You are on page 1of 9

Composites Science and Technology 123 (2016) 125e133

Contents lists available at ScienceDirect

Composites Science and Technology


journal homepage: http://www.elsevier.com/locate/compscitech

Effect of the orientation of graphene-based nanoplatelets upon the


Young's modulus of nanocomposites
Zheling Li a, Robert J. Young a, *, Neil R. Wilson b, Ian A. Kinloch a, Cristina Valle
s a,
Zheng Li a
a
National Graphene Institute and School of Materials, University of Manchester, Oxford Road, Manchester M13 9PL, UK
b
Department of Physics, University of Warwick, Coventry, CV4 7AL, UK

a r t i c l e i n f o a b s t r a c t

Article history: The spatial orientation of the reinforcing elements in polymer-based composites plays a vital role in
Received 27 August 2015 controlling mechanical properties but there is no generally-accepted way of quantifying the spatial
Received in revised form orientation at the nanoscale of plate-like fillers in nanocomposites. A previous study found that the
29 November 2015
intensity of scattering of the Raman band is dependent on the axis of laser polarization when the laser
Accepted 12 December 2015
Available online 15 December 2015
beam is parallel to the surface of the graphene plane and demonstrated that this could be used to
quantify the spatial orientation of the graphene. Based on this approach, polarized Raman spectroscopy
is used here to quantify, as an example, the level of spatial orientation of graphene oxide (GO) flakes in
Keywords:
Graphene oxide
different nanocomposites. Furthermore, the spatial orientation of nanoplatelets is quantitatively linked
Mechanical properties to the stress transfer to the reinforcement in nanocomposites and, in particular, a method for deter-
Raman spectroscopy mining the Krenchel orientation factor for these plate-like fillers directly from the Raman data is
Spatial orientation demonstrated. This approach is employed to estimate the effective Young's modulus of the reinforce-
ment in graphene-based nanocomposites.
© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).

1. Introduction considerable effort is often needed for specimen preparation [12].


Since the orientation of the reinforcement affects the mechanical
The geometry and spatial orientation of the reinforcement play properties of nanocomposites significantly [4,6,7], a generally-
vital roles in determining the mechanical properties of composites applicable method of quantifying the spatial orientation of 2D re-
[1,2], especially for composites reinforced by nano-fillers. It has inforcements such as graphene is urgently needed.
been reported that the spatial orientation of graphene in nano- The technique of Raman spectroscopy is used widely to char-
composites varies significantly with the preparation method [3,4] acterise graphene [13,14]. The stress-sensitive Raman band shift
and graphene concentration [5,6]. Even for graphene-based [15,16] has been employed to monitor the deformation mechanics
poly(vinyl alcohol) (PVA) nanocomposites prepared using a of graphene in nanocomposites [17e19], also revealing how the
similar drop casting method, different degrees of spatial orienta- size and aspect ratio of the reinforcement can affect the mechanical
tion of the graphene have been found [7,8]. Unlike the situation properties of nanocomposites [20]. However, the spatial orienta-
with conventional micron-size fibres, however, the spatial orien- tion also plays a critical role as reinforcement of the Youngs
tation of nano-fillers is not easily characterised since traditional modulus only occurs for strain in the direction of the plane of the
techniques such as optical or electron microscopy provide only graphene. There is one previous report of graphene orientation in a
limited information at this scale. The techniques used for nano- film being studied using polarized Raman spectroscopy [21],
fillers are limited to X-ray diffraction (XRD) [3,9], transmission although the depolarization ratio for nanotubes was used for the
electron microscopy (TEM) [10], and fast Fourier transforms analysis [22]. More recently, polarized Raman spectroscopy has
together with scanning electron microscopy (SEM) [11], and been applied to quantify the three-dimensional spatial orientation
of graphene in terms of a generalized spherical expanded har-
monics orientation distribution function (ODF) [23], based on the
model of Liu and Kumar developed to characterise the orientation
* Corresponding author.
of carbon nanotubes [24] but refined for the two-dimensional
E-mail address: robert.young@manchester.ac.uk (R.J. Young).

http://dx.doi.org/10.1016/j.compscitech.2015.12.005
0266-3538/© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
126 Z. Li et al. / Composites Science and Technology 123 (2016) 125e133

