materials

Article
Removal of Hexavalent Chromium in Portland
Cement Using Ground Granulated
Blast-Furnace Slag Powder
Sungchul Bae 1, * ID , Fumino Hikaru 2 , Manabu Kanematsu 2 , Chiaki Yoshizawa 3 ,
Takafumi Noguchi 4 , Youngsang Yu 5 ID and Juyoung Ha 6
1 Department of Architectural Engineering, Hanyang University, Seoul 04763, Korea
2 Faculty of Science and Technology, Tokyo University of Science, 2641 Yamazaki,
Noda, Chiba 278-8510, Japan; waraimaker@gmail.com (F.H.); manabu@rs.noda.tus.ac.jp (M.K.)
3 Research Institute of Science and Technology, Nihon University College of Science and Technology, Chiyoda,
Tokyo 101-8308, Japan; chiaki.yoshizawa@orsc.co.jp
4 Department of Architecture, Graduate school of Engineering, The University of Tokyo,
Bunkyo, Tokyo 113-8654, Japan; noguchi@bme.arch.t.u-tokyo.ac.jp
5 Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA; ysyu@lbl.gov
6 School of Environmental and Sustainability Sciences, Kean University, Union County, NJ 07083, USA;
juyoung@gmail.com
* Correspondence: sbae@hanyang.ac.kr; Tel.: +82-2-2220-0302

Received: 23 October 2017; Accepted: 20 December 2017; Published: 22 December 2017

Abstract: Using ground granulated blast-furnace slag (GGBS) under different alkaline conditions,
we studied the mechanisms and extents of Cr(VI) reduction and sorption and compared them to
reactions with Portland cement (PC). We also investigated the effects of mixing PC/GGBS ratios on
Cr(VI) dissolution after carbonating the substrates. We observed a complete sorption and reduction
of Cr(VI) to Cr(III) in a GGBS-in-Ca(OH)2 solution (pH > ~12.5) after 10 h, whereas in distilled water
(pH = ~11.5) GGBS exhibited only marginal sorption and reduction (20%). Cr reactions with dissolved
ions in supernatants derived from GGBS indicated that the anions dissolved from GGBS act as a
reducing agent for Cr(VI) in a Ca(OH)2 solution. Soft X-ray absorption microscopy identified a partial
reduction of Cr(VI) to Cr(III) on the GGBS surface. The carbonation of pure PC paste substantially
increased the amount of dissolved Cr(VI) in a solution phase whereas a 5 wt % replacement of PC
with GGBS significantly reduced the amount of dissolved Cr(VI). We concluded that in the mixed
paste during the early curing stage GGBS reduced a significant fraction of Cr(VI) to Cr(III) and that
the Cr(III) adsorbed in the GGBS-PC mixture’s hydration products does not readily dissolve, even
under carbonation conditions.

Keywords: hexavalent chromium; reduction; immobilization; slag; Portland cement

1. Introduction
Hexavalent chromium [Cr(VI)] is found in concrete as chromate, which results from the
oxidation process of trivalent chromium [Cr(III)] during Portland cement (PC) clinker production [1,2].
Chromium (Cr) can exist in several different oxidative states and its physical and chemical
properties, such as solubility, toxicity, and sorption affinity, vary with environmental conditions.
For example, Cr(III) is likely to precipitate in alkaline solutions, whereas Cr(VI) is amphoteric
with a concentration-dependent solubility [3]. Cr(VI) is highly mobile but very toxic under aerobic
conditions [1,4,5]. Cr(III) has a strong affinity to soils and other substrates; while it is less soluble and
toxic than Cr(VI), its potential for oxidation to Cr(VI) should be carefully examined in order to fully
understand Cr transport and distribution in the environment. Therefore, examining Cr(VI) dissolution,

Materials 2018, 11, 11; doi:10.3390/ma11010011 www.mdpi.com/journal/materials

Materials 2018, 11, 11 2 of 17

sorption, and reduction in PC clinker is of the utmost important for environmental concerns and for
the long-term sustainability of PC materials in construction materials [3,6].
In the past, stabilization and solidification (s/s) have been practiced on immobilized PC-associated
Cr(VI) [7–9]. Stabilization involves a reductive transformation of Cr(VI) to Cr(III) and a subsequent
chemical transformation into a highly insoluble compound, such as calcium chromate [10,11].
Solidification involves the co-precipitation of Cr(VI) with dissolved PC hydration products during
hydration reactions [10]. Calcium silicate hydrate (C-S-H), which is the primary hydration and
binding product in PC-based materials, contains most of the microporosity, resulting in a large specific
surface area that provides physical sorption and co-precipitation sites for heavy metal ions like
Cr(VI). It should also be noted that Zamorani et al. [12] observed co-precipitated phase of CaCrO4
when Cr(VI) reacted with C-S-H, suggesting that molecular-level surface reactions are important in
determining the reaction rates and pathways when hydrated products of PC-based materials react with
Cr(VI) [3]. Ettringite (3CaO·Al2 O3 ·3CaSO4 ·32H2 O) has also been identified as one of PC’s hydration
products and, with its highly reactive and needle-like column structure, has been proposed as a surface
site for Cr(VI) immobilization via anion substitution, where CrO4 2− is exchanged for SO4 2− [13,14].
Another study identified that reductive reactions occur during the curing phases and that the resulting
Cr(III) also replaces Al(III) in the ettringite crystalline structure [15]. Ionic substitutions in ettringite
were also observed to be Cr(VI)–ettringite and Cr(III)–ettringite phases that further limit the mobility
of Cr [8,16]. Cr(VI) can be substituted for SO4 2− in the interlayer of calcium aluminate monosulfate
(hereafter referred to as monosulfate, 3CaO·Al2 O3 ·CaSO4 ·12H2 O) [16]. However, the pH stability
range of reactions for Cr(VI)–ettringite and Cr(VI)-hydration products of PC is very narrow and varies
significantly [16,17].
The alkalinity of the PC-based matrix is important for decreasing the dissolution of Cr ions [3,11].
However, the carbonation process neutralizes the alkaline conditions and increases the Cr(VI) solubility
for PC pastes [18]. A decrease in the pH conditions due to carbonation causes the decomposition of
Cr(VI)-bearing hydration products in cementitious materials, which leads to more dissolved aqueous
Cr(VI) [19].
Ground granulated blast-furnace slag (GGBS) is a well-known by-product of the pig-iron
manufacturing process, and up to 60–70% of GGBS is often incorporated into concrete in order to
recycle industrial waste materials [20]. GGBS is capable of immobilizing Cr(VI) via a reductive process:
dissolved S2− ions from GGBS transform Cr(VI) into Cr(III) [3]. On the contrary, Macias et al. [18]
observed an increased dissolution of Cr(VI) in GGBS-PC pastes due to the Cr(VI) that was released
into the cement materials’ pore solution. The effect of GGBS on Cr(VI) reduction and immobilization
has not yet been quantitatively studied, and the exact mechanism of GGBS remains unclear.
The aim of this study is to elucidate the mechanism of GGBS on Cr(VI) speciation under different
alkaline levels. Ca(OH)2 was utilized in order to maintain a high alkaline condition for the activation of
GGBS. We also studied GGBS’s capacity to reduce and immobilize Cr(VI), comparing it to that of pure
PC. We performed quantitative determinations of the Cr(VI) using a UV-visible spectrophotometer.
The dissolved cations and anions in Cr(VI)’s reductive transformation in the solution phase were
quantified under different alkaline conditions using ion chromatography. The phases for each
sample that was made to react with different Cr(VI) concentrations were identified using X-ray
diffraction (XRD). Synchrotron-based X-ray microscopy and spectroscopy were used to identify
the detailed chemical speciation of Cr that had reacted with the Ca(OH)2 -activated GGBS particles.
Finally, an accelerated carbonation test was performed to examine the effects of carbonation on Cr(VI)
leaching from a GGBS-PC paste.
Materials 2018, 11, 11 3 of 17

2. Experimental Methods

2.1. Material and Sample Descriptions

The chemical compositions of PC and GGBS were determined using X-ray fluorescence (XRF, Bruker,
Karlsruhe, Germany) (Table 1). Research-grade PC was used instead of commercial PC to minimize any
possible impurities and their subsequent interference with our experimental results.

Table 1. The chemical compositions of the raw materials.

Chemical Compositions (%)

Materials
SiO2 Al2 O3 Fe2 O3 CaO MgO SO3 Na2 O K2 O TiO2 P2 O5 MnO Total
PC 20.94 5.45 2.83 64.96 1.5 2.05 0.32 0.48 0.27 0.31 0.08 99.22
GGBS 33.03 14.82 0.42 41.44 6.51 2.07 0.27 0.27 0.53 0.04 0.18 99.9

Table 2 shows the mix proportions and experimental measurements for the different samples.
Four different series of experimental samples were prepared, and they are denoted as S1, S2, S3, and S4.
In the S1 samples, we examined Cr(VI) reduction and immobilization using PC or GGBS. A Cr(VI) solution
(0.5 mg/L) was prepared by dissolving potassium dichromate (K2 Cr2 O7 ) in distilled water. The GGBS or
PC powder (40 g) was added into the Cr(VI) solutions, and the suspensions were agitated at 20 ◦ C in 60%
relative humidity (RH) conditions. The supernatant of the S1 samples was filtered, and the concentration
of the Cr(VI) remaining in the filtered solution was determined. The changes in pH of the S1 suspensions
were measured via a hand-held pH meter (D-52, Horiba, Tokyo, Japan).

Table 2. The mix proportions of the specimens.

Mix Proportions
Sample Distilled GGBS CH PC Cr(VI) Concentration Measurements
Water (L) (g) (g) (g) (mg/L)
PC-S1 - - 40 UV-vis, XRD, pH
GGBS-S1 2 40 - - 0.5 UV-vis, XRD, pH
GGBS-CH-S1 40 20 - UV-vis, XRD, pH, Soft X-ray absorption microscopy
PC-S2 - - 48
GGBS-S2 2.4 48 - - 0.5 UV-vis
GGBS-CH-S2 48 24 -
GGBS-CH-S3 2 40 20 - 0.5~100 UV-vis, XRD
GGBS-S4 2 40 - - -
Ion chromatography
GGBS-CH-S4 2 40 20 - -
GGBS = Granulated ground blast-furnace slag; CH = Ca(OH)2 ; PC = Portland cement; XRD = X-ray diffraction;
UV-vis = Ultraviolet-visible spectrophotometry.

S2 samples were prepared in order to verify what effects the dissolved ions from PC or GGBS had
on Cr(VI) reduction in the solution phase. The PC or GGBS powder (48 g) was mixed with distilled water
(2.4 L), and the substrates were equilibrated in distilled water for 24 h to ensure the dissolution of the
substrates. After 24 h, the supernatant was filtered, and 1.2 mg of K2 Cr2 O7 was added to the filtered
solutions containing the dissolved ions from the paste samples. Changes to the initial Cr(VI) concentration
(0.5 mg/L) was measured at 0, 1, 3, 6, 12, and 24 h in the reacted solutions to identify the extent of the
reductive reaction that was due to the dissolved ions from the substrates in the solutions.
The S3 experiments series aimed to determine the Cr(VI) sorption capacity of GGBS. Different amounts
of Cr(VI) ranging from 0.5 to 100 mg/L were added to a GGBS suspension. After agitating the suspension,
the remaining Cr(VI) concentration was determined at 120 and 150 h. Lastly, the S4 series was used to
determine the different chemical species of the dissolved free ions from the GGBS samples under different
alkaline condition. Ion chromatography (Dionex ICS-2000, Sunnyvale, CA, USA), using suppressed
conductivity detection, was performed to quantify the ions that were dissolved in S4 samples. The Cr(VI)
concentration in S4 samples was measured using a UV-visible spectrophotometer (DR 5000™, Hach,
Materials 2018, 11, 11 4 of 17

London, ON, Canada) and the EPA 7196A method with 1,5-diphenylcarbazide (Cr(VI) Detection limit, Max:
1.0 mg/L, Min: 0.03 mg/L). All the solutions were filtered through a 0.45-µm membrane filter, and the
alkaline conditions of the solutions were maintained using pure Ca(OH)2.

