Journal of Colloid and Interface Science 256, 201207 (2002)doi:10.1006/jcis.2001.8009

oth

rid

p

Titracrease wchain. Tacids wa spreadto relatthe sprefirst dissubsequteristicvalues ofound tpKa valfatty acmoleculdecreasgroups.

Key WpH; pKmoleculperatur

Thean interand insof a suran adsoand chapolar gthese faoverall

In 19acids ofindicatithe mol(10). Itin the c

1 To w

l

falhmEffect of Degree, Type, and Positionof Long-Chain Fat

James R. Kanicky and DinesCenter for Surface Science & Engineering, NSF-Engineering Research Center fo

Engineering and Anesthesiology, University of Flor

Received July 12, 2001; accepted October 1, 2001;

tion of a series of C18 fatty acids yields pKa values that de-ith an increasing degree of unsaturation in the fatty acidhe pKa values of stearic, elaidic, oleic, linoleic, and linolenic

ere studied and compared to values of area per molecule inmonolayer of these acids. The decrease in pKa was found

e to melting point temperature and area per molecule inad fatty acid monolayer. The pKa value was determined bysolving the fatty acid in a high pH solution (pH > 10) and

acids wan alkyand (2)found,acid) h(e.g., o

Furting ofently titrating the solution with HCl to obtain the charac-S-shaped curves used to calculate the pKa values. The pKaf stearic, elaidic, oleic, linoleic, and linolenic acids were

o be 10.15, 9.95, 9.85, 9.24, and 8.28, respectively. Theseues were in the same order as area per molecule values ofids in spread monolayers. This suggests that as area pere increases the intermolecular distance increases and pKaes due to reduced cooperation between adjacent carboxyl

C 2002 Elsevier Science (USA)ords: fatty acids; stearic; elaidic; oleic; linoleic; linolenic;

a; unsaturation; monolayer; molecular packing; area pere; intermolecular distance; limiting area; melting point tem-e.

INTRODUCTION

manner in which surfactant molecules align and pack atface is an important factor in systems involving solubleoluble interfacial films. Among the important propertiesfactant that dictate the strength, elasticity, and stability ofrbed film are surface activity, chain length compatibility,in cohesion (1, 2), as well as the interaction between

roups of the molecules in the monolayer (39). All ofctors contribute to the intermolecular distance and thestability of the monolayer.17, Irving Langmuir showed that monolayers of fattyvarious chain lengths compress to the same limiting area,ng that the different acids must all form films in whichecules are orientated vertically with respect to the surfacewas further shown that fatty acids with a cis double bondhain showed limiting areas different than those of fatty

hom correspondence should be addressed. E-mail: shah@che.ufl.edu.

in the lelimitingmonolayacid restems semonolay

The ctem obtdently v(6). Shaphospholecithinabove ato blockcreasedlecithinspondeddegree oments (7that incrper molhave shoterol tostriking(13). Faniques taffects hhow momacrosc

The paffects ifor instathe solu

201f Unsaturation on the pKay AcidsO. Shah1

Particle Science and Technology, Departments of Chemicala, Gainesville, Florida 32611

ublished online June 11, 2002

ith a saturated chain. The addition of a double bond tochain (1) limits flexibility of the chain in this region

decreases adhesion between molecules (11, 12). It wasor example, that a saturated fatty acid chain (e.g., stearics a limiting area of 21 A2, while an unsaturated chaineic acid) has a limiting area of 32 A2 (2, 8).er evidence as to the effect of unsaturation on the pack-olecules at surfaces was shown by Shah and Schulmancithincholesterol system (4, 5). They showed that theareas of synthetic (dipalmitoyl), egg, and yeast lecithiners depend on the degree of unsaturation of the fatty

idues and that the addition of cholesterol to these sys-rves to increase the area per molecule and fluidize theer.ompression data and limiting areas of the lecithin sys-ained by surface pressure measurements were indepen-erified by an enzyme-catalyzed hydrolysis experimenth and Schulman (6) measured the action of snake venomlipase A on lecithin monolayers. It was shown that amonolayer does not hydrolyze when it is compressedcritical pressure. The critical surface pressure requiredthe penetration of the enzyme into the monolayer in-

