Heavy metals speciation in soakaways sediment
and evaluation of metal retention properties of
M.A. Hossain*, H. Furumai**, F. Nakajima*** and R.K. Aryal****
*Department of Urban Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656,
Japan (E-mail: firstname.lastname@example.org; email@example.com)
**Research Center for Water Environment Technology, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo 113-8656, Japan (E-mail: firstname.lastname@example.org)
***Environmental Science Center, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
****Department of Environmental Science and Technology, Kathmandu University, Dhulikhel, Kavre, Nepal
(GPO BOx #6250) (E-mail: email@example.com)
Abstract Heavy metals speciation analysis was carried out on sediment samples accumulated within
soakaways in an old stormwater infiltration facility in Tokyo, Japan and on a soil core sample collected near
the facility. Heavy metals content in soakaways sediments were much elevated compared to nearby surface
soil with the content for Zn, Pb and Cd reaching about 5 to 10 times the content in surface soil. Speciation
results revealed that significant amount of the accumulated heavy metals were present in potential mobile
fractions, posing threat of release to underlying soil with changing environmental conditions. Detail analyses
of soil characteristics indicated significant heterogeneity with depth, especially between the surface soil and
underlying soil at site. Decrease in potential adsorption sites with depth was observed in case of underlying
soil. Reduced adsorption capacity for heavy metals was evidenced for underlying soil when compared with
surface soil. Furthermore, less capability of the soil organic matter to bind heavy metals was evidenced
through speciation analyses, which raises concern over the long-term pollution retention potential of the
underlying soil receiving infiltrated runoff.
Keywords Adsorption; heavy metals; infiltration; sediment; soil; speciation
Infiltration facilities have been in operation for several decades in the developed countries
to control urban flooding. Secondary benefits include recharge of groundwater and
reduction of non-point pollutant loads from urban surfaces to surface waters. Heavy
metals are ubiquitous toxic substances in urban runoff and are often observed at elevated
concentrations in road runoff (USEPA, 1983; Barrett et al., 1998). The most common
heavy metals found in urban runoff are chromium (Cr), nickel (Ni), copper (Cu), zinc
(Zn), lead (Pb) and cadmium (Cd). They are distributed to the urban road surface mostly
through traffic activities (Ball et al., 1991). Heavy metals in runoff are mostly associated
with particulate matters and accumulated within infiltration facilities through trapping
and adsorption to soakaways sediments (Sansalone et al., 1996). However, concern for
pollution of subsoil and groundwater from stormwater infiltration exists (Mikkelsen,
1997) and there have been few reported incidences of groundwater pollution from infiltra-
tion (Pitt et al., 1999).
In Tokyo, infiltration facilities were constructed in the early 1980s over approximately
1400 hectares area at Nerima ward by Tokyo Metropolitan Sewerage Bureau as part of
the ‘Experimental Sewer System or ESS’ (Fujita, 1984). The primary aim was to reduce
Water Science & Technology Vol 56 No 11 pp 81–89 Q IWA Publishing 2007
runoff discharge to combined sewer system, and thus decrease frequency of CSO events
(i.e., less urban flooding).
The infiltration system consisted of soakaways, trenches, infiltration LU-curb, etc
(Figure 1). Even though the infiltration performance of the facilities was found
satisfactory after more than two decades of operation (Furumai et al., 2005), possible
heavy metals pollution through infiltration of runoff water needs to be assessed. A field
survey on soakaways showed differences in accumulated sediment depth (Aryal et al.,
2006). The study confirmed significant accumulation of heavy metals in soakaways sedi-
ments. These accumulated metals can mobilize through weakening of bonds (Charles-
worth and Lees, 1999) with aging, long dry periods, formation of organic chelates and
complexation (Becker, 1997), and through mobilization of colloids, etc.
Vertical profiles of heavy metals in some sediment core samples from soakaways
with high accumulated depth showed the bottom layers having lower content than the
upper layers, indicating possible desorption/release of heavy metals from the anoxic
bottom layers (Aryal et al., 2007). Once the metals reach the soil system, the fate of the
metals are dictated by the soil properties (e.g., pH, soil organic matter, cation exchange
capacity, etc.), the character of the leachate and speciation of the adsorbed metals
(Evans, 1989; Pitt et al., 1999). The study on the content and speciation of accumulated
heavy metals within the facility, detail soil characterization and understanding of heavy
metals adsorption behaviour in underlying soil is essential to assess the sustainability of
such facilities. The present study focuses on depth-wise soil characterization, comparative
evaluation of metal speciation in soil and soakaways sediments, and assessment of soil
Materials and methods
Soil and sediment samples
The study area for the study was the infiltration facilities at Nerima ward, Tokyo.
