Interaction of Preservatives With Macrvnlolecules IV*
`Binding of Quaternary Ammonium Compounds by Nonionic Agents
`
`By PATRICK P. DELUCAt and H. B. KOSTENBAUDER
`
`Equilibrium dialysis studies utilizing a semi-
`permeable nylon membrane indicate a high
`degree of association and accompanying in-
`hibition of quaternary ammonium germi-
`cides such as cetylpyridinium chloride and
`benzalkonium chloride with nonionic sur-
`faaants such as Tween 80. Cetylpyridinivm
`chloride was also found to bind to rnethyl-
`cellulose, but not to PVP or Polyox. Benzal-
`konium chloride was not bound to methyl-
`cellulose, PVP, or Polyox under the condi-
`tions of this study.
`
`capable of neutralizing the bacteriostatic effect
`of cetyltrimethylammonium bromide and Ritter
`(11) indicated that Tween 80 was capable of
`neutralizing the bactericidal effect of cetylpyr-
`idinium chloride on tubercle bacilli. Barr and
`Tice (12) investigated seven quaternary am-
`monium compounds in the presence of 5% poly-
`oxyethylene 20 sorbitan monostearate and found
`t’ at only benzalkonium chloride was effective in
`a concentration of 0.1%. ivedderburn (13) in-
`dicated that benzalkonium chloride also mas sub-
`ject to inactivation by nonionics, including su-
`crose ester surfactants, under the conditions of
`her studies.
`iMoore and Hardwick (1) presented
`microbiological data on combinations of quater-
`nary ammonium compounds and nonionic sur-
`factants and devoted considerable discussion to
`the relative effectiveness of such combinations.
`Other reports of inactivation are discussed in the
`recent review by Beckett and Robinson (14).
`Studies cited above (1, 9) indicate that in very
`dilute solutions of the nonionic surfactants there
`is an enhancement of the effectiveness of the
`quaternary ammonium compound, but at higher
`concenfrations of the surfactant the activity of
`the germicide is greatly diminished. This is typi-
`cal of the behavior of many germicides in the
`presence of surface-active agents (1, 15, 16).
`In the present study two typical quaternary
`ammonium compounds, cetylpyridinium chloride
`and benzalkonium chloride, were selected to il-
`lustrate the relative magnitude of any interaction
`which might occur between these agents and
`nonionics such as Tween SO, methylcelldose,
`polyvinylpyrrolidone, and high molecular m-eight
`ethylene oxide polymers. The particular qua-
`ternary ammonium compounds used in this study
`were selected principally because both agents
`show absorption in the ultraviolet and thus can
`readily be determined spectrophotometrically.
`
`A sented microbiological data indicating that
`
`LTHOUGH SEVERAL. investigators have pre-
`
`nonionic surface-active agents can interfere with
`the activity of cationic germicides such as the
`quaternary ammonium compounds, textbooks
`and reference works generally do not emphasize
`the high degree of inactivation which can some-
`times occur in these systems. It has been sug-
`gested that the observed inactivation is attrib-
`utable to a preferential association or binding
`of the cationic agent with the nonionic surfactant
`(l), although no previous data have been pub-
`l’shed which would indicate the magnitude of this
`association. The present work was undertaken
`to obtain quantitative data for the degree of
`binding of several cationic agents by some typical
`nonionin and to compare the degree of binding
`with the degree of inactivation of the cationics.
`Phospholipids such as lecithin have long been
`popular as neutralizing or inactivating media for
`quaternary ammonium compounds in germicidal
`testing (2-i). Quisno, Gibby, and Foter sug-
`gested the use of an inactivating medium con-
`sisting of lecithin and Tween 80, indicating that
`the lecithin vr-as the primary inactivator with the
`In 1949
`Tween acting as a dispersing agent (S).
`Gershenfeld and Stedman (9) reported their
`observations on the activity of several cationics,
`including cetylpyridiniuni cliloride and cetyl-
`trinicthylan~nioniurii bromide. in the presence of
`varying concentrations of a nonionic surfactant.
`noting enhancrrnent of activity at low surfact-
`ant concentration and inhihition at hixher surf-
`nctant concentrations: Dax-irs (10) reported
`that a pdyethylenc glycul cetosteaql ether was
`.
