molecules
Article
Experimental Investigation and Simplistic
Geochemical Modeling of CO2 Mineral Carbonation
Using the Mount Tawai Peridotite
Omeid Rahmani 1, *, James Highfield 2 , Radzuan Junin 3,4 , Mark Tyrer 5 and
Amin Beiranvand Pour 6
1
2
3
4
5
6
*
Department of Petroleum Engineering, Mahabad Branch, Islamic Azad University, Mahabad 59135-433, Iran
560 Yishun Avenue 6 #08-25 Lilydale, Singapore 768966, Singapore; [email protected]
Department of Petroleum Engineering, FCEE, Universiti Teknologi Malaysia (UTM), Skudai 81310, Johor,
Malaysia; [email protected] or [email protected]
Universiti Teknologi Malaysia-Malaysia Petroleum Resources Corporation (UTM-MPRC),
Institute for Oil and Gas, Universiti Teknologi Malaysia (UTM), Skudai 81310, Johor, Malaysia
Mineral Industry Research Organisation, Wellington House, Starley Way, Birmingham International Park,
Solihull, Birmingham B37 7HB, UK; [email protected]
Geoscience and Digital Earth Centre (Geo-DEC), Research Institute for Sustainability and
Environment (RISE), Universiti Teknologi Malaysia (UTM), Skudai 81310, Johor, Malaysia;
[email protected] or [email protected]
Correspondence: [email protected] or [email protected]; Tel.: +98-914-442-2009
Academic Editor: Derek J. McPhee
Received: 16 January 2016 ; Accepted: 8 March 2016 ; Published: 16 March 2016
Abstract: In this work, the potential of CO2 mineral carbonation of brucite (Mg(OH)2 ) derived
from the Mount Tawai peridotite (forsterite based (Mg)2 SiO4 ) to produce thermodynamically stable
magnesium carbonate (MgCO3 ) was evaluated. The effect of three main factors (reaction temperature,
particle size, and water vapor) were investigated in a sequence of experiments consisting of aqueous
acid leaching, evaporation to dryness of the slurry mass, and then gas-solid carbonation under
pressurized CO2 . The maximum amount of Mg converted to MgCO3 is ~99%, which occurred at
temperatures between 150 and 175 ˝ C. It was also found that the reduction of particle size range from
>200 to <75 µm enhanced the leaching rate significantly. In addition, the results showed the essential
role of water vapor in promoting effective carbonation. By increasing water vapor concentration
from 5 to 10 vol %, the mineral carbonation rate increased by 30%. This work has also numerically
modeled the process by which CO2 gas may be sequestered, by reaction with forsterite in the presence
of moisture. In both experimental analysis and geochemical modeling, the results showed that the
reaction is favored and of high yield; going almost to completion (within about one year) with the
bulk of the carbon partitioning into magnesite and that very little remains in solution.
Keywords: CO2 sequestration; forsterite; ex-situ; in situ; mineral carbonation
1. Introduction
Carbon dioxide (CO2 ) is the principal greenhouse gas released into the atmosphere during fuel
combustion, particularly due to the extensive use of fossil fuels for energy production from coal, oil,
and natural gas since the industrial revolution [1]. Moreover, atmospheric CO2 has recently surpassed
400 ppm, and is predicted to increase to nearly 1000 ppm by the end of the 21st Century [2,3]. The
related global temperature rise (exceeding 2 ˝ C) [3] will almost certainly result in irreversible climate
change with potentially disastrous consequences.
Researchers have studied ways to mitigate the amount of greenhouse gases released into the
atmosphere by sequestration of CO2 through different approaches, including aquifer storage, deep
Molecules 2016, 21, 353; doi:10.3390/molecules21030353
www.mdpi.com/journal/molecules
Molecules 2016, 21, 353
2 of 19
sea storage, and mineral carbonation in particular [1,4–14]. Minerals and rocks rich in magnesium
(Mg2+ )/calcium (Ca2+ ) are commonly considered as candidates due to their wide availability, low cost,
and environmentally benign nature. [8,12,15] During mineral carbonation, CO2 reacts with Mg2+ or
Ca2+ -rich minerals (e.g., olivine and gypsum) to form solid carbonates, which are expected to be stable
over geologic time periods. The minerals olivine [(Mg,Fe)2 SiO4 ] and forsterite [Mg2 SiO4 ], containing
up to 33.6 wt. % Mg, have the highest capacity to trap CO2 as magnesium carbonates, and a high
rate of dissolution among rock-forming silicate minerals [16]. Formation of magnesite from forsterite,
the Mg-end member of the olivine solid solution series, is thermodynamically favorable based on the
negative Gibbs free energy of Reaction (1).
1{2Mg2 SiO4 ` CO2 ÑMgCO3 ` 1{2SiO2 ` 95 kJ{mol
(Rx. 1)
According to Lackner et al. [17] the chemical reactions in mineral carbonation process can be
very slow under ambient conditions and, therefore, activation processes such as exposure to acid,
heat [8], and water [18–21] have been used to accelerate the carbonation rate. These processes
are tedious and require additional energy input, which has made acid/heating approach less
attractive than some others. The rates of carbonation can be raised by increasing surface areas
of the mineral or its intermediates and by elevating the temperatures, resulting in lower kinetic
constraints [15,16,22–29]. Moreover, an in situ CO2 sequestration system with mineral carbonation
can be treated as a fully-coupled problem between rock deformation, pore-fluid flow, heat transfer,
mass transport, and chemical reaction processes. Three types of models are commonly employed in
computational petroleum geoscience and engineering research methodologies. These approaches are
geological/geochemical (conceptual), mathematical, and/or numerical simulation.
The purpose of this study was to do ex situ (laboratory) studies of the factors affecting the rates of
(forsterite) mineral carbonation (i.e., particle size, water vapor, and reaction temperature) in support of
an in situ geochemical conceptual model. A transport reaction modeling software PHREEQC (version
2.18, US Geological Survey (USGS), Reston, WV, USA), was applied for simulating the chemical
reactions and transport processes in the forsterite mineral carbonation process. Pre-treatment of
forsterite in a quantitative equivalent of mineral acid (HCl), i.e., dissolution to neutrality, was taken as
a suitably fast process (in extracting most of the Mg ion) for practical experimentation. One advantage
of this process is that the formation of magnesium carbonate (MgCO3 ) releases heat, which can in
principle be cycled back to other endothermic steps (see Reactions (2)–(4) in Section 3.3) through heat
integration in a commercial process.
2. Results and Discussion
2.1. Mineral Characterization
The elemental composition of the peridotite mineral was determined by XRF analysis as MgO
(51.9%), SiO2 (41.1%) as major components with minor levels of FeO, Al2 O3 , Na2 O, K2 O, and CaO (see
details in Table 1).
Table 1. Chemical composition of fresh peridotite mineral (wt. %) as determined by XRF analysis.
Al2 O3
0.204
CaO
0.061
FeO
5.969
MgO
51.921
K2 O
SiO2
Na2 O
Cr2 O3
Volatiles
C + CO2
H2 O
0.005
41.072
0.083
0.034
<0.352
0.291
Mg
1.812
K
0.004
Si
0.995
Na
0.008
Cr
-
-
Number of ions on the basis of O
Al
0.001
Ca
0.004
Fe
0.161
Na
0.001
Molecules 2016, 21, 353
Molecules 2016, 21, 353
3 of 19
3 of 19
Molecules 2016, 21, 353
3 of 19
XRDanalysis
analysisof
ofthe
theHCl-cleaned
HCl-cleanedstarting
startingmaterial
material(Figure
(Figure1)1)shows
showsaacharacteristic
characteristicpattern
patternof
of
XRD
˝
˝
˝
˝
˝
˝
˝
olivine
(2θ=
= 11.8
11.8°,, 23.6
23.6°,
29.3°,
some
quartz
olivine
(2θ
, 31.1°,
31.1 , 33.6°,
33.6
,40.8°,
40.8
,and
and43.6°),
43.6
),with
with
somecontamination
by
quartz
XRD
analysis
of the, 29.3
HCl-cleaned
starting
material
(Figure
1) shows
acontamination
characteristic by
pattern
ofor
˝
˝
˝
˝
free
silica
(2θ
=
20.9°,
26.5°,
50.5°
and
68.7°).
In
view
of
the
fairly
low
level
of
Fe
and
its
own
capacity
orolivine
free silica
= 20.9
, 26.5
). and
In view
of with
the fairly
level of Feby
and
its own
(2θ =(2θ
11.8°,
23.6°,
29.3°,, 50.5
31.1°, and
33.6°,68.7
40.8°,
43.6°),
some low
contamination
quartz
orfor
carbonation
(as
FeCO
3
),
modeling
studies
(vide
infra)
were
based
for
simplicity
on
the
pure
forsterite
capacity
for(2θ
carbonation
(as FeCO
studies
(vide
infra)
were
based
simplicity
on the pure
free silica
= 20.9°, 26.5°,
50.5° and
68.7°). In view
of the
fairly
low
level
of Fefor
and
its own capacity
for
3 ), modeling
composition
(MgFeCO
2SiO4(Mg
Mg-end
member
of
the
olivine
solid
solution
series.
forsterite
composition
SiO
),
the
Mg-end
member
of
the
olivine
solid
solution
series.
carbonation
(as
3),
), the
modeling
studies
(vide
infra)
were
based
for
simplicity
on
the
pure
forsterite
2
4
composition (Mg2SiO4), the Mg-end member of the olivine solid solution series.
8000
Olivine
Quartz
Olivine
Quartz
Intensity,
Lin (CPS)
Intensity,
Lin (CPS)
8000
7000
7000
6000
6000
5000
5000
4000
4000
3000
3000
2000
2000
1000
1000
0
0 10
10
20
30
20
30
40
2 Theta - Scale
40
2 Theta - Scale
50
50
60
60
70
70
Figure1.1.XRD
XRDpattern
patternof
ofthe
thestarting
startingperidotite
peridotitemineral.
mineral.
Figure
Figure 1. XRD pattern of the starting peridotite mineral.
Morphological and
andstructural
structuralchanges
changes inthe
the
olivine (nominal
composition
Mg1.84
0.16SiO4) at
Morphological
olivine
composition
Mg
FeFe
1.84
0.16 SiO4 ) at
Morphological
and
structural
changesinand
in the
olivine(nominal
(nominal
composition
Mg
1.84Fe0.16SiO4) at
various
stages
during
chemical
pretreatment
carbonation,
viz.,
after
leaching
in
HCl,
neutralization
various
stages
pretreatment
viz., after
after leaching
leachingin
inHCl,
HCl,neutralization
neutralization
various
stagesduring
duringchemical
chemical
pretreatmentand
andcarbonation,
carbonation, viz.,
2+ [as Mg(OH)
and
precipitation
of
Mg
2], and exposure to humid CO2 at 4.8 bar and 150˝°C, are best
2+
and
], and
and exposure
exposure to
to humid
humid CO
CO22atat4.8
4.8bar
barand
and150
150°C,C,are
arebest
best
andprecipitation
precipitationofofMg
Mg2+ [as
[as Mg(OH)
Mg(OH)22],
seenby
bySEM
SEMmicrographs
micrographsand
andpowder
powderXRD
XRDininFigures
Figures22and
and33,
respectively.
seen
respectively.
seen by SEM micrographs and powder XRD in Figures 2 and 3, respectively.
BySEM,
SEM, thefresh
fresh olivinesample
sample (Figure2a)
2a) consistedtypically
typicallyofofpolycrystalline
polycrystallinegrains
grainsin
inthe
the
By
By SEM,the
the fresholivine
olivine sample(Figure
(Figure 2a) consisted
consisted typically of
polycrystalline grains
in the
millimetersize
size range.After
After acidleaching
leaching andneutralization
neutralizationof
ofthe
thesieved
sievedfraction
fraction<75
<75 µm
µm(Figure
(Figure2b),
2b),
millimeter
millimeter sizerange.
range. Afteracid
acid leaching and
and neutralization of
the sieved
fraction <75
µm (Figure
2b),
the
grains
became
finer
(<5
µm)
and
the
structure
more
amorphous.
