Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Oxidation-resistant catalyst supports for proton exchange membrane fuel cells (PEMFCs) 2006

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2006-0402.pdf
ubc_2006-0402.pdf [ 9.58MB ]
Metadata
JSON: 1.0078962.json
JSON-LD: 1.0078962+ld.json
RDF/XML (Pretty): 1.0078962.xml
RDF/JSON: 1.0078962+rdf.json
Turtle: 1.0078962+rdf-turtle.txt
N-Triples: 1.0078962+rdf-ntriples.txt
Citation
1.0078962.ris

Full Text

OXIDATION-RESISTANT C A T A L Y S T S U P P O R T S F O R P R O T O N E X C H A N G E M E M B R A N E F U E L C E L L S (PEMFCs) by H A R M E E T CHHINA B . S c , Univers i ty of V ic tor ia , 2001 A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Mater ia ls Eng ineer ing) T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A A u g u s t 2006 © Harmee t C h h i n a , 2 0 0 6 A B S T R A C T Pro ton e x c h a n g e m e m b r a n e fuel ce l ls ( P E M F C s ) a re e lec t rochemica l energy conve rs ion d e v i c e s that react hydrogen and oxygen to p roduce electr icity. P E M F C s c a n be u s e d for power generat ion in the portable, s tat ionary and transportat ion sec to rs . S e v e r e per fo rmance degradat ion dur ing ex tended operat ion is h inder ing commerc ia l i za t i on of P E M F C s . O n e of the m e c h a n i s m s caus ing per fo rmance degradat ion inc ludes catalyst suppor t co r ros ion . Tungs ten carb ide ( W C ) and indium tin ox ide (ITO) were se lec ted a s suppor ts of cho i ce . P la t inum w a s d i spe rsed on commerc i a l s a m p l e s of W C and ITO. Both the thermal and e lec t rochemica l stabil i ty of the suppor ted cata lyst w a s de te rm ined . T h e stabil i ty of the suppor ts w a s c o m p a r e d with both commerc i a l catalyst (H i spec 4000 ) and ih -house Pt cata lyst . T h e in -house Pt cata lyst w a s suppor ted on c o m m o n l y used high sur face a r e a ca rbon V u l c a n X C - 7 2 R cata lyst support . T h e e lec t rochemica l test ing involved app ly ing oxidat ion cyc l es be tween +0.6 V to +1.8 V and monitor ing the loss in activity of the suppor ted cata lyst over 100 oxidat ion c y c l e s . Tungs ten carb ide w a s found to be ex t remely s tab le . However , direct c o m p a r i s o n of its stabil i ty to V u l c a n X C - 7 2 R is current ly difficult. T h e e lec t rochemica l stabil i ty of 4 0 wt% Pt d i spe rsed on W C w a s c o m p a r e d with that of 40 wt% P t / C . E v e n though there are large d i f ferences in the dens i t ies of C and W C s o m e c o m p a r i s o n s are poss ib le . A n al ternat ive method to more directly c o m p a r e the stabi l i t ies of W C and C involved d ispers ing a s imi lar amount of Pt on s imi lar powder v o l u m e s of W C and C . T h e sol id dens i t ies of both ca rbon and W C were a l so used to d i spe rse s imi lar v o l u m e s of Pt on both suppor ts in order to more directly c o m p a r e the e lec t rochemica l stabil ity. Indium tin ox ide lost approx imate ly 5 0 % of its activity after 30 ox idat ion cyc les , w h e r e a s V u l c a n X C - 7 2 R lost a lmost 1 0 0 % of its activity after 10 ox idat ion cyc l es . Both the e lec t rochemica l ox idat ion and thermal stabil i ty tests s h o w e d that ITO is ex t remely s tab le c o m p a r e d to V u l c a n X C - 7 2 R . T A B L E O F C O N T E N T S A B S T R A C T . ii T A B L E O F C O N T E N T S . . . . . iii LIST O F F IGURES v LIST O F S Y M B O L S A N D ABBREVIAT IONS viii A C K N O W L E D G E M E N T S x DEDICATION xi 1 Literature Review 1 1.1 Catalyst support materials- Introduction 1 1.2 Studies of catalyst supports 6 1.2.1 C a r b o n . -. 6 1.2.2 S u m m a r y :. 14 1.3 Carbides 14 1.3.1 T u n g s t e n carb ide and e lec t rocata lys is 15 1.3.2 S u m m a r y '. 1.9 1.4 Oxides 19 1.4.1 S u m m a r y •. : , .22 1.5 Conclusions 22 2 Introduction to the three electrode electrochemical cell, voltammetry and rotating disc electrode (RDE) 23 2.1 Three electrode electrochemical cell 23 2.2 Voltammetry : 23 2.2.1 C y c l i c vo l tammetry (CV) . . . . 23 2.3 Convection and rotating disc electrode (RDE) 25 3 Objectives 26 4 Experimental Procedure 26 4.1 Pt addition using chlorplatinic acid (method I) 26 4.2 Pt addition using Pt (II) pentan-2,4-dionate (method II) 27 4.3 Pt addition to equal powder volumes of support (for WC studies only) 27 4.4 Pt addition to carbon and tungsten carbide by solid density method (for WC studies only) 28 4.5 Thermogra vimetric analysis (TGA) 28 4.6 Rotating disc electrode (RDE) and electrochemical test set-up 28 4.7 X-Ray Diffraction (XRD) 29 4.8 Scanning electron microscopy (SEM)/Transmission electron microscopy (TEM) and Energy Dispersive X-Rays (EDX) 30 5 Determining potential and time for electrochemical stability tests 30 6 Results and d iscuss ion for tungsten carbide studies 33 6.1 Alfa Aesar WC with Pt deposition using chlorplatinic acid.. : 33 6.2 Alfa Aesar WC with Pt deposition using Pt (II) pentan-2,4-dionate 37 6.3 Comparing activities of Pt deposited using Pt (II) pentan-2,4-dionate on similar volumes of both Alfa Aesar WC and Vulcan XC-72R v : 49 6.4 Pt addition in equal solid volume ratios to carbon and tungsten carbide.. 53 7 Results and d iscuss ion for indium tin oxide (ITO) studies 57 7.1 ITO with Pt deposition using chlorplatinic acid 57 7.2 ITO with Pt deposition using Pt (II) pentan-2,4-dionate 68 8 Conc lus ions 70 8.1 Tungsten carbide , 70 8.2 Indium tin oxide ; 71 9 Future Work 72 9.1 Tungsten carbide 72 9.2 Indium tin oxide 72 10 References 73 iv LIST O F F IGURES Figure 1: S c h e m a t i c showing the pr inciple of operat ion of a proton e x c h a n g e m e m b r a n e fuel cel l ( P E M F C ) ..1 F igure 2: S c h e m a t i c show ing potent ials in different reg ions a long the fuel and ox idant s i d e s of a P E M F C before startup (A) and during startup (B). F igure redrawn f rom work by R e i s e r et a l 1 0 . 5 F igure 3: Structure of a) qu inone; b) hydroqu inone .. .7 F igure 4 : C y c l i c vo l t ammogram of Pt G D E in 1.0 m o l / d m 3 in H 2 S 0 4 at 25°C, v = 3 0 m V / s . . . .25 F igure 5: Cur ren t v s . t ime plot obta ined f rom oxidat ion cyc les for H i s p e c 4 0 0 0 cyc led be tween +0.6 V (held for 60s ) and +1.8V (held for 20s) ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2 0 0 0 R P M . . 3 2 F igure 6: No rma l i zed activity at different potent ia ls and t imes a s a result of repeated cyc l ing for H i s p e c 4 0 0 0 . A v e r a g e of three s a m p l e s with limits of error for e a c h condi t ion are plotted... 33 F igure 7: X R D spec t rum of Pt d i spe rsed on V u l c a n X C - 7 2 R 34 F igure 8: X R D spec t rum of A l fa A e s a r W C s a m p l e 35 F igure 9: W C with f lakes of unknown mater ial . . . . .35 F igure 10: C y c l i c vo l t ammogram of A l fa A e s a r W C support ing Pt, both before and after 10 oxidat ion cyc l es ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2000 R P M 36 F igure 11: C h a n g e in anod ic activity at 1.8 V a s a result of repeated cyc l ing for 40 wt% P t / A A W C and 40 wt% P t / V u l c a n X C - 7 2 R : 36 F igure 12: X R D pattern for A l fa A e s a r W C with Pt depos i t ion by Pt (II) reduct ion to Pt . .37 F igure 13: X R D pattern for 40wt % Pt on V u l c a n X C - 7 2 R with Pt depos i t ion by Pt (II) reduct ion to Pt . .; : : 38 F igure 14: T G A da ta for A l fa A e s a r W C , 40 wt% Pt o n A l fa A e s a r W C , and 40 wt% Pt on V u l c a n X C - 7 2 R under air at 40ml /m in , temperature r amped f rom 50°C to 1000°C at 2°C/ min 39 F igure 15: C h a n g e in anod ic activity at 1.8 V a s a result of repeated cyc l ing for A l f a A e s a r W C , 40 wt% Pt on V u l c a n X C - 7 2 R and 40 wt% Pt on A l fa A e s a r W C . . A v e r a g e of three s a m p l e s with limits of error for e a c h material are plotted 41 F igure 16: C y c l i c v o l t a m m o g r a m s for A l fa A e s a r W C both before and after 100 oxidat ion c y c l e s ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2000 R P M 4 2 F igure 17: C y c l i c vo l t ammograms for 40 wt% Pt on W C both before and after 100 oxidat ion cyc l es ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2000 R P M 42 F igure 18: C y c l i c vo l t ammograms for 40 wt% Pt on V u l c a n X C - 7 2 R , both before and after 100 oxidat ion c y c l e s ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2000 R P M 4 3 F igure 19: S E M i m a g e s of A l fa A e s a r W C . 44 V Figure 20 : S E M i m a g e s of 40 wt% Pt on A l fa A e s a r W C , (A), (B) s e c o n d a r y e lec t ron m o d e ; (C) , (D) mixed s e c o n d a r y and backsca t te red e lect ron m o d e 4 4 F igure 2 1 : S E M images of 40 wt% Pt on V u l c a n X C - 7 2 R depos i ted us ing chlorplat in ic ac id ; images taken us ing m ixed s e c o n d a r y and backsca t te red e lect ron m o d e 4 5 F igure 22 : S E M i m a g e s of 40 wt% Pt on V u l c a n X C - 7 2 R depos i ted us ing Pt (II) pentan-2 ,4- d ionate; ( A - C ) mixed s e c o n d a r y and backsca t te red e lect ron m o d e , (D) s e c o n d a r y e lect ron m o d e : -46 F igure 23 : T E M i m a g e s and E D X of Pt d i spe rsed on A l fa A e s a r W C us ing Pt (II) pentan-2 ,4- d ionate. E lemen ta l spec t rum and weight concent ra t ion for spots 1 and 2 on the T E M image of Pt d i spe rsed o n A l fa A e s a r W C 48 F igure 24: T E M image and Pt m a p for Pt d i spe rsed on V u l c a n X C - 7 2 R us ing Pt (II) pentan-2 ,4- d ionate 4 9 F igure 25 : C y c l i c vo l t ammograms for Pt d i spe rsed on V u l c a n X C - 7 2 R ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2000 R P M 50 F igure 26 : X R D pattern for V u l c a n X C - 7 2 R with Pt depos i t ion by Pt (II) reduct ion to Pt . .51 F igure 27 : C y c l i c vo l t ammograms for Pt d i spe rsed on A l fa A e s a r W C , init ial lyand after 100 oxidat ion cyc les ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2 0 0 0 R P M 52 F igure 28: X R D pattern for A l fa A e s a r W C with Pt depos i t ion f rom Pt (II) reduct ion to Pt. . . .53 F igure 29 : C y c l i c vo l t ammograms for Pt d i spe rsed on A l fa A e s a r W C , initially and after 100 oxidat ion cyc l es ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2000 R P M 55 F igure 30: X R D pattern for A l fa A e s a r W C with Pt depos i t ion f rom Pt (II) reduct ion to Pt 56 F igure 31 : T G A data for H i s p e c 4 0 0 0 and V u l c a n X C - 7 2 R ; under air at 40ml /m in , temperature r amped f rom 50°C to 1000°C at 2 °C /m in . 58 F igure 32: T G A data for H i s p e c 4 0 0 0 , V u l c a n X C - 7 2 R , Pt on V u l c a n X C - 7 2 R , Pt o n ITO, and ITO; under air at 40ml /m in , temperature r amped f rom 50°C to 1000°C at 2°C/min 59 F igure 33 : X R D pattern for 40 wt% Pt on ITO depos i ted us ing method I 5 9 F igure 34: No rma l i zed activity at different potent ials a s a result of repeated cyc l ing for different 40 wt% Pt ca ta lys ts . A v e r a g e of three s a m p l e s with limits of error for e a c h mater ial are plotted. •. 61 F igure 35 : C y c l i c v o l t a m m o g r a m s for ITO both before and after ox idat ion cyc l es at 1.8V. 100 oxidat ion cyc les were run; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2000 R P M 61 F igure 36: C y c l i c v o l t a m m o g r a m s for 40 wt% Pt on ITO both before and after ox idat ion c y c l e s at 1.8V. 100 ox idat ion cyc l es were run. E lec t rochemica l stabil i ty.with no c h a n g e in the C V cu rves is o b s e r v e d f rom cyc le 30 onwards ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2 0 0 0 R P M 6 2 vi Figure 37: C y c l i c vo l t ammograms for H i s p e c 4 0 0 0 , both before and after 100 oxidat ion cyc l es ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2000 R P M . 63 F igure 38: C y c l i c vo l t ammograms for 40 wt% Pt on V u l c a n X C - 7 2 R , both before and after 50 oxidat ion cyc l es ; 0.5 M H 2 S 0 4 , 30°C, 100 m V / s , 2 0 0 0 R P M 64 F igure 39 : S E M images of ITO 65 F igure 40 : S E M images of 40 wt% Pt on ITO . . .66 F igure 4 1 : T E M / E D X of 40 wt% Pt on ITO 67 F igure 4 2 : X R D pattern for Pt depos i ted on ITO by reduct ion of Pt (II) pentan-2 ,4-d ionate 68 F igure 4 3 : C y c l i c vo l t ammograms of Pt depos i ted on ITO by reduct ion of Pt (II) pentan-2 ,4- d ionate after ox idat ion cyc l es 2 and 4. T h e inset s h o w s the C V at initial point . . .69 F igure 44 : No rma l i zed activity at different potent ials a s a result of repeated cyc l ing for Pt depos i ted us ing chlorplat in ic ac id (method I) and Pt (II) pentan-2 ,4-d ionate (method II) on ITO. A v e r a g e of 3 sepa ra te s a m p l e s with limits of error for e a c h mater ial a re plotted 70 vii LIST O F S Y M B O L S A N D ABBREVIAT IONS P E M F C Pro ton e x c h a n g e m e m b r a n e fuel cel l Pt P la t inum > W Tungs ten ITO Indium tin ox ide C C a r b o n W C Tungs ten carb ide A A W C A l fa A e s a r tungsten carb ide V u l c a n X C - 7 2 R C o m m e r c i a l ca rbon suppor t w ide ly u s e d in P E M F C s (Cabot , Inc.) P A F C P h o s p h o r i c ac id fuel cel l Ox idat ion R e m o v a l of e lec t rons E lec t rochemica l Ox idat ion R e m o v a l of e lec t rons occu rs at e lec t rode of interest. E lec t rons p a s s through external circuit to ca thode w h e r e they carry out reduct ion. Co r ros ion Unl ike e lec t rochemica l ox idat ion, in cor ros ion both the a n o d e and ca thode are on the s a m e sur face and no external power supp ly is required for cor ros ion to occur . Both ox idat ion and reduct ion occu r on the s a m e sur face , whe re a potential d i f ference ar is ing f rom a local ly vary ing chem ica l env i ronment dr ives the react ion. R D E Rotat ing d i sc e lec t rode C V C y c l i c Vo l tammet ry viii M e m b r a n e e lec t rode a s s e m b l y A C K N O W L E D G E M E N T S I wou ld l ike to e x p r e s s my s incere grat i tude to my superv iso rs Dr. O l i ve ra Kes le r and Dr. S t e p h e n C a m p b e l l for their suppor t and gu idance throughout my project. T h e y both prov ided good laughs and a very sa fe and st imulat ing a tmosphe re . I wou ld a l so like to thank Dr. S t e p h e n C a m p b e l l , Dr. P a u l Beatt ie and Dr. S i lv ia W e s s e l (Bal lard P o w e r S y s t e m Inc.) for their gu idance and a l lowing m e to f ind a w a y to do my graduate work whi le cont inuing to work on other projects at Ba l la rd P o w e r S y s t e m s , Inc. I wou ld a l so like to thank M a r y Mage r , for help ing m e with the T r a n s m i s s i o n E lec t ron M i c r o s c o p e and H igh -Reso lu t ion S c a n n i n g E lec t ron M i c r o s c o p e , and An i ta L a m for helping m e with X - R a y Dif fract ion. Final ly , I wou ld l ike to a c k n o w l e d g e Natura l S c i e n c e and Eng ineer ing R e s e a r c h Counc i l of C a n a d a ( N S E R C ) for Industrial Pos tg radua te Scho la rsh ip , Ba l la rd P o w e r S y s t e m s Inc., and resea rch funding f rom the A d v a n c e d S y s t e m s Institute of Bri t ish C o l u m b i a . X DEDICATION For my h u s b a n d and fami ly who were a l w a y s there for m e . xi i Literature Review 1.1 Catalyst support materials-Introduction Pro ton e x c h a n g e m e m b r a n e fuel ce l ls ( P E M F C s ) are e lec t rochemica l energy conve rs ion dev i ces that react hydrogen and o x y g e n to p roduce electr icity. T h e y c a n be used for power genera t ion in portable, stat ionary, and t ransportat ion app l ica t ions . L o s s of pe r fo rmance dur ing ex tended operat ion is a major prob lem hinder ing commerc ia l i za t ion of P E M F C s . P r o p o s e d m e c h a n i s m s that contr ibute to pe r fo rmance degradat ion inc lude catalyst part icle s inter ing, cata lyst d isso lu t ion , m e m b r a n e degradat ion , a n d cata lyst suppor t c o r r o s i o n 1 . T h e ma in componen t of the P E M F C is the m e m b r a n e e lec t rode a s s e m b l y ( M E A ) . T h e M E A cons is t s of two porous e lec t rodes ; e a c h ca ta l yzed on one s ide and bonded with a thin m e m b r a n e . O n the a n o d e s ide , the hydrogen ox id i zes to p roduce .p ro tons and e lec t rons . T h e protons p a s s through the electrolyte to the ca thode and the e lec t rons are fo rced through an external circuit, wh ich leads to the ca thode s ide . T h e e lec t rons travel to the ca thode , whe re they c o m b i n e with oxygen to p roduce water. W h e n the e lec t rons are forced through the external circuit, direct e lec t rochemica l energy conve rs ion is a c h i e v e d (F igure 1), resul t ing in lower pollut ion and h igher ef f ic iency c o m p a r e d to c o m b u s t i o n - b a s e d ene rgy conve rs ion . • 0 2 Cathode I'l-M H ' transport H 2 Anode Vi 0 2 + 2 H + + 2e- — H 2 0 H 2 — 2H + + 2e" Overall: H 2 + l / 2 0 2 — H 2 0 Figure 1: Schematic showing the principle of operation of a proton exchange membrane fuel cell (PEMFC) 1 T h e ca thode or oxygen reduct ion e lec t rode is held at relat ively oxidat ive potent ials at e leva ted tempera tures , whe re dur ing the oxygen reduct ion p r o c e s s a toms genera ted by the Pt part ic les m a y react with ca rbon a toms to genera te g a s e o u s products s u c h a s C O and C 0 2 , result ing in a loss of ca rbon . Th i s degradat ion m e c h a n i s m is very difficult to s tudy b e c a u s e it o c c u r s very s lowly c o m p a r e d with the in tended catalyt ic react ion of oxygen with protons and e lec t rons to form water. In the l i terature, ca rbon catalyst suppor t co r ros ion has been predominant ly reported in P h o s p h o r i c A c i d Fue l C e l l s ( P A F C s ) . A l though the operat ing temperature range of P E M F C s is lower than that of P A F C s , degradat ion in pe r fo rmance due to cata lyst suppor t cor ros ion has been obse rved in P E M F C s dur ing duty cyc l ing . T h e t he rmodynamics s h o w that a l though the rate of ox idat ion is lower in the c a s e of P E M F C s , it canno t be avo ided 2 | 3 . Deve lopmen t of M E A s with longer l i fet imes and h igher power ef f ic ienc ies and m a d e from the lowest cos t c o m p o n e n t s is ongo ing . T h e e lect rocata lyst on either e lec t rode (anode and ca thode) is usua l ly impregnated into the porous structure of a ca rbon suppor t mater ia l . T h e s e suppor t mater ia ls c a n be chemica l l y or phys ica l ly act ivated ca rbons , ca rbon b lacks , and graphi t ized ca rbons . The i r ro les a re : • T o provide high sur face a rea over wh ich smal l metal l ic part ic les c a n be d i spe rsed and stabi l ized • T o a l low fac i le m a s s transport of reactants and products to and f rom the act ive s i tes • T o prov ide e lect ronic and thermal conduct iv i ty A l s o crit ical are propert ies s u c h a s porosity, pore s i ze distr ibut ion, c rush strength, su r face chemist ry , and microstructural and morpho log ica l stabil i ty that need to be cons ide red before se lec t ing a sui table support . Cata lys t suppor t ox idat ion has been o b s e r v e d a s a ser ious prob lem that l eads to ex tens ive M E A deg rada t i on 4 , l imiting M E A l i fet imes. E lec t rochemica l ox idat ion p roduces microstructural degradat ion and su r face chem ica l c h a n g e s , wh ich lead to lost catalyt ic activity or e v e n catast roph ic e lec t rode fai lure. It is known that ca rbon ox id i zes in a q u e o u s solut ion, by the fo l lowing reac t ion 5 : C + 2 H 2 0 -> C 0 2 + 4 H + + 4e" Equation 1 T h e s tandard e lec t rode potential for this react ion at 25°C is 0 .207 V vs . S H E . C a r b o n is therefore thermodynamica l l y unstab le a b o v e this potent ial . S i n c e fuel cel l ca thodes opera te at h igher potent ials than 0 .2V, the ca rbon cata lyst suppor t will ox id ize dur ing u s e . 