polylactic acid blends

Accepted Manuscript Double percolation of multiwalled carbon nanotubes in polystyrene/polylactic acid blends Giuseppe Nasti, Gennaro Gentile, Pierfrancesco Cerruti, Cosimo Carfagna, Veronica Ambrogi PII:

S0032-3861(16)30544-4

DOI:

10.1016/j.polymer.2016.06.058

Reference:

JPOL 18795

To appear in:

Polymer

Received Date: 14 April 2016 Revised Date:

21 June 2016

Accepted Date: 23 June 2016

Please cite this article as: Nasti G, Gentile G, Cerruti P, Carfagna C, Ambrogi V, Double percolation of multiwalled carbon nanotubes in polystyrene/polylactic acid blends, Polymer (2016), doi: 10.1016/ j.polymer.2016.06.058. 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.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Double percolation of multiwalled carbon nanotubes

RI PT

in polystyrene/polylactic acid blends Giuseppe Nastia,b, Gennaro Gentileb,*, Pierfrancesco Cerrutib,

Department of Chemical, Materials and Production Engineering, University of Naples Federico II,

P.le Tecchio 80, 80125 Napoli, Italy. b

M AN U

a

SC

Cosimo Carfagnab, Veronica Ambrogia

Institute for Polymers, Composites and Biomaterials (IPCB-CNR), via Campi Flegrei 34, 80078 Poz-

TE D

zuoli (Na), Italy.

AC C

EP

* Corresponding author: [email protected] (Gennaro Gentile)

KEYWORDS: multiwalled carbon nanotubes, polystyrene, polylactic acid, bicontinuity, double percolation, electrical conductivity.

1

ACCEPTED MANUSCRIPT

Abstract Selective localization of multi-walled carbon nanotubes (MWCNT) in polylactic acid/polystyrene

RI PT

(PLA/PS) bicontinuous blends is presented in this work. A masterbatch of PS with a 2 wt% content of MWCNT is prepared and then melt mixed with different amounts of PLA. This two-step process results in a double percolated morphology where PS percolates the PLA phase, and MWCNTs percolate the PS phase. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) evidence, re-

SC

spectively, the bicontinuous morphology and the high selectivity of MWCNT localization in the system. Double percolation leads to a very low electrical percolation threshold of about 0.45 vol% MWCNT on

AC C

EP

TE D

M AN U

total volume and an electrical conductivity of 10-9 S/cm.

2

ACCEPTED MANUSCRIPT

1. Introduction Blending different polymers is a common strategy in order to modulate or improve the properties of a

RI PT

polymeric material or to introduce new properties [1]. Due to the high molecular weight, achieving compatibility and miscibility of polymers is not a trivial task. In most cases, the mixed polymers undergo phase separation, since reducing interfacial area leads to minimization of the system free energy. The morphologies of phase separated immiscible polymer blends can be divided into four categories: 1)

SC

globular inclusions (of the minor phase) in a continuous matrix, 2) fibers (of the minor phase) in a con-

M AN U

tinuous matrix, 3) alternating superimposed lamellae of the two phases, and 4) bicontinuity [2-4]. Great interest has the bicontinuous morphology, which is characterized by the formation of two connected and continuous complementary phases within the whole volume of the material. In such a morphology it is not possible to define a majority phase in which the minority phase is included. The main strategy used to determine the formation of a bicontinuous morphology is selective extraction of a phase. With the use

TE D

of an orthogonal solvent is possible to selectively remove one of the two phase leaving intact the other one, and forming a porous self-standing structure. From the weight difference measured before and after extraction the bicontinuity coefficient Φ can be calculated [5]:   

EP

Φ=

(1)

AC C

where  and  are initial and final weight of the sample, and  is the weight ratio of the extracted

phase. If the sample maintains a self-standing structure after extraction, and the value of Φ is close to unity, then the material is bicontinuous. Only internal regions of the extracted phase connected to the surface are available to the solvent for the extraction. After the extraction a macro-porous material is formed, with connected pores usually in the dimension range of 1-100 µm.

3

ACCEPTED MANUSCRIPT

Another important technological aspect regarding polymer blends is the addition of fillers, with the aim of reducing total cost or imparting functional properties to the materials. Nanocomposites belong to this

RI PT

last category, and they can be obtained by adding nanosized particles with specific property to one or more polymer with the objective of conferring these properties to the final material [6,7].

Nanocomposites can be produced using solvent mediated technologies or by melt mixing. Solvent assisted approaches combined with proper chemical functionalization of the nanofillers can lead to their

SC

good dispersion within polymer matrices [8-12]. On the other hand, melt mixing is one of the most used

M AN U

industrial process being fast, cheap and easily scalable. Moreover it also avoids the issue of waste solvents disposal. For multiphasic immiscible blends, the addition of nanoparticles can result in a remarkable enhancement of thermal and electrical properties [13,14]. From a morphological point of view, the fillers can show an affinity with one of the components of the blend, tending to concentrate selectively into one of the phases or at the interface between two phases [15-18].

TE D

Among nanoparticles, carbon nanotubes (CNT) possess many interesting properties (e.g. elastic modulus, tensile strength, electrical conductivity). Under proper conditions, dispersion of single walled (SWCNT) or multi walled (MWCNT) carbon nanotubes in a polymer matrix can lead to an improve-

EP

ment of elastic modulus or give new functional properties. When the amount of CNT filler in the polymer matrix is enough to create a percolated network within the whole volume of the material, then the

AC C

polymer can show an increase in electrical conductivity of several orders of magnitude. CNTs are characterized by high form factor (103÷105), high electrical conductivity (106 S/m), similar to copper [19], theoretical Young modulus and tensile strength of 4 TPa and 220 GPa [20]. The coincidence of all these astonishing properties allows to obtain in nanocomposites higher electrical and mechanical performance improvements compared to those obtained using other nanosized particles.

4

ACCEPTED MANUSCRIPT

The dispersion state of the nanoparticles in the polymeric matrix is of fundamental importance in order to impart these properties to the final material. To this aim, it is necessary to counterbalance the attrac-

RI PT

tion energy between particles due to Van der Waals interaction [21-23]. For these systems, a concentration percolation threshold is defined that is the critical concentration of nanoparticles that leads to the formation of a continuous network within the material. When this critical concentration is reached, most or all the properties of the material show a step-like variation. The percolation threshold is shown as be-

SC

ing inversely related to the form factor of CNTs, and is strictly related to the dispersion state of the par-

conductivity) the following law is true [27]:

M AN U

ticles [24-26]. It is possible to demonstrate that for a generic property of the material (e.g. electrical =   −  

(2)

where  is a scale factor dependent on the nature of the nanoparticles,  is the volume fraction of filler,

 is the percolation threshold and  is the critical exponent. Theoretical calculations for cylindrical par-

TE D

ticles state that the minimum value for the percolation threshold should be around 0.05 vol% [28,29]. Thresholds of 0.04 and 0.05 wt% have been reported for epoxy resin/MWCNT[30,31] and polypropylene (PP)/MWCNT composites, respectively [32]. Except for these examples, the typical percolation

EP

threshold values found in literature are comprised between 1 and 5 wt% [33]. Thanks to computer as-

