ANALYST. APRIL 1993, VOL. 118 3 17 Tutorial Review Optical Chemical Sensors: Transduction and Signal Processing* Ramaier Narayanaswamy Department ot Instrumentation and Analytical Science, UMIST, P.O. Box 88, Manchester, UK M60 7QD The development of optical chemical sensors is a rapidly growing technology, and it combines the advantages of optical fibres with the selectivity and specificity of chemical transduction systems. The chemical transducer is the heart of such sensors and considerable effort is being made in their development. The analyte concentrations are determined through the measurement of chemically encoded optical signals. The optical signal processing is carried out using standard components and instrumentation t o produce quality and meaningful data. This paper reviews the principles of chemical transduction and the signal processing systems that are used in conjunction with optical chemical sensors.Keywords: Optical sensor; chemical transduction; optical instrumentation Optical methods have been used for several years in various fields of analyses. Some of the oldest and best known applications include the use of pH indicator strips for sensing pH and oxygen measurements by quenching the phosphor- escence of the dye trypaflavin adsorbed on silica gel.' However, a major achievement occurred when conventional optical sensing techniques were coupled with optical fibres. Although optical fibres were originally manufactured mainly for use in the communication industry they have been adapted to optical sensing devices. Optical fibres themselves allow transmission of light over long distances, and for chemical sensing typically 1-100 m distances are currently used.The advent of high quality, low optical fibres meant that this distance could be much greater. Additionally, the develop- ment of high performance and high quality optical com- ponents, including light sources (lasers, light emitting diodes), photodetectors, amplifiers, etc'., that can be used in conjunc- tion with optical chemical sensors has promoted rapid progress and great interest in this sensing technology. Optical fibre chemical sensors rely on an interdisciplinary approach to their development with expertise from the fields of analytical chemistry, physics (optics) and optoelectronics. Progress in this field of chemical analysis is mainly dependent on the ease and manner in which the chemical transduction system can be developed and interfaced with fibre optics.Furthermore, use of chemome trics and microprocessors, which allow storage and rapid analysis of data, has signifi- cantly contributed to the state-of-the-art of optical chemical sensors. Advantages and Disadvantages Chemical analysis is currently carried out in almost all areas of technology, and optical chemical sensors offer several advan- tages over conventional devices in a wide range of applications including process control, and environmental and biomedical fields. Optical sensors are electronically passive and are not subject to electrical and electromagnetic interferences. They are also tlexible, easily miniaturized, inexpensive and of rugged construction.They can be corrosion resistant and are * Prcsentcd at the Scnsors and Signals Symposium at The Royal Society of Chemistry Autumn Meeting, Dublin, Ireland, September 16-18, 1992. capable of real-time monitoring of samples. As they are non- electrical, optical sensors are intrinsically safe and capable of operation in hostile environments. Referencing can be carried out optically and internally and, unlike potentiometric sensors, they do not require a separate system. The use of low optical loss fibres in sensors allows transmission of optical signals over distances of up to about 1 km, and this makes them useful in remote sensing applications. Substantial economic advantages are provided by these sensors due to the possibility of multiplexing several sensors to a single, often costly, instrumentation.As these are small in size, they can be used to measure analyte concentrations without significantly perturbing the sample. Optical sensors can be developed for chemical and biological analytes for which other sensing devices are not suitable or available. Notwithstanding numerous advantages optical sensors also have some disadvantages. Ambient light can interfere with the measurement of optical signals. Therefore, either the sensors must be used in dark environments or the optical signal must be modulated to resolve it from the ambient light. In the latter case, the detector would also be modulated to receive only those modulated optical signals. Long-term stability could be a problem with optical sensors based on the use of indicator phases as a result of photodegradation or leaching effects.These can be compensated to some extent by the use of the ratio of optical signals of two measurement wavelengths. As the chemical indicator and the analyte are in different phases, the consequent mass-transfer