`~ R e c e i v d A U C I I ~ I 21. IlJil. Irwn the % h a d o f Pharmacy,
`Tcmyle I.nivrr.rity. I’tiiiadc~l,h~n Pa Hrvisrrl
`jnnrierv
`I Y I H )
`I’rrscntaci to the Szicntific Scctirm. A. 1’11. A.. Cincinnati
`mcctinq. Au~u.t I ! I . M
`t N‘altrr (i K,arr F e l l ~ ~ r . Smith Klinc si I-rcwh 1.ahora-
`torte\ Philadelphia. I’a
`
`EX PER1 M ENTAL
`Reagents.-Cct\-l1)?ridirliurii chlilritlc ( I-hesadec-
`ylpyridiniurn chloride);r benzalkoniuur chloride
`LT. S. P. (a tnisture o f a l k ~ l d i n i e t h ~ l b e n z ~ l ~ r ~ i -
`moriiurri chlorides in which the alkyls range from
`C.1Il7 to C16€I~:
`1 ;* c e t ~ l ~ l i r i i e t l ~ ~ l h e n z ~ l a ~ ~ ~ n ~ ~ ~ i i ~ ~ n r
`-
`~-
`’ Ceepryn Chloride. The \Vm. S. hierrell Co.. Cincinnati,
`Ohio.
`Zephiran Chloric!e, Winthrop.Stearns Inc.. h-ew \-ark,
`s. v.
`430
`
`~
`
`IPR2015-01099 IPR2015-01097
`IPR2015-01100 IPR2015-01105
`
`Lupin EX1180
`Page 1
`
`

`
`43 1
`
`necessary for cetylpyridinium chloride soiutirms
`and twn to thrce days for benzalkonium chlrritle
`solutions. The presence of Tween 80. however,
`reduced the equilibration time for cetylpyridinium
`chloride to one to two days.
`After equilibration, aliquots were removed from
`both sides of the membrane and concentrations of
`quaternary ammonium compounds were determined
`spectrophotometrically at a wavelength of 2%
`m r for cetylpyridinium chloride and 261.5 mr
`for benzalkonium chloride. using a Beckman DV
`spectrophotorneter. Any interference due to the
`macromolecule was eliminated by using the appro-
`priate concentration in the reference cell.
`Microbiological Studies.-Observations
`'were
`made on the influence of several nonionic surface-
`active agents on
`the bacteriostatic activity of
`cetylpyridinium chloride and benzalkonium c h b
`ride, using Aerobacter aerogenes,' a Gram-negative
`organism. The methods and culture medium
`previously described by Pisano and Kostenbauder
`(19) were employed. Approximate inhibitory con-
`centrations for cetylpyridinium chloride and benzal-
`konium chloride in the presence of Tween 80.
`Triton X-lOO,lo and Pluronic L62.I' were obtained
`through visual observation of samples each day for
`a period of two weeks.
`
`SCIENTIFIC EDITION
`July 1960
`chloride;' Tween 80,' a commercial sample and
`also a sample passed through an ion exchange
`cohmn; methylcellulose. 15 c. p. s.? polyvinyl-
`pyrrolidone;' Polyos IVSR-35.'
`To assure that any observed inactivation of
`quaternary ammonium compounds by Tween 80
`would not be attributable to the presence of un-
`esterified oleic acid in the Tween sample. the Tween
`solution was passed through a mixed-bed ion ex-
`change column* to remove traces of oleic acid which
`might be present. Quantitative removal of oleic
`acid was established by adding known amounts of
`free oleic acid to Tween 80 samples and determining
`oleic acid content by potentiometric titration with
`approximately 0.1 N SaOH before and after pas-
`sage through the ion exchange resin. Subsequent
`studies indicated that the contribution of any traces
`of oleic acid in the commercial sample of Tween 80
`to the inactivation of quaternary ammonium com-
`pounds was probably insignificant.
`Before the polyvinylpyrrolidone was employed
`in the dialysis studies, the sample was extracted
`with ether in a Soxhlet extractor for forty-eight
`hours.
`establish ap-
`Conductivity Measurements.-To
`proximate critical micelle concentrations for the
`samples of cetylpyridinium chloride and cetyldi-
`methylbenzylammonium chloride used
`in
`this
`study, conductivity measurements were obtained
`for aqueous solutions at room temperature (23-
`Industrial
`Instruments model R C
`25'). An
`M15 conductivity bridge and a dipping
`type
`conductivity cell with a cell constant of approx-
`imately 1.00 cm.-I were employed. The water
`used was that used throughout the study and
`had a- specific conductance of
`approximately
`2 X 10-1 ohm-!.
