Colorimetric Methods for Bromine Frank W. Sollo, Thurston E. Larson, and Florence F. McGurk Chemistry Section, Illinois State Water Survey, Champaign, Ill. 61820
Reportedly bromamines are far more effective disinfectants than the chloramines and approach free bromine in disinfecting power. This property of the bromamines could eliminate the need to reach breakpoint for disinfection of many waters. The major objective of this study was to develop suitable colorimetric methods for determining free bromine and bromamines in water. The possibility for use of chlorine and bromide ion to produce bromine necessitated examination of methods suitable for distinguishing between the halogens. Numerous reagents were screened on the basis of their reaction with HOCl, HOBr, NH2CI, and NH2Br. Five reagents (methyl orange, DPD oxalate, phenol red, brom cresol purple, and phenosafranin) were selected for detailed investigation. These five reagents, test methods, and a possible application of the test methods are discussed. ~~~
U
nder certain circumstances neither chlorine nor iodine is entirely satisfactory as a viricidal and germicidal agent in the preparation of potable water. Many water sources contain ammonia in such quantity that chlorination results in the formation of chloramines, which, in waters of alkaline pH, are far less effective as disinfectants than free chlorine; and the effectiveness of iodine decreases with decreasing p H and increasing concentration of iodide ion. It seemed possible that bromine had properties which could avoid these problems. In comparing bromine with chlorine o n the basis of disinfection ability, free bromine and free chlorine are approximately equal on a molar basis. However, bromamines are reportedly far more effective than chloramines, approaching free bromine and free chlorine in effectiveness. Although the bromamines are generally less stable than the chloramines, at the higher p H conditions bromamine stability is increased sufficiently to produce greater disinfection efficiency than that of the chloramines. The availability of reliable analytical tests is a primary requirement in any evaluation of bromination for disinfection of water. Few such tests are available. Stenger and Kolthoff (1935) worked with phenol red as a method for the determination of the bromide ion, oxidizing the bromide ion to free bromine in the presence of phenol red. A similar method is given in “Standard Methods for the Examination of Water and Wastewater” (American Public Health Association, 1965). Hashmi and Ayaz (1963) and Anderson and Madsen (1965) presented useful methods for determining the proportions of hypobromite, bromite, and bromate present in stock bromine solutions. Only one test has been available for distinguishing free and combined bromine. This is the amperometric titration as proposed by Johannesson (1958), based on the assumption that the 240 Environmental Science & Technology
free and combined forms do not coexist. Addition of ammonia causes a reduction of the diffusion current if the bromine is present in the free state and no change if it is combined. By the usual titration, the total bromine can be determined. Chlorine and chloramines may be determined in the same way if potassium iodide is added to permit titration of the chloramine, either present originally or formed upon addition of the ammonium salt. Johannesson (1958) also used the neutral orthotolidineferrous ammonium sulfate titration for total bromine, and Palin (1961) has adapted the N,N-diethyl-p-phenylenediamine (DPD)procedure for this purpose. The ideal method should permit the determination of either free or combined bromine in the presence of free or combined chlorine. The reason for this requirement is that bromine may be applied by adding a soluble bromide, followed by chlorine or a hypochlorite. Depending upon p H and ammonia content of the water, the result may be a mixture of either free chlorine and bromine or chloramine and bromamine. Distinguishing between free chlorine and free bromine is probably not of the utmost importance, but because of the difference in germicidal activity, it would be necessary to determine bromamine in the presence of chloramine. In this study, 27 reagents were evaluated for possible use in the determination of free and combined bromine. The problems considered in preliminary evaluations of the 27 reagents were sensitivity and reproducibility of tests, stability of reagents, and interference from naturally occurring substances in water. Five reagents that appeared to be most promising were selected for investigation : methyl orange, DPD oxalate, phenol red, brom cresol purple, and phenosafranin. Although not one of these is completely free of interferences, each serves a particular purpose, and a combination of the tests may be used to determine both the quantity and form of the total residual halogen present. The methods and the species of bromine and chlorine determined by each are shown in Table 1. Table 1. Proposed Methods for the Determination of Bromine and Chlorine Residuals
Method Methyl orange BrMethyl orange
