EP1393053A1

EP1393053A1 - Ion sensitive electrodes based on oxa-azamacrocycles as ionophore for the determination of nitrate, salicylate, or periodate anions

Info

Publication number: EP1393053A1
Application number: EP02730426A
Authority: EP
Inventors: Paul Kong Thoo Lin; Albert N. Araujo; Maria Da Conceicao Branco S. Mendonga Montenegro
Original assignee: Robert Gordon University
Current assignee: Robert Gordon University
Priority date: 2001-06-05
Filing date: 2002-06-05
Publication date: 2004-03-03
Also published as: WO2002099409A1

Abstract

An apparatus and method for determining the presence and/or the concentration of nitrate ions is disclosed. The apparatus includes an ion selective electrode coated with a macrocyclic free base such as 1,7,11,17-tetraoxa-2,6,12,16-tetraazacyclolicosane (RN=235092-95-2) or 1,6,10,15-tetraoxa-2,5,11,14-tetraazacyclooctodecane (RN=236104-63-5). The apparatus can determine the presence of nitrate at low concentrations and has good selectivity towards nitrate ions when in the presence of certain other ions. The invention also provides a method and apparatus for determining the presence and/or concentration of salicylate and periodate ions.

Description

ION SENSETIVE ELECTRODES BASED ON OXA-AZAMACROCYCLES AS IONOPHORE FOR THE DETERMINATION OF NITRATE, SALICYLATE OR PERIODATE ANIONS

The present invention relates to a method and apparatus for determining ion presence and, preferably, concentration. The present invention is particularly applicable to determine the presence and/or concentration of nitrate, periodate or salicylate ions.

Nitrates are found in many environmental sources. The most important sources of contamination by nitrates are from the excretions of farm animals, artificial fertilisers, and industrial and domestic waste. Foodstuffs and beverages can thereby accumulate great quantities of nitrates depending on the fertilisation procedures.

Nitrate has also traditionally been used as an additive conferring positive benefits in the food industry, for example, to prevent the botulism produced by the bacteria Bacillus Clostridium Botulinum, to slow lipid oxidation and the production of unpleasant smells, and to prevent the swelling of cheese.

Nitrate's most frequent toxicological risk arises from the reduction of nitrates to nitrites which oxidise Fe(II) of haemoglobin to Fe (III) , thus preventing the haemoglobin from forming a bond with oxygen; this can lead clinical symptoms such as cyanosis, hypoxia, dyspnoea, tachycardia, unconsciousness, and metahaemoglobinaemia. Nitrites are also liable to react with specific amines, giving nitrosamines, which have been proved to be highly carcinogenic.

Nitrates have been previously measured by spectrophotometry . This is typically based on the measurement of the absorption of the end products after the nitration of phenolic compounds such as those used in the xylenol orange method, the chromotopic acid and phenolsulphonic acid methods. Such nitrate determinations are indirectly performed since they actually measure the presence of ammonia or nitrite, the reduced form of nitrate, and are generally time consuming.

Ion chromatography (IC) is another instrumental technique commonly used. A mixture is injected into a column and the retention times of the various constituents of the mixture are measured and compared with retention times for known ions/constituents. With this technique the determination of nitrates in different matrices (water, beers, infant formulas, soil, vegetables and tooth products) has been developed in the last few decades. The success of conventional analytical techniques for the determination of nitrates depends on high reproducibility of the results or on the simultaneous determination of nitrate in the presence of other anions such as nitrites, chlorides, fluorides, sulphates, etc.

In known nitrate ion-selective electrodes belonging to the liquid membrane group an organic ion- exchanger is dissolved in an appropriate solvent, which is then placed inside a glass or plastic tube. At the bottom of it there is a porous thin membrane (e.g. cellulose ester, glass etc.) that separates nitrate sensor from the sample solution to be analysed. In contact with the membrane a second electrode (e.g. AgCl/Ag) is immersed in a reference solution containing nitrate ion which acts as an inner reference electrode.

Polymeric membranes were introduced to improve the mechanical robustness of the liquid membrane electrodes, and polymeric supports such as polyacrylates, silicon rubber, polyurethane foams and PVC .

A common characteristic found in every electrode described before is related to the use of an inner reference solution and the corresponding reference electrode. Attempts to eliminate the use of this solution and electrode include coating a conductive wire with a polymeric membrane .

The main problems encountered with these conventional nitrate detecting electrodes are the determination of very low nitrate concentrations (below 5.10^"5 - 1.10^"4 mol/L) and their poor selectivity.

The determination of the presence and concentration of periodate and salicylate ions is also of use. Periodate ions have been measured before by techniques such as spectrophotometric, polarographic, potentiometric and more recently by capillary electrophoresis . Salicylate ions had been measured by ion chromatography and GC/ S (gas chromatography and mass spectrometry.

