j ourna l ho me page: www.elsev ier .com/ locate /e lec tac ta
ffect of Polytetrafluoroethylene on Ultra-Low Platinum Loadedlectrospun/Electrosprayed Electrodes in Proton Exchange Membraneuel Cells
uhai Wang, Francis W. Richey, Kevin H. Wujcik, Roman Ventura,yle Mattson, Yossef A. Elabd ∗
epartment of Chemical and Biological Engineering, Drexel University, Philadelphia, Pennsylvania 19104, United States
r t i c l e i n f o
rticle history:eceived 2 May 2014eceived in revised form 19 June 2014ccepted 20 June 2014vailable online 11 July 2014
ey words:lectrospinninganofiber
uel cellafionlatinum
a b s t r a c t
In this study, catalyst layers (CLs) were fabricated using a simultaneous electrospinning/electrospraying(E/E) technique to produce unique nanofiber/nanoparticle membrane electrolyte assemblies (E/E MEAs)evidenced by scanning electron microscopy. Specifically, the effect of polytetrafluoroethylene (PTFE) inthese E/E MEAs on polymer electrolyte membrane (PEM) fuel cell performance was evaluated. E/E MEAsresult in high fuel cell performance at ultra-low platinum (Pt) loadings with higher electrochemicalsurface areas as evidenced by cyclic voltammetry experiments. Without PTFE, an E/E MEA operatedat 172 kPa (25 psi) back pressure results in a maximum power density of 1.090 W/cm2 (H2/O2) and0.647 W/cm2 (H2/air) with only 0.112 mgPt/cm2 total Pt MEA loading. Introducing PTFE (at only 1 wt%)to the electrospinning process results in an E/E MEA operated at the same back pressure (172 kPa (25psi)) with an even higher maximum power density of 1.240 W/cm2 (H2/O2) and 0.725 W/cm2 (H2/air) ata lower total Pt MEA loading of 0.094 mgPt/cm2. This corresponds to a significant reduction in Pt loading(16% of control) with only a modest reduction in power density (∼86-87% of control), where the controlMEA was produced using a conventional coating method and resulted in maximum power density of
2 2 2
1.420 W/cm (H2/O2) and 0.839 W/cm (H2/air) at a Pt MEA loading of 0.570 mgPt/cm (172 kPa (25 psi)).An excellent total MEA platinum utilization of 0.076 gPt/kW (∼13.2 kW/gPt) was achieved with the E/EMEA with PTFE at only a 0.094 mgPt/cm2 total Pt MEA loading. The improvement in E/E MEA with PTFEwas a result of increased hydrophobicity of the nanofibers evidenced by contact angle measurementsand improved PEM fuel cell performance at higher limiting current density in the mass transport region.
Proton exchange membrane (PEM) fuel cells are an excellentlternative energy source for both stationary and mobile applica-ions. However, currently the requirement for a significant amountf platinum (Pt), the most active catalyst for PEM fuel cells, lim-ts the mass commercialization of PEM fuel cells [1,2]. In previous
ork in our laboratory [3], an alternative electrode design wasxplored to reduce the amount of required Pt, while still producing
dequate PEM fuel cell power output. A simultaneous electrospin-ing/electrospraying (E/E) technique was employed to producenique nanofiber/nanoparticle electrodes as evidenced by scan-
ning electron microscopy (SEM). Specifically, Nafion nanofibers andPt/C nanoparticles were introduced separately and simultaneouslyby two different needles using electrospinninig and electrospray-ing, respectively, to produce nanofiber/nanoparticle electrodes andsubsequently membrane electrode assemblies (MEAs). In this pre-vious study, only cathodes were produced via E/E process andanodes were produced using a conventional coating technique.The E/E MEAs in this previous study resulted in ultra-low Ptcathode loadings of 0.052 and 0.022 mgPt/cm2, where maximumpower densities (at 172 kPa (25 psi) back pressure) of 1.090 and0.936 W/cm2 (H2/O2) and 0.656 and 0.625 W/cm2 (H2/air) wereachieved at these two Pt loadings, respectively. This was compared
to a conventional control MEA at a 0.42 mgPt/cm2 cathode cata-lyst loading with maximum power densities (at 172 kPa (25 psi)back pressure) of 1.420 and 0.839 W/cm2 for H2/O2 and H2/air,respectively. These results correspond to a significant reduction in
t cathode loading (5-12% of control) at only a modest reductionn power density (∼66-78% of control) for the E/E MEAs. Excellentlatinum utilization in the cathode of 0.024 gPt/kW (∼42 kW/gPt)as achieved for the E/E MEAs at 0.022 mgPt/cm2 cathode loading.
