Thermal stability of CaCO 3 /polyethylene (PE) nanocomposites

Abstract

Calcium carbonate (CaCO₃) nanoparticles in polymer matrix cause to improvement in polymer performance, including thermal stability and mechanical properties. The main goal of this article is to investigate the effect of different weight percentage of nanoparticles of CaCO₃ on thermal stability and mechanical properties of polyethylene (PE) nanocomposites. The morphological structure of CaCO₃ nanoparticles and nanocomposites was investigated by transmission electron microscopy and scanning electron microscopy. The thermal stability of PE and its nanocomposites was also determined by differential scanning calorimetry (DSC), thermogravimetric analysis (TGA), and thermomechanical analysis. Nonisothermal crystallization experiments by DSC test showed that the incorporation of nanoparticles increased the crystallinity, glass transition temperature, and the effective energy barrier for crystallization process. Besides, degradation behavior was evaluated by TGA. The onset mass loss temperature shifted to higher value in the presence of nanoparticles.

Introduction

In recent years, an interesting field for researchers is to investigate the properties of nanocomposites. Among the nanofillers, nanoparticles cause to improvement in mechanical and thermal properties of polymer nanocomposites. Calcium carbonate (CaCO₃) is one of the most useful and abundant materials which was used in polymer composites. It has three crystal morphologies, calcite, aragonite, and vaterite. The structure of calcite finds in nature easily, and it is more stable than other polymorphs of CaCO₃ at ambient temperature and atmosphere pressure. CaCO₃ with a few weight percentages loading in the thermoplastic polymer nanocomposites strongly effects on properties of polymers. ^{1 –7} The effects of addition of precipitated CaCO₃ (PCC) to thermoplastic polymers have been investigated by many researchers. The results showed that the significant improvements have been shown in thermal and mechanical properties, such as higher heat distortion temperatures, an enhancement in flame resistance, increase in modulus, stiffness, better barrier properties, enhancements of viscosity, and dimensional stability. ^{1 –7}

Among the thermoplastic polymers, polyethylene (PE) is one of the most widely used polymers, which possess relatively low costs but poor thermal and mechanical properties. In order to improve the weakness points of PE, nanosized CaCO₃ has been incorporated in PE. ^{1 –9}

Wang et al. investigated the effect of CaCO₃ on mechanical properties of low-density PE nanocomposites. They showed that the tensile property and flexural modulus of nanocomposite evidently increased by the addition of CaCO₃ as higher crystallinity index of nanocomposite compared to pure polymers. ¹

Tanniru and Misra demonstrated that the CaCO₃ nanoparticles in high density polyethylene (HDPE) matrix increased the bulk crystallinity and modulus, while the nucleating effect decreased the spherulite size. ⁸ The addition of CaCO₃ to PE increased impact strength at temperature range of −40 to +70°C. ⁸ The reinforcement of HDPE with nanosized CaCO₃ retains adequately high strength in the temperature range of −40 to +20°C. ⁹ High impact strength was accompanied by increase in modulus. The reinforcement of neat HDPE with nanosized CaCO₃ altered the deformation micro mechanism from crazing–tearing in HDPE to fibrillated fracture in polymer nanocomposite. ⁹

In the present study, CaCO₃/PE nanocomposites have been synthesized. The fracture surface of nanocomposite was confirmed by scanning electron microscopy (SEM). Furthermore, thermal degradation, crystallization kinetics, and mechanical properties of the CaCO₃/PE nanocomposites are also investigated.

Experimental

Materials

PE with the density (0.937 g cm⁻³), vicat softening point (117°C), and melting flow index (MFI) at 190°C with 2.16 kg mass is measured 4.2 g/10 min which was supplied by Tabriz Petrochemical Complex, Iran. The CaCO₃ nanoparticles (Socal® PCC) were obtained from Solvay Advanced Functional Materials, Salin Degiraud, France. CaCO₃ have a rhombohedra structure which related to the calcite phase. Figure 1 shows the transmission electron microscopic (TEM) micrographs of nanosized CaCO₃. As seen, nanoparticle has an irregular morphology and an average size of about 60 nm was measured using particle size analyzer (PSA), which is shown in Figure 2.

