Peter Hugo

Frontal analysis method to determine competitive adsorption isotherms

In order to design and to optimise preparative separations, the knowledge of the underlying thermodynamic functions, i.e., the adsorption isotherms, is the most essential information. Usually these functions cannot be predicted and various methods have been suggested to determine them experimentally. In particular, dynamic methods are attractive regarding time requirements and reliability. Frontal analysis (FA) is frequently applied to measure single solute isotherms. The theoretical background of this method is offered by the classical equilibrium theory of adsorption. Although this theory also explains the shape of multicomponent breakthrough curves, FA is only seldom applied to determine competitive isotherms. In this work FA was used to measure adsorption isotherms of three components as single solutes and in binary and ternary mixtures in a reversed-phase system. The obtained equilibrium data were correlated using the competitive Langmuir isotherm equation, a bi-Langmuir model, the ideal adsorbed solution theory and the real adsorbed solution theory. No substantial improvement of the predictions was achieved using the more complex models instead of the most simple Langmuir model.

Quantification of single solute and competitive adsorption isotherms using a closed-loop perturbation method

A modification of the classical method of perturbation chromatography for measuring isotherms of the adsorption of dissolved components is suggested. The general principle of the method consists in analyzing responses of the chromatographic system to small perturbations at different equilibrium concentrations. Essential advantages of the method are: (a) only retention times or volumes have to be measured and no detector calibration is required and (b) experiments with mixtures can be performed and analyzed efficiently. The modification suggested in this paper is the application of a closed-loop arrangement allowing the efficient exploitation of the sample. Experimental data for four different chromatographic systems are presented to illustrate the method. With the determined adsorption isotherms elution profiles could be predicted satisfactorily.

Scale up in preparative chromatography

The subject of the paper is the application of a theoretical concept to scale up preparative chromatography. As an experimental example, the separation of the α- and β-isomers of a steroid compound using silica as the stationary phase is discussed. The applied concept is mainly based on measuring the adsorption isotherms in preliminary investigations using an analytical column. With these thermodynamic parameters it is attempted to simulate the separation behaviour in larger columns using the equilibrium dispersive model. To check the accuracy of the approach, experimental studies have been carried out using three other columns with increasing diameters (0.8, 5 and 30 cm). On the basis of the determined adsorption isotherms and a phase ratio adjustment, the main features of the separation process can be predicted for the larger columns. Thus, the presented concept can be effectively used in the process of optimizing preparative chromatography.

Calculation of the maximum temperature in stirred tank reactors in case of a breakdown of cooling

Abstract Safety is an important aim of chemical reactor design. If the reactor is equipped with a cooling device, the temperature rise at a cooling breakdown must be predictable. In the present paper this problem is discussed for continuous stirred tank reactors (CSTR) and semi-batch reactors (SBR). At breakdown of cooling, runaway of the reactor temperature occurs. The maximum temperature T m reached under adiabatic conditions, depends on the process conditions at the time of breakdown. The process temperature T s must be selected in such a way that in case of a cooling breakdown the maximum temperature T m is low compared to a temperature at which decomposition is critically active. For highly exothermic reactions it will be shown by theoretical considerations that, depending on the reaction number B, an optimal process temperature T s exists, for which the temperature T m has its lowest value. For a second order reaction if B exceeds 5.83 in the CSTR and 2 in the SBR the plot of T m versus T s will present a minimum. So a larger number of reactions will be concerned by this optimisation problem for the SBR than for the CSTR. General relationships will be derived to calculate the maximum temperature T m from T s and other operating conditions. It will be shown further how the operating conditions can be changed to get a sufficiently low runaway of temperature.

