Table of contents

Molecular electronics is a new, exciting, interdisciplinary field of research. The subject broadly concerns the exploitation of organic materials in electronic and optoelectronic devices. There are many current commercial applications, including liquid crystal displays, conductive polymer sensors and pyroelectric plastics. Longer term developments might include molecular computational devices. In this review, the scope of molecular electronics is first discussed. Three examples of ongoing research that could have an impact on instrumentation and measurement science are then described. This is followed by some speculation on the possibilities for `molecular scale' electronic systems.

Simple schemes for gas microsensor packaging and testing are presented. The packaging procedures are developed for mostly hand operations without reliance on sophisticated equipment. Special flow-cell features and compact design enable easy configuration for testing in gas streams by flow-injection. Automation improves the efficiency of the sampling of gas microsensors and hence provides a higher compilation rate for the test data.

A dew-point meter capable of measuring humidity at high vapour pressure and high temperature has been constructed and tested. Humidity measurements in pure steam were made over the temperature range 100 - 150C and a vapour pressure range of 1 - 4 bar. The dew-point meter performance was assessed by comparing measurements with a pressure transmitter and agreement between the two was within % relative humidity. Humidity measurements in steam - air mixtures were also made and the dew-point meter readings were compared to those of a zirconia oxygen sensor. For these tests the dew-point meter readings were generally within % relative humidity of the oxygen sensor measurements.

The measurement of the stress - strain properties of materials at high strain rates is principally carried out using the split Hopkinson pressure bar (SHPB) technique, in which high compressive strain rates, typically in the range , are achievable with small solid discs as the test materials. Torsional SHPB systems have also been developed for test specimens in the form of small hollow cylinders. Another version of the SHPB technique has enabled high-strain-rate tensile measurements to be made, but inertial effects tend to limit the strain rate to about for the dogbone-shaped specimens. To achieve higher tensile strain rates, the authors have developed an expanding ring technique. This involves placing a thin ring of polymer as a sliding fit around a hollow, thick-walled, polymeric cylinder, the internal wall of which is loaded with a blast wave generated by an exploding wire. The tensile stress - strain properties of the ring material have been determined, at strain rates up to , by observing the movement of the ring as it freely expands away from the cylinder. Several polymers have been studied and comparisons made with compressive stress - strain behaviour.

A technique has been developed for measurement of the contact angle of liquid metals with solid materials. The liquid metal, when contained in a small cup (inner diameter <4 mm) forms a concave or convex mirror surface. For the present technique, an optical method was devised to measure the radius of curvature of this mirror surface, from which the contact angle can be determined. The surface of the liquid metal approaches sphericity as the diameter of its container is decreased. The optical system formed in this method is non-paraxial. An analytical formula, valid for a spherical non-paraxial mirror, for calculation of image size was deduced. The technique has been demonstrated using the mercury - graphite (15.9% total porosity) system at room temperature and the contact angle was determined to be .

A cell to keep samples in controlled atmosphere for Mössbauer studies of catalysts and small particles is described. The cell keeps the atmosphere of the reactor in which treatments like reduction, chemical reaction, or gas adsorption have been performed on a sample that afterwards is transferred to be studied by Mössbauer spectroscopy at low temperatures. The cell has been adapted to work inside a closed-cycle refrigerator that works in the temperature range 12 - 298 K. Its application is illustrated by studies of Fe small particles supported on in atmosphere.

The magnetoelastic properties of steel wires are used for measuring a mechanical load in tension. A microcomputer-controlled measuring instrument was designed. The dependence of magnetic behaviour on a previous state is eliminated by the suggested method of commutated measurement of amplitude permeability from saturation to the demagnetizing state. It was observed that the change of permeability with tension was linear during technical saturation of the specimen: low-carbon steel wire used in building industry. Temperature influenced the permeability curve and was found to be positive. The effect decreased with magnetizing field, but it increased under mechanical loading. The coefficient of relative change of permeability with both tension and temperature was compared. The resolution and the reproducibility of the method is described. The dispersion of magnetic properties of specimens has been shown to be less than 1.5% in the determination of resulting stress. Because of the internal stress of specimen evaluation this method allows one to measure the tension of loaded wire, especially in civil engineering applications.

