A polymer electrolyte membrane fuel cell consists of bipolar plates, gas flow channels, elastomer seals and membrane electrode assembly which is made up of a polymer electrolyte membrane sandwiched between two catalysts and two electrode backing layers (anode and cathode). A fuel cell stack is formed by combining multiple PEMFC as shown in Figure 1 and may include cooling plates to maintain uniform stack temperature for optimum performance by removing heat from the fuel cell stack. [1-3]
In a PEMFC, the hydrogen gas enters the anode electrode backing layer of the fuel cell and splits into two H+ ions and two electrons upon contact with the catalyst. The electrons are used to provide electricity while the two H+ ions diffuse through the membrane and combine with an oxygen atom and the two electrons on the cathode electrode backing layer to form a water molecule. [2]
The purpose of the elastomer seals is to isolate the reactant gases (hydrogen and oxygen) in their respective compartments. Hence, any failure or degradation of the elastomer seals would result in leakage of the reactant gases which would greatly reduce the performance of the fuel cell and could finally end up with a catastrophic failure or even explosion. [4, 5]
1.1 Environment within a fuel cell
Elastomer seals in fuel cells are constantly exposed to different types of thermal, chemical and mechanical environments. Within the thermal environment, elastomer seals are often required to operate in quick temperature cycles between -40oC to 100oC as well as to withstand a maximum long term service temperature of 100oC. [5]
For the chemical aspect, elastomer seals are continuously exposed to an acidic environment from the acidic electrolyte membrane as well as the various media such as air, water, hydrogen and oxygen gas, methanol, formic acid and hydrogen fluoride. Finally, elastomer seals are also subjected to constant mechanical compressive load. [5]
1.2 Requirements of fuel cell seals
In order to ensure long term fuel cell performance, elastomer seals are required to withstand the different types of environments present in the fuel cell. Firstly, elastomer seals should be able to function over a wide temperature range while retaining their properties. At the same time, they should compensate for the different dimensional variations of the components within the fuel cell when the temperature changes. [5]
Next, elastomer seals should possess sufficient chemical resistance against the chemically active fuel cell environment. In addition, it should be highly flexible to allow compression of the fuel cell stack when compressive load is applied.
Furthermore, elastomer seals should not contain potential catalyst poisons which would contaminate and deactivate the catalyst layer or any components which might leech out of the seal and deposit on other areas of the fuel cell that would result in cell contamination. Finally, other desirable attributes of elastomer seals include ease of processing and fabrication, electrical insulation and superior relaxation behaviour which allow the elastomer seals to retain their original shape. [3, 5]
1.3 Types of elastomeric sealing material - Silicone rubber
Elastomeric sealing materials are selected over other polymer classes such as thermoplastics, duromers or thermoplastic elastomers mainly because of their superior relaxation behaviour coupled with the availability of low hardness.
Generally, the types of elastomeric sealing materials used in fuel cells include silicone rubber, ethylene propylene diene monomer (EDPM), butyl rubber or fluoroelastomer. Out of these elastomers, silicone rubber is the most frequently used sealing material probably due to its ease of processing and widespread availability. [5, 6]
Figure 2: Chemical structure of silicone rubber
The properties of silicone rubber depend on its chemical structure which is shown in Figure 2. Firstly, high chain flexibility of silicone rubber results in an extremely low glass transition temperature while the high dissociation bond energy contributes to the thermal stability of silicone rubber. [7]
Secondly, silicone rubber has excellent resistance to attacks by oxygen and ozone because of the saturated bonds present. In addition, small interaction forces between the chains result in low viscosity hence ease of processing and fabrication is achieved. [7]
However due to the siloxane (Si-O) bond, silicone rubber is polar. [8] Therefore it is prone to hydrolysis and attacks by acids and bases. Moreover, silicone rubber has high gas and liquid permeability because of its high molar volume. [7]
Figure 3: Crosslinking reaction for silicone rubber [9]
Figure 3 shows the crosslinking reaction for silicone rubber using polydimethylsiloxane with vinyl functionalities and polydimethylsiloxane with hydrosilylation functionalities. The above formulations are combined in equal stoichiometric ratio and heated to crosslink under the platinum catalysed reaction with vinyl functionalities as the crosslinking mechanism to form silicone rubber. [9]
2. Degradation of silicone elastomer seals in fuel cell environments
Silicone elastomer seals are widely used as sealing materials due to its ease of processing and widespread availability. In open literature, there are many reports published based on the degradation of silicone elastomer seals in fuel cell environments. For instance, Schulze et al. [4] investigated the degradation of seals in a polymer electrolyte membrane fuel cell (PEMFC) during fuel cell operation whereas Tan et al. [9-12] studied the degradation of silicone elastomer seals in a simulated fuel cell environment.
