The Behavior of Nonionic Surface Active Agents in Salt Solutions
The Behavior of Nonionic Surface Active Agents in Salt Solutions
The Behavior of Nonionic Surface Active Agents in Salt Solutions
IN SALT SOLUTIONS
INTRODUCTION
This investigation was carried out to determine whether alkali metal
ions and alkaline earth metal ions have different effects on the properties
of solutions of nonionic surface-active agents. It had already been observed
(8,9) that in the presence of the alkaline earth metal ions these surface-
active materials are capable of stabilizing suspensions of clay in oil and
water media for use as drilling fluids, whereas the alkali metal ions had a
coagulating effect.
The nonionic surface-active agents used in this investigation are those
produced by reacting hydroxylie or acidic water-insoluble compounds
with ethylene oxide under such conditions that the ethylene oxide con-
denses to form a polyethylene oxide chain attached to the nonpolar
residue. The products have formulas of the type: R--O--(C2I-I40)~--
C2HsOH, where n may have average values between 3 and 20, and R
may be (a) an alkyl-aryl residue, such as the iso6ctyl-phenyl group in
Triton X-100 (Rohm and Haas Co.) ; (b) a fatty acid residue, such as the
mixed fatty and rosin acids derived from Tall Oil in Renex (Atlas Powder
Co.) ; or (c) a partial ester of hexitol anhydride and a fatty acid, in which
case there are several polyethylene oxide chains attached to each molecule
of the ester, as in the Tweens (Atlas Powder Co.).
These surface-active agents owe their solubility in water to the hydra-
tion of the polyethylene oxide chain; solubility increases and surface
activity decreases with increasing length of the chain. At temperatures
above 60°C. solutions of most of these compounds become turbid due to
dehydration and consequent association of the molecules into larger col-
loidal aggregates. It has been demonstrated that the osmotic coefficient of
solutions of these materials exhibits similar behavior to that of colloidal
electrolytes (ii); and,.further, that solutions of these nonionic compounds
solubilize w a t e r - i n s o l u b l e dyes (12), a n d a long x - r a y s p a c i n g is o b t a i n e d
1Part of this work was performed under the auspices of a National Lead Company
(Baroid Division) Fellowship, and was submitted as par~ of the Master of Science
thesis. Present address: Harvard University, Cambridge, Mass.
2 Present address: Crest Laboratories, Burbank, Calif.
223
224 DOSCtIER, MYERS AND ATKINS
~ O]~gNaCl/cc.
1
L25 QOSgNaCl/cc.
0.06
0.75
o.~0
0.2s
o
o 002 004 o.os o.oe
(Fig. 2), significant differences are observed in the slopes of the curves.
The small slope for the sodium chloride solution is indicative of the fact
that the sodium chloride solution is a comparatively poor solvent for the
Triton X-100, whereas concentrated calcium chloride solutions appear to
be better solvents than pure water (7). The dissymmetry coefficients at
concentrations of Triton X-100 above 0.5% are zero, but at lower concen-
trations were observed to vary in an erratic manner. It is therefore not
feasible at this time to extrapolate the turbidimetric results to zero concen-
tration and report molecular weights of the aggregates in the various
solutions.
The viscometric results, shown in Fig. 3, are in accord with those
obtained turbidimetrically. The viscosities of solutions of all of the non-
ionic surface-active agents are abnormally raised by moderate concentra-
226 DOSCHER, MYERS AND ATKINS
;0.¢
•o 015g.C=~
4.G
I ! ! m ~0 m .
0,02 0,04 0 6 0,08 0,10
tions of sodium chloride, but calcium chloride has only a slight effect.
Other salts were studied in less detail than the sodium and calcium chlo-
rides: the nitrates of these cations, on the basis of chemical equivalents,
have the same effect as the chlorides; lithium is considerably less effective
in raising the turbidity and viscosity than sodium; and potassium Salts
are only slightly less effective than sodium salts. Barium and magnesium
produce slightly higher viscosities and turbidities than does calcium, but
aluminum salts give somewhat lower values.
5.0 .~ael
4.o ~ ~aCt
I.G
4.o ~Cl
TRITON X-lO01 2 X
1.0
! I I
0 5.0 IOJO 15.0
SoluSilization
The solubilization of the dye, Orange OT, was studied in solutions of
Triton X-100 containing calcium and sodium chlorides. The determina-
tions were carried out by shaking a suspension of the solid dye in the
solution, settling and centrifuging at approximately 2000 ) g, and finally
analyzing the supernatant liquid with a colorimeter. The results are shown
in Fig. 4. Both sodium and calcium raise the solubility in 0.5% solutions
of Triton X-100, but sodium chloride does so only slightly more than the
44
, /,,oo, .,
FIG. 4. Solubility of Orange OT, rag./100 ee., vs. salt concentration, weight per cent.
