Environmental Pollution 218 (2016) 168e175

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Competitive adsorption of Pb and Cd on bacteriaemontmorillonite

composite*
Huihui Du a, b, Wenli Chen a, *, Peng Cai a, b, Xingmin Rong b, Xionghan Feng b,
Qiaoyun Huang a, b, *
a
State Key Laboratory of Agricultural Microbiology, Huazhong Agricultural University, Wuhan 430070, China
b
Key Laboratory of Arable Land Conservation (Middle and Lower Reaches of Yangtze River), Ministry of Agriculture, College of Resources and Environment,
Huazhong Agricultural University, Wuhan 430070, China

a r t i c l e i n f o a b s t r a c t

Article history: The characteristics and mechanisms of competitive adsorption of trace metals on bacteria-associated clay
Received 6 June 2016 mineral composites have never been studied, despite their being among the most common organic
Received in revised form emineral complexes in geological systems. Herein, competitive adsorption of Pb and Cd on Pseudomonas
27 July 2016
putidaemontmorillonite composite was investigated through adsorptionedesorption experiment,
Accepted 5 August 2016
Available online 25 August 2016
isothermal titration calorimetry (ITC), and synchrotron micro X-ray uorescence (m-XRF). From the
experiment, stronger competition was observed on clay mineral than on bacteriaeclay composite
because more non-specic sites accounted for heavy metal adsorption on clay mineral surface at the
Keywords:
Pseudomonas putida
studied pH 5. Both competing heavy metals tended to react with bacterial fractions in the composite,
Montmorillonite which was veried by the higher correlation of Cd (and Pb) with Zn (R2 0.41) elemental distribution
Composite than with Si (R2 0.10). ITC results showed that competitive adsorption exhibited a lower entropy
Heavy metal change (DS) at the metal-sorbent interfaces compared with single-metal adsorption, revealing that Cd
Competitive adsorption and Pb are bound to the same types of adsorption sites on the sorbent. The competitive effect on bacteria
eclay composite was found to be helpful for a better understanding on the xation, remobilization and
subsequent migration of heavy metals in multi-metal contaminated environments.
2016 Elsevier Ltd. All rights reserved.

1. Introduction ascribed to both bacterial clay authigenesis (Konhauser and Urrutia,

Bacteria are widely considered as efcient bio-sorbents for
heavy metals (Ding et al., 2015; Huang et al., 2005; Guine
et al.,
2006). In soil systems, EPS/biolms formed by microbes largely
affect the speciation, toxicity and mobility of heavy metals, which
are important components of their biogeochemical cycling (Gadd,
2010; Priester et al., 2006). Clay minerals are also a key compo-
nent which governs trace metal immobilization and transport, due
to their large specic surface area, high cation exchange capacity
and reactive amphoteric hydroxyl groups (Celis et al., 2000). Clays
are spatially associated with bacterial communities in natural en-
vironments (Huang et al., 2015; Marshall, 1975). This association is
*
This paper has been recommended for acceptance by B. Nowack.
* Corresponding authors. State Key Laboratory of Agricultural Microbiology,
Huazhong Agricultural University, Wuhan 430070, China.
E-mail addresses: wlchen@mail.hzau.edu.cn (W. Chen), qyhuang@mail.hzau.
edu.cn (Q. Huang).
1999) and bacterial biolm formation on clay mineral surfaces
(Alimova et al., 2006; Lnsdorf et al., 2000). These processes give
rise to bacteria-associated clay minerals that are composite solids,
which consist of clay minerals intermixed with intact and partially
degraded bacterial cells. Properties of the bacteriaeclay complexes
differ considerably from their end-member phases, exerting a
diverse inuences on the chemical reactivity of heavy metals (Chen
et al., 2008; Du et al., 2016; Fang et al., 2010a; Flemming et al., 1990;
Ohnuki et al., 2007).
Trace metals usually coexist in a range of concentrations in
geological environments, e.g., soils and sediments exposed to
leaching water from mine tailings and smelters (Buchauer, 1973).
Under such circumstance, competitive adsorption effects need to
be addressed. Researches dealing with the competitive adsorption
of metals on clay minerals have been published over the past de-
cades. For instance, Srivastava et al. (2005) reported that the se-
lective adsorption sequence for multi-metal adsorption on
kaolinite was Pb > Cu > Zn > Cd, which was positively correlated

