SiO2 composites

SiO2 composites

European Polymer Journal 48 (2012) 803–810 Contents lists available at SciVerse ScienceDirect European Polymer Journal journal homepage: www.elsevie...

1MB Sizes 3 Downloads 22 Views

European Polymer Journal 48 (2012) 803–810

Contents lists available at SciVerse ScienceDirect

European Polymer Journal journal homepage:

Confinement effect of SiO2 framework on phase change of PEG in shape-stabilized PEG/SiO2 composites Huazhe Yang a,b, Lili Feng a,c, Chongyun Wang a, Wei Zhao a, Xingguo Li a,⇑ a Beijing National Laboratory of Molecular Sciences, State Key Laboratory of Rare Earth Materials Chemistry and Applications, College of Chemistry and Molecular Engineering, Peking University, Beijing 100871, PR China b Department of Biophysics, College of Basic Medical Sciences, China Medical University, Shenyang, Liaoning 110001, PR China c Key Laboratory of Urban Stormwater System and Water Environment (Beijing University of Civil Engineering and Architecture), Ministry of Education, Beijing 100044, PR China

a r t i c l e

i n f o

Article history: Received 8 August 2011 Received in revised form 4 December 2011 Accepted 24 January 2012 Available online 9 February 2012 Keywords: Polyethylene glycol Shape-stabilized phase change materials Confinement effect PEG/SiO2 composites

a b s t r a c t PEG/SiO2 shape-stabilized phase change materials with various mass fractions and molecular weights of PEG were prepared by the sol–gel method. Polyethylene glycol (PEG) and tetraethyl orthosilicate (TEOS) were chosen as the phase change substance and the silica framework precursor, respectively. The as-prepared samples were characterized by X-ray diffraction (XRD), differential scanning calorimetry (DSC), Fourier transform infrared (FTIR) spectroscopy and scanning electron microscope (SEM) techniques. It is shown that the silica framework strongly confined the crystallization of PEG. The crystallinity and thermodynamic performance of the composites were undesirable for PEG with molecular weight of 1500 even when the PEG content reached 80 wt%. The crystallinity and thermodynamic performances of the PEG/SiO2 composites first decline then improve with the increase of the PEG molecular weights, owing to the different confinement behaviors of the silica framework. Finally, we investigated the phase change mechanism of the PEG/SiO2 composites under the different confinement of the silica framework. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Phase change materials (PCMs), a kind of materials utilizing latent heat during the phase change process, are very attractive for their applications in thermal energy storage and temperature control [1–3]. Nowadays, PCMs have attracted more and more attention as a promising candidate to deal with the energy crisis [4–6]. However, the development of PCMs is restricted as most traditional PCMs, such as inorganic salt hydrates and organic acids, face problems with corrosion, chemical stability, leakage and thermal conductivity, among others. Increasingly, shape-stabilized composite PCMs are being investigated to increase their viable applications [7–12]. Polyethylene glycol (PEG), a non-toxic and non-corrosive polymer, is considered as a promising PCM owing to ⇑ Corresponding author. Tel.: +86 10 62765930. E-mail address: [email protected] (X. Li). 0014-3057/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2012.01.016

its relatively larger heat of fusion, suitable melting point and chemical stability. However, as a solid–liquid PCM, the leakage of PEG during its phase transition cannot be neglected; in addition, PEG has low thermal conductivity. Therefore, proper composite techniques must be adopted to overcome these obstacles. The PEG/SiO2 composite, a novel shape-stabilized composite PCM, has attracted the interest of many researchers [13–15]. Grandi et al. [13] prepared PEG/SiO2 composites with different molecular weight (Mw) of PEG (PEG200/SiO2 and PEG600/SiO2) as shape-stabilized phase change materials by the sol–gel method, and analyzed the distribution of organic components in the inorganic ones as well as the thermal stability of the composite. Wang et al. [14] prepared PEG10000/SiO2 composites by a blending method and discussed the structure and thermal conductivity of the composites. Jiang et al. [15] observed the confined crystallization behavior of PEG19000 in silica networks. However, to the best of our knowledge, there is no systematic study about the confinement effect


