ISEs measure the activity of free ions in aqueous solutions.

Activity is the effective concentration - i.e. that portion of the ions which are free to take part in a given reaction - in this case coming into contact with the membrane surface. Activity is always less than concentration due to inter-ionic interactions in the solution, which inhibit the movement of ions and prevent some of them from reaching the membrane. It becomes proportionately less as the concentration increases. In practice this effect is negligible (within the error limits of the measurement) below about 0.01M for monovalent ions and 0.001M for divalent ions. The difference between activity and concentration is expressed as the Activity Coefficient.

Ionic Strength Adjustment Buffer is added equally to samples and standards to minimise any errors due to differences in ionic strength between samples and standards which will cause differences in activity coefficents which can cause the concentration to be understimated by up to 50 or 60% in the worst cases. In some cases ISABs can also include ingredients which minimise interference effects, and ensure that the pH is optimum for the ISE measurement. Furthermore, for some ions, the addition of ISAB can help to reduce the time required to reach a stable reading after immersing the electrodes in a new solution.

ISAB is not normally necessary if the total ionic strength of the samples is less than 0.01 Molar for monovalent ions or 0.001 M for divalent ions (unless required for control of pH or interference or stabilisation time) and may not be necessary at higher IS, depending on precision requirements. For more details see Electrode Operating Instructions

The recommended ISAB for each ISE can be found in the electrode specification sheet. However, it must be noted that most ISABs only increase the IS to about 0.1 Molar so will not be effective for samples which already approach this level. In this case there is no point adding more strength to the samples and it is necessary to bring the standards up to the same level by making them with a matrix similar to the samples but not containing the target ion or any that would interfere with it. Alternatively, high Ionic Strength samples can be analysed by the Standard Addition or Sample Addition method.

Ion-selective electrodes are not completely ion-specific. All are sensitive to some other ions to some extent. The degree of sensitivity to another ion is given by the Selectivity Coefficient, where a value of 0.1 indicates that the electrode is ten times more sensitive to the primary ion than to the interfering ion. A value of 1 indicates equal sensitivity to both ions.

The reference electrode outer filling solution, which is in contact with the external test solution, must not contain any ions that will interfere with the ion being detected.

The recommended reference electrode can be found in the ISE specification.

All ELIT ISEs are all-solid-state and contain no liquid or gel solutions - so there is nothing to replace.
ELIT reference electrodes contain only long-lasting gel electrolytes. The units are completely sealed and the electrolyte(s) cannot be replaced. The only need for reference electrode filling solutions is in order to keep a small drop in the protective cap to prevent the electrolyte and ceramic frit from drying out during storage.

If the signal is very erratic and jumps by tens or even hundreds of millivolts then this is probably due to minute bubbles in the reference electrode electrolyte. These can develop during transport or prolonged storage. This can normally be cured by holding the electrode firmly with the active tip pointing downwards and shaking down several times with a flick of the wrist, as with old medical mercury thermometers ( i.e. down vigorously but up gently so that the gel is propelled towards the ceramic frit and any bubbles away from it).

Large deviations may also be due to poor connections in the wiring or moisture on the contacts and in this case all connections should be checked and cleaned. Random deviations of a few millivolts may be due to contamination of the ISE membrane.
(see "How do I clean contaminated membranes" ) or to external electrostatic fields. In the latter case, if the operator or passers by are wearing static-producing clothing, it may help to ensure that every one remains still for a few seconds whilst the reading is taken.

If the slope is only a few millivolts outside specification but is stable and reproducible then the ISE can still be used satisfactorily, although the lower the slope the higher the errors on the measurement of activity (or concentration).
See "What precision can I expect from an ISE measurement".

If the slope gradually becomes lower each time it is measured then the membrane is probably contaminated and should be rejuvenated by cleaning.
See "How do I clean contaminated membranes".

Crystal membranes can be cleaned by washing with alcohol to remove any organic deposits and gently polishing the crystal surface with fine emery paper to remove any obvious markings or discoloration. Then wash with de-ionised water and wipe dry with a low lint tissue, ensuring that no particles are left on the membrane. The crystal surface should be smooth and shiny. PVC membranes must not be abraded, or even touched. These can also be washed in alcohol and then regenerated by soaking in the appropriate 1000 ppm standard solution overnight, or even for several days. They must only be gently dabbed dry after washing (not wiped) to avoid scratching the surface. The ceramic frits at the tips of reference electrodes can also be cleaned with alcohol to remove organic contamination, and soaked in 0.1Molar HCl to remove other deposits.

Most electrode systems require about three or four minutes to reach a completely stable reading. Nevertheless, most electrode combinations get to within one or two millivolts of the final value in less than thirty seconds, so it depends on your precision requirements as to whether you wait for complete stabilisation or not.

