Blends of polyamide 12 and maleic anhydride grafted paraffin wax as potential phase change materials (original) (raw)

Provided for non-commercial research and education use. Not for reproduction, distribution or commercial use.

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues.

Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited.

In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit:
http://www.elsevier.com/copyright

Author’s personal copy

Material Behaviour

Blends of polyamide 12 and maleic anhydride grafted paraffin wax as potential phase change materials

A.S. Luyt a,v{ }^{\mathrm{a}, \mathrm{v}}, I. Krupa b{ }^{\mathrm{b}}, H.J. Assumption c,d{ }^{\mathrm{c}, \mathrm{d}}, E.E.M. Ahmad a{ }^{\mathrm{a}}, J.P. Mofokeng a{ }^{\mathrm{a}}
a{ }^{a} Department of Chemistry, University of the Free State (Qwaqwa Campus), Private Bag X13, Phathaditjhaba 9866, South Africa
b{ }^{\mathrm{b}} Polymer Institute, Slovak Academy of Sciences, Dúbravská cesta 9, 842 36 Bratislava, Slovakia
c{ }^{c} Central Analytical Facility, Stellenbosch University, Private Bag X1, Matieland 7602, South Africa
d{ }^{\mathrm{d}} Sasol Technology REvD, Klasie Havenga Road, Sasoburg 1947, South Africa

A R T I C L E I N F O

Article history:
Received 5 August 2009
Accepted 15 September 2009

Keywords:

Polyamide 12

Functionalized paraffin wax
Phase change materials
Morphology
Thermal and mechanical properties

A B S T R A C T

Polyamide 12 (PA12) was blended with a high content of a maleic anhydride functionalized paraffin wax in an attempt to prepare shape stabilized phase change materials for energy storage. The morphology of the immiscible blends, and possible interactions between the blend components, were characterized through scanning electron microscopy (SEM), Fourier-transform infrared spectroscopy (FTIR) and nuclear magnetic resonance spectroscopy (NMR) analyses. Although it could not be conclusively established, the obtained results indicated a strong possibility of hydrogen bonding interaction between the anhydride −C=O-\mathrm{C}=\mathrm{O} groups on the wax and the −N−H-\mathrm{N}-\mathrm{H} groups in the amide. Such an interaction was clearly evident in the thermal, thermo-mechanical, and mechanical behaviour of the blends, as investigated through differential scanning calorimetry (DSC), dynamic mechanical analysis (DMA), stress relaxation measurements, and thermogravimetric analysis (TGA).
© 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Phase change materials (PCM) have received great interest in many applications, including energy storage materials, thermal protection systems, as well as in active and passive cooling of electronic devices [1,2]. Different inorganic as well as organic substances have already been employed for the creation of phase change materials; among the most common ones are various salts and their eutectics, fatty acids and n-alkanes [3].

Paraffin waxes seem to belong to the prospective candidate group because of many desirable characteristics, such as high latent heat of fusion, negligible super-cooling, low vapour pressure in the melt, chemical inertness and stability [4]. The number of carbon atoms in the chains of

[1]paraffin waxes with melting temperatures between 30∘C30^{\circ} \mathrm{C} and 90∘C90^{\circ} \mathrm{C} ranges from 18 to 50 (C18-C50). Increased length of the carbon atom chains increases the molecular weight and results in a higher melting temperature of the material. The specific heat capacity of latent heat paraffin waxes is about 2.1 kJ kg−1 K−12.1 \mathrm{~kJ} \mathrm{~kg}^{-1} \mathrm{~K}^{-1}. Their melting enthalpy lies between 180 and 230 kJ kg−1230 \mathrm{~kJ} \mathrm{~kg}^{-1}, which is quite high for organic materials. The combination of these two values results in an excellent energy storage density. Finally, these materials are produced in substantial quantities by the chemical industry and they are, therefore, readily available and inexpensive [5,6].

Traditionally, paraffin waxes are kept in closed tanks or containers during heating to suppress wax leakage. Another possibility to keep waxes in a stable shape during application is to blend them with convenient polymers. Paraffin waxes, blended with appropriate polymers, seem to be the best candidates for the preparation of smart polymeric phase change materials for applications such as

1. a Corresponding author. Tel.: +27 587185314.
   E-mail address: LuytAS@qwa.ufs.ac.za (A.S. Luyt). ↩︎

thermal storage of solar energy, thermal protection of electronic devices, thermal protection of food and medical goods, passive storage in bioclimatic buildings, use of off-peak rates and reduction of installed power, and thermal comfort in vehicles [7-9].