From the research conducted by Tan et al. [9-12], silicone elastomer seals were exposed to an accelerated durability test (ADT) solution comprising 48% hydrogen fluoride and 98% sulphuric acid (H2SO4) dissolved in balance reagent grade water at 60oC and 80oC with pH value of 3.35. The final composition of the ADT solution is 12ppm sulphuric acid and 1.8ppm hydrogen fluoride with 18MΩ resistance of reagent grade water. Thereafter, the silicone elastomer seals were analysed at selected time intervals using different techniques such as ATR-FTIR, XPS, atomic absorption spectrometry, microidentation, DMA, optical microscopy and weight change. The specific techniques and their findings shall be discussed in the following sections.
2.1 Methods of analysing degradation in silicone elastomer seals
2.1.1 Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR)
ATR-FTIR analysis was employed by Tan et al. [9-12] to study the surface chemistry or chemical degradation of the silicone elastomer seals before and after exposure to the simulated fuel cell environment. The spectra for unexposed sample (a), after exposure for 3 weeks (b), 5 weeks (c), 7 weeks (d), 10 weeks (e) and 12 weeks (f) at 60oC and 80oC are shown in Figures 4 and 5 respectively.
Figure 4: ATR-FTIR test results at 60oC [9] Figure 5: ATR-FTIR test results at 80oC [9]
From the above spectra, six main peaks were identified. The peak at 866 cm-1 resulted from the rocking vibration of Si-CH3 whereas peaks occurring between 1015 and 1080 cm-1 were due to the stretching vibrations of Si-O-Si present in the silicone elastomer backbone. Additionally, a peak at 1260 cm-1 was from the bending vibration of Si-CH3 while a peak near 1418 cm-1 was attributed to the rocking vibration of -CH2-. Lastly, a peak at 2960 cm-1 was from the stretching vibration mode of CH3.
With reference to Figure 4, it was noticed that the intensity of above peaks decreased sharply after 5 weeks of exposure and almost disappeared after 12 weeks of exposure at 60oC. Similar but more severe trends could be observed for exposure at 80oC in Figure 5. On the other hand, prolonged exposure resulted in a new peak emerging at 1040 cm-1 which could be due to the stretching vibration of Si-O when the chemical structure changed.
From the above trends, it was concluded that the higher exposure temperature resulted in more severe degradation which could be due to de-crosslinking via hydrolysis of crosslink sites and chain scissioning in the polymer backbone. [9-12]
2.1.2 X-ray Photoelectron Spectroscopy (XPS)
X-ray photoelectron spectroscopy (XPS) is a surface sensitive technique used by Tan et al. [10, 11] for obtaining quantitative information on the chemical surface of the silicone elastomer seals. The atomic concentration of carbon (C), oxygen (O), silicon (Si) and fluorine (F) detected on the silicone elastomer seals together with the concentration ratios for C/Si and O/Si at various times of exposure to accelerated durability test (ADT) solution at 80oC are listed in Table 1.
Table 1: XPS analysis of silicone elastomer seals at various exposure time
to ADT solution at 80oC [10]
From Table 1, it was observed that an increase in the exposure time led to a decrease in the concentration of carbon and a reduction in the C/Si ratio. Simultaneously, longer exposure time also resulted in an increase in the concentration of oxygen and the O/Si ratio. Hence, the above results indicated that more severe degradation occurred as a result of longer exposure time.
The decrease in carbon concentration and the simultaneous increase in oxygen concentration occurred because the methyl group on the silicon atom was attacked and oxidised to form the Si-O bonds while the chain in the silicone backbone (Si-O-Si) was broken due to hydrolysis. [10, 11]
2.1.3 Atomic absorption spectrometer
Atomic absorption spectrometry was performed by Tan et al. [9-12] to analyse the amount of foreign materials leeched out from the silicone elastomer seals into the ADT solution at 60oC and 80oC at various times. Silicon (Si), calcium (Ca) and magnesium (Mg) were identified using the atomic absorption spectrometer.
