APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept. 1987, p. 2082-2089 Vol. 53, No. 90099-2240/87/092082-08$02.00/0Copyright © 1987, American Society for Microbiology

Bactericidal Activities of Selected Organic N-HalaminesDELBERT E. WILLIAMS,' S. D. WORLEY,'* S. B. BARNELA,1 AND LARRY J. SWANGO2

Department of Chemistry, College of Science and Mathematics,' and Department of Microbiology, College of VeterinaryMedicine,2 Auburn University, Auburn, Alabama 36849-3501

Received 15 December 1986/Accepted 18 June 1987

The bactericidal efficacies of three organic N,N'-dihalamine disinfectants in the class of compounds termedimidazolidinones were determined for combinations of pH, temperature, and water quality treatments by usingStaphylococcus aureus and Shigella boydii as test organisms. The compound 1,3-dibromo-4,4,5,5-tetramethyl-2-imidazolidinone was found to be the most rapidly acting bactericide, especially under halogen-demand-freeconditions. The mixed N,N'-dihalamine 1-bromo-3-chloro-4,4,5,5-tetramethyl-2-imidazolidinone was found tobe intermediate in terms of rate of disinfection, while the compound 1,3-dichloro-4,4,5,5-tetramethyl-2-imidazolidinone was observed to be the slowest acting bactericide. When overall effectiveness was judged on thebasis of stability of the disinfectants along with rates of disinfection, the mixed halamine was considered toexhibit great potential for use as a disinfectant in an aqueous solution.

During the past several years our laboratories have beeninvolved in the syntheses and testing of organic halamines aspossible alternatives to free chlorine as broad-spectrumdisinfectants. Much of the research effort has been focusedon but not limited to the use of these halamines as waterdisinfectants.The compounds which are being studied here should not

be confused with the organic halamines that are produced asa result of chlorinating or brominating water containingnondefined mixtures of nitrogenous compounds. As hasbeen pointed out by White (15), many of these halogenatedderivatives exhibit minimal efficacies as disinfectants. Onthe other hand, several prudently synthesized halamines thathave considerable disinfecting or sanitizing activities exist.The organic N,N'-dihalamines under consideration here

have several attributes that make them potential candidatesfor use in the applications mentioned. Many of these at-tributes have been discussed in publications dealing with asimilar N-halamine, 3-chloro-4,4-dimethyl-2-oxazolidinone,henceforth referred to as compound I (see Fig. 1), which hasbeen studied extensively in our laboratories. Compound Iwas first synthesized and shown to be bactericidal byKaminski et al. (8) and Kosugi et al. (10). Further work inour laboratories has demonstrated its efficacy as a bacteri-cide (3, 16, 21); its stability in water as a function of pH,temperature, and water quality (17, 22); its lack of tendencyto produce trihalomethanes (18); its lack of toxicity forchickens drinking water containing it (13); some possiblemechanisms of action of the disinfectant (1, 9); and othergeneral properties of compound I as a water disinfectant (20,23, 24). As a result of this extensive study of compound I, itbecame evident that organic chloramines (and probablyhalamines in general) might have been rather hastily over-looked as useful biocides.Some possible shortcomings in the overall effectiveness of

compound I as a biocide were identified, such as the longercontact time required for disinfection than that needed forfree chlorine for some organisms and the absence of a freechlorine residual, which was contrary to the accepted guide-lines for drinking water in distribution systems. With thesedeficiencies identified, a series of experiments was under-

* Corresponding author.

taken to synthesize new organic halamine compounds whichwould not have the usual problems associated with organichalamine disinfectants. Previously, the bromine-containinganalog of compound I, 3-bromo-4,4-dimethyl-2-oxazol-idinone (compound IB), which exhibited many desirableproperties of a soluble disinfectant (7, 22a, 23), was pre-pared. Compound IB, while an excellent bactericide undermost test conditions, was not as stable in solution as desired,so further halamine compounds were synthesized andtested.

Generally, the new halamines which have been studiedhere are crystalline solids with adequate solubility in waterto allow a residual halogen concentration in excess of thatrequired for disinfection. The compounds are also quitestable under conditions normally encountered in water treat-ment facilities. The possible use of the halamines in sewagetreatment has not been examined at this time. The com-pounds do not form an appreciable free halogen residualupon dissolution, except for 3-bromo-4,4-dimethyl-2-oxazol-idinone, and as a result are stable in solution at moderatetemperatures. A secondary effect of the combined-halogennature of the compounds is that the solutions are typicallynoncorrosive, a factor of concern in the disinfection of waterin cooling towers and other heat exchangers.

This paper will report on the bactericidal activities of threenew halamines under laboratory-controlled conditions ofpH, temperature, and water quality.

