GEOCHEMICAL PARTITIONING OF HEAVY METALS IN THE URBAN SOIL, KIRKUK, IRAQ

This work focused on anthropogenic influences of the trace metals distribution in the soils of Kirkuk city. Sequential extraction technique was used to determine the distribution of the chemical fractions of Ag, Cd, Co, Cu, Ni, Pb, Zn, As, Cr and V in soil of Kirkuk city. This area is affected mainly by burning oil trash. Results show that these heavy metals were primarily restricted to surface horizons and mostly associated with the residual fraction (28.8 – 50%). The remnant fractions (13.8 – 33.1%) linked to the organic matter, 7.9 – 27.2% was bound to Fe-Mn oxide, 0.7 – 27.9 was bound to carbonate. Only a small amount of the total metals in the soil is exchangeable (0.5 – 4.2%) and water soluble (0 – 4.1%) fractions. Ag, Cd, Cu, As, Cr and V mainly associated to organic matter fraction; Co, Ni and Zn mostly bound to Fe-Mn oxide fraction; Pb primarily bound to the carbonate fraction. Metals that are bound to the organic matter fraction could be released under oxic conditions, while those associated with Fe-Mn oxide and carbonate fraction could be leached out by changes in the ionic composition and pH. The mobility factors for the metals in the surface soil ranged from 0 to 36 for Ag, 22.2 to 43.6 for Cd, 5.3 to 20.8 for Ni, 16.1 to 41.2 for Pb, 7.3 to 37.9 for Zn, 10.4 to 22.9 for As, 3.2 to 12.3 for Cr and 2.4 to 9.4 for V. The high level of metals remnant as residual fraction coupled with low values of mobility factors, indicate that these metals do not cause any environmental risk or hazard


INTRODUCTION
Geochemical partitioning of trace metals in soils with different land use categories is an important to understand behavior of metals, which derived from anthropogenic activity and directly influenced to human and animal health (Davidson et al., 2006;Gyekye, 2013;Brevik and Sauer, 2015).Metal toxicity is a function of the concentrations of specific metal species, not of the total metal (Stumm and Morgan, 1996;Langmuir, 1997a).Sequential extraction schemes are widely applied procedures to gain a better understanding of geochemical processes by a series of soil leaching steps to investigate metal content in different fractions of soils (Zeien and Brümmer 1989;Hall et al., 1996;Rauret, 1999;Filgueiras et al., 2002;Langmuir et al., 2005;Abollino et al., 2005;Laing, 2010;Howard et al., 2013).The fixation of heavy metals are controlled in clay minerals, organic matter, iron and manganese oxides and hydroxides by adsorption and in poorly soluble sulphide, carbonate, and phosphate minerals by precipitation (Bourg and Loch, 1995 in Van der Perk, 2006).Kirkuk city contain one of the most important oil field sources in Iraq.Oil trash burning from oil refining, exhaust emissions from vehicles under actual traffic conditions, in addition to private electrical generators, these all represent as an important sources of air pollutants that precipitate on the soil surface as an aerosol thin layer.The volumes of released pollutants from these sources were assumed to be in millions of tons per year (Saleh et al., 2014).Thus, it is necessary to perform an exploration of heavy metals in soils of the urban area of Kirkuk city.The objective of the present study was to determine the geochemical behavior of the heavy metals (Ag, Cd, Co, Cu, Ni, Pb, Zn, As, Cr and V) in the soil environment of Kirkuk city via their relative bonding strength in different phases and mobility.

