Poly(butylene succinate) bionanocomposites: a novel bio-organo-modified layered double hydroxide for superior mechanical properties

Abstract

Bionanocomposites based on poly(butylene succinate) and a novel organo-modified layered double hydroxide have been prepared by in situ polymerization. In order to enhance the compatibilization of the inorganic filler with the polymer matrix, an oligomer of PBS was intercalated between the layers of the clay. Composites with different percentages of filler (1, 3, 5, 10 wt%) and two different divalent cations (Zn or Mg) were prepared. The thermal, rheological and mechanical properties of the samples were investigated. The results showed that the materials feature a high thermal stability, more specifically so for the composites containing Zn²⁺ cations. Rheological investigations highlighted a significant chain extender effect of the filler toward the matrix, thus revealing a huge reinforcing effect imparted by the clays, especially for composites with Mg²⁺ cations. Such findings were also supported by an increase up to 30% of the tensile and flexural strength for a PBS composite loaded with 3 wt% of Mg/Al-LDH, thus suggesting a uniform level of dispersion and a pronounced interfacial interaction between the filler and polymer.

---

1 Introduction

Biopolymers are nowadays capturing the global plastic market at a growth rate of 30% annually. The definition of biopolymer is wide-ranging, but in general it concerns materials derived from nature and/or that can return to nature, as they are biodegradable.¹ Considerable efforts were carried out to improve these materials, since their technical performances are hardly comparable to fossil-derived polymers. More specifically, brittleness, low heat distortion temperature, high gas and vapour permeability and poor resistance to long term processing operations have all up to now strongly limited their applications.

Among biopolymers, poly(butylene succinate) (PBS) is very interesting. It is an aliphatic polyester generally obtained by direct polymerization of succinic acid and 1,4-butanediol, both potentially obtainable by renewable sources. The acid, for example may be prepared from biological feedstocks such as corn starch, whey, cane molasses, glycerol, lignocelluloses, cereals and bacteria.¹ On the other hand, biobased 1,4-butanediol mostly focuses on the production of biobased succinic acid, which can be subsequently converted to 1,4-butanediol through catalytic reduction. Recently Novamont and Genomatica have created the first industrial plant for 1,4-butanediol in Europe from renewable feedstocks (sugars), employing genetically engineered E. coli.² Various manufacturers are working to produce succinic acid including BioAmber, DSM-Roquette, BASF, Myriant Technologies and Mitsubishi chemicals.¹ Moreover BioAmber is working with NatureWorks and Mitsubishi chemicals to commercialize biobased PBS.¹

PBS presents properties such as biodegradability, melt processability, thermal and chemical resistance. Indeed it can be processed in the field of textiles by melt blow, multifilament, monofilament, flat and split yarn and also in the field of plastics injection molded products, therefore it is a promising polymer for various potential applications.³ However, insufficient stiffness, low melt strength and viscosity, low gas barrier properties are some of the deficiencies of PBS.⁴ These inherent shortcomings could be overcome by applying nanocomposite technology which permits to increase barrier properties, mechanical strength and heat resistance.⁵ Such systems consist of low volume (≈5%) additions of nanofillers into the polymer matrix. Examples of fillers include, but are not limited to, clays, carbon nanotubes (CNTs), graphene, silica, hydroxyapatite, silver and organic nanofillers, which in turn include completely green fillers such as nanocrystal/nanowhiskers cellulose, which have recently emerged in the field of bionanocomposite technology.⁶

In this field, layered double hydroxides (LDHs) are very interesting because of their versatility and simplicity of preparation. Moreover LDHs as synthetic layered open-framework minerals present great advantages, such as chemical purity and tunable composition, crystallinity, particle size control (i.e. form factor associated to exfoliation ability) and easy functionalization as well as a marked biocompatibility. This is largely exemplified by their role as carriers for gene, DNA or drug delivery reservoir, as well as photobiocides with porphyrins.^7–10

LDH materials, also referred as hydrotalcites (HT) or anionic clays, are described according to the ideal formula

[M^II_1−xM^III_x(OH)₂]_intra^x+[A_x/m^m−nH₂O]_inter (1)

where M^II and M^III are divalent and trivalent metallic cations, A is the anion, intra and inter denote the interlayer domain and interlayer space respectively. The structure is built of brucite like layers with edge-sharing M(OH)₆ octahedra. Partial M^II to M^III substitution induces positive charges within the layers, counterbalanced by anions.^11,12

There is almost no limitation to the nature of the anions which can be intercalated in LDHs. One of the objectives of the organic modification is to enlarge the interlayer distance between the layers so that the intercalation of large species like polymer chains becomes feasible. Suitable organic molecules can decrease the surface energy of LDHs and notably increase the thermodynamic compatibility with polymer materials.¹³ Moreover, if the molecules present functional groups able to react during the in situ intercalative polymerization, a stronger interaction between the polymer matrix and the nanoclay surface through its tethered molecules can be achieved, thus leading to better mechanical performances.¹⁴ The anions may be intercalated in the LDH through various methods widely reported in literature: coprecipitation, ion-exchange, acid-based treatment or regeneration.¹³

The novelty consists here in introducing an oligomer of PBS as organic modifier for LDH: as already mentioned, the organic nature of the interlayer guest molecule is of great importance in endowing the platelets with the suitable organophilic character and, in this case, the anion can also further act as reactive species.

With regard to the preparation of nanocomposites, many routes have been explored, such as the in situ polymerization, which involves reactions between nanoscale-modified additives and monomers, and subsequent polymerization.^15,16 Moreover, a direct comparison of melt intercalation and in situ polymerization recently reported,⁴ proved that the latter method is more effective for the production of the nanocomposite structures, probably due to a longer-lasting mixing with the monomers. Hence, the degree of dispersion of nanoparticles inside the polymeric matrix and the interfacial interaction are targeted, with a view to achieve top performances.