AC C

sisted calculations Kirkpatrick [34] and Balberg [35] valuated that the critical exponent should be equal to 1.3 for bi-dimensional systems and to 2 for tri-dimensional systems. In spite of this, much more scattered values of the critical exponent, between 1 and 10, are found in literature [36]. The final properties of the nano-composites are strictly related to the dispersion state of the filler. When two or more polymers are mixed together, specific interactions between particles and polymer components could result in particles to accumulate in one phase or to localize at the interface. This behavior depends on the system thermodynamics (i.e. relative wettability) as well as on kinetic factors (i.e. mix-

5

ACCEPTED MANUSCRIPT

ing parameters). In equilibrium conditions, the localization of the filler will be driven by the minimization of the energy. To predict the final position of the particles, a wettability coefficient can be calculated, defined as [37]:  

RI PT

=

(3)



where  is the interfacial energy between components, A and B representing the two polymers and C

the filler. The value of  determines which is the equilibrium position of the nanoparticle. If  > 1,

SC

particles will be localized in the phase A; if  < −1 in phase B; otherwise if −1 <  < 1, the interface

M AN U

will be the preferred region of dispersion. To calculate the   interfacial energy, two different approaches are possible [38]. For high surface energy material the mean harmonic approximation is used:   =  +  − 4 "

# $#

# %$#

& &

 $

+ &%& ' 

$

(4)

while for low surface energy materials the geometric mean approximation holds:

TE D

  =  +  − 2 )* + + + * , , -

(5)

In both equations,  + and  , are the dispersive and polar components of surface energy of the compo-

EP

nents, respectively.

In 1991, Sumita et al. introduced for the first time the concept of double percolation [39], for multi-

AC C

phasic nanocomposites for which the following conditions are satisfied at the same time: 1) uneven distribution of the filler in the phases; 2) achievement of the percolation threshold in one of the phases; 3) structural continuity of the percolated phase within the whole volume of the material. If all these conditions are satisfied the material is defined double percolated. The interest in double percolated systems lies on the need of much lower filler concentration, on the total volume, to reach high performance (e.g. electrical conductivity) [40-42].

6

ACCEPTED MANUSCRIPT

In this work we present the selective localization of MWCNT in PLA/PS bicontinuous blends. We prepared a masterbatch of PS with 2 wt% content of MWCNT, which was melt mixed with different

RI PT

amounts of PLA. Thanks to this two-step process we obtained a double percolated morphology where the MWCNT percolate the PS phase, forming a co-continuous network together with PLA. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) of neat and PS-selectively extracted blends were used to verify, respectively, the formation of a bicontinuous morphology and the se-

M AN U

about 0.34 vol% MWCNT on the total volume.

SC

lective dispersion of the filler. Double percolation led to a very low electrical percolation threshold of

2. Experimental 2.1. Materials

PS (Edistir 2982; Mw = 180,000 g/mol, Mw/Mn = 2.3, Melt flow index 7.5g/10min @ 200°C/5 kg) was

TE D

purchased from Polimeri Europa (Italy), PLA was the 2002D grade from Natureworks (Minneapolis MN, USA; Mw/Mn = 1.89; Mw = 187,000; Melt flow index 2g/10 min @ 200°C/5kg). MWCNT grown by chemical vapor deposition (nominal purity > 95%) were purchased from Cheap Tubes Inc. (Brattle-

EP

boro VT, USA). Their diameter was between 30 and 50 nm. 2.2. Sample preparation

AC C

PS/PLA blends at different composition (20/80, 30/70, 40/60, 50/50, 60/40 and 70/30 PS to PLA weight ratio) were prepared introducing proper amounts of PS and PLA in the chamber (15 cm3) of a twinscrew microcompunder (DSM Xplore MC15). Mixing was performed at 180 °C for 5 minutes, at the speed of 150 rpm, under an inert nitrogen atmosphere in order to protect polymer from degradation. In order to obtain a double percolated system a two-step process was chosen. In the first phase, three different masterbatches of polystyrene, charged with 1, 2, or 6 wt% MWCNT were prepared in a 50 cm3

7

ACCEPTED MANUSCRIPT

blending chamber of a two-roll Brabender Plastograph mixer at 180 °C for 10 minutes, under an inert nitrogen atmosphere. Compression molded disks (1 mm thick) were obtained from the different blends

berg, Germany) progressively increasing the pressure up to 100 bar.

RI PT

by compression molding at 180 °C for 10 min using a P200E bench top press (Dr Collin GmbH, Ebers-

The system containing 2 wt% of MWCNT (coded as PS2CNT) was chosen as reference masterbatch since this was the minimum content of MWCNT leading to electrical percolation (see 3.X. Electrical

SC

conductivity). Therefore, PS2CNT was used to produce blends with PLA (20/80, 30/70, 40/60, 50/50

M AN U

and 60/40 PS2CNT to PLA weight ratio). Mixing of PS2CNT with PLA was performed at 180 °C for 5 minutes at 150 rpm under nitrogen using the DSM Xplore MC15 twin-screw microcompounder. The same procedure used to prepare the PS masterbatch was applied to realize a masterbatch of PLA containing 2 wt% of MWCNT (coded as PLA2CNT). By adding the appropriate amount of PS, this

sample PS/PLA2CNT 50/50.

TE D

masterbatch was used to realize, through the DSM Xplore MC15 twin-screw microcompounder, the

In order to evaluate the effect of thermal annealing on the electrical properties of the blends, thermal post-treatments were performed on the molded PS2CNT/PLA disks at different temperatures (120, 140,

AC C

EP

160 and 180 °C) for 1 h in oven.

2.3. Rheological measurements Melt rheological properties of PLA, PS and PS2CNT were measured using a Thermo Scientific RheoStress 6000 rotational rheometer at 180 °C under nitrogen. Circular disks of 25 mm diameter obtained from granules were placed between parallel plates setting a gap of 1 mm. Frequency sweep analysis was performed in a 0.1 to 100 rad/s range (backward and forward) in the linear viscoelastic regime, with a constant stress of 50 Pa.

8

ACCEPTED MANUSCRIPT

2.4. Bicontinuity coefficient The bicontinuity coefficient was evaluated selectively removing the PS phase using hot cyclohexane.

RI PT

Samples of about 1 g were weighted and then immersed in vials containing cyclohexane at 60 °C for 24 hours. Afterwards, the samples were put in an oven at 70 °C, under vacuum, for 24 hours in order to remove the residual solvent, and then weighed again. This procedure was repeatedly performed until constant weight was reached (or until the loss of structural resistance), and the bicontinuity coefficient

M AN U

2.5. Morphology and selective localization of MWCNT

SC

was calculated as in Equation 1. Measurements were performed in triplicate.

The formation of a bicontinuous morphology was verified performing SEM analysis of extracted samples. Extruded strands of the melt mixed samples were cryo-fractured and then subject to solvent extraction as above described. SEM analysis was performed on a FEI Quanta 200 FEG SEM in high vacuum using a secondary electron detector and an acceleration voltage ranging from 10 to 30 kV. Before analy-

TE D

sis, samples were coated with a 15 nm thick Au/Pd layer with a sputter coating system. Dynamic temperature-controlled SEM experiments were carried out on the above mentioned SEM using a hot stage and a gaseous secondary electron detector (GSED) at 30 kV acceleration voltage and PH2O

rate 50°C/min).

