Surface properties and fibre-matrix adhesion of man-made cellulose epoxy composites – Influence on impact properties

Abstract

Previously conducted studies showed that UD regenerated cellulose fibre-reinforced thermoset composites can obtain specific Charpy impact strength values in the range of glass fibre reinforced composites. Composites of two different viscose fibre types, each with and without an oily avivage, were investigated. Despite similar mechanical properties of the fibres the impact strength of the CR fibre composites was about twice as high as that of the standard fibre composites.

To reveal a possible explanation for this effect the fibre surface properties were investigated more closely. AFM measurements showed no differences in fibre surface topologies. However the physico–chemical properties of the fibre types differ. IGC measurements showed that the standard Cordenka fibre without avivage (“std wo a.”) possesses a slightly higher specific surface energy and base number (K_b) than the CR fibres without avivage (“CR wo a.”) resulting in a better adhesion to the highly polar epoxy. This is also shown by the pair specific interaction parameters (I_sp) and the work of adhesion. Both are clearly higher for the epoxy – “std wo a.” pair. Accordingly the measured fibre pull-out lengths of the CR fibres are one order of magnitude higher than of the std suggesting a weaker interfacial shear strength between the CR fibres and epoxy. Within the same fibre type the samples without avivage show longer pull-out lengths. As a weaker fibre-matrix adhesion causes stronger crack deflection and energy dissipation these results correspond well with the previously measured Charpy impact strengths.

Introduction

Due to the growing environmental awareness the research in bio-based fibres and polymers for composites has grown remarkably in recent times. Besides the use of natural fibres regenerated cellulose fibres are an interesting alternative. Cellulose is dissolved and formed into endless fibres [1]. During this process important advantages of natural fibres such as the bio-based character and the low density are maintained [2] while some disadvantages of natural fibre like variability in quality or the limited length of the fibres are overcome. Usually bast fibres are used as reinforcing materials in composites as they exhibit a high stiffness and strength. However they are brittle in nature resulting in a low impact strength [3]. Regenerated cellulose fibres combine remarkable stiffness values and high elongations at break and thus can be used to produce fibre-reinforced composites that possess interesting impact and energy absorption properties. A previously published study [4] showed that UD regenerated cellulose fibre-reinforced thermoset composites can obtain specific Charpy impact strength values in the range of glass fibre composites with the same fibre content by mass. Two different viscose fibre types with identical Young's modulus and slightly different tensile strengths were compared and significant differences in Charpy impact strengths were measured. The impact strength of samples with the fibre type developed especially for composite applications (CR) is twice as high as the values of samples with the common (standard) viscose fibres. This gap cannot be explained solely by the higher tensile strength of the CR fibres and the analysis of SEM images of composite fracture surfaces leads to the assumption that the adhesions of the fibres to the matrix differ.

It is well known that besides the properties of the constituents the interfacial shear strength between fibre and matrix plays an important role in the mechanical properties of composites. It provides the structural integrity of composites and determines the ability of the interphase to transfer load from the matrix to the embedded fibres. A higher interfacial shear strength usually leads to a higher tensile and flexural strength [5], [6], [7], [8]. However in the case of impact strength not only the modulus and strength but also the pull-out of fibres is an important property to control the fracture energy of a composite. During impact the important mechanisms of energy absorption are the debonding, the pull-out and the fracture of fibres. So in the case of impact strength a poor fibre-matrix adhesion can lead to an improvement of the properties. In a composite where the elongation at break of the fibres is greater than the one of the matrix a crack originates in the matrix and propagates in it until it reaches a fibre. With increasing load the crack extends around the fibre and along the fibre-matrix interface causing fibre debonding and crack extension. Eventually the fibre breaks at a random weak spot some distance away from the crack. During further composite failure the loose end of the fibre is pulled out of the matrix and energy is dissipated due to frictional forces. A high shear strength between fibre and matrix inhibits fibre crack deflection and thus reduces the fibre pull-out. Therefore a weaker fibre-matrix adhesion can cause higher impact strength values [9], [10]. However it has to be kept in mind that a minimum of fibre-matrix adhesion is required in order to allow the transfer of stresses from the matrix to the fibre and ensure the reinforcing effect of the fibres.

