nanomaterials

Article
Effect of Graphene Oxide on Mechanical Properties
and Durability of Ultra-High-Performance Concrete
Prepared from Recycled Sand
Hongyan Chu 1, * , Yu Zhang 2 , Fengjuan Wang 2 , Taotao Feng 2 , Liguo Wang 2 and
Danqian Wang 3
1 College of Civil Engineering, Nanjing Forestry University, Nanjing 210037, China
2 Jiangsu Key Laboratory of Construction Materials, School of Materials Science and Engineering,
Southeast University, Nanjing 211189, China; tgyuzhang@outlook.com (Y.Z.);
fengjuan19921118@sina.com (F.W.); taotao_feng@yeah.net (T.F.); wlg_seu@sina.com (L.W.)
3 Advanced and Innovative Materials (AIM) Group, Department of Civil, Environmental and Geomatic
Engineering, University College London, London WC1E 6BT, UK; danqian.wang.16@ucl.ac.uk
* Correspondence: chuhongyan@njfu.edu.cn




Received: 27 July 2020; Accepted: 28 August 2020; Published: 30 August 2020


Abstract: Ultra-high-performance concrete (UHPC) has been used as an advanced construction

material in civil engineering because of its excellent mechanical properties and durability. However,
with the depletion of the raw material (river sand) used for preparing UHPC, it is imperative to
find a replacement material. Recycled sand is an alternative raw material for preparing UHPC, but
it degrades the performance. In this study, we investigated the use of graphene oxide (GO) as an
additive for enhancing the properties of UHPC prepared from recycled sand. The primary objective
was to investigate the effects of GO on the mechanical properties and durability of the UHPC at
different concentrations. Additionally, the impact of the GO additive on the microstructure of the
UHPC prepared from recycled sand was analysed at different mixing concentrations. The addition of
GO resulted in the following: (1) The porosity of the UHPC prepared from recycled sand was reduced
by 4.45–11.35%; (2) the compressive strength, flexural strength, splitting tensile strength, and elastic
modulus of the UHPC prepared from recycled sand were enhanced by 8.24–16.83%, 11.26–26.62%,
15.63–29.54%, and 5.84–12.25%, respectively; (3) the resistance of the UHPC to penetration of chloride
ions increased, and the freeze–thaw resistance improved; (4) the optimum mixing concentration of
GO in the UHPC was determined to be 0.05 wt.%, according to a comprehensive analysis of its effects
on the microstructure, mechanical properties, and durability of the UHPC. The findings of this study
provide important guidance for the utilisation of recycled sand resources.

Keywords: ultra-high-performance concrete; mechanical properties; durability; recycled sand;

graphene oxide; microstructure

1. Introduction
Ultra-high-performance concrete (UHPC) is a new construction material with excellent mechanical
properties and durability [1], which makes it an innovative material in the engineering field [2].
However, its preparation requires high-quality raw materials [3] such as quartz sand or high-grade
river sand. With the continuous modernisation and urbanisation progress in China, the domestic
construction industry has undergone vigorous development, accompanied by the rapid depletion
of raw construction materials, particularly river sand. Currently, there is a shortage of river sand
in most parts of China, significantly increasing the price of river sand. In certain areas, there is no
river sand at all. In fact, the entire world is facing a shortage of river sand. Therefore, it is critical to

Nanomaterials 2020, 10, 1718; doi:10.3390/nano10091718 www.mdpi.com/journal/nanomaterials

Nanomaterials 2020, 10, 1718 2 of 17

find an alternative fine aggregate to replace river sand. With the increasing urbanisation in China,
a large amount of construction waste has been generated there in recent years during the processes
of construction, reconstruction, expansion, and demolition. In 2017, a total of 2.38 billion tons of
construction waste was generated, and this number is expected to exceed 2.6 billion tons by the end of
2020. However, the utilisation rate of construction waste in China is <5% and is significantly lower
than that for Japan and Germany [4]. Therefore, transforming construction waste into reclaimed sand
and using it as the raw material for preparing UHPC will not only allow the utilisation of construction
waste but also turn the waste into a valuable resource that contributes to energy savings and emission
reduction. These merits will provide significant social, environmental, and economic benefits.
The preparation of concrete using recycled sand has been widely investigated. Kumar et al. [5]
reported that the performance of concrete prepared by using recycled sand remained unaffected when
the replacement level of recycled sand was <20 wt.%. Xiao et al. [6] reported that the elastic modulus
of concrete will be reduced by 45% if the replacement level of recycled sand is 100%. In general, using
recycled sand in concrete degrades its mechanical properties [7–11] and durability [7–9,12–17]. In the
case of concrete prepared from recycled sand, a large interfacial transition layer is formed on the surface
of the recycled sand during the hydration process. Additionally, the old mortar formed at the surface
of the recycled sand exhibits poor mechanical properties [18]. These features are the primary reasons
for the poor mechanical properties of concrete prepared from recycled sand. Compared with normal
concrete, concrete prepared using recycled sand has a larger chloride-ion diffusion coefficient [19],
a greater carbonation depth [20], and a higher water-absorption rate [21]. These properties indicate
that the use of recycled sand degrades the durability of the concrete. Because recycled sand has a
high porosity, concrete prepared using recycled sand has a high permeability, eventually degrading its
durability. A recent study indicated that the mechanical properties of UHPC deteriorate significantly
when the replacement level of recycled sand for river sand exceeds 50% [22]. Thus, it is critical
to develop methods that can improve the performance of concrete prepared from recycled sand.
Furthermore, few studies have been performed on the preparation of UHPC using recycled sand.
Additionally, research on the durability of UHPC prepared from recycled sand is scarce.
The use of nanomaterials can improve the properties of the matrix in cement-based materials
prepared from recycled sand, enhancing the material performance. With the advancement of
nanotechnology, the mixing of nanomaterials into cement-based materials has been increasingly
researched. A wide variety of nanomaterials have been investigated, including nano-silicon
oxide [23–25], nano-titanium oxide [26], carbon nanotubes [27–29], sulfonated graphene [30–32],
and graphene oxide (GO) [33–37]. GO is a new type of nanomaterial with excellent mechanical
properties and good dispersion properties. The elastic modulus and inherent strength of GO are
as high as 300 and 112 GPa, respectively [38]. The performance of cement-based materials can be
improved via mixing with GO for the following reasons: (1) the filling effect [39], (2) an increasing
number of nucleation sites during the hydration of cement [40], and (3) the ability to tune the
morphology of cement hydration products [41]. Therefore, the use of GO presents great potential for
enhancing the mechanical properties and durability of cement-based materials [33]. Of course, there
are some other methods to improve the properties of cement-based materials [42,43]. In particular,
nanomaterials have attracted increasing attention, owing to their capability to significantly improve
the properties of cement-based materials at a low mixing concentration. However, nanomaterials are
expensive, which prevents them from being used in large-scale applications in the field of cement-based
materials. With the advancement of nanotechnology, GO can be produced at an industrial scale at low
manufacturing costs. Additionally, only an extremely small quantity of GO is required to be mixed
in the cement-based material. Consequently, it is economically feasible to use GO for improving the
properties of cement-based materials. Thus far, the use of GO to improve the properties of UHPC
prepared from recycled sand has not been investigated. Furthermore, most of the previous research
focused on the effects of GO on the mechanical properties of concrete made from recycled sand.
Few studies have investigated the influence of GO on the concrete durability. Therefore, it is imperative
Nanomaterials 2020, 10, 1718 3 of 17

to analyse the effects of GO on both the mechanical properties and durability of UHPC prepared from
recycled sand.
In this study, the properties of UHPC prepared from recycled sand were enhanced via mixing
with GO additives. The primary objective was to investigate the effects of GO as an additive on the
mechanical properties and durability of recycled-sand-based UHPC (RS-UHPC) at different mixing
concentrations. Specifically, the effects of the GO additive on the mechanical properties (compressive
strength, flexural strength, splitting tensile strength, and elastic modulus) and durability (chloride-ion
penetration resistance and freeze–thaw resistance) of the RS-UHPC were systematically analysed at
three different concentrations. Additionally, the effects of the GO additive on the microstructure of the
RS-UHPC were investigated. The results provide important guidance for the application of UHPC
made from recycled sand.

2. Experiment

2.1. Experimental Materials

The following materials were procured to prepare the concrete sample in this study: (1) P·II52.5
grade cement produced by Jiangnan-Xiaoyetian Cement Co., Ltd. (Nanjing, China); (2) 95# micro-silicon
powder produced by Aiken International Trading (Shanghai) Co., Ltd. (Shanghai, China); and (3) grade
I fly ash produced by Zhuhai Minghui Trading Co., Ltd. (Zhuhai, China). The chemical compositions
of the cement, silicon powder, and fly ash used in this study are presented in Table 1. The recycled
sand used in this study was provided by a local construction waste disposal company. The physical
properties, grading curve, and chemical composition of the recycled sand are presented in Table 2,
Figure 1, and Table 1, respectively. The polycarboxylic acid superplasticiser used in this study was
produced by Jiangsu Subote New Materials Co., Ltd. (Nanjing, China). The solid content and
water-reducing rate of the superplasticiser were 40% and 33%, respectively. The hooked-end steel
fibres used in the experiments were purchased from Aiken International Trading (Shanghai) Co., Ltd.
(Shanghai, China). They had a length of 20.0 mm, a diameter of 0.35 mm, and a tensile strength of
approximately 3000 MPa. The CaO-type expansion agents were purchased from Jiangsu Subote New
Materials Co., Ltd. The chemical composition of the expansion agents is presented in Table 1. Finally,
the GO used in this study was produced by Shanghai Carbon Source Huigu New Material Technology
Co., Ltd. (Shanghai, China). The physical and chemical properties of the GO are presented in Table 3.
The real particle size distribution of liquid GO is shown in Figure 2. As shown in Figure 2, the particle
size distribution of the liquid GO was between 0.34 and 9.53 µm, and the mean size was 2.24 µm.
The water used in the experiments was tap water.

