- Email: [email protected]
Enhanced interfacial adhesion of glass fibers by tetravinylsilane plasma modification
Composites: Part A 58 (2014) 84–89
Contents lists available at ScienceDirect
Composites: Part A journal homepage: www.elsevier.com/locate/compositesa
Enhanced interfacial adhesion of glass fibers by tetravinylsilane plasma modification V. Cech a,⇑, A. Knob a, H.-A. Hosein b, A. Babik a, P. Lepcio a, F. Ondreas a, L.T. Drzal b a b
Institute of Materials Chemistry, Faculty of Chemistry, Brno University of Technology, Purkynova 118, CZ-612 00 Brno, Czech Republic Composite Materials and Structures Center, Michigan State University, 2100 Engineering Building, East Lansing, MI 48824-1226, USA
a r t i c l e
i n f o
Article history: Received 22 March 2013 Received in revised form 21 November 2013 Accepted 8 December 2013 Available online 15 December 2013 Keywords: A. Glass fibers A. Polymer–matrix composites (PMCs) B. Interface/interphase E. Surface treatments
a b s t r a c t Plasma-polymerized tetravinylsilane was used to surface modify glass fibers to improve interfacial adhesion of a GF/polyester composite. Plasma polymer films of controllable thickness and physicochemical properties were deposited on unsized glass fibers by RF pulsed plasma using an effective power of 0.1–5 W. The interfacial adhesion of unsized, industrially sized, oxygen plasma treated, and plasma polymer coated fibers embedded in polyester resin was determined by microindentation. The plasma modification of the glass fibers enabled a considerable increase in the interfacial shear strength compared to unsized fibers. The interfacial shear strength for the optimized plasma coating was 26% higher than that for the industrial sizing. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction The development of high-performance fiber-reinforced plastics (FRP) is closely connected with progress in engineered interfaces [1] through controlled interphases [2]. Theoretical and experimental studies have shown that the composite interphase can markedly influence the performance of composites with respect to their strength and toughness [3,4]. Sizing, i.e., functional coating (interlayer), is therefore tailored [5] to improve the transfer of stress from the polymer matrix to the fiber reinforcement by enhancing fiber wettability, adhesion, compatibility, etc. The key parameters affecting the performance of composite materials are the sizing thickness and modulus, the interaction at interfaces, and the reinforcement and matrix materials [4]. However, industrially produced sizing of glass fibers (GF) is not of a consistent thickness and uniformity [5]. Using the wet chemical process, the molecules of silane coupling agents have a tendency to self-condense, forming siloxane oligomers rather than complete bonding with the glass surface [6,7]. Only 10–20% of the total sizing is bonded to the fiber surface, and this amount is directly related to the composite shear strength [8]. The plasmachemical process is another way to improve composite performance via a controlled interphase. Low temperature plasma may be used as a gentle but powerful tool for surface treatment (chemical and/or physical altering of the fiber surface) or coating (plasma polymerization) of fibers, retaining their ⇑ Corresponding author. Tel.: +420 541149304; fax: +420 541149361. E-mail address: [email protected] (V. Cech). 1359-835X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.compositesa.2013.12.003
mechanical properties if appropriate plasma conditions are used. Plasma surface modification of fibers and its application in FRP has been studied since the 1980s, see the review in Ref. [9]. The oxygen plasma may increase the surface roughness and introduce functional groups such as AOH, CAO, C@O, and OAC@O into the surface layer of fibers, resulting in improved wettability. Argon, air, O2, CO2, H2O, and NH3 plasmas may be used for plasma treatment of fibers to improve interfacial adhesion, but the resulting shear strength is still lower compared to industrially sized fibers. The plasma coating method, unlike the plasma treatment, seems to be one of the most effective methods of achieving both high strength and high toughness, when an appropriate material is chosen for coating [10]. Plasma-polymerized organosilicones constitute a class of materials with a rich and varied scientific background [11,12]. This class of materials possesses special characteristics, distinguishing it from other plasma polymers: the ability to vary and control its organic/inorganic character (i.e., the carbon content) and polymer cross-linking by the appropriate choice of fabrication variables [13]. This allows one to control many physicochemical properties over wide ranges, resulting in an extraordinary potential for engineered interlayers in polymer composites [14,15]. In this study, we examined chemical, mechanical, and surface properties of plasma-polymerized tetravinylsilane films controlled by RF power and their application as engineered interlayers to enhance interfacial adhesion of surface modified glass fibers in GF/polyester composites. Fiber microindentation measurements were used to evaluate the interfacial shear strength [16].
V. Cech et al. / Composites: Part A 58 (2014) 84–89
2. Experimental Plasma-polymerized tetravinylsilane (pp-TVS) films were prepared by plasma-enhanced chemical vapor deposition (PECVD) employing an RF helical coupling system [17], using a pulsed regime. The films were deposited on IR-transparent silicon wafers (100) (0.8 10 10 mm3; ON Semiconductor, Czech Republic), special microscope slides (1.0 26 76 mm3; Knittel Glaser, Germany), and unsized glass fiber bundles (E-glass, 1200 tex, mean diameter 19 lm; Saint-Gobain Adfors CZ, Czech Republic). Tetravinylsilane, SiA(CH@CH2)4 (TVS, purity 97%, Sigma Aldrich), was used as the monomer. The effective power (Weff) of the pulsed plasma was controlled by changing the ratio of the time the plasma was switched on (ton) to the time it was switched off (toff), Weff = ton/T Wtotal, where the period was defined as T = ton + toff, Wtotal = 50 W, and the duty cycle, DC, as DC = ton/T 100%. The substrates were pretreated with O2 plasma (5 sccm, 4 Pa, 25 W) for 10 min to clean the surface from contaminants and improve film adhesion. Pulsed plasma was operated at conditions given in Table 1. The deposition rate increased from 10 nm min1 (0.1 W) to a sharp peak of 142 nm min1 (2.5 W), followed by a descent to 82 nm min1 (10 W). Film is deposited at steady-state plasma and thus has consistent thickness and uniformity. The substrate is hidden in a loadlock, separated from the deposition chamber, until steady-state plasma is reached. The low-temperature plasma was used as a gentle tool using a low power density (Table 1) and retaining mechanical properties of fibers. Selected chemical, mechanical, and surface properties were investigated for the plasma polymer films deposited on the planar substrates. The elemental composition of the thin films was studied by conventional and resonant Rutherford Backscattering Spectrometry (RBS) and Elastic Recoil Detection Analysis (ERDA) using a Van de Graaf generator with a linear electrostatic accelerator. The RBS spectra were evaluated by computer code GISA 3 [18] and the ERDA spectra by SIMNRA code [19], both using cross-section values from SigmaBase. Infrared measurements in the wavenumber range of 500– 4000 cm1 were made using a VERTEX 80 vacuum Fourier transform infrared (FTIR) spectrometer (Bruker Optics). Transmission spectra were obtained on films deposited on infrared-transparent silicon wafers. An absorption subtraction technique was used to remove the spectral features of silicon wafer, and background correction was carried out before each measurement. The spectral resolution was 4 cm1. Approximately 256 scans were recorded to achieve a reasonable signal-to-noise ratio. The mechanical properties (Young’s modulus, hardness) and adhesion (scratch test) of the pp-TVS films were investigated using 2D TriboScope (Hysitron) attached to an NTegra Prima Scanning Probe Microscope (NT-MDT). A Berkovich tip with a radius of curvature of about 100 nm was used. The Young’s modulus and hardness of films were determined from unload–displacement curves using the Oliver–Pharr method [20]. The scratch parameters were as follows: linear, 10 lm at 0.33 lm/s, load 1 lN–5 mN. The normal and lateral forces were measured simultaneously together