H. Yang et al. / European Polymer Journal 48 (2012) 803–810

of silica frameworks on phase change of PEG with different Mw and different mass fractions in the composites. In order to further explore the PEG/SiO2 composites for thermal energy storage, it is necessary to investigate not only the crystallization and thermal behavior of the composites, but also the inner mechanism of the phase change process of PEG in PEG/SiO2 composites and the interactions between silica framework and PEG under different conditions. In this paper, PEG/SiO2 composites with various mass fractions and Mw of PEG were prepared by the sol–gel method, and the confinement effect of silica frameworks on the shape-stabilization, crystallinity and thermodynamic properties of the composites was investigated. Furthermore, the schematic phase change models of PEG/SiO2 composites were proposed, which pave the way for further investigation on PEG/inorganic framework composites. 2. Experimental

Fig. 2. XRD patterns of PEG/SiO2 composites with different mass fractions of PEG: 1# 50%; 2# 60%; 3# 70%; 4# 80%.

2.1. Preparation of PEG/SiO2 composites The sol–gel method was adopted to prepare PEG/SiO2 composites through the hydrolysis of tetraethyl orthosilicate (TEOS). PEG1500 in separate amounts of 2.5, 3, 3.5, 4 and 4.5 g was dissolved in 30 ml of ethanol. Then, different amounts of TEOS (approximately equivalent to 2.5, 2, 1.5, and 1 g SiO2, respectively) were added (with PEG:SiO2 mass ratios of 50:50, 60:40, 70:30, 80:40 and 90:10, respectively) into the solution. Thereafter, a catalytic amount of HCl was added to adjust the pH to 1. The gels can be formed after stirring the solution for 4 h followed by aging at room temperature for over 1 week. Afterwards, the gels were heated at 40 °C in a water bath for 24 h. The solvent was evaporated in an oven for 3 days at 80 °C keeping PEG in the melted state and then the PEG/SiO2 composites were grinded with an agate motor. The flowchart of the procedure is shown in Fig. 1. PEG/SiO2 composites with different Mw (4000, 6000 and 10,000) of PEG were obtained with the same procedure. 2.2. Characterization X-ray diffraction (XRD) patterns were collected at a scanning rate of 4°/min on a DMAX 2400 Rigaku diffractometer with CuKa radiation and operating at 40 kV and 100 mA. The melting point and the phase change enthalpy were determined using a differential scanning calorimetry (DSC, Q100, TA, USA) in the range from 30 to 100 °C in a nitrogen atmosphere, and the scanning rate was at 10 °C/

Fig. 1. The experimental flowchart.

Fig. 3. DSC curves of samples 1#–4#.

min. Fourier transform infrared (FTIR, 8400, Shimadzu, Japan) spectroscopy and Hitachi S4800 (Hitachi, Japan) cold field emission scanning electron microscope (FE-SEM) techniques were adopted to investigate the structure, thermodynamic behaviors and morphology of the composites. 3. Results and discussion 3.1. PEG1500/SiO2 composites phase change materials Among the PEG1500/SiO2 composites with different mass fractions of PEG, a shape-stabilized composite can be obtained at 80 °C when the PEG content is within the range of 50–80 wt.%, while liquid PEG occurs outside of the silica framework if the content reaches 90 wt%. Fig. 2 shows XRD patterns of the stabilized composites, neat PEG1500 and silica derived from the hydrolysis of TEOS, where the mass fractions of PEG1500 in stabilized composites (samples 1#–4#) are 50%, 60%, 70% and 80%, respectively. In the patterns, only an amorphous state was found if the mass fraction of PEG1500 is lower than 70%, blunt diffraction

H. Yang et al. / European Polymer Journal 48 (2012) 803–810


Fig. 4. Schematic phase and structural transformation for PEG1500/SiO2 composites with different mass fractions: (a) interpenetrating network structure [25], PEG1500/SiO2 composites with mass fractions (b) much lower than 70%, (c) near to 70%, (d) 70%, (e) 80%, and (f) 90%.

peaks were seen when the mass fraction reached 70%, while the peaks became sharp if the content of PEG1500 increased to 80 wt%. According to the fitting of diffraction peaks and calculation of area ratio of crystallized section to amorphous section, crystallinity of samples 3#, 4# and neat PEG1500 was 5.1%, 15.4% and 50.8%, respectively. Therefore, the crystallinity of the composite was optimal in sample 4# among the series of PEG1500/SiO2 composites.

