Spectroscopic Investigation of Ni
Speciation in Hardened Cement
M . V E S P A , *, † , ‡R . D A ¨ H N ,†
D . G R O L I M U N D ,†E . W I E L A N D ,†A N D
A . M . S C H E I D E G G E R† , ‡
Paul Scherrer Institute, Laboratory for Waste Management,
5232 Villigen PSI, Switzerland, and Department of
Environmental Sciences, Swiss Federal Institute of Technology
(ETH), Zu ¨rich, Switzerland
Cement-based materials play an important role in multi-
barrier concepts developed worldwide for the safe disposal
of hazardous and radioactive wastes. Cement is used to
condition and stabilize the waste materials and to construct
the engineered barrier systems (container, backfill, and
liner materials) of repositories for radioactive waste. In this
study, Ni uptake by hardened cement paste has been
investigated with the aim of improving our understanding
of the immobilization process of heavy metals in cement on
the molecular level. X-ray absorption spectroscopy
(XAS) coupled with diffuse reflectance spectroscopy
(DRS) techniques were used to determine the local
were prepared at two different water/cement ratios (0.4,
1.3) and different hydration times (1 hour to 1 year) using a
sulfate-resisting Portland cement. The metal loadings
and the metal salts added to the system were varied (50
up to 5000 mg/kg; NO3-, SO42-, Cl-). The XAS study showed
that for all investigated systems Ni(II) is predominantly
immobilized in a layered double hydroxide (LDH) phase,
which was corroborated by DRS measurements. Only a
minor extent of Ni(II) precipitates as Ni-hydroxides
(R-Ni(OH)2and ?-Ni(OH)2). This finding suggests that
Ni-Al LDH, rather than Ni-hydroxides, is the solubility-
limiting phase in the Ni-doped cement system.
Assuring safe disposal and long-term storage of hazardous
and radioactive wastes represents a primary environmental
task of industrial societies. The long-term disposal of the
hazardous waste is associated with landfilling of cement-
stabilized waste (1), whereas deep geological disposal is
waste (2). For example, more than 90 wt% of the near-field
material of the planned Swiss geological repository for
(HCP) and cementitious backfill materials. The HCP is used
to solidify the radioactive waste. For this reason, an under-
cement is essential to predict the long-term fate of con-
taminants in the geosphere. From a chemical standpoint,
HCP is a very heterogeneous material with discrete particles
of mainly calcium (aluminum) silicate hydrates, portlandite
(calcium hydroxide), and calcium aluminates. The im-
mobilization potential of HCP originates from its selective
binding properties for metal cations and anions (3). Thus,
it appears that immobilization processes in cement systems
are highly specific with respect to the mineral components
and mechanisms involved.
Ni is among the most important contaminant in waste
materials resulting from a variety of industrial processes.
For example, Ni radioisotopes associated with irradiated
in cement-stabilized radioactive waste. In this case, this
information is of major importance for predicting the long-
term fate of Ni in the cementitious matrix of a disposal site.
In connection with the disposal of non-radioactive waste,
molecular level information will allow more detailed as-
sessment of the leachability of heavy metals, for example,
Ni, from landfills into the environment. Earlier experiments
on the Ni uptake by blended and Portland cement indicated
that, under highly alkaline conditions (pH > 12.5), poorly
crystalline Ni(OH)2(4) and Ni-Al layered double hydroxide
(Ni-Al LDH) phases (5, 6) may be formed. LDH phases can
where the MIIposition can be at least partially filled with Ni,
MIIIposition with Al, and the An-with different anions
such as CO32-, NO3-, Cl-, and SO42-. A natural occurring
Ni-Al LDH mineral is Takovite, also named Eardleyit
The objective of the present study was to investigate Ni
immobilization during cement hydration. The hydration
process was started by adding Ni salt solution to the
unhydrated cement. The sulfate-resisting Portland cement
as starting material for the present study. Note that the
(5, 6) where the Ni uptake by hydrated cement was
A combination of wet chemistry experiments, X-ray
absorption spectroscopy (XAS), and diffuse reflectance
spectroscopic (DRS) measurements was used to gain a
molecular level understanding of the immobilization pro-
to provide information on the solubility-limiting phase,
chemical speciation, and the structural coordination envi-
ronment of Ni in hydrated cement.
