Optimisation and robustness analysis of a hydrophobic interaction chromatography step

Process development, optimisation and robustness analysis for chromatography separations are often entirely based on experimental work and generic knowledge. The present study proposes a method of gaining process knowledge and assisting in the robustness analysis and optimisation of a hydrophobic interaction chromatography step using a model-based approach. Factorial experimental design is common practice in industry today for robustness analysis. The method presented in this study can be used to find the critical parameter variations and serve as a basis for reducing the experimental work. In addition, the calibrated model obtained with this approach is used to find the optimal operating conditions for the chromatography column. The methodology consists of three consecutive steps. Firstly, screening experiments are performed using a factorial design. Secondly, a kinetic-dispersive model is calibrated using gradient elution and column load experiments. Finally, the model is used to find optimal operating conditions and a robustness analysis is conducted at the optimal point. The process studied in this work is the separation of polyclonal IgG from BSA using hydrophobic interaction chromatography.     

Automated off-line optimisation of programmed elutions in reversed-phase high-performance liquid chromatography using ternary solvent mixtures

A new system for the unattended optimisation of gradient elutions in reversed-phase high-performance liquid chromatography has been developed. The system is based on the simulation of retention times under conditions of ternary solvent mixtures and ternary gradient programmes. This model is constructed departing from a few experimental data obtained in isocratic elutions and validated against experimental gradient elutions. Once validated, this retention model can be used for the unattended search of an optimum separation. The optimisation process is driven by an evolutionary algorithm (EA), specially developed to map the ternary gradient elutions problem. The development and characteristics of the retention modelling as well as those of the EA and its particularities are discussed and some real world examples of separation presented.     

Optimisation-based simulation of a pressure swing adsorption process

In this paper, an optimisation-based approach is developed for the determination of the cyclic steady-sate (CSS) of a pressure swing adsorption process (PSA). It consists in treating the simulation problem as a single dynamic optimisation problem where the performance index is the CSS condition, the decision variables are the state variables at the start of the cycle and the constraints are given by the process model equations with associated initial conditions. The resulting optimisation problem is solved using a gradient-based non linear programming (NLP) method, e.g. SQP method, where the gradients are computed by means of four different methods: finite differences, numerical and analytical sensitivities and adjoint system methods.     

The telomerisation of 1,3-butadiene and carbon dioxide: process development and optimisation in a continuous miniplant

The telomerisation of 1,3-butadiene and carbon dioxide is one of the first homogeneously catalyzed reactions using carbon dioxide as a C1-building block. In this article we describe the process development for a miniplant applying this telomerisation in a continuous scale. Through repeated optimisation of the plant setup combined with parallel laboratory batch experiments the overall space-time-yield of the plant was enhanced significantly.     

Turbidity of pulpy fruit juice: A key factor for predicting cross-flow microfiltration performance

To determine the technical and economic feasibility of cross-flow microfiltration on an industrial scale, the expected decline of permeation flux must be predicted taking into account the variability of juice's fouling potential. However, the main difficulty is finding representative parameters. Two semi-empirical models – gel polarisation and mechanistic – were used to fit experimental permeation flows using initial juice turbidity as surrogate for the volumetric concentration of particles in the feed juice. The experimental data of different banana, pineapple, and blackberry juices fitted well in both models. Although the mechanistic model more accurately estimated the permeation flux density, for practical application, the simpler polarisation model was preferred. Because this method uses a factor (i.e. turbidity) that reflects juice's fouling potential, it allows the optimisation of processing parameters and the prediction of permeation flux range in real industrial conditions.     

Automatic initialisation of buffer composition estimation for on-line analysis of unknown buffer solutions

An automatic initialisation procedure for extracting useful information about buffer composition from a titration experiment is presented in this paper. The initialisation procedure identifies which buffering components are present in the sample from a relatively long list of buffers expected in the system monitored. The procedure determines approximate pK _a values of the buffers and evaluates their maximum and minimum concentrations. This information is then used to start an optimisation procedure to fit the model of the buffer components to the titration data and to accurately determine buffer concentrations and pK _a values. The procedure has been integrated as a software layer around the buffer capacity optimum model builder (BOMB) that fits a buffer-capacity model to a measured buffer-capacity curve to estimate model properties (pK _a values and concentrations). The reliability and robustness of the resulting buffer capacity software (BCS) were tested using a titrimetric analyser simulator (TAS). The BCS was then validated off-line and on-line.     

