Poly(lactic acid) foams reinforced with cellulose micro and nanofibers and foamed by chemical blowing agents

Abstract

Biodegradable polymeric foams have gained increasing attention as an alternative to conventional polymeric foams, whose recycling is economically unviable due to its low density. Based on this, this article discusses the development of poly(lactic acid) foams produced with the insertion of four and eight parts hundred resin (phr) of long and short cellulose fibers and nanofibers. Short fibers of nanocellulose were obtained by mechanical defibrillation and dried by lyophilization, and long fibers by CO₂ supercritical fluid extraction. The poly(lactic acid) foams were produced by adding a chemical blowing agent with a pressure-free expansion method. In general, short fibers of cellulose act as nucleating agents during the expansion of the foam, which is observed by its greater number of smaller-size cells than the non-reinforced poly(lactic acid) foams. The insertion of long fibers of cellulose restricts the mobility of the polymer matrix during the expansion, thus hindering the foam its growth and formation of bubbles.

Keywords

Poly(lactic acid)foam nanocellulose chemical blowing agent

Introduction

The development of sustainable green materials combining fully compostable and biodegradable matrices derived from renewable resources is more than ever in the center of industrial and public interests.¹ PLA is getting more and more attention, because it is environmentally friendly, it presents good mechanical properties and a lower price than other biodegradable materials. It is promising to produce materials that replace expanded polystyrene (EPS) and polyurethane (PU) foams, materials that have been widely used for packaging and insulation, but are generally non-biodegradable or non-recyclable.² PLA is an aliphatic polyester that can be made from various renewable resources, such as corn or potato starch, and cane or beet sugar.^3,4

PLA foams are usually prepared using a physical foaming method, such as the extrusion foaming process with physical blowing agents or a supercritical CO₂ foaming method. These methods require expensive equipment and industrial installations associated to products with a limited thickness and volume. The physic-chemical compression-molding is traditional foaming method and has the advantages of using simple equipment and being easily performed.⁴ The expansion process with chemical blowing agents (CBAs) by hot compression can be basically divided into the method with applied pressure and the one free of pressure (free expansion). In the conventional method, for the production of foam with closed cells using the applied pressure method, a given mass of sample is placed in the mold, completely filling the cavity and the hot press is closed under a constant closure pressure; the expansion occurs with the opening of the hot press mold. The free expansion process, which uses a hot press or heated furnace, consists in placing a smaller quantity of polymer in a bigger volume mold, and by the heating action, the decomposition of the blowing agent and the expansion of the polymer occur, causing the foam to fill the mold. This technique enables the control of the volume and density of the final product, as the product must acquire the shape and the volume of the mold.

To prepare stable foams by the chemical compression-molding method, usually it is necessary to increase the melt viscosity and the strength of the polymer matrix to support the action of gases from the blowing agent during the foaming.⁴ PLA rheology is not well suited for the foaming process because it has a relatively low melt strength and typically does not exhibit any strain-hardening.⁵ Owing to the low complex viscosity and melt strength of PLA, the foaming by compression-molding method can cause the rupture of the cells, promoting the coalescence of the foam. Due to this PLA complex melt viscosity, various studies have recently been conducted to improve the melt rheological properties of PLA using chain extenders,^5–7 blending others polymers^8,9 or reinforcing the matrix with a filler or fiber.¹⁰ Some studies found that the incorporation of nanofillers can assist in the control of the foam structure and increase cell density by inducing heterogeneous nucleation.¹¹

Different cell morphologies on polymeric foams can be obtained depending on the type, size, and concentration of the filler used in the formulation of polymeric foams and, in accordance with the obtained morphology, different properties can be ascribed to the foam.¹² The interest in the use of natural fibers as reinforcement elements in polymer composites has increased significantly in recent years due to their potential in replacement inorganic fibers and fillers, such as glass fiber and calcium carbonate. In comparison to inorganic fibers, lignocellulosic fibers have interesting properties, for instance, they come from renewable sources, they are biodegradable, non-toxic, and have a lower density and abrasiveness.^13–15

Cellulose nanofibers can be considered a suitable reinforcing agent for biocomposites or nanostructured foams applications, coupled with its exceptional high specific strength and modulus, low density, possibility of chemical modification, renewable nature, and relatively low cost.¹⁰

Nanotechnology applied to polymeric nanocomposites is an emerging field and has aroused interest due to the fact that a low content of nanofillers can be used to increase the mechanical properties without impairing the elastic properties (elongation at deformation) of the composites. The reasons for these improvements owes to the fact that the nanometric size of the reinforcement element presents a much larger surface area than the one conventionally used, usually in micrometer scale.¹⁶

Cellulose nanostructures can be produced in the form of whiskers (short fibers) or nanofibers (long fibers), and can they be obtained by different methods, such as chemical or mechanical processes. The drying process of the cellulose nanostructure is also an important factor in obtaining nanofibers. Since all the processes related to obtaining cellulose nanostructures are carried out in an aqueous medium by conventional drying, with solvent evaporation, it can occurs the agglomeration of the fibers due the high concentration of hydroxyl groups, that promote secondary bonds between the fibers, consequently leading to the loss of nanoscale cellulose nanofibers. Methods such as lyophilization and supercritical extraction using CO₂ have often been reported as efficient methods for drying nanofibers without a loss of the nanoscale.^17,18

Based on this, the objective of this article is to study the influence of different sizes of cellulose fibers (CFs) and nanofibers in biodegradable PLA foams, produced by an unconventional expansion method, which uses a CBA.

