Journal of Building Engineering 66 (2023) 105856

Contents lists available at ScienceDirect

Journal of Building Engineering

journal homepage: www.elsevier.com/locate/jobe

Influence of carbonation curing and nano-silica incorporation on

compressive strength and micro-structural development of binary
RMC-based systems
Khalilullah Taj a, *, Busra Akturk b, Serhan Ulukaya a
a
Yildiz Technical University, Department of Civil Engineering, Istanbul, Turkey
b
Istanbul Bilgi University, Department of Civil Engineering, Istanbul, Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: Fresh-state properties, compressive strength, and microstructural properties of reactive MgO
Reactive magnesium cement cement-based (RMC) systems cured under accelerated carbonation or ambient conditions were
Pozzolans examined in the paper. RMC, which is an environmentally greener alternative to ordinary Port­
Nano-silica land cement (OPC), was incorporated in the reference series to contrast its mechanical properties
Accelerated carbonation with other cementitious materials. To assess the synergy of RMC with other binders such as micro-
silica, slag, and fly ash, as well as to find a way to incorporate more sustainable materials in the
mortars, binary systems were designed. Lastly, to evaluate the influence of ultra-fine nano-silica
(NS), on the composite binders, tertiary systems were formulated. In the carbonated series, a
significant pH decrease, as well as a striking increase in compressive strength were observed.
Accelerated carbonation induced the formation of magnesium calcite, which is supported by
eclectic microstructural analyses. The highest compressive strength was measured in the
carbonated reference series, binary and tertiary series also demonstrated respectable strength
which was accompanied by compact microstructure. The addition of NS correlated with a
decrease in the number of pores, especially large capillary pores, whereas no perceivable change
in reaction products was discovered. In conclusion, this study demonstrated that reactive MgO-
based cement could be used as a complete or partial replacement for conventional binders in
structural concrete; the promising qualities of carbonated specimens ensure its usage as high-
strength concrete.

1. Introduction
Sustainability, which is a topic gaining more traction by the day in the sphere of construction materials, is not without a justifi­
cation. It is well known that anthropogenic CO2 engenders environmental issues such as climate change and resource depletion [1,2].
Cement production leads to greenhouse gas emissions and overconsumption of natural resources, which is unsustainable. CO2
emissions from ordinary Portland cement (OPC) production constitute approximately 8.6% of the amount of CO2 emitted by humans
annually and are assessed to be one ton per ton of clinker [3,4]. Moreover, PC produced from calcinating clay minerals requires a
temperature of around 1400 ◦ C and energy expenditure of 3500–5000 MJ per ton of clinker [5]. To resolve the problem of near
depletion-level natural cementitious materials sources and ever-increasing energy expenditure and emission of greenhouse gases, the

* Corresponding author.
E-mail address: [email protected] (K. Taj).

https://doi.org/10.1016/j.jobe.2023.105856
Received 21 October 2022; Received in revised form 25 November 2022; Accepted 6 January 2023
Available online 10 January 2023
2352-7102/© 2023 Elsevier Ltd. All rights reserved.
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

discovery and development of a novel and more sustainable material are crucial.
Reactive magnesium cement (RMC) is generally produced by calcining magnesite (MgCO3) [6]. RMC is a cementitious material
obtained at lower temperatures than OPC (700 ◦ C vs. 1400 ◦ C), which can gain strength by sequestering CO2 [7]. It is proven that the
CO2 emission rate of cement industry can be abated by the utilization of RMC. In addition to recycling via recalcination, partially
substituting reactive MgO with ground and carbonated RMC appears to be feasible [8,9]. Moreover, producing RMC using magnesium
silicate ores is could become highly practical, because it doesn’t involve a high-energy calcination process [10–12]. Additionally, there
are other types of magnesium-containing binders as sorel cement (magnesium oxychloride) that on top of attaining strength rapidly,
they possess other sought-after qualities such as high strength to weight ratio, fire resistance, low thermal conductivity, and abrasion
resistance [13]. However, the most prevalent drawback of using sorel cement is its low water resistance [14]. Sorel cement, which is a
mixture of fine ground magnesia and a solution of magnesium chloride, is a two-part binder. It mainly hardens via the hydration of
amorphous phase. Formation of water-soluble hydration products is the leading cause of low water resistance [15]. Generation of
insoluble carbonates as the reaction products via the incorporation of additives such as phosphates, borax, resins, and formaldehyde
has been suggested [16]. Guo et al. investigated the effect of silica fume and fly ash on the water resistance of magnesium oxychloride,
wherein they discovered that 15% silica fume and 15% fly ash addition led to impressive water resistance in the specimens. The
specimens retained 100% and 95% of their ultimate strength after 28 days and 56 days of submersion in water. The increased water
resistance was attributed to the filling effect and increased impermeability provoked by the pozzolans [17]. Reaction products of MgO
cement on top of displaying a high water resistance are known to be cured submerged in water [18]. The impact of high temperatures
(up to 400 ◦ C) on the stability of hydration products reactive MgO-based systems has been studied [19]. It is reported that heating
under 200 ◦ C did not degrade the hydration products, on the contrary, it increased the compressive strength. Dehydroxylation of
brucite (main hydration product) occurred around in 300–400 ◦ C range, which resulted in degradation and strength loss.
In pastes, mortars, and concretes, where RMC is used as the binder, hardening occurs by both hydration and carbonation reactions.
Henceforth, MgO is used in synonymity with reactive MgO. The hydration in MgO-based materials occurs according to the aqueous
reaction detailed in Eq. (1), which illustrates the dissolution of MgO followed by precipitation of magnesium hydroxide (Mg(OH)2),
also known as brucite. Since brucite expands drastically in aqueous media, it does not allow the material to gain high compressive
strength [20]. Xing et al. [21] presented the dissolution (Eqs. (2)–(4)) and precipitation (Eqs. (5)–(8)) steps of hydration reaction of
MgO at temperatures below 90 ◦ C as the following:
MgO(s) + H2 O(l) 2 = Mg(OH)2 (s) (1)

MgO(s) + H2 O(l) → Mg(OH)+ −

surface + OH(aq) (2)

Mg(OH)+ − + −
surface + OH(aq) → Mg(OH) • OHsurface (3)

Mg(OH)+ • OHsurface
−
→ Mg2+ −
(aq) + 2OH(aq) (4)

Mg2+ − 4−
(aq) + 6OH(aq) → Mg(OH)6 (aq) (5)

( )
Mg(OH)4−6 (aq) + Mg+ 2−
2 (aq) → Mg(OH)2 2 (OH)2 (aq) (6)
( ) ( )
Mg(OH)2 2 (OH)2−2 (aq) + Mg+
2 (aq) → Mg(OH)2 3(s)
(7)

( ) ( )
Mg(OH)2 3n− 1
(OH)2−2 (aq) + Mg+
2 (aq) → Mg(OH)2 3n (s) (8)

Mg2+ necessary for the Mg(OH)2 formation is released from the dissolution of MgO. When the concentration of Mg2+ and OH−
reaches supersaturation, precipitation of Mg(OH)2 begins on the surface of MgO grains. The formation of magnesium hydroxide creates
a boundary that impedes further hydration reactions by isolating the particles’ nuclei from water [22]. According to the boundary
nucleation and growth (BNG) model proposed by Thomas et al. [23], precipitation occurs at the boundary surface of MgO grain, which
explains the decrease in the reaction rate with the progress over time. Aphane et al. [24] concluded that with the progress of time,
solids formed at the boundary of MgO grain disintegrate due to the stress caused by it, instead of perennial growth of the Mg(OH)2
layer. This breakage leads to the exposition of fresh surfaces for the subsequent hydration and breakage cycles. The rate and degree of
this reaction and the type of phases formed at the end of carbonation determine the range of strength gain. Some studies have sug­
gested that the hydration of RMC is limited to approximately 80% under ambient conditions [25]. Moreover, hydration kinetics of
MgO is also temperature-dependent, relatively higher temperatures influence both dissolution and precipitation processes [26].
Further, another component that increases the dissolution and precipitation rate is the hydration agent (HA) [27–30]. There are
various types of HAs reported in the literature, magnesium acetate (MA) and sodium hexametaphosphate (SHMP) being at the
forefront. MA helps the hydration mechanism by providing more magnesium cations for the precipitation of brucite as well as acetate
anions for the dissolution of MgO [27]. Besides, SHMP operates as both an HA and a disperser by super-saturating the MgO ions and
momentary inhibition of brucite formation and having a deflocculant effect which allows for a lower water to binder ratio [29,30].
RMC hardens chiefly by arresting CO2 present in the ambient conditioned room or accelerated carbonation chamber. RMC gains its
strength mostly by turning the brucite formed in the hydration reactions into hydrated magnesium carbonates (HMCs) [31]. The

2
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

carbonation reactions which occur for the synthesis of the main HMCs i.e., nesquehonite (MgCO3⋅3H2O), artinite
(Mg5(CO3)4(OH)2⋅3H2O), hydromagnesite (Mg5(CO3)4(OH)2⋅4H2O), dypingite (Mg5(CO3)4(OH)2⋅5H2O), are presented by Eqs. (9)–
(12), respectively [9,32,33]. However, in presence of CaO-containing cementitious materials, magnesium calcite is reported to
crystallize [34–36]. The CO2 concentration in the atmosphere is approximately 0.04% [37], which is orders of magnitude lower than
the concentration required to adequately carbonate the priorly formed brucite. Thus, the utilization of accelerated carbonation
chambers is common to ensure substantial strength gain [38]. According to Dung et al. [39], curing under 10% concentration of CO2
was the optimum condition for strength gain (>60 MPa) of MgO-based specimens if cured for 28 days, while 20% CO2 was beneficial
for rapid strength gain and 5% CO2 achieved similar strengths under longer periods of curing.
Mg(OH)2 + CO2 + 2H2 O → MgCO3 • 3H2 O (9)

2Mg(OH)2 + CO2 + 2H2 O → Mg2 CO3 (OH)2 • 3H2 O (10)

5Mg(OH)2 + 4CO2 → Mg5 (CO3 )4 (OH)2 • 4H2 O (11)

5Mg(OH)2 + 4CO2 + H2 O → Mg5 (CO3 )4 (OH)2 • 5H2 O (12)

The concentration of CO2 dissolved is of paramount interest due to the ionic CO2 being a prerequisite of carbonation reactions [22].
Dissolution of CO2 occurs by its reaction with H2O and OH− , steps of which are delineated through Eqs. (13)–(16). Subsequently,
previously formed brucite reacts with the CO2 to yield HMCs.
CO2(g) + H2 O(l) ⇋CO2 (aq) (13)

CO2(g) + OH − ⇋HCO−3 (14)

CO2(g) + H2 O(l) ⇋H + + HCO−3 (15)

HCO−3 + H + ⇋CO2−3 + 2H + (16)

