J Mater Sci (2013) 48:6753–6761 DOI 10.1007/s10853-013-7480-9
Effect of high energy ball milling on the structure and mechanical properties of cross-linked high density polyethylene E. Roumeli • K. M. Paraskevopoulos D. Bikiaris • K. Chrissafis
•
Received: 8 April 2013 / Accepted: 25 May 2013 / Published online: 5 June 2013 Ó Springer Science+Business Media New York 2013
Abstract The effects of high energy ball milling (HEBM) on the structure and some key-properties of crosslinked high density polyethylene (PEX) have been thoroughly examined with a combination of X-ray diffraction analysis, IR and Raman spectroscopy, differential scanning calorimetry, gel content measurements, and tensile properties tests. A structure–property relationship, which provides a reasonable explanation for the studied case has been developed based on the experimental results and their analysis. It is proposed that the HEBM provides some of the silane-grafted macromolecular chains, which have a specific orientation, with sufficient energy in order to crosslink and form small crystalline-like areas. The arrangement of chains in the ‘‘reformed’’ domains leads to a total increase of the overall crystallinity, but also a decrease of the crystalline size. The proposed model can also support the fact that by increasing the milling time, the overall crystallinity of PEX and some important mechanical properties are found to increase.
Introduction High energy ball milling (HEBM) is a relatively modern technique which is used mainly to incorporate nanoparticles into polymer matrices. Many studies have shown that E. Roumeli K. M. Paraskevopoulos K. Chrissafis (&) Solid State Physics Section, Physics Department, Aristotle University of Thessaloniki, 541424 Thessalonı´ki, Greece e-mail:
[email protected] D. Bikiaris Laboratory of Polymer Chemistry and Technology, Department of Chemistry, Aristotle University of Thessaloniki, 54124 Thessalonı´ki, Greece
HEBM prior to the final processing not only improves mixing and filler dispersion, but also leads to composites with better properties at lower filler content [1–6]. Very few studies, however, have focused on the effects of solidstate HEBM in the polymer matrix before the incorporation of fillers [7–10]. Polyethylene is considered probably the most widely consumed macromolecule today, and since it was discovered that some of its properties can be greatly enhanced by cross-linking, several technologies have been explored and industrially employed in order to produce superior polyethylene products. Cross-linked high density polyethylene (PEX) is found to have better physical properties and mechanical performance among other attractive features, including the higher upper use temperature [11–14]. Three methods are mainly employed to produce PEX using: peroxide, irradiation, and silane crosslinking agents. In each of these methods, a dense 3D network is always formed by the molecular chains of polyethylene. The degree of crosslinking directly affects the material’s physical properties. PEX has a superior performance compared to conventional high density polyethylene (HDPE) especially under stress in higher temperatures. It also has a higher stress crack resistance, higher impact strength, better chemical resistance, and elastic properties [11, 14–21]. Due to these properties PEX is used for the production of pipes appropriate of hot water supply. One of the most widely used methods to produce PEX products is the two-step silane grafting and moisture crosslinking method, which has been employed in this study [22–24]. This method involves silane grafting in polyethylene’s chains and their subsequent condensation in order to form a dense 3D network, in which the chains are linked with each other via siloxane bonds. Silane crosslinking is considered a solid state crosslinking method, in
123
6754
which the crosslinks are located mainly in the amorphous regions as water diffusion is faster and easier in them and therefore, the crystalline structure is maintained [25, 26]. Crosslinking of the amorphous regions of the polymer leads to a stronger material. Even though PEX is a known and profitable industrial product, its complex structure has not been thoroughly examined yet. Very few research articles have been dealing with the structural characteristics of PEX, and the subject of investigation has mainly been peroxide or irradiation crosslinked low density polyethylenes (LDPE) [16, 23, 27, 28]. The aim of the present study was the investigation of HEBM effects on the structure and properties of PEX. Studies of HEBM effects on high density polyethylene (HDPE) [7] have revealed that the macromolecular chain length and chain conformations remain unaffected, as evidenced by the absence of differences in the molecular weight distribution profiles. Other studies discuss polyethylene’s phase transformations due to HEBM [7, 8]. As far as the authors know, no previous literature reports have been made for the HEBM effects on PEX properties prepared by the silane method.