Materials and Methods
Sample Preparation. The cement samples were prepared
from a commercial sulfate-resisting Portland cement (CEM
I 52.5 N HTS, Lafarge, France) used to condition radioactive
cement. The metal salts were dissolved in deionized water
to obtain stock solutions with concentrations of 0.3, 0.03,
and 0.003 mol/L (pH ) 4.5-5). The solutions were mixed
with the unhydrated cement at two different water/cement
(w/c) ratios (0.4, 1.3) using a standard procedure (7). The
degree of hydration is enhanced with increasing w/c ratio
(8). The final metal concentrations of the pastes were 50,
500, and 5000 mg/kg dry (Table 1). The cement pastes were
placed into Plexiglas moulds, which were closed with
For short hydration times up to 6 hours, the slurries were
filtered (0.2 µm pore size) to separate the solid from the free
* Corresponding author phone: +41-56-310-2966; fax: +41-56-
310-4595; e-mail: email@example.com.
†Paul Scherrer Institute.
‡Swiss Federal Institute of Technology (ETH).
Environ. Sci. Technol. 2006, 40, 2275-2282
10.1021/es052240q CCC: $33.50
Published on Web 03/02/2006
2006 American Chemical SocietyVOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY92275
water. The solid materials were washed with acetone for 15
min to stop the hydration process (8), filtered, and dried in
a glovebox under controlled N2 atmosphere (CO2, O2 < 2
ppm, T ) 20 ( 3 °C). The samples hydrated for longer time
periods were stored in closed containers at 100% relative
to obtain size fractions <100 µm using a tungsten/carbide
X-ray absorption fine structure (EXAFS) and DRS measure-
ments as well as for wet chemistry experiments. For EXAFS
holders and sealed with Kapton tape.
Wet Chemistry Experiments. The wet chemistry experi-
for 30 days with Ni loadings of 5000 mg/kg (Ni_cem_30d)
chemistry experiments, the HCP material was crushed and
sieved to collect the size fraction of <63 µm. The material
liquid (S/L) ratio of 5 g/L and shaken end-over-end for 1, 14,
30, and 60 days in the glovebox under controlled N2
separated by centrifugation (60 min at 95 000g). Aliquots
spectroscopy (ICP-MS, detection limit ) 0.05 µg/L). Prior to
the ICP-MS measurements, the concentration of the main
elements (Na, K, Ca) was determined using inductively
take into account matrix effects.
The composition of ACW corresponds to the chemical
composition of a solution in equilibrium with a freshly
prepared HCP (8). Under these conditions, the ACW is a
(Na,K)OH rich fluid saturated with respect to portlandite
and calcite (pH ) 13.3). The basic preparation and the
chemical composition of ACW are described elsewhere (9,
the Ni K-edge were collected at the Swiss Norwegian Beam
Line (SNBL) and at the Dutch Belgium Beamline (DUBBLE)
at the European Synchrotron Radiation Facility (ESRF) in
crystal monochromator. The measurements were collected
at room temperature in transmission (ionization chambers)
and in fluorescence mode (SNBL, Lytle detector; DUBBLE,
9 channel monolithic Ge-solid-state detector). The mono-
chromator angle was calibrated by assigning the energy of
8333 eV to the first inflection point of the K-absorption edge
spectrum of Ni metal foil.
EXAFS data reduction was performed using the WinXAS
energy was converted to photoelectron wave vector units
(Å-1) by assigning the origin E0to the first inflection point
obtained by Fourier transforming the k3-weighted ?(k)
functions between 3.2 and 10.9 Å-1with a Bessel window
function with a smoothing parameter of 4. Multishell fits
shells (∆R ) 0.8-3.5 Å). Theoretical scattering paths for the
fit were calculated using FEFF 8.2 (12) and the structure of
?-Ni(OH)2 as a reference. The amplitude reduction factor
(S02) was determined to be 0.85 from the experimental
?-Ni(OH)2 EXAFS spectrum (5). Errors on the structural
parameters were estimated from the analysis of a series of
reference compounds (?-Ni(OH)2, R-Ni(OH)2; see Table 2).
Several reference spectra (?-Ni(OH)2, R-Ni(OH)2, synthetic
Ni-Al LDH (Ni:Al, 2:1; Ni2Al(OH)6(CO3)1/2(13)), Ni-phyllo-
silicate (14), neo-formed Ni-Al LDH formed from Ni-doped
pyrophyllite (15)) were used in support of the identification
of the Ni speciation in the cement matrix.
Wavelet transform analysis (WT) of EXAFS spectra was
used to complement the Fourier transform. WT enables a
2D visualization of the Fourier transform, with resolution in
both k (Å-1) and R space (Å) (16, 17). WT allows atoms to be
distinguish, which are located at the same distance R but
yield contributions at different k ranges. Recently, Mun ˜oz et
al. (17) applied a continuous Cauchy WT to simultaneously
decompose the EXAFS spectra of geochemical and envi-
the short-range structure of Zn-Al LDH. The Morlet WT
allows the optimization of the parameters η and σ. The
frequency of the sine and cosine functions, η, determines
how many oscillations of the sine wave are covered by a
Gaussian envelope with the half-width σ ) 1. Funke et al.
