Flow-injection analysis of nitrate by reduction to nitrite and gas-phase molecular absorption spectrometry

Transcription

1 Fresenius J Anal Chem (2001) 371 : DOI /s ORIGINAL PAPER Behzad Haghighi Abdollah Tavassoli Flow-injection analysis of nitrate by reduction to nitrite and gas-phase molecular absorption spectrometry Received: 22 May 2001 / Revised: 20 July 2001 / Accepted: 28 July 2001 / Published online: 9 November 2001 pringer-verlag 2001 Abstract Two flow-injection manifolds have been investigated for the determination of nitrate. These manifolds are based on the reduction of nitrate to nitrite and determination of nitrite by gas-phase molecular absorption spectrophotometry. Nitrate sample solution (300 µl) which is injected to the flow line, is reduced to nitrite by reaction with hydrazine or passage through the on-line copperized cadmium (Cd Cu) reduction column. The nitrite produced reacts with a stream of hydrochloric acid and the evolved gases are purged into the stream of O 2 carrier gas. The gaseous phase is separated from the liquid phase using a gas liquid separator and then swept into a flow-through cell which has been positioned in the cell compartment of an UV visible spectrophotometer. The absorbance of the gaseous phase is measured at nm. A linear relationship was obtained between the intensity of absorption signals and concentration of nitrate when Cd Cu reduction method was used, but a logarithmic relationship was obtained when the hydrazine reduction method was used. By use of the Cd Cu reduction method, up to 330 µg of nitrate was determined. The limit of detection was 2.97 µg nitrate and the relative standard deviations for the determination of 12.0, 30.0 and 150 µg nitrate were 3.32, 3.87 and 3.6%, respectively. Maximum sampling rate was approximately 30 samples per hour. The Cd Cu reduction method was applied to the determination of nitrate and the simultaneous determination of nitrate and nitrite in meat products, vegetables, urine, and a water sample. Introduction Nitrogen oxides (NO x ), including nitric oxide (NO), nitrogen dioxide (NO 2 ), nitrite (NO 2 ), and nitrate (NO 3 ) are B. Haghighi ( ) A. Tavassoli Department of Chemistry, Institute for Advanced tudies in Basic ciences, Zanjan, Iran, haghighi@iasbs.ac.ir regarded as suspected carcinogens, environmental pollutants, and precursors of acid rain. Nitrate is naturally present in vegetables and its concentration varies enormously, ranging from 1 to 10,000 mg kg 1 fresh weight. The nitrate content of vegetables can be influenced by factors related to the plant (variety, species, and maturity) and the environment (temperature, light intensity, and deficiency of some nutrients and use of fertilizers). Furthermore, nitrate and nitrite have been added intentionally during the curing of certain meat products, because they inhibit the growth of spores of Clostridium botulinum and impart characteristic color and flavor to this kind of foodstuff. The significance of nitrate to human health is that nitrate, after being metabolized or reduced to nitrite, can react with secondary or tertiary amines to form N-nitroso compounds, which are potent carcinogens. Nitrite can also interact with hemoglobin influencing the oxygen transport mechanism, giving rise to methemoglobinemia [1]. A variety of analytical methods has been developed for the determination of nitrate and nitrite and applied to the analysis of food, water, biological material, soil, plants, animal feed, and other matrices. These include kinetic analysis [2], FIA [3], HPLC [4], gas chromatography [5], capillary electrophoresis [6], amperometry [7], potentiometry [8], voltammetry [9], and chemiluminescence [10]. However, the most frequently employed methodology for the determination of nitrite, and nitrate after reduction to nitrite, is still spectrophotometry [11, 12, 13]. Among the spectrophotometric methods adopted as an AOAC official method for the determination of nitrite and nitrate is the reduction of nitrate to nitrite and reaction of nitrite with N-(1-naphthyl)ethylenediamine 2HCl and sulfanilamide [14]. Although the sensitivity of the diazotization coupling reaction is excellent, the method suffers from numerous limitations. olution ph during diazotization, amount of nitrite, reaction time for diazotization, need to destroy excess nitrite, and adjustment of ph for the coupling reaction are important factors that determine the usefulness of individual azo-dye formation methods [15, 16].