MATERIALS AND METHODSChemistry. The compounds tested were 1,3-dichloro-

4,4,5,5-tetramethyl-2-imidazolidinone (A), 1,3-dibromo-4,4,5,5-tetramethyl-2-imidazolidinone (AB), and 1-bromo-3-chloro-4,4,5,5-tetramethyl-2-imidazolidinone (ABC). Thestructures of the compounds are shown in Fig. 1 along withthe structures of compounds I and IB. The compounds weresynthesized by halogenation of the appropriate precursorimidazolidinones (2a).The water used for the disinfectant compound solutions

was made either chlorine demand free (CDF) by standardprocedures (16, 17, 19) or, in some experiments, to con-tain controlled chlorine demand (synthetic-demand water;SDW). Solutions were buffered to pH 4.5, 7.0, or 9.5 by theuse of 0.05 M sodium acetate-acetic acid, 0.05 M sodiumphosphate, or 0.01 M borate-NaOH, respectively. SDW

2082

on April 20, 2020 by guest

http://aem.asm

.org/D

ownloaded from

BACTERICIDAL ACTIVITIES OF ORGANIC N-HALAMINES

cl Br

IB

,Br

0

A AB

Br

:0

ABCFIG. 1. Structures of the N-halamine compounds considered in

this study.

contained 375 mg of each of the inorganic salts NaCl, KCl,MgCl2, and CaCl2 per liter; 50 mg of bentonite clay per liter;30 mg of humic acid per liter; 0.01% (final concentration)heat-inactivated (1 h at 60°C) horse serum; and S x 105 cellsof heat-killed (20 min at 100°C) Saccharomyces cerevisiae(ATCC 7754) per ml. The inorganic salts and bentonite claywere purchased from Fisher Scientific Co. and were usedwithout further purification. Humic acid was a product ofAldrich Chemical Co., Inc., and was used without furtherpurification. Horse serum was collected from healthy horsesby the staff of the Large Animal Clinic, Auburn UniversitySchool of Veterinary Medicine. The SDW solutions were

always buffered at pH 9.5 (0.01 M borate-NaOH) and kept at4°C during experiments.

Microorganisms and media. Cultures of Staphylococcusaureus (ATCC 25923) were obtained as lyophilized disks(Bactrol Disks; Difco Laboratories). Shigella boydii (ATCC9207) was purchased from American Type Culture Collec-tion. S. aureus was maintained on tryptic soy agar (Difco),and S. boydii was cultured on nutrient agar (Difco). Twenty-four-hour cultures were used as the inocula in each experi-ment.

Predicted disinfection times. Data for calculation of thepredicted disinfection times were derived from the normalmethod of testing as previously described (16) by usingconcentrations of total halogen ranging from 2.82 x 10-5 Mto 2.82 x 10' M (equivalent to 1 to 10 mg of Cl+ per liter)under either CDF or SDW conditions. Briefly, cells wereswabbed from the surfaces of solid media, suspended inapproximately 5 ml of sterile CDF saline, and adjusted to a

stock cell density of 1 x 108 to 2 x 108 CFU/ml by using acalibrated Klett-Summerson colorimeter equipped with agreen filter. Aliquots (0.5 ml) of the stock cell suspensionwere added to flasks containing either buffered CDF orbuffered SDW solutions and were allowed to equilibrate for5 min in the case of CDF test conditions or 15 min in the caseof SDW test conditions in a temperature-controlled orbitalshaker. Timing of the exposure to a biocidal agent wasinitiated concurrent with the addition of a disinfectant. Thefinal volume of inoculum, buffer, and disinfectant was 50 ml,with a consequent initial cell density of 1 x 106 to 2 x 106CFU/ml. Disinfectant was always added as a small volume(approximately 0.05 to 0.4 ml) of concentrated stock suchthat the dilution of the cell density was minimal. At prede-termined contact times, 1.0-ml aliquots were withdrawn andadded to an equal volume of 0.02 N sodium thiosulfate in0.05 M sodium phosphate buffer (pH 7.0). Log serial dilu-tions were made in sterile CDF saline, and the dilutionsalong with the thiosulfate-quenched aliquots were plated byusing a microdrop procedure (16). Samples were collectedfor up to 240 min of contact time unless previous replicateexperiments indicated that shorter contact times were suffi-cient for complete inactivation. In most cases the viable celldensity declined to a level such that no survivors wererecovered by the enumeration procedure used. Plates wereincubated for a minimum of 48 h to allow the growth ofslow-growing or injured organisms.The potential for halogen demand because of the introduc-

tion of low levels of medium contaminants by the inoculumpreparation method was measured. A blank inoculum insterile CDF saline was prepared by wiping the sterile surfaceof a tryptic soy agar plate and immersing the swab in 5 ml ofsaline. A solution of approximately 1 mg of Cl+ per liter wasprepared by diluting commercially available NaOCl in CDFphosphate (pH 7.0) and then titrating this solution for freechlorine by using amperometric titration. The solution wassplit into two equal volumes; one aliquot received salinecontaining material introduced from the media, and the otheraliquot received an equal volume of untreated saline. Thesaline volumes were of the same relative volume as would beintroduced through inoculation of reaction flasks (1:100dilution of the saline suspension). The free chlorine concen-trations were determined at 0.5, 1, and 17 h after the additionof saline to both flasks. The free chlorine concentrationswere not found to be different in the two solutions eitherinitially or after 17 h. The flasks were maintained open on thelaboratory bench and occasionally were swirled vigorously.A loss of approximately 0.1 mg of chlorine per liter wasfound in both solutions after 17 h. The fact that no differencewas found in the free chlorine concentrations with or withoutmedium introduction led us to adopt an inoculation methodthat does not include a wash step. We believe that the keypoint in this method of inoculation is the huge dilution usedby suspending cells in saline and then diluting the saline1:100 in the reaction solutions.The organisms chosen for these experiments were S.

