4
Supportless PdFe nanorods as highly active electrocatalyst for proton exchange membrane fuel cell Wenzhen Li a, * , Pradeep Haldar b a Department of Chemical Engineering, Michigan Technological University, Houghton, MI 49931, USA b College of Nanoscale Science and Engineering (CNSE), State University of New York (SUNY) at Albany, NY 12206, USA article info Article history: Received 11 March 2009 Received in revised form 29 March 2009 Accepted 31 March 2009 Available online 9 April 2009 Keywords: Nanorod Oxygen reduction Catalyst Palladium Fuel cells abstract The PdFe nanorods (PdFe-NRs) with tunable length were synthesized by an organic phase reaction of [Pd(acac) 2 ] and thermal decomposition of [Fe(CO) 5 ] in a mixture of oleyamine and octadecene at 160 °C. They show a better proton exchange membrane fuel cell (PEMFC) performance than commercial Pt/C in working voltage region of 0.80–0.65 V, due to their high intrinsic activity to oxygen reduction reaction (ORR), reduced cell inner resistance, and improved mass transport. Ó 2009 Elsevier B.V. All rights reserved. 1. Introduction Highly active and low cost electrocatalysts are of great impor- tance for large-scale commercial applications of PEMFCs [1,2]. One-dimensional (1-D) carbon nanotubes (CNTs) supported elect- rocatalysts have shown improved activity towards oxygen reduc- tion reaction (ORR), durability and mass transport [3–6], however, still suffer the carbon corrosion issue [7]. Supportless 1-D metallic and multi-metallic electrocatalysts have recently at- tracted enormous attention [8–10], since they can offer unique benefits, such as (1) anisotropic morphology and (2) thin and dense metal catalyst layer, which lead to higher mass transport of reac- tants, (3) high aspect ratio, which is immune to surface energy dri- ven coalescence via crystal migration. Pt and PtPd nanotubes with a diameter of 50 nm and length of 5–20 lm were reported to have improved ORR specific activities [9]. However, their relatively small electrochemical surface areas (ECSA) (<15 m 2 /g), due to its large wall thickness of 4–7 nm and inaccessible inner tube surface area, greatly limit the mass activity enhancement. In order to meet the fuel cell cost requirements, the research on decreasing Pt con- tent in electrodes and replacing Pt by other inexpensive metals, such as Pd, Fe, Co, etc. were extensively studied [11–15]. Pd is con- sidered less expensive than Pt (the price of Pd is 1/4–1/5 as that of Pt), but less active for the ORR. In an acid electrolyte the ORR ex- change current density of Pd is 10 10 A/cm 2 , which is one magni- tude lower than that of Pt (10 9 A/cm 2 ) [16]. Carbon supported Pd–M alloy nanoparticles [12,13], such as PdFe, PdCo and PdCoAu, have been demonstrated a comparable ORR activity as Pt. In this paper, we present a convenient synthesis method for preparing PdFe nanorods with a large ECSA of ca. 60 m 2 /g, and they show high performance in both three-electrode-cell and PEM single cell tests. 2. Experimental The synthesis procedures of PdFe-NR are described as follows: a mixture of 153 mg Pd(acac) 2 (0.5 mmol) and 20 mL oleylamine (OAm) under a blanket of nitrogen was quickly heated to 110 °C and held for 30 min. After 120 lL Fe(CO) 5 was injected into the synthesis system, the temperature was raised to 160 °C and held for 30 min. The solution was then cooled down to room tempera- ture by removing the mantle heater. A mixture of 10 mL hexane and 50 mL ethanol was added and the product was separated by centrifugation (6000 rpm for 10 min). The product was cleaned by redispersing in a mixture of 10 mL hexane and 15 mL ethanol and separating by centrifugation for three times. The final PdFe- NR product was dispersed in 10 mL hexane. 20 ml mixtures of OAm and octadecene (ODE) with volumetric ratio of 1:1 and 3:1 were used for synthesis of PdFe-NR(b) and PdFe nanoparticles (PdFe-NP), respectively. Pd nanoparticles (Pd-NP) with a diameter of 2–3 nm were synthesized by reduction of Pd(acac) 2 in the pres- ence of OAm and oleic acid at 200 °C for 30 min. These Pd-based NR and NP powders were obtained by filtrating their hexane solutions 1388-2481/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.elecom.2009.03.046 * Corresponding author. Tel.: +1 906 487 2298. E-mail address: [email protected] (W. Li). Electrochemistry Communications 11 (2009) 1195–1198 Contents lists available at ScienceDirect Electrochemistry Communications journal homepage: www.elsevier.com/locate/elecom

