War. Sci. Tech. Vol. 37. No. 4-7, pp. 379-384. 1998.
0 1998 IAWQ. Published by Ekvier Science Ltd
Pergamon
PII: SO273-1223(98)00221-2
P&ted in Great Britain.
0273-1223/98 $1900 + O+lO
IMMOBILIZATION OF HEAVY METALS
IN SEDIMENT DREDGED FROM A
SEAPORT BY IRON BEARING
MATERIALS
I. Miiller and E. Pluquet
Geological Survey of Lower Saxony, Institute of Soil Technology, Friedr. Missler Str.
46/50,28211 Bremen. Germany
ABSTRACT
Pot and field trials were carried out using sediment dredged from the port of Bremen (Germany) and
deposited in a settling basin near Bremen; the sediment is polluted with Cd and Zn. Five iron-bearing
materials were added to the soil (1% pure Fe in soil dry matter) to immobilize the heavy metals: ‘Red mud
from the aluminium industry, sludge from drinking-water treatment, bog iron ore, unused steel shot and steel
shot waste. The pH and CEC were little influenced by any of these treatments, but the NHdN03 and DTPA
extractable amounts of Cd and Zn, and particularly the uptake of Cd and Zn by plants, were markedly
reduced. It was demonstrated that red mud and precipitated Fe-bearing sludge were the most effective
materials. They caused an increase in the adsorption capacity of the dredged sediment with respect to Cd of
about 50%. In the pot trials, NH4N03 soluble amounts of Cd and Zn in samples of soil treated with these
materials were reduced by 50% (DTPA -ZO%), while the uptake by plants was reduced by 20-501. In the
field trial, Cd and Zn were immobilized in the soil to a certain extent, but less effect was observed on the
concentrations in plants and soil extracts compared with the pot trials. In practice, red mud is unsuitable as it
contains large amounts of Cr and A13+ ions. Therefore, only sludge from drinking-water treatment, as long as
the As concentration in it is low, remains as a useful material for immobilizing heavy metals in polluted
sediment dredged from a seaport. 0 1998 IAWQ. Published by Elsevier Science Ltd
KEYWORDS
Dredged sediments; iron oxides; heavy metals; immobilization; iron-bearing residues; pot trials; field trial.
INTRODUCTION
Altogether, 700 000 m3/a of fresh sediment has to be dredged from the port of Bremen. Some of the
sediment in the River Weser is strongly enriched with heavy metals and is disposed of in settling basins. Not
only the total content of heavy metals, but more important the mobile fraction should be known for risk
assessment, if the sediment in the settling basin is to be reused or recultivated. One possible remedial
measure is to immobilize the heavy metals, thus reducing their mobility and uptake by plants. A common
method is to apply dolomitic lime to increase the pH to neutral. Depending on the extent of pollution and the
chemical properties of the soil, optimization of the pH alone may not be sufficiently effective in reducing the
environmental impact of heavy metals (Ku&e et al., 1984).
319
380 I. MiiLLER and E. PLUQUET
Furthermore, the mobile portion of the total heavy-metal content can be reduced by strongly adsorptive
materials like clay, organic matter and iron oxides in soils (Brtimmer et
al.,
1986). The immobilizing effects
of an application of clay minerals and organic matter to contaminated soils have been described by
Vangronsveld er
al.
(1995) and Stolzer er
al.
(1994), respectively. Laboratory experiments have shown that
iron oxides adsorb heavy metals from solution (Grimme, 1968) and occlude them (Gerth & Brtimmer, 1983).
It has also been verified by electron-microprobe studies that heavy metals in contaminated soils accumulate
in iron oxides (Hiller & Brtimmer, 1995). Initial tests treating contaminated soils with iron oxides (Forster et
al.,
1983; Mench et
al.
1994, 1995) have shown that this is a promising way to immobilize heavy metals in
soils.
Investigations on immobilization of heavy metals have been carried out since 1994 on dredged sediment in a
settling basin and two other contaminated sites. Laboratory and pot trials were carried out using five iron-
bearing materials to study the changes in mobility and particularly the uptake of heavy metals by plants; the
results were then tested in field trials. Several iron-bearing residues were also investigated in the tests. The
emphasis of these investigations is on the effectiveness of this method for long term immobilization of
heavy metals and its practical application.
METHODS
Test sites and soils
The settling basin investigated is located near Bremen (northern Germany) west of the River Weser, where
the latter is joined by the River Hunte. Up to 1992, sediment from the port of Bremen was pumped into the
settling basin, forming a layer of sediment 1.80 m thick composed of loamy sand alternating with clayey silt
and ribbons of locally pure sand. Having been subsequently drained, the sediment, which is contaminated
with Cd and Zn, is now ready to be recultivated without any soil cover. The soil used for the pot trials was
taken from a depth of O-15 cm near the field trial plots. This soil showed considerable differences in clay
content and therefore in heavy metal content and other pedological parameters (Table l), underlining the
typical heterogeneity of sediment in settling basins (Herms er al., 1984).