In
agreement
with
the
findings
of
the
Inagreement
agreementwith
withthe
thefindings
findingsofof
thegrains
grainsbecame
becamefiner
finer(<5
(<5µm)
µm)and
andthe
the structure
structure more
more amorphous.
amorphous. In
Bearatetetal.
al. [30]and
and Kwonetetal.
al. [18],this
this islikely
likely due to anamorphous
amorphousSiO
SiO2residue
residueafter
afterleaching,
leaching,any
any
Bearat
Bearat et al.[30]
[30] andKwon
Kwon et al.[18],
[18], thisis
is likely due to an amorphous
SiO22residue
after leaching,
any
(hexagonal)
brucite
likely
being
nanocrystalline.
Under
humid
CO
2, the development of polyhedral and
(hexagonal)
beingnanocrystalline.
nanocrystalline.Under
Under
humid
the development
of polyhedral
(hexagonal)brucite
brucitelikely
likely being
humid
COCO
2, the
of polyhedral
and
2 , development
more
sheet-like
crystallites
(Figure
2c)2c)was
evident,
probably
representing
(hydrated) magnesium
magnesium
and
more
sheet-like
crystallites
(Figure
was
evident,
probably
representing
(hydrated)
more sheet-like crystallites (Figure 2c) was evident, probably representing (hydrated) magnesium
carbonate,
and
finally
(Figure
2d),the
therhombohedral
rhombohedralhabit
habittypical
typicalof
ofmagnesite
magnesite(MgCO
(MgCO
3) was evident.
carbonate,
of
magnesite
(MgCO
) wasevident.
evident.
carbonate,and
andfinally
finally(Figure
(Figure2d),
2d),
the
rhombohedral
3) was
3
Figure2.2.
2.SEM
SEMimages
images showing
showing (a)
(a)
the
fresh
olivine
mineral,
Figure
SEM
images
showing
(a) the
thefresh
fresholivine
olivinemineral,
mineral,and
andmorphological
morphologicalchanges
changesduring
during
Figure
and
morphological
changes
during
chemical
pretreatment
and
carbonation:
(b)
the
leached/neutralized
sample
in
the
presence
of
humid
chemical
pretreatment
and
carbonation:
(b)
the
leached/neutralized
sample
in
the
presence
of
humid
chemical pretreatment and carbonation: (b) the leached/neutralized sample in the presence of humid
2 at 150˝ °C during 15 min; (c) 90 min; and (d) 120 min.
CO
at150
150 °C
during 15
15 min;
min; (c)
(c) 90
90min;
min;and
and(d)
(d)120
120min.
min.
CO22at
CO
C during
(chemically un-pretreated) mineral, for which evidence has been reported elsewhere [31]. The abundance
of quartz (as co-product) cannot be taken as a reliable indicator of the progress of carbonation because
it is also a phase contaminant in the original mineral (see Figure 1). Furthermore, similar chemical
treatment (flux extraction of Mg2+) from serpentinites did not produce quartz but instead a silica
residue 2016,
of unusual
Molecules
21, 353 structure [27]. The unsystematic peak intensity of quartz seen here by XRD4 may
of 19
be due merely to local inhomogeneity in the samples.
M: Magnesite
HM: Hydromagnesite
Ol: Olivine
B: Brucite
Q: Quartz
Ol
Intensity, Lin (CPS)
Q
HM
B
Q
Q
B
20
M
B
M
(a)
HM
M
Q
10
B
30
(b)
M
M
40
M
50
(c)
60
2 Theta - degree
Figure 3.
3. XRD
olivine mineral
mineral during
during progressive
progressive
Figure
XRD patterns
patterns showing
showing structural
structural changes
changes in
in the
the olivine
carbonation: (a)
(a) after
and initial
initial exposure
exposure of
of the
the damp
damp residue
residue to
to humid
humid
carbonation:
after acid
acid leaching,
leaching, neutralization,
neutralization, and
2 (PCO2 = 4.8 bar, T = 150 °C)
during
15
min;
(b)
after
90
min;
and
(c)
after
120
min.
CO
˝
CO2 (PCO2 = 4.8 bar, T = 150 C) during 15 min; (b) after 90 min; and (c) after 120 min.
2.2. Effect of Particle size on the Carbonation of Olivine
Figure 3 shows XRD data obtained after the same stages of treatment in parallel with the
leaching
rate was found
be much
slower
than the
optimalolivine
carbonation
rate
(achievable
at
SEM The
analyses.
Diffractogram
3atoreveals
that
although
unreacted
(Mg1.84
Fe0.16
SiO4 main
˝
175
°C
vide
infra),
such
that
the
production
of
magnesite
in
these
experiments
reflects
mainly
variations
reflection at 2θ = 11.8 ) is predominant after only 15 min exposure to humid CO2 , hydromagnesite
2+ ion
˝ ) and
in the= rate
of 2extraction
of soluble
by acid
treatment.
Figure
4 shows
that
(HM
Mg5and
(CO3efficiency
)4 (OH)2 ‚4H
O main reflection
at 2θMg
= 31.3
a little
magnesite
(M = MgCO
3 main
2+ leaching to
˝
decreasing
the
grain
size
from
>200
µm
to
<75
µm
caused
the
limiting
degree
of
Mg
reflection at 2θ = 43.15 ), had already formed from brucite [Mg(OH)2 ] precursor, itself created in the
increase from 35% to 99%,
respectively.
It exposure
is well-known
intuitively
obvious that
the
rate
leaching/neutralization
stage.
At longer
timesand
(Figure
3b,c), reflections
due
toreaction
brucite and
˝
˝
˝
and carbonation were
degree
can be raised
by increasing
surface area,
e.g.,
by ,grinding/sieving,
as
hydromagnesite
progressively
replaced
by those the
of magnesite
(2θ =
32.25
43.15 , 54.25 ). The
shown
for
example
by
Garcia
et
al.
[32].
In
this
work,
the
smallest
particle
size
(d
<
75
µm)
is
the
only
ultimate disappearance of olivine is intriguing since it implies direct steam-activated carbonation
onethe
offering
the prospect
of full leaching
in a practical
time.
a rule
of thumb,
1–2 h for
a
of
(chemically
un-pretreated)
mineral,(99%)
for which
evidence
hasAs
been
reported
elsewhere
[31].
terrestrial
(ex-situ)
to achieve
>90%cannot
conversion
is taken
as a realistic
practical
target
to limit
The
abundance
of process
quartz (as
co-product)
be taken
as a reliable
indicator
of the
progress
of
the scale (and
associated
costs)
of anycontaminant
future installation
for CO2 mineral
sequestration.
This 1).
particle
cut and
carbonation
because
it is also
a phase
in the original
(see Figure
Furthermore,
2+
leachingchemical
procedure
(2 h at 60
°C)extraction
were therefore
standard fordid
allnot
samples
in subsequent
similar
treatment
(flux
of Mgapplied
) fromas
serpentinites
produce
quartz but
tests
described
below.
instead a silica residue of unusual structure [27]. The unsystematic peak intensity of quartz seen here
by XRD may be due merely to local inhomogeneity in the samples.
2.2. Effect of Particle Size on the Carbonation of Olivine
The leaching rate was found to be much slower than the optimal carbonation rate (achievable at
175 ˝ C vide infra), such that the production of magnesite in these experiments reflects mainly variations
in the rate and efficiency of extraction of soluble Mg2+ ion by acid treatment. Figure 4 shows that
decreasing the grain size from >200 µm to <75 µm caused the limiting degree of Mg2+ leaching to
increase from 35% to 99%, respectively. It is well-known and intuitively obvious that the reaction
rate and carbonation degree can be raised by increasing the surface area, e.g., by grinding/sieving,
as shown for example by Garcia et al. [32]. In this work, the smallest particle size (d < 75 µm) is the
only one offering the prospect of full leaching (99%) in a practical time. As a rule of thumb, 1–2 h for a
terrestrial (ex-situ) process to achieve >90% conversion is taken as a realistic practical target to limit
the scale (and associated costs) of any future installation for CO2 sequestration. This particle cut and
leaching procedure (2 h at 60 ˝ C) were therefore applied as standard for all samples in subsequent
tests described below.
Molecules 2016, 21, 353
5 of 19
Molecules 2016, 21, 353
5 of 19
Mg-leaching
(%)
Mg-leaching
(%)
5 of 19
Effect of particle size on CO2 uptake and Mg-leaching
100
1
99
90
100
80
90
70
80
60
70
50
60
40
50
30
40
20
30
10
20
0
10 0
0
0
Effect of particle size on CO2 uptake and Mg-leaching
99
64.83
64.83
43.01
35
43.01
35
15
30
15 <75 µm
30
45
60
75
90
Time (min.)
75-125 µm
45
60
75125-20090µm
105
120
µm
105>200120
0.9
1
0.8
0.9
0.7
0.8
0.6
0.7
0.5
0.6
0.4
0.5
0.3
0.4
0.2
0.3
0.1
0.2
0
0.1
CO2 input
CO(mol/g)
2 input (mol/g)
Molecules 2016, 21, 353
0
Time (min.)
Figure 4. Effect of particle size<75
onµmMg2+2+ leaching
a range
as measured by the
75-125 µm over125-200
µm of time
>200intervals
µm
Figure 4. Effect of particle size on
Mg leaching
over a range
of time
intervals as measured by the
volume of CO2 uptake in the subsequent mineral carbonation process (at 175 °C
and 2 h).
˝
volume
in thesize
subsequent
mineral carbonation
175 Casand
2 h). by the
2 uptake
Figureof4.CO
Effect
of particle
on Mg2+ leaching
over a rangeprocess
of time (at
intervals
measured
volume
of CO2 uptakeon
inthe
theCarbonation
subsequent mineral
carbonation process (at 175 °C and 2 h).
2.3. Effect
of Temperature
of Olivine
Mg2+ conversion
(%)
Mg2+ conversion
(%)
to magnesite
(carbonation
yield) yield)
to magnesite
(carbonation
2.3. Effect of Temperature on the Carbonation of Olivine
The temperature
of the of
carbonation
process was studied over the interval from
2.3. Effect
of Temperaturedependence
on the Carbonation
Olivine
The temperature
dependence
of the carbonation
process
studiedafter
overstandard
the interval
from
ambient
to 175 °C. The
effect of temperature
on the amount
of Mgwas
conversion
leaching
˝ C. Thedependence
The
temperature
of
the
carbonation
process
was
studied
over
the
interval
from
ambient
to
175
effect
of
temperature
on
the
amount
of
Mg
conversion
after
standard
treatment (99% extraction) is illustrated in Figure 5. As expected, the temperature had an important
ambient
to the
175 °C.
The effect
of temperature
on
the amount
Mg
conversion
after
standard
leaching
leaching
treatment
(99%
extraction)
is
illustrated
in Figure
5. extraction
As
expected,
the
temperature
effect on
mineral
carbonation
process
consisting
of of
Mg
and
subsequent
MgCOhad
3
treatment
(99%
extraction)
is
illustrated
in
Figure
5.
As
expected,
the
temperature
had
an
important
an precipitation.
important effect
on the mineral
carbonation
process
of Mg
extraction
subsequent
The amount
of Mg converted
at the
desiredconsisting
temperatures
was
measuredand
continuously
effect
on
the The
mineral
carbonation
process
consisting
of the
Mgdesired
extraction
and subsequent
MgCO
MgCO
The
amount
of
Mg converted
at
temperatures
was
measured
several
times.
maximum
extent of
carbonation
of (99%)
was
attained
at temperatures
in excess
of3
3 precipitation.
precipitation.
The amount
of
converted
at theof
desired
temperatures
measured
continuously
150 °C. Theseveral
quantification
ofMg
the
MgCO3 formed
in
the mineral
carbonation
was
carried out
continuously
times. The
maximum
extent
carbonation
of (99%)was
wasprocess
attained
at
temperatures
several
times.
The
maximum
extent
of
carbonation
of
(99%)
was
attained
at
temperatures
in
excess of
˝
by titration
against
HClquantification
at room temperature.
Pokrovsky
et al.in
[33]
that the dissolution
in excess
of 150
C. The
of the MgCO
thedemonstrated
mineral carbonation
process
was
3 formed
150
°C.
The
quantification
of
the
MgCO
3 formed in the mineral carbonation process was carried out
rate ofout
magnesite
at 150
°C is lower
at 25temperature.
°C whereas the
rates at 100
150 demonstrated
°C in acidic solutions
carried
by titration
against
HCl than
at room
Pokrovsky
et and
al. [33]
that the
by
against
HCl
room temperature.
Pokrovsky
et al.