Supportless PdFe nanorods as highly active …djchadde.public.iastate.edu/pdf/10.pdfSupportless PdFe nanorods as highly active electrocatalyst for proton exchange membrane fuel cell

  • Upload
    lamdan

  • View
    225

  • Download
    6

Embed Size (px)

Citation preview

Page 1: Supportless PdFe nanorods as highly active …djchadde.public.iastate.edu/pdf/10.pdfSupportless PdFe nanorods as highly active electrocatalyst for proton exchange membrane fuel cell

Electrochemistry Communications 11 (2009) 1195–1198

Contents lists available at ScienceDirect

Electrochemistry Communications

journal homepage: www.elsevier .com/ locate /e lecom

Supportless PdFe nanorods as highly active electrocatalystfor proton exchange membrane fuel cell

Wenzhen Li a,*, Pradeep Haldar b

a Department of Chemical Engineering, Michigan Technological University, Houghton, MI 49931, USAb College of Nanoscale Science and Engineering (CNSE), State University of New York (SUNY) at Albany, NY 12206, USA

a r t i c l e i n f o

Article history:Received 11 March 2009Received in revised form 29 March 2009Accepted 31 March 2009Available online 9 April 2009

Keywords:NanorodOxygen reductionCatalystPalladiumFuel cells

1388-2481/$ - see front matter � 2009 Elsevier B.V. Adoi:10.1016/j.elecom.2009.03.046

* Corresponding author. Tel.: +1 906 487 2298.E-mail address: [email protected] (W. Li).

a b s t r a c t

The PdFe nanorods (PdFe-NRs) with tunable length were synthesized by an organic phase reaction of[Pd(acac)2] and thermal decomposition of [Fe(CO)5] in a mixture of oleyamine and octadecene at160 �C. They show a better proton exchange membrane fuel cell (PEMFC) performance than commercialPt/C in working voltage region of 0.80–0.65 V, due to their high intrinsic activity to oxygen reductionreaction (ORR), reduced cell inner resistance, and improved mass transport.

� 2009 Elsevier B.V. All rights reserved.

1. Introduction

Highly active and low cost electrocatalysts are of great impor-tance for large-scale commercial applications of PEMFCs [1,2].One-dimensional (1-D) carbon nanotubes (CNTs) supported elect-rocatalysts have shown improved activity towards oxygen reduc-tion reaction (ORR), durability and mass transport [3–6],however, still suffer the carbon corrosion issue [7]. Supportless1-D metallic and multi-metallic electrocatalysts have recently at-tracted enormous attention [8–10], since they can offer uniquebenefits, such as (1) anisotropic morphology and (2) thin and densemetal catalyst layer, which lead to higher mass transport of reac-tants, (3) high aspect ratio, which is immune to surface energy dri-ven coalescence via crystal migration. Pt and PtPd nanotubes witha diameter of 50 nm and length of 5–20 lm were reported to haveimproved ORR specific activities [9]. However, their relativelysmall electrochemical surface areas (ECSA) (<15 m2/g), due to itslarge wall thickness of 4–7 nm and inaccessible inner tube surfacearea, greatly limit the mass activity enhancement. In order to meetthe fuel cell cost requirements, the research on decreasing Pt con-tent in electrodes and replacing Pt by other inexpensive metals,such as Pd, Fe, Co, etc. were extensively studied [11–15]. Pd is con-sidered less expensive than Pt (the price of Pd is 1/4–1/5 as that ofPt), but less active for the ORR. In an acid electrolyte the ORR ex-change current density of Pd is 10�10 A/cm2, which is one magni-

ll rights reserved.

tude lower than that of Pt (10�9 A/cm2) [16]. Carbon supportedPd–M alloy nanoparticles [12,13], such as PdFe, PdCo and PdCoAu,have been demonstrated a comparable ORR activity as Pt. In thispaper, we present a convenient synthesis method for preparingPdFe nanorods with a large ECSA of ca. 60 m2/g, and they showhigh performance in both three-electrode-cell and PEM single celltests.