According to the present invention there is provided apparatus for the determination of at least one of the presence and concentration of ions in a sample, the apparatus having an electrode comprising a macrocyclic compound.

The invention also provides a method of determining at least one of the presence and concentration of ions in a sample, the method comprising the steps of exposing the sample to a macrocyclic compound and determining potentiometric changes in the macrocyclic compound. Preferably, the apparatus and method can determine the amount of ions present .

The ions are typically nitrate ions. However the invention can also be used to determine the presence of periodate or salicylate ions.

Typically, the macrocyclic compound is a macrocyclic base.

The macrocyclic base is typically supported on a conductive substrate, typically a conductive metal such as copper, although other metals are also useful for this purpose. Good substrates also include non-metallic substances such as graphite.

Preferably, the macrocyclic base comprises a solid, more preferably a solid membrane.

The base is typically layered onto the conductive support. There may be a single layer or alternatively a plurality of layers. The base is typically dissolved in a solvent such as dibutylphthalate and mixed with an immobilising agent such as PVC, so that when the base-solvent- immobiliser is layered onto the substrate the solvent can evaporate and leave a thin layer or "membrane" of immobilised base on the surface of the conductive substrate.

In preferred embodiments the macrocyclic base is dissolved in an appropriate mediator solvent such as dibutylphthalate with a lipophilic additive such as tetraoctylammoniumchloride . The resulting mixture can be immobilised on an inert polymeric matrix such as PVC dissolved in anhydrous THF. The solution can then be applied dropwise on a conductive support (e.g. graphite glued on with Araldite or silver epoxy resin) if an electrode without an inner reference solution is required.

The substrate is typically housed in a plastic housing along with the necessary electrical connectors to facilitate determination by either conventional means i.e. manually measuring conductivity, or by flow procedures i.e. measuring conductivity while solution is flowing.

The base is typically a heteromacrocyclic base.

Preferably, the heteromacrocyclic base includes a nitrogen heteroatom, more preferably two nitrogen heteroatoms .

Preferably, the heteromacrocyclic base includes an oxygen heteroatom, more preferably two oxygen heteroatoms.

In alternative embodiments, other suitable heteroatoms may be used, e.g. P and S.

In particularly preferred embodiments the heteromacrocyclic base is an oxa-aza cyclic alkane. The base may be of a general formul : -

wherein m and n are independently, 2, 3, 4, 5 or 6 , In preferred embodiments n is 3 and m is 2 or 3.

Preferably, the base is of the formula :-

In other embodiments the base is of the formula:

Heteromacrocyclic bases show very good working characteristics with low detection limit of 10^"7 M. The selectivity of such electrodes to nitrate ions is very good since interference from other ions such as chloride, bromide, sulphate and nitrite does not affect measurement of nitrate concentration in solution.

An embodiment of the present invention will now be described by way of example and with reference to the accompanying drawings, in which,

Fig 1 is a schematic representation of a sequence of steps for construction of an electrode housing; , Fig 2 is a schematic representation of a sequence of steps for construction of a conductive substrate for the fig 1 electrode; Fig 3 is a schematic representation of constructed electrode; Fig 4 is a schematic representation of an electrode showing the application of the base; Fig 5 is a graph showing typical calibration curves for two 3333 free base electrode units used to detect the presence and/or concentration of nitrate ions; Figs.6a-6d are graphs showing typical calibration curves for four 3232 free base electrode units used to detect the presence and/or concentration of nitrate ions; Figs. 7a-7d are repeat results of the calibration curves shown in Figs. 6a-6d; Figs. 8a-8d are a second set of repeat results of the calibration curves shown in Figs 6a-6d; Fig. 9 is schematic representation of the synthesis of the 3232 free base; Figs. lOa-lOd are graphs showing typical calibration curves for two 3333 free base electrode units used to detect the presence and/or concentration of salicylate ions; Figs, lla-lld are graphs showing typical calibration curves for two 3333 free base electrode units used to detect the presence and/or concentration of periodate ions; Fig 12 is a Reilley diagram showing the effect of pH on the response of a 3333 free base electrode; and, Fig. 13 is a graph demonstrating the apparatus and method of determining nitrate concentration according to the present invention compared with a known technique.

Referring to the drawings, an electrode housing 1 comprises a set of Perspex cylinders. A middle 130mm long body Cm of outer diameter 10mm and inner diameter 7mm houses an inner plate support cylinder Ci of outer diameter 6.5mm and inner diameter 3.5mm, and is capped by an external 20mm long cap Ce with inner diameter 10mm and outer diameter 15mm as shown in fig Id. Before insertion of the substrate support Ci into the body Cm, a plate 5 is attached to the support with adhesive (e.g. Araldite TM) and a wire 6 is soldered to the plate 5, or is alternatively attached with a conductive adhesive such as a mixture of Araldite and graphite. The plate is typically copper but can be any other conducting material e.g. graphite. The wire 6 extends through the top of the cylinder Cm and the cap Ce and is secured to the cap Ce with Araldite. A BNC terminal is connected onto the other end of the wire 6 which is shielded by a cable (not shown) outside the cylinder Cm.

