- Email: [email protected]
Understanding the formation of pinholes in PFSA membranes with the essential work of fracture (EWF)
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/he
Understanding the formation of pinholes in PFSA membranes with the essential work of fracture (EWF) E. Moukheiber, C. Bas, L. Flandin* LEPMI, UMR 5279, CNRS, Grenoble INP e Universite´ de Savoie e Universite´ J. Fourier, LMOPS e INES e Baˆt. He´lios, Campus de Savoie-Technolac, F-73376 Le Bourget-du Lac Cedex, France
article info
abstract
Article history:
One of the most harmful degradation process in PEM fuel cell is the development of pin-
Received 24 January 2013
holes in the membrane. There is therefore a need for an effective experimental charac-
Accepted 6 March 2013
terization to allow ab initio membrane comparison. In this paper, the mechanical fracture
Available online 29 March 2013
resistance of various PFSA membranes was studied using the essential work of fracture (EWF) and tensile tests. PTFE reinforced membrane better resists pinholes formation due to
Keywords:
its high resistance to crack initiation and propagation. Additionally, energy partitioning
Polymer electrolyte fuel cells
showed that the necking and tearing stage of the layered structure membrane accounts for
Reinforced composite polymer elec-
the main part of the total fracture energy due to enhanced plastic deformation of PTFE.
trolyte membrane
Moreover, cracks were found to initiate and propagate easily in the direction parallel to the
Nafion membrane Short-side-chain
perfluorosulfonic
polymer chains which suggest that the fracture control could be optimized by pointing the direction of the gas channel perpendicularly to the orientation of the polymer chains, i.e. to
acid (SSC PFSA) membranes
rolling process during manufacturing. Finally, EWF technique was found to be more rele-
Long-side-chain
vant for assessing the differences in the mechanical behaviour of the membranes
perfluorosulfonic
acid (LSC PFSA) membranes
compared to standard tensile tests.
Essential work of fracture (EWF)
Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
To secure long-term durability and efficiency of proton exchange membrane (PEM) for fuel cells, perfluorosulfonic acid (PFSA) based membrane must demonstrate long-term electrochemical and mechanical integrities [1e4]. One of the failure modes that limits the lifetime of the fuel cell involves the fracture of the membranes. The mechanisms of pinhole formation and subsequent crack growth and propagation are complex and not fully understood. Nonetheless, those defects are likely to result from the combination of chemical and mechanical effects. Variations in the temperature and humidity during operation cause hydrothermal stresses in constrained membranes and MEA [5e8]. A
considerable thinning of the membrane was reported under fuel cell operation or OCV due to massive ionomer loss throughout the active area caused by radical attacks [9,10]. Mud cracks of different depths, typically present in the electrodes can cause delamination and/or cracking of the PEM membranes [11]. In a recent work, SEM photomicrographs around the detected flaws revealed linear cracks in the membrane essentially oriented in the direction of the gas path which points out the sharp edges of the gas channels [12]. Platinum catalyst dissolution and recrystallization, cationic contaminants are also believed to contribute to the embrittlement of polymer electrolyte membranes [10]. Consequently, this can lead to reactant gases crossover, localized heating and ultimately the failure of the membrane and thereby
* Corresponding author. E-mail addresses: [email protected], [email protected] (L. Flandin). 0360-3199/$ e see front matter Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijhydene.2013.03.031
2718
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
Fig. 1 e Chemical structures of PFSA membrane e (a) Short Side Chain (SSC) type membrane, (b) Long Side Chain (LSC) type membrane.
of the entire system. One of the possible ways to prevent pinhole formation and extend the lifetime of PEM fuel cells relies on the use of reinforced membranes. Mechanical reinforcement with expanded PTFE sheets combined usually with thinner membranes provided high membrane conductance, improved water distribution in the operating fuel cell, less dimensional variation and improved performance without scarifying durability [13]. In reinforced membrane, H2 crossover rate increased gradually while non reinforced membrane displayed a sudden and drastic jump [14]. While mechanical reinforcement can extend the lifetime, there is no clear understanding of the mechanisms that govern the mechanical behaviour of the reinforced material compared to non reinforced ones. Mechanical properties of PFSA membranes are usually characterized using tensile tests [9], however few works dealt with the essential work of fracture (EWF) tests, a method widely used to characterize fracture of polymers, related blends and composites [4,15e19]. Does the better mechanical durability of reinforced membrane result from better tensile properties or fracture toughness? Is it the young modulus or the tear resistance that controls the mechanical behaviour of PEM during fuel cell operation?
2.
Experimental
2.1.
Membranes
Membranes used in this study were provided by Solvay Speciality Polymers and Ion power Inc. All samples were pretreated in an aqueous 10 wt.% in HNO3 solution for at least 3 h at 80 C followed by a treatment for 1 h in UHQ water at 80 C. Three types of PFSA membranes can be distinguished: (i)
Table 1 e Property of commercially available membranes. Commercial reference
Aquivion E110
Nafion 111 Nafion XL100
Supplier Solvay Speciality Polymers Ion Power
EW (g eq1)
Type
1100 20
SSC
SSC-110
1100 30 1100 [17]
LSC LSC/ PTFE
LSC-110 XL-100
homogeneous LSC membrane, (ii) homogeneous SSC membrane, (iii) PFSA/PTFE composite membrane. The chemical structures of SSC and LSC membranes are shown in Fig. 1. XL100 composite consists of microporous PTFE impregnated on both sides with an LSC type PFSA solution with an equivalent weight of 900 g/mol. The thickness of the different layers is 9/12/9 mm [20]. The ion exchange capacity of the different layers was estimated by different techniques [20] however in this paper, only the overall equivalent weight was mentioned in Table 1.