Materials and methods

Materials

Poly(lactic acid) (PLA), traded under the name PLA CP 5001, was provided by Cereplast Inc. It has a melt flow index (MFI) (190℃/2.16 kg) of 3 g/10 min and a density of 1.25 g.cm⁻³. The polymer was dried overnight at 65℃ under vacuum prior to all processing. Two types of CFs were used in this study: microcrystalline cellulose (MC) supplied by Sigma-Aldrich, named Sigamacell-20 (cod. S3504), with an average diameter of 20 µm (short fiber). Bleached CFs (long fiber), obtained from eucalyptus trees, provided by Celulose Riograndense S.A. (Brazil/RS). The cellulose pulp was acquired in the form of sheets and before its use, it was ground in a MARCONI knife grinder, with a Ø 1 mm sieve. The CBA used was the azodicarbonamide (ACA), activated with zinc oxide (ZnO), provided by Inbra Industria Química Ltda. The decomposition temperature of the CBA was around 190℃.

Processes to obtain the nanocellulose fibers (NCF)

MC and CF were mechanically defibrillated in a super mass colloider grinder (MKCA6-2, Masuko Sanguo). MC was ground using a suspension containing 4.5 wt% of MC in distilled water, and ground for 200 times at a rotation rate of 2500 rpm. CF grinding was performed in a solution containing water and 3 wt% cellulose. The equipment was coupled to a recirculating pump and the grinding time was 5 h at a speed of 1500 rpm. The cellulose with water passed through a static and a rotating stone in which the nanofibers that compose the fibers were individualized.

The MC suspension was dried by a lyophilization process (Liotop, L101) in which the suspension was frozen at −80℃ and lyophilized at −40℃ for 3 days at an initial pressure of 500 µHg and a final pressure of approximately 50 µHg. The resulting cellulose nanostructures (NC) were separated by grinding (Cadence, MDR301) and classified using a 400 mesh to remove any agglomerates resulted from the drying process.

The suspension drying of the NCF was performed in a supercritical extractor, with CO₂. The equipment used was supplied by Supercritical Fluids Technologies, model FTS 150, with a 1 L pressure vessel. For the drying process, first a solvent exchange solution of water/cellulose to alcohol/cellulose was performed. For this, the cellulose was filtered in vacuum, washed by acetone and alcohol several times. After that, the NCF were redispersed in an alcohol suspension with a concentration of 5% cellulose. After the solvent exchange, the alcohol/cellulose solution was stirred in a mechanical shear propeller stirrer for 1 h and sonicated for 15 min using a Sonic Vibracell equipment. The alcohol/cellulose solution was added to the pressure vessel and CO₂ was injected to a pressure of 2000 psi and at a temperature of 45℃ for 1 h. Then, the dynamic extraction valve and the feed of CO₂ were opened to promote the solvent extraction. The total extraction time was 6 h. The cylinder depressurization lasted 20 h.

Preparation of the foams

Before the mixing, all samples (PLA and CFs) were dried at 60℃ for 10 h. PLA was mixed with the cellulose nanofibers with four and eight parts per hundred of resin (phr); 4 phr of ACA and 0.25 phr of ZnO using a Haake Rheocord Torque Rheometer. The mixing temperature was set at 160℃ and the rotor speed was 100 rpm. Initially, the PLA was added to the rheometer until the plastification of the polymer (100 s) and then, the cellulose was added and mixed during 300 s and the ACA/ZnO was added and mixed. The total mixing time was 500 s.

After mixing, the samples were formed into preforms using a Schulz thermal press, at 160℃ with 5 tons of clamping force for 5 min using a 140 × 140 × 3 mm³ mold. The average mass of the preform was 74 ± 2 g. The expansion process was performed with the free pressure method, using a hot oven, at a temperature of 190℃ during 18 min. During the foaming, a thickness limiter of 100 mm was inserted to control the free expansion of all samples. Figure 1 shows the PLA foam before (preform) and after the foaming. The physical structure of the foams which were obtained had the characteristics of a semi-rigid foam. Based on its rigid structure, this material could be applied in the replacement of rigid and semi-rigid PU foams, such as protective packaging, with a short application time.

