Illinois Fertilizer
Conference Proceedings
January 26-28, 1998
| Home | 1998 Index | Search |
|
|
| Home | 1998 Index | Search |
J.W. Stucki and K. Lee 1
The aim of this study is to develop an improved soil test for plant-available potassium (K). Many methods have been tried in the past. In this study we are revisiting a method, known as electrodialysis, that was tried many decades ago, but with only limited success. The reason to revisit this method is because it was based on sound concepts. Its limitations were based on experimental problems. We believe we can overcome those limitations. Such limitations are nothing new because both chemical and electrochemical methods have proven to carry significant weaknesses. The most widespread application of electrodialysis is in the Electro-ultrafiltration (EUF) method, which has been employed to differentiate the chemical dynamics of soil K as a plant nutrient (Grimme and Nemeth, 1979; Nemeth, 1979). This method allows fluids and solutes to move into and out of the saturated soil system by virtue of a hydraulic gradient, causing the system to be chemically open. This is a fatal flaw in the method from the perspective of providing adequate book-keeping of chemical constituents and energies, which is fundamental to rigorous solution to this problem. Application of proper principles of chemical kinetics and thermodynamics is, therefore, impossible in an EUF system. Electrodialysis methods, on the other hand, can be preserved as chemically closed systems and have been utilized to extract exchangeable cations from soil and clays (Bradfield, 1927; Löddesöl, 1932; Wilson, 1928). Results from this method closely approximate those by extraction with various weak acid and neutral salt solutions. Undesirable effects, however, can accrue during electrodialysis due to changes in pH and lengthy equilibration times.
The underlying principle for electrodialysis is found in the concept of the partial molar Gibbs free energy ((G bar)i) and the total potential (fi), which for soil and clay systems were shown by Low (1951) to be identical. The subscript i refers to any chemical species, and in this case we will let that be K+ since it is the chemical species of primary interest to us in this study. The chemical potential (µi) alone is insufficient to describe the forces governing the movement of solutes in a system where force fields other than pressure are present (Gibbs, 1948). In a soil system other such forces are ever-present, and include the electrically charged clay surfaces, which introduce an electrical force field that operates on solutes such as K+. Low (1951) expressed the partial molar Gibbs free energy, which includes all of these forces, as:
(G bar)i =
µi + qi
where qi is the potential energy of component i due to its interaction with external force fields. Since the principal external force field that is present in the soil system is due to electrical interactions between clay surfaces and the exchanged cations, the term qi can be written as ziFy, where zi represents the charge or valence on the exchangeable cation, F is a constant (Faraday's), and y is the electrostatic potential. In the soil system, charges on ions generate the electrical potential, y. Equation (1) can, thus, be rewritten as
(G bar)i = µi + ziFy
While these equations may appear to be somewhat complicated, the concept is rather simple. We can think of the fertility of the soil with respect to K as being determined simply by whether K prefers to remain at the clay surface or if it prefers to migrate to the plant root. This depends on the energy of the K, which is measured by the term (G bar)i. Movement is always in the direction of decreasing energy, like water running downhill. So if the soil is to be fertile with respect to K, the value of (G bar)i for K at the plant root must be less than it is at the clay surface. The objective of this study is to determine the threshold value of (G bar)i that separates K into its plant-available and non-available forms, and to use that information to devise a soil test to measure it routinely in soil samples.
In order to accomplish the objectives of this study, values of µi and y must be measured accurately. µi is easy because it is related simply to the amount, mi, of K+ that is present, according to the equation
µi = RTlnmi
where all other terms are constants and the solution is assumed to behave ideally.
The value of y, on the other hand, is rather difficult. Following are the results from our first phase of experiments to determine y in this context. We are seeking to obtain plots of (G bar)i versus y, and to correlate those plots with empirical results from a Greenhouse study of wheat grown under various regimes of K fertilization ranging from sufficient to deficient.
This research was performed to investigate the effects of time and pH factors on the electrodialysis method and to investigate the electrical potential that should be applied to the Gibbs free energy equation as a vital variable. An electrodialysis apparatus, which consists of two cells, was employed to obtain fundamental information for extraction of available plant K before beginning the main research. In this experiment, 20 mg/L K solution was used instead of soil and clay samples. The anode was attached to the end of the left cell that contained 20 mg/L K solution. The right cell, which retained the cathode, was separated from the left compartment by a micropore filter, and was filled with deionized water. Three electroprobes were inserted between electrodes to monitor the voltage drop at the distances 13.5 cm, 18 cm, and 22.5 cm from the anode.