http://dx.doi.org/10.1016/j.envpol.2016.08.022
0269-7491/ 2016 Elsevier Ltd. All rights reserved.
 H. Du et al. / Environmental Pollution 218 (2016) 168e175
with their hydrolytic ability. de Pablo et al. (2011) observed the
selectivity sequence adsorption of
Hg2 > Zn2 > Ba2 > Cd2 > Ni2 > Mn2 on montmorillonite. The
selective sequence of Cu > Ni for binding on montmorillonite was
proposed by Yang et al. (2005a,b) which can be interpreted by their
metal properties such as electronegativity, Misono softness
parameter, rst hydrolysis constant, and electron congurations. A
few studies on the competitive adsorption effect of heavy metals by
bacterial biomass are available. For instance, Claessens and Van
Cappellen (2007) observed that Zn (59 mM) had no measurable
effect on Cu binding to Shewanella putrefaciens cells at a pH of 5.5.
Micheletti et al. (2008) showed that Cyanothece strain 16Som 2 is
an effective sorbent for the selective removal of Cr and Cu from
multiple-metal solutions. Moreover, binding was either non-
interactive, synergistic, or competitive between metal ions. Wang
et al. (2014) recently reported that EPS extracted from Desulfovi-
brio desulfuricans exhibited a higher binding ability with Zn than
Cu. FTIR analysis showed that Cu bind with protein, whereas Zn
combined with both protein and polysaccharides. In contrast, Yang
et al. (2015a,b) observed that Cu decreased signicantly Zn
adsorption on EPS extracted from Klebsiella sp. J1, but the uptake of
Cu was not inuenced by Zn. They attributed this nding to the
stronger complexation ability between Cu and the protein-like
substances in EPS as shown by using FTIR and uorescence
spectrophotometry.
Most of aforementioned studies examined competitive
adsorption behaviors of trace metals by individual soil components
(e.g., mineral and bacteria). However, soils are more complex.
Investigating competitive adsorption of metals on composite solids
(e.g., bacteria-associated mineral complexes) is critical. Only one
report on the competitive adsorption of heavy metals on goethi-
teeBacillus thuringiensis composite was found; the competitive
sorption intensity between Cu and Cr was observed to be greater on
the goethiteeB. thuringiensis complex than on goethite (Zhu et al.,
2012). The competitive effect of heavy metals on bacteriaeclay
mineral composites has never been studied despite their being the
most common organicemineral complexes in soils and sediments.
The present work aimed to investigate the competitive
adsorption of Cd and Pb on bacteriaemontmorillonite composite.
Isothermal titration calorimetry (ITC) was employed for the rst
time to determine the mechanisms of competitive adsorption from
a thermodynamic aspect. Synchrotron micro X-ray uorescence (m-
SRXRF) was used to quantify the distribution of Cd and Pb within
the composite. New insight on the competitive binding of Cd and Pb
at the bacteriaemineral interface enhances our fundamental un-
derstanding of the retention and release and thus increasing our
knowledge on biogeochemical cycling and endpoint of heavy
metals in soil micro-environments, where bacteria-associated clay
mineral complexes are dominant or make up a signicant propor-
tion of reactive solid phases.
2. Materials and methods
2.1. Sorbents
Pseudomonas putida X4 is an aerobic soil gram-negative bacte-
rial strain. Bacterial cells were cultured at 28  C in a Luria broth
medium (pH 7.2), containing 5 g/L of bacto-yeast extract, 10 g/L of
bacto-tryptone and 5 g/L of NaCl, to the late-exponential growth
phase as described by Wu et al. (2014). Bacterial cells were obtained
by centrifugation and were washed three times with the electrolyte
(0.01 M KNO3) to remove excess media. Fresh bacterial biomass was
used for the experiments. Montmorillonite (Mont) is a represen-
tative 2:1 type clay minerals, and is widely distributed in soil en-
vironments. Montmorillonite used in this work was purchased
169
from Zhejiang Sanding Group Co. Ltd. (Shaoxing, Zhejiang, China).
Clay-sized fractions of montmorillonite (<2 mm) were isolated
based on the procedures outlined by Rong et al. (2007). Montmo-
rillonite and bacterial cells were all suspended in 0.01 M KNO3
electrolyte. The bacteriaeMont composite was obtained by mixing
bacterial and mineral suspension in a 1:1 ratio (dry-weight basis)
and incubating it for 4 h at pH 5 to achieve equilibrium.
2.2. Adsorption isotherms of Cd and Pb in single and binary systems
Batch adsorption experiments were carried out by mixing the
sorbent (Pseudomonas putida, Mont, and P. putidaeMont compos-
ite) suspension with different amounts of Cd and Pb solutions
separately (single system) or together (competitive system; initial
molar ratio of Cd to Pb was 1:1) in 0.01 M KNO3 background elec-
trolyte. The sorbent concentration was 1 g/L (dry-weight basis), and
the initial metal concentrations ranged from 0 to 0.5 mM. All the
sorbent concentrations shown below refer to dry mass unless it is
specied otherwise. All pH measurements were made with a digital
pH meter (PB, Sartorius, Germany) which has been calibrated
successively by standard buffer solution (6.86, 4.01 and 9.18). The
pH of the mixed suspensions was adjusted to pH 5 using minor
additions of 0.1 M HNO3 or 0.1 M KOH solution. The mixtures were
then incubated on a rotary shaker for 4 h (28  C). During this period,
pH adjustments of the suspensions were made every 30 min. After
equilibration, suspensions were centrifuged at 8000 g, superna-
tants were ltered through 0.45 mm nylon lter membranes and
acidied with HNO3. Residual Cd and Pb were measured through
atomic absorption spectrometry (AAS; Varin AAS240FS). Solid
phases of the bacteriaeMont composite after adsorbing Cd and Pb
in the binary-metal systems were collected for m-SRXRF
experiments.
2.3. Exchange of pre-adsorbed Cd and Pb
Pre-adsorption experiments were performed by mixing 1 g/L
sorbent (P. putida, Mont, and P. putidaeMont composite) suspen-