Table 1 Deposition conditions used for preparation of the pp-TVS films. Frequency ton, toff Period Duty cycle Effective power Effective power density Basic pressure Process gas pressure Monomer vapor flow rate
13.56 MHz 1 ms, 4–499 ms 5–500 ms 0.2–20% 0.1–10 W 2 103–1 101 W cm3 8 104 Pa 1.3 Pa 0.80 sccm
85
with the normal displacement. The root-mean-square (RMS) roughness of the plasma polymer films deposited on planar substrates was computed using AFM images (scan area 5 5 lm2) measured in semicontact mode. The glass fibers coated by ppTVS film were examined using a scanning electron microscope (SEM) (Philips XL 30/EDAX/Microspec). The sessile drop method (tangent method) employing an OCA 10 goniometer (DataPhysics) was used to measure the equilibrium contact angles. Water and diiodomethane were used as probe liquids. The surface free energy as well as the dispersion and polar components were evaluated using the Owens–Wendt–Kaelble geometric mean method [21–23]. Industrially sized GF bundles (E-glass, 1200 tex, mean diameter 19 lm; Saint-Gobain Adfors CZ, Czech Republic) were designated for GF/polyester composites. The commercial sizing based on silane coupling agents was tailored by the glass fiber manufacturer. Unsized, oxygen plasma treated, industrially sized, and plasma polymer coated glass fibers were embedded into unsaturated polyester resin (isophthalic) Viapal HP 349 F (Sirca S. p. A., Italy) to form a GF/polyester composite. A bundle of fibers was impregnated with the resin, and extra resin was carefully wiped from the bundle. The impregnated bundle was positioned axially in a silicon rubber mold, which was filled with resin and cured at 140 °C to form a polymer disk 14 mm in diameter and 5 mm in height. Details on sample fabrication can be found in Ref. [24]. The disk was embedded in a metallographic specimen mount with the fibers normal to the specimen surface and the surface was polished using conventional metallographic techniques. The microindentation test [16] was carried out on the individually selected glass fibers on a polished cross-section of GF/polyester composite using an Interfacial Testing System (ITS) (Dow Chemical Company) [25]. A diamond tip with a diameter of 12 lm was applied to push single fibers from their surrounding matrix. The interfacial shear strength, IFSS (MPa), was determined from the debond load, P (g), using a generalized empirical equation [25] (adapted for IFSS in (MPa))
" 1 # Gm 2 d E ; IFSS ¼ A 2 B C log D Ef D P
ð1Þ
where D is the fiber diameter (lm), Gm and Ef are the matrix shear modulus and fiber axial modulus, respectively, d is the matrix thickness between the tested fiber and its nearest neighbor (lm), and A = 1.249 104, B = 0.8757, C = 0.01863, E = 0.02650. A matrix shear modulus of 1.3 GPa and a fiber axial modulus of 73 GPa correspond to the polyester resin used and the glass fiber, respectively. 3. Results and discussion Tetravinylsilane molecules are activated and fragmented during the plasma process, forming free radicals due to collisions with high-energy electrons, and the highly reactive radicals recombine at the surface of growing film. An increasing number (0–4) of vinyl groups bonded to the silicon atom are eliminated from TVS molecules at enhanced effective powers, as was monitored by mass spectroscopy. If the effective power is further increased, the eliminated vinyl is fragmented into smaller carbon species in a form of mono- and bi-radicals. Therefore, by changing the effective power one is able to control not only the plasma species but also the elemental composition and chemical structure of the deposited film. The elemental composition of pp-TVS films was determined from RBS and ERDA spectra. Atomic concentrations were 9–5 at.% (silicon), 36–42 at.% (carbon), and 55–53 at.% (hydrogen), depending on the power used. The carbon to silicon ratio, which characterizes the organic/inorganic character of plasma polymer, increased from about 4 (0.1 W) to 8 (10 W) with enhanced power, as shown in
86
V. Cech et al. / Composites: Part A 58 (2014) 84–89
Fig. 1a. Typical infrared spectra of pp-TVS films deposited at different powers are given in Fig. 1b. The plasma polymer is formed mostly of a carbon network with incorporated SiAC bonding species and side vinyl groups. Absorption bands corresponding to vinyl vibrations are indicated by arrows: 3048 cm1 (CH stretching), 1589 cm1 (C@C stretching), 1404 cm1 (CH2 deformation), 1010 cm1 (CH wagging), and 950 cm1 (CH2 wagging), assigned following Ref. [26]. The concentration of vinyl groups decreased with enhanced power, as clearly shown by a descent of bands at 1404 cm1 and elsewhere; this trend corresponds to a reduction in vinyl groups, monitored by mass spectroscopy. To evaluate film adhesion by the scratch test, the pp-TVS film with a thickness of 0.1 lm was deposited on planar glass substrate at different powers. The test consists of drawing a tip over a film under increasing normal loads. The value of the load at which adhesion failure is detected is known as the critical load. The failure events were examined by atomic force microscopy (AFM). The critical load increased significantly with enhanced power and was almost three times higher for the film deposited at 10 W than that deposited at 0.1 W (Fig. 2a). We expect that chemical bonding at the film/glass interface is realized through SiAOAC and SiAOASi bonding species due to recombination of free radicals at SiAOH groups on glass surface. An increased density of interfacial bonds with enhanced power is responsible not only for the growth of adhesion but also for perfect durability of plasma polymer films on glass substrate in aqueous environments as it was demonstrated by Prikryl et al. [27]. A higher density of interfacial bonds results in a lower diffusion rate of water molecules in contrast to
Fig. 2. Mechanical properties of pp-TVS films controlled by effective power: (a) critical load characterizing adhesion of film on glass substrate; (b) Young’s modulus and hardness.
Fig. 1. Chemical properties of pp-TVS films deposited at different effective powers: (a) carbon to silicon ratio evaluated from RBS spectra; (b) FTIR spectra. The arrows indicate the positions of absorption bands corresponding to vinyl vibrations.
a relatively low concentration of bonds at industrial sizing/glass interfaces that are hydrolytically unstable [28]. Nanoindentation measurements enabled us to characterize selected mechanical properties (Fig. 2b) of 1-lm-thick films deposited on silicon wafers; the thick film was used to eliminate influence of the stiff substrate on the determined mechanical parameters. The Young’s modulus (full symbol) increased from 9.4 to 23 GPa when power was raised by two orders of magnitude (0.1–10 W). A similar trend was observed for hardness, with the values increasing from 0.9 to 3.9 GPa with enhanced power. Both the mechanical parameters increased with enhanced power due to a higher crosslinking of the plasma polymer network [29]. If the plasma energy (power) increases, monomer molecules are more activated and fragmented, forming a higher density of free radicals; the reactive species result in a highly crosslinked polymer. A very low power (power density) was used to deposit tough but flexible films of elastic deformations up to 10% [29] suitable for composite interlayers. The surface topography of pp-TVS films with a thickness of 1 lm was determined by AFM and the RMS roughness was counted from a scan area of 5 5 lm2. The RMS roughness increased with enhanced power from 2.0 nm (0.1 W) to 5.8 nm (10 W) (Fig. 3a). The roughness was influenced by kinetics of film growth at enhanced plasma energy (power). The results of the Owens– Wendt–Kaelble geometric mean method are given in Fig. 3b, where the total surface free energy and its polar and dispersion components are plotted as a function of the effective power. The total surface free energy (filled symbol) increased from 40 mJ m2 to a saturated value of 49 mJ m2, reached at a power of 5 W. The
V. Cech et al. / Composites: Part A 58 (2014) 84–89
87
Fig. 4. Scanning electron micrograph of plasma polymer coated glass fibers: (a) multi-fiber view of uniformly covered fibers; (b) detailed view of film damage at the fiber surface.
