The Polymer Processing Society 23rd Annual Meeting

TAILORING OF CELLULAR POLYMERS BY BLENDING

Holger Ruckdäschel

1, Volker Altstädt

1*, Axel H.E. Müller

2

1Department of Polymer Engineering, University of Bayreuth, Universitätsstrasse 30, 95447 Bayreuth, Germany –

altstaedt@uni-bayreuth.de; 2Department of Macromolecular Chemistry, University of Bayreuth, Universitätsstrasse 30,

95447 Bayreuth, Germany – axel.mueller@uni-bayreuth.de

The foaming behaviour of polymer blends was systematically investigated by using batch processing techniques and

carbon dioxide as a blowing agent. Binary blends over a wide compositional range were selected as reference materials,

including miscible (poly(2,6-dimethyl-1,4-phenylene ether)/poly(styrene) (PPE/PS) as well as immiscible PPE/SAN

(poly(styrene-co-acrylonitrile)) blends. Furthermore, the influence of compatibilization of PPE/SAN blends by triblock

terpolymers was investigated. The resulting foam characteristics were correlated to processing-relevant properties, in

order to elucidate the relationships between processing, structure and properties. This detailed investigation identified

additional factors determining the foamability of blends, besides the commonly accepted key criteria of neat polymers.

For example, the significant influence of the initial blend morphology, of the rheological as well as of the thermal prop-

erties was established. Furthermore, the compatibilization of these PPE/SAN blends by triblock terpolymers could be

identified as a technique to establish micro- and nanocellular foam morphologies with nanostructured cell walls. In view

of these results, new pathways for the controlled tailoring of cellular blends are derived.

Introduction

Today’s foams can cover an enormous range of applications such as load bearing, energy absorption, thermal as well as

acoustic insulation. However, the demand for fine-tuned, but also novel property profiles such as controlled cell mor-

phologies, improved barrier or mechanical properties, is steadily growing. Although the foaming of polymeric materials

has evolved into a well-developed technology for providing cellular materials with enhanced properties over the last

decades, a further improvement can only be achieved the development of new processing techniques or advanced mate-

rials. Here, the foaming of multiphase blend systems in particular can be identified as an approach that could potentially

satisfy these requirements and, thus, can be regarded as a topic of significant interest, both from the scientific as well as

from the commercial point of view.

There is an extensive body of literature on regarding the foaming behaviour of polymers as well as that of filled poly-

mers based using a variety of processing techniques [1,2]. Notwithstanding the capability to develop new classes of

cellular materials by polymer blends, the literature, in particular on foamed multiphase systems, is still limited. Under-

standing and controlling the foaming process of blends could offer completely new possibilities with regard to fine-

tuning both the processability and resulting properties as well as to developing new foam morphologies [3], as demon-

strated by first approaches [4-6]. Nevertheless, the potential of polymer blends, demonstrated in numerous fields of

applications [7], has not been fully exploited for foaming yet.

In this study, the batch foaming of polymer blends is systematically evaluated. Various binary mixtures of amorphous

components are selected as reference materials, including immiscible poly(2,6-dimethyl-1,4-phenylene

ether)/poly(styrene-co-acrylonitrile) (PPE/SAN), in particular. Furthermore, the influence of compatibilization with

triblock copolymers is investigated for PPE/SAN.

Experimental

Polystyrene (PS, BASF, grade 145D, glass transition temperature 104 °C), polystyrene-co-acrylonitrile (SAN, BASF,

VLL19, 114 °C) and poly(2,6-dimethyl-1,4-phenylene ether) (PPE, MEP, PX100F, 216 °C) were used as-received for

melt-blending. Miscible PPE/PS and immiscible PPE/SAN blends were prepared by twin-screw extrusion (Brabender,

DSE 20/40) and were subsequently pelletized. Moreover, a PPE/SAN 60/40 blend was melt-compatibilized by different

contents of polystyrene-b-polybutadiene-b-poly(methyl methacrylate) triblock terpolymers (SBM, 33 wt% PS,

34 wt% PB, 33 wt% PMMA, Mn=94 kg/mol, MWD=1.02). Such a compatibilization step leads to nanostructured blend

morphologies, also referred to as ‘raspberry morphology’ [8]. Further material characteristics have been discussed in

detail elsewhere [8,9].