(16) demonstrated that the optimum resolution at a given
distance (ropt) of interest is achieved by ηopt = 2ropt for
σ ) 1.
Diffuse Reflectance Spectroscopy (DRS). DRS has been
used in the past to study Ni and other transitional metals in
catalysis (18) and in environmental science (19, 20). The
of the first coordination shell. For example, DRS is more
different Ni phases based on the energy position of the ν2
band. This band corresponds to the3A2gf3T1gtransition of
the crystal field.
a Varian Cary 5 UV-Vis-NIR spectrophotometer described
elsewhere (21). White reflectance standard BaSO4 (Kodak)
was used to record the baseline. Processing of the spectra
included subtraction of the baseline and calculation of the
TABLE 1. Experimental Parameters for the Ni-Doped HCP Samples
in solution (M)w/ca
in HCP (mg/kg)
aw/c ) water/cement ratio.
22769ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
Kubulka-Munk function. Blanks, that is, samples prepared
in the same way as the Ni-doped HCP samples but without
metal addition, were used for spectral background subtrac-
tion. The spectra of reference compounds (synthetic
were also recorded at the same time.
Results and Discussion
Wet Chemistry Data. Wet chemistry experiments were
HCP and ACW, and compare the results with earlier
first assessments of the binding mechanism of Ni in the
cement matrix to be made. In this study, the Ni partition
between the Ni-doped HCP samples with Ni loadings of
5000 mg/kg dry HCP (Ni_cem_30d) and 500 mg/kg dry HCP
(Ni_cem_500) was determined after 30 days equilibration of
time required to achieve constant Ni concentrations in
was chosen to allow direct comparison of the aqueous Ni
concentrations in these systems with the concentration
measurements reported by Wieland et al. (10).
Figure 1 shows the Ni concentration in solution upon
equilibration of the Ni_cem_30d and Ni_cem_500 sample
material in ACW together with the sorption isotherm data
reported in Wieland et al. (10). The figure reveals that the
trend to increasing aqueous Ni concentrations with increas-
source, that is, whether Ni was released from the Ni-doped
HCP or added to solution and subsequently sorbed onto
in ACW lies above the Ni concentration of pristine HCP
equilibrated in ACW and below the solubility of ?-Ni(OH)2,
It should be noted that for the experimental setup used in
the present study (adding a highly concentrated Ni solution
to the alkaline cement system) and based on the available
thermodynamic data (ref 22 and therein), ?-Ni(OH)2 was
expected to be the only Ni phase formed during cement
hydration. The good agreement of the wet chemistry data
obtained in this study and those reported by Wieland et al.
used for the sample preparation. It is worth emphasizing
study Ni was taken up by the cement matrix during the
hydration process. Wieland et al. (10) concluded that the Ni
immobilization by HCP could not be interpreted in terms of
an adsorption-type process. The authors rather suggested
that the Ni concentration in solution is controlled by a
solubility-limiting process due to the formation of a solid
phase with varying composition (solid solution). This find-
ing is further supported by the spectroscopic studies of
Scheidegger et al. (5, 6), which suggested that a Ni-Al LDH-
type phase is the solubility-limiting phase.
TABLE 2. Structural Information Obtained from EXAFS Ni K-edge Data Analysisa
Ni-Al LDH (LDH)
neo-formed Ni-Al LDHc(N-LDH)
R(Ni-O)( 0.02 Å, CN(Ni-O)( 0.02 Å, CN(Ni-Ni)( 20%. res% ) deviation between experimental data and fit given by the relative residual in percent.
bDa ¨hn et al., 2002.cScheidegger et al., 1997.dCorrelated parameters.eFixed parameters during fitting procedures.
FIGURE 1. The concentration of Ni taken up by HCP shown as a
function of the concentration of dissolved Ni in ACW (pH ) 13.3).
b, 5000 and 500 mg/kg metal loading (this study); 0, data from
Wieland et al. (10). All samples were equilibrated for 30 days. The
shaded area indicates the Ni concentration in ACW; the gray line
indicates the solubility limit of ?-Ni(OH)2.
VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY92277
Formation of a Ni-Al LDH Phase. Figure 2a shows the
spectra of HCP samples with Ni loadings of 5000 mg/kg and
spectra (Table 1). For the following discussion, samples
with hydration times longer than 30 days (Ni_cem_30d,
Ni_cem_150d, Ni_cem_1y) will be considered.
