Accepted Manuscript Tailored glass fiber interphases via electrophoretic deposition of carbon nanotubes: Fiber and interphase characterization Qi An, Sandeep Tamrakar, John W. Gillespie, Jr., Andrew N. Rider, Erik T. Thostenson PII:
S0266-3538(17)32625-8
DOI:
10.1016/j.compscitech.2018.01.003
Reference:
CSTE 7029
To appear in:
Composites Science and Technology
Received Date: 20 October 2017 Revised Date:
12 December 2017
Accepted Date: 4 January 2018
Please cite this article as: An Q, Tamrakar S, Gillespie Jr. JW, Rider AN, Thostenson ET, Tailored glass fiber interphases via electrophoretic deposition of carbon nanotubes: Fiber and interphase characterization, Composites Science and Technology (2018), doi: 10.1016/j.compscitech.2018.01.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
TAILORED GLASS FIBER INTERPHASES VIA ELECTROPHORETIC DEPOSITION OF CARBON NANOTUBES: FIBER AND INTERPHASE CHARACTERIZATION
2
3
4 5
Center for Composite Materials, University of Delaware, Newark, DE 19716 USA Department of Materials Science & Engineering, University of Delaware, Newark, DE 19716 USA Department of Civil & Environmental Engineering, University of Delaware, Newark, DE 19716 USA Department of Mechanical Engineering, University of Delaware, Newark, DE 19716 USA Defence Science and Technology Group, Fisherman’s Bend, Victoria 3207, Australia
M AN U
ABSTRACT
SC
1
RI PT
Qi An1,2, Sandeep Tamrakar1,3, John W. Gillespie Jr.1-4, Andrew N. Rider5, Erik T. Thostenson1,2,4*
Electrophoretic deposition (EPD) has been used to deposit carbon nanotubes (CNTs) onto glass fibers to strengthen the fiber/matrix interphase of glass/epoxy composites. CNTs were functionalized using ultrasonicated ozonolysis followed by polyethyleneimine (PEI) treatment,
TE D
creating a stable, aqueous dispersion suitable for EPD. Single E-glass fiber filaments were successfully coated with the functionalized CNTs using EPD. Single fiber tensile and microdroplet debond tests were conducted to investigate the tensile properties of the CNT
EP
modified E-glass fibers and their interfacial structure and properties in an epoxy matrix, respectively. Weibull analysis of the fiber testing revealed no detrimental effects resulted from
AC C
the CNT coating, with some evidence to suggest slightly improved strength. Microdroplet tests revealed changes in the fracture modes due to the application of the CNT coating. The interfacial shear-sliding shifted from the fiber/resin interface to the CNT/resin interphase for the CNT modified fibers. Higher effective interfacial shear strength corresponded to the results where fracture propagated deeper into the CNT-rich interphase layer, confirming the trends observed in previous model interphase studies.
*
Corresponding author. Tel: +1 302 831-8789. Email address:
[email protected]
ACCEPTED MANUSCRIPT
1. Introduction Advanced fiber composites have been widely used in structural applications due to their high specific strength, outstanding corrosion resistance and good design flexibility as compared
RI PT
with conventional structural materials, such as aluminum and steel. Mechanical and physical properties of composites are influenced by the constituent materials, the interface between the fiber and the polymer matrix, and the fiber architecture. In terms of mechanical properties, the
SC
interfacial region plays a critical role in the stress-transfer mechanisms between the fibers and matrix and affects the overall structural properties [1]. Therefore, an understanding of the
M AN U
mechanisms influencing fiber/matrix adhesion is essential for the development of improved composite materials. Carbon nanotubes (CNTs), with their high stiffness and strength combined with the nanoscale size – three orders of magnitude smaller diameter than the fiber – offer unique opportunities to strengthen the fiber/matrix interface.
TE D
To explore ways of modifying the interphase properties of advanced composites, systematic changes to the fiber coating systems have been made to initially establish the influence on the intrinsic fiber mechanical properties. A recent review summarizes the scientific
EP
and technological achievements in interphase tailoring via fiber sizing, nanomaterials, fiber surface modification and matrix modifications [2]. Tanoglu and co-workers [3] experimentally
AC C
investigated the sizing effects on the strength and energy absorption of the E-glass-fiber/epoxyamine interface using a dynamic micro-debonding technique. They found that the compatible sizing on the glass fiber increased chemical bonding and adhesion between the inorganic glass fibers and the organic epoxy resin and strengthened the fiber/matrix interphase. However, the sizing reduced the fiber surface roughness and decreased the energy absorbing capability as a result of the reduced frictional stresses in the interface. Iglesias et al. [4] studied sizing effects on
1
ACCEPTED MANUSCRIPT
the mechanical performance of glass fiber composites by SEM fractographic analysis. A series of glass fibers were pretreated with different silane coupling agents (SCA), infused with epoxy resin and cured and finally tested to failure in tension. Adhesive failure modes were observed
RI PT
and they concluded that crosslink density in the coupling region increased as the accessibility of the functional groups of coupling layer increased. Feih et al. [5] investigated different SCAbased fiber sizings on the effect of the strength and fracture toughness of glass fiber composites.
