Back to list of projects

Project Title:

Electrocatalysis and transport processes in water elecrolysis and menbrane fuel cells

Ref.No.: 134

Project Type and Category:

Basic research

Project Duration:

# (See Ref.no.:73)

Project Participants:

M. Fischer, W. Schnurnberger, R. Reißner, M. Schulze, Institut für Technische Thermodynamik
Deutsche Forschungsanstalt für Luft- und Raumfahrt , 70569 Stuttgart

Sponsor:

#

Project Budget and
Funding:

#

Project Description and Objectives:

For an energetic use of hydrogen, electrochemical processing techniques are most promising. Purpose of the subproject A1 is to improve high performance electrodes which reach high current densities at low overvoltages for water electrolysis and fuel cells. The work is directed towards a profound understanding of single electrochemical reaction steps, adsorption-desorption processes at the electrode surface as well as transport processes and thermal behavior of gas diffusion electrodes. Based on our present research, the emphasis will be directed towards an investigation of reaction steps of membrane fuel cell reactions. Here, apart from highly dispersive noble metal alloys, oxidic catalysts are of special interest. These investigations are closely related to our research on hydrogen and oxygen production on oxidic catalysts. These results will be incorporated into the production of fuel cell electrodes at a reasonable cost using a rolling technique. This rolling process was adapted and modified for PEFC electrodes from a simular process for the production of electrodes for batteries and alkaline fuel cells. To understand the processes in the fuel cell, heat and material transport within the fuel cell are computer-simulated.

Technical Goals:

Chemical reactions in water electrolysis and fuel cells are running directly at the electrode surface. To understand the reactions occuring at the catalyst, not only i-U-characteristics can be analyzed with the help of impedance spectroscopy, but knowledge on adsorption and desorption processes is also important. A special interest of electrocatalysis research is therefore on the investigation of surface properties as well as the energetics at the interface electrode/electrolyte. There are more simple model surfaces under investigation in the beginning because of the complexity of the surfaces of real catalysts. Therefore, single crystal surfaces (pure or modified with adsorbates) can be selected for a model surface.
The interaction of the adsorbates (H2, O2, H2O, CO) with metal oxide surfaces is under investigation, focusing on the reduction of the susceptibility to poisoning with CO. Apart from electrochemical methods, surface analytical techniques are used for these investigations. These are photoelectron spectroscopy (XPS, UPS), thermal desorption spectroscopy (TDS) and electron diffraction (LEED).
Focus for theoretical modelling of electrode/membrane composite structures is on heat and material transport. The investigations on molecular transport of hydrogen (project A2) transportation processes in gas diffusion electrodes and membrane fuel cells are simulated and experimentally verified. The electrode/membrane composite units developed and characterized in collaboration with project A7 are used in the experiments.

Project Status

#

Preliminary or Final Results:

Adsorption and desorption of water on the pure Ni(111) surface and the Ni(111) single crystal surface precovered with potassium or oxygen was investigated with XPS and TDS. The kinetic parameters of water desorption (desorption energy and preexponential factor) could be determined with high accuracy by means of adsorption-desorption-equilibrium measurements.
On the pure single crystal surface water adsorbs without dissociation. For coverages up to 0.42ML (ML=monolayers, with respect to the number of surface nickel atoms) we find a desorption energy of 52 kJ/mol and a preexponential factor of 1016 s-1 (static). For higher coverages the desorption kinetics of ice multilayers is observed, i.e. a desorption energy of 39 kJ/mol and a preexponential factor of 8*1012 s-1 (static). With a potassium precoverage on the crystal, water molecules dissociate to OH+H at an amount proportional to the potassium precoverage. In addition we find one molecular H2O per potassium atom that is bonded more tightly to the surface than on the non-precovered surface. At a potassium precoverage below 0.15ML, the rest of the water adsorbs the same as on the pure Ni(111) surface. At about 0.15ML of preadsorbed potassium, the amounts of dissociatively (static) and more tightly bonded water ( static) add to about 0.42ML, leaving no more sites for the water to directly adsorb at the surface ( static). At higher potassium precoverages, a new adsorption site is observed showing the same kinetics as water adsorbed directly at the pure Ni surface [1 - 5]. These findings are summarized in the phase diagram in Fig. 1.
Oxygen as a precoverage on the nickel surface adsorbs atomically up to about 0.33ML. With higher amounts of oxygen, NiO islands are formed which grow laterally eventually joining together. At about 2ML of oxygen precoverage, a closed epitaxial NiO-layer is formed which shows many defects because of the lattice constant of NiO being 18% larger than that of Ni. Water adsorbs onto the surface precovered with atomically adsorbed oxygen, only in its molecular state. An amount proportional to the oxygen precoverage is bonded more tightly to the surface than on the non-precovered sample. When the oxygen precoverage is large enough for NiO islands to be formed, a part of the water adsorbs dissociatively (OH+H). With the oxygen precoverage increasing, i.e. the NiO-islands growing, the amount of more tightly bonded water decreases, and the amount of dissociated water increases [5,6]. Figure 1: Phasediagram of the coadorption system H2O/K/Ni(111)
For the computer simulation, the polymer membrane fuel cell has been subivided into different components; these are gas-distributor, gas diffusion layer, catalytic layer and membrane. Each of these layers is described by a mathematical model which accounts for the physical phenomen a arising in this structure as there are energy and mass balances and mass transport equations. Coupling these modules results in an overall model describing a single electrode-membrane unit or even a complete fuel cell stack. [7]
The water being transported as a hydrate shell with the protons results in a drying of the membrane on the anode side. As a result there are important changes with regard to dynamic Current-voltage characteristics. Starting with a completely humidified membrane, current/voltage curves are recorded with different speed of changes in the load. For a slow change of load the characteristics decrease with regard to the power output at high current densities due to the progressive water deficit in the membrane. On the other hand a very fast change in load results in a negligible change of humidity in the membrane: the characteristics is linear up to high current densities. This fact requires a presentation of load changing speed together with the current-voltage characteristic. The entity of the transportation processes can be recorded by the shape of the i-U-characteristic only approaching a quasi-stationary modus.
The electrode-membrane-electrode-structure is produced by a rolling procedure. In the PEFC the interface of electrode and electrolyte is a solid-solid transition consisting in the worst case of a high number of single point contacts. This interface can therefore produce high ohmic losses on charge transition. The electrode is generally brushed with a solution of solid electrolyte and aliphatic alcohol to increase the catalytically active area and to achieve the penetration of electrolyte into the electrode as it is inherent in porous systems with liquid electrolyte. Thus the starting point for our electrode development was replacement of this brushing procedure and integration of the formation of three-phase-boundaries in the electrode preparation process.
Figure 2: Current-voltage characteristics for different speeds of load change To achieve this development goal, we deposit solid electrolyte material in the catalytic layer of the electrode by reactive mixing of raw materials (i. e. Vulcan XC72, PTFE and pulverized polymer electrolyte). The produced electrodes are evaluated by their i-U-characterisics and compared to standard electrodes. The electrochemical results indicate, that the rolling process is a promising method to produce low cost electrode structures as applied in PEFC [8].

Related Reference Papers and Other Publications:

[1] W. Kuch, W. Schnurnberger, M. Schulze, K. Bolwin; J. Chem. Phys.101 (1994) p 1687[2]W.
Kuch; Ph.D. thesis, Universität Stuttgart, 1993[3]W. Kuch, M. Schulze, W. Schnurnberger, K. Bolwin; Surf. Sci. 287/288, p 600[4] W. Kuch, M. Schulze, W. Schnurnberger, K. Bolwin; Ber.
Bunsenges. Phys. Chem. 97 (1993) No.3, p 356 [5] R. Reißner, N. Wagner, W. Kuch, M. Schulze, W. Schnurnberger; X-Ray Photoelektron Spectroscopy and Electrochemical Studies of Nickel Surfaces Interacting with Water and Alkaline Solution; *) [6] M. Schulze, R. Reißner,
W. Kuch, K. Bolwin; Fresenius J. Anal. Chem. 353, p 661 [7] M. Wöhr, K. Bolwin, W. Schnurnberger; Dynamic modeling and simulation of a polymer membrane fuel cell (PEFC) including mass transport limitation; *) [8]D. Bevers, N. Wagner, M. von Bradke; Innovative Production Procedure for Low Cost PEFC Electrodes and Electrode Membrane Structures; *) Hydrogen ´96 - 11th World Hydrogen Energy Conference- Stuttgart, Germany