aureus and S. boydii. S. aureus was used because in previ-ous experiments (16, 19, 20) it was found to be the mosthalogen-resistant organism in our collection. Even thoughEscherichia coli is the normal indicator of fecal contamina-tion and is therefore commonly used for water disinfectionstudies, we feel that there might be applications for halaminedisinfectants outside of potable water treatment, such asrecreational water disinfection, in which pathogens such asS. aureus are extremely important. S. boydii was used as amodel species of a human enteric pathogen. In addition,

VOL. 53, 1987 2083

I

on April 20, 2020 by guest

http://aem.asm

.org/D

ownloaded from

2084 WILLIAMS ET AL.

prior work has demonstrated that E. coli is readily inacti-vated by compound I (21).

Total chlorine residuals were measured at the end ofexposure cycles by forward amperometric titration or stan-dard iodometric titration (2). lodometric titration was foundto be best for critical evaluation of the halogen content of theimidazolidinone derivatives, owing to the fact that amper-ometric titration only indicated partial halogen concentra-tions for compounds A and ABC. lodometric titration wasfound to be quantitative for the total halogen concentrationsof all of the halamines. Limited testing with a commerciallyavailable diethyl-p-phenylenediamine kit (Hellige Inc.) alsoindicated the necessity for iodometric titration unless drasticmodifications were made in the use of the test kit.

Predicted times necessary for 6-log declines in viablebacterial densities (NINO = 99.9999% decline, where No =106 CFU/ml, the cell density at time zero, and N = CFU permilliliter at time t, the contact time in minutes) were calcu-lated from transformed cell density data. The loglo(CFU permilliliter + 1) transformation was used to code the data (14),and the General Linear Models procedure of the SASInstitute, Inc., was used to predict loglo(CFU per milliliter +1) by using contact time as the independent variable in theregression equation. The treatment variables were tested fortheir effect on calculated disinfection rates by testing for theheterogeneity of the slopes of the generated regressionequations. In this test, the treatment was considered as acovariate (5) and was tested for significant interaction withthe independent variable, contact time. The significancelevel chosen for these comparisons was P > 0.05.

Rechallenge experiments. The resistance to repeated re-inoculation of bacteria was evaluated in a series of rechal-lenge experiments for each of the disinfectants tested. Thedisinfectants were buffered to pH 7.0 with 0.05 M phosphatebuffer (CDF), and solutions were stored in 100-ml screw-captubes tempered at 22°C. The initial total halogen concentra-tion was 7.05 x 10-5 M, equivalent to 2.5 mg of total chlorineper liter. The rechallenge experiments were conducted byperiodically reinoculating the disinfectant solution with afresh challenge of S. aureus at approximately 106 CFU/mland removing aliquots at predetermined timed intervals forthe determination of survivors. The aliquots were added toequal volumes of 0.02 N sodium thiosulfate buffered to pH7.0 with 0.05 M phosphate buffer to quench all the halogenpresent, and then 25-pd microdrops were plated on trypticsoy agar. After 24 and 48 h, subjective estimations ofsurvivorship were made and scored on a scale of 0 to 4, with0 representing no detectable survivors, 1 representing lessthan 50 CFU/25 pul, 2 representing more than 50 CFU/25 pulbut with distinct colonies, 3 representing colonies too nu-merous to count but not confluent, and 4 representingconfluent growth. Preliminary experiments had indicatedthat incubations beyond 48 h were unnecessary for thegrowth of weakened or injured organisms. Special mediawere not used as a rescue for injured organisms (11, 12).Rechallenging was continued until the contact time requiredfor an estimated 6-log reduction in viable bacteria exceeded24 h.

Disinfection by aged compound ABC. The ability of com-pound ABC to kill bacteria after long-term exposure to SDWconditions was determined. A sample of compound ABCwas added to SDW to a final total halogen concentration ofapproximately 2.82 x 10-4 M, and the solution was kept at4°C. Aliquots were removed and titrated for total halogen bystandard iodometric titration. After approximately 100 h ofstorage, a sample of the ABC-containing SDW solution was

L 4 tVBLL ABC

0

0 T_

o 30 60 90 120

Contact Time (min)FIG. 2. Inactivation of S. aureus by 282.1 ptM total halogen from

compounds A (A), AB (E), and ABC (O) in pH 7.0 CDF buffer at220C.

removed, and the rate of disinfection of S. aureus wasdetermined in the usual manner. Controls for this determi-nation consisted of fresh ABC, fresh A at the same totalhalogen concentration as the aged ABC, and a disinfectant-free SDW control culture.