Regional setting of the study area
The urban area of Kirkuk city is located 250 Km north of Baghdad between latitudes 35° 23 '6.33" N -35° 30' 24.94" N and longitude 44° 19' 41.28" E -44° 26' 17.33" E, covering an area of 100 km² approximately (Fig. 1).The climate of this area is humid subtropical with a mean annual rainfall of 1200 mm and a mean annual temperature (22.7 °C) (FAO, 2009).Rainfall is seasonal falls in the winter from November to April.The area is located within the Unstable Shelf in the Foothill Zone, which called Hemrin -Makhul Subzone.The area is dominated by a flat topography in the southwest with a smooth relief in the northeast (Sissakian, 1993in Majeed, 2004).
Stratigraphically, Miocene and Pliocene rocks are represented by Fatha (M.Miocene), Injana (U.Miocene), Al-Mukdadyia (L.Pliocene) and Bai Hassan (U.Pliocene) exposed in the syudy area and covered by Quaternary deposits, periodically flood and soil deposits (Buringh, 1960).The soils of the area consiste mainly of mixture of gravel, boulders, sands and clay (Jassim and Goff, 2006).Alluvial formed by the deposition and erosion periodicity during the different stage of river flooding is commonly consisted of clayey silt sandy and gravel (Buday, 1980;Kassab and Jassim, 1980;Jassim et al., 1984).

Sampling and analysis
Soil samples were taken systematically with a density of one sample per square kilometer to get a uniform coverage of the study area (De Vivo et al., 2008).Each one of collected sample was defined according to types of land (residential, commercial, industrial, roadside, green  A total of ninety nine soil sample were collected from 81 site representing surface and sub soil zones.There are 75 surface soils were taken from the uppermost layer of the soil profile (0 -20 cm), where industrial dusts and emissions from vehicles have been deposited (Lu and Bai, 2009).Subsoil samples were taken from 30 -50 cm depth in fifteen sites of urban area to correlate with topsoil samples.Collected soil samples were air-dried, thereafter breaking the agglomerates by used a plastic hammer, then sieved through a 2 mm sieve and stored in plastic bags.Consequently, soil samples were ground and sieved through < 0.5 mm in the laboratory of Soil Mechanics in the Kirkuk Construction Lab.Pseudo-total metals were calculated by using digestion method with Aqua Regia.This procedure gives an assessment of maximum potential hazard that could occur in long term, as well as in the case of environmental metal contaminants, usually not bound in silicates (Rao et al., 2007).Based on the microwave-assisted acid digestion method that proposed by the US Environmental Protection Agency (USEPA, 1998) (EPA 3051).Microwave digestion was carried out in microwave oven (MLS-ETHOS plus).Prepared 0.1 g and transferred to digestion vessels and consequently 10 mL of Aqua Regia (McGrath and Cunliffe, 1995; HCl:HNO3, 3:1, v/v) were added and left to sit under reflux conditions (Abbruzzini et al., 2014).The block was gradually heated from ambient temperature to 180 °C and the samples were processed until 1 mL of acid remained.The digests were diluted by 10 mL of distilled water and filtered by cellulose filter paper (pore diameter of 0.45 µm).Digestion was performed with triplicates for one sample; three blanks and standard samples were digested in the microwave.After that the concentration of heavy metals including Ag, Cd, Co, Cu, Ni, Pb, Zn, As, Cr and V were measured using an ICP-MS (Inductively Coupled Plasma Mass Spectrometry).