Figure 1.

TEM micrograph of CaCO₃ nanoparticles.

Figure 2.

Size distribution of CaCO₃ nanoparticles using PSA.

Sample preparation

To evaluate the role of nanosized CaCO₃ on mechanical and crystallization behavior of PE, a series of PE/CaCO₃ nanocomposite were fabricated. For this purpose, nanosized CaCO₃ and PE were first mechanically mixed in different weight ratios to achieve the desired powder composites which was done to decrease agglomeration size of nanoparticles and have good dispersion of nanoparticles in PE powders. The premixed nanocomposites powder was suspended in methanol as a media. In order to avoid agglomeration of nanoparticles and improve the dispersion of nanosized CaCO₃ in matrix, the solution was placed in an ultrasonic bath for 45 min and was dried at 60–70°C. The nanocomposite powders were compressed under 20–30 kPa at 170–180°C for 5 min, followed by cooling to room temperature. Table 1 shows the sample notations of used materials.

Table 1.

The notations of samples.

Notation of samples	Weight percentage of CaCO3	Weight percentage of PE
NP0	0	100
NP5	5	95
NP10	10	90

CaCO₃: calcium carbonate; PE: polyethylene.

Figure 3 presents the SEM micrograph obtained from the fracture surface of NP5. Good dispersion of nanoparticles in polymer matrix provides evidence for adequate sample fabrication.

Figure 3.

(a) SEM micrographs of fracture surface of NP10 at magnification (×20,000). (b) SEM micrographs of fracture surface of NP10 at magnification (×40,000).

Differential scanning calorimetry

The nonisothermal crystallization of PE and PE/CaCO₃ nanocomposites was carried out by DSC60 (Shimadzu, Japan). Temperature calibration was carried out using an indium standard (T_m = 156.6°C and ΔH_f = 28.5 J g⁻¹). In order to evaluate the nonisothermal crystallization kinetics, about 2 mg of PE and its nanocomposites were encapsulated into aluminum pans and were initially heated from room temperature up to 200°C at heating rate of 10°C min⁻¹, at this temperature sample was held for 10 min to eliminate the thermal history of the samples. Then, the samples were cooled down to room temperature at four cooling rates of 1, 5, 10, and 15°C min⁻¹. After that, the samples were reheated from room temperature to 200°C at a heating rate of 10°C min⁻¹ to observe the melting behavior of PE and PE/CaCO₃ nanocomposites.

Thermogravimetric analysis

The thermogravimetric analysis (TGA) was studied using a thermal analysis system (TGA50, Shimadzu) in air atmosphere. Nonisothermal experiments were performed in the temperature range 25−1000°C at heating rates 10°C min⁻¹. The mass of the samples for each test was about 4 mg.

Transmission electron microscopy

TEM was done by a LEO 912 AB microscope. The nanoparticles of CaCO₃ were dispersed in ethanol by using sonication for 30 min, after that dropped onto a copper grid.

Scanning electron microscopy

The fracture surface of nanocomposite was observed by SEM (LEO 1450) at an accelerating voltage of 10 kV. In order to reduce the plastic deformation of PE, breaking the sample in liquid nitrogen was done.

Particle size analyzer

Measurement of the particle size and size distribution of nanosized CaCO₃ was taken using a PSA (SALD 2101, Shimadzu).

Results and discussion

Relative crystallinity, X(t), can be determined by using power used data at any crystallization temperature, as follows ⁹

$X(t)=\frac{{\displaystyle {\int }_{T_0}^T\left(\frac{\text{d}H}{\text{d}T}\right)\times \text{d}T}}{{\displaystyle {\int }_{T_0}^{T_{\infty }}\left(\frac{\text{d}H}{\text{d}T}\right)\times \text{d}T}}$

1

where X(t), T₀ and T_∞ are the relative crystallinity index, the onset and endset crystallization temperature, respectively, and $\left(\frac{d\text{H}}{d\text{T}}\right)$ is the heat flow rate. T is the crystallization temperature at time t.