step required to achieve chem- ical equilibrium will limit the response times of optical chemical sensors. In general, sensors will have a limited dynamic range of measurement of analyte concentrations. One way this range can be extended would be by the use of multiple sensors that can mcasure different levels of analyte (see Conclusions). Despite a few limitations, optical sensors have the potential to provide an alternative to other sensing systems due to the many advantages these sensors impart in analytical measure- ments.Principle The basic concept of an optical fibre based chemical trans- ducer is simple (Fig. 1). Light from a suitable source is launched into the optical fibre and guided to a region where it interacts with a measurand system or with a chemical318 ANALYST, APRIL 1993, VO12. 118 - Incident I Source I (4 \ Reagent phase I J light To detector Amplifier Readout ( b)/ - Incident light - To detector Lr Fig. 1 Schematic diagram of the principle of an optical fibre chcmical sensor Table 1 Some examples of plain fibre sensors Species Method of detection Ref. - To detector - Incident light / Phenolic compounds Direct ultraviolet cxcitation and measurement of fluorescence 2 Reagent phase Uranium (as U02’+) Direct fluorescence 3 Fig.2 Optical sensors with chemical transducer placed at the cnd of NADH (in bioreactions) Direct fluorescence ( a ) a bifurcated optical fibre bundle and ( b ) a single optical fibre, or (c) coated on the outside of a single fibre (from ref. 5 ) 4 transducer. This interaction results in a modulation of the optical signal, and the modulated light, which is encoded with chemical information, is collected by the same or another optical fibre and guided to a light-measuring system. Depend- ing on the particular measurement device and the optical principles employed, the optical signal measured can be absorbance, reflectance, luminescence or scattering. The chemical transducer usually consists of immobilized reagent(s) that is analyte specific and is often capable of measuring trace levels of analyte.Sensor Design Two types of optical fibre sensing devices for chemical species have been developed. One type, which can be described as a ‘spectroscopic’ or ‘plain-fibre’ sensor, detects the analyte species directly through their characteristic spectral proper- ties. In such devices, the optical fibre functions as a light-guide conveying optical energy from the source to the sampling area and from the sample to the detector. In the sample, the light interacts with the chemical species being monitored (see Table 1, for some examples). In the second type of sensing device, a ‘chemical transducer’ is interfaced to the optical fibre. The analyte species interacts with the transducer, whose modified optical properties are monitored through the optical fibre.This type of device is the more common one, because a large number of analytical species are themselves colourless or non-luminescent. Furthermore, great specificity can be im- parted to these sensors as a consequence of the incorporation of the chemical transduction system. cladding material [Fig. 2(c)]. Hence, the transmission charac- teristics of the optical fibre are modified as a result of the change in the refractive index of the coating when the reagent in the transducer undergoes changes in optical characteristics owing to its interaction with the analyte. Such devices employ the principle of the evanescent wave technique. For plain-fibre sensors, the configuration of the optical fibre can be either as shown in Fig.2(a) and ( b ) or the fibre can be placed in the transmission mode whereby the length of the fibre is interrupted by an area where the analyte sample comes into the optical view of the fibre.6 As mentioned earlier, the chemical transducer consists of immobilized chemical reagents, which are normally analyte specific. The reagent systems are often employed in the solid phase for convenient handling. The reagent phase is localized in the sensing region of the optical fibre either by direct deposition on the fibre or by encapsulation with a polymeric membrane. Immobilization of chemical reagents can be carried out either physically or chemically.” The physical methods of immobilization include gel entrapment, adsorp- tion and electrostatic attraction, and use simple and econ- omical procedures. Chemical immobilization is based on the formation of a covalent bond between the reagent molecule and an activated or functionalized form of the polymeric solid support.This method is the most irreversible of the immobi- lization techniques, but requires several steps in the synthesis of the immobilized reagent phase. The simplest type of optical chemical transducer uses a single step reaction between the reagent phase and the analyte species, and a number of optical sensors studied to date have The interaction between the light and the analyte takes ciearly demonstrated this- approach. A more complicated place