Sequential extraction procedure
Water-soluble (F1) this fraction represent metal ion activity in the soil solution phase has been shown to be a better indicator of bioavailability and toxic response than is the total soil metal content (Tye et al., 2003).This fraction was extracted by adding 20 mL of deionized water.Shake for 16 h (Howard et al., 2013), 20 °C, on rolling table.
Supernatant (F1) was separated by high-speed centrifugation for 30 min at 10000 rpm.
The residue was washed by adding 10 mL of distilled water, shaking for 15 minutes and centrifuging 20 minutes at 3000 rpm.Washes were collected and analyzed with Supernatant (F1).
Exchangeable (F2) representing metals are adsorbed on clay minerals.The residue from fraction 1 (F1) was extracted by adding 20 mL of 1 mol Ammonium acetate (1 M NH 4 OAc) at pH 7. Shake for 2 h.Supernatant (F2) was separated by high-speed centrifugation for 30 min at 10000 rpm.The residue was washed by adding 10 mL of distilled water, shaking for 15 minutes and centrifuging 20 minutes at 3000 rpm.The washes were collected and analyzed with Supernatant (F2).
Carbonate (F3) has been extracted from residue (F2) by adding 20 mL of 1 M NH 4 OAc at pH 5. Thereafter , placed in Shaker for 2 h.Supernatant (F3) was separated by high-speed centrifugation for 30 min at 10000 rpm.The residue was washed by adding 10 mL of distilled water, shaking for 15 minutes and centrifuging 20 minutes at 3000 rpm.Washes were collected and analyzed with Supernatant (F3) Fe-Mn Oxide (F4): These metals associated mainly on iron and manganese oxides.The residue from fraction 3 (F3) extracted with 20 mL of 0.04 mole Hydroxylamine hydrochloride (NH 2 OH•HCl) in 25% Acetic acid (HOAc) for 6 h (water bath, 60 °C).Supernatant (F4) was separated by high-speed centrifugation for 30 min at 10000 rpm.The residue washed by adding 10 mL of distilled water, shaking for 15 minutes and centrifuging 20 minutes at 3000 rpm.Washes were collected and analyzed with Supernatant (F4) Organic matter (F5) is represented the fraction that is strongly complexed by organic matter.The residue from fraction 4 (F4) extracted with 15 mL of 30% Hydrogen peroxide (H 2 O 2 ) at pH 2 (adjusted with HNO 3 ) for 5.5 h (water bath, 80 °C), after cooling, add 5 mL of 3.2 M NH 4 OAc in 20% HNO3, Shake for 30 min, dilute to 20 mL with water.Supernatant (F5) was separated by high-speed centrifugation for 30 min at 10000 rpm.The residue was washed by adding 10 mL of distilled water, shaking for 15 min and centrifuging 20 min at 3000 rpm.The washes were collected and analyzed with Supernatant (F5) Residual (F6): This fraction was calculated as the difference between the Pseudo-total metals and the sum of extracted metals (Zufiaurre et al., 1998).

Fig. 3: Sequential extraction procedure
All solutions were analyzed using ICP-MS.Experimental results obtained on replicate soil samples demonstrate that the relative standard deviation of the sequential extraction procedure is generally acceptable (Table 1).

RESULTS AND DISCUSSION
Psedo-total metal concentration and metal partitioning the geochemical fractions with pseudo-total concentration of the heavy metals in the soil surface and profile are presented in the Table 2a and Table 2b which are expressed as a percentage in the Table 3a and Table 3b respectively.Figure 4 shows the distribution of the average content of metal fractions in soil surface.
Most of metals that present in soil associated mainly with the residual fraction (28.8 -50%) except Cd (Table 2a; Fig. 4).This related with pH value is mostly above 7 contributing to decrease metals in the water-soluble and exchangeable fractions relative to residual fraction (Chen et al., 2003).Figures 5a, 5b, 6a and 6b show the distribution of metal fractions in soil profiles, it is clear that the residual fraction for all metals increase with depth with an average value exceed 61%, this result arise from the fact that residual fraction is more stable and strongly bound to soil components relative to other fractions (Kabala and Singh, 2001).2a; Fig. 4).The distribution of Ag in various fractions was in the order residual > organic > Fe-Mn oxide > water soluble > exchangeable > carbonate.All fractions are higher in soil surface except that was bonding with Fe-Mn oxide (Table 2b; Figs.5a and 6a).
2-Cd was mainly bound to the organic matter with an average of 33.1% of total extractable content.Moreover, a considerable amount of Cd (26%) is bound to carbonate (Table 2a; Fig. 4).This result also reported in the speciation was done by the Tessier method (Tessier et al., 1979).The distribution of Cd in various fractions was in order organic > residual > carbonate > Fe-Mn oxide > exchangeable.The relative contribution of the exchangeable cadimium fraction (F2) decreased significantly with soil depth (Table 2b; Figs.5a and 6a).
3-Co was primarily bound to the Fe-Mn oxide with a median of 26.3% of total extractable content, where the high proportion of the Co in soil is bound to Mn oxides (Table 2a; Figure 4).This result were also reported by Childs (1975) and Taylor & McKenzie (1966).The distribution of Co in various fractions was in the order residual > Fe-Mn oxide > organic > carbonate > exchangeable > water soluble.
With the exception of residual fraction, Co was decreased with depth (Table 2b; Figs. 5a and 6a).
4-Cu was primarily bound to the organic matter with an average of 29% of total extractable content (Table 2a; Figure 4).The abundance of Cu in the organic fraction is also reported by several workers (Tessier et al., 1979;Zeien and Brümmer 1989).
Additionally, a considerable amount of Cu (15.8%) is also found in the fraction bound to Fe-Mn oxide.The distribution of Cu in various fractions was in order residual > organic > Fe-Mn oxide > carbonate > exchangeable > water soluble.All Cu fractions are decreased with depth except residual fraction (Figs.5a and 6a).

