Gheorghe Maria

MODEL REDUCTION AND KINETIC PARAMETERS IDENTIFICATION FOR THE METHANOL CONVERSION TO OLEFINS

Abstract In order to reduce a complex kinetic model (33 reactions), which describes the methanol conversion to olefins, to a model easier to use, the discretization procedure combined with the integral procedure and ridge regression analysis were used. Because few experimental data are available, new points were generated using either the cubic spline approximation, or the extended model prediction, identified with a nonlinear regression procedure. The latter alternative, although it requires more computational effort, is also recommended for a subsequent kinetic analysis. The final refinement of the simplified model parameters may be accomplished by a multimodal optimization procedure.

Kinetic model discrimination via step-by-step experimental and computational procedure in the enzymatic oxidation of d-glucose

The enzymatic oxidation of D-glucose to 2-keto-D-glucose (D-arabino-hexos-2-ulose, D-glucosone) is of prospective industrial interest. Pyranose oxidase (POx) from Peniphora gigantea is deactivated during the reaction. To develop a kinetic model including the main reaction and the enzyme inactivation, possible side-reactions of the non-stabilised enzyme with D-glucosone, hydrogen peroxide, and peroxide radicals were considered. A developed step-by-step combined experimental and computational procedure allowed to discriminate among alternative inactivation mechanisms and provides an increased model reliability. The most probable scheme is the enzyme inactivation by hydroxyl radicals formed from produced H2O2 in the presence of Fe2+ ions. This .OH reaction is supported by matrix assisted laser desorption ionisation-mass spectrometry (MALDI-MS) measurement. The estimated kinetic parameter values for the main reaction are of the same order of magnitude as those reported in the literature. The identified model allows a satisfactory process simulation and highlights measures to prevent the enzyme activity loss.

Derivation of optimal operating policies under safety and technological constraints for the acetoacetylation of pyrrole in a semi-batch catalytic reactor

A large number of optimization strategies have been developed for semi-batch reactors (SBR) with a fixed or free terminal time, with or without considering various sources of uncertainty. Such strategies account for safety constraints rather empirically determined in the form of parameter thresholds, while safety indices are seldom integrated in the optimization objective function. The present work illustrates how the runaway boundaries, and their confidence region associated to the parameter uncertainty, can be evaluated using the process model and a generalized sensitivity criterion, and how they can be included in the SBR optimization. A concrete example is provided for the SBR used for the acetoacetylation of pyrrole with diketene in homogeneous catalysis, a process known to be of high risk due to the very exothermic side-reactions. While previous studies approached the isothermal SBR, the present work is focused on optimizing the non-isothermal SBR operation.

Enzymatic reactor selection and derivation of the optimal operation policy, by using a model-based modular simulation platform

Abstract Design and optimal operation of an enzymatic process depend on the characteristics of the free- vs. immobilized-enzyme, purification problems, materials and operation costs. When the enzyme stabilization is not fully successful, the optimal choice of the reactor type and operation mode is depending on the process and enzyme deactivation characteristics. Quick simulation of operation modes based on the process kinetics and a library of reactor models can help in comparing alternatives for getting the final decision. The analysis is exemplified for two enzymatic processes, by determining the optimal operating policy for several tested reactors: classic batch, batch with intermittent addition of enzyme, semi-batch reactor, continuously operated packed-bed column, or a mechanically agitated reactor with immobilized enzyme in small beads. Prediction of reactor performances points out the best choice for the given enzyme characteristics, and also conditions allowing to switch the production from large-volume batches to the continuous operation mode.

Model-based heuristic optimized operating policies for d-glucose oxidation in a batch reactor with pulsate addition of enzyme

Abstract Model-based optimization of batch enzymatic/bio-reactors is an intensively studied problem due to the increased economic benefits offered by the predicted optimal operating policies. The enzymatic oxidation of d -glucose (DG) in a batch reactor is a process of industrial interest, leading to production of “rare sugars” (keto-derivatives) from monosaccharides. The enzyme activity is sensitive to operating conditions, deactivation occurring due to some of the products and impurities. To keep a high catalytic activity during the batch, a pulsate addition of enzyme solution is performed under some restrictions. By imposing the maximum batch operation time and injected volume, operating policies maximizing the DG-conversion can be derived for every batch case. In the present study, by accounting the process kinetic characteristics, a pulse-injection control function, of exponential type, is proved to be easily adaptable for the optimization of the batch reactor. The heuristic near optimal solution is simple, flexible and can efficiently replace the optimal impulse-fed-policy derived by means of the time-consuming classical direct methods. The heuristic rule, based on the process kinetic model and on a library of semi-empirical functions, can be easily applied to optimize various batch enzymatic processes.

Precautions in using global kinetic and thermodynamic models for characterization of drug release from multivalent supports

Building-up a detailed kinetic model for drug release from various supports is a difficult task, especially when chemical reactions take place, accompanied by adsorption-desorption and diffusion steps. Often, semi-empirical release models derived from theoretical formulations of the transport process and system characteristics are employed. Their parameters have limited validity as they are dependent on the support, drug-ligand properties, and release conditions. However, they are often used for a quick simulation and design of drug delivery systems with a controlled release correlating the model parameters with the system characteristics and release conditions. Detailed information allows elaboration of an extended mechanistic model; the bias in the predictions introduced on various levels of model simplification is presented in this paper. A case study of a chemically activated ligand release in human plasma from a multivalent dendrimeric support is approached, pointing out the imprecision introduced by the gradual simplification of an extended model as well as the low reliability of the prediction when using various semi-empirical global models.

