Environ. Sci. Technol. 2010, 44, 4472–4478
Ni(II) Sorption on Biogenic
Mn-Oxides with Varying Mn
Octahedral Layer Structure
MENGQIANG ZHU,*
MATTHEW GINDER-VOGEL,† AND
DONALD L. SPARKS
Department of Plant and Soil Sciences, Delaware
Environmental Institute, University of Delaware, 152
Townsend Hall, Newark, Delaware 19716
Received March 29, 2010. Accepted April 26, 2010.
Biogenic Mn-oxides (BioMnOx), produced by microorganisms,
possess an extraordinary ability to sequester metals. BioMnOx
are generally layered structures containing varying amounts of
Mn(III) and vacant sites in the Mn layers. However the
relationship between the varying structure of BioMnOx and
metal sorption properties remains unclear. In this study, BioMnOx
produced by Pseudomonas putida strain GB-1 was synthesized
at either pH 6, 7, or 8 in CaCl2 solution, and Ni(II) sorption
mechanisms were determined at pH 7 and at different Ni(II)
loadings, using isotherm and extended X-ray absorption fine
structure (EXAFS) spectroscopic analyses. Our data demonstrate
that Ni(II) sorbs at vacant sites in the interlayer of the
BioMnOx and the maximum Ni(II) sorption capacity increases
as the formation pH of BioMnOx decreases. This relation
indicates that the quantity of BioMnOx vacant sites increases
as formation conditions become more acidic, which is in
good agreement with our companion study. Contents of the
vacant sites were quantitatively estimated based on maximum
Ni(II) sorption capacity. Additionally, this study reveals that
imidazole groups are involved in Ni(II) binding to biomaterials,
and have a higher Ni(II) sorption affinity, but a lower site
density compared to carboxyl groups.
determines the maximum Pb(II) sorption capacity of synthetic birnessite (12). In addition, layer Mn(III) that surrounds
vacant sites in birnessite may affect heavy metal (e.g., Zn(II)
and Pb(II)) sorption geometry for better charge compensation
(8).
As a biogeochemically relevant metal, Ni(II) is often
associated with birnessite in the environment and has been
found incorporated into Mn octahedral layers (2, 3). It also
adsorbs at vacant sites via triple-corner sharing (TC) and at
edge sites via tridentate-edge sharing (TE) and double-corner
sharing (DC) (2, 13). The distribution of each Ni sorption
species depends on the presence of vacant sites and
experimental conditions. High pH favors Ni(II) structural
incorporation whereas low pH favors Ni(II) TC adsorption
at vacant sites (2, 10, 13). Ni(II) TE sorption only occurs at
the edge sites of triclinic birnessite which does not contain
vacant sites (10); whereas, Ni(II) DC sorption is also detected
at the edge sites of δ-MnO2 at high Ni(II) loadings (2).
Biomaterials, including microbial cells and/or extracellular
polymeric substances (EPS), have abundant anionic functional groups, such as carboxyl, phosphoryl, sulfuryl, hydroxyl, and amine functional groups (14). Extended X-ray
absorption fine structure (EXAFS) spectroscopy has demonstrated that carboxyl and phosphoryl groups are generally
responsible for heavy metal binding on biomaterials (15–17)
although other functional groups occasionally contribute (18).
Heavy metal binding to specific functional groups depends
on the type of microorganism, metal loadings, pH, and metals
(15, 16). To the authors’ knowledge, spectroscopic data
describing Ni(II) sorption on bacterial biomaterials has not
been reported.
Since Ni(II) is commonly associated with naturally occurring birnessite, we chose Ni(II) to investigate sorption on
BioMnOx produced by Pseudomonas putida strain GB-1 at
pH 6, 7, or 8 in CaCl2 solutions using sorption isotherms and
Ni K-edge EXAFS spectroscopy. The three BioMnOx contain
varying amounts of vacant sites and concentrations of layer
Mn(III) according to their structural characterization (6). This
study provides direct spectroscopic analyses to describe the
correlation of Ni(II) sorption to the number of vacant sites
and layer Mn(III) concentration. Additionally, we describe
the types of organic functional groups involved in Ni(II)
binding to biomaterials.
Introduction
Birnessite is the predominant naturally occurring Mn-oxide
in most environmental settings (1). Birnessite in soils and
sediments is often enriched with heavy metals and alkaline
and alkali earth metals (1–3). Naturally occurring birnessite
minerals have been believed to result from microbial Mn(II)
oxidation which produces biogenic Mn-oxide (BioMnOx) (4).
Naturally occurring birnessite minerals tend to be hexagonal; however, they are not identical and differ in Mn
octahedral layer structure, that is, varying amounts of Mn(III)
and quantity of vacant sites in Mn layers (1, 5). Coexisting
cations (i.e., H+, Na+, Ca2+, and Ni(II)) present during bacterial
Mn(II) oxidation affect the Mn octahedral layer structure of
BioMnOx (6, 7). Due to positive charge deficiency, vacant
sites are generally the most common sites on birnessite for
heavy metal sorption; however, in the case of Ni(II) edge
sites may play an important role in sorption (2, 8–11). Wet
chemical studies suggest that the quantity of vacant sites
* Corresponding author phone: (302) 831-1230; fax: (302) 8310605; e-mail: mzhu@udel.edu.
†
Current address: Calera Corporation, 14600 Winchester Blvd.,
Los Gatos, CA 95030.
