Elsevier

Polymer

Volume 193, 10 April 2020, 122351
Polymer

Molecular origin of the foam structure in model linear and comb polystyrenes: I. Cell density

https://doi.org/10.1016/j.polymer.2020.122351Get rights and content

Highlights

  • Low cell density in commercial PS is related to formation of ionic thin film layer.

  • Presence of low molecular weight components has little effect on cell density.

  • Treatment and purification of commercial PS results in higher cell density.

  • Molecular weight and dispersity has no effect on cell density of final foam.

  • Long chain branching in synthesized model comb PS slightly increases cell density.

Abstract

The effect of the molecular architecture of anionically synthesized linear and comb atactic polystyrenes (PS), as well as lab-scale emulsion polymerized PS and commercial PS on the cell density of foam (number of cells per unit volume of neat polymer) was investigated using CO2 as foaming agent in a batch foaming setup. The effects of molecular weight, Mw, dispersity (Ð), low molecular weight components, i.e. oligomers and residual surfactant, and the number of long chain branches per molecule, Nbr, in a series of comb-PS with the same entangled backbone, Mw,bb ≈ 290 kg/mol, Zbb ≈ 20 entanglements, and similar branches, Mw,br ≈ 44 kg/mol, Zbr ≈ 3 entanglements, but different numbers of branches, 3 ≤ Nbr ≤ 190, on the cell density were investigated. Specimens for foaming have been prepared with three different methods, which resulted in non-purified, treated, and purified samples. Well-purified samples produced foams with cell density above 109 cell/cm3, while foams out of the non-purified PS had one to three orders of magnitude lower cell density. However, artificial addition of 6 wt% oligomer PS or 3 wt% surfactant to the purified samples did not reduce the cell density significantly. Treated samples prepared by only dissolving the PS in a solvent followed by removing the solvent, produced a foam with slightly lower cell density, ~5 × 108 cell/cm3, than the purified PS. Lower cell density in non-purified PS was supposed to be related to the formation of a continuous ionic thin film layer (e.g. surfactant) rather than only the presence of low molecular weight components as impurity in the polymer. In the absence of such low molecular weight components, other molecular parameters, i.e. Mw, Ð, and Nbr in these series of comb-PS had surprisingly no distinct effect on the cell density.

Introduction

From a molecular point of view, polymer characteristics e.g. chemical repeating unit, end functional groups, crystalline or amorphous structure, molecular weight, molecular weight distribution as well as the branch architecture and topology of the polymer have a distinct effect on the cell structure, i.e. cell size, cell density and volume expansion ratio (V.E.R.) of the polymer foams [[1], [2], [3]]. Higher cell density in polymer foams with the same V.E.R. results in smaller cells and therefore better mechanical and ultimately thermal properties. Nanocellular foams with cell density higher than 1013 cell/cm3 substantially improve the thermal insulation properties [4]. Costeux et al. [[5], [6], [7]] synthesized copolymers based on polymethyl methacrylate (PMMA) with different CO2-philic acrylic comonomers via free radical polymerization. They reduced the cell size to the nanoscale domain (ca. 100 nm) and enhanced the cell density to ca. 1015 - 1016 cell/cm3 of unfoamed polymer by design of the molecular structure of the copolymer matrix to achieve high CO2 solubility. Similar approaches were used by Forest et al. [4,8] through the blending of PMMA and acrylate-based copolymers, and Otsuka et al. [9] by in situ polymerization of MMA monomers in a PS matrix to produce nanocellular foam. The effect of branches in a polymer on the cell density was investigated via blending of linear and branched isotactic polypropylene (iPP) [[10], [11], [12], [13], [14]].

However, determining the exact effect of each of these structural parameters on the cell density of foams of free radical polymerized or commercial polymers is difficult due to the high dispersity, unknown branched topologies, the dependency of the crystalline structure on the branched architecture, and the presence of small amounts of additives and residual impurities which might change the nucleation density.