sion with 0.5 mM of Cd or Pb solutions separately, with 0.01 M
KNO3 as the background electrolyte. The pH of the suspension was
buffered to 5, and then the suspension was incubated for 4 h
(28  C). After equilibration, suspensions were centrifuged at 8000 g,
supernatants were ltered through 0.45 mm membranes and
acidied with HNO3 for AAS analysis. The solid phase was re-
suspended in 0.01 M KNO3 background electrolyte, then an equal
amount of the second metal (Pb or Cd, 0.5 mM) solutions were
added to the mixture and incubated for another 4 h at pH 5. The
mixtures were then centrifuged at 8000 g, and the concentrations
of Cd and Pb remaining in solution were measured with AAS (Varin
AAS240FS). During the adsorption-desorption experiments, the pH
of the mixed suspensions was adjusted every 30 min. The experi-
mental settings of the exchange experiments were to simulate
the competitive-ITC experiment shown below, i.e., one metal was
adsorbed (titrated) on the sorbent in the presence of another metal.
2.4. Isothermal titration calorimetry
A Thermal Active Monitor III (TAM III) system (Thermometric
AB, Sweden) was used to measure the heat effect at 28  C. For
single-metal systems, the adsorbent (P. putida, Mont, and
P. putidaeMont composite, 10 g/L, pH 5) were placed in a 1 mL
stainless steel ampoule. Automatic titration with metal ion (Cd or
Pb) in the syringe (2  102 M) was performed. For competitive-ITC,
it was unable to obtain the competitive thermodynamic parame-
ters if the two metals were titrated simultaneously. Therefore a
new method was established according to Chekmeneva et al.
170 H. Du et al. / Environmental Pollution 218 (2016) 168e175
(2008). Briey, the binary adsorbent (P. putida, Mont, and
P. putidaeMont composite, 10 g/L, pH 5) and metal (Cd or Pb, 5 mM)
were placed in the reaction cell and then titrated with the second
metal ion in the syringe (Pb or Cd, 2  102 M). Actually this
experimental setting of the competitive-ITC was similar to the
exchange experiments as mentioned above. Therefore the data of
competitive-ITC and the exchange experiments were compara-
ble. For a typical experiment, 10 mL of titrant was injected into the
reaction cell solution at a rate of 1 mL/s with 360 s intervals between
injections. The heat of dilution, measured by injecting the titrant to
the background electrolyte (0.01 M KNO3, pH 5), was subtracted
from each titration to obtain net reaction heat values. This sub-
traction of background heat could be found elsewhere (Fang et al.,
2010b; Du et al., 2016). Metal and adsorbent suspensions were
suspended in 0.01 M KNO3 (pH 5) media. Data processing was
performed using the program NanoAnalyze software version 3.1.2
(TA Instruments).
2.5. X-ray imaging
Synchrotron micro X-ray uorescence (m-SRXRF) enables in situ
mapping of abundance of elements at a sub-micrometer scale with
negligible sample damage and high sensitivity (Majumdar et al.,
2012). The solid phase of bacterialeMont composite was trans-
ferred after absorbing Cd and Pb in the binary-metal system to
quantify their distribution within the composite through X-ray
imaging at the beamline 15U1 of the Shanghai Synchrotron Radi-
ation Facility (Shanghai Institute of Applied Physics, Chinese
Academy of Science). During measurement, storage ring energy
was 3.5 GeV, and the maximum stored current was approximately
300 mA. Fluorescence signals were recorded using a Si(III) solid-
state multi-element detector. Excitation of energy was xed at
16 keV, and uorescence spectra were collected with spot size of
3 mm  3 mm. In addition to Cd and Pb, Zn, Si, S, P, Al, and Ca ele-
ments were detected. Trace elements of Zn, S and P are derived
from the Luria-broth nutrient solution, and they are essential ele-
ments for bacterial growth. Si and Al are most abundant in silicon-
oxygen tetrahedron and aluminium-oxygen octahedron of clay
Fig. 1. Adsorption isotherm of Cd (up) and Pb (below) in single (solid) and binary (hollow) systems on montmorillonite (Mont), Pseudomonas putida, and P. putidaeMont composite
at a pH of 5 in 0.01 M KNO3. Solid lines represent Langmuir model tting.
minerals. Signals were collected simultaneously, and Igorpro 6.10
(WaveMatrics, Inc.) software was used to process the images (Chen
et al., 2011).
3. Results
3.1. Metal adsorption isotherms
Adsorption isotherms of Cd and Pb in single and binary systems
on Mont, P. putida, and P. putidaeMont composite are shown in
Fig. 1. The adsorption data for single and binary-metal systems can
be well tted to a Langmuir model (Srivastava et al., 2006), i.e.
qm KCe
qe (1)
1 KCe
where qe is the amount of metal adsorbed at an equilibrium metal
concentration of Ce, qm is the maximum of adsorbed metal, and K
represents an afnity constant. It must be pointed that K values are
conditional ones, in this study, they only account for a 1:1 M ratio
condition.
In single-metal systems, the high correlation coefcients
(R2  0.98) obtained from Langmuir isotherm indicate that the
adsorption of Cd and Pb follows monolayer coverage on the sorbent
surface. P. putida and Mont have the highest and lowest metal
adsorption capacity, respectively (Langmuir parameter qm). The
binary P. putidaeMont composite exhibits metal adsorption
behavior that is intermediate to P. putida and Mont with qm at
135.5 mmol Cd/kg and 215.6 mmol Pb/kg. The equilibrium con-
stants for Cd and Pb adsorption on Mont (Langmuir parameter K,
Table 1) are lower in P. putida and P. putidaeMont composites,
indicating a higher binding afnity between heavy metals and the
bacterial fractions. In binary-metal systems, Langmuir model also
yields good ts with R2  0.97 (Table 1). The adsorption capacities
of Cd and Pb decrease compared with those in single-metal sys-
tems, indicating competition between Cd and Pb ions for the
adsorption sites. However, K values for the binary systems are
higher than in single systems. This result implies that competition
 H. Du et al. / Environmental Pollution 218 (2016) 168e175 171

