Pergamon
PII: S0043-1354(97)00204-2
IVat. Res. Vol. 32, No. 2, pp. 430-440, 1998
© 1998 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
0043-1354/98 $19.00 + 0.00
HEAVY METAL CATION
UNCONVENTIONAL SORBENTS
ASHES)
RETENTION BY
(RED MUDS AND FLY
RE,AT APAK*, ESMA TOTEM, MEHMET HLIGI]L and JI]LIDE HIZAL
Department of Chemistry, Faculty of Engineering, Istanbul University, Avcdar, 34850, Istanbul,
Turkey
(First received April 1996; accepted in revised form June 1997)
Al~raet--Toxic heavy metals, i.e. copper (II), lead (II) and cadmium (II), can be removed from water
by metallurgical solid wastes, i.e. bauxite waste red muds and coal fly ashes acting as sorbents. These
heavy-metal-loaded solid wastes may then be solidified by adding cement to a durable concrete mass
assuring their safe disposal. Thus, toxic metals in water have been removed by sorption on to inexpen-
sive solid waste materials as a preliminary operation of ultimate fixation. Metal uptake (sorption) and
release (desorption) have been investigated by thermostatic batch experiments. The distribution ratios
of metals between the solid sorbent and aqueous solution have been found as a function of sorbent
type, equilibrium aqueous concentration of metal and temperature. The breakthrough volumes of the
heavy metal solutions have been measured by dynamic column experiments so as to determine the sat-
uration capacities of the sorbents. The sorption data have been analysed and fitted to linearized adsorp-
tion isotherms. These observations are believed to constitute a database for the treatment of one
industrial plant's effluent with the solid waste of another, and also to utilize unconventional sorbents,
i.e. metallurgical solid wastes, as cost-effective substitutes in place of the classical hydrous-oxide-type
sorbents such as alumina, silica and ferric oxides. © 1998 Elsevier Science Ltd. All rights reserved
Key words---cadmium (II), lead (II), copper (II), sorption, red muds, fly ashes
INTRODUCTION
Cadmium (II), lead (II) and copper (II) are well-
known toxic heavy metals which pose a serious
threat to the fauna and flora of receiving water
bodies when discharged into industrial wastewater.
In spite of strict regulations restricting their careless
disposal, these metal cations may still emerge in a
variety of wastewaters stemming from catalyst, elec-
trical apparatus, painting and coating, extractive
metallurgy, antibacterials, insecticides and fungi-
cides, photography, pyrotechnics, smelting, metal
eleetroplating, fertilizer, mining, pigments, stabil-
izers, alloy industries, electrical wiring, plumbing,
heating, roofing and building construction, piping,
water purification, gasoline additive, cable covering,
ammunition and battery industries (Buchauer, 1973;
Low and Lee, 1991; Periasamy and Namasivayam,
1994) and sewage sludge (Bhattacharya and
Venkobachar, 1984). The acute toxicity of these
heavy metals have caused various ecological cata-
strophes in human history, such as the "itai-itai"
disease due to cadmium (Riley and Skirrow, 1975).
Prolonged effect may cause other chronical dis-
orders (Huang and Ostovic, 1978).
*Author to whom all correspondence should be addressed.
Various treatment technologies have been devel-
oped for the removal of these metals from water.
The hydrometallurgical technology extracts and
concentrates metals from liquid waste using any of
a variety of processes, such as ion exchange, electro-
dialysis, reverse osmosis, membrane filtration,
sludge leaching, electrowinning, solvent stripping,
precipitation and common adsorption (LaGrega et
al., 1994a).
Both powdered (Sorg et al., 1978) and granular
activated carbon (Huang and Smith, 1981) have
been used for the adsorptive removal of Pb, Cd and
similar "soft" heavy metals, especially when associ-
ated with common organic particulate matter in
water. Activated carbon from cheaper and readily
available sources, such as coal, coke, peat, wood,
nutshell (Freeman, 1989) and rice husk (Srinivason,
1986), may be successfully employed for the
removal of heavy metals from aqueous solutions.
Hydrous oxides such as alumina, iron oxides
(hematite and goethite) (Cowan et al., 1991; Gerth
and Bruemmer, 1983) manganese (IV) oxide
(Hasany and Chaudhary, 1986) and titanium (IV)
oxide (Koryukova et al., 1984) have also been used
for the adsorption of the indicated heavy metals.
The cost of the adsorptive metal removal process
is relatively high when pure sorbents (either acti-
430
Sorption of heavy metal cations 431
vated carbon or hydrated oxides) are used.