+
oxalate oxalate K I (KI is premixed with sample) Phenol red Brom cresol purple (arsenite added at 1 min) Phenosafranin (arsenite added at 30 sec) DPD
DPD
+
Test
5.0 9.5
Residuals determined Clz Brp 4-bromamine CI2 Brp bromamine chloramine Clz Brz bromamine C12 Br2 bromamine chloramine Br? bromamine BrB
9.5
Br2
pH
2.0 2.0
6,3 6.3
+ + + + + + +
+ + +
Preparation of'Stundurd Bron~ineand Chlorine Solutions
Bromine and chlorine demand-free water was used in the preparation of all reagents and samples. This water was prepared by deionizing tap water and then distilling from acid in a n all-glass apparatus. All glassware and other equipment in contact with bromine samples were either soaked in a strong bromine solution or cleaned with chromic acid and rinsed thoroughly with the bromine and chlorine demand-free water before use. Bromine Stock Solution. These solutions were prepared by adding 2 ml liquid bromine to 1 liter of 0.15M NaOH. The final p H of the solution is 12. Standard Bromine Solutions. A standard free bromine solution was prepared daily by dilution of the stock bromine solution and standardized by iodometric titration with sodium thiosulfate. Bromamine Stock Solutions. These solutions were prepared by methods similar to those of Galal-Gorchev and Morris (1962). Standard bromine solution, equivalent to 3 mg/liter available bromine in the final dilution, was added, with mixing, to buffered ammonium chloride solutions and then diluted to a final volume of 200 ml. The resulting 200-ml sample contained approximately 0.6 mg of available bromine. ( a ) Monobromamine solutions were prepared by adding standard bromine solution to ammonia at p H 8.8. The resulting solution has a NH,/Br?. ratio of 20 :1 by weight. Ultraviolet absorption spectra indicate that this solution contains exclusively monobromamine. (0)Dibromamine solutions were prepared by stirring equal quantities of bromine and ammonia together at a p H of 7.3. Ultraviolet absorption spectra indicate that this solution is a mixture of monobromamine and dibromamine. I t is not possible to prepare solutions which are exclusively dibromamine. (c) Tribromamine solutions were prepared by adding bromine to ammonia a t pH 5. The resulting solution has a NHj/BrLratio of 1 :3 by weight. Ultraviolet absorption spectra indicate that the main bromine-containing species is tribromamine. Bromamine solutions were prepared as indicated above to check bromamine response with the proposed reagents (Figure 1). Solutions of greater bromamine concentration may be prepared by increasing the ammonia and free bromine concentrations while maintaining the same weight ratio. At the time each bromamine sample was used, an identical sample was tested with methyl orange. The results of this test were used as a measure of the total bromamine present at that time. Standard Chlorine Solutions. A standard free chlorine solution was prepared daily by dilution of Clorox (5.25 NaOCl), and standardized iodometrically against sodium thiosulfate. Chloramine Stock Solution. The solution was prepared from 109.2 mg/liter ammonium chloride and 1 ml of Clorox with
168 mg/liter sodium bicarbonate to adjust the p H to 7.8 to 8.2. The solution was standardized immediately before use by the methyl orange test for total halogen. Such a solution was exclusively monochloramine. The NH,/Cl, ratio was approximately 1:1.5 by weight. Experimenral
Calibration curves for bromine and chlorine were prepared with use of a series of dilutions of iodometrically standardized bromine or chlorine solutions. For every method the proper portion of these dilutions was added to the buffered reagent and the resulting absorbance was plotted against the free halogen concentration. F o r the methyl orange, DPD, and phenol red reagents, the tests were performed as for calibration, by adding the sample to a premixed buffer and reagent, and then determining the absorbance at the proper time interval. This order of addition of reagents is necessary for obtaining reproducible results. For the brom cresol purple and phenosafranin reagents, the premixed buffer and sample were added to the reagent. This order of addition of reagents minimizes bromamine interference. However, in the preparation of free bromine calibration curves, this order of addition is not important since bromamines should not be present and the sample may be added to the combined reagent-buffer mixture. Interference from nitrite and ferric ions in concentrations up to 10 mgiliter, manganic ion in concentrations up to 3 mg/ liter, and chloramine in concentrations up to 4 mg/liter (as Cln) was determined for each reagent. Absorbance readings were made with a Beckman Model DB spectrophotometer with a 1.0-cm cell at the wavelength of maximum absorbance for the reagent tested. A Sargent water bath and cooler was used in controlled temperature studies. Test Methods- Discussion, Reagents, Procedures Methyl Orange (MO), pH 2. Methyl orange can be used as a quantitative reagent for total bromine plus free chlorine. If sufficient bromide ion is added, total bromine and total chlorine can be determined. The