A Novel Antiflooding Electrode for Proton Exchange Membrane Fuel

Corresponding author. E-mail: [email protected]. Telephone: +86 23 60891548. Fax: +86 23 65106253., †. State Key Laboratory of Power Transmission Equ...
0 downloads 5 Views 1MB Size
Download PDF
J. Phys. Chem. C 2009, 113, 765–771

765

A Novel Antiflooding Electrode for Proton Exchange Membrane Fuel Cells M. B. Ji,†,‡,§ Z. D. Wei,*,†,‡,§ S. G. Chen,§ and L. Li§ State Key Laboratory of Power Transmission Equipment & System Security and New Technology, School of Chemistry and Chemical Engineering, and School of Material Science and Engineering, Chongqing UniVersity, Chongqing 400044, China ReceiVed: September 2, 2008; ReVised Manuscript ReceiVed: October 26, 2008

Water management in proton exchange membrane fuel cells (PEMFCs) is critical. Although water flooding has been extensively studied, all measurements that have been reported so far only involve the water removal accumulated at the gas flow channels of a bipolar plate. There are few reports that aimed to overcome water flooding happening in the pores of a porous electrode so far. This study is thus devoted to solve the water flooding happening in the pores of a porous electrode by using an antiflooding electrode (AFE) which contains water-proof oil, dimethyl silicon oil (DMS). The success of the AFE in antiflooding is due to the following: (1) DMS, in which the solubility of oxygen is over 10 times higher than that in water, supplies an unoccupied channel for oxygen transportation, (2) it solves the flooding in pores with a diameter of 20-70 nm, in which flooding often happens and is not easily eliminated by conventional means, and (3) the flooding to the porous electrode itself rather than the gas channels in the bipolar plate is principally responsible for the deterioration of the PEMFC performance. The AFE solves the flooding happening within the porous electrode, and thus, it solves the foremost problem about water flooding. 1. Introduction Proton exchange membrane fuel cells (PEMFCs) are expected to become a major power source of electric vehicles. Figure 1 shows a schematic of the PEMFC and processes happening in the PEMFC operation. The operation principle of the PEMFC is as follows: At the anode, fuel H2 is oxidized, liberating electrons and producing protons. The electrons via the external circuit and protons via the proton exchange membrane inserted tightly between the anode and the cathode flow to the cathode, where they combine with the dissolved O2 to produce water and heat. Obviously, thermal management is necessary to remove the heat in order to prevent overheating and dehydration of the membrane. Proper water management is also important, which ensures the membrane remains fully hydrated and therefore maintains good ionic conductivity and hence cell performance.1 Otherwise, if water was not removed in a timely manner, it would accumulate at the cathode of the PEMFC and further flood the cathode.2-7 With water accumulation and the cathode flooded, oxygen ingress into the catalyst surface in the porous cathode will be hindered.8-11 The lack of oxygen reaching the catalyst leads to oxygen under-stoichiometry or “starvation” at the cathode.12 Under steady state conditions, the net mass flow rate of oxygen into the system is equal to the oxygen consumed by the electrochemical reaction. In transient conditions, in particular, a sudden increase in power requirement for the fuel cell, oxygen supply to the system lags behind the demand and causes a shortage of oxygen and local oxygen starvation for the reaction. Worse still, proton (H+) reduction Figure 1. Schematic of processes happening in the CPE and the AFE. * Corresponding author. E-mail: [email protected]. Telephone: +86 23 60891548. Fax: +86 23 65106253. † State Key Laboratory of Power Transmission Equipment & System Security and New Technology. ‡ School of Chemistry and Chemical Engineering. § School of Material Science and Engineering.

reaction (PRR) instead of oxygen reduction reaction (ORR) would occur at the cathode if oxygen were depleted at the cathode.

10.1021/jp807773m CCC: $40.75  2009 American Chemical Society Published on Web 12/19/2008

766 J. Phys. Chem. C, Vol. 113, No. 2, 2009

Ji et al.

