Relative Bactericidal Effectiveness of Hypochlorous Acid and ...

SpeciesHOClClO –HClCy –Cl 2Cy –l maxnm235290215220Table 1. Absorption Maxima forMain Chlorine Specieswell below 290 nm, they are relatively stable to sunlight.Studies with chloroisocyanurate solutions (CA =26 ppm and av. Cl = 14 ppm) showed maximumstability at pH 6.8 which corresponds to minimalhydrolysis (O’Brien et al 1975). Stability decreases aspH is increased due to increased ionization of HOClto ClO – .Shortly after the commercialization ofchloroisocyanurates, it was shown that cyanuric aciddecreases the kill rate of bacteria such as S. faecalisand S. aureus by available chlorine in distilled water(Andersen 1965). Later studies confirmed these findings(Stewart and Ortenzio 1964; Fitzgerald andDerVartanian 1967; Swatek et al 1967). However, noattempts were made to correlate the observed killtimes with the concentration of biocidal species.In the absence of cyanuric acid, the rate of disinfectionby free av. Cl decreases with increased pH inaccordance with the ionization pattern of HOCl, demonstratingthat HOCl is the active biocidal agent andthat hypochlorite ion is a poor disinfectant (Fair et al1948). Disinfection by HOCl is believed to be due todiffusion through the cell wall of bacteria followed byoxidative attack resulting in disruption of metabolicprocesses (Green and Stumpf 1946). The poor bactericidalproperties of hypochlorite ion are attributed toslower diffusion through the cell membrane comparedto the neutral molecule HOCl.DISCUSSIONChloroisocyanurate Equilibria – The variouscyanurate equilibria in the cyanuric acid – availablechlorine system are shown in Figure 2. Thecomplex equilibrium system consists of six ionizationand six hydrolysis reactions, shown horizontally andvertically, respectively. Table 2 lists the reportedvalues of the equilibrium constants. The equilibriumconstants for the first 12 reactions were measured at24–25°C (O’Brien et al 1975). Other values of some ofthe chloroisocyanurate equilibrium constants havealso been reported (Brady et al 1963; Gardner 1973;Pinsky and Hu 1981). The equilibrium constant forionization of hypochlorous acid was determined overthe 4 to 34 °C temperature range (Morris 1966).Mathematical Model and Computer Program– A computer program (similar to O'Brien et al1975) was employed for calculation of the concentrationof all species in the cyanuric acid – availablechlorine system. Material balance yields the followingequations for total free av. Cl (Cl T) and total cyanurate(Cy T):ReactionpK NumberpKCl 3Cy + H 2O HCl 2Cy + HOCl 1 1.8 ± 0.2HCl 2Cy H + + Cl 2Cy – 2 3.75 ± 0.03HCl 2Cy + H 2O H 2ClCy + HOCl 3 2.93 ± 0.07H 2ClCy H + + HClCy – 4 5.33 ± 0.05H 2ClCy + H 2O H 3Cy + HOCl 5 4.07 ± 0.08H 3Cy H + + H 2Cy – 6 6.88 ± 0.04Cl 2Cy – + H 2O HClCy – + HOCl 7 4.51 ± 0.09HClCy – H + + ClCy – 8 10.12 ± 0.02HClCy – + H 2O H 2Cy – + HOCl 9 5.62 ± 0.05H 2Cy – H + + HCy – 10 11.40 ± 0.10ClCy 2– + H 2O HCy 2– + HOCl 11 6.90 ± 0.11HCy 2– H + + Cy 3– 12 13.5HOCl H + + ClO – 13 7.54Table 2. Equilibrium Constants for Reactions in theCyanuric Acid – Available Chlorine System @ ˜25ºC106 The Chemistry and Treatment of Swimming Pool and Spa Water

Hydrolysis àCl 3CyWhere: m and the sum of n + m vary from 0 to 3 andb nmis the reciprocal of the product of the appropriatedichloroisocyanuric acids are essentially completely[H nCl mCy] = [H + ] n+m [HOCl] m [Cy 3– ]b nmionized at pool pH to HClCy – and Cl 2Cy – .K 1HCl 2CyK 3H 2ClCyK 5K 2K 4Cl 2Cy –K 7HClCy –K 9K 8ClCy 2–K 11equilibrium constants. Expressions for calculatingthe concentration of cyanurate and chloroisocyanuratespecies are listed in Table 3.The following working equation was employed:Cl T= Cy TN/D + [HOCl](1 + K 13/[H + ])Where: N is a 6–term summation representing theH 3CyK 6K 10K 12 various