- Email: [email protected]
boron composites: Effect on thermal stability
Surface & Coatings Technology 239 (2014) 70–77
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Atmospheric plasma torch treatment of polyethylene/boron composites: Effect on thermal stability J. Abenojar a,⁎, M.A. Martínez a, F. Velasco a, M.A. Rodríguez-Pérez b a b
Materials Science and Engineering Department, IAAB, Universidad Carlos III de Madrid, Av. Universidad, 30-28911 Leganes, Spain Cellular Materials Laboratory (CellMat), Universidad de Valladolid, P.° Belen, 7-47011 Valladolid, Spain
a r t i c l e
i n f o
Article history: Received 16 August 2013 Accepted in revised form 9 November 2013 Available online 18 November 2013 Keywords: Composite materials Thermogravimetric analysis Surface properties Thermal properties Epoxy resin Boron
a b s t r a c t Low density polyethylene (LDPE) and its boron composites (LDPEB) are typically used as coatings with functional nuclear properties. However, they present low surface energy and adhesion problems. Atmospheric pressure plasma torch (APPT) processes improve polymer wettability. The main objective of this work is the evaluation of the effect of APPT on the thermal stability of LDPE and its composites reinforced with 15 and 30% (by wt.) of boron. Physical and chemical changes on the surface of treated materials are analysed using X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM). The influence on thermal stability is evaluated using differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) techniques. The decomposition rate is measured by means of the Kamal equation, and the activation energy of the reaction is calculated from an Arrhenius equation and the Kissinger method. Results show that the APPT treatment is adequate to treat these materials (LDPE and its boron containing composites), as it does not degrade the material, but it modifies the chemistry and nanoroughness of the surface, increasing its wettability. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Polyolefins are a family of thermoplastics with a good balance of physical and chemical properties. Due to their low cost, light-weight, easy processing and recycling characteristics, they are suitable for the manufacture of composites. Low density polyethylene/boron composites (LDPEB) are highly efficient neutron shielding materials: polyethylene (PE), due to its high hydrogen content, and boron due to its neutron absorption ability [1–3]. Depending on boron amount, LDPEB can be used in different components, from capture-gamma radiation applications (5% boron) [4] to radioactive waste immobilization [5] (when boron ranges from 30 to 50%). However, the application of polyolefins and their composites is sometimes restricted by their poor wettability, which causes adhesion problems. Plasma techniques are fast and environmentally friendly processes that significantly increase surface energy (γs) and therefore, the adhesion properties of polymers [6]. In particular, atmospheric pressure plasma torches (APPT) [7,8] are non-local thermodynamic equilibrium plasma devices which produce cleaning and surface activation by means of the breakdown of pollutants and the introduction of different moieties of polar nature [9–11]. The advantages of APPT on PE to improve wettability and adhesion ⁎ Corresponding author. Tel.: +34 916248374; fax: +34 916249430. E-mail address: [email protected] (J. Abenojar). 0257-8972/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.surfcoat.2013.11.020
properties of the substrate have been demonstrated [10]. But it is necessary to assure the thermal stability of those materials after treatment to use them in nuclear plants. These polar groups are (C_O, C\O, N\O, B\O, etc.) active species on the polymer surface. According to Mueller and Jakob, these species sometimes can accelerate the degradation process at high temperature [12]. The ageing process occurs as a consequence of diverse phenomena such as cross-linking and free radical formation. Sanchis et al. have demonstrated that low pressure oxygen plasma changes PE thermal transitions [13], appearing an additional peak in differential scanning calorimetry (DSC) curve at 65–95 °C, possibly related to an increase in the crystallinity of the polymer [14]. This will also accelerate the degradation process of PE. Corona discharge [15] and other low-pressure plasma processes [16] may increase the crystallinity of the polymer due to the high temperatures achieved during surface treatment. Unfortunately, there is no evidence in the literature about the influence of the polar moieties created by APPT [10] on the degradation of LDPE. The main objective of this work is to determine the influence of APPT and the addition of boron in the chemical and thermal properties (melting temperature, enthalpy of fusion, crystallinity and decomposition temperature) of LDPE. The study evaluates not only the changes in thermal properties, but also the mechanism, the rate and the activation energy of decomposition reactions. This calculation provides information on possible
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
changes that may occur in the LDPE by the addition of boron and APPT, dramatically important for its thermal stability and safe thermal destruction. 2. Experimental procedure
71
Thermogravimetric analysis (TGA) was carried out to evaluate the decomposition temperature and residue in all materials. A dynamic cycle between 50 and 600 °C was carried out at 5, 10 and 20 °C/min (non-isothermal decomposition). The equipment used was a TGA/ SDTA851e Mettler Toledo (Greifensee, Switzerland), under nitrogen (60 ml/min). Three samples were tested for each condition.
2.1. Composites preparation 2.6. Kinetic analysis LDPE and LDPEB composites were prepared using commercial grade LDPE powder (0.92 g/cm3 density) and boron amorphous powder (2.34–2.37 g/cm3 density, 0.4–0.7 μm particle size, from Strem Chemicals Inc. Bischheim, France). LDPE and boron particles were blended in a Rheodrive 5000 from Haake Fision, Waltham, MA, USA. Different concentrations of boron powder (15 and 30% by wt.) were used, being materials labelled as LDPE15B and LDPE30B respectively. Uniform mixtures were prepared at 130 °C, using a screw speed of 50 rpm for 6 min. Precursors were compression moulded in a two hot-plate press at 170 °C and 5.5 MPa (Remtex, Barcelona, Spain). For each formulation 1-mm thick films were prepared. Neat LDPE was processed in the same way for comparison purposes. 2.2. Atmospheric pressure plasma torch (APPT) treatment An APPT device from Plasmatreat GmbH (Steinhagen, Germany) was used to treat both LDPE and LDPEB. The setup operated at a frequency of 17 kHz and a high tension discharge of 20 kV, and it was provided with a rotating torch ending in a nozzle (1900 rpm) through which plasma was expelled. The system contains an electronically speed-controlled platform where the samples were placed. The air plasma was generated at a working pressure of 2 bars inside the rotating nozzle by a non-equilibrium discharge and expelled through a circular orifice onto the samples. The speed of the platform was set at 10 m/ min, and the distance between the sample and the plasma nozzle was fitted to 6 mm. The size of the treated samples was 3 × 2 cm2. 2.3. X-ray photoelectron spectroscopy (XPS) APPT treated samples were analysed with a VG Scientific Microtech Multilab (VG Scienta, Hastings, United Kingdom) spectrometer using a Mg-Kα X-ray source (1253.6 eV) operating at 15 keV and 300 W. The take-off angle was 45°. The analysis was performed on the surface of 1 × 1 cm2 samples at a residual pressure below 5 × 10−8 Torr. A survey scan encompassing the 0–1200 eV region was obtained for each sample. High resolution spectra were obtained in a 20 eV range. All binding energies were referred to as the C 1s core level spectrum position for C\C and C\H (hydrocarbons) species at 285.0 eV. Atomic concentrations were calculated using VGX900-W system. 2.4. Atomic force microscopy (AFM) The AFM studies were performed using a MultiMode Nanoscope® IV (Digital Instruments, Veeco Metrology Group, Santa Barbara, CA, USA). The AFM measurements were carried out at room temperature, operating in the tapping mode, employing silicon tips with a force constant of about 40 N/m and a resonance frequency close to 300 kHz, recording simultaneously height and phase images. 2.5. Thermal characterization Differential scanning calorimetry (DSC) was used to evaluate the melting temperature and crystallinity in all materials. Heating dynamic cycles between −40 and 200 °C at 10 °C/min was carried out, followed by cooling at 20 °C/min. The first one was done to remove the thermal history effects, and the measurements were done in the second run. The equipment used was a DSC822 Mettler Toledo (Greifensee, Switzerland), using nitrogen (80 ml/min) as a purge gas.
