Accepted Manuscript Synergistic effects of janus particles and triblock terpolymers on toughness of immiscible polymer blends Ronak Bahrami, Tina I. Löbling, Holger Schmalz, Axel H.E. Müller, Volker Altstädt PII:
S0032-3861(16)31135-1
DOI:
10.1016/j.polymer.2016.12.044
Reference:
JPOL 19280
To appear in:
Polymer
Received Date: 24 September 2016 Revised Date:
13 December 2016
Accepted Date: 15 December 2016
Please cite this article as: Bahrami R, Löbling TI, Schmalz H, Müller AHE, Altstädt V, Synergistic effects of janus particles and triblock terpolymers on toughness of immiscible polymer blends, Polymer (2017), doi: 10.1016/j.polymer.2016.12.044. 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
Synergistic Effects of Janus Particles and Triblock Terpolymers
RI PT
on Toughness of Immiscible Polymer Blends Ronak Bahrami,1,‡ Tina I. Löbling,2,3,‡ Holger Schmalz,2* Axel H. E. Müller, 4* Volker Altstädt1* 1
Department of Applied Physics, Aalto University School of Science, FI-02100 Espoo, Finland. 4
Institute of Organic Chemistry, Johannes Gutenberg University, 55099 Mainz, Germany
M AN U
3
Macromolecular Chemistry II, University of Bayreuth, 95440 Bayreuth, Germany,
SC
2
Faculty of Engineering Science and
KEYWORDS: Polymer blends, Janus particles, compatibilization, micromechanics, fracture toughness
TE D
ABSTRACT. By influencing both the interfacial adhesion and the morphology, compatibilizers determine the mechanical properties of polymer blends. Here, we study the mechanical
EP
properties, in particular the fatigue crack propagation (FCP) of immiscible blends of poly(2,6dimethyl-1,4-phenylene ether)/poly(styrene-co-acrylonitrile) (PPE/SAN), compatibilized with
AC C
Janus nanoparticles (JPs) and polystyrene-block-polybutadiene-block-poly(methyl methacrylate) (SBM) linear triblock terpolymers. Synergistic effects of a mixture of both compatibilizers improve the FCP behavior and reveal the important role of interface stiffness and flexibility on the mechanical properties of polymer blends. The triblock terpolymer and JPs allow at the same time an elastic and stiff linkage at the blend interface and induce multiple deformation mechanisms such as crack bridging and matrix fibrillation that can dissipate energy and contribute to an improved FCP behavior. The presented concept allows tailoring macro-
-1-
ACCEPTED MANUSCRIPT
mechanical properties of immiscible polymer blends by adjusting blend morphology and interfacial properties.
RI PT
Introduction Polymer blends are mixtures of two or more polymers without the formation of new chemical bonds to create a material with advantageous properties of both polymeric components. For
SC
example, poly(2,6-dimethyl-1,4-phenylene ether) (PPE) is often blended with poly(styrene-coacrylonitrile) (SAN) to combine its ductility with the good chemical resistance of SAN and at the
M AN U
same time to overcome its high viscosity and low processability.[1–9] However, most immiscible polymer blends (that are not specifically impact modified) suffer from inhomogeneous morphologies and deteriorated mechanical properties as the blend interface acts as stress concentration point that causes materials failure. Modifying the blend interface[10] is a
TE D
common approach to improve the mechanical properties[11] and ensure a good dispersion of the dispersed phase into the matrix phase.[12–14] Especially, controlling the blend morphology using nanoparticles that can selectively localize and self-assemble at the interface has recently
EP
gained interest.[15–18] A uniform blend morphology is desired to prevent inhomogeneous mechanical properties within one sample but likewise interfacial adhesion between the phases
AC C
contributes tremendeously to an improved macro-mechanical behaviour. Fatigue crack propagation (FCP) measurements have proven to be a strong and sensitive tool to study both the role of interfacial adhesion and blend morphology in blend systems. FCP tests are performed under cyclic stresses and simulate vibrations that materials are subjected in e.g. machine compounds or aircrafts. Materials often have a significantly different mechanical behavior under dynamic loadings and fatigue conditions as compared to static load and materials failure often occurs at much lower stress values compared to the tensile strength of the material.
-2-
ACCEPTED MANUSCRIPT
Understanding the molecular motion and energy dissipation processes in blend structures is of significant importance and could be directly correlated to their macroscopic properties such as ductility.[19–25] It was previously shown that in fine blend morphologies, where the dispersed
RI PT
phase forms droplets smaller than a certain size (< 1-2 µm), crack propagation appears to be uniform across the crack front.[26] However, it remains still a challenge to obtain morphologies with sufficiently small droplets to study in depth the deformation mechanisms during fatigue
SC
crack propagation.[27,28] The deformation speed (local fatigue crack propagation rate) is an important parameter influencing the blend interface and underlying deformation mechanisms.