RESULTS

Predicted disinfection times. Predicted inactivation timeswere consistent within treatments (combinations of pH,temperature, disinfectant, concentration of disinfectant, andtest organism) when CDF test conditions were used, i.e.,increasing the halogen concentration resulted in less timebeing needed for a 6-log inactivation. The experimentalprocedure resulted in estimates of approximately 10% errorin predictions of the time required for a 6-log decline inviable organisms, on the basis of the regression modelspecified. Observed inactivation curves were generally lin-ear for treatments with either compound AB or compoundABC conducted in CDF neutral pH buffer at 22°C (Fig. 2 and3). The presence of a "shoulder" in the curve plottedthrough the datum points for compound A was noted inexperiments conducted with both S. aureus and S. boydii asthe test organisms in neutral pH buffers at 22°C. Inactivationof S. aureus by all three halamines was more rapid and wasfound to have much less of the shoulder effect for compoundA when the treatment conditions were made alkaline and thetemperature was kept at 22°C (Fig. 4). In experiments inwhich controlled chlorine demand was imposed with SDW,the rate of disinfection depended on a nonconstant interac-tion between halogen concentration and contact time (i.e.,the halogen demand of the system). Inactivation times forthe three halamines predicted by the model log(CFU permilliliter + 1) = time are presented in Tables 1 to 3. The dataindicate no consistent pH effect for any of the compounds.

APPL. ENVIRON. MICROBIOL.

on April 20, 2020 by guest

http://aem.asm

.org/D

ownloaded from

BACTERICIDAL ACTIVITIES OF ORGANIC N-HALAMINES

0 1 2 3 4 5 6 7 8 9 10Contact Time (min)

FIG. 3. Inactivation of S. boydii by 70.52 p.M total halogen fromcompounds A (A), AB (E), and ABC (K) in pH 7.0 CDF buffer at220C.

The lowest disinfection rates were observed in SDW testsituations. In these experiments the halogen demand of thesystem coupled with the low temperature resulted in 8- to550-fold increases in inactivation times relative to thosecalculated at test conditions of neutral pH and 220C. Theeffect of the high halogen demand alone can be estimated bycomparing the predicted inactivation times for the pH 9.5,40C CDF treatments with those for the pH 9.5, 40C SDWtreatments. In these comparisons, the only differences arethe SDW additions to the buffer. High halogen demandcaused a 12- to 312-fold increase in inactivation time forcompound AB but only a 2-fold increase for compound Aand a 5- to 25-fold increase for compound ABC. Similarresults were obtained previously when a related chloramine,compound I, was evaluated in our laboratories (16).

Limited testing was done with S. boydii as the testorganism. S. boydii was not as difficult to kill as was S.aureus under any of the conditions tested except for neutralinactivation at 220C by compound AB. In previous experi-ments, this organism was also found to be less resistant todisinfection by compound I and Ca(OCl)2 (16, 19). In thoseexperiments, S. boydii was found to be intolerant of theacetate buffer system used to achieve the acidic treatmentpH (19), so pH 4.5 comparisons were not attempted with thisorganism in this study.The effects of the treatment variables-pH, type of disin-

fectant, concentration of disinfectant, temperature, and wa-ter quality-were assessed by using the General LinearModels procedure (unbalanced design). All treatments werefound to be significant at the 0.05 probability level.

Rechallenge experiments. The response of the disinfectantsolutions to repeated additions of viable bacteria was mon-itored over a period of greater than 15 weeks (2,856 h). Theage of the compound was the total storage time since thesolution was first dosed with bacteria. Control cultures wereincluded for all rechallenge experiments, but the data are not

presented, as control viability exceeded the maximum con-tact time of 1,440 min in all cases. There was no significantchange in the rate of disinfection by any of the threecompounds through 336 h with the fourth rechallenge with S.aureus (Table 4). With the fifth rechallenge at 504 h, therewas a marked decrease in the rate of killing by ABC and alesser decrease in the disinfection efficacy of AB. There wasno noticeable change in the rate of inactivation by A until theeighth rechallenge at 1,704 h.

Disinfection by aged compound ABC. Figure 5 shows acomposite graph indicating the kinetics of disinfection byaged ABC in SDW compared to fresh ABC or fresh A inSDW, with the stability curve generated by monitoringhalogen concentration as a function of storage time for theABC solution. The boxed area indicates the approximatetime at which the sample was removed for bactericidaltesting. The fresh A and ABC solutions were prepared by theaddition of a sample of A or ABC to the SDW solution at thebeginning of the disinfection experiment. A 6-log decline wasachieved in 2, 24, and 24 h with fresh ABC, aged ABC, andfresh A, respectively.

