... Li, Investigation On The Reactivity of Recycled Brick Powder

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

Cement and Concrete Composites 139 (2023) 105042

Contents lists available at ScienceDirect

Cement and Concrete Composites


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

Investigation on the reactivity of recycled brick powder


Shujun Li a, b, Gaofeng Chen a, b, Yasong Zhao a, b, Zhenhai Xu a, b, Xu Luo a, b, Cheng Liu a, b,
Jianming Gao a, b, *
a
School of Materials Science and Engineering, Southeast University, Nanjing, 211189, China
b
Jiangsu Key Laboratory for Construction Materials, Nanjing, 211189, China

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

Keywords: Knowing the reactivity of recycled brick powder (RBP) is fundamental for its correct and efficient utilization. In
Recycled brick powder this work, the dissolution behavior of RBP in NaOH solution and the degree of reaction in cement matrix were
Dissolution kinetics studied to investigate the reactivity of RBP. The results revealed that the main active species of RBP, Si and Al,
Degree of reaction
showed obvious lasting dissolution with a significant non-steady state dissolution stage, indicating that the re­
Reaction rim
action proceeds slowly. The dissolution mechanism controlled by the phase boundary reaction, with activation
energies of 82.74 kJ/mol and 75.79 kJ/mol for Si and Al dissolution, respectively, further confirmed this finding.
The strength activity index, thermogravimetric analysis, selective dissolution, BSE-IA-Mapping and MAS NMR all
confirmed reactivity of RBP. Additionally, reaction rims were observed around RBP particles in cement paste.
The low Ca/Si and Ca/Al ratios and their significant increases in a direction from the particle boundary to the
paste demonstrate that the rim was a hybrid phase composed of the original RBP components and the reaction
products.

1. Introduction grinding power consumption of WCBs is very low, approximately


one-twentieth of that of cement clinker [13]; and (iii) there is a clear
Using construction and demolition waste (CDW) to produce con­ need for RBP as access to traditional SCMs becomes unsustainable with
struction materials has become an attractive way to reduce the amount increasingly stringent carbon emission regulations [14].
of CDW and the carbon footprint of the construction industry [1]. In recent years, many researchers have conducted studies on the
Recycling waste clay bricks (WCBs) has been a key component in this feasibility of RBP, covering almost all the macroscopic properties of
field [2]. As early as the 1990s, the reuse of WCBs has received attention RBP-containing mortar and concrete [15–17]. These findings show that
due to its detectable pozzolanic activity [3–5]. Due to the increasing introduction of less than 25% of RBP has the capacity to enhance the
number of WCBs, recycling has become more urgent. According to the matrix resistance to chloride ions and sulfate attack and gains significant
available statistics [6,7], the amount of WCBs has become enormous; in mechanical properties at a later stage due to RBP’s pozzolanic activity,
China and India, WCBs account for more than 30% of CDW. How to which also accelerates the carbonization rate at a relatively high RBP
promote the reuse of WCBs reasonably and efficiently becomes replacement [18], a common issue with pozzolanic SCMs. Overall, RBP
increasingly important. presents great potential for practical application based on its observable
At present, WCBs recycling is broadly divided into two aspects. One behavior. However, the unsolved critical issues, such as reactivity and its
is coarse and fine aggregates after screening and crushing [8], and the mechanism of RBP, still hinder the further understanding of its
other is to be further destroyed to produce supplementary cementing application.
materials (SCMs) [9]. The former is mainly used in low-strength and RBP is similar to calcined clay in raw materials but exhibits differ­
light-weight concrete due to its high porosity and low strength [10,11], ential activities due to the calcination temperature. Calcined clays are
while the latter is more widely used. When WCBs are prepared as SCMs, typically burned ranging from 500 to 900 ◦ C to produce more active
or recycled brick powder (RBP), it has several advantages: (i) due to the components, i.e., the amorphous form [19–22]. The most common
reduced particle size, the pozzolanic activity of the material is more fully baking temperatures for clay bricks range between 850 and 1150 ◦ C, and
utilized, and its low-strength defects are further limited [12]; (ii) the they are usually baked near or above 1000 ◦ C to obtain high-performing

* Corresponding author. School of Materials Science and Engineering, Southeast University, Nanjing, 211189, China.
E-mail address: Jmgao@seu.edu.cn (J. Gao).

https://doi.org/10.1016/j.cemconcomp.2023.105042
Received 2 November 2022; Received in revised form 9 March 2023; Accepted 19 March 2023
Available online 24 March 2023
0958-9465/© 2023 Elsevier Ltd. All rights reserved.
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

products. Whereas, at baking temperatures near 1000 ◦ C, most clay Table 2


minerals undergo partial recrystallization, leading to a reduction in Chemical compositions of cement and RBP (wt.%).
amorphous content [23]. As a result, clay bricks exhibit lower activity Cement RBP
than calcined clay but still show a pozzolanic activity [4]. Currentlly,
SiO2 19.5 65.0
there has been numerous studies on the reactivity of RBP, but the ma­ Al2O3 4.5 19.3
jority of studies focused on determining RBP’s activity using the CaO 62.7 1.8
compressive strength index method and the Frattini testing method Fe2O3 3.0 6.5
[24], or the evaluation of RBP activity through the calcium hydroxide MgO 1.6 1.7
Na2O 0.1 1.2
(CH) content in blended cement paste [25], leading to a lack of quan­ K2O 0.8 2.6
titative and qualitative understanding of RBP reactivity. SO3 2.3 0.4
As is widely acknowledged, the reaction mechainisms of traditional Others 4.0 1.3
SCMs such as fly ash (FA), granulated blast furnace slag (GGBS), and LOI 1.5 1.6
silica fume (SF) have been well ascertained. This knowledge has pro­
vided technical and economic guidance for their utilization, successfully
promoting their practical application. However, systematic and in-depth
investigation on the RBP reaction mechanisms, and even the evaluation
of its reaction degree, are lacking, impeding the further development of
RBP as an SCM.
Therefore, this paper aims to investigate the reactivity and mecha­
nism of RBP, specifically focusing on RBP produced by WCBs with
baking temperatures exceeding 1000 ◦ C due to its wide distribution and
generation. The reactivity of SCMs generally refers to their chemical
activity, that is, participating in chemical reactions to form hydration
products and following the solubility-precipitation reaction mechanism
[26]. Thus, the RBP reactivity investigation was divided into two parts:
i) investigating the dissolution behavior of RBP in alkaline solutions
with a pH similar to that of pore solution in concrete; ii) quantifying the
reaction degree of RBP at different ages in hydrated cement, iii) estab­
lishing the RBP reaction mechanism base on above exploration. This will
provide a comprehensive and thorough understand the reactivity of
RBP.

2. Materials and methods

2.1. Materials Fig. 1. Particle size distribution and Blaine specific surface.

Portland cement (PC) P•I 42.5, provided by Fushun Cement Co., Ltd,
China was used in this study. Its physical properties and chemical
composition are presented in Tables 1 and 2, respectively. WCBs were
collected from a construction waste disposal site in Nanjing. These WCBs
were screened, crushed, sieved, and ground to produce RBP. The
chemical composition, mineral composition, and particle size distribu­
tion of RBP are shown in Table 1 and Figs. 1 and 2, respectively. The fine
aggregate used was the standard sand produced by Xaimen ISO Standard
Sand Co., Ltd, with a fineness modulus ranged between 2.3 and 3.0. Tap
water was used for preparing mortar and paste. The sodium hydroxide
used was an analytical reagent (>95 wt% pure) produced by Macklin
Biochemical Co., Ltd. Deionized water, produced by Aladdin Bio-Chem
Technology Co., Ltd, was used for the dissolution test.
During the baking of clay bricks, the minerals (e.g., kaolinite, illite,
and montmorillonite) were gradually sintered, and mullite crystallized
at above 1000 ◦ C. In this work, Mullite and Kyanite were detected in
WCBs after several tests, and a representative result of one sample is
shown in Fig. 2. Therefore, it can be confirmed that the baking tem­
perature of these clay bricks was higher than 970 ◦ C. In addition, the

Table 1
Physical properties of cement and RBP.
Specific surface areaa Density (g/ Flexural Compressive Fig. 2. X-ray diffraction pattern of RBP.
(m2/kg) cm3) strength strength
(MPa) (MPa)

3d 28d 3d 28d

Cement 344.2 3.11 5.5 8.7 28.4 46.1


RBP 427.1 2.72 - - - -
a
: Specific surface area was measured by the Blaine method.