with molecular area (i.e., dioleoyl lecithin > soybean> egg lecithin > dipalmitoyl lecithin), which corre-to their area per molecule as well as to the increase inf unsaturation of the molecule. This and other experi-) corroborated previous limiting area results indicatingeasing the degree of unsaturation yields increased area

ecule in a spread monolayer. Furthermore, Rao and Shahwn that the addition of a minimum of 20 mol% choles-

an arachidyl alcohol monolayer fluidizes the film andly increases the evaporation of water through the filmng and Shah showed again using infrared imaging tech-hat the molecular packing at the interface significantlyeat transfer at the interface (14). These studies illustratenolayer packing and fluidity can influence importantopic phenomena.resence of a net charge on a surfactant polar group alsots performance in an adsorbed monolayer. Fatty acids,nce, can become ionized by an increase in the pH oftion. If all the fatty acid molecules become ionized,

0021-9797/02 $35.00C 2002 Elsevier Science (USA)

All rights reserved.

202 KANICKY AND SHAH

e

repulsilayer c(15, 16changetion ofand HupH relaionizedproposbulk is

Mos4.8. Foacid (Care fourepresehydrogexistingby attacarboxreleasinis true,increasever, becarboncarbons, the pKa tends to level off. Increasing the chain lengthfrom pentanoic to hefrom 4.82 to 4.83. Tteractions (i.e., the eof the carbon chain)the alkyl chain.

However, in a prevalues of long-chainfatty acid molecule athose for the shortevalues higher thanbeyond. Heikkila et a(i.e., palmitic acid) ofor the same fatty aci

tlr

hy

fr

dine value, 192) was supplied by Pfaltz and Bauer (Waterbury,were

tionstilled.rwise

listedxcess

tratedoxideacid.

umedsalts

d saltxanoic acid, for example, increases the pKaherefore, we know that intramolecular in-ffects on the carboxylate anion by the restbecome negligible beyond four carbons in

vious work (21) it was shown that the pKafatty acids depend on the chain length of thend can increase to values much higher thanr chained analogs. These pKas can reach9 as the chain length increases to C16 andl. (22) reported a pKa for hexadecanoic acidf 9.7, and Peters (23) reported a pKa of 8.5

d. We find the pKa of palmitic acid to be 8.6

CT). Potassium hydroxide pellets and hydrochloric acidsupplied by Fisher Scientific (Fair Lawn, NJ). All soluwere prepared using water that was both deionized and disAll experiments were carried out at 20 1C unless othestated.

Solution preparation. Each of the C18 fatty acidsabove were placed in a clean beaker containing 10 mol% esolid potassium hydroxide pellets. Water was very slowly tito the mixture until all of the fatty acid and potassium hydrhad dissolved, producing a clear solution for each fattyBecause C18 fatty acids are insoluble in water, it was assthat this solution was composed of water-soluble potassiumof these fatty acids. The concentrations of these fatty aciFIG. 1. Effect of chain length on the intermolecular distance betwe

on between similarly charged molecules in the mono-an result in an expansion of the monolayer at high pH), which in turn can lead to a weak and unstable film. Thein monolayer characteristics of fatty acids as a func-

pH of the bulk solution was first reported by Schulmanghes (17), who found that the surface potential (V )tion resembled an acidbase titration curve. For weaklymonolayers such as those created by fatty acids, it was

ed that the difference in pKa between the surface and thesmall (18).t short-chain carboxylic acids have a pKa value of ca.r example, when acetic acid (CH3COOH) and propionicH3CH2COOH) are dissolved in water, their pKa valuesnd to be 4.74 and 4.87, respectively (19). The pKa valuents the ionic environment of the solution where 50% ofen atoms are removed from the carboxyl group by the

