The potential of an electrode may be related to the concentration of a particular species if a number of conditions are met—the ‘‘clever’’ chemistryof electrode membrane manufacture make
Trang 1in the Karl Fischer determination of water), but not often in the field of environmental analysis, and therefore will not be discussed further The potential of an electrode may be related to the concentration of a particular species if a number of conditions are met—the ‘‘clever’’ chemistry
of electrode membrane manufacture makes it possible to construct probes with useful sensitivity and selectivity to individual species in solution While the tendency for a chemical reaction to proceed is measured, no current is actually allowed to flow Potentiometry using ion selective electrodes is the technique in question, and using the glass electrode for the measurement of
pH is one special (and the best known) example.
A great deal more information can be obtained by electroanalytical methods if one parameter is varied and a second measured—so-called two- dimensional measurement In this case a signal pattern rather than one measured value is used to identify, as well as quantify, one or more species in
a solution The current can be monitored as the potential across a cell is
Trang 2scanned: the approach is termed amperometry When the current is controlled by diffusive processes in solution, the technique is termed
mercury electrode these days, the modern variants of anodic and cathodic
very low concentrations of metal ions in natural waters, even in seawater Since the electrical conductivity of an aqueous solution depends on the concentration of ions dissolved in it, this parameter serves as a useful indicator of salinity, adequate for rough measurements in the field and for more precise determinations if care is taken with calibration and temperature compensation While a current is allowed to flow in this technique, the potential is very small and is reversed at millisecond intervals
so that there is no net chemical reaction on the electrode surface.
Finally, we should consider the factors that continue to make electroanalytical techniques attractive for routine measurements They are generally simple, their limitations are well understood, and, above all, they provide an electrical signal that is very easy to transmit via telemetering systems, to store electronically, and to process in a computer Thus they make possible a number of useful determinations at low concentrations in real samples.
Trang 3Figure 1 indicates that we require a means of connecting our measuring device, such as a digital voltmeter, to the two solutions The two electrodes, shown as silver wires in Fig 1, serve this purpose, and when coated with silver chloride and kept in a potassium chloride solution of constant concentration, they maintain constant potentials Any difference in the two potentials is then taken up as a small constant included in Eq (1).
potential-There are a number of practical points which arise from this simple idea:
The membrane separating the two halves of the cell should be ideal and respond only to the ion of interest (here the hydrated proton, Hþ
aq ) The success of modern ion-selective electrodes derives from the ingenuity of chemists in developing a range of membrane systems that come close to meeting this requirement (Vesely et al., 1978).
No current should flow through this cell, as this would entail flow of the ions through the membrane The digital voltmeter must therefore have a high input impedance, taking no more than a few pA of current from the cell.
Sufficient ions should be available in the two solutions to enable the membrane to respond to them Measurements in very dilute waters are, for this reason, rather difficult and will certainly entail a longer equilibration time, say 1–2 minutes, till a stable potential is obtained.
Two electrodes are always required for potentiometry, even when it appears, at first sight, that the measurement can be made with only one, as is the case with combination pH electrodes (see below) The internal reference electrode is usually not accessible to the user, but the external one must be maintained by topping up with the electrolyte from time to time A saturated calomel electrode (see below) is frequently used as a separate external reference electrode, and if the presence of traces of chloride is undesirable, then the sulfate version can be used instead.
Figure 1 Schematic diagram of a Nernstian concentration cell with membrane.
Trang 4A The Glass Electrode—Measurement of pH
Much research has gone into the development of special glasses for the glass electrode, but the key to modern electrodes lies in the substitution of lithium
in the glass for sodium, to avoid the electrode responding better to the sodium ions at high pH than to the very dilute hydrogen ions Few users these days ever think of the ‘‘alkali error’’ associated with electrodes fabricated from Macinnes and Dole’s soda glass in 1930, and we can expect
a working range from pH 1 to pH 13, with deviations becoming significant only outside this range, using the lithia–lime glass developed by Cary and Baxter in 1949 Everything you could possibly need to know about the glass electrode has been summarized excellently by Galster (1991).
