J Polym Res (2012) 19:9814 DOI 10.1007/s10965-011-9814-0
ORIGINAL PAPER
Transport properties of natural gas through polyethylene nanocomposites at high temperature and pressure Jimoh K. Adewole & Lars Jensen & Usamah A. Al-Mubaiyedh & Nicolas von Solms & Ibnelwaleed A. Hussein
Received: 21 August 2011 / Accepted: 14 December 2011 # Springer Science+Business Media B.V. 2012
Abstract High density polyethylene (HDPE)/clay nanocomposites containing nanoclay concentrations of 1, 2.5, and 5 wt% were prepared by a melt blending process. The effects of various types of nanoclays and their concentrations on permeability, solubility, and diffusivity of natural gas in the nanocomposites were investigated. The results were compared with HDPE typically used in the production of liners for the petroleum industry. Four different nanoclays—Cloisite 10A, 15A, 30B and Nanomer 1.44P—were studied in the presence of CH4 and a CO2/CH4 mixture in the temperature range 30–70 °C and pressure range 50–100 bar. The permeability and diffusivity of the gases were considerably reduced by the incorporation of nanoclay into the polymer matrix. Addition of 5 wt% loading of Nanomer 1.44P reduced the permeability by 46% and the diffusion coefficient by 43%. Increasing the pressure from 50 to 100 bar at constant temperature had little influence on the permeability, whereas increasing the temperature from 30 to 70 °C significantly increased the permeability of the gas. Additionally, the effect of crystallinity on permeability, solubility, and diffusivity was investigated. Thus, the permeability of the CO2/CH4 mixture in Nanomer 1.44P nanocomposite was reduced by 47% and diffusion coefficient by 35% at 5 wt% loading, 50 °C, and 100 bar, compared with pure HDPE. J. K. Adewole : U. A. Al-Mubaiyedh : I. A. Hussein (*) Department of Chemical Engineering, King Fahd University of Petroleum & Minerals, Dhahran 31261, Saudi Arabia e-mail:
[email protected] L. Jensen : N. von Solms Department of Chemical and Biochemical Engineering and Center for Energy Resources Engineering, Technical University of Denmark, Lyngby, Denmark
Keywords Polyethylene . Organoclay . Nanocomposites . Permeability . Diffusion . Solubility
Introduction Polymer-based nanotechnology is an active area of research and development in both basic and applied research [1–3]. Polymers with good barrier and reinforcement properties are of interest to some researchers. In particular, polyolefins are the largest polymer family by consumption and production. Within the polyolefin family, polyethylene is the most widely used ethylene-based commodity polymer [4–8]. Today, polyethylene (PE) is used in both industrial and domestic applications owing to its inherent properties as well as its ability to be processed in order to improve its properties. For example, the offshore petroleum industry still uses steel pipes as risers and flowlines for transporting crude oils and gases from well heads to separation stations. Such pipes are coated with an internal polymeric liner (sleeve) in order to avoid corrosion of the steel by sour gases and ensure the integrity of the pipe network. Unfortunately, these liners tend to allow gas to go through at high pressure and temperature. Natural gas is mainly methane but also contains carbon dioxide and hydrogen sulfide in varying amounts. Permeated gases collect within the annulus between the sleeve and the steel wall. Permeation of acidic gas such as H2S can cause corrosion of steel pipes. Permeation of CH4 can cause pressure build up in the annulus between the liner and the pipe to an extent that can result in liner failure. Investigation of this problem has revealed that pipe failure might result purely from mechanical failure due to continuous stress cycles and creep, from transport problems resulting from gas permeation through the liner at high pressure and temperature,
Page 2 of 11
or a combination of these two factors. In this study, incorporation of nanoclay into the PE is proposed to improve barrier properties of the liner. Recently, we found that the addition of nanoclay reduced creep in PE nanocomposites [9]. Here, we assess the impact of the clay on the permeation of gas through the nanocomposite. There has been limited previous work on the transport properties of gases through high density polyethylene (HDPE) nanocomposites. Most of the studies have considered morphology, thermal and mechanical properties [7, 8], flammability and creep [6, 9–15]. Although there has been work on the transport of pure gases in pure HDPE [16], HDPE nanocomposites [17, 18], and LDPE nanocomposites [19, 20], none of these studies were performed at high pressure or high temperature. The investigation of transport properties of natural gas in PE nanocomposites at typical operating conditions of high pressure and high temperature has yet to be carried out and is therefore the subject of this work. We investigate the influence of various nanoclays on transport properties of pure and mixed gases in PE nanocomposites. Four different organoclays were used: Cloisite® 10A, Cloisite ®15A, Cloisite® 30B, and Nanomer® 1.44P (C10A, C15A, C15B, and N1.44P). Pure CH4 and a CH4/CO2 mixture were examined in the temperature range 30–70 °C and for pressures from 50 to 100 bar. The temperature and pressure ranges selected were based on typical operating conditions in the Gulf region. In addition to the permeability measurements, solubility and diffusion coefficients were also estimated. In order for a nanoclay to be successful as an additive for liner applications, it should improve both the creep and barrier properties of the PE nanocomposites. The addition of an optimum amount of nanoclay has previously been shown to reduce creep [9]. The impact of clay additives on gas transport properties of the liner, particularly permeability, is the subject of this study. Although the presence of these additives may limit the diffusion of natural gas in the liner, there is a need to screen several clays of different structures for potential applications. Thus, while it is desirable to decrease the permeability of the resulting PE–clay liner, this should not compromise other essential properties such as mechanical and thermal resistance. Results of the investigation of the thermal, bulk, and surface mechanical properties, with special emphasis on creep, were recently published [9].