SC
Results indicated that a strong interface provided higher transverse strength and crack initiation loads, while a weaker interface led to an enhanced degree of fiber bridging resulting in a higher
M AN U
interlaminar fracture toughness. Other studies examined the interfacial shear strength [6] and fiber strength [7] with the addition of different sizings. The tensile strength of the unsized fibers was 10 times lower than theoretical values due to surface defects. The study was able to determine the failure probability for unsized and sized fibers, concluding that sizing helped
TE D
reduce the influence of the surface defects, leading to higher strength. More recently, a fiber pullout study was used to determine how different particle size and distribution of sizing agents affected both carbon fiber strength and interfacial shear strength (IFSS) [8].
EP
Additionally, Gao and co-workers [9] demonstrated how the use of nanoscale silica
AC C
incorporated in the fiber sizing could be used to design interfaces that possessed high energy absorption and good structural properties. Their study detailed the role of sizing chemistry and surface roughness on glass fiber/epoxy composites performance using the microdroplet test. The chemical bond formed with compatible SCAs helped increase the IFSS while the texture created by the presence of nano-scale silica particle on the fiber surface increased energy absorption. In separate studies a series of fiber-sizing formulations were chosen to compare the effect of the SCA and film former on the properties of glass fiber/epoxy interphases [10]. It was concluded 2
ACCEPTED MANUSCRIPT
that, in addition to the SCA and sizing chemistry, that film formers played a critical role in glass fiber-matrix adhesion by controlling resin wetting of the fiber. The fiber /matrix interfacial
turn, controlled the homogeneity of the interphase region.
RI PT
properties could be adjusted by choosing different SCA / film former combinations which, in
Based on advancements in nanotechnology over the past two decades, nanoscale
SC
reinforcing fillers such as CNTs may be synthesized and integrated into the traditional composites to modify the interphase region, leading to enhancements in the mechanical and
M AN U
electrical properties of composites [11, 12]. Specifically, due to the exceptionally high surface area and aspect ratio of nanofillers, such as CNTs, the area of the newly formed interface in the nanocomposites is dramatically modified and typically an order of magnitude greater than in traditional composites. As a result, by introducing the nanomaterials into the composite, it is possible to tune the level of adhesion at the fiber/matrix interface through adjusting both the
TE D
heterogeneity of surface morphology and chemical reactivity [9,13,14] and then potentially increase the overall load-carrying capacity of the composite. For example, Godara et al. [15] have experimentally investigated the interfacial strength of glass fiber/epoxy composites that
EP
were modified by dispersing CNTs in the fiber sizing formulation or the matrix. They observed
AC C
over 90% increase in IFSS in the composite with the CNT-modified sizing. In addition, our previous studies have successfully applied the electrophoretic deposition (EPD) method to coat CNTs on both carbon and glass fibers using functionalized CNTs [17, 18]. Both CNT-modified carbon and glass/epoxy composites showed between 70% and 80% increase in the in-plane shear strength. While many of these experimental studies have proven the unique enhancements nanomaterials offer, these studies often neglect to study the interfacial region [11, 19]. Some of 3
ACCEPTED MANUSCRIPT
the studies are theoretically based [20,21], creating high fidelity models of the interfacial region. However, due to the complex nature of the interfacial and interphase regions modified with nanomaterials, it is often difficult to design experiments that can confirm the specific interactions
RI PT
examined in the models. For similar reasons, most practical engineering models of composite structures do not include the influence of nanoscale interactions on macro-scale mechanical properties [22-24]. Based on the successful interface research of conducting micro-droplet test
SC
on glass fiber/epoxy composite materials by Gao et al. [9], in this study we have used similar test
mechanisms at the fiber/resin interface.
M AN U
setup for CNT-modified glass fibers and focused on investigating the microscale load-transfer
Two main techniques are commonly utilized to integrate CNTs onto raw fibers; including both CVD [16,25-29] and EPD [17,18,30-33]. The high temperature CVD process offers a direct method for perpendicular CNT growth on the fibers [28,29] and complete coating coverage with
TE D
high concentrations of CNTs [25,26,34]. However, the CVD approach can reduce the fiber strength and modulus at the elevated temperature [35]. EPD offers a commonly applied industrial two-step coating process in which charged particles migrate under an applied field to form a
EP
consolidated film on the electrode surface [36]. EPD is energy efficient and capable of homogenously coating a variety of practical materials with compact films [37-39]. Successful
AC C
EPD relies on the functionalization of the dispersed material, which enables a surface charge to develop. The surface charge, or zeta-potential, is dependent on the solution pH and helps to aid dispersion and mobility under applied electric fields [40-43]. The EPD process has the benefit that the deposition may be carried-out under ambient conditions and allows for manipulation of the nanomaterial chemistry for the specific application. Our recent research has applied this efficient EPD process for depositing functionalized CNTs onto conductive carbon fiber [18] and