Another important factor is the role of the cation composition of the LDH filler platelets, Zn²⁺ or Mg²⁺, that was found not benign for polymer chains under oxidative exposure.¹⁷ A correlation between the divalent metal in aluminum-containing oleate modified LDHs on the morphology, thermal and fire properties of different polymer composites has also been provided by Manzi-Nshuti et al.¹⁸ Authors reported that the Zn²⁺ cation is more thermally stable as compared to Mg²⁺, and that there is a qualitative difference in the morphology between both LDHs as observed when dispersed into poly(ethylene) (PE) or poly(methyl methacrylate) (PMMA). ZnAl-based LDH platelets are better dispersed in PE while MgAl ones are better dispersed in PMMA. Therefore, authors concluded that ZnAl-based LDHs perform well with non-polar polymeric systems while MgAl-LDHs are more compatible with polar matrices.¹⁸

In view of such considerations and to further address the critical issues mentioned above, LDHs based on Zn/Al and Mg/Al cations have been prepared via coprecipitation with a PBS oligomer as organic modifier. Bionanocomposites with 1–3–5–10 wt% of the organo-modified hydrotalcite have been synthesized by the in situ polymerization process. Both clays and composites have been deeply studied in terms of molecular, morphological and chemical characteristics by GPC, XRD and FTIR. Thermal properties have been investigated by TGA and DSC. Viscoelastic and mechanical properties have been evaluated by rheology, DMTA and tensile measurements.

2 Experimental

2.1 Materials

1,4-Butanediol (BD), succinic acid (SA), dimethyl succinate (DMS), sodium hydroxide, aluminum nitrate Al(NO₃)₃·9H₂O, zinc nitrate Zn(NO₃)₂·6H₂O, magnesium nitrate Mg(NO₃)₂·6H₂O, acetonitrile and titanium tetrabutoxide (TBT) were purchased from Aldrich Chemical. All the materials were used as received.

2.2 LDH organo-modifier synthesis

In order to obtain a poly(butylene succinate) with low molecular weight (PBS_o, where “o” stands for oligomer), the procedure followed the first stage of a classic two-stage polycondensation process. A round-bottomed, wide-neck glass reactor (250 mL capacity) was charged with BD (40.0 g, 444 mmol), SA (52.8 g, 444 mmol) and TBT (0.0950 g, 0.279 mmol).

The reactor was closed with a three-necked flat flange lid equipped with a mechanical stirrer and a torque-meter. The system was connected to a water-cooled condenser and immersed in a thermostatic salt bath preheated at 190 °C, while the stirrer was switched on at 300 rpm. A heating band set at 110 °C was applied around the lid. After one hour at atmospheric pressure under nitrogen atmosphere, all the water distilled off and the distillate was checked by FTIR. Almost 75 g of a white solid were obtained and the product was used as LDH modifier. The molecular structure of PBS_o was confirmed by ¹H NMR and ¹³C NMR.^19a,b The spectra are reported in ESI (Fig. S1 and S2†).

¹H NMR (400 MHz, CDCl₃, δ): 1.60–1.80 (m, 4H; C^dH₂), 2.55–2.70 (s, 4H; C^bH₂), 3.65–3.70 (t, 2H; C^c′H₂), 4.06–4.16 (t, 4H; C^cH₂).

¹³C NMR (400 MHz, CDCl₃, δ): 25.20 (C^dH₂), 29.01 (C^bH₂), 64.16 (C^cH₂), 172.35 (C^aO).

2.3 LDH synthesis

LDH was prepared via coprecipitation, that is in the slow and simultaneous precipitation of aqueous solutions of both metals and anion, followed by an aging period.

More in detail, 61.0 g (355 mmol) of PBS_o were added in a 2 L three-necked, round-bottomed flask equipped with a reflux condenser and a pH-meter. In order to dissolve the oligomer, 300 mL of acetonitrile were used and the solution was stirred 30 min at 80 °C. Once the salt was dissolved, the temperature was decreased to 75 °C and the pH was raised to 10 with NaOH 1 M. At this point 57.1 g (192 mmol) of Zn(NO₃)₂·6H₂O and 32.4 g (85.3 mmol) of Al(NO₃)₃·9H₂O, previously solubilised in 200 mL of water, were added dropwise, under nitrogen atmosphere and vigorous mechanical stirring, to the solution containing the organic modifier. The pH was kept constant at 10 during the whole coprecipitation (30 min) by adding dropwise NaOH 1 M (total amount 740 mL). The final mixture was aged for almost 24 hours at 75–80 °C, under vigorous stirring in nitrogen atmosphere. The white solid was then filtered, washed with 1600 mL of water and 800 mL of acetonitrile, and then dried at 100 °C obtaining 66.0 g of finely ground white powder characterized by X-ray diffraction and FTIR analyses. Decarbonated water was used for all solutions and washings. The LDH prepared is named ZnAl-PBS_o. The synthesis of the LDH containing magnesium–aluminum as metal cations, followed the same procedure described above by replacing the zinc nitrate with magnesium nitrate Mg(NO₃)₂·6H₂O. The molar ratio Mg/Al was analogous to Zn/Al, that is 2.25/1. The sample is denoted MgAl-PBS_o.

2.4 In situ polymerization

Bionanocomposites with different amounts of the organo-modified LDHs (1, 3, 5, 10 wt%) were prepared by in situ polymerization. The samples are named PBS-ZnAl-PBS_o-X% or PBS-MgAl-PBS_o-X% where X% denotes the weight percentages of the LDH employed. As an example, the procedure for preparing PBS-ZnAl-PBS_o-3% is reported: a round-bottomed, wide-neck glass reactor (250 mL capacity) was loaded with BD (29.6 g, 328 mmol), TBT (0.0585 g, 0.172 mmol) and 1.41 g of ZnAl-PBS_o (previously dried overnight at 105 °C), corresponding to 3 wt% with respect to the polymer theoretical yield.

The reactor was closed with a three-necked flat flange lid equipped with a mechanical stirrer and a torque-meter which gave an indication of the viscosity of the reaction melt. The system was then connected to a water-cooled condenser and immersed in a thermostatic salt-bath at 190 °C, while the stirrer was switched on at 360 rpm. After one hour, the salt-bath was cooled to 180 °C and DMS (40.0 g, 274 mmol) was added to the reaction mixture; the temperature was then increased to 190 °C and kept at this value until all the methanol distilled off (1 hour approximately). The distillate recovered during the first stage in the condenser was collected and analyzed by FTIR. The lid was heated at a temperature of 110 °C with a heating band and the reactor was connected to a liquid nitrogen-cooled condenser. Dynamic vacuum was then applied in 60 minutes down to 0.1 mbar while the temperature was increased up to 230 °C. When the torque of the melt was around 7–8 mN, a highly viscous, light brown and transparent melt was discharged from the reactor. For comparison, PBS homopolymer was also synthesized without filler. The molecular structure of PBS was confirmed by ¹H NMR.