2
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

amorphous content was determined between 15 and 20% using a semi- continuously for 0.5 h. This is to avoid the impact of the test apparatus
quantitative method. probe (silica) on the test results and to dissolve possible precipitation.
Given that the dissolution of the active component in RBP in NaOH
2.2. Sample preparation solution follows the following reactions:
[ ]−
Al2 O3 + 3H2 O + 2OH− →2 Al(OH)4 (2)
2.2.1. Mortar samples
The moulding process for mortar specimens followed the standard [ ]−
SiO2 + H2 O + OH− → SiO(OH)3 (3)
GB/T 17617-2021 (equivalent to EU 196–2005). Each testing group
consisted of three specimens with dimensions of 40 × 40 × 160 mm3. [ ]2−
The mix proportion, specifing the quantity of materials employed in SiO2 + 2OH− → SiO2 (OH)2 (4)
moulding each group of mortar specimens, has been tabulated in The maximum degree of reaction of RBP in the NaOH solution under
Table 3. The maintenance procedure of the specimens was also followed the test conditions was calculated from Eq. (5):
the aforementioned standards.
61
MSi + 102 MAl
RRBP = 29 54
(5)
2.2.2. Paste samples 50
The moulding process of the paste samples complied with the stan­
dard GB/T 1346–2011, and the water-to-binder ratio was 0.5. The Where RRBP is the reaction fraction of RBP in solution (%); MSi is the
maintenance procedure for these sample was the same as that for the concentration of Si in the solution (ppm); MAl is the concentration of Al
mortar samples. Until the specified age, the central part of the samples in the solution (ppm).
was collected and prepared into blocks with a side length of less than 10
mm, which were then impregnated in anhydrous ethanol to stop the 2.4. RBP in cement paste
hydration process. The samples were dried in a vacuum oven at 60 ◦ C for
72 h before microscopic testing. To quantify the reactivity of RBP in cement paste, thermogravimetric
analysis (TGA), selective dissolution, magic angle spinning nuclear
2.2.3. Samples for backscattered electron images analysis magnetic resonance (MAS NMR), and BSE-IA were used.
The preparation of backscattered electron images analysis (BSE-IA)
samples is described below. The dried paste samples were the first cast in 2.4.1. TGA
molds with epoxy resin. After that, the molds containing the samples TGA was performed using a Simultaneous Thermogravimetry-
were placed in a vacuum chamber for 30 min to allow the epoxy resin to Differential Scanning Calorimetry instrument (STA 449 F5 Jupiter).
fully penetrate the pores of the samples to mitigate microstructure The method for semi-quantification of portlandite was adapted from the
damage during the polishing. After 24 h, the demolded hardened sam­ method proposed by Taehwan et al. [27]. The detailed calculation
ples were removed for sanding and polishing. Five types of sandpaper process is shown in Fig. 3, and the formula for calculating CH content is
(P400, P800, P1200, P2500, and P7000) were used for the sequential Eq (6):
sanding of the samples. The speed of the polishing machine was 1000 r/ ( )
min. The samples were polished with silk flannel containing 0.25 μm PCH =
WLCH 74.1 WLCC 74.1
× + × × 100% (6)
diamond suspension. The grinding and polishing times were determined W50 18 W50 44
by observing the sample surface with a crystallographic microscope.
Anhydrous ethanol was selected as a coolant in the grinding and pol­ Where PCH is the content of hydration products (CH) in the sample,
ishing process. WLCH is the weight difference corresponding to the CH peak determined
by the tangential method, WLCC is the weight difference corresponding
2.3. RBP in NaOH solution to the Calcium carbonate (CC) peak determined by the tangential
method, W50 is the mass of sample at 50 ◦ C.
RBP was dissolved in 0.1 mol/L NaOH solution (pH = 13), roughly
equivalent to the pore solution in concrete. A solid-liquid ratio of 1:200 2.4.2. Selective dissolution
was used to ensure adequate dissolution of RBP. Polypropylene conical The selective dissolution method was widely used to determine the
flasks with alkali resistance and deionized water were used to ensure the degree of reaction of SCMs due to their time-saving and convenience.
accuracy of the test results. 20 ◦ C, 60 ◦ C, and 80 ◦ C were selected to
investigate the effect of temperature on RBP dissolution. The release rate
of the ions (r in mol/m2/s) was denoted as Eq (1),
C×V
r= (1)
m×S×x×t

Where C is solubility of the (element in the solution (mol/L); m is mass of


the sample (g); S is initial specific surface area of the sample (m2/g); V is
Volume of the solution (L);x is atomic fraction of the element in RBP
(calculated from the chemical composition of RBP).
At each age (0.5, 1, 2, 4, and 8 h and 3, 7, 28, and 90 days), an in­
dividual testing sample (200 mL NaOH + 1 g RBP) was acidified using
HCl, the pH of the solution was adjusted between 4 and 5, and stirred

Table 3
Mixing proportions of mortar samples (g).
Sample codes W/B Cement RBP Water Sand

MC 0.5 450 225 1350


Fig. 3. Calculation example of CH in a cement paste (90-day cement paste

MB 0.5 315 135 225 1350
containing RBP).

3
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

The principle is that the hydration products and unhydrated cement


were chemically dissolved in preference, leaving unreacted SCMs [28].
However, many researchers found errors in the results obtained by se­
lective dissolution, mainly due to the insufficient dissolution of cement
clinker (alite) and CH in solvent [29]. To obtain reliable results, the
selective dissolution of paste samples containing RBP was carried out
according to the method proposed by Dong et al. [30], and the calcu­
lation formula was Eq. (7) [31].
Wur
− FC Fuc
DoRRBP = 1 − 1− Wn
× 100 (7)
FRBP Fur

Where
WL − Lc
wn = (8)
1 − Lc

m0 − m950
wL = (9)
m0

Lc = (1 − β)Lp + βLb (10)

Wur is the mass fraction of the undissolved residue in the blended cement
paste at each age, Wn is the non-evaporated water content in hardened
paste, FC is the mass fraction of cement in blende cement, Fuc is mass
fraction of undissolved cement residue, Fur is mass fraction of undis­
solved RBP residue, FRBP is the mass fraction of RBP in the blended
cement, Lc is LOI of unhydrated blended cement containing RBP, Lp is
LOI of cement, Lb is LOI of RBP, β is the mass fraction of RBP in blended
cement, wL is LOI of blended cement paste containing RBP, m0 is the
mass of the blended cement paste before calcination. m950 is the mass of
the blended cement paste after being fired at 950 ◦ C. Fig. 4. Extraction process for RBP particles.

