Fu Fuel l Ce Cell lls R s Revisi visited ted Lo Lost t in - - PowerPoint PPT Presentation

fu fuel l ce cell lls r s revisi visited ted
SMART_READER_LITE
LIVE PREVIEW

Fu Fuel l Ce Cell lls R s Revisi visited ted Lo Lost t in - - PowerPoint PPT Presentation

Fu Fuel l Ce Cell lls R s Revisi visited ted Lo Lost t in in Sp Spac ace e or Fin inally lly dow own n to o ear arth th? DOMINIC GERVASIO Department of Chemical & Environmental Energy 1 Fu Fuel el Cel ells ls


slide-1
SLIDE 1

1

Fu Fuel l Ce Cell lls R s Revisi visited ted

Lo Lost t in in Sp Spac ace e

  • r

Fin inally lly dow

  • wn

n to

  • ear

arth th?

DOMINIC GERVASIO

Department of Chemical & Environmental Energy

slide-2
SLIDE 2

Fu Fuel el Cel ells ls Revisited: visited: Lost

t in Sp Spac ace e or final ally ly down wn to to ea earth th?

?

Dominic F. Gervasio

Department of Chemical and Environmental Engineering University of Arizona, Tucson, Arizona USA

2

slide-3
SLIDE 3

3

  • Fuel cell versus battery
  • Evolution of fuel cells … in space…size and weight reduction
  • Special consideration for earth-bound fuel cells… air not O2
  • DU

DUUHH!! !! … A fuel cell power source is a syste tem

  • The critical component – the electrolyte membrane
  • Prospects for future development

Outli line ne

slide-4
SLIDE 4

4

Batteries: use dissimilar metals to generate cell voltage… simple, but metals are not energy rich Fuel Cells: dissimilar fuels generate voltage…. fuels are energy rich, but not simple to use on Earth

Energy rich fuels in fuel cells give lower size and weight with increasing application duration

Application Duration (Energy Use Requirements) System Mass Fuel + Stack Batteries

Fuel Ed (Wh/kg)

Hydrogen 33,000 Diesel Fuel 13,200 Methanol 6,200 NaBH4-30% 2,500 TNT 1,400

Battery: Ed (Wh/kg)

Rechargeable (est. max.) 200 Li/SO2 Battery (primary) 176 Alkaline Battery (primary) 80 Nickel-Cadmium (secondary) 40 3lb

H2 fuel cell

vs

Li battery

Fuel cells and batteries are both DC power supplies…what’s the difference?

slide-5
SLIDE 5

5

Consequences ?

slide-6
SLIDE 6

6

Gas tank Battery Battery Battery Battery Battery Battery Battery Battery Battery Battery

Gas Tank is 1/10 the size of car for 400 mile trip with quick refill (minutes) Battery “Gas Tank” is size of car for 400 miles

Ga Gasoline

  • line versus

sus batt ttery ery pow

  • wer

ered ed car

  • 1. Gasoline is cheap compared to batteries

One gallon of gasoline equals 6.25 lbs 6.25 / 2.2046 lb/kg = 2.835 kilograms 13,200 Whr/kg x 2.835 kg/gal = 37422 Wh/gal x $3/gal = < $0.08 per kWh Lithium battery > $20/kWh ( or is equivalent to $750 per “gallon” of electricity) ; NOT counting cost of charging from grid

  • 2. Gasoline gives more energy per kilogram compared to batteries

Battery only cars just won’t do !

There is nowhere

  • to put your golf clubs
  • or even for passengers to sit!

with slow refill (hours)

slide-7
SLIDE 7

7

Question: If a battery won’t do as the primary power source substitute for an ICE powered car … then what can you use? Answer: a fuel fed electric-powered car

… a Fuel-cell cell ca car

Q & A Q & A

slide-8
SLIDE 8

8

Hydr drog

  • gen

en fue uel l us used d in in fu fuel l cell lls s is is co comp mpar arable le to

  • gasoli

soline ne

Volumes of different Fuels equivalent to ~10 Watt

  • hours of Electrical Energy

at 100% Chemical to Electrical Conversion Efficiency. 3.33 liter H2-uncompressed

H 2

Liquid compressed hydrogen 0.5 ml H2 Solid AB5 Chemical Hydride 6 ml 2 ml Methanol-50% 1 ml Gasoline Li - ion Battery 30 ml

3 ml NaBH4-30%

Gas or Hydrogen tank

slide-9
SLIDE 9

9

Ano nother ther advanta antage e of

  • f fue

uel l cells lls … reduced pollution

slide-10
SLIDE 10

10

yields ds no added CO2 in the atmos

  • sph

pher ere Plant nts s re-abso sorb b CO2 from m atmosph mospher ere adds CO2 to the atmos mospher ere CO CO2 level els s conti tinua ually lly increase ease

Power from non-renewable fossil fuel

Power from renewable bio fuel

10

CO CO2 added ed CO CO2 remo moved ed

No CO2 removed

CO CO2 added ed

So an added benefit of hydrogen

Low pol

  • llu

luti tion

  • n

Hydrogen can be derived from… Renewable sources

  • Biomass
  • Solar
  • Wind

… or from

  • nuclear driven electrolysis
  • efficient reforming of

hydrocarbons

slide-11
SLIDE 11

11

Next xt … evolution of fuel cell design

slide-12
SLIDE 12

❑ Sir William Grove demonstrated the first fuel cell in 1839,

  • addition of electricity to Pt catalyzed electrodes in acidic water to make H2 and O2
  • and reaction of hydrogen anode and oxygen cathode to make water and generate

electricity ❑ In 1959, British engineer Francis Thomas Bacon successfully demonstrated a 5 kW stationary fuel cell using hydrogen and oxygen feeds with alkaline (aqueous KOH) electrolyte operating at 205 ° C (400 °F) and 414 N/cm2 (600 psi)

