Prediction of pH Errors in Soil-water Extractors Due to Degassing

Moisture samples taken from the unsaturated zone with soil water extractors undergo degassing and an upward shift in pH. The measured pH values from commercially available extractors are usually sufficiently in error that they cannot be used in a quantitative manner. A model has been developed that predicts the extent of CO2 degassing and the resulting pH error. Using this model measured pH values can be corrected back to in situ soil water pH provided that precipitation has not occurred in the extractor. Extractors are classified into two groups—single chamber and multichambered. The extractors are evaluated for both operation under constant vacuum (open to the source) and decreasing vacuum (evacuated and then sealed). Analysis of the data and model predictions indicates that the major factor controlling the pH error is the ratio of liquid volume to total extractor volume. Additional factors exerting major influence are the initial extractor gas composition and the total pressure in the extractor when sampled. Variations in soil solution composition and differences in soil CO2 concentrations in carbonate buffered systems had a major effect on pH values but a negligible effect on the extractor induced pH error. Under typical field conditions the multichambered extractor is predicted to give the most satisfactory results; the pH errors were sufficiently small that no corrections for degassing were necessary. Additional Index Words: suction lysimeter, tension lysimeter, ceramic cup, soil moisture, unsaturated zone, carbon dioxide, sampler. Suarez, D.L. 1987. Prediction of pH errors in soil-water extractors due to degassing. Soil Sci. Soc. Am. J. 51:64-67. V EXTRACTORS are commonly used to obtain soil solution samples from the unsaturated zone. They are easy to use and mostly sample water 1 Contribution from the U.S. Salinity Laboratory, USDA-ARS, Riverside, CA 92501. Received 18 Feb. 1986. 2 Geochemist, U.S. Salinity Laboratory, USDA-ARS 4500 Glenwood Dr., Riverside, CA 92501. in the larger pores. Thus the solution samples obtained relate better to solute fluxes than solution extracts taken from soil cores. However, vacuum extractors can extract water only under relatively wet conditions and do not represent average soil water compositions. From the study by Hansen and Harris (1975) it can be concluded that a constant vacuum is preferable in order to minimize the variation in pore sizes sampled. Errors caused by the ion exchange capacity of the extractor are well documented for ceramic (Grover and Lamborn, 1970) and can be minimized by the use of teflon. Available extractors with teflon tips, however, have low bubbling pressures and thus cannot be evacuated to low pressures. An additional problem is that soil solutions buffered by carbonate chemistry undergo an upward shift in pH when collected by extractors. This pH shift is due to CO2 loss from solution during sample collection. When vacuum is applied to an extractor, the partial pressure of CO2 in the extractor is reduced proportionately to the reduction in total pressure. As soil water enters the extractor, the solution degasses and CO2 is released. The loss of dissolved H2CO3 causes an increase in pH as well as potential precipitation of carbonates, phosphates, and oxides in the extractor. Accurate pH measurements are especially necessary when the data are used to determine potential mineralogical controls on solution compositions or when trace species such as heavy metals or phosphate are being measured. Suarez (1986) described the design of a multichambered extractor that reduces the pH error by flushing the sampling chamber with solution and minimizing the relative air volume in the extractor. This study contains an evaluation of the factors contributing to the discrepancy between the pH inside extractors and the pH of the soil water. Additionally, the predicted pH effects are compared with measureSUAREZ: PREDICTION OF pH ERRORS IN SOIL-WATER EXTRACTORS 65 ments taken with a multichambered extractor under controlled conditions. DESCRIPTION OF THE MODEL AND EXPERIMENTAL PROCEDURES A computer program was written to predict the pH shift due to degassing of CO2 from solution. The program uses the gas law (PV = nRT), kH values for O2, N2 (Handbook of Chemistry and Physics, 1957), and CO2 (Harned and Davis, 1943) and a solution chemistry subroutine containing the carbonic acid dissociation constants and ion pairs given in Suarez (1977). The model inputs are soil PC02, soil solution composition, initial PC02 in the extractor, extractor configuration, and initial vacuum applied. The model simulates an extractor by allowing discrete increments of liquid to enter the extractor. For extractors that are closed to the vacuum source, the partial pressure of each gas is initially increased proportionately to the decrease in gas volume for each added increment of liquid. The concentrations of dissolved O2, N2, and CO2 are recalculated to account for mixing of each new liquid increment with the existing solution in the sampler. After each mixing the gas and liquid are equilibrated using the following relationships Py = Py, + (Aqyi Py.KHy).