Table 1. The chemical compositions of cement, fly ash, silica fume, recycled sand, and expansive
agent (wt.%).

Chemical Composition Cement Fly Ash Silica Fume Recycled Sand Expansive Agent
CaO 64.78 18.93 0.22 20.95 68.21
Al2 O3 5.92 7.26 0.39 14.32 7.46
SiO2 20.45 73.15 96.21 55.09 6.37
Fe2 O3 3.13 0.09 0.65 3.48 4.15
MgO 1.42 0.17 0.17 1.69 1.25
SO3 3.07 1.47 12.56
K2 O 0.74 0.06 1.60
Na2 O 0.18 0.22 0.65
TiO2 0.31 0.05 0.54
P2 O5 0.07 0.21
 Nanomaterials 2020, 10, x FOR PEER REVIEW 4 of 18

Nanomaterials 2020, 10, x FORTable 2. The physical properties of the fine recycled sand.
PEER REVIEW 4 of 18
Nanomaterials 2020, 10, 1718 4 of 17
Crushing Index (%) Apparent Density (kg/m3) Water Absorption (%) Fineness Modulus
Table 2. The physical properties of the fine recycled sand.
22.6 2548 6.13 2.45
Crushing Index (%) Table 2. The physical
Apparent Densityproperties
(kg/m3of) the fine recycled
Water sand. (%) Fineness Modulus
Absorption
Table 3. The physical and chemical properties of graphene
6.13 (%) oxideFineness
(GO).
Crushing 22.6
Index (%) 2548
Apparent Density (kg/m3 ) Water Absorption 2.45
Modulus
Thickness
22.6 Flake Diameter
2548 Carbon Content 6.13 Oxygen Content 2.45
Purity Table 3. The physical and chemical properties of graphene oxide (GO). Dispersant
(nm) ( ) (%) (%)
98% Thickness Flake
~1Table 3. The Diameter
0.2–10
physical and chemicalCarbon~46Content
properties of grapheneOxygen
oxide ~53Content
(GO). water
Purity Dispersant
(nm) ( ) (%) (%)
Thickness Flake Carbon Oxygen
98%
Purity ~1 (nm) 0.2–10
Diameter (µm)
~46
Content (%) Content (%)
~53 Dispersant water

98% ~1 0.2–10 ~46 ~53 water

100
1.18 2.36 4.75
100
80 2.36 4.75
rate/(%)

1.18
80
60 0.6
rate/(%)
Passing

60
40 0.3 0.6
Passing

0.3
40 0.15
20
0.075
0.15
20
0 0.075
0 1 2 3 4 5
0 Sieve diameter/(mm)
0 1 2 3 4 5
Figure 1. The particle
Sievesize distribution of recycled sand.
diameter/(mm)
Figure 1. The
Figure particle
1. The sizesize
particle distribution of recycled
distribution sand.
of recycled sand.
25
2.24 μm
25
20 2.24 μm
Intensity/(%)

20
15
Intensity/(%)

15
10

10
5

5
0
0.01 0.1 1 10 100
0 Particle size/(μm)
0.01 0.1 1 10 100
Figure 2. The particle sizesize
distribution of GO viavia
dynamic light scattering test.test.
Figure 2. The particle distribution of GO
Particle size/(μm)
dynamic light scattering

2.2. Mixing Ratio in Experiments

Figure 2. The particle size distribution of GO via dynamic light scattering test.
Table 4 presents the mixing ratio of the different components investigated in this study. The initial
mixing ratio (labelled as RU0) corresponded to that of RS-UHPC without GO. For the other three
mixing ratios, different contents of GO were used in the sample. RU25, RU50, and RU75 labels were
Nanomaterials 2020, 10, 1718 5 of 17

used, corresponding to GO contents of 0.025, 0.05, and 0.075 wt.%, respectively. The GO content
was evaluated with respect to the mass of the cementing material (mixture of cement and minerals).
The steel fibre content in the RS-UHPC sample was fixed at 2.5 vol.% for all the mixing ratios. Different
amounts of water reducers were added to the RS-UHPC sample, in order to ensure a consistent working
performance among the different sample groups. In particular, the slump expansion was maintained
at approximately 560 mm for each group of RS-UHPC samples.

Table 4. The mix proportions of recycled-sand-based ultra-high-performance concrete (RS-UHPC) (kg/m3 ).

Mixture RU0 RU25 RU50 RU75

Cement 690 690 690 690
Fly ash 345 345 345 345
Silica fume 115 115 115 115
Recycled sand 1265 1265 1265 1265
Expansive agent 34.5 34.5 34.5 34.5
Steel fibre 195 195 195 195
Water 207 207 207 207
Water reducer 23 24.6 29.4 36.1
GO 0 0.2875 0.5750 0.8625

2.3. Preparation and Curing of UHPC

Prior to the preparation of the UHPC, the recycled sand was soaked in water for 12 h, fully
saturated with water, and then the recycled sand was washed by water, to remove mud. Eventually,
the recycled sand was dried naturally in air at ambient temperature. These procedures ensured that
the inside of the recycled sand was saturated with water, while the surface was still dry. Subsequently,
an ordinary mixer was used to prepare RS-UHPC. The detailed preparation process was as follows.
(1) Cement, silica fume, fly ash, and an expansion agent were mixed in the mixer, with stirring for 4 min.
(2) Recycled sand was added, and the stirring was continued for 3 min. (3) Approximately 3⁄4 water
was added to the water reducer, and thorough mixing was performed, using a glass rod. The product
was then added to the mixture obtained in Step (2), followed by 3 min of stirring. (4) The glass rod
was rinsed, along with the container of the water reducer with the remaining water. Then, the entire
solution was added to the mixture obtained in Step (3). Stirring was continued for 5 min. (5) Steel
fibres were added uniformly to the mixture obtained from Step (5), with stirring for 5 min.
Different types of UHPC test pieces were prepared from recycled sand, depending on the objective
of the experiment. After the test pieces were moulded, the moulds were wrapped with plastic film and
cured for 48 h, at room temperature. Subsequently, the moulds were dissembled, and the RS-UHPC
test pieces were further cured in a standard curing room, for 28 days. The temperature and relative
humidity of the curing room were maintained at 20 ± 1 ◦ C and >95%, respectively.

2.4. Experimental Method

To examine the microstructure of the RS-UHPC, we measured the porosity and pore-size
distribution of RS-UHPC samples, using AutoPore IV mercury intrusion porosimetry (Micromeritics
Instrument Corporation). This instrument can measure the pore size in the range of 3.6 nm to 400 µm,
with a maximum pressure of 414 MPa.
The tensile stress of the RS-UHPC was measured by using a universal testing machine at a loading
rate of 0.80 MPa/s. The dimensions of the test piece were 100 × 100 × 100 mm3 . There were eight
RS-UHPC test pieces for each component mixing ratio. The tensile strength of the RS-UHPC was
determined as the average of five experimental measurements. The flexural strength of the RS-UHPC
was measured, using a universal testing machine and a four-point flexural experimental device at a
loading rate of 0.08 MPa/s. The dimensions of the test piece were 100 × 100 × 400 mm3 . There were
10 RS-UHPC test pieces for each component mixing ratio. The flexural strength of the RS-UHPC
was determined by taking the average measurements for five splitting-tensile-strength experiments.
Nanomaterials 2020, 10, 1718 6 of 17

The splitting tensile strength of the RS-UHPC was measured by using a universal testing machine and
a splitting tensile strength experimental device at a loading rate of 0.08 MPa/s. The dimensions of the
test piece were 100 × 100 × 100 mm3 . Eight RS-UHPC test pieces were prepared for each component
mixing ratio. The splitting tensile strength of the RS-UHPC was determined as the average of five
experimental measurements. Finally, the elastic modulus of the RS-UHPC was measured, using a
universal testing machine and a micro-deformation experimental device at a loading rate of 0.08 MPa/s.
The dimensions of the test piece were 100 × 100 × 300 mm3 . There were 10 RS-UHPC test pieces for
each component mixing ratio. The elastic modulus of the RS-UHPC was determined as the average of
six experimental measurements. All four mechanical properties of the RS-UHPC were measured in
accordance with the Chinese standard “GB/T, 50081-2002” [44].
The chloride-ion penetration resistance and freeze–thaw resistance of the RS-UHPC were analysed
in accordance with the Chinese standard “GB/T 50082-2009” [45]. The test pieces used for the
durability test were first cured under standard curing conditions, for 91 days. The microstructure of
the RS-UHPC became more stable during this curing process. The chloride-ion permeability of the
RS-UHPC was characterised, using the rapid chloride-ion migration coefficient method. Such a test
requires a test piece with a diameter of 100 mm and a height of 50 mm. The frost resistance of the
RS-UHPC was characterised via a rapid freeze–thaw experiment. The dimensions of the test piece
were 100 × 100 × 100 mm3 . The mass and dynamic elastic modulus of the test piece were measured
prior to the freeze–thaw experiment and every 30 test cycles during the experiment. A total of 10
RS-UHPC test pieces were prepared for the durability test. The experiments were performed five times,
and the average values of the measurements were recorded.