`Dialysis Studies.-Dialysis membranes employed
`in the studies involving Tween 80 were nylon
`membranes as described previously by Patel and
`Kostenbauder (18). These membranes were pre-
`viously shown to be impermeable to Tween 80
`(18). For dialysis studies
`involving methyl-
`cellulose. polyvinylpyrrolidone, and Polyox, Visking
`cellulose casings were employed. The general
`procedure for these studies consisted of placing
`inside the dialysis bag 20 ml. of a solution containing
`the quaternary ammonium compound and, after
`tightly knotting the open end, placing the bag into
`a 12jmm. bottle containing 40 ml. of Tween or
`other polymer solution. A polyethylene film and a
`screw cap were employed as the closure. and the
`bottles were rotated a t 9 r. p. ni. in a constant
`temperature bath a t 30'. As might be expected
`for the long chain quaternary ammonium com-
`pounds, equilibration times of five to seven days were
`
`RESULTS
`Critical Micelle Concentration for Cationics.-
`Approximate critical micelle concentrations for
`cetylpyridinium chloride and cetyldimethylbenzyl-
`ammonium chloride (CDBAC) were obtained from
`the conductivity data illustrated in Fig. 1. The
`critical micelle concentration for cetylpyridinium
`chloride was found to be approximately 1.0 X
`lo-' M, while that for the CDBAC was found to be
`approximately 1.0 X lo-' Ji.
`chloride was found
`Tween SO.-Cetylpyridinium
`to interact to an extremely high degree with Tween
`80 when the cationic was employed in concentra-
`tions comparable to those which might be en-
`countered in product formulations. Figures 2 and 3
`illustrate typical adsorption
`isotherms for
`this
`system, and in Fig. 4 these data are presented in a
`manner which indicates that in an aqueous system
`containing Tween 80 and cetylpyridinium chloride.
`the ratio of total to free cetylpl~dinium chloride is
`primarily a function of the concentration of T m e n
`80. As illustrated in a previous publication (19).
`the ratio of total to free germicide can often be
`employed to predict required preservative concen-
`trations in the presence of the Tween. Figure 4
`indicates that at a concentration of 1% Tween 81.
`approximately 9Scc of the cetylpyridinium chloride
`present is bound to the Tween and thus inactivated.
`Even at a concentration of O.lyc Tween, the d.ita
`that approximately SOYc of
`suggest
`the total
`cetylpyridinium chloride would be
`inactivated.
`The dialysis studies iridicnted no significant dif-
`ference in binding of cAylp!:ridinium ch!nride iy a
`coniniercial sample of Tween 80 and a sample
`passed through an ion eschange resin to remove any
`oleic acid present.
`
`3 A I ii :iu. S.i"ij.
`10 Triton 5.100 is an alkylaryl polyether alcohol. Rcdmt
`and Haas Co.. Philadelphia. Pa.
`is a polyoxseth>-leiie polyoxypropylene
`Pluronic 1.61'
`surfactant. Wysndotte Chemicals Corp.. Wyanrlotte, Xicia.
`
`*Sample obtained through the courtesy of Dr. George D.
`Wessinger. Sterling-Winthrop Research Institute. Rensse-
`laer. h'. Y.
`4 Tween 80 is polyoxyethylene 20 sfrbitan monooleate,
`Atlas Powder Co.. Wilmington. Del.
`.
`.
`
`8 Methocel. I5 c. D. s.. the Dow Chemical C o . . Nidland.
`hlich. ' Flasdone. Antara Chemicals Division of General Aniline
`and Film Corp.. S e w York. N. Y.
`7 Palyox WSR-35 is a poly(ethy1ene oxide) of exceptionally
`high molecular weight (17), Union Carbide Chemicals C o . .
`Feu. Ynrk. S. Y.
`' The ion exchange column contained a niistiire i,f (he
`strongly basic quaternary amine ion exchange resin Amber-
`lite IRA--I(X). previously treated with SaOH solution. and
`the cation exchange resin Amberlite IR-110. previously
`trealed with HCI solution.
`
`Page 2
`
`

`
`431
`
`0 . u
`
`JOURNAL OF THE AMERICAX PHARMACEUTICAL
`ASSOCIATION
`
`Vol. 49, No. 7
`
`0 CLTILPYRIDIWIUU CHLORIDE
`0 BENZALKONIUW CHLORIDE
`(CDBAC)
`
`0
`
`1:o
`
`io w i o
`
`da
`
` lo
`
`C P C I Y l L X 10.)
`F R E E
`Fig. 2.--;\ci.;trotion isothermi for hinrlinc nf w t y l -
`pyridinium chloride by Tween SO in aqueous solution
`-4. [J.I',L; B. 0.5''L; C. l.lj';t; D. 2.5';b
`a t :B)'.