final test p H is 1.8 to 2.1 for samples containing up to 1000 mg/liter of alkalinity. This test relies on a bleaching reaction, and for this reason, accurate measurement of the reagent is necessary. The sample must be added with rapid mixing to a n excess of the acidified MO reagent. Beer's law is followed if a reasonable excess of MO is present. If methyl orange is added to the sample, or if the sample is added without adequate mixing, the bromine present will not bleach the MO quantitatively. When addition of the sample to MO produces a colorless solution, a larger quantity of MO should be used. The test must be repeated with a new diluted sample or with the larger quantity of Mo required. The calibration curves at temperatures of 2" to 40°C are linear and cover the range of 0.0 to 4.0 mg/liter as bromine
INITIAL COkOlTIONS: pH 7 PHOSPHATE BUFFER 3 m g . per liter NH 3 m g , p e r 1 1 ter Br3
I N I T I A L CONDITIONS: P H 5 ACETATE BUFFER 1 m g . p e r liter NH 3 mg. p e r 1 i t e r ~ r :
FINAL SMPLE p H = 7 . 3
F I N A L SANPLE pH=5
Figure 1. Percent response of DPD (A), phenol red (0), brom cresol purple (O), and phenosafranin (A) reagents to bromamines us. time at 20°C
20 0
E
4
6
8 1 0 0
2
4 6 8 1 0 0 TIME (MINUTES)
2
4
6
8 1 0
Volume 5, Number 3, March 1971 241
0.5
y
0.4
t
4 m
I
0
I
I
1 2 3 BROMINE ADDED (mg. per l i t e r )
I
I
4
Figure 2. Free bromine calibration curve for MO at varying sample temperatures ("C): 40" (0), 30" (A),20" (O),10" (A), and 2" ( 0 ) 1-cm cell, X = 505 mp
(Figure 2). The following volumes of standard MO solution generally apply in the indicated bromine range: 5 ml for the range of bromine below 1.9 mgiliter and 10 ml for the 1.8 to 4.2 mg/liter range. With 5 ml of standard MO solution, the absorbance for the blank is about 0.60. The response of MO to the halogens is the same on an equivalence basis for chlorine and bromine. The absorbance of MO solutions is slightly temperature dependent. For greatest accuracy, calibration should be performed at a temperature near that of the test samples. Stable calor standards may be prepared by dilution of the MO reagent. Ferric and nitrite ions do not interfere, and interference due t o manganic ion may be reduced by use of an arsenite modification. The low p H required for the MO test promotes other possible interferences: (a) chloramines, if present in the sample, react slowly with MO; (6) if the sample contains bromide ion, this chloramine interference increases; (c) if the sample contains a high degree of organic pollution, the results for all species may be low because of rapid reaction of the halogens with the organics at low pH. To limit these interferences, absorbance should be determined 1.5 min after the addition of the sample which is the time necessary for the complete reaction of bromine. Reagents. METHYLORANGE.(Powder, lot no. 82438, J. T. Baker Chemical Co., Phillipsburg, N.J., was used in this work. Other samples tested gave somewhat different calibration curves, but all produced satisfactory results. If insoluble particles are apparent in the 0.05 % stock solution, these should be removed by filtration prior to preparation of the standard 0.005 % MO solution.) Standard 0.005 % methyl orange solution is prepared by diluting a 0.05% stock solution and adding 1.65 g of sodium chloride per liter of the final dilute solution. This low chloride concentration is needed to "swamp" the effect of chlorides present in the samples. The MO reagent solutions are stable indefinitely. CHLOROACETIC ACID. Practical grade. ARSENITE-BUFFER. 0.125 g of sodium arsenite, 14.5 g of sodium citrate, and 0.375 g of citric acid, ground together as a dry powder. Procedures. To determine total bromine plus free chlorine, a 50-ml sample is added, with mixing, to a mixture of 5 ml 0.005 % MO and 1 ml of 91 % chloroacetic acid solution. The absorbance is determined at 505 mp, 1.5 rnin after preparation. If the absorbance is less than 0.10, the test is repeated with use 242 Environmental Science & Technology
of either 10 ml of 0.005 MO or a sample diluted with bromine and chlorine demand-free water. If manganese is present, a second determination is required. In this case, 0.10 g of the arsenite-buffer reagent is first dissolved in the 50-ml sample to reduce the residual halogen present. This is then added to the acidified MO and the absorbance is determined a t 505 mp, after 2.5 min. The difference in the apparent bromine or chlorine concentration found in the first determination and that due to manganese in this determination gives the true residual halogen concentration. To determine total bromine and total chlorine (total halogen), 0.5 ml of 2.6% sodium bromide solution is added to the sample after the first determination has been completed. The solution is mixed and, after 10 min, the absorbance is again determined. If manganese is present, the above-mentioned correction must be applied. N, N-Diethyl-p-PhenylenediamineOxalate (DPD) pH 6.3. The DPD oxalate test is suitable for the determination of total bromine plus free chlorine. If potassium iodide is added to the sample prior