2.2. Accelerated Carbonation and Tank Leaching Test

GGBS-PC pastes (ø50 × 100 mm) were made at GGBS replacement ratios of 0, 5, 10, and 50 wt %.
The water to total solid (GGBS + PC) ratio (w/s) of the pastes was maintained at 0.6 for all the samples.
The total initial Cr(VI) concentration was at 0.5 mg/L per 100 g of paste. During the first 24 h of curing,
plastic sheets covered the exposed surfaces of the specimen in the mold. The specimens were demolded
after 24 h and were cured for an additional 28 days at 20 ◦ C in 60% RH conditions. For accelerated
carbonation, the final sample was finely grounded and placed in a carbonation chamber at 20 ◦ C with
5% CO2 and 60% RH for 28 days. After the carbonation period was completed, the specimens were kept
in an ethanol suspension and stored in a vacuum drying chamber until further analysis. The hydration
products were identified both before and after 28 days of carbonation. After carbonation, the remaining
Cr(VI) concentration in the samples was measured according to Regulation No. 49 (The Ministry of
Land, Infrastructure and Transport, Japan).

2.3. Phase Identification

The raw material phases, the hydration products in the filtered solids derived from the S1 and
S3 suspensions, and the accelerated carbonation test specimens before and after carbonation were all
identified via powder XRD using monochromatized CuKα radiation (λ = 1.54 Å, at a voltage of 30 kV,
a current of 10 mA, in the range of 2θ = 5◦ –75◦ , D2 Phaser, Bruker, Karlsruhe, Germany). The residues
collected after filtering were dried and used for the XRD analysis. Prior to the sample preparation, both
the carbonation of the raw materials and the purity of Ca(OH)2 , which was used as a reference compound,
were checked using XRD. Although great care was taken to avoid carbonation of the pure Ca(OH)2 ,
very small peaks of CaCO3 were seen in the XRD pattern, and the amount of CaCO3 was quantified as
~2 wt % via Rietveld refinement using the Topas software (version 4.3, Bruker, Karlsruhe, Germany).
This result may be due to the unavoidable carbonation of pure Ca(OH)2 during the XRD measurements.

2.4. Scanning Transmission X-ray Microscopy (STXM)

Spatial correlations between the chemical compositions and their oxidation states were mapped by
a combination of scanning transmission X-ray microscopy (STXM) and X-ray absorption spectroscopy
(XAS) across Cr L2,3 -edges (560–600 eV) and the O K-edge (525–560 eV), which allows us to image
distributions of the chemical species with spectral sensitivity from very small spots (~35 nm).
STXM measurements were performed at the bending magnet beamline (5.3.2.1) at the Advanced
Light Source (ALS) in the Lawrence Berkeley National Laboratory, Berkeley, CA, USA [21]. To obtain
transparency in the soft X-ray region, the diluted mixtures (dried leachate of GGBS-CH-S3 and 10 mg/L
of added Cr(VI)) were drop-casted with isopropanol onto a 50-nm-thick Si3 N4 membrane (Silson Inc.,
Warwickshire, England). STXM measurements utilized a 25-nm outer-width Fresnel zone plate for
focusing the X-ray. A raster scanned onto 8 × 8 µm2 field-of-views with a square scan grid of 35 nm
steps. Image spectra (i.e., repeated images at different energies), whose finest energy step was 0.1 eV
near the absorption resonance, were recorded by a phosphor photomultiplier tube (PMT). The intensity
transmitted through the samples (I) was converted into the optical density (OD) by taking the log ratio
of the incident (I0 ) and transmitted (I) intensities (Equation (1)):
 
I
OD = − ln = σ·t = µ·ρ·t, (1)
I0

where I0 : incident X-ray photon intensity, t: specimen thickness, µ: mass absorption coefficient, and ρ:
density of the specimen.
Materials 2018, 11, 11 5 of 17
Materials 2018, 11, 11 5 of 17

To overcome
To overcome spectral
spectral distortions
distortions resulting
resulting from
from the
the acquisition
acquisition time
time gap
gap between
between the
the II00 and
and II
measurements, the
measurements, theincident
incidentbeam
beamintensity
intensitywas
wassimultaneously
simultaneously measured
measured through
through an open
an open areaarea of
of the
the sample. After image registration by a phase-correlation algorithm, the XAS spectra of each
sample. After image registration by a phase-correlation algorithm, the XAS spectra of each pixel as a pixel
as a function
function of energy
of energy were corrected.
were corrected. All dataAll data processing
processing was out
was carried carried
usingout
theusing the Axis2000
Axis2000 software
software8.3,
(version (version 8.3, Univ.,
McMaster McMaster Univ.,ON,
Hamilton, Hamilton,
Canada)ON,[22].Canada) [22]. Further
Further details details
regarding regarding
applications of
applications of STXM to the study of cementitious systems can be found
STXM to the study of cementitious systems can be found elsewhere [23–26]. elsewhere [23–26].

3. Results and Discussion

3.1. Cr(VI) Removal Using PC and GGBS-PC Samples

Figure 1a shows the time profiles of the pH values for the suspensions retrieved from the S1
samples over
over 120
120h.
h.The
ThepH pH value
value forfor PC-S1
PC-S1 increased
increased from
from 12.712.7 to 13.2
to 13.2 in 12inh12
and h decreased
and decreased to
to 12.5.
12.5. This
This is is in good
in good agreement
agreement with with PC paste,
PC paste, which which
has ahas a water
water to binder
to binder ratioratio
(W/B)(W/B)
of 0.3of[27].
0.3 [27].
No
No significant
significant changes
changes in pH
in pH were
were observed
observed afterwards.On
afterwards. Onthe
theother
otherhand,
hand,the
thepHpHvalue
value for
for GGBS-S1
GGBS-S1
increased from
from 11.1
11.1toto11.7
11.7inin
2424 h and
h and remained
remained in 11.6–11.8
in the the 11.6–11.8
rangerange untilh 120
until 120 had h had passed.
passed. When
When compared to PC-S1 and GGBS-S1, GGBS-CH-S1,
compared to PC-S1 and GGBS-S1, GGBS-CH-S1, to which Ca(OH)2 2was added,to which Ca(OH) was added, exhibited
comparatively small changes in pH within the first 20 h. It had a constant value of 12.7 until 120 h
had passed. Past studies indicate that GGBS’s GGBS’s activation
activation isis strongly
strongly dependent
dependent on on pH
pH values
values [28,29].
[28,29].
Higher pHs lead to improved GGBS activation, with a value higher than 11.5 leading to an effective
activation [30]. Therefore,
Therefore, alkaline
alkaline conditions
conditions for for GGBS
GGBS activation
activation can be maintained by adding
Ca(OH)
Ca(OH)22 (GGBS-CH-S1).
(GGBS-CH-S1). The Theeffect
effectofofalkalinity
alkalinityononGGBS
GGBSactivation waswas
activation further confirmed
further in theinXRD
confirmed the
results (see Section 3.4).
XRD results (see Section 3.4).

14 0.9
(a) (b)
Concentration of Cr(VI) in Solution (mg/L)

0.8 PC-S1
GGBS-S1
13 0.7 GGBS-CH-S1
0.5 mg/L
0.6
12
0.5
pH

0.4
11
0.3

PC-S1 0.2
10
GGBS-S1
0.1
GGBS-CH-S1
0.0
9
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Time (Hours) Time (Hours)

Figure 1.1.Time
Figure Time profiles
profiles ofthe
of (a) (a)pH
the pH of
values values of S1 suspensions
S1 samples’ samples’ suspensions and (b)
and (b) the Cr(VI) the Cr(VI)
concentrations
concentrations in the S1 samples’ filtered solutions measured
in the S1 samples’ filtered solutions measured from 0 to 120 h. from 0 to 120 h.

Figure 1b 1b depicts
depicts the
theCr(VI)
Cr(VI)concentration
concentrationofofthe thesupernatants
supernatants([Cr(VI)]
([Cr(VI)] supernatants)) retrieved
supernatants retrieved from
samplesas
the S1 samples asaafunction
functionofoftime.
time.Within
Within1 1h,h,[Cr(VI)]
[Cr(VI)] supernatants increased
supernatants increased from 0.50 to 0.70 0.70 mg/L
mg/L for
samples. Following
PC samples. Following this,this,[Cr(VI)]
[Cr(VI)]supernatants
supernatants gradually
graduallydecreased
decreasedover over120 120hh to to below
below the detection
(0.03mg/L).
limit (0.03 mg/L).The The initial
initial increase
increase in theinCr(VI)
the Cr(VI) concentration
concentration is likelyisdue likely
to thedue to dissolution
initial the initial
dissolution
of the cement ofclinkers
the cement
in theclinkers
PC samples. in theIn PCthe samples. In the GGBS-S1
GGBS-S1 without Ca(OH)2 ,without the [Cr(VI)] Ca(OH) 2, the
supernatants
[Cr(VI)] decreased from 0.5 to 0.4 mg/L within 5 h and remained
decreased from 0.5 to 0.4 mg/L within 5 h and remained relatively constant at 0.40 mg/L over 120
supernatants relatively constant at 0.40
h.
mg/L
In over 120
contrast, a rapidh. In contrast,
decrease wasa rapid
observed decrease
for thewas observed for
GGBS-CH-S1 the GGBS-CH-S1
in Ca(OH) 2 solution, within Ca(OH)
its the2
solution,supernatants
[Cr(VI)] with its the [Cr(VI)]
reaching 0 supernatants
mg/L within reaching24 h.0 This
mg/Liswithin 24 h. This
an interesting is an interesting
observation becauseobservation
increasing
because increasing the pH condition is expected to increase [Cr(VI)]
the pH condition is expected to increase [Cr(VI)]supernatants . Instead, we observed a [Cr(VI)]supernatantsa
supernatants. Instead, we observed
[Cr(VI)]supernatants
decrease decreaseof
in the presence inthe
theGGBS
presence of the
paste. OurGGBS
resultspaste.
suggestOurthat results
GGBS suggest
has anthatimpactGGBS on has an
Cr(VI)
impact on
reaction Cr(VI) in
pathways reaction
Ca(OH)pathways
2 conditions in Ca(OH)
and that conditions
2alkali-activated and that
GGBS alkali-activated
can remove and GGBS
immobilizecan
removemore
Cr(VI) and immobilize Cr(VI) more
than the untreated GGBSthan the untreated
in distilled GGBS in distilled water can.
water can.
Materials 2018, 11, 11 6 of 17
Materials 2018, 11, 11 6 of 17

3.2.
3.2. Cr(VI)
Cr(VI)Reduction
Reductionvia
via Dissolved
Dissolved Ions
Ions from
from GGBS
GGBS
Figure
Figure 2 shows the evolution of the Cr(VI) Cr(VI) concentration
concentration in in the
thefiltered
filteredsolutions
solutions([Cr(VI)]
([Cr(VI)]aqueous
aqueous))

of
of the
the S2
S2 samples
samplesas asaafunction
functionofoftime.
time.[Cr(VI)]
[Cr(VI)]
aqueous decreases
aqueous when
decreases it
when reacts
it with
reacts dissolved
with ions
dissolved
from PC and GGBS. Water-S2, the control sample, is a mixture composed
ions from PC and GGBS. Water-S2, the control sample, is a mixture composed of distilled water of distilled water and
K 2Cr2K
and O27Crmixed with awith
2 O7 mixed pre-determined Cr(VI)Cr(VI)
a pre-determined concentration of 0.5ofmg/L,
concentration brought
0.5 mg/L, to reaction.
brought The
to reaction.
initial [Cr(VI)]
The initial [Cr(VI)]
aqueous for all
aqueous the
for all samples
the was
samples 0.5
was mg/L.
0.5 mg/L.During
Duringthe first
the hour,
first hour, a aslight
slight increase
increase in
in
[Cr(VI)] aqueousfor
[Cr(VI)]aqueous forthe
thePC-S2
PC-S2sample
sample (~0.52
(~0.52 mg/L),
mg/L),when whencompared
compared to to Water-S2 (~0.47
(~0.47 mg/L),
mg/L), waswas
likely
likely due
due toto Cr(VI)
Cr(VI) dissolving
dissolving from
from the
the cement
cement clinker,
clinker, which we also also observed
observed in in Figure
Figure 1b1b because
because
PC
PC inherently
inherently contains a small amount of trace Cr(VI).