2.4.3. MAS NMR


RBP after time t.
MAS NMR was used to investigate the degree of reaction of RBP in
cement pastes, and its corresponding mode was deconvolution trans­
3. Results and discussion
formed [32,33]. The resonance frequency and spin rate of the 29Si NMR
instrument (Bruker 400 M) were set to 79.48 MHz and 10 kHz, respec­
3.1. Dissolution of RBP in NaOH solution
tively. The pulse length and relaxation delay used were 4.97 μs and 6.5
μs, respectively. Deconvolution of the spectrums was performed in Peak
3.1.1. Species dissolved in RBP
Fit using a Gaussian model. Considering the inevitable peak drift in 29Si
The results of the dissolution of RBP in NaOH solutions at different
or 27Al spectra, each deconvolution calculation aimed at obtaining the
temperatures (20 ◦ C, 40 ◦ C, and 80 ◦ C) for 4 h are presented in Fig. 5.
best fit with an equal number of peaks (software calculations determined
The results showed that the dominant dissolved element species of RBP
peak locations).
observed in NaOH solutions at 20 ◦ C, 60 ◦ C, and 80 ◦ C were Si, Al, and
2.4.4. BSE-IA
BSE-IA allows the identification of individual phases in a blended
cement paste according to their brightness, which depends on their
average atomic number [34]. This method has been widely used to
quantitatively evaluate the degree of reaction of SCMs in recent years
[35]. However, the RBP particles in BSE images exhibited similar grey
values to those of the C-(A)-S-H gels, leading to unreliable statistics of
the degree of reaction of RBP. Here, a method for marking RBP particles
in BSE images, referring to the elemental labeling method proposed in
Ref. [36], was developed to calculate the degree of reaction of RBP and
is shown in Fig. 4. The 1000× magnification of BSE images were used for
the labeling of RBP particles, and each pixel in the image corresponds to
approximately 0.27 μm in the real No. 15 BSE images were used for the
results statistics, corresponding to 90 elemental images. Image J was
used for binarization and statistics of the output images, and the
calculation formula is shown in Eq. (11).
Afanhydrous RBP (t = 0) − Afanhydrous RBP (t)
RRBP = (11)
Afanhydrous RBP (t = 0)

Where RRBP is the area fraction of RBP (equivalent to the volume frac­
tion), Afanhydrous RBP (t = 0) is remaining area fraction of the initial Fig. 5. Results of RBP dissolution in NaOH solution at different temperatures
unreacted RBP, Afanhydrous RBP (t) is remaining area fraction of unreacted for 4 h.

4
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

Ca, followed by Fe, K, and Mg after 4 h of dissolution. It is important to


mention that the dissolution of Na was not discussed due to the use of a
NaOH solution. The dissolved Si and Al were primarily derived from the
dissolution of aluminosilicate glass in RBP, as confirmed by subsequent
MAS NMR tests. Additionally, the dissolution of Ca was also noteworthy.
When the dissolution temperature was 20 ◦ C, the dissolution amount of
Ca was greater than that of Al but slightly lower than that of Si. When the
temperature was further increased to 80 ◦ C, the amount of Ca dissolved
only slightly increased. This indicates that Ca in RBP had an extremely
high dissolution rate during the initial period, but it decreased rapidly.
The above phenomenon has been attributed to the presence of "weak"
sites in the polymerized network [37]. The non-bridging oxygen com­
bined with cations (e.g., Ca2+) in the form of ionic bonds was the weak
link in the dissolution process. This causes the eraly release of Ca into the
solution. As a result, an early rapid release of Ca was observed.

3.1.2. Effect of dissolution time


Section 3.1 indicates that the mian dissolved elemental species in
RBP were Si, Al, and Ca. The amount and rate of dissolution of these
elements are therefore investigated in this section. Fig. 6 illustrates the
Fig. 7. Logarithm of the release rate of Si, Al, and Ca in RBP NaOH solution.
release amounts of Si, Al, and Ca in the RBP NaOH solution at different
times. During the initial period (<1 h), the release amounts of Ca were
higher than those of Si and Al. The release amounts of Ca in the solution observation time. However, these release rates displayed a remarkable
were still higher than those of Al until 3 h but increased slowly in the phase variation, decreasing rapidly before 7 days, but the trend was no
following time. Concerning Si and Al, their release amounts in solution longer significant for the rest of the period. Generally, the dissolution
exhibited an obvious increasing trend with time. However, Si release stage with a rapid change of the release rate was called the " non-steady
was significantly higher than Al release, which is similar to conventional state" dissolution stage, and the subsequent dissolution stage with a
SCMs [38,39]. relatively stable release rate was called the "steady state" dissolution
The release rates of Si, Al, and Ca in RBP NaOH solution were stage. Under similar experimental conditions, the duration of the non-
calculated using equation (1), as shown in Fig. 7. During the entire steady stage of Si and Al dissolution for GGBS and FA is very short,
testing period, the release rate of Ca was the highest, followed by Al and generally ending within a few hours [42]. The longer duration of the
Si, indicating that the dissolution ratio of Ca in RBP relative to the non-steady stage indicates a longer time required for the activity
original Ca content was significantly higher than that of Si and Al, which development of RBP compared with GGBS and FA. This explains the
further verified the preferential destruction of Ca in the glass network. continuous and slow strength growth of the RBP-containing cementi­
The preferential release of Al compared to Si can be explained by the tious material in the later stage from a dissolution perspective [43].
electron density polarization in the network covalent bond. Si4+ has a
charge/radius ratio more than twice that of Al3+ (Al3+: 0.39 Å and Si4+: 3.1.3. Effect of dissolution temperature
0.26 Å), and thus the electron density polarization of neighboring oxy­ Figs. 8 and 9 show the concentration of Si and Al in RBP NaOH so­
gen ions is stronger [40]. Therefore, the Al–O site in a Si–O–Al linkage is lution at different temperatures. As expected, the elevated temperature
more easily destroyed, leading to a preferential release of Al [41]. significantly increased the concentration of Si and Al. The concentration
Notably, a valuable characteristic was observed in the dissolution of of Si increased from approximately 1 mmol/L to 4 mmol/L, and the
RBP. It can be noticed that under the experimental conditions, the concentration of Al increased from approximately 0.5 mmol/L to 2
release rate of Ca, Al, and Si decreased continuously throughout the mmol/L in 8 h when the temperature was raised from 20 ◦ C to 80 ◦ C.
Also, it is worth noting that higher temperatures resulted in higher
amounts of dissolved Si and Al. When the temperature was elevated
from 60 ◦ C to 80 ◦ C, the concentration of Si and Al increased signifi­
cantly, from approximately 4 mmol/L to 7.4 mmol/L and from 2.4
mmol/L to 3.44 mmol/L, respectively. This suggests that it is feasible to
increase RBP reactivity using high-temperature curing. Additionally, the
calculated results of Si and Al release rates show that the elevated
temperature significantly increased the Si and Al release rates and
greatly reduced the duration of the non-steady phase of Si and Al
dissolution. The dissolution of the glass in alkali solutions is governed by
the surface chemistry controlled by the surface product layer and liquid
film. Increasing temperature can increase the diffusion coefficient,
thereby promoting the release of Si and Al [44]. Furthermore, the in­
crease in temperature can also increase the activity of OH− and reduce
Al(OH)-4 adsorption, thereby increasing the dissolution of Si [45].

3.1.4. Dissolution kinetics


The dissolution of RBP in alkali solutions is a typically heterogeneous
non-catalytic solid-liquid reaction. Clarifying the dissolution behavior of
RBP facilitates the understanding of its reaction mechanism. The
shrinking core model (SCM) has been widely employed to investigate
the reaction behaviour of solid reactants in solution. It is described by
three mechanisms: (1) The dissolution behavior of particles is controlled
Fig. 6. Concentration of Si, Al, and Ca in RBP NaOH solution.

5
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

Fig. 8. Concentration of Si and Al in RBP NaOH solution at different temperatures.

Fig. 9. Release rates of Si and Al in RBP NaOH solution at different temperatures.

by the diffusion of liquid film on the surface, which is expressed in Eq. chemical reaction rate constant; CAg is the concentration of A in the
(12); (2) The dissolution of particles is controlled by the chemical re­ filtrate; n is the order of reaction concerning A; De is diffusion coefficient
action on the surface of reactants, which is expressed by Eq. (13); (3) The through the product layer.
dissolution of particles is controlled by phase boundary reactions and Fig. 10 illustrates the fitting results of dissolved Si and Al in RBP
product layer diffusion, which was expressed by Eq. (14). NaOH solution at 20 ◦ C using an SCM, and the linear behavior was
employed to evaluate the fitting results. It can be found that compared
2bkc CAg
(12) with Eqs. (11) and (12), Eq. (13) shows the best fitting value of 0.982 for
2
1 − (1 − α)3 = kt k=
ρb R2
the experimental data of Si and Al dissolution from RBP in NaOH solu­
bkc CAg tion at 20 ◦ C. Similarly, the dissolution of Si and Al in RBP exhibits the
(13)
1
1 − (1 − α)3 = kt k=
ρb R same trend at 60 ◦ C and 80 ◦ C, as shown in Table 4. However, it should
be noted that this trend was no longer distinct with increasing temper­
2bkDe CAg atures. Therefore, it can be confirmed that the dissolution of Si and Al
(14)
2
1 − 3(1 − α)3 + 2(1 − α) = kt k= from RBP in NaOH solution was controlled by the phase boundary re­
ρb R2
actions and the product layer diffusion.
Where α is the fraction of B converted to a product in the solid particles To further understand the dissolution kinetics of Si and Al in RBP
at time t; ρb is the density of solid particles; r is the initial radius of solid NaOH solution, ln (k) versus 1/T was plotted based on Arrhenius Eq.
particles, b is moles of B consumed per mole of reacted A; kc is the (15), and the activation energy was calculated as shown in Fig. 10.