Many models were suggested to evaluate the nonisothermal crystallization of polymers. The crystallization kinetics of polymers can be analyzed by using the Avrami equation which is the most common approach ^{10 –12}

$\text{ln}[-\text{ln}\left(1-X_t\right)]=\text{ln}\left(Z_t\right)+n\text{ln}\left(t\right)$

2

where X(t) is the relative crystallinity at time t, the Z_t value is the crystallization rate constant (min⁻¹), and n value is the Avrami exponent. n value depends on the nucleation type and growth process parameters. The Avrami equation is based on the isothermal crystallization system. To study the characterization of nonisothermal process, Jeziorny ¹³ suggested that the crystallization rate constant was corrected by cooling rate φ

$\text{ln(}{Z}_c\text{)= ln}\left(Z_t\text{/φ}\right)$

3

Thus, the Avrami equation can be used to analyze the nonisothermal system.

Also, Ozawa ¹⁴ extended the Avrami theory from isothermal crystallization to the nonisothermal case by considering that crystallization occurs at a constant cooling rate. According to Ozawa’s theory, the relative crystallinity X_t at temperature T can be calculated as following equation

$\text{ln}[-\text{ln}\left(1-X_t\right)]=\text{ln}(K(T))-m\text{ln}(\phi )$

4

where K(T), m, and φ are the function of cooling rate, Ozawa exponent similar to Avrami exponent, the cooling rate, respectively. Finally, a linear graph of $\text{ln}[-\text{ln}\left(1-X_t\right)]$ as a function of ln(φ) was obtained. The Ozawa exponent and the kinetic parameter K(T) as slope and intercept of line were determined.

Mo’s group evaluated the crystallization behavior. ¹⁵ The nonisothermal crystallization was investigated by combining Avrami and Ozawa equations. In order to determine the relation between t and φ, they suggested that these two equations were connected together, as follows

$\text{log}(\phi )=\text{log}(F(T))-a\text{log}(t)$

5

where

$F(T)={\left[\frac{K(T)}{Z_t}\right]}^{\frac{1}{m}}a=\frac{n}{m}$

The physical meaning of the rate parameter F(T) refers to the cooling rate to reach a defined degree of crystallization at specific time. According to equation (5), at a given degree of crystallinity, the linear graph of log φ versus log t was plotted to determine slope, a, and intercept, F(T).

Figure 4 shows the nonisothermal crystallization differential scanning calorimetry (DSC) curves of PE, NP5, and NP10 at cooling rates of 10, 20, and 30°C min⁻¹. As seen in these figures, a single crystallization peak could be observed at each cooling rate. Besides, the exothermic peak becomes wider and shifts toward a lower temperature when the weight percentage of CaCO₃ nanoparticles in composite and cooling rate increased. Some useful parameters can be read from these curves directly. They are the peak temperature, T_p, the starting crystallization temperature, T_o, crystallization enthalpy, ΔΗ_C, also glass transition temperature, T_g, which are listed in Table 2. T_p, T_e, T_o values shift to lower temperature when the cooling rates increase for PE and its nanocomposites. Besides, at each cooling rate, T_p slightly increases with an augmenting the weight percent of CaCO₃ nanoparticles. It is implying that the presence of nanoparticles increases the crystallization rate. Moreover, T_g values for HDPE and its nanocomposites have similar trend by increasing the weight percent of nanoparticles. This effect can be related to the interaction of PE chains and nanoparticles and increase in HDPE chains rigidity. ¹⁶ When the amount of nanosized CaCO₃ in HDPE matrix increases, inter-particle distance decreases, so larger inter-particle distance of NP5 respect to NP10 causes to a weak network. Therefore, NP10 has minimum T_g value compared to NP5 and NP0.

Figure 4.

(a) DSC thermograms of nonisothermal crystallization of NP0 at different cooling rate. (b) DSC thermograms of nonisothermal crystallization of NP5 at different cooling rate. (c) DSC thermograms of nonisothermal crystallization of NP10 at different cooling rate.