The rheological behaviour of the blends under uniaxial elongational flow was characterized by the so-called Rheotens

method [10], using the compounded pellets as prepared. The experimental set-up for the Rheotens measurements in-

The Polymer Processing Society 23rd Annual Meeting

cludes a device for continuously providing an extruded melt strand, in the present case a high-pressure capillary rheo-

meter (Rheograph 6000, Goettfert), combined with the Rheotens-apparatus for the uniaxial elongation of the melt. Us-

ing a tubular die (diameter 2 mm, length 30 mm) and a constant piston speed of the capillary rheometer (0.3 mm/s), the

melt strand had an average die exit velocity of 10.8 mm/s (wall shear rate of 43.2 s-1

). The melt strand was subsequently

fed to the Rheotens unit and was continuously drawn down between the two counter-rotating wheels of the device,

mounted on a balanced beam. The distance between the die and the wheels was set at 96 mm. The starting velocity of

the wheels was adjusted to the die exit velocity of the melt strand. The speed of the wheels was linearly accelerated at

24 mm/s2 until rupture of the melt strand occurred. The force required to draw-down the melt strand was continuously

measured as a function of the angular speed of the wheels. The obtained force-velocity-curves were converted into

elongational viscosity-elongation rate-curves by fitting the experimental Rheotens data using a Levenberg-Marquart

routine and by applying the analytical Wagner model [11].

Compression-moulding was used to prepare specimens (20 mm x 20 mm, thickness 2 mm) for both the carbon dioxide

solubility measurements as well as for the batch foaming experiments. Batch-foaming was performed according to the

process first outlined by Suh et al. [12]. Prior to the batch foam process, the vacuum-dried specimens were saturated

with carbon dioxide in a pressure vessel at a pressure of 5 MPa and a temperature of 40 °C. A minimum of 6 days was

allowed to ensure uniform gas saturation. Following a fast pressure release (~ 25 bar/s), the samples were rapidly sub-

merged in a silicon oil bath maintained at the desired foaming temperature between 140 °C and 220 °C. Foaming times

were set at 10, 20, and 30 s, respectively; the foamed specimens were subsequently quenched in a water bath in order to

stabilize the foam morphology. The solubility of the material was analyzed by evaluation of the desorption behaviour

following saturation according to the method of Berens et al. [13].

Analysis of the foamed specimens included an investigation of the foam morphology as well as an evaluation of the

mean cell size, the cell size distribution, and the foam density. A scanning electron-microscope (SEM) (Jeol IC 848,

Leo 922), operating at acceleration voltages of 15 kV and 2 kV, respectively, was used to obtain micrographs of the

cellular morphology of the core region of the foamed specimens. For the microscopy, the foam samples were freeze-

fractured in liquid nitrogen and the fracture surfaces were sputter-coated with either gold (12 nm) or platinum (2 nm).

Furthermore, transmission electron microscopy (Zeiss 922, acceleration voltage 80 kV) was carried out, using ultra-

thin, ultra-microtomed foams and a staining technique established previously [8]. The foam density was measured by

the buoyancy method according to ISO 1183. Three specimens for each material and set of parameters were evaluated

and the average value showing a standard deviation of less than 4% is reported.

Results and Discussion

The foaming of polymers using physical blowing agents such as carbon dioxide or nitrogen relies on the same princi-

ples, regardless of the processing technique applied: (1) saturation of the polymer with the blowing agent at high pres-

sures, (2) supersaturation of the polymer-gas mixture by reducing the pressure and/ or increasing the temperature, (3)

nucleation of foam cells by the induced thermodynamic instability, (4) growth of the cells, driven by the diffusion of the

blowing agent until either cell stabilization or cell wall rupture occurs. The cell growth rate is therefore primarily con-

trolled by the diffusion rate, the rheological properties of the polymer, the temperature, the gas solubility, and the proc-

essing conditions. One of these factors, the solubility of carbon dioxide under the processing conditions selected for this

study, is summarized in Fig. 1. As can be seen, PS shows a significantly lower solubility than PPE, while the miscible

PPE/PS blends reveal an intermediate behaviour with a slight negative deviation from linearity. In a similar manner, the

solubility of the immiscible PPE/SAN blends scales with the blend composition, although the difference between neat

PPE and SAN is rather small. In all cases, the gas solubility is sufficiently high to allow foaming of the materials.

Figure 1 – Solubility of carbon dioxide in PPE/PS and PPE/SAN blends (at 40 °C and 50 bar).