Blending of paraffin waxes with polymers provides an opportunity to utilize phase change materials with a unique, controlled structure. A polymeric matrix fixes the paraffin wax in a compact form, even after melting, and suppresses wax leakage. Such materials are also easily shaped, and the polymeric phase provides its own specific properties. A variety of polymer matrices are available with a large range of chemical and mechanical properties. Thermosetting resins, for example, are particularly useful for applications where blending with paraffin waxes at room temperature is required, so that the low transition temperature PCM can stay in the solid phase during composite processing [10].

Polyethylene seems to be the most frequently used polymer for blending with paraffin waxes to obtain shapestabilized PCM [4,11]. For example, Inaba and Tu [11] investigated a shape-stabilized paraffin wax system based on paraffin wax blended with high-density polyethylene (HDPE). The resulting composite consisted of 74wt.%74 \mathrm{wt} . \% paraffin and 26wt.%26 \mathrm{wt} . \% HDPE. Our own results based on the investigation of LDPE/paraffin wax PCM were not so exciting [12]. The DMA analyses of the blends pointed out an important aspect, which in most cases is neglected. It is the question of which component forms the continuous and which the discontinuous phase. Although LDPE still forms the continuous phase up to 50wt.%50 \mathrm{wt} . \% wax, the amount of LDPE at such a high wax content is not enough to keep the material structure in a consistent shape. Controlled force ramp testing on DMA confirmed poor material strength, especially at temperatures above the wax melting temperature, i.e. temperatures that are interesting for energy storage applications. The highest concentration able to sustain the external forces as well as the thermal cycling was 40wt.%40 \mathrm{wt} . \% of paraffin wax. This value was even lower when polypropylene was used as the matrix [13]. The PCMs based on isotactic polypropylene (PP) blended with soft and hard Fischer-Tropsch paraffin wax both showed a strong phase separation. It was found that paraffin waxes melted separately in the PP matrix (up to 40wt.%40 \mathrm{wt} . \% of wax), whereas the material remained in the solid state [13].

Other potential polymer matrices are styrene-ethylene-butylene-styrene (SEBS), styrene-isoprenestyrene (SIS), and styrene-ethylene-propylene-styrene (SEPS) [14,15]. Obviously, an increase in the various applications of PCM increases the need for other potential polymers that could be used as shape stabilizing matrices for paraffin wax. In this paper we report results on the physical properties of polyamide 12 blended with a Fischer-Tropsch paraffin wax. Polyamide 12 is linear aliphatic polyamide which is a semicrystalline polymer that usually exhibits a relatively high modulus, toughness and strength, low creep and good temperature resistance [16]. In order to improve the compatibility of the components and to prevent wax leakage from the PA12 matrix, the paraffin wax was grafted with maleic anhydride [17].

2. Experimental

The grafted paraffin wax was prepared by mechanically mixing a hard Fischer-Tropsch paraffin wax (supplied by Schümann-Sasol, average molecular weight 800 g mol−1800 \mathrm{~g} \mathrm{~mol}^{-1}, melting point 90∘C90^{\circ} \mathrm{C}, density 0.94 g cm−30.94 \mathrm{~g} \mathrm{~cm}^{-3} at 25∘C25^{\circ} \mathrm{C} ), maleic anhydride (MA) and dibenzoyl peroxide (DPP) (both supplied by Sigma-Aldrich). This was followed by heating the mixture in an oil bath at 140∘C140^{\circ} \mathrm{C} for 10 min in a nitrogen atmosphere, cooling down and grinding the sample, washing in boiling water to dissolve any unreacted MA, filtering and drying in an oven at 50∘C50^{\circ} \mathrm{C} [17]. Polyamide 12 (PA12) in powder form (Vestosint, Degussa, Germany) was used as the matrix. The blends were prepared by mixing the components in the 30 ml mixing chamber of a Brabender Plasticorder PLE 331 (Brabender, Düsseldorf, Germany) at 200∘C200^{\circ} \mathrm{C} for 10 min at a mixing speed of 35 rpm . The maximum amount of wax that could be practically mixed with PA12 was 30wt.%30 \mathrm{wt} . \%. The samples for further characterization were subsequently compression molded at 200∘C200^{\circ} \mathrm{C} for 2 min .