EP

= 0.90 torr. During SEM observation, samples were heated from room temperature to 140°C (heating

AC C

The selective localization of MWCNT in the polymer blend was evaluated by TEM analysis. Ultrathin sections of selected non-extracted samples were obtained with a Leica UC7 ultramicrotome system operating at room temperature. The sections were then placed on copper grids and observed in bright field mode on a FEI Tecnai G12 Spirit Twin TEM operating at 120 kV acceleration voltage.

9

ACCEPTED MANUSCRIPT

2.6. Thermal characterizatiom Differential scanning calorimetry (DSC) was performed on a Mettler Toledo DSC822e calorimeter. The calorimeter was calibrated in temperature and energy using indium. Dry nitrogen was used as purge gas

RI PT

at a rate of 30 cm3/min during all measurements. Small piece of samples, weighting about 10 mg, were encapsulated in standard aluminum 40 µl pans, then heated at a heating rate of 20 °C/min from 0 to 220 °C. All measurements were repeated three times to improve accuracy.

SC

Thermogravimetric analysis (TGA) was performed by a Perkin Elmer Pyris 1 TGA. Samples were put in a platinum crucible and heated at a scan rate of 20° C/min from 50 °C to 750 °C under a nitrogen gas

M AN U

flow rate of 20 cm3/min. 2.7. Electrical conductivity

Volume conductivity measurements were performed using a Keithley Electrometer/High resistance meter (Model 6517A) coupled with the Resistivity Test Fixture (Model 8009). In particular, circular disks

TE D

(6 cm diameter and 1 mm thickness) were obtained from the different blends by compression molding at 180 °C for 10 min using a P200E bench top press (Dr Collin GmbH, Ebersberg, Germany). The disks were then placed between the test fixture electrodes and the conductivity was measured using an alter-

EP

nating voltage between -10V and +10V, switching from positive to negative potential every 30 seconds.

AC C

The conductivity was measured 8 times for each sample discarding the highest and the lowest measures.

3. Results and Discussion 3.1. Rheology

The formation of bicontinuous domains during polymer melt mixing depends on the ratio of viscosities at the adopted processing condition, for a given volume ratio. As demonstrated by Pötschke and Paul

10

ACCEPTED MANUSCRIPT

[4], a bicontinuous morphology is formed when the following equation is satisfied, under certain assumptions: $ .

=1

(7)

RI PT

 .$

where / ,1 are the volume fractions of the polymers and 2/ ,1 are the viscosities. In general, to obtain a bicontinuous morphology at a low volumetric content of one of the polymers, a high viscosity ratio is

SC

needed. For this reason, rheological tests were performed on neat PS and PLA, as well as on the PS2CNT masterbatch. In Figure 1a, the viscosities of PS, PLA and PS2CNT are reported. The complex

M AN U

viscosity of PLA showed a clear Newtonian plateau at low shear rates. It is clearly visible that complex viscosity is increased for the PS2CNT sample compared to neat PS. While PS almost reached the Newtonian plateau, PS2CNT showed a mixed viscoelastic/solid-like behavior suggesting that the MWCNTs network was not completely formed. In Figure 1b, the elastic and dissipative moduli are reported. The

TE D

crossover between elastic and dissipative moduli for PLA occurs at ~25 s-1 with a wide Newtonian plateau at low frequencies. A completely different behavior is shown by PS with a very short plateau and a much lower crossover frequency at ~1 s-1 that shifts to ~0.5 s-1 for the PS2CNT sample, due to the start-

EP

ing transition to a less dissipative system at low frequencies [43]. From these data we could estimate that phase inversion could occur at 60 wt% of PS. The prevision is made approximating the shear rate,

AC C

expressed in s-1, experienced by the polymers during the mixing with the velocity of twin screws, expressed in rpm (~102 s-1), as proposed by Wu in 1987 [44]. 3.2. Bicontinuity coefficient

Extracting one of the phases is a common characterization technique for bicontinuous blends. As both polymers are continuously connected within the whole volume of the material, a selective solvent can extract completely one of the two phases. Through weight measurements performed

11

ACCEPTED MANUSCRIPT

before and after extraction it is possible to determine if a bicontinuous morphology is formed and the amount of the extracted phase that is part of the continuous network. In Figure 2 we notice that

RI PT

the bicontinuity coefficient of the PS/PLA (Figure 2, red line) system starts to change at 40-50 wt% of PS, going from zero to 0.1, and then abruptly increases to 0.7 at 60 wt% of PS. It is worth to be noted that for the sample containing 70 wt% of PS the blend has no longer a bicontinuous morphology but presents spherical inclusions of PS in a continuous matrix of PLA (Figure S1 in

SC

Supporting Information), and a complete loss of structural integrity was recorded after extraction,

M AN U

so that this point should not be taken into account. The PS2CNT/PLA blends showed a linear in-

AC C

EP

TE D

crease of bicontinuity from 30 wt% to 60 wt% (Figure 2, blue line).

12

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

Figure 1: Complex viscosities (a), elastic and dissipative moduli (b) of PLA, PS and PS2CNT as a function of frequency at 180 °C.

13

SC

RI PT

ACCEPTED MANUSCRIPT

M AN U

Figure 2: Bicontinuity coefficient for PS/PLA and PS2CNT/PLA blends after selective extraction of PS and PS2CNT phase in warm cyclohexane.

3.3. Morphology

Selective extraction allowed to determine the obtained morphology from SEM imaging. In Figure 3, SEM images of PS/PLA blends at different weight fractions, namely 40/60 (Figure 3a,b), 50/50 (Figure

TE D

3c,d), and 60/40 (Figure 3e,f) are reported. It is possible to notice that the average size of PS cavities is almost constant (in the range 2-5 µm), whereas the PLA domain dimension decreases as the PS weight fraction increases.

EP

Then we proceeded to verify the effective localization of MWCNT in the blends prepared with the

AC C

PS2CNT masterbatch. As stated above, it should be possible to predict, through the evaluation of the wettability coefficient, which is the more favorable phase for the nanoparticles. For the PS/PLA/MWCNT ternary system the wettability coefficient is calculated using equation 3, where A is PS, B is PLA and C are MWCNTs. In Table 1 different literature values of surface energies of the three phases are reported. These values have been obtained at room temperature, except for PS2 and PS3, measured at 180 °C [45-48]. It is noticed that significantly different values for polar and dispersive components of the surface energies are reported by different authors. Moreover, polar and dispersive

14

ACCEPTED MANUSCRIPT

components show a strong variation with temperature and values for different temperatures are hardly available, therefore it is difficult to choose a proper set of surface energies to calculate the wettability

RI PT

coefficient. Unfortunately, direct calculation of these parameters is not trivial and cannot be easily performed in order to establish the correct values of surface energies. For this reasons, in Table 2 the wettability coefficient was calculated for any possible combinations of data from Table 1, for both harmonic and geometric approximations. All the values smaller than -1 are reported in red, indicating that

SC

MWCNTs should prefer the PLA phase; the only value between -1 and +1 is reported in black, suggest-

M AN U

ing that the particles should localize at the PS/PLA interphase; all the values bigger than +1 are reported in green, indicating PS as the favorite phase for the nanotubes. It is clear that this calculation is strongly

AC C

EP

TE D

dependent on the set of values chosen.