Experimental section Materials High density polyethylene grafted with vinyl trimethoxysilane (g-HDPE) was kindly supplied by Sioplas with a Mn of 28000, Mw 120200 and intrinsic viscocity 1.54 dl/g. A catalyst masterbatch containing the same HDPE along with dibutyltin dilaurate (DBTDL), internal lubricants, stabilizers, and various antioxidants was also supplied by Sioplas. This catalyst batch can accelerate the hydrolysis reaction of the grafted silyl groups with water to form silanols, the condensation of silanols to form siloxane bonds or both. High energy ball milling (HEBM) process In this study, mixtures of 95 parts of g-HDPE and 5 parts of catalyst masterbatch were physically mixed for 10 min, and inserted into a Retsch centrifugal ball mill (model S 100) for solid-state mixing. A cylindrical stainless steel jar of 50 ml with 6 steel balls of 10 mm diameter were used, and a rotation speed of 500 rpm was applied. The four studied materials were milled for 0, 2, 4, and 6 h, respectively. Melt mixing and crosslinked procedures The materials after ball milling were melt-mixed in a Haake-Buchler Rheomixer (model 600) with roller blades and
123
J Mater Sci (2013) 48:6753–6761
a mixing head with a volumetric capacity of 69 cm3. During the mixing period the melt temperature and torque were continuously recorded. For this case a 10 min mixing at 200 °C with a torque speed of 60 rpm were used. Following that, the prepared materials immediately after preparation were hot pressed using an Otto Weber, Type PW 30 hydraulic press connected with an Omron E5AX Temperature Controller, at a temperature of 180 ± 5 °C, in order to prepare films of 10–30 lm and 350–450 lm thickness appropriate for each type of measurement. The films were rapidly cooled by immersion in water at 25 °C. All prepared films were then exposed to a hot bath (90 °C water for 24 h) to complete the crosslinking process in the bulk of the polymer.
Characterization methods Gel content The gel fraction is a direct measure of the crosslinking degree of PEX and it has been determined according to ASTM D 2765 in a 8-h period Soxhlet extraction cycle with p-xylene at 140 °C. For these measurements, 200 mg of each material with average thickness less than 50 lm were microtomed and placed in a stainless steel fine wire mesh 100, previously weighted. After the extraction cycle, the sample was weighted in order to obtain the gel content according to the following relationship: %gel ¼
mgel 100 minitial
ð1Þ
where minitial and mgel correspond to the masses of the initial and extracted sample. X-ray diffraction X-ray diffraction patterns were collected using a watercooled Rigaku Ultima? diffractometer with CuKa radiation, a step size of 0.02° and a step time of 3 s, operating at 40 kV and 30 mA. Raman spectroscopy Raman spectroscopy measurements were performed using a Raman spectrometer (LabRAM HR; Horriba Jobin–Yvon Ltd, UK, with LABSPACE software) with a CCD camera detector. An excitation wavelength at 632.8 nm was provided by a He–Ne laser. The spectral range was 3500–700 cm-1 with confocal hole at 1000 lm, slit at 100 lm, grating at 1800, microscope objective 950, while exposure time and accumulation for measuring were 5 s and two times, respectively.
J Mater Sci (2013) 48:6753–6761
6755
Mechanical properties
Table 1 Mechanical properties of prepared PEX materials after HEBM for different times
Mechanical properties testings were performed on an Instron 3344 dynamometer, in accordance with ASTM D638 using a cross-head speed of 50 mm/min. Sheets of about 350–450 lm thickness were used, prepared as described previously. In order to measure the mechanical properties from these sheets, dumb-bell-shaped tensile test specimens (central portions 5 9 0.5 mm thick, 22 mm gauge length) were cut in a Wallace cutting press. At least five measurements were conducted for each sample, and the results were averaged to obtain a mean value. The values of Young’s modulus, tensile strength at yield and at break point and elongation at break were determined.
