Utilization of urea as an accessible superplasticizer on the Moon for lunar geopolymer mixtures

Abstract

When developing materials for lunar construction, it is essential to minimize the weight of components that have to be brought in from Earth. All necessary ingredients for geopolymers could potentially be sourced on the lunar surface, which is why the material might be an efficient construction material for infrastructure on the moon. Finding a chemical admixture that can be easily obtained on the moon, which can increase the workability while utilizing less water, would be highly beneficial for utilizing lunar regolith geopolymers for lunar 3D printing. Urea can break hydrogen bonds, and therefore reduces the viscosities of many aqueous mixtures. Since urea is the second most abundant component in urine (after water), it is readily available anywhere there are humans. We have therefore explored the possibility of utilizing urea as a chemical admixture for lunar geopolymers. Addition of urea has been compared with polycarboxylate and naphthalene based superplasticizers, and with a control mixture without superplasticizer. When curing the sample containing urea at 80 °C, the initial setting time became longer. The samples containing urea or naphthalene-based superplasticizers could bear heavy weights shortly after mixing, while keeping an almost stable shape. Samples without superplasticizer or containing the polycarboxylate-based admixture were too stiff for mold-shaped formation after casting. Samples containing urea and naphthalene-based admixtures could be used to build up a structure without any noticeable deformation. Initial compressive strength of the samples with urea was higher than for the two other specimens containing superplasticizers, and it continued to rise even after 8 freeze-thaw cycles. Microstructural studies revealed that superplasticizers can influence the formation of additional air voids within the samples.

Full text

Utilization of urea as an accessible superplasticizer on the moon for
lunar geopolymer mixtures
Shima Pilehvar
a
, Marlies Arnhof
b
, Ram
on Pamies
c
, Luca Valentini
d
,
Anna-Lena Kjøniksen
a
,
*
a
Faculty of Engineering, Østfold University College, P.O. Box 700, 1757, Halden, Norway
b
Advanced Concepts Team, ESA European Space Research and Technology Centre, Keplerlaan 1, TEC-SF, 2201AZ, Noordwijk, Netherlands
c
Department of Materials Engineering and Manufacturing, Technical University of Cartagena, Cartagena, Murcia, Spain
d
Department of Geosciences, University of Padua, 35131, Padua, Italy
article info
Article history:
Received 2 July 2019
Received in revised form
24 September 2019
Accepted 4 November 2019
Available online xxx
Handling editor: Panos Seferlis
Keywords:
Geopolymer
Lunar regolith simulant
Urea
Superplasticizer
Lunar construction
abstract
When developing materials for lunar construction, it is essential to minimize the weight of components
that have to be brought in from Earth. All necessary ingredients for geopolymers could potentially be
sourced on the lunar surface, which is why the material might be an efficient construction material for
infrastructure on the moon. Finding a chemical admixture that can be easily obtained on the moon,
which can increase the workability while utilizing less water, would be highly beneficial for utilizing
lunar regolith geopolymers for lunar 3D printing. Urea can break hydrogen bonds, and therefore reduces
the viscosities of many aqueous mixtures. Since urea is the second most abundant component in urine
(after water), it is readily available anywhere there are humans. We have therefore explored the possi-
bility of utilizing urea as a chemical admixture for lunar geopolymers. Addition of urea has been
compared with polycarboxylate and naphthalene based superplasticizers, and with a control mixture
without superplasticizer. When curing the sample containing urea at 80

C, the initial setting time
became longer. The samples containing urea or naphthalene-based superplasticizers could bear heavy
weights shortly after mixing, while keeping an almost stable shape. Samples without superplasticizer or
containing the polycarboxylate-based admixture were too stiff for mold-shaped formation after casting.
Samples containing urea and naphthalene-based admixtures could be used to build up a structure
without any noticeable deformation. Initial compressive strength of the samples with urea was higher
than for the two other specimens containing superplasticizers, and it continued to rise even after 8
freeze-thaw cycles. Microstructural studies revealed that superplasticizers can influence the formation of
additional air voids within the samples.
©2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
With a growing interest in the human exploration of celestial
bodies, establishing in-situ habitats on the lunar surface may
facilitate surveys further away from the Earth by providing an
important extra-terrestrial base (Happel, 1993). However, colo-
nizing the moon poses several problems such as high radiation
levels, extreme temperatures and large temperature fluctuations,
vacuum, and meteoroids (Benvenuti et al., 2013). According to the
National Aeronautics and Space Administration (NASA),
transporting one pound (0.45 kg) of materials into orbit costs
around $10,000 USD (Qiu and Park, 2001). Therefore, to keep the
transportation costs feasible, both NASA and the European Space
Agency (ESA) promote the use of in-situ materials (Leach, 2014).
Utilizing lunar surface materials to fabricate cement/concrete
for in-situ construction has been proposed previously (Buchner
et al., 2018; Cesaretti et al., 2014). Cesaretti et al. (2014) proposed
a D-shape 3D printing process with Sorel cement for construction
on the lunar surface. Unfortunately, the process required substan-
tial amounts of consumables (chemicals and water) to produce the
binder. Buchner et al. (2018) developed a rock-like material by
using phosphoric acid as a liquid binder. For lunar applications,
considerable amounts of water and phosphoric acid would have to
be transported to the lunar surface. However, this material seems to
*Corresponding author.
E-mail address: anna.l.kjoniksen@hiof.no (A.-L. Kjøniksen).
Contents lists available at ScienceDirect
Journal of Cleaner Production
journal homepage: www.elsevier.com/locate/jclepro
https://doi.org/10.1016/j.jclepro.2019.119177
0959-6526/©2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Journal of Cleaner Production xxx (xxxx) xxx
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177