Characterizations

The apparent density of the cellular composites was obtained by the ratio of the mass (g) and volume (cm³) of the sample. The density of seven specimens of each sample was assessed according to ASTM: D1622-08. The density test was performed at a temperature of 21℃.

The MFI of the unfoamed PLA and composites was measured without adding the blowing agent. The experiment was conducted using a Dynisco Kayeness Test Systems instrument, model D4001 Hv, at a temperature of 190℃ and 2.16 kgf, according to ASTM D1238-13.

The mechanical testing by compressive tests was performed at an EMIC DL 3000 testing machine, at a compressive speed of 50 mm.min⁻¹ using test specimens with 50 × 50 × 10 mm³.

The morphology of the fractured surface of the foams was evaluated by scanning electron microscopy (FEG-MEV) using a Tescan equipment, Mira3. All samples were previously coated with Au. The observation area in all foams was analyzed in the vertical direction of the expanded sample. The software used to measure the cell size was ImageJ. The cell population density per unit volume of the foamed composites (Nf ) was determined from the SEM micrographs using equations (1) and (2).^19,20

Vf = (1 - (\frac{ρ f}{ρ p}))

(1)

Nf = (\frac{{nM}^{2}}{A})^{3 / 2} . (\frac{1}{1 - Vf})

(2)

where Vf is the void fraction (%), ρf is the density of the foam (g.cm⁻³), ρp is the unfoamed polymer density (g.cm⁻³), Nf is the cell density (cell.cm⁻³), A is the area of the micrograph (cm²), M is the magnification factor for micrograph, n is the number of bubbles in the SEM micrograph (cells). The cell size was measured by the Ferret diameter.

The thermal properties were evaluated using differential scanning calorimetry (DSC) in a SHIMADZU model DSC-60 instrument at a heating rate of 10℃.min⁻¹ and scanned from 0 to 200℃ with a 3-min isotherm.

The water and soy oil sorption test was conducted using test specimens with 25 × 25 × 10 mm³, at a temperature of 23℃, submerging the sample completely for 24, 48, and 72 h. The water/oil sorption was determined by equation (3):¹

% sorption = \frac{Mt - Mo}{Mo} \times 100

(3)

where M_t is the mass of the samples at time t, and M₀ is the mass of the samples before the immersion in water. The water absorption was plotted as a function of time. Each sample was analyzed threefold.

The glass transition (T_g), cold crystallization (T_cc), and crystalline melting (T_m) temperatures were determined. The crystallinity index (X_c) was calculated from the second heat, according to equation (4):

Xc (%) = \frac{(Δ Hm - Δ Hc)}{(W \times Δ Ho)} \times 100

(4)

where W is the X_c correction by the PLA content in the compound, ΔHm is the melting enthalpy of the sample (J/g), ΔHc is the cold crystallization enthalpy (J/g), and ΔHo is the melting enthalpy for 100% crystalline PLA, which corresponds to 93.6 J/g.²¹

The dynamic mechanical analysis was performed in a TA Instruments Q800 equipment. The storage modulus (E′), the loss modulus (E″), and the loss tangent or loss factor (tan δ) were obtained using a compression clamp geometry with a parallel plate measurement system, with test specimens of 10 × 10 × 10 mm³. The test was performed in a non-isothermal mode with the temperature ranging from 30℃ to 140℃, at a rate of 5℃/min, with a deformation of 0.1% and 1 Hz of frequency.

Results and discussion

Figure 2 shows the SEM micrographs at different magnifications of CFs (long and short fibers) before and after the mechanical defibrillation and the drying processes. In both fibers after grinding and drying, it is possible to observe a reduction in fiber size compared to the original fiber, and the presence of fibers in micro and nanometric scale is observed. The presence of fibers in the micrometer range can be a result of a partial mechanical defibrillation or agglomeration during the drying process. During the nanocellulose drying, an attraction among the CFs can occur, promoted by the hydrogen bonds due to the presence of hydroxyl groups on the cellulose surface, which promotes the formation of fiber agglomerates and the loss of the nanometric scale in some parts of the dried fiber.^17,22

Figure 3 shows the curves obtained in the torque rheometer during the mixing of fillers and additives into the polymer. The first event, in which there was a sharp increase in the torque close to 100 s, is related to PLA insertion and melting. After this event (between 100 and 200 s), it is possible to observe an increase (second event) in the torque, but less intense than the first event, which is related to the insertion of the cellulose during the mixing process. This increase in the torque is associated with an increase in the viscosity of the polymer with the insertion of fillers, being more evident and more intense in the composites made with the NFC 8. In thermoplastic composites reinforced with fibers with smaller thickness, usually have a higher viscosity level when compared to composites produced with fiber with bigger thickness, due to the increase in the contact area of the fiber–matrix interface. Nevertheless, it is necessary to consider that long fibers tend to entangled and aggregate in the polymer melt during the mixing process, which significantly hinders the mobility of the composite during processing.^23,24

In the polymer foam processing, the melt viscosity is an important parameter for the nucleation, growth, and stabilization of the cells on the foamed cellular structure.¹⁰ The MFI values, shown in Figure 3, show that the PLA MFI was higher than that reported by the manufacturer. This may be due to degradation processes resulting from the processing. In addition, the inclusion of short fibers did not seem to significantly alter the polymer flow in molten state, even with an increase in the fiber content. However, in the presence of long fibers, the MFI decreases with an increase in the fiber content and a reduction of the fiber size, most evidently in PLA/NCF 8.