Materials and methods Permeability measurement The composition of natural gas was represented as either pure CH4 or a mixture of CH4 and CO2 [20]. The determination of permeability of pure CH4 and the mixture of CO2/CH4 in PE, PE/PE-graft-maleic anhydride (PE-g-MA), and PE/PE-g-MA
J.K. Adewole et al.
nanocomposites was performed using a high pressure 2-D permeation cell. The PE/PE-g-MA nanocomposites were prepared using different nanoclay types and with different nanoclay loadings. A detailed description of the preparation of the nanocomposites was presented elsewhere [9]. The permeation experiments were conducted at the temperatures 30, 50, and 70 °C and at pressures 50 and 100 bar. The relative standard deviation of the measured permeability was estimated to be 10%. The error is thought mainly to arise from a relatively large variation in the thickness within each single polymer sample (the thickness was measured at various spots on the sample including the edge and middle). However, in general, the average thickness of different samples was quite similar. Materials An HDPE (HE3490-LS) with a melt flow index of 0.25 g/10 min (190 °C/2.16 kg) and a density of 959 kg/m3, supplied by Borouge Company, UAE, was used as matrix. This grade is usually used for liner applications by Saudi Aramco. PE-g-MA was used as a compatibilizer and acquired from Sigma Aldrich. PE-g-MA contained approximately 3 wt% maleic anhydride; its viscosity was 1,700–4,500 cP and melt temperature was 105 °C. The commercial organoclays Cloisite® 10A, Cloisite ®15A, and Cloisite® 30B were supplied by Southern Clay Products, USA; Nanomer® 1.44P was supplied by Nanocor, USA. The compositions of the polymers and the polymer–clays used were: & & &
Pure PE PE/PE-g-MA (mass ratio 98:2 by weight) PE/PE-g-MA / nanoclay with clay loading of 1, 2.5, or 5 wt%
Natural gas is mainly composed of CH4 with variable amounts of CO2. In this study, two compositions were used: pure CH4 (AGA, 99.5%) and a mixture of 80 mol% CH4 and 20 mol% CO2 (AGA, 99.5%). Description of equipment Permeability coefficients of the gases were measured in a 2-D permeation cell. The high pressure 2-D permeation cell was designed and manufactured by the Department of Chemical and Biochemical Engineering at Technical University of Denmark. The operating conditions of the cell are up to 150 °C and 700 bar absolute. The cell consists of two chambers: a high pressure chamber (upstream) and a low pressure chamber (downstream) (Fig. 1). The polymer sample, a disc of about 10 cm in diameter and 1–2 mm thick, is sandwiched between the two chambers and supported by two porous plates. The plates allow gas to freely contact the sample, while at the same time preventing sagging of the polymer sample. The
Transport properties of natural gas
Page 3 of 11
The pressure in the downstream chamber increases as the gas permeates from the up- to the downstream side. Consequently, the pressure will also increase in the upstream chamber because an automatic controller ensures that the experiments are performed at a constant differential pressure across the polymeric membrane. In practice this is done by running the ISCO pump in constant pressure mode, using the up- and downstream pressure difference as an input to the pump. The thickness of the polymer samples was measured using a micrometer screw. The thickness was measured at 15 different spots on the sample and averaged accordingly. Theory and data analysis Fig. 1 Schematic of the 2-D permeation cell with pressure and temperature monitoring and control and data acquisition points
The permeability coefficient (Pe) is calculated as [16]: Pe ¼
ð1Þ
where Q is the volume of gas permeated through the membrane (converted to STP, i.e., standard conditions of 0 °C and 1 atm), l is the average thickness of the membrane, A is the cross-sectional area of the membrane exposed to the gas (38.48 cm2), t is the elapsed time, and ΔP is the pressure difference across the membrane. The volume of gas Q is given as: Q¼
V ðρ final ρ initial Þ ρSTP
ð2Þ
where V is the volume of the downstream chamber (22.057 cm3), ρSTP is the density at standard conditions, ρfinal is the density of the gas at the final (downstream) temperature
Pressure (Bar)
downstream chamber has a free internal volume of 22.057 cm3, whereas the upstream chamber has a variable volume due to the metal bellow placed inside the chamber. It is possible to decrease or increase the volume of the upstream chamber by injection or removal of water from the bellow using an ISCO pump (Teledyne 100DX syringe pump). As a result of this construction, permeability experiments at constant differential pressure across the polymer membrane are possible. The two chambers are held together by two stainless steel flanges that are securely fastened by eight heavy-duty stud bolts. The temperature in the chambers is controlled by circulating hot glycerine into two heating jackets surrounding the chambers. The temperature is measured by a Pt-100 thermocouple (RS Components, ±0.1 °C) placed in the gap between the two chambers close to the location of the polymer membrane. The pressures in the two chambers are measured by two Fisher-Rosemount pressure transducers. For safety reasons, the upstream side is connected to a pressure transmitter–switch system which shuts down the heating bath and the ISCO pump in case the pressure exceeds the cell limit. The entire cell after assembly is mounted on a frame in a fume hood. The up- and downstream pressures, the temperature, and the amount of liquid injected by the ISCO pump are recorded continuously on a computer during the whole experiment.
Ql A t $P
Method Any residual gas in the polymeric membrane was evacuated by applying vacuum to both sides of the membrane for several hours. The test gas (CH4 or CH2/CO2) was then charged to the upstream side; in this case the pressure difference between the up- and downstream side was chosen to be around either 50 or 100 bar. The temperature of the cell was set to 30, 50, or 70 °C.
