Doklady Chemistry, Vol. 392, Nos. 1–3, 2003, pp. 233–237. Translated from Doklady Akademii Nauk, Vol. 392, No. 3, 2003, pp. 350–355. Original Russian Text Copyright © 2003 by Fedotov, Tret’yakov, Pozdnyakov.

CHEMISTRY

Residual Polarization Appearing during Unsteady Water Infiltration through Colloidal Soil Structures G. N. Fedotov, Academician Yu. D. Tret’yakov, and A. I. Pozdnyakov Received March 20, 2002

The properties of soils as natural colloidal systems are substantially dictated by the organomineral gel (OMG) spread over the surface of coarse particulates [1, 2]. The electrical properties of soils are basically due to a salt soil solution in them and a double electrical layer at the soil particle–soil solution interface. The soil OMG and soil minerals bear a negative charge everywhere except for tropical climatic regions [3]. Therefore, when water and aqueous fluids move through the soil, the positive ions of the diffuse layer must shift in the same direction. It has been shown [4, 5] that a positive charge is accumulated in the lower soil layers during long-term water infiltration. However, under certain unsteady regimes of water movement through soils (rains, watering), we observed a fundamentally different situation: the lower soil layers accumulated a negative, rather than a positive, charge. Here, we report experimental evidence of this phenomenon and attempt to interpret its origin. The organomineral greenhouse substrate is the most suitable object for studying solution dynamics. It contains abundant colloidal particles; the greenhouse substrate does not differ from soils in this respect, although it is not a soil per se. Because the greenhouse substrate has a high porosity and a minimal content of hydrophobic compounds (in contrast to peat), it provides a high water flow rate over a wide range of water contents. The properties of the greenhouse substrate, which we determined using standard procedures [6, 7], are compiled in Table 1. In laboratory runs, 650 g of the soil was placed in a crimped plastic tube (56 mm in diameter, 400 mm in height) and vibrocompacted into a bed 350 mm high. The potential difference (PD) was measured between electrodes positioned 50 and 300 mm from the soil surface. The measuring electrodes used were standard Ag/AgCl electrodes connected to the soil through agar salt bridges. Water or an aqueous solution was Moscow State University, Vorob’evy gory, Moscow, 119992 Russia Moscow State Forest University, Mytishchi-5, Moscow oblast, 141005 Russia

poured in the amount of 200 mL onto the soil for 2–3 s. The measurement error in the PD was within 10%. Greenhouse experiments were carried out as follows. A trench was excavated in a greenhouse. Electrodes in the trench were also positioned 50 and 300 mm deep. The PD was measured during watering at a flow rate of 80 L/m2. The dc and ac electrical resistivities of soils were measured using the four-point probe technique [8]. A GFG-8217A audio-frequency oscillator and a B5-48 dc power supply were used as voltage sources. The voltage and current were measured by Mastech M890 digital multimeters with an internal resistance of 10 M Ω. The measurement error was within 1%. Distilled water, aqueous solutions of various concentrations, and a colophony sol were used. The colophony sol was prepared by dropping a 2% alcoholic solution of colophony into distilled water under continuous stirring. The sol particle size was determined by measuring the light scattering at various wavelengths [9]. The PD measurements between the electrodes during soil watering in laboratory runs (Fig. 1) show that a negative charge is accumulated in the lower soil layer at low salt concentrations; the appearing PD rapidly annihilates with time. An increase in the salt concentration decreases the bias and then changes the bias sign. A flow potential should appear in soils in similar experiments. Consider how the flow potential changes when run conditions are changed. The soil adsorption complex of greenhouse soils contains dominant amounts of divalent calcium cations (Table 1). These cations, which have a high adsorptive capacity, are mostly located in the Stern layer and also contract the diffuse layer, decreasing the ξ potential of soils. An increase in the univalent cation concentration should shift the equilibrium of the adsorption layer to lower calcium concentrations and, accordingly, increase the ξ potential and flow potential. Therefore, if the flow potential is low because of a low ξ potential, the lower soil layer accumulates a negative charge. When the flow potential rises with increasing univalent cation concentration, the electric charge changes its sign. We intensified the substitution of univalent for divalent cations by using solutions of salts whose anions

0012-5008/03/0009-0233$25.00 © 2003 åÄIä “Nauka /Interperiodica”

