The Ethanol Oxidation Reaction and Product Distributions in a Direct Ethanol Fuel Cell

The Ethanol Oxidation Reaction and Product Distributions in a Direct Ethanol Fuel Cell

by

David D. James

B.Sc. (l- hmours), Dalhousie Un iversi ty Halifax, Nova Scotia 2 008

A thesis sub mitted to the School of Graduate Studies in partial fulfi ll ment of the requirements for the degree of

Doctor of Ph iloso phy

Depm 1ment of Chemistry Memorial Uni versity St. John ·s. Newfoundland

September 23

. 20 13

Abstract

A novel o nline system for qu anti fy ing produ ct di stributions of ethanol oxidati on in di rect etha nol fu e l ce lls was deve l oped . Thi s system was shown to ha ve many advantages over prev ious l y re ported meth odo log i es s uch as: ease of operation, quick response time, economi cal, and real time analysis. Thi s system consisted of a non-di spers ive infrared C0

detector coupled with a flo w-through conductivity detector for C0

and aceti c acid measurements, respective ly. T he third ethano l ox idation product, acetald ehyde, was shown to be accurately calcul ated using a Faradaic charge balance.

Th e ef fects of fu el and product crossover were close ly examined. It was shown

that the use of oxyge n at th e cathode can lead to a n overestim ation of ethanol ox idation

products, ma inly aceti c acid, due to ethanol crossing through the membrane and reacting

chemi call y with oxygen at the cathode. Furthermore, it was shown th at s ign ificant

amo unts of acetaldehyde produced du ring ethano l oxid atio n were lost due to crossover,

l eadin g to an und erest imati on of its yield . To obtain more acc urate product distributions,

the fu el ce ll was operated us ing N

/H

gas at the cathode (whi ch eliminated the

chemi call y oxidized ethanol reaction). To furth er improve the accuracy of product yield s,

a ·'crosso ver mode" approach to operating a fuel cell was examined. It was fo und that this

meth od in creased selecti vity toward s complete ox idation (carbon di ox ide) by red ucing

poisonin g of the electro de.

A kineti c and mechani stic study on th e ethanol oxid ation reacti on was also carried out us ing an el ectrode stripping techniqu e. It was fo und that appl ying a potential in the range where CO is ox idized, fo llowed by a llow ing the ce ll to return to o pen circuit and then re-appl ying the potential led to significant increases in C0

y ield s. It was found that the C0 2 yie ld was dependent on the length of the pulse appl ied, the shorter the pul sing inte rval, the higher the yie ld. T his suggests t hat the majo rity of the C02 produced was attributed to th e oxid atio n of CO adsorbates. The tim e i nterva l between pul ses was also examined. It was fo und that th e C0

yield was ind epe ndent of the resting time, suggestin g a rapid dissociati on of etha nol on th e el ectrode, whi ch s upports prev i ous I iterature findin gs.

A Pt-RuSn0

/C anode catalyst was developed and tested in our system. This

catalyst was found to increase both the performance and the selectivity towards C0

in

co mpari so n to a Pt electrode, whi ch i s rarely reported in the literature .

Acknowledgements

I wo uld first like to express my sin cere gratitude to my supervisor Dr. Peter Pickup for the continuous s upport throughout my program. Hi s patience, moti vation, e nthusiasm and immense knowledge had made my Ph.D an enjoya ble and un forgettable experi ence. I co uld not have as ked for a better advisor for my program .

I would like to thank my superv i sory committee, Dr. Ray Poir ier and Dr. Dave Tho mpson for the adv i ce, s upport and time they have dedicated throughout my program.

I want to thank all the fa culty members of the chemi stry department who sat on my comprehensive exam committee, examined my semin ars and taught me courses throughout the yea rs.

I acknow ledge the support staff that have advised and trained me on th e various instruments throughout Memorial Un i versi ty, especially the staff members of C-CART a nd the technic ians workin g out of the Earth Sc ience Department. I'd also like to thank the administration and physical science stores staff for all the ir he lp. I thank Dr.

Guangc hun Li for sharing hi s ex pertise with me and th e Pickup group for all of their input and support toward s my proj ect.

I would like to extend my s incere gratitud e to the funding agencies th at made my

program possible: Schoo l ofGraduate St udies, the Chemi stry Department, 3M Innovat ion

and the Na tural Sci ence and Engineerin g Research Council of Canada. A spec ial than ks to NSE RC for their generous fundin g through the Post Graduate Scholarship.

I would also like to thank all the friends I have made duri ng my time

Newfoundl and for the ir s upport a nd compani onship. My deepest apprec iati on to my

parents, who have taught me the valu e of hard work and determinat ion whi ch was needed

to push me through my program. To my partner Le igh, I thank you fo r the l ove and

support and for always lifting my spirits throughout the hard times .

Contents

Abstract ii

Acknowledgements iv

Table of Contents vi

List of Tables xii

List of Figures xv i

List of Abbreviations xxiii

List of Symbols xxvi

1.

Introduction

1.1 Int roduction .. ... .... ... ... ... ... .. .. ... .. .. .. ... .... ... ... .. ... .. ... ... 2

1.2 Type of Fuel Ce lls .... .. .. ... .. .. ... .... .... ... .. .. .. ... ... 8

1 .2. 1 Fuel Ce ll Composi ti on .... .. ... .... .... .... .... .... .. .. ... ... .. ... ... .. . 8

1 .2.2 Proton Exchange Membrane Fuel Cell ( PEMFC) .... ... .... ... 9

1 .2 .3 Direct Alcoho l Fuel Ce ll (DAFC) .... .. ... ... .... .. . I 5 1.2.3.1 Direct Methanol Fue l Ce ll (DM FC) .. .... ... .... ... 1 6 1.2.3.2 Direct Ethanol Fuel Ce ll (DEFC) .. ... .. .... ... ... 17

1.2.4 Alka line Ani on Exchange Membrane Fue l Ce ll (AAE MFC) .. ... 20

1.3 Fue l Ce ll Theo ry .. .. .. ... .. ... ... .. .... .. ... .. ... ... ... .. .. .... . 23

1 .3.1 Fuel Ce ll Performance ... ... .. ... ... ... ... ... .. .. 23

1.3.2 Fuel Ce ll Efficiencies ... ... ... ... .... ... ... ... 27

1.3.3 Energy and Power Density ... ... ... ... .... ... ... ... ... ... ... ... .. 30

1.3.4 Farada ic Processes .. .... ... ... .. ... ... ... ... .... .. ... .. 32

1.4 Limitation of the EOR in DEFCs ... ... ... ... ... ... ... 33

1.4.1 Poor EOR Kine tics ... ... ... ... .. ... ... ... ... .... . 33

1.4.2 Incomplete Oxidation of Ethanol ... ... ... ... ... .. 38

1.4.3 Effects of Fue l Crossover ... ... .. ... ... .. ... ... .. ... ... . 42

1.4.4 Cost Ana lysis of PEM-DEFCs ... ... ... .... .... .... ... . 43

1. 5 Research Obj ectives .. .. .... ... ... ... .. ... .... ... .. .. .. .. ... ... ... 45

I .6 References ... ... ... .. .. ... ... .... ... .. .... ... ... .. ... .... ... .. 4 7 2. Experimental 56 2.1 Chemicals and Materi als ... .... ... ... ... ... ... ... .. ... .. 57