1.1
0.6
Concentration of Cr(VI) in Solution (mg/L)

1.0 PC-S2 0.5

GGBS-S2 0.4

Cr(VI) (mg/L)
0.9 GGBS-CH-S2
0.3
0.8 Water-S2
0.2
0.5 mg/L
0.7 0.1

0.0
0.6 0 2 4 6 8
Time (Hours)
0.5

0.4

0.3

0.2

0.1

0.0
0 2 4 6 8 10 12 14 16 18 20 22 24
Time (Hours)
Figure
Figure2.
2.Time
Timeprofile
profile of
of the
the Cr(VI)
Cr(VI) concentration
concentration in
in the
the S2
S2 specimens’
specimens’ solutions
solutions from
from 00 to
to 24
24 h.
h.

The
The [Cr(VI)] aqueousininGGBS-S2
[Cr(VI)]aqueous GGBS-S2decreased
decreasedtoto 0.38 0.38 within
within 11 hh before
before reachingreaching aa constant
constant
concentration.
concentration. On On the
the contrary,
contrary, due due toto the presence of
the presence of Ca(OH)
Ca(OH)22,, GGBS-CH-S2
GGBS-CH-S2 showed showed aa significant
significant
and
and continuous
continuous Cr(VI)
Cr(VI) reduction
reduction overover 66 h before finally
h before finally reaching
reaching 0.05
0.05 mg/L
mg/L at at 24
24 h,h, which
which suggests
suggests
an
an almost
almost aa complete
complete removal
removal of of Cr(VI)
Cr(VI) from
from the solution. Based
the solution. Based on on the
the results,
results, we we attributed
attributed the
the
reduced amount of Cr(VI) in the solution to the solution chemistry between
reduced amount of Cr(VI) in the solution to the solution chemistry between the dissolved ions in the the dissolved ions in the
GGBS-CH-S2
GGBS-CH-S2 samplesample with withCr(VI).
Cr(VI).InIncontrast,
contrast, ourour experimental
experimental data
data indicated
indicated thatthat the dissolved
the dissolved ions
ions
from from
the PCthedidPCnotdid not the
reduce reduce theOur
Cr(VI). Cr(VI).
resultsOur results that
suggested suggested that theCr(VI)
the significant significant Cr(VI)
concentration
concentration
decrease for the decrease
GGBS for the GGBS
in Ca(OH) in Ca(OH)
2 solution was2 solution
primarily was
dueprimarily due to the
to the reducing reducing
reaction reaction
between the
between
dissolvedthe dissolved
ions from theionsGGBSfrom theCr(VI).
and GGBS and Cr(VI).
Figure
Figure 33shows
showsthe theevolution
evolution of of
thethe
concentrations
concentrations of the
of dissolved
the dissolved ions from ions fromthe GGBS in water
the GGBS in
(GGBS-S4) and GGBS in Ca(OH) solution (GGBS-CH-S4) within
water (GGBS-S4) and GGBS in Ca(OH)2 solution (GGBS-CH-S4) within 25 h. In both GGBS-S4 and
2 25 h. In both GGBS-S4 and
GGBS-CH-S4
GGBS-CH-S4 samples,
samples, the concentrations of
the concentrations of Al
Al[III]
[III] were relatively lower (lower than 1 mg/L) than
were relatively lower (lower than 1 mg/L) than
those recorded for the other cations. For
those recorded for the other cations. For GGBS-S4, the GGBS-S4, thethree
threecations
cations(Al (Al [III], Si
[III] [IV], and Ca2+
, Si[IV] , and Ca 2+)) showed
showed
similar dissolution trends.
similar dissolution trends.We Wepresumed
presumed that
that thethe concentration
concentration changes
changes in those in those
cationscations resulted
resulted from
from the dissolution
the slow slow dissolution
of GGBSof GGBS in water.
in water. Theconcentration
The Si[IV] Si[IV] concentration in GGBS-CH-S4
in GGBS-CH-S4 was verywaslowvery
duelowto
due to the rapid consumption of Si in GGBS’s alkali-activation
the rapid consumption of Si in GGBS’s alkali-activation process. process.
Materials 2018, 11, 11 7 of 17
Materials 2018, 11, 11 7 of 17

1.4 1.0 16 1.0 22 700 1.0

GGBS-S4
(a) (b) (c)

Ca2+ (mg/L) (GGBS-CH-S4)

1.2 GGBS-CH-S4 14
20 600

Ca2+ (mg/L) (GGBS-S4)

GGBS-CH-S2 (Cr(VI)) 0.8 0.8 0.8
1.0 12

Cr(VI) (mg/L)
10 18 500

Cr(VI) (mg/L)

Cr(VI) (mg/L)
0.6 0.6 0.6
Al[III] (mg/L)

Si[IV] (mg/L)
0.8
8 16 400
0.6
0.4 0.4 0.4
6
0.4 14 300
4
0.2 0.2 0.2
0.2 12 200
2
0.0 0.0 0 0.0 0.0
10 100
0 5 10 15 20 25 0 5 10 15 20 25
0 5 10 15 20 25
Time (Hours) Time (Hours)
Time (Hours)
24 1.0 24 1.0 24 1.0
(d) (e) (f)
20 0.8 20 20
0.8 0.8

16 16 16

Total S (mg/L)

Cr(VI) (mg/L)
Cr(VI) (mg/L)

Cr(VI) (mg/L)
S2O32- (mg/L)
SO42- (mg/L)

0.6 0.6 0.6

12 12 12
0.4 0.4 0.4
8 8 8
0.2 0.2 0.2
4 4 4

0 0.0 0.0 0.0

0 0
0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25
Time (Hours) Time (Hours) Time (Hours)

Figure 3. Time profiles for the ion dissolution from GGBS in distilled water (GGBS-S4) and GGBS in
Figure 3. Time profiles for the ion dissolution from GGBS in distilled water (GGBS-S4) and GGBS
Ca(OH)2 solutions (GGBS-CH-S4) for 0 to 24 h. The Cr(VI) concentration in the GGBS-CH-S2 solution
in Ca(OH)2 solutions (GGBS-CH-S4) for 0 to 24 h. The Cr(VI) concentration in the GGBS-CH-S2
(Figure 2) is presented in the same plot. The maximum value of the y-axis was adjusted to the
solution (Figure 2) is presented in the same plot. The maximum value of the y-axis was adjusted to the
concentration value of each ion. (a) Al[III]; (b) Si[IV]; (c) Ca2+; (d) S2O32−; (e) SO42−; and (f) total S.
concentration value of each ion. (a) Al[III] ; (b) Si[IV] ; (c) Ca2+ ; (d) S2 O3 2− ; (e) SO4 2− ; and (f) total S.

The initial
The initial Ca Ca2+2+ concentration in GGBS-CH-S4 (~200 mg/L) was much higher than for GGBS-S4
concentration in GGBS-CH-S4 (~200 mg/L) was much higher than for GGBS-S4
(~12 mg/L)
(~12 mg/L) because
becauseofof thetheaddedaddedCa(OH) Ca(OH) 2. After 1 h, 2+ the Ca2+ concentration in GGBS-CH-S4
2 . After 1 h, the Ca concentration in GGBS-CH-S4 increased
increased significantly
significantly to ~650 mg/L to ~650 mg/Ldecreased
but then but then to decreased
400 mg/L to at
400 mg/L
3 h. Afteratthis,
3 h. we
After this, we
recorded norecorded
changes
no changes 2+ in the Ca 2+ concentration in GGBS-CH-S4. The initial rapid increase and the subsequent
in the Ca concentration in GGBS-CH-S4. The initial rapid increase and the subsequent decrease
decrease
of the Caof 2+ the Ca2+ concentration
concentration represent,represent, respectively,
respectively, the dissolution
the dissolution of Ca2+offromCa2+ from the GGBS
the GGBS and and
the
the ion’s
ion’s consumption
consumption via the via the formation
formation of alkali-activation
of alkali-activation products. The products.
constant The Ca2+ constant Ca2+
concentration
concentration
that we recorded that we3recorded
after h resultedafterfrom3 the
h resulted
balance from
betweenthe the
balance
GGBSbetween the process
dissolution GGBS dissolution
on the one
process on the one
2+ hand and the Ca 2+ consumption via the precipitation
hand and the Ca consumption via the precipitation of alkali-activation products on the other. of alkali-activation products
on the other. 2 − 2 −
The S2 O3 and SO4 concentrations were measured to represent the anions’ dissolution form
The S2O32−and
the GGBS-S4 andGGBS-CH-S4
SO42− concentrationssamples.were measuredatorelatively
In GGBS-S4, representsmallthe anions’
amountdissolution form
of anions were
the GGBS-S4
dissolved fromand GGBS, GGBS-CH-S4
and this amount samples.was In GGBS-S4,
constant for 25a h.
relatively
Conversely, small amount
in the of anions
GGBS-CH-S4 were
sample,
dissolved 2 − from GGBS, 2 − and this amount was constant for 25
the S2 O3 and SO4 dissolution rapidly increased after 7 h. Most importantly, in contrast to the h. Conversely, in the GGBS-CH-S4
Ssample,
2 O3
the S2O
2− (0~0.5 32− and SO42− dissolution
mg/L) and SO4 2− (1~2 mg/L) rapidly increased after
concentrations, we7observed
h. Most importantly, in contrast
that a significantly to
greater
the S2O3of(0~0.5
amount
2− mg/L)(S)
total sulfur and(7~13SO4mg/L)
2− (1~2 mg/L) concentrations,
was dissolved within 5 we observed
h (the that a removal
rapid Cr(VI) significantly greater
duration for
amount of total sulfur (S) (7~13 mg/L) was dissolved within
GGBS-CH-S2, in Figure 2) in GGBS-CH-S4 samples than in GGBS-S4. Previous XAS study found that5 h (the rapid Cr(VI) removal duration
for S
the GGBS-CH-S2,
in GGBS exists inmostly
Figurein 2)the
in GGBS-CH-S4
form of S2− with samples
a minor than in GGBS-S4.
amount of SO4 2−Previous
[31]. OurXAS study
results found
therefore
that the S in GGBS exists mostly in the form of S 2− with a minor amount of SO42− [31]. Our results
suggested that GGBS samples in Ca(OH)2 solution the significant Cr(VI)-concentration decrease in
therefore
the solution suggested
phase was that GGBS samples
primarily caused byinthe Ca(OH) 2 solution the significant Cr(VI)-concentration
fact that the reducing reaction between the dissolved
decrease
ions (S2− )inacted
the solution
as a reducing phase agent
was primarily caused
for the Cr(VI) in by
thethe fact that
solution, as the
hasreducing
previously reaction between
been found by
the dissolved ions (S 2−) acted as a reducing agent for the Cr(VI) in the solution, as has previously
others [3,11]. We concluded that the dissolved ions from the GGBS activated in Ca(OH)2 solution play
abeen found
critical roleby others
in the [3,11].transformation
reductive We concludedofthat the into
Cr(VI) dissolved
Cr(III) in ions from the GGBS activated in
solution.
Ca(OH)2 solution play a critical role in the reductive transformation of Cr(VI) into Cr(III) in solution.
3.3. Cr(VI) Adsorption onto GGBS-PC Samples
3.3. Cr(VI) Adsorption onto GGBS-PC Samples
We measured the changes in the Cr(VI) concentration for the S3 samples at 0, 120, and 150 h in order
We measured
to further evaluate the thecapacity
changesofinGGBS the Cr(VI) concentration
to immobilize Cr(VI) forin athe S3 samples
Ca(OH) at 0, 120, and 150 h in
2 solution (Figure 4). For each
order to further evaluate the capacity of GGBS to immobilize Cr(VI)
specimen, the initial Cr(VI) concentrations varied from 0.5 to 100 mg/L. After 120 h of agitation, in a Ca(OH) 2 solution we(Figure
observed4).
Fortrace
no eachofspecimen, the initial Cr(VI)
Cr(VI) in GGBS-CH-S3 concentrations
that had varied fromof0.5
a Cr(VI) concentration up to
to 100 mg/L.which
10 mg/L, After indicated
120 h of
agitation, we observed no trace of Cr(VI) in GGBS-CH-S3 that had a Cr(VI) concentration of up to 10
mg/L, which indicated the complete removal of Cr(VI) from that solution. The GGBS specimen
containing 50 mg/L of Cr(VI) had approximately 0.7 mg/L of Cr(VI) after 120 h. The Cr(VI)
Materials 2018, 11, 11 8 of 17

