Bromate determination in water using chlorpromazine after correction of chlorinating agents and humic substances interference

ISSN 1061-9348, Journal of Analytical Chemistry, 2007, Vol. 62, No. 11, pp. 1055–1063. Pleiades Publishing, Ltd., 2007. ARTICLES
Bromate Determination in Water Using Chlorpromazine
after Correction of Chlorinating Agents and Humic Substances
Interference1
M. G. Mitrakas
Laboratory of Analytical Chemistry, Chemical Engineering Department, School of Engineering, Aristotle University of Thessaloniki, Thessaloniki, 54124 Greece Received April 28, 2006; in final form, February 16, 2007 Abstract—The presence of soluble humic substances and chlorinating agents interfered positively with the
spectrophotometric determination of bromate ( BrO ) using chlorpromazine. Removal of the soluble humic substances through their precipitation by a basic lead acetate (15.9 g/L Pb(CH3COO)2 · 3H2O—4.7 g/L PbO)solution corrected their interference effectively. In addition, the use of NaHSO 2–OCl , and Cl2–NH2Cl, when present in concentrations of up to 1.5, 3.5, and 3.5 mg/L, respectively. Thus, the spectrophometric method was rendered suitable for the direct bromate deter-mination in natural, chlorinated, and ozonated waters, since the application to such samples resulted in the accu- rate and precise determination of bromate. The method’s detection limit was estimated as 1.6 µg /L the linear range of the calibration curve was extended up to 700 µg BrO /L. The method also gave results com- parable to those obtained by the well-established ion chromatographic method and had the additional advantageof being simple, rapid, low cost, and suitable for brackish water.
DOI: 10.1134/S1061934807110093
Ozonation can improve the odor and taste of drink- method. IC, however, is not free from weaknesses and ing water effectively, remove coloration, oxidize fer- difficulties, the main one being chloride interference [5, rous and manganous ions, and mainly destroy microor- 6]. Weinberg (1994) overcame this problem by using a ganisms. Thus, it appears as a promising alternative dis- silver cation resin and a chelation column to remove infection method for drinking water. Ozone, however, leached silver, in order to protect the separation col- also oxidizes bromide (Br–) to bromate (BrO– ). The umn. The detection limit of the method was 0.5 µg/L after the application of a preconcentration technique.
latter has been considered as a potential carcinogen and Since the established maximum contaminant level of has been classified in Group 2B by the International bromate has been set at a very low value (10 µg/L), all Agency of Research on Cancer (IARC). The World proposed methods for bromate quantification focus on Health Organization (WHO) recommended the provi-sional guideline value of 25 µg/L, which is associated achieving high selectivity, sensitivity, and low detectionlimit [7–9].
with an excess lifetime cancer risk of 7 × 10–5. Sincebromate ion is considered as a hazardous substance, the Due to certain difficulties involved with the IC European Community [1] and USEPA [2] have estab- method, simpler and cheaper spectrophotometric BrO–3 lished the value of 10 µg BrO– /L as the maximum con- determination was sought and proposed. The method employed chlorpromazine [10] and other phenothiaz- The impact of bromate on human health has resulted ines [11] as color-producing reagents and seemed to be in the appearance of several publications during the last an interesting and attractive alternative method. In an decade dealing with the analytical methods of bromate acidic environment, phenothiazines are oxidized by determination in potable waters and possible interfer- BrO– to form stable, colored cations. This method [10– ences [2–6]. The commonly used method for bromate determination is ion chromatography (IC), which has 12] had a low detection limit and no interference by Cl– been standardized and considered as a reference and other anions and cations commonly present in nat-ural water. Since its development, the method was 1 The text was submitted by the author in English.
applied in only one case to natural water samples as post-column bromate determination following IC sepa- water in a 1-L cylinder-shaped vessel equipped with a ration [13]. Recently, Mitrakas and coworkers [12] porous glass diffuser. Ozone was generated from 99.5% tried to apply the method to natural water samples, but pure oxygen. The distilled-deionized water was found that the presence of soluble humic substances adjusted to pH < 3 with H2SO4 to prevent ozone decom- resulted in positive interference giving high pseudo- position. The ozone concentration in the stock solution bromate values, thus rendering the method unsuitable (typically about 18 mg/L) was measured by the iodo- for bromate determination in natural water and restrict- metric method [16]. An aliquot of the ozone stock solu- ing its use in pure bromate solutions. The positive inter- tion (10–50 mL) was mixed with 2 L of the water sam- ference was attributed to the electron acceptor groups ple. Bromate was determined at least 24 h after sample invariably existing in the humic molecules serving as ozonation when any molecular ozone, hydroxyl, and bicarbonate radicals were not present.