These lower Pt loadings were achieved by increasing the triplehase boundaries (TPBs), which are the junction points whereatalytic and electron conduction sites, reactant gases (pores),nd proton conducting Nafion ionomer meet. Cyclic voltammetryesults confirm these findings, where increased electrochemicalurface areas were observed for the E/E MEAs compared to theontrol, i.e., more accessible Pt in the cathode. This E/E techniqueiffers from electrospinning [4,5] or electrospraying [6–9] alone,here a mixture of Nafion and Pt/C are expelled from the sameeedle. Contrastingly, the E/E technique allows for a higher level ofontrol over fiber size and Pt loading compared to other electrodeabrication techniques.
In this study, we explore the impact of having both cathodend anode catalyst layers (CLs) prepared via the E/E tech-ique and the effect of polytetrafluoroethylene (PTFE) in theseanofiber/nanoparticle CLs. Previously, the effect of PTFE in gas dif-
usion layers (GDLs), micro-porous layers (MPLs), and conventionalLs on fuel cell performance has been investigated [10–19]. Pre-ious work [10,11] has shown that GDLs loaded with 5-30 wt%TFE results in improved fuel performance due to an increasen hydrophobicity, which improves mass transport of the reac-ants from the gas flow channels to the CLs. However, excessiveTFE loading in GDLs results in poor fuel cell performance dueo reduced electrical conductivity and gas permeability [12]. Forome MEAs, a hydrophobic MPL is supported on the GDL sub-trate. Previous results have shown that ∼10-30 wt% of PTFE inhe MPL results in optimal fuel cell performance [13–15] due toncreased pressure on cathode side, which improves water man-gement in the fuel cell [11,13,16,17]. The effect of PTFE in CLn the performance of the fuel cell has been previously investi-ated for conventionally painted electrodes [18,19]. Uchida et al.18] showed that the PTFE had an effect at high current densi-ies, where flooding can limit fuel cell performance. Friedmannnd Nguyen [19] optimized their MEAs by varying the ratios ofafion, Pt/C, and PTFE in a two-step method to prepare the elec-
rodes, where their results showed that a PTFE content of 10-24t% produced the best fuel cell performances. This previous work
10–19] demonstrates that a relatively high PTFE content (>10 wt%)n the GDL, MPL, and CL is required to improve fuel cell perfor-
ance. Subsequently, the fuel cell performance decreases whenxcessive PTFE is added due reductions in pore size and electricalonductivity.
In this study, the effect of including PTFE in the E/E processnd subsequently on the fuel cell performance/Pt loading ratio wasxplored. Both anodes and cathodes in the MEAs were produced via/E process. This is the first study to investigate the effect of PTFEn nanofiber/nanoparticle CLs.