Thermogravimetric analyses were carried out in a Perkin Elmer TGA7 from 30 to 650∘C650^{\circ} \mathrm{C} at a heating rate of 10−6 min−110^{-6} \mathrm{~min}^{-1} in flowing nitrogen (20mlmin−1)\left(20 \mathrm{ml} \mathrm{min}^{-1}\right). Differential scanning calorimetry was carried out in a Perkin Elmer DSC7 in flowing nitrogen (20mlmin−1)\left(20 \mathrm{ml} \mathrm{min}^{-1}\right). Samples were heated from 20 to 220∘C220^{\circ} \mathrm{C} at a heating rate of 10−6 min−110^{-6} \mathrm{~min}^{-1}, and then cooled and re-heated under the same conditions. Melting and crystallization temperatures, as well as melting and crystallization enthalpies, were determined from the cooling and second heating runs. DMA analyses were conducted in a Perkin Elmer Diamond DMA using a dual cantilever in the bending mode. The sample dimensions were 50 mm×10 mm×1 mm50 \mathrm{~mm} \times 10 \mathrm{~mm} \times 1 \mathrm{~mm}. The samples were heated at 3∘Cmin−13^{\circ} \mathrm{C} \mathrm{min}^{-1} from 30 to 165∘C165^{\circ} \mathrm{C} at a frequency of 1 Hz . The strain amplitude was 20μ m20 \mu \mathrm{~m} and the force amplitude initial value was 100 mN . The temperaturestrain analyses were done in TMA mode from 30 to 190∘C190^{\circ} \mathrm{C} at 5∘Cmin−15^{\circ} \mathrm{Cmin}^{-1} on the same instrument. The sample dimensions were 50 mm×10 mm×1 mm50 \mathrm{~mm} \times 10 \mathrm{~mm} \times 1 \mathrm{~mm} and the applied stress was 250 mN .

SEM analyses of the samples were performed using a JEOL WINSEM-6400 electron microscope. The probe size was 114.98 nm , the probe current 0.02 nA , the noise reduction 64 Fr and the AC voltage 5.0 keV . The fractured surfaces of the samples were coated with gold by an electrode deposition method to impart electrical conductivity before recording the SEM micrographs. Stress relaxation measurements were done on a Hounsfield H5KS universal tester. Samples with a width of 10 mm and a thickness of 1 mm were stretched to 20%20 \% strain, after which the change in stress was monitored for 100 min . The FTIR analyses were done using an ATR in a Perkin Elmer Spectrum 100 spectrometer. Eight scans per sample were done at a resolution of 4 cm−14 \mathrm{~cm}^{-1}.

The solid-state NMR spectra were acquired on a Varian VNMRS 500 MHz two-channel spectrometer using 6 mm zirconia rotors and a 6 mm Chemagnetics cmT3HX{ }^{\mathrm{cm}} \mathrm{T} 3 \mathrm{HX} probe. All cross-polarization (CP) spectra were recorded at ambient temperature with proton decoupling, a 3.75μ s90∘3.75 \mu \mathrm{~s} 90^{\circ} pulse, and a recycle delay of 5 s . The power parameters were

optimised for the Hartmann-Hahn match. The contact times for cross-polarization was 1 ms . The free induction decay was 640 points, Fourier transformed with 1280 points and 0.5 Hz line broadening. Magic-angle-spinning (MAS) was performed at 7 kHz , and Adamantane was used as an external chemical shift standard. For solution NMR experiments, the samples were dissolved in 0.6 mL deuterated trifluoroacetic acid ( dd-TFAA) ( 6wt.%6 \mathrm{wt} . \% ). 13C{ }^{13} \mathrm{C} NMR experiments were performed at 150 MHz on a 5 mm PFG switchable/ broadband probe ( 1H−19 F,15 N−31P1 \mathrm{H}-19 \mathrm{~F}, 15 \mathrm{~N}-31 \mathrm{P} ) on a Varian UNITY INOVA 600MHz{ }^{\text {INOVA }} 600 \mathrm{MHz} spectrometer at 25∘C25^{\circ} \mathrm{C}.