Removal of soluble humic substances can, in theory, be achieved either through the use of membrane filtersof the appropriate pore size or by means of various floc- Total Organic Carbon (TOC)
culants or precipitants (inorganic salts), which have A Shimadzu 500 TOC analyzer was used.
been used successfully in various cases for the removalof suspended organic materials from liquids. In addi-tion, chlorpromazine and other phenothiazines are also 2 [11], which can be effectively removed by sulfite Bromate. A 1000 mg/L BrO stock solution was prepared by dissolving the appropriate amount of thereagent-grade KBrO in distilled-deionized water and Since the spectrophotometric method of BrO– stored at 4°C. Working standards were daily prepared determination using chlorpromazine seems to be a sat- by proper dilution of the stock solution. Chlorprom- isfactory alternative to the IC method, this study was azine (CLP). A 1200 mg/L solution was prepared by undertaken with the following objectives: dissolving the appropriate amount of reagent in dis- tilled-deionized water. Keeping the CLP solution —various techniques (filtration, precipitation) in refrigerated in a dark bottle resulted in a lifetime greater than 3 months. Nitrite. Standard and working solutions —sulfite ion in removing chlorinating agents; and were prepared, standardized, and assayed as describedby the standard methods [16]. Sulfamic acid. A 1 M • to propose a procedure which would permit the solution of NH2SO3H was used. Sodium bisulfite. A accurate determination of BrO– , thus rendering the 0.2 g/100 mL solution (pH 4.2) was daily prepared by spectrophotometric method suitable for application to dissolving reagent-grade NaHSO3 in oxygen-free cool ozonated, chlorinated, and natural water. The results water and kept in the dark. Natural organics. The will be substantiated by using IC as a reference method.
sodium salt of humic acid (Aldrich, H1675–2, LotNumber 16308) was used for simulating the presenceof soluble natural humic substances. Hydrochloric acid. Analytical grade HCl from Baker, stored in a glass Natural water samples. Ten natural water samples
bottle, was used, because it gave the lowest bromate were collected from Northern Greece and will be referred to as samples N1–N10 henceforth. Treatmentof the samples included filtration through a 0.45-µmpore-size membrane filter, ozonation, and/or chlorina- tion. Chlorinating agents were added to ozonated water Iron chloride [FeCl ], aluminum sulfate [Al (SO ) ], after excess ozone removal by rapid filtration through granular activated carbon. Bromate, however, was and basic lead acetate at different concentrations were determined at least 24 h after any disinfection treat- used. The latter is used in sugar factories as an extract- ant and to remove suspended organic matter fromsugar-beet sap. Since the former two reagents did not Artificial water. Background ions included sodium
control the interference effectively, while the latter did bicarbonate 5 × 10–3 M, calcium chloride 5 × 10–4 M, (see Results and Discussion), it was adopted as the pre- and sodium sulfate 5 × 10–4 M were dissolved in cipitant in the proposed procedure and composed of organic-free distilled-deionized water (pH 7.8 ± 0.1) to 15.9 g/L Pb(CH3COO)2 · 3H2O and 4.7 g/L PbO. For its simulate representative freshwater conditions.
preparation, the reagents were dissolved in the appro- Ozonation procedure. An Ozonia® Triogen Model
priate volume of distilled water under continuous stir- Compact TOGC2A ozone generator was used. The ring for approximately 24 h, the suspension was left for ozone stock solution was prepared batchwise by bub- sedimentation, and the supernatant clear solution was bling ozone through organic-free distilled-deionized JOURNAL OF ANALYTICAL CHEMISTRY Vol. 62 No. 11 2007 BROMATE DETERMINATION IN WATER USING CHLORPROMAZINE prepared by proper dilution and standardized using theDPD colorimetric method [16].
Sodium hypochlorite. A Riedel de Haen stock solu-
tion was used. A chlorine solution of about 100 mg/L was Sodium chlorate. A 1000 mg/L ClO stock solu-
daily prepared by proper dilution of the stock solution tion was prepared by dissolving the appropriate amount in distilled-deionized water and standardized by a of reagent-grade NaClO3 in distilled-deionized water.
