Direct observation of droplet structure in a vitrified microemulsion - The

E. A. Hildebrand, I. R. McKinnon, and D. R. MacFarlane. J. Phys. Chem. , 1986, 90 (12), pp 2784–2786. DOI: 10.1021/j100403a046. Publication Date: Ju...
0 downloads 0 Views 2MB Size
J . Phys. Chem. 1986, 90, 2784-2786

2784

whole studied range of NBF concentrations supports the pertinence of the chosen thermodynamic model of self-association and the assumed structure of NBF multimers.

Conclusions In the present and two preceding papers we described the results of our extensive study of N-methylacetamide,' 2-azacyclotridecanone,2 and N-benzylformamide in carbon tetrachloride solutions. The three amides manifest some strong similarities: (i) the peptide groups -NHCO- are in the trans configuration, (ii) the values of the dipole moments of their monomers are almost identical (3.8 f 0.1 D), (iii) the equilibrium constants and the thermodynamic parameters (AHand AS) are close to each other (Table 11). On the basis of spectroscopic (and calorimetric) studies one can conclude that in all three cases the self-association process proceeds in a quite similar way. No further information, beside the chain character of the association, can be gained from these methods.

In dielectric studies these three amides differ considerably (Figure 4). Because of the common features listed above, the observed differences in the pap: dependences on concentration must result exclusively from the various polarity of the formed multimers, Le., from their different structure. We have shown how the size of the peptide group substituents is decisive for association as it determines the rotational freedom of some particular elements of the multimer chain. In multimers of 2-azacyclotridecanone large spatial hindrances practically inhibit the rotation around hydrogen bonds, and therefore their structure is "rigid" and linear. In the case of N-benzylformamide,due to free rotation, the multimers have a statistically random conformation, whereas N-methylacetamide makes an intermediate case in which the multimers have an elongated shape.

Acknowledgment. This work was supported by the Polish Academy of Sciences within the framework of Project MR-1.9. Registry No. N-Benzylformamide,6343-54-0.

Direct Observation of Droplet Structure in a Vitrified Microemubion E. A. Hildebrand, I. R. McKinnon,* and D. R. MacFarlane Department of Chemistry, Monash University, Clayton, Victoria, Australia 31 68 (Received: May 20, 1985; In Final Form: January 8, 1986)

The recent development of glass forming microemulsionshas opened the way to direct probing of their structure using electron micrmcopic techniques without the danger of the structure being modified by crystallization. In this work electron micrographs obtained after vitrification and fracturing of the sample are compared with those obtained by normal cold stage electron microscopy.

Introduction Several years ago it was reported'**that glass-forming microemulsions could be prepared by choosing a co-surfactant component of the microemulsion which would also act to inhibit the nucleation and growth of ice in the oil-in-water microemulsion when the temperature was lowered below 273 K. In the cases studied to date this co-surfactant component has always been propane- 1,2-diol (propylene glycol, PG), although there is no reason to expect that other simple polyols would not act similarly. Propylene glycol is known' to render aqueous solutions glass forming at concentrations in excess of 12 mol % PG. As long as the dispersed oil phase is glass forming and provided that there is no lower temperature phase instability with respect to a twophase mixture the microemulsion as a whole can be vitrified. No such phase separation has been observed in the PG-based systems studied to date. For example, a sample cooled quite slowly was obtained in vitrified form without any increase in turbidity,2 and a H,O/PG/Tween 8O/CCl4 (Tween 80 = polyoxyethylenesorbitan monooleate) microemulsion has been reported as stable to at least -90 O C for a number of hours.3 However, at temperatures near the glass transition temperature, Tg,of the higher Tgcomponent, kinetic slowing down of equilibration processes could inhibit any thermodynamically expected phase change. The discovery of these microemulsion systems has opened up the important possibilities that the microemulsion structure can (1) MacFarlane, D. R.; Angell, C. A. J . Phys. Chem. 1982, 86, 1927. (2) Angell, C. A.; Kadiyala, R. K.; MacFarlane, D. R. J . Phys. Chem. 1984, 88, 4593. (3) Angell, C . A., private communication, 1985.

