Sage Journals: Discover world-class research

Abstract

In this study, we investigated the effect of recycling process on the molecular structure, viscoelasticity and foaming behavior of low density polyethylene (LDPE). A series of LDPE samples with different recycling process was prepared by multiple extrusion using a twin-screw extruder. The molecular weight distribution (MWD) was characterized by gel permeation chromatography (GPC). Wider MWD indicated the generation of higher molecular weight products. Small-amplitude oscillation rheology showed reduced loss factors, indicating that the chain entanglement was more difficult to relax. Moreover, nonlinear viscoelasticity was investigated using elongational rheology and molecular stress function (MSF) model. The results showed a steeper strain hardening exhibited in recycled LDPE. The correlated parameter β in the MSF model indicated that the recycling did not significantly change the branches regularity in LDPE, while the increasing $f_{\max}^{2}$ , the other correlated parameter, indicated that the chain entanglement was enhanced, which was corresponded to the improvement of high molecular weight component. The foaming results revealed that the recycled LDPE had finer cellular structure and higher nucleation density. Moreover, despite adding PP and active CaCO₃ to simulate the impurities, the foamability loss of these mixed samples was well restricted and still valuable. Recycled LDPE is instead better than its corresponding virgin one in foaming performance, exhibiting the application potential for further developments.

Keywords

Low density polyethylene recycling degradation viscoelasticity foaming

Introduction

Low density polyethylene (LDPE) is one of the most produced polymers in the world, which total output of the global market reached 6.8 × 10⁷ tons in 2019.¹ It is mainly applied in the field of plastics packaging, such as packing bags, films, cushioning foams and other disposable products.^2–4 Compared to polymers in building blocks and mechanical devices, packaging materials have an average life time of 6 months,⁵ which makes them subject to only limited chemical variation at end-of-life. These wastes have a high potential for secondary utilization, but there is still a large proportion not involved in effective recycling.⁶

The environmental pollution will be reduced^7,8 and the consumption of oil resources will be saved if a suitable use can be found for the waste LDPE. Considering that the polymer foams are more profitable with the market demand having been expanding in recent years^9,10 and the stable manufacturing of LDPE foams has been well developed in industry, using recycled LDPE to replace virgin materials in foams sector is a potentially viable option. For now, most studies have reported some particular treatment options (such as filling,¹¹ capacitating,¹² etc.) for wasted LDPE to meet the specific use requirements. Only a few studies investigated the chemical foaming.^13,14 However, few literatures have reported the physical foaming of recycled LDPE. Furthermore, there were considerable differences in waste sources in previous studies, which made the experimental results difficult to summarize and promote to other solutions. The contribution of recycling needs to be further evaluated to apply recycled LDPE more universally in manufacturing foams.

Although virgin LDPE have been frequently used in foaming,^15,16 the influence of molecular structure changes from recycling still needs to be reevaluated. During recycling,^17–19 free radicals are generated when exposed to air at high temperature. The molecular topology is changed due to random free-radical collisions. Products can be divided into two categories based on the perspective of molecular weight variations: (1) free-radical reactions will be terminated after chain scission to obtain low molecular weight (LMW) components; (2) the combination of free radicals will cause grafting reaction²⁰ to form high molecular weight (HMW) components. The competitive relationship between these two reactions determines the final molecular weight distribution (MWD) and molecular topology. Thus, it is necessary to reevaluate the rheological properties and foamability after recycling.

Rheological assessment has become a common tool in the study of polymers foamability.^21–24 The viscoelasticity was used to assess the foaming process. Long branched chain (LCB) structures^25–28 are beneficial in foaming as providing more entanglement sites and can inhibit overgrowth of bubbles to regulate the cellular structure.^29,30 LDPE is an excellently expandable material,^31,32 which is also related to the existence of LCB in its molecules. The recyclability of LDPE on foaming can be highly expected. Therefore, to identify if the recycled LDPE is still suitable for foaming, the changes in molecular structure, rheological behavior and foamability need to be clarified.

In this study, recycling behavior was simulated using multiple extrusions in a twin-screw extruder and co-blended with additives to investigate the influence of recycling on LDPE foamability. First, LDPE samples with different degradation time were prepared and the molecular weight changes were measured using gel permeation chromatography (GPC). The crystallization behaviors were analyzed by differential scanning calorimetry (DSC). Second, linear and nonlinear rheology was measured to evaluate the contribution of molecular structure changes towards chain entanglement. Thirdly, foam samples were prepared using batch foaming so that the influence of chain entanglement behavior on viscoelasticity could be analyzed from the microcellular structure. At last, the essential function of chain entanglement in evaluating the foamability of recycled LDPE was clarified.

Experiments and simulations

Materials and sample preparation

LDPE (Purell 2420F, MFR = 0.75 g/10 min at 190°C) was purchased from LyondellBasell Inc., Netherlands. CO₂ (purity: 99.995 wt.%) and N₂ (purity: 99.995 wt.%) were purchased from Air Liquide Co.. LDPE samples with extrusion times of 0, 2, 4, 6, 8, 10 were prepared using a twin-screw extruder (Nanjing Giant SJ-20, L/D = 30), named as E-0 (virgin material), E-2, E-4, E-6, E-8 and E-10, with the screw zones temperatures: 160°C, 190°C, 220°C, 220°C and 220°C. The extrusion head temperature was 215°C, main screw speed 160 rpm, and feeder screw speed 16 rpm.

