Direct Measurement of Gaseous Diffusion in Soils

Several studies (1, 5, 6, 7, 8, 9, 12, 13, 14, 15, 17, 22, 26, 27)° have dealt with the movement of gases through soils. For the most part these investigators have dealt with the characterization of soils on the basis of their air permeability, having measured the flow of the gas through soil under a pressure gradient. Most of the studies were carried out on artificially packed soil columns in the laboratory; some however, were carried out on soils in situ (12, 13, 15). The factors affecting flow characteristics, such as porosity, nature of the pores, temperature and moisture. content of the soil and of the gas, soil cover, freezing of the soil, pressure of the flowing gas, etc., were studied in a quantative manner by the early German workers (1, 6, 8, 22, 26, 27). Much of this work apparently has been overlooked by recent investigators. A knowledge 'of the laws governing the flow of gases in soils is of interest. Gaseous flow under a pressure differential, however, differs in several respects from the transfer of gases by diffusion. Buckingham (4) pointed out that gaseous flow varies as the sixth or seventh power of the porosity whereas the same author stated diffusion to vary with the second power. Other workers (10, 11, 21, 24) state that diffusion varies directly with porosity. Movement by diffusion differs from flow induced by pressure in that factors such as streamline flow, friction, and turbulence are absent where gases move by diffusion. Heinrich's work (12, 13) indicates that there may also be a rupture value for flow of gases through soils at low pressures and that the pressure required for flow reaches a maximum, and then decreases with increasing pressure to a certain minimum pressure. Romell (23) stated that when one seeks to use the permeability of a soil to air under applied pressure as a measure for its aeration, one does so on the hypothesis that mass flow plays the principal role and that diffusion is secondary. Since air permeability varies with the method by which it is determined and since it differs in its nature from diffusion, it is evident that these measurements are of limited value in making inferences regarding the diffusion mechanism in soils. Hannen (11) was probably the first to study diffusion as such in soils. He used artificially packed soil columns about 10 inches long. Below the soil column he placed a chamber of carbon dioxide. After allowing diffusion to occur for a period of about 10 hours, he analyzed the remaining gas in the chamber. He concluded that the quantity of gas diffusing varied directly with the total pore space, or Q p= kS, where Q is the amount of gas that diffused through a soil having a pore space of S, and k is a constant. Buckingham on the other hand concluded from his data that the diffusion constant varied as the square of the pore space, or Q = kS. Smith and Brown (24) attempted measurements of CO2 diffusion in moist undisturbed soil samples in the laboratory. They stated that "accurate determination of the rate of diffusion could not be made," because of complications resulting from production of CO3 in the soil while the measurements were being made. With samples of disturbed soil they found that, "the rate of diffusion of carbon dioxide through air dry soil is a linear function of porosity of the soil within the limits of porosity studied." These limits were 36.4% to 64.5% porosity. Two other studies have found a linear relationship between porosity and diffusion. Both used artificially packed soil columns of varying porosity and both used the vapors of liquid carbon disulfide as the diffusing gas. Hagan (10) studied the factors affecting diffusion and expressed his results in terms of permeability units. An interesting feature of his study is his conclusion that "the permeability of these artificially packed soil columns has been found to approach zero, not at zero free porosity, but in a porosity range of 26 to 29%." Hagan used disturbed soil samples with moisture additions to reduce the air-space porosity. Penman (21) carried out his experiments under very carefully controlled conditions in the laboratory. He used several air dry solids which, in addition to soils, included steel wool, mica, sand, and glass spheres. He expressed his results by the relationship, D/D0 = 0.66S where D is the coefficient of diffusion through the material having a pore space S, and D0 is the diffusion coefficient through free air in the apparatus used, that is, where S = 1. Penman extrapolated his curve of diffusion vs porosity to the origin, and it had a slope of 0.66 up to a porosity of about 60%. Above this porosity the slope was greater than 0.66. The above mentioned experiments on gaseous diffusion were all carried out on soil samples whose natural structure had been greatly altered (Smith and Brown's attempt, to measure diffusion of natural structure samples failed). Special effort was made in most cases to obtain uniform porosity throughout the sample. It is well-known that soils which appear to be uniform on the basis of soil type and past history, etc., vary widely in porosity when samples are taken within a few inches of one another in the field. Furthermore, there is seldom uni. form porosity within any given sample even when undisturbed samples are carefully taken. The purpose of this study was to examine the diffusion process as it occurs on undisturbed soils in situ in the absence of applied overall pressure differentials;. It is believed that the only way of fully understanding the process as it occurs under natural conditions is to carry out studies on soils in the field where physical homogeneity does not exist. Relationships between diffusion and porosity of different soils are discussed. Diffusion rates are shown on soils under different rotations and on soils where different tillage practices have been carried out.