TABLE I
Surface Tension in Salt Solutions
(dynes/cm.)
Rencx, 0.02% 40.3 I 39.8 39.3 39.1 40.3 40.4 40.9 41.3 42.6
2.0% 38.5 38.5 38.4 38.4 38.5 39.2 40.0 40.8 41.2
6.0% 38.0 38.0 38.1 37.9 38.0 38.1 38.4 39.5 41.0
,~¢ L
!
~c ~ CaCt2
!
e.c ffl
2£
/
0 .~..-.o,--,-P--'~ ° m m o I I
LO 2.0 3.0 4D
FIG. 5. Salt concentration, weight per cent, to coagulate gold sol vs. Renex
concentration, per cent X 104.
ide. The use of kaolin had several advantages: observations could be made
in a mieroelectrophoretic cell (3) with normally incident light, and the
kaolin has definite exchange sites at which cations may be held. The results
at various concentrations of Renex and added salts are shown in Fig: 6.
Z.O
Ca +÷ lO'lM
0 I z|o ! I | I I I
4.0 6.0 8.0
6,6(
620
- s~
5.$¢
'% , x O M UNDERSATURATION
5.4(:
5.0©
4,60
420 a a * | * * * * , | •
o 4.0 e~) ,~o ,6~ zoo
FIG. 7. Solubility of CaSO4.2H20 (g. Ca/g. H20) × 103 vs. Renex concentration,
weight per cent.
Unfortunately, the salts of the alkali metals salt out the nonionic colloid
before the solution is saturated with the salt, so that direct comparison of
the decrease in solubility of sodium and calcium salts cannot be made.
When 30% or more calcium chloride is added to moderately concen-
trated (about 5%) solutions of several of the nonionic surface-active
materials, a solid material slowly precipitates. The two materials which
produce such a precipitate most readily are Renex and JH-331 (a homo-
log of Triton X-100, which contains a greater number of moles of ethylene
oxide/mole nonpolar residue). The precipitate is readily soluble in water,
indicating that it is not a product of hydrolysis. It is also soluble in alco-
hols (n-amyl was the highest used), but is insoluble in acetone, dioxane,
and ethyl ether, whereas the original colloids are themselves soluble in
all of these liquids. The precipitates were filtered, recrystallized from
isopropanol, and finally washed with ether to remove any excess isopro-
panol. After two or three such recrystallizations, the solid loses its amor-
phous appearance and a pale-yellow nonhygroscopic powder is obtained.
Analysis of the solid revealed that calcium and chlorine are present in
NONIONIC SURFACE ACTIVE AGENTS 231
The diffraction pattern obtained from the anhydrous material was not as sharp
as that obtained from the system containing 2% water. The halos were, however,
centered at 3.9, 4.7, and 55.
bVery weak.
sharp spacings were restored and the long spacing increased. The dehy-
drated material was also dissolved and reprecipitated in absolute iso-
propanol; however the solid gave no long spacing and very faint short
spacings at d/n Values of 4.30 and 5.98. These two spacings were deter-
mined to be due t o the presence of a calcium chloride-isopropanol com-
plex by dissolving calcium chloride in absolute isopropanol and evaporat-
ing the solution to dryness at room temperature. The white, crystalline
solid gave spacings of 4.30 and 5.98.
DISCUSSION OF I:~ESULTS
The viscometric and turbidimetric data demonstrate that the nonionic
surface-active agents are salted out of aqueous solution by sodium chlo-
ride. Desolvation of the colloidal molecules and aggregates is presumably
the cause of this phenomenon, and judging by the increase in viscosity,
micellar growth must accompany the desolvation. The observed order of
salting out for the alkali metal ions, viz., Na, K, Li, is not the usual
order for the lyotropic series (13) in which lithium precedes sodium. The
observed effect of the heavier metal cations is also contradictory to the
usually accepted order for the lyotropic series. These cations, which have
relatively high heats of hydration and high electrical charge densities,
are usually far stronger coagulants than the alkali metal ions. It may be
concluded therefore that the alkaline earth metal ions interact with the
polyethylene oxide chain of the nonionic colloids, and thereby product a
salting-in rather than salting-out effect. This conclusion is further sub-
stantiated by the surface tension data, which show that at high ratios of
calcium chloride to colloid the surface tension is increased; and also by
the fact that the turbidity in these calcium chloride solutions is actually
less than that observed in pure water.
