Effects of surface modification and ultrasonic agitation on the properties of PHBV/ZnO nanocomposites

ZnO particles synthesis

ZnO crystals were synthesized by chemical precipitation method [15] from aqueous solution of zinc nitrate hexahydrate (Zn(NO₃)₂.6H₂O, Fluka Analytical, 99%) and hexamethylenetetramine ((CH₂)₆N₄, Merck, 99%). Crystallization reaction was carried out in a 2 L-capacity, triple-neck, jacketed reactor at 95°C reaction temperature and for 90 min. Inlet concentration of the reaction species were 0.06 M.

Surface modification was applied to ZnO particles using 3-(trimethoxysilyl)propyl methacrylate silane (CH₂=C(CH₃)CO₂(CH₂)₃Si(OCH₃)₃, MPS, Sigma-Aldrich, 98%) as coupling agent. MPS was first hydrolyzed in ethanol-water solution at pH=5.0 in an ultrasonic bath. The amount of MPS used for 1 g of ZnO was 1.67 g. ZnO particles were added into the solution to react with silane molecules for 2 h. The particles were then separated by filtration and drying.

Ultrasonic dispersion of particles in ethanol-water solution (pH=5.0) was performed using Sonics Vibra Cell ultrasonic processor with maximum power input of 750 W, with an ultrasonic probe of 13 mm diameter, at 20 kHz frequency and 40% amplitude. ZnO particles were then separated from the solvent by filtration and drying.

Characterization of ZnO particles

Scanning electron microscopy (SEM) images of the particles were taken using Zeiss EVO-LS at an accelerating voltage of 10.0 kV. Samples were coated by Au-Pd prior to examination.

Fourier transform infrared spectroscopy (FT-IR) analysis was performed on untreated and silane-treated particles, using Bruker Optics Alpha-P, in the 400–4000 cm⁻¹ region and at a resolution of 4 cm⁻¹. Spectra were obtained by collecting the average of at least 20 scans.

Fabrication of PHBV/ZnO biocomposites

PHBV (valerate content of 8 wt%) was kindly supplied by AdMajoris Company, France with the trade name of Maj’Eco FN000HA. Incorporation of ZnO particles into PHBV matrix was performed using a twin-screw extruder (Rondol, 10 mm Microlab, L/D=20) with a temperature profile of 90-135-160-160-150°C from feed to die and a screw speed of 80 rpm. PHBV/ZnO biocomposite pellets obtained after extrusion were compression molded into biocomposite sheets of 0.8 mm thickness using a mechanical press at 165°C temperature and 30, 70, and 90 bar consecutive pressures.

Characterization of PHBV/ZnO biocomposites

SEM images of PHBV composites were taken by Philips XL30 ESEM-FEG at an accelerating voltage of 5.0 kV. Samples were coated with Au prior to examination.

Thermal properties of PHBV and its composites were evaluated by differential scanning calorimetry (DSC) using Perkin Elmer (Pyris 1) under nitrogen atmosphere. Samples with 10 mg mass were heated from 0°C to 190°C (1^st heating run), kept at this temperature for 2 min to erase thermal history, cooled from 190°C to 0°C (cooling run), kept at this temperature for 2 min, and then heated to 190°C (2^nd heating run). The heating/cooling rate was 10°C/min for all three steps. Melting temperature (T_m) and enthalpy of fusion (ΔH_m) values were determined from the heating curves and crystallization temperature (T_c) and crystallization enthalpy (ΔH_c) values were determined from the cooling curves. Degree of crystallinity (X_c) was calculated from the equation:

where ΔH⁰ is the melting enthalpy of 100% crystalline PHBV (taken as 146 J/g [16], considering the low valerate content of PHBV in this study) and w is the weight fraction of PHBV in the blend.

Thermal stability of PHBV and its composites were analyzed by thermogravimetry analysis (TGA) using Perkin Elmer (Pyris Diamond) thermal analysis instrument. The samples (10 mg mass) were heated from room temperature to 550°C with a heating rate of 10°C/min under nitrogen flow. Decomposition peak temperatures (T_max) of the samples were determined from derivative TG curves as the temperatures corresponding to the highest weight loss rate.

Uniaxial tensile testing was performed according to ASTM D882-12 standard, using Instron 5982 Universal testing machine. An initial load of 1 N was used and the cross-head speed was 5 mm/min. Tensile strength at break, Young’s modulus, and elongation at break values were determined from the stress-strain curves. Toughness value was calculated as the area underneath the stress-strain curve.