[33] demonstrated
that the
dissolution
˝C
aretitration
almostrate
the
same
[34],atsuggesting
strong
decrease
activation
energy
above
100
°C.˝ C
dissolution
of magnesite
at 150 a˝ C
is lower
thanofatthe
25apparent
whereas
the rates
at 100
and
150
rate
of
magnesite
at
150
°C
is
lower
than
at
25
°C
whereas
the
rates
at
100
and
150
°C
in
acidic
solutions
According
to
Saldi
et
al.
[34]
the
tendency
for
dissolution
rates
of
MgCO
3 to decrease with increasing
in acidic solutions are almost the same [34], suggesting a strong decrease of the apparent activation
are
almost thecould
samebenefit
[34], suggesting
a strong decrease
of
the apparent
activation
energy above
100 °C.
temperature
CO2 sequestration
making
magnesite
more resistant
toof
dissolution
energy
above 100 ˝ C.
According
to Saldi et efforts
al. [34]by
the
tendency
for dissolution
rates
MgCO3 to
According
to
Saldi
et
al.
[34]
the
tendency
for
dissolution
rates
of
MgCO
3 to decrease with increasing
in deeper
strata,
thus preserving
the petrophysical
integrity
of deep
carbonate-rich
decrease
with(hotter)
increasing
temperature
could efforts
benefit
efforts
by making
magnesite
2 sequestration
temperature
could benefit
CO2 sequestration
byCO
making
magnesite more
resistant
to dissolution
confining reservoirs.
more
resistant
to
dissolution
in
deeper
(hotter)
strata,
thus
preserving
the
petrophysical
integrity
in deeper (hotter) strata, thus preserving the petrophysical integrity of deep carbonate-rich of
deep
carbonate-rich
confining reservoirs.
confining
reservoirs.
100
Effect of temperature on mineral carbonation
95
98.5
Effect of temperature on mineral carbonation
95
100
98.5
95
87
90
95
84
85
80
87
90
78
84
80
75
85
80
72
75
78
70
80
75
67
70
65
72
75
70
65
67
70 25
50
75
100
125
150
65
Temperature (°C)
65
25
50
75
100
125
150
Temperature (°C)
99
99
175
175
Figure 5. The effect of reaction temperature in gas-solid carbonation of fully extracted Mg2+ ion
(as Mg(OH)2) to yield MgCO3 (PCO2 = 4.8 bar, humid CO2, t = 2 h).
2+ ion
Figure 5. The effect of reaction temperature in gas-solid carbonation of fully extracted Mg
Figure 5. The effect of reaction temperature in gas-solid carbonation of fully extracted Mg2+ ion
(as
(as Mg(OH)2) to yield MgCO3 (PCO2 = 4.8 bar, humid CO2, t = 2 h).
Mg(OH)2 ) to yield MgCO3 (PCO2 = 4.8 bar, humid CO2 , t = 2 h).
Molecules 2016, 21, 353
6 of 19
It is clear from Figure 5 that several competing factors in the carbonation process reach equilibrium
in the temperature range between 150 and 175 ˝ C. Probably the key factor is the relative humidity
(RH). Previous work has shown that carbonation accelerates on cooling below 200 ˝ C in a fixed
partial pressure of steam or RH ě 20% [35]. Since the presence of liquid water has been shown
to be important, it can also be argued that the effect of temperature on the solubility of CO2 may
influence the carbonation rate. Chen et al. [36] indicated that increasing temperature from 25 ˝ C to
150 ˝ C generally helps the precipitation of magnesite. The carbonate solubility product (Ksp ), Henry’s
constant (KH ), and the first- and second-order dissociation constants of carbonic acid (Ka1 , Ka2 ) are all
functions of temperature. Increasing the reaction temperature decreases the value of Ksp and increases
the KH value, which lowers the amount of CO2 gas in solution at a given pressure. The dissociation
constants of carbonic acid (Ka1 and Ka2 ) are also increased by increasing temperature and promoting
solution speciation and carbonation. Thus, increasing temperature has conflicting effects, lowering the
level of dissolved CO2 gas via its effect on KH but promoting magnesite precipitation via effects on
Ka1 , Ka2 , and Ksp . Furthermore, temperature impacts on the kinetics of magnesite precipitation [37]
and also affects the type of Mg-carbonate formed. It is well-known that magnesite precipitation
kinetics at ambient temperatures are exceedingly slow, and that metastable hydrated carbonates such
as hydromagnesite, dypingite, and nesquehonite almost invariably form instead.
2.4. Effect of Water Concentration
Formal carbonation of forsterite [Reaction (1)] does not involve water molecules explicitly, so any
beneficial effect of added water is clearly a kinetic effect. As shown in Figure 6, an indication of the
importance of water is seen in the modest level of CO2 removal in its absence. By analogy with recent
works on brucite and serpentinites, almost regardless of the CO2 pressure utilized, the presence of water
vapor in high relative humidity appears crucial to obtaining practical carbonation rates [20,31,38],
evidently by establishing a highly polar thin-film aqueous overlayer that facilitates CO2 ingress
into the bulk particle. This is supported by independent studies simulating in situ or geochemical
carbon sequestration where CO2 was assumed to be in the supercritical state. Felmy et al. [39] studied
high-surface-area forsterite in the presence of water-saturated scCO2 . They concluded that the nature
of the water in contact with the reacting surface is a key factor in the enhanced magnesite formation.
When excess water was added to the forsterite particles, a thin water film was formed on the forsterite
surface promoting magnesite formation. Loring et al. [40] declared that this water film provides a
distinctive situation for the magnesite formation by decreasing the effective Mg2+ dehydration energy
and simplifying the transformation of nesquehonite to magnesite. Otherwise, the presence of liquid
water can allow the formation of magnesium bicarbonate in solution that decomposes upon drying
to magnesium carbonate. Moreover, Schaef et al. [41,42] revealed that the addition of water to the
saturated system noticeably increases the rate of mineral carbonation, facilitating the overall conversion
of nesquehonite to magnesite. No evidence of further carbonation was observed under unsaturated
conditions below 50 ˝ C. A similar promoting effect of water on brucite carbonation under scCO2 was
reported by Loring et al. [43] using in situ Fourier-Transform Infrared (FTIR) spectroscopic experiments.
To investigate the effect of humidity, the concentration of water vapor was set at various levels,
5, 10, and 20 vol % prior to carbonation of the brucite extract at 175 ˝ C. As illustrated in Figure 6,
above 5 vol % steam, the degree of carbonation increases by almost double in the presence of water
vapor such that within 5 min, complete removal of CO2 (15 vol %) was achieved. However, levels
of water vapor exceeding 20 vol % had no additional effect on the rate of removal of CO2 . It can be
concluded that water vapor is able to solvate CO2 , generate carbonate ions and protons [44], and
increase the carbonation degree of Mg(OH)2 as derived from olivine. Moreover, the aforementioned
modest uptake of CO2 (~7 vol %) under “dry” conditions may be due to adventitious water not fully
removed from the Mg(OH)2 -containing residue. Vitillo [45] declared that in the presence of water
vapor, MgCO3 crystalline phase reappeared increasingly, while the magnesium oxide periclase (MgO)
phase gradually disappeared. These observations are well in agreement with the thermodynamic
Molecules 2016, 21, 353
Molecules 2016, 21, 353
7 of 19
7 of 19
data on MgO, Mg(OH)2 , and MgCO3 systems. The promoting effect of water may be attributed to
faster
reaction
kinetics
by offering
alternative
to magnesite
via hydrocarbonate
intermediates
reaction
kinetics
by offering
alternative
routesroutes
to magnesite
via hydrocarbonate
intermediates
such
such
as dypingite.
as dypingite.
Figure 6.
6. Effect
Effect of
of the
the presence
presence or
or absence
absence of
of water
water vapor
vapor on
on the
the rate
rate of
ofCO
CO2 absorption
absorption at
at 175
175 ˝°C
on
Figure
C on
2
2SiO4 (d < 75 µm) activated by chemical pre-treatment and exposed to 0.5 L/min CO2 gas
10
g
Mg
10 g Mg2 SiO4 (d < 75 µm) activated by chemical pre-treatment and exposed to 0.5 L/min CO2 gas
bar, PH2O = 1.6–6.4 bar or 18%–72% RH).
(PCO2 ==4.8
(P
CO2 4.8 bar, PH2O = 1.6–6.4 bar or 18%–72% RH).
As a comparison to related work on Mg(OH)2 in the literature, Siriwardane and Stevens [46]
As a comparison to related work on Mg(OH)2 in the literature, Siriwardane and Stevens [46]
reported good absorption kinetics and reasonable capacity
for CO2 in plug-flow reactor experiments
reported good absorption kinetics and reasonable capacity for CO2 in plug-flow reactor experiments
over a promoted brucite (~3 mol CO2/kg or 20 mol %) in “moist” helium at 200 °C. However, the
over a promoted brucite (~3 mol CO2 /kg or 20 mol %) in “moist” helium at 200 ˝ C. However,
the
promotional effect of water per se was not explored and the sorbent surface area was low (~2.5 m2·g−1).
promotional effect of water per se was not explored and the sorbent surface area was low (~2.5 m2 ¨ g´1 ).
This compares with our thermogravimetric work [38] showing that Mg(OH)2 extracted from the
This compares with our thermogravimetric work [38] showing
that Mg(OH) extracted from the
2·g−1). These are 2evidently important
mineral is typically obtained in high-surface-area form (~25 m
2
´
mineral is typically obtained in high-surface-area form (~25 m ¨ g 1 ). These are evidently important
factors in attainment of almost quantitative (~100%) carbonation at lower temperature (150–175 °C)
factors in attainment of almost quantitative (~100%) carbonation at lower temperature (150–175 ˝ C) in
in this work, specifically the higher water levels utilized and the better dispersion of brucite derived
this work, specifically the higher water levels utilized and the better dispersion of brucite derived from
from the mineral. Based on the thermodynamic equilibrium of Mg(OH)2 formation from MgO,
the mineral. Based on the thermodynamic equilibrium of Mg(OH)2 formation from MgO, Siriwardane
Siriwardane and Stevens [46] showed that is likely to form Mg(OH)2 under the high steam environment,
and Stevens [46] showed that is likely to form Mg(OH)2 under the high steam environment, which
which accounts for the subsequent CO2 uptake. It is important to note that the Mg(OH)2 system has the
accounts for the subsequent CO2 uptake. It is important to note that the Mg(OH)2 system has the far
far lower heat of sorption. This confirms that the regeneration heat (input) needed to displace CO2 from
lower heat of sorption. This confirms that the regeneration heat (input) needed to displace CO2 from
MgCO3 by water (to form Mg(OH)2) is significantly lower than that required for the decomposition of
MgCO3 by water (to form Mg(OH)2 ) is significantly lower than that required for the decomposition of
MgCO3 (to MgO + CO2).
MgCO3 (to MgO + CO2 ).
2.5. Kinetic Analysis of Mg Extraction by HCl
2.5. Kinetic Analysis of Mg Extraction by HCl
Different kinetic analyses including expressions for product layer diffusion, film diffusion, chemical
Different kinetic analyses including expressions for product layer diffusion, film diffusion,
reaction control, and a combination of chemical reaction control (Equations (4)–(7), respectively, in
chemical reaction control, and a combination of chemical reaction control (Equations (4)–(7),
Section 3.5), were used to evaluate the integral rate data. The extent of forsterite dissolution, XE, is
respectively, in Section 3.5), were used to evaluate the integral rate data. The extent of forsterite
taken as fitting parameter but this is actually measured from the amount of magnesite, i.e., the extent
dissolution, XE , is taken as fitting parameter but this is actually measured from the amount of
of carbonation (= RCO2 in Equation (2)), because dissolution is much slower than carbonation (of the
magnesite, i.e., the extent of carbonation (= RCO2 in Equation (2)), because dissolution is much slower
Mg(OH)2 extract). Direct carbonation of unreacted forsterite is probably even slower. Thus, dissolution
is rate-determining in the overall process. The two best-fit results are illustrated in Figure 7a,b, but
the first, a combination of chemical reaction control and product layer diffusion (Equation (7)) provided
Molecules 2016, 21, 353
8 of 19
than carbonation (of the Mg(OH)2 extract). Direct carbonation of unreacted forsterite is probably
even slower. Thus, dissolution is rate-determining in the overall process. The two best-fit results
are illustrated in Figure 7a,b, but the first, a combination of chemical reaction control and product layer
Molecules 2016, 21, 353
8 of 19
diffusion (Equation (7)) provided the highest correspondence with the measured data. Thus, it can be
concluded
a combination
of chemical
reaction
control
product
layer diffusion
is rate-limiting
the highestthat
correspondence
with
the measured
data.