2. Experimental

The synthesis procedures of PdFe-NR are described as follows: amixture of 153 mg Pd(acac)2 (0.5 mmol) and 20 mL oleylamine(OAm) under a blanket of nitrogen was quickly heated to 110 �Cand held for 30 min. After 120 lL Fe(CO)5 was injected into thesynthesis system, the temperature was raised to 160 �C and heldfor 30 min. The solution was then cooled down to room tempera-ture by removing the mantle heater. A mixture of 10 mL hexaneand 50 mL ethanol was added and the product was separated bycentrifugation (6000 rpm for 10 min). The product was cleanedby redispersing in a mixture of 10 mL hexane and 15 mL ethanoland separating by centrifugation for three times. The final PdFe-NR product was dispersed in 10 mL hexane. 20 ml mixtures ofOAm and octadecene (ODE) with volumetric ratio of 1:1 and 3:1were used for synthesis of PdFe-NR(b) and PdFe nanoparticles(PdFe-NP), respectively. Pd nanoparticles (Pd-NP) with a diameterof 2–3 nm were synthesized by reduction of Pd(acac)2 in the pres-ence of OAm and oleic acid at 200 �C for 30 min. These Pd-based NRand NP powders were obtained by filtrating their hexane solutions

Page 2: Supportless PdFe nanorods as highly active …djchadde.public.iastate.edu/pdf/10.pdfSupportless PdFe nanorods as highly active electrocatalyst for proton exchange membrane fuel cell

1196 W. Li, P. Haldar / Electrochemistry Communications 11 (2009) 1195–1198

through a Nylon filter (pore size: 0.2 lm) and being dried at 60 �Cin a N2-saturated oven.

The composition, morphology and structure of thesePd-based nanocatalysts were analyzed by transmission elec-

160

180 e

Pd (1

11)

40

60

80

100

120

140

Pd (2

00)

P

Pd

PdF

Inte

nsity

/ a.u.

30 35 40 45 50 55

0

20PdF

2 theta deg

Fig. 1. TEM images of PdFe-NP (a) and NR with a length of 10 nm (b), 50 nm (c)

tron microscopy (TEM, JOEL2010, 200 kV), X-ray diffraction(XRD, SCINTAG-XDS2000, Cu Ka radiation, k = 1.5418 Å) andInduced couple plasma–atomic emission spectroscopy(ICP–AES).

2.9 nm

Pd (2

20)

-NP

NR

60 65 70 75 80

2.7 nme-NR

ree / o

, and 50 nm in one bundle (d), and XRD patterns of Pd-NP and PdFe-NR (e).

Page 3: Supportless PdFe nanorods as highly active …djchadde.public.iastate.edu/pdf/10.pdfSupportless PdFe nanorods as highly active electrocatalyst for proton exchange membrane fuel cell

-1

0

PdFe-NR Pt/C

a

W. Li, P. Haldar / Electrochemistry Communications 11 (2009) 1195–1198 1197

A conventional three-electrode-cell setup (electrolyte cell AF-CELL3, Pine) with a glassy carbon working electrode, a reversiblehydrogen reference electrode (Hydroflex�) and a Pt wire counterelectrode, was used for cyclic voltammetry (CV) and rotating diskelectrode (RDE) tests of the PdFe-NR and Pt/C (20 wt%, BASF-Fuel-Cell) samples at 65 �C. The working electrode was prepared bydropping 10 lL of the ink on the glassy carbon electrode. The inkwas prepared by ultrasonically mixing 3 mg of PdFe-NR in 5 mLethanol. 10 lL of 0.05 wt% Nafion was added on top to fix the elect-rocatalyst. For the Pt/C, 20 lL of the ink with 1 mg Pt/C/ml ethanolwas dropped on the surface of glassy carbon and covered by 10 lLof 0.05 wt% Nafion. The CV test was conducted at 50 mV/s from0.05 V to 1.0 V (vs. RHE). The ORR polarization curve was obtainedat 10 mV/s from 0.35 V to 0.95 V (vs. RHE), using a rotating speed of1600 rpm under oxygen (99.999%) bubbling. The working electro-lyte is 0.5 M H2SO4 and 0.1 M HClO4 for CV and ORR tests,respectively.