  • stationary power supply for a welder, saw and lift
  • but too heavy and too high pressure for space

❑ In 1965, the Gemini V spacecraft used the first fuel cell in space

  • 1 kilowatt proton exchange membrane (PEM) fuel cell
  • eliminated excess weight and volume lifted in orbit for 2 things:
  • i. batteries for required electrical power
  • ii. drinking water for astronauts

first t fuel l cell l in space Bacon Grove (PEM) M) made by GE

12

Evolut

  • lution

ion of

  • f fue

uel l ce cells lls

slide-13
SLIDE 13

First st fuel el cell ll in space: ce: PEM made e by GE for r the NASA A Gemini mini program am

  • in 1965, the Gemini V spacecraft was the first spacecraft to use fuel cells
  • pure hydrogen and oxygen reactants came from the propulsion system
  • sulfonated polystyrene resin polymer electrolyte membrane (PEM); with wicks to draw

water out of cell for astronauts’ drinking water (1 kWh electrical energy → 1/2 L liquid H2O)

  • so solved 2 problems: eliminated weight and volume of i) batteries & ii) water lifted into orbit
  • fuel cell used 28 mg of platinum catalyst per cm2 of electrode
  • maximum power output per was about one kilowatt operated at about 65 °C
  • 32 individual cells in series produced 23 to 26 volts per stack (efficiency = 60 to 70%) per

module and 3 modules were used in each mission, needed 2 for mission and 1 to land

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20040010319.pdf Fuel Cells For Space Science Applications , Kenneth A. Burke , National Aeronautics and Space Administration, Glenn Research Center, Cleveland, Ohio 44135 (2003)

Heat generated by the fuel cell stack was removed by a circulating coolant. P(H2) 1.7 psi above P(water), P(O2) 0.5 psi above H2 anode and cathode Ti screens with unsupported Pt catalysts with PTFE.

Module: 47 x 37.5 x 63.5cm (18 1/2 x 14 3/4 x 25 in) 112 liter = ~ 225 kg = ~ 500 lb 13

slide-14
SLIDE 14

Lesson of space program – reduce the size and weight of fuel cells

500

Gemini -1 kW

Year

Hypothetical State of Art alkaline Modified Bacon alkaline PEM acid

Specific Weight (lbs./kW) →

3) 7 kW average weighs 250 lb. (Pd = 0.062 kW/kg) in Space Shuttle

and Orbiter 32 percent KOH in an asbestos matrix operating pressure is 60 psia

by using 20 mg/cm2 Au Pt alloy catalyst on cathode and l0 mg/cm2 Pt on anode.

  • Note: internal combustion engine is about 450 lbs for 150 hp or Pd =0.7 kW/kg
  • cell temperature reduced to 93 °C. required gold plated Ni screens upon which

a catalyst layer and PTFE for gas passages to catalyst.

  • heat generated is transferred to a fluorinated hydrocarbon dielectric liquid to the

Orbiter's heat exchangers to a Freon coolant system

  • 3’rd system still used TODAY;

4) A fourth hypothetical “regenerative system” (fuel cell/electrolyzer)

  • has a marginally better alkaline fuel cell
  • photovoltaic (PV) power to electrolyze water to regenerate hydrogen and
  • xygen so alkaline fuel cell can make electrical power during sundown.

2) 1935 Bacon cell patent acquired by Pratt & Whitney (United Technologies) and modified for Apollo mission used Ni anode and Li-Ni cathode and KOH electrolyte

  • the pressure was lowered from 600 to 50 psi; KOH concentration increased from 30 to 75%
  • temperature was raised to 260 °C; a current density of 150 A/ft 2 the voltage was 0.87 V/cell
  • Heat and water removed by hydrogen circulation; with glycol-water secondary coolant loop
  • 220 lb. for 1420 W (Pd = 0.014 kW /kg) Peak power was 2295 W at 20.5 V.
  • KOH-H20 electrolyte solution was pressurized to 53.5 psia while each reactant gas cavity was

maintained at 63 psia. 1) The 1960 Gemini Pt catalyzed proton exchange membrane (PEM )fuel cell system

  • power density : 500 pounds for 1 kW or 1 kW/250kg =0.004kW/kg; 1 kW/100 liter = Pd = 0.01 kW/liter

increasing power density decreasing size and weight

14

slide-15
SLIDE 15

But alkaline fuel cells

❑ require pure O2; cannot use air with acid CO2 which reacts with basic KOH making carbonate …which is not efficient at 200oC ❑ require pure H2, compact liquid H2 from cryogenic rocket fuel available in the cold of space … but H2 gas voluminous on earth ❑ liquid electrolytes can leak under shock and vibration on earth … ionically shoring stack with drastic power reduction ❑ earth-bound fuel cells need to operate ten of thousands of hours,

  • alkaline fuel cell catalysis is more active in the short term…BUT decays after about 500 hours !

After 60 years of fuel cell research for space… the 3rd alkaline fuel cell is still the best in space

Diagram of an Alkaline Fuel Cell. 1: Hydrogen 2:Electron flow in wire 3:Load (light bulb, radio, etc.) 4:Oxygen 5:Cathode 6:Electrolyte (liquid aqueous KOH solution) 7:Anode 8:Water (produced by cell) 9:Hydroxyl Ions (OH- from KOH)

15

Liquid alkaline fuel cell is not acceptable for earth-bound use

slide-16
SLIDE 16

16

Let’s get down to earth

slide-17
SLIDE 17

For use on Earth …. its back to PEM fuel cells

To make medium sized (100kW) fuel cells for automotive and distributed generators, we need :

  • hydrogen storage for the anode and an air-fed CO2-tolerant cathode
  • therefore a polymer electrolyte membrane (PEM) is required, which is either
  • an acidic (proton conducting) membrane electrolyte to reject CO2 and have more stable Pt catalysts
  • or an alkaline (anion exchange) membrane electrolyte with low CO2 solubility to preserve OH-conduction in air