Materials 2018, 11, 11 8 of 17

the complete removal of Cr(VI) from that solution. The GGBS specimen containing 50 mg/L of Cr(VI)
concentration
had approximatelyin the0.7
specimen
mg/L ofcontaining
Cr(VI) after100 120mg/L
h. TheofCr(VI)
Cr(VI)concentration
was reduced in tothe
8.6 specimen
mg/L. Furthermore,
containing
for mg/L
100 all samples,
of Cr(VI)wewasobserved
reducedno further
to 8.6 mg/L. change in the Cr(VI)
Furthermore, concentrations
for all samples, even no
we observed after 150 change
further h. It is
possible
in that concentrations
the Cr(VI) all GGBS’s sorption sites150
even after forh.Cr(VI) are saturated
It is possible that allat this concentration,
GGBS’s sorption sites whichfor Cr(VI)would
are
explain why
saturated no concentration,
at this further removal of Cr(VI)
which wouldfrom thewhy
explain solution was observed.
no further removal of It Cr(VI)
is also from
possible that the
the solution
saturation
was of GGBS’s
observed. It is alsosurface
possiblesites
thathinders any further
the saturation reduction
of GGBS’s surfaceofsites
Cr(VI) to Cr(III).
hinders Basedreduction
any further on there
being
of no to
Cr(VI) further changes
Cr(III). Based on in the
thereCrbeing
concentrations
no further after
changesoverin150
theh,Cra concentrations
pseudo-equilibrium after overcondition
150 h,
awas attained under our
pseudo-equilibrium experimental
condition condition.
was attained under Weourdoexperimental
not have the condition.
experimental We data
do not to have
measurethe
activated complexes
experimental at the transition
data to measure state for the
activated complexes overall
at the Cr transformation
transition reaction
state for the overall mechanisms,
Cr transformation
and as amechanisms,
reaction consequence andof as
this we cannot calculate
a consequence of this wethe pseudo-equilibrium
cannot constant for theconstant
calculate the pseudo-equilibrium reactions.
for
Further
the detailed
reactions. Furtherwork is needed
detailed work isto confirm
needed the exact
to confirm mechanism
the exact mechanismat play
at playduring
duringsurface
surface site
saturation,along
saturation, alongwithwithitsits impact
impact on on Cr(VI)
Cr(VI) uptake
uptake and reduction
and reduction by GGBS.
by GGBS. Additionally,
Additionally, GGBS inGGBS in
distilled
distilled
water water (GGBS-S1)
(GGBS-S1) exhibited aexhibited a limited
limited capacity forcapacity for Cr(VI)
Cr(VI) removal removal
within 120 h:within 120 h: per
0.025 mg/L 0.025 mg/L
gram of
per gram
GGBS. of GGBS.
Related to thisRelated
result isto this
the factresult is the fact
that GGBS’s that GGBS’s
capacity capacity
to immobilize andtostabilize
immobilize
Cr(VI) and stabilize2
in Ca(OH)
Cr(VI) inwas
solution Ca(OH) 2 solution
particularly was
large: particularly 2.3
approximately large: approximately
mg/L 2.3 mg/L
per gram of GGBS. This per gram
result of GGBS.
indicates thatThis
the
result indicates
alkaline conditionsthat the alkaline
improve conditions
the dissolution improve
of GGBS. the dissolution
This results of GGBS.
in the enhancement of This resultscapacity
the GGBS’s in the
enhancement
to remove Cr(VI), of in
the GGBS’s
addition capacity
to its potential totoremove
do so inCr(VI),
distilledin addition
water to itsAll
conditions. potential
the reactionto do so in
products
distilled
of water conditions.
the experimental samples were All the reaction
further analyzedproducts
using XRDof thein experimental
order to examine samples were surface
the possible further
analyzed using
precipitate phase.XRD in order to examine the possible surface precipitate phase.

0.5mg/L
Concentration of Cr(VI) in Solution (mg/L)

100 1mg/L
3mg/L
5mg/L
50 10mg/L
50mg/L
100mg/L

10

0 20 40 60 80 100 120 140 160

Time (Hours)

Figure 4. A comparison of the Cr(VI) concentrations in the S3 specimens’ solutions at 0, 120, and
Figure 4. A comparison of the Cr(VI) concentrations in the S3 specimens’ solutions at 0, 120, and 150 h
150 h of reaction.
of reaction.

3.4. XRD Characterization of the Reacted Phases

3.4. XRD Characterization of the Reacted Phases
Figure 5 shows the diffraction patterns of raw materials and of the S1 samples, measured using
Figure 5 shows the diffraction patterns of raw materials and of the S1 samples, measured
XRD. We recorded no diffraction peaks for K2Cr2O7 in S1 specimens, indicating the complete
using XRD. We recorded no diffraction peaks for K2 Cr2 O7 in S1 specimens, indicating the
dissolution of K2Cr2O7. PC-S1 exhibited very strong crystalline peaks for the Ca(OH)2 formed from
complete dissolution of K2 Cr2 O7 . PC-S1 exhibited very strong crystalline peaks for the Ca(OH)2
the PC hydration, but such crystalline peaks were absent in GGBS-S1’s XRD pattern, indicating that
formed from the PC hydration, but such crystalline peaks were absent in GGBS-S1’s XRD pattern,
here GGBS was mostly amorphous and had not been effectively activated under a pH of circa −11.5
indicating that here GGBS was mostly amorphous and had not been effectively activated under
(Figure 1a). In the XRD patterns of PC-S1 and GGBS-CH-S1 samples, we observed a formation of
a pH of circa −11.5 (Figure 1a). In the XRD patterns of PC-S1 and GGBS-CH-S1 samples,
calcium aluminate hemicarbonate (hereafter referred to as hemicarbonate,
we observed a formation of calcium aluminate hemicarbonate (hereafter referred to as hemicarbonate,
2Ca4[Al(OH)6]2(CO3)0.5OH·5.5H2O) and calcium aluminate monocarbonate (hereafter referred to as
2Ca4 [Al(OH)6 ]2 (CO3 )0.5 OH·5.5H2 O) and calcium aluminate monocarbonate (hereafter referred to as
monocarbonate, 3CaO·Al2O3·CaCO3·10.68H2O) while only the PC-S1 sample formed ettringite.
monocarbonate, 3CaO·Al2 O3 ·CaCO3 ·10.68H2 O) while only the PC-S1 sample formed ettringite.
Materials 2018,11,
Materials2018, 11, 11
11 9 of
9 of 1717
Materials 2018, 11, 11 9 of 17

■ : Ca(OH)2 ◆ : PC Clinkers
◇
■ : Ca(OH)
Ettringite ● : C-S-H
2 ◆ : PC Clinkers
◇ : Ettringite
▽ Hemicarbonate ▲ : Monocarbonate
● : C-S-H
▽ : Hemicarbonate ▲ : Monocarbonate
◆
◆ ◆◆
■ ◆
◆ ◆◆ ◆
◆
■

(A.U)
◆ ◆ ◆ ◆◆ ◆

(A.U)
◆ ◆ ◆◆
▲
◇ PC-Raw
Intensity
◆
▲ ◇ ◇
●
◇▽ ◇ PC-Raw
Intensity ◇ ◇
●
◆
▽ ■◇ ■
▽▲ ■ ■ ■ ■ ■
PC-S1
■● ■ ■PC-S1
▽▲ ■
● ■
GGBS-CH-S1
GGBS-CH-S1
GGBS-S1
GGBS-S1
GGBS-Raw
GGBS-Raw
10 15 20 25 30 35 40 45 50 55 60
10 15 20 25 30 35 40 45 50 55 60
2Theta
2Theta
Figure 5. XRD patterns for raw materials (PC and GGBS), PC-S1, GGBS-CH-S1, and GGBS-S1.
Figure5.5.XRD
Figure XRDpatterns
patternsfor
forraw
raw materials
materials (PC
(PC and
andGGBS),
GGBS),PC-S1,
PC-S1,GGBS-CH-S1,
GGBS-CH-S1,and GGBS-S1.
and GGBS-S1.
Figure 6a shows the XRD patterns for GGBS-CH-S3 with Cr(VI) concentrations ranging
Figure
Figure 6a 6a shows
shows the the
XRDXRD patterns
patterns for GGBS-CH-S3
for GGBS-CH-S3 with Cr(VI)
with Cr(VI) concentrations
concentrations ranging ranging
between
between 0.5 and 100 mg/L. Regardless of the initial Cr(VI) concentration, all the specimens
0.5between
and 100 0.5 andRegardless
mg/L. 100 mg/L. of Regardless
the initial of the
Cr(VI) initial Cr(VI)
concentration, all concentration,
the specimens all the specimens
exhibited very strong
exhibited very strong crystalline peaks due to the Ca(OH)2. In specimens with a Cr(VI)
exhibited
crystalline very
peaksofdue strong
to crystalline
the Ca(OH)26b), peaks due
. In specimens to the
with aand Ca(OH)
Cr(VI) 2. In specimens with a Cr(VI)
concentration of 0.5 mg/L (Figure in 6b),
concentration
concentrationand of 0.5
0.5 mg/L
mg/L (Figure 6b), hemicarbonate
(Figure formed, hemicarbonate and monocarbonate
monocarbonate formed,
formed, asobserved
as observed in
hemicarbonate
GGBS-CH-S1. monocarbonate as observed in GGBS-CH-S1. No ettringite or monosulfate
GGBS-CH-S1. No No ettringite
ettringite or or monosulfate
monosulfate phases phases werewere found
found in in any
any of
of the
the S3S3 specimens.
specimens. InIn
phases were
contrast, in found
the in any ofwith
GGBS-CH the S350 specimens.
and 100 mg/LIn contrast,
of Cr(VI), inthe
thehemi-
GGBS-CH or with 50 and 100
monocarbonate XRD mg/L
peakof
contrast, in the GGBS-CH with 50 and 100 mg/L of Cr(VI), the hemi- or monocarbonate XRD peak
Cr(VI), the
intensity hemi- or monocarbonate XRD peak intensity was very low. The addition of 100 calciumof
mg/L
intensitywas wasvery
very low.
low. TheThe addition
addition of of 100
100 mg/L
mg/L of of Cr(VI)
Cr(VI) encouraged
encouraged the the formation
formationof ofcalcium
Cr(VI) encouraged
aluminum the formation of calcium aluminum Cr oxide hydrates ((Ca Al O (CrO 4 )·9H2of O),
aluminumCr Croxide
oxide hydrates
hydrates ((Ca((Ca44Al
Al22O
O33(CrO
(CrO44)·9H O), CAC-9),
)·9H22O), CAC-9), which which suggests
suggests4thatthat2 the
3 amount
the amount of
CAC-9),
Cr(VI) which suggests that the amount of Cr(VI) influences the 2formation
activation.ofPrevious
GGBS products under
Cr(VI)influences
influencesthe theformation
formation of of GGBS
GGBS products under Ca(OH)
products under Ca(OH) studies have
2 activation. Previous studies have
Ca(OH)
shown activation.
shown2 that
that Cr(VI)
Previous
Cr(VI) can can be
studies
be immobilized
have
immobilized by
shown
by pure
that
pure PC
Cr(VI) can
PC hydrates, be
hydrates, such immobilized
such as by
as ettringite, pure
ettringite, mono-, PC hydrates,
mono-, and and
such as ettringite,
hemicarbonate mono-,
[11,18]. and hemicarbonate
Additionally, high-surface [11,18].
area Additionally,
hydration high-surface
by-products,
hemicarbonate [11,18]. Additionally, high-surface area hydration by-products, such as C-S-H from such asarea hydration
C-S-H from
by-products,
PCPChydration
hydration such as C-S-H
reactions,
reactions, can from PC
can remove
remove hydration
Cr(VI)
Cr(VI) from reactions,
from solutions
solutions viacan
via remove
physical
physical Cr(VI) from
adsorption
adsorption solutions
[3].Based
[3]. Based ononour
our via
physical
findings, adsorption
i.e., the [3].
rapid Based
Cr(VI)on our findings,
reduction and i.e., the rapid
immobilization Cr(VI)
findings, i.e., the rapid Cr(VI) reduction and immobilization through GGBS under alkaline reduction
through GGBS and immobilization
under alkaline
through
conditionsGGBS
conditions under alkaline
maintained
maintained by conditions
by Ca(OH)
Ca(OH)22,, we maintained
we concludebythat
conclude Ca(OH)
that the2 ,by-products
the we conclude produced
by-products that the by-products
produced by the
by the
alkali-activated
produced by the GGBS
alkali-activated GGBScan cancontribute
alkali-activated GGBS
contribute to bothcontribute
to can the immobilization
to both the
immobilization of Cr(VI)
of Cr(VI) ions
ionsvia
immobilization viathe
the complexation
ofcomplexation
Cr(VI) ions via
with
the dissolved
dissolved anions
complexation
with resulting
with dissolved
anions from
from the
resulting anions GGBS from
resulting hydration
the GGBSreaction,
reaction, and the
hydration
and the Cr(VI) reduction
reaction,
Cr(VI) reduction
and toto
the Cr(VI)
Cr(III).
reduction
Cr(III). to Cr(III).