0.01 N standard sodium thiosulfate titrant. Working Working standards were prepared by proper dilution of cial water were prepared from a chlorine solution and Sodium perchlorate. A 1000 mg/L ClO– stock
standardized using the DPD colorimetric method [16].
solution was prepared by dissolving the appropriate Chloramine stock solution. A 200 mg/L Cl
deionized water. Working standards were prepared by 2Cl nominal concentration solution was prepared on the day of the experiments by mixing NaOCl and proper dilution of the stock solution.
(NH4)2SO4 stock solutions in distilled-deionized waterat 4 : 1 chlorine to ammonium—nitrogen mass ratio (0.8 : 1 on a molar basis). To obtain the highestmonochloramine yield and minimize ammonia volatil- Spectrophotometry. Instrumentation. A Lambda 2
UV/VIS spectrophotometer version 3.7 Perkin Elmer adjusted to pH 8.3 with either 0.5 M H2SO4 or 1 MNaOH before mixing and the NaOCl solution was Procedure. (1) A 40-mL water sample was mixed
with 3 mL of a basic lead acetate solution (correspond- ing to the final lead concentration of about 865 mg/L), preparation of the chloramine stock solution, pH was agitated gently for 5 min, and the suspension was fil- maintained at 8.3. Working standards in the range of 1– tered through a 0.45-µm pore-size membrane filter.
5 mg/L Cl2–NH2Cl in artificial water were prepared by Stirring gently, the following reagents were added to proper dilution of the chloramine stock solution and the clear sample solution: (2) 0.1 mL of 0.5 M HCL and standardized using the DPD colorimetric method [16].
0.1 mL of sodium bisulfite solution, followed by 5 mincontact time for chlorinating agents removal; (3) 0.1 mL Sodium chlorite. A 100 mg/L ClO– nominal con-
of 1.5 M KOH, followed by 5 min contact time for oxi- centration solution was prepared by dilution of 168 mg dation of a significant part of the excess sodium sodium chlorite of AGROS (80% in NaClO2) in dis- bisulfite; (4) 0.5 mL sulfamic acid solution, followed by tilled-deionized water. Solution pH was maintained at 3 min contact time for nitrite removal; and (5) 5 mL of higher than 7, thereby minimizing the potential conver- CLP solution and 4 mL acid catalyst (concentrated sion of ClO– to ClO– . This stock solution was stored HCl), followed by 5 min contact time for color devel- opment. The absorbance was measured at 527 nm in in an amber bottle in a cool, dark location. The titer of 10-cm path-length measurement cells.
the stock was quantified daily by a 0.01 N standardsodium thiosulfate titrant [16]. Working standards in Ion chromatography. The analysis was carried out
using a Dionex 4500I Ion Chromatograph under the the range of 1–5 mg/L ClO– in artificial water were below 10 µg/L. An overlap of bromate by chloride innatural water samples was observed. This overlap was attributed to the high injection volume, since bicarbon- Certain comments concerning the IC reference ate content of the samples influenced the ratio of method need to be mentioned. The linear regression equation of the calibration curve, in the range of 5– (HCO /CO ) in the eluent significantly, which in turn decreased the retention time of chloride. This overlap- ping was overcome by converting HCO of the natural The detection limit was about 5 µg/L and the method water samples to CO with the addition of an equiva- showed poor repeatability for BrO– concentrations lent amount of 1 M NaOH (meq OH– = meq HCO– ).
JOURNAL OF ANALYTICAL CHEMISTRY Vol. 62 No. 11 2007 water after its ozonation. All this implies that, if the interference was controlled, the spectrophotometricdetermination of bromate could become a successful and an easy-to-apply method compared to the alterna- Tests pertaining to the elimination of the interfer- ence were conducted on ten natural water samples prior to their ozonation. These samples analyzed using the reference IC method showed no bromate concentration (below the detection limit of 5 µg/L), while the directspectrophotometric method gave unrealistic high pseudo-bromate concentrations ranging between 10and 40 µg/L, due to the presence of soluble humic sub- stances, as was the case in a previous study [12].