2.5 Measurements

¹H and ¹³C NMR spectra were recorded on a Varian Mercury 400 spectrometer (chemical shifts are in part per million downfield from TMS); the solvent used was CDCl₃.

Gel permeation chromatography (GPC) measurements were performed on a HP 1100 Series using a PL gel 5 μm Minimixed-C column with chloroform as eluent and to dissolve polymer samples. A refractive index detector was used and a calibration plot was constructed with polystyrene standards.

FTIR analysis was conducted over the wavenumber range of 650–4000 cm⁻¹ using a Perkin Elmer Spectrum One FTIR spectrometer equipped with a Universal ATR Sampling Accessory.

Thermogravimetric analysis (TGA) was performed in nitrogen atmosphere using a Perkin Elmer TGA4000 apparatus (gas flow 40 mL min⁻¹) at 10 °C min⁻¹ heating rate from 50 to 800 °C for the nanocomposites. The heating rate for the LDHs was 20 °C min⁻¹. The onset degradation temperatures (T_onset) were set at the intersection of the tangents between the initial and inflection points. The 5% mass loss temperatures (T_D⁵) were also measured. Moreover anion and water contents were estimated by TGA.

The calorimetric analysis (DSC) was carried out by means of a Perkin-Elmer DSC6. Measurements were performed under nitrogen flow. To erase any previous thermal history, the samples (ca. 10 mg) were first heated at 20 °C min⁻¹ to 150 °C, kept at high temperature for 5 min, and then cooled down to −60 °C at 10 °C min⁻¹. The samples were then analyzed by heating from −60 °C to 150 °C at 10 °C min⁻¹ (2nd scan). During the cooling scan the crystallization temperature (T_C) and the enthalpy of crystallization (ΔH_C) were measured. During the 2nd heating scan the glass transition temperature (T_g), the melting temperature (T_m) and the corresponding enthalpy (ΔH_m) were measured. T_g was taken as the midpoint of the heat capacity increment associated with glass-to-rubber transition.

XRD measurements in steps of 0.07° over 2θ range of 2.2–80° for clay sample and over 2θ range of 2.1–35° for nanocomposite samples, were carried out at room temperature with Bragg/Brentano diffractometers (Philips PW1710 and XPERT-PRO for clay and nanocomposite samples respectively) with Cu K_α radiation (λ = 0.154 nm, monochromatisation by primary graphite crystal) generated at 40 mA and 40 kV. A flat sample holder, 1.5 mm deep, was filled with sample powder. The crystallographic parameters a (lattice parameter) and c′ have been calculated. The brucite-like sheets were stacked one over the other with two different symmetries (rhombohedral or hexagonal). Hydrotalcite (Mg/Al) crystallizes usually in rhombohedral 3R symmetry with the unit cell being a and c = 3c′ (where c′ is the thickness of one layer constituted by a brucite-like sheet and one interlayer and is calculated from the first reflection d₀₀₃). The reflection indexed as (110) can be used for the calculation of the parameter a = 2d₁₁₀, which depends on the nature of the cation and on the value of

and can be taken as an index of the non-stoichiometry with respect to the formation of the expected theoretical HT.²⁰

Oscillatory shear measurements were carried out with an ARES rotational rheometer (Rheometric Scientific TA Instruments). Parallel plates with a diameter of 8 mm were used for frequency sweeps. The range of frequency sweeps went from 0.1 to 100 rad s⁻¹. In all cases, the strain amplitude was checked to ensure that all measurements were conducted within the linear viscoelastic region. The shear dynamic measurements were carried out at a constant temperature of 120 °C and a gap height of 1 mm. From the data, storage modulus (G′) and loss modulus (G′′) were calculated as a function of the frequencies. The curve representing the real component of the complex viscosity (η′ = G′′/ω) against the imaginary component (η′′ = G′/ω) is described by the Cole–Cole plot.

Physical and mechanical properties were determined using a Rheometric Scientific DMTA IV Dynamic Mechanic Thermo analysis instrument with a dual cantilever testing geometry. Typical test samples were bars obtained by injection moulding at 140 °C using a Minimix Molder. The analysis was carried out at a frequency of 3 Hz, at 3 °C min⁻¹ in the −150 °C to 80 °C temperature range.

Tensile and flexural testing was performed at room temperature at a constant rate of 10 mm min⁻¹ and 1 mm min⁻¹ respectively, using a INSTRON 5966 series test instrument, equipped with a 10 kN load cell. Five rectangular strip specimens (100 × 10 × 3 mm) were tested for each sample.

3 Results and discussion

3.1 Organo-modified LDHs

3.1.1 Morphological characteristics. In order to obtain a stronger interaction between the PBS polymer matrix (see the molecular structure in Fig. 1) and the nanoclay surface, a poly(butylene succinate) oligomer (PBS_o) was synthesized and subsequently intercalated between the LDHs lamellae. The number-average molecular weight M_n of PBS_o, resulting from ¹H NMR, and confirmed also by GPC, is almost 955 (M_w/M_n = 4.1).

Fig. 1 Molecular structure of PBS.

In Fig. 2 the FTIR spectrum of MgAl-PBS_o, is compared to Mg/Al HT containing carbonate (MgAl-CO₃²⁻) and PBS oligomer spectra. First of all, the characteristic absorption band of carbonate at 1360 cm⁻¹ is clearly visible in the MgAl-CO₃²⁻ curve. The main vibration bands of PBS_o are the aliphatic CH₂ stretching mode at 2940 cm⁻¹, the C=O stretching vibration at 1713 cm⁻¹, the carboxyl acid C–O stretching at 1330–1312 cm⁻¹ and the C–C–O stretching bands at 1046 and 1157 cm⁻¹.^21,22 All these bands also appear in the spectrum of MgAl-PBS_o, with the addition of a broad band at 3441 cm⁻¹, attributed to the OH stretching vibration of water molecules and hydroxyl groups that belong to brucite layers. Moreover the carboxylic asymmetric and symmetric stretching (Fig. 2b), typical of the carboxylate ion, are visible at 1554 cm⁻¹ and 1387 cm⁻¹ respectively. Comparing the profiles, the oligomer seems to be present at the vicinity of the brucite-type layers.