Wish sh list

  • faster and temporally-stable and hopefully low-cost catalysts for H2 oxidation and O2 reduction
  • water-free high-temperature ion-conductors (electrolytes) to

1) minimize system bulk by shrinking radiator size 2) further elimination of bulk, weight and parasitic losses by eliminating the use of a humidifier

17

The 2 kinds of polymer electrolyte membrane (PEM)

1) proton-exchange PEM, proton (H+) passes through a cation exchange membrane 2) hydroxide-conducting PEM, hydroxide (OH-) passes through anion exchange membrane

cation exchange membrane

Anode: 2H2 → 4H+ + 4e- Cathode: O2 + 4H+ + 4e-  2 H2O

Much has been learned from space fuel cells … where does this leave us for earth bound fuel cells? in short… a tall order…seems impossible…or is it?

slide-18
SLIDE 18

18

Co Comparison mparison of

  • f PE

PEM fue uel l ce cell ll system stems

SOFC High Temp. PEM Nafion PEM

Operating Temperature 25 to 80oC 140 to 200oC 700 to 900oC Ambient Temp. Tolerance freezes/dries out invariant invariant Power (at ambient P) 0.1 to 0.35 0.12 to 0.2 (new membranes) 0.5 to 1 W/cm2 Water management tricky easy easy Gas feeds needs humidified dry dry Thermal management difficult (dry out) easy moderate (900oC) Radiator large small small Thermal Mechanical Stress shrink/swell small large stress Shock and vibration tolerant tolerant intolerant Cell Efficiency 60% 50% - 80% (new membranes) 65% Electrolyte (ions) Fixed ions Fixed ions (new membranes) Fixed ions Electrolyte (solvent) needs water none none Membrane technology mature developing developing Material Cost Nafion $700/m^2 < $ 1/10 Nafion stack < $ 1/10 Nafion stack Processing cost low low high (seals) Catalyst cost $250/kW (Pt)

$250/kW(Pt) $10 (if non Pt) $250/kW (Pt) $10 (if non Pt)

Combustion toxic non-toxic does not burn Operating life 1000 to 40,000h 20,000 h ? Thermal & Mechanical shock Contaminant Sensitivity sensitive insensitive insensitive CO sensitivity 50ppm 2% 40% Stack size >100 W/kg > 360 W/kg (metal BPP) >100W/kg Stack volume > 200 W/liter > 690 liter (metal BPP) > 500 W/liter Reformer size large (CO clean up) small small t (cold start to full power) minutes (hydrate) minutes (heat up in metal BPP hours (heat up to Temp.) Overall system large, heavy simpler, smaller larger, heavy

Attribute red is bad green is good yellow is neutral

slide-19
SLIDE 19

A fuel el ce cell PO POWER ER SO SOURCE CE is a a SYSTEM

H2 loop Air loop

  • there is a need for humidifier to keep ion exchange

membrane ionized and conductive.

  • A very large radiator is needed for rejecting waste heat

from low temperature fuel cell to the ambient For low temperature PEM fuel cell (LT-PEM FC) system For high temperature PEM fuel cell (HT-PEM FC) system

  • higher temperature fuel cells tolerate CO in

hydrogen from natural gas can be used instead of purified hydrogen.

  • heat management simplified and minimized
  • water management is simplified or eliminated
  • electrode reaction kinetics are faster at higher

temperatures (higher power) which opens door to higher power and cheaper (less) catalyst material

System considerations lead to HT PEM (NASA 250oC is best) BIG QUESTION: Is there a suitable ionomer membranes for a HT PEM fuel cell?

19

Parasitic losses

slide-20
SLIDE 20

20

Su Summary ary of ad advan anta tages es of HT PE PEM T Tec echnology nology

❑ Lower temperature operation for HT PEM FC (170°C) than turbine (> 800oC) or solid oxide fuel cell (> 800oC) ❑ Smaller and lighter than room temperature fuel cell: small radiator, no humidifier ❑ Quieter, more efficient and lower thermal signature than internal combustion engine ❑ Lower cost fuel cell membrane (hydrocarbon) than room temperature fuel cell membrane (fluorocarbon) ❑ Higher system efficiency due to good thermal integration to fuel source (e.g., hydrocarbon reformer) ❑ Higher system efficiency due to one-phase (gas) fluidics ❑ Shock and vibration tolerant, made with plastics and metals not brittle ceramics ❑ Faster startup time to full power than solid oxide fuel cell… even low temperature PEM ❑ Simplified system design (no membrane humidification, low working pressures)

slide-21
SLIDE 21

21

Operating at elevated temperatures is not solely beneficial however, because higher temperatures can result in more rapid component degradation, including

  • corrosion of bipolar plates *
  • catalyst support,*
  • degradation of seals, *
  • degradation of membranes**

Su Summary ary of possible le disad advanta antages es of HT PE PEM Tec echnology nology

slide-22
SLIDE 22

22

Ot Othe her r Com

  • mponen
  • nents

ts for

  • r PE

PEM M Fu Fuel l Cell ll St Stac ack Assemb sembly ly

  • Fuel cell stack has several hundred layers of thin materials with its own dimensional tolerances

and assembly and material cycling tolerances can result in poor sealing and stack failures.