234

FEDOTOV et al. PD, mV –7.5 PD, mV –5.0

1 2 3

–20 –2.5 1 2

–10

0

3

4 8 12 16 20 24 28 32 36 Time, h 100

0

200

Time, s

4 5, 6 10 Fig. 1. Potential difference that appears upon watering of disturbed soil structures vs. solution composition: (1) distilled water, (2) tap water, and (3–6) KCl solutions in distilled water of concentration (mol/L) (3) 0.02, (4) 0.05, (5) 0.1, and (6) 0.2 M. The inset: plots of PD between soil layers (depths are specified) vs. time elapsed after watering: (1) 5–50, (2) 5–30, and (3) 5–20 cm.

form insoluble compounds with calcium cations. The results confirmed our expectations (Fig. 2). The lower the solubility product of the resulting compound, the larger the bias, with a positive charge accumulating in the lower soil layer. Thus, the strengthening influence of the flow potential on the infiltration potential changes the sign of charges accumulated in the lower soil layer from negative to positive. Therefore, we may suggest that water infiltration through the soil involves an interplay of two mechanisms of charge separation. One mechanism is the familiar flow potential, while the other mechanism, causing the opposite charge separation, is unknown. Clearly, the flow-potential model of a fixed negatively charged network surrounded by a diffuse cation layer fails to explain the displacement of negative

charges in the direction of fluid movement. In this context, the model needs refining. Water and true solutions, when infiltrating through the greenhouse substrate bed, become turbid and give the Tyndall cone in oblique lighting [10]. Therefore, presumably, colloidal particles covering the network are partially peptized when they interact with water or solutions. If we presume that moving water converts part of the colloidal particles to a sol, then the negative charge can shift to the lower layer due to the retardation of the diffuse atmospheres of sol particles with respect to colloidal particles in the direction of movement. In fact, the observed phenomenon is similar to the Dorn effect [10], but the diffuse atmospheres around moving colloidal particles are retarded due to their interaction with the diffuse atmospheres of the fixed network, which is responsible for the observed charge displacement.

Table 1. Properties of an organomineral greenhouse soil Water content, % Kutilek total specific surface area, m2/g Solid-phase density, g/cm3 Nitrate content, mg/kg Potassium content, mg/100 g Calcium content, mg/100 g Magnesium content, mg/100 g Hydrolytic acidity, mg-equiv/100 g Base saturation, % Density, g/cm3

79.7 97.6 1.96 116 164 1200 31 1.8 98.0 0.42

Porosity, % Aqueous extract pH Salt extract pH Ammonium content, mg/kg Phosphorus (P2O5) content, mg/100 g Organic content, loss on ignition, % Sodium content, mg/100 g Total exchangeable bases, mg-equiv/100 g Cation exchange capacity, mg-equiv/100 g Salt content, mg-equiv/100 g DOKLADY CHEMISTRY

Vol. 392

Nos. 1–3

78 6.9 6.5 41 25.1 28.3 74 82.3 84.1 0.9 2003

RESIDUAL POLARIZATION

If the assumption that the peptization of soil colloids is responsible for the appearance of electrical fields with an orientation opposite to that of the flow potential is correct, then the use of a ready colloidal solution for soil watering should enhance this effect. To check this, we watered the greenhouse soil with a colophony sol (particle size, 140 nm) with a concentration of 500 mg/L. As a result, the bias increased almost threefold (from 35–38 to 90–100 mV) compared to that after watering with distilled water. These results confirmed our idea concerning the appearance of an unknown potential. We called this potential “colloidal solution infiltration factor” (CSIP). The above results were obtained from laboratory experiments on samples with a disturbed soil structure. Under the conditions of a greenhouse experiment (where the soil structure is undisturbed), the situation is somewhat different (Fig. 1, inset). The orientation of electric fields persists, but their intensity and lifetime are markedly different. The results from greenhouse experiments on watering disturbed and undisturbed soils with distilled water are compiled in Table 2. When analyzing the appearance of the PD upon water movement through soils under unsteady conditions, we proceeded from the common idea that soils are complex dispersed systems that contain aqueous salt solutions and that these solutions control the electrical properties of soils, above all, the electrical conductivity. However, the above model fails to interpret the experimental results; in soils with higher conductivity (in which the PD should disappear more rapidly), the lifetime of the PD is three orders of magnitude longer. Long-lived polarized states are well known from physics; they are associated with “freezing” of an ensemble of aligned dipoles that cannot misalign spontaneously. A similar mechanism can be presumed for soil processes but with colloidal particles as the dipoles [11]. The results are explicable from this standpoint. Where the colloidal network is strongly linked (undisturbed soils), it is more difficult to build a polarized structure from dipoles, but dipole misalignment is slower. In soils with less linked colloidal networks, dipoles are more mobile and are aligned more readily, but their initial state is recovered more rapidly. To validate the model, we attempted to recover the colloidal network in disturbed soil structures. A routine procedure of soil science, namely, filling a sample with water followed by spontaneous water removal over a day, did not give a positive result. The lifetime of residual polarization in such samples after watering was a little longer, but no more than 5–6 min. Since, in our opinion, long periods of time and a directed movement of colloidal particles are required for the network to recover, we fed distilled water to a soil sample at a flow rate of 5–8 mL/min for 6–8 h. Changes in the state of the system were monitored by measuring the PD across the sample (Fig. 3). The PD DOKLADY CHEMISTRY