M AN U
FCP tests in low- and high-speed regions for High Impact PS (HIPS) showed that polymer chains in thermoplastic materials disentangle mainly at low deformation speeds, whereas plastic deformations dominate at higher speeds.[29]
For the PPE/SAN system, the impact of an elastomeric middle block on the mechanical
TE D
properties of the blends has been thoroughly investigated using polystyrene-blockpolybutadiene-block-poly(methyl
methacrylate)
(SBM)
triblock
terpolymers
as
compatibilizers.[7–9,19–21] In this system, PS is fully miscible with PPE, PMMA is miscible
EP
with SAN at low acrylonitrile contents, while the PB block is selectively located at the blend interface.[30,31] The formation of a “raspberry” morphology (patchy spherical domains
AC C
embedded in a matrix)[19] improves the toughness of the co-continuous blends without sacrificing its modulus and FCP measurements showed an improved resistance against crack propagation.[2,8,9,29]
Synthetic access to polymeric Janus nanoparticles (JPs) on a large scale[22,24] has provided a new class of compatibilizers for PPE/SAN blends with high interfacial activity, as JPs unite the Pickering effect of nanoparticles and the amphiphilicity of block copolymers. JPs made from
-3-
ACCEPTED MANUSCRIPT
SBM by crosslinking the PB middle block[22] were proven to be more effective compatibilizers with regard to blend morphology than SBM in both solution state and during melt blending.[5,23,24] In fact, JPs dramatically decreased the average droplet size by a factor of six
RI PT
as compared to linear SBM.[5,6] However, the influence of JPs on the mechanical properties of the blends has not been addressed so far, even though the blend interface together with blend morphology (blend domain size)[9] is among the most important parameters controlling the
SC
mechanical properties.
In this study we correlate JP compatibilized blend morphologies to the micro-mechanical and
M AN U
finally macro-mechanical properties of the blends by studying the FCP behavior. In particular, the effect of JPs at the interface on micro-mechanics is thoroughly characterized with special emphasis on the fracture toughness. PPE/SAN blends (60/40 w/w) were compatibilized with 2, 5, and 10 wt% of JPs and their mechanical properties were analyzed and compared to neat and
TE D
SBM compatibilized blends. To promote synergistic effects, a blend containing both SBM and JP compatibilizers (5 wt% of each) was compounded to selectively tailor the interface and
AC C
EP
improve the mechanical properties by tuning the blend morphology.
-4-
ACCEPTED MANUSCRIPT
Experimental Section Materials. Poly(2,6-dimethyl-1,4-phenylene ether), PPE, was obtained as powder from Mitsubishi Engineering Plastics Europe (commercial grade PX100F), Düsseldorf, Germany with
RI PT
Mw = 12.9 kg/mol. The commercial poly(styrene-co-acrylonitrile), SAN, has an acrylonitrile content of 19 wt%, and was obtained as pellets from BASF AG, Ludwigshafen, Germany (SAN VLL 19100, Mw = 97.1 kg/mol). A 2:1 mixture of Irganox 1010 and Irgafos 168 (BASF,
SC
Germany), 0.1 wt% in total, were used as stabilizers during compounding. The SBM triblock terpolymer (S32B36M3297, subscripts denote the weight fraction of the respective block and
M AN U
superscript mass average molecular weight in kg/mol) and PS/PMMA JPs (synthesized from S40B20M40111) were prepared according to previously reported methods (details of the polymers including critical entanglement length of PS and PMMA are given in table S1).[5,7,22] Melt processing of the blends. Melt blending of the polymers was performed in a co-rotating
TE D
twin-screw extruder (Brabender DSE 20/40) based on the previously reported method.[5,9] The PPE/SAN weight ratio was kept constant at 60/40 to produce PPE droplets embedded in the continuous SAN matrix. The amounts of SBM and JP compatibilizers were varied (10 wt% for
EP
SBM; 2, 5, and 10 wt% for JPs) or a combination of both compatibilizers (5 wt% SBM and 5 wt% JPs) was used.
AC C
Morphological characterization. Ultrathin sections (ca. 50-80 nm) from the blended materials after extrusion were cut at room temperature using an ultra-microtome (Reichert-Jung Ultracut E) equipped with a diamond knife. The ultrathin sections were stained with OsO4 vapor for 2 h at ambient pressure to enhance the contrast between the phases (SAN appears bright, PPE dark and the PB segments of SBM and JPs appear black).[19] Bright field transmission electron microscopy at an acceleration voltage of 200 kV was carried out using a Zeiss 922 Omega
-5-
ACCEPTED MANUSCRIPT
EFTEM (Zeiss NTS GmbH, Oberkochen, Germany) equipped with a Gatan Ultrascan 1000 CCD-camera with GMS image processing (Gatan Digital Micrograph 3.9 for GMS 1.4). The number-averaged diameter of the PPE droplets and their size distribution were obtained by
RI PT
measuring at least 500 droplets in TEM micrographs using ImageJ software. First, the area of each PPE droplet was measured and then, assuming that all droplets have a spherical shape and
was back calculated from the corresponding area.