In the thermal degradation of a system, the reactions are considered irreversible and the kinetic parameters can be determined from the non-isothermal thermogravimeric data. In order to analyse the degradation mechanisms of LDPE and LDPEB, kinetic parameters such as the activation energy (E), the pre-exponential factor (A) and the conversion function f(α) or α(T) were evaluated from TGA curves at different heating rates. A differential kinetic equation (Eq. (1)) of n-order can be applied, where α is the advance degree of reaction (0 b α b 1) at time t, n is the order of reaction, and k is the rate constant (defined by the Arrhenius expression, Eq. (2)). From Eqs. (1) and (2), Eq. (3), a differential kinetic equation which combines E and α, can be obtained. In those equations, Tp is the decomposition temperature (in K), corresponding to the inflection point in TGA curve (maximum decomposition rate), R is the gas constant (J·mol−1·K−1). dα n ¼ kð1−α Þ dt ln k ¼ lnA−
ð1Þ
E RT p
dα ¼ A exp −E=RT p f ðα Þ: dt
ð2Þ
ð3Þ
From Eq. (3), Kissinger [17] defined Eq. (4), which allows calculating the activation energies from TGA at different heating rates (β) in K/min, being C a constant. This method has been applied to other polymers, such as polyester/polycarbonate systems [18]. Furthermore, an autocatalytic mechanism can be taken into account in the kinetic equation (Eq. (5)), being n and m order of reactions. The Kamal equation (Eq. (6)) [19] reflects the n-order (Eq. (1)) and the autocatalytic mechanisms (Eq. (5)). ln
β T p2
! ¼−
E þC RT p
ð4Þ
dα n m ¼ kð1−α Þ α dt
ð5Þ
dα n n m ¼ k1 ð1−α Þ þ k2 ð1−α Þ α : dt
ð6Þ
In Eq. (6), k1 is the rate constant of the n-order reaction, k2 is the rate constant of the autocatalytic reaction and n and m are the reaction orders. Eq. (6) is used in thermoset polymers to calculate the kinetic parameters in the curing reaction [20–22]. In this way, the activation energy (E) can be determined through various analytical methods: it can be obtained from the slope of the plots of ln k (Eq. (2)) or ln(β / Tp2) (Eq. (4)) vs. the inverse of Tp. The activation energy is the parameter used to estimate the lifetime of a material at different temperatures for quality control purposes or for the specification of a material. Although there are numerous methods to calculate E [23–27], ASTM E1877 or ASTM E1641 standards suggest the use of Eq. (1), being the decomposition an n-order process, with n = 1. This implies that only one species is involved in the reaction, with only one mechanism of decomposition taking place. This approach is very similar to that of Kissinger (Eq. (4)).
72
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
This research has used the Kamal approach (Eq. (6)) to calculate rate constants and to study the contribution of both mechanisms (autocatalytic and n-order). However, as two constants cannot be used to calculate activation energies in the Arrhenius equation (Eq. (2)), the selected k will be the one giving higher activation energies according to the results. 3. Results 3.1. Chemical and topographical characterization As XPS shows (Table 1 and Fig. 1), the surface has been effectively modified with APPT treatment. Fig. 1(a and b) shows the bonds existing in untreated LDPE30B. The mean carbon band (labelled as (1)) is related to C\C and C\H bonds (at 284.6 eV) [28,29]. Those bands are typical from LDPE due to its structure. Moreover, a small band (2), located at 286.3 eV, appears due to C\OH bonds, possibly related to slight oxidations during forming process. The band ((4) at 187.3 eV) is due to B [29], expected as boron particles are present in the material. After APPT treatment, the carbon bands (Fig. 1c) change. Band (2), corresponding to C\OH bonds, strongly increases due to the effect of plasma. Moreover, a signal from C\C(O)\OH band (labelled as (3), at 288.6 eV) appears. Those bands indicate that an interaction is taking place between LDPE and plasma species. Plasma also affects boron signal (Fig. 1d): a band appears at 192.3 eV (5), being assigned to B\O (B2O3) bonds, indicating that boron particles are oxidated. Nitrogen is also introduced in all the materials after APPT treatment (Fig. 1e). Its presence is logical as air was used in APPT process, and a small amount has been incorporated to the polymer. R–CN ((6), at 400.0 eV) and NH+ 4 ((7), at 401.9 eV) are present, showing how nitrogen is incorporated in LDPE chains. Table 1 summarises atomic percentages in all materials. Boron increases after APPT treatment. XPS is focused on the analysis of the surfaces and hence the detected amount of boron is higher. Boron is close to the surface due to erosion of LDPE with APPT. Moreover, oxygen contents of the composites increase due to its introduction from the atmosphere, and as a consequence, the measured carbon content is lower. After APPT treatment, the O/C ratio (Fig. 2) dramatically increases about 6 times from PE initial value and 5 times for composites. The insertion of oxygen and nitrogen atoms in the surface can only proceed via rupture of C\C/C\H linkages of LDPE (as Fig. 1 shows) and further reactions with plasma and environmental species. This fact verifies the creation of polar groups (carbonyl, carboxylic, etc.) as Encinas et al. [6] have proved on other APPT treated polyolefins. The great increase in the polar groups activates the surface of studied materials, increasing its wettability and surface energy as it has been demonstrated [6,10,30]. The surface energy changes from 24 mN/m (untreated LDPE) to 65 mN/m (APPT treated LDPE), being all increment taken place in the polar component. Fig. 3 shows selected images of the AFM study. It can be clearly appreciated that nanoroughness increases when the APPT treatment is carried out. Peaks and valleys appear in treated LDPE. The effect of boron addition can be seen in the image of APPT treated LDPE30B. It seems that wider peaks are formed in this material, possibly due to the presence of boron particles, supporting the appearance of boron in XPS analysis (Table 1). The roughness values (Table 2) support the
Table 1 Atomic percentage of the different elements in LDPE and composites, untreated and after APPT treatment. % atomic
Boron Carbon Oxygen Nitrogen
Untreated
APPT
LDPE
LDPE15B
LDPE30B
LDPE
LDPE15B
LDPE30B
0.00 97.64 2.36 0.00
0.34 96.28 3.38 0.00
0.51 98.07 1.42 0.00
0.00 75.68 22.31 2.00
1.85 78.12 17.91 2.12
2.57 74.10 21.25 2.08
increase in nanoroughness. Both the average roughness (Ra, integral of the absolute value of the roughness profile height over the evaluation length) and the root mean square roughness (Rms, an average of peaks and valleys of a materials surface profile) clearly increase with APPT treatment. Specially noticeable is the increase in roughness in LDPE from as manufactured to as treated condition, possibly related to the high deformability of this material, in particular compared to its composites. This increase in nanoroughness is in agreement with the findings of other researchers in polymer materials under similar treatments [31,32]. So APPT treatment can lead to materials with more roughness to help to adhesive bonding if required as mechanical anchoring points increase, together with an increase in wettability of the surface for a selected adhesive. 3.2. Melting temperature and crystallinity After confirming the chemical and topographical effect of APPT on LDPE and LDPEB, the main objective is to evaluate the possible influence of APPT on the thermal properties of studied materials. Fig. 4 shows the melting temperature and crystallinity for the treated and untreated materials. For untreated materials, the melting temperature does not change when boron is added (Fig. 4a), and a similar effect is observed in the crystallinity (Fig. 4b). The slight variations of melting temperature can be related to the effect of boron particles, that makes difficult the disentanglement of LDPE chains during melting. The slight increase in crystallinity could be connected to a nucleating effect of the boron particles in the LDPE crystallization, as it has been also observed in LDPE/silica nanocomposites [33]. When the APPT treatment is carried out, the changes in the melting temperature are in the range of the experimental data error. The crystallinity (Fig. 4b) diminishes in all cases, although the error bars overlapped for untreated and treated composite materials. This slight decrease can be related to cross-linking during APPT treatment, as it has been suggested for high density polyethylene (HDPE) [34]. The decrease in crystallinity and no occurrence of a second peak around 65 to 95 °C in DSC curve [13] reveal no degradation in the material with APPT, as it is has been observed with thermal ageing [14,35] or low pressure oxygen plasma treatment [13]. 3.3. Decomposition temperature The TGA curves for the untreated LDPE15B composite are shown in Fig. 5a, showing changes on the decomposition temperature with heating rate. As expected, when the heating rate increases (from 5 to 20 K/min), the decomposition temperature increases. The solid residue obtained is in accordance with the boron amount in the material (15%). The percentage of the residue does not change with heating rate, but it does accordingly to the percentage of added boron. This clearly indicates that the residue after thermal treatment is pure boron, as at the maximum temperature achieved in these tests (600 °C) and they are carried out in nitrogen, avoiding boron oxidation. Therefore, the obtained mixture is perfectly homogeneous. LDPE and the composites present a relatively good thermal stability, since no significant mass loss (b 0.5%) occurs until 300 °C. The first derivative of the TGA curve (DTG, Fig. 5b) provides the decomposition temperature (peak of the curve), and its integration with time gives the degree of conversion of the reaction. Fig. 6 shows the results for the samples heated at 10 K/min. The values obtained for LDPE agree with those obtained by Paik and Kar for commercial PE [36]. These values present a small increase for the 30% boron containing composites, due to the incorporation of the particles, and it has been found for all heating rates. The same effect was found by Chrissafis and Bikiaris [37] adding particles (montmorillonite, multi-walled carbon nanotubes and SiO2) to HDPE. When APPT treatment is carried out, the decomposition temperature does not substantially change (Fig. 6). The slightly increases
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
73
Fig. 1. XPS spectra for carbon, boron and nitrogen of LDPE30B, both untreated and after APPT treatment.