The experimental error in the angle of incidence strongly affects the film optical constants and thickness sought from the ellipsometric data, particularly when the film is thinner than 200 Å. This angle should therefore be accurately measured and this has been a frequently faced difficulty, especially in fixed-angle-of-incidence and in situ ellipsometers. We describe in this paper two numerical calibration procedures evaluating the angle of incidence from the ellipsometric data and measured on well-identified film - substrate reflecting structures either with a known or with an unknown film refractive index. The procedures rapidly determine the angle of incidence independently of the film thickness. An error analysis shows that the angle of incidence can thus be determined with an uncertainty of a few hundredths of a degree from ellipsometers inducing the same order of angular errors. The thickness ranges of films yielding such lower uncertainties are deduced from the error analysis. The effect of substrate real index error on this uncertainty is studied in particular. The supplementary results produced therein may be used in data interpretation while checking the angle of incidence if it is accurately measured by some other method. The procedures may also be easily adopted to variably damped least-squares (VDLS) algorithms employed in real-time monitoring of films during growth. The calibration procedures are illustrated for all these aspects with experimental data from real reflecting systems.

A fringe shadow (FS) method was developed for measuring the flapping angle and torsional angle of a dragonfly wing during beating motion. This new method involves two sets of fringe patterns projected onto the wing from orthogonal directions. The torsional angle is determined using the length of the shadow of the wing chord that is cast by the two sets of fringe patterns. The flapping angle is determined using the shadowgraph of the wing projected by a laser. The advantages of the FS method are its capability (i) to measure the flapping angle and torsional angle of a dragonfly wing simultaneously using only one high-speed camera and (ii) to recognize the spanwise position of a section from the number of fringes, without having to use diffuse marks that are common in current methods. The resolution of the FS method can be changed easily by adjusting the fringe spacing. The measurement results for the torsional angle and flapping angle of a dragonfly wing prove the effectiveness of the FS method in studying the flight performance of dragonflies.

Resolution characteristics of neutron spectrometers using Larmor precession of the neutron spin are limited by magnetic field homogeneities of a special type. The line integral of the modulus of the magnetic induction along a neutron trajectory of length L is a measure of the amount of precession performed by the neutron. Hence it should be precisely the same for all neutron trajectories in a diverging beam. We present an analytical solution to the variational problem , for the case of cylindrical magnets (better than any lower symmetry geometry) coaxial to the beam axis. This solution describes the best irrotational (rot B = 0) field shape along the beam axis z. It can be obtained in practice by superposing a number of solenoids of different lengths. The optimal homogeneity is significantly better than for a simple solenoid of comparable dimensions, the only magnets used until now. For realistic lengths L and beam radii r, it is however not good enough for very high-resolution measurements. We therefore introduce a technique to correct both for the residual inhomogeneities of optimized cylinder magnets and the line integral variations due to path length differences resulting from finite angular beam divergence. Such corrections can only be done by introducing current distributions in the beam. Their optimal distributions can also be calculated analytically. Until now only the residual inhomogeneities have been corrected by in beam currents. With the two concepts of optimal field shape (OFS) and path length corrections described here, the resolution properties of Larmor precession techniques can be pushed to their intrinsic limits. As a further result of the correction technique introduced here, wider angular divergences can be used, for example using multi-detectors, resulting in substantially improved neutron economy. Several neutron spin echo (NSE) spectrometers based on above ideas have in the meantime been constructed by different working parties. The experimental results confirm the calculations reported here; no significant polarization drop is observed at maximum field.

For the investigation of dispersive elementary excitations in solids where the neutron energy change depends on the momentum transfer, a special type of gradient coils is needed. We describe their design for OFS precession magnets.