2.2.
Double edge notched tensile test (DENT)
The concept of EWF tests was well described by Cotterell and Reddel [21]. Based on Broberg [22e24], the non-elastic crack tip can be divided into two parts: - An inner “fracture zone”, where the tearing process occurs. - An outer “plastic zone”, where the plastic deformation and dissipative process occurs. During the crack propagation, a significant amount of energy dissipated in the plastic zone is not directly associated with the fracture process. The total fracture energy can be divided into two components. One component We corresponds to the term that characterizes the process zone and the other Wp, the plastic zone as described in Eq. (1) Wf ¼ We þ Wp
(1)
where Wf is the total fracture energy, We is the essential work of fracture used in the process zone and Wp is the nonessential work of fracture dissipated in the outer plastic zone. The EWF test should preferably be applied to films, which can be assumed to be in plane stress conditions. We is then proportional to the ligament length, L, while Wp is proportional to the square of the ligament length, L2 as described in Eq. (2)
Acronyms Wf ¼ we Lt þ bwp L2 t
(2)
where we is the specific essential work of fracture (work per unit area), wp is the non-essential work of fracture (plastic work density), L is the ligament length of the specimen Fig. 2a, t is the thickness of the specimen, and b is the shape factor that is related to the formation of the plastic zone. The specific work of fracture, wf ¼ Wf =Lt is given by Eq (3):
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
2719
Fig. 2 e (a) DENT geometry showing the fracture and the plastic zones, (b) Typical load-displacement curve under EWF test and partitioning based on the point of maximum load (c) experimental determination of EWF parameters [15].
wf ¼ we þ bwp L
(3)
Following Eq. (3), we can be obtained by conducting DENT tests of different ligaments lengths and extrapolating wf vs. L line to zero L. To maintain the linearity between wf and L as expressed by Eq. (4), it is recommended that the ligament length varies between the minimal Lmin ¼ 3e5t and the maximum ligament length Lmax ¼ B/3. Other authors [25,26] suggested that the EWF obtained on polymeric films can be separated into two distinct contributions at the maximum load (Fig. 2b). In this model, the yielding
Fig. 3 e Definition of the notch direction with respect to polymer chain direction.
corresponds to blunting of the complete ligament whereas the necking and tearing where crack propagation and rupture of the yield ligament occurs. Therefore the specific total work of fracture can be rewritten as following: wf;y ¼ we;y þ bwp;y L
(4)
wf;n ¼ we;n þ bwp;n L
(5)
where the subscripts y and n represent yielding and necking, respectively. For the measurements of the EWF, (DENT) specimens used in the present study were cut from rectangular shape (a gage length of 50 mm and 25 mm wide) in the machine (MD) and transverse direction (TD). Initial notches were made perpendicularly to the traction direction with a fresh razor blade (Fig. 3). The length of the uncut ligament, ranged from 1 to 8 mm, was measured precisely using an optical microscope. The specimens were loaded on an ADAMEL Lhomargy tensile machine (100 N) at a crosshead rate of 5 mm/min. The measurements were performed in a controlled temperature (23 1 C) and humidity (50e60% RH) environment and examined between crossed polarizers at a 45 angle against the plane of polarization.
2.3.
Tensile tests
Uniaxial tensile tests were performed with an ADAMEL Lhomargy tensile machine (100 N) at ambient temperature
2720
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
a)
60
SSC-110
TD
LSC-110
Stress, σ (MPa)
50
XL100
40 30 20 10 0
Stress, σ (MPa)
b)
60
0
100
200
300
400
500 SSC-110
50
LSC-110 XL100
40 MD
30 20 10 0
0
100
200
300
400
500
Strain, ε (%)
Fig. 4 e Typical Stressestrain curves of samples under tensile tests, (a) MD tensile samples, (b) TD tensile samples at 25 C and 60% RH.
(23 1 C) and controlled humidity (50e60%) with a crosshead speed of 5 mm/min the membranes were cut into standard dog-bone samples with a gauge length of 2.2 cm and a width of 0.5 cm (ASTM D638). For each sample, three replicates were performed, for both EWF and tensile tests.
3.