SC
the TEM cuts have gone through the middle of each droplet, the corresponding particle diameter
Fatigue crack propagation. The FCP behavior was determined on compact tension (CT)
M AN U
specimens9 with width and thickness of 33 and 2 mm, respectively (see Figure S1a for specimen geometry and dimensions). The tests were performed based on ISO 15850/ASTM E647 at 23 °C and a relative humidity of 50 %. The samples were loaded dynamically (frequency of 10 Hz) in tension-tension mode using a servo hydraulic testing machine (IST Hydro Pulse MHF) from
TE D
Schenck, Germany. The amplitude of the cyclic stress intensity factor (∆K = Kmax – Kmin) at the crack tip was increased as a function of crack length (Figure S1b). The minimum to maximum load ratio, R, was set at 0.1. Prior to the measurement, an initial natural sharp pre-crack was
EP
introduced into the machined V-notch of the specimen by a sharp razor blade. The compliance was continuously measured by the crack opening displacement method using a transducer
AC C
(632.13F-20, MTS, Germany) fixed to the front of the CT specimen with rubber bands. From this the crack length was calculated continuously by equations published by Saxena and Hudak (see Supporting Information).[32] The advanced test system is capable to detect fatigue crack propagation rates over 7 decades in the range between 102 down to 10-4 µm/cycle. Each measurement was repeated at least three times to minimize the experimental errors and a middle curve is generated to be shown here. A detailed description of the methodology can be found
-6-
ACCEPTED MANUSCRIPT
elsewhere.[32–35] Rheological characterization. Specimens for shear rheology and dynamic mechanical analysis (DMA) were compression molded according to the previously published protocol[9] to eliminate
RI PT
any effects of orientation of the PPE particles during the injection molding process.[2] Rheological properties were investigated by a stress controlled dynamic-mechanical rheometer (RDA III from Rheometrics Scientific) under nitrogen atmosphere applying a plate-plate
SC
geometry. For the measurements disc shaped samples with a diameter of 25 mm and a thickness of 1.5 mm were used. The complex viscosities of the blends were measured as a function of
M AN U
frequency within the range of 0.1-500 rad/s. The linear viscoelastic region of the deformation was determined prior to the test as explained previously.[9] Each measurement was repeated at least three times to minimize the experimental errors. Results and discussion
TE D
FCP tests are powerful experiments to probe the toughness of a material at different rates of fatigue crack propagation. Investigating the morphology of the fractured specimen surface after the FCP test allows determination of the deformation mechanisms at different crack propagation
EP
speeds that contribute to materials toughness. Commonly, this behavior is divided into three separated regimes (Scheme 1): In region I, also known as threshold region and represented by the
AC C
lower threshold value of the stress intensity factor for crack propagation, ∆Kth, the crack growth is initiated and the crack propagation rate is very slow.
-7-
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Scheme 1: Schematic illustration of the FCP curve at slow (region I), medium (region II) and fast (region III) crack propagation. ∆Kth and ∆Kcf represent the threshold values of crack initiation and critical failure respectively. n is the exponent in the Paris law (eq. 1), which describes the crack growth in the noncritical regime.
In the stable crack growth region (region II), the crack propagates faster than in the previous
described by the Paris law:
∆
TE D
region. This region is also known as the Paris regime, as it shows a power law behavior
(1)
EP
where a, is the crack length, N number of load cycles, ∆K the stress intensity factor, C and n are
AC C
constants that depend on material, stress ratio, temperature, etc. (for details, see Supporting Information) [35]. In region III there is instable crack growth and the crack propagates very fast, represented by the upper critical value for materials failure, ∆Kcf. A shift in the curve to higher ∆K values indicates improved resistance of the material against crack growth, as at the same stress intensity factor the crack propagates slower. As the crack propagates through each region, the speed of crack growth increases with the stress intensity factor ∆K. Fatigue crack propagation behavior of JP compatibilized blends.
-8-
ACCEPTED MANUSCRIPT
The fatigue crack growth behavior of the JP compatibilized PPE/SAN (60/40 w/w) blends in comparison to the neat blend is shown in Figure 1a. In region I and II the JP compatibilized blends show comparable values to that of the neat blend. At higher crack propagation rates
RI PT
(region III), the crack propagates faster (shift to lower ∆K = Kmax – Kmin values) in blends with increasing JP contents indicating deterioration of materials’ behavior after compatibilization. Even though JPs have proven to be highly efficient compatibilizers during melt blending with
SC
respect to blend morphology,[5,23,24] the mechanical properties of the blends (specifically toughness) at fast crack propagation rates diminish. This is also confirmed by analysis of the
M AN U
stress intensity factors, ∆K, of the JP compatibilized blends compared to the neat blend in the threshold (region I) and critical fracture regions (region III, Figure 1b). The ∆Kth and ∆Kcf values represent the first and last measured point of the curves, respectively. While in region I (slow crack propagation speed), ∆Kth is not influenced by addition of JPs irrespective of JP content, in
TE D
region III (high crack propagation speed) the ∆Kcf values decrease with increasing JP content
AC C
EP
implying deterioration of the FCP behavior in this region.
-9-
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 1. a) FCP curves of the investigated PPE/SAN (60/40 w/w) blend systems compatibilized with JPs
EP
(2-10 wt%) and without compatibilizer (neat blend 60-40). b) ∆Kth and ∆Kcf values of the blends as a function of compatibilizer content.
