202
ANALYTICAL CHEMISTRY, VOL. 50, NO. 2, FEBRUARY 1978
may be statistically dubious at higher or lower concentrations. For those instances where natural water samples failed to provide an aqueous cobalamin extract with a minimum 200 pg/mL concentration, the extract was lyophilized to dryness and reconstituted to a smaller volume with 0.9% saline solution prior to CIFB analysis. Aside from the fact that the CIFB assay method is less time consuming and tedious than bioassays, it permits the analysis of cobalamins in a more representative size water sample and substantially eliminates interference of noncobalamin compounds in quantitative assays through implementation of selective benzyl alcohol extraction and concentration procedures. Moreover, the CIFB assay method described in this paper provides a means for evaluating cobalamin concentrations in natural waters on a routine basis using a sensitive analytical approach other than traditional bioassays.
ACKNOWLEDGMENT Thanks are expressed to Charles Zapsalis and Richard F. Milaszewski for helpful discussions. LITERATURE CITED (1) L. Provasoli, "Algal Nutrition and EvbophicakJn" in "Eutrophication: Causes, Consequences, Correctives", Proceedings of a Symposium, National
Academy Science, Washington, D.C., 1969, pp 574-593. J. Cairns, G. R. Lanza, and B. C. Parker, "Pollution Related Structural and Functional Changes in Aquatic Communities with Emphasis on Freshwater Algae and Protozoa", Proc. Acad. Nat. Sci. Philadelphia, 124 (5),79 (1972). A. F. Carlucci and S.B. Silbernagel, Limnol. Oceanogr., 11, 642 (1966). S. R. Hoover, L. Jasewicz, J. B. Pepinsky, and N. Porges, Sewage Ind. Wastes, 24, 38 (1952). H. Y. Neujahr, Acta, Chem. Scand., 9, 622 (1955). R. Strohecker and H. M. Henning, "Vitamin Assay-Tested Methods", Chemical Rubber Co. Press, Cleveland, Ohio, 1965. W. H. Sebreli and R. S.Harris, "The Vitamins", Vol. 11, Academic Press, New York, N.Y., 1968, pp 120-258. E. L. Smith, "Vitamin B,,", John Wiley and Sons, New York, N.Y., 1960. W. Shive, J. M. Ravel, and W. M. Harding, J. Bbl. Chem., 176, 991 (1948). R. A . Beck and J. J. Brink. Environ. Sci. Techno/.. 10. 173 (1976). K-S. Lau, C. Gottlieb, L. R. Wasserman, and V. Herbert, Blood, 26, 202 (1965). L. Wide and A. Killander, Scand. J . Ciin. Lab. Invest., 27, 151 (1971). C. Rosenbium, Talanta, 11, 255 (1964). S. Gutcho, J. Johnson, and H. McCarter, Clin. Chem. ( Winston-Salem, N.C.),19, 998 (1973). H. H. Fricke, U.S. Patent 2,582,589 (1952). RIA Products, Inc., "In Vitro Quantitative Measurement of Serum Vitamin B,, by Radioassay", Waitham, Mass., 1977. Y . K. Liu and L. W. Sullivan, Blood, 39, 426 (1972). 6.C. Parker, J . Phycoi., 5 , 124 (1969).
RECEIVED for review September 14,1977. Accepted November 21, 1977.
Determination of Chlorine Dioxide in Sewage Effluents J. Ross Knechtel" Wastewater Technology Centre, Burlington, Ontario L4R 4A6
Edward G. Janzen and Edward R. Davis University of Guelph, Guelph, Ontario N 1G 2 W 1
A spectrophotometric method is developed for the determination of chlorine dioxide in sewage treatment plant effluents. The decrease in absorbance at 550 nm of acid chrome violet K (ACVK) enables the direct spectrophotometric determination of C102in sewage effluent samples. Centrifugation is employed to remove suspended solids. I n a NH4CI-NH3 buffer of pH 8.1 to 8.4, no Interference from active chlorine, hypochlorites, chlorltes, chloramines, and nitrites was observed. The results obtained using the ACVK technique were verified against electron spin resonance spectrometry.