15

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 3: SEM micrographs of the bicontinuous morphology of selectively extracted PS/PLA blends at different weight fractions: (a, b) = 40/60, (c, d) = 50/50, (e, f) = 60/40.

3[mJ/m2] 43.5 36.6 40.2 40.7 40.6 40.7 45.3

AC C

PLA1 PLA2 PLA3 PS1 PS2 PS3 MWCNT

EP

Table 1: Surface energy coefficients, and dispersive and polar surface components for PS, PLA and MWCNTs. 34 [mJ/m2] 39.6 36.6 29.7 40.1 32.5 33.9 18.4

35 [mJ/m2] 3.9 0.0 10.5 0.6 8.1 6.8 26.9

Ref. (1) (1) (2) (1) (3) (3) (4)

16

ACCEPTED MANUSCRIPT

Table 2: Wettability coefficients of PS/PLA/MWCNT blends, calculated according to harmonic and geometric approximation. Wettability values above +1 (green) are obtained when PS is the energetically preferred phase for MWCNTs, values below -1 (red) indicate PLA as the preferred phase for the nanotubes, and the value between -1 and 1 accounts for nanoparticles which tend to segregate at the interphase. PLA3 -2.25 -9.54 -6.30

PLA1 -6.46 6.18 8.57

RI PT

PLA1 -3.42 5.03 6.79

Geometric PLA2 9.35 2.74 3.06

PLA3 -2.69 -10.68 -7.17

SC

PS1 PS2 PS3

Harmonic PLA2 -0.37 2.27 2.37

In spite of the largely scattered values of wettability coefficients, SEM and TEM images clearly show

M AN U

an asymmetrical distribution of MWCNTs. In Figure 4SEM images of the 50/50 PS2CNT/PLA blend are reported at three magnifications. The bicontinuous structure of the system and the smoother appearance of the PLA surface with respect to the PS surface are clearly shown (Figure 4(a,b) ). The magnification in Figure 4cshows that the inhomogeneity of the PS surface is due to the presence of MWCNTs,

AC C

EP

TE D

not present in the PLA.

Figure 4: SEM images of non-extracted PS/PLA 50/50 blends showing the bicontinuous morphology and the selective localization of MWCNTs in the PS phase.

17

ACCEPTED MANUSCRIPT

This is confirmed by the TEM images in Figure 5(a,b) where we can notice that MWCNT are present in PS (dark areas) and not in PLA (bright areas). Although the mixing time has a relatively long duration,

RI PT

MWCNTs remain segregated in the PS phase, in which they were originally incorporated, and PLA remains almost completely free of particles. In order to get further insight on this finding, a masterbatch containing 2 wt% MWCNTs in PLA was prepared (PLA2CNT). Comparing the dispersion state of MWCNTs in both PS2CNT and PLA2CNT masterbatches using TEM imaging (Figure S2 in SI), both

SC

nanocomposites evidenced homogeneous dispersion of nanotubes, with very few agglomerates. Mixing

M AN U

the PLA2CNT masterbatch with PS in a 50/50 weight ratio resulted in a great amount of particles migrating from the starting PLA phase to PS (Figure 5(c, d)). However, this process does not lead to the complete asymmetry obtained using the PS2CNT masterbatch.

Although the dispersion state of MWCNTs in PLA is very good, the obtained results indicate the pres-

TE D

ence of thermodynamic forces driving the particles from PLA to PS, or keeping them into PS. Moreover, another interesting behavior is the partial accumulation of MWCNT at the PS/PLA interface for the PS2CNT/PLA system, as shown in Figure 5(e, f) for PS2CNT/PLA. This can be due to a blocking effect

EP

of the interface during the diffusion of the particles. The force arising from the concentration gradient that should favor diffusion of the particles into PLA is balanced by the opposite force due to the stronger

AC C

affinity with PS. In Figure 6 it is also shown that PS inclusions in PLA phase incorporate a considerable number of nanotubes compared to the surrounding matrix. In conclusion, all these data suggest a much stronger affinity of MWCNTs for the PS phase.

18

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

Figure 5: (a,b) TEM images of the 50/50 weight ratio PS2CNT/PLA blend, showing the asymmetric dispersion of MWCNTs in the PS phase. The dark phase (PS) contains almost all the MWCNTs while few isolated particles are visible in the clear phase (PLA). (c, d) TEM images of PLA2CNT masterbatch melt mixed with PS in a 50/50 weight ratio. It is clearly shown that a great amount of MWCNTs migrate from the starting PLA phase (bright areas) to the PS (dark areas) during processing.

19

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

3.4. Thermal analysis

EP

TE D

Figure 6: a) TEM images of the 50/50 PS2CNT/PLA blend, showing the accumulation of MWCNT stopped at the PS/PLA interface during the migration from the PS phase to PLA; b) TEM images of the 50/50 weight ratio PS2CNT/PLA blend, showing the extremely selective localization of MWCNTs in PS small inclusions in the PLA phase.

Thermal properties of neat PS and PLA, masterbatches, and PS/PLA 50/50 blends neat or containing MWCNTs, were analyzed by differential scanning calorimetry. DSC thermograms are reported in Figure S3a and results are summarized in Table 3. Neat PS only shows the presence of a glass transition at 90 °C (Tg1 in Table 3). As concerning neat PLA, the DSC trace reveals a glass transition (Tg2) centered at 64 °C associated to an enthalpic relaxation [49], an exothermic transition ascribed to cold crystallization of PLA upon heating, centered at Tc 95 °C, and a final melting endotherm at Tm 170 °C [50-52]. By

20

ACCEPTED MANUSCRIPT

subtracting the area under the exothermic peak (∆Hc) from that under endothermic peak (∆Hm), and considering a melting enthalpy of 100% crystalline PLA of 142 J/g [53], it can be evidenced that PLA is

RI PT

characterized by a crystallinity of 0.05. Regarding the masterbatches, MWCNTs induce a slight increase in the glass transition temperature of polystyrene in PS2CNT, possibly due to a restricted mobility of the polymer chains in presence of the rigid filler, whereas in the case of PLA2CNT the effect of the nanofillers on Tg is negligible. In the case of PLA2CNT the exothermic transition ascribed to cold crystalli-

SC

zation of PLA is centered at Tc 98 °C, and the final melting endotherm occurs at Tm 169 °C, with a crys-

M AN U

tallinity of 0.11. Similar results were found for the PS/PLA 50/50 blend and for the PS2CNT/PLA and the PS/PLA2CNT blends. In all these cases, the Tg of the PS phase is 94 °C, very close to the value shown by the masterbatch PS2CNT, and no significant differences were observed for the Tg of the PLA phase, close to 65 °C for all samples. Furthermore, in all these blends, the presence of PS induces an increase of the temperature at which cold crystallization of PLA occurs, centered between 108 and 113

TE D

°C, while the melting temperature of the PLA phase (centered at 171 °C) and the crystallinity values were not affected, independently from the presence of MWCNTs in the blends. Thermal stability of neat PS and PLA, masterbatches, and PS/PLA 50/50 blends neat or containing

EP

MWCNTs were evaluated by TGA analysis. The onset of the thermal degradation (Td), calculated at the