AC C
Fractured surfaces of the samples in the relevant regions I (threshold) and III (instable fast crack growth) after the FCP tests were analyzed via scanning electron microscopy (SEM) to investigate the deformation mechanisms. Before the experiment the neat PPE/SAN blend exhibits a random, partially co-continuous morphology with irregular, large droplets (Figure S2). Figure 2 shows the fractured surface of the neat blend in regions I and III. Large PPE particles (RPPE = 540±300 nm) dispersed in the SAN matrix and large cracks and macro deformations
-10-
ACCEPTED MANUSCRIPT
(indicated by orange arrows) propagating mainly in the SAN matrix are visible (Figure 2a-b, Figure S3). These large cracks originate from the PPE/SAN interface, where debonding is observed due to an insufficient linkage between the phases. It occurs partially in region I (Figure
RI PT
2a) and develops into fully debonded PPE particles in region III (Figure 2b) at increased crack propagation speeds. Additionally in region I, embryonal crazes on the surface of the PPE domains form due to inherent ductility of this polymer (Figure 2a). Several energy dissipating
TE D
M AN U
in the neat blend contribute to its plastification.
SC
deformation mechanisms (macro cracks, debonding and undeveloped embryonal crazing in PPE)
Figure 2. SEM micrographs of the fractured surface of the neat 60-40 PPE/SAN blend in a) region I
AC C
by orange arrows.
EP
(threshold) and b) region III (instable crack growth). Large cracks and macro deformations are indicated
In contrast to the neat blend with its characteristic large debonded PPE particles, the SEM overview images of the fractured surfaces of the JP compatibilized blend (10 wt% JPs) in region I and III show a fine-textured structure (Figure 3). No macro deformations or pronounced surface roughness after plasticization are visible (see Figure S5 for overview images) and the small PPE particles (RPPE = 155±85 nm; Figure S6) are completely embedded in the SAN matrix (Figure 3a,b).
-11-
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 3. SEM micrographs of the fractured surface of a JP compatibilized blend (10 wt% JPs) at a, c)
TE D
region I (threshold region), and b, d) region III (instable crack growth).
Ligaments are visible at the interface (Figure 3c,d), where JPs link the PPE and SAN phases to
EP
each other, indicating crazing in these regions, but no debonding occurs in both regions. The enhanced adhesion between the blend phases, manifested by the presence of ligaments, suggests
AC C
a high entanglement density between the PS and PMMA chains of the JPs hemispheres with the PPE and SAN phases, respectively. This behavior arises from the high interfacial activity of JPs (combination of the Pickering effect with the biphasic structure (amphiphilicity) of the JP corona) which requires high desorption energies to desorb the JPs from the PPE/SAN interface.[5] In region I (Figure 3c), the craze ligaments at the interface show a clear fibrillation, where JPs bind the PPE particles to the matrix and the strong JP interface linkage hinders interface tear up.
-12-
ACCEPTED MANUSCRIPT
As the crack propagation rate increases, the stiff nature of the cross-linked JPs at the interface hinders further transfer of the force between the phases and limits the deformation of either phase. As a result, a large amount of fibrils are visible between the PPE particles and the SAN
RI PT
matrix (Figure 3d). The craze fibrils are under stress to stretch further, but are hindered due to their low elasticity (partially cross-linked PB core) and craze fibrils begin to turn into crazes. At higher crack propagation speeds, the stiff interface appears more brittle under stress and the size
SC
of the crazing area around the PPE particles decreases. The stress concentration effect of the particles is still visible as they are sitting in craters due to the fact that the causing particles were
M AN U
just below the final crack plane. This is term is also known as „secondary cracking" [36] and is associated in this case with crazing/fibrillation. The strong interface linkage prevents debonding that generally consumes a lot of energy and therefore, compared to the neat blend, the FCP behavior and in turn the toughness of the JP blends decreases. On the other hand, the advantage
TE D
of the strong bond is that it makes measurements up to very high crack propagation rates of up to 10 µm/cycle possible. This is uncommon for thermoplastic materials due to their high ductility, resulting in chain scission (complete failure) at these propagation rates. The fact that the material
EP
withstands such fast crack propagations again underlines the high efficiency of the JPs at the interface. However, on a macro scale they cause deteriorated FCP behavior as the deformation
AC C
mechanism is limited to a small crazing area around the PPE domains. Fatigue crack propagation behavior of SBM compatibilized blends. Although the blend morphology is one important parameter that governs the mechanical properties, we have demonstrated that the interfacial properties also have a crucial impact. Even though the PPE droplet size in JP compatibilized blend is with RPPE = 155±85 nm very small and homogenous as compared to the random co-continuous morphology in the neat blend, the change
-13-
ACCEPTED MANUSCRIPT
in deformation mechanisms from neat to JP compatibilized blend leads to a decrease in toughness of the material. In contrast, we have previously shown that 20 wt% SBM as compatibilizer improves the fatigue crack growth behavior in co-continuous PPE/SAN blends.[2]
RI PT
The improved FCP behavior was attributed to the compatibilizer efficiency, i.e., accumulation of SBM at the interface of the blend phases. The reduced PPE droplet size with SBM triblock terpolymer mediating between the interfaces facilitates deformation mechanisms such as
SC
debonding and plastic deformations that dissipate lot of energy.