DISCUSSIONPredicted disinfection times. The observed pH effect for

compound A was contrary to the well-known response offree chlorine to an alkaline pH (4, 6, 15) but has beendocumented previously for other N-halamine disinfectants(16). Two possibilities have been suggested to cause thiseffect (by a referee): the increased pH favors the formationof an OCl- anion as a result of hydrolysis of the N-Cl bondor the halamine undergoes a change in net charge whichallows a more efficient disinfectant mechanism to be opera-ble. A series of experiments in which the stability of thehalamines was monitored in solutions of controlled pHindicated a somewhat accelerated rate of loss of halogen at

7

6-

E~5-NS)4-

LL

0

0 30 60 90 120

Contact Time (min)FIG. 4. Inactivation of S. aureus by 141 ,uM total halogen from

compounds A (A), AB (O), and ABC (O) in pH 9.5 CDF buffer at220C.

VOL. 53, 1987 2085

on April 20, 2020 by guest

http://aem.asm

.org/D

ownloaded from

2086 WILLIAMS ET AL.

TABLE 1. Predicted inactivation times for compound A treatments

Temp Dose Time 2No. ofOrganism pH (OC) Demand (11M)a (min)b SEc R datum points

considered

S. aureus 4.5 22 CDF 141 8.95 1.12 0.904 54.5 22 CDF 282.1 4.82 0.44 0.964 47.0 22 CDF 28.21 305.90 10.36 0.992 87.0 22 CDF 70.52 184.3 35.69 0.662 77.0 22 CDF 282.1 125 15.42 0.872 99.5 22 CDF 141 123.1 6.32 0.980 79.5 4 CDF 141 684 73.89 0.922 89.5 4 CDF 282.1 492 62.41 0.887 89.5 4 SDW 141 1,336 39.8 0.994 9

S. boydii 7.0 22 CDF 70.52 19.43 0.59 0.982 59.5 4 CDF 141 30.65 2.37 0.947 69.5 4 CDF 282.1 10.45 0.47 0.984 69.5 4 SDW 141 52.34 4.73 0.939 79.5 4 SDW 282.1 28.02 2.49 0.951 6

a Total residual halogen.b Contact time predicted for a 6-log reduction in CFU per milliliter (see text).c Standard error of contact time estimate.

pH 9.5 (S. D. Worley, D. E. Williams, and S. B. Barnela, compound ABC. However, compound ABC was less af-Water Res., in press). These data suggest that the reason for fected by the addition of halogen demand to the low-enhanced biocidal activity is most likely base-catalyzed temperature test solutions than was compound AB (comparehydrolysis of the N-Cl bond. Thus, even though free the predicted inactivation time for S. aureus, pH 9.5, CDF,chlorine in the form of OCl- is a less effective bactericide 4°C, to that for S. aureus, SDW, in Tables 2 and 3). Fromthan is HOCI at a lower pH, the increased concentration of these comparisons it would seem that temperature had aOCl- from compound A at pH 9.5 enhances the bactericidal greater affect than organic load on the disinfection kinetics ofefficacy as compared with nondissociated compound A at compounds ABC and A. For compound AB, the mostpH 7.0. The mixed halamine exhibited characteristics of dramatic effect was caused by the addition of organic load toboth the chloramine and the bromamine in terms of its the test solutions. The three halamines responded in similarresponse to pH treatments. In experiments designed to fashion when treatments were conducted at 22°C, with theexamine the effect of temperature on the inactivation of S. lowest rates of disinfection occurring at a neutral pH.aureus the mixed halamine exhibited a much more dramatic Rechallenge experiments. The trend established in theseincrease in the time required for a 6-log decrease when the experiments was for the bromine-containing compound ABtreatment temperature was reduced from 22 to 4°C than did to exhibit rapid disinfection rates over several rechallengescompound A. The other halamines also had decreased but then to begin to lose disinfection capability as a result ofefficacies, but these losses were less than that recorded for exhaustion of the bromine moiety. The chloramine (com-

TABLE 2. Predicted inactivation times for compound AB treatments

Temp Dose Time 2 ~~~~~~~~~~~~~~~~~No.ofOrganism pH Temp Demand Dose (Time) SEC datum points

considered

S. aureus 4.5 22 CDF 141 0.46 0.027 0.992 37.0 22 CDF 70.52 4.68 0.54 0.928 57.0 22 CDF 141 1.88 0.40 0.850 47.0 22 CDF 282.1 0.94 0.09 0.982 39.5 22 CDF 28.21 10.7 1.47 0.865 69.5 22 CDF 70.52 1.71 0.66 0.907 49.5 22 CDF 141 0.92 0.01 0.999 39.5 4 CDF 70.52 8.19 0.95 0.904 79.5 4 CDF 141 4.66 0.63 0.909 69.5 4 SDW 70.52 2,558 596 0.527 79.5 4 SDW 141 56.8 2.38 0.988 89.5 4 SDW 282.1 26.76 3.09 0.924 6

S. boydii 7.0 22 CDF 28.21 4.85 0.37 0.967 67.0 22 CDF 70.52 1.92 0.28 0.940 39.5 4 CDF 70.52 9.45 0.65 0.976 59.5 4 CDF 141 5.36 0.04 0.999 49.5 4 SDW 70.52 122.7 5.06 0.986 89.5 4 SDW 141 29.42 1.06 0.992 6

a See Table 1, footnote a.b See Table 1, footnote b.c See Table 1, footnote c.