Precision is the reproducibility of repeat measurements of the same sample. This is different from accuracy which is how near is the result to the true value. Ideally, results of any experimental measurement should give values that are accurate within the precision limit but this is not always true for ISE measurements. Assuming that there are no variable systematic errors from interference or activity coefficient effects (which also affect the accuracy), the precision depends on the error in the measurement of the electrode potential (mV) and the slope of the calibration line (mV/decade of concentration). These errors are unlikely to be better than about ± 1 mV and are equivalent to ~4% error in the concentration for monovalent ions (slope~54) and ~8% error for divalent ions (slope~27), when working in the normal linear range of the electrodes.
However, if there are added problems of interfering ions or high ionic strength or variable temperature or variable sample motion (as may be encountered in direct field measurements in natural samples) then the errors may be higher and a value of ± 10 or 15% may be more realistic. Nevertheless, it has been shown that, under optimum conditions in the laboratory, a precision of better than ± 2% can be achieved by making frequent recalibrations and taking multiple readings of the electrode signals. For highest possible precision and accuracy the Standard Addition Method should be used.

Drift is the gradual change in electrode response over a period of time. The rate and extent of drift can vary depending on which particular ISE and reference electrodes are being used and the age and degree of contamination of the electrodes. If a series of samples are measured repeatedly it will be observed that each successive reading of the same solution will be slightly different to the previous one. This problem can be overcome by frequently measuring one calibration solution in between sample measurements, then re-calibrating when the mV value has drifted beyond an acceptable level - depending on precision requirements. Once the extent of drift is known then it should be possible to find the optimum time of use or number of samples analysed before a calibration is necessary.

13.) What is the effect of temperature change on ISE measurements ?

Unfortunately this is a complicated relationship which cannot be simply quantified in terms of mV change per °C since the effect is different at different concentrations and actual value of the mV. Moreover, the electrode slope, the liquid junction potential, and solubility of the salts in the reference system all vary with temperature. However, the magnitude of the effect of temperature change on the slope can be calculated from a modified form of the Nernst equation (Slope = 2.303RT/nF = T x Constant) to be about 3.4% per 10°C - ie: if the slope is about 55mV/dec then a 10°C rise will increase this to about 57mV/dec. Thus, in order to avoid any errors due to temperature change it is advisable to recalibrate with standards at the same temperature as the sample solutions if the sample temperature deviates by more than about 2°C from the original calibration temperature.

This depends on the precision requirements for the results and the rate of drift of the electrode system. If only an ‘order of magnitude’ measurement is required then it may only be necessary to calibrate once a day or even less frequently. Apart from drift, any large temperature changes (greater than 2°C) will cause a change in the calibration. For the most precise results, it may even be beneficial to calibrate between every sample.

These figures are quite different depending on which ion is being measured. The lowest detection limit is about 0.005ppm for Sulphide ion and the highest about 1ppm for Chloride. The total measuring range for each ISE is given in the individual specification sheets where the lower figure is the detection limit - but note that these limits are necessarily only rough estimates of what is possible since the errors escalate dramatically in the non-linear range as the slope reduces and the detection limit is approached. The best precision can only be achieved in the linear range of the electrode.

The lower limit of linearity is also difficult to define precisely and will vary slightly depending on the individual ISE/Reference electrode combination and particular laboratory conditions. In some cases it can be an order of magnitude higher than the detection limit. If users wish to analyse samples near to this value then it is necessary to make measurements with their own set of electrodes to determine what is the lowest limit for acceptable results in their own particular application.

The upper concentration limit is often quoted as 1 Molar, but in practice it is difficult to obtain reliable results above about 0.1 Molar because of uncertainties in the effect of high ionic strength on the activity coefficient.

In general, for overnight or longer storage, ISEs should always be rinsed with de-ionised water and gently dabbed dry with a low-lint tissue, and the black plastic cap should be replaced to protect the membrane from atmospheric oxidation/corrosion.

Reference electrodes must be prevented from desiccation whenever they are not in use by replacing the protective cap containing a small quantity of the appropriate outer filling solution.

However, if the electrodes are being used on a daily basis, with several analytical sessions each day then it is not necessary to replace the protective caps as long as they are left hanging in the electrode holder and protected from physical damage. In this situation it is probably most convenient to keep both the electrodes immersed in a standard solution which lies within the concentration range of the samples being analysed when not in use.

BUT NOTE that no electrode should be left soaking in pure water for more than a few minutes at a time.

ELIT Ion Selective Electrodes are exceedingly robust and durable. The shelf life of a new unused solid-state ISE is many years - providing that the protective cap is kept in place to limit any atmospheric corrosion. They are guaranteed for six months during normal operation, but in practice have been found to last much longer than this.

The maximum useable life of Ion Selective electrodes depends very much on whether they are used for continuous monitoring or intermittent measurement of individual samples (in this case also on the frequency of use) and also on the nature of the samples. Prolonged exposure to solutions outside the optimum pH range for the electrode or to pure water or to solutions which do not contain any of the target ion will shorten the life. Also if the electrode membranes become coated with a deposit or organic film. Crystal membranes can be rejuvenated by gentle polishing with fine emery paper. Deposits on PVC membranes can be washed off with alcohol.