4472
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 12, 2010
Material and Methods
Macro-Scale Experiments. Biosorbent Preparation. Biosorbents refer to the solid precipitates at the conclusion of
bacterial Mn(II) oxidation, containing both BioMnOx and
biomaterials (bacterial cells + EPS). Biosorbents were
prepared by resuspending harvested and washed P. putida
GB-1 biomaterial in 50 µM MnSO4 solutions containing 16.67
mM CaCl2 at pH 6, 7, and 8, respectively (6). Biosorbents
were then centrifuged at 3000g to remove the majority of the
EPS (6). The biosorbents are denoted as bs6, bs7, and bs8,
BioMnOx in the biosorbents denoted as BMnO6, BMnO7,
and BMnO8, and biomaterial in the biosorbents denoted as
biom6, biom7 and biom8 for pH 6, 7, and 8, respectively.
Prior to Ni(II) sorption, the biosorbents were washed four
times in 10 mM HEPES buffer (pH 7) to remove residual
ions. Dissolved Ca2+ in the biosorbent stock suspensions
was analyzed by inductively coupled plasma atomic emission
spectrometry (ICP-AES) and was less than 0.05 mM. The
BioMnOx concentration in the biosorbent stock suspensions
was colorimetrically measured after dissolving Mn-oxides
with NH2OH · HCl (19). Harvested biomaterial, without Mn(II)
addition, was also used to study Ni(II) sorption. EPS was not
10.1021/es9037066
 2010 American Chemical Society
Published on Web 05/14/2010
FIGURE 1. Sorption isotherms (Figure 1a) and linear Langmuir
plots (Figure 1b) of Ni(II) sorption at pH 7 in 50 mM NaNO3
solution on biosorbents preformed at pH 6 (bs6), 7 (bs7), and 8
(bs8), and dissolved Ca2+ and Mn(II) at sorption equilibrium
(Figure 1c). The sorption densities are normalized by dry-weight
concentration of the biosorbents. The samples (a, b, c) circled
in Figure 1a are subject to XAFS and XRD analyses.
separated from the bacterial cells prior to sorption experiments. The dry-weight concentrations of the biosorbent and
biomaterial stock suspensions were measured and used to
normalize Ni(II) sorption capacities.
Ni(II) Sorption Isotherms. Ni(II) sorption isotherms were
determined at pH 7 with 10 mL total reaction volume in 15
mL polypropylene centrifuge tubes. Biosorbent stock suspensions were added to achieve final [Mntot] as 0.5 mM; the
biosorbent dry-weight concentrations were 1.20, 1.33, and
1.55 g L-1, respectively, for bs6, bs7, and bs8. Sodium azide
(1.5 mM) was added to inhibit potential physiological and
oxidizing activities of P. putida GB-1 (20). The background
electrolyte contained 50 mM NaNO3 and 10 mM HEPES. The
buffer was used to prevent possible structural modifications
of BioMnOx during pH adjustment (6, 21). An aliquot of
Ni(NO3)2 stock solution (9.48 mM) was added to achieve initial
Ni(II) concentrations ranging from 0.048 to 0.479 mM. For
samples with high Ni(II) initial concentrations, the Ni(II)
stock solution was added incrementally to prevent possible
bulk precipitation. All experiments were performed in
triplicate. The suspensions were shaken at 120 rpm on an
orbital shaker at room temperature (∼23 °C) for 36 h. The
pH values were adjusted twice using 1 M NaOH prior to
stabilization. At the end of the reaction, final pH values
were measured to be 7.05 ( 0.04. The same procedure
used in the Ni(II) sorption on biosorbents studies was used
for Ni(II) sorption on the biomaterial study which is
denoted as Ni-biom.
The Ni(II)-sorbed biosorbents and biomaterial were
collected by centrifugation and stored at -80 °C prior to
further analyses. The supernatants were filtered with 0.22
µm cellulose filters, and the filtrates acidified with 5% HNO3
for Ni(II) and Ca2+ analyses using ICP-AES. Dissolved Mn(II)
and Ni(II) from experiments with low initial Ni(II) were
remeasured by inductively coupled plasma mass spectrometry (ICP-MS).
Molecular-Scale Measurements. Ni(II) Sorption Reference
Materials. A Ni(II) sorption standard (NiTr) on triclinic
birnessite (TcBir), that is, Na-birnessite, was made by reacting
1 mM Ni(II) with 1 g/L TcBir in 0.1 M NaNO3 at pH 7. A Ni(II)
sorption standard on δ-MnO2 (NiMnO2) was prepared by
reacting 1 mM Ni(II) with 2 g/L δ-MnO2 in 0.1 M NaNO3 at
pH 4. Synthesis of TcBir and δ-MnO2 were previously
described in Zhu et al. (6). EXAFS spectra of a marine nodule
(FeMn5DSRB) with Ni(II) incorporated into the Mn octahedral layers was provided by Prof. Sherman at the University
of Bristol, UK (3). A Ni(II)-acetate (1:4) solution was prepared
by mixing 0.2 M Ni(NO3)2 with 0.8 M NaAc at pH 6. Two
Ni(II)-phytate (1:2 and 1:4) solutions were made by adding
0.1 M Ni(NO3)2 to 0.2 and 0.4 M sodium phytate at pH 4,
respectively. Preparation of other references used in this
study, such as a Ni(II) histidine solution and a Ni(II) aqueous
solution, are described in McNear et al. (22). The Ni(II)
histidine solution was used to represent Ni(II) bound to
imidazole groups.