Figure 2a reveals that the first oscillation at ∼4 Å-1of the
experimental spectra is at a position similar to that of the
neo-formed LDH phase, which forms upon Ni uptake by
pyrophyllite (N-LDH) (15). The experimental spectra show
a small feature at ∼5 Å-1on the left side of the oscillation
at 6 Å-1, which is well reproduced in both synthetic and
neo-formed Ni-Al LDH spectra. The amplitude of this
spectral feature is clearly damped as compared to the Ni-
phyllosilicate spectrum. The beat pattern at ∼8 Å-1of the
Ni-doped HCP samples shows a splitting of the oscillation.
at ∼8 Å-1suggests the presence of Ni-Al LDH. In fact, this
the other reference compounds (R-Ni(OH)2, ?-Ni(OH)2, and
Ni-phyllosilicate) show an elongated upward oscillation
ending in a sharp tip at ∼8 Å-1. Thus, the presence of the
beat pattern at ∼8 Å-1together with the observed spectral
features at ∼4 and ∼5 Å-1indicate that a Ni-Al LDH phase
has formed in Ni-doped HCP samples.
The corresponding Fourier transforms (FT) of the k3-
weighted EXAFS spectra are shown in Figure 2b. The
amplitude of the first peak (R + ∆R ) ∼1.6 Å) is similar for
all samples and references. The imaginary part of the FT
spectra of the Ni-doped HCP samples at R + ∆R ) ∼2 Å
hydrated for 150 days (Ni_cem_150d) and for 1 year
(Ni_cem_1y) is slightly shifted to lower values (R + ∆R )
∼2.71 Å) as compared to the sample hydrated for 30 days
(Ni_cem_30d) (R + ∆R ) ∼2.78 Å). The latter distance is
comparable to those determined for the ?-Ni(OH)2and Ni-
phyllosilicate reference compounds (R + ∆R ) ∼2.75 Å).
However, the shorter distance corresponds to those deter-
mined for R-Ni(OH)2or Ni-Al LDH, respectively (R + ∆R )
∼2.71 Å). The amplitude of the second peak in the experi-
mental spectra of all Ni-doped HCP samples and the Ni-Al
LDH reference compounds is clearly reduced as compared
in the cement samples.
The structural parameters derived from multi-shell analy-
sis (∆R ) 0.8-3.5 Å) are summarized in Table 2. The first
The second coordination shell was fitted solely using Ni-Ni
pairs, because the discrimination of Ni-Ni and Ni-Al
backscattering pairs in Ni-Al LDH is problematic (5, 6). To
be able to compare the coordination numbers of the Ni-Ni
backscattering pairs (CNNi-Ni) of all samples and references,
for the ?-Ni(OH)2. Data analysis reveals similar CN and
interatomic distances (R) for all cement samples (Table 2).
are strongly reduced as compared to R-Ni(OH)2 and
?-Ni(OH)2. However, the CNNi-Ni of the HCP samples are
similar to that determined for Ni-Al LDH. The CNNi-Ni is
reduced as Ni is partly substituted by Al in Ni-Al LDH,
resulting in a significant destructive interference between
Ni and Al backscattering contributions and causing an
amplitude cancellation of the Ni and Al shells (6, 14). The
CNNi-Niin ?-Ni(OH)2is close to 6 as expected from literature
data (24), whereas the CNNi-Niof R-Ni(OH)2is slightly lower
(4.9). The difference in the CN is attributed to Ni vacancies
in the brucite-like Ni(OH)2layer (25).
Although the CNNi-Niof the Ni-doped HCP samples and
Ni-Al LDH agree very well, the overall Ni-Ni distances of
FIGURE 2. Experimental spectra of Ni-reference compounds and Ni-doped HCP samples with 5000 mg/kg Ni concentration and hydrated
EXAFS function for the Fourier-backtransform spectra obtained from Figure 2b (range: R + ∆R ) 0.8-7 Å). Dashed lines indicate spectral
features explained in detail in the text. N-LDH ) neo-formed Ni-Al LDH (15), LDH ) synthetic Ni-Al LDH (Ni:Al, 2:1) (13), r ) r-Ni(OH)2,
? ) ?-Ni(OH)2, Ni-phyl ) Ni-phyllosilicate (14).