be promising for use on the Martian surface, as phosphoric acid and
water are available on Mars. Since water is a highly valuable
resource on the lunar surface and not readily available (Hauri et al.,
2015), using building materials with high water demand is practi-
cally impossible. Geopolymers consist of silico-aluminates in an
amorphous to semi-crystalline three-dimensional structure. Geo-
polymers exhibit an excellent performance such as quick control-
lable setting and hardening (Hardjito et al., 2008; Li et al., 2004),
high compressive strength (Allahverdi et al., 2016; Ryu et al., 2013),
freeze-thaw resistance (Fu et al., 2011; Sun and Wu, 2013), excellent
durability in sulfate environment and superior resistance to acid
and salt attacks (Bakharev, 2005a, b), high fire resistance and low
thermal conductivity (Cheng and Chiu, 2003; Kong and Sanjayan,
2010), small shrinkage (Ana M. Fernandez-Jimenez and Cecilio,
2006) and adequate radiation shielding (Montes et al., 2015).
Geopolymerization of lunar regolith has therefore been proposed
due to the high amounts of aluminosilicates at a similar ratio as the
main components of traditional geopolymer binders, and the
favorable shielding properties for a crew inside the lunar con-
struction (Montes et al., 2015). Additionally, the presence of alkali
metals on the moon might be used as a source of the alkaline so-
lution for geopolymerization (Matta et al., 2009). Utilization of lu-
nar regolith and alkali metals as the components of geopolymer
composites can thereby facilitate lunar construction without the
need of bringing materials in from the Earth at extreme cost.
Recently, 3D printing technology has been developed in large
scale to allow direct construction of buildings (Buswell et al., 2018;
Ceccanti et al., 2010; Ngo et al., 2018). For lunar construction, 3D
printing minimizes the involvement of human labor in the building
process, reducing health hazards and inefficiencies during con-
struction (Benvenuti et al., 2013). However, in order to utilize 3D
printing to build a layer-by-layer structure, the geopolymer com-
posite should exhibit high workability for extrusion, keep its shape
after printing, high early concrete strength, and optimal setting
time (Le et al., 2012). Fresh geopolymer composites have poor
workability due to the higher viscosity of the alkaline solution.
Improved workability can be obtained by adding extra water to the
mixture. However, water on the moon is not readily available,
therefore adding more water to the geopolymer mixture is not
realistic (Chua and Johnson, 1998). Additionally, extra water will
reduce the compressive strength of geopolymers (Aliabdo et al.,
2016). A better way to achieve a good workability is therefore to
utilize a low dosage of a chemical admixture. For fly ash based
geopolymers, a polycarboxylate-based superplasticizer is preferred
when the calcium ion content is relatively high (Xie and Kayali,
2016), while a naphthalene based superplasticizer is effective for
low amounts of calcium (Jang et al., 2014; Xie and Kayali, 2016).
However, utilizing these superplasticizers in lunar geopolymers
would involve transportation from the Earth at great cost. It is
therefore preferable to find a superplasticizer that would be
available on the moon. The lunar surface is lacking in suitable
materials for use as superplasticizers. However, if we assume hu-
man presence during construction, we also have access to human
waste materials. Human urine contains about 9.3e23.3 g/L urea
(Putnam, 1971). It is well known that urea is capable of breaking
hydrogen bonds, thereby reducing the viscosities of many aqueous
mixtures (Usha and Ramasami, 2002). It is therefore reasonable to
assume that urea might work as a superplasticizer to reduce the
water demand of the geopolymers.
The objective of this research is to investigate whether urea
obtained from humans on the moon can be utilized as a chemical
admixture to reduce the water needed to obtain a good workability
of lunar regolith geopolymers for 3D printing. The effectiveness of
urea as a superplasticizer will be compared to more traditional
naphthalene and polycarboxylate based geopolymer
superplasticizers. The aim is to develop a lunar regolith geopolymer
mixture that can meet the severe curing conditions on the lunar
surface (extreme temperature cycles, little available water, and
vacuum) without the extreme cost of importing any of the com-
ponents from the Earth.
2. Experimental
2.1. Materials
DNA-1 lunar regolith simulant developed for ESA as a chemical
and mineralogical analog to the lunar mare regolith, has been
produced by Dini Engineering srl for Monolite UK ltd in the pre-
mises of Cascine di Buti (Pisa), Italy. The chemical composition of
this lunar regolith simulant is given in Table 1. The chemical
composition of regolith is similar to class F fly ash (Norcem, Ger-
many), which is used for terrestrial geopolymers. However, the
percentage of crystalline phase of regolith simulant is 75% whereas
37% crystalline phase was observed for fly ash. As can be seen from
Table 1, most of the components of the lunar regolith simulant is
within the range found in lunar regolith samples. Fig. 1 show
microscopic images (SEM) of the supplied lunar regolith simulant
and fly ash (FA) to evaluate the shape difference of the particles. The
particle shape of the lunar regolith simulant is angular, simulating
the shape of lunar regolith which has been bombardment by me-
teorites in the lunar environment (Colwell et al., 2007). FA is
generally spherical in shape. The size distributions of the lunar
regolith simulant and FA determined by low angle laser light
scattering (Malvern Mastersizer 2000) are displayed in Fig. 2. The
regolith particles have a higher particle size distribution than FA.
However, about 30% of the cumulative volume of lunar regolith
simulant particles overlap with the size range of FA.
Sodium hydroxide pellets purchased from VWR, Norway, was
used for preparing the alkaline solution. In this study, urea (in
powder form) supplied by VWR, Norway, was used as an accessible
lunar superplasticizer to reduce the amount of water needed to
achieve the desired the workability of lunar geopolymer (LG)
mixtures. Additionally, naphthalene and polycarboxylate based
chemical admixtures (in powder form) were utilized for compari-
son, since these have been shown to work for other geopolymers
previously (Xie and Kayali, 2016). These chemical admixtures are
named FLUBE CR 100 F (a poly-naphthalene sulfonate polymer) and
SUPLA PDP 2 SA (a polycarboxylic modified polymer), both pro-
vided by Bozzetto Group, Italy.
2.2. Mixing, casting and curing procedures
For all LG mixtures, a sodium hydroxide solution 12 M (480g/L)
was selected as an alkaline solution. The alkaline solution was
prepared one day in advance to dissolve the NaOH pellets
completely in the water and to release the exothermic reaction
heat. Laboratory trials of workability and buildability showed that
an alkaline solution to regolith ratio of 0.35 and a chemical
admixture dosage corresponding to 3% of the lunar regolith mass
were optimal. By considering the water limitation and transporting
cost for lunar constructions, this step was repeated several times to
gain the minimum amounts of alkaline solution and super-
plasticizer, while keeping the workability and strength at accept-
able levels. Accordingly, while about 64wt% of the alkaline solution
is water, the ratio of water to the total mass of geopolymer solids
(lunar regolith simulant, NaOH pellets and chemical admixture) is
0.19. The compositions of the mixtures are shown in Table 2.
For specimen preparation, regolith, alkaline solution and one
type of chemical admixture were mixed together for 10 min to
reach a homogenous and uniform mixture. After mixing, the
S. Pilehvar et al. / Journal of Cleaner Production xxx (xxxx) xxx2
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177