The recycled LDPE in industry is inevitably mixed with other materials, thus the additives existing in the system also need to be taken into account. In this experiment, PP (K8003, MFR = 2.2 g/10 min at 190°C, Sinopec Yangzi Petrochemical Co., China) and activated light CaCO₃ (1250 mesh, Suzhou Shenba Fine Chemical Co., Ltd, China) were used as impurities in the recycled LDPE. Under the same conditions, LDPE and PP were separately extruded 7 times to prepare mixed samples. LDPE and PP were mixed according to the ratio of 95/5, 90/10, 80/20 and then to extrude for the 8th time, named: LD95/P5, LD90/P10, LD80/P20, as LDPE/PP samples extruded 8 times. LDPE, PP and CaCO₃ were mixed according to the ratio of 92/5/3, 95/5/5, 85/5/10 and then to extrude for the 8th time, named: LD95/P5, LD90/P10, LD80/P20, as LDPE/PP/CaCO₃ samples extruded 8 times.

Rheological measurements and model correlation

Small-amplitude oscillation measurements were performed using a rotational rheometer (Haake Mars III, Thermo Fisher Scientific Co., Ltd) with a parallel plate geometry (35 mm in diameter, 2 mm in thickness). The linear viscoelasticity was measured at 180°C using the frequency range of 0.1 to 100 rad/s and the stress of 20 Pa. The geometric samples were prepared by compression molding at 160°C and 6 tons pressure.

Uniaxial elongational rheology measurements were performed using the same rheometer equipped with the SER universal test platform and sheet samples (20 mm × 10 mm × 1 mm) to obtain the nonlinear viscoelasticity at 180°C. Each sample was respectively subjected to four strain rates (0.1 s⁻¹, 0.3 s⁻¹, 1 s⁻¹, 3 s⁻¹). Measurements are all performed in N₂ atmosphere to prevent oxidation.

The MSF model was used to calculate the nonlinear viscosity, given by a history integral of this form:

σ (t) = \int_{- \infty}^{t} m (t - t^{'}) S_{m} (t - t^{'}) d t^{'}

(1)

where σ (t) is the additional stress tensor;

m (t - t^{'})

is a memory function related to the linear viscoelastic relaxation time spectrum

G (t - t^{'})

. Practical measurements are calculated using the relaxation modulus g_i and the relaxation time λ_i obtained from linear viscoelasticity as substitutions:

m (t - t^{'}) = \frac{d G (t - t^{'})}{d t^{'}} = \sum_{i = 1}^{n} (\frac{g_{i}}{λ_{i}}) \exp \frac{- (t - t^{'})}{λ_{i}}

(2)

The strain is calculated according to the following equation:

S_{m} (t - t^{'}) = f^{2} S_{D E}^{I A} (t - t^{'})

(3)

S_{D E}^{I A}

represents the generated strain corresponding to the extra stress tensor;

f^{2}

represents the square of relative tension in the chain segments which is associated with the free energy stored during deformation of polymers systems. For the free energy change of a polydisperse randomly branched polymer like LDPE, the equation for

f^{2}

is given by the following in uniaxial stretching experiments:

\frac{\partial f}{\partial t} = \dot{ε} \frac{β f^{2}}{1 + \frac{β - 1}{f^{4}}} (S_{11} - S_{33} - \frac{f^{2} - 1}{f_{\max}^{2} - 1} \sqrt{S_{11} + \frac{1}{2} S_{33}})

(4)

\dot{ε}

is the elongation rate in the 1-coordinate direction; S₁₁ and S₃₃ are the components of the orientation tensor

S = \frac{1}{5} S_{D E}^{I A}

in the directions parallel to stretching and perpendicular to stretching. In strain deformation, the MSF model contains two parameters, β and

f_{\max}^{2}

. For LDPE, a polydisperse randomly branched polymer, β is fitted to the steepness of strain hardening and

f_{\max}^{2}

is fitted to the maximum strain viscosity. More theoretical details about the MSF model can be found in related studies.^33–35

Molecular weight measurement

The molecular weight and molecular weight distribution (MWD) of all samples were measured using a gel permeation chromatograph (PL-GPC220, Agilent Technologics Inc., USA). The experimental temperature was 150°C and 1,2,4-trichlorobenzene was used as the eluent. The flow rate was 1 mL/min, with 0.025 wt.% BHT (2,6-di-tert-butyl-4-methylphenol, also known as butylated hydroxytoluene) added as an antioxidant. Linear polystyrene standards were used to calibrate the curve.