Attempts to remove soluble humic substances, by pass-ing the samples through a 5,000-Dalton pore-size mem- An example (sample N3 prior to ozonation) of the effect of brane filter, failed to screen them out effectively. The different precipitants on the interference of soluble humic results also showed that neither ferric nor aluminum substances in the spectrophotometric determination of bro-mate. The quantity of the precipitant is shown in the salts at various concentrations were effective means in abscissa in terms of the corresponding metal concentration.
precipitating humic substances quantitatively, sincebromate pseudo-concentrations still persisted in allsamples after this treatment. An example of this effect Chlorinating agents NH2Cl and NaOCl were removed on sample N3 is shown in the figure. On the contrary, with the addition of sulfite ion in amounts stoichiomet- the use of a basic lead acetate solution at the quantity rically equal to their concentration in water samples mentioned in the proposed procedure (see Experimen- (see section on chlorinating agents—background).
tal) resulted in an effective control of humic substanceinterference in the determination of bromate by CLP, since pseudo-bromate levels were suppressed below thedetection limit in all samples. An example of this sup- Interference and their correction. Humic sub-
pression in sample No 3 is shown in the figure. The tests stances. A recent study has shown that soluble humic with the basic lead acetate solution were extended to substances, invariably existing in natural water, inter- pure humic acid solutions prepared in distilled-deion- fered positively, giving high pseudo-bromate values, ized water to simulate the presence of natural organics thus rendering the method unsuitable for bromate in water (Table 1). All humic acid solutions showed determination in natural water [12]. In addition, this pseudo-bromate concentrations when the CLP method interference would also overestimate bromate in drink- was directly applied without any treatment with a pre- ing water, since humic substances can persist in the cipitant. When, however, the proposed procedure wasapplied to artificial humic acid solutions in concentra-tion up to 5 mg/L, no bromate concentrations were Table 1. Pseudo-bromate concentrations determined in stan-
determined (Table 1). Consequently, these experiments dard humic acid solutions using CLP and their correction by provided strong evidence that the use of a basic lead acetate solution could effectively control the interfer-ence of humic substances in bromate determination Pseudo-bromate concentration, µg Br O /L using phenothiazines. The superiority of Pb relative to Fe and Al in removing humic substances can be attrib- uted to the nature of the cation and to the maintenanceof the sample pH in the range 5.5 to 6.5 upon the addi- tion of the basic lead acetate solution (Table 2). In con- trast, the addition of the other two metal salts lowers thesample pH in a more acidic region. It is known that divalent and trivalent metals form chelates with humic substances and the stability constant of Pb chelates ishigher than for the other two metals at the pH range 5 to 6.5 [17]. In addition, most of Pb chelates at this Pb concentration are water insoluble [17, 18]. Therefore,they are effectively removed from the samples.
Nitrite. Nitrite exhibits a strong interference in
BDL: Below detection limit = 1.6 µg Br O /L.
bromate determination with CLP (Eq. (2)). This JOURNAL OF ANALYTICAL CHEMISTRY Vol. 62 No. 11 2007 BROMATE DETERMINATION IN WATER USING CHLORPROMAZINE Table 2. pH changes in the various steps of the proposed method as a dependence of bicarbonate concentration of artificial
water samples
Table 3. Control of nitrite interference in the proposed method and the influence of contact timea and bicarbonate concen-
tration
a Time lapsed between sulfamic acid and acid catalyst addition.
c 25 ± 1 indicate the result within the absolute precision.
interference is usually controlled with sulfamic radicals were not present. Preliminary experiments, • The reaction of sulfamic acid with nitrite is favored × 10–3, R2 = 0.999. [2] in an acidic environment.
• A significant loss in sensitivity was observed when In the proposed procedure with ozonated water sam- the acid catalyst (HCl) was added to the sample prior to ples, sulfamic acid eliminated nitrite interference when the addition of CLP, in agreement with Farrel et al. [11].
present in concentrations of up to 500 µg/L (Table 3),without any influence on the sensitivity of spectropho- • Nitrite will still interfere when CLP, HCl, and sul- tometric bromate determination. Ozonated water sam- famic acid are added simultaneously.
ples were spiked with nitrite 24 h after their ozonation In conclusion, the addition of 0.5 mL 1 M sulfamic when any molecular ozone, hydroxyl, and bicarbonate acid in a water sample after the removal of humic sub- JOURNAL OF ANALYTICAL CHEMISTRY Vol. 62 No. 11 2007 stances and before the addition of the acid catalyst, The rate of reaction {4} is significant for practical resulted in a pH value of about 2.5 ± 0.1, which favored purposes in pH values lower than 4 [21].
nitrite destruction with no further loss of sensitivity, Optimization of NaHSO3 addition. Preliminary
due to the CLP addition at this low pH value (Table 3).