Results and discussion
The mechanical properties of the neat samples were determined first by stress-strain tests, under controlled conditions of temperature and relative humidity. The strainestress curves are shown in Fig. 4. A summary of the extracted parameters is shown in Table 2. On a qualitative point of view, all membranes showed anisotropy. Reinforced membrane XL100 exhibits anisotropy that is slightly different from the other two membranes. The mechanical quantities derived from the analysis of the stress and strain curves were similar for TD
tensile specimens. For MD tensile specimens, XL100 displayed the highest yield stress and the lowest tensile strength in the rubbery state. Nevertheless, none of these properties stands out as significantly different from the other two membranes that suggest it would exhibit superior durability during fuel cell operation. In addition, all membranes displayed quite similar elastic young modulus. Therefore enhanced mechanical properties and durability of reinforced XL100 could not be explained considering the sole tensile tests. Fig. 5 presents the load-displacement curves of the DENT specimens with various ligament lengths in both notch directions. Like tensile tests, all EWF samples showed anisotropy. The loadedisplacement curves of the TD-notch specimens showed high energy consumption compared to MD-notch specimens at a given ligament. Based on the load-displacement curves, the specific total work of fracture and its components, the specific total work of yielding and that of necking and tearing can be calculated. Fig. 6 shows, as expected in Eq (3), the linear relationship between the specific work of fracture and the ligament length under EWF tests. As depicted in Eqs. (2) and (3), two key parameters were extracted, i.e. the intercept between the specific total work of fracture and the ligament length (we) and the slope of the linear fit (bwp). In Tables 3 and 4 are summarized the essential work of fracture we, the slope of the linear fit (bwp) and their components. As shown in Fig. 7, the composite XL-100 displayed the highest essential work of fracture and the highest slope bwp. Based on their attribution [15], XL100 has the highest resistance for crack initiation and crack propagation compared to non reinforced membranes. No significant differences were found between SSC and LSC membranes. Moreover, the essential work of fracture we and the slope bwp, were more pronounced for all TD notch specimens compared to MDnotch specimens. It appears in this case that the polymer chains prevent the initiation and propagation of the flaws while the cracks were found to initiate and propagate more easily in the direction of polymer chain and require less energy. These results are consistent with the essential work of fracture of thermoplastic polyester elastomers TPEEs of two directions specimens based on the resin flow during the injection moulding process [18]. Furthermore, the contribution of the yielding on the essential work of fracture for all samples is revealed to be larger than that of necking and tearing with both MD and TD notch specimens. Therefore the resistance to crack initiation is more likely controlled by the loaddisplacement behaviour before the yield point. Concerning the slope bwp, the contribution of the yielding was quite
Table 2 e Mechanical parameters extracted from tensile test curves. Properties
SSC-110 MD
Tensile stress (MPa) at 3 ¼ 50% Tensile strength (MPa) Elongation to break (%) Young Modulus (MPa)
16 49 290 127
LSC-110 TD
1 2 6 7
12 39 419 111
MD 0.3 1 13 20
19.7 46 200 110
0.7 1 6 20
XL-100 TD 11.6 30 405 105
1.0 4 12 15
MD 30.1 37 309 126
0.6 1 9 14
TD 14.3 34 407 127
0.6 1 13 7
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
2721
Fig. 5 e Loadedisplacement curves of the DENT specimens with various ligament lengths in both notch directions.
Fig. 6 e Relationship between the specific total work of fracture and the ligament length under EWF tests.
identical to that of necking and tearing for non reinforced membranes; however a larger contribution of the necking and tearing was observed for the reinforced membrane XL-100 which indicates that the resistance to crack propagation of this particular membrane is controlled by the loaddisplacement behaviour after the yield point. A deeper investigation of the fracture process was performed between crossed polarizers. Fig. 8 compares the load
displacement curves of LSC110 in both notch direction and the propagation of the fracture as function of the displacement. It appears that the initiation of rupture of both membranes occurs before the yield point (85% of sy). Therefore, below this point, the material presents an elastic behaviour. Moreover two steps of fracture propagation can be distinguished: A slow propagation of the flaw mainly in the necking region followed by a rapid tearing. Nevertheless,
2722
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
Table 3 e Calculated EWF parameters. Sample ID
SSC-110 LSC-110 XL-100
we(kJ/m2)
bwp(MJ/m3)
MD
TD
MD
TD
6.7 0.8 5.6 0.4 19.4 1.1
11.6 1.2 10.9 0.6 23.3 3.6
2.0 0.2 0.6 0.1 3.6 0.2
1.6 0.2 1.7 0.1 10.3 0.8
Table 4 e Calculated EWF parameters for yielding and necking processes. Sample ID
SSC-110 LSC-110 XL-100
Sample ID
SSC-110 LSC-110 XL-100
we,y(kJ/m2)
bwp,y(MJ/m3)
MD
TD
MD
TD
5.3 0.4 4.4 0.4 11.3 0.9
8.9 0.9 7.8 0.7 15.7 0.9
1.0 0.4 0.4 0.1 1.1 0.2
1.0 0.1 0.2 0.1 3.6 0.5
we,n(kJ/m2)
bwp,n(MJ/m3)
MD
TD
MD
TD
1.4 0.6 1.1 0.1 8.1 1.4
2.6 0.9 3.1 0.7 7.5 1.9
0.6 0.1 0.10 0.01 2.5 0.3
0.9 0.1 0.5 0.2 6.7 0.4
TD-notch specimens displayed much slower propagation of the fracture in the necking and tearing region than MDnotch ones which could be attributed to the polymer chain resistance.
Fig. 8 e Load displacement curves and propagation of fracture of LSC110 in both notch directions.