  • To address these issues, novel developments include:

1.

metal bipolar plates with parallel flow field

2.

stable gaskets with asymmetric design to prevent parabolic flow

3.

membrane electrode assemblies (MEAs) which need no water for proton conduction. One metal BPP geometry PEM stack with conventional milled graphite BPP

slide-23
SLIDE 23

23 Weight %

  • f Metals vs Exposure Time in 85%

H3PO4 at 150oC in air

10 20 30 40 50 60 70 80 90 100 200 400 600 800 1000 1200

Time (hrs) Weight Percent %

Ti Ni SS316 SS310 C22 C276

Corrosion testing of met etal allic ic bipola

  • lar

r plate te mate terial ial

  • C276 Alloy selected from corrosion data
  • C276 bipolar plates are conductive and stable with 50 nanometer of gold plate
  • Resistance only 0.6 Ohms as measured on a 1 x 2 x 0.1 cm plate with probes 1 cm apart

Top 2 metal passivate and stop corroding corroding

slide-24
SLIDE 24

24

Fuel cell AMREL Electronic Load Control and Recording Computer Set up to monitor power

Tes est t sta tati tion n for HT PE PEM Fuel el Ce Cell St Stac ack wi with th Met etal al Bipol

  • lar

ar Pl Plate te

End plate

20 cells with

  • metal BPPs
  • seals
  • no cooling plates

End plate

slide-25
SLIDE 25

25

100 150 200 20 40 60 80 100 120 140 160

Total 5-cell Stack Power (Watts) Time (minutes)

MEA is BASF C3 phosphoric-acid-loaded polybenzimidazole (PA PBI) with porous graphite gas-fed electrodes with 0.5 mg Pt/cm2 Area per cell = 250 cm2, Stack voltage 3 volt. Total current =58A. Gas feeds: 3.6 LPM hydrogen 1.8 LPM oxygen. P =30psig. Cell voltage 0.6 volt. Current density = 235 mA/cm2. Raw Pdensity per cell = 140 mW/cm2 (not IR corrected). T = 172oC.

St Start t up power er for r a passiv ssively ely cooled led 5-ce cell l HT PE PEM st stack k

  • phosphoric acid (H3PO4) loaded polybenzimidazole (PA PBI) housed in metallic bipolar plates with PFA gaskets at 172oC.

Decent stack performance with PA PBI membrane … but H3PO4

  • can wash out of PBI membrane giving:
  • voltage drop from resistive losses
  • voltage loss due to short circuiting
  • limit temperature: 210oC > T > 140oC
  • too low water wash
  • too high H3PO4 condensation
slide-26
SLIDE 26

26

Sc Scal ale up e up of fuel el-cell cell sta tacks hous use e in met etal al bipol polar ar plate tes

100 200 300 400 500 600 5 10 15 20 25 40 80 120 160 200 240

Polarization Curve per cell for last 20-cell stack in Metal BPPs ; 3-17-2011 A = 250 cm2.,T = 145C, F (H2) = 9LPM,7psig; F (O2)=7LPM,7psig, OCV 21.2 volt

Stack Voltage Stack Power Stack Voltage (volts) Stack Power (Watts)

Current Density (milli-Amps per cm2)

10 cell - cooled 20 cell – NOT cooled

produced almost 700 Watts. passively-cooled with metal fins; produced > 350 Watts Pdensity/cell = 0.2 A/cm2.

Need an actively-cooled stack using a flowing liquid in a metal BPP

Can be done by laminating 1 extra metal foil to each bipolar plate to allow rapid thermostatting for:

  • heating stack (by burning some H2 and O2) for rapid start-up (less than 1 minute) of fuel cell power-source system
  • cooling stack to keep MEA and cell temperature even to stop (uneven “thermal run-away”) for safe high-power operation of the stack.
slide-27
SLIDE 27

27

What’s next?

slide-28
SLIDE 28

28

Specific and Volumetric power densities with metal versus graphite bipolar plates. Metallic Graphite Specific Power (Watts / kg) 365 169 Power Density (Watts/liter) 691 396

De Design of 1.

  • 1. th

thermos ermosta tatted tted met etal alli lic c bipola

  • lar

r plate te wi with th 2.

  • 2. PF

PFA gas asket et

Top of gasket Bottom of gasket 22 channels @ ~7mil deep and 4-5mil wide Inlet ¼” 0.100” ring

1.Ther hermo mo-sta statt tted ed metal alli lic c bi bipo pola lar pl plate te (BPP PP)

  • with brazed internal cooling channels (or 3-D metal printing)

2.PFA A gasket with th radia ial l rib desig ign Ribbed gasket pattern (alternating land and channel) for mechanical support and low pressure drop from gas inlets to outlets

Tested with CDA@4-5psi, Flow @ inlet= 98sccm, Flow @

  • utlet= 92sccm

Using metal instead of graphite can:

  • Lower cost from $100 to $1 per plate
  • Lower weight and volume leading to higher specific and

volumetric power densities as shown in Table

4 mil thick metal stamped and brazed, for thin light BPP.

Using Perfluoroalkoxy (PFA) sheet in an asymmetric gasket :

  • evenly spreads gas with low pressure drop in parallel flow field
  • gives a stable and tight seal to 350oC
slide-29
SLIDE 29

29

Operating at elevated temperatures is not always beneficial, because higher temperatures can result in more rapid component degradation, including:

  • corrosion of bipolar plates Metal bipolar plates
  • catalyst support, graphite and metal oxide supports
  • degradation of seals, stable PFA Teflon seals have been demonstrated
  • degradation of membranes …polymer electrolyte membrane (PEM) is key to HT

PEM FC system development

Sum umma mary ry of

  • f sta

tatus tus of

  • f HT PEM Tec

echnology hnology

Dur urabil bility ity of

  • f HT-PEM

EM is is a a primary imary foc

  • cus

us for

  • r HT PEM

M fue uel l cel ell

slide-30
SLIDE 30

30

Alt lterna ernativ tive e Hi High gh temper peratur ture e PE PEM m M membr mbranes anes

Qu Questi stion

  • n – Any

nyth thing ing bett tter er tha han n pho hosphoric sphoric aci cid d lo loaded ded PB PBI? I? Ans nswer er -Yes …protic salt membranes!

slide-31
SLIDE 31

31

HT- PEM membrane electrode assemblies (MEAs)is the the heart of the fuel cell system

BASF recently announced exiting its high-temperature proton exchange membrane fuel cell business BASF was vendor of phosphoric-acid-loaded polybenzimidazole (PA loaded PBI) membranes

  • US-based Advent Technologies (East Hartford, Connecticut) novel membrane technology for HT-PEMFC.
  • Advent claims a proprietary membrane material for its HT-PEMFC based upon phosphoric acid loaded

pyridine with cross-linking for greater mechanical strength versus PA loaded PBI.