OH ions in the solution. The pKa can be decreasedching an electron-accepting substituent to stabilize theylate anion (20). The converse effectthat of electron-g substituents lowering the acidity and raising the pKathough less dramatic. The pKa can also be raised by

ing the carbon chain length of the carboxylic acid. How-cause electronic effects are not felt beyond two to three

s, when the chain length is increased beyond about four

8.8 bypKa va

Figuaffect itreases,

jacent meach otthe fattatom bethe morthe high

Thisof C18to descby the dof the a

Matepurity)Elaidicwere su

Linolenn the fatty acid molecules in an adsorbed film.

itration (21). Christodoulou and Rosano (24) estimatedues of 9.0 for octadecanoic (i.e., stearic, C18) acid.e 1 shows how the chain length of the fatty acid mights pKa . It is known (25, 26) that, as the chain length inc-van der Waals interactions between the chains of ad-olecules increase, bringing these molecules closer toer. When this happens, the carboxylic acid groups ofacids are also packed closer, shielding the hydrogen

tween the two oxygen atoms. The closer the molecules,e strongly shielded the hydrogen atom and consequentlyer the pKa (21).paper presents our studies on the observed pKa valuesatty acid solutions or aqueous dispersions and attemptsibe how the pKas of these fatty acids can be influencedegree of unsaturation and subsequent packing behaviorlkyl chains.

EXPERIMENTAL

rials. Stearic acid (97+% purity) and oleic acid (97+%were supplied by Fisher Scientific (Fair Lawn, NJ).acid (99% purity) and linoleic acid (95% purity)

pplied by Sigma Chemical Company (St. Louis, MO).ic acid (acid value, 203; saponification value, 200; io-

EFFECT OF UNSATURATION ON pKa OF FATTY ACIDS 203

solutions varied (stearic acid, 0.30 M; elaidic acid, 0.293 M; oleicacid, 0.35 M; linoleic acid, 0.66 M; linolenic acid, 0.61 M). Thesolutiodilutedtitratio

Deteabovedrochloneutralinflectiquentlytion votion enand mebars w

It wacid msolutiothe pHions kepKa , aare proionizedgroups

Figuincreaspletelybe packat the nmolecuOleic,cis doupositiobecom

Stea(27, 28packs nbecausproximdo notstate fooleic aiting arof theThis mgroup.low thesufficiemay inmonolaof Ref.

2

r

uaosyf4nns were filtered with a 0.22 m Cameo syringe filter andwith water to 100 mmol for determination of pKa by

n.rmination of pKa. The pKa values of the acids listedwere determined by titration at 20C with 0.1 M hy-ric acid using a METROHM 726 titroprocessor. The

ization endpoint was first determined. This is the point ofon on the S-shaped titration curve. The pKa was subse-calculated as the pH of the solution at half the neutraliza-

lume (i.e., half the volume needed to reach the neutraliza-dpoint). Five titrations were performed on each solutionan pKa values for each fatty acid were calculated. Error

ere calculated using a 95% confidence interval.

RESULTS AND DISCUSSION

as shown previously (21) that the molecules of a fattyonolayer are packed closest together when the pH of then is close to the pKa of the fatty acid (Fig. 2). Whenis very high, ionic repulsion between the carboxylateeps the molecules apart. When the pH is close to the

pproximately 50% of the acid groups in the adsorbed filmtonated. The added iondipole interaction between theand unionized acid groups is absent when all of the acidare protonated (i.e., pH is very low).re 3 shows the structures of the C18 fatty acids with theing degree of unsaturation studied. Stearic acid is com-saturated and consequently has a straight chain that caned tightly in a monolayer. Elaidic acid has a double bondumber 9 carbon, but because the double bond is trans thele remains straight and can still pack well in a monolayer.linoleic, and linolenic acids have increasing numbers ofble bonds, resulting in chains that kink and bend at then of the cis double bond (2). These molecules, therefore,e more and more difficult to pack easily in a monolayer.ric acid has a limiting area of approximately 20 A2) and packs into a solid state. Elaidic acid, even though iticely, does not become fully condensed like stearic acid