Sample solutions were collected to obtain the concentration of K that migrated from the K solution, due to an electrical charge gradient, through the membrane toward the cathode in the right cell. Samples were taken at distance intervals of 13.5 cm, 18 cm, 22.5 cm, and 26 cm. The voltage drops at these points were also recorded. Each fraction was simultaneously obtained by using a vacuum siphon technique at 10-min intervals over 80 min. This procedure was repeated after altering the voltage from 50 to 400 V, in various steps. The concentration of K was measured by flame emission using a Perkin-Elmer model 2380 atomic absorption spectrophotometer.
The change of pH and voltage drop in the reaction cell also was measured. When 50 V was applied, the pH change was measured after 0, 80, and 280 min, and the change of the voltage drop with distance was measured simultaneously. At 200 V, the pH of the sample solution was recorded after 0, 50, and 90 min, and at 400 V, these values were checked after 0, 30, 60, and 90 min. The change of voltage drop with distance was measured by a voltammeter that was attached to each electroprobe. The pH was measured using an Accumet pH meter model 15.
Changes in pH and electrical potential were also monitored colorimetrically. Bromocresol purple, a kind of pH indicator, was employed to investigate the electrical potential at the cathode. A solution of 0.1 mL of 1% bromocresol purple (5',5"-dibromo-o-cresol-sulphonephthalein) was added to the system at the site of the membrane in the cell, and 200 V was applied to the electrodes. The solution was sampled at the cathode every 20 min over 240 min. The absorbance of the solution was measured by scanning the wavelength range of 210 nm to 750 nm using a Varian Cary 5E UV-Visible spectrophotometer. The concentrations of the sample solution at 275 nm, 425 nm, and 590 nm were calculated from their absorbances. The electrical potential was calculated from the concentration using the equation
| y = ± | ( |
|
) |
|
± | ( |
|
) | ( |
|
) |
where, e0 is the permittivity of a vacuum, e is the relative permittivity (dielectric constant) of the solvent, a is the distance of closest approach of an ion and another species from the atmosphere, and 1 + ka/k is an effective radius of the ion atmosphere. The term k contains the value of the concentration (mi) and is expressed as
| k = | ( |
|
) | ½ |
( |
|
) | ½ |
where N is the Avogadro constant and the second term on the right-hand side is the ionic strength of the solution.
The field strength depended on the voltage and the distance between the electrodes. Because the distance between electrodes was kept constant, the field strength was raised only by increasing the voltage. Increased voltage caused the the ions to be transported more rapidly to the electrodes. To investigate the proper voltage and time factors, K concentrations were measured at different distances in the cell with a time interval of 10 min (Figure 1a and Figure 1b). At both 50 and 200 V, the K concentration generally increased with distance and time. Also illustrated in Figure 1 is that the concentration of K at the cathode increased with time within 80 min. Thus, more time or a greater voltage will be required to completely desorb K and/or to reach an equilibrium K concentration.
When an electrical field is applied to a KCl solution, a separation of charge within the solution is invoked, causing K+ and Cl- ions to separate and H2O to hydrolyze. The latter reaction is responsible for a change in pH. These reactions are
H2O = H+ +
OH-
K+ + Cl- = KOH
KCl = K+ + Cl-
Oxidation and reduction reactions could also take place where electrons are transferred, creating oxygen or hydrogen gas, depending on the voltage and polarity of the electrode, viz.,
H2O = ½ O2gas +
2H+ + 2e-
2H2O + 2e- = H2gas +
2OH-
Some doubt exists, however, as to whether or not voltages are high enough to cause these reactions to occur at the electrodes.
The pH in the electrodialysis cell was monitored because the pH may exert an influence on the desorption and solubility rates. The pH in the right cell (in the direction of the cathode) ranged from 5.8 to 9.6 (Figure 2a, Figure 2b, and Figure 2c), corresponding to reactions noted in equation (6). The greater the voltage that was applied, the faster the pH increased. At 400 V especially the change of pH was rapid. So the pH at the cathode increased with increasing time and voltage. The pH at the center region of the cell, however, should not change greatly during these processes. Nevertheless, the pH at each distance increased with the passage of time. Hydrolysis may effect a replacement of the exchangeable K+ by H+ if the clay suspension is placed near the anode.