However, Reference electrodes cannot be expected to last as long as all-solid-state ISEs because of the eventual drying out or dilution of the internal gel electrolyte which is in direct contact with the sample.

In the Nico2000 test laboratory we have reference electrodes and ISEs which are at least 5 years old but still working fine. But these are only used intermittently for small batches of individual samples, sometimes with several months storage between use. ISEs are stored clean and dry with tight fitting plastic caps which prevent mechanical damage and atmospheric corrosion/oxidation. Reference electrodes are stored with the plastic cap containing a little solution with the same composition as the internal electrolyte absorbed on a small cotton wool pad. This is checked frequently to ensure they do not dry out.

Many electrodes have been sold to customers for continuous monitoring. For example, one user in the USA has installed our Chloride ISEs at 50 different locations in California and New Mexico to continuously monitor concentrations in mountain streams. These have performed well for more than two years without any problems. Thus we are confident that even with continuous immersion they at least last as long as the warranty period (6 months).

ISE operation is based on ion-exchange / charge-transfer principles. Thus there is hardly any consumption of materials or wear and tear on components when samples are analysed. Assuming that the electrode is not exposed to damaging chemicals, the only deterioration is due to the gradual corrosion (oxidation, hydration, leaching) of the membrane, the plastic body, and the adhesive used to attach the membrane to the body (or to physical damage to the membrane surface or to the pin connecting to the electrode head). The extent of the corrosion is dependent on the length of time the electrode is immersed or exposed to the atmosphere rather than the number of samples analysed.

The measured mV is essentially the potential difference between the membrane potential (generated by the passage of the target ions) and the reference electrode potential. But this is complicated by the fact that there are also contributions from other potential differences that may be generated at any of the metal-metal, metal-liquid or liquid-liquid junctions in the circuit. In the case of All-Solid-State ISEs, one of the internal Junction Potentials is stable but can be hundreds of mV and is different for different ions.

The Liquid Junction Potential of the Reference Electrode can be quite variable and unstable and it is re-established each time the electrodes are immersed in a new solution. Thus it is quite common to see a small difference in reading even when the electrodes are re-immersed in the same solution. This is the main source of random error in ISE measurements. Large variations can also occur between different ISEs of the same ion (particularly when they are old and well used). The reading is also different with or without ISAB and with different reference systems.

When the electrodes are immersed in pure water or any other solution which does not contain the target ion there is no membrane potential. The measured voltage is the difference between the various junction potentials in the ISE, and the reference voltage. These can be highly variable and unstable and do not represent a "blank" as in many other chemical measurements. Thus the measured voltage is never zero when no target ion is present, and can be negative even when it is present. It all depends on the difference between the ISE membrane voltage and the sum of all the other voltages in the circuit.

Thus it is not possible to predict with any accuracy what mV reading will be seen when a particular ISE is immersed in a particular concentration of the target ion. It must be stressed that the actual mV reading is not important and can vary over tens of mV. Only the difference in mV between two solutions with different concentrations is the critical factor in ISE calculations. This defines the slope of the electrode which is essential for calibration and measurement and can be used to asses the quality of measurements. The easiest way to determine the slope is to find the difference in mV between two solutions with concentrations (strictly, activities) which are one order of magnitude apart. This is expressed as mV/decade.

As a rough guide, please see below a list of the most common ISEs together with typical mV values for 1 and 10 mMol solutions obtained during initial testing of new electrodes. For comparison, the equivalent ppm values are also given. The electrode Slope is calculated from these data. All measurements were made using a gel-filled Single Junction reference electrode at 20^oC with no ISAB.

Ion	Symbol	Ion Wt.(g)	mV in 1 mMol	ppm	mV in 10 mMol	ppm	Slope
Ammonium	NH4+	18	360	18	417	180	+57
Barium	Ba++	137	232	137	257	1370	+25
Bromide	Br-	79.9	45.4	79.9	-10.8	799	-56
Calcium	Ca++	40.1	423	40.1	449	401	+26
Cadmium	Cd++	112.4	-120	112	-92.6	1124	+27
Chloride	Cl-	35.5	188	35.5	133	355	-55
Copper	Cu++	63.5	226	63.5	252	635	+26
Fluoride	F-	19	-422	19	-477	190	-55
Iodide	I-	127	-175	127	-233	1270	-58
Lead	Pb++	207.2	-113	207	-88.3	2072	+25
Nitrate	NO3-	62	380	62	326	620	-54
Nitrite	NO2-	46	275	46	218	460	-57
Perchlorate	ClO4-	99.5	414	99	358	995	-56
Potassium	K+	39	402	39	458	390	+56
Silver	Ag+	107.9	422	108	479	1079	+57
Sodium	Na+	23	320	23	378	230	+58
Sulphide	S--	32	423	32	479	320	-56
Thiocyanate	SCN-	58	65.5	58	4.4	580	-61

Top

CCR/07/08/03 - Last Update 28 Sept. 2016