The low-temperature operation of distributed feedback semiconductor lasers in the vicinity of is investigated. The room temperature properties of a sample DFB laser are given and its power and frequency characteristics at 77 K are compared. In this way, the linearity of the temperature against frequency relation down to liquid nitrogen temperature is confirmed and a reduction in the threshold current to near zero is also shown. For spectroscopic purposes, it was found that stabilization of the laser temperature at 77 K is possible to within 50 mK over a range of about 77 - 87 K. This corresponds to a frequency change of about 120 GHz. The Schawlow - Townes relation predicts a reduction in linewidth with temperature; some evidence for this is given.

We demostrate a common-mode rejection scheme for a Faraday current sensor that can eliminate optical noise induced by fibre-link vibration.

Maximum likelihood estimation (MLE) techniques were used in the design of an autocorrelation processor for estimating the period (frequency) of a photon-resolved laser Doppler anemometer (LDA) signal. Implementation of these techniques required the derivation and numerical computation of the covariance matrix for the autocorrelation function of the photodetector signal. The processor is insensitive to the particle number density of the system and will work with photon-resolved and non-photon-resolved signals. The processor provides an accurate estimate of the error in each period estimate. Examples are given for the use of a Fisher matrix, derived as part of the MLE procedure, to predict the optimal processor performance of a LDA system as a function of the processor parameters and the system parameters. For periods of time over which the flow is steady the processor can be configured such that improved accuracy for each subsequent estimate can be achieved by utilizing the results from previous estimates. Computer simulations were used to test the processor design.

Pulsed laser deposition (PLD) of thin films is widely regarded as the best growth technique for the development of novel multilayer structures and devices; in particular, high-temperature superconducting (HTS) devices. To achieve this, it is essential to have the capability to deposit material from different targets sequentially in situ, namely without opening the deposition chamber. Here we present details of a multi-target carousel for multilayer depositions. It has been designed to allow target rotation and selection without any latching mechanisms or in-vacuum motors and is also suitable for use in automated systems. Additionally, it is robust, relatively inexpensive, compact, scalable and simple to build. Single-layer (YBCO) thin films grown using the multi-target carousel show no reduction in quality compared to ones grown using a single-target system. Moreover, the carousel has successfully been used to deposit Ag onto YBCO in situ to realize low-resistance YBCO - Ag - YBCO resistor structures; which would have been impossible with ex situ metallization. In addition, in situ homo-epitaxial MgO buffer layers on (100) MgO substrates prior to the deposition of YBCO have been investigated as a means of improving HTS film quality.

Currently pursued technologies and requirements for process tomography are briefly reviewed with emphasis on comparing existing methods and identifying roles for magnetic resonance techniques. It is concluded that fundamental studies of transport phenomena are among the beneficial applications of magnetic resonance techniques. After a brief review of the theory of magnetic resonance and a description of modern hardware for magnetic resonance imaging, specific examples of magnetic resonance investigations of mass and heat transfer are presented including studies of thermal processing, multiphase distributions, polymerization, and diffusion and flow. We conclude by speculating on future roles of NMR imaging for process developement and monitoring.

A genetic algorithm has been used to design a 35 cm diameter transverse gradient set for human brain imaging. Despite strict constraints on the maximum coil length (33 cm), gradient performance comparable to that from far longer sets of conventional geometry is achieved; thus the gradient efficiency is with a 24% standard deviation from this within a 26 cm sphere. Phase maps obtained using a gradient-acquisition for shimming sequence were used to test the linearity of the gradients without requiring a complex phantom.

A Dornier 28-channel magnetoencephalography (MEG) system equipped with RF-SQUIDs and control electronics made by CTF (Port Coquitlam, BC, Canada) was upgraded by replacing the RF-SQUIDs with DC-SQUIDs developed at the University of Jena, Germany. New preamplifiers were designed that emulate the function of the system's original RF amplifiers. This allowed the continuing use of the system's control electronics without any modification. Retaining these control devices was instrumental for preserving the system's unique ability to derive software-defined gradiometer measurements from its magnetometer signals, and also saved a considerable investment of control software tailored to the characteristics of these devices. The performance of the upgraded system compares favourably to that of other multichannel instruments; its software gradiometer channels typically attain an overall sensitivity of (an improvement by a factor of three over the original RF system) and suppress residual noise inside the shielded room by 40 to 60 dB. Also described are the procedures used to calibrate the system, to eliminate crosstalk between neighbouring channels, and to determine the coefficients for optimum suppression of ambient noise by the software gradiometer.

