Biomonitoring of Environmental Pollution Using Dielectric Properties of Tree Leaves

In the present work, dielectric measurements were performed in plane-tree leaves collected from a polluted urban site and a natural unpolluted one, in order to investigate the sensitivity of dielectric relaxation spectroscopy to the detection of
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  Biomonitoring of Environmental Pollution Using DielectricProperties of Tree Leaves V. Saltas  &  D. Triantis  &  T. Manios  &  F. Vallianatos Received: 7 June 2006 /Accepted: 26 October 2006 / Published online: 14 December 2006 # Springer Science + Business Media B.V. 2006 Abstract  In the present work, dielectric measure-ments were performed in plane-tree leaves collectedfrom a polluted urban site and a natural unpollutedone, in order to investigate the sensitivity of dielectricrelaxation spectroscopy to the detection of heavymetals pollution. Although heavy metal concentra-tions at the urban site are not found considerablehigher than those at the natural site, the two samplesexhibit different features in the recorded dielectricspectra. Evaluation of experimental data suggests that the dielectric modulus (  M  *( ω )) representation is themost suitable for accenting the different dielectricrelaxation processes of each sample. The imaginary part of dielectric modulus  M  00 ( ω ) was fitted using athree-term Havriliak   –   Negami relaxation function,with fitting parameters, which depend on the concen-trations of heavy metals. The lower frequencyrelaxation process is attributed to the ionic conduc-tivity of the samples, while the two others are due todifferent charge transport mechanisms of   α -response.The investigation of plane-tree leaves in terms of their dielectric properties can be considered as a promising biomonitoring for environmental pollution. Keywords  Biomonitoring.Dielectricspectroscopy.Heavymetals.Leaves.Pollution 1 Introduction In recent years, analysis of trace elements in treeleaves has been used as a potential biomonitoringmethod for identifying spreading of pollution fromindustrial, domestic and vehicle emissions (Alfani,Baldantoni, Maistro, Bartoli, & Virzo De Santo, 1997,2000; Alfani, Bartoli, Rutigliano, Maistro, & VirzoDe Santo, 1996; Alfani, Maistro, Iovieno, Rutigliano,& Bartoli, 1996; Maisto, Baldantoni, De Marco,Alfani, & Virzo De Santo, 2003; Swaileh, Hussein,& Abu-Elhaj, 2004). However, the above chemicalanalyses are expensive and time consuming processesand alternative methods such as magnetic surveys of tree leaves have been proposed to identify anddelineate high polluted areas, in addition to the clas-sical air quality monitoring systems (Georgeaud, Environ Monit Assess (2007) 133:69  –  78DOI 10.1007/s10661-006-9560-7V. Saltas ( * ) : F. VallianatosDepartment of Natural Resources and Environment,Technological Educational Institute of Crete,3 Romanou St, Chalepa,73 133 Chania, Crete, Greecee-mail: D. TriantisMaterials Research Laboratory, Department of Electronics,Technological Educational Institute of Athens,Athens, GreeceT. ManiosSchool of Agricultural Technology,Technological Educational Institute of Crete,Chania, Crete, Greece  Rochette, Ambrosi, Vandamme, & Williamson, 1997;Moreno, Sagnotti, Diranès-Turell, Winkler, & Cas-cella, 2003; Urbat, Lehndorff, & Schwark, 2004). Magnetic properties of tree leaves are strongly af-fected by atmospheric particulate matter, which con-sists of spherules and grains of magnetite, metalliciron or other heavy metals (Matzka & Maher, 1999;Muxworthy, Schmidbauer, & Petersen, 2002).It is obvious that all the above particulate matter will also affect the electric properties of tree leaves,which act as natural heavy metals accumulators.Dielectric spectroscopy has been applied successfullyto investigate pollution (organic or inorganic), whichappears in soils and porous materials, and promisingresults were obtained concerning the role of ionicstrength, moisture content and organic liquids (Kaya,2001; Kaya & Fang, 1997; Rowe, Shang, & Xie, 2001; Saltas, Vallianatos, Soupios, Makris, & Triantis,2006; Shang, Ding, Rowe, & Josic, 2004). Dielectric spectroscopy has been also proposed by Hill, Dissado,and Pathmanathan (1987) and Hill et al. (1986) as an informative investigative technique in complex sys-tems such as tree leaves, but according to our knowledge, studies that have been carried out tocorrelate dielectric properties with changes in traceelements concentrations due to environmental pollu-tion, are very limited (Czuba & Kraszewski, 1994).We have to mention that electrical conducting particles incorporated in an insulating material, mayaffect the electrical behaviour of such mixtures,leading to many interesting applications (Lux, 1993;Roldughin & Vysotskii, 2002). It has been reported byseveral authors that deciduous trees, including plane-trees, can accumulate a variety of compounds fromthe atmosphere and can therefore be used in the de-tection of aerial heavy metal pollution (Alaimoet al., 2000; Granier & Chevreuil, 1992; Luyssaert, Raitio, Vervaeke, Mertens, & Lust, 2002; Piczak,Le ś niewicz, & Zyrnicki, 2003).So, in the present contribution, we are focussing onthe dielectric properties of plane-tree leaves collectedfrom an urban and a natural site, in order to inves-tigate the applicability of dielectric spectroscopy tech-nique to detect environmental pollution and itssensitivity to different concentrations of heavy metals.Specifically, dielectric measurements over the fre-quency range from 10 − 3 to 10 6 Hz were carried out, inorder to reveal changes in dc-conductivity andrelaxation mechanisms related to the pollution. 2 Experimental Procedure and Analysis 2.1 Dielectric spectroscopy techniqueDielectric properties relate to the ability of a materialto polarise under the influence of an electric field. Thedielectric permittivity, which is a measure of thespecimen ’ s response to the applied field, is afrequency dependent complex function that providesinformation on the dynamical processes of thematerial. Its real part ( " 0 ) represents the energystorage to the material, while the imaginary part ( " 00 ) represents the energy losses due to polarizationand ionic conduction. There are four main types of  polarization mechanisms, namely electronic, atomic,orientation and interfacial or space charge polarization(Jonscher, 1983). Each of these mechanisms domi-nates a certain frequency range with a characteristicresonant frequency or relaxation frequency.In the present study, dielectric and conductivitymeasurements were carried out by means of a high-resolution broadband spectrometer, model Alpha-NAnalyzer, connected with a sample holder, modelBDS1200 supplied by Novocontrol (Schaumburg,1999). The frequency range of the applied ac electricfield was between 10 − 3 and 10 6 Hz. The specimenwas mounted in a sample cell between two parallelelectrodes forming a sample capacitor. Good electro-magnetic shielding was implemented to the wholesample holder in order to diminish noise problemsthat are common especially at low frequencies.2.2 Representation and modelling of dielectric dataThe dielectric sample placed in the capacitor can beconsidered as an equivalent electrical circuit com- prised of a capacitance,  C  ( ω ), in parallel with aresistance,  R ( ω ). These values are the output of thedielectric analyzer and are associated to the real andimaginary part of the complex dielectric permittivitythrough the relation: " ✻ ω ð Þ¼ " 0  i " 00 ¼  C   ω ð Þ d  " o  p  r  2   i d  ω  R  ω ð Þ " o  p  r  2  ð 1 Þ where  d   is the distance between the electrodes,  r   istheir radius,  ω =2 π  f    and  " o  is the permittivity of thevacuum. 70 Environ Monit Assess (2007) 133:69  –  78  In the case of a media with high concentration of free ions, the dc-conductivity has to be taken intoconsideration. This conductivity manifests itself in theimaginary part of the complex dielectric permittivity " ✻ ¼ " 0  i  σ o " o ω    N  þ " 00 ( )  ð 2 Þ where  σ o  is the specific dc-conductivity and the expo-nential factor,  N  , in most cases, equals to 1. Thespecific conductivity  σ * is related to the dielectric permittivity by the equation: σ ✻  ¼ σ 0  i σ 00 ¼ i ω " o  " ✻  1 ð Þ¼ ω " o " 00  i ω " o  " 0  1 ð Þ ð 3 Þ Alternatively, the reciprocal permittivity or dielec-tric modulus representation can be used to describerelaxation processes. Dielectric modulus  M  *, which isan electrical analogue to the mechanical shear modulus, is defined as (Donth, 1992),  M  ✻ ω ð Þ¼  1 " ✻ ω ð Þ ¼  M  0 ω ð Þþ iM  00 ω ð Þ¼  " 0 " 0 2 þ " 00 2  þ i  " 00 " 0 2 þ " 00 2  ð 4 Þ The advantage of the above representation is that the relaxation times obtained from the electricmodulus ( τ   M  ✻ ) representation as compared to thoseof dielectric permittivity ( τ  " ✻  ) scale as  τ   M  ✻ = τ  " ✻   " 1 =" s , (Frohlich, 1958) giving substantial differencesin systems with high dielectric strengths,  $ " ¼ " s  " 1 , where  ε s  and  ε ∞  are dielectric permittivitiesat zero and infinite frequency, respectively. Thismeans that a