2 1114 A rapid and simple UV-spectrophotometric method [17] has been adapted for the determination of nitrate in aqueous solutions. This technique is used only for screening samples with low organic matter content, i.e. uncontaminated natural waters and potable water supplies. This paper describes the development and application of an FIA GPMA method for the determination of nitrate in different matrices, without need of previous extraction and clean-up procedures, as recommended by the reference spectrophotometric method. In the proposed method, nitrate sample solution was reduced to the nitrite in the flow line by reaction with hydrazine or passage through an on-line copperized cadmium (Cd Cu) reduction column. The nitrite produced was reacted with a stream of hydrochloric acid and the evolved gases were purged into the stream of O 2 carrier gas. The gaseous phase was separated from the liquid stream in a homemade gas liquid separator [18] and then swept by the carrier O 2 gas into a home-made flow-through cell [18] positioned in the space of the cell compartment of a UV visible spectrophotometer. The transient absorbance of the gaseous phase was measured at nm. The methodology presented was applied to the determination of nitrate and the simultaneous determination of nitrate and nitrite in meat products, vegetables, urine, and water samples. Experimental Apparatus chematic diagrams of the flow systems used for determination of nitrate are shown in Fig. 1. Instruments used were a Watson Marlow peristaltic pump (Model 505Du), and a single-beam Pharmacia UV visible spectrophotometer (Ultrospec, model 4000) with a home-made flow through cell. The sample solution was injected via a Rheodyne type 50 six-way Teflon rotary valve into a carrier stream. The flow lines were made from upelco Teflon tubing (0.5 mm i.d.). The gaseous phase was separated from the liquid phase by use of a home-made gas liquid separator. The copperized cadmium (Cd Cu) reduction column was made from a glass column (3 mm i.d. 8 cm) filled with the copperized cadmium (particle size: mm). The transient absorption data were collected at 1-s intervals by use of wift II software. Chemicals All reagents were of analytical reagent grade from Fluka and were used without further purification. Working solutions were freshly prepared by accurate dilution of stock solutions. Deionized, triply distilled water was used for preparation and dilution of all solutions. Nitrate standard solution A stock solution (5000 µg ml 1 ) was prepared by dissolving g sodium nitrate in water to give a 100 ml solution. Acid solution Hydrochloric acid solutions of different concentration were prepared by appropriate dilution of concentrated HCl (36%, 1.18 g ml 1 ). Fig. 1 chematic FIA diagrams for the determination of nitrate using (a) the hydrazine reduction method and (b) the Cd Cu reduction method; and (c) for the simultaneous determination of nitrate and nitrite. P, pump; I, injection valve (300 µl); T 1, T 2, T 3, threeway connectors; D, spectrophotometric detector with a homemade flow cell (204.7 nm); GL, gas liquid separator; W, waste. Conditions: (a) R 1, CuO 4 (0.04%); R 2, NaOH (0.07 mol L 1 ) and hydrazine (0.3%); R 3, HCl (2 mol L 1 ); C 1, reduction coil (0.5 mm i.d. 150 cm); C 2, reaction coil (0.5 mm i.d. 4 cm); C 3, purging coil (0.5 mm i.d. 30 cm); O 2, carrier gas (8.1 ml min 1 ). (b) R 1, NH 4 Cl (1 mol L 1 ); R 2, HCl (2 mol L 1 ); RC, Cd Cu reduction column (3 mm i.d. 7 cm); C 1, reaction coil (0.5 mm i.d. 30 cm); C 2, purging coil (0.5 mm i.d. 25 cm); O 2, carrier gas (9.2 ml min 1 ); A, disconnection point for column regeneration. (c) All conditions as for (b); V, selection valve; B, Teflon tubing (0.5 mm i.d. 5 cm) Copper sulfate solution A stock solution of copper sulfate (1%) was prepared by dissolving 1.00 g CuO 4 5H 2 O to 100 ml with water. Hydrazine sulfate solution A stock solution of hydrazine sulfate (2%) was prepared by dissolving 5 g N 2 H 4 H 2 O 4 to 250 ml with water. odium hydroxide solution A stock solution of sodium hydroxide (1 mol L 1 ) was prepared by dissolving g NaOH to 250 ml with water. Ammonium chloride solution A stock solution of ammonium chloride (2 mol L 1 ) was prepared by dissolving g NH 4 Cl to 250 ml with water. Procedures Determination of nitrate using the hydrazine reduction method Nitrate sample solution was injected at I into the carrier stream containing copper sulfate (Fig. 1a). The nitrate zone was mixed

