(8) Denny, R. C. “Named Organic Reactions”; Plenum Press: New York, 1969; p 143. (9) Yurawecz,-M. P. J . Assoc. Off. Anal. Chem. 1979,62,36. ( 10) Saeger, V. W.; Thompson, Q. E. .Fnciron.,+Sci. Technol. 1980, 14,705. (11) Neely, €3. Enuiron. Sci. Technol. 1979, 13, 1506.
(12) Veith, G.;Detoe, D. L.; Bergstedt, V. B. J . Fish. Res. Hoard Can. 1979,36,1040.
Received for review February 2, 1981. Revised Manuscript Receiucid J u n e 15, 1981. Accepted July 2 , 1981. W e extend our thanks and appreciation to the U.S. Environmental Protection Agency for supporting this research under grants H-806350 and H-808865.
NOTES
Sodium Alteration of Chlorine Equilibria. Quantitative Description Charles N. Haast Department of Chemical Engineering and Environmental Engineering, Rensselaer Polytechnic Institute, Troy, New York 12181
Previous data on sodium alterations of hypochlorite equilibria have been reanalyzed under the assumption of existence of the ion pair NaOCl. Computed ion-pair dissociation constants range between 0.9 and 2.3 M.
Introduction Investigations on the efficiency of microbial inactivation by chlorine have indicated that high salt concentrations can increase rates of inactivation (1-4). In particular, the observed enhancement of inactivation by sodium chloride has been observed to be more significant a t p H values above the pK, for dissociation of hypochlorous acid. Jensen e t al. (3)have hypothesized that this observed salt effect is due to formation of ion pairs involving OC1-. In earlier work, Sugam and Helz ( 5 ) have demonstrated the alteration of the apparent ~ K for A hypochlorous acid dissociation in the presence of high concentrations of sodium chloride and developed an empirical Correlation for this effect. T o my knowledge, no information has yet been presented describing the formation constant for a putative sodium hypochlorite ion pair. In this communication, the data of Sugam and Helz ( 5 )and Jensen et al. (3)will be reanalyzed under the assumption of the existence of the species NaOCl per se, and a n apparent dissociation constant for this complex will be calculated. Theory Since the salt effect inherently operates under conditions of nonnegligible ionic strength, the explicit consideration of activity coefficients is necessary. T h e HOCl dissociation equilibrium constant may be defined by eq 1. KA= r*.IH+l[OCl-]/yo[HOCl]
(1)
Square brackets are used to denote molar concentrations while braces are used to denote activity. T h e variables yi and yo are the activity coefficients for univalent ions and for uncharged HOC1, respectively. On a similar basis, a dissociation constant for the complex NaOCl may be written as eq 2 if it is assumed that the activity coefficient for this complex is identical with that of HOCl. K D = yi2[Na+] [OC1-]/yo[NaOC1] Sugam and Helz provide data for a parameter, K A * * ,defined on the basis of HOCl and “total hypochlorite” concentrations. Present address: Pritzker Department of Environmental Engineering, Illinois Institute of Technology, Chicago, IL 60616. 0013-936X/81/0915-1243$01.25/0
@ 1981 American Chemical Society
If the latter term is interpreted as the sum of OC1- and NaOCl concentrations, it can be shown, by utilization of 1 and 2 that eq 3 must hold K A * * ~ * /= ~ oK A + K~Kny+’[Na+]/yo
(3)
Jensen et al. ( 3 )present data on the fraction of hypochlorite involved in ion-pairing as determined spectrophotometrically. If this fraction is defined as x and if the concentration of HOCl is negligible (Le., if the p H is much greater than ~ K A then ), it can also be shown that eq 4 must hold.
KD = yi2[Na+](1 - x)/yox
(4)
Analysis
Sugam and Helz ( 5 )conducted experiments to determine values of KA** a t various temperatures and ionic strengths. Ionic strengths were adjusted by using a synthetic sea-salt mixture, containing the cations sodium, magnesium, and calcium, with sodium being the predominant cation. According to eq 3, these data may be used to determine values for K A and Kr, providing that activity coefficients are known. T h e activity coefficient for monovalent ions a t low ionic strengths (below 0.2 M) was calculated by using the Davies correlation (6).At higher ionic strengths, the mean salt assumption was used in this calculation, and data of Miller0 (7) and Whitfield (8) for the activity coefficients of NaCl and KCl, respectively, were used. Additionally, as did Sugam and Helz ( 5 ) ,I assumed that the activity coefficient of the hypochlorite anion was identical with that of KCI. T h e activity coefficient for HOCl (and, it is assumed, NaOCl) was obtained from interpolation of data of Weiss (9);this latter correction was always less than 6%. For each temperature range (11.5 f 1, 18.3 f 0.5, 25.2,34.5, and 39.5 “C) values for the two equilibrium constants were calculated. At 25.2 “C, the calculation proceeded by least squares; a t all other temperatures constants were determined algebraically. Values for the equilibrium constants thus calculated are shown in Table I. With regard to the data of Jensen e t al. ( 3 ) ,in which a Table 1. Equilibrium Constants from Data ofsugam and Helz temp, ‘ C
11.5 18.3 25.2 34.5 39.5
KA, M
2.128 X 2.638X 2.999X 3.992 X 4.138 X
loM8
PKA
KD, M
PKD
7.667 7.567 7.520 7.389 7.371
0.705 0.863 0.541 0.939 0.637
0.152 0.064 0.267 0.207 0.196
Volume 15,Number 10,October 1981 1243