Synthesis of ZnO particles and incorporation into PHBV matrix

ZnO crystals were synthesized by chemical precipitation method from aqueous solution of zinc nitrate hexahydrate ((Zn(NO₃)₂.6H₂O) and hexamethylenetetramine ((CH₂)₆N₄). The chemical reaction started with decomposition of (CH₂)₆N₄ into formaldehyde and ammonia, and then to ammonium cations and hydroxide anions [15]. The formation of ZnO precipitate takes places according to the following reaction:

The ZnO crystals obtained at the end of the reaction were rod-shaped particles, coalescing into asterisk-like structures at their centers (Fig. 1a). Formation of rod-shaped ZnO arrays at high reaction temperatures by self-aggregation mechanism was previously reported [17]. In fabrication of polymer-inorganic particle composites, particle aggregation is known to affect composite properties negatively. Therefore, surface modification and ultrasonic dispersion methods were applied to the crystals aiming the breakage of particle agglomerates into singular, discrete particles.

Fig. 1:

SEM images of synthesized ZnO particles (a) before and (b) after silane modification.

Surface modification was applied using MPS as coupling agent. Modification of ZnO nanostructures using MPS in ethanol solution and in acidic conditions was previously reported to produce hydrophobic ZnO surfaces [18]. SEM image of the crystals after silane modification is given in Fig. 1b. Regular array structures observed in the SEM images of crystals before silane modification seem to be broken by silanization. Nevertheless, severe particle agglomeration could not be avoided.

FT-IR spectra of untreated and silane-treated particles (Fig. 2) denote that silane was chemically bonded to the particles after silanization procedure. The characteristic peak related to Zn-O stretching can be observed between 400 and 500 cm⁻¹, for both untreated and silane-treated particles. The peaks observed around 1020–1090 cm⁻¹ were associated with Si-O-Si bonds. The broad peak around 3200 was associated with H-bonded OH stretching and could be originating from water or ethanol. The peaks at 2949, 1713, 1630, and 1290 were related to –CH₂ stretching vibration, C=O stretching, –C=C stretching, and C–H in-plane bending, respectively. The broad peak around 700 cm⁻¹ was related to cis–C–H out-of-plane bending [19].

Fig. 2:

FT-IR spectrum of untreated and silane-treated ZnO particles.

For the purpose of deagglomeration of silane-treated particles, ultrasonic agitation was applied using an ultrasonic horn. A previous study on effectiveness of ultrasonic dispersion on ZnO suspension in water [12] reported that majority of deagglomeration and size reduction occurred within the 20–30 min of ultrasound application. Ultrasonication times in this study were selected as 15 and 60 min, whereas SEM images (Fig. 3a d b) revealed that complete deagglomeration could not be achieved even after 60 min of ultrasonication. The chemical bonds forming between particles after silane modification might have prevented deagglomeration by ultrasonication. Ultrasonic dispersion was also applied to untreated ZnO particles for comparison purpose. When untreated particles were used, 60 min of ultrasonication was very effective in deagglomeration and singular rod-shaped particles could be obtained (Fig. 3c).

Fig. 3:

SEM images of ZnO particles after (a) silane treatment and 15 min ultrasound application, (b) silane treatment and 60 min ultrasound application, and (c) 60 min ultrasound application (no treatment).

Particle size distribution was analyzed by statistical analysis of SEM image of deagglomerated particles after ultrasound application. Mean value of the length of ZnO rods were calculated as 3590±1880 nm and the mean diameter of the particles were calculated as 394±124 nm. Particle diameter was defined as the highest distance between opposite corners across the hexagonal face, measured at the thickest part of the crystals. Mean aspect ratio (L/D) of the particles was determined as 9.1.

ZnO particles were incorporated into PHBV matrix after silanization and/or ultrasonication in order to compare the effects of these processes on the thermal and mechanical properties of PHBV. ZnO content of 1% (w/w) was selected, since dispersion of particles is easier at lower particle loadings. The formulations and sample coding of the PHBV/ZnO biocomposites are given in Table 1.

Characterization of PHBV/ZnO biocomposites

Distribution of ZnO particles within the PHBV matrix was evaluated by SEM (Fig. 4). All composite samples showed considerably good particle dispersion regardless of the degree of particle agglomeration observed before incorporation into the matrix, emphasizing the effectiveness of melt-extrusion process on homogeneous particle distribution. However, appearance of regular rod-shaped particles rather than irregular ZnO particulates was higher in PHBV/ZnO(U) composite sample, as complete deagglomeration was achieved with these particles prior to inclusion into polymer.

Fig. 4:

SEM images of (a) PHBV/ZnO(S), (b) PHBV/ZnO(S-U), and (c) PHBV/ZnO(U) composites.