Back to list of projects

Project Title: Electrocatalysis and transport processes in water elecrolysis and menbrane fuel cells		Ref.No.: 134
Project Type and Category:	Basic research
Project Duration:	# (See Ref.no.:73)
Project Participants:	M. Fischer, W. Schnurnberger, R. Reißner, M. Schulze, Institut für Technische Thermodynamik Deutsche Forschungsanstalt für Luft- und Raumfahrt , 70569 Stuttgart
Sponsor:	#
Project Budget and Funding:	#
Project Description and Objectives:	For an energetic use of hydrogen, electrochemical processing techniques are most promising. Purpose of the subproject A1 is to improve high performance electrodes which reach high current densities at low overvoltages for water electrolysis and fuel cells. The work is directed towards a profound understanding of single electrochemical reaction steps, adsorption-desorption processes at the electrode surface as well as transport processes and thermal behavior of gas diffusion electrodes. Based on our present research, the emphasis will be directed towards an investigation of reaction steps of membrane fuel cell reactions. Here, apart from highly dispersive noble metal alloys, oxidic catalysts are of special interest. These investigations are closely related to our research on hydrogen and oxygen production on oxidic catalysts. These results will be incorporated into the production of fuel cell electrodes at a reasonable cost using a rolling technique. This rolling process was adapted and modified for PEFC electrodes from a simular process for the production of electrodes for batteries and alkaline fuel cells. To understand the processes in the fuel cell, heat and material transport within the fuel cell are computer-simulated.
Technical Goals:	Chemical reactions in water electrolysis and fuel cells are running directly at the electrode surface. To understand the reactions occuring at the catalyst, not only i-U-characteristics can be analyzed with the help of impedance spectroscopy, but knowledge on adsorption and desorption processes is also important. A special interest of electrocatalysis research is therefore on the investigation of surface properties as well as the energetics at the interface electrode/electrolyte. There are more simple model surfaces under investigation in the beginning because of the complexity of the surfaces of real catalysts. Therefore, single crystal surfaces (pure or modified with adsorbates) can be selected for a model surface. The interaction of the adsorbates (H2, O2, H2O, CO) with metal oxide surfaces is under investigation, focusing on the reduction of the susceptibility to poisoning with CO. Apart from electrochemical methods, surface analytical techniques are used for these investigations. These are photoelectron spectroscopy (XPS, UPS), thermal desorption spectroscopy (TDS) and electron diffraction (LEED). Focus for theoretical modelling of electrode/membrane composite structures is on heat and material transport. The investigations on molecular transport of hydrogen (project A2) transportation processes in gas diffusion electrodes and membrane fuel cells are simulated and experimentally verified. The electrode/membrane composite units developed and characterized in collaboration with project A7 are used in the experiments.
Project Status	#
Preliminary or Final Results:	Adsorption and desorption of water on the pure Ni(111) surface and the Ni(111) single crystal surface precovered with potassium or oxygen was investigated with XPS and TDS. The kinetic parameters of water desorption (desorption energy and preexponential factor) could be determined with high accuracy by means of adsorption-desorption-equilibrium measurements. On the pure single crystal surface water adsorbs without dissociation. For coverages up to 0.42ML (ML=monolayers, with respect to the number of surface nickel atoms) we find a desorption energy of 52 kJ/mol and a preexponential factor of 1016 s-1 (static). For higher coverages the desorption kinetics of ice multilayers is observed, i.e. a desorption energy of 39 kJ/mol and a preexponential factor of 8*1012 s-1 (static). With a potassium precoverage on the crystal, water molecules dissociate to OH+H at an amount proportional to the potassium precoverage. In addition we find one molecular H2O per potassium atom that is bonded more tightly to the surface than on the non-precovered surface. At a potassium precoverage below 0.15ML, the rest of the water adsorbs the same as on the pure Ni(111) surface. At about 0.15ML of preadsorbed potassium, the amounts of dissociatively (static) and more tightly bonded water ( static) add to about 0.42ML, leaving no more sites for the water to directly adsorb at the surface ( static). At higher potassium precoverages, a new adsorption site is observed showing the same kinetics as