3. Results and Discussion

3.1. Microstructure
The porosity of the RS-UHPC samples containing different concentrations of GO is shown in
Figure 3. The porosities of all the RS-UHPC samples containing GO were lower than those of the
samples without GO. This suggests that the addition of GO can reduce the porosity of RS-UHPC.
The porosities of RU0, RU25, RU50, and RU75 were measured to be 2.47%, 2.36%, 2.19%, and 2.25%,
respectively. The porosities of these samples decreased in the following order: RU0 > RU25 > RU75
> RU50. These values indicate that an increasing GO content leads to a nonlinear reduction in the
porosity of RS-UHPC. The porosity of the RS-UHPC prepared in this study was lower than those
of low-heat concrete [46], sacrificial concrete [47,48], and UHPC mixed with coarse aggregates [49].
The lowest porosity of RS-UHPC was observed when the GO concentration was 0.05 wt.% Therefore,
from the perspective of the porosity, the optimum GO concentration to be mixed in RS-UHPC is
0.05 wt.%. The porosities of RU25, RU50, and RU75 were lower than that of RU0 by 4.45%, 11.34%,
and 8.91%, respectively. This suggests that adding GO can improve the pore structure of RS-UHPC.
Because GO can enhance the degree of cement hydration [50], adding GO increases the amount of
hydration products in cement-based materials. Furthermore, the addition of GO provides a denser
interfacial transition layer in cement-based materials [51]. These features result in a reduced porosity
in RS-UHPC samples.
 Nanomaterials 2020, 10, x FOR PEER REVIEW 7 of 18

Nanomaterials 2020, 10, x FOR PEER REVIEW 7 of 18

Nanomaterials 2020, 10, 1718 7 of 17
3.0
0.14
0.11
2.5
3.0 0.17 0.07
0.14
2.0 0.11
2.5 0.17 0.07

Porosity/%
1.5
2.0
Porosity/%
1.0
1.5

0.5
1.0

0.0
0.5
RU0 RU25 RU50 RU75
0.0 Different kinds of UHPC
RU0 RU25 RU50 RU75
Different
Figure 3. The porosity kindswith
of RS-UHPC ofdifferent
UHPC contents of GO.
Figure 3. The porosity of RS-UHPC with different contents of GO.
The pore-size distributions of RS-UHPC
Figure 3. The porosity containing
of RS-UHPC with different concentrations
different contents of GO. of GO are shown in
Figure
The pore-size distributions of RS-UHPC containing different concentrationspatterns.
4. Different RS-UHPC samples exhibited similar pore-size distribution of GO areAshown
typicalin
peak
was4.
Figure observed
The pore-size
Different indistributions
the pore-size
RS-UHPC samples distribution
of RS-UHPC
exhibited curve
containing
similar for alldistribution
different
pore-size the RS-UHPC
concentrations samples.
patterns.of GO areThis
shown
A typical peak
peak in
was corresponded
Figure to
4. Different
observed the most
RS-UHPC
in the pore-size probable pore size.
samples exhibited
distribution As indicated
curve forsimilar by the results,
pore-size distribution
all the RS-UHPC the
samples. This most probable
patterns. pore size
A typical peak
peak corresponded
to the most probable pore size. As indicated by the results, the most probable pore sizemost
was
was 11.42,
observed11.04,
in7.38,
the and 8.23
pore-sizenm for RU0,
distribution RU25,
curveRU50,for and
all RU75,
the respectively.
RS-UHPC The
samples. was probable
This peak
11.42,
pore sizes
corresponded decreased
to the mostin the order
probable of
poreRU0size.> RU25
As > RU75
indicated by > RU50,
the in
results,
11.04, 7.38, and 8.23 nm for RU0, RU25, RU50, and RU75, respectively. The most probable pore sizesagreement
the most with
probablethe porosity
pore size
measurements.
was 11.42,
decreased 11.04,
in the
Soliman
7.38,
order of and
and
RU08.23
Tagnit-Hamou
nm for
> RU25 RU0, >
> RU75
[52]
RU25,
RU50,
reported
RU50, that
and RU75,
in agreement
the most probable
respectively.
with the porosity
pore
The size of UHPC
most probable
measurements.
is
poreapproximately
sizes decreased 10 nm,
in which
the order isofalso
RU0 consistent
> RU25 with
> RU75our > experimental
RU50, in
Soliman and Tagnit-Hamou [52] reported that the most probable pore size of UHPC is approximately
finding.
agreement Additionally,
with the
the porosity
most probable Soliman
10 measurements.
nm, which is also
pore sizes of with
and
consistent
RU25, RU50, and
Tagnit-Hamou RU75
[52]
our experimental
are all
reported
finding.
smaller
that thanprobable
the most
Additionally,
that of RU0.
the mostpore This
size behaviour
probableof pore
UHPC
indicates that
is approximately the pore structure of RS-UHPC can be improved by adding GO.
sizes of RU25, RU50,10 and nm, which
RU75 is also
are all consistent
smaller than that with our experimental
of RU0. This behaviour finding. Additionally,
indicates that the porethe
most probable pore sizes of RU25, RU50,
structure of RS-UHPC can be improved by adding GO. and RU75 are all smaller than that of RU0. This behaviour
indicates that the pore structure of RS-UHPC can be improved by adding GO.
0.20
-1
intrusion/mL⋅g

RU0
0.20
0.16 RU25
-1
intrusion/mL⋅g

RU50
RU0
0.16 RU75
RU25
0.12 RU50
RU75
Log differential

0.12
0.08
Log differential

0.08
0.04

0.04
0.00
100 101 102 103 104 105
0.00 Pore size diameter/nm
100 101 102 103 104 105
Figure 4. The pore-size distribution of RS-UHPC with different contents of GO.
Figure 4. The pore-size distribution of RS-UHPC with different contents of GO.
Pore size diameter/nm

Figure 4. The pore-size distribution of RS-UHPC with different contents of GO.

 Nanomaterials 2020, 10, x FOR PEER REVIEW 8 of 18

3.2. Mechanical Properties

Nanomaterials 2020, 10, 1718 8 of 17

3.2.1. Compressive Strength

3.2. Mechanical Properties
The compressive strengths of the RS-UHPC samples containing different concentrations of GO
areCompressive
3.2.1. presented in Strength
Figure 5. As shown, the RS-UHPC samples containing GO generally had a higher
compressive strength than those without GO. This suggests that adding GO can improve the
The compressive
compressive strengths
performance of the RS-UHPC
of RS-UHPC. samples
Specifically, containing strengths
the compressive different of
concentrations
RU0, RU25, RU50, of
GOand areRU75
presented in Figure 5. As shown, the RS-UHPC samples containing GO
were measured to be 156.21, 169.08, 182.50, and 174.23 MPa, respectively. The compressive generally had a
higher compressive strength than those without GO. This suggests that adding
strength of RU0 was higher than the minimum strength requirement of 150 MPa for UHPC. Thus, GO can improve the
compressive performance
from the perspective of of
theRS-UHPC.
mechanical Specifically,
properties,the compressive
recycled sand can strengths
be usedofasRU0, RU25,
the raw RU50,for
material
andpreparing
RU75 were measured to be 156.21, 169.08, 182.50, and 174.23 MPa, respectively.
UHPC. The compressive strengths of RU25, RU50, and RU75 were 8.24%, 16.83%, and The compressive
strength
11.54%ofhigher,
RU0 was higher than
respectively, thanthethat
minimum
of RU0. strength requirement
These results of 150
indicate that theMPa for UHPC.
compressive Thus, of
strength
from the perspective of the mechanical properties, recycled sand can be used as
UHPC does not increase linearly with the increasing GO concentration. In contrast, there exists the raw material for an
preparing UHPC. The compressive strengths of RU25, RU50, and RU75 were
optimum GO concentration at which the optimal mechanical properties are achieved. From the 8.24%, 16.83%, and 11.54%
higher, respectively,
perspective of thethan that
tensile of RU0.the
strength, These resultsconcentration
optimum indicate that oftheGOcompressive
to be mixed strength of UHPC
in RS-UHPC is 0.05
does not increase linearly with the increasing GO concentration. In contrast,
wt.%. In this study, the enhancement in the tensile strength of RS-UHPC was attributed to thethere exists an optimum
GOfollowing
concentration at which(1)the
two factors: Theoptimal
presence mechanical
of GO can properties
enhanceare theachieved.
degree ofFrom the perspective
hydration of
of the cement-
thebased
tensilematerial
strength,and
theimprove
optimumitsconcentration
microstructure of GO[50],toand
be mixed in RS-UHPC
(2) adding GO can ischange
0.05 wt.%. In this
the hydration
study, the enhancement
products of cement [53]. in the tensile strength of RS-UHPC was attributed to the following two factors:
(1) The presence of GO can enhance the degree of hydration of the cement-based material and improve
its microstructure [50], and (2) adding GO can change the hydration products of cement [53].