4
ACCEPTED MANUSCRIPT
non-conductive glass fiber substrates [17]. For depositing onto non-conductive glass fiber substrates In our previous composite and model interphase study, significant improvements in
RI PT
strength and toughness have been achieved for the EPD-CNT reinforced fiber/epoxy composites. CNT modification the interphase contributed to these significant improvements by forcing the fracture path away from the matrix/fiber interface an into the CNT-rich interphase region. The
SC
aim of the current study is to characterize the micro-mechanical properties of E-glass fibers coated with CNTs via EPD. The initial measurements examined the intrinsic effect of the EPD
M AN U
treatment on glass fiber strength and micro-droplet tests provided insight into the influence of the CNT-modified interface on the micro-mechanical shear failure mechanisms occurring in glass/epoxy composites. 2. Experimental
TE D
2.1. Materials and Processing
To study the influence of CNTs on the fiber/matrix interface, single E-glass fibers were extracted from a tow in a unidirectional E-glass fabric (style 7721, 203 g/m2, aminopropylsilane,
EP
APS, sizing, Thayercraft Inc., USA). The measured average diameters of the fibers were 10 µm. Multi-walled CNTs (Hanwha, Nanotech, Korea) were functionalized using a novel
AC C
ultrasonicated-ozonlysis (USO) process [17, 18] to produce a stable 1 g/L aqueous dispersion. Moistrure-free oxygen gas with a concentration of 90 - 95% was generated by a an oxygen concentrator (OxyMax 8, Longevity Resources Inc., Canada) as a continuous oxygen source and connected to an ozone generator (EXT120-T, Longevity Resources Inc., Canada) to treat the aqueous CNT solution maintained at 5°C. 16 hours of USO treatment breaks down CNT agglomerates and provides a stable dispersion. The ozone treated CNTs were further
5
ACCEPTED MANUSCRIPT
functionalized with a polyethyleneimine (PEI) dendrimer (Sigma Aldrich, Canada) with a Mw of 25,000. PEI was added at 1 g/L to the USO treated CNT solution and adjusted to a pH of 6 with
RI PT
acetic acid, followed by a futher a 4 hours of sonication. 2.2. Coating of Glass Fibers with Nanotubes
EPD of the CNTs was performed on individual glass fibers that were attached to an
SC
isolated window frame using double-sided tape, as shown in Figure 1a, with a coated length of 7 cm. The fixture side view from Figure 1b shows the window frame with single fibers sandwiched
M AN U
in the middle was pressed down touching the cathode to ensure intimate contact between the non-conductive E-glass fiber and the cathode. As demonstrated in our previous work on depositing carbon nanotubes onto E-glass fibers [17] the carbon nanotube film forms on the fiber surface by first depositing on the cathode and then the conductive network forms on the surface of the fiber as an extension of the electrode, requiring the need for intimate contact with the
TE D
cathode. Teflon spacers were used to maintain the electrode spacing. Stainless steel electrodes were connected to a DC power source and the whole fixture was immersed vertically in 140 mL
AC C
12 min.
EP
of the CNT dispersion. Cathodic deposition was conducted under field strengths of 64 V/cm for
6
ACCEPTED MANUSCRIPT
RI PT
Copper Tape
Single E-glass Fibers Spacers Isolated Window Frame
(a)
Stainless Steel Electrodes
Anode
M AN U
Cathode
SC
Window Frame with Single Fibers
(b)
Figure 1 Single electrophoretic deposition approach (a) front view of fiber holder (b) side view of the entire fixture. 2.3. Single Fiber Tensile Characterization
TE D
To test the fiber tensile strength, the single-fiber specimens were mounted onto cardboard sample holders with a window that matched the gage length of 25 mm. The fiber ends were attached to the cardboard under slight tension with a drop of instant-drying super glue and paper
EP
tape based on ASTM D3822M-14 [44]. Single-fiber tensile testing was performed using an
AC C
Instron 5848 Micro Tester with a 5N load cell. After the cardboard ends were gripped in the clamps of the testing machine, the sides of the window holder were cut off to make sure only the fiber itself was under load during the test. The fiber diameter was the average of five measurements taken at different locations along the fiber axis in the gage section using a Nikon UDM Eclipse LV1000 Optical Microscope. Determinations of fixture compliance were performed following ASTM D 3379-75 [45], using gage lengths of 15, 20 and 25 mm. Specimens were tested under quasi-static tensile loading at a loading rate of 0.1 mm/min and all
7
ACCEPTED MANUSCRIPT
test parameters were maintained constant for all fibers. For each fiber type 50 to 60 samples were tested. Single fiber statistical strength was analyzed by a 2-parameter Weibull distribution [46].
RI PT
The Weibull probability distribution was used to determine the cumulative distribution function of the strength of E-glass fibers with and without CNT coatings [46]. The cumulative probability of failure (Pf) is defined as
(2)
SC
= 1 − exp −
where σf is the fiber tensile strength, σ0 is the scale parameter and m is the Weibull shape
M AN U
parameter, which indicates the distribution in fiber strength. The scale and shape parameters of the Weibull distribution function for fiber strength can be estimated from experimental measurements of single-fiber tensile testing. Taking double logarithms of Equation (2) provides, ln −ln
= ln
+
ln
−
ln
(3)
ln −ln
is a linear function of ln .