found in all treated materials can be related to the presence of carbonyl and carboxyl groups in APPT treated materials. This suggests that thermal stability is not affected when APPT treatment is applied. The same behaviour was found for the other heating rates.
decomposed. Then, the degree of conversion α reached at a time t can be calculated as shown in Eq. (7). α¼
ΔWLt ΔWLTotal
ð7Þ
3.4. Kinetic analysis In order to work with the Kamal equation and to apply the Arrhenius equation, a partial integration of DTG versus time (Fig. 5b) must be done. The reaction progress will be directly proportional to the amount of weight loss (WL) and the maximum degree of conversion will be reached at the end of the reaction, when the entire polymer has been
Fig. 2. Variation of the O/C ratio with boron content and APPT treatment.
where ΔWLt is the weight variation for time t and ΔWLTotal is the total weight variation of the decomposition process. In all materials, heating to 650 °C ensures complete decomposition of the polymer, where the conversion degree is the maximum (100%). An example of the curve obtained for the conversion degree with time is shown in Fig. 7a. Fig. 7b plots the conversion rate (derivate of Fig. 7a curve) vs. the conversion degree. When this plot is fitted to the Kamal equation (Eq. (6)), the kinetic parameters are obtained (Table 3). Table 3 shows n values around 1.0 and m values around 0.6 for all materials. The goodness of fitting (R2) of the Kamal model presents values higher than 0.998. These m values indicate that more than one species is involved in the decomposition reaction of LDPE, instead of one species. The increase of both n and m values in APPT treated materials would be in agreement with the presence of more species in the material after the treatment, as XPS has shown (Table 1 and Fig. 2). Moreover, the n values around 1 are in the range of those of Baloch et al. [38] who found for commercial PE when only n-order mechanism is considered. The presence of different species would also agree with possible decomposition mechanisms proposed for polymers (randomchain scission, end-chain scission, chain-stripping or cross-linking [39]), that would possibly not follow rigorously a first order kinetics. Moreover, n and m values tend to increase when APPT treatment is done. This would imply that more species are present in the decomposition reaction. This is in agreement with XPS results (Table 1 and Fig. 2),
74
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
Fig. 3. AFM images of some materials.
that show the formation of moieties during APPT treatment that can affect decomposition process. Table 3 data indicates that the k2 constant is higher than k1 in all materials (untreated and treated), therefore the autocatalytic reaction is faster than n-order reaction. The reaction is governed by n-order mechanism, since it is the slowest reaction. For the calculations of the activation energy, only one constant was used. Two different approaches were followed: the use of the Arrhenius equation (Eq. (2)), using k2 parameters (as the autocatalytic component of the reaction will lead to higher energies) and the use of the Kissinger method (Eq. (4)). Both methods yield a linear relationship (Fig. 8), being the slope used to calculate the activation energy. Regardless of the method used to calculate the activation energy, (Arrhenius, Fig. 8a, and Kissinger, Fig. 8b), the curves are similar for the three materials, both treated and untreated. The activation energies shown in Table 4 are in the range of those obtained by Baloch et al. [38] and Paik and Kar [36]: 300.5–374.1 kJ/ mol for commercial PE, depending on the calculation method used. Values shown in Table 4, with both models, are within the ranges found for PE [38]. However, all the energies in Table 4 show differences Table 2 Roughness of LDPE and its composites, untreated and after APPT treatment. Roughness
Ra (nm) Rms (nm)
Untreated
APPT
LDPE
LDPE15B
LDPE30B
LDPE
LDPE15B
LDPE30B
23.51 29.73
28.46 36.47
41.02 50.04
66.51 84.14
69.56 85.85
60.87 72.06
of less than 10%, indicating that the effect of APPT on stability is very low, remaining its value above neat LDPE. All the energies in Table 4, calculated by both methods (Arrhenius and Kissinger) show differences of less than 10%, indicating that the effect of APPT on stability is very small and within the experimental error for these measurements. Therefore, the most important conclusion obtained from these experiments is that the presence of both the boron and the plasma treatments has a minor effect on the thermal stability of LDPE. In any case, a closer analysis of the data shows some unexpected trends. For instance, for both the pure material and the material containing 15% boron, the values of the activation energy for the treated materials are higher than those for the non-treated ones, being the difference higher for the pure material. However the contrary effect is observed for the material containing 30% boron. APPT treatment of polymers provokes the formation of oxygen containing active species. The dissociation energy of a C\O bond is higher than that of a C\H bond. Moreover, APPT can promote the formation of C_C bonds due to crosslinking. Then higher activation energy for the treated materials should be expected, independently of the boron content (Table 4). Due to this, the values of activation energy of LDPE30B, lower for treated material, cannot be related to the modifications in the composition of the polymer, but to effects related to the precision of the experiments. In this sense, one possible explanation is based on the modifications of the thermal conductivity of the materials by the presence of the particles; a higher amount of particles should increase the conductivity and then the heat transfer between the crucible and the
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
75
Fig. 4. Variation of (a) the melting temperature and (b) the crystallinity of materials, both untreated and with APPT.
sample during the TGA tests. This in turn could introduce some artefacts in the comparison of the TGA curves at high heating rates. The same type of experimental artefacts could also be the reason for the non-clear trends observed as a function of the boron content for both the non-treated and treated materials. As it is observed in Table 4, the maximum activation energy is obtained for the LDPE15B material for both calculation methods.
Fig. 5. (a) Variation of the TGA curves with heating rate, and (b) decomposition curves (TGA) and the first derivative of the TGA curve (DTG) at a heating rate of 10 K/min of untreated LDPE15B.
4. Conclusions
Acknowledgements
The APPT treatment produces chemical and morphological changes in the materials surface, as XPS and AFM show: O/C ratio increases, as a consequence of the formation of polar groups and nanoroughness is increased for all materials. The highest evaluated boron amount (30% by wt.) slightly increases the crystallinity and the decomposition temperature of LDPE. These slight variations are the consequence of the nucleation effect of boron particles on LDPE crystals. APPT treatment does not affect the melting temperature, the crystallinity and the decomposition temperature of materials. This implies that APPT does not affect the thermal stability of studied materials. In all cases, the kinetic parameters show that the decomposition reaction is mainly autocatalytic as this is the faster reaction. Reaction order is always higher than 1, indicating the presence of more than one species in the reaction. From activation energy calculations, it can be concluded that the thermal stability of LDPE and LDPEB composites is not substantially modified by APPT. The APPT treatment is adequate to treat these materials (LDPE and its boron containing composites), as it does not degrade the material, but it modifies the chemistry and nanoroughness of the surface, increasing its wettability.
Authors wish to acknowledge the Spanish Ministry of Science and Innovation (through project MAT2011-29182-C02-02) for its financial support in this research.
Fig. 6. Variation of the decomposition temperature at 10 K/min heating rate for all the tested materials.
76
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
Fig. 8. Application of (a) the Arrhenius equation (Eq. (2)) for APPT treated materials and (b) the Kissinger method (Eq. (4)) for untreated materials to calculate the activation energy.