. Experimental
.1. Materials
Isopropanol (99.5%, Sigma-Aldrich), ethanol (99.5%, Deconabs, Inc.), Nafion solution (1000 EW, 5 wt% in a 3/1 v/vf isopropanol/water, Ion Power), poly(acrylic acid) (PAA;V = 450,000 g/mol, Aldrich), 20 wt% Pt on carbon catalyst (Pt/C;ulcan XC-72, Premetek Co.), gas diffusion layer (GDL; SGL-25BC,
uel Cells Etc.), and Nafion NR-212 membrane (1100 EW, ∼50 �m0.002 in) dry thickness, Ion Power) were used as received. 60 wt%olytetrafluoroethylene (PTFE) dispersion in water (Aldrich) wasiluted to 5 wt% with 3/1 v/v of isopropanol/water before use.
cta 139 (2014) 217–224
Ultrapure deionized (DI) water with resistivity ∼16 M� cm wasused as appropriate. Ultra high purity grade N2, H2, O2 and ultrazero grade air were all purchased from Airgas and used for all fuelcell experiments.
2.2. Two-needle electrospinning/electrospraying (E/E) system
A custom-designed E/E apparatus was used and consists oftwo high-voltage power supplies (Model PS/EL50R00.8, GlassmanHigh Voltage, Inc. and Model ES40P-10 W/DAM, Gama High Volt-age Research), two syringe pumps (Model NE-1000, New Era PumpSystems), two syringe needles (i.d. = 0.024 in., Hamilton), tubing (Pt.No. 30600-65, Cole-Parmer), and a grounded collector (aluminumfoil coated cylindrical drum, o.d. = 4.85 cm). The collector drum isconnected to a motor (Model 4IK25GN-SW2, Oriental Motor) toallow for rotation during the E/E process. The rotational speed ofthe collector drum was set to 100 rpm. A gas diffusion layer (GDL)was adhered to the collector drum, where nanofibers/nanoparticlescould be directly collected via the E/E process and catalyst inkis electrosprayed and polymer solution is electrospun simulta-neously. The needle tip to collector distances, applied voltages, andsolution flow rates were 15 and 9 cm, 10.5 and 12.5 kV, and 0.3and 3 ml/h for the electrospinning and electrospraying processes,respectively. 0.25-0.7 ml and 2.5-7 ml of solutions were electrospunand electrosprayed, respectively, resulting in catalyst layers of sev-eral micormeters in thickness on average. More details regardingthe two-needle E/E apparatus are described in Wang et al. [3].
2.3. Membrane electrode assembly (MEA)
Catalyst ink used in the electrospraying portion of the E/E elec-trodes consisted of 20 mg Pt/C catalyst, 0.248 ml DI water, 0.043 mlNafion solution, 0.171 ml isopropanol/water (3/1 v/v), and 1.970 mlethanol. This mixture was sonicated for 3 min (Model CL-18, Qson-ica Sonicator) prior to electrospraying. The polymer solution usedin the electrospinning portion of the E/E electrodes consisted of79.2/19.8/1 wt/wt/wt Nafion/PAA/PTFE. A 5 wt% polymer (Nafion,PAA and PTFE) solution was prepared by combining 131.1 mg PAA,10494 mg of Nafion solution (5 wt%), 131.2 mg of PTFE solution (5wt%), and 2491.7 mg isopropanol/water (3/1 v/v). This solution wasstirred at ∼70–80 ◦C for ∼12 h to ensure complete dissolution. Thesolution was cooled down to ambient temperature before electro-spinning. The catalyst ink and the polymer solution were used tomake E/E electrodes as described in the previous section. After theE/E process, the E/E electrodes were annealed at 135 ◦C for 5 min.
MEAs were fabricated by sandwiching the Nafion NR-212 mem-brane between two catalyst-coated GDLs (anode and cathodecatalyst layers) and hot pressing (heat press, Carver) for 5 min at135 ◦C and 1.5 MPa (213 psi). All anode and cathode catalyst lay-ers in this study were E/E electrodes, unless otherwise specified.Fuel cell performance of E/E anode/cathode MEAs were comparedto MEAs where only cathodes were fabricated using E/E and whereboth anode and cathodes were prepared using a standard hand-painting GDL procedure (control). Details regarding the preparationof these MEAs can be found elsewhere [3].