studies showed that NaHSO , in concentrations of up to It is self-obvious, however, that in water samples that 20 mg/L, when added simultaneously with CLP did not do not contain nitrite, the addition of sulfamic acid can interfere. These experiments have provided strong evi- be omitted. The pH of the sample after sulfamic acid dence that bromate selectively oxidizes CLP in the addition is influenced by its bicarbonate concentration presence of a sulfite ion. However, the suppression of (Tables 2 and 3). Bicarbonate concentration in water bromate sensitivity observed when NaHSO was added samples between 1 and 8 × 10–3 M resulted in pH values in water samples before sulfamic acid, was attributed to lower than 2.6, promoting the fast destruction of nitrite.
partial bromate destruction (reaction (4)). Optimization Higher bicarbonate levels resulted in pH values above of the procedure at that step proved that: 2.7, which, in turn, significantly increased the reactiontime for nitrite removal (Table 3, line 8). A small paral- • The addition of up to 5 mg/L NaHSO3 (Eq. (3)) did lel displacement of the calibration curve was observed not suppress the sensitivity of bromate (compare linear when bicarbonate concentrations were greater than regression Eqs. (3) and (10)), as well as the nitrite 10−2 M and it is, therefore, recommended that the stan- removal presented in Table 3, when the time lapse dards should contain similar bicarbonate levels.
between sulfamic acid and acid catalyst addition waskept at 3 min: Background of chlorinating agents removal. The
most effective and widely used dechlorinating agent is the sulfite ion. Since high pH values increase the reac- • By increasing either the time lapse between sul- tion rate of a sulfite ion with oxygen, NaHSO3 (stock famic acid and acid catalyst addition at 5 min (Eq. (4)) solution pH 4.2) was selected as a dechlorinating agent and 10 min (Eq. (5)) or NaHSO addition at 7.5 mg/L 2SO3 (stock solution pH 8.3), to minimize (Eq. (6)), a suppression of bromate sensitivity was the sulfite ion–oxygen reaction during the day.
observed because of the partial bromate destruction Data in the literature indicate that reaction {1} is immeasurably fast [19]. Reaction {2} of a sulfite ion with monochloramine, as well as with NHCl 2 is completed in a matter of seconds [15]: Optimization of Chlorinating Agent Removal During the disinfection of drinking water, chlorine As expected, chlorinating agents OCl– (Eq. (7)) and dioxide (ClO2) reacts primarily by a one-electron oxi- NH2Cl (Eq. (8)) interfered in bromate determination dative pathway and, thus, the principal inorganic with CLP. From oxychlorine residuals, the chlorite ion byproduct is almost invariably chlorite ion (ClO– ) in a (ClO ) also interfered (Eq. (9)), while the chlorate ion nearly 70% yield [20]. Consequently, the chlorinating agent’s interference control should mainly include the (ClO ) and the perchlorate ion (ClO ) in concentra- effective chlorite ion removal. The main reaction of the tions up to 1 mg/L did not. The USEPA recommends chlorite ion with the sulfite ion in the pH 4–7.5 region that the combined residuals of ClO– , ClO– , and ClO– must not exceed 1 mg/L in finished water.
This corresponds to a stoichiometry of two moles of SO2– consumed for every mole of ClO– removed. In the range of pH 4.5–5.5, the total removal of the chlo- rite ion (i.e., >99%) was completed in less than 5 min[14] and the loss of the sulfite ion from the competing By adding NaHSO3 before (pH region 7–8) or after sulfite ion–oxygen reaction was minimized.
