Fuel Cells in Energy Technology (6) Werner Schindler Department of Physics Nonequilibrium Chemical Physics TU München summer term 2013

Transcription

1 Fuel Cells in Energy Technology (6) Werner Schindler Department of Physics Nonequilibrium Chemical Physics TU München summer term 2013

2 Energy losses in a hydrogen economy Energy Losses in a Hydrogen Economy Energy efficiency Regeneratively produced electricity 0% % conversion to DC Gaseous hydrogen produced by electrolysis compressed for transport after transport transfer into hydrogen tanks (for internal combustion engines) transfer and conversion to electrical energy in fuel cells - in DC - in AC share of energy spent for conversion and distribution after Ulf Bossel 2

3 Energy density of selected fuels (secondary energy carriers) LH 2 = liquid hydrogen Without container 33.3 kwh/kg ~3 times diesel fuel 3

4 PEM-FC and Direct Alcohol Fuel Cells CH 3 OH + H 2 O 6H + + 6e - + CO 2 3/2 O 2 + 6H + + 6e - H 2 O Source: J. Larminie, A. Dicks: Fuel cell systems explained. 2nd ed., Wiley,

5 The direct methanol fuel cell (DMFC) CH 3 OH + H 2 O 6H + + 6e - + CO 2 3/2 O 2 + 6H + + 6e - H 2 O Source: J. Larminie, A. Dicks: Fuel cell systems explained. 2nd ed., Wiley,

6 The direct methanol fuel cell (DMFC) G. Hoogers, Fuel cell technology Handbook 6

7 The direct methanol fuel cell (DMFC) for portable and mobile applications sole fuel cell with liquid reactant (fuel) much lower current and power density than PEMFC (10%) Transport of liquid methanol through GDL! PtRu-catalysts (15-50% Ru): Methanol is not adsorbed at Ru more water transported through the membrane Anode-GDL: Transport of methanol and water to the catalyst layer removal of CO 2 7

8 The direct methanol fuel cell (DMFC) Pt Pt/Ru lower overpotentials with Pt/Ru nevertheless large overpotentials for oxidation G. Hoogers, Fuel cell technology Handbook 8

9 The direct methanol fuel cell (DMFC) Losses: 50% activation losses! mass transport: liquid methanol in, CO 2 out! 2-4 wt.% Methanol in water at anode due to methanol-crossover: mixed potential! G. Hoogers, Fuel cell technology Handbook 9

10 The direct methanol fuel cell (DMFC) rate determining step O Indirect reaction path: formation of formaldehyde (CH 2 O) and formic acid () G. Hoogers, Fuel cell technology Handbook 10

11 Mixed potential due to CH 3 OH Crossover Boundary conditions at open circuit for positive terminal ( cathode ): total current density = 0 one electrode potential E(j=0) E eq two redox couples present partial current densities for both partial reactions unique function of electrode potential (e.g. Butler-Volmer-equation) CH 3 OH + H 2 O 6H + + 6e - + CO 2 ln j j cath <0 3/2 O 2 + 6H + + 6e - H 2 O j an +j cath =0 mixed potential j an >0 E eq (CH 3 OH/CO 2 ) E eq (O 2 /H 2 O) 11 E

12 Different approaches for the methanol feed This system uses micro-pumps and micro-pipes increasing the size, complexity and cost of the system. This system design, carrying water with the fuel severely reduces the system's energy density. Source: (June 9, 2009) 12

13 Different approaches for the methanol feed Source: T.S. Zhao et al. / J. Power Sources 195 (2010)

14 A vapor fed direct methanol fuel cell (passive design) Source: B. Xiao, A. Faghri: Numerical analysis for a vapor feed miniature direct methanol fuel cell system. Int. J. Heat Mass Transfer 52 (2009)

15 Direct methanol fuel cells (DMFCs) DMFC systems are commercially available System efficiencies are sufficient for the application but can be considerably improved Need of research Improved anode catalysts for higher cell voltage (at the same current density) Improved cathode catalysts that tolerate cross-over methanol 15

16 Disadvantages of direct methanol fuel cells (DMFCs) Sluggish kinetics of methanol oxidation Methanol crossover from anode to cathode Sluggish kinetics of oxygen reduction at cathode Permeated methanol that is oxidised at the Pt-based cathode catalyst leads to a mixed potential Challenge to find highly catalytically active and methanol tolerant electrocatalysts for the oxygen reduction reaction. One candidate is ruthenium selenide, RuSe* (*) N. Alonso-Vante et al., J. Electrochem. Soc. 138 (1991)

17 A more detailed look on some commercial DMFCs The energy content of pure methanol is 4.4 kwh/l or 5.5 kwh/kg (without container) Efficiency 1.11/ Source: (May 17, 2009) SFC Smart Fuel Cell AG (Brunnthal) 17

