Optical measurement and assessment of liquid distribution in shake flasks

  • Optische Messung und Auswertung der Flüssigkeitsverteilung in Schüttelkolben

binti Azizan, Amizon; Büchs, Jochen (Thesis advisor); Spiess, Antje C. (Thesis advisor)

Aachen (2018, 2019)
Dissertation / PhD Thesis

Dissertation, Rheinisch-Westfälische Technische Hochschule, 2018


Biotechnological development in shake flask necessitates vital engineering parameters e.g. volumetric power input, mixing time, gas-liquid mass transfer coefficient, hydromechanical stress and effective shear rate. This work provides ample experimental data for future validations representing shake flask fluid flow hydrodynamics i.e. in-phase (IP) and out-of-phase (OP) (phase phenomenon). Phase phenomenon was earlier expressed based on power input and that can now be expressed via liquid distribution. Optical fluorescence technique measurement producing 7644 data curves and 546 three-dimensional (3D) liquid distribution plots (hydrophilic), resulted from leading edge of bulk liquid (LB) and tail of bulk liquid (TB) analysis from 15-40 mL filling volume (VL), 150-450 rpm shaking frequency (n), 0-120 g/L of polyvinylpyrrolidone in fluorescent (fluorescent-PVP) solutions at 25 mm shaking diameter. The toroidal shapes of LB and TB are clearly asymmetrical and the measured TB differed by the elongation of the liquid particularly towards the torus part of the shake flask. Four new evaluation criteria proposed from 2184 evaluated cases, shed some light on how we closely relate 3D liquid distribution to phase phenomenon. The evaluation criteria on measured data are film thickness of the rotating bulk liquid, slope relative percentage of LB curves, circular angle of LB positions and 3D bulk liquid distribution’s peak shape change ($\geq$270$^\circ$ from 3D plot). A new compensated critical phase number (Phcrit) from theoretical to measured IP and OP has been revised from 1.26 to 0.91. Thus, phase change of liquid flow in 250-mL shake flask at 25 mm shaking diameter occurs at a little higher viscosity level of the fluorescent-PVP solutions than expected by the theoretical (calculated) phase number values. Additional to this, a non-viscosity model which serves as a simulated data is compared to measured liquid distribution investigated, illustrating range from similar to varying maximum liquid height (Hmax). Effect of hydrophobicity to liquid distribution angular position strengthen the fact to further understand the film and bulk liquid parameters. In conclusion, the 3D liquid distribution data collected at varying filling volume and shaking frequency, comprising of LB and TB values relative to the direction of the centrifugal acceleration are essential for validating future numerical solutions using CFD to predict vital engineering parameters in shake flask.