The fibre-matrix adhesion is divided into a physico–chemical and a frictional component. The later one is due to mechanical interlocking at the interface. The physico–chemical adhesion between fibre and matrix is based on molecular interactions, as e.g. covalent and hydrogen bonds, intermolecular forces or transcristallinity [7], [8], [11]. In the case of composites with a polymer matrix the physico–chemical contribution is important and it is governed by the surface properties of the fibre and the matrix [7]. Important characteristics are the surface energies, the acid-base interactions and the thermodynamic work of adhesion. To better understand and tailor the adhesion between fibres and matrix the physico–chemical properties of various fibres and polymers were investigated with regard to their contribution to the fibre-matrix adhesion. Several reviews focus on the surface properties of natural or cellulose fibres in combination with polymeric matrices [5], [7], [11], [12]. Due to the hydrophilic character of natural or cellulose fibres, which is given by the hydroxyl groups of the cellulose, their bond to commonly used non-polar, hydrophobic matrices is low. Therefore different surface modifications of cellulose fibres in order to reduce the hydrophilic character were the focus of several studies [5], [11], [12], [13], [14].

From the surface energetics of the fibre and the matrix the thermodynamic work of adhesion and the pair specific interaction parameter (I_sp) can be calculated. The surface energy of the fibres is directly related to the thermodynamic work of adhesion, which is directly correlated to practical adhesion. A possibility to enhance the fibre-matrix-adhesion is therefore to increase the surface energy of the fibre [11]. Also the acid-base interaction is an important factor as, if the fibre and the matrix would both be acidic or neutral only van der Waals forces would bond the fibre to the matrix [5]. An increase in acid-base interaction results in an higher interfacial shear strength [7]. Various authors found a correlation between the work of adhesion or the pair specific interaction parameter (I_sp) and the interfacial shear strength of composites [15], [16] or investigated the contribution of the I_sp to the tensile properties of composites [5]. Tze et al. [7], Schultz et al. [15] and Mukhopadhyay et al. [17] found a linear correlation between the I_sp and some mechanical properties of the composites like interfacial shear resistance (τ).

To examine the surface properties of the fibres and the matrix, in order to investigate the adhesion potential of different fibres to a matrix, inverse Gaschromatography (iGC) can be used. IGC has been used before to characterize the surfaces of various different fibres [11], [18]. Especially in the development of cellulose-polymer composites iGC has been used to analyse the interface in composites. The method of iGC is better suited for the study of cellulosic fibre surfaces than wetting or contact angle measurements, where the surface roughness, the heterogeneity of the probe or bulk penetration can cause a contact angle hysteresis [5].

IGC is a gas phase technique, first developed in the 1950s, to study surface and bulk properties of particulate and fibrous materials [19]. Apart from its high versatility and speed, the main benefit of iGC is its sensitivity at the surface of the sample. The iGC is the reverse of the analytical gas chromatography. The adsorbent under investigation is placed into a column while a known adsorptive is used in the gas phase. As in analytical gas chromatography, the retention time is obtained as the fundamental parameter measured. The retention time can be converted into a retention volume, which is directly related to several physico–chemical properties of the solid (absorbent). These properties can be thermodynamic parameters, such as surface energy or heat of sorption and kinetic parameters, such as the diffusion constant or the activation energy of diffusion. It is also possible to determine the uptake for both physisorption and chemisorption processes. In the first case, a sorption isotherm is obtained, which allows the computation of the surface area and heterogeneity profiles [20].