mixture was cast into molds at a size of 3 33cm
3
. A vibration
machine was used for 1 min to remove air trapped inside the
specimens. After casting, LG samples were pre-cured in a heating
chamber with a temperature of 80

C for 6 h. After demolding, the
samples were exposed to freeze-thaw cycles. In a single freeze-
thaw cycle, the specimens were first placed in the heating cham-
ber at 80 ±2

C for 48 h and then the samples were left in a freezer
at a temperature of 80 ±2

C for the next 48 h. The samples were
subjected to 0, 2, 4 and 8 freeze-thaw cycles.
2.3. Testing methods
2.3.1. Shape deformation and layer-by-layer buildability
The shape deformation was calculated to examine the effect of
different chemical admixtures on the shape retention of LG. After
molding 200 g mixture with a conical mold, immediately after
demolding, 1000 g and 200 0 g metal weights were put on top of the
fresh samples to evaluate the percentage of sample shape defor-
mation under pressure.
Shape deformation ¼D
after
D
before
D
before
100
where D
before
and D
after
are the bottom diameter of the sample
before and after demolding, respectively. D
before
is 7 cm for all
samples, which is the inside diameter of conical mold bottom. After
the shape deformation evaluation, only mixtures with useable
workability and with high shape retention were selected for the
layer-by-layer buildability test. The buildability of LG mixtures was
Table 1
Chemical composition of DNA-1 lunar regolith simulant and fly ash class F, compared to the composition of lunar regolith samples (the highest and lowest values of each
component from 19 analyzed lunar samples is shown).
Chemical Lunar regolith simulant DNA-1 (wt. %) Fly ash class F (wt. %) Lunar regolith soil samples range (wt. %) (McKay et al., 1991)
SiO
2
47.79 ±0.05 50.83 ±0.04 40.6e48.1
Al
2
O
3
19.16 ±0.07 23.15 ±0.06 12.0e28.0
Fe
2
O
3
8.75 ±0.01 6.82 ±0.01 4.7e19.8
CaO 8.28 ±0.03 6.87 ±0.02 10.3e15.8
K
2
O 3.52 ±0.02 2.14 ±0.01 0.04e0.55
Na
2
O 4.38 ±0.03 1.29 ±0.01 0.31e0.70
MgO 1.86 ±0.01 1.70 ±0.01 5.6e13.0
TiO
2
1.00 ±0.01 1.01 ±0.01 0.47e8.4
Fig. 1. SEM images of (a) lunar regolith simulant and (b) fly ash.
10
-2
10
-1
10
0
10
1
10
2
10
3
0
20
40
60
80
100
Particle size
(
µm
)
Regolith
Fly ash
C
umulative volume
(
vol
%)
Fig. 2. Particle size distributions of lunar regolith simulant and fly ash.
Table 2
Composition of mixture design for LG. The percentage of chemical admixtures is calculated according to the mass of lunar regolith simulant.
Name of mixture NaOH
(aq)
/Regolith Chemical admixture
W/O 0.35 e
U 0.35 Urea - 3%
C 0.35 polycarboxylate based (SUPLA PDP 2 SA) - 3%
N 0.35 naphthalene based(FLUBE CR 100 F) - 3%
S. Pilehvar et al. / Journal of Cleaner Production xxx (xxxx) xxx 3
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177

examined by means of a high-pressure syringe pump (Fusion 6000,
Chemyx, Inc.) with a constant pump rate of 20 ml/min (Fig. 3).
2.3.2. Setting time
To characterize the initial and final setting times of LG after
adding three different chemical admixtures, a Vicat needle test was
performed by a manual Vicat needle apparatus in accordance with
EN 196-3. After placing the fresh mixture in the conical metal mold,
the mold was transferred to the heating chamber at 80

C. Setting
time measurements were carried out at 80

C with an interval of
15 min. The initial setting time is the time when the needle pene-
tration is less than 39 mm whereas the final setting time is the
moment when the needle penetrates the sample to a depth of
0.5 mm.
2.3.3. Compressive strength
The compressive strength tests for LG specimens after 0, 2, 4,
and 8 freeze-thaw cycles were performed at 20

C in accordance
with EN 12190, using a digital compressive strength test machine
(Form þTest Machine).
2.3.4. FTIR
LG specimens after 0, 4 and 8 freeze-thaw cycles were charac-
terized via Fourier-Transform Infrared Spectroscopy (FTIR). FTIR
analysis was conducted with a spectrometer (PerkinElmer Spec-
trum BX) in transmittance mode from 4000 to 500 cm
1
. Samples
were dried before measurement at 80