5-
Ni was bound to Fe-Mn oxide and organic matter with an averages of 21.9% and 16.5% of total extractable content, but the greatest percentage of nickel was found in the residual fraction (47.1%) as shown in Table (2a) and Figure (4).The availability of Ni in residual fraction proved in many papers (Gupta and Chen, 1975;Tessier et al., 2002;Hickey and Kittrik, 1984;Ma and Rao, 1997;Moral et al., 2005;Iwegbue, 2007;Chaudhary and Banerjee, 2008;Ogundiran and Osibanjo, 2009;Osakwe, 2010;Iwegbue, 2011;Osakwe, 2013).The distribution of Ni in various fractions was in the order residual > Fe-Mn oxide > organic > carbonate > exchangeable > water soluble.Figure 5a and 6a show that surface soil was dominated by Ni fractions relative to deep soil, with the exception of residual fraction that was increased with soil depth (Table 2b).
6-Pb was mostly bound to the carbonate (Chen et al., 2003) with an average of 27.9% of total extractable content.The next abundant fraction of Pb was bound to Fe-Mn oxide (17.4%) and bound to organic matter (13.8%) (Table 2a; Figure 4).The distribution of Pb in various fractions was in the order residual > carbonate > Fe-Mn oxide > organic > exchangeable.With the exception of residual fraction, Pb was dominated in the surface horizons of soils by the other fractions (Table 2b; Figs.5b   and 6b).
7-Zn was mainly bound to Fe-Mn oxide with an average of 27.2% of total extractable content (Table 2a; Fig. 4).This result is reliable with the results of many researchers who found the greatest percentage of Zn associated with Fe-Mn oxide (Hickey and Kittrick, 1984;Kuo et al., 1983;Tessier et al., 1980).The distribution of Zn in various fractions was in order residual > Fe-Mn oxide > organic > carbonate > exchangeable > water soluble.Zn in subsurface horizons of soils was concentrated in the residual, organic and soluble fractions, while in surface layers, Zn was concentrated in Fe-Mn oxide, carbonate and exchangeable fractions (Table 2b; Figs. 5b and 6b).

8-
As was primarily bound to the organic matter (Varsanyi and Kovacs, 2006) with an average of 20.3% of total extractable content.The second most abundant fraction of As is in bound to Fe-Mn oxides, which contains 18.4% of As (Table 2a; Fig. 4).The distribution of As in various fractions was in the order residual > organic > Fe-Mn oxide > carbonate > exchangeable > water soluble.With the exception of residual fraction, organically bound As was dominant in the surface horizons of soils (Table 2b; Figs.5b and 6b).
9-Cr was mainly bound to the organic matter (Guerra et al., 2007) with an average of 27.2% of total extractable content (Table 2a; Figure 4).The distribution of Cr in various fractions was in the order residual > organic > Fe-Mn oxide > carbonate > exchangeable.All Cr fractions beside residual are concentrated in surface horizons (Table 2b; Figure 5b and 6b).
10-V was mostly bound to the organic matter with an average of 32.8% of total extractable content.Subsurface soil was dominated by vanadium associated with both of residual and Fe-Mn oxide fractions (Table 2a; Figure 4).Since V is bound to organic matter, V would be limited use as a reference element (Plantinga, 1997).The distribution of V in various fractions was in order residual > organic > Fe-Mn oxide > carbonate > water soluble > exchangeable.All soil portions in surface horizons enriched by V relative to deep horizons of soil except for residual (Table 2b; There are some metals increased with soil depth that depicted in Figure 5b, where Zn that associated with soluble fraction was increased with depth (from 0 -20 to 35 -50 cm) in 34%, while Ag associated with Fe-Mn oxide enhance with soil depth in 35%.In the Figure 6b, The metals that associated with soluble fractions increase with depth (from 35 -50 to >100 cm) in order Pb, Cr and V in 90%, 37% and 19% respectively; with exchangeable fractions increase in the same depth in order Zn and Cu in 60%, 18%; with carbonate arranged in order Ag and V in 44% and 27%; with Fe-Mn oxide arrange in order V and As in 20% and 12% (Figs.6a and b).High contents of some metals in deeper horizons relative to surface horizons of soil are probably caused by a gradual migration of labile forms of heavy metal.The same conclusion proved by Vaněk et al. (2005).