Thermodynamic properties of the samples were characterized by DSC measurement, as shown in Fig. 3. It should be noted that the phase change enthalpy of all composites is much lower than neat PEG1500 in spite of the high mass fraction of PEG (80 wt%). The presumed reason for this may be that most of the PEG chains were embedded in the silica network, which were confined by the silica framework and could not experience the phase change from the


H. Yang et al. / European Polymer Journal 48 (2012) 803–810

this confinement influence the thermodynamic behaviors of PEG/SiO2 composites? In fact, the thermodynamic and kinetic behavior of polymer such as polyethylene (PE) and PEG have been extensively investigated [16–24]. Early in the 1990s, Keller and his co-workers found that mesomorphic phase can play an important role in phase change. The mesomorphic phase is metastable for macroscopic systems, but stable for microscopic system with the size in the nanometer range, i.e. stability of phase is determined by the thickness of polymer crystallities. Furthermore, melting point of polymer is a function of crystal thickness, which can express as formula (1) [24]

  b T m ¼ T 0m 1  Lc Fig. 5. FTIR spectra of PEG/SiO2 composites: (a) PEG1500/SiO2 composite with mass fraction of PEG is 80%, (b) neat PEG1500 and (c) silica.

Fig. 6. XRD patterns of samples 4#–7#, and the molecular weight of PEG for them are 1500, 4000, 6000 and 10,000, respectively.

amorphous phase to the stable phase (crystal). Accordingly, the phase change of the composite from the crystal state to melting state during heating was hindered to a large extent. As a result, the enthalpy of sample 4# was much lower than that of neat PEG. In spite of the undesirable thermodynamics properties of the PEG1500/SiO2 composites, a remarkable phenomenon was found: there is no endothermic peak when the PEG mass fraction was under 70% (sample 1# and 2#), which was attributed to the loss of crystallizable PEG; a small endothermic peak occurred in sample 3#, while the peak shifted to a lower temperature in sample 4#. Sample 4# possessed a lower phase change temperature (Tm) and higher phase change enthalpy comparing with other composites, which agreed with the XRD results. Meanwhile, it is interesting that Tm of composite decreased with the increase of PEG content. In contrast, Tm of neat PEG is higher than all of the composites. The phenomenon can be attributed to the different confinement of silica framework on PEG1500 at different mass fractions. So how does


where Tm is melting temperature of polymer, T 0m is the equilibrium melting temperature, Lc is the lamella thickness and b  2cV c =DH0m , of which c is the surface free energy of the interface between the crystal and surrounding liquid, Vc is the molar volume of a crystallizable repeat unit, and DH0m is the enthalpy of melting for a 100% crystalline polymer. Therefore, the unique thermodynamic behavior in our experiment may firstly be attributed to the size confinement effect of SiO2 framework on the phase change of PEG, i.e. most of phase in sample 1#–4# were amorphous or mesomorphic owing to the confinement of silica framework. As a result, the phase change from crystallization state to melting state was inhibited during heating. In addition, SiO2 framework also inhibited the sliding motion of polymer chains outside the SiO2 framework, which thereafter inhibited the perfect crystallization of PEG. Combined with the results of XRD and DSC, we speculate the mechanism model of phase change and crystallization behaviors for PEG/SiO2 composites under various mass fractions of PEG1500, as is shown in Fig. 4. According to reference [25], the structure of PEG/SiO2 composite is interpenetrating network (Fig. 4(a)). Therefore, it is presumed that if the mass fraction of PEG1500 was much lower than 70% (Fig. 4(b)), most of PEG was thoroughly embedded inside the network of the silica framework. The gelation process occurred in liquid, PEG chains in the liquid were confined by the silica framework, which inhibited the thickening of thin lamellar polymer crystallites into stable crystal. Therefore, most PEG in the PEG/ SiO2 composites was amorphous phase or mesomorphic phase, and there was hardly any latent heat of phase change detected by DSC; With the increase of PEG content (approximate to 70 wt%), part of PEG chains were out of the silica framework and possessed higher energy than neat PEG because the association of PEG molecular chains of composite was same as PEG chains in amorphous phase or mesomorphic phase. The end of the chains outside of the framework was defined as the ‘‘active point’’ (Fig. 4(c)); When the mass fraction of PEG increased to 70%, some of the PEG chains were completely out of the framework. These chains, different from both the free neat PEG chains and fully confined chains, were confined by the active points or surface of framework. Accordingly, the association of PEG chains was different from chains confined inside the framework