at one end of the optical fibre. A variety of transducer transducer might use a multi-step reaction to produce a configurations has been employed in the sensing region.measurable optical signal. However, this type of device is Transducer based sensing devices can involve the use of either normally more difficult to implement in practice. The present bifurcated optical fibre bundles or single optical fibres5 as range of optical sensors tends to be more chemically or shown in Fig. 2. In bifurcated optical fibre based sensors, biochcrnically specific, because so much indicator chemistry i s separate fibres are employed to transmit incident and detected available for adaptation into this promising new technology. optical radiation. However, in single optical fibre based Chemistry of Sensor Response sensors, it will be necessary to distinguish between the incident and detected radiation.In both types of designs, the chemical transducer is placed at the distal end of the optical fibre [Fig. The sensor response function in optical chemical sensors 2(a) and ( b ) ] . With single optical fibres, the chemical depends on the manner in which the analyte interacts with the transducer can also be a coating on the surface of the fibre as a reagent phase with the chcmical transducer. For example, aANALYST, APRIL 1003, VOL. 118 319 reagent (R) reacting with an analyte species (A) forming a product (AR) can be represented as A + R G A R (1) Here, either R or AR is usually absorbing or luminescent, and can therefore be measured optically. The reagent employed is usually selective, producing a distinct optical change for the given analyte species.The chemical transduction is normally based on the equilibrium established during the chemical reaction between A and R, and this equilibrium can be described as where K is the equilibrium constant and the square brackets indicate the equilibrium concentration of the species involved. During the chemical reaction, the reagent R is consumed and, therefore, any absorbance or luminescence due to it de- creases, or as the product AR forms its absorption or luminescence increases. Either way, the change in optical property of R or AR can be related to their concentration and, in turn, related to the concentration of the analyte A causing changes in the measured optical property. These are described below. If the reagent is consumcd during the reaction, then its total initial concentration (cR) at any time will be given by CK = [AR] + [R] (3) Thus, from eqns.(2) and ( 3 ) , the analyte concentration can be expressed as [ A ] = p 1 (G- 1) (4) If the optical property of AR is being measured then the analyte concentration can be rclated to the concentration of AR The concentrations of R in eqn. (4) and of AR in eqn. (5) are related to the measured optical property. The reaction represented in eqn. (1) is a reversible one and thus provides sensing with a direct indicator, which requires an appropriate equilibrium constant for the desired analyte concentration range. The sensor response also depends on the total amount of the indicator. Furthermore, any uncontrolled variable that affects the equilibrium constant will be a potential source of error.For example, for any reaction involving ions, variations in ionic strength will affect the K values. If the measured optical parameter is dependent on the ratio of the concentrations of the two forms of the indicator, i.e., [AR]/[R], then the response no longer depends on the total amount of the indicator, although the dependence on the equilibrium constant remains. Indicator reactions are most familiar to the analytical chemist in the context of titration procedures and indicator strips commonly used for semi-quantitative estimation of pH. The concentration of indicator is much lower than that of the analyte so that there is no perturbing effect in the sample. For example, in optical pH sensing, there is a multitude of pH indicators with different K values.Therefore, it will be relatively easy to find a pH indicator to cover virtually any desired pH range. Because of its importance, the theory and practice of optical pH sensing has been the subject of a critical study by many workers. Indicators showing reversible reactions with analytes are generally preferred for use in optical sensors because they can provide continuous and unperturbed measurements. The equilibrium response of indicators does not depend on mass transfer; however, the response time (ie., the time required to reach the equilibrium) is dependent on mass transfer. Revers- ible interactions of analytes with indicators, which do not involve chemical reactions, can provide short response times.For example, the optical oxygen gas sensor based on the dynamic quenching of fluorescence of an indicator does not involve a chemical reaction .7,8 The fluorescence quenching effect takes place by energy transfer from indicator to the oxygen molecules when they are in contact, and there is no consumption of oxygen. However, there are only a few examples known of fully reversible analyses and sensing by optical measurements. Many reagent phases can be regener- ated after reaction with analytes. For cxample, the lead sensor based on immobilized dithizone can be regenerated with HCP whereas the aluminium sensor based on immobilized Erio- chrome Cyanine R can be regenerated using fluoride solution (Fig. 3).