Combined in-situ diagnostic tools for detection of liquid water - - PowerPoint PPT Presentation

• 1 -

Combined in-situ diagnostic tools for detection of liquid water evolution in PEMFC

23.06.2009

• R. Kuhn, Ph. Krüger, T. Arlt, I. Manke, Ch. Hartnig

Zentrum für Sonnenenergie- und Wasserstoff-Forschung, Ulm

Robert.kuhn@zsw-bw.de, +49/731/9530 203

• 2 -

• 3 -

two main factors influence the distribution and the balance between anode and cathode:

• electro-osmotic drag (protonic current)
• back diffusion (liquid water gradient)

anode cathode liquid water gradient

back diffusion electro-osmotic drag

protonic current

water distribution in fuel cells

• 4 -

presented methods for fuel cell research

spatial resolution 200 μm neutron tomography 3 μm synchrotron radiography neutron radiography 80 μm

• 5 -

X-rays

photo electron absorption scattering

core core

neutrons

absorption scattering

20 40 60 80 100 10

• 2

10 10

2

Mass attenuation coefficient, (cm

2/g)

Atomic number X-rays (100 keV)

comparison between X-rays and neutrons

20 40 60 80 100 10

• 2

10 10

2

Mass attenuation coefficient, (cm

2/g)

Atomic number X-rays (100 keV) Thermal neutrons

• 6 -

why neutrons?

20 40 60 80 100 10

• 2

10 10

2

Fe

Mass attenuation coefficient, (cm

2/g)

Atomic number X-rays (100 keV) Thermal neutrons

Gd

113Cd 10B

Li

1H

Co Ni Pb Au Al

light elements metals heavy elements

• 7 -

neutron radiography

• 8 -

resolution sample size 30 cm – 2 cm 200 µm – 80 µm

neutron tomography and radiography

1 µm 0.1 mm 1 m 1 cm 0.1 µm

• 9 -

normalization of images: water distribution map

• riginal radiography

ratio: water filled cell /empty cell water distribution 100 mm

neutrons can ‘see’ water in fuel cells

• 10 -

evolution of liquid water on anode side

neutrons can ‘see’ water in fuel cells

100 mm 50 mm 8 mm H2 out H2 in Air in Air out

• 11 -

mask for anode channel H2 out H2 in Air in Air out

neutrons can ‘see’ water in fuel cells

single cell water distribution

200 mA/cm2, 60% u.C, 20% u.A cathode outlet 100% r. H. , anode inlet 0% r. H.

• 12 -

time scanline 38 min real time

evaporation of liquid water in a channel

neutrons can ‘see’ water in fuel cells

200 mA/cm2, 60% u.C, 20% u.A cathode outlet 100% r. H. , anode inlet 0% r. H.

• 13 -

mask for anode chanel H2 out H2 in Air in Air out

neutrons can ‘see’ water in fuel cells

single cell water distribution

500 mA/cm2, 60% u.C, 20% u.A cathode outlet 100% r. H. , anode inlet 0% r. H.

• 14 -

neutron tomography

• 15 -

neutron tomography

anode cathode H2 in H2 out air in air out

• 16 -

neutron tomography

• 17 -

cell 1 cell 2 cell 3

neutron tomography: triple stack

• separate analysis of cells in a multi-stack
• differentiation between anode and cathode

Water amount

• 18 -

synchrotron radiography

• 19 -

synchrotron radiography

resolution sample size 1 cm – 0.3 mm 7 µm – 0.3 µm

1 µm 0.1 mm 1 m 1 cm 0.1 µm

• 20 -

neutron radiography

A C

100 times higher spatial resolution + 1000 times higher water sensitivity compared to neutron-based methods

[I. Manke, Ch. Hartnig et al. Applied Physics Letters 90, 174105 (2007)]

high resolution in-situ synchrotron radiography

in-situ synchrotron radiography

7 mm 100 mm

normalization

• 21 -

through plane view

integral imaging: summarized presentation of the cathodic and anodic channels and gas diffusion media

• 22 -

time dependence of liquid water evolution

close look at three different spots: 1: in the channel 2: under the rib (next to point #1) 3: under the rib, no periodic pumping observed

 time dependence of the ‘filling degree’

anodic gas channels

A A C C C

cathodic gas channels

• 23 -

dynamics of liquid water evolution

(time lapse movie)

• 24 -

time dependence of liquid water evolution

• strong correlation of water content of the two adjacent positions

in the channel and under the rib  transport pathway

• continous filling of the GDL + eruptive emptying of the pores

1 2

• 25 -

cross sectional view

[Ch. Hartnig, I. Manke et. al, APL 92 (2008) 134106]

uncorrected image after normalization with respect to an ‘empty cell’ components

• 26 -

cross sectional insights

• differentiation between MPL and GDL (substrate)
• small water clusters in the nano-litre range detectable

• 27 -

depth profile of water distribution

• quantification and differentiation of separate areas possible
• formation of water clusters next to the MPL (temperature difference)
• formation of a second diffusion barrier from i0 > 400 mA/cm2 onwards

MPL

• 28 -

augmented reality

proposed model for liquid water transport: liquid water collects under the lands of the flow fields and is transferred to the channel in an eruptive mechanism two different contributions:

• liquid water can evaporate and diffuse as vapour through the GDL
• liquid water collects in the GDL and erupts to the channel

• 29 -

conclusion / lessons learned

neutron radiography:

• determination of wet/dry cell regions
• evolution of water cluster on the anode side
• over all cell effects of flooding and opening of channels

neutron tomography:

• 3D distribution of liquid water in fuel cell stacks
• layer-by-layer determination of water distribution:

synchrotron radiography:

• initial formation of liquid water under the lands of the flowfield
• n the cathode side
• visualization of micro structure transport pathways

• 30 -

acknowledgements

DFG (Le 1433/1-2) BMBF (03SF0324) team at HZB: Nikolay Kardjilov André Hilger team at ZSW: Ludwig Jörissen Joachim Scholta Frank Häussler RuNPEM-partners

thank you for your attention!