**3. Degradation Diagnoses using Accelerated Stress Test (AST) Procedures**

To analyze degradation data of the PEMFC components and system, the traditional observation of durability information involves the degradation data as a function of time. PEMFC typically show a continuing degradation in power output during their operations, and the cumulative influence of the gradual degradation is still acceptable. The degradation will be improper if the cumulative impacts of continuing degradation become too high. Regarding the characterizations, the modifications of accelerated stress test (AST) protocols have been created for diagnosing the PEMFC degradation behaviour to decrease those restrictions. ASTs are regularly created depending on a specific application, since the performed PEMFC degradations associate to the different cell components, the origins of the stresses, and their influences. Elevated temperature, reduced humidity, open-circuit voltage, and cycling conditions are the foremost accelerated factors ordinarily used for accelerated life diagnosing. The dynamic conditions consist of relative humidity (RH), temperature, potential, freeze/thaw, or start/stop [35]. The crucial conditions for transportation requirements are dynamic load cycling, startup-shutdown, and freeze–thaw [36].

The protocols were created to represent working behaviour in each application. The protocol can be divided into 2 sub-protocols such as static driving cycle and dynamic driving cycle. The Department of Energy (DOE) created the first protocol for the 2000-hour test. After that, the New European Drive Cycle (NEDC) was created, and it was created under the assumption of the European driving behaviour at the maximum speed as 50 km/h (20 A) [37, 38]. In our previous work [39] driving protocol which is the combination of load cycling and start-stop behaviors was designed as presented in **Figure 5**. The proposed load cycling approach intended to accelerate the influences caused by the real operating conditions correlated to a generic dynamic load. The primitive concept was to design the AST protocol right on the real working conditions, assuming the load profile as the major degradation source. Thus, working conditions were accelerated stressing the real load cycles magnitude and frequency. To assure the similar degradation mechanisms the progress of the real load value was evaluated and kept during the accelerated cycle. The methodology allowed accomplishing the AST profile steady with the actual load dynamics, but amplified in magnitude and scaled in the time domain. Finally, the scaled cycles were repeated in a loop [39]. The profile of start-stop cyclic represented starting and shutting down a vehicle in a short time. At the starting situation, electrochemical reactions were speedily fed into the fuel cell to generate the desired power. On the other hand, the reactions were terminated by stopping the reactant supply.

This dynamic behaviour severely involved the operating condition changing that would cause material degradation. This created protocol presented an overview of ordinary approaches adopted in hybrid FCEV for power management, being the starting point for the load profile set-up for PEMFC operations [39].

The results of voltage degradation indicated that the voltage drop produced by load cycling gradually increased that corresponded to slowly operating condition differentiation. In contrast, the voltage degradation rate was significantly increased. This occurrence corresponded to the voltage lost in driving the chemical reaction at both on anode and cathode. The reduction of oxygen is a much slower reaction than the oxidation of hydrogen, therefore the system requires higher activation polarization, anode side losses can be neglected. This ageing regarded to damaging electrochemical surface areas of catalysts. The resistance to an electron flow through the electrically conductive PEMFC components and to an ion flow through the membrane caused a voltage drop as well. This loss generally happens in main components; bipolar plates, gas diffusion layers, catalyst, and membrane. The consumption of reactant gases at the catalyst layers leads to concentration

**Figure 5.** *The driving behaviors protocol [39].*

## *Hydrogen Fuel Cell Implementation for the Transportation Sector DOI: http://dx.doi.org/10.5772/intechopen.95291*

gradients and the partial pressure of the reactants changing, affecting a decrease in fuel cell voltage [40]. **Figure 6** illustrates the case study of sudden load variation behaviour when the car speed varies from a low speed like driving to suddenly adding acceleration and overtaking cars in front. At high load demand, the system requires high power and voltage leading to high energy to drive electrochemical reactions thermodynamically. Under this situation system temperature increases significantly, while relative humidity decreases. This operating condition may result in membrane degradation; chemical and/or mechanical degradation [41]. In terms of chemical degradation, radicals such as peroxide or/and hydrogen peroxide radically react with the backbone of the membrane (Nafion: polytetrafluoroethylene). On the other hand, this reaction cannot occur if the backbone is not fluorinated [42–44].