Vol. 392

Nos. 1–3

2003

235

PD, mV

–20 0.1

0

0.2

0.3

C, mol/L 0.4 1

20 2

40

3

60

Fig. 2. Plots of the PD during soil watering with alkalimetal salt solutions vs. salt concentration for (1) potassium chloride, potassium bromide, and potassium iodide; (2) sodium hydrophosphate; and (3) potassium fluoride.

acquires a constant level after 6 h of watering (curve 1). After stopping watering, polarization decays for several hours (curve 2). The polarization direction obtained under relatively steady-state conditions corresponds to the flow potential. Ordinary watering with water after the PD has vanished results in CSIP charge separation, and polarization persists for several hours (curve 3). This substantiates our idea of the orientational–polarizational character of the appearing PD. It is also worth noting that, regardless of the mechanism of PD generation, the type of soil colloidal network controls the lifetime of a polarized state. Recall that, if soil ions responsible for electrical conduction had the same mobility as they have in solution, the polarization effect could not exist for a long period of time. The interactions of colloidal particles with one another and the associated formation of coagulation structures have been comprehensively studied by colloid chemistry [11]. Evidently, the counterions of colloidal sol particles behave differently from the colloidal particles of a structured system (gel) since the former occur in the force field of an ensemble of particles rather than of a single, autonomous particle. The diffuse atmospheres of colloidal particles are no longer autonomous, and the bond energy of separate counterions in the structure exceeds the energy of interaction of Table 2. Electrical effects appearing during soil watering with distilled water Parameter Effect value, mV PD lifetime Soil resistivity, Ω cm

Disturbed soil Undisturbed structure soil structure 35–40

5–8

1–2 min

2–3 days

8000

3000–6000

236

FEDOTOV et al.

threshold intensity can detach particles from the structure. We believe that similar mechanisms in nature are unlikely, because in fact the OMG structure is fully destroyed. In our opinion, the interpretation of the electrical properties of soils with allowance for structural rearrangements in the OMG is much more promising.

PD, mV –15

–10 3

–5

Time, h 0

1

2

3

4

5

Low-frequency ac and dc measurements were carried out. The resistivity of the greenhouse substrate was measured as a function of field intensity using the fourpoint probe technique (Fig. 4).

5 2 10

1

15

Fig. 3. Plots of the PD in the greenhouse substrate vs. time (1) during continuous watering at 8.5 mL/min, (2) after stopping watering, and (3) after watering a sample prepared in this manner.

ρ, kΩ cm 100

1 50 3 2 4 5, 6, 7, 8 0

It is hardly possible that the observed residual polarization in soils arising under the action of a water flow is the only result of colloidal structuring. Such a structuring must be expressed in soil conductivity. Therefore, we studied soil conductivity, keeping in mind that the network can restructure when current flows.

