Full Paper
Preparation and Properties of Natural Sand
Particles Reinforced Epoxy Composites
Gang Sui, Soumen Jana, Amin Salehi-khojin, Sanjay Neema,
Wei-Hong Zhong,* Hui Chen, Qun Huo
An epoxy composite using Cancun natural hydrophobic sand particle as filler material was
fabricated in this study. Three point bending tests demonstrated an enhancement of 7.5 and
8.7% in flexural strength and flexural modulus, respectively, of epoxy composite containing
1 wt.-% sand particles without any chemical treatment involved, compared to the pristine
epoxy. Scanning electron microscopy (SEM) studies revealed that the fracture toughness of the
epoxy matrix was enhanced owing to the presence of sand particles in an epoxy/sand
composite. Through dynamic mechanical analysis (DMA) and thermal mechanical analysis
(TMA) methods, it was found that the storage
modulus (E0 ), glass transition temperature (Tg)
and dimensional stability of the sand particles/
epoxy composites were increased compared to
the pristine epoxy. The friction behavior of
epoxy/sand system reflected that the microstructure of epoxy composites was steady. These
experimental results suggest that Cancun sand,
as a freshly found natural micron porous material, may find promising applications in composite materials.
Introduction
Epoxy resins are used widely due to their good mechanical,
thermal, and electrical properties.[1–4] Many types of
epoxy resins have been developed, including bisphenol-A,
bisphenol-F, aliphatic cyclic, novolac types, etc. To further
G. Sui, S. Jana, A. Salehi-khojin, S. Neema, W.-H. Zhong
Department of Mechanical Engineering and Applied Mechanics,
North Dakota State University, Fargo, ND 58105, USA
Fax: (þ1) 701 231 7139; E-mail: katie.zhong@ndsu.edu
H. Chen, Q. Huo
Center of Nanoscience and Nanotechnology, University of Central
Florida, Orlando, FL 32826, USA
G. Sui
College of Materials Science and Engineering, Beijing University
of Chemical Technology, Beijing 100029, China
Macromol. Mater. Eng. 2007, 292, 467–473
ß 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
strengthen the properties of epoxy resins, the use of an
additional phase has been a common practice. Epoxy
resins modified with inorganic particles such as carbon,
TiO2, SiO2, Al2O3, clay and so on have shown improved
performances.[5–10] For inorganic/organic composites, the
size of particles and the interfacial adhesion have great
effect on the properties of the resin matrix. The welldispersed inorganic fillers in polymer matrices and compatibility between inorganic and organic phases are
important to achieve an overall good performance.[11–14]
The application of nanometer materials to the thermosetting resin for property modification is a promising
channel. Compared to conventional inorganic/polymer
composites that need over 30 wt.-% loading of microscale
fillers, the same level of enhancements may be achieved
with less than 10 wt.-% loading of well-dispersed nanoscale inorganic fillers.[15–17] However, this research is still
DOI: 10.1002/mame.200600479
467
G. Sui et al.
at an embryonic stage.[18–20] It usually takes complicated
procedures to prepare and treat the filler particles in order
to achieve a good dispersion and adhesion between the
filler and polymer matrices. Filler materials that can give
good dispersion and interfacial adhesion between the filler
and polymer matrices directly without sophisticated treatment protocols are highly desired for composite material
fabrication. The efficient, convenient, and natural additives are more attractive for their economical value and
broad applications.
In this work, we fabricated an epoxy composite using a
kind of natural sand as the filler material. A recent study
from our group discovered that beach sand from the area
of Cancun, Mexico, is a macro- to mesoporous material
with hydrophobic surface property. A polymer composite
made from mixing Cancun sand particles with poly(methyl
methacrylate) showed a good interfacial adhesion as
revealed from scanning electron microscopy (SEM) studies,
owing to the hydrophobic chemical structure on the sand
particle surface and the porous morphology.[21] These
results prompted us to surmise that this porous natural
sand may be a good filler material for epoxy composite
materials. In this study, we prepared an epoxy/sand resin
matrix composite with 1 wt.-% loading of sand particles
and conducted a systematic testing on the mechanical,
thermal, and other related properties of the composite.