Although a cursory evaluation of MgO carbonation can lead one to assume it to be carbon neutral, even carbon negative, at the end
of careful consideration, Shen et al. concluded that magnesia cement is not carbon-neutral, at least in the short or intermediate-term,
due to the limited natural carbonation of aforementioned cement [40]. Significant factors that may affect the carbonation efficacy of
magnesia such as curing conditions, materials, mix properties, and compositions are being investigated by researchers, which might
shift the consensus on its contribution to CO2 emission in the future [31,34,41–43].
As reported by Yu et al. [44], the addition of MgO into the slag-based grouts, besides increasing the w/b, decreased the initial and
final setting time by 82.2% and 60.6%, respectively. According to He et al. [45], reactive MgO, when used as an activator for slag,
lowered the flowability as well as setting time. The compressive strength among other properties of MgO-based composites has been
studied [46], where the substitution of MgO for pozzolans such as fly ash and slag always led to a decrease in the early age strength and
often in ultimate compressive strength. This reduction was attributed to the lower carbonation rates due to their reduced porosities.
Unluer et al. [47] investigated the effects of various additives on the MgO-based pastes, where the effect of slag was summarized as
repressive on the compressive strength. Additionally, the water demand of composites incorporating slag was the lowest among the
mixes.
The hydration of a combination of reactive MgO and silica as a binder has been studied and the pertinent strength-providing phase
has been conjectured in several studies [48–51]; magnesium silicate hydrate (M-S-H) gel is the asserted hydration phase. In the “MgO
+ Silica” pastes, M-S-H forms concurrently with Mg(OH)2, which can also react with silica to form even more M-S-H [52], and behaves
as the main binding phase [29]. In contrast to ordinary cementitious binding materials, a magnesium-containing silicate layer form
strikingly quickly when either RMC or brucite is mixed with amorphous silica (micro silica, silica fume, etc.) and water [51,53], though
it was revealed that M-S-H generation rate is noticeably higher in the systems where MgO is the precursor instead of brucite [51].
Several studies [29,51,54] have revealed that the composition of M-S-H was dependent on the ratio of Mg/Si of the reagents, in which
most researchers produced M-S-H gel mostly resembling talc (Mg3(Si2O5)2(OH)2) and serpentine (Mg3(Si2O5)(OH)4). According to
Zhang et al. [29], the formula of M-S-H gel of a mixture with a 40% MgO/60% SF and a w/b of 0.8 was calculated as
Mg8Si8O20(OH)8⋅12H2O.
Based on the unsatisfactory results of the flowability of trial productions, the necessity of a dispersing agent that is uniquely well-
fitted for the MgO-based systems was understood. SHMP, which is an additive that is reported to improve the workability of MgO-
containing mixes [30], was chosen as the dispersant. It is reported that SHMP enhanced the fresh-state properties of all the
MgO-containing blends; it suits MgO-based systems best because it accelerates the hydrolysis of MgO and retards the brucite formation
[30,52,55].
Although the impact of SF on MgO-based mortars was examined, the effects of other sources of silica, such as nano-silica (NS) are
scarce in the scientific literature. This study investigated the influence of pozzolanic materials, as well as NS, on the fresh state,
mechanical and microstructural properties of MgO-based systems exposed to different curing conditions. X-ray diffraction (XRD),
Fourier transform infrared spectroscopy (FTIR), thermogravimetric/derived thermogravimetric (TG/DTG) analyses, and scanning
electron microscopy (SEM) were utilized to characterize the phases formed as a result of hydration and carbonation reactions.
Expressly, to comprehend the full effect of NS on the median pore diameter and pore-size distribution of the specimens, mercury

3
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

intrusion porosimetry (MIP) test was conducted. The effects of two types of curing, air-curing and accelerated carbonation curing, were
contrasted using said analyses.

2. Materials and methods

2.1. Materials
The RMC which is the main ingredient of the binder was obtained by calcining of magnesite at 600–700 ◦ C from KÜMAŞ Magnesite
Inc. A modest portion of this cement consisted of lime (CaO) and silica (SiO2), which may induce the formation of different crystals
than a purer RMC. The slag, a certain type of pozzolan that contains high amounts of SiO2, alumina (Al2O3), and CaO, was supplied
from KARDEMIR Iron and Steel Plant. The type F fly ash was supplied by AKCANSA, the composition of which is in line with ASTM
C618 [56]. The micro-silica, which was used as a secondary precursor material to discern the effect of SiO2 on the hydration and
carbonation products of MgO-based pastes, was procured from ELKEM. Additionally, nano-silica was added to the blends in order to
evaluate the effect of an ultra-fine silica-rich powder. SHMP was used as a disperser that also functioned as the hydration agent in the
mixes. The chemical composition and physical properties of the RMC, slag, fly ash, micro-silica, and nano-silica utilized in this study
were determined by X-ray fluorescence (XRF) and are presented in Table 1.
The particle size distribution of the materials, which is illustrated in Fig. 1, was determined using the mastersizer analysis. The SEM
images of raw materials is given in Fig. 2. The silica-based sand employed in this study was supplied by CELIKTAS AS. According to the
test results, the water absorption ratio was determined as 0.6%, however, the moisture condition of the sand during the experimental
study was SSD. It was manually size-graded via sieve analysis, where the maximum grain size was determined to be less than 2 mm.
The gradation of fine aggregate was comparable to CEN standard sand and for its estimation; BS EN 196-1 [57] was followed.
The phase identification was assessed via XRD analysis with a scanning angle of 5–65 2θ degrees (Fig. 3). For RMC, the sharp peaks
at approximately 42.5 and 62 ◦ 2θ are representative of the reactive mineral, periclase. The slag showed a rather broad peak between 25
and 35 ◦ 2θ, which suggests the existence of minerals similar to gehlenite, fukalite, and melilite. Whereas for the fly ash, the major peaks
at 20.5 and 26.3 ◦ 2θ are assigned to quartz and the minor peaks at 16 and 30–35 ◦ 2θ are ascribed to mullite (an aluminosilicate glassy
phase, Al4.44Si1.56O9.78). The only peak seen at 23.5 ◦ 2θ in the XRD spectrum of micro-silica (MS), which is a rather broad one, affirms
its purity and amorphous structure. XRD pattern of NS reveals the existence of the same peak as MS, but with a higher intensity.

Table 1
Chemical composition and physical properties of raw materials.

Compounds (%, by mass) RMC Slag Fly Ash Micro-Silica Nano-Silica

SiO2 6.0 39.0 40.2 92.8 99.0

Al2O3 0.7 12.4 8.0 – –
Fe2O3 0.3 1.1 29.2 – –
CaO 7.9 33.7 16.1 – 0.1
MgO 83.6 7.5 0.2 – –
SO3 0.2 0.8 0.7 0.6 –
Na2O 0.4 0.4 – 1.1 –
K2 O – 0.6 – 0.4 –
TiO2 – 0.9 3.1 – –
Mn2O3 – 3.1 – – –
Cl− – – – – –
LOIa (%) 0.8 0.4 2.4 5.0 0.7
Specific density (g/cm3) 3.6 2.9 2.3 2.3 2.2
Specific surface area (cm2/g) 18700 4670 3690 203000 2020000
a
Loss on ignition.

Fig. 1. Particle size distribution of RMC, slag, fly ash, micro-silica, and sand.

4
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

Fig. 2. SEM images of RMC (a), slag (b), fly ash (c), micro silica (d), and nano-silica (e).

5
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

Fig. 3. XRD pattern of RMC, slag, fly ash, micro-silica, and nano-silica (P: periclase, Ge: gehlenite, Q: quartz, Mu: mullite).

2.2. Mix proportions and sample preparation

The binary and ternary binder systems were formulated, and the prescription in Table 2 was decided for mortars. A total of 7
different blends of mortars and pastes were produced. Distinct samples from each blend were cured at either constant ambient
temperature and relative humidity conditions or accelerated carbonation curing, totaling 14 separate series. The presence or absence
of the subscripted letter “c” at the end of the code of each series indicates the curing type; the series with the subscripted letter were
subjected to carbonation curing, while the rest were subjected to ambient curing conditions. The carbonation curing condition is
defined as a constant CO2 concentration of 10%, the temperature of 30 ◦ C, and an RH of 55 ± 5%, whereas the ambient curing
condition has identical characteristics with exception of CO2 concentration; the samples cured under ambient conditions were sub­
jected to the concentration of CO2 that exists in the atmosphere (i.e., 0.03–0.04%). Hereafter, the terms ambient curing, atmospheric
curing, and non-carbonated curing will be used interchangeably. The reference mix (R, Rc) contained RMC as the only binder, whereas
the rest contained only RMC at 70% of the binder by mass. 30% of the binder was comprised of slag, fly ash, or MS in the binary binder
systems (S, Sc, FA, FAc, MS, and MSc), whereas in the ternary binder systems (SN, SNc, FAN, FANc, MSN, MSNc) NS constituted 4% of
binder in all of them, while 26% of the binder was comprised of slag, fly ash, and MS. The water/binder ratio was kept constant at 0.7
for the mortars and 0.6 for the pastes. All the mortar series contained sand at 50% volume of the mixture. To facilitate the dispersion of
MgO, sodium hexametaphosphate (SHMP) was utilized at 4% of the total binder in all of the mixes.
To prepare mortar samples, SHMP was dissolved in the water using a Protech ultrasonic dispersing device for 10 min at a tem­
perature of 30 ◦ C. The same procedure was followed for the mixes that contained NS. Immediately after preparation of dispersant and
NS, previously weighed precursors were dry-mixed for 15 s. Next, the ready SHMP (and NS in its respective blends) solution was/were
poured into the mixer. The paste was mixed for 30 s at low speed (140 ± 5 rpm), then fine aggregate was added to the mixture at the 30
s mark and the blend was mixed for an additional 30 s at the same speed. Following a 15-s removal of mortar stuck to the walls of the
mixer by a trowel, it was mixed at the high speed (285 ± 5 rpm) for a final 60 s. The same method was adhered to in the production of
pastes, apart from the inclusion of fine aggregate. Then, mortar and paste mixtures were poured into the triplicate cubic molds of size

Table 2
Mix proportions of MgO-based mortars.

Series Precursor (% of total binder mass) Curing Condition

RMC Slag Fly ash MS NS

R 100 – – – – Ambient
Rc Carbonation
S 70 30 – – – Ambient
Sc Carbonation
MS 70 – – 30 – Ambient
MSc Carbonation
FA 70 – 30 – – Ambient
FAc Carbonation
SN 70 26 – – 4 Ambient
SNc Carbonation
MSN 70 – – 26 4 Ambient
MSNc Carbonation
FAN 70 – 26 – 4 Ambient
FANc Carbonation

6
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

50 × 50× 50 mm and plastic Falcon tubes with a volume of 50 ml, respectively. Finally, mortar molds were jolted on the vibration table
and the Falcon tubes were vibrated manually to remove the trapped air inside the mixtures.
The stretch film-wrapped molds filled with mortars were immediately placed in the moisture cabinet at a relative humidity (RH) of
55 ± 5% and a temperature of 30 ◦ C for 24 h. After 24 h, all of the specimens were demolded, and finally, mortar and paste samples of
each series were placed either in the carbonation chamber or in the moisture cabinet under the same temperature and RH.

2.3. Methodology
2.3.1. Flow diameter
The experiments on the flowability of mortar were performed using a flow table, which is according to ASTM C1437 [58]. Freshly
prepared mortars were placed in the standard cone with an internal base radius of 50 mm in two lumps, each filling approximately half
the cone. Each layer was tamped 20 times with the tamper. The test was performed with a standard flow table and is subjected to 25
blows (up and down movement of the table). After 25 drops the spread in diameter of mortar was noted down.