In a PPE/SAN blend with 10 wt% SBM as compatibilizer, the size of the dispersed PPE
M AN U
droplets was measured to be RPPE = 670±230 nm (Figure S8) which is significantly larger than for the JP compatibilized blend. Still, the FCP behavior (material´s toughness) improves in region I at low crack propagation speeds (Figure S9). Comparable to the neat blend, macro deformations and a rough surface structure due to macro cracks are observed in regions I and III,
TE D
but, instead of crazing, the elastic PB block at the interface of the SBM compatibilizer tends to tear apart or pull out from one phase upon application of stress (Figure S10 and S11).[37] This tearing dissipates considerable amount of energy and leads to an improved FCP behavior at low
EP
propagation speeds. In region III (critical fracture region) where the crack propagation speed increases, more PPE particles are already detached or have torn up SBM interfaces and the blend
AC C
behaves similar to a non-modified blend. This different behavior of SBM compared to JPs is a result of different mechanical properties of the blend interface that influence the macromechanical properties in the three regions of crack propagation speeds. The Tg of the PB segments that accumulate at the interface play a major role in determining the blend interface properties and can be measured using dynamic-mechanical analysis. Figure S12 shows that the partially crosslinked PB core in the JPs has a Tg of around 0 °C while the elastomeric PB block
-14-
ACCEPTED MANUSCRIPT
in SBM has a Tg of -80 °C. The elastic blend interface formed by SBM promotes tearing as energy dissipating mechanism. This improves the materials toughness in low and medium crack propagation speeds (region I and II; Figure S9) and is in contrast to the JPs compatibilized
RI PT
blends. On the other hand, JPs provide a fine blend morphology, introduce a stiff blend interface (higher modulus of JPs compared to SBM) and are very effective to withstand high propagation rates allowing measurements up to very high crack propagating speeds in region III. The two
SC
distinct different impacts of SBM and JPs as compatibilizers on the FCP behavior of PPE/SAN blends lead us to the hypothesis if a synergistic effect of both compatibilizer might be achieved
M AN U
by preparing blends with both compatibilizers. This concept would allow to specifically tailor the interfacial properties of the blends and connected to that the toughness of the material. Synergistic effect of combining JPs with SBM triblock terpolymer. The combination of SBM and JP (5 wt% each) produces a blend morphology with very small
TE D
PPE droplets (RPPE = 100 ± 50 nm, histogram available in S15) dispersed in the SAN matrix (Figure 4). These droplets are even smaller than in the blends compatibilized with either 10 wt%
AC C
EP
SBM (RPPE = 670 ± 230 nm) or 10 wt% JPs (RPPE = 155 ± 85 nm).
-15-
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
TE D
Figure 4. a) TEM overview of the 60/40 PPE/SAN blend with 5 wt% SBM + 5 wt% JPs. b) The compatibilizers are visible as black rim around each PPE droplet (PB stained with OsO4).
EP
This -on a first glimpse- counterintuitive result can be explained by means of the blend viscosity during extrusion. Compared to the other blends, the JP and SBM compatibilized blend
AC C
after compounding has the highest viscosity (twice as high than with JPs alone; Figure S13) and the high shear forces during extrusion allow formation of a finer blend morphology due to effective droplet break-up (and simultaneous coalescence prevention via presence of effective JPs). The PPE droplets are densely packed at the interfaces and come into close contact with each other without fusion (Figure 4b), suggesting a highly efficient compatibilization. The lack of SBM micelles or SAN inclusions in the PPE domains further corroborate that both the SBM triblock terpolymer and the JPs are located exclusively at the interface. The dark black line
-16-
ACCEPTED MANUSCRIPT
around each PPE particle thereby corresponds to the stained PB phase of the compatibilizers (JP and SBM). Although it is difficult to distinguish between the two types of compatibilizers, the chemical similarity and high intrinsic compatibility should support homogeneous mixing of both
RI PT
components.
Figure 5a summarizes the FCP curves of the 60-40 blend with both compatibilizers (5 wt% SBM + 5 wt% JPs), no compatibilizer (neat), 10 wt% JPs and 10 wt% SBM. The FCP behavior
SC
of the blend, compatibilized with both JPs and SBM, outperforms all other blends. Both the threshold region (region I) and the critical fracture region (region III), show significant
M AN U
improvements (43 % and 20 %, respectively) over the neat blend (Figure 5b) suggesting an synergistic effect of JPs and SBM. In addition, similar to the JP compatibilized blends, measurement up to very high crack propagation speeds of around 7 µm/cycle is possible, which
AC C
EP
TE D
confirms the presence of the strong JP mediated linkage at the interface.
-17-
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 5. a) FCP behavior of the blends with both compatibilizers (5 wt% SBM + 5 wt% JPs) compared to the other discussed blends, b) comparison of threshold (∆Kth) and critical fracture (∆Kcf) stress intensity
EP
values for the different blends.
AC C
Figure 6 shows SEM images of the fractured surface after the FCP test on the blend with both compatibilizers (5 wt% each) in region I and III. Both regions show macro deformations with large cracks and a rough surface structure indicating massive plastic deformations (overview images in Figure S14). The blend has an apparent continuity and shows a homogenous structure, where the PPE particles and the interface are not recognizable in the blend structure anymore. However, it is assumed that the initiation of large cracks takes place at the PPE/SAN interface, where SBM triblock terpolymer chains are located (Figure 6a,b), i. e., in analogy to the SBM
-18-
ACCEPTED MANUSCRIPT
compatibilized blend (Figure S11). A closer look at the fractured surface in Figure 6c shows premature broken vertical crazes all over the blend surface. In addition, crack bridging is visible for a very long crack of around 10 µm (Figure 6c). The fibrils bridging the crack are similar to
RI PT
the crazing ligaments found at the interface for the JP compatibilized blend (Figure 3). This type of deformation of the fibrils bridging such a large crack dissipates a lot of energy, contributing to
AC C
EP
TE D
M AN U
SC
the improved performance in crack growth and thus toughness in the threshold region (region I).