HEBM (h)
Young’s modulus (MPa)
Tensile strength at yield (MPa)
Tensile strength at break (MPa)
Elongation at break (%)
0
588 ± 17
19.8 ± 0.7
22.0 ± 2.1
385 ± 10
Differential scanning calorimetry (DSC) Differential scanning calorimetry (DSC) measurements have been performed in a DSC141 (Setaram). Temperature and energy calibrations of the instrument were performed for different heating rates, using the well-known melting temperatures and melting enthalpies of high-purity Zinc, Tin, and Indium supplied with the instrument. PEX samples of 6 mg were placed in aluminium sealed crucible, while an identical empty crucible was used as reference in each measurement. The samples were heated from ambient temperature (25 °C) to 220 °C in a 50 ml/min flow of N2. The samples were heated at 5 °C/min, held at 200 °C for 5 min, cooled to 60 °C with a cooling rate of 5 °C/min, and then heated again with the same heating rate. The second heating data were used for analysis in the study. Fourier transform infrared spectroscopy (FTIR) Fourier transform infrared spectroscopy (FTIR) in transmittance mode was also used to monitor the crosslinking process on the samples. The spectra were obtained with a Spectrum 1000 Perkin-Elmer spectrometer in the spectral area 4000–400 cm-1, with a resolution 2 cm-1 and 64 scans.
Results and discussion The mechanical properties of all PEX samples are presented in Table 1. As can be seen a progressive increase in all the measured properties with HEBM time was found for the studied polymer. PEX prepared without HEBM has Young’s Modulus 588 ± 17 MPa while after the application of HEBM Young’s modulus increases with milling time reaching an 152 MPa increase after 6 h of HEBM. A similar increase can be seen in tensile strength at yield as well as at break point, both of which were increased by 3.5
2
699 ± 12
22.2 ± 2.2
22.8 ± 3.6
407 ± 11
4
693 ± 18
22.3 ± 1.9
25.2 ± 4.2
500 ± 69
6
740 ± 9
23.3 ± 0.9
28.2 ± 4.9
544 ± 30
and 6.2 MPa, respectively, after 6 h of HEBM. The same trend with increasing milling time was also revealed for elongation at break, which was also increased with milling time from 385 to almost 544 %. This is an indication that ball milling has a positive effect on the preparation of PEX, and since a simple and reasonable explanation for this is not availiable yet, the materials were characterized using several complementary techniques in order to attempt to explain the observed behavior. The molecular weight increase after HEBM could be a reason for mechanical properties increase. However, as was found from a previous study [7], the molecular weight of HDPE remains unaffected by ball milling. The extent of crosslinking and crystalline properties of the final materials could be the main reasons for this behavior, since the materials with high crosslinking content and higher degree of crystallinity usually have higher strength. Results from gel content measurements, which are presented in Fig. 1, suggest that HEBM has only a small effect to the crosslinking reactions of PEX, since as the HEBM time increases a slight reduction of gel content was found. The changes that were detected are not very high (about 1.5 %) and therefore, the ball milling process, which is performed before crosslinking procedure, seems to have only a limited effect on the cross-liking of PEX. However, this small reduction of crosslinked content cannot explain the increased mechanical properties of the samples. Another factor that could lead to increased mechanical properties is the degree of crystallinity of the samples. The XRD patterns of all prepared samples have been stacked in Fig. 2 maintaining their original scale. It is well known that polyethylene obtains an orthorhombic crystal structure (space group Pnam) with different a, b, and c unit cell parameters. The diffraction patterns in Fig. 2 show typical peaks corresponding to (110), (200), and (020) crystallographic planes of polyethylene’s orthorhombic structure superimposed on the amorphous halo [28, 29]. Analysis of the data revealed that HEBM seems to affect the unit cell parameters and inter-planar distances of PEX (results presented in Table 2). Initially, a is notably reduced, but as milling time increases, a tends to increase and ultimately reaches the same level as the un-milled polymer. It seems
123
6756
J Mater Sci (2013) 48:6753–6761 Table 2 Unit cell parameters, inter-planar spacings and crystallinity values of all measured samples BM (h)
Unit cell parameters ˚) (A
d110 ˚) (A
d200 ˚) (A
˚ 3) V (A
Xc (%)
57.2
a
b
c
0
7.493
4.962
2.549
4.138
3.732
94.772
2
7.339
4.929
2.534
4.091
3.691
91.665
59
4
7.45
4.95
2.544
4.123
3.717
93.816
58.4
6
7.492
4.935
2.537
4.123
3.717
93.801
65.1
Fig. 1 Gel content of the prepared materials using different HEBM times
Fig. 3 Lamellar thickness values on [110] and [200] directions for all samples
Fig. 2 XRD patterns of um-milled (0 h) and milled polymers for 2, 4, and 6 h
that HEBM initially distorts the unit cell in that direction, but as milling time increases the a parameter ‘‘relaxes’’ and becomes almost the same as in the un-milled PEX. The other two unit cell parameters, b and c, present a similar behavior as they are initially reduced, then ‘‘relaxed’’ to values similar to the un-milled PEX, and finally for the 6 h milled material these parameters are reduced once again. The inter-planar spacing parameters d110 and d200 follow the same trend as the a parameter. They are initially reduced but then, as milling time increases, they become higher tending to reach the values of the un-milled polymer. The same observation can be made for the unit cell volume. In conclusion, reduced unit cells can be found in the 2 h milled polymer, while as milling time increases, the unit cell tends to enlarge and approaches its initial volume. From XRD analysis the size of crystalline domains (L) in two main crystallographic directions [110] and [200] was also calculated using Scherrer’s equation [30]: Bð2hÞ ¼