Figures 4 and 5 show SEM micrographs of the PLA and PLA foams reinforced with long and short CFs, respectively, and Table 1 shows the physical and morphological properties of the foams.

Table 1.

Foam characteristics of PLA and cellulose-reinforced PLA.

Sample	Foam density (g.cm⁻³)	Average cell diameter (mm)	Number cells per micrograph	Nf Cell density (cell.cm⁻³)
PLA	0.16 ± 0.01	0.62	142 ± 15	3.06 × 10⁶
PLA/MC4	0.17 ± 0.02	0.55	121 ± 27	2.26 × 10⁶
PLA/MC8	0.15 ± 0.02	0.39	268 ± 58	8.45 × 10⁶
PLA/NC4	0.18 ± 0.01	0.66	96 ± 35	1.51 × 10⁶
PLA/NC8	0.14 ± 0.01	0.33	232 ± 78	7.29 × 10⁶
PLA/CF4	0.19 ± 0.03	0.59	132 ± 19	2.31 × 10⁶
PLA/CF8	0.21 ± 0.03	0.78	54 ± 15	0.54 × 10⁶
PLA/NCF4	0.18 ± 0.02	0.48	163 ± 30	3.47 × 10⁶
PLA/NCF8	0.25 ± 0.04	0.97	46 ± 10	0.36 × 10⁶

In general, the foams have the cell structure composed by closed cells, with the exception of the foams reinforced with 8 phr of long fibers (CF and NCF), where it is possible to observe the presence of some deformed and open cells, with the deformation resulting from the cell collapse during expansion. The foams reinforced with short CFs (MC and NC) generally had a lower cell size with a higher cell density population per unit area (Nf ) with the increasing of fiber content, which indicates a better efficiency in the cells nucleation by the presence of short fibers. The formation of cells in polymer foams is directly influenced by the characteristics of the filler or fiber used. The existence of micropores or voids in the polymer–fiber interface usually enables the migration of the gas generated by the blowing agent to these regions and thereby promotes the cell initiation and propagation from these sites.²⁵ In the foams reinforced with long fibers of cellulose (CF and NCF), this phenomenon is not observed, since the increase of fiber content promoted an increase in the size of the cell and a decrease of the Nf.

The presence of long CFs may have restricted the mobility of the PLA matrix during expansion (by the possible entanglement between the fibers), thereby inhibiting the expansion of the cells. Long CFs creates a network within the polymer matrix, and during the foaming agent decomposition, the long CFs network impedes the mobility of the polymer matrix, and therefore hinders the foam expansion and growth. The restriction of the polymeric matrix mobility, with the inclusion of long CFs, was also observed in torque rheometer as shown in Figure 3. Since CF and NCF fibers are longer than those of MC and NC, the actual surface area is much smaller, and consequently, the CF and NCF fibers have less regions that favor cell initiation. With the restriction in the mobility of the polymer matrix, imposed by the presence of a high content of long fibers of cellulose (8 phr) during the decomposition of the blowing agent, the gas which is generated rather than creating new growth sites for the formation of new cells, tends to migrate to cells that are already in formation, promoting an increase in its volume, which facilitates the bubble coalescence phenomenon, as observed in Figure 5.

The contraction of the cell arises from their coalescence, which is favored by the formation of large cells. The polymer matrix in the molten state does not support the formation of the void (air cell) and deforms. This phenomenon is more evident in foams produced with low-viscosity polymers in the molten state, with the presence of fillers and/or with the absence of crosslinking agents.

The type and size of fibers used in the composition of the PLA foams impacted on the final density of the foams. As the process used was carried out with a volume limiter, the density of the samples is conditioned to the volume of the mold. Thus, the variation in the density of the foams is due to the shrinkage of the matrix during the cooling or to the lower expansion capacity caused by the presence of long CFs. The shrinkage of the foam occurs during the cooling concomitantly with the coalescence of the cells, a phenomenon strongly influenced by the crystallization of the PLA.

The presence of short fibers (MC and NC) promoted lower density values when compared to the pure PLA foam, while the addition of long fibers (CF and NCF) promoted a higher density in the foams. All foams reinforced with CF and NCF, presented larger cells due to the coalescence and contraction of the cells, more evidently in the PLA/NCF 8 sample, in whose micrograph is possible to observe the presence of oval cells, due to the shrinkage and coalescence of the foam in the vertical direction.