3 1115 with the carrier stream containing hydrazine sulfate and sodium hydroxide at T 1 and the reduction reaction occurred subsequently in C 1. The nitrite produced was combined with the stream of hydrochloric acid at T 2 and reaction was completed during flow in C 2. Oxygen gas was introduced into the flow line at T 3. The evolved gaseous products diffused into the O 2 segments, during flow in C 3. The gaseous phase was separated from the liquid phase by use of a gas liquid separator and was then swept into the flowthrough cell, which had been positioned in the optical path of the UV visible spectrophotometer. The molecular absorbance of the gaseous phase was recorded at nm and at 1-s intervals. Preparation of the Cd Cu reduction column The Cd reduction column was prepared and conditioned by an online procedure. For this purpose, the glass column (3 mm i.d. 8 cm) was filled carefully with new or used Cd granules (particle size mm). The filled Cd column was positioned in the proposed manifold (Fig. 1b) and the flow line was disconnected at A. The waste manifold in on-line reductor column generation was A. The Cd column was washed (1 min) and rinsed (2 min) with HCl solution (5 mol L 1 ) and distilled water as carrier solutions, respectively. The carrier solution was replaced by a CuO 4 solution (2%) which flowed into the column for 4 min or until a brown colloidal precipitate developed. Finally, the Cd Cu column was washed with distilled water (as the carrier for 3 min) to remove all colloidal precipitated Cu [19]. The flow line was reconnected again at A for nitrate analysis. The flow rate of carrier solution in all steps was 2 ml min 1. The Cd Cu reduction column was regenerated every day by the procedure mentioned. Determination of nitrate using the Cd Cu reduction method Nitrate sample solution was injected at I into the carrier stream containing ammonium chloride, and flowed into the Cd Cu reduction column, where nitrate was reduced to nitrite (Fig. 1b). The nitrite zone was mixed with the stream of hydrochloric acid at T 1 and reaction between them was completed during flow in C 1. Oxygen gas was introduced into the flow line at T 2. The evolved gaseous products diffused into the O 2 segments during flow in C 2. The gaseous phase was separated from the liquid phase by use of a gas liquid separator and then was swept into the flow-through cell, which had been positioned in the optical path of the UV visible spectrophotometer. The molecular absorbance of the gaseous phase was recorded at nm and at 1-s intervals. imultaneous determination of nitrate and nitrite using the Cd Cu reduction method Figure 1c shows the manifold used for the simultaneous determination of nitrate and nitrite. ample solution containing nitrate and nitrite was injected at I into the carrier stream containing ammonium chloride. The switch in the a position resulted in bypass of the Cd Cu column by the analyte zone and direct treatment with the hydrochloric acid. In this mode of operation only nitrite reacted with the acid and gave an absorption signal. The switch in the b position passed the analyte zone through the Cd Cu column and the absorption signal arising from the sum of both nitrite and nitrate was measured. Nitrate was determined from the difference between intensities of the absorption signals. Preparation of real samples for analysis The literature reports many