Table 2.

Values of ΔΗ_C, T_o, T_e, T_g, and T_m at various cooling rates.

Cooling rate (°C min−1)	ΔΗC (J g−1)	To (°C)	Te (°C)	Tp (°C)	Tg (°C)	Tm (°C)
PE
1	130.6	120.8	118.2	119.2	−118.5	131.0
5	148.5	118.9	115.4	117.5		131.0
10	152.5	117	110.9	115.3		130.2
15	166.0	114.6	104.1	111.5		129.5
NP5
1	116.2	119.5	116.5	118.2	−117.7	129.1
5	131.2	117.9	114.0	116.6		130.4
10	127.9	118.4	110.7	116.1		129.9
15	136.8	116.0	104.8	112.8		130.6
NP10
1	103.6	122.2	119.1	120.8	−117.5	131.6
5	131.2	120.3	115.4	118.4		130.4
10	127.9	118.4	110.7	116.1		129.9
15	136.8	116.0	104.8	112.8		129.7

PE: polyethylene.

The ΔΗ_C shifts to higher values when the cooling rate increases. Also, the percentage of CaCO₃ loading in nanocomposites causes to decrease in ΔΗ_C values. This effect can be related to the imperfect crystallization of nanocomposite compared to pure PE. Therefore, lower energy is needed to complete the crystallization process compared to pure PE.

Figures 5 represents the relative crystallinity of PE and its nanocomposites as a function of temperature at various cooling rates. As it can be seen, the plots of X(t) versus T for PE are similar to its nanocomposites as the sigmoid shape, it is implying that only the lag effect of the cooling rate on crystallization is observed. At the same crystallinity, there is incremental trend for crystallization temperature when the cooling rate increases. The reason for this effect is related to good fluidity and diffusibility of molecular chains at lower cooling rates, so increase in crystallinity at higher temperatures with respect to lower cooling rate is shown. Also, nanosized CaCO₃ in PE matrix causes to increase in the crystallinity under higher temperatures with respect to pure PE.

Figure 5.

(a) Relative crystallinity index as a function of temperature for NP0 at different cooling rates. (b) Relative crystallinity index as a function of temperature for NP5 at different cooling rates. (c) Relative crystallinity index as a function of temperature for NP10 at different cooling rates.

The relative crystallinity versus time at different cooling rates for PE, NP5, and NP10 are shown in Figures 6. It can be observed, the duration of the crystallization process decreases when the cooling rate increases. NP5 and NP10 exhibit a similar behavior to pure PE.

Figure 6.

(a) Relative crystallinity as a function of time for NP0 crystallized nonisothermally at different cooling rate. (b) Relative crystallinity as a function of time for NP5 crystallized nonisothermally at different cooling rate. (c) Relative crystallinity as a function of time for NP10 crystallized nonisothermally at different cooling rate.

Based on Avrami equation, the plots of ln(−ln(1 − X)) as a function of ln(t) are shown in Figure 7. Z_c value and Avrami exponent, n, are obtained from the intercept and slope of a straight line, which are fitted in the Avrami plot. The values of Z_c, n, and r² parameters signify the quality of the fitting which are listed in Table 3. It is shown that all n values are in a limited range of 2.6–2.8. So, the crystallization mode of PE does not change with increasing the nanoparticles. Also, it is indicated that crystallization occurs at the three-dimensional growth for all samples. Based on the values of the r² parameter, the Avrami model is suitable model to describe the nonisothermal crystallization of PE and its nanocomposites.

Figure 7.

(a) ln(−ln(1 − X_t)) versus ln(t) for NP0 at various cooling rate. (b) ln(−ln(1 − X_t)) versus ln(t) for NP5 at various cooling rate. (c) ln(−ln(1 − X_t)) versus ln(t) for NP10 at various cooling rate.

Table 3.

Values of ln(Z_t), Z_c, n, and activation energy of crystallization at various cooling rate for PE and its nanocomposites.