A modern combination pH electrode is shown in Fig 2, and a schematic diagram of the glass membrane in Fig 3 The body may be glass or glass sheathed in plastic for greater robustness The bulb is sometimes surrounded by a plastic protecting shield to minimize the chances of breaking
it by rough contact against sample containers However, this impairs the accessibility of the electrode surface to the ions in the solution and calls for either good stirring of the solution or longer waiting time until the reading is taken Further, great care is then necessary to ensure that the electrode is thoroughly washed, e.g., with a jet of distilled water, between samples Glass electrodes will give excellent service for a working life of one to two years if looked after A few important points should be remembered: The glass membrane is thin (ca 50 mm) and is easily scratched or broken, especially if used to stir crystals when making up solutions The membrane will change irreversibly if allowed to dry out, though if caught in time, an overnight soak in 1 M hydrochloric acid might rejuvenate the very thin hydrated gel coating.
Figure 2 Combination glass electrode.
Trang 5The potential should stabilize in 15–30 seconds after immersion in a sample solution Slower response may indicate that the electrode is nearing the end of its useful life At the same time the Nernstian response is also beginning to be lost For this reason, pH meters that permit the meter plus electrode to be checked and adjusted in two different buffers are always to be preferred However, apparently slow response may be due to the behavior of the sample: outgassing of CO 2 from some water samples may cause the pH
to drift upwards because the pH really is changing.
A pH electrode should always be checked against two buffers—even
an electrode with a hole in it can be made to read pH 4, but will read that same value in all solutions.
The 0.05 M potassium hydrogen phthalate (10.2 g L1 ) buffer with
pH ¼ 4.00 and the 0.05 M sodium tetraborate buffer (19.1 g L1 ) buffer with
pH ¼ 9.20, both at 20C, are reliable and easy to prepare one-component buffers Make them up fresh each week, as they are likely to deteriorate owing to bacterial growth and to absorption of CO 2 from the air, respectively.
Transistors are three-electrode electronic devices in which a small current flowing between two of the electrodes (emitter to collector) is controlled by a second, very small, current flowing between the emitter and a third, intermediate, electrode called the base In a field-effect transistor (FET), the controlling electrode responds to potential, normally generated by the adjacent electronic circuitry, but sometimes by the external signal to be measured, for example a potential generated by a chemical electrode Thus
pH meters now use a metal oxide semiconductor FET (MOSFET) operational amplifier to measure the potential of the glass electrode without taking any current from it The extension of the MOSFET concept has been to coat the metal oxide layer with a chemically selective coating – the ‘‘membrane’’ of a chemical electrode – and to allow a carefully
Figure 3 Schematic diagram of the glass membrane in a pH electrode Shaded area: dry glass membrane electrode for fluoride.
Trang 6chosen chemical system to control the transistor current Such a device has been termed a CHEMFET, a chemically sensitive field-effect transistor, but now is more usually given the name ISFET, ion-selective field-effect transistor (Bergveld, 1972) A silicon nitride coating, for example, deposited
on the metal oxide gate results in a device that has near ideal response to hydrogen ions, with a working range of pH 1–13, and that is a great deal tougher than any glass electrode Mettler-Toledo and Thermo-Russell market ISFET pH electrodes for demanding environmental applications, but they can be used only with appropriate ISFET meters and not with conventional pH meters.
Frant and Ross’s announcement in 1966 that doping LaF 3 with a little EuF 2 resulted in a single crystal with sufficient electrical conductivity to be used as
an electrode membrane, and one that would respond ideally to fluoride ions
in solution, represented a major breakthrough in the area of ion-selective electrodes, much valued because of the difficulty at that time of determining this anion by any other route (Frant and Ross, 1966) The construction is shown in Fig 4 One crucial practical problem was how to cement the LaF 3 single crystal to the plastic body and at the same time to guarantee perfect electrical insulation Users should be warned that single crystals are brittle and will shatter if dropped on a hard surface.
The behavior of this electrode is worth discussing because it illustrates problems common to most other electrodes, and also ways of overcoming these problems.
Hydrofluoric acid, HF, is a weak acid, with pK a¼3.5, and the electrode responds to the hydrated fluoride ion Therefore all solutions must
be adjusted to a pH of 5 or greater, where the acid is effectively fully dissociated An acetate buffer is therefore added to all solutions, standards and samples alike.