Thus,and
it can
be concluded
that a combination
of
for
Mg
extraction.
chemical reaction control and product layer diffusion is rate-limiting for Mg extraction.
Combination of chemical reaction control
and product layer diffusion
0.8
a
175 °C
0.7
y = 9E-05x + 0.0243
R² = 0.9955
125 °C
0.6
75 °C
0.5
25 °C
y = 5E-05x + 0.0135
R² = 0.9925
0.4
0.3
y = 3E-05x + 0.0101
R² = 0.983
0.2
0.1
y = 1E-05x + 0.0028
R² = 0.9823
0
0
2000
4000
Time (s)
6000
8000
0.8
b
175 °C
The product layer diffusion
0.7
y = 9E-05x + 0.063
R² = 0.9756
125 °C
0.6
75 °C
0.5
25 °C
0.4
y = 4E-05x + 0.0244
R² = 0.9716
0.3
y = 2E-05x + 0.0153
R² = 0.9714
0.2
0.1
y = 1E-05x - 0.0014
R² = 0.9703
0
0
2000
4000
6000
8000
Time (s)
Figure 7.
7. Kinetic
of olivine
olivine dissolution
dissolution rate
rate (in
(in HCl)
HCl) by
by plotting
plotting the
the combination
combination of
of chemical
chemical
Figure
Kinetic analysis
analysis of
reaction
control
and
product
layer
diffusion
(a)
and
the
product
layer
diffusion
(b)
vs.
time
at
various
reaction control and product layer diffusion (a) and the product layer diffusion (b) vs. time at various
reaction temperatures.
temperatures.
reaction
Activation energies (Ea) for mineral dissolution were determined from simple log/log plots of
Activation energies (E ) for mineral dissolution were determined from simple log/log plots of the
the time-independent ratea k at various temperatures. These values are presented in Table 2, from
time-independent rate k at various temperatures. These values are presented in Table 2, from which
which the Arrhenius plots shown in Figure 8 were obtained. Once again, the quality of fit was best
the Arrhenius plots shown in Figure 8 were obtained. Once again, the quality2 of fit was best for the
for the combination of product layer diffusion and chemical reaction control (R = 0.9917) as compared
combination of product layer diffusion and chemical reaction control (R2 = 0.9917) as compared to
to product layer diffusion only (R = 0.9764). Considering these models as the controlling mechanisms
product layer diffusion only (R = 0.9764). Considering these models as the controlling mechanisms
during the dissolution of forsterite, the Ea value was 15.5 kJ/mol for product layer diffusion control, and
during the dissolution of forsterite, the Ea value was 15.5 kJ/mol for product layer diffusion control,
16.0 kJ/mol for the combination of product layer diffusion and chemical reaction control. It is likely that
and 16.0 kJ/mol for the combination of product layer diffusion and chemical reaction control. It is likely
the chemical reaction is initially rate-limiting but product layer diffusion gradually becomes rate
limiting as the product layer of silica builds up and the unreacted surface area decreases. According
to Gharabaghi et al. [47] a low value of Ea indicates that product layer diffusion is rate-controlling,
Therefore, considering the values of multiple regression coefficients for different models and
calculated Ea for two selected models, it could be concluded that the dissolution rates of forsterite are
kinetically regulated by the combination of chemical reaction control and product layer diffusion.
Molecules 2016, 21, 353
9 of 19
that the chemical reaction is initially rate-limiting but product layer diffusion gradually becomes rate
limiting as the product layer of silica builds up and the unreacted surface area decreases. According
to Gharabaghi et al. [47] a low value of Ea indicates that product layer diffusion is rate-controlling,
Therefore, considering the values of multiple regression coefficients for different models and calculated
Ea for two selected models, it could be concluded that the dissolution rates of forsterite are kinetically
regulated by the combination of chemical reaction control and product layer diffusion.
Molecules 2016, 21, 353
Table 2. The rate constant calculation for every experiment at different temperatures.
9 of 19
Table 2. The rate constant calculation for every experiment at different temperatures.
k
ln k
T (˝ C)
T (K)
1/T
ln k
T (°C)
T (K)
1/T
k
´5
´9.313 175 175 448.15
448.15
9.0243
−5 10
9.0243
× 10ˆ
−9.313
0.002230.00223
´5
Combination
of chemical
Combination
of chemical
reactionreaction
control and
´9.9007
125
398.15
0.00251
5.0135
ˆ
10
5.0135 × 10−5 ´5−9.9007
125
398.15 0.00251
productcontrol
layer diffusion
and product layer
´10.4109
75
348.15
0.00287
3.0101 ˆ
10
3.0101 × 10−5 ´5
−10.4109
75
348.15 0.00287
´11.5101
25
298.15
0.00335
1.0028 ˆ 10
diffusion
1.0028 × 10−5
−11.5101
25
298.15 0.00335
´5
´9.3086 175 175 448.15
448.15
9.0635
−5 10 −9.3086
9.0635
× 10ˆ
0.002230.00223
´5
´10.1205
125
398.15
4.0244
ˆ
10
−5
4.0244 × 10
−10.1205
125
398.15 0.002510.00251
Product layer diffusion
´5
Product layer diffusion
75
348.15
´10.8121
2.0153
ˆ
10
−5
2.0153 × 10
−10.8121
75
348.15
0.002870.00287
25
298.15
0.00335
1.0014 ˆ
10´5 ´11.5115
−5
1.0014 × 10
−11.5115
25
298.15 0.00335
ln k
Kinetic Analysis
Kinetic Analysis
-9.2
-9.4
-9.6
-9.8
-10
-10.2
-10.4
-10.6
-10.8
-11
-11.2
-11.4
-11.6
-11.8
Combination of chemical reaction
control and product layer diffusion
Product layer Diffusion
y = -1.9199x - 5.0231
R² = 0.9917
y = -1.9266x - 5.1592
R² = 0.9764
2.15
2.3
2.45
2.6
2.75
2.9
1/T
(×10-3)
3.05
3.2
3.35
3.5
Figure 8. The Arrhenius plots for the extraction of Mg from forsterite using two selected models of the
Figure 8. The Arrhenius plots for the extraction of Mg from forsterite using two selected models of the
combination of product layer diffusion and chemical reaction control and the product layer diffusion.
combination of product layer diffusion and chemical reaction control and the product layer diffusion.
The rate constants generated in this study on forsterite are faster than corresponding rate
constants
reported generated
by Pokrovsky
andstudy
Schotton[48],
who only
worked
at the
temperature rate
of 25constants
°C.
The
rate constants
in this
forsterite
are faster
than
corresponding
˝
Comparison
of
photomicrographs
showing
the
surface
of
forsterite
presented
by
Pokrovsky
and
reported by Pokrovsky and Schott [48], who only worked at the temperature of 25 C. Comparison of
Schott [48] and the
samples
from
this study
(see Figure
2) demonstrates
they consist
euhedral
photomicrographs
showing
the
surface
of forsterite
presented
by Pokrovsky
and of
Schott
[48]and
and the
larger crystals due to small adhering particles and their agglomeration in acid solution at the
samples from this study (see Figure 2) demonstrates they consist of euhedral and larger crystals due to
temperature of 25 °C. In the present study, the rate constants are increased because of the far higher
small adhering particles and their agglomeration in acid solution at the temperature of 25 ˝ C. In the
density of activated sites per unit surface area at higher temperatures.
present study, the rate constants are increased because of the far higher density of activated sites per
unit surface
area at higher
temperatures.
2.6. Thermodynamic
Considerations
on the Mineral Carbonation
As regardsConsiderations
the effect of temperature
on Carbonation
the system, two physical mechanisms interact; the
2.6. Thermodynamic
on the Mineral
increase of olivine solubility with temperature and, conversely, the reduction of magnesite solubility
As
regards
the effect
of temperature
on thewas
system,
two by
physical
mechanisms
increase
with
increasing
temperature.
Heat treatment
reported
O’Connor
et al. [49] tointeract;
remove the
sorbed
water solubility
and activatewith
the mineral
surface.and,
Theyconversely,
suggested that
carbonation
phase occurssolubility
quickly in with
of olivine
temperature
thethe
reduction
of magnesite
the olivine
powders Heat
at high
temperatures,
from which
we expectet an
increase
in the sorbed
availablewater
increasing
temperature.
treatment
was reported
by O’Connor
al. [49]
to remove
reactivitythe
of mineral
with
removal
of sorbed
water
and CO2 uptake
by the
mineral
surfacein the
and activate
mineralsurface
surface.
They
suggested
that
the carbonation
phase
occurs
quickly
and, therefore, more rapid reaction.
olivine powders at high temperatures, from which we expect an increase in the available reactivity of
Additionally, elevated temperature causes an increase of the olivine dissolution rate, which
enhances the overall reaction kinetics. Our results concur with this (see Figure 5) up to the maximum
temperature studied (175 °C). The lowest CO2 concentration considered was 0.5% (volume fraction
of CO2 unconsumed) indicating olivine samples have efficiently precipitated CO2 as carbonate phase,
which is in agreement with the findings of Kwon et al. [18]. Interestingly, the CO2 sequestration capacity
of olivine mineral seen here over the temperature range of 150–175 °C was higher than in previous
Molecules 2016, 21, 353
10 of 19
mineral surface with removal of sorbed water and CO2 uptake by the mineral surface and, therefore,
more rapid reaction.
Additionally, elevated temperature causes an increase of the olivine dissolution rate, which
enhances the overall reaction kinetics. Our results concur with this (see Figure 5) up to the maximum
temperature studied (175 ˝ C). The lowest CO2 concentration considered was 0.5% (volume fraction of
CO2 unconsumed) indicating olivine samples have efficiently precipitated CO2 as carbonate phase,
which is in agreement with the findings of Kwon et al. [18]. Interestingly, the CO2 sequestration capacity
of olivine mineral seen here over the temperature range of 150–175 ˝ C was higher than in previous
studies [50–55]. We tentatively suggest that this effect is due to either an increase of the available
surface area or available moisture (or a combination of both factors). At the highest temperatures,
(e.g., the range of 150–175 ˝ C), Hänchen et al. [56] suggest the olivine samples released more Mg into
the solution, which seems a reasonable assumption to adopt here and is confirmed by the calculation
shown below. At lower temperatures (less than 100 ˝ C), King et al. [12] found that MgCO3 precipitation
was kinetically hindered due to the high activation energy for the de-solvation of the strongly hydrated
Mg2+ ions (Reaction (2)). According to Hänchen et al. [57] during low temperature experiments,
hydromagnesite [Mg5 (CO3 )4 (OH)2 ‚4H2 O] was precipitated from the solution (Reaction (3)) instead
of magnesite (MgCO3 ). When the temperature was increased (up to 125 ˝ C), hydromagnesite was
transformed to MgCO3 albeit slowly (more than 15 h).
Mg2 SiO4
(sol)
` 2CO2
(gas)
` 2H2 O (liq.) Ñ2 MgCO3
(sol)
` H4 SiO4
(aq.)
5Mg2 SiO4 (sol) ` 8CO2 (gas) ` 20H2 O (liq.) Ñ2r4MgCO3 ‚MgpOHq2 ‚4H2 Os (sol q ` 5H4 SiO4 (aq.)
(Rx. 2)
(Rx. 3)
The work of Hänchen et al. [57] is intriguing, as we found the evidence of hydromagnesite
formation in our samples, suggesting that this phase may be as a metastable “intermediate” phase [38]
with respect to magnesite, but kinetically favored as a first reaction product under certain conditions
(e.g., at low temperature conditions).
2.7. Geochemical Modeling
Equilibrium simulations were performed using PHREEQC (version 2.18—Parkhurst and
Appello [58] with data from the Lawrence Livermore National Laboratory (LLNL, Livermore, CA,
USA, supplied with the code). The first, very simple calculations examined the saturation state of
a solution (initially pure water) when equilibrated with either magnesite, or both magnesite and
carbon dioxide gas. Figure 9 shows the saturation index of both solutions with respect to magnesium
phases which could be formed from the elements in the simulated system. In each case, the solutions
suggest saturation with brucite, Mg(OH)2 , would be reached above 25 ˝ C, so the simulations were
repeated, but allowing brucite to precipitate when it would otherwise be oversaturated at ambient
pressure and it is these conditions which are shown in Figure 9. It is interesting to note that the
solution remains undersaturated with respect to hydromagnesite at all temperatures and the influence
of CO2 saturation is negligible. This reinforces the view that hydromagnesite is an artifact of rapid
precipitation, i.e., a process under kinetic control, in preference to magnesite (in line with experiments
by Konigsberger et al. [59] and Power et al. [60]), which, under equilibrium conditions, is the dominant
magnesium-bearing phase.