Table 4 Activation energies obtained by the Arrhenius (Eq. (2)) and Kissinger (Eq. (4)) methods. Arrhenius
Fig. 7. LDPE15B after APPT treatment, heated at 5 K/min rate. (a) Conversion degree vs. time. (b) Fit to the Kamal equation (Eq. (6)). LDPE LDPE APPT LDPE15B LDPE15B APPT LDPE30B LDPE30B APPT
Kissinger
E (kJ/mol) Ao (min−1)
R2
E (kJ/mol) Constant
311 354 331 339 320 314
0.9345 0.9988 0.9990 0.9945 1.0000 0.9821
321 345 322 330 312 304
1.23E 7.19E 1.93E 5.79E 2.88E 1.07E
+ + + + + +
19 22 21 21 20 20
4.46E 1.88E 5.34E 1.50E 8.16E 2.22E
+ + + + + +
R2 17 19 17 18 16 16
0.9961 0.9841 0.9998 0.9992 0.9999 0.9808
Table 3 Kinetic parameters calculated with Eq. (6) of the untreated materials and APPT treated ones (shadowed data).
LDPE
β (k/min)
TP (K)
k1
k2
n
m
R2
5
739.2 740.0 749.8 750.7 758.5 757.9 739.9 741.1 749.2 750.9 759.3 760.1 741.8 741.5 751.5 753.9 761.9 761.7
0.00007 0.00008 0.00011 0.00012 0.00028 0.00017 0.00006 0.00006 0.00007 0.00008 0.00012 0.00013 0.00006 0.00006 0.00009 0.0001 0.0001 0.00011
0.00714 0.00734 0.01448 0.01539 0.02586 0.02879 0.00768 0.00779 0.01516 0.01535 0.03028 0.03076 0.00788 0.00781 0.01551 0.01566 0.03101 0.03102
1.03 1.07 1.11 1.14 0.89 1.17 0.99 0.97 1.01 1.02 1.10 1.10 0.92 0.92 0.98 1.01 1.04 1.04
0.56 0.56 0.59 0.62 0.53 0.64 0.58 0.59 0.61 0.61 0.65 0.65 0.57 0.56 0.59 0.59 0.63 0.63
0.99817 0.99818 0.99852 0.99698 0.99807 0.99847 0.99950 0.99979 0.99930 0.99938 0.99972 0.99982 0.99963 0.99958 0.99962 0.99974 0.99989 0.99989
10 20 LDPE15B
5 10 20
LDPE30B
5 10 20
References [1] X. Colin, C. Monchy-Leroy, L. Audouin, J. Verdu, Nucl. Instrum. Methods B 265 (2007) 251–255. [2] C. Harrison, E. Burgett, N. Hertel, E. Grulke, in: S. Mathur, M. Singh, T. Ohji, A. Wereszczak (Eds.), Polyethylene/boron containing composites for radiation shielding applications, Nanostructured Materials and Nanotechnology II, 29 (8), John Wiley & Sons, Inc., New Jersey, 2009, pp. 77–84. [3] Z. Li, X. Xue, T. Jiang, H. Yang, M. Zhou, Adv. Mater. Res. 201–203 (2011) 2767–2771. [4] http://www.deqtech.com/Shieldwerx/Products/swx207hd.htm June 2012. [5] S. Cota, T.V.S. de Oliveira, M. Senne Jr., G. Pacheco, M. de Albuquerque, Prog. Nucl. Energy 51 (2009) 236–239. [6] N. Encinas, J. Abenojar, M.A. Martínez, Int. J. Adhes. Adhes. 33 (2012) 1–6. [7] C. Tendero, C. Tixier, T. Pascal, J. Desmaison, P. Leprince, Spectrochim. Acta B 61 (2006) 2–30. [8] A. Schütze, J.Y. Jeong, S.E. Babayan, J. Park, G.S. Selwyn, R.F. Hicks, IEEE Technol. Plasma Sci. 26 (1998) 1685–1694. [9] A.N. Bhoj, M.J. Kushner, IEEE Technol. Plasma Sci. 33 (2005) 250–251. [10] N. Encinas, B. Díaz-Benito, J. Abenojar, M.A. Martínez, Surf. Coat. Technol. 205 (2010) 396–402. [11] J. Abenojar, R. Torregrosa-Coque, M.A. Martínez, J.M. Martín-Martínez, Surf. Coat. Technol. 203 (2009) 2173–2180.
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77 [12] W. Mueller, I. Jakob, Polym. Degrad. Stab. 79 (2003) 161–172. [13] M.R. Sanchis, V. Blanes, M. Blanes, D. Garcia, R. Balart, Eur. Polym. J. 42 (2006) 1558–1568. [14] J.V. Gulmine, P.R. Janissek, H.M. Heise, L. Alcelrud, Polym. Degrad. Stab. 79 (2003) 385–397. [15] N.F. Belkacemi, R. Coelho, J. Phys. D: Appl. Phys. 28 (1995) 1001–1008. [16] S. Bhowmik, H.W. Bonin, V.T. Bui, T.K. Chaki, J. Adhes. 82 (2006) 1–18. [17] H.E. Kissinger, Anal. Chem. 29 (1957) 1702–1706. [18] A. Al-Mulla, J. Mathew, L. Al-Omairi, S. Bhattacharya, Polym. Eng. Sci. 51 (2011) 2335–2344. [19] M.R. Kamal, Polym. Eng. Sci. 14 (1974) 231–239. [20] S. Sourour, M.R. Kamal, Thermochim. Acta 14 (1976) 41–59. [21] S.Y. Pusatcioglu, A.L. Fricke, J.C. Hassler, J. Appl. Polym. Sci. 24 (1979) 937–946. [22] M.J.M. Abadie, D. Sakkas, Eur. Polym. J. 28 (1992) 873–879. [23] T. Ozawa, J. Therm. Anal. Calorim. 2 (1970) 301–324. [24] T. Ozawa, Bull. Chem. Soc. Jpn. 38 (1965) 1881–1886. [25] J.H. Flynn, L.A. Wall, J. Res. Natl. Bur. Stand. 70 (1966) 487–523. [26] H.L. Friedman, J. Polym. Sci. Polym. Symp. 6 (1964) 183–195. [27] J. Palacios, R. Perera, C. Rosales, C. Albano, J.M. Pastor, Polym. Degrad. Stab. 97 (2012) 729–737. [28] http://srdata.nist.gov/xps/DataDefinition.aspx May 2013.
77
[29] J.F. Moulder, W.F. Stickle, P.F.E. Sobol, K.D. Bomben, in: J. Chastain, R.C. King (Eds.), Handbook of X-ray photoelectron Spectroscopy, Physical Electronics, Inc., Minnesota, 1995. [30] N. Encinas, R.G. Dillingham, B.R. Oakley, J. Abenojar, M.A. Martínez, M. Pantoja, J. Adhes. 88 (2012) 321–336. [31] K. Tamargo-Martínez, A. Martínez-Alonso, S. Villar-Rodil, J.I. Paredes, M.A. Montes-Morán, J.M.D. Tascón, J. Chromatogr. A 1218 (2011) 3781–3790. [32] S. Ercegovíc Razíc, R. Cunko, V. Svetlicíc, S. Segota, Mater. Technol. 26 (2011) 146–152. [33] C. Saiz-Arroyo, M.A. Rodríguez-Pérez, J.I. Velasco, J.A. de Saja, Compos. Part B Eng. 48 (2013) 45–50. [34] I. Novák, A. Popelka, I. Krupa, I. Chodák, I. Janigová, T. Nedelcev, M. Spirková, A. Kleinová, Vacuum 86 (2012) 2089–2094. [35] M. Omastová, S. Podhradská, J. Prokös, I. Janigová, J. Stejskal, Polym. Degrad. Stab. 82 (2003) 251–256. [36] P. Paik, K.K. Kar, Mater. Chem. Phys. 113 (2009) 953–961. [37] K. Chrissafis, D. Bikiaris, Thermochim. Acta 523 (2011) 1–24. [38] M.K. Baloch, M.J.Z. Khurram, G.F. Durrani, J. Appl. Polym. Sci. 120 (2011) 3511–3518. [39] C.L. Beyler, M.M. Hirschler, in: P.J. DiNenno (Ed.), Thermal Decomposition of Polymers. Chapter 7: Handbook of Fire Protection Engineering, 3rd ed., National Fire Protection Association, Quincy, Massachusetts, 2002.