Kk L cos h
123
ð2Þ
where B is the integral breadth of the corresponding peak, K is a geometrical factor that for polyethylene is taken as 0.9 [28], and k is the wavelength of CuKa radiation (kCuKa = 1.5418) [30, 31]. Some important differences were revealed by the crystalline size analysis for the two main crystallographic planes that are illustrated in Fig. 3. A small reduction of crystalline size on the main crystallographic directions [110] and [200] is caused by 2 h of HEBM, while for higher milling times this reduction becomes much higher. It seems that even though the unit cell tends to approach the volume of un-milled PEX, the crystalline domains become even smaller. Crystallinity percent of the studied materials was obtained by fitting the XRD profiles at the area of 10° \ 2h \ 40° with Gaussian–Lorentzian cross product (Amplitude) curves and incorporating the fitting parameters in Eq. 3 [32], where the crystalline and amorphous peak areas are noted as Qcr and Qam, respectively. Xc % ¼
Qcr 100 Qcr þ Qam
ð3Þ
The calculated values are presented in Table 2 and the fitting process is illustrated in Fig. 4 for the un-milled polymer. By carefully observing the raw data it can be seen that the un-milled polymer seems to have higher amount of
J Mater Sci (2013) 48:6753–6761
Fig. 4 XRD profile fitting process using Gaussian–Lorentzian cross product curves for the calculation of crystallinity in un-milled PEX
amorphous content. The crystalline area and the total area are both larger than in the other patterns and therefore, their quotient is smaller. It seems that milling time leads to higher degrees of crystallinity, which reach almost 7.5 %. The higher crystallinity percentage in the milled samples suggests the formation of more crystalline domains in the bulk of the polymer. The increased crystallinity of HEBM samples can be explained also from the slightly reduced gel content of the samples, since linear PE macromolecules tend to crystallize more easily compared to crosslinked. Due to this reason high density polyethylene, which is consisted from linear macromolecules, has higher degree of crystallinity compared with low density polyethylene, which has a lot of branched and crosslinked macromolecules. The crystallinity percent was also calculated using Raman spectroscopy in order to verify the results. Raman spectra of all studied samples are presented in Fig. 5. Polyethylene’s spectrum has been known for a long time and mainly it consists of three regions [33, 34]. Assuming a two-phase model (one crystalline phase and one amorphous) the spectral data can be evaluated to estimate the crystalline content of each sample [34]. In the first region (I) the C–C skeletal stretching vibrations occur with two crystalline components, counting for in and out of phase vibrational modes at 1130 and 1062 cm-1, respectively, and one broader peak at 1080 cm-1 due to the amorphous content stretching vibrations. In region (II) the CH2 twisting vibrations can be observed resulting in one sharp peak at 1295 cm-1 due to the vibrations of the crystalline components of polyethylene, and a broad amorphous band with a maximum around 1303 cm-1. The 1295 cm-1 peak has been shown to be independent of conformational changes and has been proposed as a reliable internal intensity standard by several authors [34, 35], and in this study it was chosen as an internal standard as well. The third region (III), for which
6757
the two-phase model cannot be valid, consists of a well distinguished triplet at 1416, 1440, and 1460–1470 cm-1. The integral intensity of the 1416 cm-1 peak is considered proportional to the crystalline content of the sample [34, 36], and along with the 1440 cm-1 they are correlated to the split doublet of in and out of phase CH2 bending mode vibrations [37, 38]. This crystal field splitting is characteristic of the orthorhombic unit cell in which two molecular chains occupy the unit cell and their inter-chain interactions result in two components for the bending Raman and IR active peaks [32, 39, 40]. The 1460–1470 cm-1 broad peaks can be associated with the amorphous or an intermediate disordered phase of the system [41, 42]. The Raman spectra were used to estimate crystallinity percent following the Strobl and Hagedorn method [34]. A complete deconvolution to the spectral areas of interest was adopted, in which the subtracted baseline was a quadratic function, and then the corrected spectra were normalized assuming a unit area of the twisting vibration of CH2 at 1295 cm-1. Then the crystallinity was estimated by calculating the ratio of normalized integral intensities of the crystalline CH2 bending vibration component of each sample at 1416 cm-1 to the corresponding intensity of that peak of 100 % crystalline polyethylene [34, 43, 44]. The results are presented in Fig. 6 along with the corresponding XRD findings. They present the same trend as the XRD results and therefore, they verify that increasing