2.3.2. Setting time

The initial and final setting times of the specimens were measured according to the ASTM C191 [59] with the help of an automatic
Vicat apparatus. The paste samples were poured into the conical ring and the penetration depth of the Vicat needle was measured every
10 min. The initial setting time was considered as the elapsed time between the moment at which the solution and binder made contact
up until the moment at which the needle penetrated no more than 25 mm. The final setting time was considered as the duration
between the first contact and the moment at which the needle could not penetrate anymore.

2.3.3. pH evolution
To conduct the pH evolution test, 10 g of powder from each paste specimen was extracted using a metal mortar and pestle. The
powder-to-water ratio was adapted as 1:5. Thus, 50 g of distilled water was weighed and added to the powder. The solution was filtered
to attain only the liquid parts of the solution, not before mixing it for 10 min at room temperature by a magnetic mixer at a speed of
1000 rpm. Finally, pH measurements were taken immediately using a glass electrode and Hanna pH meter with an accuracy of 0.01. To
better grasp the influence of accelerated carbonation and selected secondary precursors on the pH of MgO-based mortars, described pH
test was performed on the 3rd, 7th, and 28th days.

2.3.4. Compressive strength

The compressive strength of mortar samples was measured in agreement with the ASTM C109 [60]. The compressive strength test
was conducted by an ALSA Hydraulic Test device on mortar samples on the 3rd, 7th, and 28th days. The mean value of triplicate
samples was considered as the compressive strength for each age.

2.3.5. Microstructural analysis

The XRD analysis was carried out using a Cu tube (45 kV, 40 mA) XPERT-PRO diffractometer on the 3rd and 28th days on the
powder extracted from the paste samples. The patterns were recorded in the range of 5 – 65 (◦ 2θ) with a step size of 0.02◦ (2θ) and a
measuring time of 1 s/step.
The Fourier transform infrared spectroscopy (FTIR) analysis was conducted by a T80+ UV/VIS spectrophotometer device. All spectra
were recorded with 32 scans from 4000 to 400 cm− 1 per spectrum. The bands between 4000 and 650 cm− 1 were regarded for analyzing
the results since the carbonation and hydration products are determined at this range. The analysis was only conducted after 28 days of
curing.
Thermogravimetric and derived thermogravimetric (TG, DTG) analyses were conducted using an EXSTAR 6000 thermal analyzer.
The range of temperature was 30–900 ◦ C with a heating rate of 15 ◦ C/min. The mass loss at distinct temperatures yielded the quantity
and quality of produced reaction products.
The scanning electron microscopy (SEM) analysis was carried out on bulk specimens, the size of approximately 1 × 1 × 0.3 cm3,
separated from the paste specimens after 28 days of curing. The samples were coated with gold before being placed in the vacuum
chamber. Subsequently, they were installed on the brass rods using a carbon tape. For the analysis, a ZEISS EVO LS10 device was
employed at an accelerating voltage of 15 kV.
Secondary electron (SE) images were captured to determine the hydration and carbonation products. Elemental ratios were
determined using energy dispersive spectroscopy (EDS) measurement on a few diverse spots, each representing a distinct formation.

2.3.6. Pore structure

Pore structure, particularly pore volume, of the MgO-based specimens was examined by using Mercury intrusion porosimetry (MIP)
on the 28th day of curing. Quanachrome-Poremaster 60 was employed to assess the porosity, which has the capability of injecting
pressure up to 55000 psi and filling cylindrical pores with diameters 0.004 μm up to 200 μm.

3. Results and discussion

Under the following headings, fresh-state properties, compressive strength, microstructural development, and porosity of the
specimens are reported and discussed.

3.1. Flow diameter

The reference mix had the lowest flow diameter of 110 mm (Fig. 4), the reason being its high fineness and its tendency to flocculate

7
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

Fig. 4. Flowability of mortars.

because of the high molecular angularity of MgO, which was in accordance with the literature [61–63]. The high water demand of
MgO-based materials is not uncommon in the literature [64], because the average particle size of MgO is as fine as 40 nm, abnormally
high water-to-binder ratios have to be set in order to achieve a workable mortar. It is worth noting that finer materials are resourceful
in terms of lowering permeability and risk of bleeding [65]. Additionally, the lower molecular weight of MgO (40 g/mol), compared to
CaO (56 g/mol), translates to a greater number of molecules per gram in the former oxide compared to the latter, thus requiring a
greater quantity of water molecules to dissolve. It is hypothesized that the rapid dissolution of MgO and precipitation of hydration
products also influence flowability.
The FA blend possessed the highest flow diameter (225 mm), which is in agreement with the particle size of fly ash compared to the
rest of the precursors. Mixes with a dose of fly ash demonstrated comparatively higher flow values due to their spherical shape and
intrinsic low fineness compared to RMC [66]. The S mix, which contained slag as the secondary precursor, demonstrated a flow
diameter of 195 mm, which is in line with its particle size. The studies that were done by Hwang et al. elucidate that the specimens with
binary binders containing both alkali-activated slag (AAS) and MgO had lower flowability than the samples containing only AAS [64,
67]. Similarly, in the present study, the flowability of MgO-based mortars was enhanced by the addition of slag. Fly ash and slag, both,
increase the flowability of the mortar mix. If curing by accelerated carbonation is aimed, RMC + fly ash and RMC + slag yield workable
mixes by sacrificing compressive strength.
Akin to the reference sample, mixes with silica, MS, or NS, exhibited lower flowability due to their extremely high fineness (as
discussed in Section 2.1). Furthermore, the existence of NS significantly affected the flowability of the mortars containing pozzolans
(with a 4% NS addition, the flowability of mixes S and FA dropped from 195 mm to 115 mm and 225 mm–130 mm, respectively). In
contrast, NS addition did not affect the flowability of the mix incorporating MS considerably, due to the concurrent presence of NS and
MS, both of which are sources of reactive silica and are very fine.

3.2. Setting time

The initial and final setting times of the mixes are demonstrated in Fig. 5. Reference mix (R) had the shortest initial setting time of
60 min, whereas the longest initial and final setting times (180 and 260 min) were observed in the FA blend. The assertion of
comparatively short initial and final setting time of MgO-based mortars is in accordance with the literature [45,67]. The utilization of
reactive MgO as the activator in alkali-activated materials yielded shorter setting times compared to the materials activated by other
activating agents [44], which puts forth the high reactivity of RMC. The addition of NS to the paste specimens seems to considerably
expedite the hydration throughout the mixes. The dominant reasoning for this effect is that NS can function as nuclei that promote
cement hydration and thus shortening the idle period of hydration. Therefore, it decreases the initial and final setting time. Another

Fig. 5. Initial and final setting time of mixes.

8
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

cause of the acceleration in the setting is the reduction in flowability, which directly results in the limited setting time [68,69]. The
decent improvement in the compressive strength of NS-containing specimens, which is discussed in section 3.4, is also owed to the
functioning of nanoparticles as nuclei. The addition of MS as a precursor increased the initial and final setting time. M-S-H, whose
formation is sustained by the Mg(OH)2 in the later stage of hydration, might also inhibit the immediate setting of the micro
silica-containing sample. Therefore, precipitation of M-S-H takes longer than the formation of brucite, all while SHMP retards the
hydration of all the mixes by delaying the precipitation of brucite. The inclusion of pozzolanic materials appears to prolong the setting
of MgO-based paste. According to He et al. [45], the concurrent presence of RMC and pozzolans displayed a smaller gap between initial
and final setting time, compared to the exclusive usage of any of them. The retardation effect of slag in the paste, which manifested
itself in the setting time test, corresponds to its high flowability effect. Therefore, it may thwart the hydration reactions of any possible
binding phase in the beginning. All of the samples containing pozzolans had greater setting times which conforms to their high flow
diameter, and it supports the lack of strong hydration products elicited by their lower compressive strength. The setting time followed
the pattern of flowability in the case of fly ash and slag addition, which makes the mixes better suited where their transportation is
needed.

3.3. pH evolution
The pH values of non-carbonated and carbonated specimens, which were measured on the 3rd, 7th, and 28th day, are presented in
Fig. 6. Similar to OPC-based materials, carbonation curing in MgO-based samples leads to lower pH values compared to their ambient-
cured counterparts [70]. However, as reported by some researchers, this reduction was minor compared to other cementitious ma­
terials such as OPC, due to the lower solubility of brucite [71]. Nevertheless, the intense variation in the pH value of ambient-cured and
carbonation-cured specimens observed in this study points to the high rate of generation of binding phases in the latter type of curing.
The impact of accelerated carbonation is conspicuous in the specimens that were subjected to such a regimen of curing.
The pH values of the MS and MSN series were lower than the rest of the non-carbonated specimens. The reduction in pH of silica-
containing samples is reported to stem from the dissolution of silica and subsequent formation of M-S-H gel [72]. Additionally, the
decline in the pH is an indication of the extent of hydration, which conforms to the high initial compressive strength gain of the MS,
MSN, MSc, and MSNc series (Fig. 7). The least pH difference was between the carbonated and non-carbonated variants of MS and MSN
blends. It was postulated that due to formation of M-S-H gel, which does not involve the binding of CO2, limited the pH drop of
specimens when subjected to accelerated carbonation. The lackluster change in pH value relative to the duration of curing of
MS-containing series is also worth elaborating on. The 3rd day and 7th day pH of the MSNc sample was measured to be around a mere
10.60 and 10.55, respectively. The dormancy in the pH also sheds light on the cause of the stagnant compressive strength evolution of
such series. The effect of NS was minimal throughout the mixes. It is perhaps because of the subpar amorphic nature of NS utilized in

Fig. 6. pH value of (a) non-carbonated (b) carbonated series on 3rd, 7th, and 28th day.

9
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

this study, which permitted the additive to act only as a filler material instead of taking an active role in the hydration reaction.
Similarly, almost in all the cases, pH decreased, even though, moderately as the curing age increased. Thus, it substantiates the
claim that free OH− ions gradually precipitate by forming brucite and subsequently M-S-H, hydrotalcite, magnesium calcite, dypingite,
hydromagnesite, and lansfordite in some cases. The greatest difference between the pH value of carbonated and non-carbonated
samples occurred after 28 days in the mix containing slag (ΔpH = 1.98). A close second is the mix containing fly ash, whose pH
value was 12.52 after 28 days of ambient curing, while a 28-day accelerated carbonation curing resulted in a pH value of 10.63. These
sharp drops in the pH are consistent with the huge disparity between the compressive strength (Fig. 7) of carbonated and non-
carbonated pozzolanic specimens. There is a marked contrast between the pH of non-carbonated reference and pozzolan-
containing blends, which can be attributed to the higher concentration of OH− released by the concurrent dissolution of MgO and
CaO. According to Zhang et al. [73], only a 0.078 wt% substitution of MgO with CaO would increase the pH from 10.5 to 11.3 at a
solution density of 100 g/L.