C for 12 h.
2.3.5. X-ray tomography
X-ray microtomography (XCT) scans were performed on drill
cores of samples after 0 and 4 freeze-thaw cycles. These cylindrical
samples have a diameter of 1 cm and a height of approximately
2 cm. The analyses were performed using a Skyscan 1172 XCT
scanner, with an energy of 100 kV, 0.9 s acquisition time and 0.3

rotation step. Tomographic reconstruction was performed using the
FDK algorithm (Feldkamp et al., 1984). The reconstructed images
consist of 1200 vertically stacked cross-sections, with a linear pixel
size of 6.5
m
m. Image processing, including binary segmentation
and particle analysis, was performed by the ImageJ software
(Schneider et al., 2012), by which the size of not interconnected
pores was measured. In order to estimate the accuracy and error
associated with the image processing procedure, a random distri-
bution of spherical pores with known particle size was generated.
After addition of gaussian noise, the image was binarized and the
total porosity and sphere radius was compared with those relative
to the original image. The average errors associated with porosity
and pore radius were 4% and 15% respectively.
3. Results and discussion
3.1. Shape deformation and layer-by-layer buildability
One of the most challenging features in 3D concrete printing is
that the materials should be able to retain their shape after
extruding. Fig. 4 depicts shape deformation of fresh mixtures after
placing a 1 kg weight on top of the samples. As can be seen from
Fig. 4a and c, fresh W/O and C samples retain a stable shape after
loading with a 1 kg weight. However, there are many fractures in
the samples, due to the stiffness of the mixtures during mixing and
molding. To have a low shape deformation, a mixture should have a
good workability and high viscosity. However, the W/O and C
samples are too stiff for casting, causing the formation of hetero-
geneous and fractured structures. It has previously been shown
that a polycarboxylate-based superplasticizer (such as the one used
in sample C) is often the best choice for fly ash class C, due to the
strong bonds between the positively charged calcium and the
negatively charged polycarboxylate (Xie and Kayali, 2016). Since
the lunar regolith simulant is similar to fly ash class F, this super-
plasticizer is not optimal for improving the workability and flow-
ability of the mixture. From Fig. 4b and d it is evident that fresh U
and N mixtures are castable after molding with a smooth surfaces
with none (U sample) or few (N sample) fractures. These samples
also retain their shape under a 1 kg external load. Accordingly,
mixture U and N were selected for further studies.
In the next step, a 2 kg weight was placed on top of the fresh U
and N mixtures for shape deformation evaluation (Fig. 5). The
percentage of shape deformation was11.4% and 13% for the U and N
mixtures, respectively. Accordingly, the sample containing urea
Fig. 3. A high-pressure syringe pump is utilized for 3D-printing the samples.
Fig. 4. Sample retention after loading a 1 kg weight over (a) mixture without any
admixture (W/O sample), (b) mixture containing 3% urea (U sample), (c) mixture
containing 3% polycarboxylate-based admixture (C sample), and (d) and mixture
containing 3% naphthalene-based admixture (N sample). The arrows show fractures
and disruptions formed during molding.
S. Pilehvar et al. / Journal of Cleaner Production xxx (xxxx) xxx4
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177

retains its shape better after loading with a weight that is 10 times
that of the 200 g sample.
In Fig. 6, layer-by-layer buildability was measured by means of a
syringe pump for selected mixtures (U and N) to see how many
layers is possible to stack without any deformation of the layers or
collapse of the structure, and without utilizing a rest time between
layers. Due to the narrow extruding tube (1 cm in diameter), only 4
and 5 layers of filament were built up for the U and F samples,
respectively. As can be seen from Fig. 6, there are slight changes of
the thickness of each layer. However, all the layers are nicely
vertically stacked, and remain steady without any obvious defor-
mation and collapse. Accordingly, urea and the naphthalene-based
admixture can contribute positively to the buildability of the LG.
3.2. Setting time
Fig. 7 shows the effect of different superplasticizers on the initial
and final setting times of lunar geopolymer at 80

C. The three
superplasticizers influence the setting times in different ways.
Incorporating 3% urea (U) postpones both initial and final setting
times in comparison with the sample without any chemical
admixture (W/O). For both mixtures containing 3%
polycarboxylate-based admixture (C) and naphthalene-based
admixture (N), there is a moderate delay of the initial setting
time, while a much longer final setting time was observed for the N
mixture. In 3D printing, the time after mixing until the fresh ma-
terial losses the workability for extruding is called the open time.
The open time is always earlier than the initial setting time (Panda
and Tan, 2018). Therefore, a longer initial setting time will help
maintain a continuous flow during pumping, and prevent the
material from becoming too hard in the 3D printer. For good shape
retention during layer-by-layer buildability, the fresh LG mixture
should gain high early strength after extrusion to tolerate subse-
quent layers resting on top (Ngo et al., 2018). Acceptable final
setting time is therefore an advantage to allow the LG layers to be
loaded onto the previous layers without shape deformation.
Accordingly, the mixture containing urea exhibits the best initial
and final setting times for 3D printing.
3.3. Compressive strength
The compressive strength of the samples after 0, 2, 4 and 8
freeze-thaw cycles is shown in Fig. 8. The sample without any
chemical admixture exhibits the highest compressive strength after
8 freeze-thaw cycles (32 MPa). As mentioned previosly, for 3D
printing high early strength is needed af ter extrusion to tolerate the
weight of subsequent layers loaded on top of the sample. The C and
N samples show a low early strength of about 1.7 MPa, which is not
optimal for 3D printing. Interestingly, the U sample exhibited a
relatively high initial compressive strength (13 MPa) after precur-
ing at 80