*() Measurements of the ratio of optical intensities at two wavelengths can be used to determine the analyte concentra- tion.Such ratio measurements are inherently more stable with respect to drift. One of the main advantages of the optical sensing approach to chemical measurements is the possibility of developing sensors with long-term stability with respect to calibration because they can be based on ratio measurements. Hence, eqn. (2) can be rearranged to As mentioned earlier, the ratio [AR] : [R] is independent of the amount of indicator. Hence, it should be possible to design ‘precalibrated’ sensors based on ratiometric intensity measurements, and the analysis can be performed directly. Most chemical sensors reported to date involve the use of indicators, and some examples are illustrated in Table 2, according to the type of reaction involved.As can be inferred, most indicator reactions involve a ground-state interaction with analyte, except in the luminescence quenching reaction, where the interaction occurs in the excited state. I - Time Fig. 3 Response of immobilized Eriochrome Cyanine R to alumi- nium ions in solution (A) and its regeneration using fluoride solution (B) Table 2 Direct indicators in optical sensors Type of reaction Analyte Acid-base p” co2 NH3 Complexation Pb2 + AP+ Ligand exchange H20 (vapour) Luminescence quenching O2 0 2 so2 Halidcs Halocarbons Ref. I I , 12,13 14 15, 16 9 10,17 18 19 7 , s 20 21 22320 ANALYST, APRIL 1993, VOL. 118 The use of irreversible chemical reactions in the develop- ment of optical sensors will result in ‘one shot’ devices that can be used only once and which must then be disposed of.Although there will be limited merit in using such reactions in fibre optic chemical sensors, they can provide high sensitivity of measurements. However, the reagent phases have to be prepared with sufficient reproducibility so that the response will be the same from sensor to sensor. Otherwise, the sensor cannot be satisfactorily calibrated. If the irreversible reagent phases are formulated as flat strips or slides, then it might be possible to achieve the necessary reproducibility in measure- ments. The reaction involved in an irreversible sensor can be represented as A + R + A R (7) Such reactions proceed towards completion, are indepen- dent of the equilibrium constant and measure the total amount o f analyte.The sensor response will be dependent on mass transfer. Examples of such sensors include those for sensing H2S’3 and CI2.2-‘ Indirect indicators have been studied for use in optical chemical sensors, which include two or more reagent com- ponents, whose interaction varies with the concentration of thc analyte. The interaction can be represented as in eqn. (1) or as R + S S R S (8) where S is the species that competes with A for the reagent R. In the absence of A, the reaction in eqn. (8) will proceed to the right. As the concentration of A increases in the sample, the reaction in eqn. (1) proceeds to the right and the reaction of eqn. (8) proceeds to the left, resulting in free S. The optical characteristics of S can be measured as a function of the concentration of A.Such interactions are referred to as ‘competitive binding reactions’ and have been used in t h e development of an optical glucose sensor based on immobi- lized Concanavalin A and fluorescein labelled dextran .Z5 In general, indirect indicators offer the potential of varying the effective equilibrium constant, which in turn determines the range of concentrations to which the sensor responds. This is n o t possible with direct indicators. In addition to the different types of chemical interactions discussed above, there are other phenomena such as catalysis, with which chemical species could be sensed by optical means. These are not discussed here. Optical Techniques The optical sensing of chemical species is based on their interaction with light. In optical chemical sensors, three of the optical techniques commonly employed for measurements are the absorption, emission and reflection or scattering of light.The quantitative aspects of the use of these optical phenomena are briefly discussed here, together with a brief explanation of thc phenomena for completeness. Absorption Absorption of optical energy gives rise to transitions in the electronic, vibrational and/or rotational energy states of the atoms and molecules, and occurs only if the difference in the energy states involved matches exactly the energy of the exciting photons. Visible and ultraviolet radiation induces