the removal of humic substances (pH region 5.5–6.5), The sulfite ion, however, also reacts with the bro- complete OCl–/NH2Cl and partial ClO removal was mate ion. The reaction for the reduction of bromate by observed (Table 4). After adjusting, however, pH at value of 5, before adding the precipitant, bromatepseudo-concentrations were measured, although no chlorine compounds were determined in any natural JOURNAL OF ANALYTICAL CHEMISTRY Vol. 62 No. 11 2007 BROMATE DETERMINATION IN WATER USING CHLORPROMAZINE water samples, whether chlorinated or not. These bro- Table 4. Influence of pH on the elimination of 0.5 mg/L var-
mate pseudo-concentrations were probably attributed ious chlorine agents in artificial water, using 5 mg/L to an ineffective humic substance removal by precipita- tion at this low pH value. Consequently, the precipita-tion (step 1) must precede the chlorinating agent removal. Experimental results showed that the addition of 0.1 mL HCl 0.5 M in water samples after precipita- tion decreased pH in the 5 ± 0.2 region, assuming thepresence of 1 to 8 × 10−3 M bicarbonate (Table 2). In this pH region, rapid and effective elimination of 2Cl, OCl , and ClO was observed (Table 4). The addition, on the other hand, of 0.1 mL KOH 1.5 M assured a pH increase to a value higher than 8. In this context, it should be noted that, at pH 8 and above, inthe presence of air, the rate of the sulfite ion–oxygen reaction significantly increases [14]. Consequently,through the reaction with oxygen, the excess sulfite ion concentration was effectively decreased, which, in turn, minimized the bromate sensitivity suppression at step 4of the proposed procedure. When, finally, the proposed * BDL: Below detection limit = 1.6 µg Br O /L.
procedure was applied to artificial solutions of chlori- nating agents, the interference of OCl– up to 3.5 mg/L(Cl2–OCl–), of NH2Cl up to 3.5 mg/L (Cl2–NH2Cl), and The detection limit of the method, calculated [22] of ClO– up to 1.5 mg/L, was effectively eliminated.
from seven replicatations of 1–5 µg BrO– /L, was esti- These values agree closely to the stoichiometry of reac- tions (1), (2), and (3), respectively, and also with data mated to be 1.6 µg BrO /L, in agreement with USEPA presented in the literature [14, 15].
method 300 involving the separation and post-columncolorimetric determination of bromate with CLP [13].
Bromate Determination in Ozonated Waters • common ions in water such as K+, Ca2+, Mg2+, The proposed procedure for bromate determination NO– , and SO2– , in concentrations up to 500 mg/L, and using CLP was also applied to ten ozonated and/or chlorinated water samples in order to test its suitability • trace elements such as F–, I–, Br–, Mn2+, Zn2+, Sr2+, for bromate-containing samples and establish its ability Al3+, and B in concentrations up to 500 µg/L, did not to control interference. The bicarbonate content of the interfere with the proposed procedure. These findings samples ranged between 1.5 and 6.5 × 10–3 M, TOC are in agreement with earlier reported results [10–12].
0.3–1.2 mg/L, while manganese and total iron were less The presence of NaCl in concentrations of up to 3% than 10 µg/L. No nitrites were measured after ozona- did not interfere with the proposed procedure. This was tion. All water samples were relatively low in total salt an additional advantage of the proposed method in content as judged by their specific electrical conductiv- comparison to the IC method, rendering the spectro- photometric method suitable for direct application inbrackish waters.
Bromate values obtained using the CLP method cor- related closely with those obtained by ion chromatog-raphy (Table 5), indicating the accuracy of the method.
Calibration Curve and Detection Limit The regression lines had a slope and an intercept not Due to the photosensitivity of blanks with chlorpro- statistically different from 1 and 0, respectively, at the mazine, distilled-deionized water was used as a blank 1% probability level. The method of bromate determi- solution. The optimization procedure for CLP and HCl nation with CLP also proved to be precise, since the showed that, by using the proposed quantities, the high- coefficient of variation (CV) was lower than 4% on est precision and sensitivity for concentrations lower average in the range of 25–100 µg BrO– /L (Table 5).
than 25 µg BrO– /L were achieved. The equation of the cal- This means that the absolute precision at bromate con- ibration curve in the linear range up to 700 µg , the reason that nitrite interference is considered to be controlled when the bromate concentration w abs = 23 × 10–4Xconc + 18 × 10–3, JOURNAL OF ANALYTICAL CHEMISTRY Vol. 62 No. 11 2007 Table 5. Determination of bromate in ozonated natural water samples by the spectrophotometric method using CLP vs anal-
ysis by ion chromatography (IC)
a Coefficient of variation, estimated from 7 replicates.
b Significant at 1% probability level.
to be 25 ± 1 µg BrO– /L (Table 2). However, for the EC • Nitrite was eliminated with sulfamic acid at pH 2.5 ± 0.1 (procedure step 4). Chlorinating agents and ozone, and USEPA MCL of 10 µg BrO– /L, the absolute preci- however, oxidize the nitrite of water to nitrate and, con- sion was also about ±1 µg BrO– /L, since the CV was sequently, they scarcely coexist, suggesting the bypass between 6–9% in the range 2–10 µg /L.