Fig. 10. Plots of 1-(1- α)2/3, 1-(1- α)1/3, and 1–3(1-α)2/3-2(1-α) versus time for the dissolved Si and Al in RBP NaOH solution.

6
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

Table 4 SCMs replacement of cement (Rc ). The correlation between SAI and Rc is
Fitting results of dissolved Si and Al in RBP NaOH solutions at a different tem­ generally satisfied: SAI ≥ 100% − Rc at given ages (7 or 28 days).
perature from the SCM model. Furthermore, the development of the pozzolanic activity of SCMs is also
Element Temperature (◦ C) R2 considered in the individual criteria and thus supplemented with SAI ≥
Eq. (11) Eq. (12) Eq. (13)
100% − Rc + Δ, Δ > 0 at a longer age (90 days). The water-binder and
binder-sand ratios of mortar specimens are commonly approximately
Si 20 0.880 0.886 0.988
1:2 and 1:3, respectively.
60 0.911 0.913 0.993
80 0.962 0.979 0.981 Figs. 12 and 13 show the compressive strength of mortar specimens
Al 20 0.857 0.869 0.982 and SAI of RBP. The investigated RBP exhibited an acceptable SAI value
60 0.882 0.885 0.962 at 28 days, slightly greater than 70%. Notably, the SAI of the RBP
80 0.957 0.961 0.988 increased steadily with an increasing curing time, indicating the
development of RBP activity at later ages. This is supported by findings
in sect. 3.2 that the dissolution of active Si and Al in RBP exhibited a
(15)
Ea
k = Ae−
significantly longer non-steady stage and continuous low dissolution
RT

rates for a longer time. Therefore, RBP-contained mortars exhibited an


Where k is the kinetics/rate constant; A is the frequency factor; Ea is the
observable development of compressive strength. The SAI line graph
activation energy; R is the universal gas constant; T is the thermody­
reflects this trend more clearly. Overall, the investigated RBP showed
namic temperature.
promising activity: SAI exceeded 70% at 28 days and continued to grow
Fig. 11 shows that the activation energies Ea of Si and Al dissolved in
after that. In addition, the longer-term SAI was worth considering for
RBP NaOH solution (pH = 13) were 82.74 kJ/mol and 75.79 kJ/mol,
evaluating RBP activity.
respectively. According to the definition of Ea , Ea can reflect the ease of
dissolution of Si and Al in RBP. However, Ea is calculated in k, which is
3.2.2. CH content
different for the same material in solutions of different solutes and pH
Investigating the variations of CH content in the matrix is a common
[44]. It is difficult to compare Ea of different pozzolanic materials, but
method to study the reactivity of SCMs with pozzolanic activity [50]. In
some evaluation and discussion can still be conducted in light of relevant
the case of CH saturation, the silicate or aluminosilicate network formers
studies. In this study, the Ea of Si and Al during the dissolution process
were depolymerized and dissolved into the solution, and the depoly­
(75.79–82.74 kJ/mol) is higher than that of the common pozzolanic
merized silicates and aluminates entered the solution following Eqs. (16)
material reaction (approximately between 30 and 50 kJ/mol) [46–48],
and (17):
and similar to that of ferronickel slag (49.45–80.47 kJ/mol) [49]. The
[ ]−
reason for the slow reaction rate of RBP is understood from the ≡ Si − O − Si ≡ + 3OH− → SiO(OH)3 (16)
perspective of dissolution kinetics. The lower Ea of Al dissolution than
[ ]− [ ]− )
that of Si dissolution indicates that Al in RBP had a faster dissolution rate ≡ Si − O − Al ≡ + 7OH− → SiO(OH)3 + Al(OH)4 (17)
under alkaline conditions and also demonstrates that the solubility of Si
is highly dependent on temperature. When Ca2+ ions are in contact with these silicate and aluminate species,
C–S–H and calcium aluminate hydrate C4AH13 are formed following Eqs.
3.2. Reactivity of RBP in cement paste (18) and (19), while AFt and AFm are formed when sulfate ions are
present following Eqs (20) and (21) [51]:
3.2.1. Strength activity index [ ]−
Y SiO(OH)3 + XCa2+ + (Z − X − Y)H2 O + (2X − Y)OH− → CX − SY − HZ
The strength activity index (SAI) is one of the most fundamental
(18)
methods for evaluating the reactivity of SCMs and has been widely used
worldwide. There are national and regional standards for SAI, such as [ ]−
2 Al(OH)4 + 4Ca2+ + 6H2 O + 6OH− →C4 AH13 (19)
GB/T 1596–2017, BS-EN 450-1-2012, and ASTM C311-2013. In the
standards mentioned above, the SAI is established based on the rate of

Fig. 11. Arrhenius plots of dissolution of Si and Al from RBP in NaOH solution
(pH = 13). Fig. 12. Compressive strength of mortar specimens.

7
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

growth rate of CH content in RBP-blended paste, the pozzolanic reaction


between RBP and CH produced by cement hydration probably occurred
at an early age (after 3 days), but this reaction was weak because of the
low amorphous content in RBP and the inhomogeneous mineral phase
composition that prevented the dissolution of active ingredients. Until
the CH in the system gradually slowed down, the consumption of CH by
RBP was only detectable when the active components of RBP were
continuously dissolved.

3.2.3. Selective dissolution and BSE-IA-Mapping


Quantitative evaluation of the degree of reaction of RBP in cement
paste is essential to investigate its reactivity. In this section, selective
dissolution and BSE-IA-Mapping were employed to evaluate the degree
of reaction of RBP at different ages. In addition, the RBP dissolution data
from the previous section are also used for the discussion in this section.
Fig. 15 shows the degree of reaction of RBP determined using se­
lective dissolution, BSE-IA-Mapping, and RBP dissolution in 0.1 mol/L
NaOH.The three methods gave similar trends in the degree of reaction of
RBP that the reaction degree increases with time. As expected, these
methods provide slightly different results. The RBP reactivity of NaOH
Fig. 13. SAI of RBP. leaching at the same age was the highest, reaching 12% at 90 days.
However, it should be added that the actual degree of reaction was
[ ]−
2 Al(OH)4 + 3SO2−4 + 6Ca2+ + 4OH− + 26H2 O + →C4 A greater than this value since other dissolved elemental species, such as
Ca, Na, and K, were not included in the statistics. The reaction degree of
• 3CaSO4 •32H2 O (20)
RBP determined using selective dissolution was slightly higher than that
[ ]− of BSE-IA-Mapping. The 180-day degree of reaction of RBP determined
2 Al(OH)4 + 6Ca2+ + • 3CaO • Al2 O3
by the two methods was 15.1% and 13.2%, respectively.
• 3CaSO4 •32H2 O + 10OH− → 3[3CaO • Al2 O3 • CaSO4 •12H2 O] + 5H2 O A higher degree of reaction can be expected using the NaOH disso­
(21) lution. The alkalinity of the NaOH solution (pH = 13) used in the
experiment was close to that of the pore solution in concrete, but the pH
Thus, the pozzolanic activity can be investigated according to the
of the pore solution can reach 13.8 with the hydration process [52]. In
variation of the CH content in the system. Fig. 14 presents the variation
theory, the polymeric network of silicate and aluminate tetrahedra in
of CH content in pastes containing RBP and QP at different ages. On the
RBP will be attacked by higher amounts of OH− in a higher alkalinity
first day, the CH content of cement paste mixed with RBP was signifi­
environment, accelerating the rate of hydrolysis and thus leading to a
cantly lower than that of pure cement paste due to the dilution of cement
higher degree of reaction. However, in the pore solution, the dissolution
clinker. After curing for 3 days, the difference in CH content between the
of Si and Al in the polymeric network will be influenced by the solution
paste containing RBP and pure cement was more obvious, and the CH
components. When Ca and Al in the solution reach a specific concen­
content increased slightly compared with that of the pure paste. The
tration, they will be enriched on the surface of SCMs particles and form a
difference in CH content between RBP-containing paste and pure cement
Ca-modified aluminate layer. This phenomenon has only been
paste became more and more significant with age because the CH con­
confirmed in the dissolution experiment of synthetic glass [35]. Current
tent of RBP-containing pastes gradually decreased after 28 days of
studies on inhibiting SCMs dissolution by Ca and Al seem to support the
curing, indicating the occurrence of the pozzolanic reaction. This is
reaction site adsorption theory [38], but the presence of Ca and Al un­
consistent with the findings in Sect. 3.1.2, verifying that the develop­
doubtedly hinders the dissolution of Si and Al in SCMs.
ment of the pozzolanic activity of RBP required a long time. From the