geometry of graphene. The model was validated on highly-ordered 3. Results and discussion
pyrolytic graphite and graphene paper but was not applied directly
to the quantitative analysis of the spatial orientation of graphene 3.1. Model of spatial orientation
and GO in nanocomposites. Moreover, the framework for using this
orientation to quantitatively determine the affect on the mechan- As explained in our previous study [23], the local orientation of a
ical properties of the nanocomposites was not derived. nanoplatelet such as graphene can be defined by its surface normal,
It is well established that the mechanical properties of polymer- shown in Fig. 2 as the z direction of the graphene in the x,y plane. By
based nanocomposites are controlled by a number of different using the polar coordinates of the z direction q and f, and the
factors due to their hierarchical structure, that spans from the nano rotation angle of the graphene x, the orientation of graphene can be
to macro length scales [25e27]. In particular it is important to related to the sample coordinates (X,Y,Z) as indicated. Conse-
consider issues such as the special properties of the polymer- quently, the ODF of the surface normal can be written in general as
matrix interface and any interphase at the nanoscale [28e31]. fN(q,f,x). On this basis, using the VV polarization, the Raman scat-
This present study deals only with geometrical factors that affect tering intensity of the sample, Isample(F), can be expressed
the micromechanics of deformation in nanoplatelet-reinforced regarding the polarization angle F (FX/FY/FZ) relative to the sam-
composites and is not concerned directly with interfacial effects. ple, as [23]:
Polarized Raman spectra have been obtained from sections of GO-   
reinforced nanocomposites and used to quantify the spatial 8 16 8
Isample ðFÞ ¼ Io $ þ hP2 ðcos qÞi  þ cos2 F
orientation of the GO within the nanocomposites. Furthermore, it is 15 21 7
 
shown how the classical Krenchel orientation factor [1] can be 8 8
determined directly from the polarized Raman spectra and used to þ hP4 ðcos qÞi  cos2 F þ cos4 F (1)
35 7
correlate the degree of spatial orientation of the reinforcement
with the mechanical properties of the nanocomposites. where Io is the amplitude and it is assumed that the surface nor-
mals are uniformly distributed around the Z-axis. By fitting the
experimental data to Eq. (1), a set of 〈P2(cosq)〉 and 〈P4(cosq)〉 values
2. Experimental can be obtained to quantify the spatial orientation of graphene in
graphitic materials [23]. When the graphene in the specimen is
2.1. Materials and sample preparation perfectly aligned, here being equivalent to it lying perfectly flat in
the XY plane, 〈P2(cosq)〉 ¼ 〈P4(cosq)〉 ¼ 1. Their values then decrease
The GO was prepared from graphite (Grade 2369, Graphexel Ltd) as the degree of alignment decreases until
using the modified Hummers method [32]. Details of its prepara- 〈P2(cosq)〉 ¼ 〈P4(cosq)〉 ¼ 0, which is characteristic of isotropic
tion and the preparation of the GO-reinforced PVA nanocomposites materials. It should also be pointed out that the parameter
(GO/PVA) can be found elsewhere [33]. The preparation methods 〈P2(cosq)〉 ¼ (3〈cos2q〉1)/2 is the Herman's orientation factor used
for the 10 wt% GO-reinforced PMMA (GO/PMMA) and 0.5 wt% GO- widely in polymer and composites science [35]. As discussed
reinforced epoxy resin (GO/epoxy) have also been described else- earlier, however, the introduction of the extra 〈P4(cosq)〉 has an
where [4,34]. advantage of reconstructing the ODF of the fillers [23]. If the Raman
scattering intensity measured in the Z direction shows no angular
dependence, which means the material is isotropic about that axis,
2.2. Characterisation then the result obtained in the X direction is similar to that in the Y
direction and sufficient to describe the spatial orientation of gra-
Atomic force microscope (AFM) images were obtained using a phene. Although GO undergoes structural distortions after oxida-
Veeco CP-II system with the ‘RTESPA’ probe (Bruker) in the tapping tion, the crystalline structure of GO is predominantly the same as
mode. SEM images of the GO flakes were obtained using XL 30 FEG that of graphene [36] but decorated by oxidative debris [37], hence
Philips. SEM images of the nanocomposites were obtained using it essentially has a structure and vibrational modes similar to those
EVO60 VPSEM (Zeiss), where samples were fractured at room of graphene. It is assumed that the Raman polarizability tensors of
temperature, and coated with gold before analysis. X-ray diffraction GO are the same as those of graphene, thus Eq. (1) will also be
(XRD) was carried out using an X'Pert DY609 X-Ray diffractometer applicable to GO. It will now be shown how this technique can be
(Philips) with a Cu Ka radiation source (l ¼ 1.542 Å). Polarized extended to quantify the spatial orientation of GO in bulk
Raman spectroscopy was carried out using Renishaw 1000/2000 nanocomposites.
spectrometers equipped with HeNe lasers (l ¼ 633 nm). The laser
beam, with a spot size ~2 mm in diameter, was aligned to the cross-
sections and the top surfaces of the nanocomposites. The laser was 3.2. GO/PVA nanocomposites
polarized in the ‘VV’ configuration, where the incident and scat-
tered radiation were polarized in the same direction. For the GO/PVA, the lateral dimensions of the GO flakes was
The Cartesian coordinate system with the X, Y and Z axes in found to be of the order of 10 mm, as shown in the SEM image
which the specimens were analysed is defined in Fig. 1(a) [23,33]. (Fig. 3(a)). The typical thickness of 1 nm of the exfoliated GO
The laser beam was aligned along the Z (perpendicular to the top monolayer determined using AFM (Fig. 3(b) and (c)) demonstrates
surface of sample, Fig. 1(b)), X and Y (perpendicular to the fracture the complete exfoliation of the GO flakes [8]. The dispersion of GO
surfaces of GO/PVA and polished surface of GO/epoxy and GO/ in the nanocomposites was characterised by XRD (Fig. 4). The
PMMA, Fig. 1(c)) directions, respectively. During the analyses the diffraction peak of the pure GO at 2q ¼ 10.5 indicates there is a d-
laser polarization configuration was kept fixed, and the specimens spacing of 0.84 nm. The similar diffraction patterns of the GO/PVA
were rotated to different angles, FX, FY and FZ for the laser beam and neat PVA sample imply the complete exfoliation of GO flakes in
parallel to the X, Y and Z axes respectively. For the materials that the nanocomposites [8]. Their exfoliation and dispersion was
were thought to be transversely isotropic, however, polarized confirmed using SEM, which also indicated that the GO flakes were
Raman spectra were obtained only with the laser beam parallel to aligned generally parallel the top surface of the nanocomposites
the Z and X directions. film (Fig. 4(a)).
Z. Li et al. / Composites Science and Technology 123 (2016) 125e133 127

