Letters to the Editors Wissenschaftliche Kurzberichte

Letters to the Editors (Short Communications) are to ensure the priority of your findings. The responsibility for its contents is absolutely yours. These papers will be published approximately within 2-6 weeks after receipt. Such papers shouid not exceed 2 pages typewritten. Please, do write your manuscript as clear as possible, especially with respect to formulas! Please, do avoid figures! Should they be inevitable for understanding, please, attach clear drawings by India ink, ready for reproduction. You will receive 75 reprints free of charge. Papers should be submitted directly to: Prof. Dr. F. H . Miiller, Haselhecke 26, D-3550 Marburg-Marbach. Kurz-Mitteilungen sichern rasch die Priorit~it. Die Verantwortung ftir den Inhalt trSgt der Autor. Erscheinungsterrain ca. innerhalb 2-6 Wochen nach Einreichung. Umfang der deutlich (Formeln!) geschriebenen Manuskripte bis zu 2 Maschinenseiten. Abbildungen vcrmeiden. Sonst klischierf~ihige Zeichnungen mit Beschriftung in weichem Blei. Sonderdruckfreiexemplare wie bei Originalarbeiten. Manuskripte an: Prof. Dr. F. H . Miiller, Haseihecke 26, D-3550 Marburg-Marbaeh. Herausgeber und Verlag

Colloid & Polymer Sci, 258, 973-976 (1980) 1980 Dr. Dietrich Steinkopff Veriag, Darmstadt ISSN 0303-402X /ASTM-Coden: CPMSB (formerly KZZPAF)

Institut Max von Laue - Paul Langevin 156X, Grenoble 38042 (France)':" and School of Chemistry, University of Bristol, Bristol':":" and Royal Signals and Radar Establishment, Malvern, Worcester***

The structure of a microemulsion droplet

D. J. Cebula*, L. Harding**, R. H. Ottewill**, and P. N. Pusey***
With 3 figures and 1 table (Received May 14, 1980) Optically transparent emulsions, usually termed microemulsions, were first investigated by Hoar and Schulman (1) in 1943 using both visual and light scattering observations. Since that time these systems have attracted considerable attention (2) and there has been some controversy about the nature of microemulsions and the origin of their stability as colloidal dispersions (3, 4). Recently we have undertaken a number of investigations on microemulsion systems with two basic objectives in view, (a) to determine the structure of microemulsion droplets and (b) to examine the physical basis of the interactions between the droplets in order to determine the reason for their stability as colloidal dispersions. Our experimental approach has been based upon the use of three scattering techniques, namely, time average light scattering, photon correlation spectroscopy (PCS) and small angle neutron scattering (SANS). The purpose of this note is to show how the use of the latter two techniques can provide fundamental information on the structure of microemulsion droplets. A more detailed report elaborating the use of these techniques to provide fundamental information on interactions will be published elsewhere (5). The technique of PCS is now well developed (6) and can be used to determine the mean diffusion coefficient of the particle, ZJ, and hence the effective mean hydrodynamic
W 145

radius R h. Under the conditions of the present experiments the measured field correlation function gl(Q,T) is expected to be close to a single exponential, namely, g l ( Q , ~)0c exp ( - D Q 2 r) [1]

where r is the correlation delay time and Q is the scattering vector defined as, 4:r . 0 Q = T sm ~ [2]

where ~. = the wavelength of the light in the microemulsion and 0 the scattering angle. In general, due to interparticle interactions, D will be a function of particle volume fraction 0. However, there is considerable evidence that, when the interactions are short-range (i. e. effectively "hard-sphere"), D does not deviate markedly from its zero concentration value for q5 < 0.05 (7). Thus the Stokes-Einstein equation can be used in the form,

R h = kT/6x~

[3]

974

Colloid and Polymer Science, Vol. 258 No. 8 (1980)

used for the measurements described here had the following volume fractions, water = 0.036, toluene = 0.865, hexanol = 0.073 and potassium oleate (as oleic acid) = 0.025. The results obtained from SANS experiments are shown in figure 1 as plots of I(Q) vs Q. One set was obtained using D 2 0 in the system in combination with H-toluene and the othm" set using D-toluene as the dispersion medium with HzO in the microemulsion droplet. The experimental data were obtained at the Institut Laue-Langevin, Grenoble, using instrument D17, with A = 12.2 ~ and with a sampledetector distance of 1.41 m. For a dispersion of identical spherical particles the intensity of scattered neutrons is given by,

where k = Boltzmann's constant, T = absolute temperature and/7 the viscosity of the solvent, taken to be 5.90 x 10 - 4 Pa.s at T = 293 K, can be used to determine R h. Intensity fluctuations in the light scattered by the microemulsion were observed at 0 = 90 relative to the incident beam; the light source was a Kr ion laser operated at X0 = 6471 ft. in vacuo. Thus, in equation (2), B. = 2~/n where n = the refractive index of the sample; n was measured to be 1.4810. For the PCS measurements, a 96-channel "one-bit" correlator was operated in the single-scaled mode (6). Data were analysed by the method of cumulants so that l), and hence R h, were obtained from the first cumulant and the normalised second cumulant; the latter figure describes departure from single exponential behaviour and was typically 0.06. The technique of small angle neutron scattering (SANS) provides a sensitive means of measuring particle size (8, 9) by measurement of the intensity of scattered neutrons, I(Q), as a function of the scattering vector, Q, as defined by equation (2), where for this technique Z = the wavelength of the neutron beam. The intensity of the scattered beam is directly related to the mean coherent scattering length, b, as given by

