rapid determination of nitrate and nitrite in drinking water samples using ion-interaction liquid...

Analytica Chimica Acta 441 (2001) 53–62

Rapid determination of nitrate and nitrite in drinking watersamples using ion-interaction liquid chromatography

Damian Connolly, Brett Paull∗National Centre for Sensor Research, School of Chemical Sciences, Dublin City University, Dublin 9, Ireland

Received 15 February 2001; received in revised form 10 April 2001; accepted 24 April 2001

Abstract

A simple ion-interaction chromatographic method for the fast determination of nitrate and nitrate at trace concentrationsin water samples is described. Using a short 3.0 cm × 0.46 cm 3 �m ODS analytical column together with a mobile phasecontaining 20 mM of the ion-interaction reagent tetrabutylammonium chloride (TBA-Cl), nitrite and nitrate could be readilyseparated from each other and other common UV absorbing anions in under 50 s. With the addition of a peristaltic pump andin-line filter, the developed method was configured for continuous monitoring of nitrate and nitrite concentrations in real tapwater samples. Up to 60 analyses/h could be carried out using the on-line system, which matches the analysis rate possiblewith traditional flow injection analysis (FIA) based methods. Standard analytical performance criteria were evaluated withnitrite being detectable at concentrations as low as 5 �g/l in actual tap water samples containing up to a 1000-fold excessof nitrate. Results obtained using the fast ion-interaction method compared well to those obtained using a conventional ionchromatographic method. © 2001 Elsevier Science B.V. All rights reserved.

Keywords: Rapid LC; Nitrate; Nitrite; Drinking water

1. Introduction

Nitrate and nitrite levels in our natural waters areimportant indicators of water quality. Nitrate and ni-trite are both intimately involved in the overall nitro-gen cycle of soil and higher plants and leaching ofnitrate from fertilisers added to soils can result in el-evated levels of nitrate in ground and surface waters.Nitrite can be formed during the biodegradation of ni-trate, ammonical nitrogen and other nitrogenous or-ganic matter, and is an important indicator of feacalpollution of natural water systems. In addition, nitrite

∗ Corresponding author. Tel.: +353-17005060;fax: +353-17005503; URL: http://www.ncsr.ie.E-mail address: [email protected] (B. Paull).

is readily oxidised to nitrate by dissolved oxygen, thus,decreasing oxygen levels in water.

When nitrate contaminated water supplies are usedas a source for drinking water adverse human healtheffects are also of great concern. Relative to nitrites,nitrates are compounds of lower toxicity, representinga danger only when ingested in excessive doses orwhen converted to nitrites. Nitrites however can haveseveral adverse effects upon human health. For exam-ple, the in vivo reaction between nitrite and secondaryor tertiary amines produces N-nitrosamines, whichare potential carcinogens, mutagens and/or terato-gens. In addition, nitrite interacts with blood pigmentto cause meta-haemoglobinemia, especially in infants(aka ‘blue baby syndrome’). This condition limits thebloods ability to carry oxygen from the lungs to therest of the body. To guard against the above effects

0003-2670/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.PII: S0003 -2670 (01 )01068 -6

54 D. Connolly, B. Paull / Analytica Chimica Acta 441 (2001) 53–62

the US EPA have set the maximum contaminant level(MCL) for nitrate in drinking water at 10 mg/l.

There are a number of analytical methodologiesapplicable to nitrite/nitrate quantification based uponeither spectrophotometric or chromatographic andelectrophoretic techniques [1]. Based upon the for-mer, numerous flow injection analysis (FIA) methodshave been developed, each exhibiting the benefitsFIA has to offer in terms of versatility, high samplethroughput, high degree of reproducibility and easeof automation [2–11]. The majority of FIA methodsbased upon spectrophotometric detection involve thewell-known reaction between nitrite under acid con-ditions and an aromatic primary amine [4–11]. Thisresulting diazotisation coupling reaction produces ahighly absorbing azo dye.

The main disadvantage of the above methods is thatto achieve simultaneous determination of both ana-lytes, the sample must be split, and a portion directedthrough a reductive column filled with copperised cad-mium [2–11]. This reduces nitrate to nitrite, such thatthe area of the first peak detected represents the con-centration of nitrite in the sample and the area of thesecond peak represents the concentration of nitrite plusnitrate. Several workers have reported problems withthe stability of the reduction column in such methods,with effects such as incomplete reduction and samplecarryover causing problems.