Z
Nanocomposite
Y

(a) X
Graphene Oxide Flake

ΦX (ΦY) Y
ΦZ Y

(b) In ‘Z’ Direction (c) In ‘X’ and ‘Y’ Direction


Fig. 1. (a) Schematic illustration of the Cartesian coordinate system with the sample geometries used in the Raman spectroscopic analysis in the three directions. (b) View with the
laser beam along the Z axis. (c) View with the laser beam along the X and Y axes. The red arrows represent the laser propagation direction and the green arrows represent the
directions of polarization of the incident and scattered radiation. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this
article.)

decreasing values of〈P2(cosq)〉 and 〈P4(cosq)〉 (Fig. 6). The possible


reason for this is that the GO flakes align by settlement after water
evaporates. At a higher concentration, however, the large amount of
the GO flakes perhaps stop this process occurring so easily resulting
in a poorer orientation of the GO.

3.3. GO/epoxy and GO/PMMA

For comparison, two kinds of other GO-based nanocomposites


were assessed using the same technique. The Raman spectra for
those two nanocomposites with the two laser polarization config-
urations are shown in Fig. 7(a). Generally they are similar to those
of the GO/PVA (Fig. 5), but with different fluorescent backgrounds.
The spatial orientation of GO in these two nanocomposite sys-
tems [4,34] was also assessed using polarized Raman spectroscopy.
As we had no prior knowledge of their state of orientation for these
materials, the variation of IG with the angle F was determined for
the direction of laser beam propagation in all the X, Y and Z di-
rections (Fig. 7(b) and (c)). The lack of variation of IG with F showed
that there was no preferred orientation of the GO in either of these
two nanocomposites. The GO/epoxy was prepared by mixing the
GO with the epoxy resin and casting into a mould [34]. In the case of
the GO/PMMA, the GO had been blended in a twin-screw extruder
Fig. 2. The relationship between the local coordinate system of the GO flake (x,y,z) and with PMMA and injection moulded before consolidation by hot-
the nanocomposite sample (X,Y,Z), which is defined by three Euler angles (q,f,x). pressing [4]. It appears that neither process led to any preferred
alignment of the GO.
It can be seen that the angular dependences of the Raman in-
The Raman spectra of the GO/PVA nanocomposites with
tensities for the different sp2-hybridized carbon materials studied
different GO loadings when the laser beam is along the X and Z
above generally follow Eq. (1). This demonstrates that it is possible
directions are shown in Fig. 5. In the similar spectra, the G band
to characterise the spatial orientation of the GO in the different
around 1600 cm1 results from the E2g vibrational mode [14]. The D
specimens in terms of their values of 〈P2(cosq)〉 and 〈P4(cosq)〉, as
band, centred around 1335 cm1, is usually assigned to the A1g
shown in Table 1. The different values of these parameters in the
symmetry, and is associated with the presence of defects [14].
Table reflect the variation of the degrees of spatial orientation of the
Given the good dispersion of GO in the nanocomposites (Fig. 4(b)),
GO in the different nanocomposites.
its spatial orientation can be assessed using the strong G band [23].
Fig. 6 shows the variation of G band intensity (IG) as the function of
3.4. Orientation distribution function (ODF)
polarization angle (FX or FZ), along with the values of 〈P2(cosq)〉 and
〈P4(cosq)〉 obtained by fitting the experimental data with Eq. (1).
Based on the 〈P2(cosq)〉 and 〈P4(cosq)〉 values determined, a best
The IG measured in the Z direction shows no angular dependence,
guess of the actual ODF can be obtained as indicated previously as
demonstrating that the materials are isotropic about that axis
[23,24,38]:
(Fig. 6). The IG measured in X direction implies a good spatial
orientation of GO at 1 wt% loading, but that their alignment tends to fN ðqÞ ¼ A exp½  ðl2 P2 ðcos qÞ þ l4 P4 ðcos qÞÞ (2)
decrease as the GO loading is increased, as indicated by the
128 Z. Li et al. / Composites Science and Technology 123 (2016) 125e133

Fig. 3. (a) SEM micrograph and (b) AFM image of the GO flakes. (c) The height profile of the lines in (b) indicate typical thickness of GO of ~1 nm showing the GO flakes have been
fully dispersed to monolayers. Dashed lines are guide to eyes.