Figure 6. SEM micrographs of the fractured surface of the blend compatibilized with both SBM + JPs at a, c) region I (threshold) and b, d) region III (fast and instable crack growth). In region III, the crack propagates faster (Figure 6d) and undeveloped vertical crazes further grow into shear bands and promote massive plastic deformation all over the surface. The
-19-
ACCEPTED MANUSCRIPT
deformation mechanisms that contribute to the improved FCP behavior of this blend system include macro cracks, interface fibrillation, crack bridging, shear yielding, and embryonal crazes all over the surface (see Table S2 for overview of deformation mechanisms in different samples).
RI PT
These multiple mechanisms are only possible by combining the strong JP mediated linkage at the interface with the high elasticity of the PB middle block in the SBM triblock terpolymer, which can promote formation of macro cracks, in the fine, homogenous morphology of the blend. This
SC
synergistic effect enables the improvement of the materials crack propagation resistance and hence toughness, but likewise it represents a general concept to tailor interfacial properties of
M AN U
blend materials.
Conclusion
The FCP behavior of 60/40 (w/w) PPE/SAN blends compatibilized with JPs, a SBM triblock terpolymer and 50/50 combinations of both was investigated and compared to the blend without
TE D
compatibilizer. The neat blend shows an irregular partially co-continuous morphology having macro cracks and debonding as main energy dissipating deformation mechanisms due to
EP
insufficient interface adhesion. While JPs are highly effective compatibilizers to provide a homogeneous blend morphology with small droplet diameters of about 300 nm, these blends
AC C
behave similar to the neat blend in threshold region (region I) but decrease mechanical properties at high crack propagation speeds (region III) due to crazing and fibril formation around the interface area as the only energy dissipating mechanisms. A strong JP mediated linkage at the blend interface is promoted by the high interfacial activity of these particles and hinders debonding of the PPE particles from the SAN matrix and initiation of macro cracks which can dissipate a lot of energy. In contrast, using SBM as compatibilizer, the elastic interface provided by the soft PB middle block allows formation of macro cracks initiating at the interface due to
-20-
ACCEPTED MANUSCRIPT
tearing of the PB middle block leading to debonded PPE particles and increased SAN matrix plasticization. These blends showed improved behavior in regions I and II as tearing dissipates
ineffective at high propagating speeds (region III).
RI PT
considerable amounts of energy, however progressed tearing makes SBM as compatibilizer
A blend containing both compatibilizers showed synergistic effects and an improved FCP behavior in all three regions because elastic and stiff interfaces are provided. The fine,
SC
homogeneous morphology of the blend with droplet diameters of only 200 nm is achieved due to the high viscosity and, as a result, higher shear forces during the extrusion process. Multiple
M AN U
deformation mechanisms acting simultaneously during fatigue crack propagation are observed, such as macro crack development, crack bridging, embryonal crazing all over the surface, and at higher crack speeds additionally shear yielding. This is a direct result of tuning interfacial adhesion and interfacial stiffness combined with an extremely high interfacial area due to
TE D
homogeneous small droplet morphology. Tailoring blend morphologies with multiple compatibilizer systems leads to synergistic effects on the macro-mechanical properties and overall improved toughness of immiscible polymer blends. Understanding these basic
EP
relationships between the blend recipe and the used compatibilizer systems allows targeting desired materials properties and may act as a helpful toolbox to design materials for long term
AC C
usage such as in pump or engine casings.
-21-
ACCEPTED MANUSCRIPT
ASSOCIATED CONTENT Supporting Information. Additional information on the FCP test, additional SEM micrographs, experimental data on DMA characterization of SBM and JPs and rheological characterization of
AUTHOR INFORMATION Corresponding Authors Prof. Dr.-Ing. Volker Altstädt:
[email protected]
M AN U
Prof. Dr. Axel H. E. Müller:
[email protected]
SC
RI PT
the blends. This material is available free of charge via the Internet at http://pubs.acs.org.
Dr. Holger Schmalz:
[email protected] Author Contributions
‡ These authors contributed equally to this work. The manuscript was written through
manuscript.
EP
ACKNOWLEDGMENT
TE D
contributions of all authors. All authors have given approval to the final version of the
AC C
The German Research Foundation (DFG) has supported this work within the AL 474/21-1 and Mu 896/39-1 grants. This work was also supported by the Academy of Finland's Centre of Excellence Programme (2014-2019). Authors would like to thank Anneliese Lang and Melanie Müller for their support for SEM and TEM measurements. The authors acknowledge BASF for donating SAN and Mitsubishi for PPE polymer.