HEBM time leads to PEX with more crystalline content. It seems that increasing the ball milling time results in an increase of the total crystallinity of the polymer. With differences reaching almost 8 % it is safe to conclude that a trend between milling time and overall crystallinity exists in this material. From careful consideration of all above discussed findings, it seems that in the case of PEX, HEBM leads to the formation of a material with higher degree of crystallinity as well as crystals that are initially smaller or distorted compared to the well-established lamellae. This ability is directly affected from milling time as by increasing milling time these new crystals become more, and the process provides sufficient amounts of energy in order for them to ‘‘relax’’ and be less distorted. The formation of new crystals can also be seen from the DSC curves (Fig. 7) of the g-HDPE, which contains silane groups, HDPE containing the condensation catalyst and their mixture after HEBM for 6 h. In this figure it can clearly be seen that the main melting peak is recorded for all samples at 130.5 °C. Furthermore, during HEBM some new crystal entities are formed which melt at higher temperatures, appearing as a shoulder in the end of the melting curve 3. The same behavior appears also for the samples that were ball milled for 2 and 4 h. The melting curves were deconvoluted by using the Fraser–Suzuki (FS)
123
6758
J Mater Sci (2013) 48:6753–6761
Fig. 5 Raman spectra of all samples Fig. 7 Melting curves of g-HDPE (1), catalyst (2), and their 6 h milled mixture (3)
Fig. 6 Crystallinity percent values for all studied samples as calculated using XRD (square points) and Raman spectra (circle points) data
function, which is often used for thermal analysis deconvolution processes and the results are presented in Fig. 8 [45–48]. The deconvolution revealed that the second peak is almost 3 °C higher than the main melting peak. The obvious second melting suggests that in the milled samples some different type of crystals has been formed, which cannot be observed in the un-milled polymer. HEBM has initially been introduced in the scientific community as mechanical alloying, since the mechanical mixing provides the blending materials with sufficient amounts of energy even at low temperatures so that they proceed to the formation of new equilibrium and nonequilibrium phases [49–51]. Several authors have reported the formation of chemical bonds between the polymer macromolecular chains during HEBM [3, 52–57]. In the present case, trimethoxy silane-grafted polyethylene and condensation catalyst were subjected to high energy mechanical milling in ambient temperature conditions. There is a possibility that in a small extend hydrolysis and condensation reactions took place in that step of the
123
Fig. 8 Deconvolution of the melting curve of the 6 h milled PEX using Fraser–Suzuki function
process resulting in siloxane bonds formation before meltmixing and curing. In the bulk of the polymer some silane grafted chains may be present parallel to each other conformation, as illustrated in Scheme 1. In such a case, when the silanol groups condensate to form a Si–O–Si network, the grafted chains come closer and therefore, can be considered as a small crystalline-like region. The proposed process is illustrated in Scheme 2. According to the proposed model, there is a possibility that some grafted chains of the amorphous regions have a parallel to each other conformation, and they are close enough so when they are provided with sufficient amounts of energy from HEBM they proceed to hydrolysis and condensation reactions. As several authors have mentioned before [49–54], ball milling can provide more than enough energy for the formation of covalent C–C bonds between the long and less mobile macromolecular chains in the crystalline regions. Moreover, during the past years experimental work and theoretical foundation have
J Mater Sci (2013) 48:6753–6761
6759
Scheme 1 Schematic representation of moisture crosslinking of PEX
Scheme 2 Model illustrating the structural changes during high energy ball milling of PEX
provided the necessary evidence that strong mechanochemical phenomena lead to covalent chemical bonds formation, which are activated by the presence of an external mechanical force [58]. Also, mechanically activated bond hydrolysis has found to selectively occur in some polymeric materials in the presence of water [58]. Therefore, the possibility of hydrolysis and condensation
reactions between the end groups of the smaller molecular chains of the amorphous regions exists. In such a case, the chains with the parallel conformation come even closer and they can be considered as crystalline-like. These small repeating units that are placed in the otherwise amorphous area constitute the newly formed crystals that count for the observed changes in PEX during HEBM. As milling time
123
6760
J Mater Sci (2013) 48:6753–6761
while the ‘‘relaxation’’ of unit cell parameters may be explained by the energy amounts that are continuously being transferred to the new crystal areas thus, allowing the chains to fold and take their most preferable places. Moreover, since the new crystals are formed in the amorphous areas next to the siloxane bonds, the presence of these bonds hinders heat transfer in these areas and therefore, their melting may appear slightly delayed and at higher temperatures, as was observed by DSC measurements.