Thermophysical properties

A differential scanning calorimeter (DSC25, TA Instruments, USA) was used. About 10 mg of the samples were weighed and warmed up from room temperature to 180°C, to eliminate the thermal history of samples. The temperature was then cooled down to 40°C at a rate of 10°C/min, and the cooling curve was recorded and held for 5 min to ensure the temperature is stable. Then heat up to 180°C at a rate of 10°C/min, hold for 10 min and record the heating curve. Crystallization temperature (T_c), melting temperature (T_m), crystallization enthalpy (∆H_c) and melting enthalpy (∆H_m) were recorded. The crystallinity (χ_c) of samples was calculated as follows:

χ_{c} = \frac{∆ H_{m}}{∆ H_{m}^{100 %}} \times 100 %

(5)

The $∆ H_{m}^{100 %}$ of blended samples is calculated as the weighted percentage of LDPE, PP:

∆ H_{m}^{100 %} = ω_{LDPE} ∆ H_{m, LDPE}^{100 %} + ω_{PP} ∆ H_{m, PP}^{100 %}

(6)

where φ_LDPE, φ_PP are the mass fraction of LDPE, PP in samples;

∆ H_{m, LDPE}^{100 %}

∆ H_{m, PP}^{100 %}

are the fusion enthalpy of 100% crystallinity in LDPE, PP.

∆ H_{m, LDPE}^{100 %}

is 293 J/g, and

∆ H_{m, PP}^{100 %}

is 209 J/g.^36,37

Batch foaming at melt state

The foaming of LDPE was carried out in an intermittent foaming system using the rapid pressure relief method. The polymer sample (10 mm × 10 mm × 2 mm) was first placed in a high-pressure foaming kettle and purged three times with CO₂, then the system temperature was stabilized at the foaming temperature, and the high-pressure CO₂ and N₂ were pumped into the kettle. 5 MPa CO₂ + 10 MPa N₂ are used for a total saturation pressure of 15 MPa to diffuse CO₂ and N₂ gas into the melt, saturated for 60 min, rapidly drop to atmospheric pressure at a pressure drop rate of 100–150 MPa/s, and then rapidly cool down to room temperature.

Foam morphology characterization

The expansion ratio of LDPE foam (R_v) is defined as the ratio of the bulk density of the LDPE solid sample to the bulk density of the LDPE foam sample, determined by the following equation:

R_{v} = \frac{ρ_{0}}{ρ_{f}}

(7)

The density of LDPE solid samples and foam samples was determined by the drainage method according to ASTM D792-00. ρ_f was calculated from precision balance measurements with equation as follows:

ρ_{f} = \frac{a}{a - b + c} ρ_{w a t e r}

(8)

where a is the mass of the sample in air, b the mass of the weight itself measured in water by pressing the sample underwater, c the total mass of the sample and the weight measured underwater, and ρ_water the density of water at test temperature.

The foam samples were quickly quenched after being immersed in liquid nitrogen for 10 min, and the fracture surfaces were coated with Pt spray to enhance conductivity. A scanning electron microscope (SEM, Nova, NanoSEM450, FEI, U.S.) was used to observe and photograph the fracture surfaces. Image Pro Plus (Media Cybernetics, Bethesda, MD) was used for cell statistics, calculated as follows:

D_{i} = \frac{\sum d_{i} n_{i}}{n_{i}}

(9)

where d_i is the perimeter equivalent cell diameter and n_i is the number of cells with diameter d_i in SEM images. The cell density (N₀) is defined as the number of cells produced per unit volume of solid LDPE, and the equation is as follows:

N_{0} = {(\frac{n}{A})}^{\frac{3}{2}} R_{v}

(10)

where n is the number of cells in SEM photographs and A the area of SEM photographs (unit: square centimeters, cm²).

Results and discussion

Molecular weight of recycled LDPE

The molecular weight of LDPE was characterized by using gel permeation chromatography (GPC). The results of GPC are shown in Figure 1. The change of MWD illustrates the effect of simulated recycling on the structure of LDPE. The MWD curves of E-2 basically overlap with E-0. Accordingly, the molecular structure is not greatly affected by less than twice recycling. Compared with virgin material, the peak molecular weight (M_p) of multiple recycled LDPE decreased from 1.09 × 10⁵ g/mol to 8.40 × 10⁴ g/mol. The number average molecular weight (M_n) generally decreased, indicating that the total number of molecules per unit mass increased. The degradation as a whole always shows molecular chain scissions. However, it can be found that since E-4, there is not only a decrease of M_n, but also the curve is higher than E-0 on the right side of about 8 × 10⁵ g/mol, which shows an obvious HMW tail.

Figure 1.

(a) Molar weight distribution of LDPE samples; (b) Average molecular weights of LDPE samples with different extrusion times.

The calculation of the heavy average molecular weight (M_w) and the z-average molecular weight (M_z) is influenced by the molecular molar mass, with higher molecular weight components having higher statistical weights. Thus, M_w and M_z are more sensitive to higher molecular weight components. And both showed significant improvement, peaking in E-8, where M_w increases from 1.57 × 10⁵ g/mol to 2.22 × 10⁵ g/mol and M_z increases from 3.78 × 10⁵ g/mol to 8.01 × 10⁵ g/mol. In the GPC curves, it can be observed that the increased HMW tails in E-6 and E-8 compared to E-0 curves are also reduced in E-10.