Based on Fig. 9, the initiation of rupture of XL100 membrane occurs before the yield point (85% of sy) and exhibited two steps propagation similar to non reinforced LSC membrane. Nevertheless, fracture propagation was much more hindered than LSC110 and could be assigned to the layered structure effect. This type of reinforcement seems to give high mechanical resistance for the membrane compared to polymer chain orientation. Indeed PTFE is considered highly resistance to fracture propagation due to its extensive blunting and plastic deformation at room temperature [27]. Finally, the combination of these two control parameters could
Fig. 7 e EWF parameters by partitioning EWF test results (a) we of MD-notch specimens, (b) we of TD-notch specimens, (c) slope (bwp) of MD-notch specimens, (d) slope (bwp) of TD-notch specimens.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
2723
in comparison to channels should be an important parameter to take into account during the manufacturing and the processing of AMEs. The best configuration would be to point the direction of the gas channels perpendicularly to the orientation of the chain polymers in the membrane in order to optimize the tear resistance. A schematic representation of the tear resistance optimization during AME process is presented in Fig. 10.
references
Fig. 9 e Load displacement curves and propagation of fracture comparisons between XL100 and LSC110 in TDnotch direction. The curves are illustrated by images taken between the crossed polarizers.
g Hi
h
t
g ou
es hn
s
Fig. 10 e Schematic representation of the tear resistance optimization during AME process.
optimize and improve the durability of membrane during fuel cell processing.
4.
Conclusion
In summary, the EWF tests were found more discriminative compared to standard tensile tests in assessing the mechanical behaviour of PFSA membrane and is more representative of membrane fracture in fuel cell. The better mechanical behaviour of composite XL-100 compared to non-reinforced membranes could be explained by a better resistance to crack initiation characterized by a higher essential work of fraction we and a better resistance to crack propagation characterized by a higher slope bwp. In addition, energy partitioning of XL100 showed that the resistance to crack propagation stage accounts for the main part of the total fracture energy due to PTFE blunting. For all samples, the TD-notch specimens displayed the highest resistance to fracture which means that the rupture of the membrane is difficult when pointing across the polymer chains. Since water and gas channels are plausible sources to the appearance of defects or fractures in the membranes due to their sharp edges under the compression of the stack, the orientation of polymer chain
[1] Mauritz KA, Moore RB. Chemical Reviews 2004;104:4535. [2] Yandrasits MA, Hamrock SJ. Fuel cell chemistry and operation, vol. 1040. American Chemical Society; 201015. [3] Di Noto V, Zawodzinski TA, Herring AM, Giffin GA, Negro E, Lavina S. International Journal of Hydrogen Energy 2012;37:6120. [4] Ghielmi A, Vaccarono P, Troglia C, Arcella V. Journal of Power Sources 2005;145:108. [5] Tang Y, Karlsson AM, Santare MH, Gilbert M, Cleghorn S, Johnson WB. Materials Science and Engineering: A 2006;425:297. [6] Kusoglu A, Karlsson AM, Santare MH, Cleghorn S, Johnson WB. Journal of Power Sources 2006;161:987. [7] Silberstein MN, Boyce MC. Journal of Power Sources 2010;196:3452. [8] Lai Y-H, Mittelsteadt CK, Gittleman CS, Dillard DA. Journal of Fuel Cell Science and Technology 2009;6. [9] Bas C, Flandin L, Danerol AS, Claude E, Rossinot E, Alberola ND. Journal of Applied Polymer Science 2012;117:2121. [10] Zhang S, Yuan X, Wang H, Merida W, Zhu H, Shen J, et al. International Journal of Hydrogen Energy 2009;34:388. [11] Gittleman CS, Coms FD, Lai Y-H. In: Polymer electrolyte fuel cell degradation. Boston: Academic Press; 2012. p. 15. [12] De Moor G, Bas C, Charvin N, Moukheiber E, Niepceron F, Breilly N, et al. Fuel Cells 2012;12:356. [13] Kolde JA, Bamdad B, Wilson MS, Zawodzinski TA, Gottesfeld S. Proceedings of the first international symposium on proton conducting membrane fuel cells I. Electrochemical Society Proceedings 1995;95-23:193e201. [14] Liu W, Ruth K, Rusch G. Journal of New Materials for Materials for Electrochemical Systems 2001;4:227. [15] Barany T, Czigany T, Karger-Kocsis J. Progress in Polymer Science 2010;35:1257. [16] Garnier G, Chehab B, Yrieix B, Brechet Y, Flandin L. Journal of Materials Science 2009;44:5537. [17] Hashemi S. Polymer Testing 2003;22:589. [18] Lee J-M, Choi B-H, Moon J-S, Lee E-S. Polymer Testing 2009;28:854. [19] Li Y, Quincy JK, Case SW, Ellis MW, Dillard DA, Lai Y-H, et al. Journal of Power Sources 2008;185:374. [20] Moukheiber E, De Moor G, Flandin L, Bas C. Journal of Membrane Science 2012;389:294. [21] Cotterell B, Reddel JK. International Journal of Fracture 1977;13:267. [22] Broberg KB. Journal of the Mechanics and Physics of Solids 1975;23:215. [23] Broberg KB. Journal of the Mechanics and Physics of Solids 1971;19:407. [24] Broberg KB. International Journal of Fracture 1968;4:11. [25] Ferrer-Balas D, Maspoch ML, Martinez AB, Santana OO. Polymer Bulletin 1999;42:101. [26] Maspoch ML, Ga´mez-Pe´rez J, Karger-Kocsis J. Polymer Bulletin 2003;50:279. [27] Brown EN, Dattelbaum DM. Polymer 2005;46:3056.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/he
Understanding the formation of pinholes in PFSA membranes with the essential work of fracture (EWF) E. Moukheiber, C. Bas, L. Flandin* LEPMI, UMR 5279, CNRS, Grenoble INP e Universite´ de Savoie e Universite´ J. Fourier, LMOPS e INES e Baˆt. He´lios, Campus de Savoie-Technolac, F-73376 Le Bourget-du Lac Cedex, France
article info
abstract
Article history:
One of the most harmful degradation process in PEM fuel cell is the development of pin-
Received 24 January 2013
holes in the membrane. There is therefore a need for an effective experimental charac-
Accepted 6 March 2013
terization to allow ab initio membrane comparison. In this paper, the mechanical fracture
Available online 29 March 2013
resistance of various PFSA membranes was studied using the essential work of fracture (EWF) and tensile tests. PTFE reinforced membrane better resists pinholes formation due to
Keywords:
its high resistance to crack initiation and propagation. Additionally, energy partitioning
Polymer electrolyte fuel cells
showed that the necking and tearing stage of the layered structure membrane accounts for
Reinforced composite polymer elec-
the main part of the total fracture energy due to enhanced plastic deformation of PTFE.