The morphology and size of crystalline regions in semi crystalline polymers affect the cell density of resulting foams. These crystalline domains can develop and grow during the gas impregnation stage, where saturation temperature and pressure strongly influence this phenomenon [15,16]. Chemical modification of a linear crystalline polymer [17,18] changes both the branching content and the crystalline nature of the blends. Therefore, distinguishing between the effect of branching content and crystal morphology on the final cell density of the resulting foams would be difficult. For instance, Spitael and Macosko [10] reported that a linear iPP had a higher cell density than LCB-iPP, and blends of the two resulted in a higher cell density than either linear iPP or LCB-iPP alone. Similar findings on the cell density were reported by Park et al. [19]. In contrast, Bahreini et al. [11] reported that the cell density gradually increased with increasing LCB-iPP content in its blends with a linear iPP. These different findings in the effects of branching on the observed cell densities might have resulted from the different crystalline properties [20,21] and morphology [22] of linear iPP with its larger spherulites, and LCB-iPP with its finer crystalline regions. These crystalline domains can be generated during either the saturation process or the depressurization stage [15]. However, these regions themselves cannot be foamed and might overall reduce the cell density of the final foam [23,24]. Therefore, the presence of LCB topologies in a crystalline polymer changes the cell density by changing the shape and amount of the crystalline regions rather than by their strain hardening effect [11].

Wang et al. [20] synthesized a series of comb polymers with different levels of LCB. The backbone of all combs was a random copolymer of propylene and p-(3-butenyl)toluene. The side chains were grafted from methyl groups of p-(3-butenyl)toluene via anionic polymerization of 1,3-butadiene and subsequently hydrogenated to poly (ethylene-co-1-butene) with an amorphous morphology due to the atactic branches of the copolymer. Comb series covered a wide range of branch lengths from 1.45 to 37.8 kg/mol and branch densities from 1.17 to 11.9 branches per 104 carbons in the PP backbone. The resulting branched polymer foams had a closed cell morphology with an increased cell density, whereas the linear iPP foam had almost open cell morphology. They attributed this difference to the higher melt strength, higher crystal density, and smaller crystal sizes in the comb series compared to the linear iPP. However, defining the separate effect of length and number of side chains on their foam properties was more difficult because the ratio of the semi-crystalline (backbone) to the amorphous (side chains) content was changed. On the other hand, rheological properties, especially shear and extensional viscosity, of the combs were strongly dependent on the length and number of side chains. It is clear that even for such well-defined semi-crystalline polymers, the presence of branched architectures can change the foam characteristics via both the crystallinity (heterogeneous nucleation mechanism) and the melt strength of the polymer. Therefore, it is reasonable to use amorphous monodisperse model polymers to investigate solely the effect of branching via melt stability on the foaming properties.

Atactic polystyrene (aPS) synthesized by free radical polymerization is widely used in bead foaming as well as extrusion and injection foaming processes. Arora et al. [25] controlled the cell size and foam density of PS in batch foaming with CO2 as a foaming agent, by varying the processing conditions i.e. saturation pressure, temperature, and depressurization rate. Stafford et al. [26] used anionic polymerization methods to synthesize well-characterized linear monodisperse aPS, covering a wide range of molecular weights from 6 to 1080 kg/mol. They compared foaming properties of these with commercial polydisperse PS. It was found that entangled (Mw > 25 kg/mol) monodisperse PS foams have a similar cell density (~109 cell/cm3) and cell diameter irrespective of their molecular weight, while polydisperse commercial PS produced significantly lower cell density (~107 cell/cm3) and larger cells. They concluded that the origin of this disparity was related to the presence of very low molecular weight components (oligomers with ~270 g/mol) rather than the dispersity of the molecular weight of the commercial PS. However, it was not proved that these low molecular weight components were PS oligomers or other chemical components which might remain in the commercial polymers.