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Atmospheric plasma torch treatment of polyethylene/boron composites: Effect on thermal stability J. Abenojar a,⁎, M.A. Martínez a, F. Velasco a, M.A. Rodríguez-Pérez b a b
Materials Science and Engineering Department, IAAB, Universidad Carlos III de Madrid, Av. Universidad, 30-28911 Leganes, Spain Cellular Materials Laboratory (CellMat), Universidad de Valladolid, P.° Belen, 7-47011 Valladolid, Spain
a r t i c l e
i n f o
Article history: Received 16 August 2013 Accepted in revised form 9 November 2013 Available online 18 November 2013 Keywords: Composite materials Thermogravimetric analysis Surface properties Thermal properties Epoxy resin Boron
a b s t r a c t Low density polyethylene (LDPE) and its boron composites (LDPEB) are typically used as coatings with functional nuclear properties. However, they present low surface energy and adhesion problems. Atmospheric pressure plasma torch (APPT) processes improve polymer wettability. The main objective of this work is the evaluation of the effect of APPT on the thermal stability of LDPE and its composites reinforced with 15 and 30% (by wt.) of boron. Physical and chemical changes on the surface of treated materials are analysed using X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM). The influence on thermal stability is evaluated using differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) techniques. The decomposition rate is measured by means of the Kamal equation, and the activation energy of the reaction is calculated from an Arrhenius equation and the Kissinger method. Results show that the APPT treatment is adequate to treat these materials (LDPE and its boron containing composites), as it does not degrade the material, but it modifies the chemistry and nanoroughness of the surface, increasing its wettability. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Polyolefins are a family of thermoplastics with a good balance of physical and chemical properties. Due to their low cost, light-weight, easy processing and recycling characteristics, they are suitable for the manufacture of composites. Low density polyethylene/boron composites (LDPEB) are highly efficient neutron shielding materials: polyethylene (PE), due to its high hydrogen content, and boron due to its neutron absorption ability [1–3]. Depending on boron amount, LDPEB can be used in different components, from capture-gamma radiation applications (5% boron) [4] to radioactive waste immobilization [5] (when boron ranges from 30 to 50%). However, the application of polyolefins and their composites is sometimes restricted by their poor wettability, which causes adhesion problems. Plasma techniques are fast and environmentally friendly processes that significantly increase surface energy (γs) and therefore, the adhesion properties of polymers [6]. In particular, atmospheric pressure plasma torches (APPT) [7,8] are non-local thermodynamic equilibrium plasma devices which produce cleaning and surface activation by means of the breakdown of pollutants and the introduction of different moieties of polar nature [9–11]. The advantages of APPT on PE to improve wettability and adhesion ⁎ Corresponding author. Tel.: +34 916248374; fax: +34 916249430. E-mail address: [email protected] (J. Abenojar). 0257-8972/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.surfcoat.2013.11.020
properties of the substrate have been demonstrated [10]. But it is necessary to assure the thermal stability of those materials after treatment to use them in nuclear plants. These polar groups are (C_O, C\O, N\O, B\O, etc.) active species on the polymer surface. According to Mueller and Jakob, these species sometimes can accelerate the degradation process at high temperature [12]. The ageing process occurs as a consequence of diverse phenomena such as cross-linking and free radical formation. Sanchis et al. have demonstrated that low pressure oxygen plasma changes PE thermal transitions [13], appearing an additional peak in differential scanning calorimetry (DSC) curve at 65–95 °C, possibly related to an increase in the crystallinity of the polymer [14]. This will also accelerate the degradation process of PE. Corona discharge [15] and other low-pressure plasma processes [16] may increase the crystallinity of the polymer due to the high temperatures achieved during surface treatment. Unfortunately, there is no evidence in the literature about the influence of the polar moieties created by APPT [10] on the degradation of LDPE. The main objective of this work is to determine the influence of APPT and the addition of boron in the chemical and thermal properties (melting temperature, enthalpy of fusion, crystallinity and decomposition temperature) of LDPE. The study evaluates not only the changes in thermal properties, but also the mechanism, the rate and the activation energy of decomposition reactions. This calculation provides information on possible
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
changes that may occur in the LDPE by the addition of boron and APPT, dramatically important for its thermal stability and safe thermal destruction. 2. Experimental procedure
71
Thermogravimetric analysis (TGA) was carried out to evaluate the decomposition temperature and residue in all materials. A dynamic cycle between 50 and 600 °C was carried out at 5, 10 and 20 °C/min (non-isothermal decomposition). The equipment used was a TGA/ SDTA851e Mettler Toledo (Greifensee, Switzerland), under nitrogen (60 ml/min). Three samples were tested for each condition.
2.1. Composites preparation 2.6. Kinetic analysis LDPE and LDPEB composites were prepared using commercial grade LDPE powder (0.92 g/cm3 density) and boron amorphous powder (2.34–2.37 g/cm3 density, 0.4–0.7 μm particle size, from Strem Chemicals Inc. Bischheim, France). LDPE and boron particles were blended in a Rheodrive 5000 from Haake Fision, Waltham, MA, USA. Different concentrations of boron powder (15 and 30% by wt.) were used, being materials labelled as LDPE15B and LDPE30B respectively. Uniform mixtures were prepared at 130 °C, using a screw speed of 50 rpm for 6 min. Precursors were compression moulded in a two hot-plate press at 170 °C and 5.5 MPa (Remtex, Barcelona, Spain). For each formulation 1-mm thick films were prepared. Neat LDPE was processed in the same way for comparison purposes. 2.2. Atmospheric pressure plasma torch (APPT) treatment An APPT device from Plasmatreat GmbH (Steinhagen, Germany) was used to treat both LDPE and LDPEB. The setup operated at a frequency of 17 kHz and a high tension discharge of 20 kV, and it was provided with a rotating torch ending in a nozzle (1900 rpm) through which plasma was expelled. The system contains an electronically speed-controlled platform where the samples were placed. The air plasma was generated at a working pressure of 2 bars inside the rotating nozzle by a non-equilibrium discharge and expelled through a circular orifice onto the samples. The speed of the platform was set at 10 m/ min, and the distance between the sample and the plasma nozzle was fitted to 6 mm. The size of the treated samples was 3 × 2 cm2. 2.3. X-ray photoelectron spectroscopy (XPS) APPT treated samples were analysed with a VG Scientific Microtech Multilab (VG Scienta, Hastings, United Kingdom) spectrometer using a Mg-Kα X-ray source (1253.6 eV) operating at 15 keV and 300 W. The take-off angle was 45°. The analysis was performed on the surface of 1 × 1 cm2 samples at a residual pressure below 5 × 10−8 Torr. A survey scan encompassing the 0–1200 eV region was obtained for each sample. High resolution spectra were obtained in a 20 eV range. All binding energies were referred to as the C 1s core level spectrum position for C\C and C\H (hydrocarbons) species at 285.0 eV. Atomic concentrations were calculated using VGX900-W system. 2.4. Atomic force microscopy (AFM) The AFM studies were performed using a MultiMode Nanoscope® IV (Digital Instruments, Veeco Metrology Group, Santa Barbara, CA, USA). The AFM measurements were carried out at room temperature, operating in the tapping mode, employing silicon tips with a force constant of about 40 N/m and a resonance frequency close to 300 kHz, recording simultaneously height and phase images. 2.5. Thermal characterization Differential scanning calorimetry (DSC) was used to evaluate the melting temperature and crystallinity in all materials. Heating dynamic cycles between −40 and 200 °C at 10 °C/min was carried out, followed by cooling at 20 °C/min. The first one was done to remove the thermal history effects, and the measurements were done in the second run. The equipment used was a DSC822 Mettler Toledo (Greifensee, Switzerland), using nitrogen (80 ml/min) as a purge gas.