e of its double bond. Elaidic acid has a limiting area of ap-ately 3133 A2 (27). Oleic, linoleic, and linolenic acidspack as tightly and as a result form a liquid expandedr which the limiting areas are approximately 41 A2 for

cid (18, 28) and 48 A2 for linoleic acid (28, 29). The lim-ea for linolenic acid is a difficult value to obtain becauseinherent characteristics of the linolenic acid molecule.olecule contains three unsaturated bonds and a carboxylTogether these polar groups seem to be sufficient to al-molecule to leave the spread film when compressed to ant pressure. Therefore, calculations of area per moleculecorrectly assume a higher number of molecules in theyer and result in a low value of area per molecule (Fig. 229).

FIG.the pKaair/wate

Figuobtainethe sol(Fig. 4it is clcrystaland crof the(Figs.endpoi

Figufatty acstearicclosely(pKa =acid (p. A strong iondipole interaction among the carboxyl groups nearvalue decreases the intermolecular distance and area/molecule at theinterface.

re 4 is a schematic diagram of a typical pH titration curved for fatty acids (30). When the pH of the solution is high,tion is clear and contains only soluble potassium salts). As hydrochloric acid is added, the pH decreases untilse to the pKa of the fatty acid. At this point (Fig. 4b),begin to appear in the solution. The pH then levels off

stals continue forming (Fig. 4c) until total protonationatty acids and precipitation of fatty acid crystals occurd and 4e). The pKa is calculated from the neutralizationt shown in Fig. 4.

re 5 shows the results of the titrations of the various C18ids with hydrochloric acid. As can be seen in this figure,acid has the highest pKa value (pKa = 10.15), followed

by elaidic acid (pKa = 9.95), and then oleic acid9.85), linoleic acid (pKa = 9.24), and finally linolenic

Ka = 8.28).

204 KANICKY AND SHAH

The rmanner

to be exof the syEven whlong-chters in o

FIG. 4. ScpKa the soluti(c) until total chematic diagram of a typical C18 fatty acid pH titration curve and the degree of aggregation vs pH expected in the aqueous solution. (a) At pH >on is clear and contains soluble potassium salts. (b) At pH pKa crystals begin to appear in solution. The crystals keep forming at lower pH valuesonversion to insoluble fatty acid (d or e).FIG. 3. Molecular structures of stearic, elaidic, oleic, linoleic, and -linolenic acids.

esults are not surprising if they are interpreted in thedescribed above. These C18 fatty acids have a tendencypelled from water in order to minimize the total energystem (i.e., to maximize entropy of the water molecules).en slightly soluble in the form of potassium salts, these

ain fatty acids arrange themselves into oligomers or clus-rder to minimize the free energy of the system. Stearic

acid, with its straight alkyl chain, packs much tighter both inan adsorbed monolayer and in the small aggregates formed insolution. Therefore it is more difficult for hydroxide ions in so-lution to strip away the acids proton. On the other end of theseries, linolenic acid contains many kinks in its chain causedby three cis double bonds. These kinks prevent the moleculesfrom packing closely, as is shown in Fig. 6. Elaidic acid, with

EFFECT OF UNSATURATION ON pKa OF FATTY ACIDS 205

FIG. 6.and the inFIG. 5. pKa values vs degree and nature of unsaturation of C18 fatty acids, obtained by acidbase titration at 20C.

Schematic representation of C18 fatty acid monolayers at the air/water interface. Note the effect of the degree of unsaturation on the area per moleculetermolecular distance, D, in the spread monolayers.

206 KANICKY AND SHAH

TABLE 1

18 Fatty Acids

C18 iac

Stearic aElaidic aOleic aciLinoleic-Linole

a Sigmb Calcc Refsd Ref.e Refsf Refs

its tranbond csults inthese mEven ining, anresult iFig. 6.