trolyte membrane
Moreover, cracks were found to initiate and propagate easily in the direction parallel to the
Nafion membrane Short-side-chain
perfluorosulfonic
polymer chains which suggest that the fracture control could be optimized by pointing the direction of the gas channel perpendicularly to the orientation of the polymer chains, i.e. to
acid (SSC PFSA) membranes
rolling process during manufacturing. Finally, EWF technique was found to be more rele-
Long-side-chain
vant for assessing the differences in the mechanical behaviour of the membranes
perfluorosulfonic
acid (LSC PFSA) membranes
compared to standard tensile tests.
Essential work of fracture (EWF)
Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
To secure long-term durability and efficiency of proton exchange membrane (PEM) for fuel cells, perfluorosulfonic acid (PFSA) based membrane must demonstrate long-term electrochemical and mechanical integrities [1e4]. One of the failure modes that limits the lifetime of the fuel cell involves the fracture of the membranes. The mechanisms of pinhole formation and subsequent crack growth and propagation are complex and not fully understood. Nonetheless, those defects are likely to result from the combination of chemical and mechanical effects. Variations in the temperature and humidity during operation cause hydrothermal stresses in constrained membranes and MEA [5e8]. A
considerable thinning of the membrane was reported under fuel cell operation or OCV due to massive ionomer loss throughout the active area caused by radical attacks [9,10]. Mud cracks of different depths, typically present in the electrodes can cause delamination and/or cracking of the PEM membranes [11]. In a recent work, SEM photomicrographs around the detected flaws revealed linear cracks in the membrane essentially oriented in the direction of the gas path which points out the sharp edges of the gas channels [12]. Platinum catalyst dissolution and recrystallization, cationic contaminants are also believed to contribute to the embrittlement of polymer electrolyte membranes [10]. Consequently, this can lead to reactant gases crossover, localized heating and ultimately the failure of the membrane and thereby
* Corresponding author. E-mail addresses: [email protected], [email protected] (L. Flandin). 0360-3199/$ e see front matter Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijhydene.2013.03.031
2718
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
Fig. 1 e Chemical structures of PFSA membrane e (a) Short Side Chain (SSC) type membrane, (b) Long Side Chain (LSC) type membrane.
of the entire system. One of the possible ways to prevent pinhole formation and extend the lifetime of PEM fuel cells relies on the use of reinforced membranes. Mechanical reinforcement with expanded PTFE sheets combined usually with thinner membranes provided high membrane conductance, improved water distribution in the operating fuel cell, less dimensional variation and improved performance without scarifying durability [13]. In reinforced membrane, H2 crossover rate increased gradually while non reinforced membrane displayed a sudden and drastic jump [14]. While mechanical reinforcement can extend the lifetime, there is no clear understanding of the mechanisms that govern the mechanical behaviour of the reinforced material compared to non reinforced ones. Mechanical properties of PFSA membranes are usually characterized using tensile tests [9], however few works dealt with the essential work of fracture (EWF) tests, a method widely used to characterize fracture of polymers, related blends and composites [4,15e19]. Does the better mechanical durability of reinforced membrane result from better tensile properties or fracture toughness? Is it the young modulus or the tear resistance that controls the mechanical behaviour of PEM during fuel cell operation?
2.
Experimental
2.1.
Membranes
Membranes used in this study were provided by Solvay Speciality Polymers and Ion power Inc. All samples were pretreated in an aqueous 10 wt.% in HNO3 solution for at least 3 h at 80 C followed by a treatment for 1 h in UHQ water at 80 C. Three types of PFSA membranes can be distinguished: (i)
Table 1 e Property of commercially available membranes. Commercial reference
Aquivion E110
Nafion 111 Nafion XL100
Supplier Solvay Speciality Polymers Ion Power
EW (g eq1)
Type
1100 20
SSC
SSC-110
1100 30 1100 [17]
LSC LSC/ PTFE
LSC-110 XL-100
homogeneous LSC membrane, (ii) homogeneous SSC membrane, (iii) PFSA/PTFE composite membrane. The chemical structures of SSC and LSC membranes are shown in Fig. 1. XL100 composite consists of microporous PTFE impregnated on both sides with an LSC type PFSA solution with an equivalent weight of 900 g/mol. The thickness of the different layers is 9/12/9 mm [20]. The ion exchange capacity of the different layers was estimated by different techniques [20] however in this paper, only the overall equivalent weight was mentioned in Table 1.