The dispersion M_w/M_n increases with the recycling number from E-0 to E-8. Two types of products exist simultaneously in the free radical reaction: the LMW component and the HMW component. The MWD broadens owing to the growth of both components. The result will undoubtedly affect the motility and crystallization behavior of molecules in the melt state. Therefore, an indirect effect on the foaming process is also expected. This is due to the randomness of free-radical reactions.^19,20 In a more similar actual recycling, that is, strong shear in a twin-screw extruder, simultaneous molecular chain scissions and random connections are still observed. As the recycling times increases, the obtained HMW fraction will likewise be irreversible oxidation and molar mass reduction. Predictably, there is an optimal time interval for the increase of M_w and M_z by recycling.

Thermophysical properties of recycled LDPE

The DSC calculation results in Figure 2 are listed in Table 1. DSC experiments showed that the crystallization enthalpy of multi-extruded E-8 was increased and the crystallization temperature was slightly shifted upward compared to E-0. It indicates that the tendency of crystallization is stronger for multi-recycled LDPE.

Figure 2.

(a) Cooling curves; (b) heating curves of samples in DSC.

Table 1.

Results of DSC calculation.

Samples	T_c (°C)	∆H_c (J/g)	T_m (°C)		∆H_m (J/g)		χ_c (%)
E-0	96.0	100.1	112.3		109.2		37.3
E-8	98.5	108.5	111.6		118.9		40.6
LD95/P5	97.9	111.6	111.9^a	166.1^b	120.9^a	2.9^b	42.9
LD90/P5/C5	98.4	94.0	111.3^a	166.0^b	107.8^a	3.0^b	40.4

^aPeak of low temperature.

^bPeak of high temperature.

After PP added, no new crystallization peak was observed near the crystallization temperature of PP. It shows that the mixed system had no obvious phase separation and then separate to crystallize. However, a newly emerging peak was observed near 80°C. It is speculated that due to its low concentration and uniform dispersion in the system, PP was difficult to nucleate at its crystallization temperature. Thus PP would be induced to crystallize only when LDPE crystals provided sufficient nucleation sites.³⁸ In the secondary heating curves of LD95/P5 and LD90/P5/C5, a minor melt peak was observed near 165°C which also indicates that the residual PP crystals melted at this temperature.

Furthermore, the addition of active CaCO₃ had no significant effect on the crystallization temperature, as well as no significant change in the peak at 82°C. However, the crystallization enthalpy was further reduced. The crystallization temperature only increased by about 0.6°C, indicating that the induction of heterogeneous crystallization nucleation was also not obvious. The effect was mainly manifested by hindering the orderly collapsation of chain segments, which reduced the crystallinity under the same cooling rate.

Linear and nonlinear rheology of recycled LDPE

In this work, the viscoelasticity of recycled samples was evaluated by using small-amplitude oscillation and uniaxial elongational rheology. Figure 3 show complex viscosity (|η*|) and loss factor (tanδ) of the melts at 180°C. Easily noticed, the viscoelasticity patterns are similar for virgin and multiple extruded samples. The trends of storage modulus G′ and loss modulus G″ are consistent, with not obvious change of |η*|. LDPE is a typical pseudoplastic fluid, so each sample exhibits significant shear thinning behavior. |η*| and shear thinning behavior did not change obviously. It indicates that the heating, oxidation, and strong shear behavior in the experimental range did not lead to serious degradation. Recycled LDPE still has a considerable utilization potential.

Figure 3.

(a) Complex viscosity |η*|; (b) loss factor tanδ of multi-extrusion samples.

The samples basically maintained essentially chemically stable in this experiment. This is related to the fact that LDPE, as a polyolefin, lacks functional groups that can be easily oxidized and decomposed. Meanwhile, the tanδ gradually drops with increasing recycling times. More notable in low-frequency range, tanδ reduced from 1.5 to 1.0 at 0.01592 Hz. This indicates that the newly formed components with more complex structures have longer relaxation time. The flow properties of the material change from viscous dominant to elastic dominant. This is supported by the relaxation time spectra of multi-extruded samples, which are detailed as Tables S1–S6 of Supporting Information and further to used in the fitting of MSF model.

The nonlinear viscoelasticity of these samples was characterized by uniaxial elongational rheology experiments. As shown in Figure 4, E-0 already exhibits noticeable strain hardening. Strain hardening was further enhanced after several extrusion cycles. Hugo Rolón-Garrido et al.³⁹ observed similar phenomena in samples after heat-oxygen treatment, and our study further verified that the enhancement of strain hardening remained under conditions closer to the actual. LDPE after oxidation treatment obtained some products with complex structures. This additional fraction contributes significantly to nonlinear rheology, which offsets the tendency of overall molecular weight shift to LMW.

Figure 4.

Elongational viscosity $η_{E}^{*}$ for different strain rates of (a) E-0; (b) E-2; (c) E-4; (d) E-6; (e) E-8; (f) E-10. Solid lines represent predictions of the MSF model.