AC C

intersection between the baseline and a tangent line drawn at the maximum rate of weight loss, the temperatures corresponding to the maximum degradation rate (Tmdr1 and Tmdr2) and the residual weight at 550 °C (RW550) are reported in Table 3. Thermogravimetric (TG) and derivative thermogravimetric (DTG) curves are shown in Figure S3(b,c). As concerning PS, the thermal degradation occurs in a single step, with an onset temperature (Td) of 408 °C and a maximum degradation rate (Tmdr2) of 433 °C. The thermal degradation of PS occurs through a chain scission mechanism that yields a mixture of saturated and unsaturated compounds, most of which are monomers, dimers or trimers [54]. The thermal degrada-

21

ACCEPTED MANUSCRIPT

tion of PLA also occurs in a single step, through a hydroxyl end-initiated ester interchange process and chain homolysis that brings to the formation of lactide, acetaldehyde, carbon monoxide and carbon di-

RI PT

oxide [55,56]. For PLA the onset temperature (Td) is 359 °C and the temperature corresponding to the maximum degradation rate (Tmdr1) is 379 °C.

As already reported, addition of MWCNTs can induce an appreciable thermal stabilizing effect to both PS and PLA [57,58]. This effect is confirmed by the TGA analysis of PS2CNT and PLA2CNT mas-

SC

terbatches, with an increase of 11-15 °C recorded for both the polymers on both Td and Tmdr. As con-

M AN U

cerning the blends, the sample PS/PLA 50/50 shows a two-step degradation mechanism, the first step attributed to the degradation of the PLA phase and the second one to the PS phase [59]. While a slight decrease of Td is recorded for this system, the temperatures corresponding to the maximum degradation rates of the two steps (Tmdr1 377 °C, Tmdr2 433 °C) are equal or very close to the corresponding temperature observed for neat PLA (Tmdr1 379 °C) and PS (Tmdr2 433 °C). This is very interesting in comparison

TE D

to what observed for both blends containing nanotubes, in which an appreciable increase of Tmdr was found for the step attributed to the thermal degradation of PS (Tmdr2 443 °C for both PS2CNT/PLA 50/50 and PS/PLA2CNT 50/50). This increase in Tmdr2 is comparable to that observed for the PS2CNT

EP

system. On the contrary, in contrast to what observed for the PLA2CNT masterbatch, Tmdr1 of the PS2CNT/PLA 50/50 is very close to that of neat PLA. Moreover, in the case of PS/PLA2CNT 50/50 a

AC C

slight decrease of Tmdr1 is observed. In agreement with the results obtained by morphological analysis, these findings are an indirect evidence of the selective localization of MWCNTs in the PS phase.

Table 3: DSC and TGA results. Glass transition temperatures (Tg1, Tg2), crystallization temperature and enthalpy (Tc, ∆Hc), melting temperature and enthalpy (Tm, ∆Hm), crystallinity (xc,PLA), onset temperature of the decomposition (Td), temperatures corresponding to the maximum decomposition rate (Tmdr1, Tmdr2), and residual weight (RW) at 550°C.

22

ACCEPTED MANUSCRIPT

Tc [°C] 95 98 113

∆Hc [J/g] 25.1 21.4 13.3

Tm [°C] 170 169 171

∆Hm

xc,PLA

[J/g] 32.6 37.2 17.9

0.05 0.11 0.06

94

110

19.1

171

21.9

0.04

94

108

-

14.5

-

3.5. Electrical conductivity

-

171

-

19.4

-

Tmdr1 [°C] 379 394

Tmdr2 [°C] 433 444

RW550 [wt%] 0.0 0.6 2.2 2.3

344

377

433

0.0

353

381

443

1.4

343

373

443

3.0

636

-

-

96.9

SC

Tg2 [°C] 90 95 94

TGA Td [°C] 408 352 421 365

0.07

-

M AN U

DSC Tg1 [°C] PS PLA 64 PS2CNT PLA2CNT 66 PS/PLA 64 50/50 PS2CNT/ 65 PLA 50/50 PS/PLA2 66 CNT 50/50 MWCNT -

RI PT

Sample

The electrical conductivities of the different PS masterbatches were compared in order to evaluate the

TE D

minimum concentration of MWCNT needed to reach the electrical percolation threshold. As it can be seen in Figure 7, the 1 wt% MWCNT masterbatch does not show any electrical conductivity difference compared to pure PS (i.e. 10-17 S/cm). For the sample containing 2 wt% MWCNT the electrical conduc-

EP

tivity was raised by 10 orders of magnitude (10-7 S/cm). The increase of MWCNT content to 6 wt% MWCNT further increased the conductivity of only one order of magnitude, indicating that the electri-

AC C

cal percolation threshold was already reached at 2 wt%. For this reason we selected the masterbatch PS2CNT for the preparation of the blends with PLA at different PS2CNT/PLA weight ratios.

23

SC

RI PT

ACCEPTED MANUSCRIPT

M AN U

Figure 7: Electrical conductivity of PS/MWCNT composites as function of the MWCNT content.

Figure 8a shows the electrical conductivity of the molded disks for the PS2CNT/PLA blends at different compositions. The red line indicates unchanged electrical conductivity values with respect to the un-

TE D

filled PLA. In 2007 Pötschke et al. [60] showed that one explanation for the loss of electrical conductivity in a MWCNT/polymer blend can be related to the destruction of the percolative network when the system is subjected to a transient shear. In this case a thermal annealing process can help to fully recov-

EP

er the electrical conductivity giving the system enough thermal energy to reconstruct the MWCNT network. Another possible explanation is related to the morphological structure of the system. Zhang et al.

AC C

[61] have shown that compression molding of a bicontinuous PS/PLA blend can result in the formation of a PLA insulating surface layer that can be reabsorbed in the bulk thanks to an annealing treatment. The formation of a superficial insulating layer was also found in the work of Gigli et al. [62] where a solvent cast film of MWCNT and polypropylene needed a plasma etching treatment to expose the nanoparticles, as they were covered by an insulating thin layer of polymer. On these bases, we performed a thermal post-treatment on the molded PS2CNT/PLA samples at different temperatures (120, 140, 160

24

ACCEPTED MANUSCRIPT

and 180 °C) for 1 h, repeating the electrical conductivity tests after the annealing. The first sample, annealed at 120 °C showed no macroscopic deformation, but at the same time no recovery in electrical conductivity was noticed. The other three samples showed, conversely, a clear increase in electrical

RI PT

conductivity of several orders of magnitudes (~109 S/cm) but different macroscopic appearances. The two samples annealed at higher temperatures (160 and 180 °C) were clearly deformed due to complete melting of the polymers and the bicontinuous morphology was lost, as proved by the samples break-

SC

down after solvent extraction of PS. Only the sample annealed at 140 °C for 1 h showed both the elec-

M AN U

trical conductivity recover and a bicontinuous structural integrity. As both of the mechanisms described above, i.e. reconstruction of the nanoparticle network and reabsorption of an external insulating polymer layer, could be responsible for the recover in the electrical conductivity, we analyzed through TEM and SEM imaging nanotube dispersion state in the bulk as well as surface morphology of solvent extracted samples. TEM analysis (Figure S4 in Supporting Information) showed no evidence of nanotube network