  • Advent claims two advantages
  • 1. operation at higher temperatures than PBI-based HT-PEMFC MEA, in some cases above 200°C and

with greater stability.

  • 2. Pressures inside the MEA can also approach those used in PAFC – a well-proven fuel cell type

EXIT BASF ENTER ADVENT

Clearly, electrolyte membranes will make or break a fuel cell system

*see: Fuel Cells Today, 28 Aug 2013; http://www.fuelcelltoday.com/analysis/analyst-views/2013/13-08-28-a-turning-point-for-high-temperature-pem-fuel-cells

phosphoric-acid-loaded polybenzimidazole MEA operates at 140 TO 190oC

slide-32
SLIDE 32

32

Old and New Proton-conducting Electrolytes

Old - Phosphoric acid ❑ Anhydrous phosphoric acid suggests how to synthesize an electrolyte that conducts proton with no water. ❑ Phosphoric acid self-ionizes and proton transfers between phosphoric acid, H3PO4 (acceptor base), and its ionic forms, such as, phosphonium ion, H4PO4

+ (donor acid).

❑ H+ transfer by rotational and vibration motions, because H3PO4 and H4PO4

+have suitable energy separation (proper ΔpK) and symmetry.

New - Protic salt ❑ A new class of proton-conducting salt electrolytes was conceived. These are called: protic ionic liquids (pILs) when in liquid form, and protic ionic membranes (pIMs) when in polymeric form. ❑ A neat protic salt electrolyte forms by transfer of proton from a Bronsted acid to a Bronsted base. ❑ pIL electrolytes, or a crystalline solid or polymeric versions, have high proton-conductivities (e.g., σ[25oC ] > 10 mS/cm), which follow Arhennius behavior temperature when the constituent acid and base have:

  • an optimal difference in pKa (~14) and
  • ions that are highly symmetrical (rotationally free).

Proton IN

H H H | | | O O O | | | O = P - OH … O = P - OH … H - O -- P = O | | | O O O | | | H H H

H+ …

1) vibrate H+ on

H + H H | | | O O O | | | H …O - P - O - H … O = P - O- H … O = P - O-H | | | O O O | | | H H H

2) vibrate H+ transfer 6) 3 rotations

H H + H | | | O O O | | | H-O - P = O … H - O - P - O - H …O = P - O-H | | | O O O | | | H H H

H+ transfer 3) vibrate

H H H

+

| | | O O O | | | H-O - P = O … H-O - P = O … H - O - P - O - H | | | O O O | | | H H H H H H | | | O O O | | | HO - P = O … HO P = O … H O -- P = O | | | O O O | | | H H H

…. H+

Proton OUT

4) vibrate H+ over

H H H | | | O O O | | | O = P - OH … O = P - OH … H - O -- P = O | | | O O O | | | H H H

5) vibrate H+ off

H H H | | | O O O | | | O = P - OH … O = P - OH … H - O -- P = O | | | O O O | | | H H H

H+ H+ Initial State Final State start over

slide-33
SLIDE 33

33

A protic ionic liquid (pIL) is made by transferring a proton from an acid to a base.

Energy Diagram for the EAN (ethyl ammonium nitrate) pIL with:

  • proton transferred (Left)
  • not transferred (Right),

E

EtNH3

+… NO3

  • HNO3… EtNH2

Proton Coordinate

E

EtNH3

+… NO3

  • HNO3… EtNH2

EtNH3

+… NO3

  • HNO3… EtNH2

Proton Coordinate Proton Coordinate

Gurney proton energy level diagram.

  • for any pair of levels, the stable entities

are upper right and lower left. pILs are a new class of solvent-free proton-conducting electrolyte. These are stable salts made when a acid transfers proton to a base. These “dry” proton conductors that can function at very high temperatures.

PR PROTIC IC IONIC C LIQUID UID (pIL) ) prot

  • ton
  • n-conducti

conducting ng Electr ctrolyte

  • lyte CONCE

CEPT PTS

EAN Δpk = 14

H+ hopping barrier

slide-34
SLIDE 34

34

Advantages of anhydrous proton-conducting pILs in HT PEM

  • pILs

Ls elimina minate e humidif idifier ier and d drastic stic size reducti duction

  • n of radia

diator tor

  • Other advantage …catalysis (next slide)
slide-35
SLIDE 35

35

Ca Cata talyti tic c Oxy xygen en Red eduction uction in anhy

hydr drous us pIL vs. Aqueo ueous us Electr ctrolytes

  • lytes

Cyclic voltammogram of Pt surface in:

  • aqueous sulfuric acid electrolyte (solid black line)
  • versus protic salt (dotted blue line).

H20

Voltammetry of Pt in aqueous and ionic liquids. Initial potential was 0.5 V vs RHE. Scan rate: 50 mV/s. Ar atmosphere. Temperature: 30 °C.

With protic salt electrolyte There is

  • no Pt-oxide formation
  • nor oxide reduction

O2 access to Pt is unimpeded and O2 reduction occurs earlier. Pt + H2O →Pt-oxide Pt ←Pt-oxide With aqueous electrolyte Pt-oxide formation

Pt + H2O → Pt-OH + H+ + e-

Pt-oxide reduction

Pt-OH + H+ + e- → Pt + H2O

Pt-oxide blocks adsorption sites delaying O2 reduction.