Figure 4 Single-crystal membrane electrode for fluoride.
Trang 7As mentioned earlier, in high salt concentrations, ion activities deviate significantly from analytical concentrations and calibrations based
on concentration, even for low fluoride concentration, and become inaccurate The answer is to add a high concentration of sodium perchlorate
to all solutions, to maintain a constant electrolyte strength for all measurements.
Certain metal ions, notably aluminum and iron(III), form very stable fluoride complexes and will effectively mask free fluoride in e.g a river water sample, so that very low fluoride concentrations will be reported if the measured potential is converted to fluoride concentration The answer here is to add a strong complexing agent, EDTA or CHDTA, to mask the metal ions.
The fluoride electrode is therefore used with TISAB (total ionic strength adjuster buffer) being added to all solutions, making possible a working range of 0.1–100 mg L1 of fluoride (Frant and Ross, 1968) As the Nernst equation is operative, the potential, in mV, is plotted against the log 10 of the concentration, either in mg L1 or as molarity (Fig 5).
The concept of using a sparingly soluble metal salt as the responsive component of a membrane lies behind the design of many types of electrode, and the silver halides are the obvious choices for making a halide ion selective electrode The problem of making a ‘‘membrane’’ that was both mechanically strong and electrically conducting was solved by Pungor et al (1966) by compressing finely powdered silver halide with a small amount of
Figure 5 Typical calibration for determination of fluoride with a LaF 3 electrode.
Trang 8silicone rubber as binder Nernstian response was obtained over useful working ranges for chloride, bromide, and iodide, but there is a degree of response to other halides best described by the interion response factor:
For good performance and little interference, the K values should be small, ideally 0.001 or less General problems with ISEs and how to characterize their performance have been discussed by Moody and Thomas (1972) The sulfide electrode presents some difficulties in use, as free sulfide ion
is obtained only at very high pH, where oxidation of the ion is facilitated The standard procedure is to use a high pH buffer (1 M NaOH, ca pH 14) with an oxygen scavenger such as cresol, but this usually is a messy solution that covers the electrode in oxidation products It is also no solution to the problem of measuring sulfide directly in sediments, where depth profiles are
of interest in investigations of the microbial mat and the pore water composition changes in composition with depth However, as the main interference is a pH effect, it can be countered by measuring the pH and correcting the sulfide electrode potential directly, with a series of calibration graphs covering the pH range of interest (Fig 6).
Liquid ion exchangers can be held on porous glass or ceramic supports to serve as membranes for ion-selective electrodes but are nowadays more commonly mixed in with a polymer to give a solid plastic membrane, enabling a large variety of chemistries to be utilized Long-chain quaternary amines are dissolved in a viscous solvent as their ion pairs, e.g., cetyltrimethylammonium cation with nitrate or perchlorate, offering good selectivity, especially for larger single-charged anions Generally one makes the assumption that interfering ions will not be present in the test sample, as for example perchlorate when an electrode is being used to monitor nitrate
Figure 6 Sulfide calibrations recorded at different pH values.
Trang 9in river water These electrodes require attention, with regular replacement
of the liquid, but are nevertheless useful for environmental monitoring purposes, covering as they do the concentration range of interest.
Early investigations in the 1960s showed that metal ions immobilized in a PVC (polyvinyl chloride) matrix, as sparingly soluble salts, could behave as selective membranes These offer the advantage of simplicity of replacement compared with the liquid ion exchanger membranes Successful calcium electrodes, for example, were made by incorporating the calcium salt of didecylphosphoric acid along with neutral dioctylphenylphosphonate as modifier, in PVC (Crags et al., 1974) Much research has subsequently gone into designing highly specific ligands for a range of metals, in which the dimensions of the chelate formed when the long-chain arms wrap around the metal ion approach the ideal for the particular metal ion.