Fig. 14. CH content in blend paste and cement paste samples at different ages. Fig. 15. Degree of reaction of RBP determined by different methods.

8
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

The resulting higher degree of reaction using selective dissolution


compared with BSE-IA-Mapping is due to an inherent error. Even
considering the loss of cement residue and RBP during dissolution, mass
loss in manual operation was unavoidable, leading to an overestimation
of the degree of reaction of RBP. It is worth noting that the results
measured by these two methods varied slightly, with a maximum dif­
ference of less than 2%, which might be attributed to the relatively low
reactivity of RBP. Overall, although the three methods gave differing
degrees of reaction of RBP, valuable reactivity could be identified: the
degree of reaction of RBP in cement paste was not high but showed long-
lasting reactivity.

3.2.4. MAS NMR


To further investigate the RBP reactivity, MAS NMR was employed.
Fig. 16 presents the 29Si MAS NMR spectrum of initial RBP and RBP Fig. 17. Deconvolution results of 29
Si MAS NMR spectra of RBP based on area
residues at different ages and their deconvolution analysis results. The percentage (%).
initial RBP and RBP residues showed a broad and overlapping peak
between − 80 and − 125 ppm, indicating the inhomogeneous presenta­ and Q4 (2Al) alter slightly. Clearly, Q4 (3Al) and Q4 (4Al) show a sig­
tion of different Q4 structures. The peaks at − 87, − 91, − 94, − 101, nificant decrease with time. The significant decrease in the content of Q4
− 107, − 112, and − 117 were assigned to Q4 (mAl, m = 0–4). Glass (3Al) and Q4 (4Al) undoubtedly indicates that Si in Q4 (3Al) and Q4 (4Al)
components and crystalline phases such as quartz and mullite contribute is more actively involved in the reaction of RBP in cement paste. In
to the formation of these sites [53,54]. contrast, the dramatic increase in the content of Q4 (0Al) shows the low
Additionally, a clear downward trend at approximately − 100 ppm to reactivity of Si in this site. Ref. [33] demonstrates that Si of Q4 (0Al) sites
− 80 ppm was observed in the 29Si spectra of samples with age, indi­ in FAis almost non-reactive with less than 0.4%, and the findings of
cating an ongoing reaction for RBP in cement paste. This spectrum Ref. [32] also support this conclusion. Based on this conclusion, Eq. (22)
segment corresponds to the peaks at the Q4 (3Al) and Q4 (4Al) sites. The is used to calculate the reactive Si in RBP in cement paste at different
significant reduction in the intensity of these peaks with time suggests ages, and the corresponding calculation results are shown in Fig. 16.
that the presence of Al–O sites contributed to the reaction of RBP. This is
attributed to the fact that Al–O sites are more easily destroyed than the SiReacted = 1 −
Q4 (0Al)Initial
(22)
Si–O sites in alkaline solutions, and thus more Al–O sites allow for a Q4 (0Al)T
higher degree of dissolution of the glassy structure. This observation is
Where SiReacted is the fraction of reacted Si in RBP (wt%); Q4 (0Al)Initial
consistent with the finding in Sect. 3.2 that Al in RBP is released more
is the fraction of Q4 (0Al) in the initial RBP from the convolution
rapidly than Si, demonstrating that the dissolution of RBP in the hard­
calculation of the 29Si spectrum; Q4 (0Al)T is the fraction of Q4 (0Al) in
ening cement is similar to that in the simulated solution (NaOH) tested.
RBP in cement paste at the time of T from the convolution calculation of
In addition, this finding that peaks in the Q4 (3Al) and Q4 (4Al) sites
the 29Si spectrum.
show a dramatical reduction in intensity at 180 days more directly
Fig. 18 shows that the amount of reactive Si from RBP in cement
confirms that the C-(A) -S-H gel is the main reaction product of RBP in
paste calculated based on the 29Si spectrum has a similar increasing
hardened cement.
trend to that in NaOH solution. The results of 29Si MSS NMR spectra and
More detailed results of the deconvolution calculations are shown in
dissolution experiments showed that the Si reactivity reached about
Fig. 17. Three different trends are observed for alterations in the content
24% and 12% at 180 and 90 days, respectively, but the former showed a
of different structures of Q4 with time. Q4 (0Al) exhibits a significant
slightly higher level at the same age. The dissolution of RBP in NaOH
increase in content with increasing age, while the content of Q4 (1Al)
solution (pH = 13) provides a nearly ideal condition for the dissolution

Fig. 18. Reacted Si in RBP subjected to cement paste and NaOH solution (pH
Fig. 16. NMR Spectra of RBP and RBP residues at different ages. = 13).