chloroisocyanurates and D is a 10–term summationrepresenting all possible cyanurates. SinceH 2Cy – HCy 2– Cy 3–Dissociation à[Cy 3– ] occurs as a factor in all terms in both theFigure 2 – Chloroisocyanurate numerator (N) and denominator (D), it cancels out.The program utilizes a Newton–Raphson convergenceEquilibriatechnique. For a given reservoir chlorine (Cl T), totalcyanurate (Cy T), hydrogen ion concentration, and estimatedCl T= [H 2ClCy] + [HClCy – ] + [ClCy 2– ] + 2[HCl 2Cy]+ 2[Cl 2Cy – ] + 3[Cl 3Cy] + [HOCl] + [ClO – ]HOCl concentration, the program iteratesthe HOCl concentration until a solution is obtained,i.e., when the calculated values of Cl Tand Cy Tareequal to the actual values. While the concentration oftricyanurate ion [CyCy T= [H 3Cy] + [H 2Cy – ] + [HCy 2– ] + [Cy 3– ] + [H 2ClCy]] is not known explicitly, it isknown implicitly, i.e., once the correct HOCl is obtained,the concentration of tricyanurate ion can be+ [HClCy – ] + [ClCy 2– ] + [HCl 2Cy] + [Cl 2Cy – ]+ [Cl 3Cy]calculated by the following equation: [Cy 3– ] = Cy T/D.While chloroisocyanurates analyze as free av. Cl Species Distribution – A plot of species distributionagainst pH for water containing ~25 ppmwith swimming pool test kits because of rapid hydrolysis,they are not free in the sense of HOCl and ClO – . cyanuric acid and ~1 ppm av. Cl is shown in Figure 3.More accurately, they represent potential free av. Cl, It is seen that at pH 7.5, typical in swimming pools, thereleasing HOCl and ClO – on demand.species distribution varies as follows:The concentration of the various cyanurate andchloroisocyanurate species can be calculated fromapplicable equilibrium constants and the concentrationsof H + (calculable from the pH), HOCl, and[H 2Cy – ] > [H 3Cy] > [HClCy – ] >> [HOCl] = [ClO – ]@ [Cl 2Cy – ] > [ClCy 2– ]tricyanurate ion by the following generalized equation:By contrast to cyanuric acid, monochloro– and[HCy 2– ] = [H + ] [Cy 3– ]/K 12[H 2Cy – ] = [H + ] 2 [Cy 3– ]/K 12K 10[H 3Cy] = [H + ] 3 [Cy 3– ]/K 12K 10K 6[ClCy 2– ] = [H + ][HOCl][Cy 3– ]/K 12K 11[HClCy – ] = [H + ] 2 [HOCl][Cy 3– ]/K 12K 11K 8= [H + ] 2 [HOCl][Cy 3– ]/K 12K 10K 9[H 2ClCy] = [H + ] 3 [HOCl][Cy 3– ]/K 12K 11K 8K 4= [H + ] 3 [HOCl][Cy 3– ]/K 12K 10K 6K 5[Cl 2Cy – ] = [H + ] 2 [HOCl] 2 [Cy 3– ]/K 12K 11K 8K 7= [H + ] 2 [HOCl] 2 [Cy 3– ]/K 12K 10K 9K 7[HCl 2Cy] = [H + ] 3 [HOCl] 2 [Cy 3– ]/K 12K 11K 8K 7K 2= [H + ] 3 [HOCl] 2 [Cy 3– ]/K 12K 10K 6K 5K 3[Cl 3Cy] = [H + ] 3 [HOCl] 3 [Cy 3– ]/K 12K 11K 8K 7K 2K 1= [H + ] 3 [HOCl] 3 [Cy 3– ]/K 12K 10K 6K 5K 3K 1Table 3 – Expressions for Calculating Concentrations ofIsocyanurate and Chloroisocyanurate SpeciesJohn A. Wojtowicz – Chapter 6.1 107

Figure 3 – SpeciesDisribution in theCyanuric Acid-Available ChlorineSystem at 25°CCA ~25 ppm,av. Cl ~1 ppmFigure 4 – 99% KillTime at 20°C v.s.HOCl Concentrationat 25°C (Organism:S. Faecalis)The presence of HOCl, in concentrations approximatingthe calculated values, was verified byroom temperature vacuum distillation of dilutechloroisocyanurate solutions at pH 7.5. The HOCl–enriched distillate was analyzed by infrared spectroscopyand by conventional av. Cl analysis. Calculationsof HOCl concentrations in the original solutions tookinto account the appropriate enrichment factors.Disinfection Data – Disinfection data (99% killtime) of Andersen (Andersen 1965) for Streptococcusfaecalis at 20°C and pH 7 and 9 for various cyanuricacid and av. Cl concentrations are listed in Table 4along with the calculated concentration of hypochlorousacid (at 25°C). The HOCl concentrations werecalculated at 25°C because the complete set of equilibriumconstants for the isocyanurates are known onlyat this temperature. The data show that the 99% kill108 The Chemistry and Treatment of Swimming Pool and Spa Water