Table 1 3.3. Thermodynamic characteristics

Langmuir parameters (qm, K and R2) for the adsorption of Cd and Pb in single and
binary systems on montmorillonite (Mont), P. putida, and P. putidaeMont composite
at pH 5 in 0.01 M KNO3.
Isothermal titration calorimetry (ITC) has emerged as a powerful
tool for acquiring the adsorption mechanisms of heavy metals on
Adsorbent Metal System Langmuir model different adsorbent from a thermodynamic aspect (Du et al., 2016;
qma (mmol/kg) Kb (L/mol) R2 Fang et al., 2010a), however, no studies concerning the competitive
Mont Cd Cd alone 73.18 9.5 0.98 thermodynamic adsorption, have been performed. In this study, a
Cd Pb 22.11 27.0 0.97 new method for evaluating the competitive thermodynamic char-
Pb Pb alone 80.01 23.0 0.95 acteristics of Cd and Pb on bacteria, mineral and their composite
Pb Cd 55.43 46.0 0.97
was established. The solid-line curves in Figs. 2 and 3 depict raw
P. putida Cd Cd alone 181.35 27.9 0.99
Cd Pb 78.15 61.1 0.97 calorimetry data for Cd and Pb adsorption on different adsorbents
Pb Pb alone 357.63 34.2 0.99 in single (orange) and binary-metal systems (blue). The upper and
Pb Cd 295.47 41.4 0.99 lower panels of each graph represent calorimetric titrations (heat
P. peMont Cd Cd alone 135.50 23.3 0.98 ow) and integrated heat values, respectively. Curve tting was
Cd Pb 58.04 55.1 0.98
conducted using the independent model (one set of sites), thus
Pb Pb alone 215.60 32.9 0.98
Pb Cd 179.58 39.9 0.97 allowing all parameters to oat (solid line). Fitting results are dis-
a played in Table 3. Some reaction systems cannot be modeled
Maximum capacity of metal adsorption on different components as predicted by
Langmuir isotherm. because of weak heat effects, and the thermodynamic data are not
b
Equilibrium constant related to the binding energy. shown. It must be pointed out that all processes (e.g, protonation/
deprotonation, metal adsorption, etc.) contribute to the thermo-
dynamic effect, in particular for competitive ITC experiment where
for adsorption sites promotes bonding of both metals on more a desorption effect is also involved. The thermodynamic parame-
specic sites in binary-metal systems. The adsorption capacity (qm) ters reect the overall effect of a reaction and are model dependent.
of Cd is reduced by 69.8% on Mont, 56.9% on P. putida, and 57.0% on In single-metal systems, the overall effects of Cd adsorption on
P. putidaeMont composite, whereas that of Pb decreases by 30.7% Mont, P. putida, and P. putidaeMont composite are clearly all
on Mont, 17.4% on P. putida, and 16.7% on the P. putidaeMont endothermic processes (Fig. 2) with DH values ranging from 12.6 kJ/
composite (Table 1). There is the occurrence of greater inhibitions mol to 16.7 kJ/mol (Table 3). Unlike Cd, Pb adsorption on P. putida is
of Pb on Cd than Cd on Pb which suggests that Pb has higher an exothermic process (9.4 kJ/mol), whereas that on Mont is
selectivity of sorbents than Cd. Under the studied mass ratio of endothermic (DH 11.3 kJ/mol) (Fig. 3 and Table 3). Negative sig-
bacteria to mineral of 1:1, bacteria-associated mineral composite nals are detected on P. putidaeMont after the rst two injections,
reacts in a manner quite similar to end-member bacteria in a whereas positive signals were detected afterward. This phenome-
competitive metal adsorption process. non suggests the preferential adsorption of Pb on bacterial cells in
the bacteriaeclay composite than on clay mineral. Entropies can
provide valuable information on the coordination environment of
complexed metals at sorbent surfaces, e.g., outer-sphere versus
3.2. Metal exchange inner-sphere complexation and metaleligand stoichiometry
(Gorman-Lewis, 2014). Table 3 shows that the entropies of Cd and
Exchange of Cd and Pb, which was previously adsorbed on Mont, Pb adsorption on the sorbents are largely positive, suggesting an
P. putida, and P. putidaeMont composite by the second metal, was increased randomness at the metal-sorbent interface. Positive en-
performed (Table 2). The percentage of Cd exchanged by Pb is tropies are assigned to inner-sphere complexes because of the
65.5%, 25.3%, and 32.1% from Mont, P. putida, and P. putidaeMont, atomic rearrangement and liberation of water molecules originally
respectively; whereas the percentage of Pb exchanged by Cd is associated with the sorbent during surface complexation (Gorman-
14.9%, 3.1%, and 3.5% from Mont, P. putida, and P. putidaeMont, Lewis et al., 2006; Ozdes et al., 2011). Consequently, inner-sphere
respectively. These results suggest that Pb has a stronger ability to complexation is probably the main adsorption mechanism of Cd
displace Cd from its complexes with the sorbent. The three and Pb on bacteria and their composite. Moreover, the calculated
different sorbent systems are compared, and results show that Cd thermodynamic afnity (K, Table 3) indicates a higher binding af-
and Pb are more easily desorbed from pure clay mineral than from nity of Pb than Cd toward the bacteria, clay mineral, and bacter-
bacteriaeclay composite. Thus, adsorption sites on the mineral iaeclay composite, which agrees with isotherm results. The
surface are readily exchangeable. Moreover, the exchange behavior thermodynamic driving force of the metal-solid surface reactions
of P. putidaeMont composite resembles that of P. putida, indicating which can be obtained by comparing jTDSj and jDHj. For Cd,
that the bacterial fraction greatly inuences the competitive jTDSj > jDHj (DH > 0) are observed on the bacteria, clay, and their
adsorption/desorption process of Cd and Pb on the bacteriaeclay composite, which indicates that the adsorption reaction of Cd are
composite. all driven by entropy. For Pb, jTDSj > jDHj (DH > 0) is observed for
clay systems, suggesting that the adsorption process is driven by
entropy, whereas those on the bacteria and bacteriaeclay com-
Table 2
Desorption of adsorbed Cd or Pb by the addition of the second metal from mont- posite are driven by both enthalpy and entropy (jTDSj < jDHj,
morillonite (Mont), P. putida, and P. putidaeMont composite at pH 5 in 0.01 M KNO3. DH < 0).
The total cumulative heats are extremely small
Adsorbent Adsorbed Adsorption Desorption Exchange
metal (mmol/kg) (mmol/kg) percentage (%) (0.05 mJe0.07 mJ) (Table 3) for the addition of Cd to P. putidaePb
and compositeePb, whereas the total cumulative heat is 0.35 mJ for
Mont Cd 59.15 38.74 65.5
Pb 78.30 11.74 14.9 the addition of Cd to MontePb, which is substantially lower than
P. putida Cd 163.35 41.32 25.3 the non-competing system. This result suggests that the adsorption
Pb 350.21 10.85 3.1 of Cd on the P. putidaePb and P. putidaeMontePb system seems
P. peMont Cd 125.50 40.28 32.1 impossible, but a small proportion of Cd can be adsorbed on
Pb 203.65 7.12 3.5
MontePb. Thermodynamic results are consistent with exchange
172 H. Du et al. / Environmental Pollution 218 (2016) 168e175

Fig. 2. Isothermal titration data for the addition of Cd to montmorillonite (Mont), P. putida, and P. putidaeMont composite in 0.01 M KNO3 background electrolyte (pH 5) at 28  C in
the absence (orange) or presence (blue) of Pb. The upper and lower panels of each graph represent calorimetric titrations (heat ow) and integrated heat values, respectively. The
solid line represents the best independent model tting results. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this
article.)