The housing 1 including the plate 5 is inverted - as shown in Fig. 2d - and a macrocyclic, electroactive free base solution, described in more detail below, is then applied to the plate 5. The solvent of the solution - in one embodiment being tetrahydrofuran (THF) - is allowed to evaporate leaving the free base layered on the plate 5 as shown in fig 4b. Typically the solution containing the electroactive free base is applied dropwise onto the plate 5 and the THF evaporates leaving a free base membrane layered on top of the plate 5.

PREPARATION OF THE FIRST ELECTROACTIVE FREE BASE

The first (3333) electroactive free base to be applied to the plate 5 is typically made as described in (Kong Thoo Lin, P. et al . Synthesis, 6, 1034-1038, 1999) which is incorporated herein by reference, with modifications as follows/ STEP 1

In a 500 cm³ round bottom flask, N- hydroxyphthalimide (36g, 0.22M) was dissolved in anhydrous dimethyl formamide (150 cm³) (DMF) followed by the addition of 1 , 3 -Dibromopropane . (17.97g, 0.089M). While stirring, triethylamine (0.225M, 22.5g, 31 cm³) was added dropwise. The reaction mixture was left stirring at 85-90°C overnight.

The solvent DMF was removed at the rotary evaporator (rfe) followed by the addition of water (300 cm³) to form a thick precipitate, which was filtered using a large Buchner funnel . The precipitate was sequentially washed thoroughly with water; saturated solution of sodium bicarbonate; dilute NaOH (1M) ; and water.

Finally the precipitate was flushed with methanol followed by diethyl ether. The precipitate was left to dry (3-4 hour) in an oven under vacuum at 60°C. The crude product (26.11g, 0.713M, 80%) was pure enough (as shown by NMR) for the next step. STEP 2

1, 3-Bisphthalimidooxypropane (26.11g, 0.07133M) was added to a solution (120 cm³) of concentrated hydrochloric acid/glacial acetic acid (20:30 ratio) to form a suspension. The latter was left refluxing overnight with stirring to form a clear solution. After cooling, all the solvent was removed at the rfe to give a white precipitate. Water (100 cm³) was added to the precipitate, which was filtered and washed thoroughly with dilute hydrochloric acid (3M) . All the filtrate was collected and the solvent evaporated at the rfe. The precipitate formed was crystallised from ethanol and a few drops of water to give colourless crystals (9.75g, 0.0645, 90%).

STEP 3

lNH ts

Mts-CI

Bisaminooxypropane dihydrochloride (5.41g, 0.0302 M) was dissolved in pyridine (120 cm³) . Mesitylene sulphonylchloride (16.55g) was added to the solution and the mixture was left stirring overnight at room temperature. All the solvent was removed at the rfe to give a precipitate, which was dissolved in chloroform (200 cm³) . The latter was transferred in a separating funnel (500 cm³) and the organic phase was washed with (I) dilute hydrochloric acid (1M) (ii) water (iii) saturated solution of sodium bicarbonate (iv) water. The chloroform layer was dried with sodium sulphate and filtered. Removal of the solvent gave a solid that was recrystallised from toluene to the final product (12.21g, 0.0259 M, 86%) .

STEP 4

MtsHN

\ D F/K₂C0₃/RT

The bismesityleneoxyamino derivative (4.7g, 0.01M) was dissolved in DMF (100 cm³) followed by the addition of l-bromopropan-3-ol (2.48g, 0.021 M) . Anhydrous Potassium carbonate (0.05 M, 6.9g, 5xs) was added and the suspension was left stirring overnight at room temperature. The solvent was removed at the rfe and the precipitate was dissolved in water, transferred in a separating funnel and extracted with chloroform. The organic layer was washed with (i) saturated solution of sodium bicarbonate (ii) water. The chloroform layer was dried with sodium sulphate and filtered. Removal of the solvent gave the diol (6.81 g) in quantitative yield. STEP 5

The diol (6.64g, 0.01133M) was dissolved in anhydrous THF (100 cm³, freshly distilled over calcium hydride) . The following reagents are then added in the following order (I) triphenyl phosphine (6.09 g, 0.0226M) (ii) bismesitylene derivative (5.33 g, 0.01133M) (iii) DIAD, diisopropylazodicarboxylate (4.699g, 0.0226M). The reaction was exothermic on addition of the DIAD and the solution was left stirring at room temperature for 2-3 hours after which a precipitate was formed. All the THF was removed at the rfe and ethanol (100 cm³) was added to the residue. The suspension was warmed and then allowed to cool on ice. The precipitate formed was filtered off and washed with cold ethanol. After drying in a vacuum oven, 7.15g (62%) of the protected macrocycle was obtained.