Wang et al. [27] chemically added low amounts of sulfonic acid groups (–SO3H, ~1 mol%) at the para position on the phenyl ring of PS, named SPS, and then neutralized this SPS with zinc acetate to make ZnSPS. They showed that the presence of low amount of ionic groups did not change the solubility of CO2 in PS, but the diffusivity of CO2 significantly decreased in ionic sample ZnSPS compared with the pure PS. Foams of ZnSPS had one order of magnitude higher cell density than the SPS foam. They concluded that the lower diffusivity of CO2 in ZnSPS resulted in higher cell density in comparison with PS and SPS. However, the crosslinking effect of zinc acetate via zinc-sulfonic complexes in ZnSPS was not investigated, where these crosslinks, similar to LCB, might increase the strain hardening and reduce the cell coalescence.

The effect of introducing of LCB topologies on the cell density of commercial aPS was studied by Liao et al. [28] using pentaerythritol triacrylate (PETA) as a multi-functional grafting agent. Different amounts of PETA resulted in different level of LCB structures qualitatively defined through the shear and extensional rheological measurements. The foam structure of samples foamed with CO2 in a batch process revealed that increasing the LCB level resulted in a slightly higher cell density. However, the presence of residual unreacted grafting agent might have itself increased the cell density, rather than the LCB.

A question of interest is whether LCB actually increases the cell density of a branched polymer compared to its linear analog. In the current work, we try to clarify the exact effect of LCB on the cell density of foams based on amorphous homopolymer model comb systems with a controlled number of branches.

In a previous work of our group [29,30], a series of atactic comb-PS with different conformational regimes from loosely grafted comb to loosely grafted bottlebrush were synthesized using anionic polymerization method and Friedel-Crafts acetylation [31,32]. These well-characterized comb-PS were nearly monodisperse, had the same backbone (Mw,bb = 290 kg/mol) and a similar branch length (Mw,br = 44 kg/mol), but with widely varying average number of branches per backbone from Nbr = 3 to 190. In the current work, these samples were foamed using a batch foaming setup at constant saturation pressure (180 bar) and pressure release rate conditions (40 bar/s) with supercritical CO2 as the physical blowing agent. The relationship between branching structure and cell density of foamed samples could be examined without complications arising from crystallization effects on the nucleation and final foam properties. A series of linear polydisperse PS polymerized with emulsion polymerization technique as well as commercial PS were also investigated in order to evaluate the impact of other molecular parameters, e.g. dispersity, the presence of residual oligomers or surfactant in PS on the cell density of resulting foams. The effect of molecular structure especially LCB on the volume expansion ratio of foams and its correlation to shear and extensional rheological properties will be published separately in part II [46].

Section snippets

Materials

Anionically polymerized polystyrenes: Several monodisperse linear PS were anionically synthesized. The atactic comb-PS series with the same backbone and branch length, but a different number of branches were synthesized by combination of anionic polymerization and the grafting-onto method [29,[31], [32], [33]]. Fig. 1 shows the four steps for synthesis of these comb-PS. In the first step, the PS backbone was synthesized via anionic polymerization. Afterward, carbonyl groups were added randomly

The molecular origin of cell density

Foaming experiments were conducted at a saturation pressure of p = 180 bar and temperature range of 125 < T < 145 °C. According to previous studies on the sorption of supercritical CO2 in polystyrene [25], the solubility of CO2 at these conditions is ca. 9 ± 1 g CO2/100 g PS, which reduces the glass transition temperature, Tg, of PS of the order of 45–50 °C [36]. This amount of CO2 in a polymer expands the PS at standard temperature and pressure (STP 25 °C and 1 bar) to a theoretical limit of

Conclusion and outlook

A detailed analysis of the effect of molecular structures, including Mw, dispersity (Ð), low molecular weight (LMW) components, e.g. oligomers or residual surfactant, and the number of long chain branches per molecule, for a series of linear and comb polystyrenes on their foam characteristics including cell size, and cell density was investigated. Different anionically and emulsion lab-scale polymerized PS, as well as commercial PS, were used for these investigations. These samples have been