H20 H20 H20 H20 H20

Consequence of this difference? … see next slide

slide-36
SLIDE 36

36

Oxy xygen en red educ uction tion on Pt Pt in in pIL ver ersus Aqueous ueous El Elec ectr trol

  • lyte

te

  • Oxygen reduction faster on Pt in 2-FPTf and starts at 1.18V at 80°C, the thermodynamic limit for the ORR, 100% efficient
  • Oxygen reduction slower on Pt in aqueous triflic acid starts at 0.9 V, 0.3 V below the thermodynamic limit, 75% efficient

Trifluoromethane sulfonic acid (TFMSA)

+

2-Fluoropyridine 2-Fluoropyridinium Triflate (2-FPTf) Protic Ionic Liquid m.p. 85°C

  • H

+

Trifluoromethane sulfonic acid (TFMSA)

+

2-Fluoropyridine 2-Fluoropyridinium Triflate (2-FPTf) Protic Ionic Liquid m.p. 85°C

  • H

+

  • H
  • H

+

CF3 SO3H + xs H2O → CF3 SO3

  • + H3O+ + xs H2O
  • 2.0
  • 1.5
  • 1.0
  • 0.5

0.0 0.5 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Current Density (mA/cm^2) Potential (V) vs. RHE

Pt voltammetry in oxygenated nonaqueous 2-FPTf

WE: Platinum 0.1 cm2 CE: Platinum RE: RHE Scan Rate: 100mV/s Temp: 80°C 1 atm Oxygen WE: Platinum 0.1 cm2 CE: Platinum RE: RHE Scan Rate: 100mV/s Temp: 80°C 1 atm Oxygen

1.2 V 0.9 V

Pt voltammetry in oxygenated aqueous TFMSA

O2 reduction current O2 reduction current

slide-37
SLIDE 37

37

Trifluoromethane sulfonic acid (TFMSA)

+

2-Fluoropyridine 2-Fluoropyridinium Triflate (2-FPTf) Protic Ionic Liquid m.p. 85°C

  • H

+

Trifluoromethane sulfonic acid (TFMSA)

+

2-Fluoropyridine 2-Fluoropyridinium Triflate (2-FPTf) Protic Ionic Liquid m.p. 85°C

  • H

+

  • H
  • H

+

H2/O2 fuel cell Performance in Fluoro-pyridinium triflate (2-FPf) pIL vs PA

  • the first synthetic liquid electrolyte that out performs phosphoric acid

I/V curves for H2 and O2 fed to Pt-catalyzed porous electrodes in 2-FPTf electrolyte at 80oC and 120C and 85% phosphoric acid electrolyte at 80oC. σ(2-FPTf) = 4x10-3 Scm-1 , A = 0.5 cm2, thicknesselectrolyte = 0.3 cm. 2-FPTf is made by mixing

  • 1 part triflic acid (TFMSA) and
  • 1 part 2-fluoropyridine

Why does 2-FPTf work better? 2-FPTf is a “water-free” electrolyte so platinum-oxide does NOT form and the

  • xygen electrode is NOT polarized.

This increases the cell voltage, efficiency and power density. 15% more efficient in 2 FPTf

slide-38
SLIDE 38

38

Solid electrolytes (PEM) only

slide-39
SLIDE 39

39

Issues making liquid fuel cells unacceptable especially for portable applications

  • liquids leak
  • liquid can seep between electrodes causing short circuits
  • liquids soften materials accelerating mechanical failure

All of these issues are corrected by using a solid polymer electrolyte

Only solid polymer electrolyte membrane (PEM) fuel cell is acceptable !

Li Liqui uid d Ele lect ctrolyt

  • lyte

e Fu Fuel l Cell ll vs Sol

  • lid

id PEM M Fu Fuel l Cell ll

slide-40
SLIDE 40

40

But not all solid electrolytes are equally good

  • Solid oxide fuel cell (SOFC ceramicmembrane) is not shock and vibration tolerant, nor

suitable for automotive

  • Water solvated membranes (like Nafion) are vibration tolerant BUT have large radiators

and need inefficient humidifiers

  • pIL addresses both issues
slide-41
SLIDE 41

41 a) Non-conducting Low water form b) Conducting High water form

  • S

SO O3

3-

H H3

3O+ +

  • S

SO O3

3H

H

H H2

2O

a) Non-conducting Low water form b) Conducting High water form

  • S

SO O3

3-

H H3

3O+ +

  • S

SO O3

3H

H

H H2

2O

Nafion has an equivalent weight is 1100 g per -SO3H group, so with too little water, there is no proton conduction,

  • pendant sulfonic acid groups are neither all ionized nor bridged by water molecules, as is shown in Fig. a.
  • with 3 waters per -SO3H group, proton conducts, water bridges ionized acid groups, as shown in Fig. b.
  • BUT… 3 waters or more per acid means bulk-like water is in the membrane.
  • so operating temperature must be < 80oC at atmospheric pressure;
  • need to humidify feed gases to retain solvent water;
  • large radiators are needed to reject waste heat from 80oC to room temperature
  • the performance of platinum catalyzed cathodes is poor (Pt-oxide),

▪ E cathode < 0.9V or lower, and fuel-cell efficiency no higher than ~60%

The “Nafion Problem” Nafion was first solid polymer electrolyte membrane…but

What to do? … solution … go to a non-aqu aqueou eous s proto ton n condu ductor ctor

slide-42
SLIDE 42

42

  • 5
  • 4
  • 3
  • 2
  • 1

1 2 3 4 1000/T (K-1) Log s (S cm-1)

membrane: PVP-H3PO4 pIL: Py-H3PO4 pIL: 2F-PyTf

Conductivity for 3 electrolyte samples as a function of temperature. Solid triangle: Solid poly vinyl pyridinium phosphate (PVP-H3PO4) membrane, Open diamond: Liquid pyridinium phosphate pIL (P- H3PO4), Open square: Liquid 2-fluoro pyridinium triflate (2-FPyTf). Electrodes E-Tek Pt/C (0.5mg/cm2) fed dry: H2 gas. T=25 to 150oC

Pyridinium Phosphate Polarization Curves 0.2 0.4 0.6 0.8 1 1.2 1 10 100 Log [Current Density (mA/cm^2)] Cell Voltage (V)

140C 120C 80C

Pyridinium phosphate Polarization Curves

Pyridinium Phosphate Polarization Curves 0.2 0.4 0.6 0.8 1 1.2 1 10 100 Log [Current Density (mA/cm^2)] Cell Voltage (V)

140C 120C 80C

Pyridinium phosphate Polarization Curves

Unoptimized fuel cells polarizations with hydrogen anode and oxygen cathode using a non- fluorinated liquid pyridinium phosphate salt (PP) .

