Computer systems for analytical chemistry

Microcomputers are becoming integral parts of all analytical instrumentation and 'personal computers' are invading the laboratory. The analytical chemist is confronted with the question of how to computerize his work and what to request from a computer system. This article attempts to provide some guidelines.     

The analytical laboratory computer as a management aid

Intelligent use of the data available in the store of the analytical laboratory computer can provide much of the information necessary for the optimization of laboratory resources. The research manager can use such information to monitor the activity of various projects and the cost effectiveness of the analytical service.     

Thermal analysis—mass spectrometer computer system and its application to the evolved gas analysis of green river shale and lunar soil samples

A thermal analysis—mass spectrometer system controlled by a computer has been developed and successfully used in the analysis of a wide variety of geochemical samples. The TA—MS computer system provides a very powerful analytical tool for the determination of volatile species released from samples over a very wide temperature and sample-size range. Use of a small laboratory computer and magnetic-tape storage units permits large quantities of analytical information to be handled and easily retrieved. The rapid scan capabilities of the quadrupole mass spectrometer operating under computer control are especially useful for following the rapidly changing composition of the evolved gases released from samples during heating under vacuum conditions. Samples of volatile-rich Green River shale and two lunar soils have been analyzed to show the utility of the TA—MS computer system.     

Laboratory computer systems and the role of the human interface

The importance of the “human interface” between the operator and a laboratory computer system is emphasized. Some rules for communication through dialogue programs and command languages as well as for the presentation of computer output are given. The degree of automation of the evaluation of analytical data versus interactive treatment is discussed.     

Artificial neural networks in foodstuff analyses: Trends and perspectives A review

Artificial neural networks are a family of non-linear computational methods, loosely inspired by the human brain, that have found application in an increasing number of fields of analytical chemistry and specifically of food control. In this review, the main neural network architectures are described and examples of their application to solve food analytical problems are presented, together with some considerations about their uses and misuses.     

Validating automated analytical instruments to meet regulatory and quality standard requirements

Analytical laboratories are more and more faced to meet official regulatory requirements as described in FDA and EPA good laboratory practice, good automated laboratory practice and good manufacturing practice regulations or to officially establish quality systems, such as specified in the ISO 9000 Series quality standards, in the ISO Guide 25 or in the EN 45001 guidelines. The impact on analytical instrumentation will be the requirement for stringent validation of analytical equipment and methods which increase the overall analysis costs. An overview is presented on the validation requirements using e.g. gas chromatography, high performance liquid chromatography, capillary electrophoresis and UV-visible spectroscopy and on the strategy to meet such needs at minimal extra costs with the help of an instrument vendor. It is recommended to use instrument hardware that has already built-in tools for self-verification and which is to be validated at the vendor's site. Performance testing in the user's laboratory is done using standard operating procedures as supplied with the instrument. If resources in the user's laboratory are limited, the performance verification is done by the vendor. Software and the entire computer system is validated prior to shipment at the vendor's site. Acceptance testing is done in the user's environment following the vendor recommendations. Analytical methods are validated automatically at the end of method development using a dedicated software. The software can be customized such that it can also be used for daily automated system suitability testing. Security and integrity of analytical data are ensured by saving the raw data together with instrument conditions and instrument log-books in check-sum protected binary register files for long-term archiving.     

The use of a micro-mini-mainframe computer network in the analytical chemical laboratory

A network of communicating and cooperating computers, consisting of a central host computer and one microcomputer per experiment, is described. The network is intended to be used as a universally applicable tool in analytical chemical research. A specially developed programming language, easy to learn, offers the possibilities of real-time clock-controlled data acquisition, control of experiments, simple logical decisions, and cooperation with the host computer, which simultaneously runs a program in a conventional higher language such as FORTRAN. The analytical application used for illustration relates to inductively-coupled plasma emission spectrometry.     

Teaching Chemometrics with a Bioprocess: Analytical Methods Comparison Using Bivariate Linear Regression

We present an advanced analytical chemistry laboratory experiment involving chemometrics. Students perform a comparison of two analytical methods by checking several analyte concentrations within a certain range by using least-squares linear regression. They obtain statistical information such as the presence of constant and proportional biases. The exercise is based on the determination of glucose levels using two colorimetric methods (enzymatic and Somogyi—Nelson) in a very simple batch system formed by an infusion of tea, glucose, and a combination of a yeast (Schizosacaromyces pombe) and a bacteria (Acetobacter xylimun), usually named Kombucha. Several samples are collected during a week of laboratory work, and measurements are performed in a subsequent four-hour laboratory class. Although commercial computer software exists for a variety of statistical applications, specific programs for the application of statistics to analytical chemistry are not prevalent. In order to solve this particular problem, a Matlab 5.3 routine is presented.     

Entanglement effects in model polymer networks

Many fundamental questions for the understanding of polymer networks are more suitably addressed by current computer simulations than by experiments. Details of the microscopic topology, such as the elastically active cluster or loop entanglements, can be identified as well as controlled. In particular, it is possible to isolate and quantify their effects on macroscopic observables such as the elastic modulus. The constraints due to connectivity and conserved topology are more clearly present for networks than for melts. Already for strand lengths between crosslinks which are relatively short, the effect of the conserved topology is important. The mode relaxation in a network is significantly different from that of a melt. For weakly crosslinked systems the melt entanglement length is the relevant scaling parameter. The elastic modulus of a long chain network under ideal conditions reaches an asymptotic value which is about 2.2 times smaller than the prediction of an affine model for a network made of strands of the melt entanglement length. An analysis of the stress reveals that in the linear regime the contribution from the excluded volume is dominant compared to that from the connectivity along the strands. For larger elongations, however, the non-linear elastic response is dominated by the chemically and topologically shortest paths through the system, where chemical crosslinks and topological entanglements between meshes of the network play a similarly important role.     

Finite size effects in tightly meshed polymer networks

Molecular dynamics computer simulations on regular, tightly meshed model networks exhibit variations of the network density with system size. We show that these variations are due mainly to network elasticity. A theoretical expression derived on the basis of the self-consistent-field approach yields finite size scaling behavior in good accord with the simulation for a wide range of thermodynamic conditions.