subsequent cooling of pure and contaminated samples, suggesting different conduction mechanisms in ... underground aquifers, with all the consequences to the environment and the society. Several ..... Solids, Wiley, London, 108. Molak A.
COMPLEX DIELECTRIC PERMITTIVITY AND ELECTRICAL CONDUCTIVITY MEASUREMENTS AS A DIAGNOSTIC TOOL FOR THE DETECTION OF HEAVY METALS ADSORBED IN BENTONITE SAMPLES V. SALTAS, N. VALLIANATOS
LYDAKIS-SIMANTIRIS,
P.
SOUPIOS
AND
F.
Department of Natural Resources & Environment, Technological Educational Institute of Crete, Chania, Greece
SUMMARY: The adsorption of heavy metals (Pb and Ni) in raw Wyoming bentonite was
investigated by means of broadband dielectric spectroscopy (BDS) in the frequency range 10 -2Hz – 106Hz and at elevated temperatures, up to 1000oC. In order to reveal various aspects of data interpretation, different formalisms of analysis were investigated, namely, complex electrical conductivity, dielectric modulus and Cole-Cole diagrams of complex impedance. The presence of different states of bound water in pure and contaminated bentonite samples results to nonreproducible recorded spectra at room temperature and it is necessary to measure at higher temperatures in order to clarify their contribution to the overall dielectric response. The observed spectral differences between pure and contaminated samples in modulus formalism are attributed to the contribution of heavy metal ions to the conduction mechanisms, which are thermally activated. A non-Arrhenius behavior of dc-conductivity is observed during heating and subsequent cooling of pure and contaminated samples, suggesting different conduction mechanisms in each case. 1. INTRODUCTION The development of effective methods for the detection, monitoring and remediation of subsurface contamination is of great importance in the scientific community, covering an interdisciplinary field of work, due to the high impact of pollution on ecosystems and the human health. Especially, heavy metals present in many industrial wastewater, cause the contamination of broad areas in industrial zones, and, even worse, the contamination of surface water and underground aquifers, with all the consequences to the environment and the society. Several methods have been developed and applied in order to address these problems, with the adsorption process being one of the most effective and adaptable (Brandl, 2005). Raw and
modified (chemically or thermally) clay minerals have been used extensively as adsorbents of heavy metals, due to their high abundance and low cost (Eren et al., 2009; Bhattacharyya and Gupta, 2008a; 2008b). The adsorption of contaminants in porous materials is a process which, to some extent, depends on the transport of the contaminants and their diffusion in active sites of the porous materials, via charge-dipole and/or dipole-dipole interactions [Lee et al., 2003; Saltas et al., 2007b]. In this sense, broadband dielectric spectroscopy (BDS) may proved to be a versatile technique to study charge transport which controls contaminant dynamics. Indeed, dielectric spectroscopy has been applied succesfully to the investigation of organic and inorganic contaminants, which appear in soils and porous materials and it has been also suggested as a bioindicator of environmental pollution (Ajo-Frankli et al., 2004; Bezzar and Ghomari, 2009; Shang et al., 2004; Saltas et al., 2007a; 2007b). In BDS, the interaction of an ac-electric field with any kind of material may be investigated over a broad frequency range (mHz up to GHz) with temperature and/or pressure as parameters, resulting in valuable information about polarization and charge transport mechanisms in the inner or grain boundaries of the material under investigation. In general, dielectric permittivity and electric conductivity of porous materials are strongly affected by several factors such as thermodynamic conditions (pressure and temperature), chemical composition of the solid and liquid phases, water content, porosity, grain size, etc. (Freund, 2003; Glover and Vine, 1992; Myounghak et al., 2007). So, in the present study, Pb- and Ni-contaminated bentonite samples were investigated by BDS over a broad frequency range and at elevated temperatures up to 1000oC. Different formalisms of data analysis were carried out, in order to study charge transport and possible relaxation mechanisms that are related to the contaminants in the solid matrix.