`T\V sll.
`
`0
`
`.2
`
`?I
`
`.6
`
`.B 1.0
`
`1.2
`
`1.4
`
`1.6 1.8
`
`F R E E C P C I M / L X 10.1
`Fig. 3.-Adsorption
`isotherms for binding of cetyl-
`pyridinium chloride by Tween 80 in aqueous solu-
`tion at 30'. D, 2.5%; E, 5.0%; F, 7.5%; G.
`10.0% TW 80.
`
`The marked scatter of points in Fig. 4 is partly a
`result of magnification of experimental error due to
`the pronounced binding which occurs. Calculation
`of r. the ratio of total to free preservative, in some
`cases involves dividing the total concentration by a
`number smaller by a factor of approximately 100;
`thus, a small error in either total or free concentra-
`tion is greatly magnified in calculation of the I
`value. Despite the inherent inaccuracy of such
`treatment of data, the convenience of a graph
`such as that presented in Fig. 4 justifies its presenta-
`tion.
`Data for the binding of benzalkonium by Tween
`80 are illustrated in Fig. 5. While the binding of
`benzalkonium chloride is considerably less than that
`exhibited by cetylpyridinium chloride, it neverthe-
`less is sufficient to cause inactivation. At a concen-
`tration of 1% Tween 80. approximately 507, of the
`benzalkonium chloride present is bound to the
`Tween. There appeared to be no significant dif-
`ference in the degree of binding exhibited by
`benzalkonium chloride U. S. P., which is described
`as consisting of alkyls in the range C? - CIS (go),
`and the binding shown by a pure sample with a C ~ S
`chain, i. e.. CDBAC.
`Figure 6 represents adsorption isotherms for the
`interactiun of cetylpyridiniunl chloride and CDDAC
`with a very dilute solution of Tween 80. This
`study was included to permit observations on the
`degree of binding in the presence of free quaternary
`concentrations both above and below the nnrrnal
`critical niicelle coricentration for the quaternary
`compmnd. These isotherins differ niarkedly from
`those obtained in higher concentrations of Tweeii
`
`Page 3
`
`

`
`July 1960
`
`SCIENTIFIC EDITIOX
`
`433
`
`45 t
`
`data illustrated in Figs. 2 and 3
`Fig. 4.-The
`plotted to give the ratio, r, of total to free cetyl-
`pyridinium chloride as a function of the concentra-
`tion of Tween 80.
`
`0
`
`I
`
`5
`
`3
`2
`4
`T W E E N 8 0 1% W I V )
`ratio, r, of total to free benzalkonium
`Fig. 5.-The
`cl:!uride as a fuxxtioxi uf Tirccn 53 caxciitiLtioxi a:
`30'.
`Total benzalkonium chloride concentration
`O.O'L5-O.lO~b. 0. benzalkonium chloride U. S. P.;
`0. cetylditnethylbenzylarnrnoniuni cliluridr.
`
`7u
`
`60
`
`30
`50,
`40
`20
`10
`FREE DRUG ( W L X 10')
`Fig. B.--Adsorption
`isotherms for binding of cetyl-
`pyridinium chloride and cetyldimethylbenzylam-
`monium chloride in 0.2% Tween 80 a t 30'. These
`curves illustrate the degree of binding occurring a t
`concentrations of the quaternary ammonium com-
`pounds both above and below their normal critical
`micelle concentrations. A , cetylpyridinium C1; B.
`benzalkonium C1( CDBaC) ; C, cetylpyridinium C1.
`0.05 M NaCI. Curve C illustrates the influence of
`electrolyte on the degree of interaction.
`
`80, showing a maximum in the adsorption isotherm
`in the region of the critical micelle concentration.
`Methylcellulose.-Figure 7 shows the interaction
`oE cetylpyridinium chloride with methyldlulose.
`The binding is highly dependent on the cetyl-
`pyridinium chloride concentration, passing through
`a maximum in the region of the normal critical
`rnicelle concentration for cetylpyridinium chloride.
`Benzalkonium chloride was found to exhibit no
`detectable binding to methylcellulose up to a con-
`centration of 2% methylcellulose and 0.125%
`cationic.
`Figure 8 shows the influence of pH and electrolyte
`on the interaction of cetylpyridinium chloride with
`methylcellulose. The binding appears to be in-
`creased somewhat in the presence of strong base.
`but is suppressed by 0.05 JI SaCl and almost
`eliminated by 0.05 JI HC1.