pHCA ppmAv. Cl ppm99% Kill Time(min) at 20°CHOCl (ppm asav. Cl) at 25°C777777777799999999999902550100255010025501000255010002550100025501000.230.250.240.240.480.480.480.980.970.970.230.240.240.240.470.480.480.470.960.980.970.950.37.211.521.73.24.710.21.62.44.13.515.529.555.31.67.712.120.40.93.35.59.40.1780.005250.002540.001270.01020.005110.002550.02140.01040.005200.007800.002090.00120.0006440.01600.004220.02410.001270.03260.008760.004920.00258Table 4 – Disinfection Data of Andersen and Calculated Hypochlorous AcidConcentrations (Organism: S. Faecalis)time: a) increased with increasing cyanuric acid concentrationat constant av. Cl, b) decreased with increasingav. Cl concentration at constant cyanuricacid concentration, c) increased with increasing pH atconstant cyanuric acid and av. Cl concentrations, andd) increased with increasing pH whether cyanuricacid was present or not. The relative importance ofHOCl and the chloroisocyanurates in disinfection canbe evaluated statistically.Statistical Analysis – A plot of 99% kill time at20°C against the calculated HOCl concentration at25°C is shown in Figure 4. The plot shows very goodcorrelation of kill times with calculated HOCl concentrations.The fact that data for cyanurate and cyanurate–freemedia fall on the same line is probablyfortuitous. They should, in fact, most likely fall onseparate lines as discussed below. However, this willnot affect statistical analysis since the cyanurate datacan be analyzed separately.Extrapolation of the calculated HOCl concentrationsto 20°C should cause the two sets of data todiverge and fall on separate parallel lines because thetemperature dependence of the HOCl concentration isdifferent for stabilized and unstabilized chlorine. Theonly parameter that changes in transforming thecalculated HOCl concentrations from 25 to 20°C is theintercept; the slope should remain the same. Forunstabilized chlorine, the HOCl concentration willincrease from 25 to 20°C due to decreased ionizationof HOCl. By contrast, the HOCl concentration instabilized chlorine solutions decreases with decreasingtemperature due to decreased hydrolysis ofchloroisocyanurates. The extent of separation of thetwo lines will depend on the temperature dependenceof ionization of HOCl and hydrolysis ofchloroisocyanurates. Thus, for the lines to be coincidentat 20°C, the calculated HOCl concentrations at25°C for the isocyanurate data would have to be a littlehigher.John A. Wojtowicz – Chapter 6.1 109

Regression analysis (Table 5) shows that the killtime correlates excellently with the free HOCl concentration(r 2 = 0.98). By contrast, correlations againstthe concentrations of the various chloroisocyanurateswere poor. The lone exception was dichloroisocyanurateion (Cl 2Cy – ) which showed a high correlation (r 2 =0.95). However, this is to be expected since the concentrationof Cl 2Cy – correlates with the concentration ofHOCl (see Figure 3). Indeed, any significant bactericidalactivity for Cl 2Cy 2– is very unlikely given the poorstatistical correlations found for the four otherchloroisosyanurates. The following equation was obtainedwhich shows that kill time at 20°C variesinversely with the calculated HOCl concentration at25°C.t 0.99= (0.0268q)[HOCl] –1.014Where: q is the relative disinfection time at 25°Cversus that at 20°C. This equation conforms to theChick–Watson Law: C n t = k, where C is the disinfectantconcentration, t the kill time, and n and k areconstants. The fact that n = 1 means that the concentrationof HOCl and time are equaly important indetermining the inactivation rate of bacteria.Effect of Temperature on Kill Time – Therate of disinfection increases with increasing temperature.Disinfection data can be extrapolated toother temperatures using the following equation:(~7.3 kcal/mol) for disinfection by hypochlorous acid(Fair et al 1948). This value