Ni EXAFS Spectroscopy. Ni K-edge EXAFS spectra were
collected from Ni(II) sorption samples in fluorescence mode
with a Lytle detector on beamline X11A at the National
Synchrotron Light Source (NSLS), Upton, NY. The beamline
was equipped with a Si (111) monochoromator. EXAFS
spectra of Ni(II) sorption standards were collected in
fluorescence mode with a 30-element Ge detector on beam
line 11-2 equipped with a Si (220) monochoromator, at the
Stanford Synchrotron Radiation Lightsource (SSRL), Menlo
Park, CA. EXAFS spectra from Ni(II) organic standards were
collected in transmission mode at the NSLS. The samples
were transferred to plastic holders sealed with Mylar films
on both sides for the measurement. EXAFS spectra of a Ni(II)
metal foil (E0 ) 8333.0 eV) were collected concurrently with
samples for internal energy calibration. A 3 µm cobalt filter
was placed between the Lytle detector and samples to reduce
elastic and Compton scattering by samples. EXAFS spectra
reduction, shell-by-shell fitting and linear combination fitting
(LCF) were performed with SIXPack (23). Theoretical phase
shift and amplitude files for EXAFS shell-by-shell fitting were
created by FEFF 8.02 from an atomic cluster with a hydrated
Ni(II) octahedral complex adsorbed at a birnessite vacant
site (10). The amplitude reduction factor (S02), 0.94, was
derived by fitting the first oxygen shell (CN ) 6, fixed) of a
Ni(NO3)2 solution and was used for all of the other shellby-shell fitting.
X-ray Diffraction. Synchrotron-based wide-angle X-ray
diffraction (XRD) patterns were collected from unreacted
VOL. 44, NO. 12, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4473
TABLE 1. EXAFS Shell-by-Shell Fitting of Ni(II) Sorption Samples and Standardsa
R
Red χ2
shell
CN
distance(Å)
σ2(Å2)
∆E(eV)
0.0088
66.34
Ni-O
Ni-Mn
Ni-Mn
6.7(0.7)
2b
6b
2.05(0.01)
3.08(0.02)
3.48(0.02)
0.007(0.001)
0.006(0.003)
0.010(0.003)
-1.99(1.34)
-1.99(1.34)
-1.99(1.34)
0.0104
77.88
Ni-O
Ni-Mn
Ni-Mn
6.6(0.7)
2b
2b
2.05(0.01)
3.06(0.02)
3.48(0.02)
0.007(0.001)
0.000(0.005)
0.004(0.001)
-2.01(1.44)
-2.01(1.44)
-2.01(1.44)
NiMnO2
0.0194
28.51
Ni-O
Ni-Mn
Ni-Mn
5.9(0.5)
7.0(1.0)
8.8(5.4)
2.05(0.01)
3.49(0.01)
5.45(0.02)
0.006(0.001)
0.009(0.001)
0.010(0.001)
3.32(0.87)
3.32(0.87)
3.32(0.87)
bs6a
0.0163
25.69
Ni-O
Ni-Mn
5.6(0.7)
4.6(2.0)
2.08(0.01)
3.51(0.02)
0.007(0.001)
0.012(0.004)
5.56(1.38)
5.56(1.38)
bs6b
0.0076
18.05
Ni-O
Ni-Mn
5.9(0.5)
4.3(1.4)
2.07(0.01)
3.50(0.01)
0.007(0.001)
0.011(0.003)
4.14(0.97)
4.14(0.97)
bs6c
0.0088
35.88
Ni-O
Ni-Mn
6.1(0.5)
3.2(1.2)
2.07(0.01)
3.50(0.01)
0.007(0.001)
0.009(0.003)
3.72(1.09)
3.72(1.09)
bs7a
0.0175
21.65
Ni-O
Ni-Mn
5.7(0.7)
3.9(2.1)
2.06(0.01)
3.49(0.02)
0.007(0.001)
0.012(0.005)
2.71(1.52)
2.71(1.52)
bs7b
0.0112
45.86
Ni-O
Ni-Mn
5.8(0.6)
3.3(1.3)
2.06(0.01)
3.50(0.01)
0.007(0.001)
0.010(0.003)
3.72(1.19)
3.72(1.19)
bs7c
0.0128
54.42
Ni-O
Ni-Mn
5.7(0.6)
3.2(1.8)
2.06(0.01)
3.50(0.02)
0.006(0.001)
0.011(0.005)
2.89(1.35)
2.89(1.35)
bs8a
0.0097
15.12
Ni-O
Ni-Mn
5.6(0.5)
5.1(1.3)
2.07(0.007)
3.51(0.01)
0.006(0.001)
0.009(0.002)
4.59(1.06)
4.59(1.06)
bs8b
0.0149
154.14
Ni-O
c
Ni-O
Ni-Mn
4.3(1.5)
2.4(1.0)
3.2(2.3)
2.04(0.03)
2.17(0.05)
3.52(0.03)
0.002(0.003)
0.002(0.003)
0.010(0.006)
5.40(2.76)
5.40(2.76)
5.40(2.76)
bs8c
0.0084
11.59
Ni-O
Ni-Mn
6.0(0.5)
3.3(1.6)
2.07(0.01)
3.50(0.02)
0.007(0.001)
0.012(0.004)
3.71(1.04)
3.71(1.04)
samples
NiTr
a
Errors are in parentheses. Only O and Mn shells are considered in the fitting without including scattering paths from
organic functional groups. The denotations of bs6a, bs6b, bs6c represent the three samples in pH 6 series (bs6), labeled in
Figure 1a, for XAFS analyses. Notations for bs7 and bs8 are in the same way as bs6. These notations are applicable in
other Tables and Figures. b Fixed during the fitting. c An additional O shell must be used to get a good fitting.
and Ni(II)-sorbed wet biosorbents in transmission mode with
an image plate on beam line 11-3 at SSRL. This method is
described in detail in Zhu et al. (6).