Removal of interfering natural organics, nitrite, chlorinating agents, and their byproducts is included in In conclusion, as all interference parameters hardly • In natural water as well as in chlorinated and ozo- ever coexist, the proposed method could always be nated water, naturally occurring humic substances simplified by skipping the respective step of the pro- strongly interfered in bromate determination with CLP, cedure. The use of the spectrophotometric method for and humic substances were effectively removed by bromate determination gave accurate and precise basic lead acetate (procedure step 1).
results that are similar to those obtained by the well- • Chlorinating agents and their byproducts were established chromatographic method. An additional advantage of the spectrophotometric determination of cedure steps 2 and 3). It is self-evident that, in nonchlo- bromate using phenothiazines stems from its simplic- rinated water, steps 2 and 3 can be omitted.
JOURNAL OF ANALYTICAL CHEMISTRY Vol. 62 No. 11 2007 BROMATE DETERMINATION IN WATER USING CHLORPROMAZINE 10. Gordon, G., Bubnis, B., Sweetin, D., and Kuo, C.A., Ozone Sci. Eng., 1994, vol. 16, p. 79.
The author is thankful to G. Kyriacou for his contri- 11. Farrel, S., Joa, F.J., and Pacey, E.G., Anal. Chim. Acta, bution with the ion chromatography determinations and V. Keramidas for his essential comments and editorialassistance.
12. Mitrakas, M., Tzimou-Tsitouridou, R., and Kerami- das, V., Int. J. Environ. Anal. Chem., 2000, vol. 78,nos. 3–4, p. 343.
13. Walters, B. and Gordon, G., Anal. Chem., 1997, vol. 69, 1. Relating to the Quality of Water Intended for Human 14. Gordon, G., Slootmaekers, B., Tachiyashiki, S., and Consumption, Directive 98/83/EC, 1998.
Wood III, D.W., J. AWWA, 1990, vol. 82, no. 4, p. 160.
2. Final Stage 1 Disinfectants and Disinfection Byproducts 15. White, G.C., Handbook of Chlorination and Alternative Rule, Fed. Reg. 63:241: 69390, 1998.
Disinfectants, New York: Wiley, 1999.
3. Kuo, C., Krasner, S.A., Stalker, G.A., and Wein- 16. Standard Methods for the Examination of Water and berg, H.S., Analysis of Inorganic Disinfection By-Prod- Wastewater, New York: APHA, AWWA, WEF, 1992, ucts in Ozonated Drinking Water by Ion Chromatogra- phy, Proc. AWWA WQTC, San Diego, 1990.
17. Scniter, M., Reaction Between Organic Matter and Inor- 4. Hautman, D. and Bolyard, M., J. AWWA, 1992, vol. 84, ganic Soil Constituents, Int. Congr. on Soil Science, 5. Weinberg, H.S., J. Chromatogr. A, 1994, vol. 671, 18. Stevenson, F.J., Humus Chemistry: Genesis, Composi- tion, Reactions, New York: Wiley, 1994.
6. Joyce, R. and Dhillon, H., J. Chromatogr. A, 1994, 19. Fogelman, D.K., Walker, M.D., and Margerum, W.D., Inorganic Chemistry, 1989, vol. 28, no. 6, p. 986.
7. Kohler, K., Novak, M., and Seubert, A., Fresenius 20. Werdehoff, S.K. and Singer, C.P., J. AWWA, 1987, J. Anal. Chem., 1997, vol. 358, no. 4, p. 551.
8. Inoue, Y., Sakai, T., Kumagai, H., and Hanaoka, Y., Anal. 21. Gordon, G., Gauw, D.R., Emmert, L.G., Walters, D.B., Chim. Acta, 1997, vol. 346, no. 3, p. 299.
and Bubnis, B., J. AWWA, 2002, vol. 94, no. 2, p. 91.
9. Seubert, A., Schminke, G., Nowak, M., Ahrer, W., and 22. Glaser, J., Foerst, D., McKee, G., Quave, S., and Buchberger, W., J. Chromatogr. A, 2000, vol. 884, Budde, W., Environ. Sci. Technol., 1981, vol. 15, no. 12, JOURNAL OF ANALYTICAL CHEMISTRY Vol. 62 No. 11 2007

Source: http://www.researchvalue.net/repository/files/document/1_Journal_of_Analytical_Chemistry_2007_Vol_62_No_11_pp_1055_1064.pdf