(a)
(a) ■■ ■ : Ca(OH)22 (b) ❖:CAC-9
(b) ❖: CAC-9
■■ ● : C-S-H ▽ ::Hemi
▽ Hemi
▲ ::Mono
▲ Mono

■ Cr(Ⅵ)concentration
Cr(Ⅵ) concentration
■■ ■ ❖
❖
❖
❖ ■ [mg/L]
[mg/L]
❖
❖ ●
●
▲
▲
100
100
(A.U)
Intensity (A.U)

❖
❖ ❖
❖ ❖
❖▽▽▲▲
50
50
▲
Intensity

▲
▽
▽
10
10
▲
▲
▽
▽
▲
55
▲
▽
▽
▲▲ 33
▽
▽

▽ 11
▽
▲
▲
0.5
0.5
10 15
10 15 20
20 25
25 30
30 35
35 40
40 45
45 50
50 55
55 60
60 10
10 15
15
2Theta
2Theta
Figure6.6.(a)
Figure (a)XRD
XRDpatterns
patterns of
of GGBS-CH-S3
GGBS-CH-S3 in in solutions
solutions containing
containing 0.5
0.5 to
to 100
100 mg/L
mg/Lof
ofCr(VI)
Cr(VI)and
and(b)
(b)
Figure 6. (a)
themagnified XRD
magnifiedXRD patterns
XRDpatternsof
patterns in GGBS-CH-S3
in the
the range
range of in
of 99 to solutions
to 15 containing
15 degrees. 0.5 to 100 mg/L of Cr(VI) and
the degrees.
(b) the magnified XRD patterns in the range of 9 to 15 degrees.
 Materials 2018, 11, 11 10 of 17
Materials 2018, 11, 11 10 of 17

3.5.Chromium
3.5. ChromiumSpeciation
Speciationvia
viaSTXM
STXM

InIn exhibiting
exhibiting different
differentCrCr oxidation
oxidation states, Cr(III)
states, and Cr(VI)
Cr(III) can be readily
and Cr(VI) can be resolvedreadily by comparing
resolved by
the sample spectra with suitable reference spectra that
comparing the sample spectra with suitable reference spectra that are taken at are taken at the Cr L -edge. This
2,3 the Cr L2,3-edge.is because
This
isnear edge X-ray
because near edgeabsorption
X-ray fine structure
absorption (NEXAFS)
fine structureis (NEXAFS)
highly sensitive
is highlyto both the oxidation
sensitive to bothstates
the
and the geometry of Cr’s first coordination shell
oxidation states and the geometry of Cr’s first coordination shell [32].[32].
Figure 77 presents
Figure presents the theX-ray
X-rayabsorption
absorption image of the
image of GGBS-CH-S3
the GGBS-CH-S3 sample taken below
sample taken Cr belowL3 -edge
Cr
at a photon energy of 560 eV, and the image contrast maps of Cr and
L3-edge at a photon energy of 560 eV, and the image contrast maps of Cr and O. Cr appeared to be O. Cr appeared to be localized
at the edge
localized of the
at the edge GGBS particles,
of the whereas whereas
GGBS particles, O was equally
O wasdistributed in the GGBS.
equally distributed in theInGGBS.
the CrIn image
the
contrast map, Area 1, as illustrated in Figure 8a, was
Cr image contrast map, Area 1, as illustrated in Figure 8a, was selected for2,3selected for the Cr L -edge NEXAFS
the Cr L2,3-edge NEXAFS analysis.
We choseWe
analysis. pure
choseK2 Cr 2 O7 K
pure as a reference for Cr(VI) spectra and used a series of Gaussian peaks to
2Cr2O7 as a reference for Cr(VI) spectra and used a series of Gaussian
fit the spectra. In Figure
peaks to fit the spectra. In Figure 8b, K28b,
Cr2K O27Crexhibits four distinct X-ray absorption features in the Cr
2O7 exhibits four distinct X-ray absorption features in the

CrL2,3L-edge spectra, indicated as a (~578.5 eV), a (~580.9 eV), b2 (~587.7

2,3-edge spectra, indicated as 2a2 (~578.5 eV), a33 (~580.9 eV), b2
(~587.7 eV), eV), and and bb33 (~589.5
(~589.5eV),eV),
where the peaks defined as a and b are induced, respectively, by
where the peaks defined as a3 and b2 are induced, respectively, by the Cr L3, and L2 absorptions. The
3 2 the Cr L 3 , and L 2 absorptions.
The additional
additional peakspeaksa2 and a2 and b2 result
b2 result fromfromthe the crystal
crystal field
field produced
produced byby thetheligands
ligandswith withdifferent
different
electronegativities,which
electronegativities, whichare arenon-linearly
non-linearlyrelated relatedtotothethemagnitude
magnitudeofofthe thecrystal-field
crystal-fieldparameter
parameter
(10Dq) [33,34]. Since the absorption shoulder features of Cr(VI) on
(10Dq) [33,34]. Since the absorption shoulder features of Cr(VI) on the low energy side are verythe low energy side are veryclose
close
to Cr(III)’s main absorption [32], the absorption peaks of a and b
to Cr(III)’s main absorption [32], the absorption peaks of a22 and b22are overlapping contributions are overlapping contributions
fromboth
from bothCr(VI)
Cr(VI)and and Cr(III).
Cr(III). TheThelowlow energy
energy sideside features
features a1 anda1 band b (the arrows in Figure 8b)
1 (the 1arrows in Figure 8b) solely
solely correspond to the octahedrally coordinated Cr(III),
correspond to the octahedrally coordinated Cr(III), which is in excellent agreement which is in excellent agreementwith withthe the
chromite(FeCr
chromite (FeCr2O 2O ) spectrum and Cr O3 [32,35,36].
4)4spectrum and Cr2O23 [32,35,36]. TheThe coexistence
coexistence of Cr
of Cr as both
as both Cr(VI)
Cr(VI) andand Cr(III)
Cr(III) is
is presumably due to the superposition of the partially-reduced Cr(VI)
presumably due to the superposition of the partially-reduced Cr(VI) and the physically adsorbed and the physically adsorbed
Cr(VI) onto
Cr(VI) onto the
the Ca(OH)
Ca(OH)22-activated
-activated GGBS.
GGBS.The Theremaining
remainingareas illustrated
areas illustrated in Figure
in Figure 8a show
8a showthe same
the
trends as Area 1, i.e., the coexistence of Cr(III) and Cr(VI). However,
same trends as Area 1, i.e., the coexistence of Cr(III) and Cr(VI). However, due to the lower Cr due to the lower Cr concentration,
the areas havethe
concentration, lower
areasX-ray
haveabsorption
lower X-ray intensities
absorption andintensities
a higher background
and a highernoise. background noise.

Figure 7. (a) Scanning transmission X-ray microscopy (STXM) image of GGBS-CH-S3 taken at 560
Figure 7. (a) Scanning transmission X-ray microscopy (STXM) image of GGBS-CH-S3 taken at 560 eV;
eV; (b) image contrast map of Cr; and (c) O for GGBS-CH-S3.
(b) image contrast map of Cr; and (c) O for GGBS-CH-S3.

The
TheOOK-edge
K-edgeNEXAFS
NEXAFSanalysis
analysisprovided
providedfurther
furtherevidence
evidencefor
forthe
thecoexistence
coexistenceofofCr(VI)
Cr(VI)andand
Cr(III) on the Ca(OH)
Cr(III) on the Ca(OH)2 -activated GGBS (GGBS-CH-S3), as shown in Figure 9. The O K-edgeNEXAFS
2 -activated GGBS (GGBS-CH-S3 ), as shown in Figure 9. The O K-edge NEXAFS
spectrum
spectrumofofthetheKK2Cr 2O7 exhibited intense pre-edge X-ray absorption peaks at 529.2 and 531.2 eV,
2 Cr2 O7 exhibited intense pre-edge X-ray absorption peaks at 529.2 and 531.2 eV,
and
and these can be attributed to
these can be attributed the electronic
to the electronic excitation
excitationfrom fromthe
theOO1s1score
corelevel
levelinto
into
thethe unoccupied
unoccupied t2g
t2g
andand
eg emolecular
g molecular orbitals,
orbitals, respectively
respectively [37].[37].
TheThe O K-edge
O K-edge NEXAFS
NEXAFS spectrum
spectrum of pureof Ca(OH)
pure Ca(OH)
2
2
differs
differs from the absorption features of 2CrO
− 4 2− [24]; in this study, the presence of crystalline Ca(OH)2
from the absorption features of CrO4 [24]; in this study, the presence of crystalline Ca(OH)2 was
was not observed
not observed in theinentire
the entire area
area of theof the specimen.
specimen. Area 1Area 1 and2 appeared
and Area Area 2 appeared
to contain to two
contain two
pre-edge,
pre-edge, low-intensity
low-intensity absorption absorption peaks
peaks arising arising
from CrOfrom2 − CrO4 . In addition to the peaks, the features
2−
4 . In addition to the peaks, the features indicated
indicated by arrows at ~533 eV in the Area 1
by arrows at ~533 eV in the Area 1 and Area 2 spectra correspondand Area 2 spectra correspond
to the to the O K-edge
O K-edge absorption peak of
absorption peak of Cr(III)-oxyhydroxide, which has previously
Cr(III)-oxyhydroxide, which has previously been observed in the spectrum of pyrite been observed in the(FeS
spectrum
2 ) of
reacting
pyrite (FeS2) reacting with Cr(VI) [35]. Area 3 did not exhibit any CrO4 or Cr(III)-oxyhydroxide2−

features because of the absence of any Cr in an oxidation state, as observed in Figure 7b.
Materials 2018, 11, 11 11 of 17

with Cr(VI) [35]. Area 3 did not exhibit any CrO4 2− or Cr(III)-oxyhydroxide features because of the
absence2018,
Materials of any Cr in an oxidation state, as observed in Figure 7b.
11, 11 11 of 17
Materials 2018, 11, 11 11 of 17

Figure 8. (a) Illustration of the analyzed areas in GGBS-CH-S3 for the Cr L2,3-edge NEXAFS in the Cr
Figure8.
Figure (a)Illustration
8. (a) Illustrationof
ofthe
the analyzed
analyzed areas
areas inin GGBS-CH-S3
GGBS-CH-S3 for the Cr L2,3
for the -edgeNEXAFS
2,3-edge NEXAFSin inthe
theCr
Cr
image contrast map (Figure 7b); (b) Cr L2,3-edge NEXAFS spectra for Area 1 in (a) and spectra from
imagecontrast
image contrastmap map(Figure
(Figure7b);
7b); (b) Cr LL2,3
(b) Cr -edgeNEXAFS
2,3-edge NEXAFSspectra
spectrafor
forArea
Area11inin(a)
(a)and
andspectra
spectrafrom
from
K2Cr2O7.
KK22Cr
Cr2O
2O7.7 .