2.2.
Double edge notched tensile test (DENT)
The concept of EWF tests was well described by Cotterell and Reddel [21]. Based on Broberg [22e24], the non-elastic crack tip can be divided into two parts: - An inner “fracture zone”, where the tearing process occurs. - An outer “plastic zone”, where the plastic deformation and dissipative process occurs. During the crack propagation, a significant amount of energy dissipated in the plastic zone is not directly associated with the fracture process. The total fracture energy can be divided into two components. One component We corresponds to the term that characterizes the process zone and the other Wp, the plastic zone as described in Eq. (1) Wf ¼ We þ Wp
(1)
where Wf is the total fracture energy, We is the essential work of fracture used in the process zone and Wp is the nonessential work of fracture dissipated in the outer plastic zone. The EWF test should preferably be applied to films, which can be assumed to be in plane stress conditions. We is then proportional to the ligament length, L, while Wp is proportional to the square of the ligament length, L2 as described in Eq. (2)
Acronyms Wf ¼ we Lt þ bwp L2 t
(2)
where we is the specific essential work of fracture (work per unit area), wp is the non-essential work of fracture (plastic work density), L is the ligament length of the specimen Fig. 2a, t is the thickness of the specimen, and b is the shape factor that is related to the formation of the plastic zone. The specific work of fracture, wf ¼ Wf =Lt is given by Eq (3):
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
2719
Fig. 2 e (a) DENT geometry showing the fracture and the plastic zones, (b) Typical load-displacement curve under EWF test and partitioning based on the point of maximum load (c) experimental determination of EWF parameters [15].
wf ¼ we þ bwp L
(3)
Following Eq. (3), we can be obtained by conducting DENT tests of different ligaments lengths and extrapolating wf vs. L line to zero L. To maintain the linearity between wf and L as expressed by Eq. (4), it is recommended that the ligament length varies between the minimal Lmin ¼ 3e5t and the maximum ligament length Lmax ¼ B/3. Other authors [25,26] suggested that the EWF obtained on polymeric films can be separated into two distinct contributions at the maximum load (Fig. 2b). In this model, the yielding
Fig. 3 e Definition of the notch direction with respect to polymer chain direction.
corresponds to blunting of the complete ligament whereas the necking and tearing where crack propagation and rupture of the yield ligament occurs. Therefore the specific total work of fracture can be rewritten as following: wf;y ¼ we;y þ bwp;y L
(4)
wf;n ¼ we;n þ bwp;n L
(5)
where the subscripts y and n represent yielding and necking, respectively. For the measurements of the EWF, (DENT) specimens used in the present study were cut from rectangular shape (a gage length of 50 mm and 25 mm wide) in the machine (MD) and transverse direction (TD). Initial notches were made perpendicularly to the traction direction with a fresh razor blade (Fig. 3). The length of the uncut ligament, ranged from 1 to 8 mm, was measured precisely using an optical microscope. The specimens were loaded on an ADAMEL Lhomargy tensile machine (100 N) at a crosshead rate of 5 mm/min. The measurements were performed in a controlled temperature (23 1 C) and humidity (50e60% RH) environment and examined between crossed polarizers at a 45 angle against the plane of polarization.
2.3.
Tensile tests
Uniaxial tensile tests were performed with an ADAMEL Lhomargy tensile machine (100 N) at ambient temperature
2720
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
a)
60
SSC-110
TD
LSC-110
Stress, σ (MPa)
50
XL100
40 30 20 10 0
Stress, σ (MPa)
b)
60
0
100
200
300
400
500 SSC-110
50
LSC-110 XL100
40 MD
30 20 10 0
0
100
200
300
400
500
Strain, ε (%)
Fig. 4 e Typical Stressestrain curves of samples under tensile tests, (a) MD tensile samples, (b) TD tensile samples at 25 C and 60% RH.
(23 1 C) and controlled humidity (50e60%) with a crosshead speed of 5 mm/min the membranes were cut into standard dog-bone samples with a gauge length of 2.2 cm and a width of 0.5 cm (ASTM D638). For each sample, three replicates were performed, for both EWF and tensile tests.
3.