Fig. 3. Isothermal titration data for the addition of Pb to montmorillonite (Mont), P. putida, and P. putidaeMont composite in 0.01 M KNO3 background electrolyte (pH 5) at 28  C in
the absence (orange) or presence (blue) of Cd. The upper and lower panels of each graph represent calorimetric titrations (heat ow) and integrated heat values, respectively. The
solid line represents the best independent model tting results. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this
article.)

experiments, showing that Pb is difcult to desorb from P. putida
and P. putidaeMont composite by the addition of Cd, and small
amounts of Pb can be exchanged from Mont.
Considerable heat effects are observed when Pb is added to the
systems of MonteCd, P. putidaeCd, and compositeeCd, with total
heats of 4.52, 40.82, and 15.87 mJ, respectively (Table 3), indi-
cating large amounts of Pb adsorbed in these systems. This result
demonstrates that a large amount of Cd can be displaced from the
sorbent surface. Heat effect changes vary with the sorbent. For
P. putida and P. putidaeMont composite system, higher exothermic
effects are observed for competing systems (9.7 kJ/mol and 6.4 kJ/
mol) compared with non-competing systems (9.4 and 5.7 kJ/mol).
For the Mont system, the competing system has a lower enthalpy
(5.1 kJ/mol) than the non-competing system (11.3 kJ/mol). Given
that the adsorption of Cd on Mont is endothermic, the desorption
process of Cd is exothermic. Therefore, the lower heat effect for
competing systems on Mont is most probably due to the
exothermic effects of Cd desorption neutralizing the endothermic
 H. Du et al. / Environmental Pollution 218 (2016) 168e175 173

Table 3
Thermodynamic parameters for Cd and Pb adsorption in single and binary systems on montmorillonite (Mont), P. putida, and P. putidaeMont composite at pH 5 and 28  C in
0.01 M KNO3 electrolyte.

Titration Total heat (mJ) K (L/mol) DG (kJ/mol) DH (kJ/mol) DS (J/mol/K)

Cd2 / Mont 5.02 0.21 (2.10 0.14)  102 13.4 0.1 12.6 0.3 86.3 2.1
Cd2 / MontePb2 0.35 0.01
Cd2 / P. putida 7.55 0.12 (4.00 0.21)  102 15.0 0.2 16.7 0.5 105.2 2.8
Cd2 / P. putidaePb2 0.05 0.01
Cd2 / Composite 6.57 0.08 (3.40 0.14)  10 2
14.6 0.2 12.8 0.8 91.0 3.0
Cd2 / CompositeePb2 0.07 0.01

Pb2 / Mont 6.79 0.11 (2.60 0.12)  102 13.9 0.1 11.3 0.1 83.8 3.1
Pb2 / MonteCd2 4.52 0.06 (1.09 0.11)  102 11.7 0.1 5.1 0.1 55.8 1.2
Pb2 / P. putida 34.41 1.32 (2.64 0.22)  104 25.5 0.2 9.4 0.2 53.6 2.0
Pb2 / P. putidaeCd2 40.82 1.07 (1.12 0.12)  104 23.3 0.4 9.7 0.7 45.1 1.9
Pb2 / Composite 8.65 0.65 (5.85 0.24)  103 21.7 0.6 5.7 0.4 53.1 1.8
Pb2 / CompositeeCd2 15.87 0.74 (2.30 0.18)  103 19.4 0.9 6.4 0.3 43.2 1.5