I(Q)= A P(Q)S(Q)

[5]

where A is a constant (dependent on geometry, scattering lengths "etc), P(Q) is the single-particle scattering factor and S(Q) the structure factor. For homogeneous particles of radius R (8), P(Q) is given by

P(Q) =
with IF(x)r =

IF(QR)t 2

[6]

b = ~ bi/V

[4]

where Y] b i = the sum of the coherent scattering lengths of the constituent atoms of the scattering molecule and V = the molecular volume. In view of the large difference in coherent scattering length between hydrogen and deuterium, it is possible to carry out experiments to examine separately the size of the aqueous core of the emulsion droplet with the use of D 2 0 and the size of the emulsion droplet using a deuterated oil phase. The system chosen for the investigation consisted of potassium oleate, hexanol, water and toluene. The region of the four-component system in which water in oil microemulsions were formed was carefully established (10) and a system with a low water volume fraction was chosen for study in order to minimise particle interactions. The system

(sin x - x cos x).

[7]

The structure factor is determined by the interparticle interactions~xSince S(Q) is unity at very low volume fractions, 0, where interactions can be neglected, in this region I(Q) o~ P(Q). For hard-sphere interactions with effective radius RI (generally with RI ~> R), S(Q) is given, to first order in ~, by (8),

S ( Q ) = 1 - 8q~IF(2Q RI) + . . .

[8]

4 x ~xx ~3 N with where the hard sphere volume fraction 01 is ~ N = number of particles per c m 3. Combining equations [6] and [8] i n e q u a t i o n [5] and expanding in powers of Q2 gives

120

L~ L~L~

I ( Q ) oc (1 - 8q5,)
90 I{Q) 60

1 -

8~bi

\ R2

+ 1

"'"

[9]

30

I
0.006 O I~ -~

0.012

Fig. 1. SANS results plotted in the form of ](Q) against Q for microemulsions containing a water volume fraction of 0.036. A, H 2 0 core in D-toluene; o, D 2 0 core in H toluene; -, least squares line as fitted to the data points

It is immediately apparent that the commonly used approach of determining R using the Guinier approximation (8), i. e. as ~ r ~ 0 , will, if interactions are ignored, give an apparent radius some tens of percent smaller than its true value for ~b~~ 4) = 0.03. This is a surprisingly large effect for such a small volume fraction. However, by evaluating equation (8) (or the more accurat~ Percus-Yevick expression for S(Q) for hard-spheres (11)) we note that, for ~ = 0.03 and Q ~> 2/R, S(Q) is within 3% of unity and I(Q) -~ AP(Q). Thus, data were fitted by the method of least-squares directly to equation (6) in the range 2/R < Q <- 4/R. The effect of polydispersity of particle size was investigated but data fitting was hardly affected. For Q ~< 4/R measured values of I ( Q ) became very small and it was judged that, due to experimental uncertainties, data m this range gave unreli-

Cebula et al., The structure o f a microemulsion droplet

975

Table 1. Results of scattering experiments on microemulsion droplets Technique SANS SANS PCS System D20/H-toluene H20/D-toluene H20/H-toluene Parameter Core radius - R c Shell radius- R s Hydrodynamic radius - R h Value//~. 40 -+ 2 48 -+ 2 52 -+ 3

I"

60

~D20
--~-D-Toluene

40

E
u

20
x

........
. . . . . . . .

l ..,.--H-Toluene

........

-~-~-Oleic Acid

-20

I
6

I
4 0

1 ........ I "--H~
4 g Radiuslnrn 8

Fig. 2. Schematic diagram to illustrate the coherent Scattering lengths of the components of the microemulsion and to demonstrate the changes in contrast produced. Left-hand s i d e - D 2 0 core in H-toluene; Right-hand s i d e - I-I20 in Dtoluene