STEP 6

The protected macrocycle (6.15g, 6.02mmol) was dissolved in freshly distilled dichloro ethane (128 ml) followed by the addition of 20% hydrobromic acid in glacial acetic acid (26g) . The reaction mixture was left stirring overnight to form an orange precipitate. The latter was filtered off using a Buchner sintered funnel . The precipitate was washed with (i) dichloromethane (ii) ethyl acetate and (iii) ethanol to give finally a pale yellow solid (2.20g, 60%) .

STEP 7

A solution of methanolic KOH was prepared was adding KOH (13.5 mM, 0.8g) to methanol (11 cm³). The macrocycle salt (1.5g, 2.4 mM) was added while stirring. The suspension was left stirring at room temperature for 2 hours. After that time, the suspension was filtered off and washed with chloroform. The filtrate was evaporated and resuspended with chloroform. The solution was filtered again and after removal of the solvent at the rfe, the 3333 free base (0.77 g) was obtained as a thick yellow oil.

PREPARATION OF THE SECOND ELECTROACTIVE FREE BASE A second free base (hereinafter the 3232 free base) has also shown good response to detecting the presence and concentration of nitrate and other ions. The 3232 base has the structure detailed below:-

The 3232 Free Base

The 3232 free base is marginally smaller than the 3333 free base. The 3232 oxazane macrocycle system (shown above) was synthesised as shown in Fig. 9 and as detailed in the aforementioned document - Kong Thoo Lin, P. et al . Synthesis, 6, 1034-1038, 1999.

In summary, an alkylation reaction between α,ω- bis (2 -mesitylsufonyl) aminooxy] propane 11 and 3- bromoethananol gave the corresponding bis-alkanol 12 which was condensed with^' another molecule of α,ω-bis (2-mesitysulfonyl) aminooxy] propane under Mitsunobu conditions to give a protected macrocycle 13. The latter was deprotected with 30% HBr/HOAc to give the macroheterocyclic as the tetrabromide salt 14 which was converted into its free base 15 with methanolic KOH ( 66% yield) . In all the steps described above, thin layer chromatography was used to monitor the reaction and also to determine the purity of products. In each step a full set of spectroscopic data (proton and carbon NMR, IR and mass spectra) was obtained for full characterisation.

Construction of the 3333 free base Ion Selective Electrode A sensor membrane incorporating the 3333 free base was prepared as follows. 0.0453 g of 3333 free base as ionophore was dissolved in 0.1517 g of dibuthylphtalate as solvent mediator. About 0.0020 g of tetraoctylammonium chloride as additive was added to the previously prepared solution.

The mixture referred above was immobilised in 0.15g of PVC previously dissolved in THF. This solution allows the coating of about 3-4 electrodes.

The solution above prepared was dropped on the conductive graphite or copper substrate plate 5 of the electrode. After THF evaporation a thin membrane was formed.

Construction of the 3232 free base Ion Selective Electrode For membrane electrode preparation a sensor solution was obtained by dissolving 0.0453 g of the [3.2.3.2] -oxazane free base in 0.1517 g of dibuthylphtalate as mediator solvent. Around 0.0020g of tetraoctylammonium chloride as additive was added to the previously prepared solution. The sensor system was immobilised in PVC (0.15g) previously dissolved in THF (6 cm³) .

Conventionally shaped electrodes were prepared as described for the 3333 free base by applying dropwise- the membrane on a support consisting of graphite powder- reated epoxy resin.

The constructed electrodes were conditionated by soaking in 10^'1 M potassium nitrate solution and when not in use were stored in a solution with a concentration of 10^"4 M.

All the reagents were of an analytical or similar quality and were used as purchased with no additional purification. The more concentrated solutions were prepared by accurate weighing of the corresponding solids followed by dilution with deionised water (conductivity less than 0.1 μS cm^"1.) The less concentrated solutions were obtained from the concentrated solutions and prepared daily. The ionic strength of the solutions was adjusted with 0.033 M potassium sulphate. For adjusting the pH and ionic strength of solutions during calibration procedures, a buffer solution of NaCH₃COO/ CH₃COOH (1=0.1 M and pH=5) was used.

The conventionally-shaped electrodes were typically prepared in accordance with a technique described elsewhere [J.L.F.C. Lima, M.C.B.S.M. Montenegro and A. M. Roque da Silva, J. Flow Injection Anal., 7, 19-33 (1990 )] by applying the base dropwise to the substrate to dry and form membranes over a graphite and epoxy resin support of the electrode.

EVALUATION OF ELECTRODES Apparatus A 2002 Crison digital potentiometer (sensitivity ± 0.1 mV) coupled with an Orion 605 switcher was typically used for potentiometric measurements. Sensitivity of 0. ImV was preferred for potentiometric measurement. These measurements determine the conductivity of the free base membrane which in turn determines the concentration of ions under investigation.