with the slope equal to m and the intercept equal to
EP
ln
TE D
Fitting the experimental data to Equation (3), for L=1 at a fixed gage length of 25 mm,
2.4. Interfacial Characterization
AC C
The microdroplet test was used to characterize the fiber/matrix interface. Single fibers were first suspended under tension, followed by application a droplet of uncured epoxy resin in the liquid state using the tip of a 6 µm diameter carbon fiber. The droplet was a mixture of a low viscosity bisphenol-A/F based epoxy resin DER 353 (Dow Chemical Company) and curing agent bis (p-aminocyclohexyl) methane (PACM-20, Air Products and Chemicals, Inc.) at a stoichiometric weight ratio of 100:28. The droplet was allowed to gel at room temperature for 5 hours, followed by a cure cycle of 2 hours at 80°C and a post cure of 2 hours at 150°C. Droplet 8
ACCEPTED MANUSCRIPT
size and shape, fiber diameter, fiber embedded length, and fiber gage length were measured by an optical microscope. Only drops with diameters of 70-150 µm and a symmetric shape were
RI PT
chosen for testing. Samples were glued to a glass tube sample holder with hanger for testing. The microdroplet test used a customized setup established by Gao [9] which consisted of an actuator driven load frame (Newport ESP 300 and Newport LTA-HS) and an open pan
SC
balance (Mettler Toledo PB303-S) as the load cell. The three–dimensional controlled loading blades were precisely positioned above the droplet with the aid of two cameras located in
M AN U
horizontal (PixeLink PL-A782, 6.6 Megapixel camera) and vertical (Allied Vision Tech GUPPY 5 Megapixel CCD machine vision camera) directions. During the test, the separation of knife edges was kept slightly larger than the tested fiber diameter with either knife approximately 5 µm from the fiber. The knife contacted the surface of the droplet to apply load and shear off the droplet along the fiber direction through the fiber/resin interphase. Specimens were tested at a
TE D
displacement rate of 0.001 mm/s and the force-displacement curve was collected by National Instruments LabView 10 software program for further data analysis. The IFSS was calculated as
AC C
debonds:
EP
the maximum average shear stress at the fiber surface at the point where the interface/interphase
=
(1)
!" #$
where τmax is the interfacial shear strength, F is the maximum fiber axial force recorded in the load-displacement curve, % is the diameter of the fiber and
&
is the embedded fiber length, as
defined in Figure 2. Over 20 specimens with embedded fiber lengths ranging from 70 to 150 µm were tested for the uncoated and CNT-PEI EPD coated fibers, with approximately half yielding valid shear strengths.
9
RI PT
ACCEPTED MANUSCRIPT
Figure 2 Microdroplet specimen configuration and definition, where Lg is the gage length.
Fracture modes after microdroplet testing were investigated using a field emission
3. Results and Discussion 3.1. Fiber Surface Morphology
M AN U
sputter deposited Pt/Au layer to prevent charging.
SC
scanning electron microscopy (AURIGA 60 Crossbeam FIB-SEM) operating at 3 kV with a
Figure 3 shows the surface morphology of as-received E-glass fiber and CNT-PEI EPD coated E-glass fiber. It can be seen from Figure 3a that the surface of the as-received E-glass
TE D
fiber was not smooth with some small heterogeneous features from the sizing agent, indicating the SCA provided some level of surface texture, which could improve the IFSS by altering
EP
interfacial sliding mechanisms. Figure 3(b) confirms that the EPD process was very effective, with evidence of a uniform PEI-CNT coating present on the fiber surface. The increased surface
AC C
roughness and porosity created by the EPD-CNT film could improve IFSS by enhancing adhesion and friction during debonding. Figure 4 shows a higher magnification micrograph of the carbon nanotube film where a section of the fiber and coating has been cut away using a focused ion beam (FIB), showing a highly-uniform, nanoporous film of carbon nanotubes on the fiber surface.
10
SC
RI PT
ACCEPTED MANUSCRIPT
EP
TE D
M AN U
(a) (b) Figure 3 SEM images of fiber surface morphology of (a) as received E-glass fiber and (b) Eglass fiber electrophoretically coated with functionalized CNTs.
AC C
Figure 4 SEM of the fiber surface where the fiber and film has been cross-sectioned using a focused-ion beam.
3.2. Single Fiber Tensile Properties Table 1 shows the elastic modulus and ultimate tensile strength for the uncoated and
CNT-PEI EPD coated E-glass fibers. The similarity in the modulus and strength for the glass fibers with or without CNT coatings illustrate that addition of EPD coated CNTs did not degrade fiber tensile properties. A Weibull distribution was created to illustrate the probability of failure 11
ACCEPTED MANUSCRIPT
at different ultimate strengths, shown in Figure 5. The results show that in all cases, the CNTcoated fibers were less likely to fail at any given ultimate strength. This further supports the conclusion that EPD did not degrade the fiber properties. The higher shape parameter m for the
RI PT
CNT-coated E-glass fiber indicated a reduced scatter, which is consistent with the CNT coating reducing the effective density of defects.
SC
Table 1 Weibull distribution parameters and fiber tensile properties. Elastic Shape Parameter, Scale Parameter, Fiber Type Modulus m σo (GPa) (GPa) Uncoated E-glass 2.81 1.09 61.1 ± 8.5 CNT-PEI E-glass 3.52 1.31 62.4 ± 6.4
1
Uncoated Fibers CNT Coated Fibers
Baseline Fibers
TE D
Ln (1/(1-P))
0 -1
-3 -4 -1.5
-1
EP
-2
-0.5
0
0.5
Probability to Failure
1
973 ± 383 1,170 ± 358
M AN U
2
Strength (MPa)
CNT Coated Fibers
0.8
0.6
0.4
0.2
0
1
0
AC C
Ln(σ σ)
0.5
1
1.5
2
Ultimate Strength (GPa)
(a)
(b)
Figure 5 (a) Weibull plot and (b) cumulative probability of failure from single-fiber tensile testing.