2. EXPERIMENTAL PROCEDURE AND DATA ANALYSIS 2.1 Samples preparation Nickel and Pb-contaminated bentonite samples were prepared by dissolving bentonite in aqueous solutions of Ni(NO3)2 and Pb(NO3)2, respectively. The mixtures were stirred at 50oC, until equilibrium was achieved, and then the samples in powder form were obtained by centrifugation and dried at 105oC. The concentrations of heavy metals (20±5 ppm for both, Ni and Pb) in the bentonite samples were estimated indirectly, by means of AAS measurements of the metals in the remaining solutions. The contaminated diatomite samples in powder form, were pressed inside a cell under high pressure and disk-shaped specimens of 20mm diameter and 2mm thickness were obtained. Each specimen was mounted in the sample cell between two platinum parallel electrodes forming a sample capacitor. Good electric contacts were achieved by Au sputtering on both sides of the sample. 2.2 Experimental setup Complex dielectric permittivity and electrical conductivity measurements were carried out by means of a high-resolution broadband spectrometer (Novocontrol Alpha-N Analyzer) in conjunction with a NovoTherm 1400 high temperature sample cell (Saltas et al., 2008). The frequency range of the applied ac-electric field was between 10 -2 Hz and 106 Hz. Measurements were carried out at vacuum (P=5x10 -3 mbar) in order to diminish the effect of oxygen fugacity to the measured dielectric and electrical properties. Good electromagnetic
CRETE 2012
2
shielding was implemented to the whole sample cell in order to diminish noise problems which are common especially at low frequencies. Isothermal spectra of electrical response were recorded from room temperature up to 1000oC, with steps of 40 or 50oC during each of the heating and cooling cycles. 2.3 Representation of dielectric measurements The dielectric sample placed between the round plate electrodes of the capacitor can be considered as an equivalent electrical circuit comprised of a capacitance, C(ω), in parallel with a resistance, R(ω). These values which are the output of the analyzer are associated to the complex impedance Z* ( ) and dielectric permittivity * ( ) of the sample through the following relations:
1 Z( )
1 i R( )
*
C( )
(1)
and *
( )
C( ) i Co
i
1
i C0 Z* ( )
C0 R( )
(2)
r 2 / d is the capacitance of the empty sample cylindrical capacitor, with where C0 o distance d between the electrodes and r their radius, ω=2πf and o is the permittivity of the *
vacuum. The specific complex conductivity permittivity by the equation: *
i
i
* o
of the sample is related to the dielectric
1
i
o
1
o
(3)
Alternatively, the reciprocal permittivity or electric modulus M * representation can be used to describe relaxation processes. It was first introduced by MćCrum et al. (MćCrum et al., 1967) and was applied successfully in cases where interfacial polarization causes large variations in dielectric permittivity at the low frequency range (Tsangaris et al., 1998; Molak et al., 2005). Electric modulus M*, which is an electrical analogue to the mechanical shear modulus, is defined as, M* ( )
*
1 ( )
M ( ) iM ( )
2
2
i
2
2
M
1
e 0
i t
d (t) dt dt
(4)
where Φ(t) describes the decay of the electric field within the sample and M ∞=1/ε΄∞ is the inverse of dielectric permittivity at high frequencies. The main advantage of the above representation is that the contribution of electrode polarization effects is negligible, allowing the evaluation of relaxation processes at the low frequency range of dielectric spectra (Tsangaris et al., 1998). All the aforementioned formalisms are equivalent to each other containing the same information, but the transformation from one dependent variable to another, may accent different relaxation process or conduction mechanisms. In the case of ionic solids such as minerals or rock-forming minerals, where no intrinsic relaxation phenomena occur, the dielectric permittivity representation is inadequate while complex impedance and complex conductivity formalisms are used, in order to distinguish between various extrinsic and intrinsic conduction mechanisms. Alternatively, the electric modulus M* representation is very useful, since, the conduction
CRETE 2012
3
processes which cause a steady increase in the imaginary part of dielectric permittivity ε*, appear as a loss peak at low frequencies in M* formalism (Elliott, 1994).
Figure 1. 3D-plane plots of imaginary part of dielectric Modulus (M¨) as a function of frequency and temperature of (a) pure (b) Ni-contaminated and (c) Pb-contaminated bentonite samples.
