The number of chemicals used in industry is increasing, and as a consequence workers in chemical industries are thought to have many opportunities for being exposed to chemicals. For organic solvents, although a number of studies have shown the toxicity of individual chemicals, there are a relatively small number of studies that have described their toxicity in terms of a quantitative structure-activity relationship (QSAR). In the present article I first introduced an outline of the method of QSAR (Hansch's method), and then reviewed the literature on QSAR of industrial chemicals, principally organic solvents and related chemicals as follows:
1) A review was made of the studies of general toxicity of chlorophenols, alcohols, amines, anilines, phenols, ethers, alkanes, ketones, acrylates, methacrylates and nitriles. In almost all cases, the general toxicity is related to log P, in which P is n-octanol/water partition coefficient, indicating the importance of the hydrophobicity of chemicals for their toxicities.
2) The QSARs of anesthetic organic solvents were reviewed. The chemicals analysed were ethers, alkanes, ketones, miscellaneous chemicals and anesthetic gases. It is shown that the relative anesthetic potency of anesthetic gases depends not only on the hydrophobicity expressed by log P but on a polar factor, while the potency of other chemicals depends largely on log P.
3) The relationship of the structure of organic solvents to skin penetration or absorption was reviewed. QSARs reveal that the potency of skin penetration or absorption of solvents is a function of water solubility, log P or hydrogen bond number, each relating to hydrophobicity.
4) Some organic solvents have an effect on the upper respiratory tract. The established QSAR models considerably resemble corresponding equations for gas/liquid partition coefficients into organic bases such as tricresylphosphate.
5) The ecotoxicity of chemicals including solvents was reviewed. Overall, the relative toxic potency of chemicals is related linearly to log P, indicating the importance of hydrophobicity in determining the toxicity of chemicals.
6) Although no QSAR has modeled the mutagenic activity of organic solvents, a few studies show the relationship between the structure and mutagenicity of haloalkenes. Finally, if QSAR could model the toxicity of a series of chemicals, it would serve our purpose of understanding the mechanism of toxicity and of predicting the toxic potency of chemicals in the same series tested.

View full abstract

Acute oral toxicity (LD₅₀-value) of organic chemicals to rats was analyzed by using solubility parameter (δ_c), a thermodynamic parameter, of the chemicals. Certain parabolic correlations were established between logarithm of LD₅₀-value (mmol/kg body weight, rats) and δ_c of all the collected chemicals (n=144, R=0.578), alcohols (n=29, R=0.587), ketones (n=7, R=0.962), aldehydes (n=9, R=0.621), ethers (n=5, R=0.890), acetates (n=7, R=0.670) and aromatics (n=84, R=0.736). Introducing molar volume (V_c) to the above equations could not improve the correlation. In the study, we assumed that as for acute toxicity, chemicals taken into the mammals through biological membrane first disturb the homeostasis, which causes certain biological reactions (i.e. death) and that amounts of the chemicals intaken are regulated by their solubility in the membrane. Based on the assumption, we drew a theoretical equation, which describes LD₅₀ by a parabolic function of δ_c. A regression analysis using the equation gave significant correlations as stated above, which incarnates the assumption. A solubility parameter of 2.30×10⁴ (J/m³)^1/2 was also determined for the biological membrane (absorption site) of rats. For comparison, log P was used to describe LD₅₀ of all the chemicals. but no correlation was established (R=0.1640.443).

View full abstract