9
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

of reactive Si, so the level of reactive Si obtained under this condition RBP particle, the 180-day reaction rims were further analyzed using
can be used as the approximate upper limit of RBP reactive Si in cement point element analysis, as shown in Fig. 21. A significant directional
paste. The level of reacted Si obtained from the deconvolution calcula­ trend of increase or decrease in the content of elements can be observed
tions of 29Si Mass NMR spectra is inevitably affected by some variables, in the reaction rim. The Si content shows a continuous decrease along
for example, the selection of site locations and used fitting function, and the RBP particle boundary to the paste, indicating that there may be a Si
the conditional qualification about the function. Moreover, the results release process. However, Ca and Al present the opposite distribution
provided by this method are coupled with slightly significant errors. characteristics to Si, demonstrating that the dissolution and reaction of
However, it can be confirmed that RBP exhibited a continuous reaction the active elements in the RBP particles are inhomogeneous. These
both in the NaOH solution and in the cement paste according to the speculations are further verified by the variation in the distribution of
experimental results; the reactivity of RBP is further clarified. the Ca/Si ratio in this region: The change of Ca/Si ratio distribution in
Fig. 19 shows the 27Al MAS spectra of initial RBP and RBP resides at this region further confirmed the above speculation: with the increase of
different ages. It is generally accepted that the 27Al MAS NMR is more the Ca/Al ratio, the RBP particle boundary extended vertically outward,
suitable for qualitative evaluation than quantitative calculations owing but the value was very small (less than 0.3), which was much lower than
to the wide peak locations, asymmetrical natural and complicated the known C–S–H gel interface, indicating that the reaction edge might
linking environments [55]. The deconvolution of Al spectra was, be a hybrid phase. It contains the original RBP particle component (i.e.,
therefore, not conducted in this study. Three distinct peaks were iden­ quartz phase) and hydration products. In addition, another special zone
tified in the spectrum of initial RBP at approximately 2, 13 and 55 ppm. (Zone 2) was found, which showed clear signs of erosion and reaction.
According to available studies, peaks at 2 and 55 ppm can be confirmed Data from the point element analysis shows that the Ca/Si ratio in this
to be associated with octahedral and tetrahedral Al, respectively, but the region was higher than that in phases from the RBP particle, supporting
Al structure represented by the peak at 12 ppm is not well defined. the idea that the rim region was a hybrid phase.
Typically, the octahedral and pentahedral Al is assigned to around In addition to particles rich in the quartz phase, another type of RBP
30–40 ppm and − 10 to 20 ppm, respectively. Thus, the peak at 13 ppm is particle is more abundant, namely the silica-aluminum phase. Similar
suspected to be tetrahedral Al with a differential structure. The RBP regions were also found around such particles, as shown in Fig. 22.
residues exhibited a peak distribution similar to the initial RBP but Although not obvious in BSE images, the elemental distribution map
decreased peak intensity of octahedral Al with age. The reduced peak clearly shows the existence of the reaction rim. The reaction rims around
intensity, confirmed by comparison with the intensity of tetrahedral Al these two types of particles exhibit similar Ca/Si values and distribution,
peak, indicates that the Al from the vitreous Al phase and the crystalline but the region located around the silica-aluminum phase particles show
mullite in RBP took part in the reaction in blend cement paste. This is an extremely high Ca/Al ratio due to the higher Al content. Comparing
also consistent with the results obtained from 29Si spectrum that Q4(3Al) the Al contents of the different localities, these Al probably belong to or
and Q4(4Al) show a significantly decreased content with increasing age, are derived from silica-aluminum phase particles. Furthermore, it is
especially after 28 days, further revealing the reactivity of RBP. noteworthy that the Na content in this region shows a clear decreasing
extending particle interface in a vertical outward direction, suggesting
3.2.5. Energy-dispersive spectroscopy (EDS) that silica-aluminum phase RBP particles abundant of Na (possibly
The combination of BSE and EDS can provide detailed information feldspar particles) can dissolve Na for reaction to produce possible
on the distribution of elements and thus provide an in-depth investiga­ C–N–S–H gels.
tion of the reactivity of RBP. Regions of the cement pastes containing the The observed reaction rim can be understood as a modified layer
RBP particles were first observed in the BSE images to obtain detailed formed by an interface-coupled dissolution-reprecipitation reaction. The
information on the reaction of RBP. The greyscale (brightness) of phases free energy difference required to drive this reaction is provided by the
in BSE pictures correlates with their atomic number, and higher difference in solubility between the parent and product phases in the
brightness indicates a higher atomic number for the phase, providing a solution. Both reactions occur at the particle interface; only a small
preliminary evaluation of the phase composition and distribution. amount of material needs to be dissolved in the boundary layer at the
Through numerous observations, it was found that “reaction rims” interface at any one time. Another important property supporting the
appeared around the RBP particles at 28 days and became more distinct continuation of the reaction is the high porosity of silica, a common
with increasing time. Typical reaction rims are shown in Fig. 20 (marked feature of silica or "gel" layers, which enables mass transfer through the
by green arrows). Similar reaction rims are frequently found in FA or corrosion zone and from the reaction interface at a wide range of pH
slag-blended cement [56,57]. Obviously, the reaction rims around RBP conditions [58,59]. Meanwhile, the porous nature of the RBP particle
particles were much less significant. This is attributed to the lower itself further enhances this possibility. Therefore, since the original RBP
reactivity of RBP than FA and GGBS, and this finding also more directly particle is not in equilibrium with the interface solution, the reaction can
confirms the existence of RBP activity. proceed through the difference in solubility between the original RBP
To obtain more detailed information on the reaction rim around the particle and the reaction product, even if the surrounding solution is

27
Fig. 19. Al NMR Spectra of RBP and RBP residues at different ages.

10
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

Fig. 20. Representative regions of cement pastes containing RBP particles at different ages.

Fig. 21. BSE image and results of elemental distribution analysis of the reaction rim at 180 days.

filled with the associated ions. However, as the thickness of the reaction the formation of the reaction rim, a schematic representation of the
rim increases, the reaction is increasingly controlled by transport of reaction is depicted in Fig. 23.
material through the pore of the reaction rim. As a result, the reaction
rims exhibit a differential Ca distribution. The difference in the distri­ 4. Conclusions
bution of Si and Al is probably due to the formation of precipitation
(products) that reduce the concentration of Si and Al in the boundary In this work, the reactivity of RBP, derived from WCBs with an
layer adjacent to its original particle constituents. To better understand original baking temperature above 1000 ◦ C, was investigated in

11
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

Fig. 22. BSE image and results of elemental distribution analysis of the reaction rim at 180 days.

• The primary reactive species in RBP, Si and Al, exhibited obvious


lasting dissolution behavior with a significant “non-steady state”
dissolution stage, indicating that the RBP reaction proceeded slowly.
This finding was verified by the dissolution mechanism controlled by
the phase boundary reaction, with activation energies of 82.74 kJ/
mol and 75.79 kJ/mol for Si and Al dissolution, respectively.
Furthermore, an increase in temperature was found to accelerate the
dissolution of RBP substantially.
• SAI, TGA, selective dissolution, BSE-IA-Mapping, and MAS NMR all
confirmed the reactivity of RBP. Regarding the SAI, it is worth
considering the complement of a 90-day observation indicator for a
more reliable assessment. Selective dissolution and BSE-IA-Mapping
Fig. 23. Schematic diagram of formation of the reaction rim. provided a comparable evaluation of the degree of reaction for RBP
in cement pastes, which was approximately 15% at 180 days. NMR
combination with its dissolution kinetics in NaOH solutions and degree tests identified reactive Si and Al in RBP primarily from the Q4
of reaction in the cement matrix. The main findings are summarized (3,4Al) and Al (IV) sites, respectively.
below.