3.3. Interfacial Properties
12
ACCEPTED MANUSCRIPT
The interfacial properties of EPD-CNT coated E-glass fibers embedded in epoxy resin were analyzed and compared by microdroplet test. In order to have a successful test, the specimen should fail through the fiber-droplet interphase and before fiber tensile failure takes
RI PT
place. A typical force-displacement response from the microdroplet test of CNT coated Eglass/epoxy specimen with an embedded fiber length of 102 µm is shown in Figure 6. The forcedisplacement plot initially increased almost linearly under elastic loading. Fiber extension
SC
occurred initially under the applied load and initial debonding between CNT coated fiber and resin led to the initial load-drop. Once complete debonding occurred, the fiber elastically
M AN U
recoiled and the frictional stress was generated by the dynamic sliding of the droplet over the fiber. The level of sliding force was dependent on the locus of fracture and the texture of the fiber surface. 200
Interfacial Debonding
TE D
F
100
EP
Force (mN)
150
Frictional Sliding
AC C
50
0
0
Elastic Loading
50
100
150
200
250
Displacement (µ µm)
Figure 6 Typical force-displacement curve for droplet debonding recorded for a microdroplet test of CNT coated E-glass/epoxy specimen with an embedded length of 102 µm.
Figure 7 shows the IFSS versus le for commercially sized and EPD-CNT coated E-glass fibers. The average IFSS for baseline E-glass fiber/epoxy specimens was 40.3 MPa with a
13
ACCEPTED MANUSCRIPT
coefficient of variation (COV) of 10.9% and 51.7 MPa with COV of 13.3% for the CNT coated E-glass fiber/epoxy specimens, respectively. This result indicates that the presence of the CNTs led to a 30% higher failure load at the point just prior to complete droplet debonding. The 30%
RI PT
increase in IFSS for the EPD-CNT fiber treatment is in good agreement with previous results from the more complex composite system [17] and the model interphase studies [47]. It should be noted that in the current work that a different epoxy resin with higher viscosity was used,
SC
which may affect the wetting and adhesion with the fiber surface, leading to reduced penetration of the porous CNT coating on the fiber surface. At a composite level, 80% improvement for in-
M AN U
plane shear strength was achieved with up to 15% volume fraction of EPD-CNTs, compared to the laminate without CNTs. The change in shear strength correlated directly with the quantity of CNTs deposited onto the glass fabric prior to resin infusion. Microscopic images confirmed that the increased toughness of the composites was due to the high-density CNT coating, which
TE D
occupied the resin-rich interstitial regions and formed a network to connect reinforcing fibers and polymer matrix. The additional PEI functionality formed localized, high cross-link density bonds with both the epoxy resin and the sized fiber surface, also assisted in film toughening. The
EP
model interphase study, where the glass fiber was replaced by a planar glass substrate with different sizing treatments, also confirmed that the formation of the PEI-CNT rich interphase
AC C
facilitated a more homogeneous, ductile fracture zone relative to the brittle glass epoxy interface, and led to a 50% higher shear strength.
14
ACCEPTED MANUSCRIPT
80
70 60 50 40 30 20
0 80
90
100
110
120
130
140
60 50 40 30 20 10 0 80
SC
10
70
RI PT
Interfacial Shear Strength (MPa)
Interfacial Shear Strength (MPa)
80
150
Embedded Length (µ µ m)
90
100
110
120
130
140
150
Embedded Length (µ µ m)
(b)
M AN U
(a)
Figure 7 Interfacial shear strength versus fiber embedded length of the (a) commercially sized E-glass/epoxy specimens and (b) CNT EPD coated E-glass/epoxy specimens.
Based on the previous findings, the increased IFSS of the CNT coated E-glass fiber can
TE D
be attributed to the change in surface roughness and morphology provided by the CNTs which facilitate increased levels of mechanical interlocking between the fiber and resin. Additionally, functionalization of CNTs enables formation of direct chemical bonding with the glass fiber and
EP
the epoxy resin. It is important to note that the aminopropylsilane (APS) based sizing chemistry for glass fibers is already well developed and that IFSS value for commercially sized E-glass
[48].
AC C
fiber in the order of 40 MPa is already 35% higher than unsized E-glass fibers (IFSS=26±3 MPa)
3.4. Failure Mode Analysis Fracture surfaces of the as-received and CNT-EPD coated E-glass microdroplet test samples were characterized by SEM in Figure 8. On the fracture surfaces a heterogeneous resin layer was left on the fiber surface after debonding, indicating inconsistent bonding between the
15
ACCEPTED MANUSCRIPT
sized fiber and epoxy resin drop. During the debonding stage, crack propagation occurred either within the epoxy resin near the interphase region or at the epoxy/glass interface. In the previous model interphase study, the APS-sized glass substrates also showed higher levels of interfacial
RI PT
failure than CNT-modified samples, and detailed surface characterization using x-ray photoelectron spectroscopy (XPS) suggests the formation of covalent bonds between the PEI-
SC
CNT and the surface [47].
Figure 9 shows SEM images of fracture surfaces from EPD-CNT microdroplet specimens.
M AN U
A comparison of the fracture surfaces at the different locations along the fiber is made for a stronger IFSS (60 MPa) and weaker IFSS (45 MPa). At the point where initial fracture occurred the stronger IFSS showed increased covering of a CNT-rich interphase region compared to the weaker IFSS, where there is a distinct separation between the CNT-rich region and the more featureless fiber surface. This suggests that the CNT-modified surface increased the level of
TE D
interphase fracture that was linked to higher strength. However, the heterogeneous failure for the stronger IFSS suggests that there could be further improvement if there was more consistent wetting across the fiber surface. In the current work the higher viscosity of the epoxy resin
EP
needed to form the microdroplets may also reduce the resin diffusion to the glass interface,
AC C
particularly for the thicker PEI-CNT coating.