to addition of the sample to the reagent, total halogen can be determined. The DPD reaction is a color development reaction in the p H range 6.2 to 6.5, and requires 1.5 min for total bromine plus free chlorine, and 5 rnin for total halogen. The calibration curve is linear in the range 0.0 to 2.5 mg/ liter free bromine, but with an appropriate calibration the range may be extended to 8 mg/liter. The calibration curve shows no temperature dependence in the temperature range 2O to 40' C (Figure 3). On an equivalence basis for bromine and chlorine, the response of DPD to these halogens is the same. Nitrite ion does not interfere with the DPD test. Ferric ion interference is negligible. Interference due to manganese may be reduced by use of an arsenite modification. In the 5-min contact time before the absorbance is measured, DPD oxalate reacts nearly completely with monobromamine (prepared at pH 8.8), dibromamine (pH 7.3), and tribromamine (pH 5) solutions (Figure 1). However, if potassium iodide is not added to the sample before the mixing of the solution with the reagent and buffer, the DPD response to tribromamine is reduced to about 65 to 70 %. Therefore, in the determination of total halogen, potassium iodide crystals are dissolved in the sample prior to testing. The procedure for total bromine plus free chlorine can be used only if the p H of the sample is greater than 8 and ammonia is absent or its concentration is several times that of the
i C
1
2 3 4 BROMINE ADDED (mg. p e r liter)
Figure 3. Free bromine calibration curve for DPD 1-cm cell, X = 552 mp
5
total bromine. If these conditions are not satisfied, complete response will not be obtained. The ferrous ammonium sulfate (FAS) titration procedure (Palin, 1961) was evaluated. Chloramine was found to interfere appreciably. During the 1 min required to complete the titration, about 10% of the chloramine present reacted with the DPD,and the continuing chloramine-DPD reaction masked an already indistinct end-point. Therefore, the DPD colorimetric method is preferred. Reagents and procedures for the DPD colorimetric methods for total bromine plus free chlorine and for total halogen are listed below. Reagents. DPD OXALATE INDICATOR SOLUTION. One gram of DPD oxalate (N,N-diethyl-p-phenylenediamineoxalate, no. 7102, Eastman Kodak, Rochester, N.Y., was used in this work) is dissolved in about 100 ml of water containing 8 ml of 1 3 sulfuric and 0.2 g disodium ethylenediamine tetraacetate dihydrate (EDTA).This solution is diluted to 1 liter and stored in a n amber glass-stoppered bottle. The reagent, if stored in a refrigerator, can be used for one month. BUFFER,p H 6.4. Dibasic sodium phosphate, 24 g, and 46 g monobasic potassium phosphate are dissolved, and this solution is combined with 100 ml of a solution containing 0.8 g of EDTA. This is then diluted to 1 liter, and 20 mg of mercuric chloride is added. Procedures. To determine total bromine plus free chlorine, 100 ml of sample is added to a mixture of 5 ml DPD indicator and 5 ml buffer, and mixed thoroughly. The absorbance is determined at 552 mp in 1.5 min. To determine total halogen (total bromine plus total chlorine), 1 g of potassium iodide (crystal, reagent grade) is dissolved in 100 ml of sample which is then added, with mixing, to a mixture of 5 ml DPD indicator and 5 ml buffer. The absorbance is determined at 552 mp in 5 min. If manganese is present, an additional determination is required. Interference due to manganese is determined by adding 100 ml of sample, with mixing, to 5 ml of buffer, 1 potassium iodide crystal, and 0.5 ml of 0 . 5 z sodium arsenite. Five milliliters of DPD indicator is then added to this solution with mixing. The absorbance is determined at 552 mp after 1.5 min. The difference in apparent bromine and (or) chlorine concentration found in the original determination and that due to manganese in this determination gives the true residual halogen concentration. Phenol Red (PR), pH 5. Phenol red is a unique test for total bromine, free or combined, in concentrations up to 5 mg/liter. The bromination of phenol red in the p H range of 4.8 to 5.0 involves a change of color from yellow to reddish violet, depending upon the concentration of bromine. The calibration curve shows no temperature dependence and is linear above 1.5 mg/liter free bromine. For the lower bromineconcentrations it exhibits a decided curvature which may be due to impurities in the reagent (Figure 4). Thus far, attempts to purify the reagent have failed. For low concentrations of bromine (0.2-0.7 mglliter), accuracy is improved by using a PR reagent of 0.1 the original strength. The PR buffer has the capacity to produce the proper pH in samples with alkalinity up to 1000 mgiliter. Ferric and nitrite ions produce negligible interferences. In samples containing both bromine and manganic ion, there appears to be a negative interference due to manganese. This apparent negative interference is perhaps not a true interference, but rather, a consumption of bromine in further oxidation of the manganese added, particularly in samples having a high PH.