2.2 Preparat ion of Electrodes and MEAs .... ... ... ... ... ... ... 58

2.3 Electrochemica l Meas urements .. ... ... .. .... ... ... ... .... ... ... 58

2.4 Product Analysis Instrumentation ... .... ... ... ... ... ... ... 59

2.4.1 Non-di spers ive In frared (ND IR) Carbon Diox ide Detector ... 59

2.4.2 Flow-Thro ug h Conduct i v ity Detector ... .. ... ... .. 60

2.4 .3 Gas Chromatograph .... .. ... .... ... ... ... .... ... ... ... .. ... 62

2.5 Cata lyst Characte rization Techniques ... .... ... ... .. ... .. ... ... 63

2.5 .1 X-Ray Diffraction (XRD) .. ... ... ... ... ... . 63

2.5 .2 Inducti vely Coup led Plas ma Mass Spectroscopy (ICP-MS) ... . 64

2.5 .3 Inductively Coupled Plas ma Opti cal Em iss ion Spectroscopy (ICP-OES) ... ... .. ... .... .... .... .... .... ... .. .. .. ... ... .. .. 64

2.5 .4 Thermogravitational Ana lys is (TGA) ... ... .... .. ... .. ... .... ... ... 65

2.5 .5 Energy Dispersive X-Ray Microana lys is (EDX) .... ... .. .. .. ... 65

2.5 .6 Transmission Electron Microsco py (TEM) ... ... .. .. .... ... 66

2.6 References ... ... ... ... .. .. .... ... ... ... .... .. .... ... ... 67

3 A Novel Methodology for Online Analysis of Products from a DEFC 68 3. 1 Introduction ... ... ... ... ... .... .. .. .. .. .... ... ... .... ... 69

3.2 Ex perin1ental ... .... ... .. ... ... ... .. .. .... ... ... ... ... ... .. .. .. . 72

3.2.1 Materials .. ... ... .. ... ... .. ... ... .. .. ... ... ... 72

3.2.2 The Fue l Cel l ... ... ... ... ... ... ... .. ... ... 73

3.2.3 Product Analysis .. ... ... ... .. .... .. ... .... .. ... . 73

3.3 Results and Discussion ... .... .. ... .. ... ... ... .. . 76

3.3 .1 Conductimetric Analysis of Acetic Ac id .... .. .. ... ... . 76

3.3 .2 C0

Measurements with an ND IR Based Detector .... ... 87

3.3 .3 Acetaldehyde Analys i s and Charge Balance ... .. ... ... 88

3.3.4 Pt and PtRu Anode Catalysts Prepa red from Dalho usie Uni vers ity ... ... ... .... ... ... .... ... .. ... .... .... ... .. 92

3.3.5 Produ ct Analysis of Homemade High Perform ance Cata l ysts .. .. 95

3.4 Conc lus ion ... ... .. ... ... ... ... ... .... .. ... 98

3.5 References .. ... ... ... ... .. ... .. ... ... ... ... .... ... 99 4 Effects of Ethanol and Product Crossover on the EOR

4.1 Introducti on ... ... ... ... .. .... .... .. .. ... ... ... .. ... . I 03

4. 2 Experimental .. ... ... .. ... ... .. ... ... ... ... ... ... ... .... ... .... ... I 06

4.2. I Materia ls ... .. ... ... .. ... ... .. ... .. ... .. I 06

4.2 .2 The Fue l Ce ll ... ... ... .... ... ... .. I 06

4.2.3 Product Analysis .. ... .. ... ... .... .... .. ... ... .. ... I 06

4.2.4 Ca libration of th e NDIR Detector with Pure C0 2 ... ... .... . 1 07 4.3 Results and Discussion ... ... .. ... .... ... ... ... ... ... ... ... ... ... I 08 4.3 .I The Effects of Ethanol Crossover ... ... ... ... ... ... . I 08 4 .3 .2 Acetic Ac id Produced by Ethanol Crossover ... ... ... ... 11 0 4 .3.3 C02 Yields .. .. ... .... .. ... ... .... ... .... ... ... .. ... ... 11 5 4 .3.4 Aceta ldehyde Crossover .. ... ... ... ... ... .. .... .... ... ... ... 11 5 4.3 .5 Product Ana lysis at the Cathode ... ... .. .. ... ... ... ... ... 123 4.4 Conclu sions ... ... ... .. ... ... ... ... ... .... ... ... 124 4.5 References .. .... ... ... ... ... ... .. .... ... ... ... ... .. ... 1 26 5 Measurement of Carbon Dioxide Yields for Ethanol Oxidation by

Operation of a Direct Ethanol Fuel Cell in Crossover Mode

5. 1 Introduction ... .... ... ... ... ... ... .. ... ... ... .. ... 1 29 5.2 Experimental .. .... ... .... ... ... .... ... ... .. ... ... .... ... .... ... . 131

5.2 .1 Material s .. ... .... ... ... .. .. ... ... ... .... .... ... .... ... .. 131 5.2.2 The Ce ll ... ... ... ... ... ... ... ... ... .. .... ... 1 31 5.2.3 C0

Analys is .. ... ... ... .. ... ... .... ... .... 1 3 1 5.3 Results and Discuss ion .. ... ... .... .... .... ... ... .. ... ... ... ... .... ... .... 1 32 5.3. 1 Testing and Cal ibration of the System with Methanol ... .... ... 1 32 5.3.2 Ethano l Oxidat ion ... ... ... ... ... ... ... .... .. .. .. .. 1 34 5.4 Conclusion s ... .. ... ... ... ... ... ... ... .. ... ... ... 1 4 1 5. 5 References ... ... ... ... ... ... .. ... ... .... ... .... ... ... .. .. 14 3

6

Kinetic and Mechanistic Study 146

6.1 Introduction ... ... .. ... ... ... ... ... .... ... ... ... 1 4 7

6.2 Experin1ental ... ... ... .... ... .. .... ... ... .... ... ... .. .. ... 151

6.2.1 Mate rials ... .... ... .... ... .. ... ... .. ... ... .. .. ... ... .. . 1 5 1 6.2.2 The Fuel Ce ll ... .... ... ... ... ... ... ... ... ... ... 151 6.2.3 C02 Anal ys is ... ... ... .. .. ... ... ... ... ... .... ... .. .. .... 154 6.2 .4 Acetic Acid Anal ysis .... .... ... ... ... ... ... ... ... .. . 1 56 6. 3 Results and Di scussio n ... ... ... ... ... ... .... .... ... 1 56 6.3.1 Linear Sweep Experiments ... ... .... ... ... ... .. .... ... ... .. .. ... ... 1 56 6.3.2 Increas ing C0 2 Yi eld Using Pul sing Tec hniques ... ... ... 1 6 1 6.3. 2.1 R esting Time Dependence on C0 2 Yi eld ... ... 1 64 6.3.2.2 C0 2 Yield D ependence on Pulse Time ... ... 1 66 6.3.2.3 Temperature Depend ence on C0

Yi eld ... .. ... .. ... 1 69 6.3.2.4 Puls ing in Crossover Mode ... ... ... .... .. .. ... .. .. ... ... 1 7 1 6.3.3 Ace taldeh yde Oxidation ... ... ... .... .... ... ... ... .. .. .... ... . 1 73 6.4 Conclu sions .... ... .. .. ... ... .... .. ... .. ... ... .... ... .... ... .. ... ... 1 76 6.5 References ... ... ... ... ... ... ... ... .. ... .... ... ... ... .. ... 178 7 Enhanced Performance and C0

Selectivity with a Pt-RuSnOz /C Catalys t

7. 1 Introdu ction ... ... .. ... ... ... ... .. ... .... ... .. .. .... 181

7.2 Ex perimental ... .. .... .... ... ... ... ... ... ... ... .. .. .... .... ... .. .... ... ... ... ... ... ... 1 83

7.2. 1 Chemical and Materi als .... ... ... ... ... .... .... .... .. ... ... .. ... .... 183