Sediment samples were collected from three soakaways in which significant sediment
accumulation (.8cm) was observed. The samples were air dried, sieved (2.0mm nylon
sieve), homogenized and kept in refrigerator in plastic bottles before analysis.
A soil core sample (5m depth) was collected from a park in the study area at a
horizontal distance of approximately 5m from the nearest soakaway. Though infiltrated
runoff reaches the subsoil at a depth greater than 1m (Figure 1), significant differences in
colour and texture observed between the first 1m soil (identified as dark brown ‘surface
andosol’) and the soil below 1m depth (identified as light brown, typical ‘Kanto loam’
soil) advocated for analysing the surface soil (,1.0m) together with underlying soil
(.1.0m). The soil core was divided into 10cm segments, air dried, sieved (2.0mm
nylon sieve), homogenized manually by hand mixing and stored for further analysis.
Figure 1 Schematic view of infiltration facility in Tokyo (constructed under ESS)
M.A. Hossain et al.
A total of 25 sliced 10cm-segments out of 50 were analyzed for pH, organic content
(i.e. ignition loss), humic content, cation exchange capacity (CEC) and heavy metal
content. Heavy metal speciation analysis was performed on 10 soil core segments and 3
Experimental protocols for soil and sediment analysis
The pH was measured following EPA method 9045D. Ignition loss was measured by
heating dry soil (oven dried at 1108C for 24 hours) at 6008C for 1 hour and observing
the weight loss. The cation exchange capacity (CEC) measured was measured from
exchangeable cations determined through extraction with 0.1M BaCl2 solution and
termed ‘effective CEC’ (CECe). The humic substances extraction and fractionation were
carried out according to International Humic Substance Society (IHSS) method of alkali
extraction (0.1M NaOH) and separation employing pH condition. The functional groups
analysis in humic substances was performed qualitatively with Fourier Transform Infra-
red (FT-IR) spectra analysis (Model: ‘JASCO FT/IR-610’).
Digestion for total metal content.
microwave (model: ‘MCS-9700’) following EPA method 3051a. However, the digestion
time was increased to 10 minutes from the original 4.5 minutes mentioned in the method
to account for the differences from the original method mentioned microwave energy and
rate of increase of temperature. The digestion procedure was calibrated with certified
reference material ‘NIST 1649a (Urban dust)’.
Sediment and soil samples were acid digested in a
sequential extraction method was applied to sediments and soil samples to look into
speciation (Hlavay et al., 2004). The ‘BCR sequential extraction method’ is
an operational extraction scheme that sequentially extracts metals in the order of
decreasing mobility (Hlavay et al., 2004). The potential mobile fractions determined in
the scheme are – 1) Acid exchangeable, 2) Reducible and 3) Oxidisable. The three
fractions usually represent the exchangeable and carbonate bound metals, the Fe/Mn
oxide bound metals and organic bound metals, respectively. The remaining amount,
termed the ‘residual metal content’ is calculated by subtracting the three fractions from
the total metal content.
Liquid-solid separation was carried out by centrifugation followed by filtration of
supernatant with 0.2mm pore PTFE filters. The metals in the extracts were all measured
with ICP-MS (Model: ‘HP 4500’). The samples were measured in triplicate except for
humic content analysis which was done in duplicate. The coefficient of variation was
lower than 10% in all cases, except for Cd content.
analysis.The ‘Community Bureauof Reference (BCR)‘three-step
The devising of the adsorption experiment
Heavy metal adsorption experiment was carried out at room temperature on air dried soil
samples with multi-element metal solution of 2mg.L21concentration. The metal solution
was prepared from standard metal solution (‘Wako W-V 113-13781’) and pH was
adjusted to 2.0. The liquid to solid ratio was 20, and the contact time was 6 hours under
150rpm horizontal shaking. Metal speciation analysis was carried out before and after the
adsorption experiment to observe the distribution.
M.A. Hossain et al.
Results and discussion
Vertical profile of soil properties
The surface soil (0–1.0m depth) showed higher pH, ignition loss, CEC and humic
content than underlying (1.0–5.0m depth) ‘Kanto loam’ soil (Figure 2.). The difference
in humic content was large even though the difference in ignition loss was comparatively
less. The humic content (extracted by NaOH) in soil was much lower compared to
ignition loss value indicating most of the humic matter in inert insoluble humin phase for
both type of soil. The difference in CEC was very significant. Soil CEC is contributed by
exchange sites in soil organic matter and clay minerals in soil. The lower CEC in soil at
greater depth may be due to less contribution from soil humic substances. The soil humic
substances play an important role in heavy metal retention in soil (Pitt et al., 1999). The
soil humic substances in underlying soil were dominated by fulvic acids (mobile at low
pH) while in surface soil humic acids (relatively immobile) were dominant (Figure 2).