2.4. Electrode characterization
Morphological characterizations of the E/E electrodes wereinvestigated with scanning electron microscopy (SEM, ModelFEI/Philips XL-30, 10 kV). SEM images of the E/E electrodeswere collected after electrospinning/electrospraying of the
nanofibers/nanoparticles on GDLs, but before MEA fabrication. Allsamples were sputter coated (Denton Desk II Sputtering System)with platinum at 40 mA for 30 s before SEM analysis. The nanofiberand nanoparticle diameters were measured using ImageJ software
mica Acta 139 (2014) 217–224 219
bS
((9eib
a(CdTta
2
fl0TafNcwcbrcciaivpfho
MaaaNms(0
3
tde(Epcar
Fig. 1. (a) SEM image of E/E electrode with 1 wt% PTFE and 0.047 mg/cm2 Pt loading,(b) higher magnification of (a), nanofiber diameter (c) and nanoparticle diameter (d)size distributions, respectively, (e) SEM image of E/E electrode with 1 wt% PTFE and
X. Wang et al. / Electrochi
y counting 30 randomly selected fibers and particles for eachEM image.
The Pt loading was measured with thermal gravimetric analysisTGA; TGA 7, Perkin Elmer). A small portion of the E/E electrode∼5–7 mg) was heated in the TGA from ambient temperature to00 ◦C at 5 ◦C/min in air at 20 ml/min. Since all components in E/Electrode degrade above 900 ◦C with the exception of Pt, the Pt load-ng was determined by comparing the weight of the E/E electrodeefore and after exposure to 900 ◦C in the TGA.
Contact angle measurements of electrodes were measured atmbient temperature. A microsyringe was used to apply a droplet12 �l) of DI water onto the electrodes. A digital camera (Nikonoolpix 7900) was used to capture the droplet image after theroplet was applied to and stable on the E/E electrode surface.he contact angle was the angle measured at the gas-liquid-solidriple point between the liquid-gas interface and the solid surfaces shown on the images [20].
.5. Fuel cell tests and cyclic voltammetry (CV)
Each MEA (1.21 cm2 area) was placed between two serpentineow field graphite plates (1 cm2 flow area) separated by two.160 mm thick Teflon coated gaskets (Pt No. 381-6, Saint Gobian).he entire fuel cell assembly consisted of an MEA, two gaskets,nd two flow plates placed between two copper current collectorsollowed by endplates all held together by tie rods (bolts) with 11.3-m (100 lb-in) of torque. The fuel cell performance (polarizationurves: voltage vs. current density) of each MEA was evaluatedith a Fuel Cell Test Station (850 C, Scribner Associates, Inc.). Fuel
ell tests were conducted at both ambient and 172 kPa (25 psi) ofack pressure with saturated (RH = 100%) anode and cathode flowates of 0.42 L/min hydrogen and 1.0 L/min air, respectively. Theathode, anode, and cell were all maintained at 80 ◦C. Polarizationurves were collected from open circuit voltage (OCV) to 0.2 V atncrements of 0.05 V/min. The fuel cell performance was recordedfter a new MEA was fully activated. The activation processncluded operating an MEA at 0.7 V for ∼1-2 hours followed byoltage scanning from OCV to 0.2 V several times. This activationrocess was repeated until the MEA reached steady state and nourther increase in current was observed when the fuel cell waseld at constant voltage. The activation process typically occursver 4-6 h before the MEA reached a steady state.
Cyclic voltammetry (CV) was performed on a two-electrodeEA with a potentiostat (Solartron SI 1287, Corrware Software)
t 20 mV/s from 0 to 0.9 V. In this configuration, the anode servess both the counter and reference electrodes. The fuel cell anodend cathode were supplied with 0.040 L/min H2 and 0.018 L/min2, respectively. Temperatures of the cathode, anode and cell wereaintained at 30 ◦C. The Pt catalyst was assumed to have an average
ite density of 210 �C/cm2 [21]. The electrochemical surface areaECSA) was determined from the hydrogen adsorption area from.1 to 0.4 V of the CV data.