〈P2(cosq)〉 and 〈P4(cosq)〉, the ODF of graphene and GO can be


reconstructed using Eq. (2) (Fig. 8). For simplicity, all the ODFs have
been normalized to their corresponding 0 values. It can be seen
that fN(q) decreases with increasing q for the GO/PVA nano-
composites showing partial orientation of the platelets. The highest
level of spatial orientation is found for the 1 wt% GO/PVA where
fN(q) decreases rapidly with increasing q, whereas it is independent
of q for the GO/PMMA and GO/epoxy, indicating the isotropic na-
ture of these two nanocomposites.

3.5. Krenchel orientation factor ho

The rule of mixtures can be used to estimate the reinforcing


effect on the elastic modulus of a composite, Ec, due to a filler with
elastic modulus Ef:

Ec ¼ hl ho Vf Ef þ Vm Em (3)

where Vf is the volume fraction of the filler, Vm the volume fraction


of the matrix, Em the elastic modulus of matrix, hl the length effi-
ciency factor, and ho the Krenchel orientation factor. The rule of
mixtures provides a simple method to estimate the elastic prop-
erties of the composite, however, it is just a first approximation
without taking into account other factors, such the Poisson's ratio.
They have been ignored here, as this work is primarily concerned
about how the orientation of a graphene-based filler will affect its
reinforcement efficiency in terms of the Krenchel factor. To be
rigorous, however, Poisson's effects need also to be taken into
consideration as suggested elsewhere [39].
Although the knowledge of 〈P2(cosq)〉 and 〈P4(cosq)〉 enables the
ODF to be obtained, one of the most useful applications of these
order parameters is in the determination of ho which defines the
effect of the orientation of materials, such as fibres [1] and carbon
nanotubes [40], upon the mechanical properties of the reinforced
Fig. 4. (a) SEM image of the cross section of the 2 wt% GO/PVA. (b) XRD patterns of GO, composites. For one-dimensional materials, ho ¼ 1 for perfectly-
neat PVA and GO/PVA nanocomposites with different GO loadings. aligned fibres or nanotubes, but 3/8 for fibres or nanotubes ori-
ented randomly in 2D plane and 1/5 for material with random 3D
fibres or nanotubes [1,40].
where the coefficients A, l2 and l4 can be solved numerically as In general, the Krenchel factor is not unique for a particular
indicated in Refs. [23,24,38]. The ODF quantifies the probability of material but depends upon the direction along which the me-
finding a platelet with surface normal at an angle q to the Z di- chanical properties are to be determined. For a single reinforce-
rection. Using the polarized Raman spectroscopy data to determine ment element, ho ¼ cos4z where z is the angle between the element
Z. Li et al. / Composites Science and Technology 123 (2016) 125e133 129

Fig. 5. Raman spectra of the GO/PVA nanocomposites with the laser beam in (a) X (blue) and (b) Z (red) directions. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)

Fig. 6. IG variation with angle F for GO/PVA nanocomposites with different GO loadings with the laser beam in X (black) and Z (red) directions. (a) 1 wt% GO, (b) 2 wt% GO, (c) 3 wt%
GO and (d) 5 wt% GO/PVA. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

and the axis along which the stress is applied. For the platelet geometry shown in Fig. 2, with a symmetrical distribution of the
130 Z. Li et al. / Composites Science and Technology 123 (2016) 125e133

Fig. 7. (a) Raman spectra of GO/epoxy and GO/PMMA with the laser beam in X (blue) and Z (red) directions. (b) GO/epoxy and (c) GO/PMMA show the corresponding IG variation
with angle F with the laser beam in X, Y and Z directions, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
this article.)

Table 1
Values of 〈P2(cosq)〉, 〈P4(cosq)〉, ho and duD/dε determined with the direction of laser beam propagation parallel to the X direction for all the materials studied.

〈P2(cosq)〉 〈P4(cosq)〉 ho duD/dε (embedded) (cm1/% strain)