Fig. 2 Data obtained from a permeation experiment. The polymer sample was PE and the gas used was CH4. The temperature was 70.1 °C and the average pressure difference between the up- and downstream side was 102.5 bar absolute
Page 4 of 11
a 3000
E DP BH
2500
5A
C1 % .5 +2
Fig. 3 a XRD patterns for clay and PE/clay nanocomposite. b SEM micrographs of PE/clay nanocomposites (left, a) magnified image (5 μm) of BHDPE +2.5 wt% C15A from Adewole et al. [9] and (right, b) BHDPE +5 wt% N1.44P (40 μm)
J.K. Adewole et al.
Intensity (CPS)
2000
N1.44P C15A
1500
1000
500
0 2
3
4
5
6
7
8
9
10
2θ
b a
and pressure, and ρinitial is the density of the gas at the starting (downstream) temperature and pressure. Figure 2 shows a run obtained from a typical permeability experiment. Specifically, data represented on this figure were obtained for methane gas permeating through PE at a pressure difference of 102.5 bar and a temperature of 70.1 °C. As shown, it took around 6,000 s before the pressure starts increasing in the downstream chamber. This is a consequence of the time taken for the gas to diffuse through the membrane. Once methane has penetrated the PE membrane, the pressure is seen to increase at a constant rate throughout the remainder of the experiment.
b
Results and discussion Morphological characterization X-ray diffraction and scanning electron microscopy analyses Details of the morphological characterization of the nanocomposites samples are described elsewhere [9]. Both X-ray diffraction (XRD) and scanning electron microscopy (SEM) techniques were employed to study the morphology. Figure 3a shows XRD patterns of BHDPE+C15A and BHDPE+N1.44P redrawn from Adewole et al. [9]. The XRD results for neat
Transport properties of natural gas
organoclay (C15A and N1.44P) are also included for comparison. The organoclay pattern shows an intense peak at around 2θ04.02° for N1.44P, corresponding to a basal spacing of 21.96 Å and 2θ04.6° which is equivalent to a spacing of 19.19 Å for C15A. Addition of 2.5 wt% of C15A to BHDPE increase the basal spacing to 34.75 Å (2θ02.54), an increase equivalent to 81.08% compared to the neat organoclay. This shift of peaks to lower angles is an indication of penetration of the polymer chains into the clay gallery as discussed in a previous publication [9] and references therein. Also, the absence of peaks for nanocomposites with 5 wt% N1.44P is an indication of finely distributed tactoids which is evident in the SEM results [9]. To complement the XRD tests, SEM analysis was performed using a Nova™ NanoSEM 230 field emission scanning electron microscope. Micrographs of the nanocomposites in Fig. 3b show that the nanoclays are well dispersed. A detailed discussion of the SEM results for these nanocomposites was given elsewhere [9].
Page 5 of 11 Table 2 Permeability coefficients obtained for PE/PE-g-MA (BHDPE) Mole fraction
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
CH4
CO2
1
0
50.6
103.1
2.09
1 0.8
0 0.2
70.0 50.47
104.8 104.5
5.09 3.13
0.8
0.2
69.9
101.9
6.98
0.8
0.2
70.1
52.0
7.03
the effect of temperature is represented by an Arrhenius equation: EP Pe ¼ Peo exp ð3Þ RT
The permeability values of pure CH4 and CH4/CO2 mixture in PE and its nanocomposite were measured as functions of pressure and temperature and the results are presented in Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, and 11 for pure HDPE and its nanocomposites. For the temperature range 30–70 °C, significant effects were noticed for all samples. For instance, increasing the temperature from 31 to 51 °C increases the permeability by a factor of three for pure HDPE as shown in Table 1. These results are expected because the increase in temperature will increase the mobility of both HDPE chain and the gas molecules resulting in enhanced diffusion [21]. In addition the solubility of methane in the polymer increases with increasing temperature, so the permeability (which is a product of the diffusivity and the solubility) must increase. The wealth of relevant literature data (see, e.g., refs. [21–24]) suggests that transport coefficients are dependent on temperature at a given pressure. For permeability,
where Peo is a temperature-independent constant, Ep is the apparent activation energy of permeation, T is the absolute temperature, and R is the universal gas constant [25]. An increase in permeability following an increase in temperature was observed for all the samples. Thus, the values of the measured permeabilities for pure HDPE were in agreement with what was reported in the literature. For example, the permeability coefficient of CH4 in HDPE at 298 K and 1 bar was reported by Alamo and Mandelkern [26] to be 0.3 × 10−8 cm3 (STP) cm−1 s−1 bar−1, whereas a value 0.58× 10−8 cm3 (STP) cm−1 s−1 bar−1 was obtained in this work for the same gas at 304 K and 106 bar. The difference in the values is due to fact that both measurements were taken at different temperature and pressure. Moreover, the effect of temperature can be further assessed by comparing our results with those of Flaconnèche et al. [16], who reported that the permeability of CH4 in HDPE was 5.1×10−8 cm3 (STP) cm−1 s−1 bar−1 at 350 K and 100 bar. In the present work, a permeability of 4.35× 10−8 cm3 (STP) cm−1 s−1 bar−1 was obtained for pure HDPE at 342.5 K and 102 bar as shown in Table 1 (confirming that the permeability is lower because at lower temperature).