The Polymer Processing Society 23rd Annual Meeting

It is well known that the cell growth behaviour as well as the final density of the polymeric foam are controlled by the

melt elongational behaviour. In the case of miscible PPE/PS blends, the elongational viscosity can be tailored by adjust-

ing the blend composition (Fig. 2a). It should be noted that both the PPE and the PS could not be measured at 220 °C

due to a too high and a too low viscosity, respectively, and only the results of PPE/PS blends are shown here. In the

case of immiscible PPE/SAN blends, the elongational viscosity of SAN at 180 °C also steadily increases with increas-

ing PPE contents up to 20 wt% - however, the physical reason is different (Fig. 2b). For the two-phase PPE/SAN blend,

the behaviour is similar to other particle-filled systems, where the PPE as ‘solid-like’ particle leads to elevated viscosi-

ties [14]. Therefore, further increasing PPE contents beyond 20 wt% lead to rapidly increasing viscosities; an effect that

is associated with the onset of percolation of the PPE phase.

Figure 2 – Elongational viscosity of PPE/PS and PPE/SAN blends. (a) PPE/PS at 240 °C. (b) PPE/SAN at 180 °C.

Similar to the rheological behaviour, the glass transition temperature of the PPE/PS blends also scales with the compo-

sition, showing values between PS (104 °C) and PPE (216 °C) [15]. As a result, a clear temperature-composition-

dependence can be obtained by evaluating the density profile of the various foams (Fig. 3a). While elevated PPE con-

tents lead to a relatively high density at low foaming temperatures, both the addition of PS as well as an increasing

foaming temperature ensure lower viscosities and, consequently, allow an enhanced cell growth and overall expansion

of the polymer [16]. Similar to the rheological properties of PPE/SAN, the density of the foamed blends remains similar

up to 20 wt%. Driven by the continuous development of a more continuous PPE phase, a further increase significantly

reduces the expandability and the attainable density reduction (Fig. 3b).

Figure 3 – Foam density of blends as a function of composition and foaming temperature [g/cm

3]. (a) PPE/PS blends.

(b) PPE/SAN blends. The foaming time was 10 sec for both blend systems.

The resulting foam morphologies of the PPE/PS blends as well as that of neat PPE foamed under identical processing

conditions are shown in Fig. 4a. As can be seen, the average cell size directly reflects the blend composition; large cells

and a pronounced density reduction are observed for PPE/PS 50/50, whereas fine-celled structures at higher foam densi-

ties are obtained when foaming blends with elevated PPE contents. The two-phase PPE/SAN blend shows a distinct

behaviour. As can be seen in Fig. 4b, the corresponding foam morphologies of SAN as well as of selected PPE/SAN

blends highlight that the presence of even small amounts of PPE significantly reduces the cell size. Simultaneously, the

cell density increases, while the density remains rather constant. Such behaviour can be attributed to heterogeneous

nucleation phenomena of the dispersed PPE phase, dominating the homogeneous nucleation at the selected experimen-

tal conditions [17]. However, steadily increasing the PPE content not necessarily improves the cellular structure, but

also deteriorates the homogeneity of the cells and, moreover, reduces the achievable density reduction. These effects

can be attributed to the increased viscosity, the increase in PPE phase size and the reduced strain hardening behaviour,

mainly driven by the presence of the dispersed, elastic PPE phase [14].

b)

The Polymer Processing Society 23rd Annual Meeting

Figure 4 – Representative SEM micrographs of the foam morphology. (a) PPE/PS blends (foaming temperature 200 °C,

foaming time 10 sec). (b) PPE/SAN blends (foaming temperature 160 °C, foaming time 10 sec).

(a) PPE/PS blends (b) PPE/PS blends

The compatibilization of PPE/SAN blends by SBM triblock terpolymers can be regarded as a promising strategy to

overcome the otherwise poor foamability at elevated PPE contents. Moreover, the toughness of the material can proba-

bly be significantly enhanced, as already demonstrated for the bulk material [18]. Following this approach leads to a

reduction of the overall PPE phase size, to a reduction the glass transition temperature of PPE, to an adjustment of the

viscosity behaviour as well as to an enhancement of the interfacial adhesion between PPE and SAN. The selective mis-

cibility of the end blocks of SBM triblock terpolymers with the blend components ensures compatibilization of the

system and, furthermore, the PS block, selectively entangled with PPE, reduces the high glass transition temperature of

PPE [8]. As demonstrated in Fig. 5, the compatibilization step clearly affects the foam morphology of a PPE/SAN 60/40

blend by 20 wt% of SBM, when compared to the uncompatilized blend. The homogeneity is strongly enhanced and the

mean cell size is further reduced to values well below 1 m; in the present case, the number-averaged cell size is in the

range of 350 nm. Moreover, the significantly improved phase adhesion leads to an incorporation of the still more dis-

persed PPE phase into the cell walls and struts, and no pronounced interfacial failure can be detected. It should be noted

that the solubility of carbon dioxide remains rather similar under the selected experimental condition, when compared to

the PPE/SAN 60/40 blend.