In the case of oxygen starvation, the original electron consuming process, ORR

O2+4H++4e-)2H2O

(E° ) 1.23 V)

(1)

would be replaced by the new electron consuming process, PRR

2H3O++2e-)H2+2H2O

(E° ) 0.00 V)

(2)

In this case, the cathode potential would drop from 1.23 to 0.00 V at current off and probably from 0.7 to -0.1 V at current on with substitution of the PRR for the ORR. Electrode polarization occurs with current flowing through the electrode. Cathode polarization causes a potential shift toward the negative direction, but anode polarization causes a shift toward the positive direction. For the hydrogen oxidation reaction (HOR) taking place at the anode,

H2+2H2O ) 2H3O++2e-

(E° ) 0.00 V)

(3)

with current flowing, the electrode potential will shift toward the positive direction, for example, 0.1 V. Thus, the cell voltage would change from the original 0.6 V by (0.7 - 0.1) to -0.4 by (-0.3 - 0.1). The output voltage of the cell, as long as it is under oxygen starvation, would be likely reversed with substitution of the PRR for the ORR from a positive value, for instance, 0.6 V, to a negative value, for instance, -0.4 V. This phenomenon is known as the voltage reversal effect (VRE) in the PEMFC. As long as the VRE appears, the output of a PEMFC stack will be seriously impaired. In a typical PEMFC design, water content is maintained by humidifying the reactant gases. At higher current densities, the excess product water is removed by convection via the air stream, and the rate of removal is controlled by judiciously adjusting moisture content, pressure drop, rate of air stream, and temperature in the flow channels. In developing a complete miniaturized power delivery system, where it is almost impossible to operate a fuel cell actively with an active control over the gas flow in and out of the device, performance enhancements can be achieved only if one is able to move water within the device from where water is produced in excess to where it is often depleted. An electrical fan tapping power from the fuel cell itself is normally used to ensure oxygen supply and remove excessive water.13,14 Although the fan can solve the problem of oxygen starvation to a certain extent, it dries out the membranes and offsets the improvement in performance due to the power consumption incurred by the fan. In addition, all means via strengthening gas convection only remove the water accumulated at the gas flow channels of a bipolar plate but do not remove the water in the pores of the catalyst layer (CL) as shown in Figure 1. Flooding of membrane electrode assemblies (MEAs) and the liquid water transport in fuel cells were observed14,15 and diagnosed.16-18 A great deal of recent modeling activity has been concentrated on in numerical simulations about water flooding in MEAs.8,9,19-21 Meanwhile, much attention has been paid to flooding in the catalyst layer and gas diffusion layer (GDL).22-30 Wang and co-workers31 added some magnetic particles inside the cathode catalyst layer to attract the paramagnetic oxygen to the catalyst layer while expelling the diamagnetic liquid water out of the MEA. Adding a sublayer microporous layer (MPL) between the catalyst layer and the GDL is considered an effective method to improve the liquid water drainage and gas diffusion.20,32,33 A MnO2-Pt/C composite electrode was designed to solve the VRE caused by oxygen starvation,34 which was based upon the fact that the electrochemical reduction of

MnO2 has almost the same Nernstian potential as the ORR. It has been found that the MnO2-Pt/C composite electrode can overcome the voltage reversal effect to a certain extent. Even though the discharged MnO2 can be recovered after a length of time at rest, the effect of discharged products Mn2+ on the MEA is still under question. Although water flooding has been extensively studied, all measurements having been reported so far only involve the water removal accumulated at the gas flow channels of a bipolar plate. There are few reports that aimed to overcome water flooding happening at the pores in the CL and the GDL of a porous electrode so far. Actually, the pores in a porous electrode would be no sooner flooded than the first drop of water was produced at the catalyst layer of a PEMFC electrode. Thus, the performance of the PEMFC reported so far was actually discounted on account of water flooding happening in the catalyst layer of the electrode. In order to shorten the distance of the gas diffusion in a water-flooded porous electrode, an ultrathin catalyst layer is often recommended in a PEMFC. That is why the catalyst coated membrane (CCM) technique, in which only a very ultrathin catalyst layer is employed, has prevailed for years. Obviously, an ultrathin catalyst layer sacrifices the electrodes’ space efficiency of the PEMFC. The objective of this work is thus devoted to solve the water flooding happening in the pores of a porous electrode with the development of a novel cathode containing water-proof oil, which was added into a fraction of pores of a cathode of a conventional Pt/C electrode. The solubility of oxygen in such water-proof oil is over 10 times higher than that in water. Thus, the pores occupied by such water-proof oil form permanent channels for oxygen transportation, in which the oil is hardly extruded by water, regardless of whether the cathode is flooded or not. The electrode containing such water-proof oil is named antiflooding electrode (AFE) hereafter. The electrode without such water-proof oil is named a conventional Pt/C electrode (CPE). The performance of the AFE was characterized electrochemically and nonelectrochemically. The single PEMFC cell with the AFE displays much better power output not only in the case of water flooding but also in the case of well-designed operation conditions than the cell only with the CPE. 2. Experimental Section 2.1. Preparation of the AFE and MEA. The CPE was prepared as described previously.35 In short, the CPE was composed of a gas diffusion layer and a catalyst layer. The gas diffusion layer was prepared on wet-proofed carbon paper. The carbon powder (Vulcan XC-72, Cabot Corp.), 30 wt % PTFE, and ethanol were ultrasonically mixed with a ratio 4:1 of carbon to solid PTFE content. The viscous mixture was coated onto carbon paper (Tony Co, Jap.) wet-proofed by PTFE and then heated at 340 °C for 30 min. A suspension consisting of catalysts 40 wt % Pt/C (Johnson-Matthey), 0.5 wt % Nafion solution (Du Pont), and isopropyl alcohol was first ultrasonically mixed for about 15 min. The ratio of Pt loading to solid Nafion was maintained at 3:1. The suspension was pipetted onto the gas diffusion layer and finally heated at 145 °C. The AFE was prepared on the basis of the above CPE. First, the water-proof oil, dimethyl silicon oil (DMS), was sprayed onto the catalyst side of the above CPE and then extracted with a water pump from the opposite side of the electrode to make the oil pass through the porous electrode and remove the surplus oil. The electrode with some DMS residue filling in part of pores in the CL and GDL was heated at 120 °C to remove the solvent in the DMS. The loading of the DMS was calculated by