In the thermal degradation of a system, the reactions are considered irreversible and the kinetic parameters can be determined from the non-isothermal thermogravimeric data. In order to analyse the degradation mechanisms of LDPE and LDPEB, kinetic parameters such as the activation energy (E), the pre-exponential factor (A) and the conversion function f(α) or α(T) were evaluated from TGA curves at different heating rates. A differential kinetic equation (Eq. (1)) of n-order can be applied, where α is the advance degree of reaction (0 b α b 1) at time t, n is the order of reaction, and k is the rate constant (defined by the Arrhenius expression, Eq. (2)). From Eqs. (1) and (2), Eq. (3), a differential kinetic equation which combines E and α, can be obtained. In those equations, Tp is the decomposition temperature (in K), corresponding to the inflection point in TGA curve (maximum decomposition rate), R is the gas constant (J·mol−1·K−1). dα n ¼ kð1−α Þ dt ln k ¼ lnA−
ð1Þ
E RT p
dα ¼ A exp −E=RT p f ðα Þ: dt
ð2Þ
ð3Þ
From Eq. (3), Kissinger [17] defined Eq. (4), which allows calculating the activation energies from TGA at different heating rates (β) in K/min, being C a constant. This method has been applied to other polymers, such as polyester/polycarbonate systems [18]. Furthermore, an autocatalytic mechanism can be taken into account in the kinetic equation (Eq. (5)), being n and m order of reactions. The Kamal equation (Eq. (6)) [19] reflects the n-order (Eq. (1)) and the autocatalytic mechanisms (Eq. (5)). ln
β T p2
! ¼−
E þC RT p
ð4Þ
dα n m ¼ kð1−α Þ α dt
ð5Þ
dα n n m ¼ k1 ð1−α Þ þ k2 ð1−α Þ α : dt
ð6Þ
In Eq. (6), k1 is the rate constant of the n-order reaction, k2 is the rate constant of the autocatalytic reaction and n and m are the reaction orders. Eq. (6) is used in thermoset polymers to calculate the kinetic parameters in the curing reaction [20–22]. In this way, the activation energy (E) can be determined through various analytical methods: it can be obtained from the slope of the plots of ln k (Eq. (2)) or ln(β / Tp2) (Eq. (4)) vs. the inverse of Tp. The activation energy is the parameter used to estimate the lifetime of a material at different temperatures for quality control purposes or for the specification of a material. Although there are numerous methods to calculate E [23–27], ASTM E1877 or ASTM E1641 standards suggest the use of Eq. (1), being the decomposition an n-order process, with n = 1. This implies that only one species is involved in the reaction, with only one mechanism of decomposition taking place. This approach is very similar to that of Kissinger (Eq. (4)).
72
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
This research has used the Kamal approach (Eq. (6)) to calculate rate constants and to study the contribution of both mechanisms (autocatalytic and n-order). However, as two constants cannot be used to calculate activation energies in the Arrhenius equation (Eq. (2)), the selected k will be the one giving higher activation energies according to the results. 3. Results 3.1. Chemical and topographical characterization As XPS shows (Table 1 and Fig. 1), the surface has been effectively modified with APPT treatment. Fig. 1(a and b) shows the bonds existing in untreated LDPE30B. The mean carbon band (labelled as (1)) is related to C\C and C\H bonds (at 284.6 eV) [28,29]. Those bands are typical from LDPE due to its structure. Moreover, a small band (2), located at 286.3 eV, appears due to C\OH bonds, possibly related to slight oxidations during forming process. The band ((4) at 187.3 eV) is due to B [29], expected as boron particles are present in the material. After APPT treatment, the carbon bands (Fig. 1c) change. Band (2), corresponding to C\OH bonds, strongly increases due to the effect of plasma. Moreover, a signal from C\C(O)\OH band (labelled as (3), at 288.6 eV) appears. Those bands indicate that an interaction is taking place between LDPE and plasma species. Plasma also affects boron signal (Fig. 1d): a band appears at 192.3 eV (5), being assigned to B\O (B2O3) bonds, indicating that boron particles are oxidated. Nitrogen is also introduced in all the materials after APPT treatment (Fig. 1e). Its presence is logical as air was used in APPT process, and a small amount has been incorporated to the polymer. R–CN ((6), at 400.0 eV) and NH+ 4 ((7), at 401.9 eV) are present, showing how nitrogen is incorporated in LDPE chains. Table 1 summarises atomic percentages in all materials. Boron increases after APPT treatment. XPS is focused on the analysis of the surfaces and hence the detected amount of boron is higher. Boron is close to the surface due to erosion of LDPE with APPT. Moreover, oxygen contents of the composites increase due to its introduction from the atmosphere, and as a consequence, the measured carbon content is lower. After APPT treatment, the O/C ratio (Fig. 2) dramatically increases about 6 times from PE initial value and 5 times for composites. The insertion of oxygen and nitrogen atoms in the surface can only proceed via rupture of C\C/C\H linkages of LDPE (as Fig. 1 shows) and further reactions with plasma and environmental species. This fact verifies the creation of polar groups (carbonyl, carboxylic, etc.) as Encinas et al. [6] have proved on other APPT treated polyolefins. The great increase in the polar groups activates the surface of studied materials, increasing its wettability and surface energy as it has been demonstrated [6,10,30]. The surface energy changes from 24 mN/m (untreated LDPE) to 65 mN/m (APPT treated LDPE), being all increment taken place in the polar component. Fig. 3 shows selected images of the AFM study. It can be clearly appreciated that nanoroughness increases when the APPT treatment is carried out. Peaks and valleys appear in treated LDPE. The effect of boron addition can be seen in the image of APPT treated LDPE30B. It seems that wider peaks are formed in this material, possibly due to the presence of boron particles, supporting the appearance of boron in XPS analysis (Table 1). The roughness values (Table 2) support the
Table 1 Atomic percentage of the different elements in LDPE and composites, untreated and after APPT treatment. % atomic
Boron Carbon Oxygen Nitrogen
Untreated
APPT
LDPE
LDPE15B
LDPE30B
LDPE
LDPE15B
LDPE30B
0.00 97.64 2.36 0.00
0.34 96.28 3.38 0.00
0.51 98.07 1.42 0.00
0.00 75.68 22.31 2.00
1.85 78.12 17.91 2.12
2.57 74.10 21.25 2.08
increase in nanoroughness. Both the average roughness (Ra, integral of the absolute value of the roughness profile height over the evaluation length) and the root mean square roughness (Rms, an average of peaks and valleys of a materials surface profile) clearly increase with APPT treatment. Specially noticeable is the increase in roughness in LDPE from as manufactured to as treated condition, possibly related to the high deformability of this material, in particular compared to its composites. This increase in nanoroughness is in agreement with the findings of other researchers in polymer materials under similar treatments [31,32]. So APPT treatment can lead to materials with more roughness to help to adhesive bonding if required as mechanical anchoring points increase, together with an increase in wettability of the surface for a selected adhesive. 3.2. Melting temperature and crystallinity After confirming the chemical and topographical effect of APPT on LDPE and LDPEB, the main objective is to evaluate the possible influence of APPT on the thermal properties of studied materials. Fig. 4 shows the melting temperature and crystallinity for the treated and untreated materials. For untreated materials, the melting temperature does not change when boron is added (Fig. 4a), and a similar effect is observed in the crystallinity (Fig. 4b). The slight variations of melting temperature can be related to the effect of boron particles, that makes difficult the disentanglement of LDPE chains during melting. The slight increase in crystallinity could be connected to a nucleating effect of the boron particles in the LDPE crystallization, as it has been also observed in LDPE/silica nanocomposites [33]. When the APPT treatment is carried out, the changes in the melting temperature are in the range of the experimental data error. The crystallinity (Fig. 4b) diminishes in all cases, although the error bars overlapped for untreated and treated composite materials. This slight decrease can be related to cross-linking during APPT treatment, as it has been suggested for high density polyethylene (HDPE) [34]. The decrease in crystallinity and no occurrence of a second peak around 65 to 95 °C in DSC curve [13] reveal no degradation in the material with APPT, as it is has been observed with thermal ageing [14,35] or low pressure oxygen plasma treatment [13]. 3.3. Decomposition temperature The TGA curves for the untreated LDPE15B composite are shown in Fig. 5a, showing changes on the decomposition temperature with heating rate. As expected, when the heating rate increases (from 5 to 20 K/min), the decomposition temperature increases. The solid residue obtained is in accordance with the boron amount in the material (15%). The percentage of the residue does not change with heating rate, but it does accordingly to the percentage of added boron. This clearly indicates that the residue after thermal treatment is pure boron, as at the maximum temperature achieved in these tests (600 °C) and they are carried out in nitrogen, avoiding boron oxidation. Therefore, the obtained mixture is perfectly homogeneous. LDPE and the composites present a relatively good thermal stability, since no significant mass loss (b 0.5%) occurs until 300 °C. The first derivative of the TGA curve (DTG, Fig. 5b) provides the decomposition temperature (peak of the curve), and its integration with time gives the degree of conversion of the reaction. Fig. 6 shows the results for the samples heated at 10 K/min. The values obtained for LDPE agree with those obtained by Paik and Kar for commercial PE [36]. These values present a small increase for the 30% boron containing composites, due to the incorporation of the particles, and it has been found for all heating rates. The same effect was found by Chrissafis and Bikiaris [37] adding particles (montmorillonite, multi-walled carbon nanotubes and SiO2) to HDPE. When APPT treatment is carried out, the decomposition temperature does not substantially change (Fig. 6). The slightly increases
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
73
Fig. 1. XPS spectra for carbon, boron and nitrogen of LDPE30B, both untreated and after APPT treatment.
