system in this study is equipped with five pH and optical dissolved oxygen (DO) sensors along the height of the PFR (Figure 1).

Static mixers included in the module are used for enhanced oxygen transfer and mixing, while the plug flow characteristics are maintained. Sampling is possible between each static mixer module (at the same heights where pH and DO are measured).

Calibration of pH electrodes was performed applying Hamilton Device Manager (HDM) software in calibration solutions of pH 7.00 and 4.01; DO sensors were calibrated in controlled N2 sparged water for 15 minutes (zero-level) and air sparged water (100 % level). After utilization, accuracy of calibration was proven by sparging water with a gas mixture of 5% O2 and a calibration in air.

Small deviations between the responses among DO sensors was normalized to one electrode, enabling a more reliable comparison between all DO sensor signals at the PFR at very low values close to zero. The sensors in the PFR were sterilized with steam for 40 min at 110 °C prior to and after each experiment.

A mineral salt medium was applied for cultivation. The sensor data was transmitted via a USB Modbus Converter RS 485 to a LabVIEW™ v. 8.2 environment.

Figure 1. Plug flow reactor (PFR) combined with stirred tank reactor (STR)


In the above system, the metabolic response of nonsporulating Bacillus subtilis mutant strains to oscillating substrate and oxygen availability were successfully examined.

With the help of the DO sensors, the dissolved oxygen concentration was monitored along the PFR. When air was supplied at the bottom, the dissolved oxygen concentration was higher at the second monitoring port (DO sensor 2) than at the first one (DO sensor 1) due to gas dispersion (Figure 2).

When no air was supplied at the bottom of the PFR, a weak signal was observed at the first electrode (DO sensor 1) (Figure 2). Based on this sensor, the oxygen content in the STR was successfully adjusted so that not all oxygen in the liquid phase was consumed by the bacteria on the way from the STR to the PFR.

Figure 2. Time course of the pH in the PFR and STR module during a B. subtilis fed-batch cultivation

No oxygen could be detected at the two top sample ports in the PFR (DO sensor 4 and 5). If the air supply relied only on the oxygen that was dissolved in the STR, a gradient of aerobic to anaerobic conditions in the PFR was observed, indicated by O2 depletion in the top phase.

The pH in the PFR and STR was monitored as well (Figure 3). In this case, the pH measurement along the PFR module did not indicate acid release as confirmed by HPLC analysis. pH monitoring is even more useful at cultivations, where acids are formed in the PFR module (e.g., E. coli cultivations).

The application of the sensors was successful, fulfilled the customer’s needs, and showed excellent measurement performance throughout all experiments.

Figure 3. Time course of the dissolved oxygen concentration in the PFR and STR module during a B. subtilis fed-batch cultivation

Stefan Junne, Ph.D. ([email protected]), Arne Klingner, Dirk Itzeck, and Peter Neubauer, Ph.D., are researchers in the laboratory of bioprocess engineering, department of biotechnology, at the Technical University of Berlin.

Previous articleNovo Nordisk Building $100M Purification Pilot Plant
Next articleTop R&D Spenders of 2013