Glass electrodes were originally explored for determination of the alkali metals, especially sodium and potassium, particularly with medical applica- tions in mind However, though a glass electrode for sodium has long been marketed and is useful in that sodium is usually present, at least in physio- logical fluids, at one hundred times the concentration of potassium, so that potassium does not cause a significant interference, the complementary problem of the determination of potassium seemed insoluble It was only when Simon and his team (1970) showed that complexes between certain antibiotics and potassium were so much more stable than the corresponding ones with sodium, that a really selective electrode could be manufactured Valinomycin, a cyclic 6-membered peptide that displays a selectivity constant with a factor of three to four thousand in favor of potassium against sodium, has formed the basis of a range of successful commercial electrodes The combination of reagents is formulated into a plastic membrane.
Electrodes are commercially available for a few gases—ammonia and carbon dioxide in particular In fact, they do not respond in the way that the ion-selective electrodes do but are pH electrodes covered with special coatings, often of silicone rubber, that offer selective permeability to the gas
in question Thus ammonia arriving at the glass electrode surface causes a rise in pH, whereas carbon dioxide causes a lowering The working ranges are much smaller than those of the true ion-selective electrodes.
Trang 10I General Comments on ISEs
Because the response of an electrode to the determinant is logarithmic, establishing the limit of detection is a little more difficult than for linear response systems Midgley (1984) has discussed this matter, showing how the intersection of the sloping line and the low-level constant potential can help Technical data in Table 1 show that a wide range of electrodes is available, for many common ions, covering concentrations of interest in environmental work as well as in many medical applications The advantages of such probes include
The simplicity of a direct reading device requiring little or no chemical sample treatment
A wide working range, typically three orders of magnitude
Reasonable tolerance to other ions in many environmental samples Suitability for continuous monitoring using data loggers to collect measurements
The possibility of being made with very small dimensions for exploring concentration profiles, e.g., in tissue or in sediment
Noble metal electrodes respond to the redox potential of their environment without actually dissolving or corroding, a fact frequently made use of for assessing the state of the chemistry in, e.g., sediment pore water A 1-mm diameter platinum wire is sealed into an insulating sheath and can then be pushed into a wet sediment with little risk of breakage Redox electrodes are
Table 1 Examples of Ion-Selective Electrodes
Trang 11small, cheap, robust, and easily cleaned Sediment cores for study in the laboratory can be filled into lengths of 6-cm plastic drainpipe, through the side of which, at regular intervals, are drilled holes just large enough for a micro redox electrode to be inserted and left for the duration of the experiment The redox potential is a property of the solution, but it will be controlled by one chemical system (the predominating one) and be at the same time indicative of all others in the same solution As iron is commonly present in sediment pore waters, has two readily acccessible oxidation states
in solution, and does not show inhibiting kinetic effects, it is probably the indicating species, so that
of the measured E h values is, however, complicated because the standard potentials of the different couples are all pH dependent An idea of the values that may be expected in soils is given by the selected potential values
in Table 2 (Cresser et al., 1993).
Note: Potentials are with respect to the standard hydrogen electrode.
Source: Cresser et al., 1993.
Trang 12the potential of which must remain invariant, and preferably known, throughout the duration of the experiment Reference electrodes make use
of a single metal—usually silver or mercury—immersed in a solution of its ions The form of the Nernst equation appears then slightly different from that for the concentration cell used to describe the behavior of ion-selective electrodes, as one of the oxidation states is now zero, that of the pure metal itself (which is not in solution):
1 A separate chamber is constructed around the silver wire, containing the salt solution
2 A sparingly soluble silver salt is chosen—normally the chloride—and the electrolyte in the chamber, potassium chloride, is maintained at a constant, relatively high (say 0.1 M or 1 M ) concentration.
The silver concentration is then governed by the chloride concentration, since
it is usually constructed as a concentric annular chamber surrounding the glass electrode itself Clearly, for correct operation, both the glass bulb and the porous plug must be below the surface of the test solution.
The second common choice for a reference electrode is the calomel electrode, taking its title from the trivial name of mercury(I) chloride,
Hg 2 Cl 2 , also a sparingly soluble salt The commonly used ‘‘saturated calomel electrode’’ owes its name to the fact that the potassium chloride filling electrolyte is saturated, as should always be apparent from the crystals
of that salt sitting inside the electrolyte chamber The advantage of this system is that it is easy to check that the solution is indeed saturated and to know that the potential of the electrode will be that expected of it; the (small) disadvantage is that the solubility of potassium chloride in water