The MEA fabrication is based on a self-developed screen-print-ing technique. Briefly, the screen printing paste was prepared byadding PdFe-NR, solubilized Nafion, cyclohexanol and ethyleneglycol to a small vial, and then was coated directly onto K+-formNafion membrane (50 lm, DuPont, NRE212) to make cathode cat-alyst coated membrane (CCM). The thickness of PdFe-NR catalystlayer was determined by a digital micrometer, five points of theCCM was randomly measured and the average thickness was re-ported. Pt/C was screen-printed on the other side of the CCM as an-ode catalyst. After the K+-form CCM was converted to H+-form, itwas assembled into a 5-cm2-PEMFC-hardware. A commercialMEA (BASF-FuelCell) with 0.5 mgPt/cm2 for both anode and cath-ode was also tested as a controlled sample. The cell temperaturewas kept at 65 �C and the relative humidity (RH) of both anodeand cathode are 100%. After the MEA was activated at 0.7 V for8 h, its PEMFC performance was tested by applying a constant volt-age and collecting the corresponding current density and in situ in-ner resistance (based on current interruption technique).

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0-7

-6

-5

-4

-3

-20.1 M HClO4, 99.999% O2

10 mV/s, 65 oC, 1600 rpm, anodic scan 25 µgPd/cm2 and 20 µgPt/cm2

i [m

A/c

m2]

E / V [RHE]

0.0

0.1

0.2

0.3

0.4

0.5

0

50

100

150

200

250

300

Specific Activity / mA.cm

-2

Mas

s Ac

tivity

/ m

A.m

g-1 SAMAPdFe-NRPt/C

0.85 V

0.9 V

b

ig. 2. (a) Polarization curves of the ORR on PdFe-NR and Pt/C (BASF-FuelCell), and) mass activity and specific activity of PdFe-NR and Pt/C in three-electrode-cell-st.

3. Results and discussion

TEM images of PdFe NP and NR are shown in Fig. 1a–d. Using anOAm and ODE ratio of 1/3, uniformly dispersed PdFe nanoparticleswith an average particle size of 2–4 nm are observed (Fig. 1a). Withthe ratio of OAm to ODE increasing to 1:1, PdFe-NR(b) with a diam-eter of 3 nm and short length of 10 nm were synthesized (Fig. 1b).When pure OAm was used for the synthesis, 50 nm-length PdFe-NRs with a similar diameter of 3 nm were produced (Fig. 1c).Therefore, we demonstrated an easy approach to control the lengthof PdFe-NR by altering the volumetric ratio of OAm to ODE. It hasbeen reported that the length of 1-D PtFe nanowire can be con-trolled by altering the surfactants ratio [17]. In our experiment, itis interesting to find that around 20–40 PdFe nanorods were self-assembled to ‘flower’-like bundles (Fig. 1d), and the distance be-tween two nanorods is around 2–4 nm. This bundle structure islikely to form a thin and dense catalyst layer in a MEA, whichcan facilitate the mass transport of reactants. The composition ofthe PdFe-NR was controlled by injecting different amount ofFe(CO)5. For example, 0.12 mL gave a Pd–Fe ratio of 52:48 (deter-mined by ICP–AES). XRD patterns of PdFe-NR and Pd-NP are shownin Fig. 1e. Broad Pd diffraction peaks are observed for both samples.The particle size based on calculation of Pd (2 2 0) diffraction peakby Debye–Sherrer formula is 2.7 nm for PdFe-NR and 2.9 nm forPd-NP. This confirms the PdFe-NR has a small diameter. It is alsoobserved that all Pd diffraction peaks of PdFe-NR shift to right,compared to Pd-NP. The Pd–Pd bond (inter-atomic) distance is0.2729 nm for PdFe-NR, which is smaller than the bond distanceof Pd-NP (0.2747 nm). It was reported the optimum Pd–Pd bond

distance corresponding to the highest ORR activity is around0.273 nm for PdFe/C catalyst [13]. In our work, the Pd–Pd bond dis-tance of the PdFe-NR is very close to this optimum Pd–Pd bond dis-tance (0.273 nm).