200 6.11 8.50

Compressive strength/MPa

5.90
8.32
160

120

80

40

0
RU0 RU25 RU50 RU75
Different kinds of UHPC
Figure 5. The compressive strength of RS-UHPC with different contents of GO.
Figure 5. The compressive strength of RS-UHPC with different contents of GO.
3.2.2. Flexural Strength
3.2.2. Flexural Strength
The flexural strengths of RS-UHPC samples containing different concentrations of GO are shown
in Figure The6. flexural
Similar strengths
to the trendof observed
RS-UHPCfor samples containingstrength,
the compressive differentthe concentrations of GOofare
flexural strengths
shown in
RS-UHPC Figure containing
samples 6. Similar toGO thewere
trend observedhigher
generally for thethan
compressive
those withoutstrength, the flexural
GO. This indicatesstrengths
that
of RS-UHPC
adding samplesthe
GO can improve containing GO were
flexural strength of generally
RS-UHPC.higher than those
The flexural without
strengths GO.RU25,
of RU0, This indicates
RU50,
andthat adding
RU75 wereGO can17.68,
15.89, improve theand
20.12, flexural
19.24 strength of RS-UHPC.
MPa, respectively. Thus,Thetheflexural
flexuralstrengths
strengthsofdecreased
RU0, RU25,
RU50,
in the andofRU75
order > RU75
RU50were > RU25
15.89, 17.68,>20.12,
RU0. and
The 19.24 MPa,
flexural respectively.
strengths of RU25, Thus,
RU50,theand
flexural
RU75strengths
were
decreased
improved in the order
by 11.26%, 26.62%,of and
RU50 > RU75
21.08%, > RU25 > compared
respectively, RU0. The withflexural
thatstrengths
of RU0. The of results
RU25, RU50,
indicateand
thatRU75 were improved
the flexural strength of byRS-UHPC
11.26%, 26.62%,
does notand 21.08%,
increase respectively,
linearly with thecompared
increasingwithGO that of RU0. The
concentration.
Theresults
flexuralindicate
strengththatofthe
theflexural
RS-UHPC strength of RS-UHPC
was maximised at adoes not increase linearly
GO concentration of 0.05 with
wt.%.the increasing
Therefore,
theGO concentration.
optimum The flexural
concentration of GO to strength
be mixedof the RS-UHPCiswas
in RS-UHPC 0.05maximised
wt.%, fromat thea perspective
GO concentrationof the of
0.05 wt.%.
flexural Therefore,
strength. the optimum
In this study, concentration
the flexural strength ofoftheGO to be mixed RS-UHPC
GO-enhanced in RS-UHPC is 0.05
ranged wt.%,
from from
15.89
the perspective
to 20.12 MPa for the of different
the flexural
GOstrength. In this study,
concentrations. the flexural
This flexural strength
strength of the
exceeds thatGO-enhanced
of previouslyRS-
reported UHPC prepared from river sand [54]. The flexural strength of UHPC is primarily affected by
 Nanomaterials 2020, 10, x FOR PEER REVIEW 9 of 18

UHPC ranged from 15.89 to 20.12 MPa for the different GO concentrations. This flexural strength
Nanomaterials
exceeds that 10,previously
2020,of 1718 reported UHPC prepared from river sand [54]. The flexural strength 9 of 17of

UHPC is primarily affected by the following factors: the degree of hydration of the cement, the
mechanical
the strengththe
following factors: of degree
the slurry, the bonding
of hydration strength
of the cement, between the steelstrength
the mechanical fibres and the slurry,
of the matrix,theand
the dispersion performance of the steel fibres [55,56]. Considering that the presence
bonding strength between the steel fibres and the matrix, and the dispersion performance of the steel of GO can
enhance
fibres theConsidering
[55,56]. degree of cement
that the hydration
presence ofand
GO improve
can enhancethe the
microstructure of the
degree of cement cement-based
hydration and
material [50], adding GO to RS-UHPC can improve its flexural strength.
improve the microstructure of the cement-based material [50], adding GO to RS-UHPC can improve its
flexural strength.

25
1.13
0.92
20
Flexural strength/MPa

0.77
0.60
15

10

0
RU0 RU25 RU50 RU75
Different kinds of UHPC
Figure 6. The flexural strength RS-UHPC with different contents of GO.
Figure 6. The flexural strength RS-UHPC with different contents of GO.
3.2.3. Splitting Tensile Strength
3.2.3. Splitting Tensile Strength
The splitting tensile strengths of RS-UHPC samples containing different concentrations of GO are
shownThe splitting
in Figure tensile to
7. Similar strengths
the trendsof RS-UHPC
observed for samples containingstrength
the compressive differentandconcentrations of GO
flexural strength,
are
the shown tensile
splitting in Figure 7. Similar
strengths to RS-UHPC
of the the trendssamples
observed for the compressive
containing strengthhigher
GO were generally and flexural
than
strength,
those of thethe splitting
samples tensile GO.
without strengths
This of the RS-UHPC
suggests samples
that adding GOcontaining
can improve GO were generally
the splitting higher
tensile
than those of the samples without GO. This suggests that adding GO can improve
strength of RS-UHPC. Specifically, the splitting tensile strengths of RU0, RU25, RU50, and RU75 were the splitting tensile
strength
13.37, 15.46, of17.32,
RS-UHPC. Specifically,
and 16.39 the splittingThus,
MPa, respectively. tensile
thestrengths
splittingoftensile
RU0, RU25,
strengthsRU50, and RU75
decreased were
in the
13.37,
order of 15.46, > RU75
RU50 17.32, > RU25
and > RU0.
16.39 MPa, respectively.
The splitting Thus, thestrengths
tensile splitting tensile
of RU25, strengths
RU50, anddecreased in the
RU75 were
order of RU50
improved > RU75
by 15.63%, > RU25
29.54%, > RU0.
and Therespectively,
22.59%, splitting tensile strengths
compared of RU25,
with that ofRU50,
RU0. andThese RU75 were
values
improved
suggest that by
the 15.63%,
splitting29.54%, and 22.59%,
tensile strength respectively,
of RS-UHPC doescompared
not increasewith that of
linearly RU0.
with theThese values
increasing
suggest that the splitting tensile strength of RS-UHPC does not increase linearly
GO concentration. The splitting tensile strength of the RS-UHPC was maximised at a GO concentration with the increasing
ofGO
0.05 concentration.
wt.%. Therefore, Thethesplitting
optimum tensile strength ofofGO
concentration theto RS-UHPC
be mixed in was maximised
RS-UHPC is 0.05at wt.%,
a GO
concentration
from the perspectiveof 0.05ofwt.%. Therefore,
the splitting the strength.
tensile optimum The concentration of GOettoal.be[57]
study of Wang mixed in RS-UHPC
revealed that the is
0.05 wt.%,
splitting fromstrength
tensile the perspective of the
of concrete splitting
made fromtensile
recycledstrength. The study
sand depends onofthe
Wang et al. [57]
strength revealed
of the new
that the
mortar, thesplitting
strengthtensile
of thestrength
recycledofaggregate,
concrete made from
and the recycled
bonding sand depends
strength of these twoon the strength ofin
components the
new mortar, the strength of the recycled aggregate, and the bonding
the concrete. Adding GO can increase the degree of cement hydration and improve the microstructure strength of these two
ofcomponents
cement-based inmaterials
the concrete.
[50].Adding
Therefore,GOadding
can increase
GO to the degree can
RS-UHPC of cement
improve hydration and of
the strength improve
new
the microstructure
mortar in the concreteofand cement-based
enhance its materials [50]. Therefore,
splitting tensile strength. adding GO to RS-UHPC can improve
the strength of new mortar in the concrete and enhance its splitting tensile strength.
 Nanomaterials 2020, 10, x FOR PEER REVIEW 10 of 18
Nanomaterials 2020, 10, 1718 10 of 17