APPL. ENVIRON. MICROBIOL.

on April 20, 2020 by guest

http://aem.asm

.org/D

ownloaded from

BACTERICIDAL ACTIVITIES OF ORGANIC N-HALAMINES

TABLE 3. Predicted inactivation times for compound ABC treatments

No. ofOrganism pH Temp Demand Dose RTie2 coints

(OC) (~~~~~~~LM)a (min)b~~~~~~~~~~consideredS. aureus 4.5 22 CDF 28.21 4.66 0.46 0.950 6

4.5 22 CDF 70.52 3.91 0.66 0.829 64.5 22 CDF 141 1.75 0.24 0.882 54.5 22 CDF 282.1 0.93 0.15 0.846 47.0 22 CDF 28.21 29.17 1.35 0.983 87.0 22 CDF 70.52 8.85 0.79 0.939 77.0 22 CDF 141 4.61 0.58 0.879 67.0 22 CDF 282.1 1.94 0.20 0.925 59.5 22 CDF 28.21 4.54 0.45 0.952 69.5 22 CDF 141 0.92 0.17 0.894 49.5 22 CDF 282.1 0.48 0.14 0.839 39.5 4 CDF 28.21 53.24 4.48 0.933 99.5 4 CDF 141 5.06 0.24 0.987 69.5 4 CDF 282.1 5.02 0.65 0.841 69.5 4 SDW 28.21 1,316 238 0.860 69.5 4 SDW 70.52 357 17.6 0.993 59.5 4 SDW 141 111.4 6.08 0.973 89.5 4 SDW 282.1 27.26 4.69 0.751 6

S. boydii 7.0 22 CDF 28.21 9.88 0.22 0.996 77.0 22 CDF 70.52 4.88 0.25 0.982 67.0 22 CDF 141 1.81 0.05 0.997 57.0 22 CDF 282.1 0.45 0.03 0.993 39.5 4 CDF 70.52 5.54 0.05 0.999 69.5 4 CDF 141 2.71 0.48 0.699 59.5 4 SDW 141 29.88 1.06 0.991 7

a See Table 1, footnote a.b See Table 1, footnote b.c See Table 1, footnote c.

pound A) remained an effective bactericide throughout thetesting period, although the contact times necessary fordisinfection were somewhat extended. The facts that thedata were collected in a subjective manner and that the timedaliquots were taken on an increasing-interval time scale(especially after 90 min of contact time) make the assignmentof error factors very difficult in this experiment. Conse-quently, in making comparisons among storage times of a

given compound, contact time differences of less than 25%were not considered as truly different. The inconsistentdisinfection rates presented (primarily for compound A)most likely represent errors resulting from the addition ofinocula with more or less than 106 CFU/ml or the detectionof very few survivors at a given contact time, resulting in thetransfer of the disinfection data to the following interval. Thepresentation of the data does not take into account thenumber of survivors present, only whether survivors were

present or absent. It is clear that the order of long-termdisinfection efficacy for the compounds was A > ABC >AB. The order of stability of these three compounds as wellas other N-halamines has been compared for aqueous solu-tions buffered to a neutral pH (Worley et al., in press), andit was observed that the stability order was exactly the same

as the efficacy order, on the basis of rechallenge data. Giventhese data, compound A would appear to have potential as a

long-term biocide when extremely short contact times are

not a necessity. The use of low-halogen residuals wouldtherefore be possible without the risk of loss of residualsbecause of halogen demand and the necessity of frequentresidual determinations.Compound ABC seemed to maintain a rapid disinfection

rate until loss of the bromine moiety. While this was notsubstantiated by an analysis capable of differentiating bro-

mine from chlorine in this compound, the rechallenge pointat which a loss of efficacy was first noted (504 h) was thesame point at which compound AB suddenly became lesseffective. Our premise is that the bromine was preferentiallybeing lost, leaving the monochloramine derivative of ABC.This is to be expected given that the N-Br bond is weakerthan the N-Cl bond. It is evident that as further rechal-lenges were performed on the monochloramine decomposi-tion product of ABC, the remaining chloramine continued toconfer biocidal activity until exhaustion at 1,704 h of storageand rechallenge. This second cascade of loss of disinfection

TABLE 4. Disinfection rates for rechallenges of halamines byS. aureus

Contact timeb for:Rechallengetime (h)a A AB ABC

0 120 (3)-1,440 (0) 1 (3)-2 (0) 1 (3)-2 (0)48 180 (1)-1,440 (0) 2 (1)-5 (0) 1 (2)-2 (0)168 108 (2)-1,440 (0) 1 (1)-2 (0) 1 (1)-2 (0)336 165 (1)-1,440 (0) 2 (2)-5 (0) 5 (1)-10 (0)504 120 (2)-1,440 (0) 10 (3)-30 (0) 60 (4)-1,440 (0)672 240 (1)-360 (0) 30 (3)-60 (0) 420 (3)-1,440 (0)