Fig. 2 FTIR spectra of (a) MgAl-CO₃²⁻, PBS_o and MgAl-PBS_o; (b) zoom region 1900–650 cm⁻¹.

The XRD patterns of the clays prepared are shown in Fig. 3, while the crystallographic data are summarized in Table 1. The patterns are characteristic of organo-LDHs with basal reflections (00l) at low angles of 2θ, due to the layered structure, and with the “saw-toothed” reflections (110 and 113) at higher angles of 2θ, indicative of the intra-layer structural ordering.²³

Fig. 3 XRD profiles of (a) LDHs and (b) comparison between the homopolymer PBS, the oligomer PBS_o and MgAl-PBS_o.

Table 1 Chemical composition and cell parameters of LDHs

Code	Experimental chemical composition^a	Interlayer distance c′ = d003^b (Å)	c = 3c′ (Å)	d110^c (Å)	a = 2d110^d (Å)	M^II/M^III ratio^e
a Anion and water content determined by TGA.b Determined by the 003 reflection in XRD analysis.c Determined by the 110 reflection in XRD analysis.d Lattice parameter a related to M^II/M^III ratio and ionic radius r of cations.e Determined on the basis of the metal cations charged during the clay syntheses.
MgAl-CO3^2−	[Mg0.75Al0.25(OH)2](CO3^2−)0.1250.48H2O	7.9	23.7	1.54	3.08	3.00
ZnAl-PBSo	[Zn0.69Al0.31(OH)2](PBSo^−)inter≈0.31(PBSo^−) ads≈0.232.29H2O	12.0	36.0	1.53	3.06	2.25
MgAl-PBSo	[Mg0.69Al0.31(OH)2](PBSo^−)inter≈0.31(PBSo^−)ads≈0.441.55H2O	12.0	36.0	1.51	3.02	2.25

---

ZnAl-PBS_o shows a narrower diffraction peak than MgAl-PBS_o indicating a higher crystallinity for the former.²⁴ The position of the reflection (110) provides information on the cell parameter a = 2d₁₁₀ which could indicate any deviation for the non-stoichiometry with respect to the formation of the theoretical HT:²⁰ the experimental results obtained are 3.06 Å for Zn/Al and 3.02 Å for Mg/Al, consistent with the fact that Zn²⁺ has an ionic radius higher (0.74 Å) than Mg²⁺ (0.65 Å) and that the cation ratio M^II/M^III is close to 2.

The interlayer distance (c′), extrapolated from the first reflection peak d₀₀₃, positioned at 2θ = 11.6° in the case of carbonate anion and 7.4° for modified PBS_o LDHs, corresponds to 7.9 Å and 12.0 Å spacing, respectively. As expected, the organic modification of the clay resulted in an increase in the basal spacing, causing a large shift in the diffraction peak towards lower values of 2θ. Such common topotactic reaction involving the sheets stacking expansion only is explained by a swelling process due to the steric volume occupied by the anion.⁴

An interlayer distance of 12.0 Å seems however not consistent with the oligomer incorporation in association with its long aliphatic chain. A much larger interlamellar spacing should have been expected in that case.

One should also note an additional response at low angle for ZnAl-PBS_o at 2θ = 3.7° (basal spacing of 23.7 Å i.e. twice the high intensity peak). This peak is not correlated to the other (this is clearly evident on the third possible contribution expected at 2θ = 11.2° (7.9 Å) while it is located at 14.6°), and therefore suggests a mixture of interleaved.

On the other hand, the reflections into the range 17° < 2θ < 30° are typical of the polymer matrix (Fig. 3b), revealing the presence of the oligomeric chains.

These apparently contradictory results are probably due to the fact that the coprecipitation procedure, occurring in a water/acetonitrile basic medium (pH = 10), most probably causes the hydrolysis of PBS oligomer chains, causing the breakage of some ester linkages and also involving the formation of the starting monomers, succinic acid (SA) and butanediol (BD), in the reaction medium. Therefore such molecules could compete to spread into LDH interlayer gap. The size of some products of PBS_o hydrolysis such as succinic acid is consistent with the observed basal spacing.²⁵ Their presence, however, cannot explain the observed diffraction lines typical of PBS_o. The latter case suggests the presence of PBS_o wrapping the formed inorganic platelets. Such an hypothesis could also explain the slight difference in the diffraction line positions for PBS_o (Fig. 3b), visualized by a small shift of their position into LDH. The observation of diffraction lines at 2θ = 19.5, 21.9, and 22.6° could be assigned to (020), (021), and (110) planes of PBS structure, respectively.^26,27

In order to better understand the composition of the intercalating agent, FTIR analyses of the samples prepared in this study were compared with the starting monomers (SA and BD) and an LDH coprecipitated with sodium succinate (MgAl-SU) (ESI, Fig. S3†). Moreover the XRD patterns of MgAl-SU, MgAl-PBS_o and PBS have been examined (ESI, Fig. S5†), with the addition of ¹H NMR analyses of the product resulting from the ageing of PBS_o, subjected to the coprecipitation procedure for LDH synthesis but without the presence of the metal cations (ESI, Fig. S4†). The FTIR analysis suggests the presence of succinate, the XRD investigation confirms the existence of the oligomeric chains and the ¹H NMR spectrometry shows the products of PBS_o hydrolysis: BD, SU and condensation products such as bis(4-hydroxydibutyl)succinate (BSB).

In brief, all such measurements suggest that the interlayer distance, obtained for the organo-modified layered double hydroxide might derive from more than one species, all simultaneously present in the nanofiller system, and identified as butanediol, succinate, BSB and some oligomeric PBS chains, once again the most probable being interleaved succinic acid having in mind that a dianion known to behave as a structuring/pillaring agent may explain the observed extended coherence domains. Hence, the intercalation reaction provides a mixed organo-modifier system, whose possible compositions are reported in Fig. 4.