Novel Antiflooding Electrode for PEMFCs

J. Phys. Chem. C, Vol. 113, No. 2, 2009 767

determining the mass difference of the AFE before and after DMS treatment. The membrane electrode assembly (MEA) was fabricated by hot pressing a sandwich of a CPE, a Nafion 112 membrane (Du Pont), and an AFE in series at 137 °C and 6 M Pa for 3 min. The Nafion 112 membrane (Du Pont) was pretreated with 3 vol % H2O2 and 0.5 M H2SO4 for 1 h to remove impurities. The membrane was then washed several times with hot ultrapure water. For the sake of effective contact between the catalyst and the membrane, Nafion solution was brushed on the catalyst layer and dried in an oven at 80 °C before being pressed into the MEA. 2.2. Test of the AFE in the Case of Complete Water Flooding. An AFE together with a saturated silver chloride reference electrode (SSCE) (0.20 V vs SHE) and a Pt wire counter electrode were immersed in 0.5 mol L-1 H2SO4 to simulate the situation of an electrode flooded completely by water. 2.3. Test of a Single Cell with the AFE. The test of a single cell with an AFE cathode or a CPE cathode was conducted on the Fuel Cell Test Station (Fuel Cell Technologies, Inc.). Pure hydrogen and oxygen were fed as the fuel and the oxidant, respectively. The long run test was conducted according to the method reported in ref 36; that is, electrical current (1 A cm-2) was drawn continuously until the cell voltage declined nearly to 0 V, after a 50 h equilibration period, and polarization curves were collected before and after long-term examination. Before each experiment, a nitrogen purge process was used to clean out the residual fuel and oxidant in the cell and pipelines. 2.4. Characterization of Porosity Distribution in a Porous Electrode. The change in internal structure of the CPE before and after introduction of the DMS was assessed using a N2 adsorption specific surface area analyzer (ASAP2010, Micromeritics Instrument Corp.). The surface area of the electrodes employed in this item is 2 × 2 cm2.

solubility of oxygen in solvent is directly proportional to its partial pressure in the gas phase. With temperature increasing, water vapor partial pressure increases and then the oxygen partial pressure decreases in the gas phase. Thus, the solubility of oxygen in water will be close to zero (0.003 mL · mL-1) at 80 °C due to increased water vapor pressure.38 While the DMS vapor pressure remains as small as ∼10-6 k Pa in a range of 0-100 °C, it means the solubility of oxygen in the DMS almost maintains unchanged in a temperature range of PEMFC operation. Besides, the relatively low freezing point (-80 °C) and quite high flash point (195 °C) of DMS are also favorable to PEMFC operation and MEA hot press. In addition, its surface tension (∼19 mN m-1) is approximately 1/3 that of water (∼66 mN m-1). It implies that DMS has much better penetrability into the deeper pores than water and can bring the dissolved oxygen into the deeper pores of a porous electrode. The change of the O2 diffusion coefficient with DMS as a substitution for water can be estimated according the Wilke-Chang’s equation39

3. Criteria for Selection of Water-Proof Oil

where Dcl is the oxygen diffusion coefficient in solution of the catalyst layer and L is the thickness of the catalyst layer. In consideration of a big difference in oxygen solubility COcl2 in water (0.003 mL · mL-1) and DMS (0.168-0.190 mL · mL-1) at 80 °C, with a half-loss in O2 diffusion velocity (2.1 × 10-5 cm2 s-1 in water vs 1.0 × 10-5 cm2 s-1 in DMS), we get a benefit of 28-32 times higher oxygen diffusion flux with substitution of the DMS for water.