electronic excitation, infrared radiation promotes vibrational excitation, and microwave radiation gives rise to rotational transit i ons .Absorption leads to a diminution of the power of the radiation as it passes through the sample. Therefore, after encountering a number of absorbing species, a light beam of initial intensity I(, will be transmitted by the sample with a reduced intensity I . It should be recognized that only those frequencies that are absorbed will be attenuated, and all other frequencies will pass through with no power loss. The decrease in the light intensity is determined by the number of absorbing species in the light path, and is related to the concentration, c , of the absorbing species through the Beer-Lambert equation A = log/o/I = EIC (9) where A is the absorbance, I is the length of the light path and E is the molar absorptivity, which is characteristic of the analyte substance at a given wavelength.Luminescence The absorption of energy from a photon causes atoms or molecules to be promoted to a higher energy state. However, the excited species is short-lived and releases its extra energy via several pathways. One of thc relaxation modes involves luminescence, wherein radiation of a lower frequency is emitted. Deactivation through luminescence can occur either from a singlet state, in which case the emission is called fluorescence, or from a triplet state, in which case the emission is called phosphorescence. Fluorescence is extremely rapid and occurs within 1-100 ns after excitation. On the other hand, phos- phorescence has a longer decay period (1-1000 ps) and persists after the removal of the excitation source.In both cases, the emitted radiation is of a different frequency to that of the exciting radiation, and its intensity, I[_, is dependent on the intensity of the exciting radiation, lo, and the concentration, c , of the luminescent species. For weakly absorbing species, i.c., A <0.05, the intensity of luminescence can be expressed by the following equation ( 10) I L = k’1(+3Elc where 1 is the length of the light path in the sample, E is the molar absorptivity, 0 is the quantum efficiency of the luminescence and k’ is the fraction of the emission that can be measured. At constant lo, eqn. (10) can be simplified to I , = kLC (11) where kL = k’IoOd. In the presence of some species (e.g., oxygen), the luminescent decay of an activated species could compete with a collisional quenching decay mode. The mean lifetime of the activated species is decreased and the luminescence intensity is reduced.In this case, the luminescence intensity, IL, is related to the concentration of the quenching species, cq, by the Stern-Volmer equation P / I L = 1 + Ksvcq ( 12) where I0 is the luminescence intensity in the absence o f the quencher and Ksv is the Stern-Volmer constant. An excited species can also be generated through a chemical reaction. The measured light that is emitted as the excited species returns to the ground state is then known as chemilumineseence, which can be quantitatively related to the concentration of the analyte species. Reflectance Reflection takes place when light strikes a boundary surface. Two distinct processes are responsible for this.The first is the mirror-type or specular reflection, which occurs at the interface of a medium with no transmission through it. The other type is diffuse reflection, wherein the radiation pene- trates and subsequently reappears at the surface of the system following partial absorption and the multiple scattering within the system. These two processes are complementary, but specular reflection can be eliminated or minimized through proper sample preparation or optical engineering.ANALYST, APRIL 1993, VOL. 118 Recorder/ display 32 1 . Lock-in amplifier Diffuse reflectance has been recognized to be dependent on the composition of the system, analogous to light absorption. Several models for diffuse reflectance have been proposed based on the radiative transfer theory, which considers all incident light to be light scattered by other particles.The most widely used model is the Kubelka-Munk theory. Here, a thick semi-infinite scattering layer is assumed and reflectance, R , is related to the concentration, c , of the absorbing species on the scattering layer through the molar absorptivity, E, and the scattering coefficient S , as follows: F(R) = ( 1 - R)'/2R = €CIS (13) where F(R) is known as the Kubelka-Munk function. Instrumentation The basic instrumentation associated with optical fibre sensors is simple and requires both optical and electronic components. A typical instrumentation system employed in conjunction with optical sensors is shown schematically in Fig. 4. Apart from the fibre, it consists of a light source, photodetector, optical couplers and monochromators, modulator, signal amplifier and readout.The source of illumination must be able to provide an intense and stable radiation. Several types o f light