2 T h e ca rbon suppor t in a P E M F C is suscept ib le to cor ros ive condi t ions s u c h as high water content, low p H , h igh oxygen concent ra t ion , temperature ranging f rom 50-90°C, and high potential (0.6-1.2 V ) 6 . T h e Pt cata lyst a lso p lays a role in acce le ra t ing the ca rbon cor ros ion . Resu l t s f rom Differential E lec t rochemica l M a s s S p e c t r o s c o p y ( D E M S ) s tud ied by R o e n et a l 6 s h o w e d that C 0 2 em i ss i on w a s direct ly proport ional to the Pt su r face a rea for ca rbon in 0.5 M su lphur ic ac i d . A t tempera tures 30, 50, and 70°C it w a s found that Pt ca ta l yzed suppor t w a s eas i l y ox id ized c o m p a r e d to the non-ca ta lyzed suppor ts . P E M F C s in automot ive app l ica t ions are expec ted to expe r i ence up to 30 ,000 s tar tup/shutdown cyc l es in their operat ing l i fetime. After shu tdown, the hydrogen is removed f rom the s tack. W h e n hydrogen is re- in t roduced dur ing start-up, e lec t rode potent ia ls in e x c e s s of 1.5V m a y be expe r i enced for short per iods of t ime. Th i s leads to a signi f icant degradat ion in the fuel cel l pe r fo rmance due to oxidat ion of the ca rbon cata lyst support . T h e catalyst suppor t must be ab le to surv ive the accumu la ted t ime at t hese high potent ia ls, up to 100 hours, in order to prov ide the n e c e s s a r y durabil i ty. Acco rd i ng to Math ias et a l 2 the ca rbon suppor ts current ly used in automot ive app l ica t ions (Vu lcan X C - 7 2 R and Ketjen) do not meet automot ive requ i rements . It is cri t ical, therefore, to have a catalyst suppor t that is more s tab le than ca rbon in P E M F C s . S e v e r e cata lyst suppor t ox idat ion has a lso been o b s e r v e d w h e n a P E M F C is dr iven into a vo l tage reversa l , wh i ch c a n occu r if a cel l rece ives an inadequate supp ly of fue l . In order to p a s s current, react ions other than fuel ox idat ion c a n take p lace at the a n o d e , inc luding water e lect ro lys is and oxidat ion of a n o d e componen t s . T h e ca rbon b lack suppor t at the a n o d e c a n severe ly ox id ize w h e n the cel l g o e s into reversa l . T h e r e have been var ious techn iques e m p l o y e d in order to m a k e the ce l ls more tolerant to cel l reversa l . T h e s e inc lude add ing a higher cata lyst load ing or c o v e r a g e on a cor ros ion- res is tant suppor t 7 . Ano the r a p p r o a c h to m a k e a reversa l tolerant a n o d e cata lyst invo lves add ing e lect rocata lyst s u c h a s R u a long with P t 8 . E v e n though ex tens ive resea rch has b e e n done in mak ing reversa l tolerant a n o d e s , s imi lar mit igation st rategies cannot be used for ca thode suppor t cor ros ion . T h e potent ia ls expe r i enced by the a n o d e dur ing reversa l a re ex t remely h igh (in e x c e s s of 2 V ) w h e n c o m p a r e d to the ca thode , and the R u addit ion to the a n o d e only a ids at t hese higher potent ials. Ru then ium addi t ion to the ca thode wou ld not m a k e the ca thode oxidat ion-resistant , s i nce the potent ials expe r i enced by the ca thode are lower (1 .5V to 1.8V) c o m p a r e d to the a n o d e 9 . There fo re , al ternat ive me thods need to be sought in order to min imize the ca thode suppor t ox idat ion in P E M F C s . R e i s e r et a l 1 0 exp la in the m e c h a n i s m of potential excu rs ion (F igure 2). Be fo re startup, both a n o d e and ca thode potent ia ls ( E a and E c ) a re at the equi l ibr ium potential of oxygen (F igure 2 A ) 3 lead ing to a ze ro cel l potential (V c e n) . W h e n hydrogen is in t roduced into the fuel e lec t rode, the a n o d e potential (Va) r e a c h e s the equi l ibr ium potential of hydrogen ( E H 2 = OV), inc reas ing V 1 c e n to ~ 1.2 V (in reg ion 1, F igure 2B) . T h e a n o d e and ca thode potent ials remain at 1.2V in reg ion 2, where oxygen is still present . There fo re , the cel l potential in region 2 is still V 2 c e n ~ 0 V . D u e to the conduct ive plates on either s ide of the M E A , the cel l potential of 1.2 V in reg ion 1 then dr ives the react ions in region 2 in order to have equa l cel l potential a c r o s s reg ions 1 and 2. D u e to h igh e lect ronic conduct iv i ty be tween reg ions 1 and 2, the ca thode cel l potential in reg ion 2, wh ich w a s a l ready at 1.2V, then r e a c h e s a m a x i m u m theoret ical potential of 2 .4V . D u e to seve ra l potent ial l o s s e s the actual potential excu rs ion on the ca thode in region B is usual ly ~ 1.5V. T h e m a x i m u m theoret ical potential of - 2 . 4 V is not o b s e r v e d by the ca thode in region 2 due to severa l k inet ic l o s s e s (high overpotent ia ls for the react ions) . T h e high potential of 1.5V on the ca thode l eads to oxygen evolut ion and ca rbon cor ros ion at the ca thode . T h e m e c h a n i s m of this potential excu rs ion is s h o w n in F igure 2 1 0 . 4 Inlet Outlet A n o d e M C a t h o d e e m A 1 b A • 1 R r I R a n V a = 1 .23 V e V c = 1 .23 V V c e l l = V c - V a = OV Inlet Anode H 2 • 2H* + 2e Va = OV 0 2 + 4H + + 4e = 2H 2 0 E a = 1.2V e- M e m b r a n e Cathode 0 2 + 4H + + 4e" = 2 H 2 0 Vc • 1.2V C + 2 H 2 0 = C 0 2 + 4H + + 4e 2H 2 0 = 0 2 + 4H + + 4e" E c = 1.2 V e- Region 1 E c e „ = V c - V a E ' c e l l = 1 . 2 V Region 2 E c e i , = V c - V a E2ce.l = OV Outlet Figure 2: Schematic showing potentials in different regions along the fuel and oxidant sides of a P E M F C before startup (A) and during startup (B). Figure redrawn from work by Reiser et a l 1 0 . Mit igat ion s t ra teg ies rev iewed by R e i s e r et al inc lude mainta in ing a reduc ing fuel in the a n o d e , e .g . , H2, CH4, etc. A n o t h e r app roach inc ludes reduct ion of cel l vo l tage by intentional ly d rawing an electr ical load dur ing s ta r tup /shu tdown 1 0 . Us ing an inert or s o m e w h a t inert g a s for purg ing, such a s N2, before int roducing fuel is another app roach s u g g e s t e d by R e i s e r et a l 1 0 . Us ing inert g a s is not an i ssue for stat ionary appl icat ions, s ince a cy l inder of an inert g a s c a n be a t tached to the modu le . However , for the automot ive sector , the extra c o m p o n e n t s to be a d d e d wou ld a d d cos t in te rms of s p a c e and weight. 5 C a r b o n ox idat ion in the ca thode is not on ly an i ssue dur ing s tar tup/shutdown, but a l so w h e n the ca r is under operat ion. Dur ing operat ion, air c r o s s e s over and reacts with hydrogen on the anode , caus ing local fuel s tarvat ion. M a n y a p p r o a c h e s have been tried to stop ca rbon ox idat ion, but s imply f inding an al ternat ive suppor t wou ld result in a s impler s y s t e m . A var iety of mater ia ls c a n be u s e d a s catalyst suppor ts . A deta i led rev iew exp la in ing the p rob lems with ca rbon suppor ts and the mit igation st rategies exp la ined by var ious resea rche rs is p resented here . 1.2 Studies of catalyst supports 1.2.1 Carbon The re are a var iety of ca rbons that a re ava i lab le a s catalyst suppor ts , and m a n y character is t ics (phys ica l , e lec t rochemica l , etc.) must be s tud ied before down-se lec t ing a suppor t for a part icular app l ica t ion. T h e fo l lowing sec t ion rev iews the ca rbon oxidat ion p rob lems encoun te red in P A F C s , the p roposed microst ructures of ca rbon that acce le ra te ox idat ion, and the modi f icat ions m a d e to ca rbon to i nc rease its stabil ity. T h e ca rbon ox idat ion prob lem in P E M F C s and the role of Pt a re a l s o exp la ined in the fo l lowing sec t ion . Studies using phosphoric acid as the electrolyte G r u v e r 1 1 s tud ied the cor ros ion of ca rbon black in phosphor ic ac id ; he exp la ins that ca rbon is the rmodynamica l l y unstab le under phosphor i c ac id fue l ce l l c a t h o d e operat ing cond i t ions . G r u v e r 1 1 d i spe rsed Pt on V u l c a n X C - 7 2 (Cabot Corp . ) and invest igated the c h a n g e in microstructure of the ca rbon by T E M both before and after operat ing for 1000 hrs at 0 .835 V at 91 °C. M ic rog raphs s h o w e d that the outermost crystal l ine shel l of the ca rbon part icle rema ined intact, whi le the central port ion, wh ich is more d iso rdered , ox id ized away . K inosh i ta et a l 1 2 s tudied the e lec t rochemica l ox idat ion of high sur face a rea ca rbon b lack ( N e o Spec t ra , C o l u m b i a C a r b o n ) with sur face a rea of 1000 m 2 / g in concent ra ted phosphor i c ac id at 135°C. T h e y exp la in qual i tat ively that two p r o c e s s e s occur initially: ox ide format ion on the sur face and C 0 2 evolut ion. A s the sur face ox ide format ion d e c r e a s e d , the C 0 2 evolut ion b e c a m e the major anod ic p r o c e s s . T h e sur face ox ide growth did not inhibit C 0 2 format ion. K inosh i ta and B e t t 1 3 later s tudied the effects of graphi t izat ion on the cor ros ion of ca rbon b lacks . T h e y u s e d three brands of ca rbon b lack : N e o S p e c t r a (1000 m 2 /g ) , G r a p h o n (100 m 2 / g , graphi t ized sphe res ) , 6 and V u l c a n (220 m 2 /g ) . Signi f icant C 0 2 evolut ion w a s obse rved for non-graphi t ized ca rbon c o m p a r e d to graphi t ized ca rbon . T h e y a l so exp la in that other r esea rche rs used potent iodynamic tests to indicate the p r e s e n c e of a qu inone /hydroqu inone sur face s p e c i e s dur ing e lec t rochemica l ox idat ion of ca rbon b lacks . Q u i n o n e is a c l a s s of a romat ic c o m p o u n d s , and the structures of qu inone and hydroqu inone are s h o w n in F igure 3. K inosh i ta and B e t t 1 3 exp la in that a w ide range of su r face ox ides is poss ib le dur ing the e lec t rochemica l ox idat ion of ca rbon . Figure 3 : Structure of a) quinone; b) hydroquinone Microstructural Studies H e c k m a n and H a r l i n g 1 4 s tud ied the microst ructures of thermal b lacks , wh ich are a type of high sur face a rea pyro lyzed ca rbon . T h e e lectron mic rographs of non-graphi t ized thermal b lacks s h o w that there is s o m e crystal l inity on the outer s ide of t hese ca rbons , and the oxidat ive attack r e m o v e s the ca rbon f rom the inside out, forming hol low c a p s u l e s . T h e y a lso c la im that not all ca rbon part ic les ox id ize . For a graphi t ized thermal b lack the ox idat ive attack is very s low c o m p a r e d to in non-graphi t ized ca rbons . A l s o , the oxidat ive attack on graphi t ized thermal b lack occu rs from the outs ide in, and the outer p lanes genera l ly f lake off as they are c o n s u m e d . H e c k m a n and H a r l i n g 1 4 exp la in that graphi t izat ion s e e m s to p roduce a thick she l l , wh ich is "airtight" and strongly res is ts ox idat ive attack. Smi th and P o l l e y 1 5 s tud ied the oxidat ion of graphi t ized ca rbon b lack. Graph i t i zed ca rbon b lacks were r esea rched main ly in the 1950 's b e c a u s e they have outs tanding sur face uniformity. Smi th and P o l l e y 1 5 report that w h e n a s tandard ca rbon b lack w a s treated with air or oxygen at tempera tures from 300-650°C, a s ix- fold su r face a rea inc rease without part icle s i z e c h a n g e w a s o b s e r v e d . T h e i nc rease in a rea without part icle s i z e c h a n g e w a s recogn ized as an i nc rease in porosi ty upon air ox idat ion. It w a s p roposed that the inc rease in a rea might be c a u s e d by preferential at tack of the oxygen at the h igh-energy e d g e s i tes of the quas i -graphi t ic paral le l layers of the ca rbon part ic les. T h e y further exp la in that the edge a toms are more suscept ib le to chemica l attack than are the a toms in the center of the basa l p lane. Heat- t reat ing ca rbon at A B 7 tempera tures a b o v e 2700°C dest roys the h igh-energy s i tes. There fo re , Smi th and P o l l e y 1 5 studied whether the preferred s i tes for oxygen attack are a l so d imin ished w h e n ca rbon is graph i t ized. T h e y heat- t reated ca rbon b lack f rom 1000-2700°C and found a d e c r e a s e in su r face a rea (determined by adsorp t ion of N 2 ) with inc reas ing temperature of heat t reatment due to s o m e part icle s inter ing. B y charac ter iz ing the s a m p l e s under the e lect ron m i c r o s c o p e , they o b s e r v e d that a s the deg ree of graphi t izat ion w a s i nc reased , the part icle s h a p e s c h a n g e d from spher ica l to regular po lyhedra . Ox ida t ion of the ca rbon b lacks led to an inc rease in a rea of 63 .5 m 2 / g for untreated ca rbon mater ia l , and an i nc rease of 4 m 2 / g in a rea w a s o b s e r v e d for the graph i t ized ca rbon . T h e inc rease in a rea of non heat- t reated ca rbon further suppor ts the hypothes is that non heat- t reated ca rbon has h igh-energy e d g e s i tes that a re suscep t ib le to o x y g e n attack. Effecf on carbon oxidation after Pt addition Wi l l sau et a l 1 6 invest igated the d e p e n d e n c e on the potential of the anod i c ox idat ion of N O R I T B R X ca rbon e lec t rodes us ing differential e lec t rochemica l m a s s s p e c t r o s c o p y ( D E M S ) in sul fur ic ac id . - T h e half w a v e potent ia ls for the O R R in.sulfur ic ac id are 888 m V ( R H E ) at the Pt -ac t iva ted e lect rode and 4 8 3 m V at the pure ca rbon e lec t rode, respect ive ly . T h e y report that no H 2 0 2 w a s detec ted in the c a s e of the Pt -ac t iva ted e lect rode, w h e r e a s perox ide w a s the ma in react ion product on pure ca rbon . Pu re ca rbon unde rgoes signif icant ox idat ion at potent ia ls h igher than 0.9V, with C 0 2 a s the ma in product. T h e Pt -act ivated ca rbon e lec t rode unde rgoes oxidat ion to C 0 2 at m u c h lower potent ia ls. T w o anod i c p e a k s in cyc l i c vo l tammetry resul t ing f rom two different ox ides are o b s e r v e d for P t - ca ta l yzed ca rbon : 1) T h e format ion of a C O - s u r f a c e layer at E > 0 .3V ( R H E ) C + H 2 0 -> C O s u r f + 2 H + + 2e" Equation 2 2) Ox idat ion of C O s u r f on Pt be tween 0.6 to 0 .8V C O s u r f + H 2 0 (with Pt catalyst) -> C 0 2 + 2 H + + 2e" Equation 3 Pt ca ta l yzes the oxidat ion of a C O - s u r f a c e layer at potent ials be tween 0 .6V and 0 .8V, and this potential range co inc ides with the reduct ion of the oxygen on these e lec t rodes , caus ing sur face destruct ion and loss of the act ive Pt s i tes . S t e v e n s et a l 1 7 s tud ied how catalyst suppor t degradat ion contr ibutes to long-term P E M F C per fo rmance degradat ion . T h e y used s a m p l e s conta in ing f rom 5 to 80 wt% plat inum suppor ted 8 on either B P 2 0 0 0 or V u l c a n X C - 7 2 ca rbons . T h e P t - l oaded ca rbons we re held at e leva ted tempera tures under either dry or humidi f ied cond i t ions for ex tended per iods of t ime. In order to test s a m p l e s under humidi f ied condi t ions, a s e a l e d humidi ty box w a s instal led in the oven with a s team generator set at 95°C, wh ich w a s fed with c o m p r e s s e d air. At regular intervals, the s t e a m f low w a s s topped and dry air w a s fed for two hours in order to dry the s a m p l e s prior to we igh ing . If the amoun t of ca rbon lost w a s sma l l after 1000h the cata lyst suppor t w a s cons ide red s tab le . B E T sur face a rea w a s m e a s u r e d us ing a 30 v o l % N 2 / 7 0 v o l % H e g a s mixture. T h e stabil i ty of the catalyst and the suppor t we re a s s e s s e d through a 1.2 V acce le ra ted test protocol . T h e 1.2 V acce le ra ted test involved test ing a 5 0 c m 2 M E A under H 2 / N 2 at 80°C for 50 hours . T h e cel l w a s held potent iostat ical ly at 1.2 V for 5 hr per cyc le and 50 cyc l i c vo l t ammograms ( C V s ) were c o l l e c t e d / o n e after e a c h 5 hour interval. T h e y a l so genera ted a polar izat ion curve by s tepp ing the cel l potential f rom 0.9 to 0 .3V, and then f rom 0.3 to 0.9 V in 0 .05V inc rements . After monitor ing the fuel cel l pe r fo rmance the s a m p l e w a s swi tched back to 1.2 V test condi t ions. It w a s found that c a r b o n s with no Pt addi t ion had very low reactivi ty and were s tab le up to 150°C under both dry and humid air condi t ions. T h e s a m p l e s conta in ing 5 to 10 wt% Pt on B P 2 0 0 0 s h o w e d no weight loss even after 2500 h of exposu re , imply ing s low ca rbon ox idat ion react ion k inet ics. It w a s reported that the number of Pt part ic les per unit a rea of ca rbon inc reased a s the Pt loading i nc reased . It w a s found that V u l c a n X C - 7 2 has less ca rbon loss than B P 2 0 0 0 . Th i s d i f ference is main ly b e c a u s e V u l c a n X C - 7 2 ( B E T = 2 2 0 m 2 / g ) is more graphit ic than B P 2 0 0 0 ( B E T = 1 3 0 0 m 2 /g ) . T h e rate of ox idat ion w a s higher for both ca rbons under humidi f ied condi t ions. C a r b o n c a n be c o n s u m e d through the wate r -gas shift react ion: C + H20<=>H2 + C O Equation 4 Th is is a conce rn for P E M F C s , a s humidi f icat ion is essent ia l for t hese types of fuel ce l ls , and the O R R has water a s its main product. T h e h igh -su r face-a rea Ket jen b lack ca rbon suppor ted cata lyst w a s e lec t rochemica l l y more act ive than the graphi t ized c a r b o n - b l a c k - b a s e d cata lyst . However , the per fo rmance of Ket jen-b lack b a s e d cata lyst deter iorated signi f icant ly with t ime, lead ing to a lmos t comp le te loss of catalyt ic activity at 8 0 0 m V after only 30 h of acce le ra ted test ing. T h e per fo rmance of the V u l c a n X C - 7 2 b a s e d cata lyst a l so deter iorated with hold t ime at 1.2 V . T h e stabil i ty of the graphi t ized ca rbon b lack m a k e s the suppor t a more attractive cand ida te for P E M F C s in tended to operate for an ex tended per iod. 9 K a n g a s n i e m i et a l 3 a l so s tud ied whether a c o m m o n ca rbon b lack, V u l c a n , ox id i zes under s imula ted P E M F C cond i t ions . T h e y reported that a var iety of ca rbon sur face ox ides have b e e n identif ied by titration and infrared spec t roscopy . T h e s e sur face ox ides m a y be pheno ls , carbony ls , carboxy l ic ac i ds , e thers, qu inones , and l ac tones 3 . T h e y depos i ted V u l c a n X C - 7 2 and 10 wt% polytetraf luoroethylene binder onto To ray ca rbon paper . A three-e lec t rode e lec t rochemica l cel l with 1 M H 2 S 0 4 w a s u s e d for the sur face oxidat ion exper iments . A cons tan t potential of 0.8, 1.0, or 1.2V w a s app l ied , and C V s were recorded after every 2 h . After 16, 60 , and 120h of oxidat ion treatment, s a m p l e s we re removed for su r face ox ide ana lys i s . T h e y found that V u l c a n expe r i enced sur face oxidat ion at potent ials > 1V at room temperature and > 0 .8V at 65°C. S u r f a c e ana lys is with X P S found an inc rease in a tomic oxygen over t ime after the 1.2 V potential ho lds. A n i nc reased current of the ca thod ic current peak at - 0 . 