procedures for the treatment of samples before the determination of nitrite and nitrate. For water samples an appropriate volume is directly examined for determination of nitrate and nitrite. Food samples can be extracted with water by heating with water [2], by heating with saturated sodium acetate in a boiling water bath for 30 min [20], with water and potassium aluminium sulfate to precipitate organic matter [21], and with water containing small amounts of sodium hydroxide [22], etc. For biological samples, it is necessary to perform pretreatment to ensure deproteinization before FIA measurement, because proteins and the other coexisting biological substances interfere with reduction of nitrate; for urine samples the dilution is sufficient and deproteinization is unnecessary [23]. In this work, water samples were examined directly using the procedure proposed, without pretreatment. The urine sample was diluted twofold before FIA measurement. Meat products and vegetables (25 g) were extracted with water (50 ml) containing small amounts of sodium hydroxide (four drops of 2 mol L 1 NaOH) [2] and the extracted solutions were filtered and diluted to 100 ml with distilled water, before FIA measurements. Table 1 elected chemical and FIA conditions for the hydrazine reduction method, obtained from the optimization experiments Condition Range Value Experimental observation studied selected on signal intensity Reagent flow rate (ml min 1 ) R Reached the plateau R R Gas flow rate (ml min 1 ) Passed over maximum Length of reaction coil (cm) C Reached the plateau C Decreased C Passed over maximum ize of sample loop (µl) Reached the plateau Temperature ( C) C Reached the plateau C 2, C Passed over maximum Concentration of reagents R 1 =CuO 4 (%) Reached the plateau R 2 =NaOH (mol L 1 ) Passed over maximum R 2 =Hydrazine (%) Passed over maximum R 3 =HCl (mol L 1 ) Reached the plateau

4 1116 Table 2 elected chemical and FIA conditions for the Cd Cu reduction method, obtained from the optimization experiments Condition Range Value Experimental observation studied selected on signal intensity Length of Cd Cu column (cm) Reached the plateau Reagent flow rate (ml min 1 ) R Passed over maximum R Gas flow rate (ml min 1 ) Passed over maximum Length of reaction coil (cm) C Passed over maximum C Passed over maximum ize of sample loop (µl) Increased Temperature ( C) C 1, C Increased (linearly) Concentration of reagents (mol L 1 ) R 1 =NH 4 Cl Increased (very slightly) R 2 =HCl Reached the plateau Results and discussion Optimization of the flow-injection manifolds The absorption spectra of the gaseous products were obtained by continuous propulsion of the carrier stream containing constant amounts of nitrate (150 µg) through the proposed Cd Cu and hydrazine reduction manifolds. The absorption spectra of the gaseous products in both reduction methods were similar to those reported elsewhere [24]. A wavelength of nm was employed, and to optimize the proposed flow-injection manifolds, the effects of hydrodynamic and chemical conditions on peak height, peak shape, and the reproducibility of the results were studied. The univariate method was employed for the optimization of the systems. Tables 1 and 2 show the results from optimization of the working conditions for nitrate determination. The optimum length of the Cd Cu reduction column was established by use of a solution containing 150 µg nitrate. Reduction column lengths 2, 3, 4, 5, 6, 7, and 8 cm were tested. For each situation, the percentage reduction was determined by comparison of the peak height achieved with that corresponding to a nitrite solution (150 µg) processed in the same way. A column length of 7 cm was chosen, because for this length the reduction efficiency was the best, approximately 90%, and reproducibility and analysis time were good. The effect of the type of acid (sulfuric and hydrochloric) was studied; hydrochloric acid was found