The cell nucleation rate is a function of the processing condition and interfacial tension at the nucleating site (namely, the cellulose-polymer interface). Dlouhá et al. (2012) states that a small amount of well-dispersed nanoparticles may serve as preferential nucleation sites with a lower energy barrier of nucleation that facilitates the cell nucleation process. However, the nucleating efficiency of the nanoparticle surface is often compromised by the aggregation of the nanoparticles or high interactions with the polymer matrix, thus making the surface of the nanoparticles energetically unfavorable for a homogenous nucleation.²⁶ Figure 6 depicts the detail of the interior of cells in the PLA foams and the dispersion of CFs in the structure of the foams. The outline of the pure PLA foam cell wall shows a smooth texture, whereas for the samples reinforced with NC and NFC, there is a texture due to the presence of fibers, which is more prominent in the sample reinforced with NCF.

Long CFs are dispersed either on the cell walls or in the middle of the cell, thereby perforating the cell. This distribution also affects the foam expansion capacity and/or the coalescence. Short CFs are preferentially deposited on the cell wall, which favors the formation of cellular structures preferably composed of closed cells.

No significant presence of fibers in nanometer scale was observed in the samples reinforced with NC and NCF. Hence, it seems that during processing, there may have occurred an agglomeration of fibers. This could explain the similarity of the morphology of the cells when compared to the foams produced with micro particles to the foams produced with nanoparticles.

Table 2 shows the PLA foam sorption values for water and soybean oil. In general, all cellulose-reinforced foams showed increase in water sorption capacity, while the oil sorption is more evident in samples reinforced with levels of 8 phr of long CFs. Samples reinforced with long fibers of cellulose exhibited a higher sorption capacity since the formation of open cells and the larger diameter of the foams, as shown in Table 1, favors the interconnection of the cells and the passage of liquids. The increase in the water sorption capacity of reinforced-cellulose foams is also due to the physicochemical nature of cellulose, which has hydrophilic characteristics.

Table 2.

Sorption of water and soy oil of PLA foams.

Sample	Water (%)			Oil (%)
Sample	24 h	48 h	72 h	24 h	48 h	72 h
PLA	13.2 ± 2.8	16.6 ± 1.5	24.1 ± 1.0	21.1 ± 1.6	25.1 ± 1.3	30.5 ± 4.5
PLA/MC4	23.8 ± 5.0	30.7 ± 7.8	35.1 ± 1.9	23.4 ± 5.9	46.4 ± 0.9	44.2 ± 7.7
PLA/MC8	17.9 ± 2.8	24.6 ± 0.6	27.2 ± 1.3	21.6 ± 2.7	28.8 ± 3.0	31.4 ± 2.1
PLA/NC4	20.9 ± 0.2	25.9 ± 0.3	33.8 ± 4.0	21.2 ± 2.0	35.1 ± 3.1	42.3 ± 6.8
PLA/NC8	20.1 ± 1.8	23.1 ± 3.7	30.0 ± 6.5	23.1 ± 0.4	30.6 ± 2.5	35.4 ± 3.9
PLA/CF4	15.6 ± 3.9	18.7 ± 1.4	29.3 ± 3.5	25.5 ± 5.0	28.6 ± 5.8	35.2 ± 2.0
PLA/CF8	21.8 ± 4.8	28.5 ± 3.0	37.9 ± 7.1	22.8 ± 3.7	39.5 ± 5.5	49.2 ± 8.6
PLA/NCF4	19.3 ± 2.5	27.1 ± 5.6	33.4 ± 1.1	24.3 ± 1.3	27.9 ± 2.5	35.2 ± 3.7
PLA/NCF8	20.1 ± 1.1	32.6 ± 7.1	45.0 ± 6.4	30.4 ± 2.3	45.0 ± 6.4	57.3 ± 15.3

An inverse behavior of oil and water sorption of foams is also observed, with the increase of the content of short and long fibers. With the increase from 4 to 8 phr of short fibers (MC and NC) in the foam, it is observed that the water and oil sorption capacity of the foams has decreased, and the largest decrease was observed in the MC-reinforced foams. As for the foams with long fibers, the behavior is the opposite; the rise from 4 to 8 phr (CF and NCF) showed that the foams have increased the sorption of water and oil, with the greatest increase in the foams reinforced with NCF fiber. This behavior is also associated with the size of the cell. Decreasing the size of the cells (a greater nucleation) hinders the migration of fluids into the foam, while the larger cells favor this phenomenon.