22789ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
the HCP samples are longer (RNi-Ni ) 3.09-3.11 Å) as
that, in addition to Ni-Al LDH phase, other Ni-containing
phases form. The longer RNi-Niis attributed to the presence
of ?-Ni(OH)2 impurities (RNi-Ni ) 3.13 Å). ?-Ni(OH)2 is
preparation is strongly oversaturated with respect to Ni-
The presence of Ni-Ni and Ni-Al backscattering con-
tributions is further substantiated using WT analysis. Figure
(Figure 3a) and the synthetic Ni-Al LDH (Figure 3b), which
were deduced using the optimized Morlet parameters η )
5.7 and σ ) 1, and a k3-weighted signal. Both samples show
a maximum at R + ∆R ) ∼1.6 Å and at k ) ∼6 Å-1with a
contribution in k-space ranging between ∼2 and ∼9 Å-1.
This corresponds to the first metal shell (Ni-O). The WT of
both samples resolves two maxima at k ) ∼6.6 Å-1and k )
based on the similarities of the WT, we infer that the second
peak in the FT of the Ni-doped HCP sample and Ni-Al LDH
consists of two identical contributions, resulting from
our hypothesis that Ni-Al LDH is formed in the cement
Independent evidence of the Ni-Al LDH formation in
Ni-doped HCP samples was further obtained from DRS
measurements. Figure 4 shows the DRS measurements for
Ni-Al LDH. The Ni-Al LDH has a ν2 band position at
15 439 cm-1, which is comparable to the position reported
by Scheinost et al. (19) (15 220 cm-1- 15 430 cm-1), and a
a ν2 band position at 15 324 cm-1and a small shoulder at
compound. This indicates that predominantly Ni-Al LDH
has formed in the cement sample.
with hydration times varying between 1 hour and 1 year
were investigated to assess the influence of the hydration
time on Ni binding in hydrated cement. Figure 2a shows the
spectra of Ni-doped HCP samples hydrated up to 1 year. In
∼8 Å-1appear. The oscillation at ∼4 Å-1becomes broader
(Figure 2a). In contrast, the position of this oscillation in the
longer hydrated samples (3 days up to 1 year; Ni_cem_3d-
the right shoulder of this oscillation for the short hydration
times (Ni_cem_1h and Ni_cem_6h) shows a small feature
shifts slightly to the right (dashed line) with increasing
upward oscillation ending at ∼8.5 Å-1for the Ni_cem_1h
and Ni_cem_6h samples, similar to R-Ni(OH)2(Figure 2a).
The EXAFS spectra of the Ni-doped HCP samples with
hydration times longer than 3 days show a splitting of the
shown in Figure 2c (∆R ) 0.8-7 Å). The broadening of the
feature at ∼4 Å-1, the shift of the feature at ∼5 Å-1, and the
of Ni-Al LDH formed in HCP increases with increasing
FIGURE 3. Wavelet analysis of the first and second shell (η) 5.7,
for 1 year), and (b) Ni-Al LDH synthetic reference compound (LDH,
Ni:Al, 2:1) (13). Circled areas indicate contributions of Ni-Ni and
the experimental spectrum and the reference compound for the
maxima at k ) ∼6.6 Å-1and ∼9.2 Å-1. This is due to the fact that
the experimental spectrum is composed of a mixture of Ni-Al LDH
and Ni-hydroxide phases, and not of a pure Ni-Al LDH, as for the
reference compound. The Ni-hydroxide phases cause the RNi-Ni
to shift to slightly longer distances R.
FIGURE 4. DRS measurement of Ni_cem_1y sample (gray) (5000
mg/kg Ni loading, hydrated for 1 year) and Ni-Al LDH synthetic
right) and that the noise level of the Ni_cem_1y sample is higher
VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY92279
shift of the feature at ∼4 Å-1and the shoulder of the feature
at ∼4 Å-1suggest that the spectra of the Ni-doped HCP
The corresponding FT of the k3-weighted EXAFS spectra
(Figure 2b) shows no evident dependence on the hydration
for all Ni-doped HCP samples. However, the CNNi-Ni de-
2). This decrease in the CNNi-Ni is caused by destructive
interference of Ni-Ni and Ni-Al backscattering pairs as
discussed earlier. The finding of a significant reduction of
CNNi-Niindicates substantial substitution of Ni by Al.
To test whether a mineral mixture would be consistent
HCP samples were fitted with linear combinations of
R-Ni(OH)2, ?-Ni(OH)2, and Ni-Al LDH. The amount of
Ni-Al LDH in the Ni-doped HCP samples was found to
increase with increasing hydration time (from ∼40% to
∼60%), whereas R-Ni(OH)2 decreases accordingly (from
to be small (∼20%) and remained constant with time.
