Automatic analytical methods for environmental monitoring and control

Jenny EMNEUS

The concept of immunoassay was first described in 1945 when Landsteiner suggested that antibodies could bind selectively to small molecules (haptens) when they were conjugated to a larger carrier molecule [1]. This hapten-specific concept was explored by Yalow and Berson in the late 1950s, and resulted in an immunoassay that was applied to insulin monitoring in humans [2, 3].

The first application of based immunological technology in the environmental field was reported in 1970, when Centeno and Johnson developed antibodies that selectively bound malathion [4]. A few years later, radioimmunoassays were developed for aldrin and dieldrin [5] and for parathion [6]. In 1972, Engvall and Perlman introduced the use of enzymes as labels for immunoassay and launched the term enzyme-linked immunosorbent assay (ELISA) [7]. In 1980, Hammock and Mumma [8] described the ELISA potential for agrochemical and environmental pollutants. Since then, the use of immunoassay for pesticide analysis has increased dramatically. Immunoassay technology has become a primary analytical method for the detection of products containing genetically modified organisms (GMOs) [9].

In the 1990s, the immunoassay laboratories were faced with many challenges, which included reengineering of instrumentation, space limitation in laboratory, limited available resources for analysis, cost compression and also increased regulation of testing laboratory. Despite these challenges, the users expected the immunoassay laboratory to provide better services. In these conditions, the immunoassay laboratory needed to be come more efficient by incorporating creative solutions and adapting to changes. One solution was automation and system reintegration. Since most immunoassay procedures are labor-intensive, automation reduces the dependency of the labor equipment. Furthermore, the analysis can be performed outside of the laboratory near to "sample source", when the automatic system is portable [10].

During the last 20 years, major advances have been achieved in the automating routine, for environmental chemistry procedures. Discrete and random access analyzers provided a wide spectrum of immunochemistry tests around the clock to meet the demands of rapid testing. The first attempt was to automate radioimmunoassay (RIA) and regarding that several automatic systems were introduced in the late 1970s. Those included the Centria (Union Carbide), Concept 4 (Micromedic), ARIA II (Becton-Dickinon), and Gammaflow (Suibb), with limited throughput and testing menu, were not as reliable and cost-effective as the users wanted [10]. Automation in immunoassay became successful when non-isotopic systems were introduced.

Generally, an automated immunoassay system has the instrument (as only one part or multiple parts), the reagents and the computer. These three components are interdependent, since the format of the reagent will determine the instrument design, while the limits of the instrument design may require modification of the reagents and immunoassay procedure. The computer program could optimize the reaction conditions, the sequence of the reagent addition and the order of the sample testing. It will expedite data processing and management as well as result reporting. A system will not be successful unless all three components are functioning well as a unit, which can be named integrated system [10].

The traditional idea of automation in immunoassay is to adapt reagent to an automated instrument for immunoanalysis. Such instrumentation mechanized all the necessary steps in immunoassay procedure (e. g. pipetting, incubation, washing and detecting the signal). The automated system performs large variety of tests, mixes at the same time with fairly high throughput. The instrument can be a general chemistry analyzer when the procedure is a homogeneous immunoassay (i. e. an assay requiring no physical separation of bound and unbound antigens) or a dedicated analyzer for heterogeneous immunoassay (i. e. an assay requiring physical separation between bound and unbound antigens).

The automated homogeneous immunoassay systems use small sample size and low reagent volume, and provide fast turnaround time. The calibration curve is stable from several days to weeks, which allows performing analysis at all hours without having a recalibration of the system. The efficiency is enhanced by saving technical time, quality control and reagent expenses. In this area, an example is the TDx analyzer (Abbott Laboratories, USA) based on the PFIA (Polarization Fluorescence Immunoassay) principle. The discrimination between bound and free analyte is read indirectly involving a fluorescent tracer, which emitted differently the polarized light in free and bound form. Therefore, TDx analyzer is useful for environmental analysis (e. g. organophosphoric pesticides [11], DDT insecticide [12], propanil herbicide [13]).