+
I
0.4
I
1
I
E 0.04
I
0* /* I I .-. 0.5 1 .o 00 BROMINE ADDED (mg. p e r liter)
Figure 4. Free bromine calibration curves for two concentrations of PR 1-cm cell, X = 588 rnw
The PR reagent is insensitive to free chlorine and chloramine in concentrations up to 6 mg/liter as chlorine. I n the 5-min contact time before the absorbance is measured, phenol red reacts nearly completely with mono-, di-, and tribromamine solutions (Figure 1). The presence of bromamine may be detected qualitatively in bromine solutions by adding l ml of l % sodium arsenite within 15 sec after addition of the sample to the phenol red in a second test. If free bromine is the only species present, results of both tests will be the same. If bromamine is present, the absorbanceof the second test with arsenite will be somewhat less than that without arsenite, because of immediate reduction of the unreacted bromamine. Reagents. PHENOLRED REAGENT SOLUTIONS. ( a ) 0.01 solution; 50 mg of phenol red (lot no. 764768 P-74, Fisher Scientific Co., Fair Lawn, N.J., was used in this work; other samples tested gave similar calibration curves) is dissolved in 5 ml of 0.1N sodium hydroxide and diluted to 500 ml. (b) 0.001 solution for bromine concentrations of 0.0 to 0.7 mg/liter. Seven drops of 0.1N sodium hydroxide are added to 50 ml of solution a and diluted to 500 ml. BUFFER,p H 5.0. Two hundred milliliters of 1.OM sodium acetate and 125 ml of 0.8M acetic acid. Procedures. To determine total bromine, a 50-ml sample is added to 2 ml phenol red reagent solution a or solution b and 5 ml of acetate buffer with mixing. After 5 min, the absorbance is determined at 588 mp. To detect bromamine, the above procedure is repeated, but within 15 sec after the sample is added to the reagent, 1 ml of 1 sodium arsenite is added. A lower absorbance reading is a qualitative indication of the presence of bromamine. Brom Cresol Purple (BCP), pH 9.5. Brom cresol purple reagent is suitable for determining free bromine in concentrations up to 5 mghiter. The BCP reaction is based o n the bleaching effect of bromine o n the reagent at p H 9.5. The BCP buffer has the capacity to avoid interference from alkalinity up to 1000 mg/liter. The
z
Volume 5, Number 3, March 1971 243
0.7
I
0.t
I
I
I
*A
0.5 w
u
4 z
0.4 Cn 0
9 0.3
0.2
0.1 1 2 3 4 BROMINE ADDED (mg. per liter)
0
5
Figure 5. Free bromine calibration curve for 3 ml of BCP 1-cm cell, X = 587 m p
sodium arsenite 1.0 min after the buffered sample is added to the reagent limits further development of color response during measurement of absorption. Free bromine and free chlorine together can be determined by adding 200 mg/liter of bromide ion to the sample before testing. Reagents. BROMCRESOLPURPLE,SODIUM SALT.(Brom cresol purple, sodium salt, catalog no. 342, Allied Chemical Corp., New York, N.Y., was used in this work.) 0.0125% solution (the reagent is dried at 105°C for 1 hr prior to weighing). BUFFER,p H 9.5. 0.042M borax with 6.0 ml of 5N sodium hydroxide added per liter. Procedure. To determine free bromine, a 50-ml sample is added with mixing to 10 ml of buffer solution. After thorough mixing, this solution is added to 1 or 3 ml of BCP reagent, and mixed. After 1 min, 1 ml of 1 % sodium arsenite solution is added. The absorbance is determined at 587 mp after 1.5 min. Phenosafranin (PS), pH 9.5. Phenosafranin reagent is suitable for the determination of high concentrations of free bromine. The phenosafranin reagent is bleached by bromine at p H 9.5. The calibration curve covers the range 0.0 to 10 mg/liter free bromine and does not appear to be temperature dependent in the range 2" to 40°C (Figure 7). The absorbance reading for the blank is about 0.70. Ferric ion, nitrite, and chloramine produce negligible interferences. PS shows the same response to manganic ion as does BCP.