5 5 V w a s obse rved in the C V after a 1.2V hold for 16h at 65°C, indicat ing a signi f icant su r face oxidat ion of V u l c a n . K inosh i ta and B e t t 1 8 de te rmined the impact of Pt on ca rbon sur face ox ides for the P t - ca ta l yzed ca rbon . T h e cha rge due to the e lec t ro -ac t i ve -ca rbon s p e c i e s is difficult to sepa ra te f rom the cha rge assoc ia ted with F a r a d a i c react ions on Pt for the Pt ca ta l yzed ca rbon e lec t rodes . T h e y used 5 wt% and 20 wt% P t -ca ta l yzed and unca ta l yzed V u l c a n X C - 7 2 graphi t ized at 2700°C ( B E T sur face a rea of 70 m 2 /g ) . T h e y exp la in that hal ide ions in solut ion strongly adso rb on Pt, and the current assoc ia ted with e lec t rochemica l react ion on Pt c a n be signi f icant ly reduced in the p r e s e n c e of hal ide ions. B y hal ide addi t ion to Pt , both F a r a d a i c and n o n - F a r a d a i c currents of Pt c a n be min im ized . A po ten t iodynamic s w e e p a l lows the determinat ion of the concent ra t ion of e lect ro-act ive ca rbon ox ide . T h e y found that the e lec t rochemica l ox idat ion current on Pt - ca ta l yzed and unca ta lyzed graphi t ized c a r b o n s in 9 6 % H 3 P 0 4 at 160°C and 1200 m V were similar, indicat ing no catalyt ic inf luence of Pt on oxidat ion of graphi t ized V u l c a n X C - 7 2 . T h e s tudy by K inosh i ta and B e t t 1 8 show ing that Pt d o e s not ca ta lyze ca rbon ox idat ion contradic ts the f indings by W i l l sau et a l 1 6 , K a n g a s n i e m i et a l 3 , and S t e v e n s et a l 1 7 . Th i s d i f ference m a y be the result of inter ference with the hal ide ions on the Pt su r face . A test with Pt addi t ion and with hal ide addit ion to Pt cou ld be c o m p a r e d to reso lve the d i sc repancy . Doping and other treatments to carbon Stud ies by K a n g a s n i e m i et a l 3 and S t e v e n s et a l 1 7 s h o w e d that the most c o m m o n ca rbon (Vu lcan) used for P E M F C s ox id izes under P E M F C cond i t ions . M a n y modi f icat ions a s d i s c u s s e d be low have been m a d e to ca rbon in order to s top ca rbon ox idat ion. T h e s tud ies exp la ined be low e x a m i n e var ious ava i lab le ca rbons , a s wel l as the effects of dop ing with boron (B) 4 a n d phospho rus (P) 5 , and of us ing g l a s s y ca rbon ( G C ) . M c D o n a l d and S t o n e h a r t 1 9 exp la in that dop ing ca rbon with B prov ides "trap s i tes" for the Pt crystal l i tes. B e n h a n c e s the rate of 10 graphi t izat ion at lower heat treatment temperature . T h e y found that Pt suppor ted on boron carb ide is more resistant to agg lomera t ion than Pt b lack (unsuppor ted Pt) or Pt on graphi te at equa l su r face c o v e r a g e in hot H 3 P 0 4 . V u l c a n X C - 7 2 R (Cabot ) is the most conduct ive fu rnace b lack commerc ia l l y ava i lab le , and has a high B E T sur face a rea (250m 2 / g ) . T h e d rawback is that it d o e s not rema in s tab le at ca thode operat ing tempera tures . S i n c e it has been used extens ive ly , it w a s c h o s e n a s the first cand ida te for dop ing with B. S i n c e the B dop ing a l lows graphi t ic charac ter ( low d 0 latt ice spac ing ) at lower temperature (1600°C) , the sur face a rea rema ins higher than the su r face a r e a result ing f rom heat-treat ing V u l c a n at 3000°C, wh ich is normal ly n e c e s s a r y for graphi t izat ion. Resu l t s s h o w e d that 1600°C is the T a b o v e wh ich no signif icant addi t ional benefit w a s ob ta ined w h e n heat ing B -doped C . T h u s B dop ing a l lows more efficient uti l ization of the noble metal e lect rocata lyst on a more s tab le suppor t . It w a s found by Y o u n g et a l 2 0 that the ox idat ion behav iour of s a m p l e s d o p e d with B and P is quite different f rom that of only B or only P -doped s a m p l e s . T h e concent ra t ion of B used w a s very low and - th i s - therefore -did not improve ca rbon crystal l inity. - P h o s p h o r u s resul ts in a proport ional i nc rease in ox idat ion inhibition a s its concent ra t ion i nc reases ; in contrast , B exhibi ts both a catalyt ic and inhibit ing effect on ca rbon ox idat ion. Bo ron ' s inhibit ing effect c a n be man i fes ted in the fo l lowing w a y s : 1) Subst i tut ional boron e n h a n c e s the graphi t izat ion of ca rbon . 2) A s the su r face ca rbon a toms are c o n s u m e d , subst i tut ional boron fo rms an ox ide film ( B 2 0 3 ) , wh ich acts a s an 0 2 d i f fusion barr ier and an act ive site b locker . Y o u n g et a l 2 0 report that B acts a s a catalyst at low B load ings and a s an inhibitor at h igher B load ings . T h e m e c h a n i s m for the B loading effect w a s not exp la ined . T h e format ion of a B 2 0 3 layer m a y impact the conduct iv i ty of the cata lyst support . N o tests we re done by Y o u n g et a l 2 0 to e x a m i n e the impact on conduct iv i ty f rom the boron ox ide layer. S a m a n t et a l 2 1 s yn thes i zed a high su r face a rea m e s o p o r o u s ca rbon suppor t for methano l ox idat ion cata lyst us ing a so l -gel techn ique. H igh-dens i ty xe roge l w a s syn thes i zed by condensa t i on of the 1 ,3-d ihydroxybenzoic ac id and fo rma ldehyde in a molar ratio of 2 :1 . T h e carbon iza t ion of the gel w a s carr ied out in a ni trogen a tmosphe re at 800°C . A B E T a rea of 724 m 2 / g w a s m e a s u r e d for the syn thes i zed ca rbon . Pt w a s ancho red us ing an incipient we tness method where the suppor t is saturated with a Pt salt with a min imal amoun t of solvent , and the reduct ion of Pt w a s carr ied out by s l o w addi t ion of 0 . 1 M sod ium formate at 60°C. T h e Pt suppor ted on the high sur face a rea ca rbon had a part icle s i ze of ~2 nm. S a m a n t 2 1 et al found 11 higher e lectrocatalyt ic activity for methano l ox idat ion in an a lka l ine med ium than in ac id so lu t ions. S i n c e the ca rbon suppor t syn thes i zed by S a m a n t 2 1 et al has low activity in ac id ic condi t ions, this suppor t canno t be u s e d for P E M F C s , wh ich are known to opera te under ex t remely ac id i c condi t ions. A s it is wel l known that the structure of the ca rbon support in f luences the structure of the cata lyst layer, ca rbon mater ia ls s u c h a s graphi te nanof ibers , ca rbon nano tubes , and ca rbon s p h e r e s are be ing r esea rched for novel suppor ts for direct methano l fuel cel l ( D M F C ) app l ica t ions. Y a n g et a l 2 2 p repared hard ca rbon spheru les ( H C S ) , us ing sugar as the precursor . S E M i m a g e s revea led H C S part ic les with a v e r a g e d iameter of ~ 2 \xm, wh ich is s igni f icant ly larger than the suppor ts current ly be ing u s e d for P E M F C s with part icle s i z e s in the 20 -100 nm range. S e r p et a l 2 3 ana l ysed literature f rom the ear ly 1990s until 2 0 0 3 and d i s c u s s e d the u s e of ca rbon nano tubes and nanof ibers a s cata lys ts and catalyst suppor ts . N a n o t u b e s are m a d e exc lus ive ly of cova lent ly bonded ca rbon a toms, and they m a y be the most oxidat ion resistant f i b res 2 3 . W h e n nano tubes are used in ca ta lys is , these conduc t i ve suppor ts present c lear d i f ferences with respect to act ivated ca rbon . C a r b o n N a n o t u b e s ( C N T ) and Graph i te Nano f ibe rs ( G N F ) we re used a s metal suppor ts for fuel cel l e lec t rodes for the oxygen reduct ion react ion ( O R R ) . T h e d rawback is that the nano tubes do not have a contro l led range of su r face a rea , and the su r face a reas are usual ly low (8 m 2 / g to 109 m 2 /g ) c o m p a r e d to that of V u l c a n , 250 m 2 / g . T h e industrial product ion of t hese mater ia ls is a lso current ly poor, and h o m o g e n e o u s geomet r ies with contro l led d iameters a re not ava i lab le . Howeve r , the lab -sca le da ta have been promis ing . T h e product ion p r o c e s s e s are mov ing towards more contro l lable a tmosphe res , but the homogene i t y in character is t ics s u c h as geomet ry , purity, etc. still n e e d s ex tens ive work. W a n g and S w a n 2 4 invest igated P t /d iamond compos i te e lec t rodes . Pt part ic les we re galvanostat ica l ly depos i ted onto a bo ron-doped polycrysta l l ine d i a m o n d thin-fi lm su r face and ent rapped within the d imens iona l l y s tab le microstructure by a s u b s e q u e n t d i a m o n d depos i t ion . Pt part icle s i z e s in the range of 10-300 nm were o b s e r v e d . Th is part icle s i ze range is too w ide ; Pt part icle s i z e s of l ess than 20 nm are preferred. T h e s e methods m a y prove to be promis ing , but the particle s i ze n e e d s to be contro l led in a lower range, and the cos t of d i amond p r o c e s s e s m a y limit the use of this techn ique in the near future. T h e ideal catalyst suppor t shou ld be g a s and water pe rmeab le , with abil i ty to conduc t both e lec t rons and protons. If a mater ial cou ld ach ieve both ionic and e lect ron ic conduct iv i ty, then it cou ld rep lace both ca rbon and ionomer in the cata lyst layer. Conduc t i ng polymer/proton 12 e x c h a n g e po lymer compos i t es have been s h o w n to exhibit high e lect ron and proton conduct iv i t ies, e .g . polypyrole and po lys tyrene su lphonate ( P P Y / P S S ) 2 5 . In this work, fuel ce l ls us ing compos i t e po lymer G D E ' s exhibi ted an o p e n circuit vo l tage of 0.63 V and a m a x i m u m s teady state current dens i ty of 97 m A / c m 2 . T h e s e va lues are low c o m p a r e d to va lues of ~ 1.0V and > 1 A / c m 2 w i t h ca rbon suppor ted cata lys ts . T h e syn thes is of P P Y / P S S compos i t e and Pt depos i t ion require more s tud ies to p roduce substant ia l i nc reases in per fo rmance . M a r i e 2 6 et al have been do ing resea rch into us ing aeroge l -suppor ted cata lys ts for P E M F C s . C a r b o n ae roge l s have high conduct iv i ty, high mesoporos i ty , a smal l deg ree of microporosi ty , and high sur face a r e a . T h e y have been cons ide red for the di f fusion layer in P E M F C s . A Rotat ing D i sc E lec t rode ( R D E ) a s s e m b l y w a s used to e lec t rochemica l l y charac te r i ze these suppor ts . W h e n syn thes iz ing the aeroge l mater ia ls , the number of large part ic les i nc reased with inc reas ing sintering temperature . Pt d i spe rsed on V u l c a n exhib i ted higher act ive a reas than Pt d i s p e r s e d o n aeroge l suppor ts . T h e authors c la im that the Pt part ic les are wel l d i spe rsed for the ae roge ls , but most of them are inact ive. A l s o , the porosi ty of ae roge ls n e e d s to be moni tored, s ince the Pt might enter pores that a re not a c c e s s i b l e for the th ree -phase boundary of ionic and e lect ron ic conductor and g a s p h a s e . Ra jesh 2 7 e t al have s tud ied hybrid mater ia ls b a s e d on transit ion metal ox ide and conduc t ing mater ial for cata lyst suppor ts . T h e hybrid cons is t s of an o rgan ic (polyani l ine) and inorganic (vanad ium pentoxide) compos i te . A T E M image of Pt loaded ( C 6 H 4 N H ) o . 4 i V 2 0 5 0 . 5 H 2 0 s h o w e d nanopar t ic les with part icle s i ze of ~ 10 nm. T h e y c la im that w h e n cyc l ing the e lec t rodes be tween - 0 . 2 and +0.8 V , exce l len t e lec t rochemica l stabil i ty w a s a c h i e v e d . T h e y a lso s tud ied the var iat ion of methano l ox idat ion current dens i t ies on Pt l oaded ( C 6 H 4 N H ) 0 . 4 1 V 2 O 5 0 . 5 H 2 O nanocompos i t es and P t A / u l c a n ca rbon b a s e d e lec t rodes . Fo r the n a n o c o m p o s i t e - b a s e d e lec t rode, as the Pt load ing i nc reases , there is a cont inuous i nc rease in activity. T h e nanocompos i t e e lec t rode exhibi ted two t imes higher activity c o m p a r e d to the P tA /u l can X C - 7 2 R ca rbon e lec t rodes. T h e y a lso found the d e c r e a s e in methano l ox idat ion activity to be 2 9 % for P t y ( C 6 H 4 N H ) 0 . 4 1 V 2 C y 0 . 5 H 2 O nanocompos i t e b a s e d e lec t rodes at the end of 2 h, and to be 7 8 % for P tA /u l can X C - 7 2 R . Both P t - W 0 3 and P t - M o 0 3 have a lso been used a s e lec t rode mater ia ls for methano l ox idat ion. T h e main i ssue with these ox ides is the seve re leach ing of the meta ls s u c h a s W and M o . T h e techn ique prov ides improved e lec t rochemica l pe r fo rmance and stabil i ty; however , leach ing effects of meta ls need to be c lose l y s tud ied. 13 1.2.2 Summary B e c a u s e ca rbon m a y ox id ize at potent ia ls a b o v e 0.2 V at 25°C, and Pt ca ta l yzes the e lec t rochemica l ox idat ion of ca rbon , t reatments of ca rbon to m a k e it more oxidat ion resistant have been sought . A var iety of suppor ts is being r e s e a r c h e d , but deta i led e lec t rochemica l s tud ies involving cyc l i c vo l tammetry, po tent iodynamic tests, and thermal stabil i ty tests need to be conduc ted for se lec t ion of novel ox idat ion resistant suppor t mater ia ls . A method for rapid eva luat ion of stabil i ty of suppor ts is a l so in high d e m a n d . A d d i n g dopants to ca rbon only de lays the e lec t rochemica l ox idat ion, but d o e s not prevent it. A e r o g e l b a s e d mater ia ls have a high sur face a rea ; however , the porosi ty c a n be a d i sadvan tage to Pt activity. Organ ic - ino rgan ic hybrid suppor ts m a y lead to leach ing of meta ls under fuel cel l condi t ions. T h e e lec t rochemica l stabil i ty of nano tubes needs to be de te rmined, and the cos t n e e d s to be d e c r e a s e d before they c a n be used a s fuel cel l cata lyst suppor ts . Overa l l , it is of great advan tage to the proton e x c h a n g e m e m b r a n e fuel cel l industry that nove l suppor ts a re be ing sought wor ldwide in order to o v e r c o m e the e lec t rochemica l oxidat ion of the current ly most wide ly used support (carbon). Little r e s e a r c h , however , has been per formed to date on the u s e of non-ca rbon catalyst suppor ts . S o m e prel iminary s tud ies by seve ra l g roups have s h o w n the stabil i ty of ca rb ides and ox ides for s o m e fuel cel l app l ica t ions. C a r b i d e s and ox ides are interest ing ca tegor ies that need to be further exp lo red for P E M F C s . However , many commerc ia l l y ava i lab le ca rb ides have low su r face a rea , and syn thes iz ing ca rb ides us ing var ious methods has most ly b e e n per formed for non-fuel cel l industr ial app l ica t ions. Tungs ten carb ide has b e e n used as an a n o d e catalyst in P E M F C s ; however , it h a s not been u s e d a s a catalyst suppor t in P E M F C ca thodes , where ox idat ive stabil i ty is a more signif icant cha l l enge . T i tan ium carb ide has a l so been u s e d as a cata lyst support . T h e s e s tud ies are rev iewed here. 1.3 Carbides T h e r e have not been m a n y s tud ies on us ing ca rb ides a s catalyst suppor ts in P E M F C s . T i C has been used a s a cata lyst suppor t in P A F C s , and a patent by J a l a n et a l 2 8 c l a ims the u s e of t i tanium carb ide a s catalyst suppor ts for e lec t rodes in fuel ce l l s . T h e c la ims were in tended to u s e this 14 catalyst suppor t in e lec t rodes for the reduct ion of oxygen in phosphor i c ac id fuel ce l ls . T h e T i C suppor t w a s c o m p a r e d with convent iona l ca rbon black catalyst suppor ts , and it exhib i ted a lower cor ros ion current (thus, better cor ros ion res is tance) than the ca rbon b lack mater ia ls . T h e stabil i ty of T i C as a cata lyst suppor t for P A F C s w a s pointed out to be remarkab le ; however , more data on longer l i fetime degradat ion wou ld have further suppor ted this c la im . T h e suppor t m a y provide cor ros ion res is tance , but the overal l activity f rom cata lys ts us ing T i C and convent iona l suppor ts w a s not c o m p a r e d . T h e catalyt ic behav iour of tungsten ca rb ide for anod i c react ions in fuel ce l ls has been s tud ied ; however , tungsten carb ide b a s e d cata lyst suppor ts have not been studied'. S o m e stud ies conduc ted on the use of tungsten carb ide in P A F C s , and its e lec t rochemica l and cor ros ion behav io r a re rev iewed here . G r o u p VI ca rb ides h a v e h igh thermal conduct iv i ty , espec ia l l y W C , wh ich has the h ighest thermal conduct iv i ty of any of the t ransi t ion-metal ca rb ides . L ike the G r o u p 29 IV and V ca rb ides , the G r o u p VI ca rb ides have low thermal e x p a n s i o n . W C has the lowest electr ical-resist iv i ty (conductivi ty.