Water-proof oil if used successfully in a PEMFC system must meet the following requirements: (1) Chemically inert and thermally stable. (2) Nonpolarity molecule structure. According to the principle “like dissolves like,” oxygen will dissolve better in a nonpolar solvent than in a polar solvent. (3) Freezing point being as low as possible but flash point as high as possible, which should cover all temperatures of PEMFC operation and MEA formation (hot-press at about 140 °C). (4) A low vapor pressure over a large temperature scope to guarantee high oxygen solubility at any temperature, low or high. (5) Low viscosity required for efficient diffusion of dissolved oxygen. (6) Small surface tension that is helpful for the oil itself to permeate into the deep pores of a porous electrode. There is somewhat a contradiction among the above criteria. For example, a macromolecule with a small molecular weight usually has a low freezing point and low viscosity but not a high flash point, low vapor pressure, or large surface tension, and vice versa. With a hard compromise, DMS with a mole molecular weight of 2000 g mol-1 comes into sight, which is a kind of transparent liquid with colorless, flavorless, and nontoxic properties. Its viscosity is as low as 15 mPa · s (25 °C). The solubility of oxygen in DMS is 0.168-0.190 mL · mL-1 (25 °C),37 which is 10 times higher than that in water (0.0171 mL · mL-1) (25 °C). Moreover, according to Henry′s law, the

° DAB ) 7.4 × 10-8

(φMB)1⁄2T µBV0.6 bA

(4)

° where DAB is diffusion coefficient of solute A in solvent B. MB and µB are mole molecular weight and viscosity (mPa · s) of solvent B, respectively. T is the absolute temperature (K). VbA is mole volume of solute A (for O2: 25.6 cm3 g-1 mol-1) at normal boiling point. φ is a factor associated with solvent B, which is 2.6 for water and 1 for DMS. Thus, the diffusion coefficient of oxygen is 2.1 × 10-5 cm2 s-1 in water and 1.0 × 10-5 cm2 s-1 in DMS at 298 K. This computed result indicates that there is a half-sacrifice of the oxygen diffusion coefficient in the DMS relative to in water. The oxygen diffusion limiting current density il is directly proportional to oxygen solubility COcl2 in water or DMS:40-43

il ) (nFCOcl2Dcl) ⁄ L

(5)

4. Results and Discussion 4.1. Performance of the AFE in a Situation of Complete Water Flooding. An experiment in which electrodes AFE and CPE were completely dipped in O2-saturated H2SO4 was employed to simulate complete water flooding of a PEMFC electrode. Figures 2-4 record chronopotentiometry of electrodes AFE and CPE dipped completely in 0.5 mol L-1 H2SO4, among which a gradually increased cathode current (Figures 2 and 3) and a constant cathode current (Figure 4) were imposed on the electrodes, respectively. In addition, in some cases, O2 was always bubbled into 0.5 mol L-1 H2SO4 before and after measurement (Figures 2 and 4), and in other cases it was only bubbled before measurement (Figure 3). Figure 5 records chronoamperometry of electrodes AFE and CPE dipped completely in 0.5 mol L-1 H2SO4, among which O2 was always bubbled into 0.5 mol L-1 H2SO4 before and after measurement. With current flow through the electrode/electrolyte interface, the electrode potential will depart from the balance value. The electrode potential is going to shift toward the negative direction

768 J. Phys. Chem. C, Vol. 113, No. 2, 2009

Figure 2. Chronopotentiometry of electrodes AFE and CPE under modulation of gradually increased current in O2-saturated 0.5 mol L-1H2SO4 (oxygen was bubbled at 10 mL min-1 during the whole experiment; Pt loading for all electrodes was 0.8 mg cm-2).

Figure 3. Chronopotentiometry of electrodes AFE and CPE under modulation of gradually increased current in 0.5 mol L-1H2SO4 (oxygen was fed at 10 mL min-1 for 20 min before measurement and then stopped after measurement; Pt loading for two electrodes was 0.8 mg cm-2).

Figure 4. Chronopotentiometry of electrodes AFE and CPE at a fixed cathode current of 10 mA cm-2 in O2-saturated 0.5 mol L-1 H2SO4 (oxygen was fed at 10 mL min-1 for 20 min before measurement and continuously fed during the whole measurement; Pt loading for all electrodes was 0.8 mg cm-2).

for cathode polarization and toward the positive direction for anode polarization. Figures 2-5 tell that, in the case of O2 continuous feeding, the AFE can sustain a much larger polariza-

Ji et al.