Analytical methods based upon separation methodsare predominantly those employing either ion chro-matography (IC) or capillary zone electrophoresis(CZE). A number of methods for nitrate and nitritedeterminations using CZE have been proposed forwater samples of varying complexity in recent years[12–16]. CZE offers a number of advantages, suchas high separation efficiency, short run-times and lowreagent costs. For example, Guan et al. recently devel-oped a capillary electrophoretic separation of nitrateand nitrite in tap and river water samples, demon-strating that a large concentration of nitrate does notadversely affect the quantification of a small amountof spiked nitrite. The minimum detection limit was1 �g/l for both nitrate and nitrite [12]. Later, Fukushiet al. demonstrated that CZE could be used to analysemore complex sample matrices, such as seawater fornitrate and nitrite by using an artificial seawater ascarrier solution [13]. This eliminated the interferenceof high concentrations of chloride ions in seawater.

Limits of detection were 0.04 and 0.07 mg/l for ni-trate and nitrite, respectively. However, in terms ofhigh-speed analysis, the run-time of just under 15 minwas relatively slow.

In stark contrast is the work of Melanson and Lucy[14], in which a shortened capillary (7 cm to detector),higher voltages and low pH values (to reduce the mo-bility of nitrite) were used to gain a separation of ni-trate and nitrite in under 12 s. Total run-time (pre-rinse,injection and separation) was just <1 min resulting ina sample throughput of up to 60 samples/h.

Chromatographic methods for nitrate and nitrite areusually based upon the use of IC [17–22]. Indeed theUS EPAs recommended method for the determinationof inorganic anions in water samples uses IC withsuppressed conductivity detection [23]. The above ionchromatographic method results in both improved re-producibility and sensitivity when compared to mostelectrophoretic methods of analysis, but does suffer interms of total analysis times, which if the sample alsocontains common anions, such as sulphate and phos-phate, can be of the order of 10–20 min. In additionto the above traditional IC methods, ion-interactionchromatographic methods have also been extensivelyinvestigated with regard to the separation of inor-ganic anions, including nitrite and nitrate. A num-ber of reviews have been published on the subjectof ion-interaction chromatography of small anions[24,25], including a recent comprehensive review byGennaro and Angelino [26], which details the advan-tages ion-interaction chromatography can introduce tothe determination of this particular group of analytes.

In a recent paper, we presented results of a studyinto the use of ion-interaction chromatography on ashort (3.0 cm×0.46 cm) ODS column for the fast sep-aration of UV absorbing anions [27]. Using an opti-mised eluent containing tetrabutylammonium chloride(TBA-Cl) as the ion-interaction reagent, the baselineseparation of eight anions, including nitrate and ni-trite, was possible in 4.5 min, with the separation offive hydrophilic inorganic anions, iodate, bromate, ni-trite, bromide and nitrate in 45 s. A brief applicationof the method to the determination of nitrate and ni-trite in both river water and drinking water sampleswas shown with total analysis times of <1 min.

In the following paper, the above system has beeninvestigated as to it’s potential for use in the rapidscreening of drinking water samples for nitrite and

D. Connolly, B. Paull / Analytica Chimica Acta 441 (2001) 53–62 55

nitrate contamination. To illustrate the potential ofrapid chromatographic methods for use in drinkingwater monitoring applications, an on-line chromato-graphic system was developed and used for the contin-uous analysis of a real drinking water sample stream.Samples were analysed at a rate of 1 injection/min,with the chromatographic system exhibiting both ef-ficiency and sensitivity comparable with standard ICmethods, and an analysis rate equal to that of the fastestFIA methods.

2. Experimental

2.1. Equipment

A Dionex DX500 ion chromatograph (DionexCorporation, Sunnyvale, CA, USA), comprising of aGP50 gradient pump, LC25 chromatography oven andan AD20 absorbance detector was used. Detectionwas by direct UV at 225 nm. The analytical columnused was a Phenomenex Hypersil, 3 �m particle size,30 mm ×4.6 mm i.d. column (Macclesfield, Cheshire,UK). The injection loop used was 50 �l. Data ac-quisition was at a rate of 10 Hz with processing ofchromatograms performed using a PeakNet 6.0 chro-matography workstation (Dionex). For the on-lineexperiments a peristaltic pump (Gilson Minipuls 312,Villiers, France) was employed to deliver the samplestream to the chromatograph injection port and anin-line 0.45 �m nylon membrane filter added (GelmanLaboratories Michigan, USA) (Fig. 2). The PeakNet6.0 software controlling the system was set-up to in-ject a sample every 60 s, beginning data acquisitionat time = 0 s and switching the injection loop backto ‘load’ position at time = 30 s. Comparative ionchromatographic experiments were carried out usingan AS17 anion exchange analytical column (Dionex)with UV detection as above.