Fig. 3. Surface properties of pp-TVS films as a function of effective power: (a) RMS roughness characterizing surface topography of deposited films; (b) surface free energy and its components analyzed by the Owens–Wendt–Kaelble method.
values of the dispersion component (empty symbol) were responsible for the increase of the surface free energy due to the decreased concentration of vinyl groups in plasma polymer with enhanced power. The polar component (half-filled symbol) with a value of about 4 mJ m2 was approximately independent of power. The pp-TVS films were deposited on bundles of unsized glass fibers under the same deposition conditions as the films deposited on planar substrates (only the substrate holder was changed). The free radicals diffuse into the central part of the bundle, forming a thin film even on the surface of central fibers during the plasma polymer deposition. However, the deposition rate decreases radially into the fiber bundle due to a shadowing effect of the surrounding fibers; thus, the film coating on the central fibers is thinner than on the fibers at the bundle edge. The glass fibers are coated by plasma polymer in a form of homogeneous and continuous film with a thickness of 1 lm, corresponding to fibers at the bundle edge (Fig. 4a), unlike industrially sized fibers with heterogeneous coatings [5] using wet chemical processing. An intentionally broken film is shown in Fig. 4b to demonstrate the presence of film on the fiber. Two sets of coated glass fiber bundles were prepared; one set was deposited at the same power of 2.5 W but with different film thicknesses (0.05, 0.1, 0.5, 1, 5, and 10 lm), and another set was deposited at the same film thickness of 1 lm but at different effective powers (0.1, 1.0, 2.5, and 5.0 W). The specified film thickness
refers to the coating on fibers at the edge of the bundle. The unsized, industrially sized, oxygen plasma treated, and plasma polymer coated fibers were embedded into polyester resin and cured to form GF/polyester composites. A polished cross-section of composite samples was subjected to microindentation measurements to evaluate the level of adhesion between the fiber and the matrix. Fibers surrounded by plasma polymer film (white area) forming an interlayer are embedded in polyester resin, as can be seen in the polished cross-section observed by optical microscopy (Fig. 5a). The thickness of the interlayer is not constant due to the shadowing effect of neighboring fibers. The tested fiber is indented under increased normal loading until adhesion failure. The fiber debonding is visible as a dark shadow around the fiber. When the shadow appears around the fiber within an angle of 90–120°, it is said to be debonded [25]. The locus of failure can be observed using a thicker interlayer, as indicated by arrows in Fig. 5b. The contrast of the image was increased intentionally to demonstrate the locus of failure. The black spots correspond to alumina nanoparticles used for cross-section polishing. The residual impression can be seen in the tested fiber. The interfacial shear strength as a function of the interlayer thickness for interlayers deposited at an effective power of 2.5 W is given in Fig. 6a. Ten fibers from different parts of the bundle were selected for testing for each surface modification. No trend was observed for the IFSS determined for central and edge fibers. The IFSS was as low as 58 MPa when the interlayer thickness was only 50 nm, likely due to fibers being not properly coated. Enhancing the interlayer thickness, the IFSS increased sharply to 102 MPa and then grew to a maximum of 136 MPa, corresponding to a thickness of 5 lm, and subsequently decreased slightly.
88
V. Cech et al. / Composites: Part A 58 (2014) 84–89
Fig. 5. Polished cross-section of GF/polyester composite with fibers coated with ppTVS film, observed by optical microscopy: (a) the cross marks the selected fiber before the microindentation test; (b) fiber debonding after the test is visible as a dark shadow around the part of the fiber marked by arrows. The residual impression can be seen in the tested fiber.
According to the nonlinear finite element analysis (NLFEA) for the microindentation test as reported by Ho and Drzal [30], the IFSS should be almost independent of the interlayer thickness. However, we suspect an influence of the fiber volume fraction, which ranged from 0.32 to 0.62 for the composite samples, as can be seen in Fig. 6a. In such case, the maximum interfacial shear stress increases significantly with enhanced fiber volume fraction according to Ref. [30]. This means that if the IFSS is invariant for fibers in composites of different fiber volume fractions, the fiber in composites with a higher fiber volume fraction will fail under a lower applied load. A slight decrease of the IFSS at an interlayer thickness of 10 lm should result from increased interfacial shear stress, due to the dominant influence of the stiffer interlayer as the volume fraction of the interphase prevails against the volume fraction of the matrix. The locus of failure was identified predominantly at the interlayer/fiber interface for a thicker interlayer, as illustrated in Fig. 5b. The IFSS was 24.3 MPa (vf = 0.52), 47.4 MPa (vf = 0.62), and 103 MPa (vf = 0.48) for unsized, oxygen plasma treated, and industrially sized glass fibers, respectively. This means that the pp-TVS film of 100 nm thickness resulted in at least a similar interfacial adhesion to the industrial sizing as the fiber volume fraction was higher (0.62). Fig. 6b shows the interfacial shear strength dependent on the effective power used for deposition of an interlayer with a thickness of 1 lm. The IFSS varies from 112 to 130 MPa at a similar
Fig. 6. Interfacial shear strength (IFSS) and fiber volume fraction for plasma polymer coated glass fibers embedded in GF/polyester composite: (a) influence of the interlayer thickness at constant deposition conditions (2.5 W); (b) influence of the effective power at constant interlayer thickness (1.0 lm). The IFSS corresponding to unsized, oxygen plasma treated, and industrially sized fibers is marked by arrows.