7.2.2 Ce lls .. ... ... .. ... ... .. ... ... ... ... ... ... .. ... . 1 84

7.2 .3 Elect roche mical Measureme nts ... ... ... ... 1 84

7.2 .4 Produ ct Analysis .... ... ... .. ... ... ... ... ... ... 1 85

7.2.5 Characte rizat i on oft he Cata lysts ... ... ... .. 1 86

7.2 .6 Preparation of Cata l ysts ... .. .. .... ... ... .... ... .... .. ... .... ... I 86

7.2.6.1 Method I (M l) ... .... ... ... ... .. ... ... ... ... 186

7.2.6 .2 Method 2 (M2) ... ... ... ... ... 187

7.3 Results and Discussion ... .... .. ... ... .... ... .. .. .. .... . 189

7.3. I Characterization of Catal ysts .. .. .... ... .... .... ... ... .. .. .... 189

7.3.2 Electrochemical Measurements in a Liquid Electrolyte Cell .... 199

7.3.3 Activities of Electrodes in a Fuel Cell .... ... .. .. .. .... ... .... 203

7.3.4 Product Analysis and C0

Selectivity .... .. .. .. .. ... ... .. .. .. .. ... . 205

7.4 Conclusions ... ... .. ... ... .... .. ... ... ... ... ... ... ... . 207

7.5 References .. .. .. ... .... .. ... ... ... ... ... ... .. .. ... .. ... ... .. ... 208

8 Summary and Future Work 210

8.1 SuJnJnary ... .. .. .. .. ... .. ... ... ... .... ... ... .... .. ... ... ... ... .... 2 1 I 8.2 Future Wo rk ... ... .. ... .. ... ... .... ... ... .. ... .. .. .. .. .. .. .. .. .... ... .. 2 13 A ^C0

and Current Traces for Rest Interval Dependence on COz Yield 215 B C0

and Current Traces for C0

Yield Dependence on Pulse Time 219

c ^C0

and Current Traces for Crossover Mode Pulsing Experiments 223

D ^C0

Traces for Aceta ld ehyde Oxidation 228

E EDX Spectra 232

F TEM Images 239

G C0

Traces for Ballard, M1 and M2 Catalysts 243

List of Tables

1.1 Energy densities per volume and per mass for many popular fuel s used in

combustion engines and fuel cells ... .... .... .. ... ... ... ... ... .. ... ... 3 I

1.2 Average price of metals used in catalysts for the EOR from July 20 II - Jul y 2012 in Europe ... .. .... .. .. ... ... .. .. ... ... ... .. .. ... .. 44

3.1 Faradaic yields from co nductivity, titrati on, GC , C0

measurements on the anode exhaust so lution from a DEFC with a Pt anode ope rated at a constant current and ambient temperature. Values in parentheses are based on the

titrati o n results ... .. ... ... .... ... .. ... ... .. .. ... ... ... ... ... ... .. ... 86

3.2 Faradai c yields of the EOR products from conductivity, titration, GC and C0

meas urements on the anode exhau st so luti on from a DEFC with a Pt anode operated at a constant current and 80 °C. Va lues in parentheses are

based on the titration res ults ... .... .... .. .... ... .... ... .... .. .. ... .... .. .. ... . 86

3.3 Faradaic yie lds of the EOR products for Pt and PtRu anode catalysts

prepared by Da lhousie Uni versity. C0

yield s were measured with the

ND IR detector, AA yields were ca l cul ated from the conductiv ity ce ll and

AA L yield s were calc ulated with equation 3.6. Experiments we re conducted at 80 °C, usin g 0 .5 mo l L-

ethanol so lution and H

gas at

the cathode .... ... ... ... ... .. ... .. ... ... .. .... ... ... ... .... .... .... ... 94

3.4 Farad aic yields for C02 a nd aceti c acid with vari ous anode catalysts, at

I 00 °C, 0. 5 mol L-

ethanol so luti on with 0

at th e cathode ... .. ... ... ... 97

4.1 Product a nalys i s for th e EOR on a Pt-Sn (9: I) anode catalyst at 80 °C, 3

bar and 2 mol L-

etha nol soluti on using 0

gas at the cathode ... .... ... ... ... I 05

4.2 Farada i c y i e ld s fro m titratio n (acetic ac id) a nd ND IR (C02 ) measurements on the anode exhaust so luti on fro m a DEFC with a Pt anode. The cathode

gas was e ither N

/H 2 or 0 2 ... .. .. .. ... ... ... ... ... .. ... .... .. .... ... . 11 4

4.3 Faradaic yields from titration, GC, and ND1R meas urements on the anode

exhaust so luti on from a DEFC with a Pt anode. The cathode gas was

I 16

4.4

Average Fa radaic yield s from titration (aceti c acid) and NDIR (C02 )

measure me nts during acetald ehyde crossover meas urements at 80 o c ... 1 20

4.5

Farada ic yield s fro m titration, GC, and ND1R measure ment s on the cathode ex haust gas from a DEFC ope rated at 80 o c wi th a Pt anode. The cathode

gas was N

/H 2 ... ... .. ... ... ... ... .... .. ... ... ... ... .... ... ... .. ... .. 124

5.1 Faradaic yields of C0

for oxidation of ethanol crossing over through a Nation® 115 proton exchange membrane at 80 o c ^{in a} DEFC. The anode

gas was N

138

6.1 Faradaic C0

yields for potentiostatic experiments at 700, 800, 900 and I 000 m V (vs. H 2 ) . C0 2 yield s were ca lculated for the entire run and for

vari ous regions of the experiment.. .... ... .. .... ... .. .... ... .. .. ... ... .. ... ... 160

6.2 Charges and moles of C0 2 obtained from graph s in Figure 6.7 .. .. .... ... ... 164

6.3 C0

yield s for pulsing experiments usi ng 2 s pulses of0.8 V of with various resting interva ls at open circuit. Ethanol vapor from I mol L-

aqueous ethanol was used as fue l. Experime nts carri ed out at roo m

te rnperature ... .. ... .. ... ... .. ... ... .. ... ... .. 1 65

6.4 Faradaic C0

yield s for pulsing experiments using 0.8 V (vs. H

pulsing

potential and a constant resting interval of 5 s at open circuit. ... .. .... .... .. .. .... 1 68

6.5 Apparent Faradaic C0 2 y i e ld calcul ated us ing the integrated moles of C02 and the charge produced in the cell us ing 2 s pulsing intervals at 0.8 V

( vs. H

and I 5 s resting intervals at open c ircuit between pulses .... .... ... .... . 171

6.6 Faradaic C0

yie lds calc ulated in crossover mode using strippin g methodo logy. The pulse times were 5 s and the rest ing tim e between

pu lses was 5 sat ope n c ircuit .... .. ... ... .... ... ... .. .... ... ... 1 73

6.7 Ace tald ehyde ox idation (0.5 mo l L- 1 in distilled water) product ana lysis at var i o us tempe ratures and currents. Wate r was passed th rough the cathode

in these experi1ne nts ... ... ... ... .... .. ... ... ... ... ... 175

7.1 Composition estimates (target) fo r RuSn0

/C and Pt-RuSn0

/C sam ples ca lcul ated from stoichiometric amoun ts of eac h element used in synthesis (see sectio n 7.2.5). Actua l mass% (found) based on TGA and ICP-OES

measuren1 ents ... ... ... ... .. .. ... .. ... ... .. 1 98

7.2 Faradaic yie l ds from titration (AA) and NDIR (C0

measurements from co mbined anode an d cathode exhausts wi th various anode cata lysts . AAL yie l ds were calcul ated based o n the Faradaic charge balance. The cathode gas was 0

Fu el cell temperature was 80 o c and 0.5 m ol L-

etha n ol so lut io n was used. T itrati on and ND IR measurements were corrected fo r

bl ank meas urements at open c ircuit where no current was fl owi ng .. ... .... .... 206