These studies revealed that indeed, the sand–epoxy
composite material exhibits a comprehensive improvement in different properties.
Experimental Part
mixture under vacuum conditions, and curing via a thermal cycle
(120 8C for 1 h and 160 8C for an additional 4 h). For preparation of
the epoxy/sand composites, 1 wt.-% of sand particles were added
into the epoxy resin after 24 h of milling and the mixture was
stirred at 60 8C for 3 h. To disperse sand in the epoxy matrix, a
Brasonic1 Ultrasonic cleaner 1210 (Branson Ultrasonics Corporation) was used as a low power sonication. Then the curing agent
was added and low power sonication was used for further mixing
of epoxy/sand at room temperature for 2 h. Subsequently, the
epoxy/sand/curing agent mixture was degassed under vacuum
condition, followed by the same thermal curing schedule as that of
pristine epoxy.
Characterizations and Measurements
A Q-test machine (MTS Co.) was used for the three point bending
test according to the ASTM-D790 (standard test methods for
flexural properties of unreinforced and reinforced plastics and
electrical insulating materials). The speed of the crosshead was
1 mm min1. The specimens were from both pristine epoxy and
epoxy/sand composites and their sizes were 2.95 0.03 mm
(thickness) 12.52 0.20 mm (width) 63.83 0.05 mm (length).
However, the span distance between supports in the three point
bending fixture was 48 mm. The data found from the tests were
recorded and processed. Flexural stress was calculated from the
following equation [ASTM D790]:
sf ¼
3PL
2bd2
(1)
where P is the peak load applied, L the length of span, and b and d
are the width and thickness of the specimens respectively.
Flexural modulus was calculated from the expression [ASTM
D790]
Materials
The diglycidyl ether of bisphenol A (DGEBA, Shell EPON1 828) and
the EPI-CURE Curing Agent W (Miller–Stephenson Chemical
Company, USA) were used to prepare the pristine epoxy resin
and the epoxy/sand composites. The Cancun sand sample was
collected from the beach area near the Gran Melia resort in Cancun,
Mexico. The sand sample was heated at 100 8C under vacuum for
10 h to eliminate residual water. Then the sand was milled for up to
24 h in a ball milling machine. Dry ball milling was undertaken in a
purified argon protected atmosphere on a planetary ball mill. Steel
balls with diameters of 4 and 6 mm were used as grinding medium.
A ball–to-powder weight ratio of 10:1 was selected to ensure a high
efficiency. Ball milling was conducted with the disk revolution
speed of 180 rpm. To minimize the oxidation during milling, we took
1 h interval after each 2 h of milling. After milling, samples were
stored in argon and taken out after 24 h. The milled sand was used
directly to prepare the epoxy/sand composites.
Preparation of the Epoxy/Sand Composites
The pristine epoxy resin was prepared by mixing the epoxy resin
with the curing agent at room temperature for 2 h, degassing the
468
Macromol. Mater. Eng. 2007, 292, 467–473
ß 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ef ¼
L3 m
4bd3
(2)
where m is the slope of the tangent to the initial straightline
portion of the load–deflection curve.
The following equation [ASTM D790] was used to calculate the
strain:
"f ¼
6Dd
L2
(3)
where D is the deflection of specimens under the applied load.
The morphology of the sand particles and fracture interface of
pristine epoxy and composite samples was examined by a JEOL
JSM-6300 model SEM.
The dynamic mechanical analysis (DMA) of the pristine epoxy
and composites was carried out with a Rheometrics Scientific
DMTA 3E dynamic mechanical analyzer at a heating rate of
10 8C min1. The coefficients of thermal expansion (CTEs) of pristine epoxy and composites were obtained with a TA Instruments
2940 thermomechanical analyzer at a heating rate of 10 8C min1
under nitrogen purge.
DOI: 10.1002/mame.200600479
Preparation and Properties of Natural Sand Particles Reinforced . . .
A tribology instrument UTM TESTER was used to measure
friction behavior of both epoxy/sand and epoxy systems. All the
friction tests were carried out at 20–25 8C and a relative humidity
of 40–60%. It was operated under 10 kg load and at 20 rpm speed.