to be best acid (in terms of peak intensity) for both systems. The results of both manifolds showed that the absorption signal intensities were dependent on temperature (Fig. 2). In the hydrazine reduction method the peaks height increased slightly with increasing temperature and passed through a maximum at approximately 40 C; in the Cd Cu reduction method peaks height increased linearly with increasing temperature. The effect of temperature on peak height Fig. 2 Effect of temperature on absorption signal intensity using (a) the hydrazine reduction manifold and (b) the Cd Cu reduction manifold. Conditions as for Fig. 1 in the Cd Cu reduction manifold was exactly similar to that reported elsewhere for nitrite determination [24]. Analytical performance of the proposed methods Calibration graphs constructed using proposed methods with injection of a series of standard nitrate solutions under the conditions given in Tables 1 and 2. The limit of detection was determined on the basis of three times the standard deviation of the measurements of blank signals [25] and reproducibility of the proposed methods was determined by repetitive injection of standard nitrate solutions. The analytical features such as calibration equation, correlation coefficients, linear dynamic range, limits of detection, and sample throughput have been summarized in Table 3. The results obtained indicate that the relation-

5 1117 Table 3 Analytical features of the proposed methods for nitrate determination Feature Cd Cu reduc- Hydrazine reduction method tion method Relationship between Y and x Y=a+b x Y=a+bln(x) a b r Reproducibility (%RD, n=6) 3.3 (x=12.0 µg) 1.7 (x=2.4 µg) 3.8 (x=30.0 µg) 2.0 (x=6.0 µg) 3.6 (x=150.0 µg) 1.9 (x=12.0 µg) 2.6 (x=24.0 µg) Dynamic range (µg) Dynamic range (µg ml 1 ) Detection limit (µg) Detection limit (µg ml 1 ) ample throughput 30 (maximum) 45 (maximum) (samples h 1 ) x=amount of nitrate (µg), Y=peak height (arbitrary unit), r=correlation coefficient ship between peak height and concentration of nitrate is logarithmic for the hydrazine reduction method but linear for the Cd Cu reduction method. The Cd Cu reduction manifold method has a wider dynamic range than the hydrazine reduction method. As was expected, the reproducibility of the hydrazine reduction method was better than that of Cd Cu reduction. Because of the good linear behavior of the Cd Cu reduction method and its good dynamic range, the Cd Cu reduction method was chosen for interference studies and analytical application. Interferences and removal of interfering ions The effect of several ions on the determination of nitrate using the proposed Cd Cu reduction method was studied for solutions containing 60 µg nitrate and different amounts of interfering ions. Maximum tolerable amounts of interfering ions, their effects, and their removal have been listed in Tables 4 and 5. Procedures such as acidification and heating eliminated interference from 2, O 3 2, NO 2 and CO 3 2, precipitation with silver sulfate or Ag + and OH [24] removed I and CN. Passage of the sample through a cation exchanger [2] (e.g. Amberlite IR-120) can eliminate interference from cations. All removal procedures should be performed before nitrate determination. Analytical applications The Cd Cu reduction procedure under conditions given in Table 2 was successfully applied to the determination of nitrate and the simultaneous determination of nitrate and Table 4 Effect of interfering ions on the determination of 60 µg nitrate Interfering ion Amount (µg) Nitrate+ Interferent Nitrate, absorption signal intensity; b, column destroyed Table 5 Effect of interfering ions, and their removal. Test solutions contained 60 µg NO 3 and 3000 µg interfering ion a see text for more details;, absorption signal intensity; b, column destroyed; n, not examined F, Cl, Br, CH 3 COO, O ±0.02 CN O NO 2, I 60 2 CO K +, Na +, Mn 2+, Ca 2+, Ba 2+, Cu 2+, Cd ±0.02 Mg Ni 2+, Zn Co Fe b Interfering ion Nitrate+ Interferent Nitrate Masking procedure a O Acidification and heating Acidification and heating 0.95 NO 2 50 Acidification and heating 0.98 I 3.9 Ag + and OH 0.99 CN 2.5 Ag + and OH 0.98 CO Acidification and heating 0.99 Mg Cation exchanger n Zn Cation exchanger n Co Cation exchanger n Ni Cation exchanger n Fe 3+ b Cation exchanger n Nitrate+ Interferent + Masking Nitrate