0.05 0
1 .o 1.5 2.0 BROMINE ADDED (mg. per liter)
0.5
2.5
Figure 6. Free bromine calibration curve for 1 ml of BCP 1-cm cell, X = 587 mp
calibration curve is nonlinear below 1 mg/liter free bromine and shows no temperature dependence in the 2" to 40°C temperature range (Figure 5). Accuracy is improved for concentrations between 0.0 and 2.0 mg/liter Brz by using one-third the quantity of BCP reagent (Figure 6). In preparing the calibration curve, the absorbance reading for the blank using 3 ml of BCP is about 0.67. Stable color standards may be prepared by dilution of the BCP reagent. Interference from nitrite or chloramine is negligible. Ferric ion produces a slight but inconsistent interference. In the absence of bromine, manganic ion produces a negligible interference. However, in the presence of bromine, manganic ion produces a negative interference. A possible explanation for this apparent negative interference may be found in the discussion of the phenol red test. Free chlorine and bromamine interfere only slightly if arsenite is added 1.0 min after sample, buffer, and reagent are mixed. The BCP response to monobromamine (prepared at p H 8.8) is about 1 %, and the response to dibromamine (pH 7.3) and tribromamine (pH 5.0) is about 2 % in 1.0 min contact time (Figure 1). In the absence of bromide ion, free chlorine produces only a slight interference (about 2.5 %) in 1.0 min. The addition of 244 Environmental Science & Technology
Free chlorine produces a significant interference. Phenosafranin shows nearly the same type of response to the bromamines as does BCP (Figure 1). In the 1.0-min contact time before the absorbance is measured, interference from monobromamine (prepared at pH 8.8) is 1 %, that fromdibromamine (pH 7.3) is about 3%, and that from tribromamine (pH 5.0) is 5%. If arsenite is not added before the absorbance is measured, the bromamine interference increases with time. Free chlorine and bromamine interference may be limited by the addition of sodium arsenite 30 sec after the buffered sample is added to the reagent. Reagents. PHENOSAFRANIN. (Phenosafranin, no. 1125, Eastman Kodak, Rochester, N.Y., was used in this work.) 0.01 solution. BUFFER,p H 9.5. 0.042M borax with 6.0 ml of SN sodium hydroxide added per liter. Procedure. To determine free bromine, a 50-ml sample is added with mixing to 10 ml of buffer. After thorough mixing, this solution is added to 3 ml of 0.01 % phenosafranin reagent and mixed. After 30 sec, 1 ml of 1% sodium arsenite is added and mixed. The absorbance is determined at 520 mp after 1 min.
0.81
I
I
I
BROMIHE ADDED
(mg.
I
per liter)
Figure 7. Free bromine calibration curve for PS 1-cm cell, X = 520 mp
6LIETII'Lpi&iZzq 5 T e s t DH = 2
F-
6 , I
LL
2
l
4
l
5
I
I
6 7 8 SAMPLE DH
l
I
9
4
5
6 7 8 SAMPLE pH
9 9 4
5
6 7 SAMPLE pH
8
9
4
5
6 7 8 SAMPLE pH
9
Figure 8. Formation of bromamine by chlorination of buffered NH3-Br- solutions. Residuals determined 2 min after addition of chlorine
Figure 9. Formation of bromamine by chlorination of buffered NH3-Br- solutions. Residuals determined 2 min after addition of chlorine
The initial conditions are 4 mg/liter NH3 and 2 mg/liter Clp with either 3(0), 5(0),or25(A)mg/literBr-
The initial conditions are 0.5 mg/liter NH3 and 2 mg/Iiter CIZwith either 3 (O), 5 (O), or 25 (A) mgiliter Br-
Application oj Test Methods
test results for samples of pH 4.3 again show that bromamines are the predominant species. For samples of pH 5.6, the PR results indicate a high concentration of bromamines, apparently tribromamine at the lower ammonia level. As the p H is increased beyond 5.6, the bromamine concentration decreases. The hgo and DPD tests for total halogen show that, at pH 7.4, the concentration of total halogen remaining in the presence of 25 mg/liter Br- is considerably lower than in the presence of 3 and 5 mg/liter Br-. Phenol red test results do indicate that some bromamine is formed under these conditions; so apparently the bromamine species formed at p H 7.4 is predominately unstable dibromamine. These preliminary studies indicate that the reaction to form bromamines is highly pH dependent. In solutions of BrNH3 Clz, bromamines are formed by a two-step process. Bromide ion must be oxidized by free chlorine before free bromine and ammonia can react. The reaction between free chlorine and ammonia is rapid over the entire range of pH values tested and competes with the reaction between free chlorine and bromide ion. Results indicate that either low p H (below 7) or high concentrations of bromide ion are necessary for appreciable formations of bromamine. With a constant bromide ion concentration greater concentrations of bromamines are formed at low pH values than at high pH values. Chloramines are the primary species formed a t pH values greater than 7 and these do not react with bromide ion unless the pH is reduced to lower values.