16
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
Figure 8 SEM images of the commercially sized E-glass/epoxy microdroplet fracture specimen after interfacial debonding. An overall view of the displacement location of the microdroplet along the fiber before and after interfacial debonding showing (top) the fracture zone of the microdroplet in its original location where it starts loading, (middle) the sliding area along the fiber during the test loading and (bottom) the microdroplet in its final location where it stops loading.
17
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 9 SEM images of the CNT coated E-glass/epoxy microdroplet fracture specimens after interfacial debonding with a higher and lower interfacial shear strength. An overall view of the displacement location of the microdroplet along the fiber before and after interfacial debonding showing (top) the fracture zone of the microdroplet in its original location where it starts loading, (middle) the sliding area along the fiber during the test loading, and (bottom) the microdroplet in its final location where it stops loading. In Figure 9 for the stronger and weaker IFSS microdroplet samples, there is a distinct
difference in the fiber surface morphologies in the regions corresponding to the droplet sliding. The high level of CNT-coating shows that the higher IFSS also corresponds to increased adhesion between the CNT coating and fiber, which also suggests improved resin wetting of the EPD coating has led to an improved interphase. In comparison the weaker IFSS sample shows
18
ACCEPTED MANUSCRIPT
much less evidence of the CNT-rich interphase coating on the fiber, which may also suggest that lower strength corresponds to poorer resin wetting which has compromised the interphase formation. The amount of interphase material removed during sliding is also consistent with
RI PT
different levels of adhesion for the stronger and weaker IFSS samples. The higher IFSS shows less of the CNT-rich interphase material has been removed from the fiber surface during sliding
SC
compared to the weaker IFSS sample.
The overall results show the sensitivity of the IFSS to the interphase formation processes
M AN U
and the resin wetting and resultant adhesion between the matrix and sized glass fiber. In cases where the resin penetrated the thick PEI-CNT modified glass fiber there was a clear increase in the IFSS, corresponding to increased levels of fracture through a CNT-rich interphase region. In weaker IFSS samples there was evidence that poor resin wetting compromised the interphase formation and the resultant fiber to matrix adhesion. The weaker CNT-modified fibers and the
TE D
commercially-sized fibers provided similar fracture surfaces, where higher levels of failure at the fiber/matrix interface led to lower strength. The current results support the experimental findings from previous work on CNT-modified macro-composite systems, in which improved
EP
performance was linked to increased fracture away from the fiber/matrix interphase and into the
AC C
CNT-rich interphase zone. 4. Conclusions
EPD was successfully applied to coat single E-glass fibers with ozone and PEI
functionalized CNTs. Single-fiber tensile testing indicated that the CNT coating did not degrade the tensile strength or stiffness of the fibers. Statistical analysis of the fiber fracture strength suggested that the EPD-CNT coating reduced the influence of surface defects on the commercially sized fiber and led to slightly improved mechanical properties. Microdroplet 19
ACCEPTED MANUSCRIPT
testing showed increased IFSS with CNT-coated fibers compared to commercially-sized fibers. SEM analysis indicated that failure mechanisms changed with addition of the EPD coated CNTs. Microdroplet specimens with higher IFSS exhibited higher levels of fracture through a CNT-rich
RI PT
interphase region. In samples with lower IFSS, the fracture surfaces exhibited larger areas of fracture at the glass/matrix interface, suggesting that resin wetting through the thicker CNT coating may have been compromised in some cases. The results confirm that increased IFSS is
SC
directly related to well-formed interphase regions which are facilitated by good resin wetting. The CNT-rich interphase encourages fracture to propagate away from the glass interface and
M AN U
leads to higher IFSS. Acknowledgement
Q.A. and E.T.T gratefully acknowledge the funding support from University of Delaware Graduate Student Office Fellowship Award and National Science Foundation (Grant CMMI-
TE D
1254540: Dr. Mary Toney, Program Director). S.T. and J.W.G. would like to acknowledge a research sponsorship by the Army Research Laboratory, which was accomplished under Cooperative Agreement Number W911NF-12-2-0022. The views and conclusions contained in
EP
this document are those of the authors and should not be interpreted as representing the official
AC C
policies, either expressed or implied, of the Army Research Laboratory or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes, notwithstanding any copyright notation herein References
[1] L. Drzal, M. Madhukar, Fibre-matrix adhesion and its relationship to composite mechanical properties, J. Mater. Sci. 28 (3) (1993) 569-610.
20
ACCEPTED MANUSCRIPT
[2] J. Karger-Kocsis, H. Mahmood, A. Pegoretti, Recent advances in fiber/matrix interphase engineering for polymer composites, Progress in Materials Science 73 (2015) 1-43.
RI PT
[3] M. Tanoglu, S. McKnight, G. Palmese, J. Gillespie Jr, The effects of glass-fiber sizings on the strength and energy absorption of the fiber/matrix interphase under high loading rates, Composites Sci. Technol. 61 (2) (2001) 205-220. [4] J. Iglesias, J. González-Benito, A. Aznar, J. Bravo, J. Baselga, Effect of glass fiber surface treatments on mechanical strength of epoxy based composite materials, J. Colloid Interface Sci. 250 (1) (2002) 251-260.