able fits. The results obtained are recorded in table 1, wherein they are compared with" the hydrodynamic radius obtained from PCS. Figure 2 gives a schematic diagram indicating that with D 2 0 in the system, it is the D 2 0 core which has the highest coherent scattering length whereas with a deuterated oil, the oil contrasts the hydrogenated material, i. e. the water drop and the surface active materials. It is generally accepted that, for a diffusing species with a well-defined outer boundary, the hydrodynamic radius R h measured by PCS is a few .~ larger than the true radius because of a layer of solvent which moves with the particle However, in the present case, where some flexibility is indicated in the outer oleate chains (see below), R h could be comparable to, or even somewhat less than, the maximum radius of the diffusing species The values obtained by SANS, using D 2 0 , correspond to the radius of the watercore of the droplet, R o although it is likely that the hydrated head groups of the surface active molecules will also be contained in this. The results indicate a distinct water phase and are not compatible with the idea of a swollen micelle. The value obtained using D-toluene as the dispersion medium, R s, iies in between the hydrodynamic radius R h and the core radius R c. The difference R s - R c is ca. 8 A, about the length of a hexano] hydrocarbon chain and head group. This suggests that this region, R~ - R c is composed of a close-packed monolayer of hexanol and oleate molecules The experiments in D-toluene, however, suggest strongly

that penetration of toluene occurs into the outer regions of the microemulsion droplet, i. e., between the spaces of the protruding oleate chains, but that it penetrates only slightly, if at all, into the close-packed shell region containing the hexanol molecules. These experiments indicate therefore that the structure of a microemutsion droplet must be close to that represented schematically in figure 3. This picture also suggests that the outer regions of the oleate chains in the toluene phase maintain some flexibility and that this aids the colloid stability by the mechanism of "steric stabilisation" which in the present context would be close to a hard-sphere interaction (12). In fact, at higher volume fractions of water, considerable interaction occurs between the droplets and a detailed study has been made of this topic (5, 13, 14, 15). The model shown in figure 3 represents a time-average structure of the microemulsion droplet. Clearly the real structure will be a dynamic one and thus there is likely to be, as in micellar systems (6), a continuous interchange between the surface active molecules in the interracial region and both the oil and water phases. For the analysis of the data the boundaries between the core and the close-packed shell of hexanol and oleate molecules and between the shell and the loosely packed outer layer of oleate chains have been taken as sharp This is an oversimplification and a more complicated model can be developed to give a more sophisticated interpretation. O n present evidence, however, it is doubtful whether this will change substantially the modei of a microemulsion droplet proposed in this communication.

976

Colloid and Polymer Science, Vol. 258. No. 8 (1980)

= Toluene

)oo c3

Hexanol Oleate

'~ ,,~

Fig. 3. Schematic diagram of the structure of a microemulsion droplet which would be consistent with the SANS and PCS experimental results

Acknowledgements
We wish to express our thanks to the Science Research Council foi~support of this work and to the Institut LaueLangevin for neutron beam facilities.

References
1) H~,;r, T. P.,J. H. Schulman, Nature t52, 102 (1943). 2) Prince, L. M., Microemulsions, Theory and Practice (Academic Press, New York 1977). 3) Overbeek, J. Th. G., Faraday Disc. Chem. Soc. 65, 7 (1978). 4) Ruckenstien, E., J. C. Chi, J. Chem. Soc. Faraday II, 11, 1960 (1975). 5) Cebula, D.J., R. H. OttewiU, P. N. Pusey, J. Ralston, to be published. 6) Cummins, H. S., E. R. Pike, Photon Correlation and Light Beating Spectroscopy (Plenum, New York 1974); Photon Correlation Spectroscopy and Velocimetry (Plenum, New York 1977). 7) Newman, J., J. L. Swinney, S. A. Berkowitz, L. A. Day, Biochemistry 13, 4832 (1974); G. K. Batchelor, J. Fluid Mech. 74, 1 (1976). 8) Guinier, A., G. Fournet, Small Angle Scattering of X-rays (Wiley, New York 1955). 9) Jacrot, B., Rep. Prod. Phys. 39, 911 (1976).

10) Harding, L., B. Sc. thesis, University of Bristol (1979). 11) Ashcroft, N. W.,J. Lekner, Phys. Rev. 145, 83 (1966). 12) Vrij, A., E. A. Nieuwenhuis, H. M. Fijnaut, W. G. M. Agterof, Faraday Disc. Chem. Soc. 65, 101 (1978). 13) Dvolaitzky, M., M. Guyot, M. Laq~ies, J.P. Le Pesant, R. Ober, C. Sauteray, C. Taupin, J. Chem. Phys. 69, 3279 (1978). 14) Cebula, D.J., L. Harding, R, H. Ottewill, to be published. 15) Cebula, D, J., D. Myers, R.H. Ottewill, to be published. 16) Aniansson, E.A. G., S.N. Wall, M. Almgren, H. Hoffmann, !. Kielmann, W. Ulbricht, R. Zana, J. Land, C. Tondre, J. Phys. Chem. 80, 905 (1976).

Authors' address :

R. H. Ottewill, School of Chemistry, University of Bristol, Cantock's Close, Bristol BS8 ITS England