Secondly, attempts were made to simulate the reaction of CO2 gas with forsterite in the presence
of water, with and without varying amounts of HCl. These results also suggest considerable
disequilibrium, in that magnesite would not be the dominant “sink” for magnesium in the presence of
the silicate ions liberated from forsterite dissolution. Examination of the saturation state of solutions in
which CO2 was sequentially added to forsterite-water mixtures, always suggested a greater tendency
for sepiolite precipitation, in preference to magnesite. Over the temperature interval 1 to 175 ˝ C, CO2
gas was sequentially “reacted” (numerically) with an excess of forsterite in the presence of water and
these simulations were repeated in the presence of HCl to represent any carry-over of the acid from
Molecules 2016, 21, 353
11 of 19
the previous processing step. Although we recognize that the presence of hydrochloric acid has little
influence on the thermodynamic equilibrium position, it has a major role in promoting a rapid increase
in the forsterite dissolution kinetics, so it was considered here. In all cases, thermodynamic equilibrium
was only reached when sepiolite was allowed to precipitate, which effectively prevented magnesite
formation. No evidence of other phases being important in these reactions was seen and the influence
of residual
HCl was negligible.
Molecules 2016, 21, 353
11 of 19
Saturation state of solution equilibrated with magnesite and CO2
1.0
-2.0
pH 10.6
-3.0
-4.0
-5.0
Brucite saturation
-1.0
Equilibrium with magnesite only
Saturation indices of solution with
respect to magnesium phases
0.0
pH 9.7
pH 8.2
pH 8.9
Magnesite
Hydromagnesite
Brucite
Artinite
CO2(g)
-6.0
-7.0
-8.0
0
25
50
75
100
125
150
175
Temperature / Degrees celsius
Figure 9. Saturation state of a solution equilibrated with magnesite and carbon dioxide gas (PCO2 = 4.8
Figure 9. Saturation state of a solution equilibrated with magnesite and carbon dioxide gas
bar), allowing brucite to precipitate above 25 °C.
(PCO2 = 4.8 bar), allowing brucite to precipitate above 25 ˝ C.
Experimentally, no evidence of sepiolite formation was seen and magnesite certainly dominates
the solid phase assemblage
afterof
carbonation
this workwas
and in
the and
previous
studies certainly
cited above.
Our
Experimentally,
no evidence
sepiolite in
formation
seen
magnesite
dominates
observations
are that forsterite
dissolves andinmagnesite
silica precipitate
as reaction
the solid
phase assemblage
after carbonation
this workand
andamorphous
in the previous
studies cited
above. Our
products are
(see that
Figures
2 and 3).dissolves
Experimental
on theand
stability
of sepiolite
haveprecipitate
demonstrated
observations
forsterite
andstudies
magnesite
amorphous
silica
as that
reaction
if the silica activity value is high in alkaline solutions, sepiolite precipitates [61–63]. From this, we must
products (see Figures 2 and 3). Experimental studies on the stability of sepiolite have demonstrated
assume that the kinetics of sepiolite formation are very slow indeed and consequently it was excluded
that if the silica activity value is high in alkaline solutions, sepiolite precipitates [61–63]. From this,
from the mineral assemblage in subsequent calculations. Figure 10 shows the solution chemistry
we must
assume
thewhere
kinetics
of sepiolite
formation
are very
indeed and consequently it
predicted
in athat
system
forsterite
is carbonated
by gaseous
COslow
2 in the presence of water. One
was excluded
from
the
mineral
assemblage
in
subsequent
calculations.
Figure
10 shows
the solution
mole of forsterite (in excess) was sequentially reacted with 0.2 mol of CO2 gas, allowing
equilibrium
chemistry
predicted
a system
where
forsterite
is carbonated
by gaseous
CO2sepiolite
in the presence
to be reached
within
both
amorphous
silica
and all possible
magnesium
phases except
(which of
prevented
forming
in the simulations).
At all temperatures
°C and
175allowing
°C
water.was
One
mole of from
forsterite
(in excess)
was sequentially
reacted withbetween
0.2 mol 25
of CO
2 gas,
carbonation
is predicted
causeboth
dissolution
of forsterite
precipitation
of magnesite
and amorphous
equilibrium
to be
reachedtowith
amorphous
silicaand
and
all possible
magnesium
phases except
silica,
with no
other
magnesium-bearing
phases
reaching
equilibrium,
sepiolite which
was25 ˝ C
sepiolite
(which
was
prevented
from forming
in the
simulations).
At allexcept
temperatures
between
allowed
to
remain
oversaturated
in
the
pore
solution.
˝
and 175 C carbonation is predicted to cause dissolution of forsterite and precipitation of magnesite
The solid phase chemistry is shown in Figure 11, which plots the ratio of forsterite dissolved
and amorphous silica, with no other magnesium-bearing phases reaching equilibrium, except sepiolite
against magnesite precipitated (mole ratio) against the number of moles of carbon dioxide consumed
whichbywas
allowed to remain oversaturated in the pore solution.
the reaction. This is in effect the “yield” of the reaction, showing the moles of product formed as
The
solid
phase
chemistry
is shown
in Figure
11, which
plots the ratio
of forsterite
dissolved
a function of
the moles
of reactant
consumed.
The theoretical
stoichiometry
(see Equation
(1), above)
against
magnesite
precipitated
ratio)one
against
the
number
of moles
of carbon
dioxide
consumed
is that
two moles
of forsterite(mole
react with
mole of
CO
2 to produce
one mole
of magnesite
and
one
by the
reaction.
This is silica.
in effect
thesimulations
“yield” of suggest
the reaction,
showing
the
product
formed
mole
of amorphous
These
that this
condition
is moles
reachedofquite
rapidly.
After reaction
of only
0.9of
mol
of CO2, consumed.
the reaction isThe
97%theoretical
complete in stoichiometry
stoichiometric terms;
that is to (1),
as a function
of the
moles
reactant
(see Equation
sayisthe
the CO
reacts to form
very
little
partitions
liquid
phase
above)
thatbulk
twoofmoles
of2 forsterite
reactmagnesite
with one and
mole
of CO
produceinto
onethe
mole
of magnesite
2 to
(cf.
Figure
10).
Figure
11
shows
that
the
temperature
does
have
an
effect
on
the
extent
of
reaction
andquite
and one mole of amorphous silica. These simulations suggest that this condition is reached
efficiency of carbonation in terms of the overall yield. The pseudo steady-state conditions established
rapidly. After reaction of only 0.9 mol of CO2 , the reaction is 97% complete in stoichiometric terms;
after reaction of 0.2 mol of CO2 are maintained if further reaction is simulated, such that the solution
that is to say the bulk of the CO2 reacts to form magnesite and very little partitions into the liquid
chemistry remains constant, as does the ratio of forsterite and CO2 consumed to the quantities of
amorphous silica and magnesite precipitated. This is shown in Figure 12, where an excess of carbon
dioxide eventually exhausts the reserve of forsterite, maintaining the stoichiometry of the reaction
throughout. It must be stated that since the simplistic geochemical reaction path modeling is used in
Molecules 2016, 21, 353
12 of 19
phase (cf. Figure 10). Figure 11 shows that the temperature does have an effect on the extent of
reaction and efficiency of carbonation in terms of the overall yield. The pseudo steady-state conditions
established after reaction of 0.2 mol of CO2 are maintained if further reaction is simulated, such that
the solution chemistry remains constant, as does the ratio of forsterite and CO2 consumed to the
quantities of amorphous silica and magnesite precipitated. This is shown in Figure 12, where an
excess of carbon dioxide eventually exhausts the reserve of forsterite, maintaining the stoichiometry
of the
reaction
It must be stated that since the simplistic geochemical reaction
Molecules
2016, throughout.
21, 353
12 of 19path
modeling
is 2016,
used21,in353this study, many physical processes, such as material deformation, pore-fluid
Molecules
12 of 19flow,
study,advection,
many physical
processes, such as material
deformation,
pore-fluid
flow, be
heat
transfer, in
heat this
transfer,
diffusion/dispersion,
are ignored,
although
they should
considered
this
study,asdiffusion/dispersion,
many
processes,
such
asalthough
material
deformation,
heat work,
transfer,
advection,
are
ignored,
they
should bepore-fluid
consideredflow,
in
as
future
work,
they physical
are expected
to be
important
during
industrial-scale-up.
In future
addition,
because
advection,
diffusion/dispersion,
are
ignored,
although
they
should
be
considered
in
future
work,
as
they are expected to be important during industrial-scale-up. In addition, because the chemical
the chemical
dissolution-front instability is also neglected in this study, many important factors, such
they
are
expected
to
be
important
during
industrial-scale-up.
In
addition,
because
the
chemical
dissolution-front instability is also neglected in this study, many important factors, such as mineral
as mineral reactive surface area, mineral dissolution ratio, solute dispersion, medium anisotropy,
dissolution-front
instability
is dissolution
also neglected
in solute
this study,
many important
factors, such
as mineral
reactive
surface area,
mineral
ratio,
dispersion,
medium anisotropy,
medium
and
medium
and
fluid compressibility,
have
also
been
ignored.
To consider
these
factors
appropriately,
reactive
surface
area, mineral
dissolution
ratio,
solute
medium
anisotropy,
medium
and
fluid
compressibility,
have also
been ignored.
To dispersion,
consider these
factors
appropriately,
more
morecomprehensive
comprehensive
chemical-transport
modeling
will
be required.
fluid
compressibility,
have also been
ignored.
To
consider
these factors appropriately, more
chemical-transport
modeling
will be
required.
comprehensive chemical-transport modeling will be required.
Ions in solution after carbonation
Ions in solution after carbonation
0.0008
Moles
of element
per dm^-3
Moles
of element
per dm^-3
0.0008
0.0007
0.0007
0.0006
pH 8.0
0.0006
0.0005
pH 8.0
0.0005
0.0004
0.0004
0.0003
0.0003
0.0002
C(4)
pH 10.5
C(4)
Si
pH 10.5
Si
Mg
Mg
pH 9.2
0.0002
0.0001
pH 9.2
0.0001
0
0 25
25
50
50
75
100
125
Temperature
/ Degrees celsius
75
100
125
150
175
150
175
Temperature / Degrees celsius
Figure 10. Solution chemistry predicted after reaction of excess forsterite (1 mol) with carbon dioxide
Figure 10. Solution chemistry predicted after reaction of excess forsterite (1 mol) with carbon dioxide
Figure
Solution
predicted after
reaction
(1 mol) with carbon dioxide
gas
(0.210.
mol
of CO2)chemistry
over the temperature
interval
(25 of
°Cexcess
to 175 forsterite
°C).
gas (0.2 mol of CO2 ) over the temperature interval (25 ˝ C to 175 ˝ C).
Moles
of magnesite
precipitated
per mole
Moles
of magnesite
precipitated
per mole
forsterite
reacted
during
carbonation
forsterite
reacted
during
'wet''wet'
carbonation
gas (0.2 mol of CO2) over the temperature interval (25 °C to 175 °C).
2.00
Yield of carbonation reaction / moles per mole of solid
Yield of carbonation reaction / moles per mole of solid
2.00
1.95
25°C
25°C
100°C
1.95
1.90
100°C
125°C
1.90
1.85
125°C
150°C
1.85
1.80
150°C
175°C
1.80
1.75
175°C
1.75
1.70
1.70 0
0
0.005
0.01
0.015
0.01 with excess
0.015
Moles of 0.005
carbon dioxide reacted
forsterite
0.02
0.02
Moles of carbon dioxide reacted with excess forsterite
Figure 11. Solid phase chemistry predicted after reaction of excess forsterite (1 mol) with carbon
Figure 11.
chemistry predicted
reaction
excess
forsterite
mol)ofwith
carbon
dioxide
gasSolid
overphase
the temperature
interval after
(25 °C
to 175of°C)
expressed
as (1
moles
magnesite
Figure 11. Solid phase chemistry predicted after reaction of excess forsterite (1 mol) with carbon dioxide
precipitated/moles
forsterite
reacted,
as
a
function
of
carbon
dioxide
consumed.
dioxide gas over the
temperature
interval
(25
°C
to
175
°C)
expressed
as
moles
of
magnesite
gas over the temperature interval (25 ˝ C to 175 ˝ C) expressed as moles of magnesite precipitated/moles
precipitated/moles forsterite reacted, as a function of carbon dioxide consumed.
forsterite reacted, as a function of carbon dioxide consumed.