(R.T.VL/VA)/ [1 + (R.T.KHy.VL/VA)} Aqy = Aqyi (Aqy, Pyi-KHy)/ [1 + (R.T.KHy-VL/VA)} where Py is the equilibrium partial pressure of gas y, Pyi is the partial pressure of gas y after mixing, Aqyi is the concentration of dissolved gas y in moles L~' after mixing, KHy is the Henry's Law constant for gas y, R is the universal gas constant (0.812 L-Pa-mole^'-deg"), T is the temperature in degrees Kelvin, VL/VA is the ratio of the liquid to gas volume, andAqv is the equilibrium concentration of the dissolved gas. For CO2 the Ky and Aq values correspond to the sum of dissolved CO2 and H2CO3. The solution pH is then calculated using the solution composition, calculated PC02 and an ion speciation subroutine. After accounting for C mass and the recomputed Aq values, the program cycles until P,, Py, and pH convergence is achieved. For extractors open to the vacuum, the total pressure is fixed by the vacuum source. Upon addition of solution, the mixed solution is allowed to degas until gas-liquid equilibrium is attained and the calculated total pressure equals the total pressure specified. The excess gas leaves the sample chamber and flows to the reservoir connected to the vacuum. After the sample chamber fills, any subsequent solution increment is assumed to displace an equal volume of liquid out of the sampler. The simulation does not consider chemical precipitation. The relationship between pH errors (due to degassing) and extractor design, soil solution composition, soil CO2 partial pressure (PC02), and the ratio liquid/total volume in the extractor, were investigated with the model, with all calculations at 25 °C. Extractors can be either single or multichambered and either connected to the vacuum source (constant vacuum) or closed off from the vacuum source (decreasing vacuum). Multichambered extractors were evaluated by placing them in a container filled with Ca, Na, Cl~, HCOj solutions at 25°C and bubbling them with a CO2-air mixture. The single chamber extractor is similar to a commercially available extractor but smaller. After evacuation, the sampler is isolated from the vacuum source and allowed to partially fill. Samples are taken by pressurizing the unit and collecting solution from the tube extending to the sampler tip. The new multichambered extractor consists of ceramic which is glued to a 50-mm long PVC tube and connected to a 10-mL closed pyrex container with 1 mm i.d. tubing. Tubing extends from the container cap to a solution reservoir which is in turn connected to a constant vacuum source. Samples are obtained by removing, replacing, and immediately capping the 10-mL sampling container (see Suarez, 1986 for construction details). RESULTS AND DISCUSSION Single Chamber Extractors For a single chamber extractor that has been evacuated and then sealed, the extent of the pH shift depends primarily on the relative quantities of liquid and gas in the extractor. The solid line in Fig. 1 shows the pH error predicted for an extractor initially filled with air, then pumped down to a total pressure of 5.05 kPa (0.05 atm), and sealed. If the soil matric potential, or pressure head, is between 0 and —101 kPa, the extractor fills until the total pressure inside the extractor equals 101 kPa + the pressure head. For this simulation the soil CO2 pressure is 1.01 kPa (0.01 atm), and the soil water contains 2 mmolc L~' alkalinity. The calculated pH error (ApH) is the pH of the solution in the extractor minus the pH of the soil solution. As the extractor fills the pH error follows the solid line in Fig. 1. The extractor should fill with water if the water content of the soil is at saturation and sufficient time is allowed after the vacuum is applied. Under these conditions and assuming that any precipitated material redissolves, the final pH error approaches zero, as shown by the solid line in Fig. 1. If the soil water pressure head is <0 (unsaturated conditions) the extractor cannot fill completely and at equilibrium a partial vacuum (with pressure equal to 101 kPa + soil water pressure head) remains in the extractor. Within a few days of an irrigation or major rainfall event the soil pressure head will commonly decrease to field capacity or —10 to —30 kPa (—0.1 to —0.3 atm). Under field conditions, extractors commonly fill until they are 10 to 80% water filled. This