-22-
ACCEPTED MANUSCRIPT
REFERENCES [1]
Du B, Handge UA, Majeed S, Abetz V. Localization of functionalized MWCNT in SAN/PPE blends and their influence on rheological properties. Polymer 2012;53:5491–
[2]
RI PT
501. doi:10.1016/j.polymer.2012.09.047. Ruckdäschel H, Sandler JKW, Altstädt V, Schmalz H, Abetz V, Müller AHE. Toughening of immiscible PPE/SAN blends by triblock terpolymers. Polymer doi:10.1016/j.polymer.2007.03.005.
Everaert V, Groeninckx G, Aerts L. Fractionated crystallization in immiscible
SC
[3]
2007;48:2700–19.
POM/(PS/PPE) blends Part 1: Effect of blend phase morphology and physical state of the
3. [4]
M AN U
amorphous matrix phase. Polymer 2000;41:1409–28. doi:10.1016/S0032-3861(99)00321-
Gutmann P, Hildebrandt K, Altstädt V, Müller AHE. Foaming of an Immiscible Blend System Using Organic Liquids as Blowing Agents. J Cell Plast 2010;46:239–58. doi:10.1177/0021955X09352138.
[5]
Bahrami R, Löbling TI, Gröschel AH, Schmalz H, Müller AHE, Altstädt V. The Impact of Janus Nanoparticles on the Compatibilization of Immiscible Polymer Blends under
[6]
TE D
Technologically Relevant Conditions. ACS Nano 2014;8:10048–56. doi:10.1002/)). Bärwinkel S, Bahrami R, Löbling TI, Schmalz H, Müller AHE, Altstädt V. Polymer Foams Made of Immiscible Polymer Blends Compatibilized by Janus Particles-Effect of Compatibilization
on
Foam
Morphology.
Adv
Eng
Mater
2016;18:814–25.
[7]
EP
doi:10.1002/adem.201500387. Ruckdäschel H, Sandler JKW, Altstädt V, Rettig C, Schmalz H, Abetz V, et al.
AC C
Compatibilisation of PPE/SAN blends by triblock terpolymers: Correlation between block terpolymer composition, morphology and properties. Polymer
2006;47:2772–90.
doi:10.1016/j.polymer.2006.02.064.
[8]
Kirschnick T, Gottschalk A, Ott H, Abetz V, Puskas J, Altstädt V. Melt processed blends of poly(styrene-co-acrylonitrile) and poly(phenylene ether) compatibilized with polystyrene-b-polybutadiene-b-poly(methyl methacrylate) triblock terpolymers. Polymer 2004;45:5653–60. doi:10.1016/j.polymer.2004.03.106.
[9]
Bahrami R, Löbling TI, Schmalz H, Müller AHE, Altstädt V. Micromechanics of “raspberry” morphology in PPE/SAN polymer blends compatibilized with linear ABC
-23-
ACCEPTED MANUSCRIPT
triblock terpolymers. Polymer 2015;80:52–63. doi:10.1016/j.polymer.2015.10.039. [10] Cerclé C, Favis BD. Generalizing interfacial modification in polymer blends. Polymer 2012;53:4338–43. doi:10.1016/j.polymer.2012.07.027.
2000;41:8455–8. doi:10.1016/S0032-3861(00)00245-7.
RI PT
[11] Karim A. Modification of the phase stability of polymer blends by fillers. Polymer
[12] Lim J, Borg L zur, Dolezel S, Schmid F, Char K, Zentel R. Strategy for good dispersion of well-defined tetrapods in semiconducting polymer matrices. Macromol Rapid Commun 2014;35:1685–91. doi:10.1002/marc.201400314.
SC
[13] Yuan J-K, Yao S-H, Sylvestre A, Bai J. Biphasic Polymer Blends Containing Carbon Nanotubes: Heterogeneous Nanotube Distribution and Its Influence on the Dielectric
M AN U
Properties. J Phys Chem C 2012;116:2051–8. doi:10.1021/jp210872w. [14] Wu S. Formation of Dispersed Phase in Incompatible Polymer: Interfacial and Rheological Effects. Polym Eng Sci 1987;27:335–43. doi:10.1002/pen.760270506. [15] Balazs AC, Emrick T, Russell TP. Nanoparticle polymer composites: where two small worlds meet. Science 2006;314:1107–10. doi:10.1126/science.1130557. [16] Kao J, Thorkelsson K, Bai P, Rancatore BJ, Xu T. Toward functional nanocomposites:
TE D
taking the best of nanoparticles, polymers, and small molecules. Chem Soc Rev 2013;42:2654–78. doi:10.1039/c2cs35375j. [17] Koning C, van Duin M, Pagnoulle C, Jerome R. Strategies for compatibilization of polymer blends. Prog Polym Sci 1998;23:707–57. doi:10.1016/S0079-6700(97)00054-3.
EP
[18] Li G, Shrotriya V, Huang J, Yao Y, Moriarty T, Emery K, et al. High-efficiency solution processable polymer photovoltaic cells by self-organization of polymer blends. Nat Mater
AC C
2005;4:864–8. doi:10.1038/nmat1500. [19] Auschra C, Stadler R. Polymer Alloys Based on Poly( 2,6-dimethyl-l,4-phenylene ether) and Poly(styrene-co-acrylonitrile) using Poly(styrene-b-(ethylene-co-butylene)-b-methyl metacryrate) Triblock Copolymers as Compatibilizers. Macromolecules 1993;26:6364–77.