Fig. 4 Molecular structures of the possible species simultaneously present in the nanofiller system: dimer of PBS (as example of oligomeric chains), bis(4-hydroxybutyl)succinate (BSB) succinic acid and butanediol.

Moreover, in order to confirm that the oligomer is not just adsorbed on the surface of LDH, a small part of LDH-PBS_o was washed with CHCl₃, in order to completely eliminate the oligomer simply physisorbed, and the sample was characterized by XRD (ESI, Fig. S7†) and TGA. The XRD profile clearly demonstrates that the oligomer is not simply physisorbed, because of the persistence of the characteristic reflections of the polymer matrix. Such persistence suggests that PBS_o is tethered to the surface of the particles.

The amount of physisorbed oligomer was quantified by TGA (ESI†) and considering the difference in residual mass loss between pristine and washed samples is almost 15–20% by weight. In any case the presence of oligomer adsorbed on the surface of LDH is a key factor in the preparation of the composites since it should improve the interfacial compatibilization with the polymer chains.

3.1.2 Thermal properties. Fig. 5a and b show the thermogravimetric and the relative derivative weight loss plots of the LDHs prepared together with PBS_o. As can be seen MgAl-CO₃²⁻ presents a low temperature decomposition step below 230 °C and a high temperature decomposition step with minimum peaks around 370 and 490 °C. The thermal behavior of the LDH phase involves a two-stage decomposition process: a low temperature dehydration stage (up to about 290 °C for Al-containing LDHs) due to the loss of physisorbed and interlayer water without collapse of the structure and a high temperature decomposition stage (≈300–500 °C) due to the loss of hydroxyl groups from brucite-like layer and to the decomposition of anions. Both the first and the second transition can occur in two stages.^20,28

Fig. 5 (a) TGA and (b) DTGA thermograms of LDHs and PBS_o.

In the case of the organo-modified clays, the presence of the organic anion determines an improvement in thermal stability respect to MgAl-CO₃²⁻, as highlighted by T_D⁵ values (Table 2), and changes the thermal decomposition behavior, which proceeds via three or four steps. It is quite difficult to calculate the exact anion content into the LDH, because of the overlapping with the transition corresponding to hydroxyl groups from brucite-like layer. In any case, taking into account the transition above 280 °C, the anion content is estimated to be lower for Zn/Al than for Mg/Al (Table 1) and this is also reflected in a higher T_D⁵ value of the latter LDH (Table 2). In this regard, Conterosito et al.,²⁹ preparing LDHs intercalated with bioactive molecules, reported that the drug loading is lower in the Zn/Al samples than in the corresponding Mg/Al samples and they ascribed such behavior to the different Al ratios in the two compounds. In this study the sample Mg/Al results fairly Al-rich with respect to Zn/Al because the lattice parameter a is slightly lower (Table 1) but more probably the reason why in our case the organic loading is higher for the Mg/Al sample should be ascribable to a grafting reaction occurred onto the surface of the LDH. In fact, the –OH groups present on the clay surface can be involved in condensation reactions. By observing the thermal decomposition profiles of the two modified clays (Fig. 5b), a significant difference is evident because the mass loss in the range ≈ 190–290 °C is noticeable in the Zn/Al sample while it is just a shoulder in Mg/Al one. Tao et al.³⁰ reported that the dehydroxylation stage begins at almost 210 °C and ends at 400 °C, therefore in our case the mass loss, in the range 190–290 °C, might also be due to dehydroxylation and could be taken as an indication of –OH loss. The lower content of –OH groups in Mg/Al can indicate the consumption of hydroxyl functionalities when the oligomeric chains, –COOH terminated, condensate with the LDH surface to form a covalent bond. The presence of the oligomer grafted on the surface of LDH is obviously an important added value of this organo-modified clay, as it can notably improve the compatibilization with the polymer matrix.

Table 2 TGA results of LDHs and PBS_o, under nitrogen flow

Code	Residual mass (wt%)	TD^5^a (°C)
a Temperature at which samples lose 5 wt%.
MgAl-CO3^2−	54	119
ZnAl-PBSo	36	196
MgAl-PBSo	29	222
PBSo	0	251

---

The residual mass losses, reported in Table 2, confirmed what stated above, because the values are 36 and 29% for ZnAl-PBS_o and MgAl-PBS_o respectively. This is consistent with the data reported by Manzi-Nshuti et al.¹⁸ Nevertheless, the thermal stabilizing effect of Zn respect to Mg, reported by the same authors,¹⁸ is not evident from these samples. This could be ascribed to the greater hydration of Zn/Al (2.29H₂O) with respect to Mg/Al (1.55H₂O), calculated by TGA considering the weight loss up to 280 °C, which can cause an acceleration of the chains hydrolysis (Table 1).

Finally, the XRD analysis too (Fig. 3a), supports the hypothesis of a grafting reaction in LDH containing Mg/Al by observing the diffractograms of the two modified clays. The relative intensity of the reflections due to the polymer (17° < 2θ < 30°) in comparison with the first reflection (2θ = 7.9°), results to be very high in the case of Mg/Al. This can indicate that the oligomeric chains are grafted onto the clay surface, but also that a crystallizing effect in the vicinity of the inorganic platelets is occurring.

3.2 Nanocomposites by in situ polymerization

3.2.1 Molecular and morphological properties. All the nanocomposites have been prepared by in situ polymerization, with different amounts of LDH (1–3–5–10 wt%). Table 3 reports the GPC results: the molecular weights are high, into the range 50 000 < M_w < 100 000 for the composites containing Mg/Al and 70 000 < M_w < 120 000 for the samples prepared with Zn/Al.