1 wt% GO/PVA 0.54 ± 0.06 0.52 ± 0.24 0.78 ± 0.04 8.8 ± 0.1 [33]
2 wt% GO/PVA 0.46 ± 0.03 0.51 ± 0.12 0.75 ± 0.02 8.3 ± 1.4 [33]
3 wt% GO/PVA 0.36 ± 0.01 0.28 ± 0.05 0.69 ± 0.01 8.3 ± 1.5 [33]
5 wt% GO/PVA 0.38 ± 0.03 0.50 ± 0.14 0.72 ± 0.02 7.5 ± 0.4 [33]
0.5 wt% GO/epoxy ~0 ~0 0.53 4.2 [34]
10 wt% GO/PMMA ~0 ~0 0.53 4.0 [4]

surface q ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
normals and forffi deformation parallel to the Y axis,
cos z ¼ 1  ðsin q sin fÞ2 . To find the Krenchel factor for a spec- 8 8 3
ho ¼ þ hP ðcos qÞi þ hP ðcos qÞi (5)
imen, this must be integrated over the ODF [1] thus: 15 21 2 35 4
Importantly ho depends only on 〈P2(cosq)〉 and 〈P4(cosq)〉 and
Z Z Z hence can be determined directly from the experimental mea-
2p 2p p h i2
surements of the angular dependence of the polarized Raman
1  ðsin q sin fÞ2 fN ðqÞsin qdqdfdx
x¼0 f¼0 q¼0 scattering intensities (without reconstructing the actual ODF).
ho ¼ Z 2p Z Z p
(4)
2p Fitting Eq. (1) to the experimental data yields the two parameters,
fN ðqÞsin qdqdfdx 〈P2(cosq)〉 and 〈P4(cosq)〉, which can then be used directly in Eq. (5)
x¼0 f¼0 q¼0
to determine ho.
Substituting in the definition of the ODF in our previous paper Eq. (5) also allows analysis of the limiting cases for nanoplatelet
[23] gives: reinforcement. For a perfectly-oriented material 〈P2(cosq)〉¼
Z. Li et al. / Composites Science and Technology 123 (2016) 125e133 131

1.5 It can be seen that the Krenchel factor decreases as the orientation
1 wt% GO/PVA 10 wt% GO/PMMA of the materials decreases until it reaches the 8/15 in the case of
2 wt% GO/PVA 0.5 wt% GO/epoxy random orientation. All of this analysis assumes that stress transfer
3 wt% GO/PVA at the filler-matrix interface is not affected by the orientation of the
5 wt% GO/PVA nanoplatelets.
1.0 Another point to bear in mind is the finite length effects: the
variation of stress across platelets of finite lateral dimensions due to
ODF fN(θ )

shear-lag effects [17] leads to the effective modulus of the filler


being reduced by a further length efficiency factor hl [41,42]. Hence
Eq. (3) above shows that the best reinforcement will be achieved
0.5
with well-aligned flat nanoplatelets that possess large lateral di-
mensions [20].

3.6. Reinforcement by graphene oxide


0.0
0 10 20 30 40 It has been demonstrated that the effective Young's modulus of
the GO, Eeff(GO) in nanocomposites can be estimated from the
Angle θ (o) Raman band shift [43], particularly, the Raman D band shift rate
with strain (duD/dε) [33]. However, the previously derived values
Fig. 8. The orientation distribution functions (ODF) for the materials studied. A con-
ignored the effect of the orientation of the GO within the composite
stant value of fN(q) indicates random orientation and it decreases more rapidly with
increasing q as the degree of orientation increases. and hence gave only a lower bound. Since the Krenchel factor ho is
determined by the orientation of GO, which also affects Eeff(GO), it
is instructive to see how ho is correlated to duD/dε. To exclude the
〈P4(cosq)〉¼1 while for 3D randomly-oriented materials 〈P2(cosq)〉¼ different Poisson's ratio of the different matrix, the duD/dε has been
〈P4(cosq)〉¼0. Eq. (5) shows therefore that the Krenchel factor is converted to the values for freestanding GO using an approach
predicted to be 1 for a perfectly-oriented material and 8/15 for a described in Refs. [16], where the Poisson's ratios of PVA, epoxy and
material reinforced with 3D-random platelets. PMMA are taken as 0.40 [44], 0.38 [45] and 0.35 [46], respectively.
Liu and Brinson [2] used the Mori-Tanaka theory to numerically Fig. 9(b) shows the variation of duD/dε with Krenchel factor, ho, for
model the elastic deformation of nanocomposites. Analysis of their the different nanocomposites [4,33,34].
data shows that for reinforcement of a polymer with nanoplatelets The linear relation between the converted duD/dε and ho vali-
of a given aspect ratio, a nanocomposite with random nanoplatelets dates the Krenchel factors determined e the stress transfer is lin-
would be expected to have a Young's modulus of around 8/15 of early related to the Krenchel factor as expected. The intercept at
that for a nanocomposite with aligned nanoplatelets deformed ho ¼ 1 allows determination of the value for perfectly-aligned
parallel to the axis of alignment e in good agreement with the exact freestanding GO, a value of duD/dε ~ 20 cm1/% strain is obtained.
prediction given by Eq. (5). As described in Refs. [33], for free-standing graphene with a
The values of Krenchel factor ho of all the studied materials are Young's modulus (Egra) of 1050 GPa reference value of (duD/
shown in Fig. 9(a). Note that the lack of any intensity variation with dε)Ref ¼ 40 cm1/% strain is obtained. Hence the Eeff(GO) can be
the angle FZ when the laser beam is parallel to the Z axis shows that calculated by comparison to this reference value [33]:
the orientation of the graphene sheets varied uniformly around this
axis as assumed above. Thus the hP2 ðcos qÞi and 〈P4(cosq)〉 values duD =dε tgra
Eeff ð GOÞ ¼   Egra (6)
derived with the laser beam in the X direction are sufficient to fully ðduD =dεÞRef tGO
describe the orientation of nanoplatelets such as graphene and GO.
where tgra and tGO represent the thickness of monolayer graphene