. Results and Discussion
Fig. 1 shows SEM images of E/E electrodes prior to MEA fabrica-ion along with nanofiber diameter and nanoparticle diameter sizeistributions. Specifically, Fig. 1 (a) shows an SEM image of an E/Electrode with 1 wt% PTFE and 0.047 mg/cm2 Pt loading and Fig. 1b) shows a magnified view of Fig. 1 (a). Fig. 1 (a) shows that the/E electrode is highly porous, which should improve gas trans-
ort throughout the electrode from the gas flow channel to theatalyst layer. The continuously connected network of nanofibersnd nanoparticles should increase the triple phase boundary, whichequires intimate junction points for combined ORR, proton and
0.013 mg/cm2 Pt loading, (f) higher magnification of (e), nanofiber diameter (g) andnanoparticle diameter (h) size distributions, respectively, (i) SEM image of E/E elec-trode with 1 wt% PTFE and 0.003 mg/cm2 Pt loading, (j) higher magnification of (i),nanofiber diameter (k) and nanoparticle diameter (l) size distributions, respectively.
220 X. Wang et al. / Electrochimica Acta 139 (2014) 217–224
F (1 wt%( /air a
ew1afitw
caattlsaFifwE
tta#cctrpe
ig. 2. Fuel cell performances of the E/E #1 (no PTFE), E/E #3 (no PTFE) and E/E #4
b) H2/O2 at ambient pressure, (c) H2/air with 172 kPa (25 psi) back pressure, (d) H2
lectron transport. Several large catalyst agglomerates (1-2 �m)ere observed in Fig. 1 (b), however, the majority were in the
00-300 nm size range as evidenced in Fig. 1 (d). Fig. 1 (c) shows broad nanofiber diameter size distribution with the majority ofbers (87%) at 100-200 nm in diameter. The nanoparticle diame-er size distribution is broader than the nanofiber size distributionith ∼30% of the nanoparticles with diameters over 800 nm.
Fig. 1 (e) shows an SEM image of an E/E electrode with a lower Ptathode loading (0.013 mg/cm2 and 1wt% PTFE) and Fig. 1 (f) shows
magnified view of Fig. 1 (e). Compared to Fig. 1 (c), Fig. 1 (g) shows broader diameter size distribution with 87% of the nanofibers inhe range of 100-300 nm. 23% of the nanoparticles in Fig. 1 (h) havehe diameters greater than 500 nm, which was attributed to theack of large agglomerates in the E/E electrodes. Figs. 1 (i) to (l)how SEM images and nanofiber/nanoparticle size distributions ofn E/E electrode with a Pt loading of 0.003 mg/cm2 and 1 wt% PTFE.igs. 1 (i) and 1 (j) show more porous and randomly oriented fibersn the E/E electrode compared to Figs. 1 (a) and 1 (b). After MEAabrication, SEM images reveal more particle-particle contacts [3],hich should result in effective electron transfer throughout the
/E electrode.Fig. 2 shows the effect of PTFE in MEAs fabricated with E/E elec-
rodes (i.e., E/E MEAs) on fuel cell performances. Fig. 2 compareshe fuel cell performance of E/E MEAs with 1 wt% PTFE (E/E #4)nd 0 wt% PTFE (E/E #3 and E/E #1). The difference between E/E3 and E/E #1 is that the E/E process was used for both anode andathode for E/E #3, while the E/E process was only used on theathode for E/E #1. The anode for E/E #1 was prepared by conven-
ional hand-painting technique described previously [3]. Detailsegarding all MEAs in this study are listed in Table 1. The fuel cellerformance for these three MEAs are compared at four differ-nt operating conditions: H2/O2 at 172 kPa back pressure (273 kPa
PTFE) with operating conditions of (a) H2/O2 with 172 kPa (25 psi) back pressure,t ambient pressure.