heat effect of Pb adsorption on Mont. In contrast, the exothermic
heat effects of Cd desorption increase the exothermic effects of Pb
adsorption on bacteria and bacteriaeclay composite.
On P. putida and P. putidaeMont, the calculated thermodynamic
afnity (K, Table 3) for competing systems is interestingly lower
than that for non-competing systems. This trend seems inconsis-
tent with the one obtained from isothermal absorption experi-
ments possibly because of two different conditions: competitive
and pre-adsorbed. For the pre-adsorbed system, the second
metal preferentially displaces pre-adsorbed metals from low-
afnity sites; thus, binding constants are lower. In contrast, in the
competitive system, heavy metals prefer to bind on more specic
sites than low-afnity sites, resulting in greater K values. Different K
values obtained from both circumstances are therefore consistent
with each other.
The entropy change (DS) data for binary-metal systems
(43.2e55.8 J/mol/K) are lower than that of single-metal systems
(53.1e83.8 J/mol/K). Chekmeneva et al. (2008) reported lower
competitive thermodynamic data for Cd and Zn on phytochelatin
(g-Glu-Cys)4-Gly (PC4) possibly because thiol groups are shared
between Cd and Zn in the intermediate mixed complex. In the
present study, lower entropies for competing systems suggest
sharing of adsorption sites between Cd and Pb on the sorbent
surface. No complete inhibition occurs in the complex system. It is
likely that Cd and Pb be adsorbed to the same types of sites on the
sorbent surface.
3.4. Elemental distribution
High-energy X-ray uorescence experiments were performed to
study Cd and Pb distributions within the bacteriaeclay composite.
m-SRXRF allows a direct method for measuring elemental distri-
butions by its synchrotron-induced X-ray uorescence signature
without any added reagents or pretreatments (Kemner et al., 2004).
X-ray uorescence intensities (element concentration) are dis-
played with different colors, from blue (minimum) to red
(maximum) for individual elements (Fig. 4). Elemental intensity
values on a pixel-by-pixel basis are compared to determine corre-
lation coefcients between different elements.
The spatial distribution of Cd in the micro-XRF maps is highly
correlated with Pb (R2 0.97). Moreover, based on elemental in-
tensities (counts), the Pb concentration in the sample is higher than
the Cd concentration. Few separate phases of Cd and Pb are present,
which demonstrates that Cd and Pb are mainly adsorbed on the
same types of adsorption sites of the composite. The m-XRF map
reveals that the Zn distribution is locally distributed with Cd with
correlation R2 0.41. The uniform distribution of Si over rather
large areas of maps shows no correlation with that of Cd (R2 0.10).
Kemner et al. (2004) showed that the elemental distribution of Zn
can represent bacterial cells. The Si signal represents mineral
fractions because silica is the basic component of montmorillonite.
Therefore, in the present study, the higher correlation of Cd with Zn
than with Si indicates that Cd (and Pb) are mainly adsorbed on
bacterial cells in the composite. Combined with the adsorption
experiments, stronger competition on clay mineral fraction results
in the combination of heavy metals to the bacterial cells at the
bacteriaeclay mineral interfaces.
4. Discussion
Our results show that competitive adsorption behaviors of Cd
and Pb on bacteriaemontmorillonite composite resemble end-
member bacteria, and competition is weaker than that on end-
member clay mineral. There is a stronger competition on clay
mineral than on bacteriaeclay composite, which may result from
diverse binding mechanisms of heavy metal ions on different sor-
bents. Metal adsorption on soil components can be either specic
or non-specic. Non-specic adsorption is believed to form outer-
sphere complexes and is generally not stronger than specic
adsorption (inner-sphere complexes) (Antoniadis and Tsadilas,
2007). Many studies document that the competitive adsorption
among heavy metals takes place mainly on low-afnity sites
because the outer-sphere complex is more reversible than the
inner-sphere complex (Antoniadis and Tsadilas, 2007; Zhu et al.,
2012). In this study, although entropies for single Cd
(86.3e105.2 J/mol/K) and Pb (53.1e83.8 J/mol/K) adsorption sug-
gest that Cd and Pb mainly form inner-sphere complexes on clay
mineral, bacteria and their composite, a relative proportion of
outer-sphere complexes cannot be excluded. The entropy effect of
the outer-sphere complexes may be concealed by the stronger
positive entropy effect of the inner-sphere. Given the adsorption
experimental results, the stronger competition on clay mineral
than on bacteriaeclay composite suggests that more low-afnity
sites account for competitive adsorption on the clay mineral sur-
face. Our exchange experiments also show easier desorption be-
haviors of Cd and Pb on clay mineral. These sites are usually
recognized as permanently charged basal surface sites (X) (Gu
et al., 2010). However, on the bacteriaeclay composite, anionic
functional groups of carboxyl, phosphoryl, and amine are abundant
on bacterial cell walls. These groups form strong inner-sphere
complexes with metal ions (Mishra et al., 2010; Sun et al., 2016).
Although inner-sphere metal complexes on bacterial surface can be
reversible, these sites might be difcult to be exchanged by another
metal than the outer-sphere metal complexes on mineral surface.
Thus, the competition on bacteriaeclay composite is weaker
because of the lower reversibility and difcultly exchangeable of
174 H. Du et al. / Environmental Pollution 218 (2016) 168e175
Fig. 4. Micro-XRF maps show the selected element distribution of Cd, Pb, Zn, S, P, Ca, Si, and Al in the montmorillnoteeP. putida composite sample after adsorbing Cd and Pb at a pH
of 5 in 0.01 M KNO3. Scanning step is 3 mm  3 mm, and the total map area is 27  30 mm2. The gure below shows correlation between counts of Cd with Pb, Zn, and Si, respectively.
adsorption sites on these complexes.
It seems that bacteria may massively outweigh the clay mineral
in terms of metal uptake under the studied mass ratio of 1:1
(bacteria to clay), and this 1:1 condition may not speak for all kinds
of bacteriaeclay composites which have different component ratios