`Polyvin ylpyrrolidone .-Seithhr
`cet y lpyridi nium
`chloride nor benzalkonium chloride was found to
`interact significantly with polyvinylp?.rrolidr,ne at a
`concentration of 2% polyvinylpyrrolidorie and 0 . N
`yo quaternary ammonium compounds.
`Po1yox.-Seither
`cetylpyridiriium chloride nor
`benzalkcrnium chlnritie ~ i l s iourid
`to
`inrwact
`significantly with this polymer at concentrations of
`0.2% Polyos and 0.0770 quaternary animoniurn
`compound.
`Studies.-Cnnrentratinns
`Microbiologifal
`of
`cetylpyridiniurn chloride and benzalkoniurn chldride
`required to inhibit growth of Arrobacfer oero:cnrs
`iii the presence of several riuniouic surfactants are
`
`Page 4
`
`

`
`434
`
`JOURNAL OF THE .\XERICAX PHARMACEUTICAL .~SSOCIATIOX
`
`Vol. 49, No. 7
`
`40
`
`35
`
`30 s
`
`X . 25 - -.
`
`5
`b;' 20
`n V
`1: 15 3
`0
`p1
`
`10
`
`5
`
`0
`
`5
`
`20 25
`15
`10
`F R E E CPC (M/L X 10')
`of cetylpyridinium chloride by
`Fig. 7.-Binding
`methylcellulose at 30°. A, 0.05%; B, 0.1%; C,
`0.5Yb; D, 1.0% methylcellulose.
`
`30 35
`
`.6resented in Table I. These studies were made
`with polymer solutions which were passed through
`ion exchange resins to remove any fatty acid or
`electrolyte present and also with some untreated
`samples of Tween 80. S o attempt was made to
`determine precise
`inhititory concentrations, but
`the increased concentrations of quaternary am-
`monium compound indicate a somewhat higher
`degree of inactivation than would be predicted from
`a consideration of Figs. 4 and 5.
`As would be predicted from the binding studies,
`benzalkonium chloride was found to be inhibited to
`a .lesser degree than was the cetylpyridinium chlo-
`ride. This observation is in agreement with that of
`Barr and Tice (12). who found that in the presence of
`Tween f30 benzalkonium chloride was eaective in a
`concentration of 0.15; while cetylpyridinium chlo-
`ride was not
`The results obtained with Pluronic L62 and Triton
`S-100 indicite that these agents show an ability to
`inactivate the quaternary aininonium compounds
`quite similar to that exhibited by Tween 80.
`
`40
`
`530
`
`35 -
`X < 25
`5
`0
`fJ 20
`n
`z
`3 15
`
`10
`
`5
`
`0
`
`30 35
`
`5
`
`15
`25
`20
`10
`FREE CPC (M/L X 10')
`of pH and electrolyte on the
`Fig. 8.-Iduence
`interaction of cetylpyridinium chloride with 0.5%
`methylcellulose in aqueous solution at 30'. A , dis-
`tilled water; B, 0.003 M XaOH; C, 0.05 iM NaCl;
`D, 0.05 Af HCI.
`
`OF SEVERAL SONIONIC SURF-
`
`TABLE I.--INFLUENCE
`ACTANTS ON CONCENTRATIONS OF CATIONIC RE-
`QUIRED TO INHIBIT Aerobacter acrogenes
`
`Nonionic
`
`Concentration-
`-Inhibitory
`Cetylpyridinium
`Bentalkonium
`CI
`CI
`No growth a t
`1-100,OOO to
`0
`1-100, OOO
`1-250, OOO
`....
`0.57, Tween 80" 1-2,500 to
`1-5.000
`2.070 Tween 80a
`1-250 to 1-500
`3.07, Tween W 1-100 to 1-2170 1-500'io-
`1-1 ,000
`....
`3.0% Triton X-
`1-100 to 1-2170
`loob
`....
`3.0% Plwonic
`1-500 to
`L63*
`1-1,OOO
`a Commercial sample. b Treated with ion exchange resin
`before use.
`
`DISCUSSION
`
`formation of mixed micelles of
`Tween 80.-The
`quaternary aniiiioiiium compound and nonionic surf-
`actant has k e n suggested as a possible mechanism
`for the assnei;itioii with iiiolerulcs such as Tween SO
`( 1 ). For such nii ixiteractiiin. the degree of binding
`\w~uld be cspccted to incrc:tse with increasiiig length
`nf the h>.dricirbq)n chAn o f the cationic. a t least so
`long ;IS the cGinccntr;itiairi of the free cationic did
`t:. .. ... .. T ! ! y c:;!i,...! !!&\...!!. c..!!cy!!:r:!ti.>!:<
`not cswed the nrirm:d critical micclle concentm-
`f,,*
`!!!C
`cctylpyridinium clilori<lc and Cl)D-\C uscd in these
`stualicx wc'rc- f i i u i i t l tii he approxiin;rtely 1.0 X
`.\I. reslwctivrly.