20
1.16

Splitting tensile strength/MPa

0.59
0.75
15 0.94

10

0
RU0 RU25 RU50 RU75
Different kinds of UHPC
Figure 7. The splitting tensile strength of RS-UHPC with different contents of GO.
Figure 7. The splitting tensile strength of RS-UHPC with different contents of GO.
3.2.4. Elastic Modulus
3.2.4. Elastic Modulus
The elastic moduli of the RS-UHPC samples containing different concentrations of GO are shown
in FigureThe8. Similar to the trends
elastic moduli of theobserved
RS-UHPC for the tensile containing
samples strength, flexural strength,
different and splitting
concentrations of tensile
GO are
strength,
shown in Figure 8. Similar to the trends observed for the tensile strength, flexural strength, of
the elastic moduli of the RS-UHPC samples containing GO were generally higher than those and
thesplitting
samplestensile
without GO. This
strength, thesuggests that adding
elastic moduli of theGO can improve
RS-UHPC the elastic
samples modulus
containing of RS-UHPC.
GO were generally
The elasticthan
higher modulithoseofofRU0,
the RU25,
samples RU50, and RU75
without GO. This weresuggests
41.96, 44.41, 47.10, and
that adding GO46.35 GPa, respectively.
can improve the elastic
Thus, the elastic moduli decreased in the order of RU50 > RU75 > RU25
modulus of RS-UHPC. The elastic moduli of RU0, RU25, RU50, and RU75 were 41.96, 44.41, 47.10, > RU0. The elastic moduli of
RU25, RU50,GPa,
and 46.35 and respectively.
RU75 were improved
Thus, the by 5.84%,
elastic 12.25%,
moduli and 10.46%,
decreased in therespectively,
order of RU50 compared
> RU75 >with RU25
that of RU0.
> RU0. TheThese
elasticvalues
modulisuggest
of RU25, that the elastic
RU50, and RU75modulus
were of RS-UHPC
improved bydoes
5.84%, not12.25%,
increase andlinearly
10.46%,
with the increasing
respectively, GO concentration.
compared with that of RU0.The elastic
These modulus of thethat
values suggest RS-UHPC wasmodulus
the elastic maximised at a GO
of RS-UHPC
concentration
does not increase of 0.05 wt.%. with
linearly Therefore, the optimum
the increasing concentrationThe
GO concentration. of GO to be
elastic mixed in
modulus ofRS-UHPC
the RS-UHPC is
0.05 wt.%, from the perspective of the elastic modulus. The elastic modulus
was maximised at a GO concentration of 0.05 wt.%. Therefore, the optimum concentration of GO to of the RS-UHPC sample
prepared
be mixed in in
this study ranged
RS-UHPC is 0.05from
wt.%, 41.96
fromtothe47.10 GPa, which
perspective exceeds
of the elasticthe value forThe
modulus. UHPC
elasticprepared
modulus
from recycled
of the RS-UHPC sandsample
[58] and is comparable
prepared to thatranged
in this study for UHPC fromprepared from river
41.96 to 47.10 sand [59],
GPa, which but isthe
exceeds
slightly lower than that of UHPC produced by Aeolian sand [60]. Ordinary
value for UHPC prepared from recycled sand [58] and is comparable to that for UHPC prepared from concrete is generally
composed
river sand of [59],
aggregates, a cement
but is slightly slurry,
lower thanand anof
that interface transition by
UHPC produced zone. The elastic
Aeolian modulus
sand [60]. of
Ordinary
concrete
concretemainly depends
is generally on the elastic
composed moduli of these
of aggregates, threeslurry,
a cement components
and an[61]. The addition
interface transition of zone.
GO can The
improve
elastic the degreeofofconcrete
modulus cement hydration
mainly dependsand enhance
on thethe microstructure
elastic of cement-based
moduli of these materials
three components [50].
[61]. The
Furthermore,
addition of the GO addition
can improveof GOthe candegree
provide of acement
denserhydration
interfacialand transition
enhance layer
theinmicrostructure
cement-based of
materials [51]. Therefore,
cement-based materialsadding GO can enhance
[50]. Furthermore, thethe resistance
addition of GOof RS-UHPC
can provideto elastic deformation.
a denser interfacial
transition layer in cement-based materials [51]. Therefore, adding GO can enhance the resistance of
RS-UHPC to elastic deformation.
 Nanomaterials 2020, 10, x FOR PEER REVIEW 11 of 18
Nanomaterials 2020, 10, 1718 11 of 17

50 3.32 1.19 1.30

2.88

Elastic modulus/MPa
40

30

20

10

0
RU0 RU25 RU50 RU75
Different kinds of UHPC
Figure 8. The elastic modulus of RS-UHPC with different contents of GO.
Figure 8. The elastic modulus of RS-UHPC with different contents of GO.
As indicated by the aforementioned results, adding GO to the UHPC sample at a concentration
of 0.075As indicated
wt.% (higherbythan
the aforementioned
the optimum GOresults,
content)adding GOa to
yielded the UHPC
slight sample
degradation inat a concentration
the mechanical
of 0.075 wt.%
properties, (higher than
as compared with the
the optimum
optimum GO content)
level. This is yielded a slight
consistent degradation
with the results of in the mechanical
Wang et al. [62].
properties, as compared with the optimum level. This is consistent with the results
It can be seen from Figures 3 and 4 that the porosity and the most probable pore size of RU75 were of Wang et al.
[62]. than
higher It canthat
be of
seen fromrespectively,
RU50, Figures 3 and 4 that the
suggesting porosity
that and the most
the microstructure probable
of RU75 waspore
worse size
thanof that
RU75
were higher than that of RU50, respectively, suggesting that the microstructure
of RU50. When the GO is mixed in UHPC at a high concentration, the large specific surface area andof RU75 was worse
thethan thatintermolecular
strong of RU50. When the GO
forces is van
(e.g., mixedderinWaals
UHPC at a associated
force) high concentration, the large
with GO cause it tospecific surface
agglomerate.
area and the the
Consequently, strong
GO intermolecular forces (e.g.,
becomes less dispersed vanRS-UHPC.
in the der WaalsTherefore,
force) associated
adding with
GO toGO causeatitato
UHPC
agglomerate. Consequently, the GO becomes less dispersed in the RS-UHPC. Therefore,
concentration higher than the optimum value slightly degrades all the mechanical properties of UHPC. adding GO
to UHPC at a concentration higher than the optimum value slightly degrades all the mechanical
3.3. Durabilityof UHPC.
properties

3.3.1.
3.3. Chloride-Ion
Durability Penetration Resistance
The chloride-ion penetration resistance of RS-UHPC samples containing different concentrations
3.3.1.
of GO is Chloride-Ion Penetration
shown in Figure Resistance
9. Generally, the chloride-ion migration coefficients of the RS-UHPC
samples Thecontaining GO were
chloride-ion smaller thanresistance
penetration those of theof RS-UHPC
RS-UHPC samples withoutcontaining
samples GO. This indicates
different
that adding GO can reduce the chloride-ion migration coefficient of RS-UHPC.
concentrations of GO is shown in Figure 9. Generally, the chloride-ion migration coefficients of the The chloride-ion
migration coefficients of RU0, RU25, RU50,smaller
and RU75 −12 , 1.08 × 10−12 , 1.02 × 10−12 ,
× 10RS-UHPC
RS-UHPC samples containing GO were thanwere
those1.16
of the samples without GO.
and 1.05 × 10 −12 m 2 s−1 , respectively. Thus, the migration coefficients
This indicates that adding GO can reduce the chloride-ion migration coefficient decreased in theoforder of RU0The
RS-UHPC. >
RU25 > RU75 > RU50. This sequence suggests that the chloride-ion migration coefficient
chloride-ion migration coefficients of RU0, RU25, RU50, and RU75 were 1.16 × 10 , 1.08 × 10 , 1.02 −12 of UHPC does
−12

not decrease
× 10 −12, and linearly
1.05 × 10with
−12 mthe2 s−1,increasing GOThus,
respectively. concentration. The chloride-ion
the migration migrationincoefficients
coefficients decreased the order of
ofRU0
RU25, RU50, and RU75 were reduced by 6.90%, 12.07%, and 9.48%, respectively,
> RU25 > RU75 > RU50. This sequence suggests that the chloride-ion migration coefficient compared withof
that of RU0.
UHPC doesTherefore, adding
not decrease GO towith
linearly RS-UHPC can significantly
the increasing improve itsThe
GO concentration. chloride-ion
chloride-ion penetration
migration
resistance.
coefficients This
of finding
RU25, RU50, is consistent with were
and RU75 the results
reduced of Guo et al. [63].
by 6.90%, The chloride-ion
12.07%, migration
and 9.48%, respectively,
coefficient
compared of with
the RS-UHPC
that of RU0. was minimised
Therefore, when
adding GOGO wastomixed
RS-UHPCin the can
UHPC at a concentration
significantly improveofits
0.05 wt.%. Thus,
chloride-ion the optimum
penetration concentration
resistance. of GOistoconsistent
This finding be mixedwithin RS-UHPC
the resultsis of
0.05
Guowt.%, from
et al. [63].the
The
perspective
chloride-ion of migration
the chloride-ion migration
coefficient coefficient.
of the RS-UHPC wasThe chloride-ion
minimised when migration coefficient
GO was mixed in theofUHPC
the
RS-UHPC sample prepared
at a concentration of 0.05 wt.%. in this study
Thus, theisoptimum
significantly smaller than
concentration of GOthattopreviously
be mixed in reported
RS-UHPC for is
UHPC [64]. Because
0.05 wt.%, from theGO can enhance
perspective thechloride-ion
of the pore structure [65] and
migration improve the
coefficient. Themicrostructure [50] of
chloride-ion migration
cement-based materials, the number of chloride ions transported through the capillary
coefficient of the RS-UHPC sample prepared in this study is significantly smaller than that previously channels is
reduced
reported withfor
theUHPC
addition of GO.
[64]. Furthermore,
Because GO canadding
enhance GO can
the improve the pore [65]
pore structure structure
and and volumethe
improve
stability of the cement-based
microstructure [50] of cement-basedmaterial;materials,
accordingly,the the number
number of large ions
of chloride porestransported
in the cement-based
through the
capillary channels is reduced with the addition of GO. Furthermore, adding GO can improve the pore
 Nanomaterials 2020, 10, x FOR PEER REVIEW 12 of 18

structure and volume stability of the cement-based material; accordingly, the number of large 12
Nanomaterials 2020, 10, 1718
pores
of 17
in the cement-based material is reduced with the introduction of GO. Consequently, the migration of
chloride ions is impeded by GO. In summary, adding GO to RS-UHPC improves its resistance to
material ischloride-ion
reduced withpenetration.
the introduction of GO. Consequently, the migration of chloride ions is impeded
by GO. In summary, adding GO to RS-UHPC improves its resistance to chloride-ion penetration.