List of Figures

1.1 Schematic of a typical single cell so lid electro lyte fuel ce ll composition .. .. .. .... ... 9

1.2 A schemati c diag ram of a typical MEA for a PEMC ... .. .... .... .. .. ... .. .. .. ... .. II

1.3 Chemi cal compos ition ofthe Na ti on® membrane deve l oped by DuPont .. .. .... .. 1 2

1.4 Mechanisti c scheme for the oxygen reduction reaction occ urring at a Pt

cath ode catalyst in an ac idic environment ... .. .... .... .. .. .. .. .. .. ... .... ... ... .. ... 15

1.5 A sche mati c of an AAEMFC for ethanol ox idati o n ... .... ... .. .. .. .... .. ... 22

1.6 A typical polarization curv e fo r a fu el ce ll. Regions: I. Acti vation

po lari zation 2. Ohmic Po l arization. 3. Concentrati on po larization .. .... ... .. .. .. ... 28

1.7 Polarization c urve fo r H

ox idati on at a 30% w/o Pt on Vulcan XC 72 anode at 80°C at vari ous steady- state tim e inte rvals. H

was doped with CO to demonstrate the effects of electrode po i so ning .. ... .. .... .... .. .... .. .... .. 34

1.8 a .) Simplified mechani sm for CO oxidation to C0

b.) Poisonin g of the

e lectrode by CO ... ... ... ... ... ... ... ... .... ... .. ... ... .... .. ... ... ... .. .... .. .. 36

1.9 A Schematic o f possible EOR pathways along with the energy barri ers

ca lcul ated at each step usin g OFT, units are in eY .... ... .... .... ... .... .... ... 40

2.1 Gas chamber tube o f DIR C0

monitor consisting of an IR lamp, optical fi Ite r and I R detector .. .. ... .. ... ... .. ... ... .. ... ... ... ... ... ... .. 60

2.2 A schemat ic o f a simple liquid flow-through co nductivity detector ... ... 61

3. 1 Schemat i c of methodo l ogy used by La my eta/. for the detecti on of ethanol ox idation products fro m a direct ethano l fu el cell ... .... ... ⁷¹

3.2 Schematic of the online syste m for product analysis from th e direct ethanol fue l cell ... .. ... .. ... ... ... ... ... ... ... ... .... 74

3.3 Co nductivity versus hydrogen ion conce ntration for acetic ac id standa rds

passe d through the fue l cell (prior to the conductivity detector) ... ... 78

3.4 a ca libration curve for fl ame ato mic absorption (FAA) testin g fo r ionic

conductivity background in direct ethano l fue l cell. .. ... ... .. ... .... .. .. .. ... .. .... 79

3.5 Co nductivity of the exhaust soluti on (so lid) and potential (dashed) vs. time at 60 mA, and roo m temperature for a DEFC with a Pt anode ... ... 82

3.6 Conductivity traces o btained for acetic acid quantificat ion at variou temperatures and cu rrents . a. ) Ambient tempe rature and 80 mA. b.) 80 o c

and 200 mA. c.) I 00 °C and 200 mA, and d .) I 00 °C and 400 mA. All

measurements were ob tained using a commercial 40% Pt anode and

cat hode (Ba llard) in a direct ethanol fuel cell ... ... ... .... .. ... 84

3.7 Carbon dioxide yie ld vs. time at 60 mA for a DEFC w ith aPt anode .. .... .... ... 89

3.8 Chromatog rams for 0.00635 mol L-

and 0.0600 mol L-

AAL standards. Peak a rea ratios were used to calibrate the GC for AAL sa mple anal ys is ... .. ... 90

3.9 Ca libration of GC for acetaldehyde analysis. AA L!EtOH peak area ratios

were used .. ... .... .... ... ... ... .... ... ... ... .. ... ... ... .. 91

3.10

Polari zation curves for the Pt and PtRu anode catalysts prepared by Dahn et a/. and a comm erci a l Pt catalyst from Ballard Power Systems .. .. ... ... .. .... .. .... .. 93

4.1

Ionic co nducti vity ofthe anode exhaust of a DEFC at open circui t a nd a mbient temperature with N

/H

at the cathode and water at the anode (0-67 min), N

/H

at the cathode and 0.5 mo l L-

ethanol at the a node (75- 1 05 min), 0

at the cathode and water at the a node ( I I 0-185 min), 0 2 at the cathode and 0.5 mo l L-

ethanol at the anode (190-250 min) ... ... Ill

4.2 Schematic of DEFC setup in crossover mode with 0.5 mol L -

acetaldehyde at the cathode and N

at the anode for determination of aceta ldehyde flux

through me1nbra ne ... ... ... ... .... ... .. ... .... ... .... ... ... ... ... ... 11 7

4.3 C urrent vs. potentia l for aceta ldehyde ox idation in crossover experiments

with 0.5 mol L

aceta lde hyde at the cathode (DH E) and N

at the anode.

Individual points show the currents (dec reasing with tim e) at 5 s interva ls

at each potential ... ... ... ... ... ... ... ... ... I 19

4.4 Current vs. time at 700 m V vs. DHE for acetaldehyde oxid ation during a crossover experiment with 0.5 mol L -

aceta ldehyde at the cathode (DH E)

and N 2 at the anode. The N 2 flow was stopped at ca. 700 s ... ... 1 22

5.1 Schematic diagram of a DEFC operated in crossover mode with C02 monitoring of the anode exhaust using a non-di spersive infrared (ND IR)

detector ... ... ... ... ... ... ... ... ... .. ... ... I 3 0

5.2 Measured C0 2 concentration in the anode exhaust vs. theoretical

concentration for a I 00% Faradaic yield for constant current (5- 80 mA) oxidation of methanol crossing through a Nation® 11 5 membrane

(T

80 °C) ... ... ... .. ... ... ... .. ... .. ... ... 1 33

5.3 Pol arization curves at 80 °C recorded in the crossover mode shown in Fig. I.

Current measurements were made (averaged over 30-100 s) foll owin g ca.

4 min at each potenti al. N

was passed over the a node at 45 mL min-

while 0.1 or 0.5 mol L-

ethanol soluti on was supplied to the cathode whi ch acted asa DH E ... .. .. ... ... ... .. ... .. ... ... ... ... .. ... ... .... l35

5.4 C0

concentrat ion and current vs. time traces recorded during the oxidation of ethano l at 80 °C and 0.70 V vs . DHE. N

was passed over the anode at

45 mL min

whi l e 0.1 mol L -

etha no l soluti on was s upplied to the cathode .. 137

5.5 Measured C0

yield vs. ce ll pote ntial for DEFCs operated in norma l a nd crossove r modes with 0.1 mol L-

EtOH at 80 °C. In each case, the C0

was measured onl y in the anode exhaust stream .... .. ... .. .. ... ... .. .... ... ... 140

6.1 Current and co nce ntration of C0 2 produced during strippin g experim ent with 2 s pulses at 0.82 V and 30 s resting intervals between pulses at open

circuit ... ... ... ... .. .. .. ... ... .. .. .. ... ... .... ... .. ... .. ... ... . 149

6.2 Schematic of system used for room temperatu re adsorbate stripping

experime nts in norm al mode with H 2 !N

as the cathode gas ... ... .. ... ... .... 152

6.3 Schemati c ofthe fu el cell setup for stripping experime nts ca rried out in

crossover tn ode .. ... .... .... ... ... .... ... .. ... ... .... ... .... .. ... ... ... ... ... . 153

6.4

I mol L-

ethano l stripping experiment in norm al mode at a mbie nt

temperature. 2 s puls ing intervals at 0.8 V (vs. H

fo llowed by 15 s rest ing intervals at open circuit potentia l. a.) Current vs . time trace,