`( M y in the
`11 and 1.11 X Ilt-'
`
`concentrations of Twcen below 0.5r%, were free
`cetylpyridiniurn concentrations of this magnitude
`approached, although in all cases the concentration
`of free benzalkrmiurn was in excess of the critical
`I t was at first suspected
`niicelle concentration.
`that the marked difference in degree o f binding of
`cetylpyridiniurn chloride and benzalkoniurn chloride
`by the Tween might be a result of heterogeneity in
`the composition nf the alkyl of the benzalkonium
`chloride. since the U. S. P. specifies C,-Cls for this
`compound. Horvever. subsequcnt binding studies
`t!Gir!g :I pure xi!iip!c , I f crt~!c!it~~t.t!!~li,enzl.l:lm!~~i~J~~-
`iuni chloride (CDI3AC) resulted in data which es-
`actly duplicutetl the degree of binding exhibited by
`the I:. S. P. product.
`
`Page 5
`
`

`
`July 1960
`
`SCIENTIFIC EDITION
`
`435
`
`If the interaction between the quaternary am-
`monium compound and the Tween can be visu-
`alized as an equilibrium between mixed micelles of
`quaternary ammonium-Tween
`and
`individual
`quaternary ammonium ion and halide ion
`
`aQ+ + bCr- + cTw * Q+.C1Twe
`
`the interaction would be expected to be dependent
`on the concentration1* of
`the long chain cation
`according to the law of mass action.la
`
`In Eq. 1, K' is an equilibrium constant, ( T w ) is the
`free Tween concentration, (Q+> is the concentration
`of individual long chain cation, (CI-) is the concen-
`tration of the individual halide ion. and (Q+.-
`cl-b?iuc) is the concentration of the mixed micelle
`or complex.
`Several investigators have observed that in an
`aqueous solution of an ionic surfactant the concen-
`the individual long chain ion passes
`tration of
`through a maximum as the total concentration of
`surfactant in the solution is increased. This max-
`imum occurs in the concentration range a t which
`micelle formation is observed. The existence of the
`maximum can be shown by the application of the
`law of mass action to micelle formation, the most
`extensive treatments being the work of Mysels
`(21) and of Sexsmith and White (22, 23). For
`the system aQ+ -I- bX-
`Q+.X-b
`where
`
`Sexsmith and White, by solving for (Q') as a
`function OF
`total quaternary concentration, (C),
`and setting d(Q+)/d(C) = 0, demonstrated that for
`several model quaternary ammonium surfactants the
`concentration of individual quaternary ammonium
`ion. (Q+). passes through a maximum if a > b 2 2 .
`The concentration of the individual halide ion,
`( X - ) , and the ion product of quaternary ammo-
`nium ion and halide ion. however, continue to in-
`crease as the total quaternary ammonium concen-
`tration is increased.
`The existence of this maximum in individual
`quaternary ammonium ion concentration provides a
`probable explanation for the observed differences
`in binding affinities of cetylpyridinium chloride and
`CQBAC.
`If
`the concentration of
`individual
`quaternary ammonium ion parallels the total free
`quaternary ammonium compound concentration
`until micelles brgiu to appear, it might be expected
`that the approximately tenfold difference in critical
`micelle concentration for the cetj-lp>-ridinium chlo-
`ride and CDBAC would result in a maximum attain-
`able coricentration of
`individtd quntcrnnry am-
`monium ion greater by approximately a factor of 10
`for the cetylpyridinium chloride solutions. Occur-
`rence of a masiinum in individual quaternary am-
`
`12 Here the assumption is made that concentrations may be
`substituted fur activities.
`'1 The interaction would. of course also be dependent on
`the concentratiun of the Tween aurl the halide iun. but these
`ccmsideratiamr are not vital t o the dimlrriun ahirh follows.
`
`moniurn ion a t extremely low mncentratirms d
`CDBAC suggests that the concentration trf single
`ions of cetyldiniethylbenzylammonium never ap-
`proaches the level encountered in cetylpyridinium
`solutions. For an interaction between a quaternary
`ammonium compound and Tween 80, such as that
`illustrated in Eq. 1. a tenfold difference in quater-
`nary ion concentration would be expected to e x m a
`considerable influence on the degree nf bindinr
`occurring.
