pH Monitoring in Dairy Starter Cultures

Characterization of acid production with optical sensors in microtiter plates

G. T. John1, D. Goelling2, I. Klimant3, H. Schneider2, and E. Heinzle1
1Technical Biochemistry, Saarland University, Saarbrücken, Germany
2Danisco Cultor Niebüll GmbH, Niebüll, Germany
3Institute for Analytical Chemistry, Chemo- and Biosensors, University Regensburg, Regensburg, Germany

Starter cultures are used to guarantee good and reproducible quality of dairy products. In this study an easy-to-use method to screen a large number of inoculum cultures as quality control for raw materials was developed. Microtiter plates with integrated chemical optical pH sensors were tested for their suitability to monitor acid production during milk fermentation. Differences in inoculum type and size of cultures of Lactococcus lactis and Streptococcus thermophilus could be clearly detected, even though there were disturbances caused by particles and fluorophores contained in the medium. The new approach was compared to the established CINAC system. This early study resulted in an improved SensorPlate - the HydroPlate.

Monitoring and screening starter cultures is of great importance in modern dairy technology to guarantee quality of products like e. g. yoghurt, cheese, or butter milk. During milk fermentation organic acids, mainly lactic acid, are produced by the microorganisms lowering the pH value. This acid production can be used to characterize milk fermentation and is therefore an important method to test the activity of starter cultures. Aim of this study was to develop a new method that allows screening of a large number of samples in microtiter plates by optical pH measurement. Preliminary tests had shown, that the optical characteristics of the plates and especially milk had influence on the pH sensor signal, as the sensors were not optically isolated. Results gained from the following fermentation experiments in microtiter plates with integrated pH sensors were compared with those from the established CINAC-method to verify the suitability of this new approach. The CINAC system is highly reliable but limited to 16 parallel experiments and quite labor intensive due to required cleaning and recalibration of the electrode. The microtiter plates with sensors should eliminate these disadvantages.

Materials & Methods

For measurements with chemical optical pH sensors 10 µL of a solution containing pH-sensitive and pH-insensitive fluorophores were dropped into each well of a 96 well round bottom polystyrene plate (Greiner, Germany). The resulting layer thickness of the so created sensor was about 5 µm. The fluorescence intensity of the pH-sensitive indicator fluorophores was measured by applying an excitation filter with maximum transmission at 492 nm. The emitted fluorescence was measured after passing through a filter of 538 nm transmission maximum. For the pH-insensitive reference fluorophores a combination of 544 nm / 590 nm filters was used. Str. thermophilus and Lc. lactis (Danisco, Germany) were used for milk fermentation experiments. They were received freeze-dried and stored at - 20 °C. Aliquots of cultures (0.5 g, 0.1 g, 0.02 g) of both strains were given to 100 mL sterilized phosphate buffer solution each. 1.2 mL of these inoculums were put into 200 mL milk medium (ultra-high temperature sterilized, homogenized milk, 0.3 % fat, Mibell, Germany) respectively. Measurements were carried out at 30 ± 0.1 °C and 37 ± 0.1 °C. Each experiment with the optical sensors was made 8-fold in parallel in a fluorescence reader (BMG Fluostar, BMG Labtechnologies, Germany). For each measurement 200 µL liquid per well were used. The pH in the microtiter plate was also measured using a pH meter (Mettler Toledo MP220, Mettler Toledo, Switzerland) with a pH-mini- electrode (Mettler Toledo Inlab® 423, Mettler Toledo, Switzerland). Comparative experiments were done with the CINAC system (Ysebaert, Frepillon, France). The system uses 16 pH-electrodes. For these measurements 200 mL glass bottles were completely filled with medium and no gas phase was left. They were kept closed throughout the entire experiment. To obtain initial calibration curves calibration solutions containing 90 % milk were used. Solutions were gained by mixing 18 mL milk with 2 mL of solutions of HCl and NaOH with varying concentrations. Measurements were taken every 10 minutes for 24 hours.

Characteristics of Milk Fermentation

The sensor material had a calibration curve as given in Fig. 1. Using milk and adjusting the pH value with acid or alkali, the resulting curve differed significantly from the calibration curve with buffer solution of identical ionic strength. This way disturbances caused by the optical properties of milk could be checked. Comparative measurements with the CINAC system were performed under oxygen exclusion and the measurement vessels were not moved. In the microtiter plates milk is exposed to air all the time and the plates are moved in the fluorescence reader. Therefore, additional off-line measurements in the plates were conducted using a pH-electrode. For experiments with the pH sensing microtiter plates, the freeze-dried inoculums of both Str. thermophilus and Lc. lactis were cultivated at 30 °C and 37 °C in sterilized milk. From the measured values of Irel pH was calculated using either of the two calibration curves shown in Fig. 1. The calibration with milk could only be used for pH values above 5 corresponding to Irel > 0.56 (Fig. 2 top). For the first 3 to 4 hours of fermentation pH values calculated like this were quite close to values measured off-line by the pH-electrode. Results from the CINAC system trended close to those taken with pH-electrode in the microtiter plates, and showed that the exposure to air and movement of the plates did not influence the fermentation process. Using the buffer calibration curve pH estimation over the whole measurements could be made, as shown in the lower graph of Fig. 2. Initially the optically estimated values were considerably higher than the electrode values but agreed reasonably well from about 4 - 10 hours of fermentation time. The reference measurements with the CINAC system were similar to the optical online sensor signal, and also the differences between inoculum sizes and cultures showed similar time courses. However with the CINAC system no minimum values could be detected like with the optical measurement. Fermentation carried out at 37 °C but otherwise identical conditions gave similar results as for 30 °C. The more slowly growing Str. thermophilus gave different curves (results not shown). Results demonstrate that fermentation power of different inoculum cultures could be compared using the pH measuring plates.


Microtiter plates with integrated chemical optical sensors (optodes) allow parallel measurement of 96 cultures. Despite some disturbances the signal obtained during fermentation was useful to distinguish between starter cultures of different activity. Also strains with different fermentation characteristics, as Lc. lactis and Str. thermophilus, showed different shapes of the pH sensor signal Irel. The major advantages of this new system compared to established methods are its simplicity, the large number of samples that can be measured in parallel using the 96 well microtiter plates, and the immense reduction of the workload. Microtiter plates with integrated chemical optical sensors are a promising method for tests and research where screening a large number of samples is necessary. Today, an improved pH sensor plate (HydroPlate) with optical isolation and less cross-sensitivity to milk ingredients is widely used for this and similar applications.

Application note adapted from John et al., 2003, pH-Sensing 96-well microtitre plates for the characterization of acid production by dairy starter cultures, J Dairy Res. 70 (3), pp. 327 - 333


Select by industry

Presens TV

Watch tutorials, webinars and informative videos about PreSens optical sensor systems.

All videos