3. Results and discussion

Fig. 1 shows a large magnification SEM photo of the fracture surface of the 80/20 w/w PA12/wax blend. The immiscibility of the PA12/wax is clearly visible in this photo. The wax is clearly the dispersed phase and seems to form crystallites of 0.5−1.0μ m0.5-1.0 \mu \mathrm{~m} diameter (some indicated with white arrows) in the semi-crystalline PA12 matrix. In some cases, it looks as if there is a weak interaction between the wax and the PA12 (open cavities around wax crystals), but in other positions on the photo there is clear evidence of strong interaction between the wax crystals and the PA12 matrix. The big difference in solubility parameters between the two components should allow an easy and effective removal of the wax phase from the blend, unless the interaction between the grafted wax and the PA12 is stronger than the interaction between the wax and the solvent. Soxhlet extractions of the wax from the 80/20 w/w\mathrm{w} / \mathrm{w} PA12/wax sample were performed in n-heptane, which is a good solvent for the wax. A mass loss of only 4%4 \% was observed, which is much lower than the amount of wax mixed into the sample. The SEM photo of the sample after extraction looked similar to the one presented in Fig. 1. This confirms that most of the wax remained in the PA12 matrix, even after extraction.

To further investigate the cause of possible interaction between the grafted maleic anhydride (MA) functional groups on the wax chains and the amide functional groups

Fig. 1. SEM picture of fractured surface of 80/20w/w80 / 20 \mathrm{w} / \mathrm{w} PA12/wax blend.
in PA12, FTIR analyses were done on the pure PA12 and wax samples, as well as on the 80/20 w/w PA12/wax blend (Fig. 2). In the spectra of both PA12 and 80/20 w/w PA12/ wax the −NH2-\mathrm{NH}_{2} bending and stretching vibrations are clearly visible at 1633 and 3285 cm−13285 \mathrm{~cm}^{-1}. For all three samples the CH2\mathrm{CH}_{2} and CH3\mathrm{CH}_{3} stretching vibrations are visible at 2850 and 2919 cm−12919 \mathrm{~cm}^{-1}, and the bending vibration is visible at 1460 cm−11460 \mathrm{~cm}^{-1}. The C−H\mathrm{C}-\mathrm{H} rocking vibration is visible for all three samples at 720 cm−1720 \mathrm{~cm}^{-1}. The amide −C=O-\mathrm{C}=\mathrm{O} vibration is visible at 1730 cm−11730 \mathrm{~cm}^{-1} for PA12 and the blend. In the spectrum of the MA-grafted wax, the presence of the absorption peak at 1780 cm−11780 \mathrm{~cm}^{-1} is due to the symmetric stretching mode of the carbonyl groups in succinic anhydride rings, which is evidence of the grafting of maleic anhydride onto the wax [18-23]. However, the weak asymmetric stretching mode of the carbonyl group is not clearly visible, and this may be attributed to the appreciable hydrolysis of succinic anhydride functional groups. The strong absorption band at 1707 cm−11707 \mathrm{~cm}^{-1}, which is characteristic of carboxylic acids and esters, is assigned to the hydrolyzed succinic anhydride groups present in the modified wax [19-23]. This peak is not visible in the spectrum of the 80/20 w/w PA12/wax blend, which is due to the conversion of the hydrolyzed succinic anhydrides to anhydrides during the process of melt blending and compression moulding of the PA12 and grafted wax at 200∘C200^{\circ} \mathrm{C} [19]. To confirm this conversion in the functionalized wax used in this project, a sample of unblended wax was heated for 24 h . The FTIR spectrum of the heat-treated sample (not presented here) clearly shows a reduction in the intensity of the 1707 cm−11707 \mathrm{~cm}^{-1} peak, and the formation of a −C=O-\mathrm{C}=\mathrm{O} vibration peak around 1860 cm−11860 \mathrm{~cm}^{-1}, which is not visible in the FTIR spectrum of the wax in Fig. 2. The absence of absorption peaks at 698 cm−1(C=C698 \mathrm{~cm}^{-1}(\mathrm{C}=\mathrm{C} bond of MA) and in the range of 1500−1670 cm−11500-1670 \mathrm{~cm}^{-1}, as well as the weak absorption band in the range of 3020−3160 cm−13020-3160 \mathrm{~cm}^{-1}, clearly proves that there was no free maleic anhydride in the wax [21,24]. The FTIR spectrum (not presented here) of the 80/20 w/w PA12/wax sample after solvent extraction was exactly the same than that of the sample before extraction. This further confirms the strong interaction