  • liquid pyridinium phosphate
  • versus solid polyvinyl pyridinium phosphate membrane

solid liquid liquid

Non

  • n-aqueous

aqueous proton

  • ton co

cond nductor uctors s

H2PO4

  • H3PO4

[ ]x

N

x

H2PO4

  • H3PO4

[ ]x

N

[ ]x

N

x

slide-43
SLIDE 43

43

H2PO4

  • H3PO4

[ ]x

N

x

H2PO4

  • H3PO4

[ ]x

N

[ ]x

N

x

Solid id polyvinylp vinylpyridinium yridinium phospha sphate te (PVPP) PP) membr brane ane

PVPP made by reacting each pyridine in polyvinylpyridine polymer with 1 phosphoric acid The solution, PVPP, a non-aqueous protic salt membrane, a“proton wire” that uses no water and leaches no ions A H2/O2 fuel cell with PVPP was run overnight under constant load of 30 mA/cm2 and the polarization (I/V test performance) did not change. This overnight stability is strong evidence that the proton only is hopping through this solid membrane which has no leachable ions or solvents. This kind of advanced HT-PEM fuel cell will be very reliable in practical use.

This sparingly soluble PVPP membrane is a crude form of Advent insoluble crosslinked “proprietary membrane material”.

I/V curve for H2/O2 fuel-cell with poly vinyl pyridine fully neutralized with phosphoric acid. Temp. = 162oC; σ =0.005 S/cm.

20 40 60 80 100 0.0 0.2 0.4 0.6 0.8 1.0 2 4 6 8 10 12 14 16 18

E / V I / mA cm

  • 2

Voltage (V) second IV test at 162

  • C

Power / mW cm

  • 2

solid

slide-44
SLIDE 44

44

Composites of inorganic and organic polymers

  • inorganic polymers , indium tin pyrophosphate (ITP)
  • organic polymer, polyvinyl pyridinium phosphate (PVPP)
slide-45
SLIDE 45

45

Inor

  • rganic

nic and organic anic composition mposition polyme mer r elect ectrolyte

  • lyte membr

brane ane (PEM) M)

  • 4
  • 3
  • 2
  • 1

2 2.4 2.8 3.2

Log σ (S cm-1) 1000 / T (K-1)

HEATING: thru-plane ITP, cold pressed at 6000 lbs COOLING: thru-plane ITP, cold pressed at 6000 lbs 150 C 120 C 90 C 60 C RT (25 C)

_|___________|_____________|________________|_______________________|_

Through-plane conductivity of a pure ITP (top) and 90%ITP/10%PVPP (bottom) membrane with electrodes under dry hydrogen atmosphere. Membrane thickness = 1.65mm, and area = 0.484 cm2. Frequency range: 50 kHz to 10 Hz. AC amplitude:

  • 50mV. Thickness of sputtered Pt el= 20nm. ETEK electrodes (0.5 mg of Pt per cm2)

used as gas diffusion layer. Pt screen current collectors. J4 ITP membrane. H2 / O2 Fuel cells made with a pure ITP membrane (top) and a membrane with an 90% ITP/ 10% PVPP blend (bottom). ETEK electrodes (0.5 mg of Pt per cm2) used as gas diffusion layer. Pt screen current collectors.

ITP ITP/PVPP

slide-46
SLIDE 46

46

Ne New w organic anic po poly lymer mers

slide-47
SLIDE 47

47

Cy Fujimoto (Sandia National Lab) and Yu Seung Kim (LosAlamos National Lab)

HT PEM Fuel l cell ll based sed on quaterna ternary ry ammonium

  • nium-bi

biphospha phosphate te ion pair irs

  • Unlike phosphoric acid (PA) loaded PBI, authors claimed the phosphoric acid does come NOT come out in presence of liquid water, with PA

loaded benzyl trimethyl ammonium (BTMA) groups on polyphenylene backbone membrane.

  • Claimed a new class of stable PEM fuel cell that can operate at low and high temperatures and retain phosphate even when water wet.
  • Has high conductivity and fuel cell performance at high temperatures to 180oC.

Fuel cell performance

Proton Conductivity of PA-doped QAPOH PEM fuel membrane

80oC 160oC

But low temperature performance suffered due to water leaching phosphoric acid (PA)

slide-48
SLIDE 48

48

Is all lost? No

slide-49
SLIDE 49

** Mandal, M., Huang, G., and Kohl, P. A., “Highly Conducting Anion Exchange Membranes Based on Cross-linked Poly(norbornene): Vinyl Addition Polymerization”, ACS Journal of Applied Energy and Materials, 2, 2447-2457 (2019).