In contrast to the above results the solubilization data show very
little difference i n the effects of the calcium and sodium salts on the
ability of solutions containing more than 0.5% of Triton X-100 to solu-
bilize a water-insoluble dye. This result, together with the turbidimetric
and viscometric results, suggests that the organization within the colloidal
micelles and the fraction of the surface-active material in the form of
micelles is not affected differently by the two salts, but that the salts
influence only the size of the micelles and the extent of solvation 'of the
micellar surfaces. The salts may cause an appreciable difference in the
amount of aggregated material at concentrations of colloid below 0.1%,
when aggregation in pure water may not be as complete, but the extremely
high suspending power of the surface-active agents in this low concen-
tration range has prevented the determination of the amount of dye
solubilized.
NONIONIC SURFACE ACTIVE AGENTS 233
solvate shell of the calcium ion may explain the observed increase in
activity coefficient and solubility of the calcium sulfate.
The increase in electrophoretic mobility upon increasing the Renex
concentration in dilute calcium solution suggests that the Tween 80
adsorbed onto the clay particle at low concentrations is stripped from the
clay surface at higher Concentrations. This may be brought about by the
growth of micelles in the bulk of the solution at larger concentrations and
their competition with the clay for the calcium ion (20). It may be noted
at this point that very stable petroleum oil-in-water emulsions have been
prepared at concentrations of Tween 80 of 0.01%, but that such emulsions
were very unstable at higher concentrations and at zero concentration
(18).
The preceding results and their interpretation indicate that the efficacy
of these nonionic materials in improving detergency in hard water (4) is
due to the following factors: (a) formation of soluble complexes with
calcium, thus preventing the formation of insoluble calcium soaps, and
(b) enhanced formation of a protective layer of the nonionic colloids on
the solid surfaces at which calcium arid other heavy metal cations are
adsorbed.
SUMMARY
1. Sodium chloride and other salts of the alkali metals have been
found to salt-out nonionic surface-active agents, whereas soluble calcium
salts and other soluble salts of heavy metals salt-in these colloids.
2 : A Crystalline complex of a nonionic surface-active agent, calcium
chloride, and water has been isolated.
3. The nonionic surface-active agents.appear to coordinate with hy-
drated calcium ions both in solution and at solid-liquid interfaces.
REFERENCES
1. AM. Soc. TASTINGMATEmALS, Card Index. Philadelphia, 1945.
2. ATLASPOWDER CO., private communication.
3. ABRAMSON,H. A. A., MOrER, L. S., AND GomN, M. H., Electrophoresis of Proteins.
Reinhold Publishing Corp., New York, 1942.
4. BARKE~, G. E., J. Am. Oil Chemists' Soc. 26, 304 (1949).
5. BOGARgOS,I-I. F., AI'~DLYNCH, C. C., J. Phys. Chem. 47, 650 (1943).
6. BRINTZINGElt,H., AND SCttALL,A., Kolloid-Z. 71, 300 (1935).
7. DEBrE, P., Ann. N. Y. Acad. Sc/. 51, 575 (1949).
8. DOSCH~R, T. M., Oil Gas J. 48, 75 (1945).
9. DOSCHER, T. M., J. Phys. & Colloid Chem. 53, 1362 (1949).
10. FUOSS, 1%. 1V[.,AND Kn~cSS, C. A., J. Am. Chem. Soe. 55, 2387 (1933).
11. GONICK,E., AND MCBAIN, d. W., J. Am. Chem. Soc. 69, 334 (1947).
12. G~EEN, A. A., AND ~%/~cBAIN,J. W., J. Phys. Chem. 51, 286 (1947).
13. KRUYT, H. 1%., Colloid Science, Vol. II, pp. 564 ft. Elsevier, New York, 1949.
NONIONIC SURFACE ACTIVE AGENTS 235
14. McBAIN, J. W., Colloid Science, pp. 133 ft. Heath, New York, 1950.
15. MARK, ]:I., in Chemical Architecture, Chap. V. Interscience, New York, 1948.
16. MARSDEN, S. S., AND McBAm, J. W., J. Phys. & Colloid Chem. 52, 110 (1948).
17. MYErs, G., M.S. Thesis, Univ. of Southern Calif., 1950.
18. MYERs, H. E., Soil Sci. 44, 331 (1937).
19. PAULI, W., A~) SzPE~, L. S., Trans. Faraday Soc. 35, 1316 (1939); and PAULI, W.,
AND Zn~TN~R, H., ibid. 35, 1234 (]939).
20. REIC~ENBERO, D., Trans. Faraday Soc. 43, 467 (1947).
21. ]~EMSOW,I., Biochem. Z. 218, 86, 170 (1930).
22. ZOm~MA, H. H., A~D WAT~aS, G. W., Ind. Eng. Chem., Anal. Ed. 13, 312 (1941).