A comparison of the limits of safe operation of a SBR and a CSTR

Abstract In the case of exothermal reactions it is necessary to calculate the limits of safe operation from stability criteria. In their general form these criteria are often difficult to apply in practice. However, under industrial conditions the temperature of a reaction is usually fixed. Unter these conditions from the stability criteria a corresponding limit of the coolant temperature can be calculated. For safe operation the actual coolant temperature must be chosen higher than this limit value. Using the well known criteria of a CSTR the principles of these calculations are demonstrated. Using empirical stability criteria of a SBR similar calculations can be performed. Taking the same reaction and reaction conditions the SBR is more critical than the CSTR.

EXPERIMENTAL INVESTIGATION AND MODELLING OF CLOSED-LOOP RECYCLING IN PREPARATIVE CHROMATOGRAPHY

Abstract The separation of a binary mixture was investigated using a modified HPLC method. Since the separation factor was low the injected sample was recycled several times through the column to simulatean increase of the column length. For the example studied the separation could be improved considerably performing four cycles. To simulate the process a dispersion model was used to describe the concentration profiles in the column. Local equilibrium for the liquid and adsorbed phase concentrations was assumed. An apparent dispersion coefficient was applied to take into account all kinetic effects causing band broadening. In order to represent mixing effects in the pump a CSTR model was included additionally. Methods to determine the free model parameters were discussed. A good agreement between the experimental data the model predictions was found for the system investigated.

Linear two-step gradient counter-current chromatography: Analysis based on a recursive solution of an equilibrium stage model

The implementation of gradients in continuously operated chromatographic counter-current processes has recently attracted considerable interest as a method to improve the performance of this effective separation method. If liquid mobile phases are applied it is advantageous to set the solvent strength in the desorbent stream higher than that in the feed stream. As a consequence. the components to be separated are more retained in the adsorption zones and more easily eluted in the desorption zones. Due to the additional degrees of freedom the design and the optimization of such a two-step gradient counter-current process is difficult. In this paper a steady state equilibrium stage model is used to simulate the process under linear conditions. A simple solution of the underlying model equations is presented capable to describe efficiently the unit for large stage numbers typically encountered in chromatographic columns. Due to the rapidity of the algorithm developed a broad range of operating conditions can be evaluated systematically for different types of gradients. The impact of (a) the functional dependence of the adsorption equilibrium constants on the solvent composition, (b) the number of equilibrium stages and (c) the specification of purity requirements is illustrated and discussed based on results of parametric calculations. The results achieved emphasize the potential of two-step gradient counter-current chromatography.

Comparison of thermokinetic data obtained by isothermal, isoperibolic, adiabatic and temperature programmed measurements

In this paper isothermal, isoperibolic and adiabatic calorimeters and a Power-Compensating DSC are compared by determining kinetic data of a simple test reaction. First, the kinetic parameters were analyzed using a conventional isothermal method, based on the analytic determination of the course of reaction. Subsequently, the kinetic data of the performed reaction were determined for the different types of calorimeters by simultaneously evaluating several measurements with identical initial conditions but different temperature courses. The kinetic parameters obtained by the different calorimeters agree reasonable well, indicating the reliability of kinetic data derived from thermokinetic methods.

Determination of chemical kinetics by DSC measurements: Part 1. Theoretical foundations

A new evaluation method for determining the kinetic parameters of a chemical reaction from DSC measurements is presented. This method evaluates a series of thermograms at different heating rates. The activation energy, the frequency factor and a general description of the concentration dependence of the reaction rate are obtained. Theorical DSC curves are also evaluated and the results obtained are compared with the prescribed parameters.

Determination of chemical kinetics by DSC measurements: Part 2. Experimental results

Abstract In part 1 (Thermochim. Acta, 225 (1993) 143) a method to evaluate DSC measurements was developed and tested with simulated data. Part 2 contains experimental results for three different chemical reactions for which the kinetic data are determined. Provided correct baseline and time lag corrections are applied, reliable kinetic data can be obtained using this evaluation method. In contrast to other methods of evaluating DSC traces to obtain kinetic data, the method described here makes no assumptions about the order of reaction. Thus complex kinetic behaviour such as autocatalysis is readily apparent.