1.0 (a) 1 wt% GO/PVA


2 wt% GO/PVA
0.5 wt% GO/epoxy
10 wt% GO/PMMA
(b) 0.5 wt% GO/epoxy
Krenchel Orientation Factor ηo

-5 10 wt% GO/PMMA
3 wt% GO/PVA
0.9 5 wt% GO/PVA
dωD/dε (cm-1/%)

0.8
-10
0.7
8
0.6
15 -15 1 wt% GO/PVA
2 wt% GO/PVA
0.5 3 wt% GO/PVA
5 wt% GO/PVA Intercept ~ -19 cm-1/%
0.4 -20
0.5 0.6 0.7 0.8 0.9 1.0
Krenchel Factor, ηο

Fig. 9. (a) The Krenchel factors of the materials studied. The dashed line represents the Krenchel value of 8/15 for materials reinforced with random nanoplatelets. (b) Correlation of
the Krenchel factor ho with the reported duD/dε (listed in Table 1) of GO/PVA (black circle) [33], GO/epoxy (blue triangle) [34] and GO/PMMA (red square) [4]. (For interpretation of
the references to colour in this figure legend, the reader is referred to the web version of this article.)
132 Z. Li et al. / Composites Science and Technology 123 (2016) 125e133

(0.34 nm) [8] and GO (0.84 nm, obtained from the XRD (Fig. 4(b)), organically modified montmorillonite in nylon-66 matrix, J. Polym. Sci. Part B
Polym. Phys. 41 (11) (2003) 1234e1243.
respectively, to account for the effect of the increased thickness on
[10] T.D. Fornes, D.R. Paul, Modeling properties of nylon 6/clay nanocomposites
the decrease of the Young's modulus. Fig. 9(b) thus predicts an using composite theories, Polymer 44 (17) (2003) 4993e5013.
Eeff(GO) of around 200 GPa, in good agreement with the values [11] X. Lin, X. Shen, Q. Zheng, N. Yousefi, L. Ye, Y.-W. Mai, et al., Fabrication of
reported from AFM studies on single sheets of GO [47], further highly-aligned, conductive, and strong graphene papers using ultralarge
graphene oxide sheets, ACS Nano 6 (12) (2012) 10708e10719.
giving confidence in the model derived above. The Young's [12] C. Zhang, S. Huang, W.W. Tjiu, W. Fan, T. Liu, Facile preparation of water-
modulus of GO is clearly significantly lower than the value of dispersible graphene sheets stabilized by acid-treated multi-walled carbon
~1050 GPa for pristine graphene, but as explained earlier [33], this nanotubes and their poly(vinyl alcohol) composites, J. Mater. Chem. 22 (6)
(2012) 2427e2434.
is due to the oxidation process causing both damage to the gra- [13] A.C. Ferrari, J.C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, et al.,
phene lattice and an increase in the effective thickness of the GO. Raman spectrum of graphene and graphene layers, Phys. Rev. Lett. 97 (18)
(2006) 187401.
[14] A.C. Ferrari, D.M. Basko, Raman spectroscopy as a versatile tool for studying
4. Conclusions the properties of graphene, Nat. Nanotechnol. 8 (4) (2013) 235e246.
[15] M. Huang, H. Yan, C. Chen, D. Song, T.F. Heinz, J. Hone, Phonon softening and
It has been demonstrated that it is possible to use the model crystallographic orientation of strained graphene studied by Raman spec-
troscopy, Proc. Natl. Acad. Sci. 106 (18) (2009) 7304e7308.
established in our previous study [23] to quantify the spatial [16] T.M.G. Mohiuddin, A. Lombardo, R.R. Nair, A. Bonetti, G. Savini, R. Jalil, et al.,
orientation of GO in a variety of different GO based nanocomposite Uniaxial strain in graphene by Raman spectroscopy: G peak splitting, Grü-
systems in terms of an ODF, and also to determine the corre- neisen parameters, and sample orientation, Phys. Rev. B 79 (20) (2009)
205433.
sponding Krenchel orientation factor from the experimental [17] L. Gong, I.A. Kinloch, R.J. Young, I. Riaz, R. Jalil, K.S. Novoselov, Interfacial stress
polarized Raman data. Apart from the quantification method for transfer in a graphene monolayer nanocomposite, Adv. Mater. 22 (24) (2010)
the spatial orientation of nanoplatelets in the nanocomposites, 2694e2697.
[18] R.J. Young, L. Gong, I.A. Kinloch, I. Riaz, R. Jalil, K.S. Novoselov, Strain mapping
another significant finding of this study is that the Krenchel factor in a graphene monolayer nanocomposite, ACS Nano 5 (4) (2011) 3079e3084.
for 3D randomly-oriented nanoplatelets is 8/15. This means that [19] L. Gong, R.J. Young, I.A. Kinloch, I. Riaz, R. Jalil, K.S. Novoselov, Optimizing the
random orientation of fillers such as graphene should reduce the reinforcement of polymer-based nanocomposites by graphene, ACS Nano 6
(3) (2012) 2086e2095.
Young's modulus of the nanocomposites by less than a factor of 2 s, A.M. Abdelkader, R.J. Young, I.A. Kinloch, The effect of flake diameter
[20] C. Valle
compared with the fully-aligned material. Compared to the on the reinforcement of few-layer grapheneePMMA composites, Compos. Sci.
reduction in the modulus of a factor of 5 going from aligned to 3D Technol. 111 (0) (2015) 17e22.
randomly-oriented fibres and nanotubes [1], it means that, from [21] Q. Liang, X. Yao, W. Wang, Y. Liu, C.P. Wong, A three-dimensional vertically
aligned functionalized multilayer graphene architecture: an approach for
solely geometrical considerations, better levels of modulus graphene-based thermal interfacial materials, ACS Nano 5 (3) (2011)
enhancement should be achievable with nanoplatelets compared 2392e2401.
with nanotubes and there is less need to ensure accurate alignment [22] H.H. Gommans, J.W. Alldredge, H. Tashiro, J. Park, J. Magnuson, A.G. Rinzler,
Fibers of aligned single-walled carbon nanotubes: polarized Raman spec-