found in all treated materials can be related to the presence of carbonyl and carboxyl groups in APPT treated materials. This suggests that thermal stability is not affected when APPT treatment is applied. The same behaviour was found for the other heating rates.
decomposed. Then, the degree of conversion α reached at a time t can be calculated as shown in Eq. (7). α¼
ΔWLt ΔWLTotal
ð7Þ
3.4. Kinetic analysis In order to work with the Kamal equation and to apply the Arrhenius equation, a partial integration of DTG versus time (Fig. 5b) must be done. The reaction progress will be directly proportional to the amount of weight loss (WL) and the maximum degree of conversion will be reached at the end of the reaction, when the entire polymer has been
Fig. 2. Variation of the O/C ratio with boron content and APPT treatment.
where ΔWLt is the weight variation for time t and ΔWLTotal is the total weight variation of the decomposition process. In all materials, heating to 650 °C ensures complete decomposition of the polymer, where the conversion degree is the maximum (100%). An example of the curve obtained for the conversion degree with time is shown in Fig. 7a. Fig. 7b plots the conversion rate (derivate of Fig. 7a curve) vs. the conversion degree. When this plot is fitted to the Kamal equation (Eq. (6)), the kinetic parameters are obtained (Table 3). Table 3 shows n values around 1.0 and m values around 0.6 for all materials. The goodness of fitting (R2) of the Kamal model presents values higher than 0.998. These m values indicate that more than one species is involved in the decomposition reaction of LDPE, instead of one species. The increase of both n and m values in APPT treated materials would be in agreement with the presence of more species in the material after the treatment, as XPS has shown (Table 1 and Fig. 2). Moreover, the n values around 1 are in the range of those of Baloch et al. [38] who found for commercial PE when only n-order mechanism is considered. The presence of different species would also agree with possible decomposition mechanisms proposed for polymers (randomchain scission, end-chain scission, chain-stripping or cross-linking [39]), that would possibly not follow rigorously a first order kinetics. Moreover, n and m values tend to increase when APPT treatment is done. This would imply that more species are present in the decomposition reaction. This is in agreement with XPS results (Table 1 and Fig. 2),
74
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
Fig. 3. AFM images of some materials.
that show the formation of moieties during APPT treatment that can affect decomposition process. Table 3 data indicates that the k2 constant is higher than k1 in all materials (untreated and treated), therefore the autocatalytic reaction is faster than n-order reaction. The reaction is governed by n-order mechanism, since it is the slowest reaction. For the calculations of the activation energy, only one constant was used. Two different approaches were followed: the use of the Arrhenius equation (Eq. (2)), using k2 parameters (as the autocatalytic component of the reaction will lead to higher energies) and the use of the Kissinger method (Eq. (4)). Both methods yield a linear relationship (Fig. 8), being the slope used to calculate the activation energy. Regardless of the method used to calculate the activation energy, (Arrhenius, Fig. 8a, and Kissinger, Fig. 8b), the curves are similar for the three materials, both treated and untreated. The activation energies shown in Table 4 are in the range of those obtained by Baloch et al. [38] and Paik and Kar [36]: 300.5–374.1 kJ/ mol for commercial PE, depending on the calculation method used. Values shown in Table 4, with both models, are within the ranges found for PE [38]. However, all the energies in Table 4 show differences Table 2 Roughness of LDPE and its composites, untreated and after APPT treatment. Roughness
Ra (nm) Rms (nm)
Untreated
APPT
LDPE
LDPE15B
LDPE30B
LDPE
LDPE15B
LDPE30B
23.51 29.73
28.46 36.47
41.02 50.04
66.51 84.14
69.56 85.85
60.87 72.06
of less than 10%, indicating that the effect of APPT on stability is very low, remaining its value above neat LDPE. All the energies in Table 4, calculated by both methods (Arrhenius and Kissinger) show differences of less than 10%, indicating that the effect of APPT on stability is very small and within the experimental error for these measurements. Therefore, the most important conclusion obtained from these experiments is that the presence of both the boron and the plasma treatments has a minor effect on the thermal stability of LDPE. In any case, a closer analysis of the data shows some unexpected trends. For instance, for both the pure material and the material containing 15% boron, the values of the activation energy for the treated materials are higher than those for the non-treated ones, being the difference higher for the pure material. However the contrary effect is observed for the material containing 30% boron. APPT treatment of polymers provokes the formation of oxygen containing active species. The dissociation energy of a C\O bond is higher than that of a C\H bond. Moreover, APPT can promote the formation of C_C bonds due to crosslinking. Then higher activation energy for the treated materials should be expected, independently of the boron content (Table 4). Due to this, the values of activation energy of LDPE30B, lower for treated material, cannot be related to the modifications in the composition of the polymer, but to effects related to the precision of the experiments. In this sense, one possible explanation is based on the modifications of the thermal conductivity of the materials by the presence of the particles; a higher amount of particles should increase the conductivity and then the heat transfer between the crucible and the
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
75
Fig. 4. Variation of (a) the melting temperature and (b) the crystallinity of materials, both untreated and with APPT.
sample during the TGA tests. This in turn could introduce some artefacts in the comparison of the TGA curves at high heating rates. The same type of experimental artefacts could also be the reason for the non-clear trends observed as a function of the boron content for both the non-treated and treated materials. As it is observed in Table 4, the maximum activation energy is obtained for the LDPE15B material for both calculation methods.
Fig. 5. (a) Variation of the TGA curves with heating rate, and (b) decomposition curves (TGA) and the first derivative of the TGA curve (DTG) at a heating rate of 10 K/min of untreated LDPE15B.
4. Conclusions
Acknowledgements
The APPT treatment produces chemical and morphological changes in the materials surface, as XPS and AFM show: O/C ratio increases, as a consequence of the formation of polar groups and nanoroughness is increased for all materials. The highest evaluated boron amount (30% by wt.) slightly increases the crystallinity and the decomposition temperature of LDPE. These slight variations are the consequence of the nucleation effect of boron particles on LDPE crystals. APPT treatment does not affect the melting temperature, the crystallinity and the decomposition temperature of materials. This implies that APPT does not affect the thermal stability of studied materials. In all cases, the kinetic parameters show that the decomposition reaction is mainly autocatalytic as this is the faster reaction. Reaction order is always higher than 1, indicating the presence of more than one species in the reaction. From activation energy calculations, it can be concluded that the thermal stability of LDPE and LDPEB composites is not substantially modified by APPT. The APPT treatment is adequate to treat these materials (LDPE and its boron containing composites), as it does not degrade the material, but it modifies the chemistry and nanoroughness of the surface, increasing its wettability.
Authors wish to acknowledge the Spanish Ministry of Science and Innovation (through project MAT2011-29182-C02-02) for its financial support in this research.
Fig. 6. Variation of the decomposition temperature at 10 K/min heating rate for all the tested materials.
76
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77
Fig. 8. Application of (a) the Arrhenius equation (Eq. (2)) for APPT treated materials and (b) the Kissinger method (Eq. (4)) for untreated materials to calculate the activation energy.