3.4. Compressive strength

Fig. 7 illustrates the compressive strength of mixes after 3, 7, and 28 days of curing. Series R achieved a 28th compressive strength
of 9.2 MPa. The substitution of MgO with pozzolans had a diminishing effect on the compressive strength, which was consistent in the
whole duration of curing. The series S and FA gained compressive strength of 5.9 and 4.2 MPa after 28 days of ambient curing,
respectively. In the case of the series MS, the substitution of RMC with silica was beneficial in terms of compressive strength. A 30%
substitution led to a 28th day compressive strength of 21.1 MPa in the series MS. A gradual increase was observed in all the non-
carbonated mixes, apart from MS and MSN. The evolution of compressive strength of the MS series was as such; it displayed an
impressive initial compressive strength on the 3rd and 7th days (i.e., 20.5 and 24.3 MPa), however, the strength diminished after the
7th day (21.1 MPa on 28th day). The reduction is possibly due to the shrinkage cracks caused by the loss of interlayer water of M-S-H
gel, which is offered as a feasible explanation for this decline by previous studies [30,74,75]. As for the carbonated series, the
carbonated reference series (Rc) gained a compressive strength of 52.1 MPa, which is the highest among the series. Similar to
non-carbonated variants, the addition of slag and fly ash curtailed the compressive strength. Sc and FAc samples attained 34.7 and 31.9
MPa of compressive strength after 28 days of accelerated carbonation curing. Not unlike its non-carbonated variant, MSc achieved
lower 7th day strength (37.1 MPa), after which the strength decreased until the 28th day.
First and foremost, the difference between the compressive strength of carbonated and uncarbonated samples is substantial in
almost all the mixes. The difference between the 28th day strength of series R and Rc is 467%. The specimens containing pozzolans (S,
Sc, SN, SNc, FA, FAc, FAN, and FANc) had reduced compressive strength compared to the reference series. However, the percentage

Fig. 7. Compressive strength of (a) NS-free series (b) NS-containing series on 3rd, 7th, and 28th day.

10
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

difference between the compressive strength of non-carbonated and carbonated variants of those mixes was significant. The series FAN
reached a compressive strength of 11.0 MPa on the 28th day, while series FANc reached 40.7 MPa, marking a 270% increase. Similarly,
the compressive strength of the series Sc increased up to 34.7 MPa on the 28th day, whereas the series S achieved 5.9 MPa. Accelerated
carbonation of OPC-RMC mortars displayed significant improvements in mechanical properties [38] while curing under natural
carbonation had a diluting effect on the mechanical strength and impermeability of mortars [72]. In the study by Taj et al. [76], the
compressive strength of an MgO-based system exposed to ambient curing conditions was investigated. The 28th day compressive
strength of the reference mix is reported to be ~12 MPa. According to Dung et al. [39], curing under 10% concentration of CO2 was the
optimum condition for strength gain of MgO-based (>60 MPa) specimens if cured for 28 days. Surprisingly, the discrepancy between
the compressive strength of the series MS, MSN, MSc, and MSNc was not as substantial as the others, notwithstanding accelerated
carbonation improved the strength of both series. For instance, the series MSN’s 28th day strength was 26.4 MPa, while the series
MSNc’s was 30.8 MPa.
Since the highest compressive strength was observed in the Rc series; It is postulated that the formation of HMCs could have been
the most significant contributor to its strength. The compressive strength of conventionally cured mortars, especially pozzolanic ones,
was less than adequate, which could have originated from the larger amounts of unreacted particles in the binder. Although in the air-
cured samples, HMC phases forming as a result of the CO2 present in the atmosphere are not completely scarce, which is supported by
the microstructural investigation; they are sorely lacking in terms of quantity in comparison to accelerated carbonation cured samples.
The compressive strength of reference samples, even those of atmospheric curing, exceeds the requirement for masonry blocks (7 MPa)
[77]. This allows the utilization of RMC as the primary binder in the production of masonry units. The lower strength values of the
series containing fly ash and slag compared to the reference series can be associated with the lessened carbonation rates within the
specimens due to their diminished MgO contents and reduced porosities, which may hinder the diffusion of CO2 through the samples

Fig. 8. 3rd day (a,c) and 28th day (b,d) XRD patterns of NS-free carbonated (c,d) and non-carbonated (a,b) series
(A: aragonite, B: brucite, Do: dolomite, Ht: hydrotalcite, M: mullite, MC: magnesium calcite, M*: M-S-H, P: periclase, Q: quartz).

11
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

[46]. In the non-carbonated pozzolanic samples, the mediocre compressive strengths may have originated from the unreactive nature
of slag and fly ash in the vicinity of RMC, which prevented the formation of continuous and dense hydration phases. In a study by
Kumar et al. [78], the compressive strength of RMC systems with pozzolans such as slag and fly ash decreased as the amount of these
materials increased. The reduction was attributed to the lower RMC content in those samples. In the carbonated pozzolanic samples,
accelerated carbonation increased the compressive strength of fly ash and slag-containing blends by more than 200% in most cases,
which is a testament to the generation of strength-providing reaction products. On the 28th day, the compressive strength of series Sc
was 588% of that of series S. The M-S-H, magnesium calcite, and hydrotalcite were observed in the XRD (Fig. 8), the formation of
which, in these types of mediums, which is also stated by others [72,79], forms the basis of compressive strength. Nevertheless, in
comparison to the reference mix, pozzolanic mixes displayed lower compressive strength values. The strength gain of carbonated and
non-carbonated MS-containing series, in general, might be caused by only hydration of precursors to form brucite and M-S-H gel. The
addition of micro-silica (MS) resulted in early-age high strength, even without carbonation. This was in complete contrast with the
reference mix. Since the formation is M-S-H gel doesn’t require CO2 or carbonation curing, the difference between the compressive
strength of MSN and MSNc is minor compared to the rest of the series. The difference can be explained by the formation of magnesium
calcite mineral, which is supported by XRD (Fig. 9).
If a comparison must be made between binary (without NS) and ternary (with NS) mixes, the effect of NS can be summarized as a
modest enhancement in the compressive strength of both carbonated and non-carbonated pozzolanic samples. The inclusion of NS
shifted the 28th day compressive strength of S, Sc, FA, and FAc, from 5.9, 34.7, 4.15, and 31.9 MPa to 9.6, 32.8, 11, and 40.4 MPa
(belonging to the series SN, SNc, FAN, and FANc), respectively. Due to the trivial change in the compressive strength, the role of NS in
the binder might only be limited to the filler effect.

Fig. 9. 3rd day (a,c) and 28th day (b,d) XRD patterns of NS-containing carbonated (c,d) and non-carbonated (a,b) series
(A: aragonite, B: brucite, Do: dolomite, Ht: hydrotalcite, M: mullite, MC: magnesium calcite, M*: M-S-H, P: periclase, Q: quartz).

12
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

3.5. Microstructural evolution

3.5.1. XRD
Figs. 8 and 9 present the XRD patterns of NS-free and NS-containing samples on days 3 and 28. In order to fully comprehend the
effect of accelerated carbonation and NS on strength development, the XRD spectrum of the reference series was compared with
carbonated and non-carbonated binary and ternary specimens separately.
For all the non-carbonated samples, regardless of age, major peaks of brucite (Mg(OH)2) and unreacted periclase (MgO) were
observed at around 18.5◦ , 38◦ , 51◦ , 58.5◦ ; and 43◦ , 62.5◦ 2θ, respectively [42,80]. The presence of these peaks indicates that the
hydration of MgO-based specimens is taking place, but not to the fullest extent. The minor peak of brucite and magnesium calcite
overlap at 29◦ 2θ, suggesting the assignment of this peak to brucite in non-carbonated and magnesium calcite in the carbonated
samples. In the series S and FA, apart from sharing the similar major peaks with the reference mix, minor peaks at 11.5◦ , 22.5◦ , 35◦ , and
60◦ 2θ corresponding to hydrotalcite were observed. In the studies discussing the durability properties, the suggestion of the formation
of hydrotalcite improving the density of the specimen was made [81,82]. However, in this study, improvement in mechanical
properties due to the formation of hydrotalcite was not noticed. Traces of unreacted mullite (20.6◦ ), which is a prominent component
of slag and more so of FA, is observed in the pozzolanic series. Peaks related to quartz (26.6◦ ), a component of fly ash, also appeared,
hinting at the existence of unreacted particles, which explains the corresponding lower strength of such series. The XRD spectra of MS
and MSN series share a broad peak between 8 and 12◦ , alongside the peak at ~36◦ 2θ, which are also identified, but to a lesser degree in
the pozzolan-containing samples. Both of these peaks correspond to M-S-H gel peaks outlined by Refs. [72,74,83,84] as the shoulder
between 10 and 13◦ , 25–30◦ , 35–39◦ , and 58–62◦ 2θ.
As for XRD spectra of the carbonated series, although they share similar major peaks with non-carbonated series, they demon­
strated signs of carbonated phases more visibly. Peaks related to nesquehonite, hydromagnesite, and dypingite, which are some of the
prominent hydrated magnesium carbonates, were not observed in the XRD pattern of the carbonated reference sample. Series Rc
displayed a heightened peak of magnesium calcite, the formation of which is made possible by the impurities present in the reactive
MgO powder. Therefore, even in the Rc series, the precipitation of crystals such as magnesium calcite, that incorporate calcium, was
conjectured. Within the series SNc and FANc, peaks of magnesium calcite intensified compared to their non-carbonated variants.

Fig. 10. 28th day FTIR spectra of NS-free non-carbonated (a), NS-free carbonated (b), NS-containing non-carbonated (c), and NS-containing carbonated (d) series.

13
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

Additionally, new but weaker aragonite and magnesium carbonate peaks were discovered in the XRD of FAc and Sc samples. Higher
contents of CaO present in slag and fly ash may have facilitated the crystallization of greater amounts of such minerals. Furthermore,
smaller but visible peaks related to aragonite are discovered at 27◦ 2θ. Surprisingly, there are peaks at similar angles in the atmosphere-
cured sample too. However, those peaks are more prominent in carbonated specimens, indicating the effectiveness of accelerated
carbonation in producing carbonated minerals [34]. Magnesium calcite and aragonite, which are mostly formed in the
pozzolan-containing samples subjected to carbonation curing, enhance the morphology and interconnectivity as well as contribute to
compressive strength gain [85,86]. Peaks of hydrotalcite were discovered in the carbonation-cured pozzolanic series too. The curing
under carbonation visibly broadened the peaks associated with hydrotalcite, which may have arisen from the expenditure of
hydrotalcite during the carbonation process [78]. Hence, it corroborates the claim that hydrotalcite is a hydration product. The XRD
patterns of MSc and MSNc samples demonstrate major and minor peaks identical to MS and MSN samples, revealing the impotent effect
of accelerated carbonation curing in this binder system particularly. In general, the palpable differences between the minor peaks of
atmosphere-cured and carbonation-cured samples indicate the formation of HMCs and carbonate phases more profusely in the latter
group of samples. Thus, a denser and stronger microstructure developed in the carbonated samples. Therefore, it is concluded that the
formation of the main carbonation product (i.e., magnesium calcite) in the calcium-rich MgO-based mortars is facilitated by a curing
regime involving 10% or more CO2 concentration.
A comparison between the 3rd day and 28th day XRD spectra of the series to assess the morphological change can be made. The
decline in the intensity of the peaks associated with periclase can be seen on the 28th day compared to the 3rd day, which can be seen
clearly throughout the samples, suggesting the continued hydration of MgO. Apart from this, there was only little to no change, with
respect to time, in the XRD pattern of the series detected. Moreover, no accountable difference in the peaks of mixes containing NS was
detected, thus, elucidating the dormancy of NS in the MgO-based mortars.