The FT-IR spectra analysis of extracted humic substance from soil core segments
revealed significant difference between the surface soil and underlying soil humic sub-
stances (Figure 3). In the underlying soil the occurrence of broad deep peaks in the fin-
gerprint region (600–1000cm21) of the spectra around wavenumber 800–1000cm21can
be attributed to alkenes and other low molecular weight compounds (Mayo et al., 2003).
The peaks corresponding to aromatic CvC bonds, i.e., double peaks around wavenumber
1400–1600cm21, tend to attenuate in underlying soil indicating a decrease in the aroma-
ticity (Corvasce, 2005). This was indicative of lower humic acid content. These results
indicated less potential of the organic matter for heavy metal retention in underlying soil.
Heavy metal contents in soil and sediments
The heavy metal content profiles in soil (Figure 4) show an overall decreasing trend with
depth. The heavy metal content in the surface soil was higher than the underlying
(1.0–5.0m) soil, especially for Pb, Ni and Cd.
Figure 2 Vertical profile of %ignition loss, pH, humic substance content and CEC in soil. p HS: Humic
M.A. Hossain et al.
Figure 3 Typical FT-IR spectra of humic substances in surface (0 – 1.0m depth) soil and underlying
(1.0–5.0m depth) soil
Figure 4 Vertical profile of heavy metal content (mg.kg21) in soil [error bar ¼ SD, n ¼ 3]
M.A. Hossain et al.
The metal content in sediment samples was much elevated compared to the surface
soil heavy metals content, especially for Zn, Pb and Cd which were almost 5 to 10 times
the content in surface soil (Table 1). The Zn and Cu content in the sediments were
comparable to content observed in other studies covering infiltration basin sediments
(e.g., Durand, 2004). The Cd and Pb content were lower than reported results by Durand
et al. (2004) while the Ni and Cr content in sediments were higher than content observed
for infiltration facilities in residential area in that study.
Speciation of heavy metals in soil and sediments
In order to assess the potential of release of adsorbed metals in soil and the soakaways
sediments, speciation analysis of heavy metals were performed on a total of 10 sliced
segments from soil core (5 from surface soil and 5 from underlying soil) and three soak-
aways sediment samples, following the BCR sequential extraction method.
Typical plot of metal speciation in one segment from surface soil, one segment from
underlying soil and in soakaways sediment samples (designated as ‘Sed-A’, ‘Sed-B’ and
‘Sed-C’) are shown in Figure 5.
In soil, insignificant amount of metals was present in acid exchangeable fraction and
the residual fraction was the major fraction. In case of underlying soil, little or no mobile
fraction was observed (Figure 5). The acid exchangeable fraction in the sediment would be
most sensitive for heavy metal desorption/release to underlying soil. In soakaways sedi-
ments, acid exchangeable fraction was the major fraction for Zn and Cd while it was a sig-
nificant portion for Cu and Ni indicating possible release in significant amount under low
pH condition. Though Cr and Cu content in surface soil and sediment were almost compar-
able (Table 1) with the ranges overlapping, the speciation was much different (Figure 5).
The sediments accumulated in the facility are contributed from nearby road dust and
wash-off from nearby surface soil. The road dusts are generated through atmospheric
deposition, crushing and grinding of road surface as well as erosion from nearby surface
soil. The large difference in content and speciation between surface soil and sediments
can be attributed to pollution from anthropogenic activities (e.g., traffic activities) and the
processes occurring within the soakaways.
Table 1 Metal content range in sediment, surface soil (0–1.0m) and underlying soil (1.0–5.0m)
Sample TypeRange of metal content (mg.kg21dry wt.)
Figure 5 Heavy metal speciation (expressed in percentage of total content) in one surface soil segment
(40–50cm), one underlying soil segment (170–180cm) and in soakaways sediments
M.A. Hossain et al.
The acid exchangeable fraction and reducible fraction for Zn and Cd in sediments
were comparable with the results of Durand et al. (2004) and Charlesworth and Lees
(1999). Very little oxidisable (i.e., organic bound) fraction was observed which was
lower than the values observed in the above mentioned studies. This was in direct con-
trast to the result observed in studies dealing with river sediments (e.g., Akcay et al.,
2003), where oxidisable fraction is often the dominant fraction.