Therefore, there is an increasing trend for substitut-
ing pure adsorbents with natural by-product or
stabilized solid waste materials for the development
of cost-effective composite sorbents capable of
treating a variety of contaminants. For example,
recent evidence on the combined use of lime, ferric
and aluminium coagulants has shown that these
substances are more effective in combination than
individually (Harper and Kingham, 1992). A num-
ber of metallurgical solid wastes such as bauxite
waste red muds and coal-fired thermal plant fly
ashes have been screened in this regard to serve as
versatile and cost-effective sorbents for heavy metals
(Apak and (0nseren, 1987; Apak
et al.,
1993) and
radionuclides (Apak
et al.,
1995; 1996). The ability
of fly ash to remove metal cations from water has
also been demonstrated in the literature
(Bhattacharya and Venkobachar, 1984; Panday and
Singh, 1985; Yadava
et al.,
1987) for a limited num-
ber of metals.
The alternative mechanism for heavy metal
removal by red muds and fly ashes (either natural
or in activated form) are assumed to comprise four
steps (Gregory, 1978; Apak and Llnseren, 1987). (i)
surface precipitation (sweep flocculation), where
most hydrolysable heavy metals are removed via co-
precipitation of their insoluble hydroxides forming
successive layers on the sorbent surface; (ii) floccu-
lation by adsorption of hydrolytic products, where
multi-nuclear hydrolysis products (formed on the
adsorbent surface as kinetic intermediates) including
[Fe2(OH)4 ]2+, [Fe3(OH)415+, [AI4(OH)s]4+ and
[AIs(OH)20] 4 + act as more effective flocculants than
their parent ions due to their higher charge and
strong specific adsorptivities; (iii) chemical adsorp-
tion based on surface-complex formation, where
metal ions are usually removed as uncharged hy-
droxides condensed on to surfaces of -OH group
bearing adsorbents (Lieser, 1975), i.e. aluminium
oxide, silica gel, ferric and titanium oxides, existing
as components of the utilized composite sorbents;
(iv) ion exchange, where the acid-pretreated sor-
bents may function as synthetic cation exchangers.
Of these mechanisms, surface precipitation and
chemical adsorption are believed to play the domi-
nant role in heavy metal ions removal (Apak and
Unseren, 1987).
The aim of the present study is to develop cost-
effective unconventional sorbents, preferably metal-
lurgical waste solids, for heavy metal removal from
contaminated water. The heavy metal (Pb, Cd and
Cu) removal capacity as well as sorption modelling
of red muds and fly ashes will be evaluated in this
regard. The irreversible nature of sorption needs to
be shown so as to guarantee non-leachability of
metals from the metal-loaded sorbents.
EXPEilIMENTAL
Materials and methods
All heavy metal solutions (divalent cations Of Pb, CA
and Cu) were prepared in stock solutions up to 10000
ppm 0a g/ml) of metal from the corresponding nitrate
salts. No further pH adjustment of these solutions was
made as their natural acidity due to hydrolysis of metals
(i.e. to form MOIl + and H +) prevented the precipitation
of the corresponding metal hydroxides. All chemicals (E.
Merck, Darmstadt, Germany) were of analytical reagent
grade.
Of the metallurgical solid wastes used as sorbents, the
red muds were supplied from Etibank Seydi~ehir
Aluminum plant, Konya, Turkey and coal fly ashes
were from TEK Af~in-Elbistan Thermal Power Plant,
Table 1. Saturation capacities of the sorbents for the metals from column and batch experiments and Langmuir parameters of equilibrium
modelling
Langmuir Parameters b
Qo (rag/8)
Equilibrium Qexp. (rag/g) Qexp. (rag/g) Theoretical
Metal ion a Adsorbent pH Column capacity Batch capacity capacity b (litre/mg) Corr. coeff. (r)
Cd (II) F 7,2 220 198.2 374.3 1.14.10 -3 0.957
Cd (II) Fw 6.7 -- 195,2 223.2 I. 17.10 -3 0.970
Cd (II) Fa 6.6 122 180,4 217.2 6.07.10 -3 0,953
Cd OI) Rw 6.0 160 66.8 113.7 0.57.10 -3 0.958
Cd (II) gab 5.9 115 66.8 112.0 0.65.10 -3 0.994
Cd (II) Ra 4.2 105 46.9 107.5 0.11.10 -3 0.989
Cu (If) F 6.0 -- 207.3 335.2 0.94,10 -3 0.968
Cu (II) Fw 5.8 264 205.8 328.2 0.75,10 -3 0.961
Cu (II) Fa 5.7 187 198.5 283.9 0.73.10 -3 0.960
Cu (II) Rw 6.0 110 75.2 90.0 0.96.10 -3 0.958
Cu (II)
Rah
5.7 100 65.2 87.8 0.79.10 -~ 0.956
Cu (II) Ra 4.5 63 35.2 65.4 1.00.10 -3 0.964
Pb (II) F 6.2 530 444.7 526.0 I.I 1.10 -3 0.948
Pb (II) Fw 6.0 -- 483.4 490.7 1.10.10 -3 0.976
Pb (II) Fa 6.0 -- 437.0 483.0 0.84.10 -3 0.958
Pb (II) Rw 6.0 161 165.8 158.9 0.66.10 -3 0.960
Pb (II) R.a 5.7 164 138.8 137.2 1.17.10 -3 0.970
Pb (II) R, 4.4 123 117.3 118.5 1.56.10 -3 0.956
"The initial aqueous metal concentrations for different metal/sorbent combinations were as follows: Cu (II) 50 mM (mmol/litre) for red
muds and 90 mM for fly ashes; Pb (If) 50 mM for red muds and 65 mM for fly ashes; Cd (II) 35 ram for red muds and 40 mM for fly
ashes,
bCalculated by the aid of iinearized Langmuir equation (4),
432 Re,at Apak
et ,7l.