`If the proposed mechanism is operative, it should
`be possible to demonstrate maxima in the adsorp-
`tion isotherms for the interaction of cetylpyridinium
`chloride and CDBAC with Tween 80. To this h d .
`the binding:data illustrated in Fig. 6 were obtained,
`using a very dilute solution of Tween 80. The ex-
`istence and relative positions of the maxima in the
`adsorption isotherms are very clearly illustrated.
`These data indicate that the interaction is dependent
`on the concentration of the single quaternary
`ammonium ion and could not be attributed solely to
`the adsorption of ion pairs or micelles. since both
`the product of (Q') ( C l - ) and the concentration of
`micelles are known to increase steadily as the total
`quaternary concentration increases (23). The de-
`crease in the degree of interaction at higher quater-
`nary concentrations is most satisfactorily explained
`thermodynamically on the basis of a decrease in con-
`centration of the individual quaternary ion.
`Observations on the interaction of Tween 80
`and cetylpyridinium chloride in the presence of
`0.05 M NaCl are indicated by the broken line in
`Fig. 6. These points suggest a shift of the maximum
`in the adsorption isotherm to lower concentrations.
`and suggest that the presence of electrolyte might
`be a rather significant factor in determining the
`degree of binding.
`The binding of organic electrolytes by nonionic
`surfactants is not limited to quaternary ammonium
`compounds; drugs such as chlorpromazine, pro-
`methazine. and tetracaine hydrochlorides are also
`bound, as are dyes and some anionic detergents
`(24).
`data for the binding
`Methylcelldose.-Available
`of cations to cellulose derivatives pertain principally
`to those agents used in dyeing textiles. Xumerous
`theories have been presented to account for the
`binding of dyes and metal ions to cellulose, some of
`( a ) ionic binding to residual car-
`which include:
`boxyls or to acidic hydroxyls by ion exchange, (b)
`ion-dipole interactions of cations with ethers or
`hydroxyls. (c) dipole-dipole interactions between
`hydroxyls of cellulose and polar groups of the dye
`molecule, ( d ) interactions between hydrocarbon
`portions of the adsorbate and the linear cellulose
`molecules.
`More recent discussions seem to indicate that
`more than one mechanism is involved in nimt
`cases of adsorption to cellulose (22). Until recently
`the importance of the hydrophobic interaction was
`not fully recngnized.
`It now appears to have been
`established that for a dye to have maximum affinity
`for a cellulose substrate the dye should be a long,
`planar nlolecule which can attain close approach to
`the linear cellulose chain. R'here this condition
`!&$ri & ~ T C L ;i i:;.X.:::;
`is m e t thcic is ari c:;:~ci:;dj
`for the dye-cellul~ae combination;
`the stability
`being attributed to interaction of the hydrophobic
`groups (2.5 ).
`
`Page 6
`
`

`
`436
`
`JOURNAL OF THE AMERICAN PHARMACEUTICAL
`ASSOCIATION
`
`Vol. 49. No. 7
`
`Figure 7 illustrates the binding of cetylpyridinium
`chloride by methylcellulose as a function of methyl-
`cellulose and cetylpyridinium chloride concen-
`tration. The interaction shows a high dependency
`on cetylpyridinium concentration and the initial
`portion of the curve, showing a rapidly increasing
`slope. is similar in appearance to the binding of
`phenol by PVP and polyethylene glycols (26) and
`to the binding of phenol (27, 28) and organic ions
`(29, 30, 31) by proteins. Curves such as this are
`sometimes considered to result when adsorption of a
`critical quantity of drug results in a change in
`configuration of tightly coiled polym’er molecules to
`make additional binding sites available. A curve
`of this nature, which seemingly indicates limitless
`binding capacity, is also similar to multilayer
`adsorption isotherms, and in this case might be
`explained on the basis of possible tail-to-tail adsorp-
`tion of quaternary after the initial adsorption of a
`critical number of ions. A mechanism of this na-
`ture has recently been demonstrated for the bind-
`ing of cationic dyes and surfactants by cotton,
`viscose rayon, and oxycellulose (22, 23, 32. 33).
`These workers demonstrated that the ion exchange
`process involved residual carboxyls in the cellulose.
`A recent report by Fishman and Miller (34)
`indicates that starch also interacts with quaternary
`ammonium compounds, and from the data pre-
`sented the mechanism for the interaction would
`appear to be quite similar to that which occurs with
`cellulose and methylcellulose.