Figure 5. Chronoamperometry of electrodes AFE and CPE as the electrode potential jumps from 0.8 to 0.2 V in O2-saturated 0.5 mol L-1H2SO4 (oxygen was fed at 40 mL min-1 for 20 min before measurement and continuously fed during the whole measurement; Pt loading for all electrodes was 0.8 mg cm-2).

Figure 6. Potential dependence of the Nyquist plots obtained on the AFE with 2.0 mg DMS cm-2 and on the CPE in O2-saturated 0.5 mol L-1 H2SO4 (6 × 104-6 × 10-3 Hz; oxygen was bubbled at 60 mL min-1 for 20 min before measurements and then stopped after measurements; Pt loading for all electrodes was 0.8 mg cm-2).

tion current (Figures 2 and 5) and a much longer time (Figure 4) than the CPE. It means that the DMS in the AFE can supply reliable and unoccupied channels for oxygen transport. Figure 3 indicates that more dissolved O2 was stored in the AFE than the CPE because the AFE can sustain a larger polarization current than the CPE in 0.5 mol L-1H2SO4, as the oxygen feed was stopped after the measurement. The marked drop in electrode potential is a symbol of O2 depletion in the electrode. With oxygen starvation, the voltage reversal effect (VRE) will soon appear in succession. For example, in Figure 4, the CPE cannot sustain longer than 1 h before the electrode potential drops to -0.1 V, but the AFE can sustain longer than 40 h without obvious potential deterioration at a polarization current density of 10 mA cm-2. The excellent antiflooding capability of the AFE in the case of complete flooding is no doubt contributed to the unoccupied channels filled by the DMS and the high solubility of oxygen in such DMS-filled channels. Electrochemical impedance spectroscopy (EIS) of electrodes AFE and CPE shown in Figure 6 further confirms the above analysis. There are two arcs in Figure 6. One at high frequency provides information about charge transfer at the electrode/ electrolyte interface, and one at low frequency provides information about the diffusion process of oxygen44,45 through

Novel Antiflooding Electrode for PEMFCs

J. Phys. Chem. C, Vol. 113, No. 2, 2009 769

TABLE 1: Fitted Parameters Based on the Nyquist Plots of Figure 6 0.40 V electrodes CPE AFE

0.30 V

Rd, Cdl, Rct, Rd, Cdl, Rct, Ω cm2 Ω cm2 mF cm-2 Ω cm2 Ω cm2 mF cm-2 13.8 12.9

23.1 16.8

36.9 39.3

8.15 6.62

70.2 49.5

61.5 64.9

the flooded pores in the porous cathode electrode structures.46 The fitted parameter based in Figure 6 is summarized in Table 1, in which Rct is the charge-transfer resistance, Rd is the diffusion resistance, and Cdl is the capacity of the double layer. According to a mathematic model built up to characterize porous electrodes,47-49 Rct and Rd have the following expressions:

Rct ) 2FaH+

{

2

1 dK1 dK2 CO2(1 - θs) + βθs dV dV

Rd ) RTd ⁄ n2F2DAcm

}

(6)

Figure 7. Cell voltage versus time of a MEA with an AFE or a CPE cathode at a current density of 1 A cm-2. Test conditions: back pressure, PO2 ) PH2 ) 180 kPa; O2 and H2 gases flowing at 150 and 160 sccm, respectively; Pt loading was 0.6 mg cm-2 at both the cathode and anode; Tcell ) 60 °C.

(7)

where R is the gas constant, T is the absolute temperature, d is the diffusion layer thickness, n is the number of electrons involved in the process, F is the Faraday constant, D is the diffusion coefficient of the dissolved oxygen, and cm is the maximum concentration of the dissolved oxygen at the electrode with a surface area A. aH+ represents the activity of protons, CO2 is the concentration of the dissolved oxygen in the electrolyte, β is the maximum concentration of an intermediate state on the electrode surface, and θs is the coverage of intermediate H2O2 on the electrode surface at a steady state. K1 and K2 are the electrochemical rate constants in the following reactions that involve the intermediate H2O2: K1

O2(dissolved) + 2H+ + 2e- 98 H2O2 K2

H2O2 + 2H+ + 2e- 98 2H2O

(8)

(9)