SC
[5] S. Feih, J. Wei, P. Kingshott, B.F. Sørensen, The influence of fibre sizing on the strength and fracture toughness of glass fibre composites, Composites Part A: Applied Science and Manufacturing 36 (2) (2005) 245-255.
M AN U
[6] J. Berg, F.R. Jones, The role of sizing resins, coupling agents and their blends on the formation of the interphase in glass fibre composites, Composites Part A: Applied Science and Manufacturing 29 (9–10) (1998) 1261-1272. [7] S.L. Gao, E. Mäder, A. Abdkader, P. Offermann, Environmental resistance and mechanical performance of alkali-resistant glass fibers with surface sizings, J. Non Cryst. Solids 325 (1– 3) (2003) 230-241.
TE D
[8] R.L. Zhang, Y. Liu, Y.D. Huang, L. Liu, Effect of particle size and distribution of the sizing agent on the carbon fibers surface and interfacial shear strength (IFSS) of its composites, Appl. Surf. Sci. 287 (0) (2013) 423-427.
EP
[9] X. Gao, R.E. Jensen, S.H. McKnight, J.W. Gillespie Jr., Effect of colloidal silica on the strength and energy absorption of glass fiber/epoxy interphases, Composites Part A: Applied Science and Manufacturing 42 (11) (2011) 1738-1747.
AC C
[10] M. Dey, J.M. Deitzel, J.W. Gillespie Jr., S. Schweiger, Influence of sizing formulations on glass/epoxy interphase properties, Composites Part A: Applied Science and Manufacturing 63 (0) (2014) 59-67. [11] G. Pandey, E.T. Thostenson, Carbon nanotube-based multifunctional polymer nanocomposites, Polymer Reviews 52 (3) (2012) 355-416. [12] E.T. Thostenson, Z. Ren, T. Chou, Advances in the science and technology of carbon nanotubes and their composites: A review, Composites Sci. Technol. 61 (13) (2001) 18991912. [13] E. Mäder, H. Jacobasch, K. Grundke, T. Gietzelt, Influence of an optimized interphase on the properties of polypropylene/glass fibre composites, Composites Part A: Applied Science and Manufacturing 27 (9) (1996) 907-912.
21
ACCEPTED MANUSCRIPT
[14] J. Berg, F. Jones, The role of sizing resins, coupling agents and their blends on the formation of the interphase in glass fibre composites, Composites Part A: Applied Science and Manufacturing 29 (9) (1998) 1261-1272.
RI PT
[15] A. Godara, L. Gorbatikh, G. Kalinka, A. Warrier, O. Rochez, L. Mezzo, F. Luizi, A.W. van Vuure, S.V. Lomov, I. Verpoest, Interfacial shear strength of a glass fiber/epoxy bonding in composites modified with carbon nanotubes, Composites Sci. Technol. 70 (9) (2010) 13461352.
SC
[17] Q. An, A.N. Rider, E.T. Thostenson, Hierarchical composite structures prepared by electrophoretic deposition of carbon nanotubes onto glass fibers, ACS applied materials & interfaces 5 (6) (2013) 2022-2032.
M AN U
[18] Q. An, A.N. Rider, E.T. Thostenson, Electrophoretic deposition of carbon nanotubes onto carbon-fiber fabric for production of carbon/epoxy composites with improved mechanical properties, Carbon 50 (11) (2012) 4130-4143. [19] E.T. Thostenson, C. Li, T. Chou, Nanocomposites in context, Composites Sci. Technol. 65 (3-4) (2005) 491-516. [20] A. Chávez-Valdez, A.R. Boccaccini, Innovations in electrophoretic deposition: Alternating current and pulsed direct current methods, Electrochim. Acta 65 (2012) 70-89.
TE D
[21] E. Bekyarova, E.T. Thostenson, A. Yu, H. Kim, J. Gao, J. Tang, H.T. Hahn, T.-. Chou, M.E. Itkis, R.C. Haddon, Multiscale carbon nanotube-carbon fiber reinforcement for advanced epoxy composites, Langmuir 23 (7) (2007) 3970-3974.
EP
[22] S.B. Lee, O. Choi, W. Lee, J.W. Yi, B.S. Kim, J.H. Byun, M.K. Yoon, H. Fong, E.T. Thostenson, T.W. Chou, Processing and characterization of multi-scale hybrid composites reinforced with nanoscale carbon reinforcements and carbon fibers, Composites Part A: Applied Science and Manufacturing 42 (4) (2011) 337-344.
AC C
[23] M. Li, M. Boggs, T.P. Beebe, C.P. Huang, Oxidation of single-walled carbon nanotubes in dilute aqueous solutions by ozone as affected by ultrasound, Carbon 46 (3) (2008) 466-475. [24] M.L. Sham, J.K. Kim, Surface functionalities of multi-wall carbon nanotubes after UV/ozone and TETA treatments, Carbon 44 (4) (2006) 768-777. [25] E. Thostenson, W. Li, D. Wang, Z. Ren, T. Chou, Carbon nanotube/carbon fiber hybrid multiscale composites, J. Appl. Phys. 91 (9) (2002) 6034-6037. [26] Q. Zhang, J. Liu, R. Sager, L. Dai, J. Baur, Hierarchical composites of carbon nanotubes on carbon fiber: Influence of growth condition on fiber tensile properties, Composites Sci. Technol. 69 (5) (2009) 594-601.