C.
During the freeze-thaw cycles, two opposing mechanisms are
affecting the compressive strength. The freeze thaw cycles are ex-
pected to reduce the compressive strength due to expansion of
water within the samples when it freezes, which may cause frac-
tures within the samples. At the same time, the geopolymerization
reaction within the samples continues, which will increase the
compressive strength. This results in a continous slight increase in
compressive strength for the U sample. For the three other samples
the competing mechanisms causes a variation in compressive
strength over the freeze thaw cycles, but with an overall strong
increase from 0 to 8 cycles. The compressive strength requirement
for lunar construction is 1/6 of the requirement of a similar struc-
ture on Earth (Montes et al., 2015), which is normally around
25e40 MPa (Kosmatka et al., 2002). Except for the C and N samples
before the freeze-thaw cycles, all samples are well above this limit
(>7 MPa).
After 8 freeze-thaw cycles, the U and N samples have approxi-
mately the same compressive strength, while the sample without
superplasticizer has a much higher compressive strength. The C-
sample has a large variation between the three tested cubes, but
seems to have a strength somewhere between the other samples.
Interestingly, the samples with the lowest compressive strength
after 8 freeze-thaw cycles (U and N) are the same ones that are best
suited for 3D-printing (Figs. 4 and 6). In order to explore the
mechanisms behind the differences, FT’IR and X-ray micro-
tomography experiments have been performed.
Fig. 5. Sample deformation after loading a 2 kg weight over (a) mixture containing 3%
urea (U sample), (b) mixture containing 3% naphthalene-based admixture (N sample).
Fig. 6. Layer-by-layer buildability of (a) mixture containing 3% urea (U sample), (b) mixture containing 3% naphthalene-based admixture (N sample).
S. Pilehvar et al. / Journal of Cleaner Production xxx (xxxx) xxx 5
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177

3.4. FTIR
Fig. 9 presents the FTIR spectra of the four different LGs after 0,
4, and 8 freeze-thaw cycles. The main band is centered around
975 cm
1
and corresponds to asymmetric stretching vibration of
SieOeT(T¼Si or Al) (Abdullah et al., 2012; Rees et al., 2007a;
2007b). This peak indicates the degree of amorphous aluminosili-
cate gel phase due to dissolution of the regolith in the alkaline
solution (Abdullah et al., 2012). After 0 freeze-thaw cycles, this peak
is strongest for the W/O and U samples, which is in agreement with
the higher compressive strengths of these samples at this stage
(Fig. 8). After 4 freeze-thaw cycles (Fig. 10b), the C sample exhibited
the highest amount of geopolymer gel formation resulting in a
broader and deeper peak. In contrast, the sample containing urea
(U) exhibited the smallest amount of gel which resulted the lowest
compressive strength after 4 freeze-thaw cycles (Fig. 8). Interest-
ingly, after 8 freeze-thaw cycles, the U sample revealed similar
intensity and depth of this peak as the C sample, which indicate a
continuing formation of geopolymeric products after 8 freeze-thaw
cycles.
The broad IR band at around 2358 cm
1
(after 4 and 8 freeze-
thaw cycles) is related to the bending vibration of HeOeH bonds
from weakly bound water molecules (Abdullah et al., 2012). Since
this peak is not evident before the freeze-thaw cycles, this is
probably from water adsorbed onto the samples during the freeze-
thaw process.
The small peaks at 1428 cm
1
has been attributed to stretching
vibrations of OeCeO bonds, which suggests that sodium bicar-
bonate has been formed due to atmospheric carbonation of the
high alkaline NaOH solution (Abdullah et al., 2012). There is a
number of smaller peaks in the 630-760 cm
1
region, which are
contributed to aluminosilicate ring and cage structures (Rees et al.,
2007a, 2007b).
3.5. X-Ray tomography
Typical 2D X-ray micro-tomography cross-sectional slices ob-
tained from the W/O, U, C, and N samples after 0 and 4 freeze-thaw
cycles are shown in Fig. 10, where cracks and air voids are displayed
in dark color (low or no X-ray attenuation). More microcracks are
evident in the U matrix after exposure to freeze-thaw cycles. This
indicates that the microcracks generated by the freeze-thaw cycles
can contribute to the deterioration of the U sample. This is probably
W/O U C N
0
2
4
6
8
Initial setting time
Final setting time
Mix
tu
r
e
S
etting time
(
h
)
Fig. 7. The initial and final setting times of W/O (without any admixture), U (urea), C
(polycarboxylate-based admixture) and N (naphthalene-based admixture) lunar geo-
polymer mixtures.
02468
0
5
10
15
20
25
30
35
4
0
W/O
U
C
N
No. of freeze-thaw c
y
cles
Compressive strength (MPa)
Fig. 8. Compressive strength of lunar geopolymer versus the number of freeze-thaw
cycles containing 0 wt% of superplasticizer (W/O) and 3 wt% of urea (U), 3%
polycarboxylate-based admixture (C), and 3% naphthalene-based admixture (N). The
samples were pre-cured for 6 h at 80 C before starting the freeze thaw cycles.
2500 2000 1500 1000 50
0
70
80
90
100
110
2500 2000 1500 1000 50
0
70
80
90
100
110
2500 2000 1500 1000 50
0
70
80
90
100
110
1428
975
0 freeze-thaw cycles
Transmittance (%)
W/O
U
C
N
(a)
760-630
1428
2358
975
760-630
4 freeze-thaw cycles
(b)
Transmittance (%)
1428
2358
975
760-630
(c)
Transmittance (%)
Wavenumber
(
cm-1
)
8 freeze-thaw cycles
Fig. 9. FTIR spectra of lunar Geopolymer specimens containing 0 wt% superplasticizer
(W/O) and 3 wt% of urea (U), 3% polycarboxylate-based admixture (C), and 3%
naphthalene-based admixture (F)(a) after 0 freeze-thaw cycles and (b) after 4 freeze-
thaw cycles, and (c) after 8 freeze-thaw cycles.
S. Pilehvar et al. / Journal of Cleaner Production xxx (xxxx) xxx6
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177