represents the combinedtemperature dependance of HOCl ionization and inactivationof bacteria. Although the concentration ofHOCl decreases with increasing temperature at constantpH due to increased ionization, the disinfectionrate in cyanuric acid–free media actually increasesdue to greater disinfection efficiency. By contrast, in acyanuric acid–available chlorine system, increasedtemperature can lead to an increase in the HOClconcentration due to increased hydrolysis ofchloroisocyanurates, resulting in an additional increasein the disinfection rate.Although the activation energy for disinfectionin a cyanurate system has not been determined, it canbe estimated. The value of 30.5 kJ/mol (7.3 kcal/mol)for a cyanurate–free system can be adjusted to that ina cyanurate system by combining it with the appropriateenergy term applicable to a cyanurate system. Theprinciple chloroisocyanurate in pool water ismonochloroisocyanurate ion. It hydrolyzes to a smallextent at pool pH, providing increased HOCl concentrationsas temperature increases. A value of 99.2 kJ/mol/mol (23.7 kcal/mol) has been reported for theenergy (i.e., heat of hydrolysis) that determines thetemperature dependence of its hydrolysis (Wojtowiczto be published). Therefore, the overall estimatedactivation energy for disinfection in a cyanurate systemis: 30.5 + 99.2 = ~129.7 kJ/mol (31.0 kcal/mol). At25°C, this will result in a decrease in kill time by afactor (q) of 2.44 compared to that at 20°C.t 2/t 1= q = exp[(–DE/R)(1/T 2–1/T 1)]Where: t 1and t 2are the 99% kill times at temperaturesT 1and T 2(in kelvins), R is the gas constant (8.314 J/deg/mol; 1.987 cal/deg/mol), and DE is the disinfectionactivation energy. For enteric bacteria (such as E. coli)in a cyanuric acid–free system, DE = ~30.5 kJ/molSpeciesHOClClO –H 2ClCyHClCy –ClCy 2–HCl 2CyCl 2Cy –Coefficient ofDetermination r 2 *0.980.070.370.590.080.600.95*r 2 is the fraction of the variability in the data explained bythe regression equation.Table 5. Regression Analysisof Disinfection DataExtrapolation of Andersen’s Data to SwimmingPools – Since hypochlorous acid andchloroisocyanurates cannot be distinguished by testkit analysis (which is typically based on diethyl–p–phenylenediamine), the total free av. Cl concentrationis the only indicator of adequate disinfection. However,because disinfection is affected by the presenceof cyanuric acid, a better indicator is the ratio of totalcyanurate to av. Cl since it correlates perfectly withthe HOCl concentration at pH 7 and 20°C in Andersen’sstudy:HOCl (pH 7 & 20°C) = 0.576(Cy T/Cl T) –1.008 , r 2 = 1.00This equation shows that the HOCl concentrationvaries inversely with the ratio of total cyanurate tototal free av. Cl.Regression analysis shows that Andersen’s datafor 99% kill time (t 0.99) at pH 7 and 20°C varies linearlywith the ratio of total cyanurate (Cy T) to total freeavailable chlorine (Cl T) as shown by the followingequation:t 0.99(pH 7 and 20°C) = 0.119 + 0.0516Cy T/Cl T, r 2 = 0.98110 The Chemistry and Treatment of Swimming Pool and Spa Water

In order to apply Andersen’s data to swimmingpool conditions, the data must be adjusted for theeffect not only of temperature but pH as well. Since killtime varies inversely with pH (which influences theHOCl concentration), the 99% kill time at pH 7 and20°C was first adjusted to pH 7.5 (typical in swimmingpools) by multiplying by the factor 1.40, which representsthe ratio of HOCl at pH 7 to that at pH 7.5 instabilized water. Cyanuric acid moderates the effect ofpH on the HOCl concentration since the factor inunstabilized water is 1.48.t 0.99(pH 7.5 and 20°C) = [0.167 + 0.0722Cy T/Cl T]/qThe factor q allows extrapolation of kill timesfrom 20°C to other temperatures. The higher thetemperature, the higher the value of q and the lowerthe kill time. At a