Figure 6: Concentration of foreign material against exposure time [9]
From the above graph, silicon ions were identified as the highest concentration followed by calcium and magnesium. As the exposure time increased, the concentration of the silicon and calcium increased while the concentration of magnesium decreased. Similar to the effect of exposure time, an increase in temperature also resulted in an increase in the amount of foreign material leeched out from the silicone elastomer seals. Therefore, it was concluded that the concentration of leeched silicone and calcium increased with temperature and exposure time.
The presence of foreign materials (e.g. silicon, calcium) could be due to the attack of fillers such as silicon dioxide and calcium carbonate by the ADT solution in which the silicone elastomer seals were submerged. [9-12]
2.1.4 Microindentation
A microindentation test was used by Tan et al. [9, 11-12] to assess the change in mechanical properties of the silicone elastomer seals before and after exposure to the fuel cell environment. Figure 5 shows the load-indentation depth curve for the unexposed silicone sample (a), and exposure at 60oC (b) and 80oC (c) for 45 weeks respectively.
Figure 7: Load-indentation depth curves for silicone elastomer seals after
45 weeks exposure to various temperatures [11]
From the above graph, it was observed that the silicone sample exposed to 80oC had the highest indentation load followed by exposure at 60oC and then the unexposed sample. Hence, it could be concluded that samples exposed to the fuel cell environment hardened over time and the effect of surface hardening was more severe for higher temperature which resulted in higher indentation load. [9, 11-12]
2.1.5 Dynamic Mechanical Analysis (DMA)
Dynamic mechanical analysis was conducted by Tan et al. [12] on the silicone elastomer seals before and after exposure to the ADT solution for 46 weeks. The storage modulus (E'), loss modulus (E") and Tan δ curves were obtained as a function of temperature for samples before exposure (a), exposure at 60oC (b) and 80oC (c) as shown in Figures 8-10.
Figure 8: Graph of storage modulus Figure 9: Graph of loss modulus (E') versus temperature [12] (E") versus temperature [12]
From Figures 8 and 9, it was noted that the storage modulus (E') and loss modulus (E") decreased drastically around the glass transition temperature (Tg) of -42oC. Further observation after the glass transition temperature showed no significant differences because the storage modulus became constant while the loss modulus decreased slightly.
Figure 10: Graph of Tan δ versus temperature [12]
Figure 10 on the other hand, illustrates that there was no shift in the peaks for all the exposed and unexposed samples which indicated that glass transition of the samples remained the same. Therefore it was concluded that the changes in dynamic mechanical properties were insignificant because the bulk properties of the material were not affected even though the surface chemistry of the sample changed as a result of the exposure to the environment. [12]
2.1.6 Optical microscopy
Optical microscopy was used by Tan et al. [9-11] to observe the surface degradation of silicone elastomer seals before and after exposure to the simulated fuel cell environment. Optical micrographs for the unexposed sample (a), exposure at 80oC for 5 weeks (b), 10 weeks (c), 12 weeks (d), and 26 weeks (e); and exposure at 60oC for 12 week (f) and 26 week (g) are shown in Figure 11.
Figure 11: Optical micrographs of silicon elastomer seal surfaces [9]
From Figure 11, the optical micrographs revealed that the surface of silicone elastomer seals deteriorated from smooth to rough to crack appearance and eventually crack propagation upon prolonged exposure. Furthermore, exposure at higher temperature resulted in more severe degradation for the same exposure time as indicated by the increased crack size.
Therefore, it was concluded that the degree of surface degradation for silicone elastomer seals increased when the temperature and exposure time increases. [9-11]
2.1.7 Weight Change
Silicone elastomer seals exposed to simulated fuel cell environment at 60oC and 80oC were measured for their weight change at selected time intervals by Tan et al. [9, 11-12]. The percentage weight loss per square centimetre at exposure times of 3, 5, 7 and 10 weeks with and without compressive load are recorded in Table 2.
Table 2: Percentage weight loss per square centimetre of test samples [9]
From Table 2, it is shown that higher temperature and longer exposure time would lead to an increase in the percentage weight loss. In addition, it was observed that similar results were obtained for an exposure temperature of 80oC with and without the compressive load. However, significant differences were recorded for the test samples at 60oC with and without the compressive load.