Ancestral population genomics using coalescence hidden Markov models and heuristic optimisation algorithms

With full genome data from several closely related species now readily available, we have the ultimate data for demographic inference. Exploiting these full genomes, however, requires models that can explicitly model recombination along alignments of full chromosomal length. Over the last decade a class of models, based on the sequential Markov coalescence model combined with hidden Markov models, has been developed and used to make inference in simple demographic scenarios. To move forward to more complex demographic modelling we need better and more automated ways of specifying these models and efficient optimisation algorithms for inferring the parameters in complex and often high-dimensional models.In this paper we present a framework for building such coalescence hidden Markov models for pairwise alignments and present results for using heuristic optimisation algorithms for parameter estimation. We show that we can build more complex demographic models than our previous frameworks and that we obtain more accurate parameter estimates using heuristic optimisation algorithms than when using our previous gradient based approaches.Our new framework provides a flexible way of constructing coalescence hidden Markov models almost automatically. While estimating parameters in more complex models is still challenging we show that using heuristic optimisation algorithms we still get a fairly good accuracy.     

Towards a portable environment for FORTRAN applications on parallel computers

Summary This paper extends the communicating sequential process model of communications on multicomputers to a virtual machine interface with skeleton algorithms for global communication of data. The software is constructed using elements of object-oriented programming. Data is divided into two types of variables: those stored locally to each process and those accessible globally. Transfer of data between processes is implicit in the library structure, providing opportunities for optimisation.The application-level interface currently resembles a vector linear algebra library. Generic communication skeletons are defined for building libraries specific to applications; they illustrate more general programming issues. An example from a wave-scattering algorithm is given.     

Optimisation & scale-up of affinity chromatography

When affinity chromatography is used on the large scale, economics dictate that the process must be optimised to ensure high throughputs and high adsorbent utilisation. The optimisation of an affinity chromatography process starts with the selection of both the ligand and matrix components of the affinity adsorbent. The relative merits of using highly selective ligands and small particle matrices in the development of high performance chromatography systems is discussed. Once the adsorbent has been chosen the parameters that characterise its interaction with the adsorbate must be characterised from small scale experiments. The method is illustrated in a series of experiments that show the influence of a number of factors involved in the preparation of immobilised monoclonal antibodies on the subsequent performance of the immunosorbents. These parameters together with computer simulations of the chromatography process can be used to design and optimise the operation of large scale systems. Attention has to be given to the optimisation of the adsorption, washing and elution phases of the process. Apart from initial process design, it is also necessary to optimise the chromatographic separation during the run. This involves the need to be able to measure the chromatographic performance in an on-line manner and a number of approaches to achieve this are described.     

A method for rapid reaction optimisation in continuous-flow microfluidic reactors using online Raman spectroscopic detection

An extremely rapid tool for continuous flow synthetic process optimisation is described. A microfluidic reaction system operating in continuous flow is used in conjunction with confocal Raman microscopy to afford rapid molecule synthesis and product quantitation. Accordingly, the approach allows for rapid reaction optimisation within a continuous flow system. Specifically, the catalytic oxidation of isopropyl alcohol (IPA) to acetone using tetra-N-propylammonium perruthanate (TPAP)/N-methylmorpholine N-oxide (NMO) in a radial interdigitated micromixer is studied as a model reaction system. The composition of the reaction effluent can be determined with great facility and information relating to catalyst/co-oxidant ratios, catalyst turnovers and reaction endpoints extracted. Specifically, variation of catalyst and co-oxidant volumetric flow rates between 0 and 50 microL min(-1) is used to vary reactant concentrations, define reaction residence times and control product conversions between 0 and 100%. The rapid nature of the system allows chemical information to be gathered and utilised on a sub-minute timescale.     

A hybrid genetic algorithm for estimating the equilibrium potential of an ion-selective electrode

Non-linear equations can be used to model the measured potential of ion-selective electrodes (ISEs) as a function of time. This can be done by using non-linear least squares regression to fit parameters of non-linear equations to an ISE response curve. In iterative non-linear least squares regression (which can be considered as local optimisers), the determination of starting parameter estimates that yield convergence to the global optimum can be difficult. Starting values away from the global optimum can lead to either abortive divergence or convergence to a local optimum. To address this issue, a global optimisation technique was used to find initial parameter estimates near the global optimum for subsequent further refinement to the absolute optimum. A genetic algorithm has been applied to two non-linear equations relating the measured potential from selected ISEs to time. The parameter estimates found from the genetic algorithm were used as starting values for non-linear least squares regression, and subsequent refinement to the absolute optimum. This approach was successfully used for both expressions with measured data from three different ISEs; namely, calcium, chloride and lead ISEs.     