Figure 9. (a) Illustration of the analyzed areas in GGBS-CH-S3 for the O K-edge NEXAFS; (b) O
Figure 9. (a) Illustration of the analyzed areas in GGBS-CH-S3 for the O K-edge NEXAFS; (b) O
Figure 9.
K-edge (a) Illustration
NEXAFS spectraof
forthe analyzed
each area inareas for K2Cr2O7. for the O K-edge NEXAFS; (b) O K-edge
in GGBS-CH-S3
(a) and
K-edge NEXAFS spectra for each area in (a) and for K2Cr2O7.
NEXAFS spectra for each area in (a) and for K2 Cr2 O7 .
3.6. Cr(VI) Dissolution and the Effects of Accelerated Carbonation
3.6. Cr(VI) Dissolution and the Effects of Accelerated Carbonation
A comparison of the dissolved Cr(VI) concentration from the specimens before and after
A comparison of the dissolved Cr(VI) concentration from the specimens before and after
accelerated carbonation is shown in Figure 10. At 56 days, all the specimens that had not been
accelerated carbonation is shown in Figure 10. At 56 days, all the specimens that had not been
exposed to atmospheric CO2 showed that relatively minimal amounts of Cr(VI) (i.e., less than the
exposed to atmospheric CO2 showed that relatively minimal amounts of Cr(VI) (i.e., less than the
detection limit of 0.03 mg/L) were released from the GGBS-PC paste, regardless of the replacement
detection limit of 0.03 mg/L) were released from the GGBS-PC paste, regardless of the replacement
Materials 2018, 11, 11 12 of 17

3.6. Cr(VI) Dissolution and the Effects of Accelerated Carbonation

A comparison of the dissolved Cr(VI) concentration from the specimens before and after
accelerated carbonation is shown in Figure 10. At 56 days, all the specimens that had not been
Materials 2018,
exposed 11, 11
to atmospheric CO2 showed that relatively minimal amounts of Cr(VI) (i.e., less than 12 of
the17
detection limit of 0.03 mg/L) were released from the GGBS-PC paste, regardless of the replacement
ratio. Conversely,
ratio. Conversely,in in thethe accelerated
accelerated carbonation
carbonation samples
samples the dissolution
the dissolution of Cr(VI)ofdepended
Cr(VI) depended
strongly
strongly on the GGBS replacement ratio for PC. In pure PC, 0.7 mg/L
on the GGBS replacement ratio for PC. In pure PC, 0.7 mg/L of Cr(VI) was leached, whereasof Cr(VI) was leached, whereas
only
only 0.1 mg/L of Cr(VI) was leached from the GGBS-PC paste with a 5 wt
0.1 mg/L of Cr(VI) was leached from the GGBS-PC paste with a 5 wt % GGBS replacement. The Cr(VI) % GGBS replacement.
The Cr(VI)
leached fromleached
the PC from thewas
paste PC paste
the sumwasofthethesum of the pre-added
pre-added Cr(VI)
Cr(VI) (0.5 mg/L) (0.5and
mg/L) and additional
additional Cr(VI)
Cr(VI) (~0.2 mg/L) contained in the PC clinker, as already observed in the dissolution
(~0.2 mg/L) contained in the PC clinker, as already observed in the dissolution experiment (Figure 1b). experiment
(Figure
For 1b). For
a sample with a sample with 50% of GGBS-replaced
50% of GGBS-replaced PC paste, we PC paste,awe
measure measure
very a very small
small amount amount
of Cr(VI) below of
Cr(VI)
the belowlimit
detection the ofdetection
0.03 mg/L.limit of solutions
Pore 0.03 mg/L. Pore solutions
extracted from PCextracted from PC
paste generally pastea pH
possess generally
in the
range of 12.4 to 13.5 [38]. Our work revealed that the alkalinity resulting from the naturallyfrom
possess a pH in the range of 12.4 to 13.5 [38]. Our work revealed that the alkalinity resulting formedthe
naturally formed Ca(OH) 2 in PC hydration was sufficient for GGBS activation and enhanced GGBS’s
Ca(OH)2 in PC hydration was sufficient for GGBS activation and enhanced GGBS’s capacity to reduce
capacity
and to reduce
immobilize and immobilize
Cr(VI), as observedCr(VI), as observed
in the solution in the
samples solution
(Figures 1bsamples (Figures 1b andin2).
and 2). Consequently, a
Consequently, in a realistic PC paste, replacing PC with a minimal amount
realistic PC paste, replacing PC with a minimal amount of GGBS had a significant impact of GGBS had a significant
on Cr(VI)
impact onand
reduction Cr(VI) reduction andeven
immobilization, immobilization,
with a 5 wt %even with a 5 wt % replacement.
replacement.

0.8

0.7
56d Curing
28d Carbonated
Concentration of Cr(VI) (mg/L)

0.6

0.5

0.4

0.3

0.2

0.1

0.0
0 5 10 50
Substitution Ratio of GGBS (wt.%)

Figure10.
Figure 10.Tank
Tank leaching
leaching test results
test results for the GGBS-PC
for the GGBS-PC paste
paste before before
and after and after carbonation
an accelerated an accelerated
test.
carbonation test.
Figure 11 shows the XRD patterns for the GGBS-PC paste, with different GGBS replacement ratios
Figurewith
and either 11 shows the XRD
or without patternsPeaks
carbonation. for the GGBS-PC
arising from paste, with different
monosulfate, Ca(OH)GGBS
2 , and replacement
C-S-H were
ratios and
detected foreither with or without
all GGBS-PC carbonation.
pastes without Peaks arising
carbonation. For the from
50 wtmonosulfate,
% GGBS-PC paste,Ca(OH) we2, noted
and C-S-H
very
weak peaks due to the presence of hydrotalcite (Mg6 Al2 (CO3 (OH)16 ·4H2 O)) and vaterite. This result
were detected for all GGBS-PC pastes without carbonation. For the 50 wt % GGBS-PC paste, we
noted very
indicated weak
that peaks due
replacing GGBS to in
thePCpresence
increasesof the
hydrotalcite
hydration(Mg 6Al2(CO
products’ 3(OH)16·4H2O))
susceptibility to and vaterite.
carbonation
Thisto
due result indicated
the reaction thatatmospheric
with replacing GGBS CO2 .inPrevious
PC increases
studies thehave
hydration products’that
also confirmed susceptibility
carbonation to
carbonation
can change the duePC to the reaction
paste’s with and
porosity atmospheric
that this COcan2. also
Previous
affectstudies have alsospecimen’s
the hardened confirmedCO that
2
carbonation can
permeability change the PC paste’s porosity and that this can also affect the hardened specimen’s
[39,40].
CO2 Monosulfate
permeability [39,40].
diffraction patterns were not detected in all samples with carbonation,
whichMonosulfate
suggests that diffraction
completepatterns were not
monosulfate detected in occurred
decomposition all samples withcarbonated
in the carbonation, which
samples.
suggests
In thatwe
addition, complete
detectedmonosulfate
the presencedecomposition occurred such
of CaCO3 polymorphs in theascarbonated samples. In
vaterite, aragonite, andaddition,
calcite;
we detected
these were thethe presence
main, of CaCO
crystalline, 3 polymorphs
accelerated such asproducts
carbonation vaterite,for
aragonite, and calcite;
all the specimens. Inthese were
particular,
thethe
as main, crystalline,
carbonation accelerated
progressed, carbonation
the peak products
intensities for all3 polymorphs
of the CaCO the specimens. In particular,
became stronger, asas has
the
carbonation progressed, the peak intensities of the CaCO3 polymorphs became stronger, as has been
previously found [41]. Presumably, the formation of aragonite is linked to the carbonated C-S-H
formed during PC clinker hydration, as has been observed in the young post-carbonation tricalcium
silicate paste [42]. In general, Ca(OH)2 and C-S-H are the main hydration products involved in the
carbonation of PC-based systems [43]; the carbonation occurs preferentially on Ca(OH)2, and
Materials 2018, 11, 11 13 of 17

been previously found [41]. Presumably, the formation of aragonite is linked to the carbonated
C-S-H formed during PC clinker hydration, as has been observed in the young post-carbonation
tricalcium silicate paste [42]. In general, Ca(OH)2 and C-S-H are the main hydration products
involved in the carbonation of PC-based systems [43]; the carbonation occurs preferentially on Ca(OH)2 ,
and precipitating
Materials 2018, 11, 11 CaCO3 polymorphs in the system [44,45]. However, carbonation in PC systems 13 of 17
that contain 75% amounts of GGBS mainly occurs on the C-S-H phases, which results in vaterite
formation [46]. C-S-H C-S-H formed
formed in GGBS-PC systems has a generallygenerally low
low Ca/Si
Ca/Si ratio
ratio because
because of the
Ca(OH)22 consumption compared
consumption compared to that of pure PC systems [47–50]. When C-S-H with aa low
[47–50]. When C-S-H with lowCa/Si
Ca/Si
ratio is exposed to CO22, both vaterite and aragonite, which are CaCO CaCO33 polymorphs, preferentially
precipitate [51].
[51].InInthethe present
present study,
study, after after carbonation,
carbonation, C-S-H C-S-H exhibited
exhibited a peak in
a peak intensity intensity
GGBS50C in
GGBS50C that was lower ◦(2θ = ~29°) than for those contained less GGBS. Furthermore,
that was lower (2θ = ~29 ) than for those contained less GGBS. Furthermore, we observed that the we observed
that
moretheGGBSmore GGBS
there was,there was, the
the higher washigher was the vaterite-related
the vaterite-related peak, whichpeak, which
formed dueformed due to the
to the carbonation
carbonation
of C-S-H with ofaC-S-H with aratio.
low Ca/Si low Ca/Si ratio.

◆ ◆
◆ ◎
■ ■ ◆
▣ ◆ ◎ ◇ ●
◇ ● ◇
◆
◆
GGBS50C

GGBS10C
After
GGBS5C Carbonation
Intensity (A.U)

GGBS0C

■ ■
✳
▣ ◆ ◆
● ✳ ◆ ■ GGBS50
✳ ● ◆■

GGBS10
Before
GGBS5 Carbonation

GGBS0

10 15 20 25 30 35 40 45 50 55 60
2Theta
Figure 11. XRD patterns of the GGBS-PC paste with GGBS replacement ratios of 0, 5, 10, and 50 wt %
Figure 11. XRD patterns of the GGBS-PC paste with GGBS replacement ratios of 0, 5, 10, and 50 wt %
before andafter
before and after
28 28 days
days of accelerated
of accelerated carbonation.
carbonation. ◆ : Vaterite,
: Vaterite, ♢: Aragonite,
♦: Aragonite, : Ca(OH)■:, }Ca(OH) 2,
: Calcite, 2
◎: Calcite, ●:
: C-S-H, C-S-H, ✳: Monosulfate,
: Monosulfate, and ⎕: Hydrotalcite.
and 2: Hydrotalcite.