Table 1 shows that the AFE has a much smaller chargetransfer resistance (Rct) and diffusion resistance (Rd) at the same electrode potential than the CPE. With the cathode polarization increasing from 0.4 to 0.3 V, Rd increases from 23.1 to 70.2 Ω cm2 in the case of the CPE but only from 16.8 to 49.5 Ω cm2 in the case of the AFE. It implies that diffusion of the dissolved oxygen at a larger cathode polarization is not as difficult in the AFE as in the CPE. The difference in the value of Rct and Rd obviously comes from contribution of the dissolved oxygen concentration CO2 in electrodes AFE and CPE according to eqs 6 and 7. Fortunately, the introduction of electrically inert DMS into the AFE did not damage the electrode/electrolyte interface because the value of Cdl, which characterizes the electrode/ electrolyte interface, in other words, the capacity of the double layer, did not reduce but slightly increased in the AFE relative to that in the CPE. 4.2. Single Cell Tests. The situation of completely flooded electrodes is an extreme case in a PEMFC. An experiment simulating a real situation of a PEMFC is conducted in a single cell with two CPEs as the anode and cathode and a cell with one CPE as the anode and one AFE as the cathode. Generally, it is hydrogen rather than oxygen that is humidified if only one reactant gas needs to be humidified in a PEMFC because water is generated at the cathode where oxygen is supplied. In order to test the antiflooding capability of the AFE, however, in this

Figure 8. Cell voltage and power density versus current density of a single cell with a CPE anode and an AFE cathode in the case of no O2 humidification (O) and O2 humidification at 156% RH (b), and a single cell with two CPEs in the case of no O2 humidification (4) and O2 humidification at 156% RH (2). Test conditions: back pressure, PO2 ) 182 kPa, PH2 ) 180 kPa; O2 and H2 gases flowing at 180-200 SCCM; Pt loading was 0.7 and 0.5 mg cm-2 in the cathode and anode, respectively; Tcell ) 60 °C.

work, oxygen was deliberately humidified. It means more water was brought into the cathode besides that generated by the PEMFC itself. Therefore, the AFE bears more serious water flooding than that in the usual case. Figures 7 and 8 display the performance of a single cell with two CPEs and a cell with one CPE anode and one AFE cathode in different experimental conditions. The experiment illustrated in Figure 7 was conducted as two sections: in the early 4 h, O2 was not humidified, and in the rest of time O2 was overhumidified at humidity of 156% RH. The voltage of the cell with an AFE cathode is always greater than the cell with a CPE cathode regardless of O2 humidification or not. The cell with a CPE cathode only sustained another 2.5 h and then collapsed after O2 was overhumidified at a humidity of 156% RH. However, the cell with an AFE cathode containing different DMS loadings sustained another 12 and 17 h before collapse. It implies that the cell with an AFE cathode wins at least 9 h more life before advent of the VRE than the cell with a CPE cathode in the case of overhumidity. Figure 8 displays cell voltage and power density versus current density of a single cell with an AFE cathode (O, b) or a CPE cathode (0, 2) in two cases, where oxygen was not humidified and was humidified with 156% RH.

770 J. Phys. Chem. C, Vol. 113, No. 2, 2009

Figure 9. Pore volume distribution of a CPE with an apparent area of 4 cm2 before (black line) and after (red line) addition of 2.5 mg cm-2 DMS.

Ji et al. hydrophobicity even though the large pores are occupied by water. Besides, due to the smaller capillary pressure in pores larger than 70 nm, it is slightly more difficult for water to condense down in such large pores. The water in the pores with a diameter from 20 to 70 nm, however, may be not easily excluded because water capillary condensation in such pores is not as small as that in large pores. The situation will be different if the pores with a diameter from 20 to 70 nm were in advance occupied by oil, such as the DMS, which has a zero contact angle with carbon, Nafion, and Teflon but has a 117° contact angle with water.50 It can be expected that the DMS can well penetrate into the pores configured by carbon, Nafion, and Teflon. Once the pores are occupied by the DMS, the DMS will be hardly extruded by water due to its water-proof property. In other words, thus, solving the problem about flooding of the pores with a diameter of 20-70 nm means solving the foremost problem that has been harassing the PEMFC for years in water management. 5. Conclusions