Process tomography systems need a long-distance high-speed serial link for data communication between sensing electronics and an image reconstruction computer, normally a PC. Using Transputer link adaptor chips (C011) and optical fibres, a low-cost serial data transmission link has been developed. A PC-XT extension card with an interfacing circuit is embedded in a PC, forming one terminal of the serial link. The other terminal can have either two 8-bit parallel ports (input and output respectively) to interface with logic circuits, or one 8-bit bi-directional port to interface with another microprocessor. The serial link with the optical fibres can transmit data at a speed of 10 or 20 Mbits per second over a distance as long as required. It has been used successfully in an electrical capacitance tomography system.

The seventh conference in the Sensors and Their Applications series proved to be its usual blend of excellent technical content together with opportunities to make new contacts and renew old ones among the UK sensors research and development community.

The organisers had arranged the contributions into nine sessions covering Optical Sensors, Sensor Arrays, Silicon Sensors, Flow Measurement and Tomography, Gas Sensors, Sensor Modelling, Acoustic Sensors and Sensor Applications. There were also 37 exhibited papers and a session was set aside for their presentation, thus avoiding the problem of finding time from the oral sessions to do justice to these papers.

Of the invited papers, M Gibbs (University of Sheffield) gave an account of piezomagnetic materials and their applications in microengineering, describing as an example a demonstration of an optically excited resonator whose resonant frequency could be varied by the application of a steady magnetic field.

D Delpy (University College London) brought us up to date with his work on spectroscopy and imaging of tissue in the near infra-red, with as examples his studies on the human breast and infants' heads. Of particular interest were the processing techniques used to minimize the effect of scattering of light by tissue.

A thought-provoking paper on environmental sensors was given by R W Bogue (Robert Bogue & Partners). In Mr Bogue's view, the opportunities for new sensors for water monitoring are not good, because most water monitoring for pollutants is done by sampling and laboratory analysis. Not only is this very cost-effective, but to replace this service with on-site sensors would be extremely difficult owing to the large number of different polluting species which need to be monitored. This is not the case with atmospheric pollution monitoring, however, as the number of pollutant species is by comparison small, and can in many cases be monitored at the point of emission (as with stack gases and the like). He also warned that the market for environmental sensors may saturate, as users who are purchasing equipment to comply with the current wave of legislation complete their purchasing.

P D Wilson (ERA Technology Ltd) described his work on a universal sensor interface chip (USIC). This device has been developed as part of the EUREKA project, JAMIE, to address problems presented in the use of application-specific integrated circuits (ASICs) for sensor interfacing. These problems are associated mainly with the extremely high cost of designing and tooling an ASIC for a specific sensor application, a cost which can only be justified when very large volumes are required. The USIC approach incorporates several input channels with analogue to digital conversion and multiplexing facilities to allow for calibration or referencing, and a reduced instruction set (RISC) processor providing either an analogue output or a serial or parallel digital output.

`An overview of electronic noses and their applications' was given by R J Elliott-Martin (Southampton University). A brief description of the human olfactory system was provided and the need for instruments to replace human experts in, for example, the food and drink industry was explained. Electronic instruments are based on arrays of chemical sensors, including semiconductor oxide sensors, polymer sensors, phthalocyanines etc, and the outputs from the sensor array are processed using pattern recognition techniques. Several instruments performing these functions are now available commercially. In addition to uses in the classification of such products as coffees, beers, whiskies, cheeses and fish, applications have recently emerged in agriculture, for assessing odour from livestock waste, and also in the diagnosis of cattle disease, a task traditionally performed by the vet smelling the beast's breath!