Figure 12
Figure 12shows
showsthethe XRDXRDpeak intensity
peak of theof001
intensity thebasal
001reflection from Ca(OH)
basal reflection for the GGBS-PC
from2 Ca(OH) 2 for the
pastes before and after carbonation. We observed a very intense peak
GGBS-PC pastes before and after carbonation. We observed a very intense peak for crystalline for crystalline Ca(OH) 2 in
the pre-carbonation pure PC paste (GGBS0). Like anticipated, as GGBS’s
Ca(OH)2 in the pre-carbonation pure PC paste (GGBS0). Like anticipated, as GGBS’s replacement replacement ratio in PC
increased,
ratio in PCCa(OH) 2 ’s peak
increased, Ca(OH)intensity decreased due to the dilution of PC and to the consumption
2’s peak intensity decreased due to the dilution of PC and to the
of Ca(OH) by GGBS. After carbonation,
consumption of Ca(OH)2 by GGBS. After carbonation,
2 more Ca(OH) 2 decreased
more in the pure
Ca(OH)2 decreased inPC
thepaste
pure PC (GGBS0)
paste
than Ca(OH) in the GGBS-PC paste. As a result, all the specimens exhibited
(GGBS0) than2 Ca(OH)2 in the GGBS-PC paste. As a result, all the specimens exhibited similar similar Ca(OH) 2 peak
intensities,
Ca(OH) 2 peak which supported
intensities, whichthesupported
findings that C-S-H is that
the findings more readily
C-S-H carbonated
is more readilythan Ca(OH)than
carbonated 2 in
GGBS-PC paste because the carbonation of C-S-H depends on the
Ca(OH)2 in GGBS-PC paste because the carbonation of C-S-H depends on the2 initial Ca(OH)2initial Ca(OH) amount present
prior to carbonation
amount present prior [46].
to Moreover,
carbonationit can[46].beMoreover,
demonstrated it canthat
be the sufficient Ca(OH)
demonstrated that the2 insufficient
pure PC
paste can
Ca(OH) 2 inbuffer
pure PCC-S-Hpastecarbonation.
can buffer C-S-H Presumably, the lower
carbonation. Cr(VI) the
Presumably, dissolution in the
lower Cr(VI) carbonated
dissolution in
GGBS-PC resulted from a pre-carbonation reductive process caused by
the carbonated GGBS-PC resulted from a pre-carbonation reductive process caused by dissolved dissolved anions such as S2−
(which are
anions such continuously
as S2− (which released from GGBS released
are continuously during hydration)
from GGBS because
during S2−hydration)
can inhibitbecause
the oxidization
S2− can
of Cr(III) to Cr(VI) [18]. It should be noted that we do not have the direct
inhibit the oxidization of Cr(III) to Cr(VI) [18]. It should be noted that we do not have the direct evidence we need to
identify the exact surface-promoted reductive transformation of Cr(VI)
evidence we need to identify the exact surface-promoted reductive transformation of Cr(VI) to to Cr(III). It is likely that
Cr(III). It is likely that S2− released during the hydration reaction is responsible for the reducing
reactions in solution and that the reduced Cr species are physisorbed to hydrated materials. Equally,
we cannot rule out the possibility that physisorbed S2− located on hydrated-materials surfaces act as
reducing agents for Cr reductive transformations. Further detailed studies are needed to distinguish
the two different reaction pathways. Moreover, C-S-H formed in the GGBS-substituted PC system,
Materials 2018, 11, 11 14 of 17

S2− released during the hydration reaction is responsible for the reducing reactions in solution and
that the reduced Cr species are physisorbed to hydrated materials. Equally, we cannot rule out the
possibility that physisorbed S2− located on hydrated-materials surfaces act as reducing agents for
Cr reductive transformations. Further detailed studies are needed to distinguish the two different
reaction pathways. Moreover, C-S-H formed in the GGBS-substituted PC system, which has a longer
mean chain
Materials length
2018, 11, 11 [52], resulting in a larger surface area that can offer better surface adsorption,
14 of 17
inclusion, and physical entrapment of Cr(VI). Based on the XRD and Cr(VI) dissolution results before
GGBS cancarbonation,
and after greatly reduce weand immobilize
concluded thatCr(VI) viareplacing
partially dissolvedPC
anions
withinto
GGBSthecan
pore solutions
greatly andand
reduce via
physical
immobilize adsorption
Cr(VI) via onto C-S-H.anions
dissolved Even though
into theC-S-H and hydration
pore solutions and viaproducts
physicalare more susceptible
adsorption onto C-S-H. to
carbonation
Even thoughin the GGBS-PC
C-S-H system
and hydration than inare
products themore
puresusceptible
PC system,toCr(VI) was not
carbonation inreadily leached
the GGBS-PC from
system
the
thanGGBS-PC
in the pure paste after carbonation.
PC system, Cr(VI) was not readily leached from the GGBS-PC paste after carbonation.

Figure 12. Comparison

Comparison ofof the
the 001 basal reflection
001 basal reflection from
from Ca(OH)
Ca(OH)2 in GGBS-PC pastes before and after
Figure 12. 2 in GGBS-PC pastes before and after
an
an accelerated carbonation. The numbers followed by GGBS indicate the
accelerated carbonation. The numbers followed by GGBS indicate the replacement
replacement ratios
ratios of
of GGBS
GGBS
for PC.
for PC.

4.
4. Conclusions
Conclusions
We found that
We found that without
without alkali
alkali activation
activation GGBS
GGBS is is limited
limited inin its
its capacity
capacity toto reduce
reduce Cr(VI).
Cr(VI).
In contrast, the Ca(OH) 2-activated GGBS demonstrated a considerably greater capacity for reducing
In contrast, the Ca(OH)2 -activated GGBS demonstrated a considerably greater capacity for reducing
and
and immobilizing
immobilizing Cr(VI).
Cr(VI). Therefore,
Therefore, we
we conclude
conclude that
that alkalinity
alkalinity isis one
one of the dominating
of the dominating factors
factors in
in
determining the extent and mechanisms of reductive Cr(VI) dissolution and uptake
determining the extent and mechanisms of reductive Cr(VI) dissolution and uptake reactions for GGBS. reactions for
GGBS. Furthermore,
Furthermore, we foundweexperimental
found experimental
evidenceevidence thatremoval
that Cr(VI) Cr(VI) from
removal the from thewas
solution solution
due towas
its
due to its reactions with the dissolved anions, 2such
− as S 2−, resulting from GGBS being under high
reactions with the dissolved anions, such as S , resulting from GGBS being under high alkaline
alkaline
conditions. conditions. X-ray absorption
X-ray absorption spectroscopyspectroscopy datathe
data identified identified the of
coexistence coexistence
Cr(III) andofCr(VI)
Cr(III)
onand
the
Cr(VI) on the Ca(OH) 2-activated GGBS. This partial reduction of Cr(VI) to Cr(III) on the surface of
Ca(OH)2 -activated GGBS. This partial reduction of Cr(VI) to Cr(III) on the surface of GGBS suggests
GGBS suggests
the presence of a the presence of a surface-site-activated
surface-site-activated reductive mechanism. reductive
The Cr(VI)mechanism.
we observed Theon Cr(VI) we
GGBS like
observed
results from on physically
GGBS like adsorbed
results from physically
Cr(VI) adsorbed
on hydration Cr(VI) on hydration products.
products.
Our
Our accelerated-carbonation experimental data indicate that
accelerated-carbonation experimental data indicate that the
the hydration
hydration products
products formed
formed in in
the GGBS-PC paste, such as C-S-H, possessed a greater resistance to Cr(VI) leaching
the GGBS-PC paste, such as C-S-H, possessed a greater resistance to Cr(VI) leaching than those formed than those
formed
in pure PC.in pure PC. Our
Our study study that
suggests suggests thatused
even the evenofthe
5% used of 5% weight-based
weight-based partial GGBSpartial GGBS
replacement
replacement for PC-based materials would be a suitable method for reducing
for PC-based materials would be a suitable method for reducing and immobilizing Cr(VI) without and immobilizing
Cr(VI)
violating without violating and
the mechanical the physical
mechanical and physical
integrity integrity
of PC [53]. of PC [53].
We hypothesize We
that hypothesize
Ca(OH) that
2 generated
Ca(OH) 2 generated during the early stage of PC hydration reactions promotes the dissolution and
during the early stage of PC hydration reactions promotes the dissolution and activation of GGBS,
activation
which of GGBS,
results which results
in an enhanced inreduction.
Cr(VI) an enhanced Cr(VI)be
It should reduction.
noted thatIt observing
should be Snoted that observing
2− concentrations in
Sthe concentrations
2− in the dissolved solution will provide useful insights for
dissolved solution will provide useful insights for determining and identifying further detailed determining and
identifying further detailed reaction pathways. Future studies that use additional information to
research similar experimental conditions are necessary to identify Cr(VI)’s exact reductive
transformation pathways in hydrated cementitious materials. In addition, in order to understand the
Cr(VI) reductive mechanism in GGBS, further detailed mechanistic and molecular research is
needed to identify the exact state of physisorbed Cr(VI) on GGBS’s hydration products, how
Materials 2018, 11, 11 15 of 17

reaction pathways. Future studies that use additional information to research similar experimental
conditions are necessary to identify Cr(VI)’s exact reductive transformation pathways in hydrated
cementitious materials. In addition, in order to understand the Cr(VI) reductive mechanism in GGBS,
further detailed mechanistic and molecular research is needed to identify the exact state of physisorbed
Cr(VI) on GGBS’s hydration products, how different types and amounts of GGBS alkaline activators
influence Cr(VI) reduction and immobilization, and the factors that control Cr(VI) reduction pathways
in carbonated GGBS-PC materials.

Acknowledgments: This research was supported by the Basic Science Research Program through the
National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning
(NRF-2016R1C1B1014179).
Author Contributions: Sungchul Bae designed the experiments and wrote most of the paper; Juyoung Ha
contributed to the analysis of the entire experiments and writing of the paper; Fumino Hikaru contributed to
the preparation the experiments and the analysis of the data of Cr reduction and XRD; Manabu Kanematsu,
Chiaki Yoshizawa, and Takafumi Noguchi contributed to the analysis of the data and the writing of the paper;
Youngsang Yu contributed the analysis of the STXM data and the writing of the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Frías, M.; de Rojas, M.I.S.; García, N.; Luxán, M.P. Contribution of toxic elements: Hexavalent chromium in
materials used in the manufacture of cement. Cem. Concr. Res. 1994, 24, 533–541. [CrossRef]
2. Klemm, W.A. Hexavalent Chromium in Portland-Cement. Cem. Concr. Aggreg. 1994, 16, 43–47.
3. Macphee, D.E.; Glasser, F.P. Immobilization Science of Cement Systems. MRS Bull. 1993, 18, 66–71. [CrossRef]
4. Costa, M. Potential hazards of hexavalent chromate in our drinking water. Toxicol. Appl. Pharmacol. 2003,
188, 1–5. [CrossRef]
5. Kjuus, H.; Lenvik, K.; Kjærheim, K.; Austad, J. Epidemiological Assessment of the Occurrence of Allergic Dermatitis
in Workers in the Construction Industry Related to the Content of Cr (VI) in Cement; National Institute of
Occupational Health: Ahmedabad, India, 2003.
6. Dresel, P.E.; Qafoku, N.; McKinley, J.P.; Fruchter, J.S.; Ainsworth, C.C.; Liu, C.; Ilton, E.; Phillips, J. Geochemical
Characterization of Chromate Contamination in the 100 Area Vadose Zone at the Hanford Site; The US Department
of Energy: Washington, DC, USA, 2008.
7. Chen, Q.Y.; Tyrer, M.; Hills, C.D.; Yang, X.M.; Carey, P. Immobilisation of heavy metal in cement-based
solidification/stabilisation: A review. Waste Manag. 2009, 29, 390–403. [CrossRef] [PubMed]
8. Rose, J.; Benard, A.; Susini, J.; Borschneck, D.; Hazemann, J.L.; Cheylan, P.; Vichot, A.;
Bottero, J.Y. First insights of Cr speciation in leached Portland cement using X-ray spectromicroscopy.
Environ. Sci. Technol. 2003, 37, 4864–4870. [CrossRef] [PubMed]
9. Malviya, R.; Chaudhary, R. Factors affecting hazardous waste solidification/stabilization: A review.
J. Hazard. Mater. 2006, 137, 267–276. [CrossRef] [PubMed]
10. Pera, J.; Thevenin, G.; Chabannet, M. Design of a novel system allowing the selection of an adequate binder
for solidification/stabilization of wastes. Cem. Concr. Res. 1997, 27, 1533–1542. [CrossRef]
11. Glasser, F.P. Fundamental aspects of cement solidification and stabilisation. J. Hazard. Mater. 1997, 52, 151–170.
[CrossRef]
12. Zamorani, E.; Sheikh, I.A.; Serrini, G. Physical properties measurements and leaching behavior of chromium
compounds solidified in a cement matrix. Nucl. Chem. Waste Manag. 1988, 8, 239–245. [CrossRef]
13. Poellmann, H.; Auer, S.; Kuzel, H.J.; Wenda, R. Solid solution of ettringites. Cem. Concr. Res. 1993, 23, 422–430.
[CrossRef]
14. Gougar, M.L.D.; Scheetz, B.E.; Roy, D.M. Ettringite and C-S-H Portland cement phases for waste ion
immobilization: A review. Waste Manag. 1996, 16, 295–303. [CrossRef]
15. Wieczorek-Ciurowa, K.; Fela, K.; Kozak, A.J. Chromium(III)-ettringite formation and its thermal stability.
J. Therm. Anal. Calorim. 2001, 65, 655–660. [CrossRef]
16. Chrysochoou, M.; Dermatas, D. Evaluation of ettringite and hydrocalumite formation for heavy metal
immobilization: Literature review and experimental study. J. Hazard. Mater. 2006, 136, 20–33. [CrossRef]
[PubMed]
Materials 2018, 11, 11 16 of 17