T h e subject of this paper is the use and measurement of chlorine dioxide in sewage treatment plant effluents. There were three good reasons for the consideration of chlorine dioxide as an alternative to chlorine ( I ) . In the first place, chlorine dioxide is a much stronger oxidizing agent than chlorine. Because of this, a smaller dosage of chlorine dioxide than chlorine should be possible for disinfection purposes. Because a smaller dosage is possible, a lower residual could result. The second reason for the choice of chlorine dioxide over chlorine was that chlorine dioxide does not react with ammonia to form chloramines as does chlorine ( 2 ) . The reaction of chlorine with ammonia does two things. It reduces the oxidizing power of the oxidant and produces a toxic by-product. The third reason for the choice of chlorine dioxide was that it oxidizes the phenolic ring in phenols ( 3 ) ,whereas, 0003-2700/78/0350-0202$01 .OO/O
chlorine forms phenolic chlorides when it combines with phenols. This could produce taste and odor problems in potable water supplies. Several published methods (4-9) exist for C102 determination in water but not for sewage effluents. These techniques are not entirely specific for C102 because the measurement includes other chlorine species ( 7 ) . Modifications in some of these procedures are suggested to mask or remove these effects. In the effluent disinfection project itself, the chlorine dioxide residues were measured amperometrically (9, IO) using phenylarsineoxide (PAO) as titrant. The chlorine dioxide values reported were not verified by a referee method. Because of the uncertainties of this technique and the results, work was undertaken to examine an alternative method. The method selected (8)t o measure chlorine dioxide depends on the selective decolorization of a dye (acid chrome violet K). The relationships between the level of chlorine dioxide initially present and the amount of decolorizing which occurs obeys Beer's law up to 200 wg C102. The maximum absorbance of the complex formed occurs a t 550 nm. A study was undertaken to investigate the feasibility of using the acid chrome violet K technique for the measurement of C102 in sewage effluent samples. The results obtained on sewage effluent samples using this technique were verified using an electron spin resonance method. The chlorine dioxide molecule is the only stable chlorine species which contains an unpaired electron and which is capable of existing in
e 1978 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 50, NO. 2, FEBRUARY 1978
solution. Thus, electron spin resonance should provide an accurate means of verifying any method for the determination of chlorine dioxide in solution. EXPERIMENTAL Apparatus. The colorimetric measurements were made on a Bausch and Lomb Spectronic 100 equipped with a 1-cm micro flow-through cell. This particular spectrophotometer has an expanded readout feature which permits the measurement of concentrations much lower than conventional spectrophotometers. The electron spin resonance (ESR) measurements were made on a Varian E-104 instrument using a Varian aqueous solution flat cell (fused quartz construction). Reagents. Sodium Chlorite. Technical grade sodium chlorite (80% NaC102, 20% NaC1) was obtained from Kingsley & Keith Ltd., (Toronto, Ontario, Canada). Chlorine Dioxide was prepared by reacting sodium chlorite with hydrochloric acid (11). Equal volumes of sodium chlorite solution (55.1 g/L NaC102) and 2 N hydrochloric acid were mixed. The gas produced was scrubbed to remove any chlorine present by passing through a saturated solution of sodium chlorite (39 g/100 cm3 of NaC102 at 17 "C). The purified chlorine dioxide was dissolved in distilled water. This reaction was allowed to proceed to completion (about 30 min). The solution was standardized using an iodometric titration with sodium thiosulfate (12). Acid Chrome Violet K . A stock solution was prepared by suspending 175 mg of acid chrome violet K (ESBE Laboratory Supplies, Toronto, Ontario) in distilled water. To the suspension were added 20 mg of purified sodium metaphosphate, 48.5 g of analytical grade ammonium chloride, and 5.3 mL of concentrated ammonium hydroxide solution (29% NH3). The suspension was finally diluted to 1 L with distilled water, thoroughly mixed, and set aside for 24 to 48 h to stabilize. The suspension was then filtered through a 0.45-pm membrane filter under vacuum. Our experience has shown that this solution is stable for at least 1 month. To prevent interference from active chlorine, hypochlorites, chlorites, and chlorates, the pH of the solution must be between 8.1 and 8.4. This is achieved after a 10-fold dilution of the ACVK solution with the sample to be analyzed. The absorbance of a 10% dilution of this ACVK stock with distilled water at 550 nm (the absorbance maximum) should be between 0.15 and 0.19 using the Bausch and Lomb Spectronic 100 (the spectrophotometer used in this study). Recommended Procedure. Standardization. Twenty milliliters of the acid chrome violet K solution are introduced into a 200-mL flask. A known amount of chlorine dioxide up to 200 fig is added, taking care that the solution drops directly into the reagent without running along the walls of the vessel. The solution is diluted to 200 mL with distilled water and the absorbance at 550 nm is read using distilled water as a blank. The absorbance of a 10% dilution of the ACVK reagent is measured also. The values of VA (absorbance of the 10% dilution of the reagent minus the absorbance of the decolorized sample) vary linearly with the amount of chlorine dioxide added up to 200 kg C102. Analysis of Residual CIOz in Samples. For the analysis of chlorine dioxide in samples, 20 mL of the ACVK reagent is diluted to 200 mL with sample. The sample is centrifuged and the absorbance at 550 nm of the clear supernatant liquid is measured (Reading 1). At the same time, 20 mL of the ACVK reagent is diluted to 200 mL with the same sample material containing no chlorine dioxide. This is accomplished by either taking a sample a t a point in the sewage treatment process just prior to the addition of chlorine dioxide or by eliminating the C102 in the sample by the introduction of 5 mg of phenol per liter. The solids are centrifuged as outlined above. The optical density of 550 nm of this second sample containing no C102 is measured (Reading 2). The difference between these two readings ( T A ) is then translated to micrograms of C102from the relationship derived in the previous (standardization)section. The concentration of chlorine dioxide in the sample, in mg/L, is obtained by dividing the micrograms of chlorine dioxide found by the aliquot of sample taken (mL).