An indicator electrode was used in conjunction with an Orion 90-02-00, silver chloride/silver double- junction electrode as the reference electrode. The external compartment of the latter electrode contained an electrolyte with low interference such as 0.033 M potassium sulphate since the sulphate anion showed reduced interference, as determined by measurement of the potentiometric selectivity coefficients. pH measurements were carried out with a Philips GAH 110 glass electrode.

ELECTRODE BEHAVIOUR The working characteristics of the conventionally- shaped electrodes constructed were evaluated by repeated calibration curves in solutions without and with ionic strength adjusted to 0.1 M by using potassium sulphate. The calibration curves were made by adding known volumes of 0. IM of nickel nitrate or other nitrate salt to water or to the ionic strength adjuster.

Table 1 below shows the calibration parameters and response characteristics obtained for the 3333 free base constructed electrodes. Characteristics Electrode Electrode I II

LID (M) 10e-7 10e-7 LLLR (M) 10e-6 10e-6 Slope (mV/dec) 5 .8±1.2 60+0.1 Reproducibility (mV/day) ± 2 ± 1 Response time (s) 30 20 LID-Limit of detection LLLR-Lower limit of linear range I-Calibration performed without ionic strength adjustment; II- Calibration performed in solutions with the ionic strength adjusted

An advantage of the electrodes shown above and certain other embodiments of the invention is that the response times for the electrodes shown above are smaller than the response times for other nitrate detectors.

Another advantage of certain embodiments of the invention is their detection of ions at a low concentration e.g. the lower detection limit for the electrodes shown in table 1 above are 10e-7M whereas known nitrate detection electrodes have a lower detection limit of around 10e-5M.

POTENTIOMETRIC SELECTIVITY COEFICIENTS

The selectively of the free base towards nitrate ions was demonstrated by Potentiometric Selectivity Coefficients. In these experiments the concentration of the nitrate ions was held constant whereas the concentration of the other interferent ions were varied. The results obtained are represented in Table 2.

Table 2 - Potentiometric Selectivity Coefficients for nitrate selective electrodes

The results show the excellent selective characteristics of the electrode in the presence of the other ions quoted, especially for chloride and nitrite.

The general working characteristics and potentiometric selectivity coefficients of the 3333 free base ion selective electrode towards nitrate 1 ions were repeated. The results are shown in Table

2 3 below.

Table 3 General working characteristics of the nitrate ISE

Characteristics ISE

Lower limit of linear response (mol L^" ) 4,2 -NT

(8,2 -10 ^■Λ--7'-) with 1=0

Practical limit of detection (mol L^" ) 3,4- 10^"6

(6,9 ^• 10-') with 1= =0

Slope (mV log^C) 60±0.1 (54.8±1.2)withl= =0

Response stability (mV d^'1) ±0.3

Response time (s) lO^-lO^molL^"1 12

10^"3-10^"2 molL^" 11

10^"3-Ϊ0^"2 molL^"1 10

Working pH range

10^" mol L^"1 2,5-6,0 lO^molL^"1 2,0-7,5 lO^molL^"1 2,0-12,0

Lifetime (months) >10

Potentiometric selectivity

Coefficients log(K^pot ,ι) X lO^molL¹ 10^"3molL^"1 10^"2 mol L^"1

SO₄ ⁼ -4.5 -5.3 -5.9

F^" -2,5 -3,3 -4,3

Cl^" -2.1 -2.5 -2.7

Br -1.2 -1.3 -1.3 r -1,4 -0,2 -1,6

HCO₃ ^" -0.8 -0.5 -0.9

Ac^" - 1.0 -1.8 -2.0

NO₂ ^" -1.8 -1.9 -18

PO₄H₂ ^" -1.1 -1.6 -1.9

PO₄rT² -1.9 -15 -1.6 The 3333 electrode response characteristics shown in table 3 were evaluated on the basis of calibration curves obtained by measuring the e.m.f .values of a series of nitrate solutions with concentrations between 10^"6 and 10^"1 mol L^"1, covering the linear and non-linear response range. The experiments were performed in solutions with or without the ionic strength being adjusted to 0.1 olL^-1 with K₂S0₄. The lower limit of linear response, the practical limit of detection, the slope and the reproducibility of the potential values of the electrodes were established according to the IUPAC convention. The data presented in table 3 corresponds to the average of six values obtained in two determinations with each of the three electrodes. Comparing the results obtained with those known from other nitrate detection techniques and devices, an improvement in the lower limit of linear range (at least one half of an order of magnitude) was obtained. The response time was determined by spiking a dilute solution (10^~5, 10^"4 and 10^"3 molL^"1) with a more concentrated one so as to obtain a 10 times more concentrated solution, and recording the time for a stable potential (+ 0.1 mV) . The records obtained showed a short response time, inferior to 12 seconds, and high stability of the potentials, which is important for direct potentiometric determinations. The electrodes showed an initial drift of potential, and the response stability appeared after the 3333 nitrate electrode had been conditioned for three days in 10^"1 molL^"1 solution of the primary ion. Theoretically, in the all-solid-state PVC membrane electrodes without internal reference solution there is no well-defined internal reference potential system. In practice, however, these electrodes provide reproducible potentials. The establishment of a constant electrode potential requires a period to stabilize the internal reference potential in the graphite conductive epoxy-PVC boundary, by means of the 0₂-H₂0 coupled as has been previously suggested.