The data presented in the current work are the averages of three
replicate measurements.
Results and Discussion
Microstructure of the Cancun Sand Particles
The diameter of the as-received Cancun natural sand
particles is about 5.5 mm, as revealed from SEM study. The
sizes of the sand particles were reduced effectively using a
ball milling method. The relationship between average
diameter of sands and milling time is shown in Figure 1.
The dimension of the sand particles decreases with increased milling time. There was only a slight change in the
particle size after 18 h of milling time. After 24 h of milling,
the average size of the sand particles had reached to about
1 mm. All the sand particles used in this study were milled
for 24 h.
From the SEM images of the sand particles milled for
24 h [Figure 2(a)], it can be seen that there is a wide
size distribution of sand particles. In the magnified image
shown in Figure 2(b), it is found that the large particles
are actually composed of many smaller nanoparticles.
These large particles are porous and brittle, and can be
broken up when subjected to external force. However, they
are small in number compared to smaller ones. The
statistical size distribution of the sand particles after 24 h
of milling is shown in Figure 3.
Figure 1. The relationship between average diameter of sand
particles and milling time.
Macromol. Mater. Eng. 2007, 292, 467–473
ß 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. The SEM images of the milled sands: (a) low resolution
and (b) high resolution.
Flexural Properties
The flexural properties of pristine epoxy and the sand
particle-filled epoxy composites are summarized in
Table 1. The flexural strength and modulus of the sand
composites were clearly improved compared to that of
pristine epoxy. The enhancement in flexural strength and
modulus of epoxy has reached to 7.5 and 8.7%, respectively. For natural sand particles without any chemical
treatment involved, these levels of enhancement are
significant for a polymer composite material at the loading
of 1 wt.-% inorganic fillers. These results indicate that the
rigid sand particles in epoxy networks directly enhance the
stiffness of composites, allowing a uniform stress distribution in the polymer, and leading to increased flexural
strength and moduli. As the rigidity of sand particles is
greater than that of epoxy resin, it can be expected that
sand particles will assist in improving the mechanical
properties of the composites. Small sand particle with
larger surface area achieve better wetting and adhesion which leads to better reinforcing ability and stiffer
composite system. Sumita et al.[22] studied the effect of
www.mme-journal.de
469
G. Sui et al.
Figure 3. The size distribution of sand particles after 24 h of
milling.
ultrafine silica particles with different diameters on
mechanical properties of polypropylene composites and
found the same kind of behavior with smaller sizes (equal
to or less than the size we used). However, a decrease of
elongation is a typical phenomenon for composites reinforced by inorganic fillers. But in this study, the elongation
of epoxy/sand composites at fracture is almost the same as
that of pristine epoxy. It may be further alleviated by
enhancing the adhesion between inorganic and organic
phases and reducing the inorganic filler dimensions.
Figure 4. Fracture morphology of the epoxy and composites after
three-points bending tests: (a) pristine epoxy; (b) epoxy/sand
composites.
Characterization of Fracture Surface
Figure 4 shows the fracture surface of the pristine epoxy
and epoxy/sand composites after three-points bending
tests. The characteristic fracture features of pristine epoxy
materials are typical of a brittle material [Figure 4(a)] and
thus the resistance to crack propagation is low. For epoxy/
sand composites, river patterns and branches are prominent in the fracture surface [Figure 4(b)]. These new surfaces were created by massive plastic deformation under
flexural loading. More external energy can be dissipated
through these solid patterns compared to pristine epoxy,
leading to increased fracture strength of composites. This
can be explained by two possible mechanisms. One is crack
bowing theory[23,24] which states that when a crack
propagates through a matrix and encounters a nanoparticle, the crack bows around this particle to by-pass the
blockage. This means that deflection of the crack tip occurs
and after passing around multiple particles, the path of the
crack tip appears as a zigzag trajectory. This leads to a
change in the mode of crack opening: mode I to mode I/II
for crack tilting and mode I to mode I/III for crack twisting,
and a high intensity load is required for crack propagation
with this condition. Another mechanism, crack tip
Table 1. The bending properties of pristine epoxy and epoxy/sand composites.