-- 1 0 5 S / c m at 20°C) of any interstitial ca rb ides and qual i f ies a s the most metal l ic ca rb ide . W C c a n be m a d e by direct carbur isat ion of the metal with ca rbon or graphi te at 1400-2000°C in hydrogen or v a c u u m . T h e carb ide format ion p r o c e s s c a n a l so use tungsten ox ide, tungst ic ac id , or a m m o n i u m tungstates a s the start ing ma te r i a l s 2 9 . 1.3.1 Tungsten carbide and electrocatalysis In order to reduce the cost by reduc ing the amount of Pt in a fuel ce l l , it is pract ical to s e a r c h for a subst i tute b a s e metal a s a catalyst . A s i d e f rom being good electr ical conduc to rs and hav ing good catalyt ic activity, e lect rocata lys ts used in ac id ic e lectro lytes shou ld a l so be ac id resistant. T h e s tandard revers ib le potent ial for the o x y g e n reduct ion react ion ( O R R ) is 1.23V, but k inet ic l imitations for the O R R lead to cel l overpotent ia l l oss of 0 .3-0 .4V. M e n g et a l 3 0 o b s e r v e d the synerg is t ic effect of the addi t ion of tungsten ca rb ides ( W 2 C / C ) to Pt ca ta lys ts on the O R R in a lka l ine med ia . T h e y exp la in that pure W 2 C with a part icle s i ze of <10 nm c a n be p repared by control l ing the ratios of W to C to less than 20 wt%. T h e y obse rved ten t imes larger current densi ty for P t - W 2 C / C than for P t / C . T h e y did not exp la in the m e c h a n i s m for the improvement of oxygen e lec t roreduct ion. For direct methano l fuel ce l ls ( D M F C s ) a major prob lem is the c r o s s o v e r of methano l f rom a n o d e to ca thode , wh ich depo la r i ses the ca thode . T h e y eva lua ted the catalyt ic effect of W 2 C / C by measu r i ng the kinetic parameters us ing Ta fe l plots with 1 M K O H solut ion, and s tud ied the impact of addi t ion of 0.1 M methano l on O R R at 25°C with a s c a n rate of 1 m V / s . T h e activity of O R R on the P t / W 2 C / C e lec t rode w a s hardly af fected by methano l at concent ra t ions up to 1 M at room temperature; however , the O R R on the P t / C e lect rode w a s 15 ser ious ly af fected by methano l . T h e P t / W 2 C / C exhibi ted a more posi t ive onse t potential of 5 0 m V c o m p a r e d to that of P t / C ( -48mV). A l s o , P t / W 2 C / C exhibi ted a higher e x c h a n g e current dens i ty (io) of 0 .180 x 10" 4 m A / c m 2 than the P t / C cata lys ts with i 0 0 .119 x 1 0 " 5 m A / c m 2 . T h e W 2 C modi f ied Pt not only s h o w e d a synerg is t ic effect to improve the activity for O R R , but a l so immuni ty to methano l . L e e et a l 3 1 s tud ied the stabil i ty and e lectrocata ly t ic activity of W C by addi t ion of T a catalyst . T h e y report that s tud ies have s h o w n the instabil i ty of W C under ac id ic and ox idat ive condi t ions. T h e W C + T a or the pure W C were depos i ted o n g l a s s y ca rbon us ing R F sputter ing. T h e stabil i ty and the e lec t rochemica l activity for the O R R were invest igated in a sol id-state cel l with Naf ion 117 a s the electrolyte. T h e cel l w a s main ta ined under ei ther oxygen or ni trogen a tmosphe res , wh i ch we re humidi f ied by pass ing through a bubbler . T h e X R D pattern had only two p e a k s at 35.8 a n d 4 1 , co r respond ing to W C and T a , respect ive ly . T h e C V ' s of pure W C at 30°C a n d 60°C s h o w e d an anod ic peak above 0.5 V at 30°C, and at 60°C larger anod i c and ca thod ic currents we re o b s e r v e d . T h e C V data s h o w e d that the T a addi t ion to W C s h o w e d no anod i c peak up to 1V at both 30°C and 60°C. T h e oxidat ion of W C by anod i c polar izat ion m a y p roceed a s 3 1 : W C + 5H 2 O = WO3 + C O 2 + 10H + + 10e" Equation 5 L e e et a l 3 1 carr ied out s l ow s c a n vo l tammetry in N 2 and 0 2 a t m o s p h e r e s . Unde r N 2 they obse rved an inc rease in the ca thod ic current for pure W C , wh ich might have been due to partial reduct ion of W(VI ) ox ide to W ( V ) ox ide. T h e i nc rease in ca thod ic current w a s not o b s e r v e d for W C + T a catalyst under the N 2 a tmosphe re . T h e onse t potential for the O R R w a s o b s e r v e d at 0 .8V (vs. S H E ) , wh ich is 0 .35V higher than that for pure W C catalyst . F rom these data they conc lude that T a addi t ion to W C e n h a n c e s its stabil i ty and electrocatalyt ic activity. T h e y c la im that the e n h a n c e d stabil i ty of W C with T a catalyst might be due to the format ion of W - T a al loy in the W C + T a catalyst . Howeve r , a key exper iment us ing a pure T a catalyst for O R R under t hese condi t ions w a s not conduc ted . S tudy ing T a catalyst a lone wou ld have a l lowed a determinat ion of whether it is the T a 2 0 5 format ion on the sur face of W C caus ing the e n h a n c e d e lectrocata ly t ic activity or whether there is an interact ion be tween T a and W C . If T a 2 0 5 is act ive under the condi t ions used by L e e et a l 3 1 then it might be that W C is not requi red; therefore, the resul ts shou ld have been c o m p a r e d to T a catalyst a s the blank. M a et a l 3 2 s tud ied the e lect ro-ox idat ion behav iour of tungsten carb ide e lec t rodes in different e lectro lytes us ing a three-e lec t rode e lec t rochemica l cel l . T h e y s tud ied the anod i c c h a r g e of tungsten carb ide and Pt in 3.5 M HCI . T h e resul ts s h o w e d that the cha rge c o n s u m e d by hyd rogen 16 adsorpt ion and desorp t ion on W C and Pt e lec t rodes w a s 0.17 C / c m 2 and 0.48 C / c m 2 , respect ive ly , prov ing that the tungsten carb ide has lower e lectrocatalyt ic activity for hyd rogen oxidat ion in HCI solut ion than Pt. W h e n they tes ted a W C e lect rode in 2 M H 2 S 0 4 under H 2 , the potent ial i n c reased signi f icant ly a s a c o n s e q u e n c e of the ox idat ion of tungsten ca rb ide . T h e co lor of the e lec t rode a l so c h a n g e d to blue ( W 2 0 5 and W 8 0 2 3 ) and ye l low ( W 0 3 ) . O x y g e n evolut ion and chlor ine evolut ion were a lso o b s e r v e d w h e n H 2 S 0 4 and HCI were u s e d a s the e lect ro ly tes, respect ive ly . T h e y a l so s tud ied the e lect ro-ox idat ion behav iour of tungsten ca rb ide e lec t rodes in 2.5 M K O H solut ion and found that the activity and stabil i ty of W C for hydrogen oxidat ion w a s ex t remely low, and that W C directly ox id ized to W 0 3 in bas i c med ia . T h e W C e lect rode exhibi ts high e lec t rochemica l stabil i ty for hydrogen ionizat ion in ac id ic e lectro lytes w h e n the potential is be low 8 0 0 m V . A b o v e 8 0 0 m V the authors c la im that the overal l e lec t rochemica l react ion of W C in ac id ic so lut ions fo l lows equat ion 5. M a et a l 3 3 a l so prepared and determined the e lectrocatalyt ic propert ies of tungsten ca rb ide e lec t roca ta lys t . . They , report that W C d isp lays "p la t inum- l ike" behav iour , in seve ra l reac t ions . T h e y a lso ment ion that a H 2 - 0 2 fuel cel l s tack with a con t inuous run l ifetime of 5000h at 150°C has been a c h i e v e d us ing H 3 P 0 4 as the electrolyte, Pt a s the ca thode , and W C as the a n o d e . T h e y syn thes i zed tungsten carb ide e lec t rocata lys ts by reduct ion of H 2 W 0 4 under a f lowing C O and C 0 2 mixture with C O : C 0 2 - 1 : 1 0 . T h e s a m p l e s were ca rbur i sed at 750-800°C. After carbur isat ion the W C e lec t rodes with su r face a r e a of 3 c m 2 were p repared by co ld p ress ing the mixtures of tungsten carb ide part icle and P T F E b inder onto metal m e s h subst ra tes . W h e n the mater ial w a s heat- t reated at 700°C, - 7 1 % W 2 C w a s present , and w h e n the temperature w a s inc reased to 750°G, ~ 6 9 % W C . w a s syn thes i zed . T h e W 2 C : W C ratios were ca lcu la ted f rom X R D peak heights. T h e B E T m e a s u r e m e n t s s h o w e d that the s a m p l e s with W C a s the major p h a s e had a higher su r face a rea (25-30m 2 /g ) than the s a m p l e s with W 2 C a s the major p h a s e (5- 10m 2 / g ) . T h e y report that the W 2 C p h a s e of tungsten carb ide exhibi ts higher e lectrocatalyt ic propert ies than W C for the hydrogen anod i c ox idat ion react ion in concent ra ted H 3 P 0 4 so lut ion. However , the cor ros ion resistivity of W 2 C der ived f rom H 2 W 0 4 at e leva ted tempera tures is poor. T h e anod ic propert ies of W 2 C and W C e lec t rodes in 1 2 % HCI solut ion were invest igated. T h e overpotent ia l of W C e lec t rodes w a s found to be m u c h lower than that of W 2 C e lec t rodes . T h e Tafe l s lope of the W C e lect rode w a s twice a s large a s that of the W 2 C e lec t rode, indicat ing the different m e c h a n i s m of hydrogen anod i c oxidat ion for the two e l e c t r o d e s 3 3 . T h e sur face a rea reported by M a et a l 3 3 is low c o m p a r e d to the ava i lab le high sur face a rea ca rbon suppor ts ( - 3 0 0 m 2 /g ) . 17 T h e r e are many patents fi led regard ing use of W or W C a s a n o d e e lec t rocata lys ts in fuel ce l l s . Joe l Chr is t ian et a l 3 4 f rom O s r a m Sy l van ia have fi led numerous patents on the u s e of W conta in ing cata lys ts . T h e syn thes is of catalyst is a c h i e v e d by low temperature e lec t rochemica l act ivat ion of W conta in ing p recursor depos i ted on a C support . A m m o n i u m meta tungstate ( A M T ) w a s d i spe rsed on C a r b o n (Vu lcan X C - 7 2 R ) , and this catalyst p recursor w a s used a s the a n o d e of a P E M F C . T h e format ion of a W-con ta in ing cata lyst layer on a C suppor t w a s conf i rmed by sput tered neutral m a s s spec t roscopy ( S N M S ) and X - ray photoelect ron s p e c t r o s c o p y ( X P S ) . N a 2 C 0 3 (2M) solut ion w a s u s e d as an electrolyte and 0.5-30 V D C w a s app l ied to act ivate the precursor to form W-con ta in ing catalyst . T h e loading of this catalyst p recursor on to C c a n be inc reased by the u s e of a surfactant, cety lpyr id in ium chlor ide. T h e y c la im that power output c a n be i nc reased by 2 0 % relat ive to Pt conta in ing ca ta lys ts . Simi lar ly they c la im that the M E A that they m a d e (5 c m 2 ; 0.5 mg ca ta lys t / cm 2 ) has s imi lar pe r fo rmance to that of Pt cata lyst with a s imi lar load ing. J o e l et a l 3 5 have a l s o been look ing at suppor t ing W C on high su r face a rea c a r b o n in order to m a k e W C conta in ing ca ta lys ts . T h e y syn thes i zed W C mater ia l by react ing a mixture of a tungsten.precursor and a high sur face a rea suppor t in f lowing hydrocarbon with or without H 2 g a s at 500°C-800°C in a tube fu rnace . T h e y report that the compos i t i on they m a d e might be WC-i.x (x = 0 to 0.5). Par t icu late s i z e s of 15 A - 30 A we re a c h i e v e d , with a su r face a rea of 6 5 m 2 / g . Major c la ims were on the p r o c e s s of mak ing W C . T h e appl icat ion of W C catalyst in a P E M fuel cel l w a s not demons t ra ted in this patent. A ma in prob lem with us ing commerc i a l ca rb ides and nitrides is their low spec i f i c su r face a r e a . T h e s e a r c h for me thods of produc ing high su r face a rea has been deve lop ing s i nce 1985, w h e n a patent w a s granted to Boudar t et a l 3 6 for high spec i f i c su r face a rea ca rb ides and nitr ides. T h e des i red ca rb ides and nitrides were p roduced by reduc ing ox ides of the des i red e lemen t in a reduc ing env i ronment at high tempera tures with ca rbon and ni trogen present in this a tmosphe re . T h e y s h o w e d that the B E T sur face a rea of mo l ybdenum carb ide m a d e us ing this method (51 m 2 /g ) w a s at least 4 t imes higher than that of the convent iona l cata lyst (12.5 m 2 /g ) . Th i s method results in i nc reased sur face a rea , but the su r face a rea a c h i e v e d by Boudar t et a l 3 6 is still lower than required for fuel cel l e lec t rode app l ica t ions . C la r i dge et a l 3 7 u s e d the method tested by Boudar t et a l 3 6 to syn thes ize high sur face a rea mo l ybdenum and tungsten carb ide b a s e d cata lys ts . T h e y u s e d tempera tu re -p rog rammed reduct ion ( T P R ) of the re levant ox ides in f lowing methane or me thane /hyd rogen mixtures. T h e format ion of the ca rb ide w a s moni tored us ing X R D . T h e y c la im that very broad p e a k s for the ox ide were s e e n , indicat ing that a very smal l amoun t of ox ide is present . However , usua l ly b road p e a k s m e a n smal le r crystal l i te s i z e . T h e B E T su r face 18 a reas of the s a m p l e s we re 30 m 2 / g for M o 2 G and 20 m 2 / g for W C ; however , t hese su r face a reas are still sma l l for fuel cel l appl icat ion. T h o m p s o n et a l 3 8 in their granted patent c l a imed u s e of noble metal cata lys ts suppor ted on electr ical ly conduc t i ve transit ion metal (G roup IV to VI) b a s e d c e r a m i c s s u c h a s ca rb ides , nitr ides, bor ides, or s i l ic ides. T h e y s h o w e d e x a m p l e s us ing W C , for wh ich X R D data were acqu i red but no fuel cel l da ta were s h o w n . T h e des i red sur face a rea for the suppor t w a s approx imate ly 40 m 2 / g , wh ich is s igni f icant ly lower than the sur face a r e a of the current ca rbon suppor t ( - 300 m 2 /g ) being u s e d ex tens ive ly for P E M F C s . 1.3.2 Summary U s e of tungsten carb ide ( W C ) a s a fuel cel l cata lyst h a s been m a d e in the past, but detai led s tudy on its use has not been reported a s a cata lyst suppor t in P E M F C s . Deta i led tests involv ing oxidat ion cyc l es under s imula ted or ac tua l fuel cel l condi t ions need to be conduc ted in order to obse rve the oxidat ion stabil i ty of tungsten carb ide . T h e possibi l i ty of tungsten carb ide ox id iz ing to tungsten ox ide and the condi t ions under wh ich this h a p p e n s need to be veri f ied in order to u s e W C a s a cata lyst suppor t in P E M F C s . H igh sur face a rea ca rb ides n e e d to be syn thes i zed before they c a n be u s e d in P E M F C s . T h e impact on stabil i ty and per fo rmance f rom us ing W C or W 2 C p h a s e s a l so n e e d s to be de te rmined. Al ternat ive suppor ts s u c h a s conduct ing ox ides are an emerg ing ca tegory of oxidat ion resistant catalyst suppor ts , and are a lso of interest in this work. T h e u s e of conduc t ing ox ides in fuel ce l ls by severa l g roups is desc r ibed in the fo l lowing sec t ion . 1.4 Oxides Meta l ox ides are one of the most important ca tegor ies of sol id ca ta lys ts or catalyst s u p p o r t s 3 9 . Meta l ox ides s u c h a s T i 0 2 , and l n 2 0 3 a re n-type sem iconduc to rs . T in doped l n 2 0 3 ( ITO) is a w ide ly u s e d t ransparent conduct ing ox ide ( T C O ) . The re are only a f ew s tud ies conduc ted us ing N b - d o p e d titania a s a cata lyst support . T h e key f ind ings of t hese s tud ies are rev iewed here. It is be l ieved that s i nce dop ing with al iovalent ions c a n e n h a n c e conduct iv i ty, both N b - d o p e d titania and ITO might be suff iciently electr ical ly conduc t i ve to be used a s cata lyst suppor ts . 19 Sing le cat ion s to ich iometr ic ox ides s u c h a s t i tanium diox ide ( T i 0 2 ) a re resist ive; however , t i tanium sub-ox ide , inc luding T i 4 0 7 and other p h a s e s , exhibit e lec t ron ic conduct iv i ty. T h e d rawback with t i tanium sub-ox ide is that under fuel cel l operat ion it b e c o m e s sto ich iometr ic and forms a res ist ive layer of T i 0 2 at the t h ree -phase react ion in te r face 4 0 . C h e n et a l 4 0 have s tud ied N b doped rutile, T io . 9 Nbo . i0 2 . Th i s mater ia l has many benef ic ia l charac ter is t ics for use a s a catalyst suppor t in P E M F C s . T h e y u s e d E b o n e x (At raverda Ltd. , U.K. ) , wh ich is an electr ical ly conduct ive c e r a m i c cons is t ing of severa l subox ides of t i tanium diox ide, main ly T i 4 0 7 and T i 5 O g . In spi te of the p r e s e n c e of reduced t i tanium, E b o n e x is e lec t rochemica l l y s tab le in both ac id ic and bas i c so lut ions. It a l s o h a s a high conduct iv i ty ( - 1 0 3 S / c m ) a n d g o o d co r ros ion res is tance . It is wel l known that the e lect ron ic conduct iv i ty of t i tan ium-based c e r a m i c s or ig inates f rom the p r e s e n c e of T i 3 + ions. The re are two w a y s to c reate the T i 3 + ions in the rutile structure: by creat ing oxygen v a c a n c i e s by heat ing T i 0 2 in a reduc ing a tmosphe re , or by introducing appropr ia te donor dopants , e .g . , N b . T h e y syn thes i zed two conduc t i ve t i tanium c e r a m i c s , T i 4 0 7 and T i 0 9 N b 0 . i O 2 , made , respect ive ly , by reduc ing and dop ing rutile t i tanium d iox ide. Both c e r a m i c - s a m p l e s - w e r e dark blue in color, and had an electr ical conduct iv i ty in the range of 0.2-1.5 S / c m (two-point m e a s u r e m e n t technique) . B E T su r face a rea w a s relatively low for all three suppor ts , 2 and 1.4 m 2 / g for the syn thes i zed T i 4 0 7 and Tio.9NBO.1O2, respect ive ly , and 1 m 2 / g for E b o n e x . T h e C V s of the c e r a m i c s s h o w e d a w ide potential w indow of s tabi l i ty . ranging f rom -0.4 to 2 V v s . R H E , whi le V u l c a n X C - 7 2 R ca rbon , a c o m m o n l y u s e d e lect rocata lyst support , expe r i ences oxidat ion current at posi t ive potent ials a s low a s 1.0 V . T h e s e suppor ts s e e m promis ing b a s e d on the e lec t rochemica l stabil i ty da ta ; however , l i fetime data are still n e e d e d to c o m p a r e with convent iona l ca rbon suppor ts to de termine the benef i ts of the ox ide suppor ts . T h e T i 4 0 7 - s u p p o r t e d cata lyst s h o w e d catalyt ic activity very s imi lar to that of the E b o n e x suppor ted cata lyst for O R R , whi le the cata lyst suppor ted on the T i 0 9 N b o . i 0 2 exhibi ted h igher currents at al l app l ied vo l tages . T h e cata ly t ic activity of E b o n e x - a n d T i 4 0 7 - s u p p o r t e d ca ta lys ts d e c r e a s e d with operat ion, and the authors exp la ined that these non-s to ich iometr ic ox ides ox id ized to resist ive s to ich iometr ic ox ides , result ing in a greater internal iR drop. T G A data s h o w e d that both E b o n e x and T i 4 0 7 a re ox id ized to T i 0 2 , beg inn ing at about 4 0 0 ° C and cont inuing to 600°C, w h e r e a s T i 0 . 