Table 1 Permeability coefficients obtained for pure HDPE samples
Table 3 Permeability coefficient obtained for BHDPE+1 wt% C15A
Influence of temperature
Mole fraction CH4
CO2
1 1 1 0.8 0.8 0.8 0.8
0 0 0 0.2 0.2 0.2 0.2
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
30.9 50.8 69.5 31 50.1 69.8 70.2
105.7 104.6 101.8 105.2 100.8 102.1 54.6
0.573 1.87 4.35 1.01 3.03 5.83 6.54
Mole fraction CH4
CO2
1 1 1 0.8 0.8 0.8 0.8
0 0 0 0.2 0.2 0.2 0.2
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
30.9 49.9 69.6 30.86 50.2 69.7 69.5
104.3 104.9 105.1 104.3 104.9 103.4 57.5
0.609 1.87 4.94 1.06 2.77 6.71 6.84
Page 6 of 11
J.K. Adewole et al.
Table 4 Permeability coefficient obtained for BHDPE+2.5 wt% C15A Mole fraction
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
CH4
CO2
1
0
50.3
105.2
1 0.8
0 0.2
69.7 50.5
0.8
0.2
0.8
0.2
Table 6 Permeability coefficient obtained for BHDPE+2.5 wt% C10A Mole fraction
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
CH4
CO2
1.95
1
0
50.30
104.8
1.86
102.9 104.5
5.19 2.91
1 0.8
0 0.2
69.7 50.20
104.8 105.0
4.86 2.76
69.8
101.5
7.13
0.8
0.2
69.7
103.8
6.47
69.7
53.5
7.11
The evolution of permeability with gas pressure has been reported to be dependent on the type of permeating molecule, although the effect of pressure on the transport coefficients of gases in polymer is not yet clear—both an increase and a decrease in permeability with increase in pressure have been reported. For instance, Naito et al. [19] studied the effect of pressure on gas permeation through LDPE and PP at constant temperature of 25 °C and a pressure in the range 1–130 atm. Their results showed that the permeability of highly soluble gases such as CH4 and CO2 increases slightly with increasing pressure. Similar results were obtained by other researchers [23, 27, 28]. However, some of the results obtained by Flaconnèche et al. [16] showed that in the range 40–100 bar, the pressure difference has no significant effect on gas permeability. Pressure increment from 40 to 100 bar at 80 °C resulted in an insignificant reduction in permeability from 29×10−8 to 28× 10−8 cm3 (STP) cm−1 s−1 bar−1. Ghadimi et al. [29] reported a reduction in permeability for both single gases and gas mixtures in LDPE and polydimethylsiloxane (PDMS). In the study by Flaconnèche et al. [16], permeability of CH4 decreased from 4.3×10−8 to 3.7×10−8 cm3 (STP) cm−1 s−1 bar−1 when pressure increased from 40 to 100 bar and then increased again when the pressure was reduced from 100 to 75 bar at constant temperature of about 40 °C. The results obtained in the current paper for pure HDPE are in general agreement with these studies. For instance, when the temperature was increased by about 20 °C (from 50.1 to 70.2 °C) and the pressure decreased by about 50 bar (from 100.8 to 54.6 bar), permeability increased by 62%
despite the decrease in pressure (Table 1). Here, however, the effect of pressure is overwhelmingly overshadowed by the effect of the temperature increase (as observed in other similar cases in the literature). The effect of pressure is clearer at constant temperature of about 70 °C. A pressure increase from 54.6 to 102.1 bar caused an 11% decrease in permeability. This is not unexpected, because a pressure increase has two opposing effects as explained by Klopffer and Flaconnèche [22]: An increase in the pressure leads to an increase in polymer density via polymer compaction which consequently reduces the free volume available for the penetrant molecule to permeate through and thus a decrease in permeability. At the same time, the increase in pressure increases the gas concentration. A critical look at the change in permeability for BHDPE (Table 2) and its nanocomposites (Tables 3, 4, 5, 6, 7, 8, 9, and 10) showed that there was very little or practically no change in permeability when the pressure was doubled. For example, permeability changes of 0.71% (for BHDPE), 1.82% (for BHDPE+1 wt% C15A), 0.28% (for BHDPE+2.5 wt% C15A), and 0.46% (for BHDPE+5 wt% N1.44P) were observed when changing the pressure by 50 bar. These results show that the influence of pressure is not significant. Published permeability data were not available for comparison with the current results for HDPE nanocomposites. However, these results were in agreement with previous studies when compared with other polymer nanocomposites. For example, the results of Merkel et al. [30] on permeability of gases through poly(1-trimethylsilyl-1-propyne) showed that pressure dependence on permeability varies with nanoparticle contents. The phenomenon that occurs in the nanocomposites can be explained by the second opposing effect. When pressure
Table 5 Permeability coefficient obtained for BHDPE+5 wt% C15A
Table 7 Permeability coefficient obtained for BHDPE+2.5 wt% C30B
Influence of pressure
Mole fraction CH4
CO2