in natural setting. However, it is speculated that the competitive
effect is not merely dependent on the total adsorption sites
assigned to each fraction, but largely depends on the surface sites
that are readily exchangeable. The composite system must show
many differences as compared to the single-component system no
matter under what mass ratio conditions. This difference are based
on the different binding mechanisms of heavy metals on the end-
member component. Competitions easily occur on clay mineral
fractions than on bacterial fraction, and heavy metals tended to
react with bacterial fractions in the composite. m-XRF also
conrmed the accumulation of metal ions on bacterial fractions in
the composites. Consequently, the conclusion obtained in this work
has the potential to apply to different component ratio systems.
Heavy metals tend to retain on bacterial component at the
organoeclay mineral interfaces which is important in governing
their xation and migration in natural environment, such as in
rhizosphere soil or bacterial clay authigenesis-dominated sedi-
ments, where a signicant proportion of bacterial biomass might be
associated with phyllosilicates. These organic-bound metals exhibit
a higher stability against replacement by other metals compared
with clay-adsorbed metals. Additionally, previous study showed
that abiotic organicemineral complexes exhibit different compet-
itive metal adsorption behavior from pure mineral phases,
depending on metal species and pH (Hizal et al., 2009; Zhu et al.,
2010). Our results, combined with those of these previous works,
demonstrate the signicance of abiotic and/or biotic organic mat-
ters in affecting competitive metal adsorptions on soil minerals.
Under this premise, mineraleorganic composite is a more suitable
analogue for a good predictor of the mobility and thus eventual fate
of heavy metals in multi-metal polluted environments. Moreover,
we should consider dynamic changes of the composite solid pha-
ses. Because these biotic organic substances (e.g., bacterial cells)
may crack into soluble macromolecules because of cell lysis,
degradation, and long-term changes of environmental factors (e.g.,
pH, temperature, and human activities), thus inuencing the
mobility and fate of heavy metals through soil and groundwater
movement. Lastly, our study shows progress in the using of ITC to
investigate the competitive mechanism of a reaction from a ther-
modynamic aspect. Due to its wide-adaptability, this approach has
the potential for wide application in the area of environmental
science, where competitive adsorption process are involved.
5. Conclusion
We present the rst attempt to achieve competitive adsorption
of heavy metals on bacteriaeclay mineral interfaces. Isothermal
titration calorimetry (ITC) combined with synchrotron micro X-ray
uorescence (m-SRXRF) has been demonstrated as an effective way
to study the competitive adsorption process. We found that bac-
teriaeclay composites showed signicant differences in the multi-
metal adsorption process compared with their pure clay mineral
 H. Du et al. / Environmental Pollution 218 (2016) 168e175
end-member (i.e., stronger competition for pure clay mineral than
for clay mineralebacteria composite). This strong competition on
clay fraction causes heavy metals to combine with bacterial factions
in the composite. The competitive adsorption process results in
lower entropy changes (DS) at the metalesorbent interfaces
compared with single-metal adsorption reactions, revealing that Cd
and Pb share the same types of adsorption sites on the sorbent
surface. This work highlights competitive adsorption behavior of
heavy metals at the interface of organoeclay mineral composites,
which is helpful for us to gain a deeper understanding of their
immobilization and transport in multi-metal polluted
environments.
Acknowledgements
We thank the National Natural Science Foundation of China
(NSFC) (No. 41230854), NSFCeRS (Royal Society) joint project (No.
4151101315), and the Fundamental Research Funds for the Central
Universities (2662015PY016, 2662015PY116) for the nancial sup-
port of this research. We are also grateful to the Shanghai Syn-
chrotron Radiation Facility (SSRF, BL15U1) for supporting the m-
SRXRF experiments.
References
Alimova, A., Roberts, M., Katz, A., Rudolph, E., Steiner, J., Alfano, R., Gottlieb, P., 2006.
Effects of smectite clay on biolm formation by microorganisms. Biolms 3,
47e54.
Antoniadis, V., Tsadilas, C.D., 2007. Sorption of cadmium, nickel, and zinc in mono-
and multimetal systems. Appl. Geochem. 22, 2375e2380.
Buchauer, M.J., 1973. Contamination of soil and vegetation near a zinc smelter by
zinc, cadmium, copper, and lead. Environ. Sci. Technol. 7, 131e135.
Celis, R., Hermosn, M.C., Cornejo, J., 2000. Heavy metal adsorption by functional-
ized clays. Environ. Sci. Technol. 34, 4593e4599.
Chekmeneva, E., Prohens, R., Daz-Cruz, J.M., Arin ~ o, C., Esteban, M., 2008.
Competitive binding of Cd and Zn with the phytochelatin (g-Glu-Cys)4-Gly:
comparative study by mass spectrometry, voltammetry-multivariate curve
resolution, and isothermal titration calorimetry. Environ. Sci. Technol. 42,
2860e2866.
Chen, G., Chen, X., Yang, Y., Hay, A.G., Yu, X., Chen, Y., 2011. Sorption and distribution
of copper in unsaturated Pseudomonas putida CZ1 biolms as determined by X-
ray uorescence microscopy. Appl. Environ. Microb. 77, 4719e4727.
Chen, X., Chen, L., Shi, J., Wu, W., Chen, Y., 2008. Immobilization of heavy metals by
Pseudomonas putida CZ1/goethite composites from solution. Colloid Surf. B 61,
170e175.
Claessens, J., Van Cappellen, P., 2007. Competitive binding of Cu2 and Zn2 to live
cells of Shewanella putrefaciens. Environ. Sci. Technol. 41, 909e914.
de Pablo, L., Ch
avez, M.L., Abatal, M., 2011. Adsorption of heavy metals in acid to
alkaline environments by montmorillonite and Ca-montmorillonite. Chem. Eng.
J. 171, 1276e1286.
Ding, C., Cheng, W., Sun, Y., Wang, X., 2015. Effects of Bacillus subtilis on the
reduction of U(VI) by nano-Fe0. Geochim. Cosmochim. Ac. 165, 86e107.
Du, H., Chen, W., Cai, P., Rong, X., Dai, K., Peacock, C.L., Huang, Q., 2016. Cd(II)
sorption on montmorillonite-humic acid-bacteria composites. Sci. Rep. 6,
19499.