source have been employed in optical chemical sensors, viz., incandesccnt lamps, gas lasers, light-emitting diodes (LEDs) and semicon- ductor injection lasers. Incandescent sources, such as tungsten lamps and quartz-halogen lamps, emit a broad spectrum of optical radiation and are used as sources of ultraviolet and visible light in short-range optical fibre sensors. Gas lasers are useful general purpose sources of highly intense and coherent radiation and are employed in long-distance (remote) sensing systems. The LEDs and injcction lasers are miniature sources of high-intensity monochromatic radiation; they produce incoherent light with a spectral bandwidth of 40---50 nm and are useful in short range fibre optic sensing (up to 1 km).On the other hand, injection lasers radiate a coherent beam of light with narrower spectral bandwidth (5-1 0 nm) and are excellent sources for transmission of light in fibres of greater lengths. The photodetection system is essentially a photon-counting device, where optical signals arc converted into electrical signals, which can be easily amplified by electronic means. The photodetector must possess a peak sensitivity at the mcasuremcnt wavelength, must generate a minimal amount of noise to the transmitted signal and must respond rapidly to variations in intensity of the incident light.Various types of photodetectors have been incorporated in optical chemical sensors. These include photomultiplier tubes, positive intrin- sic negative (PIN) photodiodes, photodiodc arrays and avalanche photodiodes. Optical couplers are required to focus the light beam to the optical fibre and also direct radiation from the return fibre to the photodetector. Focusing light from a source into an optical fibre is a more tedious operation. Laser sources produce a coherent beam of light with a cross-section almost equal to the cross-sectional area of the fibre, and can therefore be coupled signal source modulator c h romator Detector Fig. 4 used in conjunction with an optical fibre chemical sensor Schematic diagram of a typical and simple instrumental set-up very effectively to an optical fibre.On the other hand, LEDs and incandescent lamps radiate divergent beams of light and require lenses to focus their optical output onto the fibre. The coupling of the photodetector to the optical fibre is more easily accomplished, as the detectors have relatively large active surface areas and large acceptance angles. However, before the light is led to the photodetector, Wavelengths other than that relating to the species of interest have to be excluded. Optical resolution has to be carried out if illumination is provided by an incandescent source or an LED. Isolation of the desired wavelength can be achieved through filters or monochromators. Simple optical filters attenuate the light significantly and could reduce the sensitivity of the optical system.Monochromators offer high efficiency and can be adapted for different wavelengths. As mentioned earlier, ambient light can interfere with the measurement by gaining entrance into the sensing region of thc optical fibre. This intcrference could be eliminated by maintaining the sampling area in a dark environment, protected by a light-tight covering, or by modulating the light source so that optical sign& derived from it can be differen- tiated from an extraneous radiation. The electronic components of the instrumentation could suffer from drift during the course of operation. For instance, the intensity of the light source can vary due to ageing or due to fluctuations in the power supply unit. The effect of these variations can be eliminated by the use of internal referencing, which involves monitoring a radiation that is not altered during the sensing process.Applications Optical fibres have been used in chemical and biochemical analyses for the transmission of optical signals both with and without chemical transducer attached. Whereas in the latter (plain fibre sensor) type the analyte itself possesses some measurable optical property, in the former (chemical trans- ducer based) type, the analyte interacts with the immobilized reagent producing a change in the optical property of the reagent. The latter type is the most commonly employed design in many optical fibre chemical and biochemical sensors because many analytes themselves do not possess measurable optical properties.In the plain fibre sensor type devices, optical fibres are normally constituted as part of the instrumentation in many commercially available systems, which are intended for use in fermentation systems, bioreac- tors, explosive atmospheres, river and reservoir monitoring, etc. The chemical transducer based optical sensor allows a much wider range of application and is of potential use in all types of analytical sciences. Typical areas are pollution and process control , biotechnology , environmental and atmos- pheric monitoring, clinical analysis, defence, water analysis, etc. Many