I n preliminary studies of the reaction of chlorine with solutions containing ammonium and bromide ions, four of the five proposed colorimetric methods were used for the determinations of bromine, bromamine, chlorine, and chloramine. The effects of pH, and ammonium and bromide ion concentrations have been examined to determine their relation to the proportion of chloramine and bromamine resulting. A buffered solution containing ammonium chloride and sodium bromide was prepared and chlorine was then added with rapid mixing. The p H values ranged from 4 to 9. Two ammonia concentrations, 0.5 mgiliter and 4.0 mg/liter, were used. The bromide ion concentrations tested were 3, 5, and 25 mg/liter. The free chlorine addition was constant at 2.0 mg/ liter. All samples were prepared at room temperature, 26OC. Two minutes after the addition of chlorine, portions of each sample were removed and the following series of tests were run as nearly simultaneously as possible (Figures 8 and 9). The brom cresol purple reagent was used to determine free bromine residuals; the phenol red reagent was used to determine total bromine exclusive of chlorine; and methyl orange reagent with bromide added to the final mixture, and DPD oxalate with prior addition of potassium iodide to the sample, were used to determine total halogen (total bromine and total chlorine). For chlorinated samples containing 4.0 mg/liter N H 3 (Figure 8) and 3, 5 , and 25 mg/liter Br-, the BCP test results show that negligible quantities of free bromine were formed. Any apparent response is probably due to interference from bromamines in the 1 min of contact prior to the addition of arsenite. Fcr these same samples, the phenol red test results indicate that bromamines are the primary species formed at a pH value of 4.3; results of the h i 0 and DPD tests for total halogen at this pH agree with the PR results. As the sample pH is increased beyond p H 4.3, the PR results indicate that the concentrations of bromamine decrease and the rate of decrease is greater for decreasing bromide levels. For these chlorinated samples containing 4 mglliter NH3, the concentrations of bromide ion and the sample pH have little effect on the final concentration of total halogen, as shown by the MO and DPD results, which were in good agreement. For chlorinated samples containing 0.5 mg/liter N H 3 (Figure 91, the BCP test again indicates that only negligible quantities of (apparent) free bromine were formed. The PR, MO, and
DPD
+
+
Conclusions The major objective of this study has been the development of suitable analytical methods for the determination of free bromine and of bromamines in water. Ideally, the methods should be capable of differentiating between free bromine, free chlorine, bromamines, and chloramines. The methods should be suitable for use in polluted waters and free of interference from the substances normally found in such waters. In addition, the methods should be suitable for field use. Free bromine, free chlorine, bromamines, and chloramines may be distinguished from one another. Free bromine alone may be determined by the brom cresol purple or phenosafranin tests. PS is the less sensitive of the two tests. Phenol red is a unique test for total bromine. The DPD and MO tests may be used to determine total bromine plus free chlorine. The DPD test serves well to indicate total halogen, if potassium iodide Volume 5, Number 3, March 1971 245
Acknowledgment
Literature Cited American Public Health Association, “Standard Methods for the Examination of Water and Wastewater,” 12th ed., New York, 1965, pp 66 and 67. Anderson, T., Madsen, H. E. L., Anal. Chem. 37, 49 (1965). Galal-Gorchev, H., Morris, J. C., Division of Water and Waste Chemistry, 142nd Meeting, ACS, Atlantic City, N.J., September 1962. Hashmi, M. H., Ayaz, A. A., Anal. Chem. 35,908 (1963). Johannesson, J. K., Analyst 83,155 (1958). Palin, A. T., Water Sew. Works 108,461 (1961). Stenger, V. A,, Kolthoff, I. M., J. Amer. Chem. Soc. 57, 831 (1935).
This research was sponsored by the U.S. Army Medical Research and Development Command, under contract no. DA-49-193-MD-2909.