C02

concentrati on vs. time ... .. .. .... ... .. ... .... ... .... .. .. .. .. .. .... .. .... ... .. .. .. .. ... .. .... . 155

6.5 Current and C0

concentration measureme nts by appl yin g a linear potential sweep from 0 to 1.1 V (vs . H

Room temperature experim ent using 1.0 mo l L-

eth ano l va por as the fue l and H

!N

at the cathode to

obta in a stabl e reference potential .. ... .. .. .... ... .. ... .... ... ... .... .... ... .... 157

6.6 C0

traces for consta nt pote ntial experiments fo r 700, 800, 900, and I 000

m V (v s. H

using 1.0 mo l L-

ethanol at room temperature .. .. ... ... .. .. .. .... 159

6.7 Pul sin g experiment us ing 5 s pulses at 0.8 V (vs . H

with 5 s resting interva ls at open c ircuit between pulses. Slow response time of C0 2

measureme nts were corrected fo r .... .. ... ... .. ... .. .. .. ... ... ... .. ... 162

6.8 C0 2 traces for stripping experiments using vario us fue l ce ll temperatures.

Vapour from I mol L-

etha no l solution, 2 s pulses at 0.8 V (vs. H 2 ) with

15 s resting intervals at o pen c irc uit .. ... .. ... .. ... .. .... 1 70

7.1 XRD patte rns of the Pt-RuSn0 2 /C cata lysts powders ... .. ... .. .... ... .. .. .. ... 190

7.2 SEM image of the surf ace of a Pt-Ru Sn0 2/ C (method I) elecrode. Po ints

used for EDX analys is are indi cated .. .. ... ... ... .. .. .. ... 1 9 1

7.3 EDX spectra for po ints # I, 2 and 3 respecti ve ly f ro m Figure 7. 2 ... 192

7.4 (a) . SE M image of the surface of the Pt-RuSn0 2 /C (M2) cata lyst. (b). EDX analys is fo r metal co ntent of the M2 cata lyst .. .. .. .. .. ... ... .... .... ... .... 194

7.5 TEM images o f the M I Pt- Ru Sn0

/C cata lyst .. .... .... .. ... ... ... .... ... .. 1 96

7.6 TGA results for (a) Pt-Ru Sn0 2/ C (MI ) and (b) Pt-Ru Sn0 2 /C (M2) ... ... ... 1 97

7.7 Blank cyc li c voltammograms fo r the Pt-RuSn0 2 /C and the Pt electrodes in 0. 1 mol L-

H 2 S0

at 298 K us ing a scan rate of 0.0 I V s-

200

7.8 CV curves fo r ethanol oxid ati o n usin g 0.2 mol L-

ethano l soluti on added

to 0.1 mo l L-

H 2 S0

at vari ous electrodes. (a) First cycle (b) seco nd cyc l e ... 20 I

7.9 Pol ari zation curves for th e M I , M2 and Pt Black (Ballard ) anode e l ectrode

in a DEFC. 0. 5 mol L-

ethanol soluti on was used as the fuel, 0

was used

as the ox id ant gas and curves were conducted at 80°C .... ... ... ... ... ... 205

List of Abbreviations

2, 4-DNPH AA

AAEMFC AAL AC AFC AOR

CFP CB

co

C02

c v

DA FC Dal DC DEFC DEMS DFAFC DFT DHE

2, 4- Dinitropheny lhydrazin e Acetic ac id

Alkaline anion exchange membrane fuel ce ll Acetaldehyde

Alternating current Alkaline fue l ce ll

Alcohol ox idation reaction Carbon fiber paper

Carbon black Carbon monox ide Carbon diox ide Cyc lic voltammetry Direct alcoho l fuel cell Dalh ousie Univers ity Direct current

Direct ethanol fuel cell

Diff erenti al el ectrochemica l mass spectrometry Direct formic ac id fuel cell

Dens ity functional th eory

Dynami c hydrogen e l ectrode

DMFC EDX EO R FAA FTIR GC HFC HPLC ICP-MS ICP-O ES Ml M2 MEA MOR NDIR OCP ORR PEFC PEM PEMFC PSA RT

Direct methanol fu el cell

Energy di spers ive x-ray mi croana lys is Ethanol oxidation reaction

Flame atomic absorption

Fourier transform infrared spectroscopy Gas chromatog raphy

Hydrogen fuel cell

High performance liquid chromatog raphy Inductivel y coupled plas ma mass spec troscopy

lnductiv ity co upled plasma optica l emi ssi o n spectroscopy Method # 1 for makin g Pt-RuSn0

/C cata lyst

Method #2 for makin g Pt-RuSn0

/C cata lyst Membrane electrode asse mbly

Methanol oxidation reaction Non-dispersive infrared Open circuit potential Oxygen redu ction reaction Polymer electrolyte fue l ce ll Proton exchange membrane

Proton exchange membrane fuel ce ll Polystyrene sulfonic acid

Room temperature

SCE Saturated calomel electrode

SEM Scanning electron microscopy

SHE Standard hydrogen e lectrode

SPEFC Solid polymer electro l yte fue l cell

TEM Transmission e lectron microsco py

TGA Thermog ravitati onal analysis

XRD X-ray diffraction

List of Symbols

A Ampere

~ Rate of fue l consumption B Peak width at half height

c ^Coulomb

ca. Approximately

D Mean particle diameter DE Energy dens ity

Dp Power density

e electron

Eo Standard pote ntial

E Potentia l

En Energy

Erev

Reversible cell potential eV Electron vo l t

F Faraday constant

G Gibb' s energy

lc Wave l e ngth

h Hour

Current

I Jim

Lim it ing current

lo

Exchange current Intensity

J Current density

mm Minute

11 Overpotential

n Electrons transferred in reacti on

N Moles of product

e Diffraction ang le

n Ohm

p Flow rate

ppm Parts per m iII ion

Q ^Charge

R Gas consta nt

R E Electronic res istivity

s Second

s ^Siemens

~s

Entropy change

(J

Co nd uctivity

T

Effic iency

T ime

T Temperature

v

v w

X

Volume Voltage (vo lts) Watts

Species adsorbed on electrode surface

CHAPTER 1

Introduction

1.1 Introduction

With the slow depl etion of foss il fuels , the search for new renewab le energy sources has been an area of intense investigation . One area of focus that has been very attract ive for many research groups is the development of fuel ce ll techn ology usin g hydroge n [ 1-4], met ha nol [5-I 0], etha nol [I 1- 18], or formic ac id [ 19, 20] as the primary fuel source. T he conversion of c hemi cal energy into e lectrica l energy using fuel ce ll s has been ach ieved and reported since the late 1830s [21 ]. The principles behind the fuel ce ll were first re - ported by German scienti st Chri stian Schonbei n in 1838, but a ce ll was not demon strated experimentally until the foll owin g year [22]. Sir William Groove's serend ipitous demon- stration of the fuel ce ll in 1839 came from connecting th e two e lectrodes of an e lectrolyz- e r together w hi ch were orig ina lly attached to a battery. He observed a consumption of hydroge n and oxygen, resulting in a reverse flowing current [2 1].

Due to the poor understanding of electrochemic al theory, very little advancement in fuel ce lls was made until the 1950s. In 1959 Francis Bacon, work ing o ut of the de- partment of c hem ica l engineering at Cambridge U niversi ty, constructed the first practical work ing fuel ce ll, using nickel based electrodes in an alkaline environment [23].