The heterogeneous immunoassay is more versatile than homogeneous immunoassay in terms of automation, since there is not limitation of analyte size (small and large molecule can be analyzed). The analyte fractions (bound and unbound) are separated by a simple washing step, which can eliminate also the interfering substances present in the sample before quantification, and in this way the method sensitivity wins several orders compared with the sensitivity of homogeneous immunoassays. But, heterogeneous immunoassays are more labor-intensive and time-consuming, which indeed requires a dedicated immunoassay analyzer, semi-automated or fully automated.

The semi-automated immunoassay system is an automated instrument built on multiple blocks. These building blocks may be either linked by computer program or mechanically attached. In most semi-automated systems, these blocks function separately (e. g. the pipetting / injecting of reagents, the incubation of the reaction mixture, the bound/free separation by washing the solid phase). ELISA is one of the batch immunoassay methods, which fit well with the semi-automated systems, but the instrument is used frequently in clinical analysis (e. g. Amerlite, Kodak [14], Photon QA, Hybritech [15], Commander, Abbott [16]). For environmental analysis, the semiautomatic systems are set up usually in flow injection systems, based on heterogeneous competitive immunoassay. J. Yakovleva et al [17] reported in 2002 a semi-automated immunoassay method for atrazine analysis, which involves offline preparation of sample (mixture of analyte and enzyme tracer) followed by automated analysis. The sample is passed over the silicon chip surface with anti-atrazine antibody immobilized and then an enzyme substrate is injected to quantify the tracer concentration bound on the immuno active surface, which gives an indication on the analyte concentration from the sample. Atrazine was also analyzed using the semi-automated system consisting from a protein G modified disc composite from monolithic metacrylate and polyethylene [18]. Alkylphenol ethoxylate surfactants [19], alkylbenzene sulfonates [20], 4-nitrophenol [21] and 2,4,6-trichlorophenol [22] are other examples of automated immunoassay systems involved in environmental analysis.

Fully automated system for heterogeneous immunoassay link all the separate components of the semi-automated systems and allow the testing to be completed from sample addition to result reporting. Depending upon the ability of the system to conduct the analysis, the fully automated system can be further subdivided into batch and "in flow". An example of a batch-automated system is the portable immunochemical sensor for field screening used in TNT, atrazine and diuron determination based on ELISA principle [23]. The instrument consists in a µ-fluidic part and ground plate with automated control and one-way chip, which hosts all immunoreagents (antibody, enzyme tracer). The user has only to add the sample and the instrument will show directly the analyte concentration detected in the sample content. Another example of fully automated immunoassay system is BIACORE (Biacore, Uppsala, Sweden), which can be named "in flow" system. The instrument function is based on the SPR (Solid Plasmon Resonance) principle. It is equipped with a continuous flow system in which four channels are coupled in series. It has an automatic sample needle to deliver buffer and sample to the sensor chip surface. The continuous flow ensures that no changes in analyte concentration occur during the measurement. It can be used for kinetic measurement as well as for environmental analysis. The "River Analyzer" (RIANA) is another example from this subclass of fully automated systems, which was applied for environmental pollutant and their metabolites analysis (e.g. atrazine, 2,4-dichlorophenoxyacetic acid, isoproturon, pentachlorophenol, alachlor) [24-26]. RIANA is based on specific immunoassays for analyte recognition and Total Internal Reflection Fluorescence (TIRF) as transducer principle. The device consists of an optical detection unit, a flow cell and an integrated fluid handling based on flow injection (FIA). The sample and anti-analyte antibody labeled with fluorescence marker are mixed together and after the incubation time, the mixture is injected into the system and the unbound antibody is retained on the transducer surface covered with immobilized antigens. The bound antibody fraction on the transducer surface will indicate the analyte concentration from the sample.

The impact of automation on immunoassay operation was especially on the mechanization of the testing procedure and the consolidation of the workstations. The random access feature of the automation will facilitate the workflow and improve the turnaround time. Automation will change the function of a technologist from technician to data manager and quality control officer. Through workstation consolidation, it will reduce the labor equipment, as well as the skill level and number of workers. The ability to perform simple as well as multiple analysis with better sensitivity, accuracy and precision are other important advantages of the automation process in immunoassay, which can prove the benefits of an automated immunoassay system.