Results
Sorption Isotherms and Dissolved Mn(II)/Ca2+. The maximum Ni(II) sorption capacity of the biosorbents decreases
as biosorbent formation pH increases (Figure 1a). Negligible
amounts of dissolved Mn(II) are observed in the system,
except for the first two sorption data points on bs7 which is
likely due to contamination during the Ni sorption experiment (Figure 1c). However, upon formation, the biosorbents
do contain Mn(II), which decreases as the BioMnOx formation
pH increases (6). The reason for the absence of Mn(II) is not
clear. It may be due to the washing procedure which could
desorb Mn(II) from the biomaterials and/or BioMnOx (11).
The dissolved Ca2+ concentration increases as biosorbent
formation pH increases (Figure 1c), implying that the
biosorbents formed at a higher pH retain more Ca2+ during
their formation. Increased Ni(II) loadings also increase the
dissolved Ca2+ concentration, particularly for bs8. Since the
dissolved Ca2+ concentration is high, even at low Ni(II)
loadings, the majority of the dissolved Ca2+ is not caused by
Ni(II) sorption but by cation exchange with Na+ (∼60 mM)
in the background electrolyte.
4474
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 12, 2010
X-ray Diffraction of Biosorbents. The presence of single
(31/02) reflections (Supporting Information S1) demonstrate that all of the reacted BioMnOx are of hexagonal
layer symmetry (5, 24). The washing procedure changes
the layer structure of the unreacted BMnO8 from orthogonal layer symmetry (splitting of 31/02 reflection) to
hexagonal layer symmetry (Supporting Information S1).
This change results from layer Mn(III) migration out of
the Mn octahedral layers and is likely caused by H+/Na+
exchange with Ca2+ during the washing steps (6, 21). Na+
may not be able to prevent migration of layer Mn(III) as
well as Ca2+ due to its weaker sorption affinity (6). The
(31/02) and (20/11) peaks broaden for the unreacted
BioMnOx as the BioMnOx formation pH increases from 6
to 8 (Supporting Information S1), suggesting different layer
structure, as characterized in the companion paper (6).
Ni EXAFS Spectroscopy. Two Mn shells at Ni-Mn
distances of ∼3.08 and ∼3.48 Å are derived from NiTr spectra
fitting (Table 1 and Figure 2), which are attributed to Ni(II)
sorption (∼3.08 Å) at edge sites (TE) (10) and Ni(II) sorption
at vacant sites (TC) or at edge sites (DC) (∼3.48 Å). The
contributions of each of these shells are 0.53 ( 0.21 for TC
and 0.47 ( 0.21 for TE and were determined by fixing the Mn
coordination numbers (CNs) at 6 and 2, respectively (Table
1). When the Mn CNs are fixed at 2, representing both TE
FIGURE 2. Shell-by-shell fitting of Ni(II) k-edge EXAFS spectra
and radial structural functions (RSF) over a k range of 3-11
Å-1. Dotted lines and solid lines represent data and fitting,
respectively. EXAFS of Ni-his, Ni-ace, and Ni-biom are not
fitted. EXAFS fitting results for FeMn5SRDB can be found in
Peacock and Sherman (10). The RSF peaks of Ni incorporation
in Mn layers (IN), Ni tridentate sorption at edge sites (TE), and
Ni triple-corner sorption at vacant sites (TC) are labeled.
and DC modes, the fits result in unreasonably small
Debye-Waller factors (σ2) (0.000 for TE and 0.004 for DC)
although other parameters are similar (Table 1). Therefore,
it is likely that the Mn shell at ∼3.48 Å is derived from a TC
species. Peacock and Sherman (10) found that Ni(II) TE
species exist only at edge sites of triclinic birnessite at pH 7.
The presence of Ni(II) TC species on TcBir in this study
suggests that our TcBir sample contains some vacant sites.
For δ-MnO2, the high CN, ∼7.0, of the Mn shell at ∼3.49
Å (Table 2) and, ∼8.8, of the Mn shell at ∼5.45 Å indicate Ni
TC adsorption at vacant sites rather than DC adsorption at
edge sites (2). This is further confirmed by a two Mn-shell
fitting of NiMnO2 spectra with CNs fixed at 2 and 6,
representing DC and TC species (data not shown), and
resulting in an unreasonable contribution factor (1.35) for
TC. A higher Ni(II) loading on δ-MnO2 (10 mM Ni(II) with
2 g/L δ-MnO2) decreases the second shell amplitude in R
space (data not shown), likely due to the Ni(II) DC sorption
at edge sites (2).
EXAFS fitting reveals a single Ni-Mn distance at 3.49 3.52 Å for Ni(II) sorption on the biosorbents (Table 1). Other
atomic shells, such as carbon, nitrogen, etc., from biomaterials are not considered in the fitting. Inclusion of additional
Mn shells in the fitting does not improve fitting results,
indicating that Ni is neither present as TE sorption complexes
nor incorporated into vacant sites of BioMnOx. The relatively
high CNs of the Mn shell indicate TC rather than DC
adsorption (Table 1). The results also demonstrate that the
CNs of the Mn shell decrease with Ni(II) loadings (Table 1).