volume fraction of fibers (0.52–0.56), with the minimum at 2.5 W. A higher power corresponding to an interlayer of higher Young’s modulus (Fig. 2b) is favorable for increased adhesion (Fig. 2a) between the interlayer and the glass fiber, enhanced RMS roughness (Fig. 3a), and improved wettability (Fig. 3b), though at the expense of a lower concentration of the vinyl groups (Fig. 1b) at the matrix/interlayer interface. Therefore, a slight IFSS increase at 0.1 W could be induced by improved adhesion between the matrix and the interlayer due to a higher concentration of vinyl groups, which are responsible for chemical bonding to the polyester resin. In contrast, a slight IFSS increase at 5 W could be induced by improved adhesion between the interlayer and the glass fiber. The interfacial shear strength corresponding to the interlayer deposited at 5 W was 26% higher than that for industrial sizing. An IFSS value of 103 MPa (vf = 0.48) for industrially sized glass fibers is relatively high with respect to an expected shear yield strength of bulk matrix (40 MPa). According to the model simulation (NLFEA), the maximum interfacial shear stress is located at about three quarters of a fiber diameter below the free surface [30] and the shear stress decreases steeply across the interphase from the matrix/fiber interface (103 MPa) down to a value below 40 MPa within a distance of 1.2 lm from the interface. A modified matrix of higher stiffness (shear strength) than that of the bulk matrix was identified between the bulk matrix and the fiber in the case of industrially sized glass fibers and the interphase thickness was determined to be 2 lm for GF/polyester composite by Hodzic et al. [31]. Thus, shear failure in the interphase could be caused by interfacial shear failure between the fiber and the modified matrix or the shear failure of the modified matrix itself but not by the
V. Cech et al. / Composites: Part A 58 (2014) 84–89
shear failure of the bulk matrix. The locus of failure was identified predominantly at the interlayer/fiber interface (Fig. 5b), where the shear yield strength of plasma polymer interlayer can be expected in range from 0.7 to 1.4 GPa based on comparison of simulated and experimental data from nanoindentation measurements [32]. Thus, IFSS values at 130–140 MPa for plasma polymer coated fibers may be realistic. The empirical Eq. (1) agrees well with the nonlinear finite element method and the best agreement was obtained at vf = 0.36. However, the equation results in overestimated values by about 20% for a fiber volume fraction of 0.6 [30]. According to the previous studies, the microindentation technique yields higher interfacial shear strength by 50% than the single fiber fragmentation technique [16] and the short-beam shear test [14]. Single-fiber techniques are used to obtain quantitative information about the IFSS. However, it is difficult to determine exact IFSS values, as there are differences in the specimen loading and mechanics of failure for the tests and IFSS calculations are based on very simplified assumptions. The microindentation test enables measurements on a real composite, unlike artificial single-fiber composite, but still we are not able to include factors such as debonding criteria, interlayer properties, properties of modified matrix, residual stresses, effect of neighboring fibers, and possible damage of the polished specimen to determine more exact IFSS data. 4. Conclusion Plasma polymer films of tetravinylsilane were deposited on planar and fibrous substrates by plasma-enhanced chemical vapor deposition operated at different effective powers from 0.1 to 10 W. Using powers within two orders enabled us to control chemical, physical, and surface properties of the deposited film in wide ranges; thus, such film could be used effectively as the functional interlayer in GF/polyester composites with a controlled interphase. The film adhesion on glass substrate could be increased almost threefold, the Young’s modulus by 140%, the RMS roughness threefold, the surface free energy (wettability) by 21%, and the concentration of functional groups (vinyl) decreased significantly by increasing the effective power from 0.1 to 10 W. Oxygen plasma treatment of the glass fibers enabled a twofold increase in the interfacial shear strength compared to the unsized fibers, but was still too low compared to industrially sized fibers. However, tetravinylsilane plasma modified (coated) fibers resulted in enhanced interfacial adhesion compared to industrial sizing for a wide variety of physicochemical properties of the deposited films. The plasma polymer films of tailored physicochemical properties have potential applications in subsequent improvement of composite interphases. The plasmachemical technology is a perspective technique for construction of gradient interlayers required for novel conception of composites without interfaces [13]. Acknowledgements This work was supported in part by the Czech Science Foundation, Grant Nos. P106/11/0738 and P205/12/J058, the Czech Ministry of Education, Grant No. ME09061, and the Technology Agency of the Czech Republic, Grant No. TA01010796. The authors would like to thank Brian Rook (Composite Materials and Structures Center) for his kind assistance with sample polishing and ITS measurements, Dr. V. Perina (Nuclear Physics Institute, Academy of Sciences) for RBS/ERDA measurements and analyses, and Dr. M. Sirovy and Saint-Gobain Adfors CZ s.r.o. for providing glass fibers.
89
References [1] Kim JK, Mai YW. Engineered interfaces in fiber reinforced composites. Amsterdam: Elsevier; 1998. [2] Drzal LT, Rich MJ, Lloyd PF. Adhesion of graphite fibers to epoxy matrices: I. The role of fiber surface treatment. J Adhes 1983;16:1–30. [3] Broutman LJ, Agarwal BD. A theoretical study of the effect of an interface on the properties of composites. Polym Eng Sci 1974;14:581–8. [4] Labronici M, Ishida H. Toughening composites by fiber coating. A review. Compos Interfaces 1994;2:199–234. [5] Dwight DW. Glass fibre reinforcements. In: Kelly A, Zweben C, editors. Comprehensive composite materials. Amsterdam: Elsevier; 2000 [chapter 1.08]. [6] Nishioka GM. Interaction of organosilanes with glass fibers. J Non-Cryst Solids 1990;120:102–7. [7] Wang W, DiBenedetto AT. A modified silane treatment for superior hydrolytic stability of glass reinforced composites. J Adhes 1999;68:183–201. [8] Thomason JL. The interface region in glass–fiber-reinforced epoxy–resin composites. 3. Characterization of fiber surface-coatings and the interphase. Composites 1995;26:487–98. [9] Li R, Ye L, Mai YW. Application of plasma technologies in fibre-reinforced polymer composites: a review of recent developments. Composites: Part A 1997;28:73–86. [10] Kim JK, Mai YW. High-strength, high fracture-toughness fiber composites with interface control – a review. Compos Sci Technol 1991;41:333–78. [11] Wrobel AM, Wertheimer MR. Plasma-polymerized organosilicones and organometallics. In: d’Agostino R, editor. Plasma deposition, treatment, and etching of polymers. New York: Academic Press; 1990. p. 163–268. [12] Segui Y. Plasma deposition from organosilicon monomers. In: d’Agostino R, Favia P, Fracassi F, editors. Proc NATO ASI plasma processing of polymers. Acquafredda di Maratea: Kluwer Academic Publishers; 1997. p. 305–19. [13] Cech V. Plasma polymer films: from nanoscale synthesis to macroscale functionality. In: Zhang S, editor. Nanostructured thin films and coatings, vol. 1. New York: CRC Press; 2010. p. 481–527. [14] Cech V. Plasma-polymerized organosilicones as engineered interlayers in glass fiber/polyester composites. Compos Interfaces 2007;14:321–34. [15] Cech V. Plasma polymer film as model interlayer for polymer composites. IEEE Trans Plasma Sci 2006;34:1148–55. [16] Drzal LT, Herrera-Franco PJ, Ho H. Fiber–matrix interface tests. In: Kelly A, Zweben C, editors. Comprehensive composite materials. Amsterdam: Elsevier; 2000 [chapter 5.05]. [17] Cech V, Prikryl R, Balkova R, Vanek J, Grycova A. The influence of surface modifications of glass on glass fiber/polyester interphase properties. J Adhes Sci Technol 2003;17:1299–320. [18] Saarilahti J, Rauhala E. Interactive personal-computer data-analysis of ion backscattering spectra. Nucl Instrum Methods: Part B 1992;64:734–8. [19] Mayer M. SIMNRA User’s guide. Julich: Forschungscentrum; 1998. [20] Oliver WC, Pharr GM. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J Mater Res 1992;7:1564–83. [21] Owens DK, Wendt RC. Estimation of the surface free energy of polymers. J Appl Polym Sci 1969;13:1741–7. [22] Kaelble DH. Dispersion-polar surface tension properties of organic solids. J Adhes 1970;2:66–81. [23] Kaelble DH, Cirlin EH. Dispersion and polar contributions to surface tension of poly(methylene oxide) and Na-treated polytetrafluorethylene. J Polym Sci: Part A 1971;2:363–8. [24] Cech V, Janecek P, Lasota T, Bursa J. A fiber-bundle pull-out test for surfacemodified glass fibers in GF/polyester composite. Compos Interfaces 2011;18:309–22. [25] The Interfacial Testing System (ITS). The Dow Chemical Company, Freeport, Texas. [26] Lin-Vein D, Colthup NB, Fateley WG, Grasselli JG. The handbook of infrared and Raman characteristic frequencies of organic molecules. San Diego: Academic Press; 1991. p. 85. [27] Prikryl R, Cech V, Kripal L, Vanek J. Adhesion of pp-VTES films to glass substrates and their durability in aqueous environments. Int J Adhes Adhes 2005;25:121–5. [28] Plueddemann EP. Silane coupling agents. New York: Plenum Press; 1991. [29] Cech V, Vanek J, Goruppa AA, Jones FR. RF-power-controlled Young’s modulus of plasma-polymerized organosilicon films. J Mater Sci 2005;40:5099–102. [30] Ho H, Drzal LT. Evaluation of internal mechanical properties o fiber reinforced composites using the microindentation method. Composites Part A 1996;27:961–71. [31] Hodzic A, Stachurski ZH, Kim JK. Nano-indentation and nano-scratch of polymer/glass interface. Part I: Experimental and mechanical analysis. Polymer 2000;41:6895–905. [32] Cech V, Hosein HA, Lasota T, Drzal LT. Mechanical properties of plasmapolymerized tetravinylsilane films. Plasma Process Polym 2011;8:138–46.
