Interphasal water

FIG. 1 Typical DSC thermograms of K-oleate-hexanol-dodecane-water microemul-sion samples. Surfactant/oil = 0.2 g/mL alcohol/oil = 0.4ml/mL water/(water + oil) = 0.222-0.4 g/g. Curve a, W/O microemulsion sample curve b, D2O/01I microemulsion sample. Endothermic peaks due to the fusion of DjO (277 K), free water (273 K), dodecane and interphasal water (263 K), bound water (233 K), and hexanol (220 K) were identified. (From Ref. 11.)... 

The simple but elegant way by which Senatra and coworkers solved the problem of identifying interphasal water in dodecane-containing systems should be noted. Both interphasal water and dodecane melt at about - 10°C, thus leading to overlapping of their fusion peaks. However, the existence of interphasal water may be clearly shown by taking the following into consideration [10] ... 

The heat of fusion, measured at - 10°C, is higher than that required for the known amount of dodecane in the system. By subtracting the enthalpy change due to the dodecane, the contribution of the interphasal water is readily calculated by the equation [10]... 

Aa DSC-ENDO study Free Interphasal water water Hexadecane Enthalpy contribution DSC-EXO smdy (A//w)fe T Type of structure... 

AH )263 = enthalpy change of interphasal water. (This type of water melts at 263 K). ° A/fh = enthalpy change of pure hexadecane. 

The variation of the temperatures of the midpoints of the peaks related to water (bound, interphasal, and free) as a function of water content followed the same pattern a gradual increase of the temperature to a less negative value and then a more or less constant temperature. Such behavior was observed for systems A [10], B [45], D [42], and E [42] for the binary system water-Ci2(EO)8 [45] and for aqueous solutions of polymers (here as a function of the sorbed water content) such as poly(4-hydroxystyrene) [40] and mucopolysaccharides [83], For example, system A has an endothermic peak at about -10°C, which is ascribed to the melting of interphasal water (and dodecane). In fact, it begins at about - 12°C, increases in height with increasing water content, and levels off at about - 10°C, when (total) water content approaches 30 wt%. For the binary system water + Ci2(EO)s, interphasal water melts between -3 and -4°C [45,84],... 

In system A the surfactant becomes saturated with water at Nweo = 3, where iVw/Eo is the number of (interphasal) water molecules per ethylene oxide (EO) group of the surfactant. For this water/surfactant molar ratio, (total) water content is again about 30 wt%. 

Alcohol also strengthens the interaction of the surfactant with water molecules adjacent to it. This is revealed by the shift of the endothermic peak (ascribed to interphasal water) toward more negative temperatures upon the addition of alcohol. We showed this behavior for system A (peak at about -10°C) compared with that of the binary system water-Ci2(EO)8, which has an endothermic peak at -3 to -4°C (see above). 

It was found that the effect of alcohol becomes more pronounced as the alcohol chain shortens. This phenomenon was observed for bound (and interphasal) water endothermic peaks in systems A and B [45] and in systems D and E [42]. However, the melting temperature of bound water in these systems levels off at a... 

We first exemplify our ideas using system A. In this system, the DSC results demonstrate that no free water is detected below about 30 wt% of the total water [10], whereas in the binary counterpart [water-Ci2(EO)8 mixtures] no free water is detected below about 60 wt% of the total water ( ) [44]. The same water content at surfactant saturation was observed in the water-Cie (EO)2o system [44]. In the binary system water-Ci2(EO)8, interphasal water indeed melts at about -3°C, but when free water forms it melts at about - 1°C [45]. It may then be concluded that the freeze-thaw process imposed on the interphasal water clearly does not cause water segregation. At higher water concentrations, free water begins to appear. The interphasal water concentration (relative to the constant total weight of the microemulsion) is not altered. Yet its fraction in total water content constantly decreases as more and more free water is formed. Thus, a gradual decrease of its peak (at -10°C) is expected and observed [10,45], The amount of total water (revealed as free water) equals, more or less, that of added water plus the fixed amount of interphasal water at saturation, for each microemulsion sample. It is most probable that free water at subzero temperature maintains the form of ice lumps within what is called (when the sample is kept at room temperature) a microemulsion core, or as an outer ice shell after the inversion to an O/W microemulsion. 