Theoretically, the chemical surface area (CSA) of a metal nano-rod is inversely proportional to its diameter, but is independentwith its length. The Pd-NR with a diameter of 3 nm has a CSA of110 m2/g. The ECSA of PdFe-NR obtained from CV test (based onthe hydrogen desorption peaks) is 60 m2/g, which gives a Pd utili-zation of 55%. The ORR activity for PdFe-NR and Pt/C catalysts mea-sured using RDE technique in the three-electrode-cell is shown inFig. 2. At 0.9 V (RHE), the PdFe-NR exhibits a lower ORR activitythan Pt/C, however, as the voltage is low to 0.85 V, the mass activ-ity (MA) for PdFe-NR is 284 mA/mgPd, which is slightly higher than265 mA/mgPt for Pt/C, this trend is even larger when the voltage isdecreased to 0.8 V: 1290 mA/mgPd for PdFe-NR vs. 805 mA/mgPtfor Pt/C. The ORR curve of PdFe-NR is steeper than that of Pt/C, thisis attributed to thin layer of PdFe-NR (no carbon, dense and thinnercatalyst on GC electrode), which can facilitate oxygen transport toPdFe-NR surface.

A 5-cm2-MEA with the PdFe-NR as cathode catalyst (0.22 mgPt/cm2) was fabricated by a self-developed screen-printing technique.The PdFe-NR based cathode catalyst layer is only around 1 lm. TheH2–air PEMFC performance comparison of this MEA and the com-mercial MEA with Pt/C cathode catalyst of 0.5 mgPt/cm2 is shownin Fig. 3. At a high working voltage of 0.9 V, the mass activity of Pt/C is 36 mA/mgPt, which is two times higher than PdFe-NR (18 mA/mgPd). However, as the working voltage moves to the practicaloperation region (0.8–0.65 V), the cell performance of PdFe-NR is

F(bte

Page 4: Supportless PdFe nanorods as highly active …djchadde.public.iastate.edu/pdf/10.pdfSupportless PdFe nanorods as highly active electrocatalyst for proton exchange membrane fuel cell

0 100 200 300 400 500 600 700 8000.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Anode catalyst: Pt/C (40wt%, BASF-FuelCell), 0.5 mgPt/cm2

Cell: 65oC, 100% RHAnode/cathode: H2/air, 0.1/0.4 slpm, 0/0 psi

Pt/C, 0.50mgPt/cm2

PdFe-NR, 0.22mgPd/cm2

i [mA/cm2 ]

E /

V

Fig. 3. Polarization curves of single PEMFCs (5-cm2-MEA) with PdFe-NR and Pt/C(BASF-FuelCell) cathode catalysts.

1198 W. Li, P. Haldar / Electrochemistry Communications 11 (2009) 1195–1198

higher than that of Pt/C e.g. at 0.75 V, the current density and themass power density of PdFe-NR is 1.65 times higher than Pt/C(245 mA/cm2 vs. 148 mA/cm2, 184 mW/cm2 vs. 111 mW/cm2),even on half precious metal loading (0.50 mgPt vs. 0.22 mgPd).When the working voltage is in the range of 0.8–0.65 V, the polar-ization curve is a combined effect of catalyst activity, inner cellresistance and mass transport. Therefore, the I–V curves showthe PdFe-NR catalyst has a better overall PEMFC performance inthis working voltage region.