absolute pressure) on anode and cathode sides, H2/O2 at ambientpressure, H2/air at 172 kPa back pressure (273 kPa absolute pres-sure) on anode and cathode sides, and H2/air at ambient pressure(Figs. 2(a) to (d), respectively). For all four fuel cell operation con-ditions, E/E #4 (with 1 wt% PTFE) results in a higher peak powerand higher limiting current density in the mass transport region.Note that the E/E MEAs with no PTFE (E/E #3 and E/E #1) havesimilar fuel cell performances, however, the total MEA Pt loadingin E/E #3 is approximately half compared to E/E #1. The cathodePt loadings are similar between these two MEAs, where both werefabricated via the E/E process. However, the anode Pt loading inE/E #1 (0.150 mg/cm2) is higher than the anode Pt loading in E/E#3 (0.056 mg/cm2), where a conventional hand-painting techniquewas used for E/E #1 and the E/E process was used for E/E #3. Thissuggests that Pt loading on the anode side is not a limiting factor andthe total MEA Pt loading can be reduced by using the E/E processon both anode and cathode sides of the MEA.
It is well documented that the addition of PTFE to the cata-lyst layer improves fuel cell performance due to improved liquidwater removal resulting in a faster transport path of reactants inthe catalyst layer, which can also be evidenced by higher oxy-gen consumption in the cathode half cell reaction. When floodingis more prevalent, oxygen consumption is reduced, which is alsoreferred to as oxygen gain. In this study, when the fuel cells wereoperated with 172 kPa back pressure (273 kPa absolute pressure),the oxygen gains at 1.5 A/cm2 for E/E #1, #3, and #4 were 0.16,0.18, and 0.14 V, respectively (Fig. 2 (a) and (c)). The lower oxy-gen gain or higher oxygen consumption of E/E #4 shows that the
catalyst layer with PTFE provides a more effective water removalrate from the catalyst layer and consequently a faster transportpath of reactants to the catalyst layer is achieved. When the backpressure was removed, the effect of PTFE on the oxygen gain was
X. Wang et al. / Electrochimica Acta 139 (2014) 217–224 221
E/E el
nTip
adpr1#obatd
Fop
Fig. 3. (left) Hand-painted electrode without PTFE, (middle)
ot as apparent as the case with back pressure (Fig. 2 (b) and (d)).his is because at lower limiting current densities water floodings not as dominant, which mitigates the PTFE effect on the fuel cellerformance.
Table 2 summarizes the differences in fuel cell performancemong the different MEAs at the first fuel cell operating con-ition of Fig. 2 (H2/O2 fuel cell performance at 172 kPa backressure (273 kPa absolute pressure) on anode and cathode sides,espectively). Maximum power densities of 1.240, 1.090, and.090 W/cm2 were measured at total MEA Pt loadings of 0.094 (E/E4), 0.112 (E/E #3), and 0.202 (E/E #1) mg/cm2, respectively. Inther words, higher performance at lower loadings can be achievedy using the E/E process on both anode and cathode catalyst layers
nd including 1 wt% PTFE to the electrospinning polymer. In rela-ion to the control MEA with no E/E process, the maximum powerensity of 1.420 W/cm2 was measured at a total MEA Pt loading of
ig. 4. Fuel cell performances of the E/E MEAs (all with 1 wt% PTFE) at various total MEA Pperating conditions of (a) H2/O2 with 172 kPa (25 psi) back pressure, (b) H2/O2 at ambiressure.
ectrode without PTFE, (right) E/E electrode with 1 wt% PTFE.
0.570 mg/cm2. This corresponds to a 87% maximum power outputat only 16% total MEA Pt loading for E/E #4 compared to the control,i.e., a platinum utilization of 0.076 gPt/kW for E/E #4 compared to0.401 gPt/kW for the control.