1 1 0.8 0.8
0 0 0.2 0.2
8
3
−1
T (°C) P (bar) Pe ×10 (cm (STP) cm
49.7 69.2 49.7 69.3
103.4 105.6 102.4 104.5
1.91 4.90 2.81 6.80
s
−1
−1
bar )
Mole fraction CH4
CO2
1 1 0.8 0.8
0 0 0.2 0.2
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
50.2 68.7 50.2 68.8
104.0 103.3 104.8 104.8
2.13 4.68 2.76 6.45
Transport properties of natural gas Table 8 Permeability coefficient obtained for BHDPE+1 wt% N1.44P Mole fraction
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
Page 7 of 11 Table 10 Permeability coefficient obtained for BHDPE+5 wt% N1.44P Mole fraction
CH4
CO2
1
0
50.10
103.6
1.92
1 0.8
0 0.2
69.90 50.2
105.2 105.3
5.23 1.70
0.8
0.2
69.90
104.9
7.17
was increased there is also a corresponding increase in penetrant concentration and the diffusing molecule can plasticize the macromolecular chains, which results in an increased free volume. Thus, the opposite effect is suspected to have been responsible for keeping the permeability constant. It can also be inferred that these two opposing effects prevail in BHDPE, BHDPE+C15A, and BHDPE+N1.44P samples while only the hydrostatic pressure effect prevails in pure HDPE. Influence of compatibilizer As explained earlier, BHDPE is a blend of HDPE and PE-gMA which was used as a compatibilizer. The permeability results for BHDPE are shown in Table 2. The presence of compatibilizer was observed to slightly increase the permeability at 50 °C and 100 bar as well as at 70 °C and 100 bar. These results are in agreement with Picard et al. [17]. Influence of nanoclay Tables 3, 4, 5, 6, 7, 8, 9, and 10 and Figs. 4, 5, and 6 represent pure CH4 and mixed CH4/CO2 permeability in HDPE containing various amounts of nanoclay loadings. The incorporation of nanoclays into PE has been found to improve gas barrier property because the nanoclays create tortuous paths that retard the gas molecules’ movement through the polymer [10]. However, this is not always the case for nanoparticle-filled polymers [31] as can be seen in the results of the current study. Permeabilities of some samples increased, some decreased, whereas others remained constant. For instance, permeability of pure CH4 dropped by about 3.6% and that of the mixture Table 9 Permeability coefficient obtained for BHDPE+2.5 wt% N1.44P Mole fraction CH4
CO2
1 1 0.8 0.8
0 0 0.2 0.2
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
50.60 69.90 50.60 69.90
103.5 103.1 103.4 103.6
1.89 4.93 4.44 6.70
T (°C) P (bar) Pe ×108 (cm3 (STP) cm−1 s−1 bar−1)
CH4
CO2
1
0
50.16
104.4
1.87
1 0.8
0 0.2
69.31 50
104.6 104.2
4.71 1.62
0.8
0.2
69.22
104.3
6.53
0.8
0.2
69.19
52.0
6.56
dropped by 11.9% with the addition of 1 wt% of C15A in comparison with pure HDPE as shown in Table 3 and Figs. 4 and 5. Generally, with increasing nanoclay content, the barrier properties are expected to improve. However, the permeability of pure gas was practically constant at low temperature but increased by 14.8% at high temperature with the addition of 2.5 wt% C15A (Table 4). For the gas mixture, the permeability decreased by 4% at low temperature and increased by about 8% at high temperature. Similarly, nanocomposites formed from N1.44P interestingly showed lower permeability than pure HDPE. For instance, the permeability of pure CH4 was reduced by 3.6% at low temperature and increased by 4.2% at high temperature (Table 10). For natural gas, the barrier property improved by about 44% (Table 8) and 46.5% (Table 10) at low temperature and high pressure by the addition of 1 wt% and 5 wt% of N1.44P, respectively. Permeability remained almost constant at high temperature and low pressure (70 °C and 50 bar) especially at 1 wt% loading for C15A (Table 3) and 5 wt% loading for N1.44P (Table 10). At high temperature and pressure (70 °C and 100 bar), there was an average increase in permeability for all samples by about 11.8% in comparison with neat HDPE. The results are in agreement with those reported from similar studies by Merkel et al. [30], Lee et al. [32], Matteucci et al. [31], and Picard et al. [17] for pure PE and other polymers. A possible explanation for the increase in permeability was offered by Matteucci et al. [31] and Merkel et al. [30]: Nanoparticles can inhibit the efficient segmental chain packaging in polymers thereby increasing free volume in the polymer phase which consequently increases the permeability. Also, in other heterogenous polymer systems such as rubbery polymer nanocomposites, voids at the polymer– particle interface or between particle aggregates cause permeability to be greater in nanocomposites than in unfilled polymers. The free volume increases with temperature; hence, permeability is increased. Influence of gas concentration Although constant gas concentration was used throughout the experiment, effects of concentration changes come into
Page 8 of 11
J.K. Adewole et al.