relaxation process in the modulusrepresentation appears at a higher frequency than thecorresponding process in  ε * and the ionic conduction process which causes a steady increase in  ε 00 , appearsas an increasing loss at low frequencies (Elliott,1994).In the  ε *( ω ) representation, the experimental dataare often modeled with a superposition of Havriliak   –   Negami (HN) dielectric relaxation functions and aconductivity term, as shown below (Havriliak & Negami, 1966). " ✻ ω ð Þ¼ " 0  i " 00 ¼ i  σ 0 " 0 ω    N  þ X nk  ¼ 1 $ " k  1 þ  i ωτ  " ✻ k  ð Þ α k  ð Þ β  k  þ " 1 k  " # ð 5 Þ where  τ   is a characteristic relaxation time. The shape parameters  α  and  β   are usually constrained to lie between 0 and 1; however   β   might exceed 1, providedthat   αβ   ≤ 1. Parameter   α  describes the broadness of the relaxation while  β   is an indicator of the peak asymmetry. For not too small values of   α  and  β  ,1/2 πτ   corresponds approximately to the frequency of the maximum peak in  ε 00 (Jonscher, 1983). Each termin the sum of Eq. 5, corresponds to different dis-tributions of relaxation processes.In the present case HN model functions wereapplied in the modulus formulation of experimentaldata. However, the analytical transformation of   ε * to  M  * by using Eq. 4 cannot be performed exactly, but in good approximation a proper model function isgiven by an expression that differs from the HNfunction mainly by the reciprocal ratio of frequencies,if we take into consideration the inverse sign of the  M   HN  relaxation increment (Kremer & Schönhals,2002; Schlosse, Schönhals, Carius, & Goering, 1993). 2.3 Samples preparationPlane-tree leaves were collected from a polluted urbanand a natural site, respectively, in order to ensure thedifferent amount of pollutants in them. The first sample was collected from the centre of Athens city, ahighly polluted area (HPA sample) due to vehicleemissions and domestic heating systems. The secondsample came from a low polluted area (LPA sample)in the countryside, far away from any industrialactivities. Sampling of fresh leaves was carried out in spring and was repeated in autumn, before the treesshed their leaves. In each case, leaves at 3 m height were removed from different branches resulting in atotal of 20 leaves.The sample material was mounted between tworound gold plated electrodes of 16 mm diameter, as a Environ Monit Assess (2007) 133:69  –  78 71  sandwich capacitor while press was applied to ensuregood electric contact of the leaf surface with theelectrode. The distance between the electrodes was0.2 mm. Prior to the placement of the leaf to thesample cell, dust was removed from its surface by asmooth cloth. For each of the samples, dielectricmeasurements were carried out separately in three tofour leaves, immediately after their collection and preparation. The measurements were repeated for twoto three different areas of the leaf surface, in order toensure the same dielectric behaviour all over the leaf surface. The resulted spectra did not show anyremarkable changes in spectral shape and absolutevalues of conductivity and dielectric constant. So, thedielectric spectra, which are presented in next section,are representative of all the collected leaves, in eachmeasured sample.Dielectric spectroscopy technique is very sensitiveto moisture and may be applied in order to determinewater content in porous materials (Rusiniak, 1998).So, the samples were dried in a furnace at 50°C for 2 days, in order to achieve the same water content inthem. Heavy metal concentrations in leaves weredetermined with an atomic absorption spectrometer (AAS). The concentrations of the heavy metals (Cd,Cu, Ni, Zn and Pb) were measured in the wholeamount of leaves in each of the samples that werecollected in autumn, and the results are shown inTable I. Considerable changes are observed in Ni,while concentration of Cd is the same in bothsamples. 