RESULTS A N D D I S C U S S I O N I n t e r f e r e n c e s . By a procedure identical t o t h a t for standardization, the acid chrome violet K reagent was allowed
203
Table I. Species Found Not to Interfere with Chlorine Dioxide Determination ClO,, mg/L
Species and concentration level (mg/L) Tested
0.2 0.2 0.2 0.2 0.2
Chlorine (Cl,) u p t o 2 0 mg/L Hypochlorite (C10,') up t o 20 mg/L Chlorite (C10;) up to 50 mg/L Chloramine u p to 1000 mg/L Nitrite (NO;) u p t o 20 mg/L
Table 11. General ESR Instrument Parameters (for 52 mg ClO,/L) Microwave frequency Field
9.495 GHz
Scanning range Modulation frequency
1000 Gauss
3370 Gauss
100 KHz
Scanning time Receiver gain Modulation amplitude Time constant
30 min 2x
lo4
1 Gauss 2s
1' Flgure 1. ESR spectrum of 52 mg C I 0 2 / L in distilled water at room
temperature t o react with chlorine, hypochlorite, chlorite, chloramine T, and nitrite. These potential interferences were allowed to react together and in various combinations with acid chrome violet K both in the presence and absence of chlorine dioxide. The concentrations of various species which did not interfere are shown in Table I. Reproducibility. A volume of a nonchlorinated final effluent from a sewage treatment plant was treated with 4.0 mg/L chlorine dioxide. The suspension was mixed using a magnetic stirrer for 5 min. A sufficient quantity of acid chrome violet K reagent was added to bring its concentration to 10% by volume and the suspension was thoroughly mixed. Ten samples of 50 mL each were withdrawn, centrifuged, and readings taken on the Bausch and Lomb Spectronic 100 in t h e expanded concentration mode. T h e range of results calculated from these 10 readings was 0.38 to 0.40 mg C102/L. Verification of Colorimetric P r o c e d u r e . The chlorine dioxide molecule contains a n umpaired electron. T h i s property makes the molecule paramagnetic. No other stable chlorine species capable of existing in solution contains a n unpaired electron. Chlorine dioxide shows no tendency t o dimerize ( 2 ) . Any dimer found (C1204) would not have a n unpaired electron and thus would not exhibit electron spin resonance. The electron spin resonance technique should thus be able t o provide an accurate measurement of the chlorine dioxide concentration in samples and a means of verifying the colorimetric procedure. As well, the ESR technique has been advocated as a primary standard method for t h e analysis of c102 (13). A characteristic ESR trace of chlorine dioxide is shown in Figure 1. This particular curve is a trace resulting from t h e ESR analysis of a solution containing 5 2 mg C102/L. T h e spectrum shape is in agreement with t h a t found by another author (13). The ESR instrumental parameters used are listed in Table 11.
204
ANALYTICAL CHEMISTRY, VOL. 50, NO. 2 , FEBRUARY 1978
T
T
A Figure 5. ESR
spectrum of sewage effluent sample No. 2 at room
temperature
Figure 2.