Influence of Other Experimental Parameters. Besides
hydration time, other experimental parameters were ex-
pected to have a potential effect on the Ni speciation, for
example, the initial Ni concentrations, the anions added to
the system, and the w/c ratio. Therefore, these parameters
were varied to assess their influence on the Ni uptake by
Figure 5 shows the normalized, background-subtracted,
and k3-weighted EXAFS spectra (a), the FT (b), and the FT-1
(c) of Ni-doped HCP samples hydrated for 30 days. These
samples were either doped with Ni(NO3)2 at different
concentrations (50, 500, 5000 mg/kg; Ni_cem_50, Ni_cem_500,
Ni_cem_30d) and different w/c ratios (0.4 and 1.3;
Ni_cem_30d, Ni_cem_w/c_1.3), or the Ni salts used for the
Ni_cem_30d, Ni_cem_SO4, Ni_cem_Cl) (Table 1). Figure
5a reveals that, for the Ni_cem_SO4, Ni_cem_Cl, and
Ni_cem_w/c_1.3 samples, the oscillations at ∼4, ∼5, and
∼8 Å-1are comparable to those previously observed for the
Ni_cem_30d sample, which was prepared at w/c ) 0.4 to
obtain 5000 mg/kg Ni loading using Ni(NO3)2. Further, the
(Figure 5b). To better visualize the contributions resulting
from the second shell, a FT-1was performed in the range
between 2.2 and 3.40 Å (Figure 5d). The contributions of
the second shell in the Ni_cem_SO4, Ni_cem_Cl, and
Ni_cem_w/c_1.3 samples are very similar to that in the
Ni_cem_30d sample. Thus, neither the w/c ratios nor the
type of anions used for the preparation have any noticeable
influence on the EXAFS spectra.
However, some changes in the EXAFS spectra appear at
a splitting, which becomes more pronounced in the sample
with Ni loading of 50 mg/kg (Ni_cem_50). The oscillation at
with decreasing Ni loading. The described spectral features
(∼4, ∼5, and ∼8 Å-1) are better resolved in the FT-1(Figure
a shift to the right and a broadening with decreasing Ni
concentration as compared to the Ni_cem_30d sample. A
shift to the left with decreasing Ni concentrations for the
Ni_cem_500 and Ni_cem_50 samples, indicating significant
changes in the backscattering contributions (Figure 5d).
Structural parameters derived from multi-shell analysis
(Table 2) show similar CNNi-O, CNNi-Ni, and RNi-O;Ni-Nifor all
varied(NO3-, SO42-, Cl-;
experimental spectra, which are further comparable to
structural parameters determined for the Ni_cem_30d sample.
The only exception is the Ni_cem_50 sample, for which the
multi-shell fit approach used throughout this study failed as
it resulted in Ni-Ni distances (3.17 Å) longer than any
time, the situation regarding the Ni speciation at the lowest
loading is unclear.
Controlling Uptake Mechanism of Ni in Cement. From
Ni phases in the Ni-doped HCP samples is controlled by
both kinetic and thermodynamic constraints. The EXAFS
results show that the formation of Ni-Al LDH, which
for 30 days at different concentrations (500, 50), using w/c 1.3 or
different anions (Cl-, SO42-) compared to Ni_cem_30d (NO3-, 5000
mg/kg, w/c 0.4) of (a) k3-weighted, normalized, background-
subtracted EXAFS spectra (the circled area indicates features
explained in the text), (b) experimental (solid line) and theoretical
(dashed line) Fourier transforms (modulus and imaginary parts)
obtained from the EXAFS spectra presented in Figure 5a, (c) k3-
weighted EXAFS function for the Fourier-backtransform spectra
obtained from Figure 5b (range: R + ∆R ) 0.8-7 Å), and (d) k3-
weighted EXAFS function for the Fourier-backtransform spectra of
the second shell obtained from Figure 5b (range: R + ∆R )
2.20-3.40 Å). Dashed lines indicate spectral features explained in
detail in the text. 50 ) Ni_cem_50, 500 ) Ni_cem_500, 5000 )
Ni_cem_30d, SO42-) Ni_cem_SO4, Cl-) Ni_cem_Cl, w/c 1.3 )
22809ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
immediately starts upon the addition of a Ni solution to
further shows that, in addition to Ni-Al LDH, R/? Ni(OH)2
wet chemistry data, however, suggest that ?-Ni(OH)2is not
the thermodynamically most stable phase in the systems.