Received for review October 13, 1969. Accepted December 4 , 1970. Presented at the Division of Water, Air, and Waste Chemistry, 157th Meeting, ACS, Minneapolis, Minn., April 1969.
is added to the sample. The MO test is also suitable for the determination of total halogen, in the presence of bromide ion, It can be concluded also, from studies of the reaction of chlorine with solutions containing bromide ion and ammonia that, to form bromamines in a water supply by addition of chlorine and bromide ion, these would have to be added to a small volume of the water at low p H to form free bromine, and subsequently this solution would have to be mixed with the remainder of the water to be treated.
Natural Synthesis of Ozone in the Troposphere Lyman A. Ripperton, Harvey Jeffries, and James J. B. Worth’ Department of Environmental Sciences and Engineering, University of North Carolina, Chapel Hill, N.C. 27514
I Observational
and experimental data suggest or demonstrate that some of the nonurban, tropospheric ozone is synthesized in situ rather than being transported from the stratosphere. Experiments showed that ozone can be synthesized from naturally occurring precursors at concentrations comparable to those found in nature. Two precursor systems which produced ozone were NO2 a-pinene hv + and CH20 a-pinene hv e.This work provides chemical evidence for the theory proposed by Frenkiel and Paetzold, on the basis of diffusion theory, that there is a “tropospheric source” of ozone.
+
+
+
+
I
t is well known that ozone (03)is synthesized in the polluted air of cities (Leighton, 1961; Wayne, 1962) and in the stratosphere (Junge, 1963). It has not been established that O3 is synthesized in the natural troposphere. There are, however, scattered references to this possibility. In 1941, Regener (Junge, 1963) proposed the theory that O3observed in the troposphere originated in the stratosphere. On the basis of diffusion theory, however, Frenkiel(l960) and Paetzold (19til), concluded that not all tropospheric O3has a stratospheric origin. McKee (1961) measured O3in Greenland and concluded that some of it was probably of local photochemical origin. Although the available evidence suggests that most tropospheric O3 (outside the polluted areas) originates in the stratosphere, there is evidence of tropospheric synthesis. In seven cases of 250 (Hering and Borden, 1967), 0 3 concentrations were greater at or below 2.5 km than at the tropopause. In 44 of 251 cases, the O3 concentration at altitudes below 5 km was greater than at 5 km. Layers of relatively high tropospheric O3concentration near the tropopause can be caused by
Research Triangle Institute, Research Triangle Park, N.C. 27705. 246 Environmental Science & Technology
stratospheric intrusion as suggested by Kroening and Nye (1962). High values near the ground, however, are more easily explained by in situ synthesis than by subsidence, through half the troposphere, of a layer of high 0 3 content. Went (1960) suggested a mechanism for tropospheric synthesis similar to that by which 0 3 is thought to be formed in polluted air, NOz olefin hv +, with “terpenoid” compounds replacing the simpler olefins of polluted air. aStephens (1962) experimented with the system NO2 pinene hv +, and observed the formation of PAN (which would register as O 3 on a KI method of analysis), but he did not determine O 3itself. Rasmussen and Went (1965) found an average of 1.0 pphm (maximum 5.0 pphm) of atmospheric terpenoid compounds in the Appalachians and the Ozarks. Lodge and Pate (1963) found an average of 0.09 pphm (maximum 0.50) of atmospheric NO2 in Panama, and Worth et a/. (1967) measured an average of 0.40 pphm (maximum 2.6 pphm) in the southern Appalachians. The experiments discussed below were part of a long-term study of the behavior of O 3 in the lower troposphere. Field data were examined for evidence of natural tropospheric synthesis of 03,and experiments were designed specifically to examine the possibility of natural tropospheric O3 synthesis. Other experiments were designed to explore mechanisms by which natural O3synthesis occurs in the troposphere.
+
+
+
+
Procedure
Synthesis in Untreated Natural Air, In the late summer of 1964 and 1965, continuous oxidant readings were made simultaneously with Mast Ozone Meters on Green Knob, N.C., and in Little Lost Cove, an adjacent valley. At a site in rural piedmont North Carolina, 0 3 has been measured with Mast Meters at 1.2, 9.2, 18.2, and 36.6 m on a tower on the campus of the Research Triangle Institute. Some data at 9.2 and 36.6 m have been obtained with Regener chemiluminescent ozone meters. At the piedmont tower site, air was blown, untreated, into a large Mylar bag with a Gelman “hurricane blower” and