Figure 7 shows the compressive stress versus the strain graphs of the pure PLA foam and PLA-reinforced foams with different types of CFs. The compressive strength behavior in polymeric foams is strongly dependent on the nature of foam (rigid or flexible), the cell opening, foam density (void fraction), the size and density of cells, and the thickness of the solid fraction in the cell borders.²⁷ The compression strength curves of all foams exhibit three well-defined regions: (I) the linear elasticity region, (II) the plateau, and (III) the densification region.²⁸ Samples reinforced with short CFs (MC and NC) show, in comparison with the pure PLA foam, that the reinforced samples with 4 phr showed superior results in the compression resistance from the elastic deformation, while the reinforced samples with 8 phr showed lower results when compared to the pure PLA foam. Samples reinforced with long CFs generally have a greater resistance to compression due to the lower degree of expansion for these samples, and, as observed, the higher density and the morphology of the cells wall contour is thicker in these samples than the pure PLA foams, as seen in Figures 4 and 5. These compressive strength behaviors reveal a tendency of decreasing the mechanical compressive strength with a decrease in the foam density and the foam cell size (increase of cell density), which is the opposite behavior of most polymeric foams. This behavior is probably attributed to the thickness of the solid fraction in the cell borders. By increasing the cell density, the thickness of the cell borders will be small, causing in a decrease in the mechanical properties of this foam.

The DSC results for pure PLA and PLA/cellulose composites without the blowing agent are given in Table 3. The thermal behavior of the thermograms of all samples were similar, and it is possible to observe the presence of the glass transition region (T_g) close to 57℃, a cold crystallization event (T_cc) between 104 and 140℃, and a melting temperature (T_m) around 157℃.

Table 3.

Thermal characteristics, on the second heat, of PLA and PLA/cellulose composites.

		Cold crystallization (T_cc)				Melting (T_m)
Sample	T_g (℃)	To	Td (℃)	Te	ΔHc (J/g)	To	Td (℃)	ΔHm (J/g)	X_c (%)
PLA	57	111	126	140	12.5	148	156	14.9	2.54
PLA/MC4	57	106	115	124	19.1	146	157	22.9	4.22
PLA/MC8	57	106	117	126	19.9	148	156	21.8	2.20
PLA/NC4	57	107	115	130	19.5	148	157	24.3	5.34
PLA/NC8	57	105	115	125	19.0	147	157	21.3	2.25
PLA/CF4	57	109	116	130	16.4	148	157	18.5	2.33
PLA/CF8	57	106	111	123	22.9	147	157	24.7	2.09
PLA/NCF4	57	106	116	126	16.7	147	158	18.8	2.34
PLA/NCF8	59	104	113	121	23.3	150	159	24.6	1.50

T_g: glass transition; To: temperature onset; Td: temperature midpoint; Te: temperature endset; ΔHc: crystallization enthalpy; ΔHm: melting enthalpy; X_c: crystallinity.

Table 3 shows that the T_g and the T_m of PLA composites with cellulose did not change significantly when compared to the pure PLA. Only a low increase in the T_m of reinforced composites with 8 phr of NCF is observed, which may result from the decreased molecular mobility of the PLA chains imposed by the presence of the fibers.

The cold crystallization temperature is commonly found in the PLA samples and is due to the low crystallization rate of PLA during the cooling, which promotes a low degree of crystallinity. Thus, during the heating, the molecules tend to organize and complete the crystallization. After the CFs are inserted, it is observed that the cold crystallization event begins early, at lower temperatures, when compared to the pure PLA. The same effect was observed by Cho et al. (2013), which assigns the presence of the cellulose as sites that facilitate the crystallization of the sample and promoting early the crystallization of the polymer. The interaction between the polymer matrix and the fillers immobilized and hindered the mobility of the PLA molecules and promoted their crystallization.¹⁰

The crystallinity of a polymeric system depends on the processing conditions and the heating–cooling cycles that the material undergoes during the processing, molding, temperature, cooling rates, nucleation density, and annealing time.²⁹ Table 3 shows that the crystallinity of the polymer is affected by the presence of CFs, most evidently in the samples reinforced with short fibers (MC and NC), where it is possible to observe an increase in the crystallinity with 4 phr of cellulose, when compared to the pure PLA. The cold crystallization and fusion enthalpy are also affected by the presence of cellulose, having an increase in both enthalpies when compared to the pure PLA.

In general, smaller particles promote an increase in crystallinity of the PLA composites when compared with the composites produced with long CFs. This phenomenon was also observed by Mathew et al. (2006) and is assigned to facility of the crystals to grow around smaller particles than on the contour of long fibers.

The melting enthalpy is associated with total fraction of crystallinity that polymers can have. This crystallinity rate has a direct influence during the PLA foaming process. Nofar and Park (2014) reported that the crystallinity of polymers during foaming could have significant effects on cell nucleation, for the cells can form both around additives (i.e. fillers) as around crystals.³⁰

Crystallization during foaming can also improve the low melt strength of PLA through the network of nucleated crystals. The ability PLA has to expand will be consequently increased by minimizing gas loss and cell coalescence. However, an excessively high crystallinity would also suppress foam expansion due to excessive stiffness in the matrix and less gas dissolution.³⁰ In this research, even with the foaming temperature being held above the PLA melting temperature, which theoretically only exists in the amorphous phase, the onset of the blowing agent decomposition occurred at lower temperatures, and impacted on the nucleation of the cells.