Fig. 2. FTIR transmittance spectra of PA12, wax and an 80/20 w/w PA12/wax blend.

between PA12 and MA-grafted wax, which should prevent leakage of the wax from the PA12 matrix.

Solution and solid-state NMR was used to try and further clarify these possible reactions. Fig. 3 shows a comparison of the 13C{ }^{13} \mathrm{C} solution NMR spectra of PA12 and the 80/20 w/w PA12/wax blend. Upon dissolution of the blend in dd-TFAA, a large part of the wax precipitated out of solution. This is reasonable since TFAA is a poor solvent for wax. By inspection, only slight differences were observed between the spectra of pure PA12 and the 80/20 w/w PA12/ wax blend. This was mainly in the region of the main chain methylene units of PA12. This is visible in the inset box and is indicated through the loss of resolution observed for the peaks in the 29−30ppm29-30 \mathrm{ppm} range in the 80/20w/w80 / 20 \mathrm{w} / \mathrm{w} PA12/wax blend, as compared to the same region in pure PA12. It is thought to be related to the overlapping methylene backbone peaks of the wax generally observed around 29−30ppm29-30 \mathrm{ppm}, as is the case for polyethylene [25]. If this is indeed so, the wax should have some association with the PA12 to have remained in solution and to be observable on the spectrum. The interaction between the -NH- of the amide linkage of PA12 and the maleic anhydride on the backbone of the wax could not be established since the concentrations were too low.

Fig. 4 represents the 13C{ }^{13} \mathrm{C} CP-MAS spectra of the pure wax, the MA modified wax, the 80/20 w/w PA12/wax blend and pure PA12. The solid-state NMR results confirm the presence of the γ\gamma and α\alpha crystal forms of PA12, as described in the paper by Mathias and Johnson [26], and are in agreement with the two melting points observed in DSC. It is clear from the ratio of the intensities of the main peaks on the 13C{ }^{13} \mathrm{C} CP-MAS spectrum that the thermodynamically stable γ\gamma-form is present in higher concentrations. This is also evidenced by the chemical shift positions associated with the γ\gamma-form, where the tallest peak at 32.8 ppm corresponds to the methylene units in the all-trans conformation and the upfield peak at 30.4 ppm with the methylene groups in the gauche conformation. The carbon α\alpha to the carbonyl appears at around 36.6 ppm , while the carbon adjacent to the nitrogen of the amide linkage

Fig. 3. Comparison of the 13C{ }^{13} \mathrm{C} solution NMR spectra of PA12 and the 80/20 w/w\mathrm{w} / \mathrm{w} PA12/wax blend.

Fig. 4. 13C{ }^{13} \mathrm{C} CP-MAS spectra of the pure wax, the MA modified wax, the 80/20 w/w PA12/wax blend, and pure PA12.
appears at 39.8 ppm . Some notable differences can be observed between the intensities and sharpness of the peaks in the compared spectra. The increased intensity of the peak at 32.8 ppm for the 80/20 w/w PA12/wax blend, when compared to that for PA12, can be ascribed to the alltrans methylene units of the wax incorporated into the PA12 matrix (also seen in the inset box where the main wax peak coincides with the main PA12 peak). The total signal from 20 to 40 ppm is significantly broadened in the spectrum of the 80/20 w/w PA12/wax blend, and the peak at 36.6 ppm is less resolved. This broadening and loss in resolution can generally be attributed to increased disorder in a system and thus, in our case, loss of rigidity of the PA12 matrix in the presence of the wax.