TE D

reconstruction because the dispersion state of MWCNTs in the PS phase after annealing was comparable to that evaluated on untreated samples. Instead, by comparing the surfaces of PS2CNT/PLA 50/50 before and after annealing (Figure 8(b,c)), it is evident that the thermal treatment can effectively cause

EP

the partial reabsorption of the insulating PLA surface layer. This phenomenon was confirmed by temperature-controlled SEM experiments, in which during heating from room temperature to 140 °C, a pro-

AC C

gressive increase of contrast was recorded due to the absorption of the PLA layer and the progressive exposure of the PS phase (dark regions in Figure 8e, see also video S5 in the Supporting Information) to the surface of PS2CNT/PLA 50/50. Once the annealing conditions were optimized, the electrical conductivity was measured on samples with different PS2CNT/PLA weight ratios. As shown by the blue line in Figure8a, after the annealing all

25

ACCEPTED MANUSCRIPT

the samples above 40 wt% of PS2CNT showed an increase of the electrical conductivity of more than 8 orders of magnitude (from 10-17 to 10-9 S/cm). It is worth to be noted that the total amount of MWCNT for the 40 wt% PS2CNT sample is 0.8 wt%,

RI PT

leading to a MWCNT percent content of about 0.45 vol% on total volume (using 2.1 g/cm3 density for MWCNT [63], 1.25 g/cm3 for PLA and 1.05 g/cm3 for PS). This value is lower or comparable to the values of other bicontinuous nanocomposites blends using carbon based nanoparticles. In 1998 Sumita

SC

et al. described the electrical percolation of short carbon fiber filled polymer blends: high-density poly-

M AN U

ethylene (HDPE)/isotactic polypropylene (iPP) [64], and HDPE/poly(methyl methacrylate) [65]. In these works the electrical percolations were respectively 1.25 phr and 1.5 phr (equivalent to 1.23 wt% and 1.47 wt%). Meincke et al. [66] studied the electrical behavior of the ternary blend of polyamide-6 (PA6) and acrylonitrile/butadiene/styrene (ABS) with carbon nanotubes that, thanks to the selective localization of CNT, lead to an onset in the electrical conductivity at 2-3 wt% of filler loading. Moreover Bose

TE D

et al. [67] studied the same PA6/ABS blend using two different types of MWCNTs. In both cases, the electrical percolation was observed 3-4 wt%, while the rheological percolation was in the range 1-3 wt%. The absolute value of the electrical conductivity can also be compared to previous works, as it was

EP

found to be 2 orders of magnitude lower than the value showed by the PS2CNT masterbatch. Similar results were obtained by Pötschke et al. [16] and Fina et al. [68], and can be explained by considering the

AC C

reduced amount of electrically conductive material (in our case the PS2CNT phase), as well as the longer electrically conductive path due to the bicontinuous structure. According to this hypothesis, electrons are forced to follow a much longer pathway to go from one side to the other side of the sample, because they must follow the percolative MWCNT network within the tortuous PS phase that percolates the entire volume of the sample.

26

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 8: electrical conductivity of PS2CNT/PLA blends as a function of the PS2CNT weight content before (blue line) and after (red line) annealing (a); false color SEM images of extracted PS2CNT/PLA 50/50 before (b) and after (c) annealing, showing both surface and bulk of the samples; temperaturecontrolled SEM analysis of PS2CNT/PLA 50/50 before (d, 20°C) and after (e, 140°C) heating (see also video S2 in the supporting information);

27

ACCEPTED MANUSCRIPT

4. Conclusions In this work we reported the preparation of double percolated blends of PS, PLA and MWCNT starting

RI PT

from a PS masterbatch containing 2 wt% MWCNT. Melt mixing of the PS masterbatch and PLA, under proper process conditions, led to a double percolated morphology. The blends showed bicontinuity and, after a thermal annealing post-treatment, an exponential electrical conductivity increase of about 8 orders of magnitude. Therefore, the possibility of obtaining electrically conductive polymers by using

SC

MWCNT selectively dispersed in a multiphasic matrix was demonstrated. The blend micro-structuring

M AN U

had a key role in significantly reduce the electrical percolation threshold of the system, and in turn the total amount of nanoparticles needed to improve the electrical performance of the samples.

Acknowledgements

TE D

GN acknowledges for financial support Campania Region (MASTRI Excellence Network – Materiali, and strutture intelligenti – POR Campania 2007/2013).

EP

References

AC C

[1] B.D. Favis, Factors influencing the morphology of immiscible polymer blends in melt processing. In Polymer Blends Volume 1: Formulation; John Wiley & Sons: New York, NY, 2000; Vol. Polymer, pp 239-289. [2] M. Hara, J.A. Sauer, Synergism in Mechanical Properties of Polymer/Polymer Blends. Journal of Macromolecular Science, Part C: Polymer Reviews, 38 (1998) 327-362. [3] L.A. Utracki, Commercial Polymer Blends; Chapman & Hall, 1998. [4] P. Pötschke, D.R. Paul, Formation of Co-continuous Structures in Melt-Mixed Immiscible Polymer Blends. Journal of Macromolecular Science, Part C: Polymer Reviews, 43 (2003) 87-141. [5] K.L. Rasmussen, J. Lyngaae-Jørgensen, E.A. Chtcherbakova, L.A. Utracki, Flow induced deformation of dual-phase continuity in polymer blends and alloys. Part I. Polymer Engineering & Science, 39 (1999) 1060-1071.

28

ACCEPTED MANUSCRIPT

[6] W. Zhao, Y. Tang, J. Xi, J. Kong, Functionalized graphene sheets with poly (ionic liquid) s and high adsorption capacity of anionic dyes. Applied Surface Science, 326 (2015) 276-284. [7] J. Gu, N. Li, L. Tian, Z. Lu, Q. Zhang, Q. High thermal conductivity graphite nanoplatelet/UHMWPE nanocomposites. RSC Advances, 5 (2015) 36334-36339.

RI PT

[8] S.H. Choi, D.H. Kim, A.V. Raghu, K.R. Reddy, H.I. Lee, K.S. Yoon, H.M. Jeong, B.K. Kim, Properties of graphene/waterborne polyurethane nanocomposites cast from colloidal dispersion mixtures. Journal of Macromolecular Science, Part B, 51 (2012) 197-207.

SC

[9] M. Hassan, K.R. Reddy, K.R. Haque, A:I, Minett, V.G. Gomes, V. G. High-yield aqueous phase exfoliation of graphene for facile nanocomposite synthesis via emulsion polymerization. Journal of colloid and interface science, 410 (2013) 43-51. [10] K.R. Reddy, B.C. Sin, K.S. Ryu, J.C. Kim, H. Chung, Y. Lee, Conducting polymer functionalized multi-walled carbon nanotubes with noble metal nanoparticles: Synthesis, morphological characteristics and electrical properties. Synthetic Metals, 159 (2009) 595-603.

M AN U

[11] K.R. Reddy, H.M. Jeong, Y. Lee, A.V. Raghu, Synthesis of MWCNTs‐core/thiophene polymer‐ sheath composite nanocables by a cationic surfactant‐assisted chemical oxidative polymerization and their structural properties. Journal of Polymer Science Part A: Polymer Chemistry, 48 (2010) 14771484.