Figure 8 shows the heating scans, obtained by DMA, of the pure PLA foam and the PLA foams reinforced with CF. At room temperature, all samples were in glassy state. DMA is a thermal-mechanical analysis technique that measures the resistance of materials as they are deformed under periodic stress. The storage modulus (E′) relates directly to the stiffness of the sample. As per Figure 7(a) and (d), in the glassy region, the E′ of all foams reinforced with long CFs is higher than the pure PLA foam. In the foams reinforced with short CFs, the E′ in the glass transition is only higher in the samples reinforced with 4 phr of MC and NC. These results are in agreement with the results of the compression modulus obtained from the mechanical compression assays (Figure 7).

The sharp decrease in the storage modulus (around 57℃) corresponds to the α-relaxation of the amorphous regions in PLA, and is associated to the T_g. Above this transition, the E′ showed an abrupt increase attributed to the cold crystallization, as also noted earlier in the behavior of DSC thermograms. On account of the cold crystallization, PLA chains get rearranged to a particular crystalline pattern that imparts some rigidity leading to enhance the E′ of the samples.^31,32 The decrease in modulus at temperatures around 130℃ indicates the softening of the sample before the onset of melting.³³

In addition, the storage modulus increased in all samples with cellulose-reinforced foams in the rubbery region at higher temperatures, but prior to cold crystallization events, such as 80℃. This increase is more evident in foams reinforced with long fibers, and this increase is explained by the fact that the matrix becomes extremely soft in the rubbery state and the reinforcement becomes much more noticeable at high temperatures.³²

The peaks of the loss modulus curves of the reinforced samples (Figure 7(b) and (e)) generally exhibited a shift to higher temperatures, such values are directly related to the T_g of the polymer matrix, and this increase could indicates a decrease in the macromolecules mobility due to the presence of the CFs. Figure 8(c) and (f) shows the effect of the CFs on the tan delta peak. It is possible to see that the tan delta peak (α-transition) is slightly shifted to a higher temperature on PLA/NCF 8. The shift to a higher temperature usually indicates a restriction in the movement of the molecules owing to of an improved interaction in filled polymers.

Conclusions

PLA foams reinforced with different types of micro- and nano-CFs were developed by the free expansion method, using CBAs. The shapes and volumes of the production of the foam are flexible in this method, and it presents a low investment for their production. The presence of CFs changes the morphology of the cells during the expansion of the foam, and an increase in the short fiber cellulose content is observed, with a tendency to decrease the cell size and increase the cell density. On the other hand, the long CF content tends to an increase in the cell size and facilitates the phenomenon of coalescence of the cells. The apparent density of the samples was also affected by the presence of CFs, in which the foams reinforced with short fiber samples showed a lower density compared to the pure PLA foams, while the foams reinforced with long fibers promoted a higher density, which an impact on the mechanical properties of foams, increasing the mechanical property of compressive strength. The presence of long CFs promoted the perforation of the cell in the foam, causing the partial formation of open cells, which increases the liquid sorption capacity when compared to the non-reinforced PLA foam. Based on this, through a simple method of foaming and insertion of different types of cellulose, it is possible to produce different cell morphologies in PLA foams.

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: The authors would like to thank the National Research and Development Council (CNPq) and the Secretariat of Science, Innovation and Development of Rio Grande do Sul (SCT/RS) for the financial support.

References

Boissard

CIR

Bourban

P-E

Tingaut

et al.

Water of functionalized microfibrillated cellulose as foaming agent for the elaboration of poly(lactic acid) biocomposites. J Reinf Plast Compos 2011; 30: 709–719.

Neagu

Cuénoud

Berthold

et al.

The potential of wood fibers as reinforcement in cellular biopolymers. J Cell Plast 2011; 48: 71–103.

Eyholzer

Tingaut

Zimmermann

et al.

Dispersion and reinforcing potential of carboxymethylated nanofibrillated cellulose powders modified with 1-hexanol in extruded poly(lactic acid) (PLA) composites. J Polym Environ 2012; 20: 1052–1062.

Lou

Zhang

et al.

Polylactide foams prepared by a traditional chemical compression molding method. J Appl Polym Sci 2013; 130: 330–337.

Mihai

Huneault

Favis

. Rheology and extrusion foaming of chain-branched poly(lactic acid). Polym Eng Sci 2010; 50: 629–642.

Pilla

Kim

Auer

et al.

Microcellular extrusion-foaming of polylactide with chain-extender. Polym Eng Sci 2009; 49: 1653–1660.

Corre

Y-M

Maazouz

Duchet

et al.

Batch foaming of chain extended PLA with supercritical CO₂: Influence of the rheological properties and the process parameters on the cellular structure. J Supercrit Fluid 2011; 58: 177–188.