3.5.2. FTIR
The FTIR spectra of all the samples at the age of 28 days are presented in Fig. 10. The largest absorption band in all the non-
carbonated specimens, being 3690 cm− 1, assigned to O–H stretching band, confirmed the formation of high amounts of brucite
[41]. In sample R, the absorption band at 1455 cm− 1 is characteristic of periclase, hinting at the incomplete hydration of MgO [87].
The Q2 Si–O stretching band at 990 cm− 1, which was observed in the S, SN, FA, FAN, MS, and MSN samples, is evident of dissolution of
silica available in fly ash, slag, and micro-silica to form of M-S-H gel [78]. The band at 1740 cm− 1 assigned to H2O bending modes of

Fig. 11. 28th day TG-DTG curves of NS-free non-carbonated (a), NS-free carbonated (b), NS-containing non-carbonated (c), and NS-containing carbonated (d) series.

14
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

vibration, which confirms the generation of hydrotalcite in the FA, S, and FAN samples, is in line with the XRD results (Figs. 8 and 9).
Within the non-carbonated series containing silica (i.e., MS, MSN, S, SN, FA, FAN), the structural OH− from M-S-H was confirmed by
the broad band between 3200 and 3600 cm− 1, which was not found in the reference sample [51,88]. The band at 1640 cm− 1, which is
missing from FTIR spectra of carbonated series, could be assigned to the absorbed water [51], which is the vibration mode of water
entrapped within pores. The FTIR results were mostly in line with the results of XRD and compressive strength, indicating a noticeable
difference between the spectra of samples based on their curing conditions.
Similar to the non-carbonated series, the OH− stretching band at 3690 cm− 1 wavelength was observed in the carbonated samples as
well, though the peak was considerably more immature, elucidating the conversion of Mg(OH)2 into carbonated products. The series
Rc, Sc, and FAc demonstrated the same bands at the same wavelength. However, the antisymmetric stretching band of CO2− 3 at 1380
cm− 1 was deeper for the pozzolanic series. Contrary to R, the absorption band at 1445 cm− 1 is missing from the FTIR spectrum of the
series Rc, which may explain the higher mechanical strength of the carbonated mixes compared to the non-carbonated ones. The
antisymmetric stretching band at ~1380 cm− 1 which is indicative of CO2− 3 [89,90], suggests the formation of carbonate-containing
phases such as nesquehonite, hydromagnesite, and dypingite in the reference series; and magnesium calcite in mixes containing fly
ash and slag. Additionally, according to Kumar et al., the band at 860 cm− 1 suggests the formation of calcite/magnesium calcite [78].
According to Kunzel et al. [91], CO2−
3 symmetric bending band at 860 cm
− 1
and CO2−
3 antisymmetric stretching band at 1100 cm
− 1
are
characteristic of HMCs and carbonate phases. Within the MSc sample, the symmetric and antisymmetric Si–O–Si bands at 800 and
1100 cm− 1, respectively, reveal the little unreacted amorphous silica [74]. The existence of this weak band confirms the overwhelming
dissolution of SiO2 and the formation of M-S-H. The results indicate the formation of carbonate phases in both carbonated and
non-carbonated specimens. The absorption bands associated with HMCs in the carbonated samples were significantly deeper
compared to non-carbonated samples, revealing the effect of accelerated carbonation on the microstructure of MgO-based materials.
The effect of NS on the FTIR spectra was not discernible, possibly due to its inertness in the hydration and carbonation reactions.

3.5.3. TG/DTG
The percentage of weight loss and the derivative of weight loss in contrast to the temperature on the 28th day are presented
(Fig. 11). The three major decomposition steps closely follow the literature [92–95], which are identified as.
• 0–350 ◦ C: Dehydration of pore water, water locked to M-S-H gel, and water bonded to HMCs.
• 350–450 ◦ C: Dehydroxylation of HMCs, decarbonation of HMCs, and decomposition of brucite.
• 450–900 ◦ C: Solely decarbonation of HMCs and magnesium calcite. There is, however, a debate about this range. Although
decarbonation is often reported to begin between 500 and 550 ◦ C [96]; in some cases, it is believed to start happening around
400 ◦ C [97,98]).
According to the DTG curves of the non-carbonated series, the peaks corresponding to the dehydration of water were observed at
slightly different temperatures of approximately 80 ◦ C in most of the mixes. It is important to notice that in the Rc series, the peaks
corresponding to water evaporation at 100 ◦ C and dehydroxylation of brucite are shallower and deeper, respectively, compared to the
rest. Thus, it conforms to the findings of the XRD and compressive strength test. For sample R, the endothermic peak representing the
decomposition of brucite was observed at the temperature of 370 ◦ C, while in the Rc mixes the endothermic peaks shifted to 400 ◦ C.
This shift in the decomposition temperature (the temperature at which the rate of mass loss is maximum) suggests that a change has
occurred in the amount and type of decomposed minerals. As mentioned before, between the temperature range of 350–450 ◦ C
decarbonation and dehydroxylation of HMCs take place in addition to the decomposition of Mg(OH)2. Since the carbonate phases tend
to decompose at higher temperatures, it is only logical to speculate that the higher amount of HMCs formed in the Rc sample altered the
temperature at which the endothermic peak materialized.
In the DTG curves of pozzolan-containing specimens, a significant peak representing the decomposition of brucite and dehy­
droxylation of HMCs was observed at 370–400 ◦ C. Similar to the reference sample, a shift in the endothermic peak of the rest of the

Table 3
Weight loss of mixes.

Series Weight loss (% of original)

0–350 ◦ C 350–450 ◦ C 450–900 ◦ C Total

R 6.6 22.7 3.6 33.0

S 13.2 13.5 3.7 30.3
FA 11.3 15.0 3.6 29.9
MS 11.9 19.9 4.5 36.4
Rc 8.6 25.1 4.1 37.7
Sc 13.2 16.8 5.1 35.2
FAc 11.9 19.9 4.5 36.4
MSc 12.5 14.4 5.2 32.1
SN 13.3 13.2 3.9 30.4
FAN 12.4 14.0 4.0 30.3
MSN 14.9 12.0 3.8 30.7
SNc 14.1 14.6 6.0 34.7
FANc 14.4 15.6 6.7 36.6
MSNc 13.4 14.3 5.2 32.9

15
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

samples also occurred, the cause of which is attributed to the same occurrence. The mass loss in the range of 170–300 ◦ C is reportedly
caused by dehydration of hydrotalcite and M-S-H, whereas mass loss until 170 ◦ C originates only from dehydration of M-S-H [29,99].
Decarbonation of HMCs, magnesium calcite, and hydrotalcite occurred in a more gradual manner. Further examination of TG, along
with the quantification of hydrate and carbonate phases within each mix, was provided by the estimation of weight losses at each
sudden change in the rate TG curves.
Table 3 presents the total weight loss and the weight loss that occurred at each range of temperatures. The results elucidate that
most of the weight loss occurred in the 350–450 ◦ C temperature range, which is associated with three main reactions: (i) dehy­
droxylation of hydromagnesite and dypingite, (ii) decarbonation of nesquehonite, and (iii) decomposition of uncarbonated brucite. Rc,
Sc, and FAc series undergo an average of 3.5% more weight loss in the 350–450 ◦ C temperature range compared to the R, S, and FA
series, which supports the existence and decomposition of carbonated phases in the former series. The specimens that were subjected to
accelerated carbonation curing experienced a 5.5% more weight loss in total compared to the specimens cured in the ambient con­
ditions. These findings are in agreement with the compressive strength evolution of the carbonated and uncarbonated series. When the
comparison between the total weight loss of the moisture-cured samples is made, MS underwent the largest weight loss due to the
decomposition of a larger amount of M-S-H gel and brucite, where the former reaction product is scarce in the rest of the mixes. Due to
the formation of a higher amount of hydration product compared to the rest, the second-largest total weight loss occurred in the
reference sample R, 33.0%, which is supported by the higher and sharper peaks of brucite seen in the XRD pattern (Fig. 8). There are
elevated weight losses in all the mixes subjected to accelerated carbonation, in contrast to those cured in ambient conditions, except
the MS mix. The weight loss of reference series R and Rc in the 0–350 ◦ C temperature range is significantly lower than in the remaining
samples, which can be attributed to the lack of free water in the pores or in between the layers of M-S-H gel in these series, which is also
substantiated with the low flowability of the mentioned series. A difference between the weight loss of R, S, FA, SN, FAN and Rc, Sc,
FAc, SNc, and FANc samples exists in the temperature range of 350–450 ◦ C, revealing the influence of accelerated carbonation on
reaction products of MgO-based materials. The extra weight loss observed in carbonated samples might as well be caused by the
dehydroxylation and decarbonation of HMCs, whereas the rest of the evaporated weight may be associated with the decomposition of
brucite. The discrepancy in the weight loss at the 450–900 ◦ C temperature range between the mixes exposed to separate curing
regiments might only stem from the generation of carbonated phases. The comparison between total weight losses of the mixes sheds
light on the amount of entirety of the reaction products, where H2O and CO2 are chemically bound within the sample.

Fig. 12. SEM images of series R (a), Rc (b), S (c), Sc (d), FA (e), FAc (f), MS (g), and MSc (h).

16
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

Fig. 13. SEM images depicting unreacted particles in series S (a), FA (b), and MS (c).

3.5.4. SEM
SEM images of carbonated and non-carbonated samples are presented in Fig. 12. In general, the samples cured via conventional
curing techniques displayed a more scattered microstructure. The series R demonstrated a discontinuous and loose network structure.
Additionally, the grainy surface exhibited by the series R may correlate with the existence of unhydrated periclase particles. Similarly,
the non-carbonated pozzolanic samples (e.g., FA and S) possessed rather porous and weak phases. Fig. 13 highlights the conspicuous
unreacted particles which may have hindered compressive strength evolution in the pozzolanic mixes. Notably, the unreacted
spherical particles of fly ash (Fig. 13b) were detected as well as particles of slag (Fig. 13a). These sightings can shed light on the weaker
compressive strength of the pozzolanic series (e.g., FA, FAN, S, SN); such that, anhydrous fly ash and slag particles may inhibit the
formation of interconnected conglomerates that were observed in the carbonated series. The series MS had a particularly smooth and
uninterrupted interphase, which elucidates the formation of M-S-H gel, as described by Kumar et al. [78]. Even so, cracking (Fig. 13c)
caused by the stress on the SiO2 particles due to the shrinkage of M-S-H is the possible reason for the diminished compressive strength
of MS-containing samples in the later ages. Fig. 14 presents the EDS quantification of some of the series, including the series R
(Fig. 14a), which highlights magnesium (Mg) and oxygen (O) as constituting most of the sample by weight. Since hydrogen (H) cannot
be detected with such a technique, the simultaneous presence of both Mg and O points to the precipitation of brucite and the presence
of unreacted periclase. On top of Mg and O, noticeable quantities of Calcium (Ca), Aluminum (Al), sodium (Na), Silicon (Si), and Iron
(Fe) were discovered in the EDS of the series FA (Fig. 14c). The detection of such elements is in line with the chemical composition of
hydrotalcite. However, due to the incohesive nature of crystals, high compressive strength ought not to be expected. The compara­
tively higher percentage of Si detected in the EDS image of the series MS also confirms the existence of M-S-H gel (Fig. 14e). However,
all the MS-containing series displayed heterogeneous particles which resemble spherical particles of silica, based on the detection of
immense disparity in the percentage of Si compared to neighboring areas, which stand out from the rest of the formations. The
prominent binding phases that were captured by the electron microscope are illustrated in Fig. 14. The laminar phases, which can also
be seen in the reference series, is be assigned to brucite (Fig. 15a).
The carbonation-cured series, in general, demonstrated a denser and more interconnected microstructure compared to non-
carbonated ones. Round-shaped assemblages were observed in the SEM image of the Rc series (Fig. 15b). Similar formations were
discovered in the MgO-based materials by Mo et al. [38], which they assigned to magnesium calcite. As a consequence of the pre­
cipitation of magnesium calcite, formation of which is also supported by the XRD spectrum (Fig. 8), the Rc series may have displayed a
much more compact morphology. Signs of Carbon were detected by EDS only in the Rc sample (Fig. 14b), which further corroborates

17
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

Fig. 14. EDS quantification of series R (a), Rc (b), FA (c), S (d), and MS (e).