LCF Analysis of Ni(II)-biomaterial. The EXAFS of the Nibiom was fit using linear combinations of Ni(II) reference
compounds (Table 2 and Figure 3). Fitting the data with
Ni-acetate and Ni-phytate (1:2 or 1:4) does not result in a
satisfactory representation of the Ni-biom EXAFS spectra
(Table 2 and Figure 3). Addition of Ni-his to the LCF analysis
(Table 2 and Figure 3) improves the fitting accuracy,
particularly in the k-range of 3.0-6.5 Å-1, and reduces Red
χ2 from 0.162 to 0.099, whereas the inclusion of Ni-glycine
leads to a Red χ2 of 0.138 (not shown). Fitting of the spectra
without Ni-phytate (i.e., with Ni-ace and Ni-his only) does
not reduce the fitting accuracy (Table 2 and Figure 3),
indicating the phosphoryl group is not a necessary functional
group to describe Ni binding in this biomaterial. Adding Niaqueous to the standards used in fitting does not further
improve the accuracy of the LCF analysis (Table 2), indicating
that Ni(II) is bound to the biomaterial in an inner-sphere
fashion. The best fit of the unknown spectra, with the least
number of components, is obtained from a combination of
Ni-ace (f ) 0.635) and Ni-his (f ) 0.379) only (Table 2 and
Figure 3).
Both histidine and glycine are amino acids containing
the NH2-CHR-CO2 molecular fragment. However, histidine
TABLE 2. EXAFS Linear Combination Fitting of Ni(II) Sorbed on the Biomaterial (Ni-biom) and Biosorbents (from bs6a to bs8c)
with Sorption Standards: Ni(II) Sorbed on δ-MnO2 (NiMnO2), Ni(II) Acetate Solution (Ni-ace), Ni(II) Histidine Solution (Ni-his),
Ni(NO3)2 Solution (Ni-aqu) and Ni Phytate (Ni-phy) Solution. Errors Are Shown in Parentheses
Ni-aqu
Ni-ace
Ni-phy(1:2)a
Ni-his
0.984 (0.011)
0.300 (0.100)
0.635 (0.028)
Ni-biom
0.074 (0.037)
0.030 (0.037)
Ni-his
bs6a
bs6b
bs6c
bs7a
bs7b
bs7c
bs8a
bs8b
bs8c
0.448
0.259
0.261
0.298
0.265
0.193
0.350
0.153
0.250
(0.044)
(0.034)
(0.026)
(0.038)
(0.028)
(0.029)
(0.026)
(0.030)
(0.031)
0.448 (0.082)
0.575 (0.041)
0.417 (0.060)
Ni-ace
0.134
0.326
0.378
0.326
0.359
0.508
(0.038)
(0.029)
(0.023)
(0.032)
(0.024)
(0.025)
0.503 (0.025)
0.508 (0.027)
0.949 (0.010)
0.661 (0.099)
0.379
0.306
0.339
0.354
0.323
0.675 (0.033)
0.216 (0.089)
0.227 (0.065)
Ni-biomb
0.582
0.585
0.639
0.624
0.624
0.701
0.350
0.656
0.758
(0.082)
(0.063)
(0.049)
(0.070)
(0.052)
(0.054)
(0.026)
(0.055)
(0.058)
(0.030)
(0.036)
(0.033)
(0.032)
(0.319)
NiMnO2
0.376
0.409
0.369
0.339
0.366
0.297
0.626
0.333
0.266
a
Ni-phytate (1:4) is not as good as Ni-phytate (1:2) in the LCF and is not used.
represent total Ni(II) bound to the biomaterials in the biosorbents.
b
(0.016)
(0.013)
(0.010)
(0.014)
(0.010)
(0.011)
(0.019)
(0.011)
(0.012)
Red χ2
sum
0.206
0.170
0.162
0.102
0.117
0.099
0.100
0.093
0.984
0.949
0.961
1.01
0.981
1.00
1.00
0.996
Red χ2
sum
0.180
0.109
0.064
0.132
0.075
0.080
0.261
0.081
0.092
0.958
0.993
1.01
0.963
0.990
0.998
0.977
0.990
1.02
Sum of Ni-ace and Ni-his fractions to
VOL. 44, NO. 12, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4475
FIGURE 3. Linear combination fitting of Ni-biom and
Ni-biosorbent EXAFS spectra. Ni-biom is fitted with Ni(II)
organic complexes: a, acetate + phytate; b, acetate +
histidine; c, phytate + histidine; d, acetate + phytate +
histidine; e, acetate + phytate + histidine + aqueous.
Ni-biosorbents are fitted with Ni-his, Ni-ace and NiMnO2. The
labels correspond to Table 2. The experimental data of the
Ni-biom were plotted in red for a better comparison with fitting
results.
has much better fitting accuracy than glycine in the LCF
analysis of Ni(II) sorption on the biomaterial. The probable
reason is that histidine possesses an imidazole group which
has high complexation affinity to heavy metals (26, 27). This
suggests that the imidazole group is involved in the Ni(II)
binding to the biomaterial. In this study, Ni(II) can bind to
cell surfaces, EPS and/or even inside of the cells because P.
Putida can uptake Ni(II) (28).
LCF Analysis of Ni(II)-Biosorbents. LCF analysis using Niace, Ni-his and NiMnO2 were performed on Ni(II) sorption
samples to evaluate the relative contributions of Ni(II) bound
to carboxyl groups, imidazole groups, and vacant sites,
respectively (Table 2 and Figure 3). Inclusion of Ni-aqu, Niphy, NiTr, and FeMn5SRDB does not improve the fitting
accuracies. Good LCF fitting accuracies (Table 2 and Figure
3) with NiMnO2 further confirms that Ni(II) sorption occurs
at vacant sites of BioMnOx. Our study does not reveal Ni(II)
DC sorption at edge sites of BioMnOx, even at relatively high
Ni(II) loadings.