According to the results from MSF model fitting, $f_{\max}^{2}$ increases from 40 to 360 and then decreases, while β always remains 1.2. The physical meaning of $f_{\max}^{2}$ is the maximum value of stored energy when the chain segment is subjected to elastic deformation. Thus, $f_{\max}^{2}$ determines the steepness of the viscosity growth curve, which reflects the full entanglement at the same strain rate. $f_{\max}^{2}$ has the same trend as M_w and M_z, and grows substantially, all reaching their maximum at E-8. This suggests that $f_{\max}^{2}$ is also sensitive to the HMW component, as its complex structure will contribute more entanglement sites to enhance the behavior of chain entanglement and store higher energy in strain deformation. The physical meaning of β is the ratio of branched chain mass to main chain mass. In the MSF curve, β determines the early or late appearance of additional increase in elongation viscosity, which can indicate the different types of branching structures, such as the proportional relationship between various types of branched structures. Considering that LDPE randomly generates side chains and grows during the synthesis process, there are not only two types of structures, the main chain alone and branches on the main chain, but higher level branches also randomly appear on the branches, eventually forming tree-like structures. So for randomly branching polymers like LDPE, the proportional relationship between different types of branching structures (e.g., branched chains on the main chain and more advanced branched chains) affects the value of β. The β remains unchanged, it can be assumed that even though the chain segment length changest, it is still a combination of structures already present in virgin materials. The proportional relationship of each level branched structure remained the same.

Based on the variation pattern of $f_{\max}^{2}$ and β, the molecular structure change can be inferred. Due to the randomness of radical sites, most of the sufficiently large products also have similar regularity in the mass ratio of branches versus main chains, approaching the random branching proportion applied to the virgin material. Instead of demonstrating selective chain scission and extension: as assumed in the extreme case, each chain scission occurs at the branching site, resulting in a significant reduction in entanglement sites (conflicting with $f_{\max}^{2}$ growing); or chain extension occurs only at specific location, changing the proportion of branched structures (conflicting with β unchanged). Therefore, within not too many recycles, the average number of free radical reaction sites per molecule is less. The chain segments containing unaltered structures between adjacent sites are large enough. So that the distribution pattern of branches inside the chain segment can be maintained. The generated HMW fractions are also internally composed of several LMW fractions obeying similar patterns. However, when stretched, the HMW components have stronger entanglement and more sections of the whole chain segment, the degree of molecular deformation is substantial, need to store more energy. While β is essentially constant, $f_{\max}^{2}$ which reflects the stored energy in deformation enhances.

Rheological analysis was also performed for the blended systems. As shown in Figure 5, after 8 times recycling, the viscosity of PP is significantly lower than that of LDPE, while the shear thinning phenomenon is weaker. On the one hand, this is related to the structure of the virgin material PP itself. On the other hand, since the degradation of PP is dominated by chain-scission. The chain scission during degradation is severe and it is difficult to produce HMW components. Therefore, it is manifested in rheology as a simple viscosity reduction.

Figure 5.

(a) Complex viscosity (|η*|); (b) loss factor tanδ of LDPE/PP samples with 8 times extrusion.

After LDPE/PP blended extrusion, the |η*| appeared slightly elevated. It indicates that the system containing less than 10% of PP can still maintain excellent compatibility. The blended system is molecularly dispersed to have the effect of polymer alloy. Consequently, blending with lower viscosity PP, LDPE instead shows increase in |η*| and decrease in tanδ. Compared to LD90/P10, LD80/P20 showed decreased |η*|, but tanδ showed opposite patterns in high and low frequency parts. The reason for this phenomenon may be that at 20% PP addition, LDPE and PP already exhibit severe phase separation forming the PP phase with lower viscosity. Thus LD80/P20 showed decrease in |η*|. Meanwhile, because its lower viscosity, PP phase has flows more easily under small shear at high frequency. Therefore, LD80/P20 has higher tanδ than LD90/P10. In low frequency part, the presence of PP phase forms phase interface, which prevents the relaxation of LDPE chain segment, making the elasticity more prominent. Therefore, tanδ is slightly lower.

As shown in Figure 6 5% PP content was added 1250 mesh CaCO₃ for simulating the inorganic filler actually contained. The linear rheological properties of LDPE/PP system were unchanged with the addition of CaCO₃ up to 10%. Small additions of CaCO₃ had no obvious influences on linear viscoelasticity. The relaxation time spectra of blended samples demonstrate this as detailed in Tables S7–S13 in Supporting Information.

Figure 6.

(a) Complex viscosity (|η*|); (b) loss factor (tanδ) of LDPE/PP/CaCO₃ samples with 8 times extrusion.

As shown in Figure 7, the addition of PP and active CaCO₃ did not change strongly on the strain viscosity curve, indicating a continuous LDPE phase in the system at a large addition of PP and active CaCO₃. PP additions within the continuous phase, or even the formation of phase interfaces have little significant impact on entanglement sites in LDPE phase. Both $f_{\max}^{2}$ and β are insensitive to PP additions, but only a reduction of $f_{\max}^{2}$ was observed in LD85/P5/C10. It is presumed that the excessive ${CaCO}_{3}$ addition in the system plays a role in reducing the number of entanglement sites at this time. Filler particles were well dispersed in matrix. The contact space among the surrounding molecular chain segments is occupied, which reduces the probability to form intermolecular entanglement sites. Only a lower stress needs to achieve the same slippage rate. Thus, high filler concentration acts as a lubricant for the substrate. A reduced strain hardening is observed.

Figure 7.