	NP0
φ (°C min−1)	r 2	ln(Zt)	Zc	n	ΔE(j)
1	0.9989	−11.918	6.67 × 10−6	2.495	32,067
5	0.999	−9.1605	0.16	2.6495
10	0.9993	−8.3038	0.435	2.7159
15	0.9985	−7.8812	0.59	2.7649
	NP5
φ (°C min−1)	r 2	ln(Zt)	Zc	n	ΔE(j)
1	0.9998	−11.556	9.58 × 10−6	2.4316	35,772
5	0.9991	−8.8651	0.169	2.6399
10	0.998	−8.4887	0.428	2.7421
15	0.9966	−8.3018	0.575	2.8087
	NP10
φ (°C min−1)	r2	ln(Zt)	Zc	N	ΔE(j)
1	0.9999	−11.034	1.61 × 10−5	2.4257	36,115
5	0.9999	−8.834	0.171	2.5309
10	0.9992	−9.1615	0.4	2.8974
15	0.9972	−7.8856	0.59	2.775

PE: polyethylene.

The Ozawa equation is also frequently used to evaluate nonisothermal crystallization kinetics. According to fail to fit a linear plot when ln(ln(1 − X)) is plotted against ln(ϕ), so the Ozawa equation is not suitable to describe nonisothermal crystallization of PE, NP5, and NP10.

In order to determine the crystallization kinetic of polymers, Mo suggested that the double logarithm plot of cooling rate versus crystallization time at different relative crystallinity, X, was drawn. Then, F(T) and a could be determined from the slope and intercept of linear plots. Figure 8 presents the plots of ln(ϕ) versus ln(t) for PE and its nanocomposites. The values of F(T) and a are listed in Table 4 for PE, NP5, and NP10. Based on Figure 8, all plots gives a straight line. So, this approach is applicable to evaluate the crystallization behavior of PE and its nanocomposites. At a certain relative crystallinity, a high value of F(T) means a high cooling rate is needed to reach this X_t in a unit time, which attributes there is difficulty for crystallization of PE. It is found that the F(T) value increases slightly with reinforcement of nanoparticles in matrix. This means that the crystallization of PE chains becomes difficult in the presence of nanoparticles. The values of a for all samples are constant at different crystallinity degrees.

Figure 8.

Plots of ln(ϕ) versus ln(t) for NP0, NP5, and NP10 samples.

Table 4.

Values of F(T) and a for NP0, NP5, and NP10.

	PE
Xt	r 2	F(T)	a
20	0.9948	5.91	1.41
40	0.9957	6.4	1.43
60	0.9954	6.72	1.43
80	0.9935	7.08	1.46
	NP5
Xt	r 2	F(T)	a
20	0.971	5.76	1.41
40	0.972	6.256	1.41
60	0.963	6.62	1.43
80	0.961	6.962	1.43
	NP10
Xt	r 2	F(T)	A
20	0.998	6.3	1.57
40	0.999	6.73	1.55
60	0.999	7.09	1.55
80	0.9979	7.54	1.58

PE: polyethylene.

Kissinger proposed a method to calculate activation energy in nonisothermal crystallization ¹⁷

$\frac{\text{d ln}\left(\frac{\varphi }{T_P^2}\right)}{\text{d}\left(\frac{1}{T_P}\right)}=\frac{-\Delta E}{R}$

6

where R is the universal gas constant and the ΔE is the activation energy required for the transportation molecules from a molten state to growth of crystal surfaces. Figure 9 shows that the plots of $\text{ln}\left(\frac{\varphi }{T_P^2}\right)$ against $\left(\frac{1}{T_P^{}}\right)$ for PE and its nanocomposites. ΔE is obtained from the slope of this line, as listed in Table 3. The values of ΔE increases with increase in nanoparticles in PE matrix. Immobility of polymer chains as high T_g value, reduction of particle distance in PE matrix, and strong network between CaCO₃ nanoparticles and PE chains are concluded that more energy is used to rearrangement of PE chains is happen when the weight percentage of nanoparticles increases in PE. Also, it means that nanoparticles influence on the PE chains to crystallize difficulty and drop in crystallization rates during the nonisothermal crystallization process.