Richards

Rizvi

Chow

et al.

Biodegradable composite foams of PLA and PHBV using subcritical CO₂. J Polym Environ 2008; 16: 258–266.

Yuan

Liu

Ren

. Preparation, characterization, and foaming behavior of poly(lactic acid)/poly(butylene adipate-co-butylene terephthalate) blend. Polym Eng Sci 2009; 49: 1004–1012.

10.

Cho

Park

Yun

et al.

Influence of cellulose nanofibers on the morphology and physical properties of poly(lactic acid) foaming by supercritical carbon dioxide. Macromol Res 2013; 21: 529–533.

11.

Gao

Wang

et al.

Effect of melt viscosity on the cell morphology and properties of poly(lactic acid) foams. J Cell Plast 2015; 52: 175–187.

12.

Doroudiani

Kortschot

. Expanded wood fiber polystyrene composites: Processing – structure – mechanical properties relationships. J Thermoplast Compos 2004; 17: 13–30.

13.

Kuboki

Lee

Park

et al.

Mechanical properties and foaming behavior of cellulose fiber reinforced high-density polyethylene composites. Polym Eng Sci 2009; 29: 2179–2188.

14.

Zimmermann

MVG

Turella

Santana

RMC

et al.

The influence of wood flour particle size and content on the rheological, physical, mechanical and morphological properties of EVA/wood cellular composites. Mater Design 2014; 57: 660–666.

15.

Kuboki

. Foaming behavior of cellulose fiber-reinforced polypropylene composites in extrusion. J Cell Plast 2014; 50: 113–128.

16.

Seydibeyoglu

Oksman

. Novel nanocomposites based on polyurethane and micro fibrillated cellulose. Compos Sci Technol 2008; 68: 908–914.

17.

Peng

Gardner

Han

. Drying cellulose nanofibrils: in search of a suitable method. Cellulose 2012; 19: 91–102.

18.

Khalil

HPSA

Davoudpour

Islam

MdN

et al.

Production and modification of cellulose nanofibrillated using various mechanical processes: A review. Carbohydr Polym 2014; 99: 649–665.

19.

Petchwattana

Sirijutaratana

. Influences of particle sizes and contents of chemical blowing agents on foaming wood plastic composite prepared from poly(vinyl chloride) and rice hull. Mater Design 2011; 32: 2844–2850.

20.

Matuana

Faruk

Diaz

. Cell morphology of extrusion foamed poly(lactic acid) using endothermic chemical foaming agent. Bioresource Technol 2009; 100: 5947–5954.

21.

Lim

–

Auras

Rubino

. Processing technologies for poly(lactic acid). Prog Polym Sci 2008; 33: 820–852.

22.

Zimmermann

MVG

Borsoi

Lavoratti

et al.

Drying techniques applied to cellulose nanofibers. J Reinf Plast Compos 2016; 35: 682–697.

23.

Guo

Lee

Rizvi

et al.

Influence of wood fiber size on extrusion foaming of wood fiber/HDPE composites. J Appl Polym Sci 2008; 107: 3505–3511.

24.

Hristov

Vlachopoulos

. Effects of polymer molecular weight and filler particle size on flow behavior of wood polymer composites. Polym Compos 2008; 29: 831–839.

25.

Rizvi

Iliev

Park

. A novel system design for continuous processing of plastic/wood-fiber composite foams with improved cell morphology. J Cell Plast 2002; 38: 367–383.

26.

Dlouhá

Suryanegara

Yano

. The role of cellulose nanofibres in supercritical foaming of polylactic acid and their effect on the foam morphology. Soft Matter 2012; 8: 8704–8713.

27.

Klemmper

Frisch

. Handbook of polymeric foams and foam technology, Chapter 1. Germany: Hanser Publishers 1992.

28.

Eaves D. Foam Fundamentals. In: Eaves D (ed) Handbook of polymer foams. UK: Rapra Technology, Chapter 1, 2004, pp.1–8.

29.

Mathew

Oksman

Sain

. The effect of morphology and chemical characteristics of cellulose reinforcements on the crystallinity of polylactic acid. J Appl Polym Sci 2006; 101: 300–310.

30.

Nofar

Park

. Poly (lactic acid) foaming. Prog Polym Sci 2014; 39: 1721–1741.

31.

Yamoum

Magaraphan

. Effect of peanut shell content on mechanical, thermal, and biodegradable properties of peanut shell/polylactic acid biocomposites. Polym Compos 2015; 38: 682–690.

32.

Suryanegara

Nakagaito

Yano

. The effect of crystallization of PLA on the thermal and mechanical properties of microfibrillated cellulose-reinforced PLA composites. Compos Sci Technol 2009; 69: 1187–1192.

33.

Mofokeng

Luyt

Tábi

et al.

Comparison of injection moulded, natural fibre-reinforced composites with PP and PLA as matrices. J Thermoplast Compos 2011; 25: 927–948.