9 N b 0 . i O 2 d o e s not s h o w a signif icant weight ga in up to 1000°C. After heat ing in air for 20 h at 500°C, E b o n e x and T i 4 0 7 turned white and their conduct iv i ty d e c r e a s e d by at least f ive orders of magni tude, cons is ten t with comp le te oxidat ion to a T i (IV) ox ide. Unde r the s a m e condi t ions, the conduct iv i ty of T i 0 . 9 N b 0 . 1 O 2 d e c r e a s e d by approx imate ly 0 . 1 % of its initial va lue and its color c h a n g e d gradual ly f rom deep blue to b lue-gray. N b d o p e d 20 t i tania is both thermal ly and e lec t rochemica l l l y s tab le c o m p a r e d to reduced t i tania; however , the syn thes is method used by C h e n et a l 4 0 d o e s not provide a high sur face a rea support . So l -ge l der ived suppor ts m a y prov ide a higher su r face a r e a . A n al ternat ive d o p e d ox ide is ITO. Detai ls on the s tud ies conduc ted by other resea rche rs on ITO are rev iewed here. ITO is w ide ly used a s a t ransparent conduc t ing ox ide for smart w indows . A conduct iv i ty of 1 0 4 S / c m is often quoted for an op t im ized ITO f i lm. S tudy of this mater ial a s a cata lyst suppor t has not been ex tens ive . E lec t ron ic character is t ics studied by other resea rch g roups are rev iewed here. However , b e c a u s e it is an n-type semiconduc to r , it m a y exhibit s imi lar behav io r to that of N b d o p e d t i tania. L iu et a l 4 1 p repared ITO f i lms by so l -ge l d ip coat ing on g l a s s subst ra tes . Hal l effect m e a s u r e m e n t s s h o w e d an i nc rease in e lectr ical carr ier concent ra t ion with inc reas ing S n 0 2 content. T h e effect of the S n 0 2 content on the shee t res is tance of 70 nm thick ITO mono laye rs s h o w e d that the shee t res i s tance w a s a m in imum at ~11 m o l % S n 0 2 . M a t v e e v a 4 2 s tud ied the e lec t rochemica l behav io r of ITO in ac id and b a s e electro lytes and noted that it d i sso l ves , espec ia l l y at pH less than 1. M a t v e e v a 4 2 found that in 1 M N a O H dur ing ca thod ic polar izat ion, the e lec t rode c o m p o n e n t s are deep ly reduced , s o that ITO is gradual ly and i rreversibly conver ted to a metal l ic mirror with a not iceable d e c r e a s e of oxygen content. At a high anod i c current densi ty , the ITO e lect rode u n d e r g o e s modi f icat ions and its conduct iv i ty d e c r e a s e s , a l so due to the c h a n g e of oxygen content in the ox ide latt ice. M a t v e e v a 4 2 u s e d e lec t rochemica l condi t ions that wou ld a l low the character iza t ion of ITO for smart w indows . T h e e lec t rochemica l stabil i ty of ITO a s a catalyst support has not b e e n s tud ied. M a s o n et a l 4 3 found that up to 6-cat ion % S n w a s so lub le in l n 2 0 3 , and the result ing mater ia ls we re conduc t i ve with high electron popu la t ions . T h e y hypo thes i zed the ex is tence of neutral reduc ib le (2Sn ' i n O" i ) x a s s o c i a t e s , wh ich form upon dop ing of l n 2 0 3 . 2 S n 0 2 => (2Sn i„0" i ) x + 3 0 x 0 Equation 6 T h e (2Sn ' i n O"i) x c luster c a n be reduced at lower tempera tures to r emove the oxygen interstit ials. It m a y be that (2Sn ' i n O"j) x a s s o c i a t e s p lay a major role in the defect chemis t ry of ITO c o m p a r e d to the intrinsic de fec ts (i.e. oxygen v a c a n c i e s ) . T h e s e a s s o c i a t e s m a y be respons ib le for the high concent ra t ions of S n that c a n be incorporated into ITO. O n c e incorporated, the e x c e s s S n t ied up in neutral a s s o c i a t e s s e r v e s as a reservoi r for the product ion of addi t ional f ree S n - d o n o r s upon remova l of o x y g e n interstitials dur ing reduct ion. 21 Within the ox ide category, the use of zeo l i tes a s catalyst suppor ts has a l so been repor ted. S c e l e l 4 4 f i led a patent in wh ich use of noble metal cata lys ts suppor ted on conduc t i ve zeol i te part iculate mater ial a s catalyst suppor t is c l a i m e d . Th i s suppor t mater ia l u s e s 65 wt% C and 35 wt% conduct ive zeol i te . T h e y c la im that the zeol i te preferably h a s a su r face a rea of be tween about 100 to about 4 0 0 m 2 / g and C be tween 10 to about 50 m 2 / g . T h e zeo l i tes have ac id ic protonic enti t ies on the sur face , mak ing them hydrophi l ic relat ive to ca rbon . Zeo l i tes cons is t of 1, 2, or 3 d imens iona l channe l s , and conduc t i ve po lymers m a y fill t hese . Th i s patent is b a s e d o n a concept , and no proof of concep t has been a c h i e v e d . T h e r e are no data to prove if the des i red sur face a rea has been a c h i e v e d , whether zeo l i tes are suff iciently conduct ive , what is the impact of hav ing 3 5 % carbon in the suppor t mater ia l , and f inal ly whether this mater ial c a n wi thstand the fuel cel l condi t ions. 1.4.1 Summary ITO is electr ical ly conduc t i ve and is relat ively s tab le in ac id with p H greater than 1. A s exp la ined above , dop ing with pentava lent ions c a n e n h a n c e the e lect ron ic conduct iv i ty of tetravalent meta l ox ides . O x y g e n def ic ient ox ides may .no t be s tab le in P E M F C s , and therefore dop ing is o n e of the methods to s tabi l ize the e lect ronic conduct iv i ty . In the N b d o p e d titania s y s t e m , l i fetime durabil i ty test ing still n e e d s to be conduc ted and the sur face a r e a n e e d s to be i nc reased before it c a n rep lace ca rbon in w i d e s p r e a d use . ITO prepared v ia so l -ge l me thods is a good cand ida te for non-ca rbon catalyst suppor ts for P E M F C s ; however , e lec t rochemica l stabil i ty and l ifetime da ta are still not ava i lab le for this mater ia l . T h e r e has been no pub l i shed work showing the e lec t rochemica l stabil i ty of ITO a s a catalyst support . 1.5 Conc lus ions C a r b o n will ox id ize at potent ia ls a b o v e 0.2 V at 25°C, and add ing dopants to ca rbon only de lays the e lec t rochemica l ox idat ion but d o e s not prevent it. Little r e s e a r c h , however , has b e e n per formed to date on the u s e of non-ca rbon catalyst suppor ts . A l ternat ive cata lyst suppor ts s u c h a s tungsten carb ide and Indium tin ox ide c a n be u s e d as cata lyst suppor t mater ia ls in P E M F C s . H igh sur face a rea ca rb ides and ox ides are not yet c o m m e r c i a l l y ava i lab le . Deta i led tests involving oxidat ion cyc l es under s imula ted or actual fuel cel l cond i t ions need to be conduc ted in order to obse rve the ox idat ion stabil i ty of tungsten carb ide and ITO. T h e r e has been no pub l ished work showing the e lec t rochemica l stabil i ty of ITO and tungsten ca rb ide a s al ternat ive catalyst suppor ts for P E M F C s . 22 2 Introduction to the three electrode electrochemical cell, voltammetry and rotating disc electrode (RDE) 2.1 Three electrode electrochemical cell A typical th ree-e lec t rode e lec t rochemica l s ys tem cons is t s of a work ing, re ference and counter e lec t rode i m m e r s e d in an electrolyte so lut ion. T h e e lec t rochemica l react ion occu rs at the work ing e lec t rode ( W E ) , leading to an e lect ron transfer. Th i s e lect ron t ransfer genera tes an e lectr ica l current te rmed F a r a d a i c current. A potentiostat, wh i ch contro ls the potent ia l , i m p o s e s a cyc l i c l inear s w e e p on the work ing e lect rode, f rom wh ich a current-potent ial cu rve is ob ta ined . A n auxi l iary or counter e lec t rode (AE) is dr iven by the potentiostat to ba lance the F a r a d a i c p r o c e s s at the W E with a n e lect ron t ransfer of oppos i te d i rect ion. Th i s c a u s e s oxidat ion to take p lace at the A E if reduct ion o c c u r s at the W E . T h e re ference e lect rode prov ides a f ixed potent ial , wh ich d o e s not vary dur ing the exper iment (potential shou ld be independent of the current densi ty) . Potent ia l of the re fe rence e lec t rode is known a n d is kept constant . T h e potent ial be tween the W E a n d the R E is contro l led by a potentiostat and any c h a n g e in app l ied potential to the cel l appea rs directly a c r o s s the W E - s o l u t i o n interface. T h e F a r a d a i c current at the W E is t ransduced to a potential output at a part icular sensit iv i ty (AA/ ) and reco rded in a digital or ana log fo rm. 2.2 Voltammetry Vo l tammet ry is an e lect roanalyt ica l method a l lowing eva luat ion of an analy te by measu r i ng current a s a funct ion of app l ied potent ial . Vo l tammet ry is used for fundamenta l s tud ies of ox idat ion and reduct ion p r o c e s s e s in var ious med ia , adsorp t ion on su r faces , and e lectron t ransfer m e c h a n i s m s o n e lec t rode s u r f a c e s 4 5 . In vo l tammtery, va r ious exci tat ion s igna l wave fo rms a re poss ib le . L i n e a r - s c a n vol tammetry, wh ich is a potent ial vs . t ime wave fo rm , is most c o m m o n l y u s e d . 2.2.1 Cycl ic voltammetry (CV) C V is a potential control led e lec t rochemica l exper iment , in wh ich the direct ion of the potential is reversed at the end of the first s c a n . A cyc l i c potential is usua l ly i m p o s e d on an e lect rode and the current r e s p o n s e is m e a s u r e d . Revers ib i l i ty is advan tageous , s i nce the product of the e lec t ron 23 t ransfer react ion that occur red in the forward s c a n c a n be p robed aga in in the reverse s c a n . It is a powerful tool for determin ing formal redox potent ials, detect ing chem ica l react ions that p recede or fol low the e lec t rochemica l react ion, and evaluat ing e lectron t ransfer k inet ics. A C V is usua l ly plotted a s current vs . potent ial . F igure 4 s h o w s a typical C V of Pt in 1.0 m o l / d m 3 H 2 S 0 4 at 25°C . Four major reg ions are evident f rom the C V . In forward s w e e p : 1) Ox idat ion of a d s o r b e d H 2 at 0.0 to about +400 m V , with the twin p e a k s co r respond ing to weak l y bound and strongly bound (at h igher posi t ive potent ials) hydrogen a toms . 2) In the center of the vo l tammetr ic curve is a region whe re only low currents (posi t ive anod i c for the posi t ive s w e e p and negat ive for the negat ive s w e e p ) c a n be found . T h i s is the doub le- layer region where only capac i t i ve p r o c e s s e s take p lace . 3) Ox ide film format ion o c c u r s at about +750 m V and cont inues to potent ials a b o v e +2000 m V . 4) O x y g e n gas evolut ion starts at ~ + 1 5 0 0 m V . In reverse s w e e p : 1) O x i d e reduct ion is o b s e r v e d b e l o w + 1 0 0 0 m V . 2) Doub le layer reg ion is fo l lowed by 3) Hyd rogen adsorpt ion at +400 m V and then 4) Hyd rogen evolut ion at 0 .0V 24 Electrode Potential /mV vs. R H E Figure 4: Cycl ic voltammogram of Pt G D E in 1.0 mol /dm 3 in H 2 S 0 4 at 25°C , v = 30mV/s 2.3 Convect ion and rotating disc electrode (RDE) Convec t i on a l lows transport of s p e c i e s due to external mechan i ca l fo rces s u c h a s mov ing of the e lec t rode or stirring of the solut ion. Convec t i on is an important form of def ined and reproduc ib le m a s s transport; dur ing wh ich current dens i t ies 3-100 t imes greater than the s teady state di f fusion l imited va lue are c o m m o n 4 5 . T h e rotating d i sc e lec t rode is the most popular s ys tem for kinetic and mechan is t i c s tud ies. A n R D E cons i s t s of a work ing e lec t rode mater ia l (usual ly g lassy ca rbon or Pt) e n c l o s e d in a Tef lon or ce ram ic shea th . A n R D E w h e n rotated in a solut ion ac ts a s a pump, pull ing the solut ion vert ical ly upwards towards the d i sc and then throwing jt outwards. T h e key advan tage with us ing an R D E is that the rate of m a s s transport to the e lec t rode may be var ied over a substant ia l range and in a control led way , without rapid c h a n g e in the e lect rode potential . S o m e of the p rob lems are that the solut ion c a n leak into any gap be tween the act ive d isk mater ia l and the insulat ing shea th . A l s o , no ise f rom poor e lectr ical con tac ts c a n lead to a p rob lem. T o avo id this p rob lem the shaft of the R D E is 25 usual ly direct ly l inked to the motor dr ive, and the contact is m a d e with a h igh qual i ty ca rbon brush contact ( A g / C material) . 3 Objectives A s stated above , acu te degradat ion of catalyst suppor ts in P E M F C s is h inder ing the commerc ia l i za t ion of t hese fuel ce l ls . T h e object ive of this s tudy is to eva lua te two types of ox idat ion-res is tant cata lyst suppor ts for P E M F C s . T h e object ives for t hese invest igat ions involve d ispers ing Pt on commerc i a l tungsten carb ide and indium tin ox ide ( ITO) and then determin ing the e lec t rochemica l and thermal stabi l i t ies of t hese mater ia ls . T h e Pt d i spers ion methods u s e d for commerc ia l mater ia ls we re a l so used to d i spe rse Pt on wide ly used ca rbon (Vu lcan X C - 7 2 R , Cabo t ) in order to c o m p a r e the activity of mater ia ls p repared by s imi lar me thods . 4 Experimental Procedure Pt w a s suppor ted on both W C and ITO, and the e lec t rochemica l stabil i ty of the suppor ted cata lyst w a s determined us ing ox idat ion cyc l es and cyc l i c vo l tammetry. T h e convent iona l method of Pt d ispers ion (referred to a s method I for Pt addit ion) by hydroxylat ion of chlorplat in ic ac id with a b a s e s u c h as sod ium hydrogen ca rbona te fo l lowed by reduct ion us ing fo rma ldehyde w a s not success fu l to d i spe rse Pt on W C . Pt addi t ion by method I invo lved us ing sod ium hydrogen carbona te , wh ich reacted with W C and fo rmed f lakes of unknown mater ia l . There fo re , an al ternat ive method n e e d e d to be d e s i g n e d . Pt addi t ion us ing Pt (II) pentan-2,4-d ionate precursor (referred to as method II for Pt addit ion) w a s d e v e l o p e d in -house, lead ing to a s u c c e s s f u l Pt d ispers ion onto W C . 4.1 Pt addition using chlorplatinic acid (method I) 3.44g N a H C 0 3 w a s d i sso l ved in 200ml H 2 0 in a 500ml round bottom f lask. T o this w a s a d d e d 0 .60g of the catalyst support . T h e mixture w a s ref luxed for severa l hours to ensu re comp le te wett ing of the support . Us ing an addi t ion funne l , 1g H 2 P t C I 6 d i sso l ved in 60ml H 2 0 w a s a d d e d drop-w ise over severa l minutes. T h e mixture w a s aga in a l lowed to reflux for two hours . 780(j.l f o rma ldehyde (methanal H C H O ) solut ion (37%) in 7.8ml H 2 0 w a s a d d e d by addi t ion funnel ove r 26 approx imate ly o n e minute. T h e mixture w a s a l lowed to react and reflux overnight before fi ltering, wash ing with water, dry ing, and gr inding. 4.2 Pt addition using Pt (II) pentan-2,4-dionate (method II) Pt (II) pentan-2 ,4-d ionate (Alfa A e s a r , - 4 8 wt% Pt) w a s used a s the Pt p recursor and then later reduced under hydrogen a tmosphe res . Pt (II) pentan-2,4-d ionate w a s found to be inso lub le in water, s o acetonitr i le w a s u s e d . Pt (II) pentan-2 ,4-d ionate T o obtain 4 0 wt% Pt on W C , 0.50 g of P t (II) pentan-2 ,4-d ionate w a s d i sso l ved into 110 m L acetonitr i le: T o this w a s a d d e d 0.36 g of A l fa A e s a r W C . T h e so lvent w a s a l lowed to evapora te and the result ing sol id w a s heat t reated in a tube fu rnace at 600°C for 5 hours under 20 v o l % A r / b a l a n c e H 2 . 4.3 Pt addition to equal powder volumes of support (for W C studies only) Th is p rocedure invo lves d ispers ing s imi lar a m o u n t s of Pt onto suppor ts that have a s igni f icant densi ty d i f ference; e.g. , ca rbon c o m p a r e d to W C , wh ich have dens i t ies of 1.8 g / c m 3 and 15.6 g / c m 3 , respect ive ly . Usua l l y 40 wt% Pt is d i s p e r s e d on ca rbon , but s ince the densi ty d i f ference between W C and C is large, direct c o m p a r i s o n be tween 4 0 wt% Pt on W C and 40 wt% Pt on C d o e s not prov ide suff icient information to direct ly c o m p a r e activity leve ls and even degradat ion rates. There fore , an al ternat ive method w a s used in wh ich 40 wt% Pt w a s d i spe rsed on C an a s imi lar vo lume fract ion of Pt w a s d i spe rsed on the W C suppor t mater ia l , us ing tapped powder dens i t ies for c o m p a r i s o n . T a p densi ty is the densi ty of a powder w h e n the vo lume receptac le is tapped or v ibrated under spec i f ied cond i t ions whi le be ing l oaded . T a p dens i ty w a s u s e d to m e a s u r e s imi lar p o w d e r v o l u m e s of the suppor t mater ia ls. T h e tap densi ty method involved pack ing 2 ml of a suppor t powde r in a g l a s s cy l inder. T h e cy l inder w a s tapped o n a coun te r top 2 0 0 t imes until 2 m l of wel l p a c k e d suppor t mater ia l w a s ach ieved . Both ca rbon and tungsten ca rb ide were tapped to a 27 vo lume of 2 m L A solut ion of Pt (II) pentan-2,4-d ionate (Alfa A e s a r , - 4 8 wt% Pt) d i sso l ved in 100 ml acetonitr i le w a s p repared . Fo r . ca rbon , 40 wt% Pt w a s a d d e d , and a s imi lar amoun t of Pt as that for ca rbon w a s a d d e d to W C . T h e Pt (II) solut ion w a s a d d e d to the 2 ml p a c k e d suppor t mater ia l , and the solut ion w a s then hea ted to dry. T h e result ing sol id w a s heat t reated in a tube fu rnace at 600°C for 5 hours under 20 v o l % A r /ba lance H 2 . T h e heat t reatment a l lowed Pt (II) reduct ion to Pt meta l . 4.4 Pt addition to carbon and tungsten carbide by solid density method (for W C studies only) In sec t ion 4 .3 ca rbon w a s tapped to a vo l ume of 2 ml with a m a s s of 0 .1140g . Us ing the sol id dens i t ies of ca rbon and W C , 1.8 g / c m 3 and 15.6 g / c m 3 , respect ive ly , the amoun t of W C having the s a m e sol id vo lume w a s ca lcu la ted to be 0 .99 g. T o 0.99 g of W C the s a m e amount of Pt w a s a d d e d a s in sec t ion 4 .3 (40wt% Pt relat ive to carbon) . Both the Pt addi t ion and reduct ion we re carr ied out a s exp la ined in sec t ion 4 .3 . 4.5 Thermogravimetric analysis (TGA) Thermograv imet r i c ana lys i s ( T G A ) w a s u s e d to determine, the stabil i ty of the convent iona l P t / C , P t / W C and Pt / ITO to oxidat ion in air (40 ml/min) as the temperature w a s ramped f rom 50°C to 1000°C at 2°C/min . T h e s a m p l e s were held at 50°C for five minutes to a l low more t ime for water to be r emoved gently. T h e s a m e procedure w a s a lso used to test cata lys ts conta in ing 40 wt% Pt on V u l c a n X C - 7 2 R and commerc ia l l y ava i lab le H i s p e c 4 0 0 0 (40 wt% Pt o n V u l c a n X C - 7 2 R , J o h n s o n Mat they) . T h e data were a n a l y s e d by either plotting the der ivat ive of the m a s s (dm/dT) a s a funct ion of s a m p l e temperature, or by plotting the normal ized weight a s a funct ion of s a m p l e temperature. 