ESR
spectrum of 1.3 mg C102/L in distilled water at room
temperature
Sample/ amplitude, standard cm Standard 1 8.4 Standard 2 17.0 Sample 1 10.2 Sample 2 12.4
T
A
Flgure 3. ESR spectrum of 0.65 mg C102/L in distilled water at room
temperature T
i\ W Figure 4,
ESR
Table 111. Chlorine Dioxide Measurement in Sewage ESR vs. ACVK Measurements
L
spectrum of sewage effluent sample No. 1 at room
temperature Before a comparison of results obtained using the acid chrome violet K test and the electron spin resonance method can be made, one important difference between the two tests must be discussed. The acid chrome violet K test gives the chlorine dioxide concentration in the sample a t the time the dye (ACVK) is added to and reacts with the chlorine dioxide. The ESR method gives a dynamic measurement-the chlorine dioxide concentration could likely be decreasing during the period of the measurement. The two tests must thus be made on the same sample as close as possible to each other in time. Figure 2 represents the spectrum of a 1.3 mg C102/L; Figure 3 represents the spectrum of a 0.65 mg C102/L. The instrumental parameters for these spectra are listed in Table 11, with the exceptions of receiver gain (5 X lo6),modulation amplitude (0.5 X lo2),and the time constant (4 s). A sample of nonchlorinated effluent from a sewage treatment plant was treated with 6 3 mg/L chlorine dioxide. The reaction was allowed to proceed for a pre-determined time period after which the sample was divided into two parts-one for electron spin resonance, and one for ACVK colorimetric analysis. The ESR scan was then started. After the first peak in the scan was obtained (see Figure 4), the ACVK colorimetric test was carried out on the other part of the sample. The time elapsed between the start of the ESR scan and the beginning ACVK test was about 5 min. The length of time required for the complete ESR scan was 21/4 min. Based on known decay curves of chlorine dioxide, a time of 21/4 minutes should not
C ~ O content ,
%
ESR
ACVK
Difference
0.65 1.30 0.78 0.95
...
... ...
0.85 0.89
9 6
amount to more than a 5% decay. Thus, the procedure used for comparing the methods as outlined in the previous paragraph should be valid. The ESR scans of two such samples are shown in Figure 4 and 5. The ESR instrumental parameters were the same as those for Figures 2 and 3. For this application of ESR, the maximum to minimum amplitude in an ESR spectrum is proportional to the concentration of the material present which exhibits the paramagnetism (14). When this measurement is made on two standards run earlier, 1.3 and 0.65 mg/L, a calibration for the analysis of samples with unknown C102 content can be set up. A similar measurement of the two samples referred to earlier can be made, and the chlorine dioxide content of these samples calculated. The maximum to minimum amplitudes along with the corresponding chlorine dioxide concentrations for standards and samples, as well as the differences between chlorine dioxide measurements obtained on the sewage samples using the two different techniques are shown in Table 111. From this last Table 111, it is clear that the measurements obtained using both techniques (ACVK-colorimetric and ESR) are comparable. The ACVK technique should be reliable for the direct colorimetric measurement of chlorine dioxide in sewage effluent.
ACKNOWLEDGMENT The authors earnestly thank Fritz Engler of the Ontario Ministry of the Environment (Pollution Control Branch) for his kind technical assistance and the use of his Bausch and Lomb Spectronic 100 during the study. LITERATURE CITED F. A. Tonelli, Ontario Ministry of the Environment,Toronto, Ontario personal communication, 1977. and Physical on D. H. Rosenblatt, "Chlorine Dioxide-Chemical Ozone/Chlorine Dioxae Oxidation of Organic Materials", Cincinnati, Ohio, November 1976. Olin Chemicals Report (Stamford, Conn.), "Treatment of Water Supplies with Chlorine Dioxide (C102)" (1975). M. A. Post, and W. A. Moore, Anal. Chem., 31, 1872 (1959). R. N. Ashton, J . A m . Water Works Assoc., 42, 151 (1950). J. E. Hailer, and S. S.Listek, Anal. Chem., 2 0 , 639 (1948). G. C. White, "Handbook of Chlorination", Van Nostrand-Reinhold, Toronto, 1972, pp 612-13. W. Masschelein, Anal. Chem., 38. 1839 (1966) Standard Methods for the Examination of Water and Wastewater, 14th Ed., 1975, APHA-AWWA-WPCF, p 357. R. W. Andrew, and G. E. Glass, "Amperometric Titration Methods for Total Residual Chlorine, Ozone, and Sulphite", U.S. EPA report unpublished, 1974.
A N A L Y T I C A L CHEMISTRY, VOL. 50, NO. 2, FEBRUARY 1978
205
and Practical Applications", McGraw-Hill, New York, N.Y., 1972, pp 462-463.
(11) L. Beuermann, Gas Wasserfach,, 106,78 (1965) (in German). (12) Ref. 9, p 342. (13)C. 8.Murphy, and E. C. Tim, Jr., "A Novel Application of Electron Spin Resonance,The Analysis of Chlorine Dioxide in Wastewater", (unpublished) Northeast Regional Meeting, American Chemical Society, Rochester, N.Y., October 1975. (14) J. E. Wertz, and J. R. Boiton. "Electron Spin Resonance-Elemental Theory
R~~~~~~~for review september 5 , 1977. ~~~~~~~d~~~~~b~~ 14, 1977.