The improved fuel cell performance of E/E #4 with the inclusionof 1 wt% PTFE compared to E/E #3 and E/E #1 (both with no PTFE),where all three MEAs have similar cathode Pt loadings (0.047, 0.056,and 0.052 mg/cm2, for E/E #4, E/E #3, and E/E #1, respectively),suggests that an increase in hydrophobicity of the nanofibers in E/E#4 enhances mass transfer and subsequently fuel cell performance.The improvement is associated to the more hydrophobic surface ofthe pores in catalyst layer. This hydrophobic surface is of criticalimportance to achieve improved water management, particularly,
in the cathode side of the fuel cell by effectively removing liquidwater to provide a faster transport path for reactants to the reactivesites.
t loadings (0.094, 0.026, 0.006 mg/cm2 for E/E #4, E/E #5, E/E #6, respectively)withent pressure, (c) H2/air with 172 kPa (25 psi) back pressure, (d) H2/air at ambient
222
X.
Wang
et al.
/ Electrochim
ica A
cta 139
(2014) 217–224
Table 1E/E MEAs.
MEA Type Cathode method Anode method PTFE content in bothanode and cathode
a Fuel cell operating conditions: H2/O2/bp; bp = back pressure of 172 kPa (25 psi); cathode/anode/cell: 80/80/80 ◦C; 100% RH; PEM = Nafion 212. b2015 DOE Target = 0.125 g/kW [22].
X. Wang et al. / Electrochimica Acta 139 (2014) 217–224 223
F utiliza1
dwattsebum
tc#cisa1wctTipr
altoscuwnt
(amoEp
ig. 5. Maximum fuel cell power output versus (a) total MEA Pt loading and (b) Pt72 kPa (25 psi) back pressure on both anode and cathode sides.
In order to test this hypothesis, the contact angles for threeifferent electrodes (hand-painted control electrode, E/E electrodeithout PTFE, and E/E electrode with 1 wt% PTFE) were measured
nd are shown in Fig. 3. When 1 wt% PTFE is added to the electrode,he measured contact angle, Fig. 3 (c), is much higher than the elec-rodes without PTFE. These results suggest that the addition of amall amount of PTFE can effectively change the wettability of thelectrode fabricated via the E/E process. The increased hydropho-icity in the pores of catalyst layer is advantageous in removing liq-id water and improving gas transport evidenced by the improve-ent in the mass transfer region of the polarization curve (Fig. 2).Fig. 4 shows the effect of Pt loading on fuel cell performance for
hree different E/E MEAs (all with 1 wt% PTFE) at four different fuelell operating conditions. The total MEA Pt loadings of E/E #4, E/E5, and E/E #6 are 0.094, 0.026, 0.006 mg/cm2, respectively. In allases, the fuel cell performance increased with increasing Pt load-ng. Fuel cell results for the first operating condition for all MEAs areummarized in Table 2. In comparison to the control MEA (no E/E),
87%, 42%, and 30% maximum power output at only 16%, 4%, and% total MEA Pt loading for E/E #4, E/E #5, and E/E #6, respectively,ere observed at the first fuel cell operating condition. The signifi-
ant reduction in Pt loading from E/E #4 to E/E #6 does correspondo a decrease in Pt utilization of 0.076, 0.040, and 0.015 gPt/kW.herefore, Fig. 4 does not necessarily suggest that higher Pt load-ng is always desired for all cases, where the trade-off betweeneak power and Pt utilization will vary based on application/costatio.
The correlation between peak power and total MEA Pt loadingnd Pt utilization is shown in Fig. 5. For both graphs, an MEA thaties in the upper-left hand corner is the desired trade-off. It is clearhat E/E #4 is the most attractive MEA in terms of balancing powerutput and Pt loading or Pt utilization. A higher power output cantill be achieved with the control MEA, but much lower Pt loadingsan be achieved without a significant reduction in power outputsing both the E/E process with 1 wt% PTFE. Higher contents of PTFEere also explored with the E/E process, however, reproducibleanofibers were not attainable at PTFE contents above 1 wt% dueo electrospinning instability.