[2] R.S. Yalow and S.A. Berson, Nature (London), 184 (1959) .

[3] R.S. Yalow and S.A. Berson, J. Clin. Invest., 39 (1960) .

[4] E.R. Centeno and W.J. Johnson, Int. Arch. Allergy Appl. Immunol., 37 (1970) 1.

[5] J.J. Langone and H.v. Vunakis, Res. Common. Chem. Pathol. Pharmacol., 10 (1975) 163.

[6] C.D. Ercegovich, R.P. Vallejo, R.R. Gettig, L. Woods, E.R. Bogus and R.O. Mumma, J. Agric. Food Chem., 29 (1981) 559.

[7] E. Engvall and P. Perlmann, J. Immunol., 109 (1972) 129.

[8] B.D. Hammock and R.O. Mumma, Pesticide Analytical Methodology, (1980) 321-352.

[9] G. Shan, C. Lipton, S.J. Gee and B.D. Hammock, Immunoassay, biosensors and other chromatographic methods, Handbook of Residue Analytical Methods for Agrochemicals, (2002) 623-679.

[10] D.W. Chan, Automation of Immunoassay, Immunoassay, (1996) 483-505.

[11] A.Y. Kolosova, J.-H. Park, S.A. Eremin, S.-J. Kang and D.-H. Chung, Journal of Agricultural and Food Chemistry, 51 (2003) 1107-1114.

[12] S.A. Eremin, A.E. Bochkareva, V.A. Popova, A. Abad, J.J. Manclus, J.V. Mercader and A. Montoya, Analytical Letters, 35 (2002) 1835-1850.

[13] A.I. Krasnova, S.A. Eremin, M. Natangelo, S. Tavazzi and E. Benfenati, Analytical Letters, 34 (2001) 2285-2301.

[14] J.D. Faix, Amerlite immunoassay system, Immunoassay automation: A practical guide, (1992) 117-127.

[15] R.F. Frye, The photon-ERA immunoassay analyzer, Immunoassay automation: A practical guide, (1992) 269-292.

[16] P.P. Chou, IMx system, Immunoassay automation: A practical guide, (1992) 203-219.

[17] J. Yakovleva, R. Davidsson, A. Lobanova, M. Bengtsson, S. Eremin, T. Laurell and J. Emnéus, Anal. Chem., 74 (2002) 2994-3004.

[18] S.R. Jain, E. Borowska, R. Davidsson, M. Tudorache, E. Pontén and J. Emnéus, Biosensors & Bioelectronics, 19 (2004) 795-803.

[19] M. Badea, C. Nistor, G. Yasuhiro, F. Shigeru, D. Shin, D. Andrei, D. Barceló, V. Francesc and J. Emnéus, Analyst (Cambridge, United Kingdom), 128 (2003) 849-856.

[20] M. Franek, J. Zeravík, S.A. Eremin, J. Yakovleva, M. Badea, A. Danet, C. Nistor, N. Ocio and J. Emnéus, Fresenius J. Anal. Chem., 371 (2001) 456-466.

[21] C. Nistor, A. Oubina, D. Barceló, M.-P. Marco and J. Emnéus, Anal. Chim. Acta, 426 (2001) 185-195.

[22] C. Nistor and J. Emnéus, Analytical and Bioanalytical Chemistry, 375 (2003) 125-132.

[23] P.M. Krämer, I.M. Ciumasu, C.M. Weber, G. Kolb, D. Tiemann, I. Frese, H. Löwe and A.A. Kettrup, A new, automated, portable immunochemical sensor system for field screening, IAEAC: The 6th Workshop on Biosensors and BioAnalytical µ-Techniques in Environmental and Clinical Analysis, ENEA - University of Rome "La Sapienza", October 8-12, 2004 - Rome, Italy (2004) .

[24] E. Mallat, C. Barzen, R. Abuknesha, G. Gauglitz and D. Barceló, Anal. Chim. Acta, 426 (2001) 209-216.