Molecules 2016, 21, 353
13 of 19
Molecules 2016, 21, 353
13 of 19
Moles of solids in equilibrium assemblage
Moles of solid present in assemblage
2.5
2.0
1.5
Forsterite
1.0
Magnesite
Quartz
0.5
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Moles of carbon dioxide reacted with 1 mole forsterite
Figure 12. Solid phase chemistry predicted after reaction of one mole of forsterite with excess carbon
Figure 12. Solid phase chemistry predicted after reaction of one mole of forsterite with excess carbon
dioxide gas at 25 °C, showing the stoichiometric relationship between moles of reactant (forsterite)
dioxide gas at 25 ˝ C, showing the stoichiometric relationship between moles of reactant (forsterite)
consumed to moles of product (magnesite and quartz) precipitated.
consumed to moles of product (magnesite and quartz) precipitated.
3. Experimental and Modeling Section
3. Experimental and Modeling Section
3.1. Materials
3.1. Materials
Three kilograms of olivine mineral were excavated and collected from different depths of the
Mont
peridotite
stratum
in Malaysia.
Although and
this collected
source is from
local, different
the results
can be
ThreeTawai
kilograms
of olivine
mineral
were excavated
depths
of the
considered
as
broadly
representative
of
peridotite
as
the
mineral
composition
is
typical.
The
samples
Mont Tawai peridotite stratum in Malaysia. Although this source is local, the results can be considered
were crushed
by a mechanical
grinder
into
four different
size ranges:
µm, 75–125
as broadly
representative
of peridotite
asand
the sieved
mineral
composition
is typical.
The<75
samples
were µm,
crushed
125–200
µm
and
>200
µm.
Subsequently,
they
were
dried
to
a
constant
weight
at
120
°C
for
2
h.
Ten
by a mechanical grinder and sieved into four different size ranges: <75 µm, 75–125 µm, 125–200 µm
grams of each size fraction were mixed separately in 500 mL of 0.01 M hydrochloric acid (HCl, QRëC
and >200 µm. Subsequently, they were dried to a constant weight at 120 ˝ C for 2 h. Ten grams of each
reagent grade, 37%) solution to remove impurities.
size fraction were mixed separately in 500 mL of 0.01 M hydrochloric acid (HCl, QRëC reagent grade,
The olivine samples were characterized using scanning electron microscopy (SEM, S-4700 Hitachi,
37%)Tokyo,
solution
to remove
impurities.
Japan)
and X-ray
diffraction (XRD, X’Pert powder, PANalytical, Almelo, The Netherlands),
The
olivine
samples
were characterized
using
electron microscopy
(SEM,
S-4700 Hitachi,
while elemental composition
was measured
by scanning
X-ray fluorescence
(XRF, PW-1410,
PANalytical,
Tokyo,
Japan)
and
X-ray
diffraction
(XRD,
X’Pert
powder,
PANalytical,
Almelo,
The
Netherlands),
Almelo, The Netherlands). The carbonation yield of the olivine was determined using a total
carbon
whileanalyzer
elemental
composition
wasCorp.,
measured
by X-ray
fluorescence (XRF, PW-1410, PANalytical, Almelo,
(TCA,
CS844, LECO
St. Joseph,
MI, USA).
The Netherlands). The carbonation yield of the olivine was determined using a total carbon analyzer
3.2.
Experimental
(TCA,
CS844,
LECOApparatus
Corp., St. Joseph, MI, USA).
The experimental equipment for CO2 mineralization consisted of a CO2 analyzer system
3.2. Experimental
mounted in aApparatus
flow system connected to a cylindrical (500 mm × 10 mm (d)) autoclave reactor with
the
to supplyequipment
diluted COfor
2 and flushing with pure N2 (Figure 13). CO2 was introduced into the
Themeans
experimental
CO2 mineralization consisted of a CO2 analyzer system mounted
reactor at different partial pressures (up to 30%). Two flow meters (FM-1050, Matheson Tri-gas,
in a flow system connected to a cylindrical (500 mm ˆ 10 mm (d)) autoclave reactor with the means to
Basking Ridge, NJ, USA) were used to control the flow rate of the inlet gases. The autoclave reactor
supply diluted CO2 and flushing with pure N2 (Figure 13). CO2 was introduced into the reactor at
was loaded with mineral and acid, then placed in a furnace where the temperature was measured
different
partial pressures (up to 30%). Two flow meters (FM-1050, Matheson Tri-gas, Basking Ridge,
and controlled using a thermocouple inserted directly into the reactor.
NJ, USA)Awere
used
togenerator
control the
flow
of the inlet
The autoclave
reactorwas
wasmeasured
loaded with
water
vapor
was
usedrate
to humidify
the gases.
gas stream.
CO2 consumption
mineral
and
acid,
then
placed
in
a
furnace
where
the
temperature
was
measured
and
controlled
using
as the difference between supply and vent level at fixed flow (with integration over time) using an
a thermocouple
inserted
directly
into
the
reactor.
optical IR-sensor (GMP221, Vaisala Oyj, Helsinki, Finland), according to Equation (1):
A water vapor generator was used to humidify the gas stream. CO2 consumption was measured
(pCO − pCO
) × ∆t × Q
as the difference between CO
supply
and vent=level at fixed flow (with integration over time) using
uptake
(1) an
R×T×M
optical IR-sensor (GMP221, Vaisala Oyj, Helsinki, Finland), according to Equation (1):
ˆ
CO2 uptake
mol
g
˙
`
˘
n
ÿ
pCO2 in ´ pCO2 out i ˆ ∆t ˆ Q
“
RˆTˆM
i
(1)
Molecules 2016, 21, 353
14 of 19
Molecules 2016, 21, 353
14 of 19
where where
pCO2pCO
pCO
aremean
mean
value
of 2pCO
(atm)
at the
inflow
outflowsupplyP
(equilibrium
out 2and
2 2inin are
out and
pCO
value
of pCO
(atm)2 at
the inflow
and
outflowand
(equilibrium
supplyP
= 4.8
∆t Q
and
areinterval
time interval
(min)
flow rate
(L/min), respectively,
and T are gas
= 4.8
bar),bar),
Δt and
areQ
time
(min) and
flowand
rate (L/min),
respectively,
R and T are gasRconstant
(0.082057
l.atm/mol.K)
and temperature
(K), respectively,
and M is the and
massM
of forsterite
(g). of forsterite (g).
constant
(0.082057
l.atm/mol.K)
and temperature
(K), respectively,
is the mass
13. Schematic diagram of experimental set-up for CO2 sequestration in chemically pretreated
FigureFigure
13. Schematic
diagram of experimental set-up for CO2 sequestration in chemically pretreated
peridotite mineral from Mount Tawai, Malaysia.
peridotite mineral from Mount Tawai, Malaysia.
3.3. Experimental Procedure
3.3. Experimental Procedure
In this study, 1 M HCl was used for dissolution of Mg from the mineral matrix of peridotite and
Inthe
this
study, 1 experiment
M HCl was
used
for dissolution
of vessel.
Mg from
thea stoichiometric
mineral matrix
of peridotite
dissolution
was
conducted
in a separate
When
amount
of HCl and
solution was
added to olivine
powder in ainreaction
vesselvessel.
and stirred
with
a magnetic stirrer
at 60 °Cof HCl
the dissolution
experiment
was conducted
a separate
When
a stoichiometric
amount
forwas
two added
h, the entire
mass formed
a slurry.
This slurry
was then
intoa the
autoclave
reactor
solution
to olivine
powder
in a reaction
vessel
andtransferred
stirred with
magnetic
stirrer
at 60 ˝ C
and neutralized by adding a base (NaOH: Fisher Chemicals reagent grade, with purity 99.999%) until
for two h, the entire mass formed a slurry. This slurry was then transferred into the autoclave reactor
the final pH was increased to above seven. Before heating the reactor, it was purged with nitrogen in
and neutralized by adding a base (NaOH: Fisher Chemicals reagent grade, with purity 99.999%) until
order to replace the air inside the reactor, then the reactor was pre-heated to 175 °C for 1 h to dry the
the final
pH and
wascooled
increased
to above
Before heating the reactor, it was purged with nitrogen in
slurry
to ambient.
COseven.
2 gas (SFE grade, with a purity of 99.99% contained in a dip-tube
˝ C for 1 h to dry the
order to
replace
the
air
inside
the
reactor,
then
was
pre-heated
to 175
cylinder and purchased from MOX Company,the
KL,reactor
Malaysia)
was
passed through
the dried
slurry at
typical
flue gas
level (15 vol
%)2 or
4.8(SFE
bar pressure
tot = 32
bar) in the
presence of
slurry aand
cooled
to ambient.
CO
gas
grade, (P
with
a purity
of absence
99.99% or
contained
inwater
a dip-tube
vapor.
temperature
of MOX
the carbonation
process
was studiedwas
overpassed
the interval
from ambient
to slurry
cylinder
andThe
purchased
from
Company,
KL, Malaysia)
through
the dried
175
°C.
at a typical flue gas level (15 vol %) or 4.8 bar pressure (Ptot = 32 bar) in the absence or presence of
The reactions involved in the extraction of Mg and carbonation are as follows:
water vapor. The temperature of the carbonation process was studied over the interval from ambient
4HCl (liquid) + Mg2SiO4 (solid) → 2MgCl2 (aqueous) + SiO2 (solid) + 2H2O (liquid)
(Rx. 4)
to 175 ˝ C.
The reactions
involved
in the
extraction
of Mg and
carbonation
as follows:
In this reaction
scheme
(Reaction
(4)), forsterite
dissolves
in HCl, are
forming
soluble MgCl2 (Mg2+
remains in solution), and leaves behind insoluble SiO2. [15] Magnesium hydroxide [Mg(OH)2] is
precipitated
by `
neutralization
(Reaction
(5)). Then,
by passing
Mg(OH)
2 is converted(Rx. 4)
4HCl (liquid)
Mg2 SiO4 with
Ñ 2MgCl
` SiO
`2,2H
2 O (liquid)
(solid)NaOH
2 (aqueous)
2 (solid)CO
to MgCO3 in a gas-solid carbonation process (Reaction (6)).
2+
In this reaction scheme (Reaction
forsterite
dissolves
in HCl, forming soluble MgCl
MgCl2(4)),
+ 2NaOH
→ Mg(OH)
2 + 2NaCl
(Rx. 5)2 (Mg
remains in solution), and leaves behind insoluble SiO2 . [15] Magnesium hydroxide [Mg(OH)2 ] is
Mg(OH)2 + CO2 → MgCO3 + H2O
(Rx. 6)
precipitated by neutralization with NaOH (Reaction (5)). Then, by passing CO2 , Mg(OH)2 is converted
of water
vapor level
on carbonation
to MgCO3 The
in a effect
gas-solid
carbonation
process
(Reactionwas
(6)).studied in the range 5–20 vol % H2O,
corresponding to a range of relative humidity 18%–72% RH. After completion of the experiment, the
samples were collected and
filtered
using <75 µm pore size Whatman filter papers. The MgCO3(Rx. 5)
MgCl
2 ` 2NaOH Ñ MgpOHq2 ` 2NaCl
MgpOHq2 ` CO2 Ñ MgCO3 ` H2 O
(Rx. 6)
The effect of water vapor level on carbonation was studied in the range 5–20 vol % H2 O,
corresponding to a range of relative humidity 18%–72% RH. After completion of the experiment,
Molecules 2016, 21, 353
15 of 19
the samples were collected and filtered using <75 µm pore size Whatman filter papers. The MgCO3
formed in the sample was quantified by titration against HCl. For the titration, a certain amount of
solid (0.5 g) was weighed in a conical flask. Then, 20 mL of 1 M HCl was added into the flask and was
allowed about two hours to react with the MgCO3 . The excess HCl was then back-titrated using 0.1 M
NaOH solution. From the difference in titrant volume, the HCl consumed was calculated from which
the content of MgCO3 was deduced.