2.2. Reagents

For preparation of the mobile phase, water usedwas obtained from a Millipore Milli-Q water purifi-cation system (Millipore, Bedford, MA, USA), tetra-butylammonium hydroxide (TBA-OH) was suppliedby Aldrich, (Aldrich, Milwaukee, WI, USA) as a 50%(w/v) solution in water, and methanol was obtained

from Labscan (Labscan Limited, Stillorgan, Dublin,Ireland). The optimised mobile phase for the rapidseparation of nitrate and nitrite consisted of 20 mMTBA-OH in 20% aqueous MeOH, titrated to pH 6.2using dilute HCl. The mobile phase was degassed andfiltered using 0.45 �m filters before use. The flow rateused was 2.0 ml/min. Column temperature was set atambient for all separations. The eluent used for theanalysis of validation samples was 15 mM NaOH, pre-pared from a 50% (w/w) NaOH solution (Aldrich).The flow rate used was 1.0 ml/min.

Stock standard solutions of concentration 1000 mg/lwere prepared weekly and working standards preparedfrom each respective stock solution on a daily basis asrequired. Nitrite and nitrate standards were preparedfrom their respective sodium salts (Aldrich).

3. Results and discussion

3.1. Ion-interaction chromatography

As mentioned previously, in an earlier study weshowed how fast and efficient separations of UV ab-sorbing anions could be achieved using ion-interactionliquid chromatography on short analytical columnspacked with a 3 �m ODS stationary phase [27]. Thedeveloped method used TBA-Cl as the ion interac-tion reagent added to a mobile phase consisting of upto 20% MeOH in water. Due to the rapid run-timesachievable on such a short column, the mobile phasecould be completely optimised to obtain the highestresolution of selected groups of UV absorbing anionsin only a few hours. The mobile phase was optimisedby varying the TBA-Cl (0.5–50.0 mM) and MeOH(0–20%) concentrations over 20 mobile phase prepa-rations and evaluating each resultant chromatogramaccording to a normalised resolution product. Un-der optimum conditions it was possible to separate8 UV absorbing anions, including thiosulphate andbenzoate, in 4 min, and five environmentally signif-icant anions, iodate, bromide, nitrite, bromate andnitrate, in a separation window of just 28 s, with atotal analysis time of under 50 s. For the latter groupof anions the optimised mobile phase consisted of20 mM TBA-Cl in 20% aqueous MeOH (pH 6.2).Using these conditions a brief application of the sys-tem to the determination of nitrate and nitrite in both


Fig. 1. Optimised separation of five inorganic UV absorbing anions, iodate, bromate, nitrite, bromide and nitrate. Mobile phase: 20 mMTBA-Cl, 20% MeOH, pH 6.2. Flow rate: 2.0 ml/min.

drinking water and river water samples was shown.In both water samples, nitrate and nitrite could bedetermined at low mg/l concentrations without anyinterference from other common matrix anions. Inaddition, the total analysis time was under 50 s, whichindicated that the method had considerable potentialfor the rapid screening of water samples for theseparticular analytes. The optimised fast ion-interactionseparation of the above five anions is shown as Fig. 1.

3.2. On-line determination of nitrate and nitrite indrinking water

The determination of nitrate and nitrite in both nat-ural waters and treated drinking waters is one of themost common analyses carried out on a daily ba-sis by water treatment companies and environmentalagencies, often on large numbers of individual sam-ples. The advantage of instrumental methods basedupon either FIA or IC is that either can (i) be usedwith an auto-sampler when dealing with large num-bers of samples or (ii) be readily configured to runon-line and monitor a flowing sample stream. In thisstudy, basic IC equipment, incorporating the above fastion-interaction chromatography, was set-up to con-tinuously sample and analyse the laboratory drinking

water supply, delivered to the IC system as a dynamicflowing sample stream. The system was configured asshown in Fig. 2. The drinking water sample was con-tinuously pumped using a basic peristaltic pump froma dynamic reservoir (being continuously filled fromthe laboratory tap) via an in-line filter, through the50 �l injection loop of an automated injection valve.

Fig. 2. Instrumental set-up for on-line flow analysis of tap water.A: tap; B: sink; C: peristalsic pump; D: 0.45 �m filter; E: injectionvalve; F: HPLC pump; G: analytical column; H: UV detector.