22
ACCEPTED MANUSCRIPT
[27] H. Qian, A. Bismarck, E.S. Greenhalgh, M.S.P. Shaffer, Carbon nanotube grafted carbon fibres: A study of wetting and fibre fragmentation, Composites Part A: Applied Science and Manufacturing 41 (9) (2010) 1107-1114.
RI PT
[28] M. Kumar, Y. Ando, Chemical vapor deposition of carbon nanotubes: A review on growth mechanism and mass production, Journal of nanoscience and nanotechnology 10 (6) (2010) 3739-3758. [29] R.J. Sager, P.J. Klein, D.C. Lagoudas, Q. Zhang, J. Liu, L. Dai, J.W. Baur, Effect of carbon nanotubes on the interfacial shear strength of T650 carbon fiber in an epoxy matrix, Composites Sci. Technol. 69 (7–8) (2009) 898-904.
SC
[30] A.R. Boccaccini, J.A. Roether, B.J.C. Thomas, M.S.P. Shaffer, E. Chavez, E. Stoll, E. Jane Minay, The electrophoretic deposition of inorganic nanoscaled materials, J Ceram Soc Jpn 114 (1325) (2006) 1-14.
M AN U
[31] A.R. Boccaccini, J. Cho, T. Subhani, C. Kaya, F. Kaya, Electrophoretic deposition of carbon nanotube-ceramic nanocomposites, Journal of the European Ceramic Society 30 (5) (2010) 1115-1129. [32] A.R. Boccaccini, J. Cho, J.A. Roether, B.J.C. Thomas, E. Jane Minay, M.S.P. Shaffer, Electrophoretic deposition of carbon nanotubes, Carbon 44 (15) (2006) 3149-3160.
TE D
[33] M.S.P. Shaffer, X. Fan, A.H. Windle, Dispersion and packing of carbon nanotubes, Carbon 36 (11) (1998) 1603-1612. [34] A.N. Rider, E. Yeo, J. Gopalakrishna, E.T. Thostenson, N. Brack, Hierarchical composites with high-volume fractions of carbon nanotubes: Influence of plasma surface treatment and thermoplastic nanophase-modified epoxy, Carbon 94 (2015) 971-981.
EP
[35] J. Cho, K. Konopka, K. Rożniatowski, E. García-Lecina, M.S.P. Shaffer, A.R. Boccaccini, Characterisation of carbon nanotube films deposited by electrophoretic deposition, Carbon 47 (1) (2009) 58-67.
AC C
[36] L. Besra, M. Liu, A review on fundamentals and applications of electrophoretic deposition (EPD), Progress in Materials Science 52 (1) (2007) 1-61. [37] K. Peng, L.Q. Liu, H. Li, H. Meyer, Z. Zhang, Room temperature functionalization of carbon nanotubes using an ozone/water vapor mixture, Carbon 49 (1) (2011) 70-76. [38] K.A. Wepasnick, B.A. Smith, K.E. Schrote, H.K. Wilson, S.R. Diegelmann, D.H. Fairbrother, Surface and structural characterization of multi-walled carbon nanotubes following different oxidative treatments, Carbon 49 (1) (2011) 24-36.
23
ACCEPTED MANUSCRIPT
[39] L.C. Tang, H. Zhang, J.H. Han, X.P. Wu, Z. Zhang, Fracture mechanisms of epoxy filled with ozone functionalized multi-wall carbon nanotubes, Composites Sci. Technol. 72 (1) (2011) 7-13.
RI PT
[40] H. Chiang, C.P. Huang, P.C. Chiang, The surface characteristics of activated carbon as affected by ozone and alkaline treatment, Chemosphere 47 (3) (2002) 257-265. [41] S. Osbeck, R.H. Bradley, C. Liu, H. Idriss, S. Ward, Effect of an ultraviolet/ozone treatment on the surface texture and functional groups on polyacrylonitrile carbon fibres, Carbon 49 (13) (2011) 4322-4330.
SC
[42] I. Zhitomirsky, Electrophoretic deposition of organic-inorganic nanocomposites, J. Mater. Sci. 41 (24) (2006) 8186-8195.
M AN U
[43] J. Sun, L. Gao, Development of a dispersion process for carbon nanotubes in ceramic matrix by heterocoagulation, Carbon 41 (5) (2003) 1063-1068. [44] ASTM D3822/ D3822M-14. Standard test method for tensile properties of single textile fibers, American Society for Testing and Materials International, West Conshohocken, PA, 2014. [45] ASTM D3379-75e1. Standard test method for tensile strength and young's modulus for high-modulus single-filament materials, American Society for Testing and Materials International, West Conshohocken, PA, 1989.
TE D
[46] T.W. Chou, Microstructural Design of Fiber Composites, Cambridge University Press, 2005.
EP
[47] A.N. Rider, Q. An, N. Brack, E.T. Thostenson, Polymer nanocomposite – fiber model interphases: Influence of processing and interface chemistry on mechanical performance, Chem. Eng. J. 269 (0) (2015) 121-134.
AC C
[48] A. Andres Leal, J.M. Deitzel, S.H. McKnight, J.W. Gillespie Jr., Interfacial behavior of high performance organic fibers, Polymer 50 (5) (2009) 1228-1235.
24