[25] E. Mallat, D. Barceló, C.G. Barzen, G. and R. Abuknesha, TrAC, 20 (2001) 124-132.

[26] E. Mallat, C. Barzen, A. Klotz, A. Brecht, G. Gauglitz and D. Barceló, Environ. Sci. Technol., 33 (1998) 965.

The most common method of detection in automated analysis is colorimetry and its close relative UV-spectrophotometry; this is a natural fact because the UV-VIS absorption measurements are by far, the most usual detector in analytical instrumentation. The well-known parts of a colorimeter or an spectrophotometer are: (a) the light source, which can be as frequent as a tungsten-filament light bulb; (b) the required optics for focusing the light into a colored filter (colorimeter) or a monochromator to disperse the electromagnetic radiation according to the different wavelengths into a “rainbow” of “colors” (UV light is included in the definition) by using a prism or a diffraction grating. By rotating the prism or grating, the wavelength of light can be selected to match the wavelength with that absorbed by the sample; (d) a sample compartment to hold a transparent (glass or silica) cell, (e) a light-sensitive detector, the photomultiplier tube (PMT) to convert the light intensity into an electric current, and (f) electronics for measuring and displaying the output from the PMT. The analytical method involves the reaction of the analyte to match with two main objectives: (a) to obtain a new chemicals presenting higher molecular absorptivity to increase the sensitivity; and, (b) to avoid interferences.

Most spectrophotometric procedures can be found within any of the related methods. Most of them are the result of “translating” a classic batch procedure into an automated method. Due to that, the chemical processes are always very similar as they start from the same sample and analyte to process and have the same goal, the spectrophotometric measurement of the same final chemical. Differences among them are mostly due to the adaptation of the chemical process to the automation type.

Due to the large number of the existing automatic analytical methodologies and the also a large number of environmental interest analyses; this chapter tries to give a panoramic vision by describing the determination of different analytes in water samples with the aid of different type of automatic procedures.
III.4.1. NUTRIENTS: ENVIRONMENTAL SIGNIFICANCE AND MEASUREMENT

Nutrients are mainly compounds of nitrogen or phosphorus, although other elements (iron, magnesium, and potassium, among others) are also necessary for bacterial and plant growth.

The nutrient concentrations are very relevant for life in natural waters because, in excess, they cause nuisance growth of algae or aquatic weeds. A higher concentration of one of them involves the abnormal growth of the population of definite algae. In some industrial wastewaters treatment plants, ammonia or phosphoric acid must be added as a supplement bearing in mind that a deficiency of nutrients limits the effectiveness of biological treatment processes.
Table III.4.1. Examples of automated spectrometric methods for pollutants determination

Analyte	Automated methodology	Chemical method
Nitrite/nitrate ammonia	Lab-on-a valve	Griess modified Salycilate
Phosphate	Colorimetric analyzer	Molybdate
Calcium	Flow Injection Analysis, FIA	Ca(II) / o-cresolphthalein at pH 10.0
Cyanide	Segmented flow	pyridine + barbituric acid
Pb (II)	Sequential Injection Analysis, SIA	Sulfide-resazaurine, Pb (II) as catalyst
Chlorine	Multi-commutation	Oxidation of dianisidine
Cr (III) and Cr (VI)	FIA and r-FIA	Cr(VI)/1,5-diphenylcarbazide

Nitrogen occurs primarily in the oxidized forms like nitrate or nitrite or the reduced form of ammonia or "organic nitrogen". The latter means, the nitrogen is part of an organic compound such as an amino acid, a nucleic acid, a protein, etc. and can be used as nutrient after the organic nitrogen decomposes to a simpler form.

The classical batch spectrophotometric methods for nitrogen compounds, which have been translated into automatic methods, are the following:

Ammonia is determined by the Nessler, phenate or salycilate methods, after distillation from an alkaline solution to separate it from interferences. In recent automatic methods, the distillation is substituted by volatilization methods and separation from the matrix by a membrane separator.

Organically-bound nitrogen can be determined by the same ways after a digestion (the Kjeldahl procedure) which converts the nitrogen into ammonia.