Table 4 Activation energies obtained by the Arrhenius (Eq. (2)) and Kissinger (Eq. (4)) methods. Arrhenius
Fig. 7. LDPE15B after APPT treatment, heated at 5 K/min rate. (a) Conversion degree vs. time. (b) Fit to the Kamal equation (Eq. (6)). LDPE LDPE APPT LDPE15B LDPE15B APPT LDPE30B LDPE30B APPT
Kissinger
E (kJ/mol) Ao (min−1)
R2
E (kJ/mol) Constant
311 354 331 339 320 314
0.9345 0.9988 0.9990 0.9945 1.0000 0.9821
321 345 322 330 312 304
1.23E 7.19E 1.93E 5.79E 2.88E 1.07E
+ + + + + +
19 22 21 21 20 20
4.46E 1.88E 5.34E 1.50E 8.16E 2.22E
+ + + + + +
R2 17 19 17 18 16 16
0.9961 0.9841 0.9998 0.9992 0.9999 0.9808
Table 3 Kinetic parameters calculated with Eq. (6) of the untreated materials and APPT treated ones (shadowed data).
LDPE
β (k/min)
TP (K)
k1
k2
n
m
R2
5
739.2 740.0 749.8 750.7 758.5 757.9 739.9 741.1 749.2 750.9 759.3 760.1 741.8 741.5 751.5 753.9 761.9 761.7
0.00007 0.00008 0.00011 0.00012 0.00028 0.00017 0.00006 0.00006 0.00007 0.00008 0.00012 0.00013 0.00006 0.00006 0.00009 0.0001 0.0001 0.00011
0.00714 0.00734 0.01448 0.01539 0.02586 0.02879 0.00768 0.00779 0.01516 0.01535 0.03028 0.03076 0.00788 0.00781 0.01551 0.01566 0.03101 0.03102
1.03 1.07 1.11 1.14 0.89 1.17 0.99 0.97 1.01 1.02 1.10 1.10 0.92 0.92 0.98 1.01 1.04 1.04
0.56 0.56 0.59 0.62 0.53 0.64 0.58 0.59 0.61 0.61 0.65 0.65 0.57 0.56 0.59 0.59 0.63 0.63
0.99817 0.99818 0.99852 0.99698 0.99807 0.99847 0.99950 0.99979 0.99930 0.99938 0.99972 0.99982 0.99963 0.99958 0.99962 0.99974 0.99989 0.99989
10 20 LDPE15B
5 10 20
LDPE30B
5 10 20
References [1] X. Colin, C. Monchy-Leroy, L. Audouin, J. Verdu, Nucl. Instrum. Methods B 265 (2007) 251–255. [2] C. Harrison, E. Burgett, N. Hertel, E. Grulke, in: S. Mathur, M. Singh, T. Ohji, A. Wereszczak (Eds.), Polyethylene/boron containing composites for radiation shielding applications, Nanostructured Materials and Nanotechnology II, 29 (8), John Wiley & Sons, Inc., New Jersey, 2009, pp. 77–84. [3] Z. Li, X. Xue, T. Jiang, H. Yang, M. Zhou, Adv. Mater. Res. 201–203 (2011) 2767–2771. [4] http://www.deqtech.com/Shieldwerx/Products/swx207hd.htm June 2012. [5] S. Cota, T.V.S. de Oliveira, M. Senne Jr., G. Pacheco, M. de Albuquerque, Prog. Nucl. Energy 51 (2009) 236–239. [6] N. Encinas, J. Abenojar, M.A. Martínez, Int. J. Adhes. Adhes. 33 (2012) 1–6. [7] C. Tendero, C. Tixier, T. Pascal, J. Desmaison, P. Leprince, Spectrochim. Acta B 61 (2006) 2–30. [8] A. Schütze, J.Y. Jeong, S.E. Babayan, J. Park, G.S. Selwyn, R.F. Hicks, IEEE Technol. Plasma Sci. 26 (1998) 1685–1694. [9] A.N. Bhoj, M.J. Kushner, IEEE Technol. Plasma Sci. 33 (2005) 250–251. [10] N. Encinas, B. Díaz-Benito, J. Abenojar, M.A. Martínez, Surf. Coat. Technol. 205 (2010) 396–402. [11] J. Abenojar, R. Torregrosa-Coque, M.A. Martínez, J.M. Martín-Martínez, Surf. Coat. Technol. 203 (2009) 2173–2180.
J. Abenojar et al. / Surface & Coatings Technology 239 (2014) 70–77 [12] W. Mueller, I. Jakob, Polym. Degrad. Stab. 79 (2003) 161–172. [13] M.R. Sanchis, V. Blanes, M. Blanes, D. Garcia, R. Balart, Eur. Polym. J. 42 (2006) 1558–1568. [14] J.V. Gulmine, P.R. Janissek, H.M. Heise, L. Alcelrud, Polym. Degrad. Stab. 79 (2003) 385–397. [15] N.F. Belkacemi, R. Coelho, J. Phys. D: Appl. Phys. 28 (1995) 1001–1008. [16] S. Bhowmik, H.W. Bonin, V.T. Bui, T.K. Chaki, J. Adhes. 82 (2006) 1–18. [17] H.E. Kissinger, Anal. Chem. 29 (1957) 1702–1706. [18] A. Al-Mulla, J. Mathew, L. Al-Omairi, S. Bhattacharya, Polym. Eng. Sci. 51 (2011) 2335–2344. [19] M.R. Kamal, Polym. Eng. Sci. 14 (1974) 231–239. [20] S. Sourour, M.R. Kamal, Thermochim. Acta 14 (1976) 41–59. [21] S.Y. Pusatcioglu, A.L. Fricke, J.C. Hassler, J. Appl. Polym. Sci. 24 (1979) 937–946. [22] M.J.M. Abadie, D. Sakkas, Eur. Polym. J. 28 (1992) 873–879. [23] T. Ozawa, J. Therm. Anal. Calorim. 2 (1970) 301–324. [24] T. Ozawa, Bull. Chem. Soc. Jpn. 38 (1965) 1881–1886. [25] J.H. Flynn, L.A. Wall, J. Res. Natl. Bur. Stand. 70 (1966) 487–523. [26] H.L. Friedman, J. Polym. Sci. Polym. Symp. 6 (1964) 183–195. [27] J. Palacios, R. Perera, C. Rosales, C. Albano, J.M. Pastor, Polym. Degrad. Stab. 97 (2012) 729–737. [28] http://srdata.nist.gov/xps/DataDefinition.aspx May 2013.
77
[29] J.F. Moulder, W.F. Stickle, P.F.E. Sobol, K.D. Bomben, in: J. Chastain, R.C. King (Eds.), Handbook of X-ray photoelectron Spectroscopy, Physical Electronics, Inc., Minnesota, 1995. [30] N. Encinas, R.G. Dillingham, B.R. Oakley, J. Abenojar, M.A. Martínez, M. Pantoja, J. Adhes. 88 (2012) 321–336. [31] K. Tamargo-Martínez, A. Martínez-Alonso, S. Villar-Rodil, J.I. Paredes, M.A. Montes-Morán, J.M.D. Tascón, J. Chromatogr. A 1218 (2011) 3781–3790. [32] S. Ercegovíc Razíc, R. Cunko, V. Svetlicíc, S. Segota, Mater. Technol. 26 (2011) 146–152. [33] C. Saiz-Arroyo, M.A. Rodríguez-Pérez, J.I. Velasco, J.A. de Saja, Compos. Part B Eng. 48 (2013) 45–50. [34] I. Novák, A. Popelka, I. Krupa, I. Chodák, I. Janigová, T. Nedelcev, M. Spirková, A. Kleinová, Vacuum 86 (2012) 2089–2094. [35] M. Omastová, S. Podhradská, J. Prokös, I. Janigová, J. Stejskal, Polym. Degrad. Stab. 82 (2003) 251–256. [36] P. Paik, K.K. Kar, Mater. Chem. Phys. 113 (2009) 953–961. [37] K. Chrissafis, D. Bikiaris, Thermochim. Acta 523 (2011) 1–24. [38] M.K. Baloch, M.J.Z. Khurram, G.F. Durrani, J. Appl. Polym. Sci. 120 (2011) 3511–3518. [39] C.L. Beyler, M.M. Hirschler, in: P.J. DiNenno (Ed.), Thermal Decomposition of Polymers. Chapter 7: Handbook of Fire Protection Engineering, 3rd ed., National Fire Protection Association, Quincy, Massachusetts, 2002.