Elongational viscosity $η_{E}^{*}$ for different strain rates of (a) LD95/P5; (b) LD90/P10; (c) LD80/P20; (d) LD92/P5/C3; (e) LD90/P5/C5; (f) LD85/P5/C10. Solid lines represent predictions of the MSF model.

Foaming behavior of recycled LDPE

As shown in Table 2, the optimal foaming temperature of E-0 was 120°C. The cell size at high temperature (125°C) was slightly improved, but the cell density was dramatically reduced, and obvious bubble aggregation and through-holes appeared in Figure 8. These defects, on the one hand, increased the escape rate of the foaming agent in the bubble hole growth and fixing steps, weakening the driving force of the foaming process; on the other hand, it would increase cell diameter and increase the open cell proportion. Therefore, the expansion ratio of E-0 at 125°C decreased clearly. The cell structure at this temperature deteriorated visibly, and the optimal temperature interval was narrow. E-8 obtained higher expansion ratio at both 115°C and 125°C. The expansion ratio and cell parameters changed smoothly with temperature, which widened the foaming temperature interval. Compared with E-0, the E-8 sample at 120°C had 28.9 times cell density and only 31.5% cell diameter. Considering the larger cell diameter and relatively high open cell proportion would cause serious loss of use properties, such as cushioning, heat insulation, and water barrier. Only after extruded, LDPE had significantly changed the cell morphology, helping to improve performance and further broaden the application area of LDPE foams.

Table 2.

Foaming results of E-0 and E-8.

Samples	Temperature (°C)	Expansion ratio	Mean diameter (μm)	Cell density (×10⁴/cm³)
E-0	115	8.2	74.5	2.10
	120	18.5	110.3	1.33
	125	11.5	111.4	0.746
	130	5.8	—	—
E-8	115	12.8	32.1	36.4
	120	17.6	34.7	38.5
	125	15.7	34.6	37.2
	130	4.9	—	—

Figure 8.

SEM images of all samples prepared at 115–125°C and LD90/P10 prepared at 120–130°C.

Crystallization and rheology work together to maintain the appropriate resistance during foaming and optimize the cell structure. After recycling, the change of molecular topology led to more pronounced elongation hardening phenomenon and also improved crystallinity. During the bubble growth period with high deformation rate, the maintenance of cells structure relies mainly on the viscosity improvement caused by strain hardening. On the one hand, it serves to inhibit the rupture of cell walls, which reduces the gas escape rate. On the other hand, the stress near the cell wall during the growth process is higher, which can reduce the threshold nucleation radius to induce cell nucleation.⁴⁰ Also the improved crystallinity will further enhance the melt strength during the cooling and molding of bubbles. In addition, the bubble wall of larger cell diameter has higher deformation degree suffering more obvious stretching effect, so the growth resistance is greater. The strain hardening phenomenon also plays a role in reducing and homogenizing the cell diameter. Arguably, for indicators of industrial interest, such as production stability and mechanical properties, recycled LDPE undergoing thermal-oxidative and mechanical degradation instead has some advantages compared with virgin materials.

Besides, the existence of impurities complicated the foaming results. As shown in Figure 9, the LDPE/PP system appeared obvious changes in expansion ratio and cell structure. Comparing LD95/P5 with E-8, the cell diameter of the sample at 120°C is enlarged, meanwhile the expansion ratio is almost constant. However, the addition of PP clearly improves the expansion ratio at 125°C. It indicates that minor PP addition is helpful to maintain the cell structure under high temperature. The upper limit of foaming temperature was therefore increased. Proper increased crystallinity helps to set the bubbles and then to rise the upper limit of foaming temperature.

Figure 9.

Expansion ratio of (a) LDPE/PP; (b) LDPE/PP/CaCO₃; mean diameter of (c) LDPE/PP; (d) LDPE/PP/CaCO₃; cell density of (e) LDPE/PP; (f) LDPE/PP/CaCO₃.

The expansion ratio curve of LD90/P10 indicates that the addition of PP also raises the lower limit of foaming temperature. Since the T_g of PP is higher than that of LDPE and they remain well compatibility at this content. While raising the crystallinity, the addition of PP will also improve the overall T_g. Accordingly, in LD90/P10, the foaming temperature shows a low temperature performance at 120°C and the optimum expansion ratio temperature moves up to 125°C. The interval is narrower compared to LD95/P5, which is negative for continuous production.

With the PP content further increased to LD80/P20, samples did not change substantially in the rheological tests, but varied enormously in the foaming experiments. The expansion ratio of LD80/P20 did not reach 2 times at all temperatures. This result has no application value for high expansion ratio materials like LDPE. LDPE can ensure some compatibility at low PP content, but as the PP content is comparable, the system without additional fillers is inevitably difficult to be gradually compatible, the system without additional fillers is inevitably difficult to be compatible. The high PP concentration leads to agglomeration and there is already obvious phase separation in the system. When the bubble is growing, the blowing agent will leak severely at the phase interface so that the driving force is insufficient to expand the melt. For the recovery of such thermodynamically incompatible systems and application in foaming industry, limiting the proportion of incompatible components, or the addition of fillers is optional, otherwise serious loss of performance is unavoidable.