Fig. 3. Real-time trend analysis of nitrate (�) and nitrite (�) concentration in 60 tap water samples plotted against sample/analysis time.Validation samples (*, �) were taken at 10 min intervals. Chromatographic conditions as in Fig. 1.

The injection valve was connected in such a way as toallow the additional injection of standard solutions atany desired point (for system calibration) within thesampling period and the IC system was set-up to au-tomatically inject from the real sample stream everyminute.

To test the performance of the system it was leftto monitor the drinking water stream unattended for1 h, analysing a total of 60 discrete drinking watersamples. At two points during this trial period nitriteand nitrate were crudely added to the dynamic samplereservoir to test the response of the system to rapidchanges in the concentrations of the two analytes. Ni-trite was added at 20 min and nitrate added at 40 min.The sample stream was also sampled at 10 min in-tervals for comparative analyses (Section 3.3). Fig. 3illustrates the results of this trial, showing the con-centrations of nitrate and nitrite in the laboratory tapwater against sampling time. Each point representsan analyte concentration determined from each of the60 individual chromatograms. As would be expectedfrom a individual water supply, the concentrations ofthe two analytes did not vary significantly over the 1 h

period. There was no detectable nitrite, and the nitrateconcentration was steady at between 7.5 and 8.0 mg/l.However, the two points at which the sample streamwas spiked with nitrite and nitrate can be clearly seen.The system responds rapidly and equally rapidly re-turns to the original concentrations, illustrating howsuch a system would identify short-term concentrationchanges that could otherwise be missed.

Chromatograms of the tap water sample and thesame sample spiked with nitrite, obtained during theabove on-line trial can be seen in Fig. 4. The threeoverlaid chromatograms show the concentrations ofnitrite and nitrate present in the three samples takenat time 21, 22 and 23 min. As can be seen from thechromatograms shown, the tap water samples resultedin no interfering peaks from other matrix anions, suchas chloride and sulphate, and even after 60 consec-utive injections there was no sign of any interferingpeaks that may have been retained for longer. The twopeaks are also well resolved from both each other anda small system peak that elutes at 0.4 min, with thesignal clearly returning to baseline well before 1 minis up.


Fig. 4. Overlaid chromatograms illustrating a nitrite spike increasing with time as it travels through the system. Samples analysed at 22,23 and 25 min. Chromatographic conditions as in Fig. 1.

Fig. 5. System precision calculated on the nitrate peak in a single tap water sample injected 30 times consecutively: upper part: areaprecision; lower part: retention time precision. Mean nitrate concentration: 3.9 mg/l. Chromatographic conditions as in Fig. 1.


Table 1Comparison of sample throughput of recently published analyticaltechniques for nitrate and nitrite determinations

Technique Sample throughput References(samples/h)

FIA 20 [3]30 [2]50 [28]

Micro-FIA (nitrate only) 20–30 [32]

CZE 10 [12]15 [29,30]60 [14]

IC 10 [31]12 [23]

Fast-IIC 60 This work

To determine system precision, a single homoge-nous tap water sample containing a slightly lower con-centration of nitrate (3.9 mg/l) was taken and put inplace of the flowing sample reservoir. This single realsample was then repeatedly injected 30 times. The

Fig. 6. Upper part: overlay of a 20, 40, 100 and 200 �g/l nitrite spike in a water sample containing 5 mg/l nitrate. Chromatographicconditions as in Fig. 1. Lower part: same spiked samples run on an AS17 anion-exchange column.

results obtained are shown as Fig. 5. As can be seenthe system precision was 0.53% for peak area (shownas nitrate concentration) and 0.42% for peak retentiontime. Therefore, if for on-line monitoring purposes aquantifiable change in concentration was defined as achange in concentration equal to 10 times the methodR.S.D.%, at the above sample concentration (∼4 mg/lnitrate), changes as small as ±0.2 mg/l could be quan-tified. The highly reproducible peak retention time isalso essential for monitoring purposes as it allows thesoftware to accurately identify and integrate the peakautomatically, thus permitting the use of ‘real time’trending software to be used in monitoring applica-tions.

In terms of sample rate per hour the fast ion-interaction method detailed here compares veryfavourably with the current range of analytical meth-ods available for simultaneous nitrite and nitrate de-terminations. Table 1 shows some of the most recentmethods published for nitrate and nitrite determi-nations and their maximum sample throughput perhour. As can be seen from the table, the developed


fast ion-interaction method can analyse more samplesper hour than the FIA based methods and matchesthe rate of the ultra-rapid CE method developed byMelanson and lucy [14]. However, the method de-scribed here has the added advantage in that it can beused to monitor flowing sample streams.