The DSC results of the PA12/wax blends are summarized in Table 1, and the heating curves are shown in Fig. 5. The DSC curve of the pure wax shows a broad melting peak between 50 and 110∘C110^{\circ} \mathrm{C} with more than one peak maximum. The strongest peak maximum appears at 76∘C76^{\circ} \mathrm{C}. This is the result of the melting of different molecular weight fractions of the wax [27]. Paraffin wax also crystallizes in various structures [28]. Pure PA12 shows two well resolved peaks. It usually crystallizes in the γ\gamma-form from the melt at atmospheric pressure [29]. This crystal structure is believed to be hexagonal or monoclinic with two strong WAXS reflections corresponding to d-spacings of about 0.42 and 1.59 nm . However, certain conditions can lead to the formation of the monoclinic α\alpha-structure which is characterized by d-spacings of about 0.37 and 0.45 nm [30]. A coexistence of the two crystal structures has also been observed in PA12 films, verified by WAXS [31]. In our case, the peak which belongs to the γ\gamma-form lies at 178.5∘C178.5^{\circ} \mathrm{C}, and the other one at 170∘C170^{\circ} \mathrm{C}.

From Fig. 5 it is clear that the components are immiscible due to the different chemical compositions and crystalline structures of PA12 and wax. The melting temperature of PA12 slightly decreases with an increase in wax content. Since Fischer-Tropsch paraffin wax is a very hard wax as a result of its high crystallinity, it may act similarly to inorganic fillers. It is known that inorganic

Table 1
DSC melting peak temperatures and enthalpies for the investigated samples.

TmT_{\mathrm{m}} - average melting peak temperature; ΔHm\Delta H_{\mathrm{m}} - average melting enthalpy; ΔHmtheor \Delta H_{\mathrm{m}}^{\text {theor }} - theoretically calculated enthalpy from Eq. (1) (values in brackets are standard deviations).
fillers reduce polymer chain mobility which causes crystallization into thinner lamellae. The small wax crystals, that also strongly interact with PA12, probably play a similar role. The decrease in the melting point of PA12 is then the result of the formation of smaller PA12 crystallites in the presence of the wax. The melting enthalpies of PA12 and wax were evaluated separately from the areas under their respective melting peaks. The melting enthalpy related to the wax portion increased with an increase in wax content, while that related to the PA portion decreased (Table 1). The experimental values are close to the theoretically expected values. The theoretically expected values (ΔHmtheor )\left(\Delta H_{\mathrm{m}}^{\text {theor }}\right) were calculated from equation (1):
ΔHmtheor =wiΔHm\Delta H_{\mathrm{m}}^{\text {theor }}=w_{i} \Delta H_{\mathrm{m}}
where wiw_{i} is the mass fraction of PA12 or wax, and ΔHm\Delta H_{\mathrm{m}} is the melting enthalpy of pure PA12 or wax. This confirms the SEM observations that the wax crystals were fairly homogeneously dispersed in the PA12 matrix, and that there was insignificant wax leakage during sample preparation.

Gordon [32], in his discussion of the glass transition behaviour of polyamides, observed appearance, disappearance and shifting of glass transition peaks to higher temperatures, depending on the thermal treatment of the polymer. He explained these phenomena in terms of the slow formation of a hydrogen-bonded network in the amorphous regions of the polymer. It is the disruption of

Fig. 5. DSC heating curves for PA12, wax and different PA12/wax blends.
this network that is normally considered to be the glass transition in nylons. The network is slow in re-forming because of problems involved in matching up potential hydrogen-bonding sites, which are, of course, distributed at intervals along the polymer chain. The temperature at which the network is disrupted is apparently dependent, not so much on the ratio of bonding to nonbonding sites, as on the temperature at which it was formed. The mechanical damping factor curves for PA12 and its blends with the wax are presented in Fig. 6. Pure PA12 shows two transitions at 50 and 141∘C141^{\circ} \mathrm{C}. It seems as if the first transition may be the typical PA12 glass transition. The position of the first transition does not change significantly when PA12 is blended with the wax, or with increasing wax content. It is, however, interesting that the position of the second transition moves to significantly lower temperatures when wax is present and with increasing wax content. If we assume that the disruption of a hydrogen-bonded network in the PA12 gives rise to the second transition, then it is possible that increasing amounts of wax, which strongly interacted with the PA12 amide groups, reduced the hydrogen bonding and the accompanying rigidity in the PA12. The number of hydrogen bonds to be disrupted will also be reduced, which will give rise to a reduction in transition temperature of the second transition until, at fairly high wax contents, there is a strong overlap of the two transitions in the temperature range of the first transition.