Table 3 GPC and TGA results of PBS and its nanocomposites

Sample	Mw × 10^−3^a	Mw/Mn^a	Tonset^b (°C)	TD^5^b (°C)	Residual mass^b (%)
a Determined by GPC in CHCl3.b Determined by TGA under nitrogen flow.
PBS	96	2.4	379	331	0.0
PBS-ZnAl-PBSo-1%	124	2.3	366	328	0.0
PBS-ZnAl-PBSo-3%	88	1.9	365	323	0.5
PBS-ZnAl-PBSo-5%	70	2.2	353	315	0.4
PBS-ZnAl-PBSo-10%	68	2.2	362	320	3.4
PBS-MgAl-PBSo-1%	94	2.3	368	330	0.0
PBS-MgAl-PBSo-3%	106	2.4	363	320	0.5
PBS-MgAl-PBSo-5%	78	2.3	340	307	0.6
PBS-MgAl-PBSo-10%	46	2.1	345	302	2.7

---

The degree of dispersion of the filler in PBS matrix has been evaluated by XRD diffraction. The results obtained are quite similar for all the samples, thus only the Zn/Al composites have been reported in Fig. 6, as representative patterns. Reflections related to the polymer are present into the region 2θ = 17–30° and the absence of the diffraction lines at lower angles, related to the pristine clay, suggests that an exfoliation process had probably occurred. This was also checked on 10 wt% composite.

Fig. 6 XRD profiles of Zn/Al nanocomposites and PBS.

3.2.2 Thermal properties. The thermal stability of the PBS/clay nanocomposites has been studied through thermogravimetric analysis. The weight loss due to the formation of volatile products after degradation at high temperatures, in an inert gas atmosphere, was monitored as a function of temperature and is presented in Fig. 7, while the TGA data are reported in Table 3. First of all, samples have similar profiles and lose weight in a single main decomposition process into the 280–420 °C range. The onset degradation temperature of the composites, as measured from the intersection of the tangent of the initial point and the inflection point, are located at lower values with respect to the homopolymer. The incorporation of organo-modified LDHs does not improve the thermal stability of the composites, but rather induces the opposite. This could be due to the catalyzing effect of hydroxide/metal cations in PBS matrix, thus inter–intra-molecular transesterifications of the polyester can occur. In addition the water molecules released during decomposition accelerate chain hydrolysis.^4,31 T_onset values of PBS-ZnAl-PBS_o (3–5–10 wt%), reported in Table 3, are higher and closer to each other as compared to PBS-MgAl-PBS_o (3–5–10 wt%), therefore in composites the thermal stabilizing effect of Zn²⁺ cations, which slow down the hydrolytic processes occurring during heating, is evident. Though there is no regular trend, proportional to the increase of the clay content, it can be observed that both types of composites (Zn/Al and Mg/Al) present the lowest thermal stability with 5 wt% of clay content. Higher loading, as 10 wt%, seems to shift the degradation to higher temperature, thus drawing a U-shape trend of T_onset as a function of the loading percentage. For the latter, greater than 5%, an extended ceramic layer, most probably thermally insulating, was formed during the thermal event, thus limiting the diffusion into the composite. For lower loading the effect is too small (below some sort of threshold) leaving the composite with no efficient barrier effect and with catalyzing degradation phenomena being predominant. The data of T_D⁵, defined as the temperature at which 5 wt% loss weight occurs, exhibit behaviors similar to T_onset. Finally all the residues are coherent with the theoretical filler content.

Fig. 7 TGA thermograms of (a) Zn/Al, (b) Mg/Al nanocomposites and PBS.

DSC experiments on PBS and nanocomposites were carried out to understand the effect of LDH on the thermal properties. The melting temperature of PBS is 116 °C and those of composites, reported in Table 4, remain fairly constant at around 114 °C. Moreover all the melting curves are similar to those of PBS. In all cases the presence of a little melting peak just before the mean melting can be observed. Phua et al.³² have ascribed the presence of this multiple melting to different types of crystalline lamellae in PBS, which correspond to two types of crystallites of closely related structures. Regarding the glass transition temperatures the values appear almost similar but a more in-depth discussion will be developed in the next paragraph, when DMTA results will be presented. Fig. 8 illustrates the DSC cooling profiles from melt for all samples. Even if there is no regular trend in T_C, the crystallization enthalpy ΔH_C tends to decrease. Therefore the filler seems to hinder the crystallization process. Furthermore the peaks become more and more narrow indicating a faster rate of crystallization, which could depend on the nucleating effect of the LDH platelets.

Table 4 DSC results of PBS and its nanocomposites

Sample	Tg^a (°C)	TC^b (°C)	ΔHC^b (J g^−1)	Tm^a (°C)	ΔHm^a (J g^−1)
a Determined by DSC during the 2nd heating scan.b Determined by DSC during the cooling scan from the melt at 10 °C min^−1.
PBS	−33	70	72	116	77
PBS-ZnAl-PBSo-1%	−37	66	56	113	59
PBS-ZnAl-PBSo-3%	−36	69	59	114	61
PBS-ZnAl-PBSo-5%	−35	72	55	113	60
PBS-ZnAl-PBSo-10%	−34	70	51	114	57
PBS-MgAl-PBSo-1%	−35	70	68	115	71
PBS-MgAl-PBSo-3%	−35	68	64	115	64
PBS-MgAl-PBSo-5%	−32	64	75	116	76
PBS-MgAl-PBSo-10%	−34	71	57	115	57

---

Fig. 8 T_C profiles obtained by DSC during the cooling scan for (a) Zn/Al, (b) Mg/Al nanocomposites and PBS.

3.2.3 Viscoelastic and dynamic mechanical properties. In order to understand the thermal processability of the PBS/clay nanocomposites, some information on the rheological behavior of the nanocomposites under molten state is very useful.