4.6 Rotating disc electrode (RDE) and electrochemical test set-up T o test the e lec t rochemica l stabil i ty of the catalyst suppor ts , 2 0 m g of e a c h suppor ted cata lyst w a s d i spe rsed in 2ml of g lac ia l e thano ic ac id us ing u l t rasound. Us ing a micropipet te, 5(xl of the s u s p e n s i o n w a s d i s p e n s e d onto the flat su r face of a po l ished g l a s s y ca rbon ( G C ) rotating d i sc e lec t rode ( R D E ) . T h e so lvent w a s r emoved gent ly with a hot air b lower, leav ing suppor ted 28 catalyst (50>g) on the d i sc . T h e commerc i a l 5 v o l % a lcoho l ic Naf ion™ (DuPont ) , with E W of 1100 w a s di luted by addi t ion of 0.5 ml of 5 v o l % a lcoho l i c Naf ion™ in 5 ml p ropono l . Us ing a micropipette, 5JJ.I of di luted a lcoho l ic Na f ion™ w a s d i s p e n s e d onto the d isc . T h e so lvent w a s a l lowed to s lowly evapora te in still air in a g l ass enc losu re s o that a coheren t Naf ion™ film w a s cas t over the cata lyst and the d isc . T h e R D E w a s then i m m e r s e d in deoxygena ted 0 . 5 M H 2 S 0 4 at 30°C and rotated at 3 3 . 3 3 H z (2000rpm). T h e e lec t rochemica l cel l compr i sed a g lass work ing compar tment with a water jacket connec ted to a c i rculat ing water bath and two s ide compar tmen ts : one conta in ing a Pt g a u z e counter e lec t rode connec ted by a g lass frit, and the other conta in ing the revers ib le hydrogen e lec t rode ( R H E ) connec ted by a Lugg in capi l lary. B a s e d on prel iminary vo l tage cyc l ing exper imenta l resul ts, the oxidat ion potential c h o s e n for the e lec t rochemica l cyc l ing tests w a s +1.8V vs . R H E . A b o v e +1.8V, cons ide rab le g a s w a s evo l ved , wh ich sepa ra ted the cata lyst / Naf ion™ depos i t f rom the d i sc . At lower potent ia ls, the ox idat ion w a s not detec tab le in a sui table exper imenta l t imeframe. T h e oxidat ion cyc l ing p rocedure w a s a s fo l lows. Us ing an E G & G 2 6 3 ( P A R , P r ince ton , N J ) potentiostat with Co r rwa re sof tware (Scr ibner A s s o c i a t e s ) , potential s teps (oxidat ion cyc les ) be tween 0.6 V and + 1.8V were app l ied . T h e e lec t rode w a s held at 0 .6V for 60 s e c o n d s and at 1.8 V for 20 s e c o n d s . A C y c l i c V o l t a m m o g r a m ( C V ) w a s reco rded be tween 0 .0V and 1.4V at 1 0 0 m V / s before the ox idat ion cyc l es began and then aga in after every 10 oxidat ion cyc l es , until a total of 100 oxidat ion cyc l es had been app l ied . T h r e e separa te tests were comple ted for e a c h s a m p l e to o b s e r v e repeatabi l i ty. 4.7 X-Ray Diffraction (XRD) X - R a y Diffraction ( X R D ) w a s used to de termine the p r e s e n c e of crystal l ine Pt on the suppor ts and the a v e r a g e crystal l i te s i z e s . Crystal l i te s i z e s of the Pt and of the suppor ts were ca lcu la ted us ing the Sche r re r e q u a t i o n 4 6 : t = 0.9 (Wb)cos6 b Equation 7 where t = crystal l i te s i ze in A , X is the wave leng th , (1 .5406 A in this c a s e for C u K a radiation), b is the full-width at half max imum ( F W H M ) of a peak in the X R D spec t rum, and 0 b is the diffraction ang le for that peak . 29 4.8 Scanning electron microscopy (SEM)/Transmission electron microscopy (TEM) and Energy Dispersive X-Rays (EDX) A 2 0 0 k V Hi tachi H-800 T E M , with the Quar t z X O n e E D X sys tem and a S - 4 7 0 0 F E S E M (Fie ld E m i s s i o n S c a n n i n g E lec t ron M i c roscope ) were used to charac te r i ze the suppor ted and unsuppor ted mater ia ls . E lemen ta l m a p s f rom E D X were used to charac te r i ze the d ispers ion of Pt on a support . 5 Determining potential and time for electrochemical stability tests S i n c e acute degradat ion of catalyst suppor t is obse rved dur ing s tar t -up/shutdown cyc l es for a P E M F C stack, there w a s a need to determine the condi t ions for ex-s i tu acce le ra ted tests s o that the stabil i ty of var ious catalyst suppor ts cou ld be tes ted. T h e ex-s i tu tests involved prepar ing R D E s a m p l e s a s exp la ined in the exper imenta l p rocedure sec t ion . Fo r the exper iments . invo lv ing the determinat ion of the potential and t ime to be used for e lec t rochemica l stabil i ty test ing, commerc i a l catalyst H i s p e c 4 0 0 0 (40 wt% Pt suppor ted on V u l c a n X C - 7 2 R , J o h n s o n Mat they) w a s u s e d . Ox idat ion cyc l es we re app l ied with different potent ials and t imes a s s h o w n in T a b l e 1. For examp le , the R D E w a s held at +0.6V for 60s and then the potential w a s s tepped up to +1.8V, where it w a s held for 20s , and then the potential w a s s tepped down to +0.6V. A total of 100 oxidat ion cyc l es were app l ied . C y c l i c vo l t ammograms ( C V s ) were reco rded before any ox idat ion cyc les and then after every 10 oxidat ion cyc l es . T h e C V s were recorded be tween 0 .0V and 1.4V at l O O m V / s . F igure 5 s h o w s an e x a m p l e of the resul ts of oxidat ion c y c l e s f rom +0.6 to +1.8V for H i s p e c 4 0 0 0 . T h e s e data we re used to ca lcu la te the normal ized activity (F igure 6), wh ich w a s ca lcu la ted by record ing the last current point f rom the data set at 1.8V just before the current b e c a m e negat ive under 0.6 V cond i t ions . T h e s e points are marked with ar rows a s points 1, 2, 3, etc. in F igure 5. T h e current at point 1 w a s taken as the initial current for the normal ized activity plots. T h e current d e c r e a s e d a s the cata lyst support ox id i zed ; therefore, the currents s u b s e q u e n t to point 1 were normal ized to the initial current va lue, and the cu rves in F igure 6 we re plotted a s the norma l i zed activity vs . the cumula t ive number of oxidat ion c y c l e s . C o m p a r i n g H i s p e c 4 0 0 0 be ing held at 1.8V for 60s , 40s and 20s s h o w s that the 1.8 V (60s) condi t ion is too destruct ive, a s a lmost all of the suppor t is ox id ized after on ly 4 ox idat ion c y c l e s . A condi t ion that wou ld a l low a drop in act ivi ty that c a n be detec ted over a " reasonab le " range of 30 exper imenta l condi t ions wou ld be preferred s o that the stabil i ty of var ious suppor ts c a n be c o m p a r e d . W h e n the condi t ions at 1.8V (60s) and 1.8 (40s) were u s e d , the activity d ropped rapidly; therefore, if the e lec t rode is held at 1.8V, a holding t ime of 20s is preferred. W h e n the e lec t rode w a s held at 1.5 V for 60 or 20 s e c o n d s , the ox idat ion of the suppor t w a s m u c h s lower , requir ing longer exper imenta l t imes to de termine the e lec t rochemica l stabil ity. There fore , the 1.8V (20s) condi t ion w a s preferred over the condi t ions us ing 1.8 V (60s or 40s ) and 1.5V (60s or 20s) . Hold time Hold time Lower at lower Higher at higher potential potential potential potential Sample (V) (s) (V) (s) 1 0.6 60 1.8 60 2 0.6 60 1.8 40 3 0.6 60 1.8 20 4 . 0.6 60 1.5 60 5 0.6 60 1.5 20 Table 1: Parameters for determining condit ions for electrochemical stability 31 Time (s) Figure 5: Current vs . time plot obtained from oxidation cycles for Hispec 4000 cycled between +0.6 V (held for 60s) and +1.8V (held for 20s); 0.5 M H 2 S 0 4 , 30°C , 100 mV/s, 2000 R P M . 32 1.1 n j 2 3 4 5 6 7 8 9 1 0 11 -0.1 -I No. of oxidation cycles Figure 6: Normalized activity at different potentials and times as a result of repeated cycling for Hispec 4000. Average of three samples with limits of error for each condition are plotted. 6 Results and d iscuss ion for tungsten carbide studies Oxidat ion stabil i ty of commerc ia l tungsten ca rb ide (Alfa A e s a r ) w a s eva lua ted in order to de termine the viabil ity of this mater ia l to be u s e d a s an ox idat ion-res is tant cata lyst suppor t for P E M F C s . 6.1 Alfa Aesar W C with Pt deposition using chlorplatinic acid In order to charac te r i ze P t / W C , 40 wt% Pt w a s d i spe rsed on W C . A n X R D pattern of Pt d i spe rsed on V u l c a n X C - 7 2 R is s h o w n in F igure 7. T h e a v e r a g e crystal l i te s i ze for Pt w a s ca lcu la ted us ing the Sche r re r equat ion. T h e a v e r a g e tungsten carb ide crystal l i te s i ze w a s ca lcu la ted to be 11.3 nm from the p e a k s at 29: 39.8° , 46 .1° , and 67.6° . Pt w a s depos i ted us ing method I a s exp la ined in the exper imenta l sec t ion . Af ter P t depos i t ion on W C us ing method I, c lea r f lakes we re o b s e r v e d in the samp le , wh i ch d id not d i sso l ve in water after w a s h i n g the s a m p l e severa l t imes. X R D did not s h o w any p e a k s other than the Pt and W C peaks . There fo re , the f lakes we re likely amorphous . T o determine whether these f lakes a p p e a r e d due to W C 33 react ing with the N a H C 0 3 , a blank react ion whe re W C w a s ref luxed in N a H C 0 3 w a s conduc ted , and these f lakes we re o n c e aga in o b s e r v e d . T h e X R D of W C ref luxed in the solut ion of N a H C 0 3 s h o w s the p r e s e n c e of W C with a broad peak at 2-theta of - 2 3 ° , s igni fy ing the p r e s e n c e of an a m o r p h o u s p h a s e 4 7 (F igure 9). M a et a l 3 2 eva lua ted e lectro-oxidat ion behav io r of tungsten carb ide e lec t rodes in different e lectro lytes. They a l so found poor stabil ity of W C in a lka l ine solut ion, s ince the W C is directly ox id i zed to W 0 3 . T h e p r e s e n c e of unidenti f ied a m o r p h o u s f lakes a long with W C p e a k s s u g g e s t s that s o m e crystal l ine W C is present , but s o m e of it t ransforms into a m o r p h o u s f lakes w h e n e x p o s e d to a bas i c env i ronment . There fore , an al ternat ive method, wh ich d o e s not use alkal i , n e e d e d to be used w h e n depos i t ing Pt on W C . T h e C V data s h o w a comp le te loss of Pt su r face a r e a after only 10 oxidat ion c y c l e s (F igure 10). T h e loss in sur face a rea w a s de te rmined f rom the loss of a rea of the Pt ox ide reduct ion peak at 0.6 V . T h e C V pattern is slightly resist ive due to p rob lems with the R D E dur ing the exper iment . T h e normal ized activity vs . the number of ox idat ion cyc l es plot (F igure 11) s h o w s that W C a s a suppor t is not s tab le, s i nce the percent l oss in activity is s imi lar for both Pt on V u l c a n X C - 7 2 R and Pt on A l fa A e s a r W C . T h e poor stabil ity might result f rom the mixture of W C and a m o r p h o u s f lakes. There fore , an al ternat ive method for Pt depos i t ion w a s d e v e l o p e d , and the resul ts f rom the Pt depos i t ion method II are reported in sec t ion 0. Pt 2-Theta - Scale r Figure 7: XRD spectrum of Pt dispersed on Vulcan XC-72R. 34 W C W C W C W C W C 9 - T h eta . S r. Figure 8: X R D spectrum of Alfa Aesar W C sample 35 < •«-» §J O.O0E*OI 3 o Potential (V) — CV Initial — CV Final Figure 10: Cycl ic voltammogram of Alfa Aesar W C supporting Pt, both before and after 10 oxidation cycles; 0.5 M H 2 S 0 4 , 30°C, 100 mV/s, 2000 R P M . 1.2 -Pt on AAWC method I -Pt on Vulcan method I 4 5 6 7 No. of oxidation cycles 10 Figure 11: Change in anodic activity at 1.8 V as a result of repeated cycling for 40 wt% Pt /AAWC and 40 wt% Pt/Vulcan XC-72R. 36 6.2 Alfa Aesar W C with Pt deposition using Pt (II) pentan-2,4-dionate In these s tud ies 40 wt% Pt w a s d i spe rsed on W C . T h e X R D spec t rum of Pt (II) reduced to Pt onto A l fa A e s a r W C is shown in F igure 12. T h e ave rage crystal l i te s i z e for W C ca lcu la ted f rom the p e a k s at 20 31.5° , 35.6° , and 48.3° w a s 36 nm. T h e a v e r a g e crystal l i te s i ze for Pt ca lcu la ted f rom p e a k s at 26 39.8° and 46.1° w a s 30 nm. The re we re no f lakes o b s e r v e d result ing f rom this syn thes is method, and the X R D s h o w s the p r e s e n c e of Pt and W C p e a k s only, with no e v i d e n c e of an a m o r p h o u s p h a s e present. T h e a v e r a g e crystal l i te s i ze for Pt suppor ted on V u l c a n X C - 7 2 R (Figure 13), ca lcu la ted from p e a k s at 26 40° and 46° w a s 33.6 nm. Pt depos i t ion on W C and V u l c a n X C - 7 2 R w a s s u c c e s s f u l , but the Pt crystal l i te s i z e n e e d s to be reduced to 7-10 nm. Pt WC WC WC Pt Pt .6 10 2 - T h e t a - S e a Figure 12: X R D pattern for Alfa Aesar W C with Pt deposition by Pt (II) reduction to Pt. 37 Pt I 1000 - q 900 - 800 - 700 - 31 40 50 60 . 70 2-Theta- Scale Figure 13: XRD pattern for 40wt % Pt on Vulcan XC-72R with Pt deposit ion by Pt (II) reduction to Pt. T G A data we re used to o b s e r v e the mater ia ls ' stabil ity to c h e m i c a l ox idat ion. T h e thermal stabil ity da ta for A l fa A e s a r W C , 40 wt% Pt on A l fa A e s a r W C , and 4 0 wt% Pt on V u l c a n X C - 7 2 R are shown in F igure 14. Pt suppor ted on V u l c a n X C - 7 2 R loses ~ 5 5 % of the mater ia l , wh ich co r responds to the loss of ca rbon . Both A l fa A e s a r W C and Pt suppor ted on W C ga ined weight a b o v e ~450°C. T h e ga in in weight m a y be attr ibuted to tungsten ox ide format ion. Tungs ten ox ide is a ye l low powder and it w a s o b s e r v e d that both A l fa A e s a r W C and Pt suppor ted on W C had turned f rom dark gray powders to ye l low powders after the T G A run w a s comple te . S i n c e P E M F C s are low temperature operat ing fuel ce l ls the tungsten ca rb ide oxidat ion m a y not be an i ssue . However , with the low pH and high potential condi t ions in P E M F C s , tungsten carb ide m a y undergo ox idat ion. Deta i led e lec t rochemica l tests wou ld help in unders tand ing whether tungsten carb ide oxidat ion is an i ssue for P E M F C operat ion. 38 1.2 1.1 40 wt% Pt on AAWC 1 N o rm al iz ed  w ei g h t o  o  o  Alfa Aesar WC 0.6 40 wt% Pt on Vulcan XC-72 R 0.5 ) 200 400 600 800 Sample Temperature (°C) 1000 Figure 14: T G A data for Alfa Aesar W C , 40 wt% Pt on Alfa Aesar W C , and 40 wt% Pt on Vulcan XC-72R under air at 40ml/min, temperature ramped from 50°C to 1000°C at 2°C/ min. T h e e lec t rochemica l stabil i ty w a s determined for A l fa A e s a r W C , 40 wt% Pt on A l fa A e s a r W C , and 40 wt% Pt on V u l c a n X C - 7 2 R (F igure 15). T h e e lec t rochemica l stabil i ty de te rmined by acce le ra ted test ing on R D E s h o w s that W C is more s tab le than Pt suppor ted on V u l c a n X C - 7 2 R w h e n cyc led be tween +0.6V and +1.8V. T h e C V for pure A l fa A e s a r W C (F igure 16) m a y form tungsten b ronze in the hydrogen adsorpt ion/desorpt ion region. T h e C V after 100 cyc l es d e v e l o p s revers ib le p e a k s at 0.65 V in the anod ic reg ion and at 0.55 V in the ca thod i c region. T h e ~ 1 0 0 m V di f ference be tween the peaks ind icates the p r e s e n c e of revers ib le solut ion s p e c i e s rather than adsorpt ion of any s p e c i e s 4 8 . T h e revers ib le peaks might be from qu inone /hyd roqu inone or other ca rbon -oxygen s p e c i e s 3 . T h e C V for Pt suppor ted on W C (Figure 17) s h o w s a sha rp anod i c peak a b o v e 1.2V, wh ich might result f rom tungsten ox ide on the sur face of the catalyst . A sha rp anod i c peak w a s not o b s e r v e d a s more C V s we re acqu i red . The oxidat ion of the W C by anod i c polar izat ion h a s been reported by L e e et a l 3 1 . T h e y report that a sha rp anod i c peak a b o v e 0 .8V for C V of pure W C has a l so been o b s e r v e d . T h e ox idat ion of W C is cons ide red to fol low equat ion 5. 39 In the current s tudy the sharp anod i c peak w a s not obse rved after the initial C V s c a n . Th i s might result f rom sur face oxidat ion of W during the first s c a n and no further ox idat ion dur ing the subsequen t C V s c a n s . T h e structure of the suppor t might have c h a n g e d f rom Pt suppor ted on W C to P t suppor ted o n a W 0 3 she l l encapsu la t i ng a W C core . A l s o , on the posi t ive s c a n a b road peak be tween 0.3 and 0.45 V with no apparent counter-part on the negat ive s c a n is o b s e r v e d . S imi la r resul ts have been reported by other r e s e a r c h e r s 4 9 , 5 0 . Tungs ten ox ide cou ld form two stable hydrogen tungsten b ronzes , H o . i 8 W 0 3 and H 0 3 5 W O 3 , and sub-s to ich iomet r ic ox ides , W 0 3 . y by react ion with h y d r o g e n 4 9 : W O 3 + x H + + x e " = H x W O 3 ( 0 < x < 1 ) Equation 8 W 0 3 + 2yH + + 2ye" = W 0 3 . y + y H 2 0 (0 < y < 1) Equation 9 A l s o , hydrogen spi l l -over f rom Pt has been repo r t ed 4 9 : x P t H a d s + W 0 3 = Pt + H x W 0 3 Equation 10 H x W 0 3 = x H + + W 0 3 + xe Equation 11 T h e inc rease in hydrogen a d s / d e s peak a r e a a long with a broad anod i c peak be tween 0.3 and 0.45 V probab ly o c c u r s due to tungsten b ronze format ion. T h e s e resul ts sugges t that W C might be ox id iz ing under the current e lec t rochemica l cond i t ions u s e d to test the stabi l i ty of this suppor t mater ia l . Tungs ten carb ide m a y not be stab le under P E M F C condi t ions and m a y ox id ize to W 0 3 with operat ion. Howeve r , the Pt reduct ion peak at - 0 . 7 5 V d o e s not d e c r e a s e in a rea e v e n after 100 oxidat ion cyc l es for P t / W C (F igure 17). E v e n if W C is ox id i zed to W 0 3 the ox id ized tungsten is suff iciently conduct ive to be u s e d a s a support . T h e extent of oxidat ion under operat ion and the impact on fuel cel l pe r fo rmance n e e d to be careful ly de te rmined before rul ing out the opt ion of us ing W C a s a catalyst suppor t in P E M F C s . T h e Pt reduct ion peak is comple te ly lost after 100 oxidat ion cyc l es for Pt suppor ted on V u l c a n X C - 7 2 R (F igure 18). It is ex t remely difficult to c o m p a r e the act ivi t ies for V u l c a n X C - 7 2 R and W C by suppor t ing Pt by we igh t% s ince the dens i t ies of these suppor ts a re signi f icant ly different (with dens i t ies of 1.8 g / c m 3 for C and 15.6 g / c m 3 for W C ) . T h e s e exper iments we re done to o b s e r v e the thermal and e lec t rochemica l character is t ics of W C ; however , an al ternat ive method shou ld be used to c o m p a r e the stabil i ty of P t /Vu l can X C - 7 2 R and P t / W C more directly acco rd ing to su r face a rea . Ano ther c o m p a r i s o n be tween the act ivi t ies of P t /Vu l can X C - 7 2 R and P t / W C c a n be m a d e 40 by suppor t ing exact ly the s a m e amount of Pt on suppor ts that have s a m e vo lume. Resu l t s for Pt suppor ted on both V u l c a n X C - 7 2 R and W C with s imi lar v o l u m e s are reported in sec t ion 6.3. 