1,008 120 (2)-240 (0) >1,440 (4) 360 (4)-1,440 (0)1,704 300 (1)-1,440 (0) >1,440 (4) >1,440 (1)2,856 360 (2)-1,440 (0) >1,440 (4) >1,440 (2)

a Length of time since the solution was dosed with halamine at a totalhalogen concentration of 7.05 x 1i-' M.

b Contact times are in minutes. Numbers in parentheses indicate thesubjective scores for numbers of survivors (see text) at the indicated contacttimes. Data were not collected for scores at contact times between thoseindicated in the ranges.

VOL. 53, 1987 2087

on April 20, 2020 by guest

http://aem.asm

.org/D

ownloaded from

2088 WILLIAMS ET AL.

7 -

IL

0)

2'

1

0

LIJ'v':

40 60 80 100

Time (hr)

0 4 8 12 16 20 24

Contact Time (hr)FIG. 5. Effect of SDW storage on ABC. Freshly prepared ABC (O), ABC stored in SDW for 100 h (*), or freshly prepared A (A) was used

as a disinfectant, and the inactivation time for S. aureus (106 CFU/ml) was predicted. The inset is the stability curve associated withcompound ABC in SDW and indicates the time at which stored compound ABC was sampled for biocidal activity. All disinfectants wereadded to a starting total halogen concentration of 1.41 x 10-4 M. +, SDW control.

efficacy coincided with the loss of disinfection efficacy ofcompound A (Fig. 5).Compound AB lost disinfection efficacy rapidly after 500 h

of storage and rechallenge. While this compound was themost active of the three tested in short-term storage, theincreased biocidal efficacy was not generated without aprice, the loss of long-term stability.

Disinfection by aged compound ABC. Figure 5 indicatesthat as total halogen was lost from compound ABC, thedisinfection kinetics of compound ABC become more similarto those of compound A. The rate of disinfection by agedABC was not significantly different from the rate of disin-fection by freshly prepared A. However, the disinfectionrate in SDW was drastically different from that observed forfreshly prepared ABC. These data support the previousstatement that compound ABC preferentially loses its bro-mine moiety and, as a result, resembles compound A in itsdisinfection efficacy as it ages. The preferential loss ofbromine is greatly accelerated in SDW storage, owing to themuch greater halogen demand of SDW than of CDF rechal-lenge cultures.

After consideration of both the rechallenge data and thedisinfection data, it can be concluded that compound ABC isan exceptional bactericide for a broad spectrum of applica-tions. The compound has adequate stability to continue as aneffective biocide for many weeks while still maintaining anadequate disinfection rate. Compounds A and AB should notbe ignored as possible alternative disinfectants, especiallycompound A for use in treatments for which extended

contact times do not present a problem, such as closed-cyclecooling or heating systems or swimming pool winterization.

ACKNOWLEDGMENTS

We gratefully acknowledge the support of the U.S. Army MedicalResearch and Development Command at Fort Detrick, Frederick,Md., and the U.S. Air Force Engineering and Services Center atTyndall Air Force Base, Tyndall, Fla., through contract DAMD17-82-C-2257.We thank the Water Resources Research Institute at Auburn

University for administration of this project.

LITERATURE CITED1. Ahmed, H., J. L. Aull, D. E. Williams, and S. D. Worley. 1986.

Inactivation of thymidylate synthase by chlorine disinfectants.Int. J. Biochem. 18:245-250.

2. American Public Health Association. 1985. Standard methods forthe examination of water and wastewater, 16th ed. AmericanPublic Health Association, Washington, D.C.

2a.Barnela, S. B., S. D. Worley, and D. E. Williams. 1987. Synthe-ses and antibacterial activity of new N-halamine compounds. J.Pharm. Sci. 76:245-247.

3. Burkett, H. D., J. H. Faison, H. H. Kohl, W. B. Wheatley, S. D.Worley, and N. Bodor. 1981. A novel chlorine compound forwater disinfection. Water Resour. Bull. 17:874-879.

4. Dychdala, G. R. 1968. Chlorine and chlorine compounds, p.278-304. In C. A. Lawrence and S. Block (ed.), Disinfection,sterilization, and preservation. Lea & Febiger, Philadelphia.

5. Freund, R. J., and R. C. Littell. 1981. SAS for linear models.SAS Institute, Inc., Cary, N.C.

6. Haas, C. N., and S. B. Karra. 1984. Kinetics of microbial

APPL. ENVIRON. MICROBIOL.

on April 20, 2020 by guest

http://aem.asm

.org/D

ownloaded from

BACTERICIDAL ACTIVITIES OF ORGANIC N-HALAMINES

inactivation by chlorine. Water Res. 18:1443-1449.7. Kaminski, J. J., and N. Bodor. 1976. 3-Bromo-4,4-dimethyl-2-

oxazolidinone. Tetrahedron 32:1097-1099.8. Kaminski, J. J., M. H. Huycke, S. H. Selk, N. Bodor, and T.