G Parry (Oxford University) reviewed recent developments in optical components using Fabry--Perot microcavities grown epitaxially in III--V materials. These comprised modulators and lasers which can be manufactured at potentially very low cost. The modulators can modulate amplitude and phase using low-voltage inputs, and are high-speed devices offering on:off ratios in excess of 100:1. The vertical cavity surface emitting lasers (VCSELs) offer the potential of low threshold currents, low cost and efficient coupling to optical fibres.

It is not simply the technical content of the presented material which makes for a good conference, but the opportunity to determine the mood of the community of fellow workers in one's field and to meet workers in one's own and related fields. To make this possible, we were hosted at a reception in Dublin's Mansion House by the Lord Mayor (truly an eccentric amongst a nation of eccentrics) and the following evening the Conference Dinner was held in the austere grandeur of the Dining Hall at Trinity College, Dublin. Before the meal, Dr N McMillan of the Regional Technical College, Carlow (Ireland) reminded us of the contribution of Trinity College to the development of optics and optical engineering through the work of such figures as Tyndall and Fitzgerald. Following this, Professor J D C Jones, as Honorary Editor of Measurement Science and Technology, presented the award for best exhibited paper to S Bourin of Robert Gordon University for her paper `A novel optical-fibre laser-induced-fluorescence detection system for capillary array electrophoresis'.

An interesting statistic emerging from this conference was the enduring interest in optical sensors. Not only were three out of the nine oral sessions devoted to optical devices but almost half of the exhibited papers also dealt with this class of sensor, in addition to which several optical techniques were described in the chemical, acoustic and sensor applications sessions.

The conference was attended by slightly more than 100 delegates, and 74 papers were presented either orally or by means of posters. Rather disappointing, though, was the small number of delegates from non-academic organisations (in which I include industrial research and development and consultancies), constituting only some 20% of all present.

In conclusion, then, the Sensors and Their Applications series of conferences continues to flourish, providing a biennial forum for developments within the UK sensors community. Its relevance could be further enhanced by encouraging more participation from manufacturing industry.

The now ubiquitous computer is becoming an important tool for the experimentalist as well as the theoretician, not only for producing the end-product of journal papers but also, increasingly, for the acquisition and processing of experimental data. Computer hardware, peripheral electronics and data analysis software have all developed considerably since the early works of authors such as Dalglish (1987 An Introduction to Control and Measurement with Microcomputers (Cambridge: University Press)) using the (now humble) Apple II and BBC computers.

This is a good introduction to the field of computer controlled instrumentation as the field now stands. It is intended for the beginner or the experienced researcher who is not computer literate and provides good comparative tables of hardware and systems as a guide to help the reader to make an informed choice. It is a well structured work and is appropriately divided into two sections. The first concerns the acquisition of data and instrument control and the second concerns the manipulation of this data.

The book is predominantly concerned with the software considerations. One-third of the first part of the book is devoted to instrumentation software and data structures. The other two-thirds of part 1 outlines the field for the first-time reader and contains chapters which provide introductions to basic data acquisition concepts presented as brief summaries. For example, data conversion techniques (both D/A and A/D) are covered in only ten pages including figures and tables. This does not detract from the work as a valuable starting point. However, the intended reader should be aware that it only provides summarized accounts of many topics since these are provided as background material to the discussion of more sophisticated software. The book is attempting to portray the complete picture for the benefit of those not experienced with either the hardware or the software and so, to keep the work to a manageable and readable size, few of the concepts are covered in depth.

The range of sophisticated software now available, such as MATLAB and LabWindows, are recent developments compared with the early work with microcomputers. Half of the book is devoted to the use of software such as these. MathCad, DSPlay, SPICE and Lotus tools are all discussed. The presentation of these tools is more detailed and includes some good worked examples. Since this is not intended to be a textbook, problems are not provided.

As an overview of the field of computer-based instrumentation, this book achieves the purpose of a good introductory text out of which the initiate can make an informed choice in building his or her own system. For a more detailed discussion of the basic concepts the reader could then refer to the extensive references for each chapter or the operating manuals of the hardware and software of their choice.