3.4. Estimation of Carbonation Yield
The extent of CO2 mineral carbonation YCO2 was estimated using the TCA method,
ˆ which
˙ is
1
based on the mineralogy of samples tested and the capacity of carbon sequestrated
[64].
RCO2
RCO2 is considered as the weight fraction of CO2 that can be trapped in a specific amount of mineral.
According to Gadikota et al. [64] the capacity of CO2 sequestration in forsterite (Mg2 SiO4 ) can be
expressed as follows:
ˆ
˙
YMg
WCO2
1
ˆ MWCO2
(2)
“
“
Wf o
RCO2
MWMg
where W f o and WCO2 are weights of forsterite before its mineral carbonation and CO2 sequestrated
in the solid phase (i.e., magnesite), respectively. YMg is the mass fraction of Mg2+ in the forsterite
(i.e., 34.55%) that can react with carbon dioxide to form stable magnesite. MWmg and MWCO2 are
the formula weights of Mg2+ (48.61 mol/g in the forsterite) and CO2 (44.01 mol/g), respectively, in
the carbonated forsterite (2MgCO3 & SiO2 —see (Reaction (1))). Therefore, YCO2 is the amount of
carbon dioxide sequestrated (as magnesite) relative to the maximum capacity of CO2 sequestration in
1
forsterite p
“ 31.27%q.
RCO2
ff
3.67 ˆ weight fraction of carbon in MgCO3
`
˘ ˆ 100%
“ RCO2 ˆ
1 ´ 3.67 ˆ weight fraction of carbon in MgCO3
«
YCO2
(3)
where 3.67 is the CO2 /C mass ratio. The carbonation yields of the forsterite and the effect on these of
key empirical variables—reaction temperature, time, and particle size, are compared below.
3.5. Modeling System and Kinetic Analysis
Thermodynamic calculations were performed with the PHREEQC program software (version 2.18)
with data from the LLNL [58]. Forsterite is the Mg-end member of the forsterite-fayalite solid solution
series, and is included in the LLNL database. This program was used to estimate the dissolution and
carbonation of forsterite samples in order to predict CO2 uptake processes and potentials. Moreover,
thermodynamic equilibrium constants for the mineral carbonation reactions of forsterite were provided
by model databases. In doing so, reaction kinetics were implemented by using a BASIC interpreter.
The possibility of implementing reactions kinetic into the code as BASIC statements was also used to
predict the reaction progress over time. Consequently, the quality and validity of the model system
and the determined rate and equilibrium parameters were verified against the results of carbonation
experiments with forsterite samples. The data from a sequence of laboratory efforts were applied for
that purpose, which were performed in the aqueous autoclave mini reactor.
The kinetic analysis of the forsterite dissolution rates was determined according to “standard
integral analysis Levenspiel’s method” [65] in Mg-rich solution using HCl. The results were set into several
heterogeneous reaction models represented by integral rate equations and then the multiple regression
coefficients (R) were calculated. The shrinking core model described by Dri et al. [66] was applied for the
constant size of forsterite particles. Based on this method, reaction rates take place at the outer surface
of the unreacted particles, and heterogeneous reactions are controlled by the product layer diffusion
(Equation (4)), film diffusion (Equation (5)) and chemical reaction control (Equation (6)). In addition, the
Molecules 2016, 21, 353
16 of 19
possibility of having a compound effect of “chemical reaction control” and “product layer diffusion” was
investigated using Equation (7).
2
kt “ 1 ´ 3p1 ´ XE q 3 ` 2 p1 ´ XE q
(4)
1
3
kt “ 1 ´ 3p1 ´ XE q
(5)
kt “ XE
(6)
2
1
3
3
kt “ r1 ´ 3p1 ´ XE q ` 2 p1 ´ XE q s ` r1 ´ 3p1 ´ XE q s
(7)
In these equations. “t” is time (s) and “k” (s´1 ) and “XE ” denote the rate constant and the extent
of reaction, respectively.
4. Conclusions
This work has experimentally and numerically modeled the process by which carbon dioxide gas
may be sequestered, in situ by reaction with forsterite and/or its extracted intermediate brucite (ex situ)
in the presence of moisture. In both cases, we have found that the reaction is favored resulting in a high
carbonate yield; going almost to completion with the bulk of the carbon partitioning into magnesite
and that very little remaining in solution. In the presence of water vapor, the degree of mineral
carbonation was increased, we suggest, due to an alternative carbonation pathway providing for faster
reaction kinetics [64]. Despite the observations made in other studies, we suggest that hydromagnesite
is an intermediate in these carbonation experiments but is converted into magnesite on the (hours)
timescale of reaction studied here, although the mechanism is as yet unclear. Hydromagnesite is
less desirable as final product as it corresponds to only 80% sequestration (relative to magnesite) on
a molar basis. Moreover, we recognize that this system is itself not at thermodynamic equilibrium,
but maybe take a very long period and recrystallization as sepiolite (Mg4 Si6 O15 (OH)2 ¨ 6H2 O) is one
possible outcome. However, we have found no evidence that this occurs over the period of one year.
The clay mineral sepiolite is known to form in nature from dolomite/silica assemblages in the presence
of water, but its formation kinetics are slow and we do not expect its formation to occur spontaneously
in an industrial carbonation plant. Thus, we propose that the carbonation of readily available forsterite
is a viable route for carbon sequestration.
From the computational perspective, a CO2 sequestration system with mineral carbonation can
be treated as a fully-coupled problem between rock deformation, pore-fluid flow, heat transfer, mass
transport, and chemical reaction processes. The results obtained from this study establish the viability
of a geochemical model that can be used in to simulate the dynamic processes involved in CO2
sequestration in the Mount Tawai peridotite, Malaysia.
Acknowledgments: The authors appreciate the Division of Research & Technology in IAU-Mahabad Branch for
financial support and the Universiti Teknologi Malaysia (UTM) for chemical and structural analyses. The authors
are also grateful to Sahar Zarza and Shahram M. Aminpour for valuable comments.
Author Contributions: “O.R. conceived and designed the experiments. J.H. assisted with manuscript draft and
consultation. R.J. did the interpretation of the due analyses. M.T. contributed in the simulation section and
consultation. A.B. prepared the samples and some parts of the revised manuscript.”
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Rahmani, O.; Junin, R.; Tyrer, M.; Mohsin, R. Mineral carbonation of red gypsum for CO2 sequestration.
Energy Fuels 2014, 28, 5953–5958. [CrossRef]
Molecules 2016, 21, 353
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
17 of 19
Sanderson, B.M.; O’Neill, B.C.; Kiehl, J.T.; Meehl, G.A.; Knutti, R.; Washington, W.M. The response of the
climate system to very high greenhouse gas emission scenarios. Environ. Res. Lett. 2011, 6, 034005. [CrossRef]
Bodman, R.W.; Rayner, P.J.; Karoly, D.J. Uncertainty in temperature projections reduced using carbon cycle
and climate observations. Nat. Clim. Chang. 2013, 3, 725–729. [CrossRef]
Bachu, S.; Gunter, W.D.; Perkins, E.H. Aquifer disposal of CO2 : Hydrodynamic and mineral trapping.
Energy Convers. Manag. 1994, 35, 269–279. [CrossRef]
Sanna, A.; Uibu, M.; Caramanna, G.; Kuusik, R.; Maroto-Valer, M.M. A review of mineral carbonation
technologies to sequester CO2 . Chem. Soc. Rev. 2014, 43, 8049–8080. [CrossRef] [PubMed]
Kaszuba, J.P.; Janecky, D.R.; Snow, M.G. Carbon dioxide reaction processes in a model brine aquifer at 200 ˝ C
and 200 bars: Implications for geologic sequestration of carbon. Appl. Geochem. 2003, 18, 1065–1080. [CrossRef]
Xu, T.F.; Apps, J.A.; Pruess, K. Numerical simulation of CO2 disposal by mineral trapping in deep aquifers.
Appl. Geochem. 2004, 19, 917–936. [CrossRef]
Maroto-Valer, M.M.; Fauth, D.J.; Kuchta, M.E.; Zhang, Y.; Andrésen, J.M. Activation of magnesium rich minerals
as carbonation feedstock materials for CO2 sequestration. Fuel Process. Technol. 2005, 86, 1627–1645. [CrossRef]
Palandri, J.L.; Kharaka, Y.K. Ferric iron-bearing sediments as a mineral trap for CO2 sequestration: Iron
reduction using sulfur-bearing waste gas. Chem. Geol. 2005, 217, 351–364. [CrossRef]
Flaathen, T.K.; Gislason, S.R.; Oelkers, E.H.; Sveinbjörnsdóttir, A.E. Chemical evolution of the Mt. Hekla,
Iceland, groundwaters: A natural analogue for CO2 sequestration in basaltic rocks. Appl. Geochem. 2009, 24,
463–474. [CrossRef]
Gislason, S.R.; Wolff-Boenisch, D.; Stefansson, A.; Oelkers, E.H.; Gunnlaugsson, E.; Sigurdardottir, H.;
Sigfusson, B.; Broecker, W.S.; Matter, J.M.; Stute, M.; et al. Mineral sequestration of carbon dioxide in basalt:
A pre-injection overview of the CarbFix project. Int. J. Greenh. Gas Control 2010, 4, 537–545. [CrossRef]
King, H.E.; Plümper, O.; Putnis, A. Effect of Secondary Phase Formation on the Carbonation of Olivine.
Environ. Sci. Technol. 2010, 44, 6503–6509. [CrossRef] [PubMed]
Qafoku, O.; Kovarik, L.; Kukkadapu, R.K.; Ilton, E.S.; Arey, B.W.; Tucek, J.; Felmy, A.R. Fayalite dissolution
and siderite formation in water-saturated supercritical CO2 . Chem. Geol. 2012, 332, 124–135. [CrossRef]
Rahmani, O.; Tyrer, M.; Junin, R. Calcite precipitation from by-product red gypsum in aqueous carbonation
process. RSC Adv. 2014, 4, 45548–45557. [CrossRef]
Lackner, K.S.; Wendt, C.H.; Butt, D.P.; Joyce, E.L.; Sharp, D.H. Carbon dioxide disposal in carbonate minerals.
Energy 1995, 20, 1153–1170. [CrossRef]
Olsson, J.; Bovet, N.; Makovicky, E.; Bechgaard, K.; Balogh, Z.; Stipp, S.L.S. Olivine reactivity with CO2 and H2 O
on a microscale: Implications for carbon sequestration. Geochim. Cosmochim. Acta 2012, 77, 86–97. [CrossRef]
Lackner, K.S.; Butt, D.P.; Wendt, C.H. Progress on binding CO2 in mineral substrates. Energy Convers. Mgmt.
