Hak Koon Yeoh

An algorithm to form balanced and diverse groups of students

We share the details of a simple algorithm which groups students based on diversity in gender and race while keeping the group average CGPA nearly equal. Application on simulated and actual class profiles gave multiple‐feasible solutions within seconds, with a maximum absolute deviation in group average CGPA of 0.03/4.00 or lower.

Interlinked population balance and cybernetic models for the simultaneous saccharification and fermentation of natural polymers.

Simultaneous Saccharification and Fermentation (SSF) is a process where microbes have to first excrete extracellular enzymes to break polymeric substrates such as starch or cellulose into edible nutrients, followed by in situ conversion of those nutrients into more valuable metabolites via fermentation. As such, SSF is very attractive as a one‐pot synthesis method of biological products. However, due to the co‐existence of multiple biochemical steps, modeling SSF faces two major challenges. The first is to capture the successive chain‐end and/or random scission of the polymeric substrates over time, which determines the rate of generation of various fermentable substrates. The second is to incorporate the response of microbes, including their preferential substrate utilization, to such a complex broth. Each of the above‐mentioned challenges has manifested itself in many related areas, and has been competently but separately attacked with two diametrically different tools, i.e., the Population Balance Modeling (PBM) and the Cybernetic Modeling (CM), respectively. To date, they have yet to be applied in unison on SSF resulting in a general inadequacy or haphazard approaches to examine the dynamics and interactions of depolymerization and fermentation. To overcome this unsatisfactory state of affairs, here, the general linkage between PBM and CM is established to model SSF. A notable feature is the flexible linkage, which allows the individual PBM and CM models to be independently modified to the desired levels of detail. A more general treatment of the secretion of extracellular enzyme is also proposed in the CM model. Through a case study on the growth of a recombinant Saccharomyces cerevisiae capable of excreting a chain‐end scission enzyme (glucoamylase) on starch, the interlinked model calibrated using data from the literature (Nakamura et al., Biotechnol. Bioeng. 53:21‐25, 1997), captured features not attainable by existing approaches. In particular, the effect of various enzymatic actions on the temporal evolution of the polymer distribution and how the microbes respond to the diverse polymeric environment can be studied through this framework. Biotechnol. Bioeng. 2015;112: 2084–2105.

Applying the least squares method for the titrimetric determination of the concentration of VFA

The titrimetric method is used for on-site measurement of the concentration of volatile fatty acids (VFAs) in anaerobic treatment. In current practice, specific and interpolated pH-volume data points are used to obtain the concentration of VFA by solving simultaneous equations iteratively to convergence (denoted as SEq). Here, the least squares method (LSM) is introduced as an elegant alternative. Known concentrations of VFA (acetic acid and/or propionic acid) ranging from to 200 to 1,000 mg/L were determined using SEq and LSM. Using standard numbers of data points, SEq gave more accurate results compared with LSM. However, results favoured LSM when all data points in the range were included without any interpolation. For model refinement, unit monovalent activity coefficient (f(m) = 1) was found reasonable and arithmetic averages of dissociation constants and molecular weight of 80 mol% acetic acid were recommended in the model for VFA determination of mixtures. An accurate result was obtained with a mixture containing more VFA (butyric acid and valeric acid). In a typical VFA measurement of real anaerobic effluent, a satisfactory result with an error of 14% was achieved. LSM appears to be a promising mathematical model solver for determination of concentration of VFA in the titrimetric method. Validation of LSM in the presence of other electrolytes deserves further exploration.