TE D

[12] S.J. Han, H.I. Lee, H.M. Jeong, B.K. Kim, A.V. Raghu, K.R. Reddy, Graphene modified lipophilically by stearic acid and its composite with low density polyethylene. Journal of Macromolecular Science, Part B, 53 (2014) 1193-1204. [13] J. Gu, Z. Lv, Y. Wu, R. Zhao, L. Tian, Q. Zhang, Enhanced thermal conductivity of SiCp/PS composites by electrospinning–hot press technique. Composites Part A: Applied Science and Manufacturing, 79 (2015) 8-13.

EP

[14] J. Gu, Q. Zhang, J. Dang, C. Yin e S. Chen, Preparation and properties of polystyrene/SiCw/SiCp thermal conductivity composites. Journal of Applied Polymer Science, 124 (2012) 132-137.

AC C

[15] F. Gubbels, R. Jerome, E. Vanlathem, R. Deltour, S. Blacher, F. Brouers, Kinetic and Thermodynamic Control of the Selective Localization of Carbon Black at the Interface of Immiscible Polymer Blends. Chemistry of Materials, 10 (1998) 1227-1235. [16] P. Pötschke, A.R. Bhattacharyya, A. Janke, Morphology and electrical resistivity of melt mixed blends of polyethylene and carbon nanotube filled polycarbonate. Polymer, 44 (2003) 8061-8069. [17] X. Zhao, J. Zhao, J. Cao, X. Wang, M. Chen, Z. Dang, Tuning the dielectric properties of polystyrene/poly(vinylidene fluoride) blends by selectively localizing carbon black nanoparticles. The Journal of Physical Chemistry B, 8 (2013) 2505-2515. [18] G. Filippone, A. Causa, M. Salzano de Luna, L. Sanguigno, D. Acierno, D. Assembly of plate-like nanoparticles in immiscible polymer blends – effect of the presence of a preferred liquid–liquid interface. Soft Matter, 10 (2014) 3183-3191.

29

ACCEPTED MANUSCRIPT

[19] T.W. Odom, J.-L. Huang, P. Kim, C.M. Lieber, Atomic structure and electronic properties of single-walled carbon nanotubes. Nature, 391 (1998) 62-64. [20] M. Treacy, T. Ebbesen e J.M. Gibson, Exceptionally high Young's modulus observed for individual carbon nanotubes. Nature, 381 (1996) 678-680.

RI PT

[21] J. Li, P.C. Ma, W.S. Chow, C.K. To, B.Z. Tang, J.K. Kim, Correlations between Percolation Threshold, Dispersion State, and Aspect Ratio of Carbon Nanotubes. Advanced Functional Materials, 17 (2007) 3207-3215.

SC

[22] G. Gentile, V. Ambrogi, P. Cerruti, R. Di Maio, G. Nasti, C. Carfagna, Pros and cons of melt annealing on the properties of MWCNT/polypropylene composites. Polymer Degradation and Stability, 110 (2014) 56-64.

M AN U

[23] V. Ambrogi, G. Gentile, C. Ducati, M.C. Oliva, C. Carfagna, Multiwalled carbon nanotubes functionalized with maleated poly(propylene) by a dry mechano-chemical process. Polymer, 53 (2012) 291299. [24] I. Balberg, Tunneling and nonuniversal conductivity in composite materials. Physical Review Letters, 59 (1987) 1305-1308. [25] J.G. Simmons, Generalized Formula for the Electric Tunnel Effect between Similar Electrodes Separated by a Thin Insulating Film. Journal of Applied Physics, 34 (1963) 1793-1803.

TE D

[26] M. Foygel, R.D. Morris, D. Anez, S. French, J.V. Sobolev Theoretical and computational studies of carbon nanotube composites and suspensions: Electrical and thermal conductivity. Physical Review B, 71 (2005) 104201. [27] S. Xu, O. Rezvanian, K. Peters, M.A. Zikry, The viability and limitations of percolation theory in modeling the electrical behavior of carbon nanotube–polymer composites. Nanotechnology, 24 (2013) 155706.

EP

[28] Y.Y. Huang, E.M. Terentjev, Dispersion of Carbon Nanotubes: Mixing, Sonication, Stabilization, and Composite Properties. Polymers, 4 (2012) 275-295.

AC C

[29] I. Balberg, Recent developments in continuum percolation. Philosophical Magazine Part B, 56 (1987) 991-1003. [30] J. Sander, M.S.P. Shaffer, T. Prasse, W. Bauhofer, K. Schulte, A.H. Windle, Development of a dispersion process for carbon nanotubes in an epoxy matrix and the resulting electrical properties. Polymer, 40 (1999) 5967-5971. [31] J.K.W. Sandler, J.E. Kirk, I.A. Kinloch, M.S.P. Shaffer, A.H. Windle, Ultra-low electrical percolation threshold in carbon-nanotube-epoxy composites. Polymer, 44 (2003) 5893-5899. [32] A. Rodney, J. David, M. Mickael, R. Terry, Fabrication of Carbon Multiwall Nanotube/Polymer Composites by Shear Mixing. Macromolecular Materials and Engineering, 287 (2002) 1439-2054. [33] W. Bauhofer, J.Z. Kovacs, A review and analysis of electrical percolation in carbon nanotube polymer composites. Composites Science and Technology, 69 (2009) 1486-1498.

30

ACCEPTED MANUSCRIPT

[34] S. Kirkpatrick, Percolation and Conduction. Reviews of Modern Physics, 45 (1973) 574-588. [35] I. Barlberg, N. Binenbaum, Computer study of the percolation threshold in a two-dimensional anisotropic system of conducting sticks. Physical Review B, 28 (1983) 3799-3812.

RI PT

[36] S. Wu. Polymer Interface and Adhesion; CRC Press, 1982. [37] A.C. Baudouin, C. Bailly, J. Devaux, Interface localization of carbon nanotubes in blends of two copolymers. Polymer Degradation and Stability, 95 (2010) 389-398. [38] P. Pötschke, S. Pegel, M. Claes, D. Bonduel, A Novel Strategy to Incorporate Carbon Nanotubes into Thermoplastic Matrices. Macromolecular Rapid Communications, 29 (2008) 244-251.

SC

[39] M. Sumita, K. Sakata, S. Asai, K. Miyasaka, H. Nagakawa, Dispersion of fillers and the electrical conductivity of polymer blends filled with carbon black. Polymer Bulletin, 25 (1991) 265-271.

M AN U

[40] J. Huang, C. Mao, Y. Zhu, W. Jiang, X. Yang, Control of carbon nanotubes at the interface of a cocontinuous immiscible polymer blend to fabricate conductive composites with ultralow percolation thresholds. Carbon, 73 (2014) 267-274. [41] S.T. Naira, P.P. Vijayan, P. Xavier, S. Bose, S.C. George, S. Thomas, Selective localisation of multi walled carbon nanotubes in polypropylene/natural rubber blends to reduce the percolation threshold. Composites Science and Technology, 116 (2015) 9-17.