Contents lists available at ScienceDirect
Composites: Part A journal homepage: www.elsevier.com/locate/compositesa
Enhanced interfacial adhesion of glass fibers by tetravinylsilane plasma modification V. Cech a,⇑, A. Knob a, H.-A. Hosein b, A. Babik a, P. Lepcio a, F. Ondreas a, L.T. Drzal b a b
Institute of Materials Chemistry, Faculty of Chemistry, Brno University of Technology, Purkynova 118, CZ-612 00 Brno, Czech Republic Composite Materials and Structures Center, Michigan State University, 2100 Engineering Building, East Lansing, MI 48824-1226, USA
a r t i c l e
i n f o
Article history: Received 22 March 2013 Received in revised form 21 November 2013 Accepted 8 December 2013 Available online 15 December 2013 Keywords: A. Glass fibers A. Polymer–matrix composites (PMCs) B. Interface/interphase E. Surface treatments
a b s t r a c t Plasma-polymerized tetravinylsilane was used to surface modify glass fibers to improve interfacial adhesion of a GF/polyester composite. Plasma polymer films of controllable thickness and physicochemical properties were deposited on unsized glass fibers by RF pulsed plasma using an effective power of 0.1–5 W. The interfacial adhesion of unsized, industrially sized, oxygen plasma treated, and plasma polymer coated fibers embedded in polyester resin was determined by microindentation. The plasma modification of the glass fibers enabled a considerable increase in the interfacial shear strength compared to unsized fibers. The interfacial shear strength for the optimized plasma coating was 26% higher than that for the industrial sizing. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction The development of high-performance fiber-reinforced plastics (FRP) is closely connected with progress in engineered interfaces [1] through controlled interphases [2]. Theoretical and experimental studies have shown that the composite interphase can markedly influence the performance of composites with respect to their strength and toughness [3,4]. Sizing, i.e., functional coating (interlayer), is therefore tailored [5] to improve the transfer of stress from the polymer matrix to the fiber reinforcement by enhancing fiber wettability, adhesion, compatibility, etc. The key parameters affecting the performance of composite materials are the sizing thickness and modulus, the interaction at interfaces, and the reinforcement and matrix materials [4]. However, industrially produced sizing of glass fibers (GF) is not of a consistent thickness and uniformity [5]. Using the wet chemical process, the molecules of silane coupling agents have a tendency to self-condense, forming siloxane oligomers rather than complete bonding with the glass surface [6,7]. Only 10–20% of the total sizing is bonded to the fiber surface, and this amount is directly related to the composite shear strength [8]. The plasmachemical process is another way to improve composite performance via a controlled interphase. Low temperature plasma may be used as a gentle but powerful tool for surface treatment (chemical and/or physical altering of the fiber surface) or coating (plasma polymerization) of fibers, retaining their ⇑ Corresponding author. Tel.: +420 541149304; fax: +420 541149361. E-mail address: [email protected] (V. Cech). 1359-835X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.compositesa.2013.12.003
mechanical properties if appropriate plasma conditions are used. Plasma surface modification of fibers and its application in FRP has been studied since the 1980s, see the review in Ref. [9]. The oxygen plasma may increase the surface roughness and introduce functional groups such as AOH, CAO, C@O, and OAC@O into the surface layer of fibers, resulting in improved wettability. Argon, air, O2, CO2, H2O, and NH3 plasmas may be used for plasma treatment of fibers to improve interfacial adhesion, but the resulting shear strength is still lower compared to industrially sized fibers. The plasma coating method, unlike the plasma treatment, seems to be one of the most effective methods of achieving both high strength and high toughness, when an appropriate material is chosen for coating [10]. Plasma-polymerized organosilicones constitute a class of materials with a rich and varied scientific background [11,12]. This class of materials possesses special characteristics, distinguishing it from other plasma polymers: the ability to vary and control its organic/inorganic character (i.e., the carbon content) and polymer cross-linking by the appropriate choice of fabrication variables [13]. This allows one to control many physicochemical properties over wide ranges, resulting in an extraordinary potential for engineered interlayers in polymer composites [14,15]. In this study, we examined chemical, mechanical, and surface properties of plasma-polymerized tetravinylsilane films controlled by RF power and their application as engineered interlayers to enhance interfacial adhesion of surface modified glass fibers in GF/polyester composites. Fiber microindentation measurements were used to evaluate the interfacial shear strength [16].
V. Cech et al. / Composites: Part A 58 (2014) 84–89
2. Experimental Plasma-polymerized tetravinylsilane (pp-TVS) films were prepared by plasma-enhanced chemical vapor deposition (PECVD) employing an RF helical coupling system [17], using a pulsed regime. The films were deposited on IR-transparent silicon wafers (100) (0.8 10 10 mm3; ON Semiconductor, Czech Republic), special microscope slides (1.0 26 76 mm3; Knittel Glaser, Germany), and unsized glass fiber bundles (E-glass, 1200 tex, mean diameter 19 lm; Saint-Gobain Adfors CZ, Czech Republic). Tetravinylsilane, SiA(CH@CH2)4 (TVS, purity 97%, Sigma Aldrich), was used as the monomer. The effective power (Weff) of the pulsed plasma was controlled by changing the ratio of the time the plasma was switched on (ton) to the time it was switched off (toff), Weff = ton/T Wtotal, where the period was defined as T = ton + toff, Wtotal = 50 W, and the duty cycle, DC, as DC = ton/T 100%. The substrates were pretreated with O2 plasma (5 sccm, 4 Pa, 25 W) for 10 min to clean the surface from contaminants and improve film adhesion. Pulsed plasma was operated at conditions given in Table 1. The deposition rate increased from 10 nm min1 (0.1 W) to a sharp peak of 142 nm min1 (2.5 W), followed by a descent to 82 nm min1 (10 W). Film is deposited at steady-state plasma and thus has consistent thickness and uniformity. The substrate is hidden in a loadlock, separated from the deposition chamber, until steady-state plasma is reached. The low-temperature plasma was used as a gentle tool using a low power density (Table 1) and retaining mechanical properties of fibers. Selected chemical, mechanical, and surface properties were investigated for the plasma polymer films deposited on the planar substrates. The elemental composition of the thin films was studied by conventional and resonant Rutherford Backscattering Spectrometry (RBS) and Elastic Recoil Detection Analysis (ERDA) using a Van de Graaf generator with a linear electrostatic accelerator. The RBS spectra were evaluated by computer code GISA 3 [18] and the ERDA spectra by SIMNRA code [19], both using cross-section values from SigmaBase. Infrared measurements in the wavenumber range of 500– 4000 cm1 were made using a VERTEX 80 vacuum Fourier transform infrared (FTIR) spectrometer (Bruker Optics). Transmission spectra were obtained on films deposited on infrared-transparent silicon wafers. An absorption subtraction technique was used to remove the spectral features of silicon wafer, and background correction was carried out before each measurement. The spectral resolution was 4 cm1. Approximately 256 scans were recorded to achieve a reasonable signal-to-noise ratio. The mechanical properties (Young’s modulus, hardness) and adhesion (scratch test) of the pp-TVS films were investigated using 2D TriboScope (Hysitron) attached to an NTegra Prima Scanning Probe Microscope (NT-MDT). A Berkovich tip with a radius of curvature of about 100 nm was used. The Young’s modulus and hardness of films were determined from unload–displacement curves using the Oliver–Pharr method [20]. The scratch parameters were as follows: linear, 10 lm at 0.33 lm/s, load 1 lN–5 mN. The normal and lateral forces were measured simultaneously together