The Ni(II) sorption loadings on BioMnOx, histidine, and
carboxyl groups are individually calculated based on LCF
fractions (Table 2) and are nonlinear square fitted with the
Langmuir adsorption equation (Figure 4 and Table 3). The
maximum Ni(II) sorption capacity (qmax) of BioMnOx decreases as the BioMnOx formation pH increases, but the
sorption affinity (k) dramatically increases on the BioMnOx
formed at pH 8 (Table 3). Compared to Ni-his, Ni-ace exhibits
a higher maximum sorption capacity, that is, a higher site
density of the carboxyl group, but a lower Ni(II) sorption
affinity (Table 3). As Ni(II) loading increases, Ni sorption
densities on biomaterials continue to increase; however, Ni
sorption densities on BioMnOx become constant for BMnO7
and BMnO8 (Figure 4a), suggesting sorption saturation. This
results in smaller molar fractions of Ni bound to BioMnOx
and subsequently causes the obtained CNs of the Mn-shell
in the EXAFS shell-by-shell fitting to decrease as Ni loadings
increase (Table 1).
Mass Balance of Ca2+ and Ni(II). Since Ni(II) sorption on
BioMnOx reaches a plateau on BMnO8 (Figure 4a), the
increase in dissolved Ca2+ concentration is likely caused by
Ni(II) sorption on the biomaterial. Ca2+ mass balance
calculations demonstrate that the increased [Ca2+] of 0.047
(from 0.218 to 0.265) mol kg-1 biosorbent (Figure 1c) is close
to the corresponding increase of 0.050 (from 0.098 to 0.148)
mol Ni kg-1 biomaterial on the biom8 (Figure 4a). Using the
same argument, it can be concluded from Figure 4b that
most of the Ca2+ replaced by Ni sorption is initially bound
to carboxyl sites rather than histidine, consistent with the
view that the tightly bound Ca2+ can be attributed to strong
binding to carboxyl groups (29). At low Ni loadings, the slight
increase in [Ca2+], 0.015 (from 0.203 to 0.218) mol kg-1
biosorbent (Figure 1c) only accounts for a small portion of
the corresponding Ni(II) sorption on bs8 (0.093 from 0.060
to 0.153 mol kg-1 biosorbent) (Figure 1a), which also occurs
for all data points for bs6 and bs7. This indicates that most
of the Ni(II) initially sorbs onto sites not occupied by Ca2+
before the BioMnOx is saturated by Ni(II). Consequently,
Ca2+ does not substantially compete with Ni(II) for sorption
on BioMnOx under the conditions of this study.
Discussion
In addition to BioMnOx, the biosorbents also contain bacterial
cells and EPS. Nevertheless, these biomaterials are unlikely
to alter the intrinsic reactivity of the negatively charged
BioMnOx (30). The data also show that neither Ca2+ nor
interlayer Mn are in substantial competition with Ni(II) for
sorption on BioMnOx. Thus, the observed Ni(II) sorption
FIGURE 4. Ni(II) sorption isotherms on BioMnOx (BMnO6, 7 and 8) and biomaterials (biom 6, 7, and 8) (a), and histidine and acetate
(b). The sorption densities are calculated based on component molar fractions of EXAFS linear combination fittings.
4476
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 12, 2010
TABLE 3. Langmuir Parameters Obtained from Least Square Fittings for Ni(II) Sorption Isotherms for Each Sorbent Component
qmaxa
kb
R2
c
qmaxa
Ni-biomaterial
bs6
bs7
bs8
0.183
0.185
0.188
36.0
27.6
20.7
a
0.066
0.049
0.040
103.2
86.4
123.4
Maximum sorption loading.
b
0.993
0.995
0.999
c
fvacd
0.232
0.203
0.159
94.5
56.0
618.0
0.841
0.953
0.985
0.104
0.092
0.074
Ni-ace
0.954
0.897
0.696
0.125
0.164
0.130
Sorption affinity parameter. c Fitting goodness.
differences among the three BioMnOx are derived from their
intrinsic structural characteristics, that is, vacant site and
layer Mn(III) content (6). Since Ni(II) sorption on BioMnOx
occurs at vacant sites only, the maximum Ni(II) sorption
capacities are apparently determined by the vacant site
quantity (12). The results also suggest that vacant site
occupancies of BioMnOx increase as the formation conditions
become more acidic, consistent with our structural study
(6).
Estimation of Quantities of Vacant Sites. Assuming both
sides of all vacant sites are occupied by Ni(II) at maximum
sorption loadings (8), the molar fractions of vacant sites are
estimated based on the following equations (Table 3),
fvac )
R2
NiMnO2
Ni-his
bs6
bs7
bs8
kb
qmax / 2
qmax / 2 + 1
in which qmax is the maximum sorption loading. All calculated
fvac lie within the range of the two end members, ideal
hexagonal (fvac ) 0.167) and triclinic birnessite (fvac ) 0) (21).
As BioMnOx formation pH increases, the vacant site fraction
decreases (Table 3), consistent with the structural characterization of BioMnOx (6). The vacant site content in BioMnOx
has been implied based on Zn(II) sorption loadings by a
similar methodology (9). This method directly measures
contents of vacant sites and is therefore not impaired by
particle size effects as compared with EXAFS multiplescattering fitting (24). However, this assumes that metals are
sorbed only at vacant sites rather than other sites (2). Zn(II)
may be a better candidate than Ni(II) because Zn has not
been found to adsorb at edge sites. Additionally, metal
sorption or experimental conditions should not create new
vacant sites.
Mn(III) Roles. Empirical bond valence calculations
showed that Mn(III)-coordinated oxygen atoms encompassing a vacant site are more undersaturated than Mn(IV)coordinated oxygen atoms (8). Consequently, vacant sites
surrounded by a mixture of Mn(III) and Mn(IV) could exhibit
stronger adsorption affinities for interlayer metal cations than
vacant sites surrounded only by Mn(IV). The dramatic change
in Ni(II) sorption affinities of the BioMnOx formed at pH 8
could be due to the high fraction of layer Mn(III), as found
in the companion paper. The similarity of Ni(II) sorption
affinity on the BMnO7 to BMnO6 is probably due to the
presence of dissolved Mn(II), which interferes with Ni(II),
thereby decreasing apparent Ni(II) sorption affinity on
BMnO7.