Results and discussion
The mechanical properties of the neat samples were determined first by stress-strain tests, under controlled conditions of temperature and relative humidity. The strainestress curves are shown in Fig. 4. A summary of the extracted parameters is shown in Table 2. On a qualitative point of view, all membranes showed anisotropy. Reinforced membrane XL100 exhibits anisotropy that is slightly different from the other two membranes. The mechanical quantities derived from the analysis of the stress and strain curves were similar for TD
tensile specimens. For MD tensile specimens, XL100 displayed the highest yield stress and the lowest tensile strength in the rubbery state. Nevertheless, none of these properties stands out as significantly different from the other two membranes that suggest it would exhibit superior durability during fuel cell operation. In addition, all membranes displayed quite similar elastic young modulus. Therefore enhanced mechanical properties and durability of reinforced XL100 could not be explained considering the sole tensile tests. Fig. 5 presents the load-displacement curves of the DENT specimens with various ligament lengths in both notch directions. Like tensile tests, all EWF samples showed anisotropy. The loadedisplacement curves of the TD-notch specimens showed high energy consumption compared to MD-notch specimens at a given ligament. Based on the load-displacement curves, the specific total work of fracture and its components, the specific total work of yielding and that of necking and tearing can be calculated. Fig. 6 shows, as expected in Eq (3), the linear relationship between the specific work of fracture and the ligament length under EWF tests. As depicted in Eqs. (2) and (3), two key parameters were extracted, i.e. the intercept between the specific total work of fracture and the ligament length (we) and the slope of the linear fit (bwp). In Tables 3 and 4 are summarized the essential work of fracture we, the slope of the linear fit (bwp) and their components. As shown in Fig. 7, the composite XL-100 displayed the highest essential work of fracture and the highest slope bwp. Based on their attribution [15], XL100 has the highest resistance for crack initiation and crack propagation compared to non reinforced membranes. No significant differences were found between SSC and LSC membranes. Moreover, the essential work of fracture we and the slope bwp, were more pronounced for all TD notch specimens compared to MDnotch specimens. It appears in this case that the polymer chains prevent the initiation and propagation of the flaws while the cracks were found to initiate and propagate more easily in the direction of polymer chain and require less energy. These results are consistent with the essential work of fracture of thermoplastic polyester elastomers TPEEs of two directions specimens based on the resin flow during the injection moulding process [18]. Furthermore, the contribution of the yielding on the essential work of fracture for all samples is revealed to be larger than that of necking and tearing with both MD and TD notch specimens. Therefore the resistance to crack initiation is more likely controlled by the loaddisplacement behaviour before the yield point. Concerning the slope bwp, the contribution of the yielding was quite
Table 2 e Mechanical parameters extracted from tensile test curves. Properties
SSC-110 MD
Tensile stress (MPa) at 3 ¼ 50% Tensile strength (MPa) Elongation to break (%) Young Modulus (MPa)
16 49 290 127
LSC-110 TD
1 2 6 7
12 39 419 111
MD 0.3 1 13 20
19.7 46 200 110
0.7 1 6 20
XL-100 TD 11.6 30 405 105
1.0 4 12 15
MD 30.1 37 309 126
0.6 1 9 14
TD 14.3 34 407 127
0.6 1 13 7
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
2721
Fig. 5 e Loadedisplacement curves of the DENT specimens with various ligament lengths in both notch directions.
Fig. 6 e Relationship between the specific total work of fracture and the ligament length under EWF tests.
identical to that of necking and tearing for non reinforced membranes; however a larger contribution of the necking and tearing was observed for the reinforced membrane XL-100 which indicates that the resistance to crack propagation of this particular membrane is controlled by the loaddisplacement behaviour after the yield point. A deeper investigation of the fracture process was performed between crossed polarizers. Fig. 8 compares the load
displacement curves of LSC110 in both notch direction and the propagation of the fracture as function of the displacement. It appears that the initiation of rupture of both membranes occurs before the yield point (85% of sy). Therefore, below this point, the material presents an elastic behaviour. Moreover two steps of fracture propagation can be distinguished: A slow propagation of the flaw mainly in the necking region followed by a rapid tearing. Nevertheless,
2722
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
Table 3 e Calculated EWF parameters. Sample ID
SSC-110 LSC-110 XL-100
we(kJ/m2)
bwp(MJ/m3)
MD
TD
MD
TD
6.7 0.8 5.6 0.4 19.4 1.1
11.6 1.2 10.9 0.6 23.3 3.6
2.0 0.2 0.6 0.1 3.6 0.2
1.6 0.2 1.7 0.1 10.3 0.8
Table 4 e Calculated EWF parameters for yielding and necking processes. Sample ID
SSC-110 LSC-110 XL-100
Sample ID
SSC-110 LSC-110 XL-100
we,y(kJ/m2)
bwp,y(MJ/m3)
MD
TD
MD
TD
5.3 0.4 4.4 0.4 11.3 0.9
8.9 0.9 7.8 0.7 15.7 0.9
1.0 0.4 0.4 0.1 1.1 0.2
1.0 0.1 0.2 0.1 3.6 0.5
we,n(kJ/m2)
bwp,n(MJ/m3)
MD
TD
MD
TD
1.4 0.6 1.1 0.1 8.1 1.4
2.6 0.9 3.1 0.7 7.5 1.9
0.6 0.1 0.10 0.01 2.5 0.3
0.9 0.1 0.5 0.2 6.7 0.4
TD-notch specimens displayed much slower propagation of the fracture in the necking and tearing region than MDnotch ones which could be attributed to the polymer chain resistance.
Fig. 8 e Load displacement curves and propagation of fracture of LSC110 in both notch directions.