Comparative evaluation of critical operating conditions for a tubular catalytic reactor using thermal sensitivity and loss-of-stability criteria

Optimal operation of a chemical reactor according to various performance criteria often drives the system towards critical boundaries. Thus, precise evaluation of runaway limits in the parametric space becomes a crucial problem not only for the reactor’s safe operation, but also for over-designing the system. However, obtaining an accurate estimate for operating limits is a difficult task due to the limited validity of kinetic models describing complex processes, as well as the inherent fluctuations of the system’s properties (catalyst, raw-material quality). This paper presents a comparison of several effective methods of deriving critical conditions for the case of a tubular fixed-bed catalytic reactor used for aniline production in the vapour phase. Even though the methods being compared are related to one another, the generalised sensitivity criterion of Morbidelli-Varma (MV) seems to be more robust, not depending on a particular parameter being perturbed, when compared to the criteria that detect an incipient loss of system stability in the critical region (i.e., div-methods based on the system’s Jacobian and Green’s function matrix analysis). Combined application of div- and MV criteria allows for an accurate evaluation of the distance from the reactor’s nominal conditions to the safety limits.

Structured cell simulator coupled with a fluidized bed bioreactor model to predict the adaptive mercury uptake by E. coli cells

a b s t r a c t A bi-level model is proposed by coupling a three-phase fluidized bioreactor (TPFB), used for mercury uptake from wastewaters by immobilized Escherichia coli cells, with a cellular simulator of the genetic regulatory circuit (GRC) controlling the mercuric ion reduction in the cytosol. While keeping a reasonable agreement with the experimental data from literature (free cell cultures with/without cell membrane permeabilization), the structured model advantages are coming from the prediction detailing degree [simulated 26 + 3 (cell + bulk) vs. 3 (bulk) variable dynamics] covering a wide range of input Hg2+ loads (0–100 mg L −1 ), and cloned E. coli cells with various amounts of mer-plasmids (3–140 nM). The model offers the possibility to predict the inner cell mercury reduction rate (different from the apparent rate observed in the bioreactor), the bacteria metabolism adaptation to environmental changes over several cell cycles, and the effect of cloning cells to modify their behaviour under stationary or perturbed conditions.

In silico Derivation of a Reduced Kinetic Model for Stationary or Oscillating Glycolysis in Escherichia coli Bacterium

Modelling bacteria glycolysis is a classical subject but still of high interest. Glycolysis, together with the phosphotransferase (PTS)-system for glucose transport into the cell, the pentose-phosphate pathway (PPP), and tricarboxylic acid cycle (TCA) characterize the central carbon metabolism. Such a model usually serves as the foundation for developing modular simulation platforms used for consistent analysis of the control / regulation of target metabolite synthesis. The present study is focused on analyzing the advantage and limitations of using a simplified but versatile ‘core’ model of mTRM) of glycolysis when incomplete experimental information is available. Exemplification is made for a reduced glycolysis model from literature for Escherichia coli cells, by performing a few modifications (17 identifiable parameters) to increase its agreement with simulated data generated by using an extended model (127 parameters) over a large operating domain of an experimental bioreactor. With the expense of ca. 8–13 % increase in the relative model error vs. extended simulation models, derivation of reduced kinetic structures to describe some parts of the core metabolism is worth the associated identification effort, due to the considerable reduction in model parameterization (e.g. 17 parameters in mTRM vs. 127 in the extendedChassM model of Chassagnole et al.), while preserving a fair adequacy over a wide experimental domain generated in-silico by using the valuable extended ChassM. The reduced model flexibility is tested by reproducing stationary or oscillatory glycolysis conditions. The reduced mTRM model includes enough information to reproduce not only the cell energy-related potential through the total A(MDT)P level, but also the role played by ATP/ADP ratio as a glycolysis driving force. The model can also reproduce the oscillatory behaviour occurrence conditions, as well as situations when homeostatic conditions are not fulfilled.

Interference of the oscillating glycolysis with the oscillating tryptophan synthesis in the E. coli cells

Abstract Autonomous oscillations of glycolytic intermediates’ concentrations reflect the dynamics of the control and regulation of this major catabolic pathway in living cells. In spite of its major role in the cell central carbon metabolism (CCM), glycolysis interference with other metabolic pathways has seldom been investigated. However, the glycolysis model module is the ‘core’ part of any systematic and structured model-based analysis of the cell metabolism. On the other hand, tryptophan (TRP) synthesis is another oscillatory metabolic process of high practical interest in the biosynthesis industry, and in medicine. Because the TRP synthesis is strongly connected to the glycolysis through the PEP (phosphoenolpyruvate) node, based on two reduced kinetic models for glycolysis and TRP synthesis, this paper performs, for the first time, an in silico analysis of the way by which the two oscillatory processes interfere in the E. coli cells. The analysis allows guidance of tryptophan synthesis optimization.