Ni Distribution Between Mn-oxides and Biomaterials.
A large fraction of Ni(II) is associated with the biomaterial
before BioMnOx is saturated by Ni(II) sorption. However,
Zn(II) is not sorbed on P. putida MB-1 biomaterial until the
BioMnOx is saturated (9). In the Zn study, BioMnOx is 25
wt.% of the total biosorbents (9), whereas the BioMnOx in
the current study is only 2.8-3.6 wt.%. The high portion of
18.8
11.7
20.7
d
0.992
0.991
0.974
Molar fraction of vacant sites.
Ni(II) binding to the biomaterials may be due to the high
biomaterial content in this study. Additionally, Ni(II) has
relatively weak sorption affinity to birnessite compared to
other heavy metals, such as Pb(II), Cu(II) and Zn(II) (31), and
thus, the high Ni(II) affinity of histidine groups can comparably compete with BioMnOx for binding Ni(II). Furthermore, the different Ni(II) sorption affinity of BioMnOx derived
from a different Mn layer structure also affects heavy metal
distributions on them. The distribution of heavy metals
between the biomaterial and minerals depends on metal
sorption affinities and the relative amounts of each component.
Environmental Implications. Mn-oxides are major metal
sequesters in many geochemical environments, such as the
ocean floor, and can affect trace metal cycling. As a dominant
environmental Mn-oxide phase, layer Mn-oxides may have
a wide range of metal sorption properties, depending on
their formation conditions, and can impose different impacts
on their surroundings. For instance, Mn-oxides are not stable
when redox conditions change in the environment. Once
Mn-oxides are reduced and dissolved, the Mn-oxides having
more vacant sites can accelerate heavy metal cycling due to
their bearing more heavy metals. Even for large tunnel
structure of Mn-oxides, such as todorokite (3 × 3 tunnel),
vacant sites/Mn(III) may play a similar role in metal sorption
because building units of the tunnel structure are also edgesharing Mn octahedral layers (1).
Acknowledgments
Mengqiang Zhu is grateful for a University of Delaware
Institute of Soil and Environmental and Quality (ISEQ)
Graduate Fellowship. The research was supported by Delaware EPSCoR with funds from the National Science Foundation Grant EPS-0447610 and the State of Delaware and USDA
Grant 2005-35107-16105. Use of the NSLS was supported by
the U.S. Department of Energy, Office of Science, Office of
Basic Energy Sciences, under Contract No. DEAC0298CH10886. The beamline X11 is supported by the Office of
Naval Research and contributions from Participating Research Team (PRT) members. Portions of this research were
carried out at the Stanford Synchrotron Radiation Lightsource, a national user facility operated by Stanford University
on behalf of the U.S. Department of Energy, Office of Basic
Energy Sciences. The SSRL Structural Molecular Biology
Program is supported by the Department of Energy, Office
of Biological and Environmental Research, and by the
National Institutes of Health, National Center for Research
Resources, Biomedical Technology Program.
Supporting Information Available
X-ray diffraction patterns of biosorbents. This material is
available free of charge via the Internet at http://pubs.acs.org.
VOL. 44, NO. 12, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4477
Literature Cited
(1) Post, J. E. Manganese oxide minerals: Crystal structures and
economic and environmental significance. Proc. Natl. Acad.
Sci. U.S.A. 1999, 96, 3447–3454.
(2) Manceau, A.; Lanson, M.; Geoffroy, N. Natural speciation of Ni,
Zn, Ba, and As in ferromanganese coatings on quartz using
X-ray fluorescence, absorption, and diffraction. Geochim. Cosmochim. Acta 2007, 71 (1), 95–128.
(3) Peacock, C. L.; Sherman, D. M. Crystal-chemistry of Ni in marine
ferromanganese crusts and nodules. Am. Mineral. 2007, 92,
1087–1092.
(4) Tebo, B. M.; Bargar, J. R.; Clement, B. G.; Dick, G. J.; Murray,
K. J.; Parker, D.; Verity, R.; Webb, S. M. Biogenic manganese
oxides: Properties and mechanisms of formation. Annu. Rev.
Earth. Planet Sci. 2004, 32, 287–328.
(5) Drits, V. A.; Lanson, B.; Gaillot, A.-C. Birnessite polytype
systematics and identification by powder X-ray diffraction. Am.
Mineral. 2007, 92, 771–788.
(6) Zhu, M.; Ginder-Vogel, M.; Parikh, S.; Feng, X.; Sparks, D. L.,
Cation effects on the layer structure of biogenic Mn-oxides.
Environ. Sci. Technol. 2010, DOI 10/1021/es1009955.
(7) Webb, S. M.; Tebo, B. M.; Bargar, J. R. Structural influences of
sodium and calcium ions on the biogenic manganese oxides
produced by the marine Bacillus sp. strain SG-1. Geomicrobiol.
J. 2005, 22, 181–193.
(8) Manceau, A.; Lanson, B.; Drits, V. A., III. Results from powder
and polarized extended X-ray absorption fine structure spectroscopy. Geochim. Cosmoichim. Acta 2002, 66 (15), 2639–2663.
(9) Toner, B.; Manceau, A.; Webb, S. M.; Sposito, G. Zinc sorption
to biogenic hexagonal-birnessite particles within a hydrated
bacterial biofilm. Geochim. Cosmochim. Acta 2006, 70, 27–43.