12
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

• Observations of reaction rims surrounding RBP particles in the [9] Q. Tang, Z. Ma, H. Wu, W. Wang, The utilization of eco-friendly recycled powder
from concrete and brick waste in new concrete: a critical review, Cem. Concr.
cement paste suggest that the reaction rim is a hybrid phase
Compos. 114 (2020), 103807, https://doi.org/10.1016/j.
composed of both original RBP components and newly-formed cemconcomp.2020.103807.
products. The significant increase in Ca/Si, Ca/Al, and Ca/Na ra­ [10] Y. Zhao, J. Gao, F. Chen, C. Liu, X. Chen, Utilization of waste clay bricks as coarse
tios from the particle boundary towards the cement paste further and fine aggregates for the preparation of lightweight aggregate concrete, J. Clean.
Prod. 201 (2018) 706–715, https://doi.org/10.1016/j.jclepro.2018.08.103.
supports this finding. [11] C. Zheng, C. Lou, G. Du, X. Li, Z. Liu, L. Li, Results Mechanical properties of
recycled concrete with demolished waste concrete aggregate and clay brick
The results in this work indicate that RBP derived from WCBs with an aggregate, Results Phys. 9 (2018) 1317–1322, https://doi.org/10.1016/j.
rinp.2018.04.061.
original baking temperature above 1000 ◦ C has low reactivity and ex­ [12] E. Navrátilová, P. Rovnaníková, Pozzolanic properties of brick powders and their
hibits a reaction mechanism that significantly differs from conventional effect on the properties of modified lime mortars, Construct. Build. Mater. 120
SCMs, such as FA and GGBS. However, RBP obtained from WCBs with an (2016) 530–539, https://doi.org/10.1016/j.conbuildmat.2016.05.062.
[13] Y. Chen, S. Yang, R. Xu, Experimental study on making mineral admixture by waste
original baking temperature below 1000 ◦ C is expected to exhibit clay brick, Brick-Tile (12) (2017) 3, https://doi.org/10.16001/j.cnki.1001-
improved activity and a more intriguing reaction mechanism. Another 6945.2017.12.005.
important aspect to note is the congruent or incongruent dissolution of [14] J.M. Paris, J.G. Roessler, C.C. Ferraro, H.D. DeFord, T.G. Townsend, A review of
waste products utilized as supplements to Portland cement in concrete, J. Clean.
RBP, which can contribute to a deeper understanding of the reaction Prod. 121 (2016) 1–18, https://doi.org/10.1016/j.jclepro.2016.02.013.
mechanisms. Further investigation of these aspects is therefore war­ [15] T. Ye, J. Xiao, Z. Duan, S. Li, Geopolymers made of recycled brick and concrete
ranted. It is believed that the pursuit of these endeavors can yield a more powder – a critical review, Construct. Build. Mater. 330 (2022), 127232, https://
doi.org/10.1016/j.conbuildmat.2022.127232.
thorough understanding of the reaction mechanism of RBP, thereby
[16] A. Alsaif, Utilization of ceramic waste as partially cement substitute – a review,
facilitating its diverse and tiered utilization, ultimately resulting in Construct. Build. Mater. 300 (2021), 124009, https://doi.org/10.1016/j.
improved utilization efficiency of RBP. conbuildmat.2021.124009.
[17] Z. He, A. Shen, H. Wu, W. Wang, L. Wang, C. Yao, J. Wu, Research progress on
recycled clay brick waste as an alternative to cement for sustainable construction
CRediT authorship contribution statement materials, Construct. Build. Mater. 274 (2021), 122113, https://doi.org/10.1016/
j.conbuildmat.2020.122113.
[18] J. Shao, Hydration Performance and Carbonation of Clay Brick Powder-Cement
Shujun Li: Software, Validation, Investigation, Resources, Data Complex Cementitious Material, 2019. https://10.27014/d.cnki.gdnau.2019.002
curation, Writing - original draft, Writing - review & editing. Jianming 304.
Gao: Formal analysis, Conceptualization, Writing - review & editing, [19] K. MacKenzie, I. Brown, R. Meinhold, M. Bowden, Outstanding problems in the
kaolinite-mullite reaction sequence investigated by 29Si and 27Al solid-state
Funding acquisition, Project administration. Gaofeng Chen: Software,
nuclear magnetic resonance: I, metakaolinite, J. Am. Ceram. Soc. 68 (6) (1985)
Data Curation. Yasong Zhao: Investigation, Zhenhai Xu: Investigation. 293–297.
Xu Luo: Supervision. Cheng Liu: Software. [20] D. Massiot, P. Dion, J.F. Alcover, F. Bergaya, 27Al and 29Si MAS NMR study of
kaolinite thermal decomposition by controlled rate thermal analysis, J. Am. Ceram.
Soc. 78 (1995).
[21] C.E. White, J.L. Provis, T. Proffen, D.P. Riley, J.S.J. van Deventer, Density
Declaration of competing interest functional modeling of the local structure of kaolinite subjected to thermal
dehydroxylation, J. Phys. Chem. 114 (14) (2010) 4988–4996, https://doi.org/
The authors declare that they have no known competing financial 10.1021/jp911108d.
[22] R. Fernandez, F. Martirena, K.L. Scrivener, The origin of the pozzolanic activity of
interests or personal relationships that could have appeared to influence calcined clay minerals: a comparison between kaolinite, illite and montmorillonite,
the work reported in this paper. Cement Concr. Res. 41 (1) (2011) 113–122, https://doi.org/10.1016/j.
cemconres.2010.09.013.
[23] M. Kornmann, Clay Bricks and Rooftiles, Manufacturing and Properties, 2007.
Data availability [24] V.F. Rahhal, M.A. Trezza, A. Tironi, C.C. Castellano, M. Pavlikova, J. Pokorny, E.
F. Irassar, O. Jankovsky, Z. Pavlik, Complex characterization and behavior of waste
Data will be made available on request. fired brick powder-portland cement system, Materials 12 (10) (2019), https://doi.
org/10.3390/ma12101650.
[25] H. Wu, C. Liang, C. Wang, Z. Ma, Properties of green mortar blended with waste
Acknowledgments concrete-brick powder at various components, replacement ratios and particle
sizes, Construct. Build. Mater. 342 (2022), 128050, https://doi.org/10.1016/j.
conbuildmat.2022.128050.
The authors deeply appreciate the financial support provided by the [26] B. Lothenbach, K. Scrivener, R.D. Hooton, Supplementary cementitious materials,
National Natural Science Foundation of China (No. 52130210) and the Cement Concr. Res. 41 (12) (2011) 1244–1256, https://doi.org/10.1016/j.
Jiangsu Key Laboratory of Construction Materials for technical support. cemconres.2010.12.001.
[27] Taehwan Kim, J. Olek, Effects of sample preparation and interpretation of
thermogravimetric curves on calcium hydroxide in hydrated pastes and mortars,
References Transport. Res. Rec. 2290 (2012) 10–18. https://10.3141/2290-02.
[28] K. Luke, F.P. Glasser, Selective dissolution of hydrated blast furnace slag cements,
Cement Concr. Res. 17 (2) (1987) 273–282, https://doi.org/10.1016/0008-8846
[1] M.S. Aslam, B. Huang, L. Cui, Review of construction and demolition waste
(87)90110-4.
management in China and USA, J. Environ. Manag. 264 (2020), 110445, https://
[29] V. Kocaba, E. Gallucci, K.L. Scrivener, Methods for determination of degree of
doi.org/10.1016/j.jenvman.2020.110445.
reaction of slag in blended cement pastes, Cement Concr. Res. 42 (3) (2012)
[2] C.L. Wong, K.H. Mo, S.P. Yap, U.J. Alengaram, T.-C. Ling, Potential use of brick
511–525, https://doi.org/10.1016/j.cemconres.2011.11.010.
waste as alternate concrete-making materials: a review, J. Clean. Prod. 195 (2018)
[30] G. Dong, Reaction Degree of Fly Ash and Blast Furnace Slag in Cement Pastes,
226–239, https://doi.org/10.1016/j.jclepro.2018.05.193.
China Building Materials Academy, Beijing, 2008.
[3] S. Wild, A. Gailius, H. Hansen, L. Pederson, J. Szwabowski, Pozzolanic properties of
[31] M.B. Haha, K. De Weerdt, B. Lothenbach, Quantification of the degree of reaction
a variety of European clay bricks, Build. Res. Inf. 25 (3) (1997) 170–175, https://
of fly ash, Cement Concr. Res. 40 (11) (2010) 1620–1629, https://doi.org/
doi.org/10.1080/096132197370435.
10.1016/j.cemconres.2010.07.004.
[4] G. Baronio, L. Binda, Study of the pozzolanicity of some bricks and clays,
[32] G. Fang, M. Zhang, Multiscale micromechanical analysis of alkali-activated fly ash-
Construct. Build. Mater. 11 (1) (1997) 41–46, https://doi.org/10.1016/S0950-
slag paste, Cement Concr. Res. 135 (2020), 106141, https://doi.org/10.1016/j.
0618(96)00032-3.
cemconres.2020.106141.
[5] S. Wild, Observations on the use of ground waste clay brick as a cement
[33] X. Gao, Q.L. Yu, H.J.H. Brouwers, Apply 29Si, 27Al MAS NMR and selective
replacement material, Build. Res. Inf. 24 (1) (1996) 35–40. https://10.1080/09
dissolution in identifying the reaction degree of alkali activated slag-fly ash
613219608727496.
composites, Ceram. Int. 43 (15) (2017) 12408–12419, https://doi.org/10.1016/j.
[6] M. Agarwal, A. Krishan, Reusability of construction & demolition waste in bricks,
ceramint.2017.06.108.
Int. Res. J. Eng. Technol. 4 (2017) 148–151.
[34] K. Scrivener, H. Patel, P. Pratt, L. Parrott, Analysis of phases in cement paste using
[7] H. Cheng, Reuse research progress on waste clay brick, Procedia Environ. Sci. 31
backscattered electron images, methanol adsorption and thermogravimetric
(2016) 218–226, https://doi.org/10.1016/j.proenv.2016.02.029.
analysis, MRS Proceedings 85 (1986), in: https://10.1557/PROC-85-67.
[8] B. Wang, L. Yan, Q. Fu, B. Kasal, A comprehensive review on recycled aggregate
[35] P.J.M. Monteiro, G. Geng, D. Marchon, J. Li, P. Alapati, K.E. Kurtis, M.J.A. Qomi,
and recycled aggregate concrete, Resour. Conserv. Recycl. 171 (2021), 105565,
Advances in characterizing and understanding the microstructure of cementitious
https://doi.org/10.1016/j.resconrec.2021.105565.