Table 11 Transport properties of CH4 in PE/nanoclay composite Difference S×102 Difference (%)a D×107 Difference Pe ×108 3 −1 −1 −1 a 3 −3 −1 (cm (STP) cm s bar ) (%) (cm (STP) cm bar ) (cm2 s−1) (%)a
Nanoclay loading Nanoclay type (wt%)
4.950
0.00
3.92
0.00
1.0 2.5
0.0 Cloisite 15A Cloisite 15A
1.87 1.95
1.94
−3.61 0.52
0
4.943 4.615
−0.13 −6.76
3.78 4.23
−3.54 7.78
5.0
Cloisite 15A
1.91
−1.55
4.573
−7.62
4.18
6.56
1.0 2.5
Nanomer 1.44P 1.92 Nanomer 1.44P 1.89
−1.03 −2.58
4.108 4.712
3.19 −4.80
3.76 4.01
−4.16 2.29
5.0
Nanomer 1.44P 1.87
−3.61
4.094
−17.29
4.57
16.52
S solubility, D diffusion coefficient a
Indicates the change in permeability, solubility, and diffusivity of nanocomposites in comparison with the pure polymer
play while considering the impact of the presence of one gas on the permeability of another. Moreover, the increase in permeability of one penetrant in the presence of another has been reported previously. For example, Jordan and Koros [33] reported an increase in CH4 and N2 permeability in PDMS in CO2/CH4 and CO2/N2 mixtures. This increase was explained (using the free volume analysis) by hypothesizing that highly sorbing CO2 plasticized the polymer matrix and this resulted in an increase in light gas permeability. In the current study, at the same pressure and temperature (50 °C and 100 bar), the permeability of mixed gas was about twice that of pure gas in HDPE. At 70 °C and 100 bar, the permeability of CH4/CO2 mixture is about 1.5 and 1.2 times that of pure CH4 in BHDPE and BHDPE+C15A respectively. We suggest that the sorbed gas plays the role of a plasticizer and this role diminishes at high temperature owing to increased polymer mobility and decreased sorption. In fact, the reduction in permeability enhancement
factor from 3 to 1.5 and 1.2 is likely due to the effect of increased temperature on the plasticization process. It has also been shown that CO2 solubility decreases at increasing temperatures [16]. In addition, the type of additive or polymer is another factor that influences the sorption process. Hence, the factors that affect permeation are not limited to plasticization but rather a complex sorption/desorption process that depends on the composition of the composite and the nature of additives and functional groups on the surface of the additives as well as the temperature. Influence of crystallinity on transport coefficients The degree of crystallinity is an important property of semicrystalline polymers that influences the transport behavior of gases in polymers [16, 23]. Transport and thermodynamic properties such as permeability, solubility, and diffusivity 4.5
2.7
2.5
1wt% C15A
2.5wt% C15A
4 5wt% N1.44P
2.4
2.3
Permeability (barrer)
Permeability (barrer)
2.6
Pure HDPE
2.5wt% C15A
Pure HDPE
1wt% C15A
3.5
3
2.2
2.5 5wt% N1.44P
2.1
2
2 Pure CH4
Fig. 4 Evolution of transport properties of pure CH4 as a function of nanoclay type and loadings at 50 °C and 100 bar
Mixed CH4/CO2
Fig. 5 Evolution of transport properties of mixed CH4/CO2 as a function of nanoclay type and loadings at 50 °C and 100 bar
Transport properties of natural gas 8
11
7
2.5wt% C15A Pure HDPE
9
1wt% C15A
Pe* 108(cm3(STP)/cm.s.bar)
12
10
Permeability (barrer)
Page 9 of 11
5wt% N1.44P
8 7 6 5
CH4 (Pure PE+1wt% C15A) CH4/CO2 Mixture (Pure PE+1wt% C15A) CH4 (Pure PE) CH4/CO2 Mixture (Pure PE)
6 5 4 3 2
4
1
3
0 0
0.1
0.2
2
0.3
0.4
0.5
a
Mixed CH4/CO2
are related to free volume present in the polymer matrix [23]. In this section we attempt to correlate permeability to the free volume. The free volume was calculated from the measured crystallinity. The detailed procedure for crystallinity measurement was presented elsewhere [9]. In the present work, the measured degree of crystallinity (Xc) was used to calculate the volume fraction of amorphous phase (6a) for the polymer samples: 6a ¼ ð1 Xc Þ
ρ ρa
ð4aÞ
where ρ is the density of the semicrystalline polymer and ρa is the density of the amorphous phase of the polymer. The density of the amorphous phase is taken to be 0.855 g/cm3 [16] and the density of PE at room temperature as provided by the supplier was 0.959 g/cm3. Densities of C10A, C15A, C30B, and N1.44P nanoclays were 1.90, 1.66, 1.98,and 1.75 g/cm3 as obtained from the manufacturers. The density of the nanocomposite was calculated using ρnc ¼ ρm vm þ ρf vf
ð4bÞ
where ρnc, ρm, and ρf are densities of nanocomposite, polymer matrix, and filler, respectively [34], and vm and vf are the volume fractions for the matrix and filler, respectively. For the density at higher temperatures a thermal expansion coefficient of 0.001003 cm3 g−1 °C−1 was used in the equation below [35]. 1=ρ ¼ a þ bT
ð4cÞ
where ρ is the density at temperature T in °C, and b is the volume expansion coefficient. The effect of crystallinity on permeability, solubility, and diffusion coefficient of pure HDPE and its nanocomposites
Fig. 7 Permeability dependence on volume fraction of amorphous phase (pure HDPE and 1 wt% C15A nanocomposite)
is illustrated in Figs. 7, 8, and 9. As shown, the permeability of gases increase with the increase in 6a, in agreement with previous reports [16]. The permeability of the CO2/CH4 mixture was higher than that of pure CH4 for all samples. The diffusion coefficients of both CH4 and CO2/CH4 mixture were almost equal in PE and PE/nanocomposite at low 6a. The results of permeability and solubility obtained from this estimation compare very well with previous literature. For example, the solubility coefficient of pure CH4 in a PE sample with 6a of 0.37 was 0.056 cm 3 (STP) cm −3 bar −1 at 77 °C whereas in our work, 0.049 cm3(STP) cm−3 bar−1 was obtained for a PE sample with 6a of 0.39 at 70 °C (Fig. 8). 0.1 CH4 (Pure PE+1wt%C15A) CH4/CO2 Mixture (Pure PE+1wt%C15A) CH4 (Pure PE) CH4/CO2 Mixture (Pure PE)
0.09 0.08
S (cm3(STP)/cm 3.bar)
Fig. 6 Evolution of transport properties of mixed CH4/CO2 as a function of nanoclay type and loadings at 70 °C and 50 bar
0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 0
0.1
0.2
0.3
0.4
0.5
a
Fig. 8 Solubility dependence on volume fraction of amorphous phase (pure HDPE and 1 wt% C15A nanocomposite)
Page 10 of 11
J.K. Adewole et al.