Changes in pH in the cell were also observed from the color change of the bromocresol purple solution. The color of the bromocresol purple (bcp) solution became more blue around the cathode and yellow around the anode with increasing time, and these colors changed rapidly when the applied voltage increased (Figure 3).
The adsorbance of the bcp was measured to investigate the electrical potential at the cathode, with the hope that it may also be useful as a meaure of electrical potential at any point in the solution. This has the advantage of avoiding the use of electrodes with their accompanying uncertainties in colloidal supsensions. Standard calibration curves at the wavelengths of 275 nm, 425 nm, and 590 nm were obtained from standard solutions of 0.25, 0.50, 1.03, 2.06, and 4.12 mg/L (Figure 4). The wavelength of 425 nm yielded the most linear standard curve. Figure 5 shows that the concentration of bcp generally increased with time. The electrical potential at the cathode was then calculated using Equation (4) and values for mi determined from each of the three wavelengths analyzed. Results are reported in Figure 6. The average value of electrical potential calculated from the three wavelength was 0.018207 V. The value of electrical potential was almost constant with the passage of time. Thus electrical potential at the cathode was virtually independent of time.
More data at each distance in the cell are required to see a whole pattern of the electrical potential in the cell. However, obtaining the actual electrical potential is somewhat problematic. First of all, in principle the term ziFy has a spatial limitation because it is generally applied to interfacial regions and potentials at electrode surfaces. In an electrolyte solution, y arises from the charge on the ions and from their interaction with surrounding electrified interfacial surfaces. It represents the potential at the surface of a charged body or ionic sphere, so applying this term to the electrodialysis cell directly, where distances are in cm rather than nm, may be a problem. An effort must be mounted to investigate the relationship between electrical potential and outer voltage potential.
Through the relationships between K+ concentration and time, between pH and voltage, and between electrical potential at the cathode and time, fundamental information of electrodialysis was collected. The continuous increase of K+ concentration showed that we have to extend the time to reach equilibrium, or greater voltages are required. The greater the voltage applied to the electrodialysis system, the larger the change in pH. But if the very high voltage, such as 400 V, is excluded from the cell, the pH change in the cell will have no effect on K desorption. A compromise between time and voltage appears imminent.
More information is required before we can apply the principles of chemical kinetics and thermodynamics to our electrodialysis apparatus. The relationships between desorbed K+ concentration and temperature, and among electrical potential, distance and applied voltage will be examined in the next experiments.
Figure 1a. Relation between K concentration and time at 50 V.
Figure 1b. Relation between K concentration and time at 200 V.
Figure 2a. The change of pH with time at 50 V.
Figure 2b. The change of pH with time at 200 V.
Figure 2c. The change of pH with time at 400 V.
Figure 4. Standard curves of the bromocresol purple solutions at the cathode.
1J.W. Stucki is a Professor and K. Lee is a Graduate Research Assistant, Dept. of Natural Resources and Environmental Sciences, University of Illinois.
Bradfield, R. 1927. The use of electrodialysis in physico-chemical investigations of soils. First Internatl. Cong. Soil Sci. 2:264-278.
Gibbs, J.W. 1948. The collected works of J. Willard Gibbs. Vol. 1. Yale University Press. New Haven, CN.
Grimme, H., and K. Nemeth. 1979. The evaluation of soil K status by means of soil testing. Proc. Congr. Int. Potash Inst. 11: 99-108.
Low, P.F. 1951. Force fields and chemical equilibrium in heterogeneous systems with special reference to soils. Soil Sci. 71:409-418.
Löddesöl, A. 1932. Factors affecting the amount of electrodialyzable ions liberated from some soils. Soil Sci. 33:187-211.
Nemeth, K. 1979. The availability of nutrients in the soil as determined by electro-ultrafiltration (EUF). Advances in Agronomy. 31:155-187
Wilson, 1928. Exchangeable cations in soils as determined by means of normal ammonium chloride and electrodialysis. Soil Sci. 26:407-421.