470
sf (Mean W SD)
Ef (Mean W SD)
ef (Mean W SD)
MPa
GPa
%
Pristine epoxy
149.6 W 4.2
10.3 W 0.2
2.5 W 0.2
Epoxy/sand composites
160.8 W 5.2
11.2 W 0.3
2.3 W 0.2
Macromol. Mater. Eng. 2007, 292, 467–473
ß 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/mame.200600479
Preparation and Properties of Natural Sand Particles Reinforced . . .
shielding[23–26] depends on many factors such as plastic
deformation, residual stress fields, microcracking, etc. A
crack tip in a plastic deformation zone faces a compressive
force which opposes further crack opening action and this
resistance in crack opening might face further obstacles
with the introduction of nanoparticles in the matrix.
Therefore, for further crack opening, an increased load is
needed which may pull out the particle from the matrix.
With increased applied load and further crack opening,
sand particles tend to pull out from the matrix. While
pulling, a kind of friction exists between the sand particle and matrix and more energy dissipation occurs in the
system. However, this energy dissipation depends on the
interaction between nanoparticles and the matrix. From
the above two theories, it can be summed up that more
energy and load were needed for crack propagation in
sand-filled epoxy composites compared to pristine epoxy.
Therefore, addition of sand to epoxy helped to improve the
toughness of the composite.
DMA
The specimens generally deform sinusoidally in response
to an applied oscillating force. The resultant strain in
specimen due to the sinusoidal load depends upon both
elastic and viscous behavior of the specimen. In this study,
the storage modulus (E0 ) and the loss factor (damping
coefficient, tan d ¼ E0 /E00 , E00 is the loss modulus) of epoxy
and composites were determined by DMA. The storage
modulus reflects the elastic modulus of the composites
which measures the recoverable strain energy in a deformed specimen, and the loss modulus (or viscous modulus)
is related to the energy lost due to energy dissipation as
heat. The storage modulus of epoxy and composites versus
temperature curves is shown in Figure 5. It is found from
Figure 5 that the storage modulus of the epoxy composites
Figure 5. The storage modulus versus temperature curves of
pristine epoxy and epoxy/sand composites.
Macromol. Mater. Eng. 2007, 292, 467–473
ß 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
exceeds that of pristine epoxy under all temperature regions. The rigid sand particles enhance the stiffness of the
epoxy matrix composite which in turn shows an increment of storage modulus of the epoxy/sand composite
specimens.
Values of tan d provide information regarding damping
of the materials. Figure 6 shows how tan d of the pristine
epoxy and the epoxy/sand composites vary with temperature. It is used for measuring the glass transition
temperature Tg of the materials as well and the peak of
tan d in tan d–temperature T curve can be identified as Tg.
From Figure 6, it is seen that Tg of the composite shifts to a
temperature 10 8C above Tg of the pristine epoxy resins.
Mobilization of macromolecules within materials in
molecular scale is related to Tg and with increasing increments of Tg, mobility of molecules diminishes. Due to good
adhesion of the sand particle with matrix and its rigidity,
mobilization of the macromolecule made of sand particle
and its surrounding matrix decreases compared to pristine
epoxy matrix and as a result increased Tg was observed in
epoxy/sand composite. Therefore, this represents that the
restriction effect of mobilization in epoxy/sand composite
provides better stability when compared to that in pristine
epoxy. The filling of sand also reduces the height of tan d
peak, which is related to the energy damping characteristics of the materials. These results are owing to the
enhancement in the stiffness of epoxy composites by the
sand particles. The DMA tests indicate that the sand
particles effectively enhance the mechanical properties of
the epoxy resins.