0022-3654/86/2090-2784$01.50/0

now be captured and studied by a variety of techniques including electron microscopy, small angle scattkring, both neutron ana X-ray, and NMR. A recent article4 in this Journal reported the vitrification in glassy microemulsion of three molecular liquids (CS2, benzene, and CClJ which had not previously been vitrified and which, on the basis of bulk glass-forming trends, would not be expected to vitrify. The glass transition temperatures (an isostructural relaxation time point) for these liquids were as expected on the basis of extrapolations from binary solution Tg data, thus indicating that the microemulsion droplets exhibit largely unaltered relaxation properties with respect to their bulk liquids. Also reported were electron micrographs of the quenched microemulsions. These clearly indicated a simple droplet structure. At higher droplet concentrations there was an apparent hexagonal ordering of the droplets in the matrix phase. The electron micrographs just described were obtained by the quenching of a thin (-0.1 Wm) film of the microemulsion on an electron microscope sample screen. It is possible that the ordering observed was the result of surface forces acting on the thin film. Also, because the samples were diluted with water before freezing the compositions studied were not those of the finger region of the pseudoternary phase diagram. In this note we report a series of electron microscope experiments in the finger region of the pseudoternary phase diagram of the same system using two types of freeze fracture techniques (described below). These, we believe, have allowed us to obtain microeraohs which are trulv reoresentative of the bulk microemulsTon'and avoid any artffact's associated with surface inter(4) Dubochet, J.; Adrian, M.; Teixeira, J.; Alba, C. M.; kadiyala, R. K.; MacFarlane, D. R.; Angell, C. A. J . Phys. Chem. 1984, 88, 6727.

0 1986 American Chemical Society

Droplet Structure in a Vitrified Microcmulsion

The Journal of Physicol Chemistry, Vol. 90, No. 12, 1986 2785

h e n 80

i Figure 1. Pseudo-three-componentphase diagram for the system propylene glycol.3H20-Twccn 80-0-xylene (volume %).

Figure 2. (a) Sample holder, (b) double-replicafracture device (i) o p e d and (ii) closed.

actions. However, they do show evidence that the size of the 'droplets" is a function of temperature.

Experimental Section Microemulsion samples were prepared by titrating an initial mixture containing the aqueous phase and &xylene (volume ratio 21) with a solution of Tween SO/o-xylene in the volume ratio 21. The aqueous phase consisted of propylene glycol and water in the mole ratio 1:3. Compositionsin the two-phase regions of the phase diagram (Figure I ) tended to form a more or less stable emulsion which became optically transparent on addition of one or other of the components sufficient to reach the microemulsion region. Freeze fracture samples of the microemulsion were obtained by using two techniques. In the first, a small drop of sample liquid was placed in the specimen holder, vitrified in Freon 22 and stored in liquid nitrogen. The specimens were then transferred to the cold stage of the freeze etch unit (Balzers BAF 300/301), which had been precooled, under vacuum, to -150 OC. The unit was then sealed and reevacuated, and the microtome arm and knife (stainless steel) cooled to -1 80 "C. The temperature of the stage was then increased to -1 10 OC and the sample cut until a flat surface was visible. The knife was then placed over the specimen, allowing sublimation from the cut surface (etching) to proceed for a predetermined length of time (2-3 min), after which the surfaces are coated with platinum/carbon as described below. In the semnd technique, "double-replica freeze fracture", a drop of liquid was deposited on a metal disk. A second disk was then placed on top and the two disks are centered as shown in Figure

Figure 3. Transmission electron micrographs obtained by double-replica freeze fracture. (a. top) Microemulsion sample at 15OOX magnification; (b. middle) Same sample at 15OOOX, (c. bottom) propylene glycol.3H20 sample a t 15000X. All micrographs were recorded on a JEOL

ZOOCXTEM. 2a. A further drop was then placed in the central bore of the specimen "sandwich", any excess liquid being removed with filter paper. The specimen was checked with an optical microscope and then frozen by plunging it into liquid nitrogen. at -210 OC. After 2&30 s the specimen was transferred to the cleavage unit (Figure 2b), also stored under liquid nitrogen. The cleavage device was then transferred to the freeze etch unit. The stage temperature was then increased, the unit opened, and the sample etched and coated.

2186 The Journal of Physical Chemistry, Vol. 90, No. 12, 1986

.'I

Figure 4. Transmission electron micrograph of replica of surface obtained by microloming of solid sample. Magnification ISOOOX.