The rheological behavior of the nanocomposites is strongly influenced by their nanoscale structure, the state of dispersion and interfacial characteristics. Therefore, information on rheological behavior of the nanocomposites also helps understand the structure–property relationship. By plotting the rheological data in Cole–Cole representation, curves related to η′′–η′(ω) can be extrapolated. Considering PBS-ZnAl-PBS_o-X% composites (Fig. 9a), the convex downward semi-circle profile at the intercept η′′ which tends to 0 and corresponds to the Newtonian zero-shear viscosity η′₀ at ω = 0, is shifted to a higher value for PBS-ZnAl-PBS_o-1% and 3% with respect to the homopolymer. Since the value of η′₀ is proportional to the molecular weight according to a power-law

η′₀ = KM_w^3.4 (3)

a chain extender effect of the filler toward the matrix is evident.^33,34 This can be related with an interface between the organoclay and the aliphatic polyester chains strongly developed, thanks to the functional groups present onto the filler. On the contrary, the composites with a larger clay content, such as 5% and 10%, do not present convex downward semi-circle profiles, and the viscosity values are low, suggesting a resulting effect more degradative than reinforcing most probably due to a chemical instability triggered above 3 wt% of filler. Concerning the PBS composites containing Mg/Al (Fig. 9b) the chain extender effect is observed in the case of 3 wt% of clay, while the semi-arc turns into a straight line, showing the presence of a gel-like structure, without apparent finite M_w, with 5 and 10 wt% of filler. This underlines the strong attritive phenomenon responsible of such behavior at low ω as well as an interface largely developed between the filler and the polymer.³⁵ Filler contents lower than 3 wt% determine a chain reduction effect. Therefore in this case LDH acts more as a plasticizer.

Fig. 9 Cole–Cole η′′–η′(ω) of (a) Zn/Al, (b) Mg/Al nanocomposites and PBS.

These results suggest the presence of a better dispersion of Mg/Al in the polymer matrix with respect to the systems containing Zn/Al. This is in agreement with literature¹⁸ and can be related to a finer crystallite size of Mg/Al with respect to Zn/Al containing clay, hence the interfacial interaction may be more developed. Indeed, as evidenced by the XRD results (Fig. 3), the former clay is less crystalline with a large number of nuclei.²⁰ The dynamic mechanical analysis substantially confirms the above discussed results. Fig. 10 shows the storage modulus E′ and tan δ (the damping factor), as a function of temperature. Almost all nanocomposites show better storage modulus than the neat PBS and the extent of the improvement depends on the clay interfacial interaction with the polymer, the aspect ratio of the clay, the processing procedure and the resultant nanocomposite structure.³⁶

Fig. 10 DMTA spectra for (a) Zn/Al, (b) some Mg/Al nanocomposites and PBS. Storage modulus (E′) and tan δ as a function of temperature.

More in detail the samples with Zn/Al present similar and close profiles of both E′ and tan δ. By considering the maximum tan δ peak, the glass transition temperature can be extrapolated.

PBS presents a T_g of almost −16 °C, which shifts to −12 °C with the addition of filler (Fig. 10a).

In general, nanofillers can reduce the chain mobility (increase of T_g) or increment the free volume due to the presence of the rigid nanosheets (decrease of T_g). Probably the first contribution predominates in our composites, even if the influence is not so significant, as already reported in many papers.^37–39

When taking the samples with Mg/Al into consideration, an analogous trend is present but the curves are less close with respect to the other composites (Fig. 10b). A more marked reinforcing effect is clearly present in PBS-MgAl-PBS_o-3% over the entire temperature range, because E′ increases of almost 50% at −150 °C with respect to the homopolymer. A slight increase in glass transition temperatures is also observed. Therefore DMTA data confirm the results obtained from rheology: the level of clay dispersion, hence the degree of interaction between the polymer matrix and the clay, is considerable and the final materials present a remarkable improvement in mechanical properties when compared to the homopolymer.

3.2.4 Tensile and flexural properties. As highlighted by the rheological and dynamic mechanical experiments, the composite PBS-MgAl-PBS_o-3% performs better than the other polymer clay compositions, therefore its tensile and flexural properties have been investigated. Table 5 shows the results obtained and, when compared to PBS, it can be observed as both strength (σ_max) and elongation at break (ε_b) increase, 30% and 98% respectively, suggesting a uniform dispersion and a good interfacial filler/polymer interaction. Moreover the improvement in strain value indicates that the filler does not cause brittleness in the material, a common shortcoming in composites.

Table 5 Mechanical properties of PBS and PBS-MgAl-PBS_o-3%

Sample	Tensile properties		Flexural properties
	Strength (σmax, MPa)	Elongation at break (εb, %)	Strength (σmax, MPa)	Elongation at break (εb, %)
PBS	25.2 ± 0.5	4.6 ± 0.2	36.5 ± 2.4	4.9 ± 0.4
PBS-MgAl-PBSo-3%	32.7 ± 1.4	9.1 ± 1.2	46.5 ± 0.5	8.4 ± 0.1

---

An analogous behavior was noticed when applying flexural tests, as the composite presents higher strength and strain values, 27% and 71% respectively, confirming the high degree of dispersion of LDHs inside the PBS matrix, which results in the high amount of stress transfer at the interface. These data are consistent with literature.³⁶

4 Conclusions

Novel and fully biobased nanocomposites based on PBS and layered double hydroxides have been prepared by in situ polymerization, with different percentages of filler and two different divalent cations (Zn²⁺ or Mg²⁺). In order to enhance the compatibilization of the inorganic filler with the polymer matrix, an oligomer of PBS was used as intercalating agent in the coprecipitation synthesis of LDH. The experimental results have suggested that the intercalation reaction provides a mixed filler system, containing most probably butanediol, succinate and some oligomeric chains. On the basis of the data reported, it can be concluded that:

- Mg-containing LDHs seem to retain more organic modifier compared to Zn-containing LDHs because of a grafting reaction occurring on the surface of the clay;

- Nanocomposites with Zn²⁺ present better thermal stability with respect to Mg²⁺;

- Nanocomposites with Mg²⁺ show superior mechanical properties with respect to Zn²⁺;

- A significant chain extender effect of the filler is highlighted.

The setting-up of these bionanocomposites makes it possible to obtain new materials with tunable and superior mechanical properties, thanks to an excellent developed interfacial interaction filler/polymer. This surely allows to widen the field of application of PBS to sectors requiring better mechanical properties as well as barrier effect. Moreover, the successful strategy of introducing an oligomer of the matrix within the filler can be simply applied and exploited for similar systems as well.