[20] Macosko CW, Gue P, Khandpur AK, Nakayama A, Marechal P, Inoue T. Compatibilizers for Melt Blending: Premade Block Copolymers. Macromolecules 1996;29:5590–8. [21] Zhuang Q, Xue Z, Yuan Y, Han Z. Molecular Simulation of Miscibility of Poly(2, 6dimethyl-1,
4-phenylene
Compatibilizer
Triblock
ether)/Poly(styrene-co-acrylonitrile) Terpolymer
-24-
SBM.
Polym
Compos
Blend
with
the
2011;32:1671–80.
ACCEPTED MANUSCRIPT
doi:10.1002/pc. [22] Gröschel AH, Walther A, Löbling TI, Schmelz J, Hanisch A, Schmalz H, et al. Facile, solution-based synthesis of soft, nanoscale Janus particles with tunable Janus balance. J Am Chem Soc 2012;134:13850–60. doi:10.1021/ja305903u.
RI PT
[23] Walther A, Matussek K, Müller AHE. Engineering nanostructured polymer blends with controlled nanoparticle location using Janus particles. ACS Nano 2008;2:1167–78. doi:10.1021/nn800108y.
[24] Walther A, Müller AHE. Janus Particles: Synthesis, Self-Assembly, Physical Properties,
SC
and Applications. Chem Rev 2013;113:5194–261. doi:10.1021/cr300089t.
[25] Hertzberg RW, Nordberg H, Manson JA. Fatigue crack propagation in polymeric
M AN U
materials. J Mater Sci 1907;5:521–6.
[26] Koster C, Altstädt V, Kausch HH, Cantwell WJ. Split rate fatigue propagation in polymer blends. Polym Bull 1995;34:243–8. doi:10.1007/BF00316402. [27] Morelli TA, Takemori MT. Fatigue crack advance mechanisms in polymers: rubber toughening mechanisms in blends of poly (2, 6-dimethy1-1, 4-phenylene oxide) and polystyrene. J Mater Sci 1983;18:1836–44.
1994;51:873–85.
TE D
[28] Wyzguski MG, Novak GE. Fatigue-Resistant Nylon Alloys. J Appl Polym Sci
[29] Ramsteiner F, Armbrust T. Fatigue crack growth in polymers. Polym Test 2001;20:321–7. doi:10.1016/S0142-9418(00)00039-8.
EP
[30] Ruckdäschel H, Fischer F, Altstädt V, Müller AHE. Fatigue Crack Growth Behavior of Multiphase Blends. J Solid Mech Mater Eng 2008;2:417–27. doi:10.1299/jmmp.2.417.
AC C
[31] Shariatpanahi H, Nazokdast H, Hemmati M. Relationship between interfacial tension and dispersed- phase particle size in polymer blends . II . PP/PA6. J Appl Polym Sci 2003;88:54–63.
[32] Saxena A, Hudak SJ. Review and extension of compliance information for common crack growth specimens. Int J Fract 1978;14:453–68. doi:10.1007/BF01390468.
[33] Fischer F, Beier U, Wolff-Fabris F, Altstädt V. Toughened high performance epoxy resin system for aerospace applications. Sci Eng Compos Mater 2011;18:209–15. doi:10.1515/SECM.2011.042. [34] Kothmann MH, Zeiler R, Anda AR De, Brückner A, Altstädt V. Fatigue crack
-25-
ACCEPTED MANUSCRIPT
propagation behaviour of epoxy resins modified with silica-nanoparticles. Polymer 2015;60:157–63. doi:10.1016/j.polymer.2015.01.036. [35] Paris P, Erdogan F. A Critical Analysis of Crack Propagation Laws. J Basic Eng 1963;85:528. doi:10.1115/1.3656900.
RI PT
[36] Karger-Kocsis J, Walter R, Friedrich K. Annealing effects on the fatigue crack propagation of injection moulded PEEK and its short-fibre composites. J Polym Eng 1988;8:221–55. doi:10.1515/POLYENG.1988.8.3-4.221.
AC C
EP
TE D
M AN U
and Morphology. Taylor & Francis Group; 2005.
SC
[37] Michler GH, Balta-Calleja FJ. Mechanical Properties of Polymers Based on Nanostructure
-26-
ACCEPTED MANUSCRIPT
For Table of Contents Use Only
RI PT
Synergistic Effects of Janus Particles and Triblock Terpolymers on Toughness of Immiscible Polymer Blends
AC C
EP
TE D
M AN U
SC
by Ronak Bahrami, Tina I. Löbling, Holger Schmalz,* Axel H. E. Müller* and Volker Altstädt*
-27-
ACCEPTED MANUSCRIPT
- Tailoring the blend on the nanoscale and generating nanostructures control the macro-mechanical properties of polymer blends - Synergistic effects of Janus nanoparticles and SBM triblock terpolymers as compatibilizers in PPE/SAN blends
RI PT
- Janus particles induce strong bonds and crazing in the raspberry morphology due to their stiff nature - SBM induces shear yielding in the raspberry morphology due the elastic nature of PB at the interface
AC C
EP
TE D
M AN U
SC
- Fatigue crack propagation as sensitive method to investigate toughness behavior of complex multi scale structures