Table II. Data of Jensen et al. NaCi added, M
total sodium, M
ionic strength, M
X
0.05
0.0656 0.1156 0.5156
0.076 0.126 0.526
0.014 0.049
0.10 0.50
0.077
spectrophotometric determination of x was made a t varying ionic strengths and sodium concentrations, data are shown in Table 11. It is unclear a t what temperature these assays were conducted. From the data in Table 11, activity coefficients computed as described above, and eq 4, an average value for K Dof 2.283 M is obtained. The average value for ~ K isD-0.33. The computed HOCl dissociation constants (KA)obtained from analysis of data of Sugam and Helz (Table I) may be compared directly with constants obtained from Morris ( I O ) , who reported that eq 5 described ~ K in A the temperature range 5-35 “C to an accuracy of 0.01 unit:
~ K =A3000.O/T(K) - 10.0686
+ 0.0253T(K)
(5)
The mean deviation between the values for ~ K inATable I1 and those a t comparable temperatures calculated via eq 5 is 0.024, with the former values averaging less than the latter values. Sugam and Helz ( 5 )indicate that their experimental values are precise to 0.02 units. The apparent deviation from the results of Morris (10)are within the sum of the reported experimental errors of the two investigators. This difference is also approximately the same as the correction between operationally defined hydrogen-ion activity and true thermodynamic activity as described by Sugam and Helz ( 5 ) . However, the correspondence between the PKA values in Table I with those calculated from eq 5 suggests that the NaOCl ion-pairing phenomenon may be satisfactorily described by eq 2. Inspection of Table I does not reveal a consistent dependency of K D upon temperature. If the values for this constant are assumed to be relatively independent of temperature, then it becomes possible to compare values for K,, as calculated from the data of Sugam and Helz ( 5 ) with those calculated D from the data of Jensen et al. ( 3 ) .Both K D and ~ K values show different variances between the data set of Sugam and Helz and that of Jensen et al. I t is therefore necessary to employ a nonparametric comparison. Using the Mann-Whitney U statistic ( I I ) , one finds that the two data sets are different a t the 5% confidence level ( U = 15, p = 0.05). The dissociation-constant values for Sugam and Helz are lower than those of Jensen et al. There are a t least several possible reasons for this discrepancy: (1)Both Sugam and Helz ( 5 )and Jensen e t al. ( 3 ) employed mixed salts in adjusting ionic strength. The
1244
Environmental Science & Technology
former study also employed calcium and magnesium in addition to sodium. In contrast, the medium employed by Jensen et al. ( 3 )contained the ca$ions sodium and potassium exclusively, the latter oFjvhich niay be a t least as effective as sodium a t complexing hypochlorite. (2) Jensen et al. ( 3 ) ,in using spectrophotometry to determine the fraction of complexed hypochlorite assumed that the species NaOCl had a negligible extinction coefficient in comparison to OC1-. This appears to be an unsubstantiated assumption. In spite of the statistical significance of the difference in K,, between the two data sets, both data sets are consistent with the existence of NaOCl as an ion pair. Taken together, the two studies permit an equilibrium constant: as defined by eq 2, to be estimated as between 0.5 and 2.8 M, or a pK,, between -0.45 and +0.30. In conjunction with experimental evidence that this complex may be more virucidal than OC1- (1-3),and perhaps as bactericidal as HOCl ( 4 ) ,it would no longer appear valid to neglect hypochlorite ion-pair formation in disinfection research investigations. Furthermore, a more precise evaluation of K,, is needed to permit a more detailed analysis of the environmental significance of this species Summary
Data of Sugam and Helz ( 5 )and Jensen et al. ( 3 )have been analyzed under the assumption that the ion pair NaOCl exists. The existence of such a species with a dissociation constant between 0.9 and 2.3 M is consistent with both sets of data, although different values are obtained for the two data sets. Further studies are suggested to define the precise value of this constant. L i t e r a t u r e Cited (1) Scarpino, P . V.; Berg, G.; Chang, S. L.; Dahling, D.; Lucas, M. Water Res. 1972,6,959. (2) Engelbrecht, R. S.; Weber, M. J.; Salter, B. L.; Schmidt, C. A. A p p l . Enuiron. Microbiol. 1980,40, 249. ( 3 ) Jensen. H.; Thomas, K.: Sharp. D. G. A.D. DEnuiron. ~. Microbiol. 1980,40,633. (4) Haas. C. N.: Zapkin, M. A., presented at the 81st Annual Meeting of the American-Society for Microbiology, 1981. ( 5 ) Sugam, R.; Helz, G. R. Enuiron. Sci. Technol. 1976,10, 384. 16) Butler. J. N. “Ionic Eauilihrium: A Mathematical Amroach”: _. Addison-Wesley: New York, 1964. ( 7 ) Millero, F. J. In ‘‘Activity Coefficients in Electrolyte Solutions”; Pytkowicz, R. M., Ed.; CRC Press: Boca Raton, FL, 1979: Vol. 2, Chapter 2. (8) Whitfield, M. In “Activity Coefficients in Electrolyte Solutions”; Pytkowicz, R. M., Ed.; CRC Press: Boca Raton, FL, 1979; Vol. 2, Chapter 3. (9) Weiss, 3. J. Z. Anorg. Allg. Chem. 1930,192,97. (10) Morris, J. C. J . Phys. Chem. 1966, 70, 3798. (11) Sokal, R. R.; Rohlf, F. J . “Introduction to Biostatistics”; W. H. Freeman: San Francisco, CA, 1973. Receiued for reuiew January 21,1981. Accepted J u n e 22,1981