TE D

[42] C. Mao, I. Zhu, W. Jiang, Design of electrical conductive composites: Tuning the morphology to improve the electrical properties of graphene filled immiscible polymer blends. ACS Applied Materials and Interfaces, 4 (2012) 5281-5286. [43] A.K. Kota, B.H. Cipriano, M.K. Duesterberg, A.L. Gershon, D. Powell, S.R. Raghavan, H.A. Bruck, Electrical and Rheological Percolation in Polystyrene/MWCNT Nanocomposites. Macromolecules, 40 (2007) 7400-7406.

EP

[44] S. Wu, Formation of dispersed phase in incompatible polymer blends: Interfacial and rheological effects. Polymer Engineering & Science, 27 (1987) 335-343.

AC C

[45] G. Biresaw, C.J. Carriere, Interfacial tension of poly(lactic acid)/polystyrene blends. Journal of Polymer Science Part B: Polymer Physics, 40 (2002) 2248-2258. [46] V. Khoshkava, M.R. Kamal, Effect of Surface Energy on Dispersion and Mechanical Properties of Polymer/Nanocrystalline Cellulose Nanocomposites. Biomacromolecules, 14 (2013) 3155-3163. [47] S. Wu, Surface and interfacial tensions of polymer melts. II. Poly(methyl methacrylate), poly(nbutyl methacrylate), and polystyrene. The Journal of Physical Chemistry, 74 (1970) 632-638. [48] S. Nuriel, L. Liu, A.H. Barber, H.D. Wagner, Direct measurement of multiwall nanotube surface tension. Chemical Physics Letters, 404 (2005) 263-266. [49] R. Avolio, R. Castaldo, G. Gentile, V. Ambrogi, S. Fiori, M. Avella, M.E. Errico, Plasticization of poly(lactic acid) through blending with oligomers of lactic acid: Effect of the physical aging on properties. European Polymer Journal 66 (2015) 533-542.

31

ACCEPTED MANUSCRIPT

[50] M.L. Di Lorenzo, P. Rubino, M. Cocca, Miscibility and properties of poly(L-lactic acid)/poly(butylene terephthalate) blends. European Polymer Journal 49 (2013) 3309-3317. [51] M.L. Di Lorenzo, P. Rubino, M. Cocca, Isothermal and non-isothermal crystallization of poly(Llactic acid)/poly(butylene terephthalate) blends. J. Appl. Polym. Sci., 131 (2014) 40372.

RI PT

[52] M. Cocca, R. Androsh, M.C. Righetti, M. Malonconico, M.L. Di Lorenzo, Conformationally disordered crystals and their influence on material properties: The cases of isotactic polypropylene, isotactic poly(1-butene), and poly(L-lactic acid). Journal Molecular Structures, 1078 (2014) 114-132. [53] Y. Baimark, P. Srihanam, Influence of chain extender on thermal properties and melt flow index of stereocomplex PLA. Polymer Testing, 45 (2015) 52-57.

SC

[54] J.D. Peterson, S. Vyazovkin, C.A. Wight, Kinetics of the thermal and thermo-oxidative degradation of polystyrene, polyethylene and poly(propylene). Macromol.Chem. Phys., 202 (2001) 775–784.

M AN U

[55] H. Zou, C. Yi, L. Wang, H. Liu, W. Xu, Thermal degradation of poly(lactic acid) measured by thermogravimetry coupled to Fourier transform infrared spectroscopy. Therm Anal Calorim, 97 (2009) 929-935. [56] S. Agustin-Salazar, N. Gamez-Meza, L. A. Medina-Juàrez, H. Soto-Valdez, P. Cerruti, From Nutraceutics to Materials: Effect of Resveratrol on the Stability of Polylactide. ACS Environmental Chemistry & Engineering, 2 (2014) 1534-1542.

TE D

[57] S.T. Kim, H.J. Choi, S.M. Hong, Bulk polymerized polystyrene in the presence of multiwalled carbon nanotubes. Colloid and Polymer Science, 285 (2007) 593-598. [58] D. Wu, L. Wu, M. Zhang, Y. Zhao, Viscoelasticity and thermal stability of polylactide composites with various functionalized carbon nanotubes. Polymer Degradation and Stability, 93, (2008) 1577– 1584.

EP

[59] A. Mohamed, S.H. Gordon, G. Biresaw, Poly(lactic acid)/polystyrene bioblends characterized by ther-mogravimetric analysis, differential scanning calorimetry, and photoacoustic infrared spectroscopy. Journal of Applied Polymer Science, 106 (2007) 1689-1696.

AC C

[60] I. Alig, T. Skipa, M. Engel, D. Lellinger, S. Pegel, P. Pötschke, Electrical conductivity recovery in carbon nanotube–polymer composites after transient shear. Physica Status Solidi B, 244 (2007) 42234226. [61] W. Zhang, S. Deodhar, D. Yao, Geometrical Confining Effects in Compression Molding of Cocontinuous Polymer Blends. Annals of Biomedical Engineering, 38 (2010) 1954-1964. [62] F. Malara, M. Manca, L. De Marco, P. Pareo, G. Gigli, Flexible Carbon Nanotube-Based Composite Plates As Efficient Monolithic Counter Electrodes for Dye Solar Cells. ACS Applied Materials & Interfaces, 3 (2011) 3625-3632. [63] J.H. Lehman, M. Terrones, E. Mansfield, K.E. Hurst, V. Meunier, Evaluating the characteristics of multiwall carbon nanotubes. Carbon, 49 (2011) 2581-2602.

32

ACCEPTED MANUSCRIPT

[64] C. Zhang, X.S. Yi, S. Asai, M. Sumita, Morphology and electrical properties of short carbon fiber‐ filled polymer blends: High‐density polyethylene/poly (methyl methacrylate). Journal of applied polymer science, 69 (1998) 1813-1819.

RI PT

[65] C. Zhang, X.S. Yi, S. Asai, M. Sumita, Selective location and double percolation of short carbon fiber filled polymer blends: high-density polyethylene/isotactic polypropylene. Materials Letters, 36 (1998) 186-190. [66] O. Meincke, D. Kaempfer, H. Weickmann, C. Friedrich, M. Vathauer, H. Warth, Mechanical properties and electrical conductivity of carbon-nanotube filled polyamide-6 and its blends with acrylonitrile/butadiene/styrene. Polymer, 45 (2004) 739-748.

SC

[67] S.Bose, A.R. Bhattacharyya, A.P. Bondre, A.R. Kulkarni, P. Potschke, Rheology, electrical conductivity, and the phase behavior of cocontinuous PA6/ABS blends with MWNT: Correlating the aspect ratio of MWNT with the percolation threshold. Journal of Polymer Science Part B: Polymer Physics, 46 (2008) 1619-1631.

AC C

EP

TE D

M AN U

[68] A. Fina, Z. Han, G. Saracco, U. Gross, M. Mainil, Morphology and conduction properties of graphite-filled immiscible PVDF/PPgMA blends. Polymers for Advanced Technologies, 23 (2012) 15721579.

33

ACCEPTED MANUSCRIPT

Double percolation of multiwalled carbon nanotubes

RI PT

in polystyrene/polylactic acid blends Highlights

Melt mixing of PS, PLA and MWCNTs results in a double percolated morphology

•

Highly selective localization of MWCNTs in the PS phase is evidenced

•

Double percolation leads to a low electrical percolation threshold

AC C

EP

TE D

M AN U

SC

•