Optimal design of batch and simulated moving bed chromatographic separation processes

Preparative chromatography, especially simulated moving bed (SMB) chromatography, is a key technology for the separation of fine chemicals on a production scale. Most of the design methods for batch and SMB processes proposed in the open literature deal with the optimisation of the operating conditions for a given chromatographic unit only. Therefore, a comparison of the process economy may lead to incorrect results. In this contribution, an effective strategy for the optimal choice of all process parameters (operation and design parameters) is proposed. The main idea of this strategy is to apply a detailed and experimentally validated process model and to reduce the number of influencing parameters by introducing and optimising dimensionless process parameters. It is shown that there is an infinite choice of design and operating parameters to achieve maximum productivity or minimum separation costs and not at the maximum pressure drop only. The detailed design of the chromatographic unit and the selection of the operating conditions can be adjusted by considering the availability of columns and packing materials. As the model system, the separation of a racemic mixture (EMD53986) on Chiralpak AD was investigated. After complete optimisation of a batch and a SMB unit, a real comparison of the process economy can be achieved. Finally, the influences of two different objective functions, productivity and specific separation cost, are analysed.     

Multiobjective evolutionary optimisation for surface-enhanced Raman scattering

In most optimisation experiments, a single parameter is first optimised before a second and then third one are subsequently modified to give the best result. By contrast, we believe that simultaneous multiobjective optimisation is more powerful; therefore, an optimisation of the experimental conditions for the colloidal SERS detection of l-cysteine was carried out. Six aggregating agents and three different colloids (citrate, borohydride and hydroxylamine reduced silver) were tested over a wide range of concentrations for the enhancement and the reproducibility of the spectra produced. The optimisation was carried out using two methods, a full factorial design (FF, a standard method from the experimental design literature) and, for the first time, a multiobjective evolutionary algorithm (MOEA), a method more usually applied to optimisation problems in computer science. Simulation results suggest that the evolutionary approach significantly out-performs random sampling. Real experiments applying the evolutionary method to the SERS optimisation problem led to a 32% improvement in enhancement and reproducibility compared with the FF method, using far fewer evaluations.     

A systematic approach for optimisation of supercritical-fluid extraction of polycyclic aromatic hydrocarbons from earthworms

A systematic approach using a mathematical model as an alternative to time-consuming empirical optimisation of a supercritical-fluid extraction (SFE) procedure is presented. The model was applied to the extraction of 15 polycyclic aromatic hydrocarbons (PAH). The selected fat-containing matrix is the earthworm used in ecotoxicological absorption studies. Settings for optimal recovery were established for the important parameters (temperature, pressure, amount of trapping sorbent, flow, and dynamic extraction time) using a D-optimal experimental design (including quadratic terms and interactions). The recoveries were modelled using a two sigmoid-model with parameters for each of the individual PAH. The main objective was to optimise the conditions for 15 PAH congeners by maximisation of the lowest recovery. The results show that for some parameters, e.g. the amount of sorbent material, optimal conditions are identical for all PAH. For other parameters, e.g. extraction time, the optimum is PAH dependent. The advantage of this optimisation procedure is that, within three days of analysis (73 experiments), optimised extraction conditions for the extraction of the set of 15 PAH were found but also optimum conditions for specific subsets can be extracted from the collected data for specific subsets.     

Response surface methodology for optimisation of 4,6-Dinitro-o-cresol quantification using hanging mercury drop electrode

A method for the pesticide DNOC (4,6-Dinitro-o-cresol) quantification using hanging mercury drop electrode (HMDE) and stripping square wave voltammetry (SSWV) optimisation is proposed. As a continuous cathodic current decrease was observed during the experiments, a waiting time, together with solution pH and SWV instrumental variables were optimised by factorial designs. From the two reduction current peaks values, only one was considered as dependent variable in the optimisation process. While the cathodic current peak and standard deviation were used in the waiting time and solution pH optimisation process, the instrumental parameters for SWV were optimised by using only current peak values. With the optimal parameters, a calibration curve from (0.01–0.55) × 10^?6 mol L^?1 with LOD of 2 × 10^?8 mol L^?1 was obtained. The proposed method was checked for DNOC quantification in different water samples obtained from Cordoba area and very good results, with recovery values around 102% were observed.     