such applications have been comprehensively reviewcd".'6.'7 and it is not intended to discuss them here; the reader is advised to source the literature as required. Conclusions Optical fibres have induced a renaissance of optical methods of chemical analysis.By its combination with suitable instrumentation , optical chemical sensors can be developed. The basic instrumentation associated with optical sensors is simple and requires only standard components, which are well advanced in their state-of-the-art. Hence progress in this field of optical sensors will be dependent on the development of appropriate reagent phases for the transducer. Ideally, optical sensing devices should be characterized by high sensitivity, selectivity and reliability. Furthermore, the ability to perform measurements in real-time, in a site-specific fashion and on a continuous basis, would be more important in322 ANALYST, APRIL 1993, VOL. 118 * Y .- ! l o _ 2 9 6 7 8 9 10 11 Fig.5 Comparison of reflectance rcsponsc of a twin-probc optical pH sensor (Bromocrcsol Purplc + Bromothymol Blue) with the responsc o f a pH electrode mctcr (correlation coefficient = 0.9998) Electrode pH meter [pH] the application of these sensors to practical analysis problems. The concept of multi-sensors for the simultaneous detection of scveral species has begun to attract considerable interest. For example, optical fibre sensors for in vivo continuous measure- ments of pH, oxygen and carbon dioxide have been bundled into a single probe, for use in medical diagnostics,28 with a view to insertion via catheters. The problem of limited dynamic ranges could be addressed with the use of multi-sensors capable of measuring different concentration levels of analyte.For example, the dynamic range of pH measurements can be extended from 2 pH units with a single sensor to about S pH units using a twin-sensor probe (Fig. S).'9 Several chemical reactions have been adapted in the development of optical sensors. In many cases, the sensors have been selective and sensitive to specific analytes, and the motivation for devcloping optical chemical sensors is thus established. I t is likely that such development will continue to be a very active area of analytical research. This development also requires expertise from several areas including indicator/ reagent synthcsis, polymer chemistry, biochemistry, ana- lytical spectroscopy, fibre optics and optoelectronics, because optical chemical sensors require an interdisciplinary approach to enhance growth in this area.References 1 Kautsky. H., and Hirsch, A., Z. Anorg. Allg. Chem., 1935,222, 126. 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Chudyk, W. A., Carrabba, M. M., and Kenny, J. E., Anal. Chem., 1985,57, 1237. Klainer, S., Hirschfeld, T., Bauman, H., Milanovich, F., Perry, D., and Johnson, D., Annuul Report, LBL 11081, Lawrcnce Bcrkelcy Laboratory, CA, 1981. Wolfbeis, 0. S . , in Sensors in Bioprocess Control, cds. Twork, J . \'., and Yacynych, A. M., Marccl Dekkcr, New York. 1990. Narayanaswamy, R., Anal. Proc., 198.5, 22, 204. Narayanaswamy, R., and Sevilla, F., 111, J . Phys. E, 1988, 21. 10. Pcterson, J . I . , Fitzgcrald, R. V., and Buckhold, D. K., Anal. Chem., 1984, 56, 62. Li, P. Y . F., and Narayanaswamy. R.. Anulyst, 1989, 114. 1191. de Oliveira, W. A., and Narayanaswamy, R., Tulanta. 1092,39, 1499. Ahmad, M. B . , and Narayanaswamy, R., unpublishcd work. Pcterson, .l. I., Goldstein. S. R., Fitzgerald, R. V., and Buckhold. D. K., Anal. Chem., 1980. 52, 864. Saari. L. A . , and Seitz, W. R., Anal. Chem., 1982, 54, 821. Kirkbright, G. F., Narayanaswamy, R., and Wclti, N. A . , Anulysr, 1984, 109, 1025. Zhujun, Z., andSeitz, W. R.,Anal. Chim. Actu, 1984, 160,305. Giuliani, J . F., Wohltjcn, H., and Jarviq, N . L., Opt. Left., 1983, 8 , 54. Caglar, P., and Narayanaswamy, R., Anulyst, 1987, 112, 1285. Saari, L. A., and Seitz, W. R., Anal. Chem., 1983, 55, 667. Russell, A. P., and Fletcher, K. S . , Anal. Chim. Actu, 19x5, 170, 209. Zhujun, Z., and Seitz, W. R . , Anal. Chem., 1984, 56, 220. Wolfbcis, 0. S., and Sharma, A . , Anal. Chim. Acta. 1988, 208, 53. Urbano. E., Offcnbacher. H., and Wolfbeis, 0. S . , And. Chem., 1984, 56, 427. Wolfbeis, 0. S . , Posch, H. E., and Kroncis, H. K . , Anal. Chem., 1985, 57, 2556. Narayanaswamy, R., and Sevilla F., 111. Fresenius' 2. Anal. Chern.. 1988, 329, 789. Momin, S . A., and Narayanaswamy. R., Anal. Chim. Acfa, 1991, 244, 71. Schultz, J. S . , Mansowi, S . , and Goldstcin, I . J . , Diubetes Cure, 1982, 5 , 245. Seitz, W. R., CRC Crit. Rev. Anal. Chem.. 1988, 19, 135. Wolfbeis, 0. S., Fiber Optic Chemical Sensors and Biosensors, CRC Press, Boca Raton, 1991, vols. I and 11. Gehrich, J. L.. Lubbers, D. W., Opitz, N . , Hansmann, D. K., Miller, W. W., Tusa, J . K., and Yasufo, M., IEEE 7runs. Biomed. Eng., 1986, BME-33, 117. Guestrin, H. R., and Narayanaswamy, R., unpublishcd work. Paper 210.57188 Received October 27, I992 Accepted November 5, I992