H. Yang et al. / European Polymer Journal 48 (2012) 803–810


Fig. 7. DSC curves of the samples with different molecular weights: (a) samples 4#–7#, the molecular weight of PEG for them are 1500, 4000, 6000 and 10,000, respectively; (b) neat PEG with different molecular weight.

Table 1 Thermodynamic property of the samples with different molecular weights. Sample

Tc (°C)

Enthalpy (J/g)

4# 5# 6# 7# PEG1500 PEG4000 PEG6000 PEG10000

44.37 42.59 55.57 59.38 49.68 62.34 63.36 64.93

7.339 1.937 71.79 74.50 148.2 202.1 172.0 167.0

as well as the free one, i.e. the ‘‘part-confined state’’ (Fig. 4(d)). In this case, the sliding motion of polymer chain was restricted owing to the confinement of active points or surface of framework and thereby the growth to stable phase (crystal) of PEG was restricted. As a result, the sizedetermined chemical potential of the melting point decreased, and the phase change was prone to take place at lower temperature, and so the composites had a lower Tm; When the PEG mass fraction was 80%, as the amount of part-confined chains increased (Fig. 4(e)), the whole energy of the composite system increased and the instability of the composite system thereby increased. As a result, Tm declined further and the enthalpy increased; If the PEG content was increased to 90 wt% (Fig. 4(f)), some of the free PEG would be around the framework and the shape-stabilized composite could not be obtained at the drying temperature (higher than melting point of PEG1500). FTIR spectra of silica gel, neat PEG1500 and composite with 80 wt% of PEG are presented in Fig. 5.The broad peak located at 3420 cm1 was assigned to stretching vibration of –OH group which indicated the presence of hydroxyl ions and/or adsorbed water, and the peaks in range of 1600–1740 cm1 also lie in the vibration range of water; a peak at 2357 cm1 indicated stretching vibration of –CH–, and the –CH– bending vibration peaks were at 1465, 1400, 1385 and 1280 cm1; the vibration peaks at 2357 and 2337 cm1 were due to CO2 in the air; peaks at 1120 and 1049 cm1 lie in the vibration range of C–O–C; The

Si–O vibration peaks were at 2357 and 460 cm1. Compared with the vibration peaks of silica gel and neat PEG1500, there were no additional peaks for the composite. Therefore, it was a simple physical composite of silica gel and neat PEG1500; no chemical reactions occurred during the composite formation.

3.2. PEG/SiO2 composites with different Mw of PEG According to the above-mentioned results, when the Mw of PEG is low (1500), the enthalpy is also low owing to the confinement effect of silica gel (despite the fact that the mass fraction of PEG1500 is 80%). In order to further investigate the interaction between PEG and silica framework, PEG with same mass fraction (80%) but different Mw of 1500, 4000, 6000 and 10,000 were adopted to prepare the PEG/SiO2 composites (sample 4#–7#, respectively). From the XRD patterns shown in Fig. 6, crystallized PEG was found in all of the samples, and crystallinity of samples 4#–7# was 15.4%, 8.0%, 47.4% and 48.3%, respectively, i.e. the crystallinity of samples first declined then improved with the increase of PEG molecular weights. DSC curves of these composites and the corresponding neat PEG are shown in Fig. 7. In Fig. 7(a), all of the samples exhibit endothermic peak of crystallized PEG, and the corresponding curves of neat PEG are in Fig. 7(b). Comparison chart of Tm and enthalpy for composites and neat PEG is listed in Table 1. With the increase of Mw, Tm and enthalpies of composites first decreased then increased. Alternatively, Tm of neat PEG increased with the increase of Mw, while enthalpy first increased then decreased, i.e. the change of thermodynamic property for composites with different Mw was different from that of neat PEG. Particularly, the crystallization behavior of sample 5# was peculiar and it was reproduced for several times to confirm the experimental result. The reason may lie in the different confinement of silica framework on PEG with different chain length. Furthermore, compared with sample 4# and 5#, enthalpies of sample 6# and 7# increased significantly,