1.5

0.033
Chloride-ion migration
2 -1
1.2 0.053 0.078

coefficient/×10 m ⋅s
0.027
-12
0.9

0.6

0.3

0.0
RU0 RU25 RU50 RU75
Different kinds of UHPC
Figure 9. The chloride-ion migration coefficient of RS-UHPC with different contents of GO.
Figure 9. The chloride-ion migration coefficient of RS-UHPC with different contents of GO.
3.3.2. Freeze–Thaw Resistance
3.3.2. Freeze–Thaw Resistance
The mass loss rates of RS-UHPC samples containing different concentrations of GO are shown in
The mass loss rates of RS-UHPC samples containing different concentrations of GO are shown
Figure 10. Increasing the number of freeze–thaw cycles resulted in a higher mass loss rate for all the
in Figure 10. Increasing the number of freeze–thaw cycles resulted in a higher mass loss rate for all
different the
types of RS-UHPC
different types of samples.
RS-UHPC Generally, the massthe
samples. Generally, loss rateloss
mass of rate
RS-UHPC
of RS-UHPCcontaining GO GO
containing was
lower than that of UHPC without GO. This suggests that GO can improve the
was lower than that of UHPC without GO. This suggests that GO can improve the freeze–thaw freeze–thaw resistance
of RS-UHPC. The mass
resistance loss rates
of RS-UHPC. of the
The massRS-UHPC
loss rates samples preparedsamples
of the RS-UHPC in this study decreased
prepared in the
in this study
followingdecreased
order: RU0 > RU25
in the > RU75
following > RU50.
order: RU0 >The
RU25 mass loss rate
> RU75 of the
> RU50. TheRS-UHPC
mass losswasrate minimised
of the RS-UHPCwhen
was minimised
GO was mixed in the UHPC when atGOa was mixed in the
concentration of UHPC at a concentration
0.05 wt.%. Therefore, the of optimum
0.05 wt.%. concentration
Therefore, the
of GO to optimum
be mixedconcentration
in RS-UHPC of GO to bewt.%,
is 0.05 mixedfrom
in RS-UHPC is 0.05 wt.%,
the perspective of from the perspective
the mass loss rate.ofThe
the mass
mass
loss rate. The mass loss rates of RU0, RU25, RU50, and RU75 after 300 freeze–thaw
loss rates of RU0, RU25, RU50, and RU75 after 300 freeze–thaw test cycles were 0.79%, 0.67%, 0.44%, test cycles were
0.79%, 0.67%, 0.44%, and 0.55%, respectively. As indicated by these results, the mass loss of the RS-
and 0.55%, respectively. As indicated by these results, the mass loss of the RS-UHPC prepared in this
UHPC prepared in this study during the 300 freeze–thaw test cycles was almost negligible. This
study during the2020,
Nanomaterials 30010,freeze–thaw test cycles was almost negligible. This finding is consistent
x FOR PEER REVIEW with
13 of 18
finding is consistent with previously reported freeze–thaw experimental results for ordinary UHPC
previously reported
[64]. freeze–thaw experimental results for ordinary UHPC [64].

1.0
RU0
0.8 RU25
RU50
Mass loss ratio/%

RU75
0.6

0.4

0.2

0.0
0 60 120 180 240 300
Number of cycles/time
Figure 10. The mass loss ratio of RS-UHPC with different contents of GO.
Figure 10. The mass loss ratio of RS-UHPC with different contents of GO.

The relative dynamic elastic moduli of the RS-UHPC samples containing different
concentrations of GO are shown in Figure 11. The relative dynamic elastic modulus of the RS-UHPC
samples decreased with the increasing number of freeze–thaw cycles. Generally, the relative dynamic
elastic modulus of the RS-UHPC containing GO was higher than that of the RS-UHPC without GO.
Thus, adding GO to RS-UHPC can improve its freeze–thaw resistance. The relative dynamic elastic
moduli of the RS-UHPC samples prepared in this study decreased in the following order: RU50 >
RU75 > RU25 > RU0. These results indicate that the relative dynamic elastic modulus of the RS-UHPC
 0 60 120 180 240 300
Number of cycles/time

Figure 10. The mass loss ratio of RS-UHPC with different contents of GO.
Nanomaterials 2020, 10, 1718 13 of 17
The relative dynamic elastic moduli of the RS-UHPC samples containing different
concentrations of GO are shown in Figure 11. The relative dynamic elastic modulus of the RS-UHPC
The relative dynamic elastic moduli of the RS-UHPC samples containing different concentrations
samples decreased with the increasing number of freeze–thaw cycles. Generally, the relative dynamic
of GO are shown in Figure 11. The relative dynamic elastic modulus of the RS-UHPC samples decreased
elastic modulus of the RS-UHPC containing GO was higher than that of the RS-UHPC without GO.
with the increasing number of freeze–thaw cycles. Generally, the relative dynamic elastic modulus of
Thus, adding GO to RS-UHPC can improve its freeze–thaw resistance. The relative dynamic elastic
the RS-UHPC containing GO was higher than that of the RS-UHPC without GO. Thus, adding GO to
moduli of the RS-UHPC samples prepared in this study decreased in the following order: RU50 >
RS-UHPC can improve its freeze–thaw resistance. The relative dynamic elastic moduli of the RS-UHPC
RU75 > RU25 > RU0. These results indicate that the relative dynamic elastic modulus of the RS-UHPC
samples prepared in this study decreased in the following order: RU50 > RU75 > RU25 > RU0.
was maximised at a GO concentration of 0.05 wt.%. Thus, the optimum concentration of GO to be
These results indicate that the relative dynamic elastic modulus of the RS-UHPC was maximised at a
mixed in RS-UHPC is 0.05 wt.%, from the perspective of the relative dynamic elastic modulus. After
GO concentration of 0.05 wt.%. Thus, the optimum concentration of GO to be mixed in RS-UHPC
300 freeze–thaw test cycles, the relative dynamic elastic moduli of RU0, RU25, RU50, and RU75 were
is 0.05 wt.%, from the perspective of the relative dynamic elastic modulus. After 300 freeze–thaw
95.85%, 96.34%, 97.38%, and 96.57%, respectively. These results indicate that the reduction in the
test cycles, the relative dynamic elastic moduli of RU0, RU25, RU50, and RU75 were 95.85%, 96.34%,
relative dynamic elastic modulus was almost negligible for the RS-UHPC after 300 freeze–thaw test
97.38%, and 96.57%, respectively. These results indicate that the reduction in the relative dynamic
cycles, which is consistent with the results of Karim et al. [64].
elastic modulus was almost negligible for the RS-UHPC after 300 freeze–thaw test cycles, which is
consistent with the results of Karim et al. [64].

100

99
elastic modulus/(%)

RU0
Relative dynamic

98 RU25
RU50
RU75
97

96

95
0 60 120 180 240 300
Number of cycles/(time)
Figure 11. The
Figure relative
11. The dynamic
relative elastic
dynamic modulus
elastic of RS-UHPC
modulus withwith
of RS-UHPC different contents
different of GO.
contents of GO.

According to the previously reported criterion [45], the freeze–thaw cycling test was terminated
once either of the following two conditions was satisfied: (1) the total mass loss rate reached 5%, and (2)
the relative dynamic elastic modulus was reduced to <60% of the initial value. As indicated by the
measurements collected in this study, neither of the conditions was met for RS-UHPC during the 300
freeze–thaw test cycles. Therefore, the RS-UHPC prepared in this study had excellent freeze–thaw
resistance. Because the presence of GO can provide a denser microstructure and interfacial transition
zone in cement-based materials [50,51], adding GO to RS-UHPC can improve its freeze–thaw resistance.
Furthermore, GO can enhance the pore structure of cement-based materials [65]. Thus, the water
transport in RS-UHPC is impeded during freeze–thaw cycling after the addition of GO. Hence, adding
GO to RS-UHPC can improve its freeze–thaw resistance.
The chloride-ion penetration resistance and freeze–thaw resistance of the RS-UHPC were both
degraded slightly when GO was mixed in the UHPC at a concentration of 0.075 wt.% (exceeding the
optimum concentration). This performance degradation is attributed to the same factors responsible
for the degradation of the mechanical properties, i.e., adding GO to UHPC at a high concentration
results in the agglomeration of GO.
In summary, the RS-UHPC samples prepared in this study had excellent mechanical properties
and durability. According to a comprehensive analysis of the microstructure, mechanical properties,
and durability of UHPC, the optimum concentration of GO to be mixed in UHPC was determined to
be 0.05 wt.%. The results indicated that adding GO can enhance the pore structure of RS-UHPC, which
Nanomaterials 2020, 10, 1718 14 of 17

was a major factor contributing to the improved mechanical properties and durability of the RS-UHPC
prepared in this study.
It should be highlighted that the mechanical properties and durability of RS-UHPC were
significantly enhanced due to the incorporation of a small quantity of GO, because the degree of cement
hydration of RS-UHPC was increased, and the pore structure of RS-UHPC was improved. However, the
mechanical properties and durability of RS-UHPC were slightly decreased, when the GO exceeded the
optimum concentration, because of the poor ability of GO to be dispersed in RS-UHPC. The optimum
concentration of GO to be mixed in RS-UHPC was determined to be 0.05 wt.%, considering the effects
of GO on the microstructure, mechanical properties, and durability of RS-UHPC.

4. Conclusions
A systematic study was performed to analyse the effects of GO additives on the mechanical
properties and durability of RS-UHPC at different mixing concentrations. Furthermore, the effects
of GO on the microstructure of RS-UHPC were investigated. According to the results, the following
major conclusions are drawn.