Conclusions Fig. 9 FTIR spectra of un-milled PEX before (1) and after (2) curing and 6 h milled PEX before HEBM (3), after HEBM and before curing (4), and after curing (5)
and provided energy increase, the newly formed crystals use this energy to ‘‘relax’’ and form better folds or be arranged in a more ‘‘preferable’’ way. To support further the proposed model, FTIR measurements were performed to the samples before and after curing in water. In Fig. 9 the peaks corresponding to silane grafting are indicated by arrows, and they are due to the grafted trimethoxy silane groups at 800, 1090, and 1190 cm-1 [11]. These peaks appear in all the spectra even though in the cured samples they are less intense indicating that less un-crosslinked active groups have remained in the cured samples. The peak at 1020 cm-1 that is attributed to the Si–O–Si vibrations [11] can also be seen in the spectra of cured PEX indicating the formation of a siloxane network. By studying the spectra of un-milled PEX before and after curing (spectra 1 and 2 in Fig. 9), it is obvious that the crosslinking step occurs during the final step of curing in water. In the case of the 6 h milled PEX in Fig. 9 the comparison of the spectra before and after HEBM and after the final curing step suggests that in the milled samples many siloxane bonds have already been formed even before the final curing step. This finding confirms our proposed effect of HEBM on PEX, since it proves that crosslinking has occurred in some extend during the milling process. Therefore, as evidenced by FTIR and crystallinity measurements, the mechanochemically induced crosslinking that occurs in PEX during HEBM is directly related to the formation of new small crystal domains. The proposed model is based on a combination of all experimental findings. The new small crystals that are formed due to mechanochemically induced crosslinking, lead to an increase of the total crystallinity, but since they are much smaller than the well-shaped existing lamellae, the overall crystalline size should be smaller. The higher crystallinity explains the improved mechanical properties,
123
In the present study, high-energy ball-milling effects on the structure and properties of crosslinked high density polyethylene were examined. It was found that some keyproperties of PEX, such as crystallinity and mechanical properties are affected by the HEBM process. A structure– property relationship was suggested as a possible explanation for the different properties of the milled polymer. It was proposed that through the HEBM process, sufficient amounts of energy are supplied to the system, which are used to crosslink the polymer through the formation of siloxane bonds. In some cases, the silane grafted molecular chains that are crosslinked by HEBM have a parallel conformation to each other and therefore, they form new small crystals in the amorphous area of the bulk of the polymer. These newly formed smaller crystals result in an increase of the overall crystallinity and therefore, an enhancement of the mechanical properties. All the observed differences between the un-milled and milled materials, including unit cell parameters and lamellar thickness values, are in agreement with the proposed model. Furthermore, the melting trance of the 6-h milled polymer revealed a second melting peak that corresponds to the delayed melting of the new crystals as they are in the amorphous areas ‘‘protected’’ by the formed crosslinks. Acknowledgements The authors would like to thank Prof. Aldo Boccaccini of the University of Erlangen-Nurnberg for allowing the use of the Raman spectroscopy facilities available at Biometerials Department, and Mr. C. Dolle for performing these measurements. This study was financially supported by the Greek General Secretariat of Research and Development (09SYN-33-484).