CRediT authorship contribution statement

Mahdi Abbasi: Conceptualization, Methodology, Investigation, Formal analysis, Writing - original draft, Writing - review & editing, Visualization. Lorenz Faust: Conceptualization, Methodology, Investigation, Resources, Writing - review & editing. Manfred Wilhelm: Conceptualization, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Mahdi Abbasi gratefully acknowledges the Alexander von Humboldt Foundation and the German Science Foundation, DFG WI1911/18–2, for financial support. The authors also gratefully acknowledge the DECHEMA's Max Buchner Research Foundation for research grant. The authors thank Jonas Keller and Dr. Jennifer Kübel for fruitful discussion and Dr. Michael Pollard for proofreading of this article.

References (46)

  • K. Riazi et al.

    Polystyrene comb architectures as model systems for the optimized solution electrospinning of branched polymers

    Polymer

    (2016)
  • M.A. Shafi et al.

    Initial bubble growth in polymer foam processes

    Chem. Eng. Sci.

    (1997)
  • M. Abbasi et al.

    Molecular origin of the foam structure in model linear and comb polystyrenes: II. Volume expansion ratio

    Polymer

    (2020)
  • S.-T. Lee et al.

    Foam Extrusion: Principles and Practice

    (2014)
  • S.-T. Lee et al.

    Polymeric Foams: Science and Technology

    (2006)
  • S.-T. Lee et al.

    Polymeric Foams: Mechanisms and Materials

    (2004)
  • S. Costeux et al.

    Experimental study and modeling of nanofoams formation from single phase acrylic copolymers

    J. Cell. Plast.

    (2015)
  • S. Costeux

    CO2-blown nanocellular foams

    J. Appl. Polym. Sci.

    (2014)
  • T. Otsuka et al.

    Nanocellular foams of PS/PMMA polymer blends

    Macromol. Mater. Eng.

    (2008)
  • P. Spitael et al.

    Strain hardening in polypropylenes and its role in extrusion foaming

    Polym. Eng. Sci.

    (2004)
  • E. Bahreini et al.

    Influence of molecular structure on the foamability of polypropylene: linear and extensional rheological fingerprint

    J. Cell. Plast.

    (2018)
  • G.J. Nam et al.

    Effect of long-chain branches of polypropylene on rheological properties and foam-extrusion performances

    J. Appl. Polym. Sci.

    (2005)
  • J. Stange et al.

    Effect of long-chain branching on the foaming of polypropylene with azodicarbonamide

    J. Cell. Plast.

    (2006)
  • Cited by (6)

    • Prediction of open porosity of polystyrene foams based on simulation of bubble wall rupture during bubble growth

      2022, Chemical Engineering Science
      Citation Excerpt :

      Method (iv) has the best controllability because only ρn and Er are involved and no heterogeneous system is required, unlike other methods. Numerous control techniques (Abbasi et al., 2020; Leung et al., 2007; Naguib et al., 2002; Zhai et al., 2006) and physical models (Ge et al., 2022, 2020; Li et al., 2011; Venerus, 2001) have been reported thus far. All the aforementioned strategies have been applied practically.

    • Large amplitude oscillatory shear flow: Microstructural assessment of polymeric systems

      2022, Progress in Polymer Science
      Citation Excerpt :

      In the most recent paper, Xiong and Yu [87] presented a semi-empirical constitutive equation with four parameters that are fully determined from the linear spectrum. They considered a coupled orientation-stretch conformation tensor for both backbone and side chains to predict the LAOS and extensional behavior of the entangled model comb polymers [54,75,98–104]. As it is shown in Fig. 5, this model predicts scaling laws and a plateau (maximum) behavior for Q0 alike Pom-Pom model.

    • Advance in the rheology study of comb polymers

      2022, Gao Xiao Hua Xue Gong Cheng Xue Bao/Journal of Chemical Engineering of Chinese Universities
    1

    Current address: Borealis Polyolefine GmbH, Innovation Headquarters, 4021, Linz, Austria.

    View full text