We attributed the improved PEMFC performance to: (1) theadvantageous Pd–Pd bond distance (0.2729 nm) for PdFe-NR,which favors oxygen double bond dissociation [13], thus improvesthe ORR intrinsic activity; (2) the unique 1-D nanorod structureand the super thin catalyst layer (�1 lm vs. 25 lm for Pt/C catalystlayer), which can reduce the overall internal resistance of PdFe-NR-based MEA (6 mX vs. 9 mX for Pt/C-based MEA) and improve themass transport. It was reported that a 60–80% mass activity (powerdensity) benefit could be gained if the ohm resistance and masstransport were effectively reduced [2]. Here we clearly demon-strated this benefit through using super-thin 1-D PdFe-NR-basedcatalyst layer. The reported supportless nanocatalyst synthesisand MEA fabrication strategies have a great potential to be ex-tended to other highly active bi-, tri-metallic 1-D electrocatalysts,such as PdCo/Ni-NR for ORR and PtRuCo-NR for methanol oxida-tion reaction. It should be noted that the PdFe-NR-based MEA

shows a performance drop when the operation voltage is lowerthan 0.65 V, this could be attributed to ‘water flooding’ inside thecatalyst layer. Future work focused on optimizing the catalyst layercompositions and the gas diffusion layer structures will be ex-pected to improve its performance in high current density region,and this is currently under study in my group.

4. Conclusions

In this study, uniform PdFe-NR with a diameter of 3 nm andcontrollable length of 10–50 nm were synthesized through a sin-gle-step organic phase reduction process. The ratio of two surfac-tants (OAm and ODE) was found to effectively control the lengthof nanorods in range of 10–50 nm. The PdFe-NR demonstrates abetter PEMFC performance than commercial Pt/C in practicalworking voltage region (0.80–0.65 V), which can be attributed totheir high intrinsic activity towards ORR, reduced cell inner resis-tance, and improved mass transport.

Acknowledgements

We thank Seth Knupp, Dr. Thomas Murray and Dr. WentaoWang for their helps with XRD and TEM analysis and MEAfabrication.

References

[1] W. Vielstich, A. Lamm, H.A. Gaisteiger, Handbook of Fuel Cells – Fundamentals,Technology and Applications, vol. 2, Wiley, NY, 2003.

[2] H.A. Gaisteiger, S.S. Kocha, B. Sompalli, F.K. Wagner, Appl. Catal. B 56 (2005) 9.[3] G.L. Che, B.B. Lakshmi, E.R. Fisher, C.R. Martin, Nature 393 (1998) 346.[4] W.Z. Li, C.H. Liang, W.J. Zhou, J.S. Qiu, Z.H. Zhou, G.Q. Sun, Q. Xin, J. Phys. Chem.

B 107 (2003) 6292.[5] G. Girishkumar, K. Vinodgopal, P.V. Kamat, J. Phys. Chem. B 108 (2004) 19960.[6] W.Z. Li, X. Wang, Z.W. Chen, M. Waje, Y.S. Yan, Langmuir 21 (2005) 9386.[7] X. Wang, W.Z. Li, M. Waje, Y.S. Yan, J. Power Sources 158 (2006) 154.[8] M. Debe, A.K. Schmoechel, G.D. Vernstrom, P. Atanasoki, J. Power Sources 161

(2006) 1002.[9] Z.W. Chen, M. Waje, W.Z. Li, Y.S. Yan, Angew. Chem. Int. Ed. 46 (2007) 4060.

[10] H. Wang, C. Xu, F. Cheng, S. Jiang, Electrochem. Commun. 9 (2007) 1212.[11] O. Savadogo, K. Lee, K. Oishi, S. Mitsushima, N. Kamiya, K.I. Ota, Electrochem.

Commun. 6 (2004) 105.[12] J.L. Fernandez, V. Raghuveer, A. Mathiram, A.J. Bard, J. Am. Chem. Soc. 127

(2005) 13100.[13] M-H. Shao, K. Sasaki, R.R. Adzic, J. Am. Chem. Soc. 128 (2006) 3526.[14] N.A. Vante, H. Tributsch, Nature 323 (1986) 431.[15] R. Bashyam, P. Zelenay, Nature 443 (2006) 63.[16] R.P. O’Hayre, S.-W. Cha, W. Colella, F.B. Prinz, Fuel Cell Fundamentals, Wiley,

NY, 2006.[17] C. Wang, Y. Hou, J. Kim, S. Sun, Angew. Chem. Int. Ed. 46 (2007) 63.