The foaming results revealed that containing a small amount of active CaCO₃ had only a little changes on expansion ratio and cellular structure. The excessive addition of CaCO₃ led to the aggregation and enlarged diameter of bubbles. The expansion ratio loss was considerable, and increased open cells were noted. It suggested that the introduction of activated CaCO₃ had stronger effect on destabilizing the cell wall than nucleation in foaming. Presumably, the interface energy barrier was not high because of the excellent compatibility of active CaCO₃, thus the nucleation effectiveness was not obvious. On the one hand, the excessive filler would further enhance the melt strength of matrix, increasing the growth resistance and reducing the multiplicity. On the other hand, CaCO₃ is selected with 1250 mesh, which the diameter of many particles is in the similar magnitude with the thickness of cell walls, which can be observed in Figure 8. The stress near CaCO₃ particles in the melt will be different from other parts, which may destroy the stress uniformity of the cell wall, then affect the stability. Therefore, the percentage of waste polymer containing inorganic fillers needs to be strictly controlled in recycling, in order to ensure that the total amount of fillers does not seriously affect the quality of foam products.

Conclusion

In this study, LDPE was repeatedly extruded in a twin-screw extruder to investigate the effect of recycling on its rheological properties and foamability from the viewpoint of molecular structure change. First, in GPC, M_n reduced slightly from 1.09 × 10⁵ g/mol to 8.40 × 10⁴ g/mol, but M_w increased from 1.57 × 10⁵ g/mol to 2.22 × 10⁵ g/mol, indicating that higher molecular weight fractions were formed during recycling. Second, in viscoelastic rheology, tanδ at 0.01592 Hz decreases from 1.5 to 1.0, which shows a transition from viscous to elastic state. Besides, for the parameters calculated using MSF model in nonlinear viscoelastic rheology, β remained constant at 1.2, but $f_{\max}^{2}$ showed a substantial improvement from 40 to 360. The almost constant β represented the regularity of branched structure in LDPE was basically maintained. While the dramatically raised $f_{\max}^{2}$ and steeper strain hardening revealed that the chain segments exhibited stronger entanglement behavior when subjected to stretching. Finally, the foaming results demonstrated that the recycled LDPE exhibits optimized foamability, thus owing potential applications in foams. Furthermore, PP and activated CaCO₃ were added to simulate impurities contained in actual systems. The losses in foaming performance were acceptable with no more than 10% addition of PP or activated CaCO₃ in this study.

Supplemental Material

Supplemental Material - Effect of thermal-oxidative and mechanical degradation of recycled LDPE on foaming

Supplemental Material for Effect of thermal-oxidative and mechanical degradation of recycled LDPE on foaming by Huanyu Zou, Jiawei Lu, Pengfei Zhou and Tao Liu in Journal of Cellular Polymers.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Key Research and Development Program of China (2016YFB0302200).

ORCID iD

Tao Liu

Supplemental Material

Supplemental material for this article is available online.

References

Faust

Denifl

Hapke

. Recent advances in catalytic chemical recycling of polyolefins. ChemCatChem 2023; 15: e202300310.

Kawecki

Gonçalves

JSV

, et al. Polymer-specific dynamic probabilistic material flow analysis of seven polymers in Europe from 1950 to 2016. Resour Conserv Recycl 2021; 173: 105733.

Kawecki

Nowack

. Polymer-specific modeling of the environmental emissions of seven commodity plastics as macro- and microplastics. Environ Sci Technol 2019; 53: 9664–9676.

Kawecki

Scheeder

PRW

Nowack

. Probabilistic material flow analysis of seven commodity plastics in europe. Environ Sci Technol 2018; 52: 9874–9888.

Geyer

Jambeck

Law

. Production, use, and fate of all plastics ever made. Sci Adv 2017; 3: e1700782.

Jiang

Wang

Jiang

, et al. Assessment of plastic stocks and flows in China: 1978–2017. Resour Conserv Recycl 2020; 161: 104969.

Hidalgo-Ruz

Gutow

Thompson

, et al. Microplastics in the marine environment: a review of the methods used for identification and quantification. Environ Sci Technol 2012; 46: 3060–3075.

Auta

Emenike

Fauziah

. Distribution and importance of microplastics in the marine environment: a review of the sources, fate, effects, and potential solutions. Environ Int 2017; 102: 165–176.

Tomin

Kmetty

. Polymer foams as advanced energy absorbing materials for sports applications—a review. J Appl Polym Sci 2021; 139: 51714.

10.

Coste

Negrell

Caillol

. From gas release to foam synthesis, the second breath of blowing agents. Eur Polym J 2020; 140: 110029.

11.

Youssef

El-Gendy

Kamel

. Evaluation of corn husk fibers reinforced recycled low density polyethylene composites. Mater Chem Phys 2015; 152: 26–33.

12.

Mészáros

Fejős

Bárány

. Mechanical properties of recycled LDPE/EVA/ground tyre rubber blends: effects of EVA content and postirradiation. J Appl Polym Sci 2012; 125: 512–519.

13.

Kabamba

Rodrigue

. The effect of recycling on LDPE foamability: elongational rheology. Polym Eng Sci 2008; 48: 11–18.

14.