(1) The porosity of the RS-UHPC ranged between 2.19% and 2.47%. Adding GO to the RS-UHPC
reduced the porosity by 4.45–11.34%. The most probable pore size of the RS-UHPC was reduced
with the GO addition. These findings indicate that adding GO to RS-UHPC can improve its
pore structure.
(2) The tensile strength of the RS-UHPC ranged between 156.21 and 182.50 MPa. Adding GO to the
RS-UHPC enhanced its tensile strength by 8.24–16.83%. This result indicates that adding GO to
RS-UHPC can improve its tensile strength.
(3) The flexural strength of the RS-UHPC ranged between 15.89 and 20.12 MPa. Adding GO to the
RS-UHPC enhanced its flexural strength by 11.26–26.62%. This result suggests that adding GO to
RS-UHPC can improve its flexural strength.
(4) The splitting tensile strength of the RS-UHPC ranged between 13.37 and 17.32 MPa. Adding GO
to the RS-UHPC enhanced its splitting tensile strength by 15.63–29.54%. This result indicates that
adding GO to RS-UHPC can improve its splitting tensile strength.
(5) The elastic modulus of the RS-UHPC ranged between 41.96 and 47.10 GPa. Adding GO to
RS-UHPC increased its elastic modulus by 5.84–12.25%. This result suggests that adding GO to
RS-UHPC can enhance its resistance to elastic deformation.
(6) The chloride-ion migration coefficient of the RS-UHPC ranged between 1.02 × 10−12 and
1.16 × 10−12 m2 ·s−1 . Adding GO to the RS-UHPC reduced its chloride-ion migration coefficient
by 6.90–12.07%. This result indicates that adding GO to RS-UHPC can improve its resistance to
chloride-ion penetration.
(7) The mass loss rate and relative dynamic elastic modulus of the RS-UHPC after 300 freeze–thaw
test cycles were approximately 0.44–0.79% and 95.85–97.38%, respectively. These results indicate
that adding GO to RS-UHPC can improve its freeze–thaw resistance.
(8) According to the comprehensive analysis of the effects of GO on the microstructure, mechanical
properties, and durability of RS-UHPC, the optimum concentration of GO to be mixed in RS-UHPC
was determined to be 0.05 wt.%.

Author Contributions: Conceptualization, H.C. and Y.Z.; data curation, F.W. and L.W.; funding acquisition,
H.C.; investigation, T.F. and L.W.; methodology, H.C. and Y.Z.; resources, F.W.; visualization, T.F. and D.W.;
writing—original draft, H.C.; writing—review and editing, F.W. and D.W. All authors have read and agreed to the
published version of the manuscript.
Funding: This research was funded by [the National Natural Science Foundation of China] grant number
[51808294], [the Natural Science Foundation of the Education Department of Jiangsu Province] grant number
[18KJB430017], [the Youth Fund from State Key Laboratory of High Performance Civil Engineering Materials]
grant number [2019CEM011], and [the Open Fund from Jiangsu Key Laboratory of Construction Materials] grant
number [CM2018-01].
Nanomaterials 2020, 10, 1718 15 of 17

Conflicts of Interest: The authors declare no conflict of interest.

References
1. Pyo, S.; Abate, S.Y.; Kim, H.K. Abrasion resistance of ultra high performance concrete incorporating coarser
aggregate. Constr. Build. Mater. 2018, 165, 11–16. [CrossRef]
2. Tian, H.; Zhou, Z.; Wei, Y.; Wang, Y.; Lu, J. Experimental investigation on axial compressive behavior of
ultra-high performance concrete (UHPC) filled glass FRP tubes. Constr. Build. Mater. 2019, 225, 678–691.
[CrossRef]
3. Zhao, S.; Sun, W. Nano-mechanical behavior of a green ultra-high performance concrete. Constr. Build. Mater.
2014, 63, 150–160. [CrossRef]
4. Tam, V.W.Y.; Soomro, M.; Evangelista, A.C.J. A review of recycled aggregate in concrete applications
(2000–2017). Constr. Build. Mater. 2018, 172, 272–292. [CrossRef]
5. Kumar, B.M.V.; Ananthan, H.; Balaji, K.V.A. Experimental studies on utilization of coarse and finer fractions
of recycled concrete aggregates in self compacting concrete mixes. J. Build. Eng. 2017, 9, 100–108. [CrossRef]
6. Xiao, J.; Li, W.; Fan, Y.; Huang, X. An overview of study on recycled aggregate concrete in China (1996–2011).
Constr. Build. Mater. 2012, 31, 364–383. [CrossRef]
7. Kisku, N.; Joshi, H.; Ansari, M.; Panda, S.K.; Nayak, S.; Dutta, S.C. A critical review and assessment for
usage of recycled aggregate as sustainable construction material. Constr. Build. Mater. 2017, 131, 721–740.
[CrossRef]
8. Silva, R.V.; de Brito, J.; Dhir, R.K. Establishing a relationship between modulus of elasticity and compressive
strength of recycled aggregate concrete. J. Clean. Prod. 2016, 112, 2171–2186. [CrossRef]
9. Fan, C.; Huang, R.; Hwang, H.; Chao, S. Properties of concrete incorporating fine recycled aggregates from
crushed concrete wastes. Constr. Build. Mater. 2016, 112, 708–715. [CrossRef]
10. McGinnis, M.J.; Davis, M.; de la Rosa, A.; Weldon, B.D.; Kurama, Y.C. Strength and stiffness of concrete with
recycled concrete aggregates. Constr. Build. Mater. 2017, 154, 258–269. [CrossRef]
11. Thomas, J.; Thaickavil, N.N.; Wilson, P.M. Strength and durability of concrete containing recycled concrete
aggregates. J. Build. Eng. 2018, 19, 349–365. [CrossRef]
12. Kou, S.; Poon, C. Long-term mechanical and durability properties of recycled aggregate concrete prepared
with the incorporation of fly ash. Cem. Concr. Compos. 2013, 37, 12–19. [CrossRef]
13. Kanellopoulos, A.; Nicolaides, D.; Petrou, M.F. Mechanical and durability properties of concretes containing
recycled lime powder and recycled aggregates. Constr. Build. Mater. 2014, 53, 253–259. [CrossRef]
14. Guo, H.; Shi, C.; Guan, X.; Zhu, J.; Ding, Y.; Ling, T.C.; Zhang, H.; Wang, Y. Durability of recycled aggregate
concrete—A review. Cem. Concr. Compos. 2018, 89, 251–259. [CrossRef]
15. Limbachiya, M.C.; Meddah, M.S.; Ouchagour, Y. Use of recycled concrete aggregate in fly-ash concrete.
Constr. Build. Mater. 2012, 27, 439–449. [CrossRef]
16. Sim, J.; Park, C. Compressive strength and resistance to chloride ion penetration and carbonation of recycled
aggregate concrete with varying amount of fly ash and fine recycled aggregate. Waste Manag. 2011, 31,
2352–2360. [CrossRef]
17. Saravanakumar, P.; Dhinakaran, G. Durability characteristics of recycled aggregate concrete. Struct. Eng.
Mech. 2013, 47, 701–711. [CrossRef]
18. Adessina, A.; Fraj, A.B.; Barthelemy, J.F.; Chateau, G.; Garnier, D. Experimental and micromechanical
investigation on the mechanical and durability properties of recycled aggregates concrete. Cem. Concr. Res.
2019, 126, 105900. [CrossRef]
19. Otsuki, W.N.; Miyazato, S. Influence of recycled aggregate on interfacial transition zone, strength, choride
penetration and carbonatation of concrete. J. Mater. Civ. Eng. 2003, 15, 443–551. [CrossRef]
20. Abbas, A.; Fathifazl, G.; Isgor, O.; Razaqpur, A.; Fournier, B.; Foo, S. Durability of recycled aggregate concrete
designed with equivalent volume method. Cem. Concr. Compos. 2009, 31, 555–563. [CrossRef]
21. Rao, M.C.; Bhattacharyya, S.K.; Barai, S.V. Influence of field recycled coarse aggregate on properties of
concrete. Mater. Struct. 2011, 44, 205–220.
22. Jiang, J.; Zhou, W.; Gao, Y.; Wang, L.; Wang, F.; Chu, H.; Xu, G.; Vandevyvere, B.; Sierens, Z.; Li, J.
Feasibility of manufacturing ultra-high performance cement-based composites (UHPCCs) with recycled
sand: A preliminary study. Waste Manag. 2019, 83, 104–112. [CrossRef] [PubMed]
Nanomaterials 2020, 10, 1718 16 of 17

