The specific surface area of a soil is an important and intrinsic property that is correlated to phenomena such as adsorption and release of chemicals, cation exchange and water retention. It is defined as the total (inner and outer) surface per mass unit (m²/g). The specific surface of soils is highly variable, depending on particle size distribution and clay mineralogy. A usual procedure to estimate the surface area is to measure the amount of gas or liquid needed to cover with a molecular monolayer the entire surface. Knowing the area covered by a single molecule on the surface, the total surface area can be easily estimated. However, experimental procedures can be highly sophisticated and time consuming.

The amount of water remaining in a soil at equilibrium is a function of the size and volume of the water-filled pores and of the amount of water adsorbed onto the particles. These amounts are dependent on the matric potential applied to the soil, the solute content and the thermal conditions, and define the potential of water with reference to a pool of distilled water at standard conditions.

The aim of this work was to correlate the specific surface of Oxisols obtained by nitrogen adsorption with soil-moisture retention at various potentials, obtained in laboratory experiments using a pressure-plate apparatus.

Samples of 27 Oxisols from Brazil, with clay content ranging from 0.170 to 0.840 kg/kg, organic matter content between 0 and 0.055 kg/kg and effective cation exchange capacity between 0.03 and 4.33 cmol_c/dm³, were used. Specific surface (SS) of the samples was determined with nitrogen as adsorbate, at 78 K, and the isotherms were interpreted by the BET model. Gravimetric water contents (W) were obtained at –33, –500, –1,000 and –1,500 kPa matric potentials (Y), using a pressure-plate apparatus.

Values of SS ranged from 8.4 to 59.5 m²/g, and those of W from 0.041 to 0.345 kg/kg, for all four water potentials.

Specific surface data were related to the amount of water retained at each potential by linear regression, resulting in four linear equations. Slopes of the straight lines varied from 174.2 (at –33 kPa) to 219.2 m²/g (at –1,500 kPa). Intercepts ranged from –7.8 (at –33 kPa) to –1.6 m²/g (at –1,500 kPa). The determination coefficient R² ranged from 0.90 to 0.93, showing a good fitting model.

Finally, it was possible to fit, by non-linear regression, the slope values and intercepts for the pressures applied in the –33 and –1,500 kPa Y range. The obtained regression equations were: Intercept = -9.15 + 0.2520 T ^0.5 – 0.001545 T (R² = 0.99); Slope = 162.17 + 2.2670 T ^0.5 – 0.02127 T (R² = 0.99), where T = –Y is the applied pressure-plate tension, in kPa.

These equations allowed estimating with good precision the SS of Oxisols, by calculation of the linear equation parameters, intercept and slope, for every potential between –33 and –1,500 kPa, and determining the soil moisture at the selected potential.