Thermo-Mechanical Analysis (TMA)
TMA is a kind of thermal analysis used to measure dimensional (physical) changes in a sample with adjustable
sample strain when temperature and/or time varies. It is
also used to determine CTE and Tg. Figure 7 plots the
dimensional change versus temperature curves for the
pristine epoxy and epoxy/sand composites. It shows that
for a long temperature range (from 25 to 250 8C)
dimensional change in epoxy/sand composite is less
compared to pristine epoxy. CTE is the ratio of change
of dimension per unit dimension and change of temperature and can be termed as (l/L)/DT; where l is the change in
length, L the original length and DT is the change in
temperature. Figure 8 shows the CTEs below and above Tg
for the pristine epoxy and epoxy/sand composites. It can
be seen from the figure that in both cases CTEs of epoxy/
sand composites are less compared to that of the epoxy
system. The change in dimension and CTE can be explained
by the concept of molecule interaction. With the increase
in temperature, molecular vibration in the material
increases, which causes an increment in the intermolecule
www.mme-journal.de
471
G. Sui et al.
epoxy molecules upon heating. The CTEs
of the epoxy/sand composite in both
below and above Tg are lower than those
of the pristine epoxy resin.
Friction Behavior
From Figure 9, it can be seen that the
friction coefficient of pristine epoxy is
about 0.12, and it increases to the value
of 0.18 on epoxy/sand composites. So
with 1 wt.-% addition of sand into
epoxy, friction coefficient increases by
50%. The filling of the sand particles enhances the coarseness of epoxy
matrix and therefore the friction coefficient is increased. The friction coefficient of epoxy composites is stable with
Figure 6. The loss factor versus temperature curves of pristine epoxy and epoxy/sand
prolonged time, and this reflects the
composites.
constant microstructure of composites.
However, at the very beginning, the
distance. However, higher restriction effect in mobilization
friction coefficient of epoxy/sand composite increases for a
of sand/composite molecules (as discussed earlier) limits
small amount of time and after that it shows stability. But
the amplitude of vibration. This means that the interthis does not happen to pristine epoxy system and it shows
molecule distances do not increase in epoxy/sand
that from the start to end of the experiment, the friction
composites as much as it increases in epoxy material.
coefficient is stable. Sand particles make the surface of a
Therefore, the dimensional change in epoxy/sand compocomposite rough which exists for less time with erosion
site is less compared to that in the epoxy system. In
under friction. Therefore it can be concluded that due to
addition, it can be mentioned that the inorganic sand
the porosity and brittleness, the sand can be broken up
additives neither deform nor relax like the organic epoxy
when subjected to external force. However, when the sand
molecules as the temperature increases. Therefore, if the
was made into composite with epoxy, an improvement in
sand can be effectively dispersed in the epoxy matrix, the
the property of the epoxy is seen. With time variation, for
rigid sand particles retard the thermal expansion of the
epoxy/sand composite and epoxy, friction coefficients vary
a negligible amount and wear rates in both cases are low.
However, the friction coefficient fluctuation in epoxy/sand
composite is less than that in the epoxy system. Sand gives
more rigidity to epoxy composite because wear rate is
Figure 7. Dimension change versus temperature curves of pristine
epoxy and epoxy/sand composites.
472
Macromol. Mater. Eng. 2007, 292, 467–473
ß 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 8. Effect of sand on the CTEs of pristine epoxy and epoxy/
sand composites.
DOI: 10.1002/mame.200600479
Preparation and Properties of Natural Sand Particles Reinforced . . .
Acknowledgements: We would like to acknowledge Dr. Joseph
Brennan for collecting the sand samples for this project. Partial
work reported here was supported by NASA through grant
NNM04AA62G and by NSF through NIRT grant 0506531.
Received: December 12, 2006; Revised: February 4, 2007;
Accepted: February 7, 2007; DOI: 10.1002/mame.200600479
Keywords: composites; epoxy resin; modulus; sand particles
Figure 9. The friction coefficient versus friction time curves of
pristine epoxy and epoxy/sand composites.
lower, whereas in epoxy system wear rate is higher,
although it has a low friction coefficient. Therefore
considering both friction coefficient and wear rate, an
epoxy/sand composite is more rigid and stable compared
to the epoxy system.