10
respectively [36]. The Lennard-Jones parameter for the gas mixture was calculated using a geometric mean combining rule: r ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi " " " ð5bÞ ¼ K CH4 CO2 K CH4 K CO2
CH4 (Pure PE + 1wt%C15A) CH4/CO2 Mixture (Pure PE + 1wt% C15A) CH4 (Pure PE) CH4/CO2 Mixture (Pure PE)
9 8
2
D*107 (cm /s)
7 6
D was then calculated from the measured Pe and the estimated S:
5 4
D¼
3 2 1 0 0
0.1
0.2
0.3
0.4
0.5
a
Fig. 9 Diffusion coefficient dependence on volume fraction of amorphous phase (pure HDPE and 1 wt% C15A nanocomposite)
Solubility and diffusion coefficient of CH4 and CO2/CH4 mixture in PE nanocomposite The permeability coefficient is the critical parameter in the design of new barrier materials [9]. Because it is the product of the diffusion coefficient (D) and solubility coefficient (S), knowledge of these two parameters can provide an insight into the mechanisms responsible for the improvement or degradation in the barrier properties of N1.44P and C15A nanocomposites. Solubility of CH4 and the CO2/CH4 mixture was estimated in order to investigate the effect of nanoclay on these transport coefficients: ρ " S ¼ ð1 Xc Þ exp 5:07 þ 0:022 ð5aÞ ρa K The Lennard-Jones parameter K" for pure CH4, CO2, and a mixture of both gases is 136.5, 190, and 161,
Pe S
ð6Þ
Equation 5a was derived by combining the Michaels and Bixler model for gas solubility [37] and a correlation for estimating solubility of gases in totally amorphous PE [22]. The solubility results obtained from this estimation are presented in Tables 11 and 12 and are found to be consistent with measured solubility [22]. Therefore, the results obtained in Figs. 8 and 9 support the use of Eqs. 5a, 5b, and 6 in estimating the solubility and diffusivity. The permeability results discussed earlier revealed that N1.44P was the best additive because it reduced the permeability throughout the nanocomposite. Tables 11 and 12 show the estimated S and D for CH4 and CO2/CH4 mixture in C15A and N1.44P. The observed trends in these tables help explain the variation in transport properties of gases with the type of nanoclay used. From these tables, the highest drop in CH4 solubility was achieved at 5 wt% loading of N1.44P, whereas the highest CH4 diffusion coefficient was obtained at 1 wt% loading of N1.44P (Table 11). For natural gas, the 5 wt% N1.44P nanocomposite is the only sample with all gas transport properties lower than that of pure HDPE. Thus, the 5 wt% N1.44P is considered a good candidate for blending with PE liners to limit the permeability of natural gas in liners. The diffusion of CH4 in N1.44P varied linearly with loading, whereas that of C15A showed a maximum at intermediate concentration of 2.5%. Generally, the decrease in the permeability of pure CH4 in C15A and N1.44P nanocomposites can be attributed to the decrease in
Table 12 Transport properties of CO2/CH4 mixture in PE/nanoclay composite Nanoclay loading Nanoclay type (wt%)
Difference S×102 Difference (%)a D×107 Difference Pe ×108 3 −1 −1 −1 a 3 −3 −1 (cm (STP) cm s bar ) (%) (cm (STP) cm bar ) (cm2 s−1) (%)a
0.0 1.0 2.5 5.0 1.0 2.5 5.0
3.03 2.77 2.91 2.81 1.70 4.44 1.62
a
Cloisite 15A Cloisite 15A Cloisite 15A Nanomer 1.44P Nanomer 1.44P Nanomer 1.44P
0.00 −8.58 −3.96 −7.26 −43.89 46.53 −46.53
8.478 8.474 7.912 7.839 8.756 8.078 7.018
0.00 −0.05 −6.68 −7.54 3.28 −4.72 −17.22
Indicates the change in permeability, solubility, and diffusivity of nanocomposites in comparison with the pure polymer
3.57 3.27 3.68 3.58 1.94 5.49 2.31
0.00 −8.52 2.95 0.37 −45.67 53.85 −35.37
Transport properties of natural gas
their solubility coefficient (Table 11). This trend is in line with results reported by Dhingra and Marand [38] on the permeability of CH4 in thermoplastic polyimide. On the other hand, the permeability behavior of N1.44P is quite different for both pure CH4 and the CO2/CH4 mixture at 5 wt% loading. For example, a decrease in the permeability of CH4 was observed to be due to the drop in solubility, whereas the decrease of the permeability of the CO2/CH4 mixture could be due to the decrease in both the solubility and the diffusion coefficient. One of the reasons for the excellent behavior of N1.44P, especially at 5 wt% loading, can be attributed to the fact that it is a surface compatibilized montmorillonite.
Conclusion HDPE nanocomposites containing different fillers at different loadings were prepared through melt blending. HDPE grades that are typically used in gas liners were selected for this study. Values for permeability of CH4 and a mixture of CO2 and CH4 in the nanocomposites were measured in a 2D permeation cell. The selected gas simulates the composition of natural gas. Effects of pressure, temperature, amorphous phase fraction, and nanoclay type were investigated. Our measurements revealed that N1.44P was the additive that led to the highest drop in permeability. The permeability of CO2/CH4 mixture in N1.44P nanocomposite was reduced by 47% and the diffusion coefficient by 35% at 5 wt% loading, 50 °C, and 100 bar, compared with pure HDPE. The decrease in the permeability of pure CH4 in N1.44P nanocomposite was attributed to the significant decrease in its solubility, whereas the decrease of the permeability of the CO2/CH4 mixture was explained by the decrease in both the solubility and the diffusion coefficient. One of the reasons for the excellent barrier properties of N1.44P, especially at 5 wt% loading, is the fact that it is a surface compatibilized montmorillonite. Thus, N1.44P nanoclay is recommended for use in liner applications following pilot testing.