Table 1. Characteristic pedological data of the soil used in pot and field trials
soil
pot trial
field trial
ChY pH CEC,, C, Fe* Mn* Cd’ Pb’ Zn’
%
w-32)
mm01 % % ____ ---------_--------_ mg& ____ _ __-_____--___ ____
lAWlO@
20.1 7.2 26.9 2.4 2.9 996 4.2 85 4.53
24.7 7.1 30.0 3.1 3.2 1150 7.1 145 790
# aqua regia digestion.
Pot and field tests
For the pot trials, soil from the settling basin was sieved; the < 2 cm fraction was mixed with the five
selected materials (1% pure Fe in soil dry matter) and placed in 7 L plastic pots with four replicates
(Table 2). Watering and application of fertilizer, fungicides and insecticides were held constant.
The field trials were limited to treatment with sludge from drinking-water treatment (DW) and iron waste
from descaling steel plate (SW) as well as lime treatment and a zero-treatment as control. Each treatment
Immobilization of heavy metals in sediment 381
Table 2. Treatment of soil from the settling basin
treatment pot trial field trial material used
co
X X
control (no treatment)
LI x x lime (2 gAc+g soi@ field trial au&: keep pH 7 + 0.5 % free carbonate
FUVI
X
red mud (20 % Fe) from aluminium industry
DW
X X
‘Fe-bearing sludge (37 % Fe) from drinking-water treatment plant
BO
X
bog iron ore (38 % Fe) as a natural Fe oxide
SN x new stee[ shot ( 98 % Fe) usu.aQ used as abrake in steel industry
SW
X X
iron waste (78 % Fe) from SN descaling of untreated steel plate
The bay barves1 an& the p\arrts used in me
pot
1ira\s Iwheat, smnach, rye grass} were an&ys&
for
heavy
metal content and yield. The mobility of heavy metals in the soil samples were determined by extraction
with NH4N0, and DTPA. As the investigated settling basin is mainly contaminated with Cd and Zn, this
study concentrates on these two elements.
The following investigations were carried out on air-dried soil samples sieved to c 2 mm mainly according
to German standards: aqua
regja digestion
lAX!XY&V, )992>, DTEA
extraction &h3say & )3oNe)), )9X3), 1
M NH4N03 extraction (E DIN 19730, 1995), cation exchange capacity (CEC,,,, DIN 19684) and
Frew&& i&k~ I&s Cd WX~~X+& %s !k+,+i&&g e< &. (IssS$ wrk.g &@I K C@KG& m &zz+k.
The plants were dried at 60°C and ground in a rotary grinder; this was followed by microwave digestion
using 0.25 g plant material and HN03 (5 ml, 65%, Suprapur) with H,O, (2 ml, 30%, p.a.). Analysis of all
elements were carried out with flame or flameless atomic absorption spectrometer, depending on the
concentration expected.
RESULTS AND DISCUSSION
None of the treatments had much effect on the pH of the soil samples, but RM, due to its high NaOH content
fram the
Al
production process (Kullmartn ecu& 19&?],
imeased the @i
kam 7.2 (controll to 7.5 (Table 3).
The CEst was also little influenced by any of the treatments; only DW caused an increase of 10%.
The total Cd and Zn contents of the soil changed little as a result of the treatment, but in some cases showed
dilution (RM, DW, BO). Only the Zn content increased from 453 to 471 mgkg when SW was added, owing
to the fact that this residue contains high Zn (1490 mg/kg). The mobile contents of these two elements were
claar>> ‘m55uenceh ]“I abe3> PSJ, anhDUS &iuc&ti~Y+Y$+Q~ s.nii&ati~i ti Ch Xrratib 0-a 5R80
(lITPA: -20%). The other treatments showed Iess impact on the extractable amounts, and SW
caused a
small
inccrease
05
exrracXatieZn.