.0 , , , , , ,
t [] CdF 5.0
A CdF w
4.0 ~ caF,
• I
3.0
Jo.'~ ~]~ Cd~
I
1.ot • "1
0.0 ~ • • I ~ ~ w
0.0 1.0 2.0 3.0 4.0 5.0 6.0
C e (mi/mL)
Fig. 1. Distribution coefficient of Cd (II) as a function of equilibrium aqueous concentration on fly
ashes and red muds.
Kahramanmara~, Turkey. The red muds, obtained as alka-
line leaching wastes of bauxite in the Bayer process of
alumina manufacture, had the following chemical compo-
sition by weight: Fe203 37.3%, A1203 17.6%, SiO2 16.9%,
TiO2 5.6%, Na20 8.3%, CaO 4.4%, loss on ignition
7.2%. Red muds, being multicomponent systems, are com-
posed of sodium aluminosilicates, kaolinite, chamosite,
iron oxides (hematite) and hydroxides. Basically, Fe is in
the form of hematite, Ti is in the form of Fe-Ti oxides
and AI is in the form of ahiminosilicates. 94% of red
muds have less than 10/an grain size.
The red muds were thoroughly washed with water to a
neutral pH, dried and sieved (R,) prior to adsorption
tests. The red muds were also acid-treated (R~). The acid
treatment was carried out according to a modified version
of Shiao's procedure (Shiao and Akashi, 1977) by boiling
I00 g of water-washed and dried red mud in 2 dm 3 of
10% (by weigh0 HCI solution for 2h, filtering off,
thoroughly washing with water, drying and sieving to
obtain the Ra-sorbents. The acid-treatment technique,
which has also been applied by Wahlberg
et al.
(1964) to
clay minerals for improving their surface properties, has
been demonstrated with success in synthesizing a better
adsorbent from red muds in phosphorus (Shiao and
Akashi, 1977) and heavy metal (Apak and Unseren, 1987;
Apak
et al.,
1995, 1996) removal. However, acid treatment
of red mud sorbents had the drawback of the partial loss
of acid-soluble fractions like hematite. The Ra fraction
was further subjected to heat treatment at 600°C for 4 h
to obtain the Rah sorbents. The red muds (partly agglom-
erated due to relative humidity) could not be classified
with respect to true grain size as most were of 200 mesh
size in wet sieving.
The specific areas of Rw, Ra and Rah samples were 14.2,
20.7 and 28.0 m2/g, respectively, measured by the BET/Nz
method (Brunauer
et al.,
1938).
Coal fly ash was recovered from the cyclones and elec-
trostatic precipitators of the power plant and had the fol-
lowing average composition: CaO 42.5%, SiO2 21.9%,
SO3 13.6%, A1203 11.8%, Fe.zO3 2.4%, MgO 1.3%, K20
1.1%, Na20 0.9%, loss on ignition 4.4%. Almost 99% of
the fly ash could pass through a 200-mesh sieve. The raw
fly ash (F) was washed with 10-fold distilled water for sev-
eral (5-6) times, filtered and dried (Fw). A part of the Fw
was further treated several times with acid using 2% (by
weight) HCI in boiling solution for 2 h. Higher acidity (as
8.0
[] oaf
5.0 A c~F w
4.0
• 0.%
2.o3°1.o •
0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
C e ling/roLl
Fig. 2. Distribution coefficient of Cu (II) as a function of equilibrium aqueous concentration on fly
ashes and red muds.