3 Experimental Results and Discussion Dielectric measurements of leaves collected in springexhibit identical behaviour, in both, real and imagi-nary part of dielectric function. This is a predictableresult, since these leaves are new and should appear the same physiological state. The situation is quitedifferent in leaves collected at the end of autumn,reflecting all the possible changes that occur, due tothe accumulation of particulate matter and the resulted physiological reactions.The variation of the specific conductivity  σ * of thetwo samples, collected in autumn, as a function of frequency is shown in Figure 1. Visual inspectionshows that conductivity decreases with decreasingfrequency and reaches values between 10 − 13 and10 − 14 S/cm at very low frequencies. These lowconductivity values are indicative of the absence of free water in both samples. However, over the wholefrequency range, the conductivity of the HPA sampleis always higher than that of LPA sample. The latter observation may be related to the higher heavy metalsconcentration in the HPA sample, as compared withthe LPA sample. It has been reported by Czuba andKraszewski (1994) that, a larger amount of free water was observed in plants grown with Ni and Cd than inuntreated ones. However, in our case we have toexclude the contribution of excess free water in HPAsample to the measured conductivity, since the water should appear only in bound states, due to the longdrying of the samples.For both samples at high frequencies (>10 2 Hz),the conductivity may be described with a power lawfunction, as it is evident from the linear variation of  σ *( ω ) with frequency, in log  –  log representation (seedashed lines in Figure 1). In this universal law, whichhas been suggested by Jonscher (1983, 1999), σ ✻ ω ð Þ σ o  /  f    n ð 6 Þ where the exponent   n  lies between 0 and 1. In the present case we found that   n  takes the value of 0.95for both samples at high frequencies (>10 2 Hz), whilefor frequencies lower than 10 2 Hz, the slope decreasesand takes the values of 0.68 and 0.72 for the LPA andthe HPA sample, respectively. Deviations from line-arity in conductivity spectrum occur at very lowfrequencies in HPA sample and may be related to the Samples Metals accumulation (mg kg − 1 d.w.)Cd Cu Ni Zn PbHPA 0.38 15.11 4.17 16.74 8.13LPA 0.36 10.98 0.92 10.32 5.62 Table I  Heavy metalsaccumulation in measuredtree leaves (mg kg − 1 d.w.)72 Environ Monit Assess (2007) 133:69  –  78  different contributions of the various mechanisms todc-conductivity.The real and imaginary part of dielectric permit-tivity  ε *( ω ) is shown in Figure 2a and b, respectively.Real part of dielectric permittivity (dielectric constant, ε 0 ) at high frequencies lies between 2 and 3 for bothsamples. These values are far away from 80, which isthe permittivity of water, suggesting the absence of free water into the samples, as a consequence of their long drying. We have to mention here, that thecytoplasmic fluid of the cells is mainly comprised of water (85  –  90%). However, the above measuredvalues are very close to the relative permittivity of cellulose (3.5) from which the cell wall structure of leaves is formed. Different dispersion is obvious between the two samples and a step-like behaviour in ε 00 , which is indicative of relaxation mechanisms,suggests that the two samples exhibit different di-electric relaxation mechanisms. However, these fea-tures are difficult to distinguish and different formulations must be used in order to accent them.Based on Eq. 4, we have changed the representationof the dielectric data from  ε 0 and  ε 00 to  M  0 and  M  00 .The obtained modulus spectra  M  0 ( ω ) and  M  00 ( ω ) aredepicted in Figure 3a and b, respectively. In thisrepresentation, different relaxation mechanisms,which contribute to the overall response, are obviousonly by sight.In order to accent the different relaxation mecha-nisms, the modulus data (  M  00 ) were fitted by thefollowing three-term HN dielectric relaxation func-tion, giving very good correlation, as it is seen inFigure 4a and b for HPA and LPA samples,respectively.  M  ✻ ω ð Þ¼ X 3 k  ¼ 1 $  M  k  1 þ  i ωτ   M  ✻ k  ð Þ α k  ð Þ β  k  þ  M  1 k  " #  ð 7 Þ The fitting parameters of the above HN relaxationfunctions are summarised in Table II. Large standarddeviations, which were observed for HN1 peak  parameters of LPA sample, could be explained bythe presence of more than one relaxation mechanismsin the higher frequency region, which however are not related directly to heavy metals, as it will becomeevident later in this section. The lower frequencyrelaxation peak, indicated as HN3, arises at   ∼ 1.4×10 − 2 Hz in HPA sample, while it is shifted to lower frequency in LPA sample. These peaks are connectedto the ionic conduction of the samples, which causes asteady increase in dielectric loss spectra, as it has beenmentioned previously (see Figure 2). Their shape and Figure 1  Conductivityspectra of the two samples.  Dashed lines  correspond todifferent values of theexponent   n  in Eq. 6.Environ Monit Assess (2007) 133:69  –  78 73
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!