References 1. Gorrasi G, Di Lieto R, Patimo G, De Pasquale S, Sorrentino A (2011) Polymer 52:1124. doi:10.1016/j.polymer.2011.01.008 2. Terife G, Narh KA (2011) Polym Compos 32:2101 3. Sorrentino A, Gorrasi G, Tortora M et al (2005) Polymer 46:1601. doi:10.1016/j.polymer.2004.12.018 4. Gorrasi G, Sarno M, Di Bartolomeo A, Sannino D, Ciambelli P, Vittoria V (2007) J Polym Sci B 45:597. doi:10.1002/polb.21070
J Mater Sci (2013) 48:6753–6761 5. Suryanarayana C (2001) Prog Mater Sci 46:1. doi:10.1016/ S0079-6425(99)00010-9 6. Jiang X, Drzal LT (2012) J Appl Polym Sci 124:525 7. Olmos D, Domı´nguez C, Castrillo PD, Gonzalez-Benito J (2009) Polymer 50:1732. doi:10.1016/j.polymer.2009.02.011 8. Ishida T (1994) J Mater Sci Lett 13:623. doi:10.1007/bf00271215 9. Castricum HL, Yang H, Bakker H, Van Deursen JH (1997) In: Fiorani D, Magini M (eds) Synthesis and properties of mechanically alloyed and nanocrystalline materials, pts 1 and 2 Ismanam-96. Trans Tech Publications Ltd, Stafa-Zurich 10. Wu H, Liang M, Lu C (2012) Thermochim Acta 545:148. doi: 10.1016/j.tca.2012.07.008 11. Oliveira GL, Costa MF (2010) Mater Sci Eng A 527:4593. doi: 10.1016/j.msea.2010.03.102 12. Wang Z, Hu Y, Gui Z, Zong R (2003) Polym Test 22:533. doi: 10.1016/s0142-9418(02)00149-6 13. Ritums JE, Mattozzi A, Gedde UW, Hedenqvist MS, Bergman G, Palmlo¨f M (2006) J Polym Sci B 44:641. doi:10.1002/polb.20729 14. Celina M, George GA (1995) Polym Degrad Stab 48:297. doi: 10.1016/0141-3910(95)00053-O 15. Venkatraman S, Kleiner L (1989) Adv Polym Technol 9:265. doi: 10.1002/adv.1989.060090308 16. Kuan H-C, Kuan J-F, Ma C-CM, Huang J-M (2005) J Appl Polym Sci 96:2383. doi:10.1002/app.21694 17. Gan Q, Qi R, Zhang J, Yu J, Huang S (2011) J Appl Polym Sci 119:2539. doi:10.1002/app.31639 18. Atkinson JR, Cicek RZ (1983) Biomaterials 4:267. doi:10.1016/ 0142-9612(83)90026-1 19. Narkis M, Raiter I, Shkolnik S, Siegmannz A, Eyerer P (1987) J Macromol Sci B 26:37. doi:10.1080/00222348708248057 20. Andreopoulos AG, Kampouris EM (1986) J Appl Polym Sci 31:1061. doi:10.1002/app.1986.070310407 21. Azizi H, Morshedian J, Barikani M, Wagner MH (2011) Adv Polym Technol 30:286. doi:10.1002/adv.20224 22. Narkis M, Tzur A, Vaxman A, Fritz HG (1985) Polym Eng Sci 25:857. doi:10.1002/pen.760251311 23. Barzin J, Azizi H, Morshedian J (2006) Polym Plast Technol Eng 45:979. doi:10.1080/03602550600718209 24. Barzin J, Azizi H, Morshedian J (2007) Polym Plast Technol Eng 46:305. doi:10.1080/03602550601155815 25. Azizi H, Morshedian J, Barikani M (2009) J Vinyl Add Tech 15:184. doi:10.1002/vnl.20194 26. Bengtsson M, Gatenholm P, Oksman K (2005) Compos Sci Technol 65:1468. doi:10.1016/j.compscitech.2004.12.050 27. Bullen DJ, Capaccio G, Frye CJ, Brock T (1989) Br Polym J 21:117. doi:10.1002/pi.4980210205 28. Rizzo P, Baione F, Guerra G, Martinotto L, Albizzati E (2001) Macromolecules 34:5175. doi:10.1021/ma010121z 29. Bunn CW (1939) Trans Faraday Soc 35:482 30. Langford JI, Wilson AJC (1978) J Appl Crystallogr 11:102. doi: 10.1107/S0021889878012844 31. Clements J, Jakeways R, Ward IM (1978) Polymer 19:639. doi: 10.1016/0032-3861(78)90116-7