swimming pool temperature of 80°F(26.7°C), q = 3.29 which corresponds to 99% kill timesof 0.6 – 0.2 minutes over the recommended free av. Clrange of 1 – 3 ppm at 25 ppm cyanuric for entericbacteria such as S. faecalis. Higher cyanuric acidconcentrations will cause the 99% kill times to increase.By contrast, higher temperatures, especiallyin spas, will decrease kill times.Effect of Swimming Pool Contaminants onDisinfection Rate – Longer kill times were observedin pool water, compared to distilled water, whethercyanuric acid was present or not (Swatek et al 1967).This was attributed to unknown variables, which inreality are swimming pool contaminants. In swimmingpools, nitrogenous compounds such as ammonia,amino acids, and creatinine, introduced by bathers,can react with free av. Cl. This forms bactericidallyineffective combined chlorine, resulting in increasedkill times. Combined chlorine compounds do not hydrolyzeto a significant extent to HOCl. Thus, it isimportant to maintain a low combined available chlorinelevel by frequent superchlorination or shock treatmentin order to oxidize chlorinated compounds suchas monochloramine, N–chloroamino acids, and N–chlorocreatinine, which are much poorer disinfectantsthan HOCl.Addition of ammonia to stabilized aqueous av. Clresults in increased kill times proportional to theamount added (Fitzgerald and DerVartanian 1967).By contrast to ammonia, urea, a major swimming poolcontaminant, does not affect disinfection in short termexperiments (Fitzgerald and DerVartanian 1967).Although urea is a potential source of ammonia, itshydrolysis is very slow under swimming pool conditions.However, since urea serves as a nutrient forbacteria and algae, it is necessary to oxidize it byperiodic shock treatment.ReferencesAndersen, J. R. “A Study of the Influence of CyanuricAcid on the Bactericidal Effectiveness of Chlorine.”American Journal of Public Health 1965, 55, 1629–37.Brady, A. P., K. M. Sancier, G. J. Sirine. “Equilibria inSolutions of Cyanuric Acid and its ChlorinatedDerivitives.” Journal of the American ChemicalSociety 1963, 85, 3101–4.Fair, J. M., J. C. Morris, S. L. Chang, I. Weil, R. P.Burden. “The Behavior of Chlorine as a WaterDisinfectant.” Journal of the American WaterWorks Association 1948, 40, 1051–1061.Fitzgerald, G. P., M. E. DerVartanian. “FactorsInfluencing the Effectiveness of Swimming PoolBactericides.” Applied Microbiology 1967, 15, 504–509.Gardner, J., “Chloroisocyanurates in the Treatmentof Swimming Pool Water.” Water Research 1973,7, 823–33.Green, D. E., P. K. Stumpf. “The Mode of Action ofChlorine.” Journal of the American Water WorksAssociation 1946, 38, 1301–5.Morris, J. C. “The Acid Ionization Constant of HOClfrom 5 to 35°.” Journal of the American ChemicalSociety 1966, 70, 3798–3805.Nelson, G. D. “Special Report No. 6862 – SwimmingPool Disinfection with Chlorinated–s–triazinetrione Products.”, Monsanto Chemical Co.,St. Louis, Missouri, March 1967.O’Brien, J. E., J. C. Morris and J. N. Butler in Chemistryof Water Supply, Treatment, and Distribution.(Ann Arbor, Michigan: Ann Arbor SciencePublishers, Inc. 1972) pp. 333–358.Pinsky, M. L.; H–C Hu. “Evaluation of theChloroisocyanurate Hydrolysis Constants.”Environmental Science and Technology 1981, 15,423–30.Stewart, L. S.; L. F. Ortenzio. “Swimming Pool ChlorineStabilizers.” Soap and Chemical Specialties 1964,40, 79–82, 112–13.Swatek, F. E.; H. Raj; and G. E. Kalbus. “A LaboratoryEvaluation of the Effects of Cyanuric Acid on theBactericidal Activity of Chlorine in Distilled Waterand Swimming Pool Water.”, Paper presented atNSPI Convention, Las Vegas, NV, 1967.Wojtowicz, J. A. “Re–evaluation of ChloroisocyanurateHydrolysis Constants.” Submitted for review andpublication to the Journal of the Swimming Pooland Spa Industry 1996.John A. Wojtowicz – Chapter 6.1 111