CRETE 2012
4
3. EXPERIMENTAL RESULTS AND DISCUSSION Repetitive dielectric and electrical measurements in Pb- and Ni-contaminated bentonite samples, that were carried out at room temperature and at various low concentrations of heavy metals (from a few ppm to several decades ppm), did not exhibit any indicative spectral feature of their presence in the solid matrix, as compared to the pure bentonite samples. In these cases, the role of water in bound form is dominant, causing the screening of any other contributions to the recorded dielectric spectra and the appearance of non-reproducible spectral features, due to small variations of water content (Saltas et al., 2008). It is well-known that BDS is very sensitive to the presence of small amounts of water in samples such as clay minerals and porous materials (Saltas et al., 2008). In the case of bentonite which belongs to smectite group with a layered crystal structure, the bound water consists of interlayer water and water adsorbed on the external surfaces of the quasi-crystals, with different dielectric response in each case. Furthermore, the contributions of the bulk and the surface conductivity to the total electrical conductivity are strongly affected by these different states of bound water. The above contributions of various relaxation and conduction mechanisms to the overall dielectric response are clearly indicated in the dielectric modulus representation of dielectric data. In Figure 1, recorded data are presented in dielectric modulus diagram form, as a function of frequency and temperature, for pure, Ni- and Pb-contaminated bentonite samples. 1800 1500 1200 900
T [K]
600
300
-5
10
-7
10
pure bentonite Ni-bentonite Pb-bentonite
-9
10
-11
dc-conductivity, σO [S/cm]
10
-13
10
(a) heating
-15
10
-5
10
-7
10
-9
10
-11
10
-13
10
(b) cooling -15
10
0,5
1,0
1,5
2,0
2,5
3,0
3,5
-1
1000/T [K ]
Figure 2. Arrhenious plots of dc-conductivity σo, of pure and contaminated bentonite samples during (a) heating up to 1000oC and (b) subsequent cooling to RT.
CRETE 2012
5
0
0
4
Z'' [Ohms]
-1x10
6
-1x10
4
-2x10
6
-2x10 4
-3x10
6
-3x10
4
-4x10
Pb-bentonite
Ni-bentonite
pure bentonite 4
6
-5x10
-4x10 0,0
4
4,0x10
Z' [Ohms]
0,0
6
6
2,0x10 4,0x10
0,0
6
6
2,0x10 4,0x10
Z' [Ohms]
Figure 3. Cole-Cole plots of complex impedance for pure, Ni- and Pb-contaminated bentonite samples at 400oC, during the cooling cycle. Note the different scale of impedances in each plot. At room temperature, the behavior is rather complicated with low values of dielectric modulus and various broad spectral features, due to the presence of different states of bound water, as it was mentioned previously. However, at temperatures higher than 200oC, two well-defined broad peaks are developed in the case of the pure sample (peaks 1,2 in Figure 1a), which are shifted to higher frequencies, as temperature increases. In the case of the contaminated samples, an additional peak is also observed at the lower frequency range, following a similar behavior with temperature for both, Ni- and Pb-contaminated samples. The high frequency relaxation peak (peak 2) is attributed to the polarization of different states of bound water, while the relaxation peak observed in the medium frequency range (peak 1) is attributed to the grains polarization, due to the migration of ions along the clay particles. The latter peak is affected by the bound water which forms conduction paths to the ions motion and is also temperature dependent. The low frequency relaxation peak (peak 3) is attributed to the total conductivity of the contaminated samples, which increases due to the presence of the heavy metal ions. All the aforementioned relaxation and conduction peaks are hardly distinguished in complex ac-conductivity spectra (not shown here) but, we may extract additional information from the dcconductivity values (measured at 10-2Hz). The Arrhenious plots (dc-conductivity versus inverse temperature) of pure and contaminated samples during heating up to 1000oC and subsequent cooling at RT are depicted in Figure 2. The contaminated samples exhibit almost the same behavior either during heating or subsequent cooling, due to similar concentrations of Ni and Pb in them. However, the dc-conductivity values of the pure sample are different as compared to the contaminated samples, over a broad temperature range (400-900K during heating). In all cases, dc-conductivity increases non-linearly with temperature due to the contribution of different conduction mechanisms which are thermally activated. At low and medium temperatures, hydrogen-related and hopping (small polaron) conduction are responsible for electrical conduction, while at high temperatures ionic conduction dominates. The step-like feature between 600-900oC may be attributed to the dehydration of bentonite and is observed only in the case of the pure bentonite sample.
CRETE 2012
6
An alternative representation of dielectric data is the Cole-Cole diagram where the imaginary part of complex impedance is plotted versus its real part, over the measured frequency range. Representative Cole-Cole diagrams of the measured samples at 400oC during the cooling cycle are shown in Figure 3. A typical spectrum consists of a depressed semicircle which is composed of overlapping arcs that are related to grain boundaries and grains interior, as well as a tail at the right side (low frequencies) due to electrode response. However, it is hard to distinguish between the contribution of heavy metals and the other relaxation and conduction mechanisms due to the similar distribution of relaxation times in each case. Lower values of impedance are recorded for the pure bentonite sample due to the hydration of the sample during the cooling cycle.