10

20

30

40

50 60 E, mV/cm

Fig. 4. Plots of the resistivity of the greenhouse substrate vs. electrical field intensity in soils measured with (1–7) ac at (1) 10, (2) 20, (3) 30, (4) 40, (5) 60, (6) 80, and (7) 100 Hz and (8) dc.

a counterion from a diffuse layer with an autonomous colloidal particle. Soil depolarization under such conditions should have some specifics. Charge carriers (counterions) can participate as autonomous particles in charge transport through soils when a threshold intensity of the internal electrical field is surpassed; this

Both dc and ac (60–100 Hz frequencies) resistivities are in fact unchanged. When the frequency decreases, the resistivity in low-intensity fields increases; at 10 Hz, the resistivity is more than one order of magnitude higher. The resistivity results can be interpreted as arising from the soil colloidal network. Let us regard the colloidal network as an ensemble of colloidal dipoles that connect electrodes. The direct currents of dipole chains do not destroy a chain; frequencies of 60–100 Hz do not have enough time to destroy it. In both cases, the electrical conductivity has the highest value. At low frequencies, chains can be in part destroyed because of realignment of colloidal particles, and the amount of chains responsible for current passage decreases. This decrease is especially notable in low-intensity fields, where the rotation of colloidal dipoles has a low speed. A growth in the field intensity increases the dipole rotation speed; the amount of broken chains per unit time decreases, and the conductivity is on the level of the dc conductivity. Charge supply and sink from an outer source do not change the system and do not destroy the colloidal network; here, the network functions merely as a transmitter. We regard an ensemble of oriented dipoles (a polarized colloidal network) as a system with a PD; such a system can depolarize only through restructuring of the colloidal network, which requires much energy. Therefore, the process in soils with a strong colloidal network is rather slow. This approach might be aimed at understanding of first- and second-order seismoelectrical effects and mechanoelectrical phenomena [12, 13]. To this end, one should bear in mind that the surface of any natural body is covered with a gel layer [14, 15] and that this layer is less pronounced than in soils. The aforesaid implies that a united soil colloidal network and its restructuring control the appearance and lifetime of polarized states in soils. DOKLADY CHEMISTRY

Vol. 392

Nos. 1–3

2003

RESIDUAL POLARIZATION

REFERENCES 1. Tyulin, A.F., Organo-mineral’nye kolloidy pochv, ikh genezis i znachenie dlya kornevogo pitaniya vysshikh rastenii (Organomineral Colloids, Their Genesis and Role in the Root Nutrition of Higher Plants), Moscow: Akad. Nauk SSSR, 1958. 2. Zolotareva, B.N., Gidrofil’nye kolloidy i pochvoobrazovanie (Hydrophilic Colloids and Soil Formation), Moscow: Nauka, 1982. 3. Rowell, D.L., Soil Science: Methods and Applications, Longman, 1994. 4. Semenov, A.S., Elektrorazvedka metodom estestvennogo elektricheskogo polya (Electric Geophysical Exploration by the Natural Electric Field Method), Leningrad: Nedra, 1980. 5. Borovinskaya, L.B., Pochvovedenie, 1970, no. 11, p. 29. 6. Praktikum po agrokhimii (Laboratory Manual on Agricultural Chemistry), Mineev, V.G., Ed., Moscow: Mos. Gos. Univ., 1989. 7. Gasanov, A.M., Praktikum po pochvovedeniyu (Laboratory Manual on Soil Science), Moscow: Mos. Gos. Univ. Prirodoobustroistva, 2000.

DOKLADY CHEMISTRY

Vol. 392

Nos. 1–3

2003

237

8. Vadyunina, A.F. and Korchagina, Z.A., Metody issledovaniya fizicheskikh svoistv pochv i gruntov (Methods of Investigation of Soils and Grounds), Moscow: Vysshaya Shkola, 1973. 9. Grigorov, O.N., Koz’mina, Z.P., Fridrikhsberg, D.A., et al., Rukovodstvo k prakticheskim rabotam po kolloidnoi khimii (Laboratory Manual on Colloid Chemistrty), Moscow: Khimiya, 1964. 10. Fridrikhsberg, D.A., Kurs kolloidnoi khimii (Course of Colloid Chemistry), Leningrad: Khimiya, 1984. 11. Efremov, I.F., Periodicheskie Kolloidnye Struktury (Periodical Colloidal Structures), Leningrad: Khimiya, 1971. 12. Ivanov, A.G., Dokl. Akad. Nauk SSSR, 1939, vol. 24, no. 1, pp. 41–43. 13. Sobolev, G.A. and Demin, V.M., Mekhanoelektricheskie yavleniya v Zemle (Mechanoelectrical Phenomena in the Earth), Moscow: Nauka, 1980. 14. Martynov, G.A., Kolloidn. Zh., 1978, vol. 40, no. 6, p. 1110. 15. Lipson, G.A. and Kolodieva, T.S., Kolloidn. Zh., 1972, vol. 34, no. 2, p. 235.