The following mechanism, presented as Eqs. 1–3 [45], illustrates membrane damage where hydrofluoric acid was produced. **Figure 7** shows the morphological feature of a membrane electrode assembly (MEA) observed by scanning electron microscope and energy-dispersive X-ray (SEM–EDX) after 888 hours of operation duration. The investigated results found that the gas diffusion layer was dissolved by fluorine leaching, and the gas diffusion layer lost weight around 30.95%wt [46]. Small spots in **Figure 8** shows the catalyst removal from catalyst supports.

**Figure 6.** *The cyclic profile represents sudden load variation.*

**Figure 7.** *SEM–EDX micrographs of gas diffusion layer damaged by hydrofluoric acid leaching.*

#### **Figure 8.**

*SEM–EDX micrograph of the catalyst layer on the PEMFC membrane.*

Chemical degradation of the membrane [45]

$$\text{R}-\text{CF}\_2\text{COOH}\_2 + \text{HO}^\* \rightarrow \text{R}-\text{CF}\_2 + \text{CO}\_2 + \text{H}\_2\text{O} \tag{1}$$

$$\text{R}-\text{CF}\_2^\* + \text{HO}^\* \rightarrow \text{R}-\text{CF}\_2\text{OH} \rightarrow \text{R}-\text{COF} + \text{HF} \tag{2}$$

$$\text{R}-\text{COF} + \text{H}\_2\text{O} \rightarrow \text{R}-\text{COOH} + \text{HF} \tag{3}$$

The relative humidity cycling related to load cyclic profile generated membrane swelling and shrinking. This phenomenon is associated with hydration state and operating temperature called mechanical degradation [27]. The postmortem analyses from the humidity cycling tests suggest that in-plane tensile membrane stresses result in cracks to initiate and propagate within the membranes in a subcritical fashion. Cyclic mechanical stresses cause if hydrophilic membranes are exposed to fluctuating hygrothermal conditions during the PEMFC operation [47]. Moreover, these phenomena may affect to catalyst degradation such as Ostwald ripening or sintering of catalyst particles. The Ostwald ripening appears from the thermodynamic driving force, while the sintering is caused by the reduction in surface energy with particle growth. The phenomenon leads to the dissolution of smaller particles and the growth of larger particles related to Eq. 4–6. In PEMFC, the platinum (Pt) transformation via a coupled process involving the transport of ions (Pt2+ and/or Pt4+) through an ionomer/aqueous medium and a parallel (coupled) transport of electrons through the carbon support [48].

Pt dissolution reactions [49]

$$\text{Pt} \rightarrow \text{Pt}^{2+} + 2\text{e} - \quad \text{(1.188 V vs. RHE)} \tag{4}$$

$$\text{Pt} + \text{H}\_2\text{O} \rightarrow \text{PtO} + 2\text{H}^+ + 2\text{e}^- \quad \text{(0.980 V vs. RHE)}\tag{5}$$

$$\text{PtO} + 2\text{H}^\* \rightarrow \text{Pt}^{2\*} + \text{H}\_2\text{O} + 2\text{e}^- \quad \text{(0.208 V vs. RHE)}\tag{6}$$

The catalyst degradation can be diagnosed via several techniques; electrochemical impedance spectroscopy (EIS) technique, cyclic voltammetry (CV) technique, SEM–EDX technique, and X-ray photoelectron spectroscopy (XPS). The investigated results from EIS explain the impact of ionic resistance, activation loss (relate to the loss of the electrochemical surface area), and mass transport losses [50]. The CV can be used to diagnose the evaluation of catalyst activity. With is analyzing, the counter and reference electrodes generally act as an anode side because the kinetics of the oxidation reaction is relatively fast. Working electrode acts as a cathode side

*Hydrogen Fuel Cell Implementation for the Transportation Sector DOI: http://dx.doi.org/10.5772/intechopen.95291*

because the reduction reaction is a rate-determining step [51]. X-ray photoelectron spectroscopy (XPS) is a quantitative spectroscopic technique applied to characterize the surface chemistry of materials or to measure the elemental composition, empirical formula, chemical state, and electronic state of the elements existing in materials [51].

At low load demand in the sudden load variation, the system requires low voltage impacting slow reaction, so the system temperature decreases related to an increase in relative humidity. This operating condition causes a flooding phenomenon that can be studied via a hysteresis loop. The different voltage values from upward current testing and downward current testing in the hysteresis loop indicate accumulated water inside the PEMFC. The water flooding effects catalyst oxidation, reactant starvation, electro-osmosis, and back diffusion [52].