Greco

Maffezzoli

Manni

. Development of polymeric foams from recycled polyethylene and recycled gypsum. Polym Degrad Stabil 2005; 90: 256–263.

15.

Escudero

Solórzano

Rodriguez-Perez

, et al. Structural characterization and mechanical behaviour of LDPE structural foams. A comparison with conventional foams. Cell Polym 2009; 28: 289–302.

16.

Rostami

Azdast

Hasanzadeh

, et al. A study on fabrication of nanocomposite polyethylene foam through extrusion foaming procedure. Cell Polym 2021; 40: 231–243.

17.

Cuadri

Martín-Alfonso

. The effect of thermal and thermo-oxidative degradation conditions on rheological, chemical and thermal properties of HDPE. Polym Degrad Stabil 2017; 141: 11–18.

18.

Salvalaggio

Bagatin

Fornaroli

, et al. Multi-component analysis of low-density polyethylene oxidative degradation. Polym Degrad Stabil 2006; 91: 2775–2785.

19.

Pinheiro

Chinelatto

Canevarolo

. The role of chain scission and chain branching in high density polyethylene during thermo-mechanical degradation. Polym Degrad Stabil 2004; 86: 445–453.

20.

Jin

Gonzalez-Gutierrez

Oblak

, et al. The effect of extensive mechanical recycling on the properties of low density polyethylene. Polym Degrad Stabil 2012; 97: 2262–2272.

21.

Rainglet

Chalamet

Bounor-Legaré

, et al. Polypropylene foams under CO₂ batch conditions: from formulation and rheological modeling to cell-growth simulation. Polymer 2021; 218: 123496.

22.

Zhao

Wang

Zhang

, et al. Lightweight and strong fibrillary PTFE reinforced polypropylene composite foams fabricated by foam injection molding. Eur Polym J 2019; 119: 22–31.

23.

Wang

Zhao

Mark

, et al. Ultra-tough and super thermal-insulation nanocellular PMMA/TPU. Chem Eng J 2017; 325: 632–646.

24.

Bahreini

Aghamiri

Wilhelm

, et al. Influence of molecular structure on the foamability of polypropylene: linear and extensional rheological fingerprint. J Cell Plast 2017; 54: 515–543.

25.

Zhao

, et al. Rheological and foaming behaviors of long-chain branched polyamide 6 with controlled branch length. Polymer 2021; 224: 123730.

26.

Yao

, et al. Improvement of poly (ethylene terephthalate) melt-foamability by long-chain branching with the combination of pyromellitic dianhydride and triglycidyl isocyanurate. Ind Eng Chem Res 2019; 58: 3666–3678.

27.

Shi

Yang

, et al. A facile methodology to effectively improve the melt strength and microcellular foamability of isotactic polypropylene. J Polym Res 2020; 27: 118.

28.

Liu

Wang

, et al. Effect of chain extension on the rheological property and thermal behaviour of poly (lactic acid) foams. Cell Polym 2013; 32: 343–368.

29.

Zhao

Wang

, et al. High-expanded foams based on novel long-chain branched poly (aryl ether ketone) via ScCO₂ foaming method. Polymer 2019; 165: 124–132.

30.

Yang

Xin

Mughal

, et al. Prediction of foamability of polyethylene terephthalate using viscous and elastic parameters. Adv Polym Technol 2018; 37: 2344–2353.

31.

Zhang

Liu

, et al. Evaluating melt foamability of LLDPE/LDPE blends with high LLDPE content by bubble coalescence mechanism. Polymer 2021; 213: 123209.

32.

Park

Lee

Wang

, et al. Strategies for achieving microcellular LDPE foams in extrusion. Cell Polym 2006; 25: 1–18.

33.

Rolón-Garrido

Wagner

. Linear and non-linear rheological characterization of photo-oxidative degraded LDPE. Polym Degrad Stabil 2014; 99: 136–145.

34.

Wagner

Hepperle

Münstedt

. Relating rheology and molecular structure of model branched polystyrene melts by molecular stress function theory. J Rheol 2004; 48: 489–503.

35.

Wagner

Yamaguchi

Takahashi

. Quantitative assessment of strain hardening of low-density polyethylene melts by the molecular stress function model. J Rheol 2003; 47: 779–793.

36.

James

. Polymer data handbook. Oxford, UK: Oxford University, 1999, pp. 962–964.

37.

Arroyo

Zitzumbo

Avalos

. Composites based on PP/EPDM blends and aramid short fibres. Morphology/behaviour relationship. Polymer 2000; 41: 6351–6359.

38.

Mofokeng

Ojijo

Ray

. The influence of blend ratio on the morphology, mechanical, thermal, and rheological properties of PP/LDPE blends. Macromol Mater Eng 2016; 301: 1191–1201.

39.

Hugo Rolón-Garrido

Zatloukal

Wagner

. Increase of long-chain branching by thermo-oxidative treatment of LDPE: chromatographic, spectroscopic, and rheological evidence. J Rheol 2013; 57: 105–129.

40.

Zhao

Wang

Zhang

, et al. Nanocellular TPU composite foams achieved by stretch-assisted microcellular foaming with low-pressure gaseous CO₂ as blowing agent. J CO₂ Util 2021; 53: 101708.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.21 MB