T he first majo r commercial use of fue l ce lls dates back to the early 1 960s with the use of solid polymer electro lyte fuel ce lls (SPEFCs - now known as polymer electro lyte fue l cells (PEFCs)) in the Gemini space missions [24, 25]. A ll space vehicles were equipped with these fue l ce lls which were used as auxiliary power sources ; generatin g power for li ghtin g, regul ating gas pressures and producing drinking water aboard the space craft.

T hese first generation fue l cells used hydrogen gas as the fue l a nd a polystyrene s ulf onic

acid (PSA) proton exchange membrane that enabled hi gh power densit ies [26]. However at the time, due to the extremely high cost of manuf acturing a nd running these fuel cell s their use was limited to space mi ssion s.

In the late 1960s, one of the biggest breakthroughs in P EFC technology was the development of a perfluoro sulfonic acid membrane manufactured by Dupont, also kn own as Nafion® [27]. The hi gh e lectronegativity of the side groups o n the Nati on® membrane a llowed for a two fo ld increase in proto n conductivity compared to that of the PSA mem- brane . T he increased conductivity of the membranes allowed for sign ifican t increases in both th e power d ensity and the cell efficiency [26, 28, 29]. The durability of the Nation®

membrane was exceptional, havin g a life time of about four ord ers of magnitude higher than that of the PSA membrane [30]. The exceptional durability of the Nation® mem- brane sparked interest into the development of direct methan ol fuel cells (DMFCs), direct etha no l fuel ce lls (DEFCs) and direct formic acid fuel ce lls (DFAFCs) in ac id e lectro-

lytes.

Over the past coupl e of decades, fue l cell technology has progressed to the point

of comm ercia lization for many small portable app lications [31-33] and also a few larger

sca le commerc ial products [34, 35]. Hydrogen fuel cells ( HFC) have been at the top of

the researc h ladder for a lmost a ll transportati on manu facturers. A lthough t here are cur-

rent ly no commercial H FC vehicles, there have been many prototype and demonstration

vehicles produced from all the large vehicle manu facturers [36-38]. Ball ard Power Sys-

tems, Toyota, H ydrogenics a nd a few other large scale fue l cell devel opers have intro-

duced a three year HFC bus trial w hi ch is e mployed in man y co untri es inc luding : Cana-

da, US, Braz il, Ge rmany, C hina a nd England [39, 40]. T hese HFC buses have bee n re- ported to be 39-141 % more ef fi c ie nt for city driving co mpared to th e diese l fuelled buses c urrentl y used [ 4 1]. Many w holesa le and serv ice companies including Sysco Foods, Fed- e ral Express Couri er and Coca-Cola have taken ad vantage of the HFC tec hnology by us- ing forklifts powered by HFCs [42]. Other HFC vehi cles recent ly manuf actured inc lude motorcycles, subma rines, bi cyc les, boats and even airplanes.

DM FCs have been the most widely commerc ia lized type of fu e l cell (31, 32].

DM FCs are unable to produce la rge a mounts of power, making them poor ca nd idates for large mac hinery a nd vehic les. However, they d o have the ca pability to store small amounts of energy over long tim e periods in sma ll spaces [43]. T his makes t hem idea l for sma ll energy applicati ons s uch as cell phones, la ptops, cameras and in many military ap- pli cati ons. In th e mid 1990s, over 20,000 meth ano l " fl ex" vehic les we re so ld in the U.S.

T hese vehi cles had the capability to run on methanol or gas [44]. Due to hi gh cost of methanol in th e mid to late 1 990s, the foc us qui ckl y shi fte d fro m m ethanol to ethanol as an altern ati ve to gaso line. T he commerc iali zation of DEFCs could be considered lac king in co mparison to HFCs and DM FCs. Few to none of the appli cations mentioned above have shown increased performance when replaced w ith D EFCs . However, due to the in- trin s ic a dvantages of ethanol as a fuel (see be low), research and deve lopment into DEFCs has gain ed conside rable atte nt ion.

A lthough hyd rogen is the best candid ate as a fuel for fuel ce lls fro m a weight en-

ergy density po int of view (energy produced per m ass of fuel = 33.3 kW h kg.

th e dan-

ger and co mplications w ith the storage, transportation and use of hydrogen severe ly lim-

its its applications [ 45 , 46]. The hydrogen fuel ce ll is th e simplist of the fuel cel ls usin g H 2 at the anode and 0

at the cathode, producing H

0 exc lusively as a product and providing four electrons per mole of 0

consumed . With water as the only reaction prod- uct, the kinetics at the anode are extremely facile l eadi ng to very high cell perf ormances.

Furthermore, since C0 2 is not a by-prod uct of HFC devices , they are a "zero-emi ssion"

(no carbon footprint) energy source, making them id ea l from an environme ntal po int of view [23].

Meth ano l a nd et hano l on the ot her hand are liquid at room temperature, maki ng the transportation and storage ofthe fuel safe and easy. The ava ilabil ity of both methano l and ethano l a lso make the m very attract ive, as heavy reformers (a dev ice that extracts a desired product from a larger product) are not needed as is the case for hydrogen produc- tion. The energy densities for methanol and ethanol are 8.01 and 6.09 kWh kg-

respec- tive l y [47, 48]. A lthough th ese alcohols use very s imilar components to the hydroge n fuel cell (anode, cathode and membrane), th eir performances are significa ntly lower due to intermediates adso rbed o n the anode during oxid ation [ 49,50].

The comm erc ial production of methano l invo l ves th e reacti on of carbon monox-

ide and H 2 gas. During the major coa l mining days, synthet ic gas was produced fro m the

reaction of methane (whi ch was a product of coa l mining) with water in a process call ed

steam methane ref o rmin g. The H 2 gas was then reacted with CO over a nic kel, z inc ox ide

or a lumina cata lyst to produce methano l. Presentl y, meth ano l can be produced via three

different method s; steam methane refo rming, direct catalyt i c conversio n of methane us-

ing Cu-zeo lites or e l ectrochemi ca lly from C02 usin g CuO or Cu02 catalysts.

Methanol has been reported to have a much more simple reaction mechanism than that of etha no l in a fuel cell [51 , 52]. In methan ol ox idation, intermediates form at the anode and oxygen containing species are needed to oxidize these to C02. To produce the oxygen containing species, 0

must be reduced at the cathode. This oxygen reduction re- action (ORR) is the rate determining step, and is extremely s l ow on the Pt cathode that is used in the hydrogen fue l cell. Although the thermodynamic reversible potential of ORR for MeOH is comparab le to that of H 2 at the cathode (1.214 Y and 1.23 Y vs. SHEre- spectively) , th e methano l oxidation reacti on (MOR) occurs at a much hi gher potential on the Pt cathode [53]. Due to the s low kinetics of the MOR, the Pt anode gets quick ly cov- ered with CO intermediates, limiting the avai lab le adsorpti on sites for oxygen containing species, hence, the oxidation to C02 [6,54,55]. This "poisonin g" of the anode is the pri- mary reason for the slugg is h kinetics of DMFCs. Recent studies have shown that all oyi ng Pt with a secondary metal (primarily Ru) has the ability to dec rease the CO poisoning at the electrode, leadin g to significant increases in MeOH ox idation kinetics [eg. 51 ,56-58].

MeOH crossing through the membrane has also caused man y problems with DMFCs. At all co ncentrati ons, MeOH has been shown to cross through the proton exchange mem- bran e (PEM) from the anode to the cathod e [59]. T hi s causes a mixed potential at the cathode l eadin g to a sign ificant decrease in potentia l at the cathode and hence a decrease in overall ce ll performance.

Ethanol can be produced both naturall y and syntheticall y. The majority of ethanol

that i s used in a l coholic beverages and as a fue l is produced natura lly by the fermentation

of sugars. Thi s fermentati on process inv olves yeast metabo lizing sugar under certa in

th ermal conditions producing ethanol and C0 2 . Ethanol used as a chemica l so lvent is typ- ically mad e syntheticall y. Thi s process is a n ac id-catalyzed hydration reacti on in volvin g ethy l ene and water producing ethanol exclusive ly.