Tablintermolayer, aIt is inspondssense s

energythe solistearicincreaspoint tetion. Fobetweeexists b

Areafatty actainties as to whether or not the fatty acids present at the surfaceare actuformedlayer pas X-raBrewsttweenhas beespreadphase tequalsistic ofteractio(Fig. 10

m

o

a

a

o

iso

h

53, 1016 (1949).

ally present in uniform spread molocules or if they haveaggregates or islands at the surface. Fatty acid mono-

hase transitions can also be studied using methods suchy diffraction (3136), FTRaman spectroscopy (35), ander angle microscopy (BAM) (36, 37). A correlation be-fatty acid monolayer phase transition and solution pHn shown using the BAM technique (37). Specifically,monolayers of fatty acids show maximum resistance toransition under compression when the pH of the solutionthe pKa , indicating that the L2 phase, which is character-fatty acid monolayers, is stabilized by the iondipole in-ns between carboxylic acid and carboxylate headgroupsof Ref. 37).

9. van Deenen, L. L. M., Houtsmuller, U. M. T., de Haas, G. H., and Mulder,E., J. Pharm. Pharmacol. 14, 429 (1962).

10. Langmuir, J. Am. Chem. Soc. 39, 1848 (1917).11. Heikkila, R. E., Kwong, C. N., and Cornell, D. G., J. Lipid Res. 11, 190

(1970).12. Seelig, A., and Seelig, J., Biochemistry 16, 45 (1977).13. Rao, Y. K., and Shah, D. O., J. Colloid Interface Sci. 137, 25 (1990).14. Fang, H., and Shah, D. O., J. Colloid Interface Sci. 205, 531 (1998).15. Adam, N. K., Proc. R. Soc. London A 99, 336 (1921).16. Adam, N. K., and Miller, J. G. F., Proc. R. Soc. London A 142, 401 (1933).17. Schulman, J. H., and Hughes, A. H., Proc. R. Soc. London A 138, 430

(1932).18. Davies, J. T., and Rideal, E. K., Interfacial Phenomena. Academic Press,

New York, 1961.19. Budavari, S. (Ed.). Merck Index, 12th ed. Merck Research Labs, White-

house Station, NJ, 1996.Selected Physical Properties of C

fatty Structure and degree Melting point Limid of saturation temp. (C)a mon

cid 18:0 6971cid 18:1; (trans)9 4445d 18:1; (cis)9 1314acid 18:2; (cis)9,12 5 1nic acid 18:3; (cis)9,12,15 11 10aAldrich catalog.

ulated from limiting area of monolayer data.. 27 and 28.27.. 18 and 28.. 28 and 29.

s double bond, packs much closer than its cis doubleounterpart oleic acid. As can be seen in Fig. 5, this re-a higher pKa value as well. One must keep in mind thatolecules are always in a state of thermal kinetic motion.a spread monolayer, fatty acid tails are spinning, rotat-

d colliding with the adjacent molecules. These collisionsn increased intermolecular distance, D, as illustrated in

e 1 contains data on melting point temperatures (Tm),lecular distances between molecules in a spread mono-nd the observed pKa values of the fatty acids studied.

teresting to note that the decrease in pKa values corre-to a decrease in melting point temperatures. This makesince the melting point temperature is an indication of therequired to disrupt the molecular packing of crystals ind phase. For molecules that can pack very closely (e.g.,acid), Tm is relatively high. As degree of unsaturationes, Tm decreases and so too does pKa . However, meltingmperature and pKa do not have such a strong correla-r example, there is an approximate 30C difference in Tmn elaidic acid and oleic acid, while only a 0.1 differenceetween their pKas./molecule and intermolecular distance calculations ofid monolayers made from -area curves leave uncer-

In sufirst timues ofthe pKlar distintermo

The auCenter foEEC 94-0

1. ShiaA., P

2. DavPres

3. ShiaGreg

4. Shah5. Shah6. Shah7. Sing8. Schnting area of Intermolecular distance Observedolayer (A 2) (A ) in monolayer b pKa

20c 4.47 10.153133d 5.575.74 9.95

41e 6.40 9.8548 f 6.93 9.24 8.28

mary, the results presented in this paper report for thee the effect of the degree of unsaturation on the pKa val-leic, linoleic, and linolenic acids and further correlatevalue of a long-chain fatty acid with the intermolecu-

nce between the fatty acid molecules. The greater thelecular distance, the lower the pKa value of the acid.

