Band 71, Heft 4

The nomenclature of natural products has suffered from much confusion, mostly for historical reasons. The isolation of a new substance, in the early days of the science, generally preceded its characterization by a lengthy period. Thus, these compounds were often assigned trivial names that gave no indication of the structure of the molecule and were often found afterwards to be misleading. Even when the original names were later revised (for example: glycerin to glycerol) the new names often expressed the structure imperfectly and were thus unsuitable for the nomenclatural manipulation that is required to name derivatives or stereoisomers. The result was a proliferation of trivial names that taxed the memory of chemists and obscured important structural relationships. The resultant disorder in the literature led to the creation of committees of specialists with the task of codifying the naming of compounds in various connected areas of natural-product chemistry, such as steroids, lipids, and carbohydrates. As far as their recommendations have been followed, their efforts have been successful in eliminating confusing or duplicate nomenclature. It is the aim of the lUPAC Commission on Nomenclature of Organic Chemistry to unite as far as possible all the specialist reports into a single set of recommendations that can be applied in most areas of natural-product chemistry. Accordingly, provisional recommendations were prepared and published as Section F of the lUPAC Organic Nomenclature Rules, first in 1976, and then in the 1979 edition of the Rules.

Introduction and Fundamental Concepts It has long been known that solvents often affect chemical reactivity, this involving, e.g., the shift of the position of chemical equilibria (thermodynamic aspect) as well as significant changes in reaction rate constants (kinetic aspect). Physical properties, particularly the frequencies and intensities of transitions in IR, UV-visible, fluorescence, NMR and ESR spectroscopies are also known to be affected by solvents. These phenomena are consequences of differences in the solvation of reagents and products (thermodynamic effects) or reagents and activated complexes (kinetic effects). Differential solvation of species in the ground and excited states accounts for the spectral phenomenology indicated above. Differences in solvation of a given solute in two different solvents determine the size of the corresponding partition coefficient. It is customary to state that these effects reflect the influence of "solvent polarity". According to Reichardt, "solvent polarity" is "the overall solvation capability (or solvation power) of solvents, which in turn depends on the action of all possible, nonspecific and specific intermolecular interactions between solute ions or molecules and solvent molecules, excluding, however, those interactions leading to definite chemical alterations of the ions or molecules of the solute (such as protonation, oxidation, reduction, chemical complex formation, etc.)". This definition underscores the extreme complexity of "solvation effects" at the molecular level. This notwithstanding, solvent effects (SE) often display some remarkable regularities that allow in many cases an "empirical treatment" that sheds light on their origin and main contributors. Consider a solute S and two different properties P 1 and P 2 , taking the values { P 10 , P 11 , ..., P 1 i , ...} and { P 20 , P 21 , ..., P 2 i , ...} respectively in solvents S 0 , S 1 , ..., S i ,... A scale of SEs is simply constructed by taking for each solvent, S i , the difference P 1 i - P 10 , S 0 being chosen as a reference solvent. If the property P 1 is a "good descriptor" of SEs on P 2 , equation (1) holds for solvent S i : P 2 i - P 20 = k ( P 1 i - P 10 ) (1) or P 2 i - P 20 = k p i (2) wherein k is a constant independent of the solvents and determined solely by P 2 . p i is the "solvent parameter" characteristic of solvent S i . More generally, the properties being compared might belong to two different solutes. For any property P m of any solute, and if P 1 is a "good descriptor", equation (3) holds: P mi - P m 0 = k m p i (3) We draw attention to the (frequently overlooked) fact that if equations such as (3) were truly general, then, all SEs would be linearly related to an extremely high degree of precision and a single, universal scale of SEs would exist. This is against all the available experimental evidence. Excellent correlations of narrower scope do exist however, that successfully link a very large amount of experimental data for a substantial variety of solvents and solutes and a relatively small number of empirical scales. In this compilation, scales are selected on the basis of criteria to be discussed below. For each of them, the most reliable values of solvent parameters are given. The physical foundations and the scope of the scales are discussed. Some suggestions regarding their use are made. For the purpose of facilitating future work in the field, some indications are given regarding the experimental determination of the various parameters. Solvent-solute interactions always involve dispersion or London's forces and, very often, dipolar and/or multipolar interactions. Current theoretical models including London's, Hildebrand's and/or reaction field (RF) theories allow to express these "non specific" parts of SEs as functions of physical properties such as the refractive index, relative permittivity and thermodynamic properties of the solvent. Quantitative rankings of solvents ("solvent scales") can thus be constructed on the basis of such properties as the refractive index, n ; electric permittivity (formerly known as dielectric constant), e r ; Hildebrand's solubility parameter, d H ; the modulus of the molecular dipole moment, μ , and various functions thereof. These scales shall be termed "model-independent". Quite generally, it is assumed that dispersive and electrostatic interactions are independent and additive. Following Palm and Koppel, it is further assumed that other contributions to solvent-solute interactions, notably hydrogen bonding are also independent from and additive to, the "non specific" contributions. Use is often made of "model-dependent" scales. They are based on the similarity principle : the ranking of the efficiency of solvents on a given property is quantitatively compared to their influence on a reference physical or chemical property of a reference solute ("molecular probe"). The associated formalism is quite simple and has been outlined above. These scales can be divided into two different categories, depending on whether they quantify the overall "polarity effect" of the solvent, in Reichardt's sense or else, they are intended to measure one or various components of the overall solvation power of the solvent. Some scales have been built on the basis of a statistical treatment of SEs on large sets of experimental data of various origins. They are absolutely "empirical" in that they try to quantify "average" SEs without specifically seeking a formal link with current theoretical concepts on solvent-solute interactions. Last we mention that the case of self-associated solvents and mixtures thereof involves a number of conceptual and experimental difficulties that, in our opinion, justify a separate treatment. Here we deal with non-hydrogen bond donor solvents, although a number of weak hydrogen bond donor solvents are also included. Part I of this compilation is organized as follows: Description of the Tables. Physical properties and model-independent scales. Table I. Model-dependent scales: a) "Overall solvation" scales. b) Scales of dipolarity/polarizability. c) Scales of hydrogen bonding accepting power (hydrogen bonding basicity). d) Scales of "hard" and "soft" Lewis acidity and basicity. Statistical scales. Table II. General comments on the scales.