A further advantage of certain embodiments of the present invention is that the concentration of nitrate, periodate or salicylate ions may be determined without the need for an inner reference solution in the measuring device. This reduces the cost of constructing the measuring device and allows it to be versatile, robust and smaller than known nitrate ion sensors.

A further advantage of certain embodiments is the low cost to manufacture the electrodes, due in part, to the small quantities of free base used for each electrode.

Figure 5 shows the calibration curve for 2 3333 free base electrodes in nitrate solution. Calibration curves plot the conductivity of the free base membrane electrodes at various concentrations of nitrate ions. A linear relationship demonstrates that the free base membrane electrodes accurately measures the nitrate concentration. A good linear relationship is shown in the Fig. 5 curves. Calibration curves using the 3333 free base electrodes for salicylate ions and periodate ions are shown in Figs lOa-lOd and lla-lld respectively. These results also show a good linear relationship and demonstrate that the 3333 free base electrode is also useful for determining the presence and/or concentration of such ions.

The method of obtaining calibration curves for the 3232 free base electrode - shown in Figs. 6a-6d, 7a- 7d and 8a-8d - was the same as that used for the 3333 electrode.

The calibration curves shown show that the ion selective electrode containing the 3232 free base also gave a linear relationship with good Nerstian gradients. The lower limit of detection is 7.99x10^" ^SM of nitrate (N0₃_) . These results were obtained in solutions with ionic strength adjusted with Na₂S0₄ (0.033M) .

Although the sensitivity is lower than the previous 3333 free base electrode, the response is very good.

It is apparent that a variety of macrocyclic and particularly heteromacrocyclic systems can be used in the preparation of membrane sensor for Ion Selective Electrodes to detect nitrate ions. They may also be used to detect periodate, salicylate or other ions. INFLUENCE OF pH To determine the influence of pH on the potential of the electrodes a Reilley diagram was performed in solutions with the ionic strength adjusted (1=0. IM). For that, the electrodes were immersed in a nickel nitrate solution or other nitrate salt of a given concentration and the pH was varied by addition of strong base (NaOH) or strong acid (H₂S0 ) . The corresponding diagram for a nitrate solution of 10^"3 M was represented in Fig 12.

The Reilley diagram indicated a pH operational working range between pH 2.3 and 12.0. However for the lower concentrations the pH operational range diminishes. Hence the adjustment of the pH samples before analysis will be probably necessary in some applications at very low nitrate concentration. Nevertheless the range of pH over which certain embodiments of the electrode may be used without adjusting the pH of the samples is ideal for biological and environmental samples. The potentiometric selectivity coefficients shown in table 3 were determined with three different concentrations of the primary ion together with the same concentration of an anion interferent (10^~4, 10^" ³ and 10^"2 mol L^"1) , following the separated solution method detailed above. Certain embodiments of the invention exhibit a better selectively coefficient towards chloride ions (the main interference of the nitrate selective electrodes) than other known nitrate sensor units. An advantage of embodiments of the present invention is that the concentration of the nitrate ions may be determined in the presence of other ions, e.g. in food where natural salt and therefore chloride ions are present .

The variation in the working characteristics with time was used as criterion for evaluating the electrode's lifetime. The electrodes when in regular use were stored in 10^~4 mol L^"1 potassium nitrate solution. All the 3333 electrodes analysed had a lifetime generally greater then 10 months, which is longer then the durability of known electrodes whose membranes are based on mobile carrier sensors but constructed using an internal reference solution. This is probably due to the presence of just a single contact surface between the membrane and solution, instead of two as in conventional electrodes; that is, the area through which leaching occurs is halved. Certain embodiments of the present invention benefit from the absence of an inner reference solution which provides a number of benefits including longer lifetime of the electrodes as described above.

Moreover in the conventional electrodes there is a net hydrodynamic pressure, favouring the loss of sensor to the external solution. Analytical Applications To evaluate the analytical uses of the constructed 3333 free base electrodes it was decided to devise a simple and quick procedure for determining nitrate concentration in different types of vegetables. Different vegetable samples were stirred in deionized water at room temperature after different extraction times followed by the nitrate analysis using the brucine method. From the plots obtained, it was concluded that thirty minutes was an adequate period of time for complete nitrate extraction. Since nitrate solution is a culture medium for bacteria and algae, boric acid could be added to prevent biological growth if sample extract needs to be stored.