1997, 38, S259–S264. [CrossRef]
Kwon, S.; Fan, M.H.; DaCosta, H.M.D.; Russell, A.G.; Armistead, G.R. Factors affecting the direct
mineralization of CO2 with olivine. J. Environ Sci. 2011, 23, 1233–1239. [CrossRef]
Larachi, F.; Gravel, J.P.; Grandjean, B.P.A.; Beaudoin, G. Role of steam, hydrogen and pretreatment in
chrysotile gas–solid carbonation: Opportunities for pre-combustion CO2 capture. Int. J. Greenh. Gas Control
2012, 6, 69–76. [CrossRef]
Fricker, K.J.; Park, A.H.A. Effect of H2 O on Mg(OH)2 carbonation pathways for combined CO2 capture and
storage. Chem. Eng. Sci. 2013, 100, 332–341. [CrossRef]
Fagerlund, J.; Zevenhoven, R. An experimental study of Mg(OH)2 carbonation. Int. J. Greenh. Gas Control
2011, 5, 1406–1412. [CrossRef]
McGrail, B.P.; Schaef, H.T.; Ho, A.M.; Chien, Y.J.; Dooley, J.J.; Davidson, C.L. Potential for carbon dioxide
sequestration in flood basalts. J. Geophys. Res. 2006, 111, 12201–12213. [CrossRef]
Oelkers, E.H.; Cole, D.R. Carbon Dioxide Sequestration A Solution to a Global Problem. Elements 2008, 4,
305–310. [CrossRef]
Matter, J.M.; Broecker, W.S.; Stute, M.; Gislason, S.R.; Oelkers, E.H.; Stefansson, A.; Wolff-Boenisch, D.;
Gunnlaugsson, E.; Axelsson, G.; Bjornsson, G. Permanent Carbon Dioxide Storage into Basalt: The CarbFix
Pilot Project, Iceland. Energy Procedia 2009, 1, 3641–3646. [CrossRef]
Schaef, H.T.; McGrail, B.P.; Owen, A.T. Basalt- CO2 –H2 O interactions and variability in carbonate
mineralization rates. Energy Procedia 2009, 1, 4899–4906. [CrossRef]
Molecules 2016, 21, 353
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
18 of 19
Tian, S.; Jiang, J.; Li, K.; Yan, F.; Chen, X. Performance of steel slag in carbonation–calcination looping for
CO2 capture from industrial flue gas. RSC Adv. 2014, 4, 6858–6862. [CrossRef]
Highfield, J.; Lim, H.Q.; Fagerlund, J.; Zevenhoven, R. Activation of serpentine for CO2 mineralization by flux
extraction of soluble magnesium salts using ammonium sulphate. RSC Adv. 2012, 2, 6535–6541. [CrossRef]
Chu, D.H.; Vinoba, M.; Bhagiyalakshmi, M.; Baek, I.H.; Nam, S.C.; Yoon, Y.; Kim, S.H.; Jeong, S.K. CO2
mineralization into different polymorphs of CaCO3 using an aqueous-CO2 system. RSC Adv. 2013, 3,
21722–21729. [CrossRef]
Gulliver, D.M.; Lowry, G.V.; Gregory, K.B. CO2 concentration and pH alters subsurface microbial ecology at
reservoir temperature and pressure. RSC Adv. 2014, 4, 17443–17453. [CrossRef]
Bearat, H.M.J.; McKelvy, M.J.; Chizmeshya, A.V.G.; Gormley, D.; Nunez, R.; Carpenter, R.W.; Squires, K.;
Wolf, G.H. Carbon sequestration via aqueous olivine mineral carbonation: Role of passivating layer
formation. Environ. Sci. Technol. 2006, 40, 4802–4808. [CrossRef] [PubMed]
Fagerlund, J.; Highfield, J.; Zevenhoven, R. Kinetics studies on wet and dry gas-solid carbonation of MgO
and Mg(OH)2 for CO2 sequestration. RSC Adv. 2012, 2, 10380. [CrossRef]
Garcia, B.; Beaumont, V.; Perfetti, E.; Rouchon, V.; Blanchet, D.; Oger, P.; Dromart, G.; Huc, A.Y.; Haeseler, F.
Experiments and geochemical modelling of CO2 sequestration by olivine: Potential, quantification.
Appl. Geochem. 2010, 25, 1383–1396. [CrossRef]
Pokrovsky, O.S.; Golubev, S.V.; Schott, J.; Castillo, A. Calcite, dolomite and magnesite dissolution kinetics in
aqueous solutions at acid to circumneutral pH, 25 to 150 ˝ C and 1 to 55 atm pCO2 : New constraints on CO2
sequestration in sedimentary basins. Chem. Geol. 2009, 265, 20–32. [CrossRef]
Saldi, G.D.; Jordan, G.; Schott, J.; Oelkers, E. Magnesite growth rates as a function of temperature and
saturation state. Geochim. Cosmochim. Acta 2009, 73, 5646–5657. [CrossRef]
Highfield, J.; Chen, J.; Bu, J.; Åbacka, J.; Fagerlund, J.; Zevenhoven, R. Steam-promoted gas-solid carbonation
of magnesia and brucite below 200 ˝ C. In Proceedings of the 4th International Conference on Accelerated
Carbonation for Environmental and Materials Engineering (ACEME 2013), Leuven, Belgium, 9–12 April 2013.
Chen, Z.Y.; O’Connor, W.K.; Gerdemann, S.J. Chemistry of aqueous mineral carbonation for carbon
sequestration and explanation of experimental results. Environ. Prog. 2006, 25, 161–166. [CrossRef]
Saldi, G.D.; Daval, D.; Morvan, G.; Knauss, K.G. The role of Fe and redox conditions in olivine carbonation
rates: An experimental study of the rate limiting reactions at 90 and 150 ˝ C in open and closed systems.
Geochim. Cosmochim. Acta 2013, 118, 157–183. [CrossRef]
Highfield, J.; Åbacka, J.; Chen, J.; Nduagu, E.; Zevenhoven, R. Overview of the ÅAU/ICES collaboration
in ex-situ CO2 mineralization. In Proceedings of the 13th International Conference on Carbon Dioxide
Utilization (ICCDU 2015), Singapore, Singapore, 5–9 July 2015. paper 251.
Felmy, A.R.; Qafoku, O.; Arey, B.W.; Hu, J.Z.; Hu, M.; Schaef, H.T.; Ilton, E.S.; Hess, N.J.; Pearce, C.I.;
Feng, J.; et al. Reaction of water-saturated supercritical CO2 with forsterite: Evidence for magnesite formation
at low temperatures. Geochim. Cosmochim. Acta 2012, 91, 271–282. [CrossRef]
Loring, J.S.; Thompson, C.J.; Wang, Z.; Joly, A.G.; Sklarew, D.S.; Schaef, H.T.; Ilton, E.S.; Rosso, K.M.;
Felmy, A.R. In situ infrared spectroscopic study of forsterite carbonation in wet supercritical CO2 . Environ. Sci.
Technol. 2011, 45, 6204–6210. [CrossRef] [PubMed]
Schaef, H.T.; McGrail, B.P.; Loring, J.L.; Bowden, M.E.; Arey, B.W.; Rosso, K.M. Forsterite [Mg2 SiO4 )]
carbonation in wet supercritical CO2 : An in situ high-pressure X-ray diffraction study. Environ. Sci. Technol.
2013, 47, 174–181. [CrossRef] [PubMed]
Schaef, H.T.; Windisch, C.F., Jr.; McGrail, B.P.; Martin, P.F.; Rosso, K.M. Brucite [Mg(OH2 )] carbonation in
wet supercritical CO2 : An in situ high pressure X-ray diffraction study. Geochim. Cosmochim. Acta 2011, 75,
7458–7471. [CrossRef]
Loring, J.S.; Thompson, C.J.; Zhang, C.Y.; Wang, Z.M.; Schaef, H.T.; Rosso, K.M. In situ infrared spectroscopic
study of brucite carbonation in dry to water-saturated supercritical carbon dioxide. J. Phys. Chem. A 2012,
116, 4768–4777. [CrossRef] [PubMed]
Kwak, J.H.; Hu, J.Z.; Turcu, R.V.F.; Rosso, K.M.; Ilton, E.S.; Wang, C.; Sears, J.A.; Engelhard, M.H.; Felmy, A.R.;
Hoyt, D.W. The role of H2 O in the carbonation of forsterite in supercritical CO2 . Int. J. Greenh. Gas Control
2011, 5, 1081–1092. [CrossRef]
Vitillo, J.G. Magnesium-based systems for carbon dioxide capture, storage and recycling: From leaves to
synthetic nanostructured materials. RSC Adv. 2015, 5, 36192–36239. [CrossRef]
Molecules 2016, 21, 353
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
19 of 19
Siriwardane, R.V.; Stevens, R.W., Jr. Novel regenerable magnesium hydroxide sorbents for CO2 capture at
warm gas temperatures. Ind. Eng. Chem. Res. 2009, 48, 2135–2141. [CrossRef]
Gharabaghi, M.; Irannajad, M.; Azadmehr, A.R. Leaching kinetics of nickel extraction from hazardous waste
by sulphuric acid and optimization dissolution conditions. Chem. Eng. Res. Des. 2013, 91, 325–331. [CrossRef]
Pokrovsky, O.S.; Schott, J. Forsterite surface composition in aqueous solutions: A combined potentiometric,
electrokinetic, and spectroscopic approach. Geochim. Cosmochim. Acta 2000, 64, 3299–3312. [CrossRef]
O’Connor, W.K.; Dahlin, D.C.; Rush, G.E.; Dahlin, C.L.; Collins, W.K. Carbon dioxide sequestration by direct
mineral carbonation: Process mineralogy of feed and products. Miner. Metall. Process. 2002, 19, 95–101.
Na, B.K.; Koo, K.K.; Eum, H.M.; Lee, H.; Song, H.K. CO2 recovery from flue gas by PSA process using
activated carbon. Korean J. Chem. Eng. 2001, 18, 220–227. [CrossRef]
Yong, Z.; Mata, V.; Rodriguez, A.E. Adsorption of carbon dioxide onto hydrotalcite-like compounds (HTlcs)
at high temperatures. Ind. Eng. Chem. Res. 2001, 40, 204–209. [CrossRef]
Ko, D.; Siriwardane, R.V.; Biegler, L.T. Optimization of pressure swing adsorption and fractionated vacuum
pressure swing adsorption processes for CO2 capture. Ind. Eng. Chem. Res. 2005, 44, 8084–8094. [CrossRef]
Reynolds, S.P.; Ebner, A.D.; Ritter, J.A. New pressure swing adsorption cycles for carbon dioxide
sequestration. Adsorption 2005, 11, 531–536. [CrossRef]
Xu, X.C.; Song, C.S.; Miller, B.G.; Scaroni, A.W. Adsorption separation of carbon dioxide from flue gas of
natural gas-fired boiler by a novel nanoporous “molecular basket” adsorbent. Fuel Process. Technol. 2005, 86,
1457–1472. [CrossRef]
Hicks, J.C.; Drese, J.H.; Fauth, D.J.; Gray, M.L.; Qi, G.G.; Jones, C.W. Designing adsorbents for CO2 capture
from flue gas-hyperbranched aminosilicas capable of capturing CO2 reversibly. J. Am. Chem. Soc. 2008, 130,
2902–2903. [CrossRef] [PubMed]
Hänchen, M.; Prigiobbe, V.; Storti, G.; Seward, T.M.; Mazzotti, M. Dissolution kinetics of forsteritic olivine at
90–150 ˝ C including effects of the presence of CO2 . Geochim. Cosmochim. Acta 2006, 70, 4403–4416. [CrossRef]
Hänchen, M.; Prigiobbe, V.; Baciocchi, R.; Mazzotti, M. Precipitation in the Mg-carbonate system—effects of
temperature andCO2 pressure. Chem. Eng. Sci. 2008, 63, 1012–1028. [CrossRef]
Parkhurst, D.L.; Appelo, C.A.J. User’s Guide to PHREEQC (Version 2) a Computer Program for Speciation,
Batch-Reaction, One-Dimensional Transport, and Inverse Geochemical Calculations; Water-Resources Investigations
Report 99–4259; U.S. Geological Survey: Reston, WV, USA, 1999; p. 312.
Konigsberger, E.; Konigsberger, L.C.; Ager, H.G. Low-temperature thermodynamic model for the system
Na2 CO3 –MgCO3 –CaCO3 –H2 O. Geochim. Cosmochim. Acta 1999, 63, 3105–3119. [CrossRef]
Power, I.M.; Harrison, A.L.; Dipple, G.M.; Wilson, S.A.; Kelemen, P.B.; Hitch, M.; Southam, G. Carbon
mineralization: From natural analogues to engineered systems. Rev. Miner. Geochem. 2013, 77, 305. [CrossRef]
Wollast, R.; Mackenzie, F.T.; Bricker, O.P. Experimental precipitation and genesis of sepiolite at earth-surface
conditions. Am. Miner. 1968, 53, 1645–1662.
Christ, C.L.; Hostetler, B.P.; Siebert, R.M. Studies in the system MgO–SiO2 –CO2 –H2 O: III. Am. J. Sci. 1973,
273, 65–83. [CrossRef]
Birsoy, R. Formation of sepiolite-palygorskite and related minerals from solution. Clays Clay Miner. 2002, 50,
736–745. [CrossRef]
Gadikota, G.; Matter, J.; Kelemen, P.; Park, A.H.A. Chemical and morphological changes during olivine
carbonation for CO2 storage in the presence of NaCl and NaHCO3 . Phys. Chem. Chem. Phys. 2014, 16, 4679.
[CrossRef] [PubMed]
Levenspiel, O. Chemical Reaction Engineering, 2nd ed.; John Wiley and Sons: New York, NY, USA, 1972.
Dri, M.; Sanna, A.; Maroto-Valer, M.M. Dissolution of steel slag and recycled concrete aggregate in
ammonium bisulphate for CO2 mineral carbonation. Fuel Process. Technol. 2013, 113, 114–122. [CrossRef]
Sample Availability: Not available.
© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons by Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).