17. Perkins, R.B.; Palmer, C.D. Solubility of Ca-6[Al(OH)6 ]2(CrO4 )3 center dot 26H2 O, the chromate analog of
ettringite; 5–75 ◦ C. Appl. Geochem. 2000, 15, 1203–1218. [CrossRef]
18. Macias, A.; Kindness, A.; Glasser, F.P. Impact of carbon dioxide on the immobilization potential of cemented
wastes: Chromium. Cem. Concr. Res. 1997, 27, 215–225. [CrossRef]
19. Wazne, M.; Jagupilla, S.C.; Moon, D.H.; Christodoulatos, C.; Koutsospyros, A. Leaching Mechanisms of
Cr(VI) from Chromite Ore Processing Residue. J. Environ. Qual. 2008, 37, 2125–2134. [CrossRef] [PubMed]
20. Richardson, I.G.; Groves, G.W. Microstructure and microanalysis of hardened cement pastes involving
ground granulated blast-furnace slag. J. Mater. Sci. 1992, 27, 6204–6212. [CrossRef]
21. Kilcoyne, A.L.D.; Tyliszczak, T.; Steele, W.F.; Fakra, S.; Hitchcock, P.; Franck, K.; Anderson, E.; Harteneck, B.;
Rightor, E.G.; Mitchell, G.E.; et al. Interferometer-controlled scanning transmission X-ray microscopes at the
Advanced Light Source. J. Synchrotron Radiat. 2003, 10, 125–136. [CrossRef] [PubMed]
22. Hitchcock, A.P. aXis 2000—Analysis of X-ray Images and Spectra. Available online: http://unicorn.mcmaster.
ca/aXis2000.html (accessed on 21 December 2017).
23. Bae, S.; Taylor, R.; Hernández-Cruz, D.; Yoon, S.; Kilcoyne, D.; Monteiro, P.J.M. Soft X-ray Spectromicroscopic
Investigation of Synthetic C-S-H and C3S Hydration Products. J. Am. Ceram. Soc. 2015, 98, 2914–2920.
[CrossRef]
24. Bae, S.; Taylor, R.; Shapiro, D.; Denes, P.; Joseph, J.; Celestre, R.; Marchesini, S.; Padmore, H.; Tyliszczak, T.;
Warwick, T.; et al. Soft X-ray Ptychographic Imaging and Morphological Quantification of Calcium Silicate
Hydrates (C-S-H). J. Am. Ceram. Soc. 2015, 98, 4090–4095. [CrossRef]
25. Bae, S.; Taylor, R.; Kilcoyne, D.; Moon, J.; Monteiro, P.J.M. Effects of Incorporating High-Volume Fly Ash
into Tricalcium Silicate on the Degree of Silicate Polymerization and Aluminum Substitution for Silicon in
Calcium Silicate Hydrate. Materials 2017, 10, 131. [CrossRef] [PubMed]
26. Bae, S.; Kanematsu, M.; Hernandez-Cruz, D.; Moon, J.; Kilcoyne, D.; Monteiro, P.J.M. In Situ Soft X-ray
Spectromicroscopy of Early Tricalcium Silicate Hydration. Materials 2016, 9, 976. [CrossRef] [PubMed]
27. Rashad, A.M.; Bai, Y.; Basheer, P.A.M.; Milestone, N.B.; Collier, N.C. Hydration and properties of sodium
sulfate activated slag. Cem. Concr. Compos. 2013, 37, 20–29. [CrossRef]
28. Fernandez-Jimenez, A.; Puertas, F. Effect of activator mix on the hydration and strength behaviour of
alkali-activated slag cements. Adv. Cem. Res. 2003, 15, 129–136. [CrossRef]
29. Chang, J.J. A study on the setting characteristics of sodium silicate-activated slag pastes. Cem. Concr. Res.
2003, 33, 1005–1011. [CrossRef]
30. Song, S.; Sohn, D.; Jennings, H.M.; Mason, T.O. Hydration of alkali-activated ground granulated blast furnace
slag. J. Mater. Sci. 2000, 35, 249–257. [CrossRef]
31. Roy, A. Sulfur speciation in granulated blast furnace slag: An X-ray absorption spectroscopic investigation.
Cem. Concr. Res. 2009, 39, 659–663. [CrossRef]
32. Grolimund, D.; Trainor, T.P.; Fitts, J.P.; Kendelewicz, T.; Liu, P.; Chambers, S.A.; Brown, G.E. Identification of
Cr species at the aqueous solution-hematite interface after Cr(VI)-Cr(III) reduction using GI-XAFS and Cr
L-edge NEXAFS. J. Synchrotron Radiat. 1999, 6, 612–614. [CrossRef] [PubMed]
33. Naftel, S.J.; Sham, T.K.; Yiu, Y.M.; Yates, B.W. Calcium L-edge XANES study of some calcium compounds.
J. Synchrotron Radiat. 2001, 8, 255–257. [CrossRef] [PubMed]
34. Stöhr, J. NEXAFS Spectroscopy; Springer: Berlin, Germany, 1992.
35. Doyle, C.S.; Kendelewicz, T.; Bostick, B.C.; Brown, G.E. Soft X-ray spectroscopic studies of the reaction
of fractured pyrite surfaces with Cr(Vl)-containing aqueous solutions. Geochim. Cosmochim. Acta 2004,
68, 4287–4299. [CrossRef]
36. Shen, C.; Chen, H.; Wu, S.; Wen, Y.; Li, L.; Jiang, Z.; Li, M.; Liu, W. Highly efficient detoxification of Cr(VI)
by chitosan-Fe(III) complex: Process and mechanism studies. J. Hazard. Mater. 2013, 244–245, 689–697.
[CrossRef] [PubMed]
37. Kucheyev, S.O.; van Buuren, T.; Baumann, T.F.; Satcher, J.H.; Willey, T.M.; Meulenberg, R.W.; Felter, T.E.;
Poco, J.F.; Gammon, S.A.; Terminello, L.J. Electronic structure of titania aerogels from soft X-ray absorption
spectroscopy. Phys. Rev. B 2004, 69, 245102. [CrossRef]
38. Andersson, K.; Allard, B.; Bengtsson, M.; Magnusson, B. Chemical-Composition of Cement Pore Solutions.
Cem. Concr. Res. 1989, 19, 327–332. [CrossRef]
39. Rimmelé, G.; Barlet-Gouédard, V.; Porcherie, O.; Goffé, B.; Brunet, F. Heterogeneous porosity distribution in
Portland cement exposed to CO2-rich fluids. Cem. Concr. Res. 2008, 38, 1038–1048. [CrossRef]
Materials 2018, 11, 11 17 of 17

40. Johannesson, B.; Utgenannt, P. Microstructural changes caused by carbonation of cement mortar.
Cem. Concr. Res. 2001, 31, 925–931. [CrossRef]
41. Puertas, F.; Palacios, M.; Vazquez, T. Carbonation process of alkali-activated slag mortars. J. Mater. Sci. 2006,
41, 3071–3082. [CrossRef]
42. Slegers, P.A.; Rouxhet, P.G. Carbonation of the hydration products of tricalcium silicate. Cem. Concr. Res.
1976, 6, 381–388. [CrossRef]
43. Morandeau, A.; Thiery, M.; Dangla, P. Investigation of the carbonation mechanism of CH and C-S-H in terms
of kinetics, microstructure changes and moisture properties. Cem. Concr. Res. 2014, 56, 153–170. [CrossRef]
44. Glasser, F.; Matschei, T. Interactions between Portland cement and carbon dioxide. In Proceedings of the
ICCC Conference, Montreal, QC, Canada, 8–13 July 2007.
45. Thiery, M.; Villain, G.; Dangla, P.; Platret, G. Investigation of the carbonation front shape on cementitious
materials: Effects of the chemical kinetics. Cem. Concr. Res. 2007, 37, 1047–1058. [CrossRef]
46. Borges, P.H.R.; Costa, J.O.; Milestone, N.B.; Lynsdale, C.J.; Streatfield, R.E. Carbonation of CH and C-S-H in
composite cement pastes containing high amounts of BFS. Cem. Concr. Res. 2010, 40, 284–292. [CrossRef]
47. Richardson, I.G.; Brough, A.R.; Groves, G.W.; Dobson, C.M. The Characterization of Hardened
Alkali-Activated Blast-Furnace Slag Pastes and the Nature of the Calcium Silicate Hydrate (C-S-H) Phase.
Cem. Concr. Res. 1994, 24, 813–829. [CrossRef]
48. Wang, S.D.; Scrivener, K.L. Hydration Products of Alkali-Activated Slag Cement. Cem. Concr. Res. 1995,
25, 561–571. [CrossRef]
49. Suda, Y.; Saeki, T.; Saito, T. Relation between Chemical Composition and Physical Properties of C-S-H
Generated from Cementitious Materials. J. Adv. Concr. Technol. 2015, 13, 275–290. [CrossRef]
50. Beaudoin, J.J.; Feldman, R.F. The significance of helium diffusion measurements in studying the removal of
structural water in inorganic hydrated systems. Cem. Concr. Res. 1978, 8, 223–231. [CrossRef]
51. Black, L.; Breen, C.; Yarwood, J.; Garbev, K.; Stemmermann, P.; Gasharova, B. Structural features of C-S-H(I)
and its carbonation in air—A Raman spectroscopic study. Part II: Carbonated phases. J. Am. Ceram. Soc.
2007, 90, 908–917. [CrossRef]
52. Richardson, I.G.; Groves, G.W. The structure of the calcium silicate hydrate phases present in hardened
pastes of white Portland cement/blast-furnace slag blends. J. Mater. Sci. 1997, 32, 4793–4802. [CrossRef]
53. Ayub, T.; Khan, S.U.; Memon, F.A. Mechanical Characteristics of Hardened Concrete with Different Mineral
Admixtures: A Review. Sci. World J. 2014, 2014, 875085. [CrossRef] [PubMed]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).