6761 32. Akovali G, Atalay A (1997) Polym Test 16:165. doi:10.1016/ S0142-9418(96)00037-2 33. Boerio FJ, Koenig JL (1970) J Chem Phys 52:3425 34. Strobl GR, Hagedorn W (1978) J Polym Sci 16:1181. doi: 10.1002/pol.1978.180160704 35. Koglin E, Meier RJ (1999) Comput Theor Polym Sci 9:327. doi: 10.1016/S1089-3156(99)00022-7 36. Kurelec L, Rastogi S, Meier RJ, Lemstra PJ (2000) Macromolecules 33:5593. doi:10.1021/ma9911187 37. Masetti G, Abbate S, Gussoni M, Zerbi G (1980) J Chem Phys 73:4671 38. Abbate S, Gussoni M, Zerbi G (1980) J Chem Phys 73:4680 39. Hagemann H, Snyder RG, Peacock AJ, Mandelkern L (1989) Macromolecules 22:3600. doi:10.1021/ma00199a017 40. Painter PC, Runt J, Coleman MM, Harrison IR (1977) J Polym Sci 15:1647. doi:10.1002/pol.1977.180150912 41. Mutter R, Stille W, Strobl G (1993) J Polym Sci B 31:99. doi: 10.1002/polb.1993.090310113 42. Taddei P, Affatato S, Fagnano C, Bordini B, Tinti A, Toni A (2002) J Mol Struct 613:121. doi:10.1016/S0022-2860(02)00141-2 43. Naylor CC, Meier RJ, Kip BJ et al (1995) Macromolecules 28:2969. doi:10.1021/ma00112a050 44. Glotin M, Mandelkern L (1982) Colloid Polym Sci 260:182. doi: 10.1007/bf01465438 45. Svoboda R, Ma´lek J (2013) J Therm Anal Calorim 111:1045. doi: 10.1007/s10973-012-2445-9 46. Fraser RDB, Suzuki E (1966) Anal Chem 38:1770. doi:10.1021/ ac60244a038 47. Fraser RDB, Suzuki E (1969) Anal Chem 41:37. doi:10.1021/ ac60270a007 48. Perejo´n A, Sa´nchez-Jime´nez PE, Criado JM, Pe´rez-Maqueda LA (2011) J Phys Chem B 115:1780. doi:10.1021/jp110895z 49. Koch CC, Cavin OB, McKamey CG, Scarbrough JO (1983) Appl Phys Lett 43:1017 50. Schwarz RB, Johnson WL (1983) Phys Rev Lett 51:415 51. Fan GJ, Guo FQ, Hu ZQ, Quan MX, Lu K (1997) Phys Rev B 55:11010 52. Smith AP, Spontak RJ, Ade H, Smith SD, Koch CC (1999) Adv Mater 11:1277. doi:10.1002/(sici)1521-4095(199910)11:15\1277: aid-adma1277[3.0.co;2-9 53. Bai C, Spontak RJ, Koch CC, Saw CK, Balik CM (2000) Polymer 41:7147. doi:10.1016/S0032-3861(00)00048-3 54. Smith AP, Shay JS, Spontak RJ et al (2000) Polymer 41:6271. doi:10.1016/S0032-3861(99)00830-7 55. Cavalieri F, Padella F, Bourbonneux S (2002) Polymer 43:1155. doi:10.1016/S0032-3861(01)00721-2 56. Pan J, Shaw WJD (1994) J Appl Polym Sci 52:507. doi: 10.1002/app.1994.070520405 57. Smith AP, Bai C, Ade H, Spontak RJ, Balik CM, Koch CC (1998) Macromol Rapid Commun 19:557. doi:10.1002/(sici) 1521-3927(19981101)19:11\557:aid-marc557[3.0.co;2-x 58. Beyer MK, Clausen-Schaumann H (2005) Chem Rev 105:2921. doi:10.1021/cr030697h
123