Table 2 also lists the measured electrochemical surface areaECSA) for each MEA in this study. The ECSAs of all E/E MEAs arell over 80 m2/gPt, which are higher than the control MEA (53.2
2/gPt). This provides a rationale for the improved Pt utilizationbserved in E/E MEAs compared to the control MEA. Overall, the/E MEAs show that only 1-36% of the Pt can deliver 30-77% theeak power compared to the control MEA.
[
[
tion for all E/E MEAs and control MEA. Fuel cell operating conditions: H2/O2 with
4. Conclusions
The effect of PTFE in nanofiber/nanoparticle electrodes fabri-cated using a simultaneous electrospinning/electrospraying (E/E)technique on PEM fuel cell performance was explored in this study.An improved fuel cell performance was observed for the E/E MEAwith PTFE compared to the E/E MEA without PTFE when the Ptloading was similar in both MEAs. Specifically, an improvementin the higher limiting current density of the fuel cell polarizationcurve (mass transfer region) was observed in E/E MEA with PTFEdue to an increased hydrophobicity of the nanofibers evidencedby contact angle measurements. As mentioned previously, resultswith conventional electrodes requires higher PTFE contents (>10wt%) to achieve improved fuel cell performance, whereas thenanofiber/nanoparticle electrodes in this study only require 1 wt%PTFE. Understanding the power/loading relationships in E/E fuelcell electrodes not only as function of morphology (nanofiber sizeand distribution, nanoparticle size and distribution, porosity, con-nectivity), but also as a function of materials chemistry (PTFE, purityof Nafion, differences in carbon supports and chemistries, otherPt-based catalysts) will be of interest for future exploration.
Acknowledgements
This work is supported in part by the Energy CommercializationInstitute under grant no. DUETRF-5.
12] D. Bevers, R. Rogers, M. von Bradke, Journal of Power Sources 63 (1996) 193.13] X.L. Wang, H.M. Zhang, J.L. Zhang, H.F. Xu, Z.Q. Tian, J. Chen, H.X. Zhong, Y.M.
Liang, B.L. Yi, Electrochimica Acta 51 (2006) 4909.14] S. Park, J.W. Lee, B.N. Popov, Journal of Power Sources 177 (2008) 457.15] G. Velayutham, J. Kaushik, N. Rajalakshimi, K.S. Dhathathreyan, Fuel Cells 7
(2007) 314.
16] X. Wang, T.V. Nguyen, Journal of The Electrochemical Society 157 (2010) B496.17] A.Z. Weber, J. Newman, Journal of The Electrochemical Society 152 (2005)
A677.18] M. Uchida, Y. Aoyama, N. Eda, A. Ohta, Journal of The Electrochemical Society
142 (1995) 4143.
[
cta 139 (2014) 217–224
19] R. Friedmann, T.V. Nguyen, Journal of The Electrochemical Society 157 (2010)B260.
20] C.-Y. Wang, Handbook of Fuel Cells-Fundamentals, in: Wolf Vielstich, A. Hubert,Gasteiger, Arnold Lamm (Eds.), Technology and Applications, 3, 2003, p. 337,Chapter 29.
21] H.A. Gasteiger, W. Gu, R. Makharia, M.F. Mathias, B. Sompalli, Handbook of
Fuel Cells - Fundamentals, in: W. Vielstich, H.A. Gasteiger, A. Lamm (Eds.),Technology and Applications, 3, 2003, p. 593, Chapter 46.
22] The US Department of Energy (DOE), Energy Efficiency and Renewable Energy,http://www.eere.energy.gov/hydrogenandfuelcells/mypp/pdfs/fuel cells.pdf,(2011).