From the results disclosed in Figures 7 and 8 the following points can be concluded: (1) Regardless of O2 humidification or not, the performance of a single cell with an AFE cathode is always better than the cell with a CPE cathode. It means that even at a well-designed operational condition the flooding of produced water to the pores of the cathode also seriously worsens the PEMFC output. Fortunately, the AFE can overcome it. (2) The performance of the MEA with a CPE cathode deteriorates very quickly after O2 overhumidification, but the cell with an AFE cathode can sustain for quite a long time without serious degradation. It means that the pores for gas transportation in the CPE will be predominantly occupied by water in the case of O2 overhumidification. Most of the loss in a cell’s performance is not from the water flooding to the gas channels in the bipolar plate but from the water flooding to the pores in the catalyst layer. (3) The cell performance at a large current density was noticeably improved with the introduction of the AFE into the MEA. For example, in the case of O2 overhumidification, the cell with two CPEs can only sustain a current density of 1.5 A cm-2 but the cell with an AFE cathode can sustain a current density as large as 3.0 A cm-2; in the case of the nonhumidified cathode, the cell with two CPEs can sustain a current density of about 2.3 A cm-2 but the cell with an AFE cathode can sustain a current density as large as 3.3 A cm-2. The larger the current density is drawn, the more water will be produced, and the more pores for gas transportation will be occupied by water. The difference in the case of the nonhumidified cathode once again proves that the flooding to the porous electrode itself rather than the gas channels in the bipolar plate is principally responsible for the deterioration of the PEMFC performance. Thus, solving the flooding happening within the porous electrode means solving the foremost problem associated with water flooding. The AFE was just invented for this purpose. 4.3. Distribution of the DMS in the AFE. Figure 9 shows the difference in pore volumes of a CPE before and after introduction of the DMS. The main difference in pore volumes appears in the pores from 20 to 70 nm. It suggests that the DMS mainly fills in the pores with meso-diameters from 20 to 70 nm. It is the water condensation in the pores with a diameter from 20 to 70 nm that cause so-called flooding of the electrode. The pores with a diameter below 20 nm may always be occupied by water due to water capillary condensation in such small pores. The pores with a diameter over 70 nm may primarily belong to the gas, and water can be easily excluded because of their

An antiflooding electrode was invented by introduction of water-proof oil DMS into the conventional Pt/C electrode. The experiments results indicate that the DMS mainly distributes in the pores with diameters ranging from 20 to 70 nm. The single PEMFC cell with the AFE cathode displays better power output not only in the case of water flooding but also in the case of a well-designed operational condition than the cell with the conventional cathode. The novel antiflooding electrode displays outstanding antiflooding capability, especially in the case of a large current density and overhumidification. The success of the AFE in antiflooding lies in that (1) it solves the water flooding to the porous electrode itself rather than the water accumulation in the gas channels of the bipolar plate, and (2) it solves the water flooding of the pores with diameters of 20-70 nm, in which water flooding frequently happens and is not easy to remove by routine methods. Acknowledgment. This work was financially supported by NSFC of China (Grant Nos. 20476109 and 20676156), by the Chinese Ministry of Education (Grant No. 307021), China National 863 Program (2006AA11A141 and 2007AA05Z124), andbytheChongqingMunicipalGovernment,China(CSTC2007AB6012). References and Notes (1) Mauritz, K. A.; Moore, R. B. Chem. ReV 2004, 104, 4535. (2) Choe, Y.-K.; Tsuchida, E.; Ikeshoji, T.; Yamakawa, S.; Hyodo, S.-A. J. Phys. Chem. B 2008, 112, 11586. (3) Yan, L. M.; Ji, X. B.; Lu, W. C. J. Phys. Chem. B 2008, 112, 5602. (4) Jang, S. S.; Goddard, W. A., III J. Phys. Chem. C 2007, 111, 2759. (5) Benziger, J. B.; Chia, E.-S.; Decker, Y. D.; Kevrekidis, I. G. J. Phys. Chem. C 2007, 111, 2330. (6) Perrin, J.-C.; Lyonnard, S.; Guillermo, A.; Levitz, P. J. Phys. Chem. B 2006, 110, 5439. (7) Pivovar, A. M.; Pivovar, B. S. J. Phys. Chem. B 2005, 109, 785. (8) Wang, C.-Y. Chem. ReV 2004, 104, 4727. (9) Weber, A. Z.; Newman, J. Chem. ReV 2004, 104, 4679. (10) Tamain, C.; Poynton, S. D.; Slade, R. C. T.; Carroll, B.; Varcoe, J. R. J. Phys. Chem. C 2007, 111, 18423. (11) Um, S.; Wang, C. Y. J. Power Sources 2006, 156, 211. (12) Piela, P.; Springer, T. E.; Davey, J.; Zelenay, P. J. Phys. Chem. C 2007, 111, 6512. (13) Song, R. H.; Kim, C. S.; Shin, D. R. J. Power Sources 2000, 86, 289. (14) Tu¨ber, K.; Po´cza, D.; Hebling, C. J. Power Sources 2003, 124, 403. (15) Yang, X. G.; Zhang, F. Y.; Lubawy, A. L.; Wang, C. Y. Electrochem. Solid-State Lett. 2004, 7, A408. (16) Feindel, K. W.; LaRocque, L. P. A.; Starke, D.; Bergens, S. H.; Wasylishen, R. E. J. Am. Chem. Soc. 2004, 126, 11436.