23. Nunes, C.; Slizkova, Z.; Stefanidou, M.; Nemecek, J. Microstructure of lime and lime-pozzolana pastes with
nanosilica. Cem. Concr. Res. 2016, 83, 152–163. [CrossRef]
24. Nili, M.; Ehsani, A. Investigating the effect of the cement paste and transition zone on strength development
of concrete containing nanosilica and silica fume. Mater. Des. 2015, 75, 174–183. [CrossRef]
25. Li, W.; Huang, Z.; Cao, F.; Sun, Z.; Shah, S.P. Effects of nano-silica and nano-limestone on flowability and
mechanical properties of ultra-high-performance concrete matrix. Constr. Build. Mater. 2015, 95, 366–374.
[CrossRef]
26. Menga, T.; Yub, Y.; Qiana, X.; Zhana, S.; Qiana, K. Effect of nano-TiO2 on the mechanical properties of cement
mortar. Constr. Build. Mater. 2012, 29, 241–245. [CrossRef]
27. Alessandro, A.D.; Rallini, M.; Ubertini, F.; Materazzi, A.L.; Kenny, J.M. Investigation on scalable fabrication
procedures for self-sensing carbon nanotube cement-matrix composites for SHM applications. Cem. Concr.
Compos. 2016, 65, 200–213. [CrossRef]
28. Zhang, W.; Wu, P.; Zhang, Y.; Zeng, W.; Yang, F. The effect of carbon nanotubes on the mechanical and
damping properties of macro-defect-free cements. Sci. Eng. Compos. Mater. 2020, 27, 28–40.
29. Stynoski, P.; Mondal, P.; Marsh, C. Effects of silica additives on fracture properties of carbon nanotube and
carbon fiber reinforced Portland cement mortar. Cem. Concr. Compos. 2015, 55, 232–240. [CrossRef]
30. Chu, H.; Ge, X.; Jiang, J.; Tang, J.; Zhang, Z. Mechanical properties and ablation behaviour of nuclear
sacrificial materials containing graphene sulfonate nanosheets. Constr. Build. Mater. 2018, 191, 69–79.
[CrossRef]
31. Chu, H.; Jiang, J.; Sun, W.; Zhang, M. Effects of graphene sulfonate nanosheets on mechanical and thermal
properties of sacrificial concrete during high temperature exposure. Cem. Concr. Compos. 2017, 82, 252–264.
[CrossRef]
32. Chu, H.; Jiang, J.; Sun, W.; Zhang, M. Mechanical and thermal properties of graphene sulfonate nanosheet
reinforced sacrificial concrete at elevated temperatures. Constr. Build. Mater. 2017, 153, 682–694. [CrossRef]
33. Peng, H.; Ge, Y.; Cai, C.S.; Zhang, Y.; Liu, Z. Mechanical properties and microstructure of graphene oxide
cement-based composites. Constr. Build. Mater. 2019, 194, 102–109. [CrossRef]
34. Luo, J.; Chen, S.; Li, Q.; Liu, C.; Gao, S.; Zhang, J.; Guo, J. Influence of graphene oxide on the mechanical
properties, fracture toughness and microhardness of recycled concrete. Nanomaterials 2019, 9, 325. [CrossRef]
35. Newell, M.; Garcia-Taengua, E. Fresh and hardened state properties of hybrid graphene oxide/nanosilica
cement composites. Constr. Build. Mater. 2019, 221, 433–442. [CrossRef]
36. Du, S.; Wu, J.; Alshareedah, O.; Shi, X. Nanotechnology in cement-based materials: A review of durability,
modeling, and advanced characterization. Nanomaterials 2019, 9, 1213. [CrossRef]
37. Sun, H.; Ling, L.; Ren, Z.; Memon, S.A.; Xing, F. Effect of graphene oxide/graphene hybrid on mechanical
properties of cement mortar and mechanism investigation. Nanomaterials 2020, 10, 113. [CrossRef]
38. Zandiatashbar, A.; Lee, G.-H.; An, S.J.; Lee, S.; Mathew, N.; Terrones, M.; Hayashi, T.; Picu, C.R.; Hone, J.;
Koratkar, N. Effect of defects on the intrinsic strength and stiffness of grapheme. Nat. Commun. 2014, 5, 4186.
[CrossRef]
39. García-Taengua, E.; Sonebi, M.; Hossain, K.M.A.; Lachemi, M.; Khatib, J. Effects of the addition of nanosilica
on the rheology, hydration and development of the compressive strength of cement mortars. Compos. Part
B Eng. 2015, 81, 120–129. [CrossRef]
40. Sonebi, M.; García-Taengua, E.; Hossain, K.M.A.; Khatib, J.; Lachemi, M. Effect of nanosilica addition on the
fresh properties and shrinkage of mortars with fly ash and superplasticizer. Constr. Build. Mater. 2015, 84,
269–276. [CrossRef]
41. Lv, S.H.; Ma, Y.J.; Qiu, C.C. Effect of graphene oxide nanosheets of microstructure and mechanical properties
of cement composites. Constr. Build. Mater. 2013, 49, 121–127. [CrossRef]
42. Wei, Y.; Zhang, Y.; Chai, J.; Wu, G.; Dong, Z. Experimental investigation of rectangular concrete-filled fiber
reinforced polymer (FRP)-steel composite tube columns for various corner radii. Compos. Struct. 2020, 244,
112311. [CrossRef]
43. Zhang, Y.; Wei, Y.; Bai, J.; Wu, G.; Dong, Z. A novel seawater and sea sand concrete filler FRP-carbon steel
composite tube column: Concept and behaviour. Compos. Struct. 2020, 246, 112421. [CrossRef]
44. GB/T 50081-2002. Standard for Test Method of Mechanical Properties on Ordinary Concrete; China Architecture
and Building Press: Beijing, China, 2002. (In Chinese)
Nanomaterials 2020, 10, 1718 17 of 17

45. GB/T 50082-2009. Standard for Test Methods of Long-Term Performance and Durability of Ordinary Concrete; China
Architecture and Building Press: Beijing, China, 2009. (In Chinese)
46. Wang, L.; Ju, S.; Chu, H.; Liu, Z.; Yang, Z.; Wang, F.; Jiang, J. Hydration process and microstructure evolution
of low exothermic concrete produced with urea. Constr. Build. Mater. 2020, 248, 118640. [CrossRef]
47. Chu, H.; Jiang, J.; Sun, W.; Zhang, M. Mechanical and physicochemical properties of ferro-siliceous concrete
subjected to elevated temperatures. Constr. Build. Mater. 2016, 122, 743–752. [CrossRef]
48. Chu, H.; Jiang, J.; Sun, W.; Zhang, M. Thermal behavior of siliceous and ferro-siliceous sacrificial concrete
subjected to elevated temperatures. Mater. Des. 2016, 95, 470–480. [CrossRef]
49. Jiang, J.; Zhou, W.; Chu, H.; Wang, F.; Wang, L.; Feng, T.; Guo, D. Design of eco-friendly ultra-high
performance concrete with supplementary cementitious materials and coarse aggregate. J. Wuhan Univ.
Technol.-Mater. Sci. Ed. 2019, 34, 1350–1359. [CrossRef]
50. Lu, Z.; Hou, D.; Ma, H.; Fan, T.; Li, Z. Effects of graphene oxide on the properties and microstructures of the
magnesium potassium phosphate cement paste. Constr. Build. Mater. 2016, 119, 107–112. [CrossRef]
51. Li, X.Y.; Li, C.Y.; Liu, Y.M.; Chen, S.J.; Wang, C.M.; Sanjayan, J.G.; Duan, W.H. Improvement of mechanical
properties by incorporating graphene oxide into cement mortar. Mech. Adv. Mater. Struct. 2018, 25,
1313–1322. [CrossRef]
52. Soliman, N.A.; Tagnit-Hamou, A. Partial substitution of silica fume with fine glass powder in UHPC: Filling
the micro gap. Constr. Build. Mater. 2017, 139, 374–383. [CrossRef]
53. Richardson, I.G.; Girao, A.V.; Taylor, R.; Jia, S. Hydration of water- and alkaliactivated white Portland cement
pastes and blends with low-calcium pulverized fuel ash. Cem. Concr. Res. 2016, 83, 1–18. [CrossRef]
54. Su, Y.; Wu, C.; Li, J.; Li, Z.-X.; Li, W. Development of novel ultra-high performance concrete: From material
to structure. Constr. Build. Mater. 2017, 135, 517–528. [CrossRef]
55. Liu, J.; Han, F.; Cui, G.; Zhang, Q.; Lv, J.; Zhang, L.; Yang, Z. Combined effect of coarse aggregate and fiber
on tensile behavior of ultra-high performance concrete. Constr. Build. Mater. 2016, 121, 310–318. [CrossRef]
56. Meng, W.; Khayat, K.H. Improving flexural performance of ultra-high-performance concrete by rheology
control of suspending mortar. Compos. Part B Eng. 2017, 117, 26–34. [CrossRef]
57. Wang, W.; Liu, Y.; Jiang, L.; Zhao, L.; Li, Z. Effect of physical properties of recycled coarse aggregate on the
mechanical properties of recycled aggregate thermal insulation concrete (RATIC). Constr. Build. Mater. 2018,
180, 229–238. [CrossRef]
58. Dimitriou, G.; Savva, P.; Petrou, M.F. Enhancing mechanical and durability properties of recycled aggregate
concrete. Constr. Build. Mater. 2018, 158, 228–235. [CrossRef]
59. Alsalman, A.; Dang, C.N.; Prinz, G.S.; Hale, W.M. Evaluation of modulus of elasticity of ultra-high
performance concrete. Constr. Build. Mater. 2017, 153, 918–928. [CrossRef]
60. Chu, H.; Wang, F.; Wang, L.; Feng, T.; Wang, D. Mechanical properties and environmental evaluation of
ultra-high-performance concrete with aeolian sand. Materials 2020, 13, 3148. [CrossRef]
61. Wang, Y.; Zhang, H.; Geng, Y.; Wang, Q.; Zhang, S. Prediction of the elastic modulus and the splitting tensile
strength of concrete incorporating both fine and coarse recycled aggregate. Constr. Build. Mater. 2019, 215,
332–346. [CrossRef]
62. Wang, Y.; Yang, J.; Ouyang, D. Effect of graphene oxide on mechanical properties of cement mortar and its
strengthening mechanism. Materials 2019, 12, 3753. [CrossRef]
63. Guo, K.; Miao, H.; Liu, L.; Zhou, J.; Liu, M. Effect of grapheme oxide on chloride penetration resistance of
recycled concrete. Nanotechnol. Rev. 2019, 8, 681–689. [CrossRef]
64. Karim, R.; Najimi, M.; Shafei, B. Assessment of transport properties, volume stability, and frost resistance of
non-proprietary ultra-high performance concrete. Constr. Build. Mater. 2019, 227, 117031. [CrossRef]
65. Yang, H.; Cui, H.; Tang, W.; Li, Z.; Han, N.; Xing, F. A critical review on research progress of graphene/cement
based composites. Compos. Part A Appl. Sci. Manuf. 2017, 102, 273–296. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).