The portion of Ni-Al LDH formed increases with increasing
hydration time, indicating that Ni-Al LDH is the thermo-
dynamically most stable Ni-containing phase in the cement
matrix. The content of ?-Ni(OH)2was found to be small and
remained constant with time, suggesting that the transfor-
into Ni-Al LDH with increasing hydration time as revealed
from a decrease of the amount of initially formed R-Ni(OH)2
with time. Note that an enhanced stability of Ni-Al LDH
over R-Ni(OH)2in Al-containing hyperalkaline solution was
of the present study further corroborate the findings of
of predominantly Ni-Al LDH and minor quantities of
?-Ni(OH)2when Ni was sorbed onto hydrated cement and
the earlier work of Scheidegger et al. (5, 6) clearly show that
both modes of Ni immobilization, that is, during cement
hydration and due to sorption onto hydrated cement, lead
to identical Ni speciations.
The formation of Ni-Al LDH requires that an Al
source is available over the entire period of the hydration
process. Lothenbach and Wieland (8) recently inves-
tigated the hydration process of the cement used in the
present study. The Al-containing clinker minerals, which
slowly dissolve during hydration, are aluminate and
ferrite. Themost important
carboaluminate (3CaO‚Al2O3.‚CaCO3‚11H2O; ∼8 wt %),
and hydrotalcite (LDH, [Mg1-xAlx(OH)2]x+(An-)x/n‚yH2O;
stops after 24 hours when the sulfate source (gypsum) is
exhausted. After that the formation of calcium-monocar-
boaluminate and hydrotalcite starts, and the portion of the
In this study, the formation of Ni-Al LDH was observed in
the first hours of the hydration process, indicating competi-
tion between ettringite and Ni-Al LDH for Al. Nevertheless,
a significant influence of the competitive reactions on the
hydration process can be excluded due to the small amount
of Ni-Al LDH formed at the given initial Ni concentrations.
The staff of the Swiss-Norwegian Beam Line (SNBL) and the
Dutch Belgium Beamline (DUBBLE) at the European Syn-
chrotron Radiation Facility (ESRF) (Grenoble, France) is
thanked for experimental assistance during the EXAFS
measurements. Thanks are also extended to Dr. E. Curti, D.
Kunz, and Dr. M. Harfouche for assistance during the
measuring campaigns. Prof. R. A. Schoonheydt and H.
Leeman from the Centre of Surface Chemistry and Catalysis
faculty in KU-Leuven provided the DRS measurements, and
Dr. C. A. Johnson (EAWAG, Switzerland) is warmly thanked
for the Ni-Al LDH reference compounds. Gratitude is
expressed to Dr. A. Ulrich (EMPA, Switzerland) for the ICP-
MS measurements, S. Ko ¨chli (PSI) for the ICP-OES measure-
ments, and D. Kunz (PSI) for his contribution to the wet
is gratefully acknowledged for useful discussion and con-
structive comments. Partial financial support was provided
by the National Cooperative for the Disposal of Radioactive
Waste (Nagra), Switzerland.
Supporting Information Available
the edges of the tetrahedra are filled with Ni and Al, and in
Cl-, SO42-). This material is available free of charge via the
Internet at http://pubs.acs.org.
(1) Schmidt, M.; Beckefeld, P.; Go ¨tz, R.; Kamsties, S.; Kretz, C.;
Immobilisierung von Schadstoffen-Recycling-Verbesserung der
Deponiefa ¨higkeit;ExpertVerlag: Renningen-Malmsheim,1995.
(2) Chapman, N.; McCombie, C. Principles and standards for the
Ltd.: Oxford, 2003.
forms; Spence, R. D., Ed.; Lewis Publishers: Boca Raton, FL,
modelling of blended cements at elevated temperatures (50 °C
to 90 °C). DoE/UMIP/PR/94.011, 1994.
(5) Scheidegger,A.;Wieland,E.;Da ¨hn,R.;Spieler,P.Spectroscopic
in cement. Environ. Sci. Technol. 2000, 34, 4545-4548.
(6) Scheidegger, A. M.; Wieland, E.; Scheinost, A. C.; Da ¨hn, R.; Tits,
J.; Spieler, P. Ni phases formed in cement and cement systems
Radiat. 2001, 8, 916-918.
(7) Do ¨hring, L.; Go ¨rlich, W.; Ru ¨ttener, S.; Schwerzmann, R. Her-
stellung von homogenen Zementsteinen mit hoher hydraulis-
cher Permeabilita ¨t. Nagra NIB 94-29, 1994.
(8) Lothenbach, B.; Wieland, E. A thermodynamic approach to the
(9) Wieland, E.; Tits, J.; Spieler, P.; Dobler, J. P. Interaction of Eu-
Res. Soc. Symp. Proc. 1998, 506, 573-578.