References

1. M. M. Reddy, S. Vivekanandhan, M. Misra, S. K. Bhatia and A. K. Mohanty, Prog. Polym. Sci., 2013, 38, 1653–1689.
2. L. Sisti, G. Totaro and P. Marchese, in Biodegradable and Biobased, Polymers for Environmental and Biomedical Applications, ed. S. Kalia and L. Avérous, John wiley & sons, Expected publication date, 2015.
3. S. Sinha Ray and M. Bousmina, Prog. Mater. Sci., 2005, 50, 962–1079.
4. L. Sisti, G. Totaro, M. Fiorini, A. Celli, C. Coelho, M. Hennous, V. Verney and F. Leroux, J. Appl. Polym. Sci., 2013, 130, 1931–1940.
5. J. W. Rhim, H. M. Park and C. S. Ha, Prog. Polym. Sci., 2013, 38, 1629–1652.
6. V. Ojijo and S. Sinha Ray, Prog. Polym. Sci., 2013, 38, 1543–1589.
7. U. Costantino, M. Nocchetti, L. Tammaro and V. Vittoria, Recent Pat. Nanotechnol., 2012, 6, 218–230.
8. J. M. Oh, S. J. Choi, G. E. Lee, J. E. Kim and J. H. Choy, Chem.–Asian J., 2009, 4, 67–73.
9. J. M. Oh, D. H. Park, S. J. Choi and J. H. Choy, Recent Pat. Nanotechnol., 2012, 6, 200–217.
10. E. Kafunkova, K. Lang, P. Kubat, M. Klementova, J. Mosinger, M. Slouf, A.-L. Troutier-Thuilliez, F. Leroux, V. Verney and C. Taviot-Gueho, J. Mater. Chem., 2010, 20, 9423–9432.
11. F. Leroux and J. P. Besse, Chem. Mater., 2001, 13, 3507–3515.
12. A. L. Troutier-Thuilliez, C. Taviot-Gueho, J. Cellier, H. Hintze-Bruening and F. Leroux, Prog. Org. Coat., 2009, 64, 182–192.
13. F. R. Costa, M. Saphiannikova, U. Wagenknecht and G. Heinrich, Adv. Polym. Sci., 2008, 210, 101–168.
14. C. Berti, M. Fiorini and L. Sisti, Eur. Polym. J., 2009, 45, 70–78.
15. C. Taviot-Guého and F. Leroux, in Structure and Bonding, ed. X. Duan and D. G. Evans, Berlin, Germany, 2006, vol. 119, pp. 121–159.
16. R. Schollhorn, Chem. Mater., 1996, 8, 1747–1757.
17. S. P. Lonkar, S. Therias, N. Caperaa, F. Leroux and J. L. Gardette, Eur. Polym. J., 2010, 46, 1456–1464.
18. C. Manzi-Nshuti, P. Songtipya, E. Manias, M. M. Jimenez-Gasco, J. M. Hossenlopp and C. A. Wilkie, Polymer, 2009, 50, 3564–3574.
19. (a) Y. J. Phua, W. S. Chow and Z. A. Mohd Ishak, eXPRESS Polym. Lett., 2013, 7, 340–354; (b) L. Ren, Y. Wang, J. Ge, D. Lu and Z. Liu, Macromol. Chem. Phys., 2015, 216, 636–640.
20. F. Cavani, F. Trifirò and A. Vaccari, Catal. Today, 1991, 11, 173–301.
21. H. Yang, L. Song, Q. Tai, X. Wang, B. Yu, Y. Yuan, Y. Hu and R. K. K. Yuen, Polym. Degrad. Stab., 2014, 105, 248–256.
22. H. S. Kim, H. J. Kim, J. W. Lee and I. G. Choi, Polym. Degrad. Stab., 2006, 9, 1117–1127.
23. H. C. Greenwell, C. C. Marsden and W. Jones, Green Chem., 2007, 9, 1299–1307.
24. E. Káfuňková, C. Taviot-Guého, P. Bezdička, M. Klementová, P. Kovář, P. Kubát, J. Mosinger, M. Pospíšil and K. Lang, Chem. Mater., 2010, 22, 2481–2490.
25. C. Taviot-Guého, A. Faour, Y. Feng and F. Leroux, Faraday Trans., 2010, 39, 5994–6005.
26. S. S. Ray, K. Okamoto and M. Okamoto, Macromolecules, 2003, 36, 2355–2367.
27. Z. Qiu and W. Wang, Polymer, 2006, 47, 6429–6437.
28. S. Mallakpour and M. Dinari, Prog. Org. Coat., 2014, 77, 583–589.
29. E. Conterosito, G. Croce, L. Palin, C. Pagano, L. Perioli, D. Viterbo, E. Boccaleri, G. Paul and M. Milanesio, Phys. Chem. Chem. Phys., 2013, 15, 13418–13433.
30. Q. Tao, H. He, R. L. Frost, P. Yuan and J. Zhu, Appl. Surf. Sci., 2009, 255, 4334–4340.
31. M. F. Chiang, M. Z. Chu and T. M. Wu, Polym. Degrad. Stab., 2011, 96, 60–66.
32. Y. J. Phua, W. S. Chow and Z. A. Mohd Ishak, eXPRESS Polym. Lett., 2011, 5, 93–103.
33. M. Hennous, Z. Derriche, E. Privas, P. Navard, V. Verney and F. Leroux, Appl. Clay Sci., 2013, 71, 42–48.
34. F. Leroux, A. Dalod, M. Hennous, L. Sisti, G. Totaro, A. Celli, C. Coehlo and V. Verney, Appl. Clay Sci., 2014, 100, 102–111.
35. Q. Zhou, V. Verney, S. Commereuc, I. J. Chin and F. Leroux, J. Colloid Interface Sci., 2010, 349, 127–133.
36. V. Ojijo and S. S. Ray, Prog. Mater. Sci., 2014, 62, 1–57.
37. Y. F. Shih, T. Y. Wang, R. J. Jeng, J. Y. Wu and C. C. Teng, J. Polym. Environ., 2007, 15, 151–158.
38. K. Okamoto, S. Sinha Ray and M. Okamoto, J. Polym. Sci., Part B: Polym. Phys., 2003, 41, 3160–3172.
39. S. Sinha Ray, K. Okamoto and M. Okamoto, J. Appl. Polym. Sci., 2006, 102, 777–785.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra24031j

---