Novel Antiflooding Electrode for PEMFCs (17) Feindel, K. W.; Bergens, S. H.; Wasylishen, R. E. J. Am. Chem. Soc. 2006, 128, 14192. (18) He, W.; Lin, G.; Nguyen, T. V AIChE J. 2003, 49, 3221. (19) Shah, A. A.; Kim, G. S.; Gervais, W.; Young, A.; Promislow, K.; Li, J.; Ye, S. J. Power Sources 2006, 160, 1251. (20) Pasaogullari, U.; Wang, C. Y. J. Electrochem. Soc. 2005, 152, A380. (21) Meng, H.; Wang, C. Y. J. Electrochem. Soc. 2005, 152, A1733. (22) Siegel, N. P.; Ellis, M. W.; Nelson, D. J.; von Spakovsky, M. R. J. Power Sources 2004, 128, 173. (23) Lin, G.; He, W.; Nguyen, T. V. J. Electrochem. Soc. 2004, 151, A1999. (24) Shimpalee, S.; Beuscher, U.; Van Zee, J. W. Electrochim. Acta 2007, 52, 6748. (25) Kong, C. S.; Kim, D. Y.; Lee, H. K.; Shul, Y. G.; Lee, T. H. J. Power Sources 2002, 108, 185. (26) Wilson, M. S.; Valerio, J. A.; Gottesfeld, S. Electrochim. Acta 1995, 40, 355. (27) Eikerling, M.; Komyshev, A. A. J. Electroanal. Chem. 1998, 453, 89. (28) Kimball, E.; Whitaker, T.; Kevrekidis, Y. G.; Benziger, J. B AIChE J. 2008, 54, 1313. (29) Natarajan, D.; Nguyen, T. V AIChE J. 2005, 51, 2587. (30) Yi, J. S.; Yang, J. D.; King, C AIChE J. 2004, 50, 2594. (31) Wang, L. B.; Wakayama, N. I.; Okada, T. Electrochem. Commun. 2002, 4, 584. (32) Pasaogullari, U.; Wang, C.-Y.; Chen, K. S. J. Electrochem. Soc. 2005, 152, A1574. (33) Weber, A. Z.; Newman, J. J. Electrochem. Soc. 2005, 152, A667. (34) Wei, Z. D.; Ji, M. B.; Hong, Y.; Sun, C. X.; Chan, S. H.; Shen, P. K. J. Power Sources 2006, 160, 246.

J. Phys. Chem. C, Vol. 113, No. 2, 2009 771 (35) Wei, Z. D.; Yan, C.; Tan, Y.; Li, L.; Sun, C. X.; Shao, Z. G.; Shen, P. K.; Dong, H. W. J. Phys. Chem. C 2008, 112, 2671. (36) Boysen, D. A.; Uda, T.; Chisholm, C. R. I.; Haile, S. M. Science 2004, 303, 78. (37) Xing, S. M.; Wang, Y. L. Synthetic Technics and Product Application of Organic Silicon; Chemical Industry Press: Beijing, 2000. (38) Bean, J. A. Lange’s Handbook of Chemistry, 15th edition; McGrawHill: New York, 2001. (39) Li, R. H. Elements of Mass Transfer; Beijing Aviation College Press: Beijing, 1987. (40) Giner, J.; Hunter, C. J. Electrochem. Soc. 1969, 116, 1124. (41) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; Wiley & Sons: New York, 1980. (42) Roche, I.; Chaıˆnet, E.; Chatenet, M.; Vondra´k, J. J. Phys. Chem. C 2007, 111, 1434. (43) Murthi, V. S.; Urian, R. C.; Mukerjee, S. J. Phys. Chem. B 2004, 108, 11011. (44) Ciureanu, M.; Roberge, R. J. Phys. Chem. B 2001, 105, 3531. (45) Raistrick, I. D. Electrochim. Acta 1990, 25, 1579. (46) Varcoe, J. R.; Slade, R. C. T.; Wright, G. L.; Chen, Y. L. J. Phys. Chem. B 2006, 110, 21041. (47) Arai, H.; Mu¨ller, S.; Haas, O. J. Electrochem. Soc. 2000, 147, 3584. (48) Alcaide, F.; Brillas, E.; Cabot, P. L. J. Electroanal. Chem. 2003, 547, 61. (49) Pyun, Su-II; Ryu, Y. G. J. Power Sources 1996, 62, 1. (50) Wu, R. H.; Sun, M. J. Surface and Interface of High Polymer; Science Press: Beijing, 1998.

JP807773M