18 A more detailed look on some commercial DMFCs Data from 18

19 A more detailed look on some commercial DMFCs The energy content of pure methanol is 4.4 kwh/l Efficiency 1.4/ Source: (June 9, 2009) MTI Micro (Albany, NY, USA) 19

20 Influence of Anode Catalyst Loading Catalyst loading: amount of catalysts on the electrode (e.g. anode) 900 Anode: 5,3 mg/cm2 800 Anode: 2,1 mg/cm2 Anode: 1,1 mg/cm ma/cm2 U / mv i / ma/cm Anode loading / mg/cm 2 i@ 500 mv 400 mv i / ma/cm 2 Membrane: Nafion 105 T c : 110 C T bef : 80 C c m = 1 mol/l Oxidant: Air Anode: JM Pt/Ru Kathode: JM Pt-black 6,4 mg/cm 2 F m : 4 ml/min F a : p m : p a : 1,5 l/min 2,5 bar 4,0 bar 20

21 Temperature dependence of the current-voltage characteristics of a direct methanol fuel cell Improved kinetics 21

22 The oxygen reduction reaction in DMFCs Methanol is able to cross-over from the anode to the cathode side This results in: Losses of fuel lower conversion efficiency higher fuel costs Parasitic oxidation of methanol at the cathode generation of heat Poisoning of Pt-catalysts and thus limited oxygen reduction efficiency Possible solutions of the problems: Using selective catalysts for ORR (catalysts which reduce oxygen to water but do not oxidize methanol) Ruthenium (Ru) is methanol tolerant but not very active Ruthenium (Ru) is less active than platinum (Pt) because it is oxidized to RuO x Candidate material: ruthenium selenide (Ru x Se y ) Selenium in Ru x Se y seams to prevent oxidation of Ru and thus enhances activity 22

23 The search for suitable oxygen reduction catalysts Performance of Pt/C deteriorates in the presence of methanol RuSe/C is not affected by the presence of methanol Dispersion of the RuSe nanoparticles on the carbon support is still a bit poor Better dispersion of RuSe nanoparticles should yield a higher active area and therefore a higher oxygen reduction activity M. Neergat, D. Leveratto and U. Stimming, Fuel Cells 2 (2002) 1 23

24 Direct ethanol fuel cell Much less toxic than methanol High energy density (including container) Ethanol: 8.0 kwh/kg or 6.3 kwh/l H 2 (30 MPa): 2.46 kwh/kg or 0.75kWh/l Gasoline: 12.7 kwh/kg or 8.7 kwh/l Available from renewable resources Problem: No good catalyst identified up to now 24

25 The search for a suitable catalyst for the ethanol oxidation Source: Antolini, E. J. Power Sources 170 (2007) 1. Dependence of the maximum power density of DEFCs with Pt/C, Pt Ru (1:1) and Pt Sn/C (3:1) as anode catalysts on cell temperature. O 2 pressure: 3 atm; ethanol solution: 1 mol l 1. Anode metal loading 1 mg cm 2. Cathode 20 wt.% Pt/C, Pt loading 1 mg cm 2. 25

26 The search for a suitable catalyst for the ethanol oxidation Pt/C, 0.1 M EtOH Source: V. Rao et al., J. Electrochem. Soc. 154 (2007) B1138-B1147. Coupling of test cell anode outlet with mass spectrometer (Differential Electrochemical Mass Spectrometry) Determination of CO 2 -efficiency 26

27 The search for a suitable catalyst for the ethanol oxidation Source: V. Rao et al., J. Electrochem. Soc. 154 (2007) B1138-B1147. Pt-Sn promising catalyst at low temperatures 27

28 Requirements for a good DEFC Operation at temperatures >> 150 C to enhance C-C bond breakage catalyst testing at elevated temperatures new membranes suitable for operation at higher temperatures non-carbon supports Catalysts with high CO 2 efficiency and high current efficiency 28

29 So far there are only prototypes of direct ethanol fuel cell stacks The "Schluckspecht", an experimental electric car powered by a direct ethanol fuel cell stack. Developed at the University of applied sciences Offenburg. 29

30 A DAFC stack developed at the ZAE Gregor Rupp (ZAE Bayern) Stack with 10 cells with 5x5 cm 2, total weight < 1 kg; HighSpec M200 MEA by JMFCs (Anode: PtRu/C, Cathode Pt/C); meander flow fields. 30