Metals mobility in the soil profiles
The initial fractions in sequential extraction procedure release the most mobile and bioavailable fractions.Therefore, a high mobility factor (MF) value represents relatively high reactivity, high lability, and high biological availability of heavy metals in soil (Karczewska, 1996;Ma and Rao, 1997;Kabala and Singh, 2001).Mobility factor has been calculated using a six-step extraction scheme (Kabala and Singh, 2001;Salbu et al., 1998;Narwal et al., 1999).The mobility factor was calculated (MF; Salbu et al., 1998;Narwal et al., 1999) based on the equation: The mobility factors (MF) of the metals for soil profile are show in Table 4.
Figure 7 shows the mobility factors in surface soil horizons.Figures 8 and 9 show the value of mobility in soil profile.The results proved the highest mobility of Cd in comparison with the other metals, because the highest average values of MF exceed 30% (Fig. 7).Pb was second mobile metal reached 28.6% (Fig. 7).The mobility of Cd increased more than Ni and Cu metals (McLean and Blendsoe, 1992).Mobility factors (MF) depicted in the figures 8 and 9 illustrate that the most of metals in the top surface were higher than that in the subsurface horizons, as indicator of heavy metals stability in the subsurface horizons.The mobility of Cu was increased in surface horizons relative to deep horizons of about 29% (Fig. 8).This could be arise from its bounding to organic matter that mentioned in the previous section, because Cu has a high affinity for soluble organic ligands and formation of these complexes may greatly increase Cu mobility in the soils (McLean and Blendsoe, 1992;Amrhein et al., 1992).The mobility of Ag is very low relative to other metals (Fig. 7), this results can be proved that silver is highly immobile in the soil environment, because it is very strongly adsorbed by clay and organic matter and precipitates as silver, AgCl, Ag 2 SO 4 and AgCO 3 , which are highly insoluble (Lindsay, 1979).Increased values of Ag and V in subsurface horizons observed in Figure 8 are proved their high mobility relative to surface soil.Moreover the increased content of Ag, Zn and V in deeper horizons in Figure 9, could be explained by partial transfer of heavy metals as a labile forms through soil profile with the increase of their total contents in the depth could not be eliminated (Borůvka et al., 1996).The high MF values for Cd and Pb indicate the high lability and biological availability in soils.This result correspond with many workers (Karczewska, 1996;Ma and Rao, 1997;Ahumada et al., 1999;Narwal et al., 1999).
Therefore, Cd and Pb represent the highest potential risk for the surrounding environment.
Mobility factor related inversely with residual fraction as observed in Figure 10 that indicates the metals under study do not cause any environmental risk.The same finding was obtained by Osakwe (2013).among soil profiles may indicates that anthropogenic sources are slightly impact on the urban environment.The mobility of heavy metals were correlated inversely with their residual fraction, indicating that the metals added by anthropogenic sources remained in relatively weakly bound forms, whereas the large portion (61%) of metals restricted in more stable and strongly bound fraction (F6) to subsoil components which is an indication that the studied metals did not pose any environmental risk.

Fig. 1 :
Fig. 1: Location map of the study area

Fig. 2 :
Fig. 2: The sampling sites and landuse categories of soil in the study area

Fig. 4 :
Fig. 4: Distribution of the heavy metal fractions in soil surface

Figure
Figure 5b and 6b).Similar distributions of relative abundance are observed in the case of Cu, As and Cr, and same pattern happened with each of Co, Ni and Zn.The major fractions are associated with residual and organic matter (fractions F6 and F5).