382 I. MijLLER and E. PLUQUET
Table 3. pH, CEC,,,, total and extractable Cd and Zn in soil at the end of pot trial using wheat
treatment pH CEC, Cd Cd Cd Zn Zn Zn
tod DTPA N&NO3 totat DTPA NIX,NOs
co
(CaCM mm01
IAE4oog ---------------------------------- mg/kg __--------___-______-----------------
7.2 23.8 4.2 1.65 0.020 453 86.3 0.31
LI 7.3 22.8 4.2 1.63 0.014 4.55 83.5 0.28
RM 7.5 24.0 4.0 1.33 0.008 433 71.3 0.12
DW 7.3 27.8 4.0 1.28 0.009 435 69.3 0.12
BO 7.2 25.0 4.0 1.50 0.013 435 77.3 0.25
SN 7.2 22.7 4.2 1.55 0.013 443 80.5 0.26
SW 7.2 23.4 4.1 1.58 0.012 471 99.0 0.37
* aqua regia digestion
When it had reached full maturity, the wheat was separated into corn and straw. The yield of corn (CO: 69
g/pot) was only influenced by DW (-7%) and SN (+13%), and the straw (CO: 58 g/pot) by SN (+17%). All
treatments caused a reduction of the Cd concentration in both corn and straw by over 30% (Fig. 1). The most
effective treatments in reducing the Cd level in corn were red mud (45%), shot waste (-44%) and sludge
from drinking water treatment (-41%). The Zn content was influenced less by the treatments than Cd: Except
SW, all the treatments caused a decrease in Zn concentration in corn of lo-15% and in straw of about 20-
30%.
60
70
60
. 50
s
I 40
8 30
20
10
0
co
LI RM DW 80 SN SW
1 i¶Cdcorn q Cd straw n Zn corn Eal l traw 1
Figure 1. Cd and Zn contents of wheat corn and straw from a pot trial using different treatments.
0.6
0.7
0.6
0.5
0
P
0.4
=
i
0.3
P
c
0.2
0.1
0
Spinach was used as a test plant from 1994 to 1996, and showed significant variations between the years.
The Cd and Zn contents of ‘control’ in 1994 was twice the concentrations in 1995 and 1996 (Table 4).
Nevertheless, a declining Cd concentration using RM, DW or SW by IO-20% in 1995 and by 30-50% in
1994 and 1996 was observed over the years. The influence of the treatments on the Zn content of spinach
was less uniform. In 1994 LI, RM and SN (all -20%) showed the best results, in 1995 only SW (-15%) and
Immobilization of heavy metals in sediment 383
in 1996 only RM (-30%) were particularly effective. The extractable Cd and Zn fractions hardly differed
between the years and equalled the figures in Table 3.
Table 4. Cd and Zn in spinach grown over three successive years in pot trials on soil from a settling basin,
using different treatments
treatment
co
LI
RM
DW
BO
SN
SW
1994 1995 1996
Cd Zn Cd Zn Cd zn
----_______----___-------------------------- mglkg dry matter ________________ - ___________________________
2.11 255 1.21 104 1.10 111
1.47 191 1.28 99.1 0.95 91.0
1.24 192 0.99 109 0.65 77.4
1.30 240 1.08 93.8 0.69 91.0
1.65 229 1.18 92.5 0.89 92.9
1.65 202 1.28 97.8 0.97 96.7
1.03 263 1.05 85.3 0.81 104
The yield of rye grass grown in pots was only influenced by RM (-10%) and SN (+17%). The Zn
concentrations in the grass hardly differ from one treatment to another, whereas the Cd content was reduced
by RM, DW, SN and SW by about 2530% (Table 5).
Table 5. Weighted means of Cd and Zn contents in grass grown on soil from a settling basin using different
treatments
treatment
co
LI
RM
DW
BO
SN
SW
pot trial field trial
Cd zn Cd zn
_________-____________________________ mg/kg dry mat&r ------- ---------- --------- ------- - - -- --
0.37 60.3 0.26 73.1
0.33 59.2 0.26 76.3
0.27 64.3
0.26 56.9 0.21 66.3
0.30 57.3
0.28 57.5
0.27 64.5 0.18 66.3
Due to the larger total contents of Cd and Zn in the soil used for the field trial (Table 1), the NH4N03 and
DTPA extractable fractions were twice those obtained in the pot trials and were reduced by DW and SW by
about 10%. Although the Cd content (total and soluble) was much higher in the soil used in the field trial,
the Cd concentrations in grass were found to be 20% lower compared with the pot trial (Table 5). The Cd
concentration in the grass from the field trial was reduced by treatment with DW (-20%) and SW (-30%) and
the Zn concentrations also fell in both cases by about 10%.
Correlation between soil extracts and concentration in grass was very low in the field trial, except the
recognizible relation between soluble Zn and its concentration in grass (rM.48). In the pot trials, the DTPA
and NH4N03 extractable fractions of Cd and Zn hardly differed between the years or between the plants
tested, and was almost the same as the data given in Table 3. Correlation between soil extracts and plant
uptake was generally poor in the pot trials too, showing two exceptions which were NH4N0, soluble Cd and
Cd in wheat corn (r2=0.88) and in grass (r2=0.76).
Many authors have found a good correlation between Cd extractable by solutions of neutral salts and Cd
uptake by plants (Sauerbeck & Styperek, 1985; Sanders et al., 1987; Prtiess, 1992). Our tests involving