Sample Preparation To accomplish nitrate determinations in real samples, an accurately weighed sample - 1 g of previously homogenized and fully dried (65°C) vegetable was extracted in 100 mL of deionized water during 30 min at room temperature. The extract was filtered over a 100 ml volumetric flask and was made up to volume with deionized water.

The brucine spectrophotometric method with slight modifications was adopted as the reference method since it has been shown that this method is adequate in the determination of nitrate in carbon black, tobacco, and meat products. As the extracted samples of vegetables are coloured, an error in the determination of the absorbance can arise. To circumvent this problem active carbon was added to the sample extracts before the analysis. The samples were filtered through an acid washed filter paper to avoid contamination by trace nitrate contained in filter papers. Taking into account the approximate ionic composition of the samples, the pH of the vegetable extracted samples and the corresponding Reilley diagrams obtained for the electrodes only, the addition of an ionic strength adjuster solution (ISA) was carried out. For the potentiometric determinations an aliquot of 50.0 ml of the extracted sample was pipetted directly into a 100 ml plastic beaker and 5.0 mL of ionic strength indicator (ISA) solution was added.

Twelve different types of vegetable samples were extracted and the resulting solutions were simultaneously analysed by the spectrophotometric method as reference and by the 3333 free base electrode using direct potentiometry. In each case, the precision was evaluated by its application to eleven samples of the same product and was expressed in terms of its standard deviation and coefficient of variation. Standard nitrate additions were used to evaluate the accuracy of the method and the average percentage of spike recovery, for three different additions, was calculated. The data obtained is shown in Table 4 and confirms that the 3333 free base ion selective electrode has good precision and accuracy. To test whether the 3333 free base electrode and the reference methods differ in their precision, a significance F test (two-tailed test) was carried out. The calculated F-values, using the electrode, were always less than the critical F-value for all the samples analysed. Hence there is no significant difference between the two standard deviations at the 95% confidence level. The limit of detection of the proposed method was obtained as recommended by the Analytical Methods Committee 48 and was found to be 0.05 g kg-1. Finally, thirty two samples of different types of vegetables were simultaneously analysed by the 3333 free base electrode and the reference methods. From the results obtained it can be concluded that the nitrate contents ranged, with a large distribution, from 0.5 gKg^"1 to 17.2 gKg^"1 with an average value of 5.3 gKg^"1. The percentages of difference obtained between the potentiometric and reference methods ranged from -5.3 to +5.6, with an average value of -0.4.

Table 4 Comparative study of the precisions and accuracies attained in the determination of nitrate in different types of vegetables by simultaneous application of the proposed potentiometric and reference methods.

Sample Reference method I.S.E. method

X^a C.V ^b R ^c X^a C.V R ^c RE^d F

Celery 15.17±0.17 1.1 101.7 15.02+0.13 0.9 101.6 -1.0 1.71

.ettuce 5.7810.09 1.5 101.7 5.6210.06 1.1 101.7 -2.8 2.25 spinach 8.07±0.12 1.5 101.6 7.93+0.10 1.3 101.6 -1.6 1.44

Cucumber 1.6910.04 2.4 101.3 1.65+0.03 1.8 101.3 -2.4 1.77

'arsley 2.1310.15 1.2 98.9 11.90+0.11 0.9 101.1 -1.9 1.86 ladish 13.1810.14 1.1 98.6 12.98+0.12 0.9 101.4 -1.5 1.36

Carrot 0.7410.002 0.3 101.8 0.7610.002 0.3 101.4 +2.4 0 r^ indive 0.6510.001 0.1 102.4 0.66+0.001 0.1 98.6 +1.5 0

Cabbage 2.1610.05 2.3 101.3 2.2210.03 1.3 100.0 +2.4 2.78

Broad bean 1.86+0.04 2.1 98.02 1.8910.02 1.1 100.3 +1.6 4.00

^>ea 6.09+0.12 2.0 99.23 6.3910.07 1.1 101.1 +4.6 2.94

'urtichoke 3.9710.15 3.8 99.75 4.0510.16 3.9 100.6 +1.9 0.88

^a Mean nitrate concentration and standard deviation (g/Kg ^-1). Coefficient of variation (%) ^c Mean percentage of spike recovery (%). ^d Relative error of the potentiometric method versus the reference method (%).

• Tmhi e c .ri't.i'cal i F τ-»- a _l iue, c _o_ns .i •d ie _ri *ng . a _ 9 n5rn%/ o -if" confi r.dJe—nc _e. l 1evel 1 and t i 10n d Jegrees o ~£f frϊ eedom, for a two- tailed test, is 3.72.