Fang, L., Cai, P., Li, P., Wu, H., Liang, W., Rong, X., Chen, W., Huang, Q., 2010a.
Microcalorimetric and potentiometric titration studies on the adsorption of
copper by P. putida and B. thuringiensis and their composites with minerals.
J. Hazard Mater. 181, 1031e1038.
Fang, L., Huang, Q., Wei, X., Liang, W., Rong, X., Chen, W., Cai, P., 2010b. Microca-
lorimetric and potentiometric titration studies on the adsorption of copper by
extracellular polymeric substances (EPS), minerals and their composites. Bio-
resour. Technol. 101, 5774e5779.
Flemming, C.A., Ferris, F.G., Beveridge, T.J., Bailey, G.W., 1990. Remobilization of toxic
heavy metals adsorbed to bacterial wall-clay composites. Appl. Environ. Microb.
56, 3191e3203.
Gadd, G.M., 2010. Metals, minerals and microbes: geomicrobiology and bioreme-
diation. Microbiology 156, 609e643.
Gorman-Lewis, D., Fein, J.B., Jensen, M.P., 2006. Enthalpies and entropies of proton
and cadmium adsorption onto Bacillus subtilis bacterial cells from calorimetric
175
measurements. Geochim. Cosmochim. Ac. 70, 4862e4873.
Gorman-Lewis, D., 2014. Enthalpies and entropies of Cd and Zn adsorption onto
Bacillus licheniformis and enthalpies and entropies of Zn adsorption onto Ba-
cillus subtilis from isothermal titration calorimetry and surface complexation
modeling. Geomicrobiol. J. 31, 383e395.
Gu, X., Evans, L.J., Barabash, S.J., 2010. Modeling the adsorption of Cd (II), Cu (II), Ni
(II), Pb (II) and Zn (II) onto montmorillonite. Geochim. Cosmochim. Ac. 74,
5718e5728.
Guine
, V., Spadini, L., Sarret, G., Muris, M., Delolme, C., Gaudet, J.P., Martins, J.M.,
2006. Zinc sorption to three gram-negative bacteria: combined titration,
modeling, and EXAFS study. Environ. Sci. Technol. 40, 1806e1813.
Hizal, J., Apak, R., Hoell, W.H., 2009. Modeling competitive adsorption of copper(II),
lead(II), and cadmium(II) by kaolinite-based clay mineral/humic acid systems.
Environ. Prog. Sustain. 28, 493e506.
Huang, Q., Chen, W., Xu, L., 2005. Adsorption of copper and cadmium by Cu and Cd-
resistant bacteria and their composites with soil colloids and kaolinite. Geo-
microbiol. J. 22, 227e236.
Huang, Q., Wu, H., Cai, P., Fein, J.B., Chen, W., 2015. Atomic force microscopy mea-
surements of bacterial adhesion and biolm formation onto clay-sized particles.
Sci. Rep. 5, 16857.
Kemner, K.M., Kelly, S.D., Lai, B., Maser, J., O'Loughlin, E.J., Sholto-Douglas, D., Cai, Z.,
Schneegurt, M.A., Kulpa, C.F., Nealson, K.H., 2004. Elemental and redox analysis
of single bacterial cells by X-ray microbeam analysis. Science 306, 686e687.
Konhauser, K.O., Urrutia, M.M., 1999. Bacterial clay authigenesis: a common
biogeochemical process. Chem. Geol. 161, 399e413.
Lnsdorf, H., Erb, R., Abraham, W., Timmis, K., 2000. Clay hutches: a novel inter-
action between bacteria and clay minerals. Environ. Microbiol. 2, 161e168.
Majumdar, S., Peralta-Videa, J.R., Castillo-Michel, H., Hong, J., Rico, C.M., Gardea-
Torresdey, J.L., 2012. Applications of synchrotron m-XRF to study the distribution
of biologically important elements in different environmental matrices: a re-
view. Anal. Chim. Acta 755, 1e16.
Marshall, K.C., 1975. Clay mineralogy in relation to survival of soil bacteria. Annu.
Rev. Phytopathol. 13, 357e373.
Micheletti, E., Colica, G., Viti, C., Tamagnini, P., De Philippis, R., 2008. Selectivity in
the heavy metal removal by exopolysaccharide-producing cyanobacteria.
J. Appl. Microbiol. 105, 88e94.
Mishra, B., Boyanov, M., Bunker, B.A., Kelly, S.D., Kemner, K.M., Fein, J.B., 2010. High-
and low-afnity binding sites for Cd on the bacterial cell walls of Bacillus subtilis
and Shewanella oneidensis. Geochim. Cosmochim. Ac. 74, 4219e4233.
Ohnuki, T., Yoshida, T., Ozaki, T., Kozai, N., Sakamoto, F., Nankawa, T., Suzuki, Y.,
Francis, A.J., 2007. Chemical speciation and association of plutonium with
bacteria, kaolinite clay, and their mixture. Environ. Sci. Technol. 41, 3134e3139.
Ozdes, D., Duran, C., Senturk, H.B., 2011. Adsorptive removal of Cd(II) and Pb(II) ions
from aqueous solutions by using Turkish illitic clay. J. Environ. Manage. 92,
3082e3090.
Priester, J.H., Olson, S.G., Webb, S.M., Neu, M.P., Hersman, L.E., Holden, P.A., 2006.
Enhanced exopolymer production and chromium stabilization in Pseudomonas
putida unsaturated biolms. App. Environ. Microb. 72, 1988e1996.
Rong, X., Huang, Q., Chen, W., 2007. Microcalorimetric investigation on the meta-
bolic activity of Bacillus thuringiensis as inuenced by kaolinite, montmoril-
lonite and goethite. Appl. Clay Sci. 38, 97e103.
Srivastava, P., Singh, B., Angove, M., 2005. Competitive adsorption behavior of heavy
metals on kaolinite. J. Colloid Interf. Sci. 290, 28e38.
Srivastava, V.C., Mall, I.D., Mishra, I.M., 2006. Equilibrium modelling of single and
binary adsorption of cadmium and nickel onto bagasse y ash. Chem. Eng. J. 117,
79e91.
Sun, Y., Zhang, R., Ding, C., Wang, X., Cheng, W., Chen, C., Wang, X., 2016. Adsorption
of U(VI) on sericite in the presence of Bacillus subtilis: a combined batch, EXAFS
and modeling techniques. Geochim. Cosmochim. Ac. 180, 51e65.
Wang, J., Li, Q., Li, M.M., Chen, T.H., Zhou, Y.F., Yue, Z.B., 2014. Competitive
adsorption of heavy metal by extracellular polymeric substances (EPS) extrac-
ted from sulfate reducing bacteria. Bioresour. Technol. 163, 374e376.
Wu, H., Chen, W., Rong, X., Cai, P., Dai, K., Huang, Q., 2014. Soil colloids and minerals
modulate metabolic activity of Pseudomonas putida measured using microcal-
orimetry. Geomicrobiol. J. 31, 590e596.
Yang, J., Wei, W., Pi, S., Fang, M., Li, A., Dan, W., Jie, X., 2015a. Competitive adsorption
of heavy metals by extracellular polymeric substances extracted from Klebsiella
sp. J1. Bioresour. Technol. 196, 533e539.
Yang, S., Ren, X., Zhao, G., Shi, W., Montavon, G., Grambow, B., Wang, X., 2015b.
Competitive sorption and selective sequence of Cu(II) and Ni(II) on montmo-
rillonite: batch, modeling, EPR and XAS studies. Geochim. Cosmochim. Ac. 166,
129e145.
Zhu, J., Pigna, M., Cozzolino, V., Caporale, A.G., Violante, A., 2010. Competitive
sorption of cooper(II), chromium(III) and lead(II) on ferrihydrite and twe
organomineral complexes. Geoderma 159, 409e416.
Zhu, J., Huang, Q., Pigna, M., Violante, A., 2012. Competitive sorption of Cu and Cr on
goethite and goethite-bacteria complex. Chem. Eng. J. 179, 26e32.