Etha no l has been the focus of many research groups due to its abi lity to be pro-

duced in large quantiti es from the ferm entation of biomass . Th is gree n, renewable fue l

a l so has adva ntages over methanol in that it is non-toxic, and has a hi ghe r boil i ng point

and hi gher energy dens ity [60, 6 1, 62] . However, the ethano l ox idation reaction (EOR) in

DEFCs follow an extreme ly complex mechani sm in volv ing many intermed iate ste ps and

path ways and is not full y understood [63 -65]. Unl ike meth ano l, ethanol conta ins a C-C

bond th at must be broken in o rder fo r it to be full y oxidized into C02 . Thi s co mplete pro-

cess requires the transfer of 12 electrons fo r two mo l es of etha no l consumed. However

ma ny groups including our own have reported onl y partia l ox idation into acetaldehyde

(requi r ing a two e l ect ron transfer) and aceti c ac id (requiring a four electron transfer),

making the EOR extremely in effi c ient [66-69]. Along with the prob lem of the C-C bond

cleavage, similar barriers to those fo r MeOH oxida tio n a re also present. CO po iso ning of

the electrode has been reported for EOR [ 1 4, 55, 70] and crossover of ethanol from anode

to cathode causing a mi xed potentia l [7 1 , 72] has also been obse rved . These prob l ems

a l ong with recent advances to their so lutions and fu ture objectives are all disc ussed in thi s

work .

1.2. Types of Fuel Cells

1.2.1. Fuel Cell Composition

A fuel ce ll is a devi ce that directly converts c hemica l energy into electrical energy by feeding the ce ll with an extern al fue l. Its main components are a n anode, a cathode and an e lectro lyte membrane . The schematic in Figure 1.1 illustrates a typical fuel ce ll design.

T he fuel cell hardw are consists of two metal support blocks w ith etc hed fl ow fi elds. T he

a node and cathode catal ysts a re designed with the exact dim ensio ns of the flow fie ld to

minimize cata lyst material. In most cases, po lymer based gaskets are placed around th e

cata lyst to limit the fue l and /or cathode gas from leaching o ut of the flow fi elds. At the

a node plate, the fu e l is passe d direct ly through the plate to the flow fi e ld . Th e cathode gas

also flows directly to the flow fi eld . T he metal plate supports mu st be non-reactive w ith

the fue l a nd oxidant in o rder to optimi ze fuel ce ll reactions. Betwee n t he anode and cath-

ode cata lyst layers, there is an electro lyte membrane. W ith the help of the membrane, the

fue l and ox idant gas are able to react and the oxidation products ex it via the anode e fflu-

ent stream . T he e lectrons released from th e fuel, pass through a n external circuit (see

F igure 1.2) from the anode to cathode and proto ns a re transferred through the ce ll mem-

brane to the cathode . At the cathode, oxygen gas accepts the transferred e lectrons and

protons pro vided from the fue l and is co nverted to H

0 . Depending on the fu el used, the

number of electrons that are transferred and th e reaction products are va ried. T he proton

exchange membra nes (P EM) used in most fue l cells a nd in a ll of the work in thi s proj ect,

are Nation® membranes (see ref. [28-30]). These membranes aid in the transfe r of pro- tons from th e anode to the cathode .

Oxidation

Anode catalyst

products

+.--. l -- - - 1 ' ---"1

Fuel

entrance

Anode

Electrolyte membrane

Cathode catalyst

•• _Cathode

• •

• •

Oxidant

••

Figure 1.1: Schemati c of a typica l singl e cell so lid electrolyte fu el cell compos ition.

The fue l cell has many sim ilarities to a primary battery. T he main distin gui sh ing feature is th at a fu el cell uses a cont inuous fl ow of both fue l and ox idant, ma king them long last- ing devices. Batteri es are considered ene rgy storage devices, as they enclose the fuel and ox idant ins ide the cas ing. The lifetime and max imum amount of energy produced from a battery depend s on th e amount of active materi a l sto red and the rate at which it is co n- sumed. Once a ll the chemical reactants have reacted, the battery can no longer produce e nergy, and is use less . There also exist second ary batteries whi ch use exte rnal e l ectr i c ity to rep l enish th e electro active-materi als. A fuel cell could be cons idered a long l ast ing primary battery ifthe fuel co nta iner is included.

1.2.2. Proton Exchange Membrane Fuel Cells (PEMFC)

PEM FC (also referred to above as PEFC) are a type of fuel cell that relies on the use of a so lid thin laye r membrane to shuttle protons from the anode to the cathode. They are ide- al for use in low to moderate temperature env iro nments (usuall y up to I 00 °C).The most distingui shing part of a PEMFC is the m embrane electrode assembly (MEA). As shown in Figure I .2, the MEA co nsists of a cathode, a proton exchange membrane and an anode.

The thin membrane is sandwiched between the anode and cathode and the asse mbly is usually hot-pressed together. T he anode and cathode typicall y use carbon fiber paper as a s uppo rt, but g lassy carbon and carbon c loth hav e also been used. The anode and ca thode cata lyst laye rs a re usually comprised of a Pt-carbon mix or Pt-based all oyed Pt-metal- carbon mix.

Acting as probabl y the most important part of a P EMFC, the proton exchange membrane serves as both a proton conductor and the barrier separating the fu el and oxy- gen gas. Since Dupont's breakthrough with the Nafion® c lass of membranes, a significant amount of research has been conducted in an attempt to o ptimize the c haracteri stics and cost of the me mbrane [73, 74].

The essentia l characteristic of a proton exchange membrane is the ability to con-

duct protons w itho ut conducting e lectron s (which would short-out the fuel ce ll). T he

me mbrane should be equipped with ma ny proton condu cting s ide groups and a very sta-

ble backbone which ca n withstand heat, the fue l and most importantl y the harsh ox idative

enviro nment at the cathode [26, 75]. The membrane must also be resistant to fuel and gas

crossover, as any leakage in the membra ne would result in reduced efficiency . A low re-

s istance to H+ mi gration must also be present in the m embrane. However, a low re-

sistance could also lead to hydrod ynamic drag of H2 0 and possibly ethanol from the an- ode to the cathode, l eading to an in crease in fuel crossover. Dupont

Proton exchange membrane

Pt based anode catalyst layer

r · - - -

I Carbon fiber paper support

l. - - ·-

~---.=-==-~---·-··----··----

--- --- ---- · · ···--

~onfiberpapersup~

Pt based cathode catalyst layer

Figure 1.2: A schematic diagram of a typical MEA for a PEMFC.

successfull y constructed membranes that fulfill a ll of the above criteria w ith the deve l- opment of the Nati on® polymers shown in Figure 1.3.

As shown in th e chemical structure, Nation® was deve l oped by the polyme riza-

ti on of a perfluoranated vinyl eth er and tetrafluroethylene monomer [26]. These po l ymers

are usua ll y deriv ed from the - S0

F precurso r which is reacted with KOH in water and

then soaked into acid to form the active sid e chain -S0

-H+. With this very acti ve s ide

cha in, the Nation® membrane acts as a superac id , with th e ability to co nduct protons from

the anode and tra ns fe r them to the cathode with ease [26].

Figure 1.3: C hemical compositi on of the Nati on® membra ne developed by Dupont [26].

A lmost all anode catal ysts for P EMFC -HFC a re Pt (or Pt based all oys). T he abil- ity of Pt to oxidize H

into H+, shown in equations 1.1 and 1.2, ma kes it ideal for prov id- ing e lectron s a nd protons to th e cathode half-reaction.