1.4 , 1.2 1 6. 11 16 21 26 31 36 No. of Oxidation cycle Figure 15: Change in anodic activity at 1.8 V as a result of repeated cycling for Alfa Aesar W C , 40 wt% Pt on Vulcan XC-72R and 40 wt% Pt on Alfa Aesar W C . Average of three samples with limits of error for each material are plotted. 41 1.00E-04 -| -2.00E-04 J Potential (V) Figure 16: Cycl ic voltammograms for Alfa Aesar W C both before and after 100 oxidation cyc les; 0.5 M H 2 S 0 4 , 30°C , 100 mV/s, 2000 R P M . 0.003 i 0.0025 -\ | Potential (V) | Figure 17: Cycl ic voltammograms for 40 wt% Pt on W C both before and after 100 oxidation cycles; 0.5 M H 2 S 0 4 , 30 °C , 100 mV/s, 2000 R P M . 42 0.001 0.0005 -0.0005 i -0.001 A -0.0015 — A f t e r 100 cyc les — Initial Potential (V) Figure 18: Cycl ic voltammograms for 40 wt% Pt on Vulcan XC-72R, both before and after 100 oxidation cycles; 0.5 M H 2 S 0 4 , 30°C, 100 mV/s, 2000 R P M . S E M and T E M images S E M i m a g e s of A l fa A e s a r W C s h o w the p r e s e n c e of d e n s e s p h e r e s p a c k e d together with s o m e i r regular ly -shaped part ic les (F igure 19). It is difficult to d is t inguish be tween Pt and W C part ic les in the high-resolut ion S E M i m a g e s of Pt on W C (Figure 20). 43 Figure 19: S E M images of Alfa Aesar W C Figure 20: S E M images of 40 wt% Pt on Alfa Aesar W C , (A), (B) secondary electron mode; (C), (D) mixed secondary and backscattered electron mode. 44 T h e S E M images of Pt on V u l c a n X C - 7 2 R depos i ted by method I (F igure 21) and method II (F igure 22) s h o w f ine Pt part ic les d i s p e r s e d on ca rbon . Pt part icle s i z e rang ing f rom 10-20 nm w a s o b s e r v e d for P t / C p repared by method I, wh ich is c l o s e to the a v e r a g e crystal l i te s i ze of 11.6 nm ca lcu la ted f rom the X R D results. P t part icle s i z e s rang ing f rom 10 to 6 0 nm are o b s e r v e d for P t / C depos i ted by method II. A v e r a g e crystal l i te s i ze for this mater ia l w a s ca l cu la ted to be 36 nm from the X R D resul ts. T h e Pt part icle s i ze n e e d s to be opt imized for method II. T h e opt imizat ion work for Pt depos i t ion is r e c o m m e n d e d for future s tud ies. $4700 5.0kV 4.3mm x60.dk SE(U,-40) 6/17/06 500nm S4700 5 OkV 4.3mm xTO.Ok SE[U-40) 5/17/06 SOOnm Figure 21: S E M images of 40 wt% Pt on Vulcan XC-72R deposited using chlorplatinic acid; images taken using mixed secondary and backscattered electron mode. 45 Figure 22: S E M images of 40 wt% Pt on Vulcan XC-72R deposited using Pt (II) pentan-2,4- dionate; (A-C) mixed secondary and backscattered electron mode, (D) secondary electron mode. 46 T E M of Pt dispersed on Alfa Aesar W C using Pt (ll)-pentan-2,4-dionate A T E M image of Pt d i spe rsed on A l fa A e s a r W C is s h o w n in F igure 23 . T h e e lementa l spec t rum a long with the concent ra t ions for spot 1 and spot 2 on the T E M image of P t d i spe rsed on A l fa A e s a r W C s h o w that the dark a rea (spot 1) has a P t : W ratio of 0 :100, and spot 2 has a P t : W ratio of 22 :78 by weight . It is difficult to d is t inguish be tween Pt and W C in T E M i m a g e s , but the e lementa l m a p s c a n be used to obtain the relat ive ratios of e a c h e lement . T h e p r e s e n c e of iron, cobal t , and coppe r result f rom the s a m p l e holder. T h e T E M images , a long with e lementa l m a p of P tA /u l can X C - 7 2 R (Figure 24), s h o w the p r e s e n c e of Pt c lus ters d i s p e r s e d o n ca rbon . T h e d ispers ion n e e d s to be opt imized s o that f ine Pt part ic les a re even ly d i spe rsed on the suppor t mater ia l . 47 1000 500 Counts 800 H 600 400 H 200 A Spot 1 W l Cu F in KB Pt Fe KA 1 W LB Ft KB I CoKA I P t l i 1 C O K | 1 L n 11 WLG W U 1 4 A / U . I J/)J \ , .AA. , . . Pt LG T " r-'r-"i r'*"i r"1"1" Spot 2 - W L V MG Cu F / MN p, F b K A 1 W LB ^ J M 1 If F b K B 1 Pt L* i pt i CD K A I I j * l n W L G | M G C o t * LJ1 P l L f f&1 MN W LI if. — • 10 15 keV Spot 1 Spot 2 Iron 4.38 wt% 5.75 wt% Cobalt 3.42 wt% 4.75 wt% Copper 38.78 wt% 35.81 wt% Tungsten 53,39 wt% 41.99 wt% Platinum 0.02 wt% 11.71 wt% 0 5 10 15 keV Figure 23: T E M images and EDX of Pt dispersed on Alfa Aesar W C using Pt (II) pentan-2,4- dionate. Elemental spectrum and weight concentration for spots 1 and 2 on the T E M image of Pt d ispersed on Alfa Aesar W C . 48 6.3 Comparing activities of Pt deposited using Pt (II) pentan-2,4-dionate on similar volumes of both Alfa Aesar W C and Vulcan XC-72R It is difficult to direct ly c o m p a r e the activit ies for V u l c a n X C - 7 2 R a n d W C suppor t ing c o m p a r a b l e weight percent of Pt , s i n c e the densi t ies of these suppor ts a re signi f icant ly different. There fo re , s imi lar v o l u m e s of the suppor ts with the s a m e amount of Pt on e a c h we re a l s o tes ted in order to further c o m p a r e the two suppor ts . S i n c e ca rbon c a n be m a d e by m a n y m e t h o d s lead ing to different morpho log ies e a c h t ime, the tap densi ty is usua l ly u s e d to cha rac te r i ze the dens i ty of t hese mater ia ls. T a p dens i ty involves pack ing mater ia l to a cer ta in v o l u m e by tapp ing it. Both tungsten carb ide a n d V u l c a n X C - 7 2 R were tapped 2 0 0 t imes until exact ly 2 ml of the mater ia l w a s ob ta ined . T o e a c h of the suppor ts w a s a d d e d the s a m e amount of Pt , wh i ch w a s 4 0 wt% relative to 2m l of p a c k e d c a r b o n . T h e method for Pt depos i t ion on s imi lar v o l u m e s of V u l c a n X C - 7 2 R a n d W C is repor ted in sect ion 4 .3 . T h e C V for P tA /u l can X C - 7 2 R (Figure 25) s h o w s the initial Pt charac ter is t ics , a n d a comp le te loss of the plat inum ox ide reduct ion peak that is o b s e r v e d after 100 cyc l es . T h e X R D pattern for P t /Vu l can X C - 7 2 R s h o w s the p resence of Pt (Figure 26) , wh ich was not o b s e r v e d for the X R D pattern for P t / W C (Figure 28) . 49 Initial •After 100 cycles Potential (V) Figure 25: Cycl ic voltammograms for Pt dispersed on Vulcan XC-72R; 0.5 M H 2 S 0 4 , 30°C , 100 mV/s , 2000 R P M . 50 2-Theta - Scale Figure 26: X R D pattern for Vulcan XC-72R with Pt deposition by Pt (II) reduction to Pt. T h e C V for P t / W C (F igure 27) d o e s not exhibit any Pt charac ter is t ics s u c h a s hydrogen adsorp t ion /desorp t ion or Pt ox idat ion/ reduct ion. E v e n though a s imi lar amoun t of Pt w a s depos i ted on both V u l c a n X C - 7 2 R and W C , no Pt character is t ics we re o b s e r v e d in the C V or in the X R D pattern (F igure 28). S imi la r pack ing techn iques we re used in pack ing both the V u l c a n X C - 7 2 R and W C powders to a vo lume of 2 ml. However , W C being very d e n s e with sma l l part icle s i z e (~ 36 nm), will pack very c lose ly and have a very high sur face a rea per unit vo lume. E v e n with s imi lar powder vo l ume (tap densi ty) , V u l c a n X C - 7 2 R h a s agg lomera tes with s i z e s up to the |im range, and therefore has a very different su r face a rea c o m p a r e d to A l fa A e s a r W C . There fore , the Pt content by sur face a rea wou ld be very low on W C c o m p a r e d to V u l c a n X C - 7 2 R w h e n the tap densi ty vo l ume method is used for Pt depos i t ion . T h e ca lcu la ted Pt weight pe rcen ts relative to A A W C and ca rbon are s h o w n in T a b l e 2. T h e Pt wt% for A A W C w a s ca lcu la ted to be only 0 . 9 1 % , wh ich is wel l be low the detect ion limit of the X R D . A n al ternat ive method involved compar ing the stabil i ty of the two suppor ts with s imi lar sol id vo l ume rat ios of Pt to suppor t mater ia l . T h e resul ts of this compar i son are p resen ted in sec t ion 6.4. 51 A m o u n t of P t (II) p recursor (g) Pt metal (g) M a s s of mater ial (g) Pt content (wt%) A A W C 0.16 0.08 8.23 0.91 C 0.16 0.08 0.11 39 .95 Table 2: Amount of A A W C and C tapped in a 2ml volume and the calculated wt% of Pt added to each support material 0.0002 0.0001 -0.0004 -0.0005 1.4 •After 100 cycles •Initial Potential (V) Figure 27: Cycl ic voltammograms for Pt dispersed on Alfa Aesar W C , initially and after 100 oxidation cycles; 0.5 M H 2 S 0 4 , 30°C, 100 mV/s, 2000 R P M . 52 W C W C W C W C 2-Theta - Scale Figure 28: XRD pattern for Alfa Aesar W C with Pt deposition from Pt (II) reduction to Pt. 6.4 Pt addition in equal sol id volume ratios to carbon and tungsten carbide Us ing the s a m e weight ratio of Pt to C (40 wt%) used in sec t ion 6.3, the amoun t of W C required to obtain equa l so l id vo lume ratios of Pt to both C and W C w a s ca lcu la ted . T h e C V for P t / W C (F igure 29) d o e s not exhibi t any Pt charac ter is t ics s u c h a s hydrogen adsorp t ion /desorp t ion or Pt ox idat ion/ reduct ion. E v e n though a s imi lar vo l ume ratio of Pt w a s depos i ted on both V u l c a n X C - 7 2 R a n d W C , no P t charac ter is t ics w e r e o b s e r v e d in the C V . Howeve r , there are low intensity P t p e a k s obse rved in the X R D pattern (F igure 30). T h e ca lcu la ted Pt weight percent relat ive to A A W C and ca rbon are shown in T a b l e 3. E q u a l sol id vo l ume ratios resul ted in 6 wt% Pt on A A W C , wh ich w a s wel l within the X R D detect ion limit. However , the effect of Pt w a s not o b s e r v e d in the cyc l i c vo l t ammogram. T a b l e 3 c o m p a r e s the Pt .suppor t rat ios for tes ts involv ing Pt d i spe rs ion us ing the s a m e weight ratio (40wt% Pt, sec t ion 0), tapped powder vo lume ratio (sect ion 6.3) and sol id vo l ume ratio (sect ion 6.4). Fo r the s a m e weight ratio method , where 40wt% Pt w a s d i spe rsed on A A W C , the 53 P t : W C sol id vo l ume ratio w a s - 0 . 4 9 , wh i ch w a s m u c h higher c o m p a r e d to the Pt : C sol id vo l ume of - 0 . 5 . Desp i te the high amoun t of Pt vo lume d ispers ion on A A W C both the cyc l i c vo l tammetry and ox idat ion cyc le data s h o w e d that the A A W C support mater ial w a s m u c h s tab le w h e n c o m p a r e d to ca rbon . However , none of the Pt d ispers ion methods used in this s tudy c a n direct ly c o m p a r e the e lec t rochemica l stabil i ty of A A W C and ca rbon . The re is a need to normal ize Pt content per unit su r face a rea of the suppor t in order to better c o m p a r e the act ivi t ies of convent iona l ca rbon suppor ts and W C . Pt with s imi lar su r face a rea c a n be suppor ted on both ca rbon and W C with s imi lar su r face a reas . Both B runauer -Emmet t -Te l l e r ( B E T ) and mercu ry poros imetry c a n be used to determine the total su r face a rea ava i lab le in both C and W C suppor ts . A more deta i led s tudy is requi red to deve lop a su r face a r e a b a s e d me thod a n d is r e c o m m e n d e d for future work. C a r b o n is known to have internal microporos i ty that c a n be m e a s u r e d by B E T ( N 2 adsorpt ion) , but the internal porosi ty will not be ava i lab le for Pt depos i t ion . There fore , the actual su r face a rea of Pt that c a n be suppor ted on ca rbon will be lower than that wh i ch c a n be m e a s u r e d by B E T . Mercu ry poros imetry c a n be used to de termine the pore vo l ume of pores with microporos i ty ; however , -h igh .p ressures-a re required for mercury to be fi l led in to the smal l pores , and this cou ld lead to ruptured ca rbon part ic les. There fo re , e v e n normal iz ing Pt content per unit su r face a rea of the suppor t canno t be done directly. Howeve r , the e lec t rochemica l stabil i ty da ta b a s e d on s imi lar amoun ts of Pt suppor ted on s imi lar su r face a rea suppor ts cou ld be more c o m p a r a b l e than the c o m p a r i s o n s m a d e in this s tudy. 54 -0.0012 -0.0014 Potential (V) 1.2 1.4 •After 100 cycles Initial Figure 29: Cycl ic voltammograms for Pt dispersed on Alfa Aesar W C , initially and after 100 oxidation cyc les; 0.5 M H 2 S 0 4 , 30°C , 100 mV/s, 2000 R P M . 55 8. o , W C W W W C W C Pt J W C 2-Theta - S c a l e Figure 30: XRD pattern for Alfa Aesar W C with Pt deposition from Pt (II) reduction to Pt. 56 A m o u n t of Pt (II) p r e c u r s o r (g) Pt (g) W C ( g ) Pt content (wt%) W C volume (cm 3) Pt vo lume (cm 3) Pt vo l : so l id W C vol W C 4 0 w t % P t 0.500 0.240 0.360 40.000 0.023 0.011 0.485 T a p powder density 0.156 0.075 8.266 0.898 0.530 0.003 0.007 Sol id density 0.156 0.075 0.994 7.003 0.064 0.003 0.055 A m o u n t of Pt (II) p r e c u r s o r (g) Pt(g) C ( g ) Pt content (wt%) C vo lume (cm 3) Pt vo lume (cm 3) Pt vo l : sol id C v o l c 40 wt% Pt 0.500 0.240 0.360 40.000 0.200 0.011 0.056 T a p powder density 0.156 0.075 0.114 39.644 0.063 0.003 0.055 Sol id density 0.156 0.075 0.114 39.644 0.063 0.003 0.055 Solid density used for calculations (g/cm 3): C = 1.8; Pt = 21.45; W C = 12.6 Table 3: Compar ison of Pt wt%, and Pt to W C volume ratio for each method used to disperse Pt on A A W C 7 Results and d iscuss ion for indium tin oxide (ITO) studies Both thermal and e lec t rochemica l stabil i ty of commerc ia l l y ava i lab le ITO ( N a n o p h a s e ) and Pt suppor ted on ITO were eva lua ted . Pt w a s depos i ted us ing two me thods (I and II) a s exp la ined in the exper imenta l p rocedure sect ion . 7.1 ITO with Pt deposit ion using chlorplatinic acid T G A Results Figure 31 s h o w s T G A da ta for H i s p e c 4 0 0 0 , V u l c a n X C - 7 2 R , and Pt depos i ted in -house on V u l c a n X C - 7 2 R . T h e peak at tempera tures be low 50°C is due to the loss of moisture. H i s p e c 4 0 0 0 , wh ich is more act ive than the cata lyst m a d e in -house (40 wt% Pt on V u l c a n X C - 7 2 R ) , starts to thermal ly ox id ize at ~300°C, whi le the in -house P t /Vu l can X C - 7 2 R starts to ox id ize at ~325°C. V u l c a n X C - 7 2 R without Pt d o e s not start to ox id ize until ~650°C . T h e thermal ox idat ion resul ts of P t suppor ted on V u l c a n X C - 7 2 R ag ree with the f indings by R o e n et a l 6 that Pt ca ta l yzes 57 the oxidat ion of ca rbon . F igure 32 s h o w s the normal ized weight l oss for seve ra l catalyst suppor t mater ia ls a s a funct ion of temperature . Af ter heat ing the mater ia ls to 1000°C under air, V u l c a n X C - 7 2 R lost 1 0 0 % of its weight, H i s p e c 4 0 0 0 lost 57 wt%, - 4 0 wt% Pt on V u l c a n X C - 7 2 R lost 5 5 wt%, ITO lost 1 wt%, a n d P t / ITO lost 0.7 wt%. T h e l o s s e s for H i s p e c 4 0 0 0 a n d 4 0 wt% Pt o n V u l c a n X C - 7 2 R c o r r e s p o n d to a comple te loss of ca rbon , and the rema in ing weight c o r r e s p o n d s to the Pt meta l left in the T G A cruc ib le . T h i s result c lear ly ind icates that the ITO suppor t is thermal ly s tab le , in cont ras t to the other suppor t mater ia ls s tud ied . -0.02 -0.04 I -0.06 • c E -0.08 Ui E TJ -0.1 E -0.12 -0.14 -0.16 -0.18 Hispec 4000 00 900 X J 1000 Vulcan 40 wt% Pt on Vulcan Sample Temperature ( °C) Figure 31: T G A data for Hispec 4000 and Vulcan XC-72R; under air at 40ml/min, temperature ramped from 50°C to 1000°C at 2 °C /min . 58 1.2 -, ITO and Pt/ITO 0 200 400 600 800 1000 Sample Temperature (T°C) Figure 32: T G A data for Hispec 4000, Vulcan XC-72R, Pt on Vulcan XC-72R, Pt on ITO, and ITO; under air at 40ml/min, temperature ramped from 50°C to 1000°C at 2°C/min . X-Ray Diffraction Results Figure 33 s h o w s the X R D pattern for 40 wt% Pt suppor ted on commerc ia l l y ava i lab le ITO powder (Nanophase ) . Us ing the Sche r re r equat ion , the crystal l i te s i ze for Pt w a s ca lcu la ted to be 13 nm, and that of ITO w a s 38 nm. -: IT O 1 I T O Pt ro I T O 10 • 20 30 40 50 60 70 2-Theta - Scale Figure 33: XRD pattern for 40 wt% Pt on ITO deposited using method I 59 Electrochemical Testing: Rotating Disc Electrode (RDE) T h e results of ox idat ion c y c l e s from 0.6 to 1.8V were u s e d to ca lcu la te the norma l i zed activity for ITO, H i s p e c 4 0 0 0 and P t / ITO, as s h o w n in F igure 34. T h e normal ized activity w a s ca lcu la ted by record ing the last current point f rom the data set at 1.8V just before the current b e c a m e negat ive under 0.6 V cond i t ions . C o m p a r i n g H i s p e c 4 0 0 0 , Pt on ITO, and Pt on V u l c a n X C - 7 2 R be ing held at 1.8V for 20 s per cyc le , the stabil i ty fo l lows the order of: P t / ITO » H i s p e c 4 0 0 0 » 40wt% Pt on V u l c a n X C - 7 2 R . Pt on V u l c a n X C - 7 2 R and H i s p e c 4 0 0 0 lost mos t of their activity after 10 cyc l es . Pt on V u l c a n X C - 7 2 R had s imi lar thermal stabil i ty and activity l oss under ox idat ion c y c l e s to those of H i s p e c 4 0 0 0 , indicat ing that the method u s e d in -house to d i spe rse Pt y ie lds cata lyst with s imi lar stabil i ty a s commerc ia l l y ava i lab le H i s p e c 4 0 0 0 . Pt on ITO after 10 cyc l es only lost ~ 2 5 % of its activity, indicat ing that ITO is a m u c h more stable suppor t than V u l c a n X C - 7 2 R . T h e ITO part icle s i ze is lower than that of V u l c a n X C - 7 2 R ; however , by us ing so l -ge l me thods , the sur face a rea of ITO c a n be e n h a n c e d . It wou ld be useful to s tudy the activity of ITO with opt imized sur face a rea , with cata lyst d i spe rsed by s imi lar methods a s u s e d in the current s tudy. Th i s work will be left for a future study. T h e slight i nc rease in activity at ox idat ion cyc l es 11 a n d 21 for ITO are obse rved b e c a u s e after 10 cyc les , a C V test w a s per fo rmed, wh ich i nc reased the activity just before the start of the next set of ox idat ion cyc l es . T h e C V for pure ITO in F igure 35 s h o w s the stabil i ty of this mater ia l . T h e norma l i zed activity plot s h o w s that the activity i nc reased to over 1 poss ib ly due to sur face roughen ing f rom ox idat ion- reduct ion cyc les . S i n c e In c a n have mult iple oxidat ion states, the su r face oxidat ion c a n i nc rease the a rea , leading to h igher activity than at the start ing point. F igure 35 and F igure 36 s h o w the cyc l i c vo l t ammograms of Pt on ITO and H i s p e c 4 0 0 0 , respect ive ly , both before and after 100 oxidat ion cyc les . Pt on ITO s h o w e d signi f icant ly better e lec t rochemica l stabil i ty, a s de te rmined by a lower loss of e lec t rochemica l l y act ive sur face a rea . Th i s sur face a rea w a s de termined f rom the a rea under the hydrogen adsorp t ion cu rves . Hyd rogen adsorpt ion p e a k s were present for the Pt on ITO even after 100 cyc l es from 0.6 to 1.8 V . O n the other hand , mos t of act ive su r face a r e a of the H i s p e c 4 0 0 0 w a s lost after 100 c y c l e s . T h e Pt ox ide reduct ion peak for H i s p e c 4 0 0 0 w a s a l so obse rved to shift to lower potent ials (0.75 V to 0 .55V) , w h e r e a s the s a m e peak for Pt suppor ted on ITO did not shift, even after 100 cyc l es . T h e 40 wt% Pt on V u l c a n X C - 7 2 R lost a lmos t all of its act ive a r e a after only 50 cyc les (F igure 38). T h e total currents in the cyc l i c vo l tammetry tests for ITO suppor t ing Pt we re m u c h lower than those for H i s p e c 4 0 0 0 b e c a u s e the act ive sur face a rea of the Pt part ic les on the ITO is m u c h lower than that in H i s p e c 4 0 0 0 . C h a n g i n g the microstructure of the cata lyst /suppor t combinat ion in future tests c a n modi fy this total activity. 