This recent addition to the Cambridge Studies in Modern Optics series aims to present the characteristic engineering problems encountered in the design of electro-optical systems concerned with signal processing. It does this through the presentation of the work of six groups of workers in the field. The book provides a good insight into the general nature of the design problems of electro-optical systems through the presentation of specific problems and their resolutions.

Chapter 1 deals with the design of coherent optical systems concerned with image processing (e.g. feature extraction using Fourier and Hough transforms). The design of the the system is shown to involve the characteristics of available optical devices, architectural considerations (e.g. of optical correlators) and the algorithms employed. Set against this scheme are the practical purposes for which the system is to be designed and constructed. The case study examined shows how novel algorithms can produce new architectures which, in turn, lead to new operations.

The potential of optical processors lies in their analogue operation: noise is thus a problem. Chapter 2 is devoted to an analysis of noise in matrix - vector multipliers, from which design criteria for optical linear-algebra processors are derived. The theoretical work is then augmented by experimental work on different types of light-emitting devices. Measurements of the fluctuating light intensities emitted from a number of different light sources which were subjected to a range of systematically varied drive currents were carried out. By this means design trade-off information was obtained.

Chapter 3 examines CAD techniques for use in the design of optical processors. Commercially available CAD packages are reviewed. Usually, the complexity is such that no single package can completely simulate the performance of a design; consequently, more than one CAD package is likely to be required for exacting designs. Additionally, packages are required to deal with limitations in performance deriving from the effects of diffraction, small-angle optical scatter and stray light. CAD packages which are capable of handling such factors over large dynamic ranges are required. Further design constraints include fabrication limitations such as assembly accuracy, glass homogeneity etc, though some CAD packages accomodate tolerancing in the design process. Design examples discussed include acousto-optic (Bragg cell) microwave spectrum analysers and space-integrating acousto-optic temporal signal correlators. It should be emphasized that from the standpoint of design it is the approaches adopted by the CAD packages that are of importance.

Chapter 4 examines the application of optical interconnections between large arrays of digital microprocessors as a means of increasing the computational capabilities of such systems. Three schemes are considered; the mesh, the hypercube and the fully-connected scheme. Using existing theory, comparisons are made between guided-wave and free-space connections in terms of density of connections, power dissipation etc. As the number of processors and the number of interconnects increases, free-space interconnects possess advantages over guided-wave approaches.

Chapter 5 describes the design of a Van der Lugt correlator which uses electrically addressable spatial light modulators. After discussing various types of SLM, those which incorporate electronic circuitry at each pixel to create arrays are concentrated on. Two types are examined: the ferroelectric liquid SLM and the silicon SLM. Important design issues are the influence of performance of pixel and device size, pixel fill factor, contrast ratio, flatness and uniformity and framing speed. Experimental results from a designed correlator are presented and conclusions drawn.

Chapter 6 is devoted to the design of a free-space digital optical logic system. After discussing the potential of the system, an overview is given of the characteristics of three optical logic devices: a Fabry - Perot etalon, the self electo-optic effect device (SEED) and light-emitting devices, such as the laser diode. The characteristics are then related to design objectives, which are essentially system-related issues, such as size of detector and modulator/emitter elements, element pitch, switching energy, power and speed. A system design using a symmetric SEED device (S-SEED) is presented, as well as test results of an operating system based on the design.

The book as a whole, through the description of different approaches to the design problems arising in different aspects of the field, presents a comprehensive view of the important issues which have to be confronted in the design of optical systems for signal processing. It can be recommended to optical researchers and designers and as an excellent introduction to the basic techniques of optical processing.

B O'Connor

The table headings were missing from tables 1 and 2. The corrected tables are given below.

Table 1. Focal properties of the immersion objective. The theoretical predictions in the zero space-charge, zero-sized aperture limit are v = -1, M = 0.5.

Table 2. Aberration coefficients.