ACKNOWLEDGMENT

thors convey their sincere thanks to the NFS Engineering Researchr Particle Science and Technology at the University of Florida (Grant2989) for partial support of this research.

REFERENCES

, S. Y., Chhabra, V., Patist, A., Free, M. L., Huibers, P. D. T., Gregory,atel, S., and Shah, D. O., Adv. Colloid Interface Sci. 74, 1 (1998).es, J. T., and Rideal, E. K., Interfacial Phenomena, 2nd ed. Academic, New York, 1963., S. Y., Patist, A., Free, M. L., Chhabra, V., Huibers, P. D. T.,

ory, A., Patel, S., and Shah, D. O., Colloids Surf. A 128, 197 (1997)., D. O., and Schulman, J. H., J. Lipid Res. 6, 341 (1965)., D. O., and Schulman, J. H., J. Lipid Res. 8, 215 (1967)., D. O., and Schulman, J. H., J. Colloid Interface Sci. 25, 107 (1967)., C. P., and Shah, D. O. , Colloids Surf. A 77, 219 (1993).

eider, V. L., Holman, R. T., and Burr, G. O., J. Phys. Colloid Chem.

EFFECT OF UNSATURATION ON pKa OF FATTY ACIDS 207

20. Sutherland, I. O., Comprehensive Organic Chemistry: The Synthesis andReactions of Organic Compounds, Vol. 2. Pergamon, Elmsford, NY, 1979.

21. Kanicky, J. R., Poniatowski, A. F., Mehta, N. R., and Shah, D. O., Langmuir16, 172 (2000).

22. Heikkila, R. E., Deamer, D. W., and Cornwell, D. G., J. Lipid Res. 11, 195(1970).

23. Peters, R. A., Proc. R. Soc. London A 133, 140 (1931).24. Christodoulou, A. P., and Rosano, H. L., Adv. Chem. Ser. 84, 210 (1968).25. Tanford, C., The Hydrophobic Effect: Formation of Micelles and Biolog-

ical Membranes, 2nd ed. Wiley, New York, 1980.26. Israelachvili, J. N., and Pashley, R. M., Nature 300, 341 (1982).27. Hifeda, Y. M., and Rayfield, G. W., J. Colloid Interface Sci. 104, 209 (1985).28. Tomoaia-Cotisel, M., Zsako, J., Mocanu, A., Lupea, M., and Chifu, E.,

J. Colloid Interface Sci. 117, 464 (1987).

29. Peltonen, J. P. K ., and Rosenholm, J. B., Thin Solid Films 179, 543 (1989).30. Rosano, H. L., Breindel, K, Schulman, J. H., and Eydt, A. J., J. Colloid

Interface Sci. 22, 58 (1966).31. Ueno, S., Miyazaki, A., Yano, J., Furukawa, Y., Suzuki, M., and Sato, K.,

Chem. Phys. Lipids 107, 169 (2000).32. Peng, J. B., Foran, G. J., Barnes, G. T., and Gentle, I. R., Langmuir 13,

1602 (1997).33. Peng, J. B., Barnes, G. T., Gentle, I. R., and Foran, G. J., J. Phys. Chem. B

104, 5553 (2000).34. Dutta, P., Colloids Surf. A 171, 59 (2000).35. Tandon, P., Forster, G., Neubert, R., and Wartewig, S., J. Mol. Struct. 524,

201 (2000).36. Fainerman, V. B., Vollhardt, D., and Johann, R., Langmuir 16, 7731 (2000).37. Johann, R., Vollhardt, D., and Mohwald, H., Langmuir 17, 4569 (2001).

INTRODUCTIONFIG. 1.

EXPERIMENTALRESULTS AND DISCUSSIONFIG. 2.FIG. 3.FIG. 4.FIG. 5.FIG. 6.TABLE 1

ACKNOWLEDGMENTREFERENCES