4. CONCLUSIONS Broadband dielectric spectroscopy measurements were carried out in pure, Ni- and Pbcontaminated bentonite samples at elevated temperatures, in order to investigate possible relaxation and conduction mechanisms that are related to the contaminants and are thermally activated. Electric modulus representation is proved to be the most suitable for the analysis of dielectric response and different spectral features were identified and attributed to bound water, grains polarization and electrical conduction, including the contribution of the heavy metal ions. Furthermore, the Arrhenious plots of dc-conductivity of pure and contaminated bentonite samples indicated different thermally activated conduction mechanisms. In conclusion, BDS is proved to be sensitive at low concentrations of heavy metals when measurements are carried out at temperatures much higher than RT but a detailed investigation is necessary in order to relate the concentration of the heavy metals with the spectral parameters.
ACKNOWLEDGEMENTS This work is partially suported by the European Social Fund and National Resources in the framework of the project THALIS (32.4.4), entitled ‘‘GEODIAMETRIS – Integrated Geoinformatics Technologies for Time-Lapse Monitoring of Land Pollution from the Disposal of Olive-Oil Mills Waste’’.
REFERENCES Ajo-Frankli J.B., Geller J.T. & al (2004) The dielectric properties of granular media saturated with DNAPL/water mixtures. Geophys. Res. Lett. 31, L17501. Bezzar A. and Ghomari F. (2009) Nondestructive test to track pollutant transport into landfill liners. Environmental Geology 57, 285-290. Bhattacharyya K. G. and Gupta S. S. (2008a) Adsorption of Fe(III), Co(II) and Ni(II) on ZrOkaolinite and Zr)-montmorillonite surfaces in aqueous medium. Colloids and Surfaces A: Physicochem. Eng. Asppects 317, 71-79. Bhattacharyya K. G. and Gupta S. S. (2008b) Adsorption of a few heavy metals on natural and modified kaolinite and montmorillonite: A review. Advances in Colloid and Interface Science 140, 114-131. Bradl H. D. (ed) (2005) Heavy Metals in the Environment: Origin, Interaction and Remediation. Interface Science and Technology, vol. 6.
CRETE 2012
7
Elliott S. R. (1994) Use of the modulus formalism in the analysis of ac conductivity data for ionic glasses, Journal of Non-Crystalline Solids 170, 97. Eren E., Afsin B. and Onal Y. (2009) Removal of lead ions by acid activated and manganese oxide-coated bentonite. Journal of Hazardous Materials 161, 677-685. Freund F., J. (2003) On the electrical conductivity structure of the stable continental crust, Journal of Geodynamics 35, 353-388. Glover P. W. and Vine F. J., (1992) Electrical conductivity of carbon-bearing granulite at raised temperatures and pressures, Nature 360, 723-726. Lee J. H., Oh M. H. & al (2003) Dielectric dispersion characteristics of sand contaminated by heavy metal, landfill leachate and BTEX (02-104B). Journal of Hazardous Materials B105, 83-102. MćCrum N. G., Read B. E. and Williams G. (1967) Anelastic and Dielectric Effects in Polymer Solids, Wiley, London, 108. Molak A., Paluch M., Pawlus S., Klimontko J., Ujma Z. and Gruszka I. (2005) Electric modulus approach to the analysis of electric relaxation in highly conducting (Na0.75Bi0.25)(Mn0.25Nb0.75)O3 ceramics, J. Phys. D: Appl. Phys. 38, 1450. Myounghak O. Yongsung K. and Junboum P. (2007) Factors affecting the complex permittivity spectrum of soil at a low frequency range of 1kHz-10MHz. Environmental Geology 51, 821833. Saltas V., Triantis D. & al (2007a) Biomonitoring of environmental pollution using dielectric properties of tree leaves. Environmental Monitoring and Assessment 133, 69-78. Saltas V., Vallianatos F. & al (2007b) Dielectric and conductivity measurements as proxy method to monitor contamination in sandstone. Journal of Hazardous Materials 142, 520525. Saltas V., Vallianatos F. and Triantis D. (2008) Dielectric properties of non-swelling bentonite: The effect of temperature and water saturation. Journal of Non-Crystalline Solids 354, 55335541. Shang J. Q., Ding W. & al (2004) Detecting heavy metal contamination in soil using complex permittivity and artificial neural networks. Can. Geotech. J. 41, 1054-1067. Tsangaris G. M., Psarras G.C. and Kouloumbi N. (1998) Electric modulus and interfacial polarization in composite polymeric systems, Journal of Materials Science 33, 2027.
CRETE 2012
8