Based on Fig. 9, the initiation of rupture of XL100 membrane occurs before the yield point (85% of sy) and exhibited two steps propagation similar to non reinforced LSC membrane. Nevertheless, fracture propagation was much more hindered than LSC110 and could be assigned to the layered structure effect. This type of reinforcement seems to give high mechanical resistance for the membrane compared to polymer chain orientation. Indeed PTFE is considered highly resistance to fracture propagation due to its extensive blunting and plastic deformation at room temperature [27]. Finally, the combination of these two control parameters could
Fig. 7 e EWF parameters by partitioning EWF test results (a) we of MD-notch specimens, (b) we of TD-notch specimens, (c) slope (bwp) of MD-notch specimens, (d) slope (bwp) of TD-notch specimens.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 2 7 1 7 e2 7 2 3
2723
in comparison to channels should be an important parameter to take into account during the manufacturing and the processing of AMEs. The best configuration would be to point the direction of the gas channels perpendicularly to the orientation of the chain polymers in the membrane in order to optimize the tear resistance. A schematic representation of the tear resistance optimization during AME process is presented in Fig. 10.
references
Fig. 9 e Load displacement curves and propagation of fracture comparisons between XL100 and LSC110 in TDnotch direction. The curves are illustrated by images taken between the crossed polarizers.
g Hi
h
t
g ou
es hn
s
Fig. 10 e Schematic representation of the tear resistance optimization during AME process.
optimize and improve the durability of membrane during fuel cell processing.
4.
Conclusion
In summary, the EWF tests were found more discriminative compared to standard tensile tests in assessing the mechanical behaviour of PFSA membrane and is more representative of membrane fracture in fuel cell. The better mechanical behaviour of composite XL-100 compared to non-reinforced membranes could be explained by a better resistance to crack initiation characterized by a higher essential work of fraction we and a better resistance to crack propagation characterized by a higher slope bwp. In addition, energy partitioning of XL100 showed that the resistance to crack propagation stage accounts for the main part of the total fracture energy due to PTFE blunting. For all samples, the TD-notch specimens displayed the highest resistance to fracture which means that the rupture of the membrane is difficult when pointing across the polymer chains. Since water and gas channels are plausible sources to the appearance of defects or fractures in the membranes due to their sharp edges under the compression of the stack, the orientation of polymer chain
[1] Mauritz KA, Moore RB. Chemical Reviews 2004;104:4535. [2] Yandrasits MA, Hamrock SJ. Fuel cell chemistry and operation, vol. 1040. American Chemical Society; 201015. [3] Di Noto V, Zawodzinski TA, Herring AM, Giffin GA, Negro E, Lavina S. International Journal of Hydrogen Energy 2012;37:6120. [4] Ghielmi A, Vaccarono P, Troglia C, Arcella V. Journal of Power Sources 2005;145:108. [5] Tang Y, Karlsson AM, Santare MH, Gilbert M, Cleghorn S, Johnson WB. Materials Science and Engineering: A 2006;425:297. [6] Kusoglu A, Karlsson AM, Santare MH, Cleghorn S, Johnson WB. Journal of Power Sources 2006;161:987. [7] Silberstein MN, Boyce MC. Journal of Power Sources 2010;196:3452. [8] Lai Y-H, Mittelsteadt CK, Gittleman CS, Dillard DA. Journal of Fuel Cell Science and Technology 2009;6. [9] Bas C, Flandin L, Danerol AS, Claude E, Rossinot E, Alberola ND. Journal of Applied Polymer Science 2012;117:2121. [10] Zhang S, Yuan X, Wang H, Merida W, Zhu H, Shen J, et al. International Journal of Hydrogen Energy 2009;34:388. [11] Gittleman CS, Coms FD, Lai Y-H. In: Polymer electrolyte fuel cell degradation. Boston: Academic Press; 2012. p. 15. [12] De Moor G, Bas C, Charvin N, Moukheiber E, Niepceron F, Breilly N, et al. Fuel Cells 2012;12:356. [13] Kolde JA, Bamdad B, Wilson MS, Zawodzinski TA, Gottesfeld S. Proceedings of the first international symposium on proton conducting membrane fuel cells I. Electrochemical Society Proceedings 1995;95-23:193e201. [14] Liu W, Ruth K, Rusch G. Journal of New Materials for Materials for Electrochemical Systems 2001;4:227. [15] Barany T, Czigany T, Karger-Kocsis J. Progress in Polymer Science 2010;35:1257. [16] Garnier G, Chehab B, Yrieix B, Brechet Y, Flandin L. Journal of Materials Science 2009;44:5537. [17] Hashemi S. Polymer Testing 2003;22:589. [18] Lee J-M, Choi B-H, Moon J-S, Lee E-S. Polymer Testing 2009;28:854. [19] Li Y, Quincy JK, Case SW, Ellis MW, Dillard DA, Lai Y-H, et al. Journal of Power Sources 2008;185:374. [20] Moukheiber E, De Moor G, Flandin L, Bas C. Journal of Membrane Science 2012;389:294. [21] Cotterell B, Reddel JK. International Journal of Fracture 1977;13:267. [22] Broberg KB. Journal of the Mechanics and Physics of Solids 1975;23:215. [23] Broberg KB. Journal of the Mechanics and Physics of Solids 1971;19:407. [24] Broberg KB. International Journal of Fracture 1968;4:11. [25] Ferrer-Balas D, Maspoch ML, Martinez AB, Santana OO. Polymer Bulletin 1999;42:101. [26] Maspoch ML, Ga´mez-Pe´rez J, Karger-Kocsis J. Polymer Bulletin 2003;50:279. [27] Brown EN, Dattelbaum DM. Polymer 2005;46:3056.