3.3. Analytical performance characteristics

Detector linearity was investigated in two ways.Firstly, standard calibration curves were constructedover the range 0.5–25 mg/l for both nitrate and nitrite(n = 5), with R2 > 0.999 in both cases. Secondly,as is likely to be the case in most water samples, thelinearity of trace nitrite in the presence of excess ni-trate was determined. A collected water sample con-taining 5 mg/l nitrate was spiked with 20, 40, 100 and200 �g/l nitrite. The linearity for nitrite spikes in thereal sample was again R2 > 0.999. Fig. 6 shows theoverlaid chromatograms obtained from the injectionsof the above spiked samples. Also shown are chro-

Fig. 7. Upper part: overlay of a 25 �g/l nitrite/nitrate standard and blank. Chromatographic conditions as in Fig. 1. Lower part: overlay ofa 25 �g/l nitrite/nitrate standard run on an AS17 anion-exchange column and blank. Mobile phase: 15 mM NaOH. Flow rate: 1.0 ml/min.Loop size: 50 �l. Detection wavelength: 225 nm.

matograms obtained for the above spiked samples us-ing the conventional size AS17 anion exchange col-umn with a hydroxide eluent. As can be seen fromthe chromatograms shown, even at such high ratiosnitrite is still well resolved from the nitrate peak andthe separation is complete in less than one-fifth timeof the conventional column. In terms of efficiency, theshort column method also compares favourably withthe conventional column, with an average of 63,100plates/m compared to 53,540, respectively.

Detection limits were also investigated and com-pared to those obtained with the conventional anionexchange method. Calculated using the peak heightequivalent to three times the baseline noise, abso-lute detection limits of 0.15 and 0.35 ng (based upon50 �l injection volume) were obtained for nitrite andnitrate respectively. Fig. 7 shows the two sets ofchromatograms obtained for standard solutions of ni-trite (25 �g/l) and nitrate (25 �g/l) at concentrationsclose to the method detection limits. The two setsof chromatograms show how the sensitivities of the


two methods were very similar, with the peak heightsfor 25 �g/l nitrite equal to 1.2 × 10−3 AU using thefast ion-interaction method and 1.3 × 10−3 AU us-ing the conventional anion exchange method. A briefstudy on the affect sample injection volume had uponchromatographic performance was carried out us-ing injection loops of 25, 50 and 75 �l. Surprisinglyneither resolution nor efficiency were significantlyaffected over the above range, indicating the largerloop could be used to further decrease detectionlimits. The large surface area of the 3 �m ODS sta-tionary phase provides a higher column capacity thanwould be expected of a short column containing 5 or10 �m particles and this allowed the injection of suchrelatively large sample volumes.

Possible interferences were also investigated. Com-mon anions that gave no detector response and alsohad no effect upon the chromatography (at concentra-tions expected in drinking waters) included chloride,sulphate, phosphate, formate and acetate. Less com-mon UV absorbing anions which were retained anddid give a detector response, but baseline resolvedfrom the nitrite and nitrate peaks included, iodate(0.40 min), bromate (0.58 min), bromide (0.70 min),iodide (∼1.7 min), thiosulphate (∼2.6 min), thio-cyanate (∼3.8 min), and benzoate (∼4.3 min). Of theabove only bromide may be expected to be present indrinking water but at concentrations well below thatof nitrate.

Finally, to determine accuracy the tap water sam-ples taken during the on-line trail were analysed bythe conventional anion exchange method for nitriteand nitrate concentration. As seen with the on-lineion-interaction method no nitrite was present in anyof the six samples taken. However, the concentra-tion of nitrate found matched that determined usingion-interaction extremely well. The concentrations de-termined for nitrate in the six tap water samples usingthe two methods were found to be within 1 and 5%of each other, with the concentrations found using theconventional anion exchange method plotted as indi-vidual points within Fig. 3.

4. Conclusions

A fast ion-interaction liquid chromatographicmethod for the determination of nitrite and nitrate

has been developed. Included in an on-line chro-matographic system, or indeed combined with anauto-sampler, the method can be used for the rapidscreening of drinking water samples for nitrite and ni-trate contamination. The method is both sensitive andselective and equals the currently used FIA methodsin terms of analysis rate.