water adsorbed directly at the pure Ni surface [1 - 5]. These findings are summarized in the phase diagram in Fig. 1. Oxygen as a precoverage on the nickel surface adsorbs atomically up to about 0.33ML. With higher amounts of oxygen, NiO islands are formed which grow laterally eventually joining together. At about 2ML of oxygen precoverage, a closed epitaxial NiO-layer is formed which shows many defects because of the lattice constant of NiO being 18% larger than that of Ni. Water adsorbs onto the surface precovered with atomically adsorbed oxygen, only in its molecular state. An amount proportional to the oxygen precoverage is bonded more tightly to the surface than on the non-precovered sample. When the oxygen precoverage is large enough for NiO islands to be formed, a part of the water adsorbs dissociatively (OH+H). With the oxygen precoverage increasing, i.e. the NiO-islands growing, the amount of more tightly bonded water decreases, and the amount of dissociated water increases [5,6]. Figure 1: Phasediagram of the coadorption system H2O/K/Ni(111) For the computer simulation, the polymer membrane fuel cell has been subivided into different components; these are gas-distributor, gas diffusion layer, catalytic layer and membrane. Each of these layers is described by a mathematical model which accounts for the physical phenomen a arising in this structure as there are energy and mass balances and mass transport equations. Coupling these modules results in an overall model describing a single electrode-membrane unit or even a complete fuel cell stack. [7] The water being transported as a hydrate shell with the protons results in a drying of the membrane on the anode side. As a result there are important changes with regard to dynamic Current-voltage characteristics. Starting with a completely humidified membrane, current/voltage curves are recorded with different speed of changes in the load. For a slow change of load the characteristics decrease with regard to the power output at high current densities due to the progressive water deficit in the membrane. On the other hand a very fast change in load results in a negligible change of humidity in the membrane: the characteristics is linear up to high current densities. This fact requires a presentation of load changing speed together with the current-voltage characteristic. The entity of the transportation processes can be recorded by the shape of the i-U-characteristic only approaching a quasi-stationary modus. The electrode-membrane-electrode-structure is produced by a rolling procedure. In the PEFC the interface of electrode and electrolyte is a solid-solid transition consisting in the worst case of a high number of single point contacts. This interface can therefore produce high ohmic losses on charge transition. The electrode is generally brushed with a solution of solid electrolyte and aliphatic alcohol to increase the catalytically active area and to achieve the penetration of electrolyte into the electrode as it is inherent in porous systems with liquid electrolyte. Thus the starting point for our electrode development was replacement of this brushing procedure and integration of the formation of three-phase-boundaries in the electrode preparation process. Figure 2: Current-voltage characteristics for different speeds of load change To achieve this development goal, we deposit solid electrolyte material in the catalytic layer of the electrode by reactive mixing of raw materials (i. e. Vulcan XC72, PTFE and pulverized polymer electrolyte). The produced electrodes are evaluated by their i-U-characterisics and compared to standard electrodes. The electrochemical results indicate, that the rolling process is a promising method to produce low cost electrode structures as applied in PEFC [8].
Related Reference Papers and Other Publications:	[1] W. Kuch, W. Schnurnberger, M. Schulze, K. Bolwin; J. Chem. Phys.101 (1994) p 1687[2]W. Kuch; Ph.D. thesis, Universität Stuttgart, 1993[3]W. Kuch, M. Schulze, W. Schnurnberger, K. Bolwin; Surf. Sci. 287/288, p 600[4] W. Kuch, M. Schulze, W. Schnurnberger, K. Bolwin; Ber. Bunsenges. Phys. Chem. 97 (1993) No.3, p 356 [5] R. Reißner, N. Wagner, W. Kuch, M. Schulze, W. Schnurnberger; X-Ray Photoelektron Spectroscopy and Electrochemical Studies of Nickel Surfaces Interacting with Water and Alkaline Solution; ) [6] M. Schulze, R. Reißner, W. Kuch, K. Bolwin; Fresenius J. Anal. Chem. 353, p 661 [7] M. Wöhr, K. Bolwin, W. Schnurnberger; Dynamic modeling and simulation of a polymer membrane fuel cell (PEFC) including mass transport limitation; ) [8]D. Bevers, N. Wagner, M. von Bradke; Innovative Production Procedure for Low Cost PEFC Electrodes and Electrode Membrane Structures; *) Hydrogen ´96 - 11th World Hydrogen Energy Conference- Stuttgart, Germany