31 The ZAE prototype running on methanol 1 mol/l methanol 31

32 The ZAE prototype running on ethanol 1 mol/l ethanol 32

33 A short comparitive summary for PEMFCs and DAFCs Polymer electrolyte fuel cell (PEMFC) running on H 2 and O 2 (or air) + Reliable and available on the market + Power density around 1000 mw cm -2 - Problem of hydrogen storage and supply not resolved completely Direct methanol fuel cell (DMFC) + Available on the market + Infrastructure for the fuel is available or relatively easy to realize + Methanol can be produced from renewable sources + Biodegradability of the fuel - Power density relatively low (around 60 mw cm -2 ) - High content of noble metals - Toxicity of methanol Direct ethanol fuel cell (DEFC) + Fuel: no severe toxicity issues, biodegradable - So far only prototypes 33

34 Electrode Reactions and Electrocatalysis 1 Taken from: Molecular Level Picture of the Hydrogen Evolution Egill Skúlason, Gustav S. Karlberg, Jan Rossmeisl, Thomas Bligaard, Jeff Greeley, Hannes Jónsson, Jens K. Nørskov; 34

35 Electrocatalysis The problems related to the increase of rates of electrochemical reactions, or, to put it another way, the decrease of overpotential, needed to perform reactions at a given rate are the subject of electrocatalysis. Both increase and decrease are of considerable practical importance since they affect the economics of electrochemical processes. Source: Encyclopædia Britannica Encyclopædia Britannica Online. 12-May-2009 < 35

36 Recent Special Issues on Electrocatalysis 36

37 Recent Special Issues on Electrocatalysis Electrochimica Acta, Volume 55, Issue 26, Pages (1 November 2010) RECENT ADVANCES IN ELECTROCATALYSIS AND PHOTOELECTROCATALYSIS 37

38 What determines the electrocatalytic activity of a fuel cell electrode? Anode reaction: e.g. hydrogen oxidation reaction (HOR): H 2 2H e - Reactant H 2 2 e - Product 2 H + 2 H + Nafion: proton conductor Pt Pt Substrate (e.g. carbon): electron conductor Efficient mass transport needed High electron conductivity needed High proton conductivity needed Catalyst particle: location of electron transfer Exchange current density (j 0 ) and thus the reaction rate depend on: Composition of the catalysts Particle size, number of defects Chemistry and physics of the substrate 38

39 Elementary steps in electrocatalytic reactions Diffusion of reactant to the electrode surface In some cases followed by adsorption (physical and/or chemical) Surface reaction (Langmuir-Hinshelwood mechanism; Eley-Ridealmechanism) in electrochemical reactions always related with an electron transfer Desorption and diffusion of the product species Binding interaction with the surface by: Physical adsorption: Van-der-Waals forces Chemical adsorption: stronger binding interaction with the surface via e.g. two centres two electron interaction 39

40 Mechanisms of heterogenous reactions Langmuir-Hinshelwood mechanism Both molecules adsorb and the adsorbed molecules undergo a bimolecular reaction A A ad B B ad A ad + B ad Products Eley-Rideal mechanism Only one of the molecules adsorbs and the other one reacts with it directly, without adsorbing A A ad A ad + B Products 40

41 Elementary steps of the hydrogen evolution reaction (HER) Taken from: Molecular Level Picture of the Hydrogen Evolution Egill Skúlason, Gustav S. Karlberg, Jan Rossmeisl, Thomas Bligaard, Jeff Greeley, Hannes Jónsson, Jens K. Nørskov; 41

42 Influence of the catalyst material: the Sabatier principle There is an optimum of the rate of a catalytic reaction as a function of the heat of adsorption - Paul Sabatier, 1905 If the adsorption of intermediates is too weak, the catalyst has little effect. If it is too strong, the adsorbates will be unable to desorb from the surface ( poisoning of the surface). Hence, the interaction of the intermediates with the surface should be neither too strong nor too weak. 42

43 How to understand catalyst materials? Computational methods are able to deliver insights into parameters controlling electrocatalytic activity DFT-calculations are able to predict key parameters of novel catalyst materials High-throughput computational screening [1] based on DFT calculations combined with an experimental screening allows identifying novel catalysts [1] J. Greeley et al.: Nature Materials 5 (2006)

44 An interesting paper on computational screening of electrocatalytic materials Nature Materials 5 (2006)

45 Current density ~ reactivity Influence of the catalyst material: the volcano curve Binding is too strong Binding is too weak Source: J. Greeley, T. F. Jaramillo, J. Bonde, I. B. Chorkendorff, J. K. Norskov, Nature Materials 5 (2006)

46 Finding novel highly efficient catalysts: DFT studies DFT calculations in order to identify binary alloys with optimal binding interactions (e.g. hydrogen) Binding interaction is an important factor DFT calculations allow to identify promising catalysts: Criterion ΔG H max. 0.3 ev Source: J. Greeley et al., Nature Materials 5 (2006) 909 PtBi looks promising 46

47 The good electrocatalytic activity of PtBi is confirmed by experimental data Source: J. Greeley et al., Nature Materials 5 (2006)