From these results, it was concluded that compressive load and temperature were the dominant factors for the weight loss at 60â-¦C and 80â-¦C respectively. [9] In addition, Cleghorn et al. [13] monitored the thinning of actual silicone elastomer seals after exposure which could be the result of weight loss. Typical thickness reduced by approximately 25 µm after 26,300h life test at conditions relevant to stationary fuel cell applications and more severe degradation was observed at active area edges between the MEA and gas flow channels.
2.2 Effect of environmental factors on degradation of silicone elastomer seals
The operating environment of fuel cells can vary in terms of temperature, pressure (compression), humidity and acidity. The effect of temperature on the degradation of silicone elastomer seals was studied by many authors including Tan et al. [9-12] and Patel et al. [14]. From their research conducted, it was apparent that higher temperature would hasten the process of degradation as revealed by the different methods of analyses described in the sections above. Similarly, higher compressive forces would also exacerbate the degradation process as pointed out by Hinds [15] and from the results obtained by Tan et al. [9].
In addition, a highly humid environment (moisture) might contribute to the degradation of silicone elastomer seals by de-crosslinking via hydrolysis of crosslink sites. However, Husar et al. [16] had reported that degradation of silicone elastomer seals occurred even in an environment without humidification at 60oC stack temperature, 4.5 bar gas pressure for 20h of operation. Therefore it can be concluded that the effect of moisture is not as significant for the degradation of silicone elastomer seals as temperature and pressure.
Lastly, due to the nature of the acidic electrolyte membrane, all the experiments were performed in an acidic environment. Although there are limited researches conducted to study the effect of acidity on the degradation of silicone elastomer seals in a fuel cell environment, Gravier et al. [17] had performed a study on the degradation of silicone elastomer in soil. The results obtained indicated that the level of acidity affects the rate of degradation along with the type of soil, organic matter, temperature and moisture content. Hence further investigation would be required to study the sole effect of acidity on the degradation of silicone elastomer seals in a fuel cell environment.
2.3 Mechanisms of silicone elastomer degradation
Silicone elastomer seals exposed to a fuel cell environment may undergo degradation resulting in network changes such as chain scissioning in the backbone or de-crosslinking via hydrolysis of crosslink sites as shown in Figure 12 and 13.
Figure 12: Process of chain scisson in silicone elastomer [18]
Chain scissoning in the backbone of silicone elastomer occurs in the presence of water. Thomas [18] reported that the rate of scisson increases when the content of water vapour in the environment increases. Therefore the rate of scisson reaches a maximum when the environment becomes saturated with water vapour. A typical process of chain scissoning is illustrated in Figure 12. [18]
Figure 13: Process of de-crosslinking reaction at crosslinked sites [9]
On the other hand, degradation of silicone elastomer seals could also happen by de-crosslinking via hydrolysis of crosslink sites with breakage of the Si-C bond through hydrolysis to form Si-OH groups. Upon prolonged exposure to the ADT solution, the Si-OH groups combined to form the Si-O-Si as shown in Figure 13.
In contrast to the above softening mechanisms, hardening of silicone elastomer seals could also occur by oxidative crosslinking where the methyl group is separated from the silicon under the influence of oxygen after prolonged exposure to high temperatures above 200oC. [8]
3. Conclusion from the literature review
In this literature review, numerous methods to analyse the degradation of silicone elastomer seals after exposure to simulated fuel cell environment have been described. These methods could be categorised under chemical, mechanical and physical analyses.
Within the methods of chemical analyses, ATR-FTIR, XPS and atomic absorption spectroscopy were employed to study the surface chemistry degradation. Under the mechanical analyses, microidentation and DMA were performed to assess the changes in mechanical properties. Lastly, optical microscopy and weight change are examples of physical analyses used to monitor the surface degradation and the effect on sample weight.
Between the various chemical analyses, ATR-FTIR would be the most useful chemical technique because of its ability to detect the vibration of peaks from specific chemical groups (e.g. Si-CH3 or Si-O-Si) when the chemical network changes. As for the mechanical methods, microindentation would be preferred over DMA because it is able to judge the degree of sample hardening after exposure to the simulated fuel cell environment. Lastly, optical microscopy was the ideal choice of physical analysis for monitoring the process of surface degradation.
In order to fully investigate the severity of degradation of silicone elastomer seals, at least one analysis technique from each of the individual categories (e.g. chemical, mechanical and physical) would be required with ATR-FTIR, microindentation and optical microscopy being the preferred choice.