Table 1 Deposition conditions used for preparation of the pp-TVS films. Frequency ton, toff Period Duty cycle Effective power Effective power density Basic pressure Process gas pressure Monomer vapor flow rate
13.56 MHz 1 ms, 4–499 ms 5–500 ms 0.2–20% 0.1–10 W 2 103–1 101 W cm3 8 104 Pa 1.3 Pa 0.80 sccm
85
with the normal displacement. The root-mean-square (RMS) roughness of the plasma polymer films deposited on planar substrates was computed using AFM images (scan area 5 5 lm2) measured in semicontact mode. The glass fibers coated by ppTVS film were examined using a scanning electron microscope (SEM) (Philips XL 30/EDAX/Microspec). The sessile drop method (tangent method) employing an OCA 10 goniometer (DataPhysics) was used to measure the equilibrium contact angles. Water and diiodomethane were used as probe liquids. The surface free energy as well as the dispersion and polar components were evaluated using the Owens–Wendt–Kaelble geometric mean method [21–23]. Industrially sized GF bundles (E-glass, 1200 tex, mean diameter 19 lm; Saint-Gobain Adfors CZ, Czech Republic) were designated for GF/polyester composites. The commercial sizing based on silane coupling agents was tailored by the glass fiber manufacturer. Unsized, oxygen plasma treated, industrially sized, and plasma polymer coated glass fibers were embedded into unsaturated polyester resin (isophthalic) Viapal HP 349 F (Sirca S. p. A., Italy) to form a GF/polyester composite. A bundle of fibers was impregnated with the resin, and extra resin was carefully wiped from the bundle. The impregnated bundle was positioned axially in a silicon rubber mold, which was filled with resin and cured at 140 °C to form a polymer disk 14 mm in diameter and 5 mm in height. Details on sample fabrication can be found in Ref. [24]. The disk was embedded in a metallographic specimen mount with the fibers normal to the specimen surface and the surface was polished using conventional metallographic techniques. The microindentation test [16] was carried out on the individually selected glass fibers on a polished cross-section of GF/polyester composite using an Interfacial Testing System (ITS) (Dow Chemical Company) [25]. A diamond tip with a diameter of 12 lm was applied to push single fibers from their surrounding matrix. The interfacial shear strength, IFSS (MPa), was determined from the debond load, P (g), using a generalized empirical equation [25] (adapted for IFSS in (MPa))
" 1 # Gm 2 d E ; IFSS ¼ A 2 B C log D Ef D P
ð1Þ
where D is the fiber diameter (lm), Gm and Ef are the matrix shear modulus and fiber axial modulus, respectively, d is the matrix thickness between the tested fiber and its nearest neighbor (lm), and A = 1.249 104, B = 0.8757, C = 0.01863, E = 0.02650. A matrix shear modulus of 1.3 GPa and a fiber axial modulus of 73 GPa correspond to the polyester resin used and the glass fiber, respectively. 3. Results and discussion Tetravinylsilane molecules are activated and fragmented during the plasma process, forming free radicals due to collisions with high-energy electrons, and the highly reactive radicals recombine at the surface of growing film. An increasing number (0–4) of vinyl groups bonded to the silicon atom are eliminated from TVS molecules at enhanced effective powers, as was monitored by mass spectroscopy. If the effective power is further increased, the eliminated vinyl is fragmented into smaller carbon species in a form of mono- and bi-radicals. Therefore, by changing the effective power one is able to control not only the plasma species but also the elemental composition and chemical structure of the deposited film. The elemental composition of pp-TVS films was determined from RBS and ERDA spectra. Atomic concentrations were 9–5 at.% (silicon), 36–42 at.% (carbon), and 55–53 at.% (hydrogen), depending on the power used. The carbon to silicon ratio, which characterizes the organic/inorganic character of plasma polymer, increased from about 4 (0.1 W) to 8 (10 W) with enhanced power, as shown in
86
V. Cech et al. / Composites: Part A 58 (2014) 84–89
Fig. 1a. Typical infrared spectra of pp-TVS films deposited at different powers are given in Fig. 1b. The plasma polymer is formed mostly of a carbon network with incorporated SiAC bonding species and side vinyl groups. Absorption bands corresponding to vinyl vibrations are indicated by arrows: 3048 cm1 (CH stretching), 1589 cm1 (C@C stretching), 1404 cm1 (CH2 deformation), 1010 cm1 (CH wagging), and 950 cm1 (CH2 wagging), assigned following Ref. [26]. The concentration of vinyl groups decreased with enhanced power, as clearly shown by a descent of bands at 1404 cm1 and elsewhere; this trend corresponds to a reduction in vinyl groups, monitored by mass spectroscopy. To evaluate film adhesion by the scratch test, the pp-TVS film with a thickness of 0.1 lm was deposited on planar glass substrate at different powers. The test consists of drawing a tip over a film under increasing normal loads. The value of the load at which adhesion failure is detected is known as the critical load. The failure events were examined by atomic force microscopy (AFM). The critical load increased significantly with enhanced power and was almost three times higher for the film deposited at 10 W than that deposited at 0.1 W (Fig. 2a). We expect that chemical bonding at the film/glass interface is realized through SiAOAC and SiAOASi bonding species due to recombination of free radicals at SiAOH groups on glass surface. An increased density of interfacial bonds with enhanced power is responsible not only for the growth of adhesion but also for perfect durability of plasma polymer films on glass substrate in aqueous environments as it was demonstrated by Prikryl et al. [27]. A higher density of interfacial bonds results in a lower diffusion rate of water molecules in contrast to
Fig. 2. Mechanical properties of pp-TVS films controlled by effective power: (a) critical load characterizing adhesion of film on glass substrate; (b) Young’s modulus and hardness.
Fig. 1. Chemical properties of pp-TVS films deposited at different effective powers: (a) carbon to silicon ratio evaluated from RBS spectra; (b) FTIR spectra. The arrows indicate the positions of absorption bands corresponding to vinyl vibrations.
a relatively low concentration of bonds at industrial sizing/glass interfaces that are hydrolytically unstable [28]. Nanoindentation measurements enabled us to characterize selected mechanical properties (Fig. 2b) of 1-lm-thick films deposited on silicon wafers; the thick film was used to eliminate influence of the stiff substrate on the determined mechanical parameters. The Young’s modulus (full symbol) increased from 9.4 to 23 GPa when power was raised by two orders of magnitude (0.1–10 W). A similar trend was observed for hardness, with the values increasing from 0.9 to 3.9 GPa with enhanced power. Both the mechanical parameters increased with enhanced power due to a higher crosslinking of the plasma polymer network [29]. If the plasma energy (power) increases, monomer molecules are more activated and fragmented, forming a higher density of free radicals; the reactive species result in a highly crosslinked polymer. A very low power (power density) was used to deposit tough but flexible films of elastic deformations up to 10% [29] suitable for composite interlayers. The surface topography of pp-TVS films with a thickness of 1 lm was determined by AFM and the RMS roughness was counted from a scan area of 5 5 lm2. The RMS roughness increased with enhanced power from 2.0 nm (0.1 W) to 5.8 nm (10 W) (Fig. 3a). The roughness was influenced by kinetics of film growth at enhanced plasma energy (power). The results of the Owens– Wendt–Kaelble geometric mean method are given in Fig. 3b, where the total surface free energy and its polar and dispersion components are plotted as a function of the effective power. The total surface free energy (filled symbol) increased from 40 mJ m2 to a saturated value of 49 mJ m2, reached at a power of 5 W. The
V. Cech et al. / Composites: Part A 58 (2014) 84–89
87
Fig. 4. Scanning electron micrograph of plasma polymer coated glass fibers: (a) multi-fiber view of uniformly covered fibers; (b) detailed view of film damage at the fiber surface.