The linear regression analysis of the results is shown in Fig. 13. The confidence limits of the slope, intercept of the regression line and the t- value for the correlation at a 95% confidence level and n-2 degrees of freedom were also determined. From these the calculated slope and intercept do not significantly differ from the ideal values of 1 and 0, respectively. Thus there is no evidence of systematic difference between the 3333 free base electrode and the reference one. From the correlation coefficient, the t-value calculated was greater then the tabulated t-value (2.04), hence showing significant correlation exists between both methods .

The reproducibility, dynamic response and lifetime of the nitrate selective electrode based on the [3.3.3.3] -oxazane system as a sensor, were good. The limits of detection and the selectively coefficient towards interference ions, such as chloride, were better than known electrodes. The free base electrodes also showed good mechanical characteristics and do not require particular maintenance.

'Advantages of certain embodiments of the invention include that the use of the macrocyclic coated electrodes offers an easy and quick method with good precision and accuracy for the determination of nitrate in different types of vegetables avoiding the use of interference suppresser solutions that usually contain relatively toxic and expensive reagents. This advantage will also be evident in other matrices with even lower nitrate concentrations and similar chloride concentrations.

Modifications and improvements can be incorporated without departing from the scope of the invention.

Claims

1. Apparatus for the determination of at least one of the presence and concentration of ions in a sample, the apparatus having an electrode comprising a macrocyclic compound.

2. Apparatus as claimed in claim 1, wherein the ions being determined are nitrate ions.

3. Apparatus as claimed in claim 1, wherein the ions being determined are one of periodate ions and salicylate ions.

4. Apparatus as claimed in any preceding claim, wherein the macrocyclic compound is supported on a conductive substrate.

5. Apparatus as claimed in claim 4, wherein the conductive substrate is housed within a plastic housing and connected to means to measure conductivity.

6. Apparatus as claimed in any preceding claim, wherein the macrocyclic compound comprises a solid polymeric membrane.

7. Apparatus as claimed in any preceding claim, wherein the macrocyclic compound is a heteromacrocyclic compound.

8. Apparatus as claimed in claim 7, wherein the heteromacrocyclic compound is an oxa-aza cyclic alkane .

9. Apparatus as claimed in any preceding claim, wherein the macrocyclic compound has the general formula :-

wherein m and n are independently, 2, 3, 4, 5 or 6.

10. Apparatus as claimed in claim 9, wherein n is 3.

11. Apparatus as claimed in claim 9 or claim 10, wherein m is 3.

12. Apparatus as claimed in any one of claims 9 to 10, wherein m is 2.

13. Apparatus as claimed in any preceding claim, wherein macrocyclic compound has the formula.--

14. Apparatus as claimed in any one of claims 1-12, wherein the macrocyclic compound is of the formula:-

15. Apparatus as claimed in claim 2 or any one of claims 4-14 when dependent on claim 2, which can detect nitrate ions with a concentration of around 10^"7 M.

16. A method of preparing an apparatus as claimed in any preceding claim, wherein the macrocyclic compound is layered onto a conductive substrate.

17. A method as claimed in claim 16, wherein the macrocyclic compound is dissolved in a solvent and mixed with an immobilising agent, and, after the compound-solvent-immobiliser is layered onto the conductive substrate, the solvent is allowed to evaporate to leave a thin layer of immobilised macrocyclic compound on the surface of the conductive substrate.

18. A method as claimed in claim 17, wherein the macrocyclic compound is also dissolved in a mediator solvent with a lipophiliσ additive.

19. A method as claimed in claim 17 or 18, wherein the conductive substrate is an inert polymeric matrix.

20. A method as claimed in any one of claims 16-19, wherein the macrocyclic compound is applied dropwise onto the conductive support.

21. A method of determining at least one of the presence and the concentration of ions in a sample, the method comprising the steps of : - exposing the sample to a macrocyclic compound and determining potentiometric changes in the macrocyclic compound.

22. A method as claimed in claim 21, wherein an electrode comprising the macrocyclic compound is inserted into the sample.

23. A method as claimed in claim 21 or 22, wherein changes in the conductivity of the electrode comprising the macrocyclic compound are recorded and used to determine the concentration of the ions.

24. A method as claimed in any of claims 21 to 23, wherein the sample is a food sample.

25. A method as claimed in any one of claims 21 to 24, wherein the sample is pre-treated with an agent to change its pH.

26. A method as claimed in any one of claims 21 to 25, wherein the ions are nitrate ions.

27. A method as claimed in any one of claims 21 to 25, wherein the ions are at least one of salicylate and periodate ions.

28. A method as claimed in any one of claims 21-27, wherein the macrocyclic compound is an oxa-aza cyclic alkane.

29. A method as claimed in any one of claims 21 to 28, wherein the macrocyclic compound has the general formula :-

in which m and n are independently, 2, 3, 4, 5 or 6.

30. A method as claimed in claim 29, wherein n is 3 and m is 2 or 3.