Fig. 15. SEM images depicting reaction products in R (a) and Rc (b) series.

18
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

the formation of carbonate phases. Such formations were not prevalent in the non-carbonated R series, confirming the precipitation of
carbonate phases mostly in samples exposed to accelerated carbonation. The porousness of the series FAc (Fig. 12f) is significantly less
than what was observed in the FA series (Fig. 12e). Furthermore, SEM figure of the series Sc (Fig. 11d) depicts that the sample is devoid
of cracks considerably compared to its non-carbonated counterpart (Fig. 12c). In the carbonated reference series, the brucite as­
semblages were significantly less conspicuous compared to the series R.
The effect of NS on the microstructure of the MgO-based mortars other than slight densification of the surface in all the NS-
containing mixes could not be discerned through SEM analysis, which elucidates the passivity of NS in their reaction mechanisms.

3.6. MIP
The pore size distribution and total porosity of carbonated and non-carbonated series cured for 28 days is given in Fig. 16. Most
notably, the total porosity and the size of large capillary pores decreased with the help of carbonation. Additionally, the effect of NS is
conspicuous; such that, it lowered the pore volume. It affirms the notion that NS functions as a filler component. It will provide high
impermeability against harmful substances and good durability. As expected, the reference mix had the lowest porosity among the
carbonated and non-carbonated series, which conforms to their high compressive strength, as presented in Fig. 7. It is worth noting that
the effect of NS in reducing porosity in the atmospheric series is much more apparent than in the carbonated ones due to drop caused by
the curing procedure. The highest total porosity of approximately 12% was observed in the FA series. The fly ash and slag-containing
mixes generally demonstrated greater pore size and pore volume. The addition of slag leading to a coarser pore size distribution has
been previously reported [43]. The change in the slope of pozzolanic series curves between the pores of diameter 0.01 μm and 0.3 μm is
most likely caused by the interlayer structure M-S-H [42].

Fig. 16. Total porosity and pore size distribution of non-carbonated (a, c) and carbonated (b, d) series, respectively.

19
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

4. Conclusions
The aim of this study was to investigate the relevant mechanical performance and microstructural evolution of MgO-based mortars
cured with two different curing techniques: ambient curing and accelerated carbonation curing. The effect of 10% CO2 concentration
for up to 28 days was contrasted with the conventional curing regimen, as well as the influence of pozzolanic and nanoscopic additives
was examined. The conclusions drawn from the results of this paper are stated in the sections below.

• The addition of pozzolans significantly improved the flow diameter. Expressly, the introduction of fly ash into the mix approxi­
mately doubled the flow diameter of the reference mortar. Fly ash particles’ spherical shape, relatively low fineness, and their
inability to quickly hydrate may have caused higher workability. The ultra-fine nature of RMC used in this study, in addition to its
predilection to coagulation, is speculated to have lowered the flowability.
• The reference and NS-containing mixes demonstrated the lowest initial and final setting times. The high precipitation rate of
brucite and catalytic effect of NS are hypothesized to have triggered the comparatively premature setting in these mixes. Since the
precipitation rate of M-S-H gel, which is a hydration product in pozzolan-containing blends, requires more time, thusly the increase
in the setting times of binary mixes can be explained.
• Chiefly, the impact of accelerated carbonation curing in lowering the pH value throughout the mixes was unmistakable. pH value,
which was measured on the 3rd, 7th, and 28th day, gradually decreased. The inclusion of slag raised the pH value the most, which
might stem from the fact that concurrent dissolution of CaO and MgO results in a higher dose of OH− . Among the non-carbonated
series, the specimen containing micro-silica indicated the lowest pH. The decrease in pH is a sign of hydration and mechanical
strength development, which corresponds to higher compressive strength of the MS series. The effect of NS was negligible in
regulating the pH.
• Compressive strength of all the samples increased until the 28th day, with the exception of the MS-containing series. Slag and fly
ash had a depreciatory effect on the compressive strength of MgO-based systems, even though their respective series demonstrated
satisfactory results under accelerated carbonation. Crystallization of M-S-H gel in the MS and MSN series, wherein accelerated
carbonation is not a prerequisite, produced the highest strength (21.1 and 26.4 MPa respectively) amid the ambient cured series.
Carbonation curing raised the compressive strength up to 488% and the highest compressive strength of 52.1 MPa was achieved in
the Rc series. NS improved the compressive strength in the ternary systems. However, the enhancement most likely arises from its
filler effect.
• The progressive growth of brucite became apparent with the comparison of the 3rd day and 28th day XRD spectra of MgO-based
mortars. In addition, noticeable quantities of brucite and periclase were observed in FTIR analysis. The formation of M-S-H gel in
silica-containing series was confirmed through XRD, FTIR, and SEM. In the microstructure of carbonated series, crystallization of a
carbonate phase is postulated, the most likely candidate being magnesium calcite. Moreover, precipitation of hydrotalcite and
aragonite was not uncommon. Based on TG/DTG results, the total weight loss of carbonated specimens was on average 5.5% higher
than non-carbonated ones; this, in fact, corroborates the claim that the generation of reaction products denser than the precursors
takes place in carbonated mediums. In general, SEM images contrasted the grainy surface of atmosphere-cured samples versus the
more compact exterior of the samples cured with carbonation. Porous phases as well as unreacted particles of fly ash and slag were
detected in the pozzolanic series, which explains the reduction in compressive strength. The decline in the 28th day strength of the
MS series was justified by the observation of the shrinkage cracks induced by the stress concentration around SiO2 particles.
• Porosimetry analysis most importantly confirmed the role of NS as a filler component. The carbonation curing and NS reduced the
cumulative porosity while decreasing the pore size of capillary pores.
In a nutshell, reactive MgO cement is a remarkably more sustainable alternative binder compared to OPC. Its sufficient strength,
satisfactory fresh-state properties, dense microstructure, and low porosity, especially in the NS-containing carbonated specimens,
warrant its usage in the structural concrete elements. However, more robust investigations on the durability properties such as
shrinkage resistance, impermeability, and soundness are necessary to form a consensus on the long-term application of such materials.
Chiefly, based on the visual inspection of hardened pastes, cracks induced most probably by shrinkage requires further examination.

CRediT author statement

Khalilullah Taj: Conceptualization, Methodology, Validation, Investigation, Writing - Original Draft, Visualization Busra Akturk:
Conceptualization, Methodology, Validation, Investigation, Resources, Writing – Review & Editing, Supervision Serhan Ulukaya:
Conceptualization, Methodology, Validation, Resources, Writing – Review & Editing, Supervision, Project administration, Funding
acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data availability

Data will be made available on request.

20
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

Acknowledgments
RMC and slag operated in this work were supplied by KUMAS AS and AKCANSA, respectively. The authors thank them for their
contributions.