The two techniques should produce similar surfaces. However, irregularities of the knife edge used in the first technique produces artifacts such as "chatter" markings and surface roughness that can seriously complicate the description and analysis of the IC. sulting micrograph. The semnd technique, double-replica freeze fracture, produces a surface that is a true representation of a fracture through the vitreous material. The fact that both of the fracture surfaces can be examined is also of use when there is m e doubt about the topography observed in the resulting micrograph. A second differencebetween the two techniqua is the difference expected in the cooling rates due to differences in the thermal properties of the specimens arising from differences in mass and differences in geometry. While it is normal to emphasize surface features by allowing sublimation ('etching") to m u r the nature of the surface was found to be independent of the sublimation time, The method used to study the fractured vitreous microemulsion surface in the transmission electron microscope (TEM) involves the creation of a platinum-carbon replica of the fracture surface. After vitrifying, fracturing, and etching, the sample is coated by evaporation of Pt-C (the R-C gun being positioned at fixed angle to the specimen) onto the specimen surface. The PtC deposition is immediately followed by the evaporation of pure carbon from a second evaporation source directly above the specimens. When deposition is complete the vacuum chamber is vented and specimens are removed. The replicas of the surface are then floated off the specimen by submerging at a shallow angle in distilled water and collected onto copper grids which are subsequently mounted in the TEM (JOEL 200 CX) for inspection. Results and Mscussion Figure 3 shows double-replica freeze fracture micrographs of the microemulsion (40% P G 3 H 2 0 27% o-xylene, 33% Tween 80 by volume) at two different magnifications and a micrograph of a propylene glycol/water mixture for comparison. Figure 4 shows a micrograph of the same microemulsion but where the surface was prepared by using the first of the techniques described above.

Hildebrand et al. While the micrographs in Figures 3 and 4 are different, the droplet structure of the microemulsion is clearly shown. Comparison of parts b and c of Figure 3 reveals a clear similarity between the continuous region of the microemulsion micrograph and the micrograph of the propylene glycol/water. This ssentially establishes a control for the microemulsion micrographs by indicating that the freeze fracture procedure does not produce artifacts in the way of surface features nor does it produce devitrification in the bulk continuous phase toward crystalline ice. That the micrographs of Figures 3a and 4 are similar to those seen previouslf removes all possibility that the previously observed droplet structures may have been induced during the quenching of the thin films. However, the hexagonal ordering observed in the previous work is not apparent in either of the micrographs of Figures 3 and 4. If it is assumed the droplets shown in Figure 3 are of reasonably uniform diameter, then the largest visible in the micrograph are those that have been totally exposed, these having droplet diameters in the region 50-60 nm. This is larger by an order of magnitude than the particles shown in Figure 4 and those observed in the previous work? The microemulsion of the previous work is at least 20 times more dilute than the present. When the composition is changed along the line of increasing water content starting from the finger region of the quaternary phase diagram the microemulsion changes from clear to slightly turbid (but stable, indicating particles >lo0 nm in diameter) and returns to optically clear. These simple visual observations indicate that there are major changes in size distribution of the particles, and hence structure of the solution, as the water content is changed. The degree of similarity between the micrograph for Figure 4 to that of previous work is perhaps somewhat surprising. The techniques used in this work essentially guarantee to produce a representation of the microemulsion in the fluid state. It is, however, not clear to which region of temperature this representationapplies. The difference between the results obtained by using the two freeze fracture techniques with their different cooling rates may well represent a temperature dependence of the microemulsion structure. It is clear from all the results and particularly Figure 3 that the microemulsion droplets represented there vary in both size and shape. Chains of 2-12 predominate over single droplets which of course may only be the exposed portion of a more complex structure. Conclusion Freeze fracture electron microscopy is a powerful tool for examining the structure of glass-forming microemulsions. The structure of the particular microemulsion studied in this work is complex with a tendency to be in the form of chains of droplets, and present indications are that both droplet size and structure are strong functions of composition and temperature.

Acknowledgment. We express our thanks to MI. L. Frawley for his technical assistance and for many useful discussions and also to Prof. C. A. Angel1 for his encouragement of this work. Financial support from Monash University Special Rscarch Grant is gratefully acknowledged. R@(ry No. Tween 80, 9005-65-6: o-xylene. 95-47-6: propylene glycol, 5 7 - 5 6 ,