Fig. 3. Surface properties of pp-TVS films as a function of effective power: (a) RMS roughness characterizing surface topography of deposited films; (b) surface free energy and its components analyzed by the Owens–Wendt–Kaelble method.
values of the dispersion component (empty symbol) were responsible for the increase of the surface free energy due to the decreased concentration of vinyl groups in plasma polymer with enhanced power. The polar component (half-filled symbol) with a value of about 4 mJ m2 was approximately independent of power. The pp-TVS films were deposited on bundles of unsized glass fibers under the same deposition conditions as the films deposited on planar substrates (only the substrate holder was changed). The free radicals diffuse into the central part of the bundle, forming a thin film even on the surface of central fibers during the plasma polymer deposition. However, the deposition rate decreases radially into the fiber bundle due to a shadowing effect of the surrounding fibers; thus, the film coating on the central fibers is thinner than on the fibers at the bundle edge. The glass fibers are coated by plasma polymer in a form of homogeneous and continuous film with a thickness of 1 lm, corresponding to fibers at the bundle edge (Fig. 4a), unlike industrially sized fibers with heterogeneous coatings [5] using wet chemical processing. An intentionally broken film is shown in Fig. 4b to demonstrate the presence of film on the fiber. Two sets of coated glass fiber bundles were prepared; one set was deposited at the same power of 2.5 W but with different film thicknesses (0.05, 0.1, 0.5, 1, 5, and 10 lm), and another set was deposited at the same film thickness of 1 lm but at different effective powers (0.1, 1.0, 2.5, and 5.0 W). The specified film thickness
refers to the coating on fibers at the edge of the bundle. The unsized, industrially sized, oxygen plasma treated, and plasma polymer coated fibers were embedded into polyester resin and cured to form GF/polyester composites. A polished cross-section of composite samples was subjected to microindentation measurements to evaluate the level of adhesion between the fiber and the matrix. Fibers surrounded by plasma polymer film (white area) forming an interlayer are embedded in polyester resin, as can be seen in the polished cross-section observed by optical microscopy (Fig. 5a). The thickness of the interlayer is not constant due to the shadowing effect of neighboring fibers. The tested fiber is indented under increased normal loading until adhesion failure. The fiber debonding is visible as a dark shadow around the fiber. When the shadow appears around the fiber within an angle of 90–120°, it is said to be debonded [25]. The locus of failure can be observed using a thicker interlayer, as indicated by arrows in Fig. 5b. The contrast of the image was increased intentionally to demonstrate the locus of failure. The black spots correspond to alumina nanoparticles used for cross-section polishing. The residual impression can be seen in the tested fiber. The interfacial shear strength as a function of the interlayer thickness for interlayers deposited at an effective power of 2.5 W is given in Fig. 6a. Ten fibers from different parts of the bundle were selected for testing for each surface modification. No trend was observed for the IFSS determined for central and edge fibers. The IFSS was as low as 58 MPa when the interlayer thickness was only 50 nm, likely due to fibers being not properly coated. Enhancing the interlayer thickness, the IFSS increased sharply to 102 MPa and then grew to a maximum of 136 MPa, corresponding to a thickness of 5 lm, and subsequently decreased slightly.
88
V. Cech et al. / Composites: Part A 58 (2014) 84–89
Fig. 5. Polished cross-section of GF/polyester composite with fibers coated with ppTVS film, observed by optical microscopy: (a) the cross marks the selected fiber before the microindentation test; (b) fiber debonding after the test is visible as a dark shadow around the part of the fiber marked by arrows. The residual impression can be seen in the tested fiber.
According to the nonlinear finite element analysis (NLFEA) for the microindentation test as reported by Ho and Drzal [30], the IFSS should be almost independent of the interlayer thickness. However, we suspect an influence of the fiber volume fraction, which ranged from 0.32 to 0.62 for the composite samples, as can be seen in Fig. 6a. In such case, the maximum interfacial shear stress increases significantly with enhanced fiber volume fraction according to Ref. [30]. This means that if the IFSS is invariant for fibers in composites of different fiber volume fractions, the fiber in composites with a higher fiber volume fraction will fail under a lower applied load. A slight decrease of the IFSS at an interlayer thickness of 10 lm should result from increased interfacial shear stress, due to the dominant influence of the stiffer interlayer as the volume fraction of the interphase prevails against the volume fraction of the matrix. The locus of failure was identified predominantly at the interlayer/fiber interface for a thicker interlayer, as illustrated in Fig. 5b. The IFSS was 24.3 MPa (vf = 0.52), 47.4 MPa (vf = 0.62), and 103 MPa (vf = 0.48) for unsized, oxygen plasma treated, and industrially sized glass fibers, respectively. This means that the pp-TVS film of 100 nm thickness resulted in at least a similar interfacial adhesion to the industrial sizing as the fiber volume fraction was higher (0.62). Fig. 6b shows the interfacial shear strength dependent on the effective power used for deposition of an interlayer with a thickness of 1 lm. The IFSS varies from 112 to 130 MPa at a similar
Fig. 6. Interfacial shear strength (IFSS) and fiber volume fraction for plasma polymer coated glass fibers embedded in GF/polyester composite: (a) influence of the interlayer thickness at constant deposition conditions (2.5 W); (b) influence of the effective power at constant interlayer thickness (1.0 lm). The IFSS corresponding to unsized, oxygen plasma treated, and industrially sized fibers is marked by arrows.
volume fraction of fibers (0.52–0.56), with the minimum at 2.5 W. A higher power corresponding to an interlayer of higher Young’s modulus (Fig. 2b) is favorable for increased adhesion (Fig. 2a) between the interlayer and the glass fiber, enhanced RMS roughness (Fig. 3a), and improved wettability (Fig. 3b), though at the expense of a lower concentration of the vinyl groups (Fig. 1b) at the matrix/interlayer interface. Therefore, a slight IFSS increase at 0.1 W could be induced by improved adhesion between the matrix and the interlayer due to a higher concentration of vinyl groups, which are responsible for chemical bonding to the polyester resin. In contrast, a slight IFSS increase at 5 W could be induced by improved adhesion between the interlayer and the glass fiber. The interfacial shear strength corresponding to the interlayer deposited at 5 W was 26% higher than that for industrial sizing. An IFSS value of 103 MPa (vf = 0.48) for industrially sized glass fibers is relatively high with respect to an expected shear yield strength of bulk matrix (40 MPa). According to the model simulation (NLFEA), the maximum interfacial shear stress is located at about three quarters of a fiber diameter below the free surface [30] and the shear stress decreases steeply across the interphase from the matrix/fiber interface (103 MPa) down to a value below 40 MPa within a distance of 1.2 lm from the interface. A modified matrix of higher stiffness (shear strength) than that of the bulk matrix was identified between the bulk matrix and the fiber in the case of industrially sized glass fibers and the interphase thickness was determined to be 2 lm for GF/polyester composite by Hodzic et al. [31]. Thus, shear failure in the interphase could be caused by interfacial shear failure between the fiber and the modified matrix or the shear failure of the modified matrix itself but not by the
V. Cech et al. / Composites: Part A 58 (2014) 84–89