Conclusion
Epoxy composites were prepared with the loading of
1 wt.-% Cancun sand. The flexural strength and flexural
modulus of epoxy had increased by 7.5 and 8.7% due to the
addition of the sand. Fracture morphology observation
showed that the fracture toughness of epoxy composites
has improved. The epoxy/sand composites possessed
obviously enhanced E0 , Tg and dimensional stability, as
well as friction resistance, compared to the pristine epoxy
resin. The experimental results showed that the microstructure of epoxy/sand composites was steady.
Although many kinds of inorganic particles have been
reported as additives for epoxy composites, Cancun sand,
without any chemical treatment involved in this work,
exhibited a comprehensive improvement in mechanical,
thermal, and other related properties of epoxy resins. The
above experimental results indicate that this natural
hydrophobic sand may be a good filler material for
polymer composite materials.
Macromol. Mater. Eng. 2007, 292, 467–473
ß 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[1] E. T. Thostenson, T. W. Chou, Carbon 2006, 44, 3022.
[2] D. Olmos, R. López-Morón, J. González-Benito, Compos. Sci.
Technol. 2006, 66, 2758.
[3] L. J. Ci, J. B. Bai, Compos. Sci. Technol. 2006, 66, 599.
[4] M. Moniruzzaman, F. M. Du, N. Romero, K. I. Winey, Polymer
2006, 47, 293.
[5] W. K. Goertzen, M. R. Kessler, Compos.: Part B 2007, 38, 1.
[6] M. Hussain, A. Nakahira, S. Nishijima, K. Niihara, Mater. Lett.
1996, 27, 21.
[7] Y. P. Zheng, Y. Zheng, R. C. Ning, Mater. Lett. 2003, 57, 2940.
[8] M. Hussain, A. Nakahira, K. Niihara, Mater. Lett. 1996, 26, 185.
[9] G. Shi, M. Q. Zhang, M. Z. Rong, B. Wetzel, K. Friedrich, Wear
2004, 256, 1072.
[10] C. K. Lam, K. T. Lau, Compos. Struc. 2006, 75, 553.
[11] B. Fiedler, F. H. Gojny, M. H. G. Wichmann, M. C. M. Nolte,
K. Schult, Compos. Sci. Technol. 2006, 66, 3115.
[12] A. Zhamu, M. Wingert, S. Jana, W. H. Zhong, J. J. Stone,
Compos.: Part A 2007, 38, 699.
[13] S. C. Zunjarrao, R. P. Singh, Compos. Sci. Technol. 2006, 66,
2296.
[14] K. Friedrich, Z. Zhang, A. K. Schlar, Compos. Sci. Technol. 2005,
65, 2329.
[15] Y. K. Choi, K. Sugimoto, S. M. Song, Y. Gotoh, Y. Ohkoshi,
M. Endo, Carbon 2005, 43, 2199.
[16] J. A. Kim, D. G. Seong, T. J. Kang, J. R. Youn, Carbon 2006, 44,
1898.
[17] Y. Zhou, F. Pervin, V. K. Rangari, S. Jeelani, Mater. Sci. Eng. : A
2006, 426, 221.
[18] J. C. Lin, Compos.: Part B 2007, 38, 79.
[19] B. Wetzel, P. Rosso, F. Haupert, K. Friedrich, Eng. Fract. Mech.
2006, 73, 2375.
[20] A. Zhamu, W. H. Zhong, J. J. Stone, Compos. Sci. Technol. 2006,
66, 2736.
[21] H. Chen, J. H. Wang, Z. Rahman, J. G. Worden, X. Liu, Q. Dai,
Q. Huo, J. Mater. Sci. (in press).
[22] M. Sumita, T. Ookuma, K. Miyasaka, K. Ishikawa, J. Mater. Sci.
1982, 17, 2869.
[23] S. M. Wiederhorn, Annu. Rev. Mater. Sci. 1984, 14, 373.
[24] R. A. Pearson, ‘‘Toughened Plastics’’, C. K. Riew, A. J. Kinloch,
Eds., American Chemical Society, Washington, DC 1993.
[25] K. T. Faber, A. G. Evans, Acta Metall. 1983, 31, 565.
[26] K. T. Faber, A. G. Evans, Acta Metall. 1983, 31, 577.
www.mme-journal.de
473