For Cu, the dominant fraction in soakaways sediments was oxidisable fraction and the
percentage contribution was comparable to Charlesworth and Lees (1999); Durand et al.
(2003) and Akcay et al. (2003). Changes in the organic matter with aging and long dry
period resulting in oxidation of organic matter are the likely environmental scenarios for
mobilization of Cu in soakaways sediments. For both Pb and Cr, the residual fraction
was the dominant fraction, while significant amount existed in the oxidisable and
reducible fractions. This was also different from observed speciation in soil. The
distribution of Pb was in contrast to what had been reported by Charlesworth and Lees
(1999) for soakaways sediments but in agreement with Durand et al. (2004) for the first
two sediment samples.
The heavy metals speciation in soakaways sediments differed significantly from one
another, especially in case of Pb and Cr (Figure 5). This difference may be due to the
difference existing at source (e.g., road dust composition, speciation, etc.) and/or different
environmental conditions existing within the soakaways. The mobility order for the
metals in sediments in terms of percentage in acid exchangeable fraction can be given as
Zn . Cd . Cu . Ni . Pb . Cr.
Evaluating adsorption potential of soil
The significant difference in heavy metal speciation between soil and soakaways sedi-
ments and possibility of metals in significant amount reaching the subsoil, made it
necessary to investigate the metal retention characteristics of the soil. In order to investi-
gate the soil heavy metal retention capacity, air dried sub-samples from two sliced
samples (40–50cm and 170–180cm) were subject to adsorption experiment with multi-
element metal solution. The pH ¼ 2.0 of the metal solution was employed to simulate
low pH condition. It was expected soil organic matter would play a key role in adsorption
at such low pH. The final pH was 2.54 and 2.10, respectively. The three step BCR
sequential extraction method performed on the residue to recover the adsorbed metals
and evaluate the change in speciation after adsorption.
The soil sample representing underlying soil (i.e., 170–180cm) exhibited less metal
adsorption potential compared to surface soil (i.e., 40–50cm) sample, particularly for Zn
Cd and Ni (Figure 6). In case of Zn, Cd and Ni, most of the adsorbed metal moved to the
acid exchangeable fraction for both soil samples. Major portion of adsorbed Cr in surface
soil moved to oxidisable fraction.
A significant amount of metals moved to the residual fraction which represents soil
mineral matter especially clay, crystalline iron oxides, etc. Apart from Cr in surface soil,
there was little adsorption of heavy metal in the oxidisable fraction (i.e. complexed with
organic matter). On the contrary desorption from oxidisable and reducible fractions were
observed in case of Cu. The organic matter in soil seemed to have less stable bond with
the metals than expected and was particularly selective for metals. It was evident from
mass balance that the desorbed metals readsorbed to the residual fractions (i.e. clay
minerals) resulting in high increment in that fraction.
Batch adsorption provides long contact time which might be the reason that readsorp-
tion could take place. However, in the natural environmental condition with infiltrating
water, readsorption by the clay minerals is expected to be lower since the adsorption to
M.A. Hossain et al.
these sites are slower when compared to adsorption to soil organic matter or oxides. In
general, metals adsorbed to organic matter in soil are regarded ‘less mobile’ compared to
metal bound to carbonates or Fe/MnO sites. Hence, less adsorption to oxidisable fraction,
indicating little adsorptive potential of organic matter for metals, is a concern for long
term metal retention capacity of soil.
The sediment samples in soakaways had 5 to 10 times Zn, Cd and Pb content compared
to that in surface soil. A significant amount of heavy metals in sediments was present in
acid exchangeable fraction posing threat of release to underlying soil at low pH,
especially for Zn and Cd. Considerable amount of Cu, Pb and Cr in sediments were
bound to organic matter which, though considered less mobile, may mobilize through
oxidation of organic matter enhanced by long dry periods, microbiological activities, etc.
Significant difference in soil characteristics was observed with depth indicating difference
in metal retention capacity at different layers especially between surface andosol soil’
and underlying ‘Kanto loam’ soil. From adsorption test results, surface soil seemed to
have greater metal adsorption capacity than underlying soil under low pH condition.
Further insights into mechanisms of leaching of metals from infiltration facilities and
enhanced understanding of heavy metals selectivity and competition for adsorptive sites
in soil through competitive adsorption tests are necessary to make an adequate assessment
Figure 6 The distribution of heavy metals in the soil core samples before and after adsorption experiment
for the surface soil segment (at left) and underlying soil segment (at right)
M.A. Hossain et al.