shear failure of the bulk matrix. The locus of failure was identified predominantly at the interlayer/fiber interface (Fig. 5b), where the shear yield strength of plasma polymer interlayer can be expected in range from 0.7 to 1.4 GPa based on comparison of simulated and experimental data from nanoindentation measurements [32]. Thus, IFSS values at 130–140 MPa for plasma polymer coated fibers may be realistic. The empirical Eq. (1) agrees well with the nonlinear finite element method and the best agreement was obtained at vf = 0.36. However, the equation results in overestimated values by about 20% for a fiber volume fraction of 0.6 [30]. According to the previous studies, the microindentation technique yields higher interfacial shear strength by 50% than the single fiber fragmentation technique [16] and the short-beam shear test [14]. Single-fiber techniques are used to obtain quantitative information about the IFSS. However, it is difficult to determine exact IFSS values, as there are differences in the specimen loading and mechanics of failure for the tests and IFSS calculations are based on very simplified assumptions. The microindentation test enables measurements on a real composite, unlike artificial single-fiber composite, but still we are not able to include factors such as debonding criteria, interlayer properties, properties of modified matrix, residual stresses, effect of neighboring fibers, and possible damage of the polished specimen to determine more exact IFSS data. 4. Conclusion Plasma polymer films of tetravinylsilane were deposited on planar and fibrous substrates by plasma-enhanced chemical vapor deposition operated at different effective powers from 0.1 to 10 W. Using powers within two orders enabled us to control chemical, physical, and surface properties of the deposited film in wide ranges; thus, such film could be used effectively as the functional interlayer in GF/polyester composites with a controlled interphase. The film adhesion on glass substrate could be increased almost threefold, the Young’s modulus by 140%, the RMS roughness threefold, the surface free energy (wettability) by 21%, and the concentration of functional groups (vinyl) decreased significantly by increasing the effective power from 0.1 to 10 W. Oxygen plasma treatment of the glass fibers enabled a twofold increase in the interfacial shear strength compared to the unsized fibers, but was still too low compared to industrially sized fibers. However, tetravinylsilane plasma modified (coated) fibers resulted in enhanced interfacial adhesion compared to industrial sizing for a wide variety of physicochemical properties of the deposited films. The plasma polymer films of tailored physicochemical properties have potential applications in subsequent improvement of composite interphases. The plasmachemical technology is a perspective technique for construction of gradient interlayers required for novel conception of composites without interfaces [13]. Acknowledgements This work was supported in part by the Czech Science Foundation, Grant Nos. P106/11/0738 and P205/12/J058, the Czech Ministry of Education, Grant No. ME09061, and the Technology Agency of the Czech Republic, Grant No. TA01010796. The authors would like to thank Brian Rook (Composite Materials and Structures Center) for his kind assistance with sample polishing and ITS measurements, Dr. V. Perina (Nuclear Physics Institute, Academy of Sciences) for RBS/ERDA measurements and analyses, and Dr. M. Sirovy and Saint-Gobain Adfors CZ s.r.o. for providing glass fibers.
89
References [1] Kim JK, Mai YW. Engineered interfaces in fiber reinforced composites. Amsterdam: Elsevier; 1998. [2] Drzal LT, Rich MJ, Lloyd PF. Adhesion of graphite fibers to epoxy matrices: I. The role of fiber surface treatment. J Adhes 1983;16:1–30. [3] Broutman LJ, Agarwal BD. A theoretical study of the effect of an interface on the properties of composites. Polym Eng Sci 1974;14:581–8. [4] Labronici M, Ishida H. Toughening composites by fiber coating. A review. Compos Interfaces 1994;2:199–234. [5] Dwight DW. Glass fibre reinforcements. In: Kelly A, Zweben C, editors. Comprehensive composite materials. Amsterdam: Elsevier; 2000 [chapter 1.08]. [6] Nishioka GM. Interaction of organosilanes with glass fibers. J Non-Cryst Solids 1990;120:102–7. [7] Wang W, DiBenedetto AT. A modified silane treatment for superior hydrolytic stability of glass reinforced composites. J Adhes 1999;68:183–201. [8] Thomason JL. The interface region in glass–fiber-reinforced epoxy–resin composites. 3. Characterization of fiber surface-coatings and the interphase. Composites 1995;26:487–98. [9] Li R, Ye L, Mai YW. Application of plasma technologies in fibre-reinforced polymer composites: a review of recent developments. Composites: Part A 1997;28:73–86. [10] Kim JK, Mai YW. High-strength, high fracture-toughness fiber composites with interface control – a review. Compos Sci Technol 1991;41:333–78. [11] Wrobel AM, Wertheimer MR. Plasma-polymerized organosilicones and organometallics. In: d’Agostino R, editor. Plasma deposition, treatment, and etching of polymers. New York: Academic Press; 1990. p. 163–268. [12] Segui Y. Plasma deposition from organosilicon monomers. In: d’Agostino R, Favia P, Fracassi F, editors. Proc NATO ASI plasma processing of polymers. Acquafredda di Maratea: Kluwer Academic Publishers; 1997. p. 305–19. [13] Cech V. Plasma polymer films: from nanoscale synthesis to macroscale functionality. In: Zhang S, editor. Nanostructured thin films and coatings, vol. 1. New York: CRC Press; 2010. p. 481–527. [14] Cech V. Plasma-polymerized organosilicones as engineered interlayers in glass fiber/polyester composites. Compos Interfaces 2007;14:321–34. [15] Cech V. Plasma polymer film as model interlayer for polymer composites. IEEE Trans Plasma Sci 2006;34:1148–55. [16] Drzal LT, Herrera-Franco PJ, Ho H. Fiber–matrix interface tests. In: Kelly A, Zweben C, editors. Comprehensive composite materials. Amsterdam: Elsevier; 2000 [chapter 5.05]. [17] Cech V, Prikryl R, Balkova R, Vanek J, Grycova A. The influence of surface modifications of glass on glass fiber/polyester interphase properties. J Adhes Sci Technol 2003;17:1299–320. [18] Saarilahti J, Rauhala E. Interactive personal-computer data-analysis of ion backscattering spectra. Nucl Instrum Methods: Part B 1992;64:734–8. [19] Mayer M. SIMNRA User’s guide. Julich: Forschungscentrum; 1998. [20] Oliver WC, Pharr GM. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J Mater Res 1992;7:1564–83. [21] Owens DK, Wendt RC. Estimation of the surface free energy of polymers. J Appl Polym Sci 1969;13:1741–7. [22] Kaelble DH. Dispersion-polar surface tension properties of organic solids. J Adhes 1970;2:66–81. [23] Kaelble DH, Cirlin EH. Dispersion and polar contributions to surface tension of poly(methylene oxide) and Na-treated polytetrafluorethylene. J Polym Sci: Part A 1971;2:363–8. [24] Cech V, Janecek P, Lasota T, Bursa J. A fiber-bundle pull-out test for surfacemodified glass fibers in GF/polyester composite. Compos Interfaces 2011;18:309–22. [25] The Interfacial Testing System (ITS). The Dow Chemical Company, Freeport, Texas. [26] Lin-Vein D, Colthup NB, Fateley WG, Grasselli JG. The handbook of infrared and Raman characteristic frequencies of organic molecules. San Diego: Academic Press; 1991. p. 85. [27] Prikryl R, Cech V, Kripal L, Vanek J. Adhesion of pp-VTES films to glass substrates and their durability in aqueous environments. Int J Adhes Adhes 2005;25:121–5. [28] Plueddemann EP. Silane coupling agents. New York: Plenum Press; 1991. [29] Cech V, Vanek J, Goruppa AA, Jones FR. RF-power-controlled Young’s modulus of plasma-polymerized organosilicon films. J Mater Sci 2005;40:5099–102. [30] Ho H, Drzal LT. Evaluation of internal mechanical properties o fiber reinforced composites using the microindentation method. Composites Part A 1996;27:961–71. [31] Hodzic A, Stachurski ZH, Kim JK. Nano-indentation and nano-scratch of polymer/glass interface. Part I: Experimental and mechanical analysis. Polymer 2000;41:6895–905. [32] Cech V, Hosein HA, Lasota T, Drzal LT. Mechanical properties of plasmapolymerized tetravinylsilane films. Plasma Process Polym 2011;8:138–46.