of nanoplatelets in composites. Beyond just graphene and GO, this troscopy, J. Appl. Phys. 88 (5) (2000) 2509e2514.
approach should be more widely applicable to the determination of [23] Z. Li, R.J. Young, I.A. Kinloch, N.R. Wilson, A.J. Marsden, A.P.A. Raju, Quanti-
the orientation of other nanoplatelet fillers for which well-defined tative determination of the spatial orientation of graphene by polarized
Raman spectroscopy, Carbon 88 (0) (2015) 215e224.
Raman spectra can be obtained. Moreover, the effect of spatial [24] T. Liu, S. Kumar, Quantitative characterization of SWNT orientation by
orientation upon the mechanical properties of composites pre- polarized Raman spectroscopy, Chem. Phys. Lett. 378 (3e4) (2003) 257e262.
dicted in terms of the Krenchel orientation factor will be valid for all [25] M. Quaresimin, M. Salviato, M. Zappalorto, Strategies for the assessment of
nanocomposite mechanical properties, Compos. Part B-Eng. 43 (5) (2012)
types of 2D platelet reinforcement. 2290e2297.
[26] T.S. Gates, G.M. Odegard, S.J.V. Frankland, T.C. Clancy, Computational mate-
Acknowledgements rials: multi-scale modeling and simulation of nanostructured materials,
Compos. Sci. Technol. 65 (15e16) (2005) 2416e2434.
[27] S. Yu, S. Yang, M. Cho, Multi-scale modeling of cross-linked epoxy nano-
This research has received funding from the European Union composites, Polymer 50 (3) (2009) 945e952.
Seventh Framework Programme under grant agreement no 604391 [28] R.K. Patel, B. Bhattacharya, S. Basu, Effect of interphase properties on the
damping response of polymer nano-composites, Mech. Res. Commun. 35
Graphene Flagship, EPSRC (award no. EP/I023879/1) and AFOSR/ (1e2) (2008) 115e125.
EOARD (award no. FA8655-12-1-2058). One of the authors (Z-L. L.) [29] I. Sevostianov, M. Kachanov, Effect of interphase layers on the overall elastic
is grateful to the China Scholarship Council for financial support. and conductive properties of matrix composites. Applications to nanosize
inclusion, Int. J. Solids Struct. 44 (3e4) (2007) 1304e1315.
[30] A. Lazzeri, V.T. Phuong, Dependence of the Pukanszky's interaction parameter
References B on the interface shear strength (IFSS) of nanofiller- and short fiber-
reinforced polymer composites, Compos. Sci. Technol. 93 (2014) 106e113.
[1] H. Krenchel, Fibre Reinforcement, Akademisk Forlag, Copenhagen, 1964. [31] H.W. Wang, H.W. Zhou, R.D. Peng, L. Mishnaevsky, Nanoreinforced polymer
[2] H. Liu, L.C. Brinson, Reinforcing efficiency of nanoparticles: a simple com- composites: 3D FEM modeling with effective interface concept, Compos. Sci.
parison for polymer nanocomposites, Compos. Sci. Technol. 68 (6) (2008) Technol. 71 (7) (2011) 980e988.
1502e1512. [32] W.S. Hummers, R.E. Offeman, Preparation of graphitic oxide, J. Am. Chem. Soc.
[3] H. Kim, C.W. Macosko, Processing-property relationships of polycarbonate/ 80 (6) (1958), 1339.
graphene composites, Polymer 50 (15) (2009) 3797e3809. [33] Z. Li, R.J. Young, I.A. Kinloch, Interfacial stress transfer in graphene oxide
s, I.A. Kinloch, R.J. Young, N.R. Wilson, J.P. Rourke, Graphene oxide and
[4] C. Valle nanocomposites, ACS Appl. Mater. Interfaces 5 (2) (2013) 456e463.
base-washed graphene oxide as reinforcements in PMMA nanocomposites, [34] Z. Li, R.J. Young, R. Wang, F. Yang, L. Hao, W. Jiao, et al., The role of functional
Compos. Sci. Technol. 88 (0) (2013) 158e164. groups on graphene oxide in epoxy nanocomposites, Polymer 54 (21) (2013)
[5] N. Yousefi, M.M. Gudarzi, Q. Zheng, S.H. Aboutalebi, F. Sharif, J.-K. Kim, Self- 5821e5829.
alignment and high electrical conductivity of ultralarge graphene oxide- [35] J.J. Hermans, P.H. Hermans, D. Vermaas, A. Weidinger, Quantitative evaluation
polyurethane nanocomposites, J. Mater. Chem. 22 (25) (2012) 12709e12717. of orientation in cellulose fibres from the X-ray fibre diagram, Recl. Des. Trav.
[6] D.R. Bortz, E.G. Heras, I. Martin-Gullon, Impressive fatigue life and fracture Chim. des Pays-Bas 65 (6) (1946) 427e447.
toughness improvements in graphene oxide/epoxy composites, Macromole- [36] N.R. Wilson, P.A. Pandey, R. Beanland, R.J. Young, I.A. Kinloch, L. Gong, et al.,
cules 45 (1) (2011) 238e245. Graphene oxide: structural analysis and application as a highly transparent
[7] J. Liang, Y. Huang, L. Zhang, Y. Wang, Y. Ma, T. Guo, et al., Molecular-level support for electron microscopy, ACS Nano 3 (9) (2009) 2547e2556.
dispersion of graphene into poly(vinyl alcohol) and effective reinforcement of [37] J.P. Rourke, P.A. Pandey, J.J. Moore, M. Bates, I.A. Kinloch, R.J. Young, et al., The
their nanocomposites, Adv. Funct. Mater. 19 (14) (2009) 2297e2302. real graphene oxide revealed: stripping the oxidative debris from the
[8] X. Zhao, Q. Zhang, D. Chen, P. Lu, Enhanced mechanical properties of graphene-like sheets, Angew. Chem. Int. Ed. 50 (14) (2011) 3173e3177.
graphene-based poly(vinyl alcohol) composites, Macromolecules 43 (5) [38] M. van Gurp, The use of rotation matrices in the mathematical description of
(2010) 2357e2363. molecular orientations in polymers, Colloid Polym. Sci. 273 (7) (1995)
[9] Z.-Z. Yu, M. Yang, Q. Zhang, C. Zhao, Y.-W. Mai, Dispersion and distribution of 607e625.
Z. Li et al. / Composites Science and Technology 123 (2016) 125e133 133