References 1. Paul DR, Robeson LM (2008) Polymer 49:3187–3204 2. Xiao J, Hu Y, Kong Q, Song L, Wang Z, Chen Z, Fan W (2005) Polym Bull 54:271–278 3. Ahmadi SJ, Hunang YD, Li W (2004) J Mater Sci 39:1919–1925 4. Nwabunma D (2008) In: Nwabunma D, Kyu T (eds) Polyolefin composite. Wiley-Interscience, New Jersey 5. Hu Y, Lu H, Song L (2008) In: Nwabunma D, Kyu T (eds) Polyolefin composite. Wiley-Interscience, New Jersey
Page 11 of 11 6. Deka BK, Maji TK, Mandal M (2011) Study on properties of nanocomposites based on HDPE, LDPE, PP, PVC and clay. Polym Bull. doi:10.1007/s00289-011-0529-5 7. Scharlach K, Walter K (2007) J Zhejiang Univ Sci A 8:987–990 8. Lomakin SM, Dubnikova IL, Shchegolikhin AN, Zaikov GE, Kozlowski R, Kim GM, Michler GH (2008) J Therm Anal Calorim 94:719–726 9. Adewole JK, Al-Mubaiyedh UA, Ul-Hamid A, Al-Juhani AA, Hussein IA (2011) Can J Chem Eng 9999:1–13 10. Deepthi MV, Sharma M, Sailaja R, Anantha P, Sampathkumaran P, Seetharamu S (2010) Mater Des 31:2051–2060 11. Gopakumar T, Lee J, Kontopoulou M, Parent J (2002) Polym 43:5483–5491 12. Zhai H, Xu W, Guo H, Zhou Z, Shen S, Song Q (2004) Eur Polym J 40:2539–2545 13. Zhang J, Wilkie CA (2003) Polym Degrad Stab 80:163–169 14. Zhao C, Qin H, Gong F, Feng M, Zhang S, Yang M (2005) Polym Degrad Stab 87:183–189 15. Chuang TH, Guo W, Cheng KC, Chen SW, Wang HT, Yen YY (2004) J Polym Res 11:169–174 16. Flaconnèche B, Martin J, Klopffer MH (2001) Oil Gas Sci Technol 56:261–278 17. Picard E, Vermogen A, Gerard J, Espuche E (2008) J Polym Sci B Polym Phys 46:2593–2604 18. Jacquelot E, Espuche E, Gerard JF, Duchet J, Mazabraud P (2006) J Polym Sci B Polym Phys 44:431–440 19. Naito Y, Mizoguchi K, Terada K, Kamiya Y (1991) J Polym Sci B Polym Phys 29:457–462 20. Hotta S, Paul DR (2004) Polymer 45:7639–7654 21. Raharjo RD, Freeman BD, Paul DR, Sarti GC, Sanders ES (2007) J Membr Sci 306:75–92 22. Klopffer M, Flaconnèche B (2001) Oil Gas Sci Technol Rev IFP 56:223–244 23. Lin H, Freeman B (2004) J Membr Sci 239:105–117 24. Safari M, Ghanizadeh A, Montazer-Rahmati MM (2009) Int J Greenh Gas Control 3:3–10 25. Stern SA, Krishnakumar B, Nadakatti SM (1996) In: Mark JE (ed) Physical properties of polymers handbook. AIP, Woodbury 26. Alamo RG, Mandelkern L (2009) In: Mark JE (ed) Polymer data handbook. Oxford University Press, Oxford 27. Flaconnèche B, Martin J, Klopffer M (2001) Oil Gas Sci Technol Rev IFP 56:245–259 28. Koros WJ, Madden CW (2003) Transport Properties. In: Mark HF (ed) Encyclopedia of polymer science and technology. WileyInterscience, New Jersey 29. Ghadimi A, Sadrzadeh M, Shahidi K, Mohammadi T (2009) J Membr Sci 344:225–236 30. Merkel CT, He Z, Pinnau I, Freeman BD, Meakin P, Hill JA (2003) Macromol 36:6844–6855 31. Matteucci S, Raharjo RD, Kusuma VA, Swinnea S, Freeman BD (2008) Macromol 41:2144–2156 32. Lee H, Jung D, Hong C, Rhee KY, Advani SG (2005) Compos Sci Technol 65:1996–2002 33. Jordan SM, Koros WJ (1990) J Polym Sci B Polym Phys 28:795–809 34. Zeng Q, Yu A (2010) In: Vittal M (ed) Optimization of polymer nanocomposite properties. Wiley-VCH, Weinheim 35. Hussein IA (1999) Molecular order, miscibility and rheology of molten polyethylenes. PhD Thesis, University of Alberta 36. Welty JR, Wicks CE, Wilson RE, Rorrer G (2008) Fundamentals of momentum, heat and mass transfer, 5th edn. Wiley, New Jersey 37. Mogri Z, Paul DR (2004) Polym 42:7781–7789 38. Dhingra SS, Marand E (1998) J Membr Sci 141:45–63