Optimisation and characterisation of silica-based reversed-phase liquid chromatographic systems for the analysis of basic pharmaceuticals

Reversed-phase liquid chromatography using silica-based columns is successfully applied in many separations. However, also some drawbacks exist, i.e. the analysis of basic compounds is often hampered by ionic interaction of the basic analytes with residual silanols present on the silica surface, which results in asymmetrical peaks and irreproducible retention. In this review, options to optimise the LC analysis of basic pharmaceutical compounds are discussed, i.e. eluent optimisation (pH, silanol blockers) and stationary phase optimisation (development of new columns with minimised ionic interactions). The applicability of empirical based, thermodynamically based and test methods based on a retention model to characterise silica-based reversed phase stationary phases, as well as the influence of the eluent composition on the LC analysis of basic substances is described. Finally, the applicability of chemometrical techniques in column classification is shown.     

Implementation of the multi-channel monolith reactor in an optimisation procedure for heterogeneous oxidation catalysts based on genetic algorithms

A multi-criteria optimisation procedure based on genetic algorithms is carried out in search of advanced heterogeneous catalysts for total oxidation. Simple but flexible software routines have been created to be applied within a search space of more then 150,000 individuals. The general catalyst design includes mono-, bi- and trimetallic compositions assembled out of 49 different metals and depleted on an Al2O3 support in up to nine amount levels. As an efficient tool for high-throughput screening and perfectly matched to the requirements of heterogeneous gas phase catalysis - especially for applications technically run in honeycomb structures - the multi-channel monolith reactor is implemented to evaluate the catalyst performances. Out of a multi-component feed-gas, the conversion rates of carbon monoxide (CO) and a model hydrocarbon (HC) are monitored in parallel. In combination with further restrictions to preparation and pre-treatment a primary screening can be conducted, promising to provide results close to technically applied catalysts. Presented are the resulting performances of the optimisation process for the first catalyst generations and the prospect of its auto-adaptation to specified optimisation goals.     

触变性含蜡原油等温胶凝过程特性及模型研究

综合利用理论推导和流变实验方法对触变性含蜡原油的等温胶凝过程进行了详细的研究, 采用结构参数法建立了静态结构恢复模型, 并经实验结果证明该模型对多种原油的结构变化曲线均有较高的拟合精度. 同时以青海原油为例, 根据实验数据和拟合参数分析了含蜡原油等温胶凝过程特性. 结果表明, 随着温度的升高、预剪切速率的增大以及预剪切时间的增加, 原油在恒温静止过程中形成的结构强度均会减弱. 测量温度越高, 预剪切时间越长, 则原油的结构参数变化速率越大; 而预剪切速率越高, 原油结构参数变化速率在初始阶段越大, 在最终阶段反而越小, 但与未经剪切油样相比, 原油受剪切后其结构参数变化速率在整个测量时间范围内均偏大.     

A Simplified Model for the Etch Rate of Novolac-Based Photoresist

The novolac-based resins used as positive-tone photoresists are frequently etched in an oxygen plasma. It is desirable to have a predictive model of the photoresist etch rate but, for process improvement, control, and analysis, the development of a rigorous mechanistic model is impractical. Instead, a simplified mechanistic model is derived, here, according to the method proposed by Hougen and Watson for the study of fluid–solid interactions. This model derivation method is employed in order to arrive at a functional form that represents chemical etching of the resist by oxygen radicals, assisted by the plasma ion flux. Values for model parameters are determined from process data by nonlinear regression. The quality of the model fit to the data is tested statistically.     

A rapid, automated approach to optimisation of multiple reaction monitoring conditions for quantitative bioanalytical mass spectrometry.

An improvement to the procedure for the rapid optimisation of mass spectrometry (PROMS), for the development of multiple reaction methods (MRM) for quantitative bioanalytical liquid chromatography/tandem mass spectrometry (LC/MS/MS), is presented. PROMS is an automated protocol that uses flow-injection analysis (FIA) and AppleScripts to create methods and acquire the data for optimisation. The protocol determines the optimum orifice potential, the MRM conditions for each compound, and finally creates the MRM methods needed for sample analysis. The sensitivities of the MRM methods created by PROMS approach those created manually. MRM method development using PROMS currently takes less than three minutes per compound compared to at least fifteen minutes manually. To further enhance throughput, approaches to MRM optimisation using one injection per compound, two injections per pool of five compounds and one injection per pool of five compounds have been investigated. No significant difference in the optimised instrumental parameters for MRM methods were found between the original PROMS approach and these new methods, which are up to ten times faster. The time taken for an AppleScript to determine the optimum conditions and build the MRM methods is the same with all approaches.