6 1118 Table 6 Determination of nitrate and nitrite in meat products, vegetables, urine, and water ample Amount of nitrate Amount of nitrite Determined Added Found Recovery Determined Added Found Recovery (µg) (µg) (%) (µg) (µg) (%) Meat sausage mg kg mg kg Mortadella mg kg mg kg Lettuce mg kg 1 None Cucumber mg kg 1 None Well water µg ml 1 None Urine µg ml µg ml nitrite in a number of meat products, vegetables, a water sample, and urine. To check the validity of the proposed procedure for nitrate and nitrite in these samples, the recoveries of added nitrate and nitrite, were determined. All the samples were prepared according the procedures proposed for the preparation of real samples, and each recovery was calculated by comparing the results obtained before and after adding the nitrate and nitrite standard solutions. The amounts of nitrate and nitrite in the samples, and their recoveries, are shown in Table 6. Conclusions The methods presented in this paper are safe, simple, and of low cost and also have several advantages, e.g. low reagent and sample consumption and good selectivity. The procedure based on the hydrazine reduction method resulted in a logarithmic relationship between concentration of nitrate and analytical signal, with good reproducibility. The procedure based on the Cd Cu reduction method resulted in a linear relationship between concentration of nitrate and analytical signal with a wide linear dynamic range. The Cd Cu reduction method was applied to the determination of nitrate and the simultaneous determination of nitrate and nitrite in meat products, vegetables, urine, and a water sample. The results reported show that the method is suitable for the routine determination of nitrate and nitrite in real samples. Acknowledgment The authors wish to express their gratitude to Institute for Advanced tudies in Basic ciences (IAB) for the support of this work. References 1. MAFF (Ministry of Agriculture, Fisheries and Food) (1992) Nitrate and nitrite and N-nitroso compounds in food: second report, Food surveillance paper, No 32. HMO, UK 2. Zatar NA, Abu-Eid MA, Eid AF (1999) Talanta 50: Ahmed MJ, talikas CD, Tzouwarakarayanni M, Karayannis KI (1996) Talanta 43: Cheng CF, Tsang CW (1998) Food Addit Contam 15: Jain, mith RM, Verma KK (1997) J Chromatogr A 760: Jimidar M, Harmann C, Cousement N, Massart DL (1995) J Chromatogr A 760: Bertotti M, Pletcher D (1997) Anal Chim Acta 337:49 8. chaaler U, Bakker E, pichiger E, Pretsch E (1994) Talanta 41: Ximenes MIN, Rath, Reyes FGR (2000) Talanta 51: Cox RD (1980) Anal Chem 52: Verma KK, tewart KK (1988) Anal Chim Acta 214: Kawakami T, Igrashi (1996) Anal Chim Acta 333: Gettha K, Balasubrramanian N (2000) Anal Lett 33: AOAC (1995) Method , Official methods of analysis, 16th edn. AOAC, pp Norwitz G, Keliher PN (1982) Anal Chem 54: Norwitz G, Keliher PN (1982) Talanta 29: Clesceri L, Greenberg AE, Eaton AD (1998) tandard methods for the examination of water and wastewater, 20th edn. American Public Health Association, Washington, pp afavi A, Haghighi B (1997) Talanta 44: Hecht F (1975) In: Kolthoff IM, Elving PJ (eds) Treatise on analytical chemistry, part I, Vol 11. Wiley, New York, pp Dhont JH (1960) Analyst 85: Coppola ED, Wickroscki AF, Hanna JG (1975) J Assoc Off Anal Chem 58: Tanaka A, Nose N, Iwasaki H (1982) Analyst 107: Higuchi H, Motomizu (1999) Anal ci 15: Haghighi B, Tavassoli A (2001) Talanta (in press) 25. Miller JC, Miller JN (1992) tatistics for analytical chemistry, 2nd edn. Ellis Horwood, New York, p 115