References
[1] N. Ying, D. Zhou, Q. Chen, Q. Ye, Z. Han, Long-term link detection in the CO2 concentration climate network, J. Clean. Prod. 208 (2019) 1403–1408, https://
doi.org/10.1016/j.jclepro.2018.10.093.
[2] N. Mohamad, K. Muthusamy, R. Embong, A. Kusbiantoro, M.H. Hashim, Environmental impact of cement production and Solutions: a review, Mater. Today
Proc. 48 (2021) 741–746, https://doi.org/10.1016/j.matpr.2021.02.212.
[3] S.A. Miller, A. Horvath, P.J.M. Monteiro, Readily implementable techniques can cut annual CO2 emissions from the production of concrete by over 20%,
Environ. Res. Lett. 11 (7) (2016) https://doi.org/10.1088/1748-9326/11/7/074029.
[4] B. Metz, O. Davidson, H. de Coninck, M. Loos, L. Meyer, Carbon Dioxide Capture and Storage, 2005.
[5] J. Deja, A. Uliasz-Bochenczyk, E. Mokrzycki, CO2 emissions from Polish cement industry, Int. J. Greenh. Gas Control 4 (4) (2010) 583–588, https://doi.org/
10.1016/j.ijggc.2010.02.002.
[6] Y. Tan, C. Wu, H. Yu, Y. Li, J. Wen, Review of reactive magnesia-based cementitious materials: current developments and potential applicability, J. Build. Eng.
40 (October 2020) (2021), 102342, https://doi.org/10.1016/j.jobe.2021.102342.
[7] M. Schneider, M. Romer, M. Tschudin, H. Bolio, Sustainable cement production-present and future, Cement Concr. Res. 41 (7) (2011) 642–650, https://doi.org/
10.1016/j.cemconres.2011.03.019.
[8] R. Hay, C. Otchere, G. Kashwani, K. Celik, Recycling carbonated reactive magnesium cement (RMC) as a building material, J. Clean. Prod. 320 (January) (2021),
128838, https://doi.org/10.1016/j.jclepro.2021.128838.
[9] C. Sonat, C.H. Lim, M. Liska, C. Unluer, Recycling and reuse of reactive MgO cements – a feasibility study, Construct. Build. Mater. 157 (2017) 172–181, https://
doi.org/10.1016/j.conbuildmat.2017.09.068.
[10] E. Gartner, T. Sui, Alternative cement clinkers, Cement Concr. Res. (2017), https://doi.org/10.1016/j.cemconres.2017.02.002.
[11] F. Winnefeld, E. Epifania, F. Montagnaro, E.M. Gartner, Further studies of the hydration of MgO-hydromagnesite blends, Cement Concr. Res. 126 (November)
(2019), 105912, https://doi.org/10.1016/j.cemconres.2019.105912.
[12] V. Luong, R. Amal, J.A. Scott, S. Ehrenberger, T. Tran, A comparison of carbon footprints of magnesium oxide and magnesium hydroxide produced from
conventional processes, J. Clean. Prod. (2018), https://doi.org/10.1016/j.jclepro.2018.08.225.
[13] O. Jankovský, et al., Carbon dioxide uptake by MOC-based materials, Appl. Sci. 10 (7) (2020) 1–11, https://doi.org/10.3390/app10072254.
[14] H. Deyi, O. David, Sustainable Remediation of Contaminated Soil and Groundwater, Butterworth-Heinemann, 2019.
[15] S.A. Walling, J.L. Provis, Magnesia-based cements: a journey of 150 Years, and cements for the future? Chem. Rev. 116 (7) (2016) 4170–4204, https://doi.org/
10.1021/acs.chemrev.5b00463.
[16] M. Hewlett, P. Liska, Lea’s Chemistry of Cement and Concrete, 2019.
[17] Y. Guo, Y.X. Zhang, K. Soe, R. Wuhrer, W.D. Hutchison, H. Timmers, Development of magnesium oxychloride cement with enhanced water resistance by adding
silica fume and hybrid fly ash-silica fume, J. Clean. Prod. 313 (April) (2021), 127682, https://doi.org/10.1016/j.jclepro.2021.127682.
[18] H.A. Abdel-Gawwad, S.A. Abo El-Enein, M. Heikal, S. Abd El-Aleem, A.A. Amer, I.M. El-Kattan, Synergistic effects of curing conditions and magnesium oxide
addition on the physico-mechanical properties and firing resistivity of Portland cement mortar, Construct. Build. Mater. 176 (2018) 676–689, https://doi.org/
10.1016/j.conbuildmat.2018.05.092.
[19] R. Hay, N.T. Dung, A. Lesimple, C. Unluer, K. Celik, Mechanical and microstructural changes in reactive magnesium oxide cement-based concrete mixes
subjected to high temperatures, Cem. Concr. Compos. 118 (January) (2021), https://doi.org/10.1016/j.cemconcomp.2021.103955.
[20] L. Mo, M. Deng, M. Tang, Effects of calcination condition on expansion property of MgO-type expansive agent used in cement-based materials, Cement Concr.
Res. 40 (3) (2010) 437–446, https://doi.org/10.1016/j.cemconres.2009.09.025.
[21] Z. Xing, L. Bai, Y. Ma, D. Wang, M. Li, Mechanism of magnesium oxide hydration based on the multi-rate model, Materials 11 (1835) (2018) 1–13, https://doi.
org/10.3390/ma11101835.
[22] G.L. Smithson, N.N. Bakhshi, Kinetics and mechanism of carbonation of magnesium oxide slurries, Ind. Eng. Chem. Process Des. Dev. 12 (1) (1973) 99–106.
[23] J.J. Thomas, S. Musso, I. Prestini, Kinetics and activation energy of magnesium oxide hydration, J. Am. Ceram. Soc. 97 (1) (2014) 275–282, https://doi.org/
10.1111/jace.12661.
[24] M.E. Aphane, E.M. Van Der Merwe, C.A. Strydom, Influence of Hydration Time on the Hydration of MgO in Water and in a Magnesium Acetate Solution, vol. 96,
2009, pp. 987–992.
[25] X. Tang, L. Guo, C. Chen, Q. Liu, T. Li, Y. Zhu, The analysis of magnesium oxide hydration in three-phase reaction system, J. Solid State Chem. 213 (2014)
32–37, https://doi.org/10.1016/j.jssc.2014.01.036.
[26] L.F. Amaral, I.R. Oliveira, R. Salomao, E. Frollini, V.C. Pandolfelli, Temperature and common-ion effect on magnesium oxide (MgO), hydration 36 (2010)
1047–1054, https://doi.org/10.1016/j.ceramint.2009.12.009.
[27] K.P. Matabola, E. Van der Merwe, C.A. Strydom, F.J.W. Labuschagne, The influence of hydrating agents on the hydration of industrial magnesium oxide, Soc.
Chem. Ind. (July) (2010) 1569–1574, https://doi.org/10.1002/jctb.2467.
[28] D. Filippou, N. Katiforis, On the Kinetics of Magnesia Hydration in Magnesium Acetate Solutions, vol. 328, 1999, pp. 322–328. October 1998.
[29] T. Zhang, L.J. Vandeperre, C.R. Cheeseman, Formation of magnesium silicate hydrate (M-S-H) cement pastes using sodium hexametaphosphate, Cement Concr.
Res. 65 (2014) 8–14, https://doi.org/10.1016/j.cemconres.2014.07.001.
[30] Y. Jia, et al., Role of sodium hexametaphosphate in MgO/SiO2 cement pastes, Cement Concr. Res. 89 (2016) 63–71, https://doi.org/10.1016/j.
cemconres.2016.08.003.
[31] E. Grünhäuser Soares, J. Castro-Gomes, Carbonation curing influencing factors of carbonated reactive magnesia cements (CRMC) – a review, J. Clean. Prod. 305
(2021) 127210, https://doi.org/10.1016/j.jclepro.2021.127210.
[32] N.T. Dung, A. Lesimple, R. Hay, K. Celik, C. Unluer, Formation of carbonate phases and their effect on the performance of reactive MgO cement formulations,
Cement Concr. Res. 125 (November 2018) (2019), 105894, https://doi.org/10.1016/j.cemconres.2019.105894.
[33] C. Unluer, A. Al-Tabbaa, Enhancing the carbonation of MgO cement porous blocks through improved curing conditions, Cement Concr. Res. 59 (2014) 55–65,
https://doi.org/10.1016/j.cemconres.2014.02.005.
[34] L. Mo, F. Zhang, M. Deng, Effects of carbonation treatment on the properties of hydrated fly ash-MgO-Portland cement blends, Construct. Build. Mater. 96
(2015) 147–154, https://doi.org/10.1016/j.conbuildmat.2015.07.193.
[35] S. Kumar, E.H. Yang, C. Unluer, Investigation of chloride penetration in carbonated reactive magnesia cement mixes exposed to cyclic wetting–drying,
Construct. Build. Mater. 284 (2021) 122837, https://doi.org/10.1016/j.conbuildmat.2021.122837.
[36] D.K. Panesar, L. Mo, Properties of binary and ternary reactive MgO mortar blends subjected to CO2 curing, Cem. Concr. Compos. 38 (2013) 40–49, https://doi.
org/10.1016/j.cemconcomp.2013.03.009.
[37] R.O. Teskey, A. Saveyn, K. Steppe, M.A. McGuire, Origin, fate and significance of CO2 in tree stems, New Phytol. 177 (1) (2008) 17–32, https://doi.org/
10.1111/j.1469-8137.2007.02286.x.
[38] L. Mo, D.K. Panesar, Effects of accelerated carbonation on the microstructure of Portland cement pastes containing reactive MgO, Cement Concr. Res. 42 (6)
(2012) 769–777, https://doi.org/10.1016/j.cemconres.2012.02.017.