Higuchi. 1976. N-halo derivatives. V. Comparative antimicro-bial activity of soft N-chloramine systems. J. Pharm. Sci. 65:1737-1742.

9. Kohl, H. H., W. B. Wheatley, S. D. Worley, and N. Bodor. 1980.Studies on the antimicrobial activity of N-chloramine com-pounds. J. Pharm Sci. 69:1292-1295.

10. Kosugi, M., J. J. Kaminski, S. H. Selk, I. H. Pitman, N. Bodor,and T. Higuchi. 1976. N-halo derivatives. VI. Microbiologicaland chemical evaluations of 3-choro-2-oxazolidinones. J.Pharm. Sci. 65:1743-1746.

11. LeChevallier, M. W., S. C. Cameron, and G. A. McFeters. 1983.New medium for improved recovery of coliform bacteria fromdrinking water. Appl. Environ. Microbiol. 45:484-492.

12. McFeters, G. A., S. C. Cameron, and M. W. LeChevallier. 1982.Influence of diluents, media, and membrane filters on detectionof injured waterborne coliform bacteria. Appl. Environ. Micro-biol. 43:97-103.

13. Mora, E. C., H. H. Kohl, W. B. Wheatley, S. D. Worley, J. H.Faison, H. D. Burkett, and N. Bodor. 1982. Properties of a newchloramine disinfectant and detoxicant. Poult. Sci. 61:1968-1971.

14. Sokal, R. R., and F. J. Rohif. 1981. Biometry, 2nd ed. W. H.Freeman & Co., New York.

15. White, G. C. 1986. Handbook of chlorination, 2nd ed. VanNostrand Reinhold Co., New York.

16. Williams, D. E., S. D. Worley, W. B. Wheatley, and L. J.Swango. 1985. Bactericidal properties of a new water disinfec-tant. Appl. Environ. Microbiol. 49:637-643.

17. Worley, S. D., and H. D. Burkett. 1984. The stability in water ofa new chloramine disinfectant as a function of pH, temperature,and water quality. Water Resour. Bull. 20:365-368.

18. Worley, S. D., H. D. Burkett, and J. F. Price. 1984. Thetendency of a new water disinfectant to produce toxic triha-lomethanes. Water Resour. Bull. 20:369-371.

19. Worley, S. D., L. J. Swango, M. H. Attleberger, C. M. Hendrix,H. H. Kohl, W. B. Wheatley, D. E. Williams, H. D. Burkett, D.Geiger, and T. Clark. 1984. New disinfection agents for water.Technical report AD A149537. Engineering & Services Labora-tory, U.S. Air Force, Tyndall Air Force Base, Tyndall, Fla.

20. Worley, S. D., W. B. Wheatley, H. H. Kohl, H. D. Burkett, J. H.Faison, J. A. Van Hoose, and N. Bodor. 1983. A novel bacteri-cidal agent for treatment of water, p. 1105-1113. In R. L. Jolley(ed.), Water chlorination: environmental impact and healtheffects, vol. 4. Ann Arbor Science Publishers Inc., Ann Arbor,Mich.

21. Worley, S. D., W. B. Wheatley, H. H. Kohl, H. D. Burkett, J. A.Van Hoose, and N. Bodor. 1983. A new water disinfectant; acomparative study. Ind. Eng. Chem. Prod. Res. Devel. 22:716-718.

22. Worley, S. D., W. B. Wheatley, H. H. Kohl, J. A. Van Hoose,H. D. Burkett, and N. Bodor. 1983. The stability in water of anew chloramine disinfectant. Water Resour. Bull. 19:97-100.

22a.Worley, S. D., D. E. Williams, S. B. Barnela, E. D. Elder, L. J.Swango, and L. Kong. 1986. New halamine water disinfectants,p. 61-79. In G. E. Janauer (ed.), Progress in chemical disinfec-tion, vol. 3. State University of New York, Binghamton.

23. Worley, S. D., D. E. Williams, H. D. Burkett, S. B. Barnela, andL. J. Swango. 1985. Potential new water disinfectants, p.1269-1283. In R. L. Jolley (ed.), Water chlorination: environ-mental impact and health effects, vol. 5, F. Lewis, PublishersLtd., London, United Kingdom.

24. Worley, S. D., D. E. Williams, H. D. Burkett, L. J. Swango,C. M. Hendrix, and M. H. Attleberger. 1984. Comparisons of anew N-chloramine compound with free chlorine as disinfectantsfor water, p. 45-60. In G. E. Janauer (ed.), Progress in chemicaldisinfection, vol. 2. State University of New York, Binghamton.

VOL. 53, 1987 2089

on April 20, 2020 by guest

http://aem.asm

.org/D

ownloaded from