13
S. Li et al. Cement and Concrete Composites 139 (2023) 105042

materials, Cement Concr. Res. 124 (2019), 105806, https://doi.org/10.1016/j. [48] L. De Windt, D. Deneele, N. Maubec, Kinetics of lime/bentonite pozzolanic
cemconres.2019.105806. reactions at 20 and 50◦ C: batch tests and modeling, Cement Concr. Res. 59 (2014)
[36] V. Kocaba, E. Gallucci, K.L. Scrivener, Methods for determination of degree of 34–42, https://doi.org/10.1016/j.cemconres.2014.01.024.
reaction of slag in blended cement pastes, Cement Concr. Res. 42 (3) (2012) [49] R. Cao, Z. Jia, Z. Zhang, Y. Zhang, N. Banthia, Leaching kinetics and reactivity
511–525, https://doi.org/10.1016/j.cemconres.2011.11.010. evaluation of ferronickel slag in alkaline conditions, Cement Concr. Res. 137
[37] K. Newlands, The Early Stage Dissolution Characteristics of Aluminosilicate (2020), 106202, https://doi.org/10.1016/j.cemconres.2020.106202.
Glasses, University of Aberdeen, 2015. [50] F. Deschner, B. Lothenbach, F. Winnefeld, J. Neubauer, Effect of temperature on
[38] C. Kuenzel, N. Ranjbar, Dissolution mechanism of fly ash to quantify the reactive the hydration of Portland cement blended with siliceous fly ash, Cement Concr.
aluminosilicates in geopolymerisation, Resour. Conserv. Recycl. 150 (2019), Res. 52 (2013) 169–181, https://doi.org/10.1016/j.cemconres.2013.07.006.
104421, https://doi.org/10.1016/j.resconrec.2019.104421. [51] C. Shi, R.L. Day, Pozzolanic reaction in the presence of chemical activators: Part II
[39] R. Snellings, Solution-controlled dissolution of supplementary cementitious — reaction products and mechanism, Cement Concr. Res. 30 (4) (2000) 607–613,
material glasses at pH 13: the effect of solution composition on glass dissolution https://doi.org/10.1016/S0008-8846(00)00214-3.
rates, J. Am. Ceram. Soc. 96 (8) (2013) 2467–2475, https://doi.org/10.1111/ [52] K. Natkunarajah, K. Masilamani, S. Maheswaran, B. Lothenbach, D.A.
jace.12480. S. Amarasinghe, D. Attygalle, Analysis of the trend of pH changes of concrete pore
[40] E.M. Gartner, D.E. Macphee, A physico-chemical basis for novel cementitious solution during the hydration by various analytical methods, Cement Concr. Res.
binders, Cement Concr. Res. 41 (7) (2011) 736–749, https://doi.org/10.1016/j. 156 (2022), 106780, https://doi.org/10.1016/j.cemconres.2022.106780.
cemconres.2011.03.006. [53] M. Criado, A. Fernández-Jiménez, A. Palomo, I. Sobrados, J. Sanz, Effect of the
[41] Y. Xiao, A.C. Lasaga, Ab initio quantum mechanical studies of the kinetics and SiO2/Na2O ratio on the alkali activation of fly ash. Part II: 29Si MAS-NMR Survey,
mechanisms of quartz dissolution: OH− catalysis, Geochem. Cosmochim. Acta 60 Microporous Mesoporous Mater. 109 (1) (2008) 525–534, https://doi.org/
(13) (1996) 2283–2295, https://doi.org/10.1016/0016-7037(96)00101-9. 10.1016/j.micromeso.2007.05.062.
[42] Y. Briki, M. Zajac, M.B. Haha, K. Scrivener, Factors affecting the reactivity of slag [54] Á. Palomo, S. Alonso, A. Fernandez-Jiménez, I. Sobrados, J. Sanz, Alkaline
at early and late ages, Cement Concr. Res. 150 (2021), 106604, https://doi.org/ activation of fly ashes: NMR study of the reaction products, J. Am. Ceram. Soc. 87
10.1016/j.cemconres.2021.106604. (6) (2004) 1141–1145, https://doi.org/10.1111/j.1551-2916.2004.01141.x.
[43] S. Li, G. Chen, Z. Xu, X. Luo, J. Gao, Particle-size effect of recycled clay brick [55] K.J.D. Mackenzie, R.H. Meinhold, B.L. Sherriff, Z. Xu, 27 Al and 25 Mg solid-state
powder on the pore structure of blended cement paste, Construct. Build. Mater. magic-angle spinning nuclear magnetic resonance study of hydrotalcite and its
344 (2022), 128288, https://doi.org/10.1016/j.conbuildmat.2022.128288. thermal decomposition sequence, J. Mater. Chem. 3 (12) (1993) 1263–1269,
[44] I. Nikolić, A. Drinčić, D. Djurović, L. Karanović, V.V. Radmilović, V.R. Radmilović, https://doi.org/10.1039/jm9930301263.
Kinetics of electric arc furnace slag leaching in alkaline solutions, Construct. Build. [56] A.V. Girão, I.G. Richardson, R. Taylor, R.M.D. Brydson, Composition, morphology
Mater. 108 (2016) 1–9, https://doi.org/10.1016/j.conbuildmat.2016.01.038. and nanostructure of C–S–H in 70% white Portland cement–30% fly ash blends
[45] B.R. Bickmore, K.L. Nagy, A.K. Gray, A.R. Brinkerhoff, The effect of Al(OH)4− on hydrated at 55◦ C, Cement Concr. Res. 40 (9) (2010) 1350–1359, https://doi.org/
the dissolution rate of quartz, Geochem. Cosmochim. Acta 70 (2) (2006) 290–305, 10.1016/j.cemconres.2010.03.012.
https://doi.org/10.1016/j.gca.2005.09.017. [57] B. Li, Q. Li, W. Chen, Spatial zonation of a hydrotalcite-like phase in the inner
[46] Q. Luo, G. Chen, Y. Sun, Y. Ye, X. Qiao, J. Yu, Dissolution kinetics of aluminum, product of slag: new insights into the hydration mechanism, Cement Concr. Res.
calcium, and iron from circulating fluidized bed combustion fly ash with 145 (2021), 106460, https://doi.org/10.1016/j.cemconres.2021.106460.
hydrochloric acid, Ind. Eng. Chem. Res. 52 (51) (2013) 18184–18191. [58] O. Deruelle, O. Spalla, P. Barboux, J. Lambard, Growth and ripening of porous
[47] N. Cristelo, P. Tavares, E. Lucas, T. Miranda, D. Oliveira, Quantitative and layers in water altered glasses, J. Non-Cryst. Solids 261 (1) (2000) 237–251,
qualitative assessment of the amorphous phase of a Class F fly ash dissolved during https://doi.org/10.1016/S0022-3093(99)00587-6.
alkali activation reactions – effect of mechanical activation, solution concentration [59] D. Rebiscoul, A.V.d. Lee, F. Rieutord, F. Ne, O. Spalla, A. El-Mansouri, P. Frugier,
and temperature, Compos. B Eng. 103 (2016) 1–14, https://doi.org/10.1016/j. A. Ayral, S. Gin, Morphological evolution of alteration layers formed during
compositesb.2016.08.001. nuclear glass alteration: new evidence of a gel as a diffusive barrier, J. Nucl. Mater.
326 (2004) 9–18. https://10.1016/j.jnucmat.2003.10.015.

14

You might also like