(10) Peacock, C. L.; Sherman, D. M. Sorption of Ni by birnessite:
equilibrium controls on Ni in seawater. Chem. Geol. 2007, 238,
94–106.
(11) Li, X.; Pan, G.; Qin, Y.; Hu, T.; Wu, Z.; Xie, Y. EXAFS studies on
adsorption-desorption reversibility at manganese oxide-water
interfaces: II. Reversible adsorption of zinc on δ-MnO2. J. Colloid
Interface Sci. 2004, 271 (1), 35–40.
(12) Zhao, W.; Cui, H. J.; Liu, F.; Tan, W. F.; Feng, X. H. The relation
between the amount of Pb(II) adsorbed by birnessites and their
average oxidation state of Mn. Clays Clay Miner. 2009, 57 (5),
513-520.
(13) Peacock, C. L. Physiochemical controls on the crystal-chemistry
of Ni in birnessite: Genetic implications for ferromanganese
precipitates. Geochim. Cosmochim. Acta 2009, 73 (12), 3568–
3578.
(14) Fein, J. B.; Martin, A. M.; Wightman, P. G. Metal adsorption
onto bacterial surfaces: Development of a predictive approach.
Geochim. Cosmochim. Acta 2001, 65 (23), 4267–4273.
(15) Sarret, G.; Manceau, A.; Spadini, L.; Roux, J.-C.; Hazemann,
J.-L.; Soldo, Y.; Eybert-Berard, L.; Menthonnex, J.-J. Structural
determination of Zn and Pb binding sites in Penicillium
chrysogenum cell walls by EXAFS spectroscopy. Environ. Sci.
Technol. 1998, 32 (11), 1648–1655.
4478
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 12, 2010
(16) Boyanov, M. I.; Kelly, S. D.; Kemner, K. M.; Bunker, B. A.; Fein,
J. B.; Fowle, D. A. Adsorption of cadmium to Bacillus subtilis
bacterial cell walls: A pH-dependent X-ray absorption fine
structure spectroscopy study. Geochim. Cosmochim. Acta 2003,
67 (18), 3299–3311.
(17) Toner, B.; Manceau, A.; Marcus, M. A.; Millet, D. B.; Sposito, G.
Zinc sorption by a bacterial biofilm. Environ. Sci. Technol. 2005,
39 (21), 8288–8294.
(18) Guine, V.; Spadini, L.; Sarret, G.; Muris, M.; Delolme, C.; Gaudet,
J. P.; Martins, J. M. F. Zinc sorption to three gram-negative
bacteria: Combined titration, modeling, and EXAFS study.
Environ. Sci. Technol. 2006, 40 (6), 1806–1813.
(19) Burle, E.; Kirby-Smith, W. W. Application of formaldoxime
colorimetric method for the determination of manganese in
the pore water of anoxic estuarine sediments. Estuaine Coasts
1979, 2 (3), 198–201.
(20) Rosson, R. A.; Tebo, B. M.; Nealson, K. H. Use of poisons in
determination of microbial manganese binding rates in seawater. Appl. Environ. Microbiol. 1984, 47 (4), 740–745.
(21) Silvester, E.; Manceau, A.; Drits, V. A. Structure of synthetic
monoclinic Na-rich birnessite and hexagonal birnessite: II.
Results from chemical studies and EXAFS spectroscopy. Am.
Mineral. 1997, 82, 962–978.
(22) McNear, D. H.; Chaney, R. L.; Sparks, D. L. The effects of soil
type and chemical treatment on nickel speciation in refinery
enriched soils: A multi-technique investigation. Geochim. Cosmochim. Acta 2007, 71 (9), 2190–2208.
(23) Webb, S. SIXPack: a graphical user interface for XAS analysis
using IFEFFIT. Phys. Scr. 2005, T115, 1011–1014.
(24) Webb, S. M.; Tebo, B. M.; Bargar, J. R. Structural characterization
of biogenic Mn oxides produced in seawater by the marine
Bacillus sp. strain SG-1. Am. Mineral. 2005, 90, 1342–1357.
(25) Yang, D. S.; Wang, M. K. Syntheses and characterization of wellcrystallized birnessite. Chem. Mater. 2001, 13, 2589–2594.
(26) Malachowski, L.; Holcombe, J. A. Immobilized poly-L-histidine
for chelation of metal cations and metal oxyanions. Anal. Chim.
Acta 2003, 495 (1-2), 151–163.
(27) Kramer, U.; Cotter-Howells, J. D.; Charnock, J. M.; Baker, A. J. M.;
Smith, J. A. C. Free histidine as a metal chelator in plants that
accumulate nickel. Nature 1996, 379 (6566), 635–638.
(28) Tripathi, V. N.; Srivastava, S. Ni2+-uptake in Pseudomonas putida
strain S4: A possible role of Mg2+-uptake pump. J. Bioscience
2006, 31 (1), 61–67.
(29) Sobecka, D. C.; Higginsb, M. J. Examination of three theories
for mechanisms of cation-induced bioflocculation. Water Res.
2002, 36, 527–538.
(30) Templeton, A. S.; Trainor, T. P.; Traina, S. L.; Spormann, A. M.;
Brown, G. E. Pb(II) distributions at biofilm-metal oxide interfaces. Proc. Natl. Acad. Sci. U.S.A. 2001, 98 (21), 11897–11902.
(31) Makenzie, R. M. The adsorption of lead and other heavy metals
on oxides of manganese and iron. Aust. J. Soil Res. 1980, 18,
61–73.
ES9037066