Sorption of heavy metal cations 433
6.0
5.0 A ~'~w
PbF,,
4.0 • ~'Rw
1.0 ~
0.0
0.0 1.0 2.0 3.0 4.0 5.0 8.0 7.0 8.0 9.0
C e
(m mLl
Fig. 3. Distribution coefficient of Pb (II) as a function of equilibrium aqueous concentration on fly
ashes and red muds.
in the activation of red mud) was avoided due to severe
losses of fly ash components by solubilization. The solid
product was thoroughly washed with water, filtered, and
oven dried at 100 + 5°C to produce the acid-treated (Fa)
sorbent.
The X-ray diffractogram (Apak
et al.,
1996) of the Fw
identified 51% calcite (CaCO3), 32% anhydrite (CaSO4),
9% quartz (SiO2) and 3% hematite (Fe203) in the crystal-
line phase. Elemental analysis of selected spots in the het-
erogenous amorphous slag particles by the XRF technique
(Apak
et al.,
1996) yielded 41-52% CaO, 27% SiO2, 13%
A1203, 2-5% FeO, 1-4% MgO and up to 2% other ox-
ides.
The BET/N2 surface area of fly ash were 10.2 and
14.3 m2/g for Fw and Fa, respectively.
250
g
@
150
100
50
0
,' o 8 o
[]
CdF
A ca%
• cd%
)
0
0.0
[]
0
• •
0
2.0
Co(msI L)
!
4.0
Fig. 4. Isotherm of Cd (II) adsorption onto fly ashes and red muds.
6.0
434 Re,at Apak
et al.
160
120
80
40
[]
[]
[]
©
[]
©
®
[]
ml,
0.0
©
A
©
O
r •
0.4 0.8
Ce(mg/mL)
1.2
Fig. 5. Isotherm of Cu (II) adsorption onto fly ashes and red muds.
CuF
C.F a
c,e,,
c,e,
Point of zero charge (PZC) measurements by potentio-
metric titration of the sorbent suspensions at different
ionic strengths (Apak
et aL,
1995, 1996) yielded approxi-
mate
PZC
values of 6.4 and 8.3 for fly ash and red mud
sorbents, respectively.
When 1 g of sorbent was equilibrated with 50 ml dis-
tilled water, the indicated sorbents showed the following
approximate final pH in their aqueous leachates:
Rw Ra R~ F Fw Fa
pH 8.1 4.8 5.3 12.0 10.8 9.3
The acid-treated sorbents contained no free HCI but 15-
20 mg Cl-/g.
Batch sorption tests were carried out by agitating a sus-
pension of I g sorbent in 50 ml metal nitrate solution for
8 h (equilibration period) at room temperature
(25 +0.1°C) in stoppered flasks placed on a thermostatic
water-bath/shaker. After centrifugation, the remaining
metal concentration in the filtrate was determined by
flame AAS (Perkin Elmer 300, Norwalk, CT, U.S.A.) and
the equilibrium pH was measured by a pH-meter
(Metrohm E-512 Herisau, Switzerland) equipped with a
glass electrode.
The metal concentration retained in the sorbent phase
(qe, mg/g) was calculated by
qe = (Co - G)V/m
(1)
where Co and C~ are the initial and final (equilibrium) con-
centrations of the metal ion in solution (mg/litre), V is the
solution volume (litres) and m is the mass of sorbent (g).
The solid/water distribution ratios (at equilibrium) of
metals for both sorption and desorption were calculated
by
Ko = qo/G (2)
where KD is the empirical distribution ratio of the metal
cation M ((mg/g)/(mg/litre)= litres/g) determined on the
approximately linear portion of the corresponding adsorp-
tion isotherm.
Batch desorption tests were carried out by agitating I g
of metal loaded sorbent with 50 ml of the desired solution
until equilibrium (8 h).
The saturation capacities of the sorbents for the uptake
of indicated metals were determined by both batch and
column tests. For the latter, 40 g of adsorbent was filled at
a height of 8-11 cm in a thermostatic (25 + 0.1°C) column
of dimensions (h = 30 cm, ~b = 3 cm), and the adsorbate
solution was fed (counter to gravity) by a peristaltic pump
through the fixed bed of sorbent at a constant rate of
0.5 ml/min. The metal concentration of the eluate was
recorded against throughput volume. The dynamic metal
uptake capacities of the sorbents were calculated by the in-
tegration technique (Apak
et al.,
1996), i.e. the area above
the curve up to the line on which the eluate concentration
was equalized with the initial concentration of metal was
calculated. The total amount of retained metal was divided
by the mass of sorbent to yield the saturation capacity
(t.)col. x
RESULTS AND DISCUSSION
In weakly acidic-neutral suspensions whose pH
was attained naturally by equilibrating aqueous
metal nitrate-sorbent mixtures, the distribution
ratios generally increased with initial aqueous