Figure 5 – Effect of compatibilization of PPE/SAN blends by 20 wt% of SBM triblock terpolymers (foaming time

10 sec, foaming temperature 160 °C).

Yet, TEM micrographs (Fig. 6) reveal further interesting features. The matrix of the foamed structure is formed by the

SAN (bright) containing dispersed PPE particles (dark). At the interface between PPE and SAN, the SBM can be de-

tected due to the presence of the PB middle block (black). These micrographs confirm the nanostructured blend mor-

phology forming the cell walls.

The Polymer Processing Society 23rd Annual Meeting

Figure 6 – TEM micrographs of the foam morphology of a PPE/SAN 60/40 blends compatibilised by 20 wt% of SBM

triblock terpolymers (foaming time 10 sec, foaming temperature 180 °C). (TEM micrographs showing the foam cells

(white) and the blend phases SAN (bright), PPE (dark) as well as PB of the SBM triblock terpolymers (black)).

Conclusion

This detailed investigation identified additional factors determining the foamability of multiphase blends, besides the

commonly accepted key criteria of neat polymers including sufficient gas solubility and diffusion rates, suitable proc-

essing conditions and balanced rheological properties. In the case of multiphase blends, the significant influence of the

initial blend morphology, of the rheological as well as of the thermal properties of the respective phases and the blend

needs to be taken into account. Furthermore, the compatibilization of these PPE/SAN blends by block copolymers could

be identified as a technique to establish micro- and nanocellular foam morphologies with nanostructured cell walls. In

view of these results on selected blend systems, additional criteria determining the foamability of polymer blends as

well as the morphology of the cell walls can be determined and, furthermore, new pathways for the controlled tailoring

of cellular blends are derived.

Acknowledgements

The authors thank the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) for funding of the pro-

ject A10 within the framework of the SFB 481. We further acknowledge the experimental support by Denise Danz and

Holger Schmalz (Macromolecular Chemistry II; synthesis of triblock terpolymers), by Anne Lang and Carmen Kunert

(Polymer Engineering/ SFB 481, Z2; electron microscopy), as well as by Julius Rausch and Christian Türke (Polymer

Engineering; batch foam processing). The authors also thank Dr. Jan Sandler (BASF AG, Ludwigshafen) for numerous

fruitful discussions in the field of multiphase foams.

References

1. S.T. Lee, Foam Extrusion, CRC Press, Boca Raton, 2000.

2. D. Klempner; K.C. Frisch, Handbook of Polymeric Foams and Foam Technology, Hanser, Munich, 1991.

3. R. Gendron, Thermoplastic Foam Processing. Principles and Development, CRC Press, Boca Raton, 2005.

4. M. H. Lee; C. Tzoganakis; C. B. Park Polym. Eng. Sci. 1998, 38, 1112.

5. K. Taki; K. Nitta, S. I. Kihara; M. Oshima J. Appl. Polym. Sci. 2005, 97, 1899.

6. S. Siripurapu; Y.J. Gay; J. R. Royer; J.M. DeSimone; R.J. Sponta; S.A. Khan Polymer 2002, 43, 5511.

7. D. R. Paul; C. B. Bucknall Polymer Blends, John Wiley & Sons, New York, 2000.

8. H. Ruckdäschel; J. K. W. Sandler; V. Altstädt; C. Rettig; H. Schmalz; V. Abetz; A. H. E. Müller Polymer 2006,

47, 2772.

9. H. Ruckdäschel; J. Rausch; J. K. W. Sandler; V. Altstädt; H. Schmalz; A. H. E. Müller Polym. Eng. Sci. 2006,

submitted.

10. J. Meissner Rheol. Acta 1971, 10, 230.

11. H. M. Wagner; A. Bernnat J. Rheol. 1998, 42, 917.

12. J. E. Martini-Vvendensky; N. P. Suh; F. A. Waldmann U.S. Patent No. 4 473 665, 1984.

13. A. R. Berens; G. S. Huvard; R. W. Korsmeyer; F. W. Kung J. Appl. Polym. Sci. 1992, 46, 231.

14. J. F. Le Meins; P. Moldenaers; J. Mewis Rheol. Acta 2003, 42, 184.

15. P. R. Couchmann Macromolecules 1978, 11, 1156.

16. F. Wöllecke, V. Altstädt, J. Sandler, D. Rakutt, Cell. Polym. 2002, 21, 445.

17. J. S. Colton; N. P. Suh Polym. Eng. Sci. 1987, 27, 485.

18. H. Ruckdäschel; J. K. W. Sandler; V. Altstädt; H. Schmalz; V. Abetz; A. H. E. Müller Polymer 2007, accepted.