H. Yang et al. / European Polymer Journal 48 (2012) 803–810

Fig. 8. Schematic phase and structural transformations for PEG/SiO2 composites whose PEG molecular weights are: (a) 1500, (b) 4000, (c) 6000 and (d) 10,000, respectively.

Fig. 9. SEM images of PEG/SiO2 composites: (a) 4# (b) 5# (c) 6# and (d) 7#.

H. Yang et al. / European Polymer Journal 48 (2012) 803–810

which indicates that the increase of crystallized PEG, and the mass fraction of crystallized PEG in the composites can be calculated with the following formula.

C r ¼ DHf =ðvA  DH0f Þ  100%


where Cr is the mass fraction of unconfined PEG, DHf is the phase change enthalpy of the composites, vA is the percentage content of PEG in the composite and DH0f is the phase change enthalpy of neat PEG [15]. According to formula (2), mass fractions of unconfined PEG in samples 4#–7# are 6.19%, 2.37%, 52.17% and 78.52%, respectively. Considering that neat PEG cannot be fully crystallized, the change of crystallized PEG derived from DSC results was in accordance with the crystallinity results of XRD. In order to explain the ‘‘irregular’’ change of crystallintiy and thermodynamic results in composites with different Mw, the speculated PEG/SiO2 composites transformation model mentioned-above was applied, as shown in Fig. 8. When Mw is 1500 (Fig. 8(a)), chain length of PEG is short and most of PEG was confined in the silica framework. As mentioned-above, PEG inside silica framework was amorphous phase or mesomorphic phase, which can hardly contribute to the phase change latent heat. While PEG chains outside the framework were part-confined; Confinement ability of active point and surface of silica frame work determined the chemical potential of melting and thereby determined the thermodynamic behavior of the composite; with the increase of Mw to 4000 (Fig. 8(b)), chain lengths outside the framework lengthen. Owing to the fixed total mass of composites, the molar number of PEG decreased with the increase of molar mass, i.e. amorphous phase or mesomorphic phase of PEG dominated in the composite and the amount of part-confined PEG (small PEG crystals) decreased. As a result, crystallinity of composite worsened and the phase change enthalpy significantly decreased. In addition, the confinement effect of silica framework weakened further with the increase of chain length outside the framework, and the part-confined PEG tended to be freed from the confinement of active points at a lower temperature, i.e. Tm of PEG4000/SiO2 composite decreased; with an increase of Mw to 6000 (Fig. 8(c)), some of PEG molecular chains could penetrate the silica framework, which made the active points closer to the framework. As a result, the confinement effect was strengthened and Tm of the composite increased. Furthermore, the quantity of part-confined PEG6000 increased, since the interspaces of the framework were occupied by the molecular chain. Therefore, enthalpy of the PEG6000/ SiO2 composite increased remarkably; When Mw increased to 10,000 (Fig. 8(d)), the quantity of part-confined PEG10000 increased further and the crystallinty and enthalpy of PEG10000/SiO2 composite improved further. In other words, the differences of crystallinity and thermodynamic properties for composites with different Mw lies in the difference of confinement effect of silica framework on PEG molecular chains in PEG/SiO2 composites. In this way, phase competition (amorphous phase, mesomorphic phase, stable phase) for different composite is different and thereby the thermodynamic behaviors are different. Research on composites with other Mw is in progress.