Melting and crystallization behavior of PHBV and its composites were investigated by DSC analysis. First heating scans of all samples (Fig. 5a) exhibited a double-melting peak behavior, indicating the existence of two crystal phases. Bimodal melting behavior was previously explained by melting, recrystallization, and remelting during the heating scan [20]. If the heating rate is sufficiently low, less perfect crystals find time to melt and recrystallize into more stable crystals. More stable crystals then melt at a higher temperature. Low-temperature melting peak (T_m1) was observed around 174°C for all samples, whereas high-temperature peak (T_m2) was around 181°C (Table 2). The intensity of the T_m2 peak was increased for ZnO-added composites, implying more crystals recrystallizing into more stable crystals probably due to the increased mobility of polymer chains. This effect was lowest for P/ZnO(S) among the composite samples, while it was the highest for the P/ZnO(U) sample. Total enthalpy of fusion (ΔH_m) value was decreased for the composite samples compared to neat PHBV, indicating a decrease in the degree of crystallinity. Lowest ΔH_m value was obtained for the P/ZnO(U) sample.

Fig. 5:

DSC thermograms of PHBV and PHBV/ZnO composites: (a) first heating, (b) cooling, and (c) second heating curves.

Double-melting peak behavior was not observed during the second heating scan (Fig. 5c). This indicated that processing conditions during the composites fabrication promoted melting-recrystallization-remelting mechanism, whereas controlled crystallization during the cooling scan hindered this behavior. T_m value did not change with the addition of ZnO particles and was around 175°C for all samples. However, ΔH_m value decreased by the addition of ZnO, implying a decrease in crystallinity. Lowest X_c value was obtained for the P/ZnO(U) composite. Similarly, T_c values were not changed with the addition of ZnO particles, but ΔH_c values were decreased, during the cooling scan (Fig. 5b). PHBV is a semi-crystalline polymer with considerably high crystallinity, resulting in impaired mechanical properties such as brittleness. Therefore, the decrease in X_c value can be important for improvement of mechanical properties of the polymer.

Hydroxyapatite particles had shown a decreasing effect on crystallinity of PHBV, in a previous study of the authors [7], explained by formation of attractive non-covalent interactions that inhibit crystallinity by constraining polymer molecules. In another study related to the influence of ZnO nanoparticles on electrospun PHBV nanofibers [21], ZnO particles were reported to disturb the mobility of PHBV chains, by formation of hydrogen bonds between the particles and PHBV, resulting a decrease in crystallinity. Xiang et al. [22] also reported that formation of hydrogen bonding induced decrease in crystallinity in tannic acid reinforced PHBV films.

Thermal degradation of PHBV and its composites were analyzed by TGA. Thermal degradation of PHBV proceeds via random chain scission, resulting in a gradual reduction in molecular weight [23]. The PHBV and composite samples fabricated in this study, exhibited a similar degradation mechanism; one-step weight loss around 300 ˚C. On the other hand, decomposition peak temperature (T_max) was slightly decreased for PHBV/ZnO composites in comparison with neat PHBV (Table 3). The composites exhibited similar T_max values, and they were about 7–9°C lower than that of neat PHBV. Slight decreases were also observed for the temperatures corresponding to 2% (T_d,2) and 10% of weight loss (T_d,10) for PHBV/ZnO composites. Fortunately, T_d,₂ values were considerably above the T_m values, and the processing window of the polymer was not significantly narrowed.

In a previous study [24], ZnO particles incorporated in PHBV by solution casting method were reported to increase thermal degradation temperatures. This behavior was explained by the creation of a barrier by ZnO particles against the transport of decomposition products from the bulk to the gas phase. Processing conditions can strongly affect the distribution, orientation, and location of the particles within the polymer matrix; resulting in diverse composite properties [24]. High thermal conductivity of ZnO particles might have also contributed to thermal degradation. Previously, increased thermal conductivity due to the addition of cellulose nanowhiskers [25] and graphene [26], [27] were reported to promote the decrease in thermal degradation temperature of polymers.

Mechanical properties of PHBV and PHBV/ZnO composite samples were evaluated by tensile testing. Mechanical properties of polymer composites are known to be strongly related to degree of filler dispersion, interfacial characteristics, and crystallinity. In general, ZnO particles affected tensile strength of PHBV positively, (Fig. 6a). For the composites fabricated with ultrasound-dispersed particles, this effect was more profound due to better particle dispersion within the matrix. Young’s modulus did not change significantly, suggesting ZnO particles were effective on strength of the polymer rather than its stiffness. Ultrasound-dispersed ZnO particles increased elongation at break and toughness values of PHBV significantly, and the increase was higher for silane-treated particles (Fig. 6b). Decrease in crystallinity is known to cause an increase in ductility, and the tensile results obtained were consistent with the DSC analysis findings. Nevertheless, ductility is also related to the shear stress at the particle-polymer interface, which could be affected positively with silane-treatment due to stronger adhesion. Better particle distribution due to ultrasound application might have also affected ductility positively.

Fig. 6:

Tensile properties of PHBV and PHBV/ZnO composites: (a) tensile strength and Young’s modulus, and (b) toughness and elongation at break.