`Data’presented in Fig. 8 illustrate the effect of
`pH and electrolyte on the degree of binding, and
`the results are indicative of an ion exchange mech-
`anism as the initial step in the interaction. The
`interaction is apparently enhanced by dilute alkali,
`greatly diminished by 0.05 M SaCI. and almost
`entirely eliminated by 0.05 M HC1.
`The maximum which appears in the adsorption
`isotherm occurs in the neighborhood of the critical
`mirelle concentration for cetylpyridinium chloride
`and is typical of adsorption isotherms described by
`Sexsmith, ct al. (22. 23. 32, 33). for the interaction
`of quaternary ammonium compounds with cellulose.
`Sexsmith, el al., demonstrated that the existence of
`the mzxinium can be attributed to the occurrence
`of a maximum in the concentration of individual
`quaternary ion and the resultant maximum in the
`ion exchange adsorption (22.23).
`to interact
`Failure of benzalkonium chloride
`with methylcellulose can be attributed to the
`relatively low concentration of individual quaternary
`ion which can m r
` in these solutions. Failure of
`Polyox and PVP to interact with the quaternary
`amnioniurn compounds can probably be attrihuted
`to the nonionic and nonionogenic nature of these
`polymers.
`estimation
`Microbiological Observations.-The
`of inhibitory concentrations of cationic agents in the
`presence o f t lie nonionics indicated an cven higher
`degree of inactivation of the cationic than would be
`predicted from the equilibrium dialysis studies.
`This discrepancy can probably be attributed to the
`presence of approximately 0.1 .\I
`total electrolyte
`in tiir cuiiuir i i d i u a .
`.\ddi:ii;n
`c;f c!cc::::lytc is
`Lmown to lowcr the critical nlicelle cwiwntration
`for quaternary ammonium compounds (3.5). and at
`the critieil miwlle nmrcn-
`mnccntraticmz bc*!tw
`tration the presence of electrolyte would also be
`
`expected to enhance the interaction between the
`cationic and the Tween. While an exact correla-
`tion between binding data and microbiological
`data would apparently require knowledge of the
`interaction as a function of electrolyte concentra-
`tion, the dialysis studies presented here do verify the
`existence of a remarkably high degree of interaction
`for some quaternary ammonium germicides with
`nonionics such as Tween 80 and methylcellulose,
`and illustrate the perils of making the general
`assumption that nonionic agents are compatible
`with cationic and anionic drugs.
`
`SUMMARY
`
`1. Equilibrium dialysis studies indicate an
`extremely high degree of association and accom-
`panying inactivation of quaternary ammonium
`germicides in the presence of nonionic surface-
`In 1 per cent aqueous solutions of
`active agents.
`Tween 80 approximately 95 per cent of the total
`cetylpyridinium chloride or 50 per cent of the total
`benzalkonium chloride present would be bound
`to the Tween and thus inactivated. In 0.1 per
`cent Tween 80 solution approximately 60 per
`cent of the cetylpyridinium chloride would be
`bound.
`2. Methylcellulose was found to interact
`significantly with cetylpyridinium chloride, but
`not with benzalkonium chloride. Polyvinyl-
`pyrrolidone and Polyox were not found to inter-
`act with these cationics.
`3. These studies indicate quite clearly that
`it is not justifiable to assume that nonionic agents
`are always compatible with cationic and anionic
`drugs.
`
`REFERENCES
`
`14) Rahn. 0.’. T&s JOURNAL, 36, 13-1(1947).
`
`(1) Mopre. C. D.. and Hardwick. R. B., Mfg. Chemirf.
`27,305(1936); 29,194(!958).
`(2) Baker. 2.. Hamson. R. W.. and Miller. B. F.. J .
`E z p f l . Med., 74, 621(1941).
`(3) hiud!er. W. S.. Seeley. D. B.. and Larkin, E. P.. Soap
`Sarrf. Chemicals 23 123(1947).
`5 ) Lawrence. C. A.. “Surface Active Quaternary Am-
`monium Germicides,” Academic Press. Inc., New York,
`N. Y.. 19.50. pp. 126 IT.
`(6) Reddish. G. F.. “Antiseptics. Disinfectants, Funpi-
`cides. and Chemical and Physical Sterilization,” 2nd ed.. Lea
`& Febiger. Philadelphia, Pa., 1957, p. 51.
`(7) Schwartz. A. JI.. Perry, J. W;. and Rerch. J.. “Sur-
`face Active Agents and Detergents. Vol. 11. Interscience
`Publishers. Inc.. New York. N. Y. 1958 p. 208.
`(8) Quisno. R.. Gibby. I. W.