Acknowledgement Download full-text
The authors greatly acknowledge the contribution of the ‘Japan Science and Technology
(JST)’ funded ‘Core Research on Evolutional Science and Technology (CREST)’ project
on ‘Risk-based Management of Self-regulated Urban Water Recycle and Reuse System’
for the financial as well as logistics support extended to this study.
Akcay, H., Oguz, A. and Karapire, C. (2003). Study of heavy metal pollution and speciation in Buyak
Menderes and Gediz river sediments. Water Res., 37(4), 813–822.
Aryal, R.K., Furumai, H., Nakajima, F. and Hossain, M.A. (2007). Vertical distribution and speciation of
heavy metals in stormwater infiltration facilities: possible heavy metals release to groundwater. Water
Prac. Technol., 2(2) doi:10.2166/wpt.2007.052.
Aryal, R.K., Murakami, M., Furumai, H., Nakajima, F. and Jinadasa, H.K.P.K. (2006). Prolonged deposition
of heavy metals in infiltration facilities and its possible threat to groundwater contamination. Water Sci.
Technol., 54(6–7), 205–212.
Ball, D., Hamilton, R. and Harrison, R. (1991). The Influence of Highway-Related Pollutants on
Environmental Quality. In: Highway Pollution, Hamilton, R. and Harrison, R. (eds), Elsevier Science,
New York, pp. 1–47.
Barrett, M.E., Irish, L.B., Jr., Malina, J.F., Jr., and Charbeneau, R.J. (1998) Characterization of highway
runoff in Austin, Texas area. J. Environ. Eng. 124:2 131–137.
Becker, U. and Peiffer, S. (1997). Heavy-metal ion complexation by particulate matter in the leachate of solid
waste: a multi–method approach. J. Contaminant Hydrol., 24(3–4), 313–344.
Charlesworth, S.M. and Lees, J.A. (1999). Particulate-associated heavy metals in the urban environment: their
transport from source to deposit, Coventry, UK. Chemosphere, 39(5), 833–848.
Corvasce, M., Zsolnay, A. and Miano, T.M. (2005). Humic Acids and Water Exctractable Organic Matter
(WEOM) along a deep soil profile. Geophys. Res. Abst., 7.
Durand, C., Ruban, V., Clozel, B. and Ambles, A. (2004). Characterisation of sludge from retention and
infiltration ponds—Potential mobility of trace metals. In: Proceedings of NOVATECH 2004—5th
International Conference on Sustainable Techniques and Strategies in Urban Water Management,
GRAIE, Lyon-France, pp. 1285–1292.
Evans, L.J. (1989). Chemistry of metal retention by soils: Several processes explained. Environ. Sci.
Technol., 23(9), 1446–1456.
Fujita, S. (1984). Experimental sewer system for reduction of urban storm runoff. In: Proceedings of 3rd
International Urban Drainage, Chalmers Univ. of Technology, Gutenberg, Sweden, pp. 1211–1220.
Furumai, H., Jinadasa, H.K.P.K., Murakami, M., Nakajima, F. and Aryal, R.K. (2005). Model description of
storage and infiltration functions of infiltration facilities for urban runoff analysis by a distributed model.
Water Sci. Technol., 52(5), 53–60.
Hlavay, J., Proshaka, T., Weisz, M., Wenzel, W.W. and Stingeder, G.J. (2004). Determination of trace
elements bound to soils and sediment fractions (IUPAC technical report). Pure Appl. Chem., 76(2),
Mayo, D.W., Miller, F.A. and Hannah, R.W. (2003). Course Notes on Interpretation of Infrared and Raman
Spectra, Wiley Interscience, John Wiley & Sons, New Jersey, p. 567.
Mikkelsen, P.S., Ha ¨fliger, M., Ochs, M., Jacobsen, P., Tjell, J.C. and Boller, M. (1997). Pollution of soil and
groundwater from infiltration of highly contaminated stormwater—a case study. Water Sci. Technol.,
Pitt, R., Clark, S. and Field, R. (1999). Groundwater contamination potential from stormwater infiltration
practices. Urban Water, 1(3), 217–236.
Sansalone, J.J., Buchberger, S.G. and Al-Abed, S.R. (1996). Fractionation of heavy metals in pavement
runoff. The Sci. Total Environ., 189/190, 371–378.
USEPA (1983). Results of the Nationwide Urban Runoff Program. Report PB 84-185552 Water Planning
Division, United States Environmental Protection Agency, Washington, DC.
M.A. Hossain et al.