Acknowledgements

The authors would like to thank Swords Laborato-ries Ltd., for their financial contribution towards thisproject.

References

[1] G. Ellis, I. Adatia, M. Yazdanpanah, S.K. Makela, Clin.Biochem. 31 (1998) 195.

[2] Z. Zhi-Qi, G. Lou-Jun, Z. Han-Ying, L. Qian-Guang, Anal.Chim. Acta 370 (1998) 59.

[3] A.A. Ensafi, A. Kazenzadeh, Anal. Chim. Acta 382 (1999)15.

[4] A. Cerda, M.T. Oms, R. Forteza, V. Cerda, Anal. Chim. Acta371 (1998) 63.

[5] D. Gabriel, J. Baeza, F. Valero, J. Lafuente, Anal. Chim. Acta359 (1998) 173.

[6] K. Horita, G. Wang, M. Satake, Anal. Chim. Acta 350 (1997)295.

[7] R.T. Masserini Jr., K.A. Fanning, Marine Chem. 68 (2000)323.

[8] R. Segarra Guerrero, C. Gomez, J. Martinez Calatayud,Talanta 43 (1996) 239.

[9] A. Kojlo, E. Gorogkiewicz, Anal. Chim. Acta 302 (1995) 2–3.[10] M.J. Ahmed, C.D. Stalikas, S.M. Tzouwaru-Karayanni, M.I.

Karayannis, Talanta 43 (1996) 1009.[11] O. Pinho, I.M.P.L.V.O. Ferreira, M. Beatriz, P. Oliveira, M.A.

Ferreira, Food Chem. 62 (1998) 359.[12] F. Guan, H. Wu, Y. Luo, J. Chromatogr. A 719 (1996) 427.[13] K. Fukushi, K. Tada, S. Takeda, S. Wakida, M. Yamane, K.

Higashi, K. Hiiro, J. Chromatogr. A 838 (1999) 303.[14] J.E. Melanson, C.A. Lucy, J. Chromatogr. A 884 (2000) 311.[15] P.A. Marshall, V.C. Trenerry, Food Chem. 57 (1996) 339.[16] M. Jimidar, C. Hartman, N. Cousement, D.L. Massart, J.

Chromatogr. A 706 (1995) 479.[17] H. Sakai, T. Fujiwara, T. Kumamaru, Anal. Chim. Acta 331

(1996) 239.[18] S.A. Everett, M.F. Dennis, G.M. Tozer, V.E. Prise, P.

Wardman, M.R.L. Stratford, J. Chromatogr. A 706 (1995)437.

[19] M.I.H. Helaleh, T. Korenaga, J. Chromatogr. B 744 (2000)433.

[20] M. Novic, B. Divjat, B. Pihlar, J. Chromatogr. A 827 (1998)83.


[21] M.N. Muscara, G. de Nucci, J. Chromatogr. B 686 (1996)157.

[22] I. El Menyawi, S. Looareesuwan, S. Knapp, F. Thalhammer,B. Stoiser, H. Burgman, J. Chromatogr. B 706 (1998) 347.

[23] EPA Method 300.0, Determination of inorganic ions byion chromatography, United States Environmental protectionAgency, Cincinnati, OH, 1993.

[24] P.R. Haddad, A.L. Heckenberg, J. Chromatogr. 300 (1984)357.

[25] M.L. Marina, J.C. Diez-Masa, M.V. Dabrio, J. Liq.Chromatogr. 12 (1989) 1973.

[26] M.C. Gennaro, S. Angelino, J. Chromatogr. A 789 (1997) 181.

[27] D. Connolly, B. Paull, J. Chromatogr. A 917 (2001)353.

[28] J.F. van Staden, M. Makhafola, South Africa J. Chem. 52(1999) 49.

[29] A.A. Okemgbo, H.H. Hill Jr., S.G. Metcalf, M.A. Bachelor,J. Chromatogr. A 844 (1999) 387.

[30] A.A. Okemgbo, H.H. Hill Jr., W.F. Siems, S.G. Metcalf,Anal. Chem. 71 (1999) 2725.

[31] I.N. Papadoyannis, V.F. Samanidou, C.C. Nitsos, J. Liq.Chromatogr. 22 (1999) 2023.

[32] P.H. Petsul, G.M. Greenway, S.J. Haswell, Anal. Chim. Acta428 (2001) 155.

rapid determination of nitrate and nitrite in drinking water samples using ion-interaction liquid...

Documents