Fig. 6. Heat dissipation as function of temperature for PA12 and the different PA12/wax blends.

Fig. 7. Tensile elongation as function of temperature under a fixed load of 250 mN for PA12 and the different PA12/wax blends.

The thermomechanical elongation of the different samples is presented in Fig. 7, and the results clearly indicate the loss in rigidity of the samples with increasing wax content. The only sensible explanation for this observation is the interruption of the hydrogen bonding in PA12 as a result of the interaction between the wax MA groups and the PA12 amide groups. Under the applied experimental conditions, the extensibility of the 20 and 30%30 \% wax containing samples has almost doubled compared to that of pure PA12. It is further interesting to see that there is not much difference between the behaviour of the 20 and 30% wax containing samples. The DMA damping factor curves (Fig. 6) show a similar trend. 20% MA-grafted wax blended with PA12 is therefore sufficient to almost completely disrupt the hydrogen bonding between the MA12 chains.

The stress relaxation results in Fig. 8 further confirm that the PA12 chain mobility was significantly influenced by the presence of wax in the polymer structure. Both PA12 and its blend samples almost immediately relaxed after the initial application of stress to obtain the set strain. Because of hydrogen bonding, which immobilizes the PA12 chains,

Fig. 8. Tensile stress relaxation curves at room temperature for PA12 and the different PA12/wax blends.

Fig. 9. TGA curves for PA12 and the different PA12/wax blends.
it is expected that pure PA12 will show a low level of stress relaxation. In our case, the blend samples relaxed much more than the pure PA12, indicating the disturbance of the hydrogen bonding in PA12 and leading to a much higher level of stress relaxation.

The TGA curves in Fig. 9 show that both PA12 and its wax blends completely decompose in a nitrogen atmosphere and that there is no char formation. The pure wax starts decomposing at about 190∘C190^{\circ} \mathrm{C}, which is much lower than the 400∘C400^{\circ} \mathrm{C} for pure PA12. The TGA curves for the blends show that the wax decomposes first, followed by the decomposition of the PA12. There is a strong relationship between the amount of wax initially mixed into the sample and the percentage mass loss at the end of the first decomposition step, which also indicates a fair distribution of the wax in the PA12 matrix and insignificant leakage of the wax during sample preparation. The onset temperatures of wax decomposition in the blends are also much higher than that for pure wax. The most probable reason for this is that the interaction between the wax MA groups and the PA12 amide groups reduces the wax chain mobility, which results in reduced free radical transfer during the degradation process. It is interesting that the presence of wax has very little influence on the PA12 degradation, despite the fact that all the other results indicate increased chain mobility of PA12 in the presence of wax. The reason for this is probably the stabilization of the wax itself through its interaction with PA12, because the end of the wax degradation is at about the same temperature as the start of the PA12 degradation (Fig. 9).

4. Conclusions

The morphology, thermal and thermo-mechanical properties of polyamide 12/maleic anhydride grafted wax blends, as well as possible interactions between PA12 and the functionalized wax, were investigated. Although the initial idea was to investigate these blends as potential phase change materials for thermal energy storage, it was not practically possible to prepare homogeneous blends containing more than 30wt.%30 \mathrm{wt} . \% wax. The results of SEM, FTIR and NMR analyses of the samples pointed to a probability

of hydrogen bonding between the anhydride groups on the wax and the amide groups in PA12, although such interaction could not be conclusively established. However, the influence of wax content on the mechanical and thermomechanical properties, as well as the thermal degradation behaviour, of the samples relates strongly to the existence of such an interaction. The DSC results confirmed that neither the crystallization behaviour of the wax nor that of the PA12 were influenced by blending the two components, that the wax crystals were homogeneously distributed through the PA12 matrix, and that there was no apparent leakage of the wax from the PA12 matrix during processing.

Acknowledgements

The South African National Research Foundation (ISTA82433 and ICD2006060100008), the Scientific Grant Agency of the Ministry of Education of Slovak Republic (VEGA 2/0063/09), and the University of the Free State are acknowledged for financial support.