(10) Wieland, E.; Tits, J.; Ulrich, A.; Bradbury, M. H. Experimental
evidence for solubility limitation of the aqueous Ni(II) con-
centration and isotopic exchange of
systems. Radiochim. Acta 2006, 94, 29-36.
(11) Ressler, T. WinXAS: A program for X-ray absorption spectros-
copy data analysis under MS-Windows. J. Synchrotron Radiat.
1998, 5, 118-122.
(12) Rehr, J. J.; Mustre de Leon, J.; Zabinsky, S. I.; Albers, R. C.
Theoretical X-ray absorption fine structure standards. J. Am.
Chem. Soc. 1991, 113, 5135-5140.
(13) Johnson, C. A.; Glasser, F. P. Hydrotalcite-like minerals
(M2Al(OH)6(CO3)0.5XH2O, where M ) Mg, Zn, Co, Ni) in the
stability. Clays Clay Miner. 2003, 51, 1-8.
(14) Da ¨hn, R.; Scheidegger, A. M.; Manceau, A.; Schlegel, M. L.;
Baeyens, B.; Bradbury, M. H.; Morales, M. Neoformation of Ni
phyllosilicate upon Ni uptake on montmorillonite: A kinetics
study by powder and polarized extended X-ray absorption fine
structure spectroscopy. Geochim. Cosmochim. Acta 2002, 66,
(15) Scheidegger, A. M.; Lamble, G. M.; Sparks, D. L. The kinetics of
nickel sorption on phyrophyllite as monitored by X-ray absorp-
(16) Funke, H.; Scheinost, A. C.; Chukalina, M. Wavelet analysis of
71, 094110, 1-7.
(17) Mun ˜oz, M.; Argoul, P.; Farges, F. Continuous Couchy wavelet
transform analyses of EXAFS spectra: A qualitative approach.
Am. Mineral. 2003, 88, 694-700.
(18) Bensalem, A.; Weckhuysen, B. M.; Schoonheydt, R. A. In situ
catalysts:Kinetics of the reduction process with carbon
monoxide. J. Phys. Chem. B 1997, 101, 2824-2829.
(19) Scheinost, A. C.; Ford, R. G.; Sparks, D. L. The role of Al in the
Acta 1999, 63, 3193-3203.
63Ni in cementitious
VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY92281
(20) Scheckel, K. G.; Sparks, D. L. Kinetics of the formation and
dissolution of Ni precipitates in a gibbsite/amorphous silica
mixture. J. Colloid Interface Sci. 2000, 229, 222-229.
(21) Verberckmoes, A. A.; Uytterhoeven, M. G.; Schoonheydt, R. A.
Framework and extraframework Co2+in CoAPO-5 by diffuse
reflectance spectroscopy. Zeolites 1997, 19, 180-189.
2003, 134, 941-973.
(23) Scheinost, A. C.; Sparks, D. L. Formation of layered single- and
double-metal hydroxide precipitates at the mineral/water
interface:A multiple-scattering XAFS analysis. J. Colloid
Interface Sci. 2000, 223, 1-12.
(24) Mansour, A. N.; Melendres, C. A. Analysis of X-ray absorption
spectra of some nickel oxycompounds using theoretical stan-
dards. J. Phys. Chem. A 1998, 102, 65-81.
(25) Bode, H.; Dehmelt, K.; Witte, J. Zur Kenntnis der Nickelhy-
droxidelektrode-I. U ¨ ber das Nickel (II)-Hydroxidehydrat. Elec-
trochim. Acta 1966, 11, 1079-1087.
(26) Wang, C. Y.; Zhong, S.; Konstantinov, K.; Walter, G.; Liu, H. K.
Structural study of Al-substituted nickel hydroxide. Solid State
Ionics 2002, 148, 503-508.
Received for review November 8, 2005. Revised manuscript
received January 25, 2006. Accepted January 30, 2006.
22829ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
SUPPORTING INFORMATION FOR
Spectroscopic Investigation of Ni Speciation in Hardened Cement Paste
M. Vespa1,2*, R. Dähn1, D. Grolimund1, E. Wieland1, A. M. Scheidegger1,2
1Paul Scherrer Institute, Laboratory for Waste Management, 5232 Villigen PSI, Switzerland
2Departement of Environmental Sciences, Swiss Federal Institute of Technology (ETH), Zürich, Switzerland
Number of pages: 2
Number of Figures: 1
Number of Tables : 0
S2 Download full-text
Figure S1. Structural model of Ni-Al LDH. The edge of the tetrahedra are filled with Ni and Al. The interlayer
position can be filled with different anions such as CO32-, NO3-, Cl-, SO42-.
Ni + Al
Ni + Al