2Pt

+

( I. I )

( 1 .2)

In H FCs, thi s is the o nl y reactio n that take place at the anode. T his reaction takes place spo ntaneo usly on the Pt electrode, makin g it extre me ly efficient. A few minor problems that can lea d to ineffi cienc ies in the a node ha lf -reaction of H FC are small amount of CO poisonin g and dehydration of the anode. It has been reported that during the reforming process (produ ction of H

sm all amount of CO are produced [76]. Even at very low con- centrati ons, CO has the abi I ity to compete fo r adsorption sites on the Pt anode, caus ing a decrease in ava ilable s ites for Ha d s, leadin g to inef ficie nc ies in the cell [53 ].

Until recent years, Pt and Pt based catalysts have been stri ctly used in the oxygen

reduction reaction (O RR) at the cathode. T hi s half- reactio n involves 0 2 be ing reduced in

the presence of protons and electrons transferred from the anode to produce

T he

reaction mec hani sm for the ORR has been extensively researched on many di ffe rent Pt

surf aces [77,78]. Vari ous mechani sti c pathways have been identified for the ORR reac-

tion, in cludin g a direct four electron tra nsfer and a multi step transfer of four electrons

[53, 79]. T he mechanism for this complex react ion is still not fu ll y understood, as theo-

retica l studi es hav e shown many different intermediate and radical formations prior to the

final products [80]. A simplified schematic of the possible pathways is depicted in Figure

1.4. T he 0

at the cat hode is first adsorbed onto the Pt s urface w ith breakage of the dou-

ble bond. Once the oxygen is adsorbed on the Pt surface, the reaction will take one of two

accepted pathways. T he desirabl e pathway, for a PEMFC, occurring at 1.23 V is a direct

four e lectron-four proton transfer to H

0. The seco nd possible pathway occ urring at a

mu ch lower t hermod ynamic e lectrode potential (0.70 V) is a two electron-two proton

transfer to produce adso rbed H

0

This mechanism is the preferred mechanism of the

ORR in industry for t he production of H

0

From this step, the adsorbed hydrogen per-

ox ide can take one of two paths. At hi gh potentials of arou nd I. 76 V , a second two elec-

tron-two proton reaction can take place to produced H

0 . The oth er possible pathway is

the desorption at the e lectrode, formi ng H

0

as a fina l product. Studies suggest that the

desorption product is not likely in PEMFC since onl y small amoun ts of

were found

in the O RR product analysis and the electron transfer number has been reported to be

close to 4. Eq uations 1.3- 1.5 summari ze the direct and indirect pathways of 0

reduction

to and H

0 and H

0

respectively. Equations 1.4 a nd 1.5 are coup led together to form

the indirect pathway [53].

( 1.3)

( 1.4)

( 1.5)

Comparing the complexity and potenti als needed of the ox idation and reduction reacti ons, it is c lear that the ORR is the rate determining step in HF C reactions.

T he hi gh cost of Pt m etal coupl ed with the large overpotent ial requi red fo r redu c- ti on at the cathode have attributed to the limitation of fuel cell s fo r many applicati ons.

O ver the past decade, many research group s have devoted ample time into the develop-

ment of new " low-Pt" [81-83] and " Pt-free" [84-86] cath ode cata lysts for ORR. These

catalysts have shown fairly good activity towards O RR in the absence of eth anol. How-

ever, when ethano l was introduced, large increases in ove rpotent ial were observed . At the

mome nt, the low ethanol tolerance coupled w ith s igni ficant decreases in perfor mance

(compared to Pt) have made these catalysts un suitab le for O RR in DEFCs .

) Pt Pt Pt Pt

o-o \ I

Pt Pt Pt Pt

2W )

H

H,

I

0

Pt Pt Pt Pt

2+•

H'o-o.-H

\ I

Pt Pt Pt Pt

l

Pt Pt Pt Pt

Figure 1.4: M echani stic scheme of the oxidation reduction reacti on occ urrin g at a Pt ca thode catalyst in an ac idic e nv ironment (adapted from [53]).

1.2.3. Direct Alcohol Fuel Cell (DAFC)

DAFCs wo rk in a very simil ar manner to that of the H FC, w ith the excepti on of sma ll a lcoho ls as the fuel at the a node instead of H

The components of DAFCs are the same as in HFCs w ith an anode, cathode and proton excha nge membrane. Many different an- ode and cathode catalysts have been developed in attempt s to optimize DAFCs efficien- c ies. As mentioned earli er, the ineffi c iencies of DAFCs ari se fr om th ree ma in prob le ms:

large ove rpotentials at t he cathode, the crossover of fuel from a node to cathode, and the

large overpotenti als for the alcohol ox idati on reaction (AOR) resul ting from adsorb ed intermed iates. T hese problems w ill be further investigated in the subsections of this work .

1.2.3.1. Direct Methanol Fuel Cell (DMFC)

DMFCs use the si mplest a lcoho l, methanol, as the fuel which is continually pumped to the anode of the fuel cell. Methanol reacts at the anode in the presence of water to pro- duce C0

6 protons and 6 e lectrons. T he ove rall half-reaction at the anode is described as:

( I .6)

Unlike the hyd rogen ox idation react ion (equati ons 1.1 and 1.2), methano l forms carbon based in termediates at the anode. These intermed iates signifi cantly lower the reacti on rate at the anode as they need oxyge n conta ining species, to be full y ox idized to C0

It can- not be determined for certa in the exact mechan ism that takes place in thi s reacti on but the most accepted mechanism based on electrochem ical t heory is described by equations 1.7 and 1.8 [6].

( I .7)

Pt - CO + Pt- OH

2Pt +

+ H + +

^{( 1.8)}

During MOR, CO is strong ly adsorbed o nto the Pt anode surface. The remova l of the CO

involves th e oxidation of water into surface bound hydroxyl radica ls wh ic h then react

with the CO to form C0

The major problem with the anode reaction is the large CO

coverage on the anode surface. Due to very stro ng interactions between th e CO a nd the Pt

electrode, the anode surface is easily and quickl y covered with CO. This poiso ns the sur- face, leaving limited Pt adsorption sites to oxidize the water, leading to very slow overall oxidation kinetics at the anode surface [51 , 56-58].

The reduction of oxygen at th e cathode is very similar to that of HFCs (equations 1.3-1.5). As mentioned in section 1.1 , the major problem in DMFCs arises from fuel crossover. Even at low concentrati ons, methanol can easily cross through th e PEM. As its concentration increases, the rate of methanol flux through the membrane is significantl y in creased. Furthermore, when the temperature of the DMFC increases, the flux in creases significantly [59]. When the methanol crosses from the a node to the cathode, it can be rapidly consumed by oxygen. Thi s not only wastes fuel but also leads to a ·'mixed poten- tial " at the cathode resulting in a s i gnificant decrease in reducti on potential fo r 0

and a decrease in overall cell perf o rmance. Although the mechanism for the oxidation of meth- anol (and ethano l) is fairl y uncertain at the cathode, the reduction of the ce ll potential is indicati ve of the oxidation tak ing place. Furthermore, since low fuel co ncentrations and temperatures are needed to limit the effects of crossover, the overal l number of electrons transferred and the kinetics of the overa ll reaction a re decreased, hence reducing the max- imum current.

1.2.3.2. Direct Ethanol Fuel Cell (DEFC)

All experim enta l res ults in this work use DEFCs, and these from here on out wi ll be the

focus of this report. DEFCs use the second sma llest molar mass alcohol, ethanol, as the

fuel at the anode and 0

norm all y in air at the catho de as th e ox idant. Since ethano l is a

larger molecule than methanol, more e l ectrons need to be transferred to ful ly oxidize it to