SEM images of samples 4#–7# are shown in Fig. 9. As one kind of polymer, PEG tends to melt during the focus of electron gun of SEM and experience the melting-solidification process. In contrast, silica can maintain its rigid framework. Therefore, PEG and silica were distinguishable in spite of the amorphous phase of the composites. When Mw is 1500 (Fig. 9(a)), the silica framework was covered with PEG layer and the structure of silica framework could be seen because the PEG layer was thin. With an increase of Mw to 4000 (Fig. 9(b)), the silica framework was more obvious since the quantity of part-confined PEG was low. When Mw increased to 6000 (Fig. 9(c)) and 10,000 (Fig. 9(d)), the silica was covered with PEG again and an irregular morphology could be seen, since the PEG layer was thicker due to the longer molecular chain. The results are in accordance with assumptions based on the phase change and crystallization model of PEG in PEG/SiO2 composites. 4. Conclusion Phase change of PEG in the PEG/SiO2 composites with different mass fractions and the molecular weight of PEG were investigated. Different thermodynamic behaviors of the composites are observed owing to the confinement effect of silica framework. For PEG1500/SiO2 composites, the crystallinty and thermodynamic behavior of the composites are strongly confined by SiO2 framework; the increase of the Mw to 6000 or 10,000 can remarkably enhance the both the crystallinity and enthalpy of the composites. The different thermodynamic behaviors for different composites can be attributed to the phase competition and crystallization behavior of PEG under different confinement of SiO2. Acknowledgements The authors acknowledge MOST of China (Nos. 2009CB939902 and 2010CB631301) and NSFC (Nos. 20971009 and 51071003). References [1] Farid MM, Khudhair AM, Razack SAK, Al-Hallaj S. Energ Convers Manage 2004;45:1597–615. [2] Jegadheeswaran S, Pohekar SD. Renew Sust Energ Rev 2009;13:2225–44. [3] Sharma A, Tyagi VV, Chen CR, Buddhi D. Renew Sust Energ Rev 2009;13:318–45. [4] Khudhair AM, Fraid MM. Energ Convers Manage 2004;45:263–75. [5] Shukla A, Buddhi D, Sawhney RL. Renew Sust Energ Rev 2009;13:2119–25. [6] Benli H, Durmus A. Sol Energy 2009;83:2109–19. [7] Zeng RL, Wang X, Chen BJ, Zhang YP, Niu JL, Wang XC, et al. Appl Energ 2009;86:2661–70. [8] Diaconu BM, Varga S, Oliveira AC. Appl Energ 2010;87:620–8. [9] Cao Q, Liu PS. Eur Polym J 2006;42:2931–9. [10] Xi P, Gu XH, Cheng BW, Wang YF. Energ Convers Manage 2009;50:1522–8. [11] Zhan YZ, Zhu P, Zhao X, Wang B. Textile Dyeing Finishing J 2007;29:1–5. [12] Zhang J, Ding YM, Chen NY. Acid J Salt Lake Res 2006;14:9–13. [13] Grandi S, Magistris A, Mustarelli P, Quartarone E, Tomasi C, Meda L. J Non-Cryst Solids 2006;352:273–80. [14] Wang WL, Yang XX, Fang YT, Ding J. Appl Energ 2009;86:170–4. [15] Jiang S, Yu D, Ji X, An L, Jiang B. Polymer 2000;41:2041–6.


H. Yang et al. / European Polymer Journal 48 (2012) 803–810

[16] Rastogi S, Hikosaka M, Kawabata H, Keller A. Macromolecules 1991;24:6384–91. [17] Keller A, Hikosaka M, Rastogi S. Physica Scripta 1996;T66:243–7. [18] Keller A, Rastogi S, Hikosaka MH. Macromol Symp 1997;124:67–81. [19] Albrecht T, Armbruster S, Keller S, Strobl G. Macromolecules 2001;34:8456–67. [20] Albrecht T, Armbruster S, Keller S, Strobl G. Eur Phys J E 2001;6:234–7. [21] Hong K, Rastogi A, Strobl G. Macromolecules 2004;37:10174–9.

[22] Rastogi S, Höhne GWH, Keller A. Macromolecules 1999;32:8897–909. [23] Strobl G. Prog Polym Sci 2006;31:398–442. [24] Pfefferkon D, Kyeremateng SO, Busse K, Kammer HW, ThurnAlbrecht T, Kressler J. Macromolecules 2011;44:2953–63. [25] Du SF. Research of PEG/SiO2 form-stable phase change materials by Sol–Gel approach [D]. Chengdu: University of Electronic Science and Technology of China; 2007.