a contributing factor to the lower compressive strength of this
sample after the freeze thaw cycles (Fig. 8).
Fig. 11 illustrates the air voids (in red color) in the 3D images of
the samples. In order to quantify the differences, the pore diameter
distributions are plotted in Fig. 12 and the volume %, mean diam-
eter, and number of air voids per mm
3
is shown in Fig. 13.Figs. 11
and 13a illustrates that before the freeze-thaw cycles the inclu-
sion of urea (U) and the polycarboxylate-based admixture (C) led to
a substantial increase of porosity in comparison with the sample
without any admixture (W/O). This is expected to affect the me-
chanical performance of the materials, and might be contributing
to the lower compressive strength of the U and C samples
compared to the W/O sample (Fig. 8). A lower flowability or higher
viscosity is expected to increase the air content of the mixtures
(Valcuende et al., 2012). Accordingly, before the freeze-thaw cycles
the better workability of the N sample leads to smaller (Figs. 12 and
13b ) and lower amounts (Fig. 13a) of air voids compared to the C
sample. When aqueous solutions of urea are heated up to 80

C,
NH
3
and H
2
gasses can be released (Jones and Rollinson, 2013).
These gasses might contribute to creating more voids in the U
sample compared to the other samples, as observed in Figs. 11 and
13a. Despite its poor workability, the W/O sample has a low amount
of air voids before the freeze-thaw cycles compared to the other
samples (Figs. 11 and 13 a). This might indicate that the inclusion of
superplasticizers enhances air void formation which provide
durability for the sample in freezing-thawing situations
(Ła
zniewska-Piekarczyk, 2012). In addition, the mean diameter
(Fig. 13b) and pore size distribution (Fig. 12) illustrates that larger
air voids are formed before the freeze-thaw cycles for the W/O and
C samples compared to the N sample. This might suggest that
poorer workability of the samples causes the formation of larger air
voids (Sakai et al., 2006; Valcuende et al., 2012).
After 4 freeze-thaw cycles, the total volume of air voids in-
creases significantly for the W/O and C samples while there is little
change for the U and N samples (Figs. 11 and 13a). Accordingly, the
Fig. 10. 2D X-ray tomography images of (a) W/O sample e0 cycles, (A) W/O sample e
4 cycles, (b) U sample e0 cycles, (B) U sample e4 cycles, (c) C sample e0 cycles, (C) C
sample e4 cycles, (d) N sample e0 cycles, and (D) N sample e4 cycles. The arrows
show the microcracks in the sample matrix. The field of view is approximately 1 cm.
Fig. 11. 3D X-ray-tomography rendering of samples containing 0 wt% superplasticizer
(W/O) and 3 wt% urea (U), 3 wt% polycarboxylate-based admixture (C), and 3 wt%
naphthalene-based admixture (N) after 0 and 4 freeze-thaw cycles.
S. Pilehvar et al. / Journal of Cleaner Production xxx (xxxx) xxx 7
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177

differences in compressive strength after the freeze-thaw process
(Fig. 8) does not seem to be directly correlated with the volume of
new air voids formed. This suggests that the effect of the super-
plasticizers on the geopolymerization reaction plays a larger role
than the formation of additional air voids after the freeze-thaw
cycles. As can be seen from Fig. 12, the freeze-thaw cycles seems
to shift the size distributions towards larger air voids. This can both
be due to an enlargement of existing voids by expansion of
entrapped water, and caused by several smaller voids expanding
into a common much larger void. The latter effect seems to be
dominant for the N-sample, which has significantly fewer voids
after the freeze-thaw cycles (Fig. 13c).
4. Conclusions
The possibility of utilizing urea (U) as a chemical admixture for
lunar geopolymers was studied. The results were compared with
polycarboxylate (C) and naphthalene-based (N) superplasticizers,
and with a control mixture without superplasticizer (W/O). The
influence of these three different powder admixtures was investi-
gated on the physical, mechanical and microstructural properties of
lunar regolith geopolymer. The following conclusions can be drawn
from this work:
1. The W/O and C samples are too stiff for casting and cause the
formation of fractured structures, whereas fresh U and N mix-
tures are castable after molding with smooth surfaces with no
(U sample) or few (N sample) fractures. These U and N mixtures
also retain their shape under external loads and can be pumped
and built layer by layer.
2. Incorporating 3% urea postpones both initial and final setting
times in comparison with mixture without any admixture. For
3D printing applications, a longer initial setting time will help
maintain a continuous flow during pumping, and prevent the
material from becoming too hard. For good shape retention, the
fresh LG mixture should gain high early strength after extrusion
to tolerate subsequent layers resting on top. In addition, urea has
acceptable values of initial and final setting times.
3. The U sample exhibits a relatively high initial compressive
strength (13 MPa) before freeze-thaw cycle which is more
practical for 3D printing purposes than the other samples.
Although, the U and N samples have the lowest compressive
strength after 8 freeze-thaw cycles, they are best suited for 3D-
printing.
4. FTIR show that the W/O and U samples have the highest amount
of geopolymer gel formation after 0 freeze-thaw cycles, whereas
10 100 100
0
0.00
0.02
0.04
0.06
0
.
08
10 100 100
0
0.00
0.02
0.04
0.06
0.08
10 100 100
0
0.00
0.02
0.04
0.06
0.08
10 100 100
0
0.00
0.02
0.04
0.06
0.08
Volume fraction (%)
0 freeze-thaw cycles
4 freeze-thaw cycles
(a) W/O
Volume fraction (%)
(b) U
Volume
f
raction (%
)
(c) C
(d) N
Volume fraction (%)
Pore diameter
(
m
)
Fig. 12. Differential size distributions of air voids inside the W/O, U, C and N samples,
obtained from image analyses of the X-ray-tomography images after 0 freeze-thaw
cycles and 4 freeze-thaw cycles.
W/OUCN
0
2
4
6
8
10
Volume (%)
0 freeze-thaw cycles
4 freeze-thaw cycles
(a)
W/OUCN
0
25
50
75
Mean diameter ( m)
(b)
W
/O U C
N
0
25
50
75
100
125
Number/mm
-3
(c)
Fig. 13. (a) Volume % of air voids, (b) mean diameter of air voids, and (c) number of air
voids per mm
3
for the W/O, U, C and N samples obtained from image analyses of the X-
ray-tomography images after 0 freeze-thaw cycles and 4 freeze-thaw cycles.
S. Pilehvar et al. / Journal of Cleaner Production xxx (xxxx) xxx8
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177

