The interactions between soil constituents can significantly alter the soil sorption activity. Consequently, relationships between sorption values and soil composition are often not additive and cannot be easily quantified. The first step in the elucidation of such a relationship is the estimation of relative sorption activities of soil clay and organic components at different moisture. At RH < 20%, based on the example of vapor-phase sorption of haloidalkanes on synthetic clay-humic complexes, Cabbar (1999) showed that humic acids (HA) are less active than clay. Moreover, he revealed the blocking effect of HA on sorption sites of the mineral soil surface. However, these results were obtained on commercial HA (the structure of such HA strongly differed from native structure of soil HA). More accurate data were presented by Poe et al. (1988), where the authors investigated vapor-phase sorption of organic compounds of different nature on soils. Unfortunately, their conclusions on the correlation between sorption activities of mineral and organic soil constituents are not entirely convincing. This is due to these conclusions being made based on the data for only two soils, which possess similar granulometric composition but small differences in soil organic matter (SOM) contents (C_org = 2.82 vs. 0.97%).

In this regard we examined native soil samples with the purpose to find moisture conditions that alter the ratio of their activities, and to establish the relation between the chemical nature of a sorbate and the presence of SOM blocking effect on sorption as a consequence of clay-SOM interactions. Vapor-phase sorption isotherms of n-nonane, p-xylene (hydrocarbons - HC) and methanol (an aliphatic alcohol) on soils in the moisture range of 0-9% were obtained with the help of static headspace gas-chromatographic analysis method. Soil samples from seven separate horizons of dark-grey forest soil (Tatarstan republic, Russia) were analyzed, which had similar granulometric composition (34-38% of clay) but were significantly different in C_org (0-7.5%). The sorption values V_s of both HC (p-xylene and n-nonane) on oven-dried soil samples strongly decreased with the increase of C_org. This showed pronouncedly that HC interact with SOM more weekly than with clay compounds, and that furthermore, SOM blocks active sorption centers of soil clay surface. The decrease of V_s values of both HC (p-xylene and n-nonane) with the increase of C_org (from lower to upper soil layers) was also observed for soil samples moistened up to 2-4%. However, this tendency became weaker when moisture content increased. When the latter reached 5%, sorption values were practically independent from SOM contents. Moreover, at 6%, we even observed a small inversion of this dependence, i.e. the growth of V_s values with the increase of C_org contents. In comparison with HC sorption, our results for methanol were different. Here, V_s values for methanol sorbed by oven-dried soil layers did not change significantly with the increase of C_org. This indicates the similarity between sorption activities of non-hydrated mineral and organic soil constituents in relation to methanol. This result is surprising since it is known that alcohols are capable to interact specifically with soil clay minerals as opposed to HC, which realize only non-specific interactions. In this connection we have proposed that the main reason of such fact is different mechanisms of interaction of alcohols and HC with SOM. To verify this possibility, we studied the sorption of HC (benzene) and aliphatic alcohol (ethanol) by dried samples of humic acids extracted from dark-grey forest soil. As a result we have seen that V_s values for benzene on HA were insignificant (close to zero) in all isotherm range, whereas V_s values for ethanol were also insignificant up to activities of P/P₀ = 0.25 but strongly increased (single-order) at activities above P/P₀ = 0.25. Such strong positive "cooperative effect" was due to the change of spatial phase structure of HA caused by the initially sorbed alcohol molecules and resulting in the formation of additional binding centers.

Thus, we have demonstrated that relative sorption activities of soil clay and organic constituents depend not only on moisture but also on the chemical sorbate nature. The effect of sorption "blocking" caused by the coating of oven-dried mineral surface with SOM, has no universal character and is dependent on the nature of organic sorbates. * The work was supported by ISTC, Project #2419; RFBR, Project 06-04-49097a; and Grant of President of Russia MK-2439.2005.5.