Nitrite is determined colorimetrically by the Griess method, which suffered many changes from different authors who proposed more suitable reagents.

The most popular method for nitrate is reducing nitrate to nitrite chemically using copperized cadmium (other reducing metals or amalgams have been also proposed) then analyzing the nitrite. Nitrate can be also converted into nitrite by photoirradiation (UV region) with a low pressure Hg lamp; or, by the homogeneous way with dissolved reducing chemicals.

Trends in Analytical Chemistry are in the way to put together automation with miniaturization. A “natural son” from miniaturization of FIA and SIA is the so called Lab-on-a-valve; an automatic and miniaturized compact set-up designed by J. Ruzicka and based on the SIA principles with the aid of a syringe pump instead of the usual peristaltic pump.

The general characteristics of this novel emergent technology are previously reported in the section Application of flow techniques in environmental monitoring and control. The miniaturization reaches to the detector; a diode spectrophotometer (LED) provided with optical fibers from light source to the sample (rest of the set) and from sample to computer. The instrumental set-up (detector excepted) is always the same for any procedure or detection way; versatility is obtained through the software.

Two spectrophotometric methods for water samples have been selected to illustrate the use of this new methodology: namely, nitrite/nitrate and ammonia.
III.4.1.1. Nitrite / Nitrate
The method is based on the reaction of nitrites with sulfanilamide to form and azo dye that will then couple with N-(1-naphthyl) ethylenediamine dihydrochloride to form magenta colored solution which can be quantified in the spectrophotometer at 540 nm.

The linear working range and sensitivity of this procedure are variable, depending on minor changes like sample size and flow rates: namely, greater sensitivity is easily obtained by using larger sample volumes (over the range 80-200 L) and minor flow rate; or a method to be applied to higher concentrations by using a smaller sample (25-80 L) and fast flow rate.

Figure III.4.1. Set-up for the spectrophotometric automated determination of nitrite and nitrate. Reproduced from J. Ruzicka, Flow Injection 2nd Edition (CD) FIAlab Instruments, with permission.

1. 
   to avoid large amounts of toxic metals to the public sewage.
   
2. 
   Suspended matter or turbid samples can clog the Cu-Cd reactor. Previous filtering is advisable.
   
3. 
   The Cu-Cd column will be degraded by samples containing soluble mercury or thiosulfate. Samples containing soluble oils can inactivate the catalytic surface. A sample suspected of containing oil or grease requires a pre-treatment; oil or grease should be extracted by a liquid-liquid process with an organic solvent layer. Cadmium should be re-copperized.
   
4. 
   The lamp performs with a remarkable stability leading to improve the reproducibility.
   

The analytical figures of merit of the procedure are as follows: Linear dynamic range: 0.04 - 10.0 mg (N) L^-1, nitrate; 0.015 - 2.0 mg (N) L^-1, nitrite. Detection limits: nitrate, 1.4 g L^-1 and nitrite, 0.7 g L^-1. Deviation: 2% and Sample throughput: Maximum of 98 h^-1 (36.6 s/assay). Sample volume, 200 μL.

The problem due to the atmospheric oxidation of nitrites can be prevented for sample stored more than 24 hours; samples should be protected form daylight and added HgCl

Figure III.4.2. Absorption spectrum of the nitrite assay. Reproduced from J. Ruzicka, flow Injection 2nd Edition (CD) FIAlab Instruments, with permission.

Nitrate can be photo-reduced to nitrite in the 200 - 300 nm regions where nitrate exhibits two UV bands; a low-pressure 8 W lamp as UV source coiled with 697 cm of PTFE coil 0.8 mm internal diameters served for the goal. The pH affects the photochemical process; it is especially favorable the alkaline region. The temperature had shown a negative influence the room temperature being the finally selected one. Is important the irradiation time to fix the optimum found time period, 3 min 11 sec, is necessary to keep a flow-rate (for the selected reactor length) of 1.1 mL min^-1. The presence of sensitizers, scavengers etc was tested only best results were obtained with the presence of 0.001 mol L^-1 Na₂-EDTA as an activator for the reaction.