Kinetic Microdetermination of Manganese in Nonferrous Alloys, and of Nitrilotriacetic Acid, Ethylenediaminetetraacetic Acid, and Diethylenetriaminepentaacetic Acid D. P. Nlkolells and T. P. Hadjlloannou" Laboratory of Analytical Chemistry, University of Athens, Athens, Greece
An automatlc spectrophotometric klnetlc method Is descrlbed for the ultramlcrodetermlnatlon of manganese, based on Its catalytlc effect on the perlodate-antimony( 111) reactlon. The method Is also used for the determlnatlon of nltrllotrlacetlc acld (NTA) and of ethylenedlamlnetetraacetlc acld (EDTA) and dlethylenetrlamlnepentaacetlc add (DTPA), on the bask of thelr actlvatlng and lnhlbltory effect, respectlvely, on the manganese-catalyzed perlodate-antimony( 111) reactlon. The tlme requlred for the reaction to consume a flxed amount of perlodate Is measured automatlcally and related dlrectly to the catalyst or actlvator or lnhlbltor concentration. Manganese In the 10-'-10-' M level, and NTA, EDTA, and DTPA In the lO-'-lO-'M level were determined wlth an average error of about 2 % . The method has been applled to the determlnatlon of manganese In nonferrous alloys.
antimony(II1) reaction t o consume a fixed amount of periodate, and thus for the absorbance to decrease by a preselected amount, is measured automatically with a solid state "double-switching" network (3,8)and related directly to the catalyst (Mn) or activator (NTA) or inhibitor (EDTA or DTPA) concentration. Rapid operation, good accuracy, and high sensitivity are t h e main features of t h e automatic method. Ultramicroamounts of manganese in the range 0.1-2 c(g (4.6 X 10-'-9.1 X lo4 M)in the absence of NTA, and 4-40 ng (1.8 X 104-1.8 X lo-' M) in the presence of NTA, were determined with an average error and relative standard deviation of about 2%. Measurement times varied from a few seconds to about 2 min. T h e method has been applied successfully to the determination of manganese in nonferrous alloys. Also, nanogram amounts of NTA and microgram amounts of EDTA and DTPA were determined with the same degree of accuracy and precision.
T h e phenomenon of catalysis has been widely used for the kinetic determination of trace amounts of various elements (I). In t h e case of metal ion catalysis, complexing agents forming highly stable metal complexes with such catalysts act as inhibitors and they can be determined on the basis of their inhibitory effect on t h e metal ion-catalyzed reactions (2-4). In other cases, the rate-increasing effect of complexing agents o n t h e metal ion-catalyzed reactions has been used for their determination (4-7). I n a study of periodate reactions, it was found t h a t t h e reactions of periodate with mercury(1) and antimony(II1) which proceed at a measurable rate a t room temperature are highly accelerated by trace amounts of manganese, in the ppb range. T h e periodate-antimony(II1) reaction was found t o be more sensitive. In this paper, a kinetic spectrophotometric method for the ultramicrodetermination of manganese(I1) is described, based on its catalytic effect on the periodateantimony(II1) reaction
EXPERIMENTAL
IO4- + Sb(II1) + 2H'
manganese catalyst
IO3- + S b ( V ) + H,O
The method has also been used for the determination of trace amounts of nitrilotriacetic acid (NTA) on t h e basis of its activating effect on t h e manganese-catalyzed periodateantimony(II1) reaction, and of ethylenediaminetetraacetic acid (EDTA) and diethylenetriaminepentaacetic acid (DTPA) on t h e basis of their inhibitory effect on the aforementioned reaction. I n all cases, t h e time required for t h e periodate0003-2700/78/0350-0205$01 .OO/O
Apparatus. As previously described (3, 8). The "doubleswitching" network was adjusted to measure the time required for the recorder pen to cross preselected positions in the chart, corresponding to 0.84 and 0.79 absorbance unit. Reagents. Water was purified by doubly distilling de-ionized water through an all borosilicate-glass still. This water passed the dithizone test (9) and was used throughout this work. Antimony Trichloride Stock Solution, 0.0041 M. SbC13, 0.9352 g, is dissolved in 1 L of 3 M HCl. More dilute solutions are prepared daily by dilution. Buffer, sodium periodate, manganese(II), and ligands solutions are prepared as previously described (3, I O ) . The very dilute NTA, EDTA, and DTPA solutions (