Figure 9.

Plots of $\text{ln}\left(\frac{\varphi }{T_P^2}\right)$ versus $\left(\frac{1}{T_P^{}}\right)$ for pure PE and its nanocomposites.

The effect of CaCO₃ nanoparticles on the thermal stability of nanocomposites is studied by means of TGA in air atmosphere. The weight loss (TG) and the derivative thermogravimetric profiles (DTG) for PE and its nanocomposites are shown in Figure 10. The incremental trend of decomposition temperature for PE is seen when the weight percentage of CaCO₃ added to PE matrix, indicating that the restriction flexibility of PE chains in the presence of CaCO₃ nanoparticles, and strong polymer network improved the thermal stability of PE. According to DTG profiles of all samples, two peaks are observed in the DTG curves at heating rate of 10°C min⁻¹. The presence of CaCO₃ nanoparticles causes to increase the amount of degradation during the second stage.

Figure 10.

Weight loss and DTG curves of PE and its nanocomposites at heating rate of 10°C min⁻¹.

The values of onset decomposition temperature (T_o), peak decomposition temperature at the maximum weight loss rate from DTG curves (T_p), endset decomposition temperature (T_e), and residual mass at 600°C (R_e) for PE and its nanocomposites are summarized in Table 5. It can be seen, CaCO₃ nanoparticles decreases the flexibility of PE matrix, it is concluded that high temperature needed to overcome the interaction between polymer chains and nanoparticle due to strong network, so the thermal stability of PE matrix increases with increase in weight percentage of nanoparticles.

Table 5.

Parameters of the TGA and DTG curves of pure PE and its nanocomposites.

Heating rate	To (°C)	Te (°C)	Tp (°C)	Residue (600°C, wt%)
NP0
10	396.74	472.15	417.1	0.744
NP5
10	412.63	487.95	426.89	5.177
NP10
10	427.27	493.64	443.95	10.133

TGA: thermogravimetric analysis; DTG: differential thermogravimetry; PE: polyethylene.

The tensile properties of PE and its nanocomposites are measured at different temperatures, that is, 30°C, 60°C and 90°C, and summarized in Figures 11 and 12. It is found that the elastic modulus and yield stress of all samples reduce when the temperature increases. It is related to increase in mobility of PE chains and decrease in interaction force between PE chains. Therefore, yield stress and elastic modulus of all samples reduce when the temperature increases.

Figure 11.

Yield stress of PE and its nanocomposites as a function of temperature.

Figure 12.

Variation of elastic modulus at different temperature versus weight percentage of CaCO₃ nanoparticles.

Moreover, with augmenting nanoparticles in matrix, elastic modulus and yield stress of composite increase compared to pure PE at all temperatures. This behavior could be explained by the immobility of PE chains, stiff network of polymer and nanoparticles. Clearly, the yield stress increases with increasing CaCO₃ nanoparticles content. It is probably due to the good interfacial adhesion between CaCO₃ nanoparticles and PE matrix.

Conclusions

Thermal stability of PE in the presence of CaCO₃ nanoparticles is studied by using DSC, TGA, and thermomechanical analysis. Also, the structure of CaCO₃/PE nanocomposites is investigated by using SEM. The results of DSC showed that the incorporation of nanoparticles to PE matrix caused to increase in the crystallization temperature as nucleation site of nanoparticles. Besides, the glass transition temperature enhances when the weight percentage of nanosized CaCO₃ increases. This effect can be related to the reduction mobility of PE chains. The Avrami and Mo models are found to describe the nonisothermal crystallization data of PE and its nanocomposites. The Ozawa equation failed to provide suitable description of the crystallization kinetics of PE, NP5, and NP10. The crystallization activation energy increases when the loading of nanosized CaCO₃ in PE matrix increases. TGA data reveal that CaCO₃ nanoparticles cause to delay the PE decomposition during thermal oxidation. PE has minimum value of yield stress and young modulus respect to NP5 and NP10 at different temperatures.