[39] P.P. Castan~ eda, P. Suquet, Nonlinear composites, in: G. Erik van der, Development of a universal stress sensor for graphene and carbon fibres, Nat.
Y.W. Theodore (Eds.), Advances in Applied Mechanics, Elsevier, 1997, pp. Commun. 2 (2011) 255.
171e302. [44] K. Urayama, T. Takigawa, T. Masuda, Poisson's ratio of poly(vinyl alcohol) gels,
[40] C.A. Cooper, R.J. Young, M. Halsall, Investigation into the deformation of car- Macromolecules 26 (12) (1993) 3092e3096.
bon nanotubes and their composites through the use of Raman spectroscopy, [45] A. Smith, S.J. Wilkinson, W.N. Reynolds, The elastic constants of some epoxy
Compos. Part A Appl. Sci. Manuf. 32 (3e4) (2001) 401e411. resins, J. Mater. Sci. 9 (4) (1974) 547e550.
[41] L. Deng, S.J. Eichhorn, C.-C. Kao, R.J. Young, The effective Young's modulus of [46] H. Lu, X. Zhang, W.G. Knauss, Uniaxial, shear, and poisson relaxation and their
carbon nanotubes in composites, ACS Appl. Mater. Interfaces 3 (2) (2011) conversion to bulk relaxation: studies on poly(methyl methacrylate), Polym.
433e440. Eng. Sci. 37 (6) (1997) 1053e1064.
[42] D. Hull, An Introduction to Composite Materials, first ed., Cambridge Uni- [47] J.W. Suk, R.D. Piner, J. An, R.S. Ruoff, Mechanical properties of monolayer
versity Press, Cambridge, 1981. graphene oxide, ACS Nano 4 (11) (2010) 6557e6564.
[43] O. Frank, G. Tsoukleri, I. Riaz, K. Papagelis, J. Parthenios, A.C. Ferrari, et al.,