Лако-красочные материалы — производство

Mass alone is not the only determinant of volatility. Thus, benzalde — hyde has an RMM of 106, close to that of methyl butanoate, but its boiling point is substantially higher (by 76 °С). Other factors are evidently important, and it becomes useful to deal with the set of molecular properties which collectively contribute to what is known as ‘polarity’. Explicit definitions of this term are rare. It relates particu­larly to the degree to which electronic charge is spread evenly through a molecule or whether certain locations have relatively high concentra­tions of positive or negative charge. As explained below, we use it here to represent the summation of physical interactions at the molecular level which may influence a molecule’s free energy and other thermo­dynamic parameters (and hence also its availability and mobility within any given matrix).

A number of molecular interactions are feasible, but not all are relevant for the types of molecules found in perfumes. For example, ion-ion forces are unlikely to feature in any direct perfume interac­tions, since fragrance components rarely bear a charge (except for materials that are influenced by extremes of pH, e. g. Schiff’s bases are protonated in strong acid, carboxylic acids are substantially anionic above pH 6). The principal physical interactions contributing to overall polarity that need to be considered are ion-dipole, dipole-dipole, dispersion forces (‘London’ forces) and hydrogen bonding. These may all play a part, depending upon molecular size, the presence of permanent dipoles and the type of functional group present.

As a general guideline, usually the affinity between a molecule and its microenvironment is higher when it is surrounded by molecules that are capable of expressing the same types of interaction, e. g. limonene dissolved in a hydrocarbon or citronellol dissolved in ethanol. Some specific examples are discussed later, but the effect may be taken as a more general form of the often expressed chemical adage that ‘like dissolves like’. The activity coefficient, y, introduced in equation (3) provides a useful way of assessing affinity. For any single volatile material, values of у greater than unity imply partial pressures in excess of that predicted by Raoult’s Law, indicating that more material is in the gas phase than would be expected, and that the material is effectively being ‘pushed’ out of the system. Conversely, values of у below unity imply lower partial pressures, less material in the gas phase, slower evaporation and, perhaps, a concomitant increase in the persistence of the material in a system. As an aside, certain high-boiling materials are well known to promote the longevity of other materials in a perfume. For the physical chemist this phenomenon (based on negative deviations from Raoult’s Law) is an example of what is termed ‘fixation’ within the fragrance world, but to the perfumer the term also conveys harmonious blending of notes throughout a per­fume’s in-use life.

Table 11.2 contains some data that exemplify the above comments. It cites values of the activity coefficient for an ester (benzyl acetate), an alcohol (heptan-2-ol) and a terpene (limonene) in different environ­ments encompassing a range of polarities: water (highly polar), an aqueous surfactant (as used in shampoos) and a moderately polar solvent used in perfumery (diethyl phthalate, DEP). It can be seen that heptanol and limonene have values of у in DEP which are similar and also greater than that of benzyl acetate, but in water limonene has a very large activity coefficient. Limonene is not able to participate in any strong, cohesive interactions with water molecules, and the high у is a consequence of this. Heptanol, on the other hand, can participate in hydrogen bonding and exhibits much lower values of y. Note that in the case of the shampoo, we may be dealing with ‘apparent’ activity coefficients since the degree of liquid phase homogeneity is not certain (owing to the presence of micelles and/or emulsion droplets).

To reinforce what this means in practice, Figures 11.1 and 11.2 depict

the equilibrium headspace profile (i. e. the gas phase concentrations of volatiles) above two systems containing the same perfume ingredients: a neat perfume oil and a cologne (typically these are alcoholic solutions containing 1-3% perfume). Figure 11.1 shows a (partial) headspace chromatogram containing labelled peaks corresponding to three ter- pene alcohols frequently used in cologne perfumes (dihydromyrcenol, linalool and citronellol). The same three peaks are marked by asterisks in Figure 11.2, which shows the headspace profile once the perfume oil has been taken up into aqueous alcohol. It can be seen that the concentrations of the terpene alcohols are relatively reduced in the polar medium (the cologne) compared with non-hydrogen-bonding molecules bearing other functional groups. Similar, but smaller, differ­ences are also present for many of the other materials, and in consequence differences in odour characteristics may well occur. This level of understanding is useful, but the perfumer needs to know how to select ingredients that provide superior performance in the target product area. To meet these needs, we need to recognize and ideally measure or calculate the molecular characteristics that govern or mediate activity coefficient behaviour.

It would be very convenient if it were possible to calculate activity coefficients for any molecule in any given environment. Unfortunately, few situations occur in which this is possible, and rarely do such situations appertain to real products. Thus, for example, we may estimate the activity coefficients of many alkanes and simple derivatives at infinite dilution in water, but the corresponding values for the same materials in a specific shower gel are not readily calculable from first principles. However, a large number of parameters are available in the literature and have been used to answer questions related to physical behaviour.

The pharmaceutical industry has for many years developed mathe­matical models to explain biological activity of drugs: these are termed quantitative structure-activity relationships (QSARs). These tech­niques may also be applied to the situations described herein, although more correctly we are often more interested in QPARs, where the P stands for ‘property’ (which may refer to macroscopic properties, such as density, melting point, or viscosity, or to molecular or sub-molecular properties, such as molecular or fragmental volumes, or atomic charges). Since these properties are always at least partially dependent upon ‘structure’ in its broadest sense, the distinction between these approaches is sometimes blurred, particularly when pure structural data are used in the same mathematical model as a melting point! But whatever the semantics, the underlying need remains the same: to understand and predict the effect of changing the structure of a perfume ingredient on its performance, so that we may reasonably estimate, for example, the concentration of a perfume ingredient above a cologne, or its affinity for cotton during a wash cycle. This whole approach is detailed elsewhere in this book (see Chapter 14), so just two parameters found to have widespread application in the perfumery area are discussed here. These parameters are the octanol/water partition coefficient (usually expressed as its common logarithm, log P), and the Hildebrand solubility parameter (sp, often designated S with or without a subscript). The Hildebrand parameter (Barton, 1985) is defined in equation (5), where AH is the molar enthalpy of vaporization, V is the molar volume, R is the gas constant and T is the temperature. The SI unit for the solubility parameter is MPa0’5, but the c. g.s. system equivalent (cal cm _3)0-5 is commonly seen.

sp = [(AH — RT)jV]0’5 (5)

The physical significance of these parameters is discussed in the next section. For now, the key point to make is that these parameters may feature explicitly in empirical QPARs developed to help ingredient selection or design. Alternatively, they may be used simply as classifica­tion variables to help identify ingredients that are likely to exhibit good performance in a specific fragranced product. In this second approach, a number of parameters are investigated for their usefulness in characterizing behaviour, e. g. a plot of log P versus boiling point for a variety of ingredients may lead to the identification of clusters of materials with good performance. Such classification variables may be used qualitatively or quantitatively, depending upon the difficulty of the problem and the statistical expertise available. The same para­meters and approaches find particular use in the development of perfumes in which fragrance longevity is a key requirement, and this is the focus of the next section.