Stable conductive ammonium hydroxide polymer** … precursor to HTPEM

*Stable AEMs under realistic

  • perating

conditions (e.g., 80 °C and 1 M KOH) have been synthesized by combining

  • all-hydrocarbon backbone
  • with tethered cations on long alkyl chains

* Lee, W. H.; Mohanty, A. D.; Bae, C. Fluorene-Based Hydroxide Ion Conducting Polymers for Chemically Stable Anion Exchange Membrane Fuel Cells. ACS Macro Lett. 2015, 4, 453−457. * Ono, H.; Kimura, T.; Takano, A.; Asazawa, K.; Miyake, J.; Inukai, J.; Miyatake, K. Robust anion conductive polymers containing perfluoroalkylene and pendant ammonium groups for high performance fuel cells. J. Mater. Chem. A 2017, 5, 24804−24812. * Lee, W. H.; Kim, Y. S.; Bae, C. Robust Hydroxide Ion Conducting Poly(biphenyl alylene)s for Alkaline Fuel Cell Membranes. ACS Macro

  • Lett. 2015, 4, 814−818.

Tetrablock copolymer, PNB-X34-Y66 TMHDA (N,N,N′,N′- Tetramethyl-1,6 hexane diamine cross-linking agent

49

slide-50
SLIDE 50

50

H O H-OP=O O H

H O H-OP=O O H

H O OP=O O H

H O OP=O O H

H3PO4 + X-linked block copolymer PNB-X34-Y66 → HT PEM

From hydroxide conductor to high temp proton conductor

  • by adding an acid (here, phosphoric acid) to hydroxide adduct of ammonium polymer

high temp proton conductor hydroxide conductor Tri-alkyl ammonium group

slide-51
SLIDE 51

51

Ben enef efits its of

  • f ne

new poly

  • lymer

mer for

  • r ma

making ing a HT a HTPEM EM fue uel l ce cell ll

  • This new polymer offers a a new platform for stable proton

conduction in a polymer electrolyte membrane (PEM) at high temperature (T < 200oC)

  • Conductivity of ammonium phosphate comparable to

phosphoric acid, expect high conductivity in the polymer membrane (conductivity around 0.1 S/cm)

  • Ionic groups are covalently and electrostatically immobilized

so TRULY WATER INSOLUBLE and NONLEACHABLE for extending stable operation at both high and low temperatures ( -55oC < T < 200oC) in the presence of liquid water.

slide-52
SLIDE 52

52

Evolution of fuel cells is system optimization by

  • reducing components to decreasing size, weight and parasitic losses
  • increase fuel cell power density with catalysts (Pt) to decrease size and weight
  • using polymer electrolyte membrane (PEM) is preferred electrolyte affording better system

designs

  • catalyst and electrolyte are intimately related for improving fuel cell

Insoluble protic salt polymers with a high density of covalently attached protic salt groups allow proton to be conducted without water at elevated temperatures following Arrhennius behavior.

  • Protoic salts membranes :
  • are Bronsted salts which contain a proton and whose acid and base moieties are

covalently and electrostacically attached to polymer

  • have pK separated by more than 4 units and less than 14. Here amines and phosphates
  • but other combinations possible

Conc nclus lusions ions and d Recommenda

  • mmendations

tions for r futur ure e resear search h

These developments give a promising path toward stable water-free proton conducting membranes for stable, compact and efficient fuel cell systems with low levels of Pt catalysts and possibly non-platinum catalyst for the hydrogen and

  • xygen fuel cell for fuel cell in the 100kilowatt range for automotive and distributed power applications
slide-53
SLIDE 53

53

  • 1. micro-CHP (emergency and off-grid residential electricity and heat)
  • 2. power production (distributed generation)
  • 3. portable power (automotive)
  • 4. hydrogen separation (electrolysis of water and hydrogen compression)
  • 5. electrical power inside airplanes
  • 6. remote communications stations and battery chargers
  • 7. Storage and use of renewables, like solar during sundown, wind during calm, etc.

Mai ain te terres estrial trial mar arkets ets suiting ting HT-PE PEMFC FC te technolog nology: :

slide-54
SLIDE 54

54

Army Research Office, ARO Project No. W911-NF-04-1-0060

  • Dr. Robert Mantz, Project Monitor

NASA-Glen NASA Project No. NC04GB06G

  • Dr. James Kinder, Project Monitor
  • U. S. Department of Energy

DoE Project No. DE-FG36-06G016029

  • Drs. Kathi Epping, Gregory Kleen Project Managers

and

Ack cknowledge nowledgements ments

  • Drs. Joanna Moore, James Kinder and Jean-Phillipe Belieres

Project Monitors

slide-55
SLIDE 55

55

EXTRA SLIDES

slide-56
SLIDE 56

56

Thermocouple Gas In/Out PTFE Cell 4 pt. wire probe

Schematic diagram (top) and real hardware (bottom) for accepting solid electrolyte membrane electrode assembly and feeding oxygen to one side and hydrogen to the other to produce fuel cell measurements inside isothermal oven.

PTFE Probe wires Viton gaskets Stainless Steel Gas ITP Membrane Pt sputtered on each face Pt screen E-TEK Electrodes

Solid id Elect ectro rolyte lyte Test t Fuel el Cell

slide-57
SLIDE 57

57

➢E-TEK ELAT V3 electrodes (Pt loading 0.5 mg/cm2) compression sealed around Ionic Liquid contained in Teflon (PTFE) cell body. ➢Liquids tend to leak and cause shorts in the fuel cell, not suitable for commercial use.

Li Liquid uid El Elec ectro trolyte lyte Fu Fuel el Ce Cell ll Tes est t Ha Hardw rdware are

slide-58
SLIDE 58

58

Schematic representation of PTFE micro fuel cell with Gas Fed Electrodes and liquid electrolyte, like phosphoric acid. Schematic representation of a micro fuel cell with Gas Fed Electrodes and PEM electrolyte, like Nafion.

Teflon block Holds electrodes Contains liquid acid MEA With electrodes

  • n membrane

Liqu quid d Fuel l Cell Soli lid d PEM Fuel uel Cell

Liquid uid vs. Polymer mer Electr ctrolyte

  • lyte Membr

brane ane (PEM) M) Fuel uel Cell

No leaks No shorts Less Corrosion

PEM Preferred