According to Riddle & Fowkes [21] the total surface energy of a material is often divided into two components: dispersive (London dispersion, van der Waals, Liftschitz interactions) and specific (acid-base, polar interactions).

$${\gamma }_s^T={\gamma }_s^{ab}+{\gamma }_s^d$$

The dispersive surface energy (γ_S^d) analysis is performed by measuring the net retention volume V_N (measured retention volume corrected with dead volume) for a series of alkane elutants. The dead-volume is determined by an unretained solute. The dispersive surface energy can be determined with the Dorris and Gray method [22], by plotting the RTln(V_N) versus the carbon number (of the alkanes) which produces a linear correlation. The dispersive component of the solid sample can be determined from the slope of the line

$$Slope=2{\left({\gamma }_{C{H}_2}{\gamma }_S^d\right)}^{1/2}\ast N_A{a}_{C{H}_2}$$

where γ_S^d is the dispersive component of the solid surface energy, $a_{C{H}_2}$ is the cross sectional area of a methylene group and N_A is Avogadro's number.

The specific contribution of the total surface energy is obtained via iGC SEA by first measuring the specific free energies of adsorption for different polar probe molecules (ΔG_SP). These values are determined by measuring the retention volume of polar probe molecules on the samples. In the polarisation approach [18], the ΔG_SP values are determined from a plot of RTln(V_N) versus the molar deformation polarisation of the probes (P_D).

$$P_D=\left\{M{W}^{\ast }(r^2-1)/D^{\ast }(r^2+2)\right\},$$

where MW is the molar mass of the probe, r is the reflective index of the probe and D is the probe liquid density. On the RTln(V_N) versus P_D plot the points representing a polar probe are located above the alkane straight line and the vertical distance between the polar data point and the straight line is equal to the specific component of the free energy of adsorption of the polar probe [23]. From the ΔG_SP values of two monopolar probes, the specific surface energy (${\gamma }_S^{\mathrm{ab}}$) can be calculated by the van Oss approach [24]. The specific contribution is subdivided into an acid γ⁺ and a base γ⁻ parameter of the surface tension of the mono-functional polar probes. In this approach, the Della Volpe scale is employed, with a pair of mono-functional acidic and basic probe molecules (dichloromethane (CH₂Cl₂) – γ⁺: 124.58 mJ/m² and ethyl ethanoate (ethyl acetate) (C₄H₈O₂) – γ⁻: 475.67 mJ/m²).

The approach of Gutmann represents the electron-accepting and electron-donating characteristics of the surface by the acid and base numbers (K_a and K_b) respectively. The K_a and K_b constants of a polymer (matrix of the composite) and fibre, may define the pair specific interaction parameters (I_sp) by the following expression [15], [18],

$$I_{sp}=K_a^f{K}_b^m+K_a^m+K_b^f$$

where f and m corresponds to the fibre and the matrix, respectively.

The surface energy is directly related with the thermodynamic work of cohesion and adhesion and it can be calculated with the following expressions [6],

$$W_{Coh}^{Total}=2\left[\left({\gamma }_s^d\right)+{({\gamma }_s^{-}·{\gamma }_s^{+})}^{1/2}+{({\gamma }_s^{+}·{\gamma }_s^{-})}^{1/2}\right],$$

$$W_{Adh}^{Total}=2\left[{({\gamma }_{s1}^d·{\gamma }_{s2}^d)}^{1/2}+{({\gamma }_{s1}^{-}·{\gamma }_{s2}^{+})}^{1/2}+{({\gamma }_{s1}^{+}·{\gamma }_{s2}^{-})}^{1/2}\right],$$

where ${\gamma }_s^d$ is the dispersive surface energy component of the solid material, ${\gamma }_s^{-}$ and ${\gamma }_s^{+}$ are the acid and base components of the specific surface energy of solid material, and the number 1 and 2 denote e.g. polymer and fibre, respectively.

Within this study the previously tested [4] viscose fibres are investigated more deeply in regard to their surface properties and the interfacial shear strength. The differences in adhesion to the epoxy matrix are quantified by measuring the fibre pull-out length with a separate experimental set up. Moreover the surface energy properties of the fibres are examined by inverse gas chromatography (iGC) to identify reasons for the differences in fibre-matrix adhesion.