60 1.6 n 1.4 1.2 Hispec 4000 ITO 40 wt% Pt on ITO - 4 0 w t % P t o n Vulcan XC-72R 16 21 Number of Oxidation Cycles 26 31 Figure 34: Normalized activity at different potentials as a result of repeated cycl ing for different 40 wt% Pt catalysts. Average of three samples with limits of error for each material are plotted. 1.00E-04 5.00E-05 0.00E+00 O -1.00E-04 4 -1.50E-04. -2.00E-04 -2.50E-04 Potential (V) — Initial — After 100 cyc les Figure 35: Cycl ic voltammograms for ITO both before and after oxidation cycles at 1.8V. 100 oxidation cycles were run; 0.5 M H 2 S 0 4 , 30°C , 100 mV/s , 2000 R P M . 61 0.001 -> Potential (V) Figure 36: Cycl ic voltammograms for 40 wt% Pt on ITO both before and after oxidation cycles at 1.8V. 100 oxidation cycles were run. Electrochemical stability with no change in the C V curves is observed from cycle 30 onwards; 0.5 M H 2 S 0 4 , 30°C , 100 mV/s, 2000 R P M . 62 Figure 37: Cycl ic voltammograms for Hispec 4000, both before and after 100 oxidation cycles; 0.5 M H 2 S 0 4 , 30°C , 100 mV/s, 2000 R P M . 63 0.001 -, -0.0015 J Potential (V) Figure 38: Cycl ic voltammograms for 40 wt% Pt on VulcanXC-72R, both before and after 50 oxidation cycles; 0.5 M H 2 S 0 4 , 30°C , 100 mV/s, 2000 R P M . T E M and S E M images Figure 39 and F igure 4 0 s h o w S E M i m a g e s for N a n o p h a s e ITO and 40 wt% Pt on N a n o p h a s e ITO, respect ive ly . A mixture of spher ica l and oc tahedra l structure of ITO crystal l i tes is v is ib le for the N a n o p h a s e ITO samp le . S m a l l Pt part ic les a re obse rved for the 40 wt% Pt on N a n o p h a s e ITO s a m p l e in the S E M images . F igure 41 s h o w s three T E M i m a g e s of 40 wt% Pt on ITO. T h e E D X data indicate that Pt c lus ters a re d i s p e r s e d on sma l l ITO part ic les. Fur ther microstructural opt imizat ion of the ITO c a n lead to further i nc reases in overal l e lec t rochemica l activity. 64 Figure 39: S E M images of ITO 66 ••• Pt In Sn Sn In Pt In Sn Figure 41: T E M / E D X of 40 wt% Pt on ITO 67 7.2 ITO with Pt deposition using Pt (II) pentan-2,4-dionate XRD data T h e X R D pattern for ITO with Pt f rom depos i t ion method II is s h o w n in F igure 4 2 . T h e Pt depos i t ion method II invo lved d ispers ing P t (II) sal t onto ITO and later reduc ing it to Pt meta l in a tube furnace. T h e reduc ing a tmosphe re c a u s e d Pt to al loy with In, forming ln 2 Pt . PtSn2/ In2Pt In2Pt In2Pt Figure 42: XRD pattern for Pt deposited on ITO by reduction of Pt (II) pentan-2,4-dionate. Electrochemical testing: Rotating Disc Electrode (RDE) T h e C V after cyc le 2 and 4 for Pt on ITO is s h o w n in F igure 43 . T h e inset s h o w s the initial C V with no Pt character is t ics . T h e a b s e n c e of Pt C V character is t ics initially might be f rom the format ion of In-Pt al loy on the sur face . A s more C V s c a n s were taken, the C V s exhib i ted Pt character is t ics s u c h a s hydrogen adsorp t ion /desorp t ion and Pt ox idat ion/ reduct ion p e a k s . T h e oxidat ion cyc l es (F igure 44) s h o w that P t depos i ted by method II on ITO h a s a lower activity than Pt on ITO depos i ted by method I. T h e Pt depos i t ion method II is not preferred for ITO s ince the In and Pt form an al loy, and the e lec t rochemica l stabil ity of this mater ial is poorer than that of the s a m p l e prepared by method I. 68 Potential (V) 1.4 - After cyc le 4 •After c y c l e 2 Potential (V) Figure 43: Cycl ic voltammograms of Pt deposited on ITO by reduction of Pt (II) pentan-2,4- dionate after oxidation cycles 2 and 4. The inset shows the C V at initial point. 69 1.2 i 4 0 w t % P t on ITO method I 40 wt% Pt on ITO method II 16 21 Number of Oxidation Cycles 26 31 Figure 44: Normalized activity at different potentials as a result of repeated cycl ing for Pt deposited using chlorplatinic acid (method I) and Pt (II) pentan-2,4-dionate (method II) on ITO. Average of 3 separate samples with limits of error for each material are plotted. 8 Conc lus ions 8.1 Tungsten carbide Pt depos i t ion involving a lka l ine mater ia ls canno t be u s e d for Pt d i spers ion on W C . A n al ternat ive method w a s c rea ted involv ing d ispers ion of Pt (II) salt, wh ich w a s later r educed to p roduce Pt d i spe rsed on W C , as conf i rmed by X R D . Direct c o m p a r i s o n of the e lec t rochemica l stabil i ty be tween ca rbon and W C is difficult due to the large d i f ferences in densi ty . Different tests with vary ing amoun ts of reactants we re s tud ied, the tests involved c o m p a r i s o n s with the s a m e weight ratio of Pt on W C and C , tapped powder vo l ume ratio, and f inal ly the sol id vo lume ratio. T h e 40wt% Pt on W C and C w a s not a s u c c e s s f u l test c o m p a r e the e lec t rochemica l stabil i ty of W C and C due to the large dens i ty d i f ferences be tween W C and ca rbon . T h e tap dens i ty method w a s u s e d to tap s imi lar v o l u m e s of the suppor t mater ia l . 70 For ca rbon that w a s tapped to a cer ta in v o l u m e 40wt% Pt w a s a d d e d and s imi lar amount of Pt a s for ca rbon w a s a d d e d to W C . Final ly , the sol id vo l ume ratio test involved d i spers ing s imi lar amoun t of Pt on ca rbon and W C , whe re the v o l u m e s of the suppor ts u s e d were ca lcu la ted b a s e d on the sol id dens i ty of the support mater ia ls . T h e Pt vo lume to W C vo lume ratio da ta is s u m m a r i z e d in T a b l e 3. T h e Pt to W C vo lume ratio for the 40wt% P t / W C w a s 0.49, wh ich w a s signi f icant ly h igher than c o m p a r e d to P t : C vo lume ratio of 0.06. B a s e d o n h igher content of Pt in c a s e of 40 wt% Pt the e lec t rochemica l stabil i ty of this suppor t w a s signi f icant ly better than ca rbon . But , s i nce the sur face a rea of W C w a s a lot lower than ca rbon and the signif icant dens i ty d i f ference, direct c o m p a r i s o n be tween ca rbon and W C cannot be m a d e . W i th the tap dens i ty vo lume ratio test the Pt wt% relative to W C w a s only 1 % and w a s not de tec ted both in X R D and in the C V s . S o m e Pt w a s detected in the X R D for the P t / W C us ing the sol id dens i ty vo l ume ratio test but no Pt charac ter is t ics were o b s e r v e d in the C V . Al ternat ive me thods n e e d to be determined in order to directly c o m p a r e the e lec t rochemica l stabil i ty of W C and ca rbon and the sugges t ions a re exp la ined in future work sec t ion 9 .1 . 8.2 Indium tin oxide Resu l t s for Pt depos i t ion by method II indicate that this method resul ts in the format ion of l n 2 P t al loy, and the Pt / ITO has a lower e lec t rochemica l stabil i ty c o m p a r e d with the P t / ITO p repared by method I. C o m p a r i n g H i s p e c 4 0 0 0 , Pt on ITO, and Pt on V u l c a n X C - 7 2 R be ing held at 1.8V for 20s per cyc le , the stabil i ty fo l lows the order of: P t / ITO » H i s p e c 4 0 0 0 * 40wt% Pt on V u l c a n X C - 7 2 R . Pt on V u l c a n X C - 7 2 R and H i s p e c 4 0 0 0 lost mos t of their activity after 10 ox idat ion c y c l e s . Pt on ITO after 10 oxidat ion c y c l e s on ly lost - 2 5 % of its activity, indicat ing that ITO is a m u c h more s tab le suppor t than V u l c a n X C - 7 2 R . T h e total currents in the cyc l i c vo l tammetry tests for ITO suppor t ing Pt we re m u c h lower than those for H i s p e c 4 0 0 0 b e c a u s e the act ive su r face a r e a of the Pt part ic les o n the ITO is m u c h lower than that in H i s p e c 4 0 0 0 . C h a n g i n g the microstructure of the cata lys t /suppor t combina t ion in future tests c a n modi fy this total activity. Overa l l , ITO has potential a s an oxidat ion-res is tant cand ida te mater ial for catalyst suppor ts in P E M F C s . 71 9 Future Work 9.1 Tungsten carbide T h e r e is a n e e d to normal ize Pt content per unit su r face a rea of the suppor t in order to more direct ly c o m p a r e the act ivi t ies of convent iona l ca rbon suppor t and W C . Both B runaue r -Emmet t - Te l le r ( B E T ) and mercury poros imetry c a n be u s e d to determine the total su r face a rea ava i lab le in both suppor ts . H igh sur face a rea tungsten carb ide n e e d s to be syn thes i zed in order to e n h a n c e the cata lyst layer activity. Encapsu la t i ng high su r face a r e a ca rbon with tungsten ca rb ide m a y lead to an ox idat ion resistant high su r face a rea W C coa ted ca rbon support . If ca rbon ox id i zes , it is lost a s ca rbon d iox ide g a s , caus i ng Pt to fall off the suppor t and lead ing to low Pt su r face a rea and signif icant pe r fo rmance degrada t ion . Howeve r , if tungsten ca rb ide ox id i zes to tungsten ox ide , it rema ins conduct ive , and little loss in Pt su r face a r e a is e x p e c t e d . Potent ia l use of tungsten ox ide as a cata lyst suppor t for P E M F C s n e e d s to be invest igated. S y n t h e s i s routes s u c h a s so l -ge l me thods that c a n provide high su r face a r e a tungsten ox ide a lso need to be invest igated. Final ly , the e lec t rochemica l stabil i ty of tungsten ox ide n e e d s to be s tud ied. 9.2 Indium tin oxide T h e ITO part icle s i ze is lower than that of V u l c a n X C - 7 2 R ; however , by us ing so l -ge l me thods , the sur face a r e a of ITO c a n be e n h a n c e d . It will be usefu l to s tudy the activity of ITO with op t imized sur face a rea , with cata lyst d i spe rsed by s imi lar methods a s u s e d in the current s tudy. Further s tud ies are requi red to opt imize the part icle s i ze distr ibution and d ispers ion of the cata lyst on the suppor t mater ia l . P e r f o r m a n c e in a work ing fuel cel l ca thode of P t suppor ted on ITO a l so n e e d s to be cha rac te r i zed . 72 10 References S t e v e n s , D A ; D a h n , J . R. C a r b o n 2 0 0 5 , 4 3 , 179-188 . Math ias , M . ; M a k h a r i a , R.; Gas te ige r , H, ; C o n l e y , J . ; Ful ler , T. ; G i t t l eman, C ; K o c h a , S . ; Mil ler, D.; Mit testeadt, C ; X i e , T. ; Y a n , S . ; Y u , P . Interface 2005 , 14, 24 -35 . K a n g a s n i e m i , K.; Cond i t , D.; Ja rv i , T . Jou rna l of the e lec t rochemica l soc ie ty 2004 , 151 , E 1 2 5 - E 1 3 2 . Interface Fal l 2 0 0 5 , 21-44 . 2005 . Re f T y p e : M a g a z i n e Art ic le K inosh i ta , K. C a r b o n ; E lec t rochemica l and P h y s i c o c h e m i c a l Proper t ies ; W i ley , N e w Yo rk : 1988; pp 316-334 . R o e n , L. M. ; Pa ik , C . H.; Ja rv i , T . D. E lec t rochemica l and sol id-state letters 2004 , 7 , A 1 9 - A 2 2 . Kn ights , S . ; Tay lor , J . ; W i l k i nson , D.; C a m p b e l l , S . [WO0115254 ] . 2 0 0 1 . Re f T y p e : Patent Y e , S . ; Beat t ie , P . ; C a m p b e l l , S . ; W i l k i nson , D.; T h e o b o l d , B.; Thompse t t , D. [ U S 2 0 0 4 0 1 3 9 3 5 (A1)]. 2 0 0 5 . Re f T y p e : Patent Ba l la rd unpub l i shed resul ts . 2006 . Re f T y p e : G e n e r i c R e i s e r , C ; Brego l i , L ; Pa t te rson , T.; Y i , J . ; Y a n g , J . ; Per ry , M . ; Ja rv i , T . E lec t rochemica l and sol id-state letters 2005 , 8, A 2 7 3 - A 2 7 6 . Gruver , G . A . Jou rna l of the e lec t rochemica l soc ie ty 1978, 125, 1719 -1720 . K inosh i ta , K.; Bett, J . C a r b o n 1973 , 11, 237 -247 . K inosh i ta , K.; Bett, J . Ef fects of graphi t izat ion on the cor ros ion of ca rbon b lacks . A d v a n c e s in instrumentat ion, P r o c e e d i n g s , 43 -55 . 1974. R e f T y p e : C o n f e r e n c e P r o c e e d i n g H e c k m a n , F.; Har l ing, D. R u b b e r C h e m . T e c h n o l . 1966, 39, 1-13. S c h m i d , K.; Ro th , J . Journa l of N u c l e a r Mater ia ls 2002 , 302 , 96 -103 . W i l l sau , J . ; He i tbaum, J . Journa l of E lec t roana ly t ica l Chem is t r y 1984, 6, 9 3 - 1 0 1 . S t e v e n s , D. A . ; H i cks , M . T.; H a u g e n , G . M . ; D a h n , J . R. Journa l of the e lec t rochemica l soc ie ty 2 0 0 5 , 152, A 2 3 0 9 - A 2 3 1 5 . K inosh i ta , K.; Bett, J . C a r b o n 1974, 12. 73 (19) M c D o n a l d , J . P . ; Stonehar t , P . Produc t ion of s tab le, h igh su r face a rea c a r b o n s e lect rocata lyst suppor t through dop ing with B o r o n . C o n f e r e n c e on C a r b o n , 16th , 150 -151 . 1983 . 1983 . Re f T y p e : C o n f e r e n c e P r o c e e d i n g (20) Y o u n g , J . ; L e e , J . ; J u b i s a , L ; R a d o v i c , R. C a r b o n 2003 , 1987 -1997 . (21) S a m a n t , P . ; R a n g e l , C ; R o m e r o , M . ; F e r n a n d e s , J . ; F igue i redo , J . Jou rna l of power s o u r c e s 2005 , 151 , 79-84 . (22) Y a n g , R.; Q i u , X . ; Z h a n g , H. ; L i , J . ; Z h u , W . ; W a n g , Z . ; H u a n g , X . ; C h e n , L. C a r b o n 2 0 0 5 , 4 3 , 11-16. (23) Se rp , P. ; Cor ra i s , M . ; Ka lck , P . App l i ed Ca ta l ys i s , A : G e n e r a l 2003 , 2 5 3 , 337 -358 . (24) W a n g , J . ; S w a i n , G . M . Jou rna l of E lec t rochemica l Soc ie t y 2 0 0 3 , 150, E 2 4 - E 3 2 . (25) Q i , Z . ; Le febvre , M . C ; P i ckup , P . G . Journa l of E lect roanaly t ica l Chem is t r y 1998, 459 , 9- 14. (26) „ M a r i e , J . ; Be r thon-Fabry , S.;. A c h a r d , P . ; Cha tene t , M . ; Pradoura t , A . ; Cha ine t , E . Journa l of non-crysta l l ine so l ids 2004 , 350 , 88 -96 . (27) R a j e s h , B. ; T h a m p i , K. R.; B o n a r d , J . M . ; Ma th ieu , H. J . ; Xan thopou los , N . ; V i s h w a n a t h a n , B. Jou rna l of P o w e r S o u r c e s 2005 , 141 , 35 -38 . (28) J a l a n , V . ; Frost , D. G . Fue l Ce l l E lec t rocata lys t Suppor t C o m p r i s i n g a n Ultra-f ine cha iny - structured T i tan ium C a r b i d e . [4,795,684], 1989. Re f T y p e : Patent (29) P i e r s o n , H. O . H a n d b o o k of Ref rac tory C a r b i d e s and Ni t r ides; N o y e s Pub l i ca t ions : 1996 ; . pp 1-115. (30) M e n g , H. ; S h e n , P . K. C h e m i c a l C o m m u n i c a t i o n 2005 , 2005 , 4 4 0 8 - 4 4 1 0 . (31) L e e , K.; Ishihara, K.; M i t sush ima , S . ; K a m i y a , N . ; O t a , K . E I e c t r o c h i m i c a A c t a 2004 , 4 9 , 3 4 7 9 - 3 4 8 5 . (32) M a , C ; G a n , Y . ; C h u , Y . ; H u a n g , H. ; C h e n , D.; Z h o u , B. T ransac t i ons of nonfer rous meta ls soc ie ty of C h i n a 2004 , 14, 11-14. (33) M a , C ; Z h a n g , W . ; C h e n , D.; Z h o u , B. T ransac t i ons of nonfer rous meta ls soc ie ty of C h i n a 2002 , 12, 1015-1019 . (34) Chr is t ian , J . B.; M e n d e n h a l l , R. G . Tungs ten-con ta in ing fuel cel l cata lyst and method of mak ing s a m e . O s r a m S y l v a n i a Inc. [US6 ,656 ,870 B2] , 1-9. 2 0 0 3 . Re f T y p e : Patent (35) Chr is t ian , J . B.; D a n g , T . A . ; M e n d e n h a l l , R. G . Suppor ted tungsten ca rb ide mater ia l . O s r a m Sy l van ia Inc. [ U S 6 . 5 5 1 , 5 6 9 B1], 1-9. 3 A . D . Re f T y p e : Patent 74 (36) Boudar t , M . ; O y a m a , S . T. ; Vo lpe , L. B o a r d of T r u s t e e s of Le land Stanford Jr .Un ivers i ty . [4,515,763]-. 4 A . D . Re f T y p e : Paten t (37) C la r idge , J . B. ; Y o r , A . P . E. ; B r u n g s , A . J . ; A l v a r e z , C . M . ; S l o a n , J . ; T s a n g , S . C ; G r e e n , M . L. H. Jou rna l of ca ta lys is 1998, 180, 8 5 - 1 0 0 . (38) T h o m p s o n , ! . T. ; W i x o m , M. ; T a r n o w s k i , D.; P u , C . [US 6 ,297 ,185 B1] . 2 0 0 1 . Re f T y p e : Patent (39) R e d d y , B. M . ; K h a n , A . Ca ta l ys i s R e v i e w s 2 0 0 5 , 47 , 257 -296 . (40) C h e n , G . ; Ba re , S . R.; Ma l louk , T. E . Journa l of E lec t rochemica l Soc i e t y 2002 , 149, A - 1092 -A -1099 . (41) L iu , J . ; Rad le i n , E. ; Fr ischat , G . H. P h y s . C h e m . G l a s s e s 1999, 2 7 7 - 2 8 1 . (42) M a t v e e v a , E . Journa l of the e lec t rochemica l soc ie ty 2 0 0 5 , 152, H 1 3 8 - H 1 4 5 . (43) M a s o n , T. O . ; G o n z a l e z , G . B.; K a m m l e r , D. R.; M a s o u r i a n - H a d a v i , N . ; Ingram, B. J . Th in .So l i d F i lms 2002 , 4 1 1 , 106-114. (44) S c e l e l , M . Fue l cel l E lec t rode compr i s ing conduc t i ve zeol i te suppor t mater ia l . [6,117,581]. 2000 . Re f T y p e : Patent (45) H iguch i , E. ; U c h i d a , H. ; Fu j inami , T. ; W a t a n b e , M . So l id S ta te Ionics 2004 , 171 , 45 -49 . (46) Cull i ty, B. D. E l e m e n t s of X - R a y Dif fract ion; 1967; pp 96 -103 . (47) Cull i ty, B. D. E l e m e n t s of X - R a y Dif fract ion; A d d i s o n W e s l e y Pub l i sh ing C o m p a n y , Inc: 1978; pp 1-30. (48) Gree f , R.; Pea t , R.; Peter , L. M . ; R o b i n s o n , J . ; P le tcher , D. Instrumental M e t h o d s in E lec t rochemis t ry ; El l is Horwood l imited: 1985 . (49) S h i m , J . ; L e e , C . R.; L e e , H. K.; L e e , J . S . ; C a i r n s , E . J . Jou rna l of power s o u r c e s 2 0 0 1 , 1 0 2 , 1 7 2 - 1 7 7 . (50) K u l e s z a , P . J . ; Fau lkner , L. R. Jou rna l of the e lec t rochemica l soc ie ty 1989, 136, 7 0 7 - 7 1 3 . 75

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
China 11 0
Germany 5 21
Japan 4 0
United States 2 3
South Africa 1 0
City Views Downloads
Beijing 9 0
Ingolstadt 4 0
Tokyo 4 0
Nanjing 2 0
Unknown 2 21
Sunnyvale 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items