References

[1] R. Zalba, J.M. Marin, L.F. Cabeza, H. Mehling, Applied Thermal Engineering 23 (2003) 251-283.
[2] A.M. Khudhair, M.M. Farid, Energy Conversion and Management 45 (2004) 263-275.
[3] M. Kenisarin, K. Mahkamov, Renewable and Sustainable Energy Reviews 11 (2007) 1913-1965.
[4] Y. Hong, G. Xin-Shi, Solar Energy Materials and Solar Cells 64 (2000) 37−4437-44.
[5] F. Asinger, Paraffins: Chemistry and Technology. Pergamon Press, 1967.
[6] RUBTHERM GmbH website.
[7] Y. Cai, Y. Hu, L. Song, Y. Tang, R. Yang, Y. Zhang, Z. Chen, W. Fan, Journal of Applied Polymer Science 99 (2006) 1320-1327.
[8] M. Xiao, B. Feng, K. Gong, Energy Conversion and Managament 43 (2002) 103-108.
[9] S. Geschwander, P. Schossig, H.M. Henning, Solar Energy Materials and Solar Cells 89 (2-3) (2005) 307-315.
[10] S. Peng, A. Fuchs, R.A. Wirtz, Journal of Applied Polymer Science 93 (2004) 1240-1251.
[11] H. Inaba, P. Tu, International Journal of Heat and Mass Transfer 32 (1997) 307-312.
[12] I. Krupa, G. Miková, A.S. Luyt, European Polymer Journal 43 (2007) 4695-4705.
[13] I. Krupa, G. Miková, A.S. Luyt, European Polymer Journal 43 (2007) 895-907.
[14] X. Py, R. Olives, R.F. Mauran, International Journal of Heat and Mass Transfer 44 (14) (2001) 2727-2737.
[15] M. Xiao, B. Feng, K. Gong, Solar Energy Materials and Solar Cells 69 (2001) 293-297.
[16] J.K.W. Sandler, S. Pegela, M. Cadek, F. Gojny, M. van Es, J. Lohmar, W. J. Blau, K. Schulte, A.H. Windle, M.S.P. Shaffer, Polymer 45 (2004) 2001-2015.
[17] I. Novák, I. Krupa, A.S. Luyt, Journal of Applied Polymer Science 93 (2) (2004) 662-668.
[18] W. Qiu, T. Endo, T. Hirotsu, European Polymer Journal 41 (2005) 1979-1984.
[19] S.H.P. Bettini, J.A.M. Agnelli, Polymer Testing 19 (2000) 3-15.
[20] Z. Aimin, L. Chao, European Polymer Journal 39 (2003) 1291-1295.
[21] C. Nakason, A. Kaesaman, P. Supasanthitikul, Polymer Testing 23 (2004) 35−4135-41.
[22] C. Li, Y. Zhang, Y. Zhang, Polymer Testing 22 (2003) 191-195.
[23] A.I. Vicente, J. Campos, J. Moura Bordado, M. Rosario Ribeiro, Reactive and Functional Polymers 68 (2008) 519-526.
[24] H. Krump, P. Alexy, A.S. Luyt, Polymer Testing 24 (2005) 129-135.
[25] M. De Pooter, P.R. Smith, K.K. Dohrer, K.F. Bennett, M.D. Meadows, C.G. Smith, Journal of Applied Polym Science 42 (1991) 399−408399-408.
[26] L.J. Mathias, C.G. Johnson, Macromolecules 24 (1991) 6114-6122.
[27] A.S. Luyt, I. Krupa, Thermochimica Acta 467 (2008) 117-120.
[28] A. Genovese, G. Amarasinghe, M. Glewis, D. Mainwaring, R.A. Shanks, Thermochimica Acta 443 (2006) 235-244.
[29] K. Inoue, S. Hoshino, Journal of Polymer Science Part B: Polymer Physics 11 (1973) 1077-1089.
[30] T. Ishikawa, S. Nagai, Journal of Polymer Science Part B: Polymer Physics 15 (1977) 1315-1317.
[31] T. Ishikawa, S. Nagai, Journal of Polymer Science Part B: Polymer Physics 18 (1980) 291-299.
[32] G.A. Gordon, Journal of Polymer Science Part B: Polymer Physics 9 (1971) 1693-1702.