21
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

[39] N.T. Dung, R. Hay, A. Lesimple, K. Celik, C. Unluer, Influence of CO2 concentration on the performance of MgO cement mixes, Cem. Concr. Compos. 115 (July
2020) (2021), 103826, https://doi.org/10.1016/j.cemconcomp.2020.103826.
[40] W. Shen, L. Cao, Q. Li, Z. Wen, J. Wang, Y. Liu, Is magnesia cement low carbon ? Life cycle carbon footprint comparing with Portland cement, J. Clean. Prod.
131 (2016) 20–27, https://doi.org/10.1016/j.jclepro.2016.05.082.
[41] N.T. Dung, C. Unluer, Carbonated MgO concrete with improved performance: the influence of temperature and hydration agent on hydration, carbonation and
strength gain, Cem. Concr. Compos. 82 (2017) 152–164, https://doi.org/10.1016/j.cemconcomp.2017.06.006.
[42] R. Zhang, D.K. Panesar, Carbonated binder systems containing reactive MgO and Portland cement: strength, chemical composition and pore structure, J. Clean.
Prod. 271 (2020), 122021, https://doi.org/10.1016/j.jclepro.2020.122021.
[43] L. Mo, D.K. Panesar, Accelerated carbonation - a potential approach to sequester CO2 in cement paste containing slag and reactive MgO, Cem. Concr. Compos.
43 (2013) 69–77, https://doi.org/10.1016/j.cemconcomp.2013.07.001.
[44] H. Yu, Y. Yi, C. Unluer, Heat of hydration, bleeding, viscosity, setting of Ca(OH)2-GGBS and MgO-GGBS grouts, Construct. Build. Mater. 270 (2021), 121839,
https://doi.org/10.1016/j.conbuildmat.2020.121839.
[45] J. He, W. Zheng, W. Bai, T. Hu, J. He, X. Song, Effect of reactive MgO on hydration and properties of alkali-activated slag pastes with different activators,
Construct. Build. Mater. 271 (2021), 121608, https://doi.org/10.1016/j.conbuildmat.2020.121608.
[46] S. Ruan, C. Unluer, Influence of supplementary cementitious materials on the performance and environmental impacts of reactive magnesia cement concrete,
J. Clean. Prod. 159 (2017) 62–73, https://doi.org/10.1016/j.jclepro.2017.05.044.
[47] C. Unluer, A. Al-Tabbaa, The role of brucite, ground granulated blastfurnace slag, and magnesium silicates in the carbonation and performance of MgO cements,
Construct. Build. Mater. 94 (2015) 629–643, https://doi.org/10.1016/j.conbuildmat.2015.07.105.
[48] J.C. Yang, The system magnesia-silica-water below 300◦ C.: I, low-temperature phases from 100◦ to 300◦ C. and their properties, J. Am. Ceram. Soc. 43 (10)
(1960) 542–549, https://doi.org/10.1111/j.1151-2916.1960.tb13610.x.
[49] G.L. Kalousek, D. Mui, Studies on formation and recrystallization of intermediate reaction products in the system magnesia-silica-water, J. Am. Ceram. Soc. 37
(2) (1954) 38–42, https://doi.org/10.1111/j.1151-2916.1954.tb14001.x.
[50] O.Y. Golubeva, E.N. Korytkova, V.V. Gusarov, Hydrothermal synthesis of magnesium silicate montmorillonite for polymer-clay nanocomposites, Russ. J. Appl.
Chem. 78 (1) (2005) 26–32, https://doi.org/10.1007/s11167-005-0225-z.
[51] J. Temuujin, K. Okada, K.J.D. MacKenzie, Role of water in the mechanochemical reactions of MgO-SiO2Systems, J. Solid State Chem. 138 (1) (1998) 169–177,
https://doi.org/10.1006/jssc.1998.7768.
[52] Z. Li, Y. Xu, T. Zhang, J. Hu, J. Wei, Q. Yu, Effect of MgO calcination temperature on the reaction products and kinetics of MgO – SiO 2 – H 2 O system, J. Am.
Ceram. Soc. (November) (2018) 1–17, https://doi.org/10.1111/jace.16201.
[53] H. Ono, S.I. Wada, Properties of layer silicates formed from MgO-SiO2-H 2O mixtures at 25◦ C, J. Fac. Agric. Kyushu Univ. 52 (1) (2007) 159–162, https://doi.
org/10.5109/9299.
[54] T. Mitsuda, H. Taguchi, Formation of magnesium silicate hydrate and its crystallization to talc, Cement Concr. Res. 7 (3) (1977) 223–230, https://doi.org/
10.1016/0008-8846(77)90083-7.
[55] G.V.P.B. Singh, C. Sonat, E.H. Yang, C. Unluer, Performance of MgO and MgO – SiO 2 Systems Containing Seeds under Different Curing Conditions, vol. 108,
2020. December 2019.
[56] ASTM C 618, Standard Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete, ASTM Int., 2014, pp. 1–5, https://doi.org/
10.1520/C0618.
[57] B. E. 196-1, Methods of testing cement Part 1: determination of strength, Br. Stand. (2016) 1–33.
[58] ASTM C1437, Standard Test Method for Flow of Hydraulic Cement Mortar, ASTM Int., 2007, pp. 6–7.
[59] ASTM C-191, Time of setting of hydraulic cement by Vicat needle, Annu. Book ASTM Stand. i (2008) 1–10, 191-04.
[60] ASTM C109, Standard test method for compressive strength of hydraulic cement mortars, ASTM Int 4 (May) (1999) 1–6.
[61] T. Gonçalves, R.V. Silva, J. de Brito, J.M. Fernández, A.R. Esquinas, Mechanical and durability performance of mortars with fine recycled concrete aggregates
and reactive magnesium oxide as partial cement replacement, Cem. Concr. Compos. 105 (September 2019) (2020) 1–10, https://doi.org/10.1016/j.
cemconcomp.2019.103420.
[62] H.M. Tran, A. Scott, Strength and workability of magnesium silicate hydrate binder systems, Construct. Build. Mater. 131 (2017) 526–535, https://doi.org/
10.1016/j.conbuildmat.2016.11.109.
[63] S. Ruan, C. Unluer, Effect of air entrainment on the performance of reactive MgO and PC mixes, Construct. Build. Mater. 142 (2017) 221–232, https://doi.org/
10.1016/j.conbuildmat.2017.03.068.
[64] D.H. Vo, C.L. Hwang, M.D. Yehualaw, M.C. Liao, The influence of MgO addition on the performance of alkali-activated materials with slag− rice husk ash
blending, J. Build. Eng. 33 (January 2020) (2021), 101605, https://doi.org/10.1016/j.jobe.2020.101605.
[65] R. Zhang, D.K. Panesar, New approach to calculate water film thickness and the correlation to the rheology of mortar and concrete containing reactive MgO,
Construct. Build. Mater. 150 (2017) 892–902, https://doi.org/10.1016/j.conbuildmat.2017.05.218.
[66] M.G. Khalil, F. Elgabbas, M.S. El-Feky, H. El-Shafie, Performance of geopolymer mortar cured under ambient temperature, Construct. Build. Mater. 242 (2020),
118090, https://doi.org/10.1016/j.conbuildmat.2020.118090.
[67] C.L. Hwang, D.H. Vo, V.A. Tran, M.D. Yehualaw, Effect of high MgO content on the performance of alkali-activated fine slag under water and air curing
conditions, Construct. Build. Mater. 186 (2018) 503–513, https://doi.org/10.1016/j.conbuildmat.2018.07.129.
[68] Y. Qing, Influence of Nano-SiO 2 Addition on Properties of Hardened Cement Paste as Compared with Silica Fume, vol. 21, 2007, pp. 539–545, https://doi.org/
10.1016/j.conbuildmat.2005.09.001.
[69] Y. Chen, Y. Deng, M. Li, Influence of Nano-SiO 2 on the Consistency , Setting Time, Early-Age Strength , and Shrinkage of Composite Cement Pastes, vol. 2016,
2016.
[70] S. Diamond, Effects of two Danish flyashes on alkali contents of pore solutions of cement-flyash pastes, Cement Concr. Res. 11 (c) (1981) 383–394.
[71] R. Hay, K. Celik, Cement and Concrete Research Hydration , carbonation , strength development and corrosion resistance of reactive MgO cement-based
composites, Cement Concr. Res. 128 (November 2019) (2020), 105941, https://doi.org/10.1016/j.cemconres.2019.105941.
[72] F. Jin, A. Al-Tabbaa, Strength and hydration products of reactive MgO-silica pastes, Cem. Concr. Compos. 52 (2014) 27–33, https://doi.org/10.1016/j.
cemconcomp.2014.04.003.
[73] T. Zhang, C.R. Cheeseman, L.J. Vandeperre, Development of low pH cement systems forming magnesium silicate hydrate (M-S-H), Cement Concr. Res. 41 (4)
(2011) 439–442, https://doi.org/10.1016/j.cemconres.2011.01.016.
[74] C. Sonat, N.T. Dung, C. Unluer, Performance and microstructural development of MgO–SiO2 binders under different curing conditions, Construct. Build. Mater.
154 (2017) 945–955, https://doi.org/10.1016/j.conbuildmat.2017.08.020.
[75] T. Zhang, et al., Control of drying shrinkage in magnesium silicate hydrate (m-s-h) gel mortars, Cement Concr. Res. 88 (2016) 36–42, https://doi.org/10.1016/j.
cemconres.2016.05.011.
[76] K. Taj, B. Akturk, S. Ulukaya, Fresh state properties and compressive strength development of reactive MgO-based systems, Mater. Today Proc. 65 (2022)
1064–1069, https://doi.org/10.1016/j.matpr.2022.04.143.
[77] BS EN 771-3, Specification for masonry units - aggregate concrete masonry units (Dense and lightweight aggregates), Br. Stand. (2003) 4–16, https://doi.org/
10.1617/s11527-010-9622-8.
[78] S. Kumar, C. Sonat, E.H. Yang, C. Unluer, Performance of reactive magnesia cement formulations containing fly ash and ground granulated blast-furnace slag,
Construct. Build. Mater. 232 (2020), 117275, https://doi.org/10.1016/j.conbuildmat.2019.117275.
[79] L.J. Vandeperre, M. Liska, Hydration and mechanical properties of magnesia, Pulverized Fuel Ash, and Portland Cement Blends 20 (May) (2008) 375–383.
[80] R. Hay, K. Celik, Hydration , carbonation , strength development and corrosion resistance of reactive MgO cement-based composites, Cement Concr. Res. 128
(January) (2020), 105941, https://doi.org/10.1016/j.cemconres.2019.105941.

22
K. Taj et al. Journal of Building Engineering 66 (2023) 105856

[81] P. Nath, P. Sarker, Effect of fly ash on the durability properties of high strength concrete, Procedia Eng. 14 (2011) 1149–1156, https://doi.org/10.1016/j.
proeng.2011.07.144.
[82] D.M. Roy, P. Arjunan, M.R. Silsbee, Effect of silica fume , metakaolin , and low-calcium fly ash on chemical resistance of concrete, Cement Concr. Res. 31 (2001)
1809–1813.
[83] D.R.M. Brew, F.P. Glasser, Synthesis and characterisation of magnesium silicate hydrate gels, Cement Concr. Res. 35 (1) (2005) 85–98, https://doi.org/
10.1016/j.cemconres.2004.06.022.
[84] J. Temuujin, K. Okada, K.J.D. MacKenzie, Formation of layered magnesium silicate during the aging of magnesium hydroxide-silica mixtures, J. Am. Ceram.
Soc. 81 (3) (1998) 754–756, https://doi.org/10.1111/j.1151-2916.1998.tb02405.x.
[85] L. Mo, F. Zhang, M. Deng, Effects of carbonation treatment on the properties of hydrated fly ash-MgO-Portland cement blends, Construct. Build. Mater. 96
(2015) 147–154, https://doi.org/10.1016/j.conbuildmat.2015.07.193.
[86] L. Mo, F. Zhang, D.K. Panesar, M. Deng, Development of low-carbon cementitious materials via carbonating Portland cement-fly ash-magnesia blends under
various curing scenarios: a comparative study, J. Clean. Prod. 163 (2017) 252–261, https://doi.org/10.1016/j.jclepro.2016.01.066.
[87] H. Ono, S.-I. Wada, Effects of active silica on stabilization reaction of copper and nickel ions by magnesia, Clay Sci. 13 (2007) 149–157.
[88] T. Perraki, A. Orfanoudaki, STUDY OF RAW AND THERMALLY TREATED SEPIOLITE FROM THE MANTOUDI AREA , EUBOEA , Greece X-ray diffraction , TG/
DTG/DTA and FTIR investigations, J. Therm. Anal. Calorim. 91 (2008) 589–593.
[89] O. Kangal, C. Firat, A. Guney, Flotation properties of unusual carbonates : huntite and hydromagnesite, Miner. Eng. 18 (2005) 631–634, https://doi.org/
10.1016/j.mineng.2004.09.006.
[90] R.L. Frost, S.J. Palmer, Infrared and infrared emission spectroscopy of nesquehonite Mg(OH)(HCO 3)⋅2H2O-implications for the formula of nesquehonite,
Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 78 (4) (2011) 1255–1260, https://doi.org/10.1016/j.saa.2010.12.059.
[91] C. Kuenzel, F. Zhang, C.R. Cheeseman, E.M. Gartner, The mechanism of hydration of MgO-hydromagnesite blends, Cement Concr. Res. 103 (October 2017)
(2018) 123–129, https://doi.org/10.1016/j.cemconres.2017.10.003.
[92] R.L. Frost, S. Bahfenne, J. Graham, W.N. Martens, Thermochimica Acta Thermal Stability of Artinite, Dypingite and Brugnatellite — Implications for the
Geosequestration of Green House Gases, vol. 475, 2008, pp. 39–43, https://doi.org/10.1016/j.tca.2008.06.007.
[93] S. Purwajanti, et al., Synthesis of magnesium oxide hierarchical microspheres: a dual-functional material for water remediation, ACS Appl. Mater. Interfaces 7
(38) (2015) 21278–21286, https://doi.org/10.1021/acsami.5b05553.
[94] G. Jauffret, J. Morrison, F.P. Glasser, On the thermal decomposition of nesquehonite, J. Therm. Anal. Calorim. 122 (2) (2015) 601–609, https://doi.org/
10.1007/s10973-015-4756-0.
[95] L.A. Hollingbery, T.R. Hull, The thermal decomposition of huntite and hydromagnesite-A review, Thermochim. Acta 509 (1–2) (2010) 1–11, https://doi.org/
10.1016/j.tca.2010.06.012.
[96] J. Lanas, J.I. Alvarez, Dolomitic lime: thermal decomposition of nesquehonite, Thermochim. Acta 421 (1–2) (2004) 123–132, https://doi.org/10.1016/j.
tca.2004.04.007.
[97] V. Vágvölgyi, M. Hales, R.L. Frost, A. Locke, J. Kristóf, E. Horváth, Conventional and controlled rate thermal analysis of nesquehonite Mg(HCO3)(OH)•2(H2O),
J. Therm. Anal. Calorim. 94 (2) (2008) 523–528, https://doi.org/10.1007/s10973-007-8844-7.
[98] M.C. Hales, R.L. Frost, W.N. Martens, Thermo-Raman spectroscopy of synthetic nesquehonite - implication for the geosequestration of greenhouse gases,
J. Raman Spectrosc. 39 (March) (2008) 1141–1149, https://doi.org/10.1002/jrs.
[99] L. Pu, C. Unluer, Durability of Carbonated MgO Concrete Containing Fly Ash and Ground Granulated Blast-Furnace Slag, vol. 192, 2018, pp. 403–415, https://
doi.org/10.1016/j.conbuildmat.2018.10.121.

23