Let us first take a hypothetical case, where excess liquid alcohol is contained in the capsule together with water and surfactant. If the alcohol could interact in any way with water or surfactant at subzero temperature, it would have undergone the same interaction, faster, at room temperature. The reason for this is that usually the oil is not directly involved in interactions with the polar headgroups of the surfactant, and even less with water. (The relatively rare case of system C, in which the oil itself interacts with surfactant or water, is analyzed in the following.) Obviously, the alcohol cannot cause a saturated surfactant to bind more water, nor can it detach (free) water from a partially hydrated surfactant. Thus, in a microemulsion sample containing 18.6 wt% hexadecane, 18.6 wt% pentanol, 37.2 wt% Ci2(EO)8, and 25.6 wt% water, only two melting peaks were observed one at -10.4°C for interphasal water and one at +15.3°C for hexadecane. This... 

FIG. 12 Endothermic contribution due to the melting of interphasal water at T = 263 K. Water-hexadecane system (Table 2), C = 0.105. Inset shows a close-up of the 263 K peak. (From Ref. 13.)... 

The analysis by DSC of the freezing behavior of liquid microemulsion samples (DSC-EXO spectra) was first undertaken for checking on whether it were possible to distinguish the interphasal water endothermic peak from that of the dodecane in the water-dodecane system. As shown in Fig. 21, the DSC-EXO recordings confirm the presence of a thermal event superimposed on that of the dodecane [13,15]. Since then, DSC-EXO analysis has been carried out as a standard procedure for all the systems studied. 

The bound water freezes at a lower temperature than the interphasal water, which is just another way of sa5dng that the surfactant-water interaction is stronger for the bound water. However, the interphasal water-surfactant hydrate is, of course, more stable because if the positive entropic contribution (to the formation of ordered crystals in the freezing process) is overcompensated in the case of bound water at a lower temperature than in the case of interphasal water, then the bound water will obviously melt at this lower temperature. For simplicity, we assume here that no supercooling occurs. Clearly, the argument also remains essentially the same when supercooling does occur. 

As both dodecane and interphasal water melt at about - 10°C (with overlapping of their fusion peaks), interphasal water peaks in dodecane-contain-ing systems should be unambiguously identified. This can be done in two ways (besides using D2O instead of water) [11] ... 

The distinction between the melting peaks of free and interphasal water enables the evaluation of their relative amounts in microemulsion samples. If the water content is gradually increased along known dilution lines, it is usually observed that the concentration of the interphasal water increases imtil it reaches a constant value (relative to a fixed amount of the microemulsion sample) that corresponds to a specific... 

It is also worth noting that the apparent plateau, characterizing the variation of interphasal water content with the total water content in Fig. 3, is absent (although it should be present if the surfactant was fully hydrated ) from the corresponding binary system (see Fig. 4) as well as from the Ci2(EO)23 (Brij 35)-water system [9]. Furthermore, the plateau is also absent from the quaternary system water-butanol + dodecane 1 1 (by weight)-the commercial sugar ester SI570 (see Fig. 5). This observation will be analyzed in more detail elsewhere (Ezrahi et al. paper in preparation). 

FIG. 3 Variation of free and interphasal water content with the total concentration of water for system A. Concentrations are calculated as a weight percentage relative to the weight of the microemulsion samples. (X) Free water (0) interphasal water. (From Refs. 11 and 20.)... 

FIG. 4 Variation of free and interphasal water content with total water content for... 

Two interrelated topics that bear most directly on the description of the hydration shell—i.e., the bound water layer(s)—are the definition of the shell and its thickness. The problem of how the bound water can be sufficiently precisely defined is discussed elsewhere [11,37,51] and we shall not pursue it further here. It is clear, however, that the extent to which water is affected by a nearby surface is a function of the distance between them, namely the thickness of the hydration shell. Second-layer water (and, obviously, multilayer water) is much less perturbed than the water adjacent to the surface. We have used several methods to evaluate the thickness of the interphasal water layer in system A (as revealed by the low-temperatme behavior of water) [2,11] and found it to be about 0.5 nm. Virtually the same value has been assessed for the thickness of the bound water layer on many organic and inorganic substrates [37,52-57]. As 0.5-0.6 nm is the thickness of two water molecules [45], we may envisage two monolayers of interphasal (or boimd) water that are loosely associated with the substrate. We have shown that Aw/eo = 3 for system A at a total water content of 30 wt%. 

---