the C sample has the strongest peak after 4 freeze-thaw cycles.
After 8 freeze-thaw cycles, the U sample exhibited similar
intensiveness and depth of the peak as the C sample, which
indicate a continuing formation of amorphous aluminosilicate
gel after 8 freeze-thaw cycles.
5. 2D X-ray images show more microcracks in the U matrix after
freeze-thaw cycles. This might be the cause of the lower
compressive strength of the U sample. The microstructural
studies also illustrate that adding superplasticizers can enhance
air void formation, and thereby provide durability for the sam-
ple in freezing-thawing situations. Mean diameter and pore size
distributions show that larger air voids are formed before the
freeze-thaw cycles for the W/O and C samples. This can be due to
the poorer workability of the samples causing the formation of
larger air voids.
6. Overall urea exhibits promising properties as a superplasticizer
for 3D-printing of lunar geopolymers.
Further studies are needed in order to assess how these lunar
regolith geopolymers will behave under the severe lunar condi-
tions, with a vacuum that can cause the volatile components to
evaporate, and large temperature fluctuations which might cause
crack formation. The ability of the geopolymers formed under these
conditions to withstand meteorite bombardment, and to shield
against high radiation levels should also be evaluated. In addition,
3D-printing the lunar regolith geopolymers under lunar conditions
is expected to be much more challenging than it would be under
normal atmospheric conditions. Additional work regarding these
aspects are currently in progress.
Acknowledgement
This work is the result of an Ariadna study, a joint collaborative
research project between the Faculty of Engineering at Østfold
University College and the Advanced Concepts Team (ACT) of the
European Space Agency (ESA). We would like to thank Rino Nilsen,
Rudi Yi Xu and Andreas Erichsen for their assistance. The authors
acknowledge Fundaci
on S
eneca Agencia de Ciencia y Tecnología de
la Regi
on de Murcia “Ayuda a las Unidades y Grupos de Excelencia
Científica de la Regi
on de Murcia (Programa S
eneca 2014)" (Grant #
19877/GERM/14), for financial support.
References
Abdullah, M.M.A.B., Hussin, K., Bnhussain, M., Ismail, K.N., Yahya, Z., Abdul Razak, R.,
2012. Fly ash-based geopolymer lightweight concrete using foaming agent. Int.
J. Mol. Sci. 13 (6), 7186e7198.
Aliabdo, A.A., Abd Elmoaty, A.E.M., Salem, H.A., 2016. Effect of water addition,
plasticizer and alkaline solution constitution on fly ash based geopolymer
concrete performance. Constr. Build. Mater. 121, 694e703.
Allahverdi, A., Pilehvar, S., Mahinroosta, M., 2016. Influence of curing conditions on
the mechanical and physical properties of chemically-activated phosphorous
slag cement. Powder Technol. 288, 132e139.
Ana, M., Fernandez-Jimenez, A.P., Cecilio, L.-H., 2006. Engineering properties of
alkali-activated fly ash concrete. ACI Mater. J. 103 (2), 106e112 .
Bakharev, T., 2005a. Durability of geopolymer materials in sodium and magnesium
sulfate solutions. Cement Concr. Res. 35 (6), 1233e1246.
Bakharev, T., 2005b. Resistance of geopolymer materials to acid attack. Cement
Concr. Res. 35 (4), 658e670.
Benvenuti, S., Ceccanti, F., De Kestelier, X., 2013. Living on the moon: topological
optimization of a 3D-printed lunar shelter. Nexus Netw. J. 15 (2), 285e302.
Buchner, C., Pawelke, R.H., Schlauf, T., Reissner, A., Makaya, A., 2018. A new plane-
tary structure fabrication process using phosphoric acid. Acta Astronaut. 143,
272e284.
Buswell, R.A., de Silva, W.L., Jones, S., Dirrenberger, J., 2018. 3D printing using
concrete extrusion: a roadmap for research. Cement Concr. Res. 112, 37e49.
Ceccanti, F., Dini, E., De Kestelier, X., Colla, V., Pambaguian, L., 2010. In: 3D Printing
Technology for a Moon Outpost Exploiting Lunar Soil, 61st International
Astronautical Congress, pp. 1e9. Prague, CZ, IAC-10-D3.
Cesaretti, G., Dini, E., De Kestelier, X., Colla, V., Pambaguian, L., 2014. Building
components for an outpost on the Lunar soil by means of a novel 3D printing
technology. Acta Astronaut. 93, 430e450.
Cheng, T.W., Chiu, J.P., 2003. Fire-resistant geopolymer produced by granulated
blast furnace slag. Miner. Eng. 16 (3), 205e210.
Chua, K.M., Johnson, S.W., 1998. Martian and lunar cold region soil mechanics
considerations. J. Aerosp. Eng. 11 (4), 138e147.
Colwell, J., Batiste, S., Hor
anyi, M., Robertson, S., Sture, S., 2007. Lunar surface: dust
dynamics and regolith mechanics. Rev. Geophys. 45 (2).
Feldkamp, L., Davis, L., Kress, J., 1984. Practical cone-beam algorithm. JOSA A 1 (6),
612e619.
Fu, Y., Cai, L., Yonggen, W., 2011. Freezeethaw cycle test and damage mechanics
models of alkali-activated slag concrete. Constr. Build. Mater. 25 (7),
3144e3148 .
Happel, J.A., 1993. Indigenous materials for lunar construction. Appl. Mech. Rev. 46
(6), 313e325.
Hardjito, D., Cheak, C.C., Ing, C.H.L., 2008. Strength and setting times of low calcium
fly ash-based geopolymer mortar. Mod. Appl. Sci. 2 (4), p3.
Hauri, E.H., Saal, A.E., Rutherford, M.J., Van Orman, J.A., 2015. Water in the Moon’s
interior: truth and consequences. Earth Planet. Sci. Lett. 409, 252e264.
Jang, J., Lee, N., Lee, H., 2014. Fresh and hardened properties of alkali-activated fly
ash/slag pastes with superplasticizers. Constr. Build. Mater. 50, 169e176.
Jones, J.M., Rollinson, A.N., 2013. Thermogravimetric evolved gas analysis of urea
and urea solutions with nickel alumina catalyst. Thermochim. Acta 565, 39e45.
Kong, D.L.Y., Sanjayan, J.G., 2010. Effect of elevated temperatures on geopolymer
paste, mortar and concrete. Cement Concr. Res. 40 (2), 334e339.
Kosmatka, S.H., Kerkhoff, B., Panarese, W.C., MacLeod, N.F., McGrath, R.J., 2002.
Design and Control of Concrete Mixtures, Seventh Canadian Edition. Cement
Association of Canada, p. 151.
Le, T.T., Austin, S.A., Lim, S., Buswell, R.A., Gibb, A.G., Thorpe, T., 2012. Mix design
and fresh properties for high-performance printing concrete. Mater. Struct. 45
(8), 1221e1232.
Leach, N., 2014. 3D printing in space. Architect. Des 84 (6), 108e113.
Li, Z., Ding, Z., Zhang, Y., 2004. Development of sustainable cementitious materials.
In: Proceedings of International Workshop on Sustainable Development and
Concrete Technology, pp. 55e76. Beijing, China.
Matta, M., Smith, S., Baumgardner, J., Wilson, J., Martinis, C., Mendillo, M., 2009. The
sodium tail of the Moon. Icarus 204 (2), 409e417.
McKay, D.S., Heiken, G., Basu, A., Blanford, G., Simon, S., Reedy, R., French, B.M.,
Papike, J., 1991. The Lunar Regolith, Lunar Sourcebook. Camebridge University
Press, Houston, Texas.
Montes, C., Broussard, K., Gongre, M., Simicevic, N., Mejia, J., Tham, J., Allouche, E.,
Davis, G., 2015. Evaluation of lunar regolith geopolymer binder as a radioactive
shielding material for space exploration applications. Adv. Space Res. 56 (6),
1212e1221.
Ngo, T.D., Kashani, A., Imbalzano, G., Nguyen, K.T., Hui, D., 2018. Additive
manufacturing (3D printing): a review of materials, methods, applications and
challenges. Compos. B Eng. 143, 172e196.
Panda, B., Tan, M.J., 2018. Experimental study on mix proportion and fresh prop-
erties of fly ash based geopolymer for 3D concrete printing. Ceram. Int. 44 (9),
10258e10265.
Putnam, D.F., 1971. Composition and Concentrative Properties of Human Urine.
NASA Contracror Report. NASA CR-1802.
Qiu, Y., Park, K., 2001. Environment-sensitive hydrogels for drug delivery. Adv. Drug
Deliv. Rev. 53 (3), 321e339.
Rees, C.A., Provis, J.L., Lukey, G.C., Van Deventer, J.S., 2007a. In situ ATR-FTIR study of
the early stages of fly ash geopolymer gel formation. Langmuir 23 (17),
9076e9082.
Rees, C.A., Provis, J.L., Lukey, G.C., van Deventer, J.S.J., 2007b. Attenuated total
reflectance Fourier transform infrared analysis of fly ash geopolymer gel aging.
Langmuir 23 (15), 8170e8179.
Ryu, G.S., Lee, Y.B., Koh, K.T., Chung, Y.S., 2013. The mechanical properties of fly ash-
based geopolymer concrete with alkaline activators. Constr. Build. Mater. 47,
409e418.
Sakai, E., Kasuga, T., Sugiyama, T., Asaga, K., Daimon, M., 2006. Influence of super-
plasticizers on the hydration of cement and the pore structure of hardened
cement. Cement Concr. Res. 36 (11), 2049e2053.
Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH Image to ImageJ: 25 years of
image analysis. Nat. Methods 9 (7), 671e675.
Sun, P., Wu, H.-C., 2013. Chemical and freezeethaw resistance of fly ash-based
inorganic mortars. Fuel 111, 740e745.
Usha, R., Ramasami, T., 2002. Effect of hydrogen-bond-breaking reagent (urea) on
the dimensional stability of rat tail tendon (RTT) collagen fiber. J. Appl. Polym.
Sci. 84 (5), 975e982.
Valcuende, M., Parra, C., Marco, E., Garrido, A., Martínez, E., C
anoves, J., 2012. In-
fluence of limestone filler and viscosity-modifying admixture on the porous
structure of self-compacting concrete. Constr. Build. Mater. 28 (1), 122e128.
Xie, J., Kayali, O., 2016. Effect of superplasticizer on workability enhancement of
Class F and Class C fly ash-based geopolymers. Constr. Build. Mater. 122, 36e42.
Ła
zniewska-Piekarczyk, B., 2012. The influence of selected new generation ad-
mixtures on the workability, air-voids parameters and frost-resistance of self
compacting concrete. Constr. Build. Mater. 31, 310e319.
S. Pilehvar et al. / Journal of Cleaner Production xxx (xxxx) xxx 9
Please cite this article as: Pilehvar, S et al., Utilization of urea as an accessible superplasticizer on the moon for lunar geopolymer mixtures,
Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.119177