Back to list of projects

Project Title:

Modification and characterisation of fuel generating photo electrodes with new organic and anorganic surface layers

Ref.No.: 25

Project Type and Category:

Hydrogen production

Project Duration:

1989 – 1998

Project Participants:

Institute f. organ. Chemie (IOC), Prof. Dr. F. Effenberger (A4.2)
Inst. f. physikal. Elektronik (IPE), Prof. Dr. J.H Werner (A4.1)
Other participants of sfb 270:
Project Manager:Prof. Dr. Werner, Prof. Dr. Effenberger

Sponsor:

DFG: 294,800 DM/y since 1989

Project Budget and
Funding:

Approx.. 2,6 Mio DM

Project Description and Objectives:

In a photoelectrochemical cell hydrogen is generated at the surface of a semiconductor electrode which is illuminated and dips in an electrolyte. Early approaches with TiO2-electrodes succeeded in principle [1], but, because of its high band-gap, this material turns out to be unfavourable for collecting of solar energy. In the other hand silicon with a band-gap of 1,1 eV has a much lower threshold of absorption but is thermodynamically instable towards photoanodic but also cathodic processes of corrosion [2]. Besides thin layers of metal or metal oxides organic polymers provide a useful means for coating and hence protection of the electrodes [3]. As thin films of conducting polymers like e.g. polyacetylene, polyaniline, polypyrrole, and polythiophene [4,5] are useful materials for passivation of photoanodes, electronmediators must be incorporated in the surface layer of the cathodes to guarantee an efficient charge transfer. Moreover from a kinetic point of view it is necessary to link catalysts like e.g. RuO2, IrO2, Pt or Pd [6] or molecular systems [7] to the surface to accelerate such multiple electron processes like that in water electrolysis and to reduce the overpotential. Due to the stability of polythiophene in aqueous solution [5b] this polymer seems to be appropriate for protection of electrodes. Consequently this polymer, provided with redox-active, catalytic sidegroups, should match the expected features. According to their redox potentials, 4,4'-bipyridine salts as well as an organometallic rhodium-complex, a real homogeneous molecular catalyst, were considered convenient for the reductive part of water electrolysis and ferrocene for the oxidative part (Scheme 1). The polymers should be built up via electropolymerization of functionalized mono- or oligothiophenes.
In the first part of this programm a convenient synthetic route to thiophenes and oligo-thiophenes with 3-(-haloalkyl)-substituents had to be developed in order to bind the redox catalysts. The second part should include the electrooxidative polymerization of the monomeric building blocks as well as investigations refering to polymer and film characteristics like redox properties, conductivity and absorption. To probe the protective and catalytic properties of these materials under working conditions films on silicon wafers were prepared by spin coating of polymer solutions followed by evaporation of the solvent.

Technical Goals:

Modification and optimization of electrode surfaces

Project Status

Final Phase

Preliminary or Final Results:

The synthesis of a series of mono- and oligothiophenes 1 - 4 with various alkyl chain length and substitution pattern was realized via nickel catalyzed Grignard coupling reaction [8]. As outlined in Scheme 2, the redox catalysts were linked by nucleophilic substitution of the terminal bromide. The 2,2'-bipyridine derivative 6 was converted to the desired rhodium complex 8 (x=2,3; n=5) in cooperation with project A5. Apart from compounds, in which the oxidation potential of the thiophene is unfavourable with respect to its catalytic moiety, the oligothiophenes in 5, 7 und 8 were electropolymerized.
In the cyclic voltammogram of a platinum electrode coated with polythiophene 5 (x=1, n=8) redox curves of the incorporated viologen (Eo[V2+/+]= -0,74 V; Eo[V+/o]= -1,2 V; gg. Ag/AgCl) are clearly visible (Figure 1).
Figur1:Cyclic volCyclic voltammogram of a P5 (n=8) film at a scan rate of 50 mV/s. Inset: electrooxidative polymerization of 5 (n=8) [5x10-2 M] in acetonitrile (0.1 M TBAHPF) in a potentiodynamic experiment.tammogram of a P5 (n=8) film at a scan rate of 50 mV/s. Inset: electrooxidative polymerization of 5 (n=8) [5x10-2 M] in acetonitrile (0.1 M TBAHPF) in a potentiodynamic experiment.
Films prepared from 5 and 8, respectively, either on silicon or glassy carbon electrodes, on which hydrogen reveals a considerable overpotential, demonstrate the suitability of these polymers by both passivating the surface of the photoelectrode and activating the generation of hydrogen catalytically. The polymers obtained by electrooxidation of 5 or 8 are soluble in polar organic solvents. While p-Si is applicable only in the reductive but not in the oxidative potential range polymer layers with thicknesses between 0.05 and 5 µm were obtained by spin coating from those solutions. As charge transfer through the layer is assumed to take place via viologen units additional platinum particle are necessary to release hydrogen. Figure 2 presents             the possible variations in distribution investigated. Type 1: precipitation of Pt directly on the semiconductor surface by electron-beam evaporation providing a homogeneous Pt-layer (A) or by galvanostatic reduction of H2PtCl6 solutions yielding island structures (B) with subsequent polymer film precipitation; type 2: homogeneous distribution of Pt particles achieved by casting with a suspension of finest Pt powder in the polymer solution; type 3: precipitation of Pt of the polymer film by sputtering (A) or by galvanostatic reduction of H2PtCl6 solutions (B).
Figure 2: Possible distribution of platinum in the system p-silicon/polymer/ platinum. The photoelectrodes were characterized via I-U-relationships as shown in Figure 3a. From comparative measurements the following results can be summerized. p-Si-electrodes, both coated and uncoated with polymer resemble in the I-U-diagram (curves 2, 3) and reveal no photopotential in absence of a catalyst. Layers of poly(3-dodecylthiophene) (P3DDT), having no viologen side groups, behave as an insulator (curve 1). Due to individual polymer absorption lamdamax = 480-500 nm) electrodes with catalytically active layers work with a gain in energy only if the layer thickness is <1µm. Thin films are preferred in long-term measurements since thick layers often peel off after short operating time. Coated and catalytically active photoelectrodes provide a greater energy gain in strong acidic medium rather than in neutral environment according to literature findings [9]. In thin films the distribution of Pt (type 1-3) hardly influences the observed photopotential (100-280 mV), but type 1 electrodes result in a higher long-term stability than type 3. Electrodes of type 1 (A) show an ohmic contact between Si and Pt and are therefore less suitable. On the contrary, the polymer in type 3 (A) acts as a Schottky barrier. Photocathodes of type 1 (B) (curve 5) provide the best result with an efficiency of 1.2 % and a high long-term stability. In a test, broken off 120 h later (Figure 3b) in which the electrode worked at the maximum power, the photocurrent practically remains constant after a small initial reduction but slightly improved photovoltage (curve a). Comparatively the photocurrent of an uncoated p-Si/Pt-electrode decreases significantly within 24 h due to photocorrosion (curve b). Within this long testing time about 0.35 l hydrogen evolved at an electrode area of 1 cm² without degradation of the polymer layer and the silicon surface, respectively.
Figur3: I-U-relationships of p-Si/polymer(P3DDT)- (1), p-Si/polymer(P5)- (2), p-Si- (3), p-Si/ Pt- (4) and p-Si/Pt/polymer(P5)-photocathode at pH 0 and 20 mV/s. Photocurrent-time-relationship of a p-Si/Pt/polymer(P5)-photocathode (a) working at maximum power at pH 0 within 120 h; (b) p-Si/Pt-photoelectrode without polymer. Rhodium complex functionalized polythiophene P8 (x=2, n=5), polymerized on glassy carbon

Related Reference Papers and Other Publications:

Internet-Site: www.uni-stuttgart/UNIuser/sfb270/A4_2_E.htm
Wasserstoff als Energieträger, ISBN 3-00-001 796-8
[1]A.Fujishima, K.Honda, Bull.Chem.Soc.Jpn. 44 (1971) 1148
[2]H.Gerischer, Pure Appl.Chem. 52 (1980) 2649
[3]M.S.Wrighton, Science 231 (1986) 32
[4.a] R.A.Simon, M.S.Wrighton, Appl.Phys.Lett. 44 (1984) 930
[4.b]K.M.Kost, D.E.Bartak, B.Kazee, T.Kuwana, Anal.Chem. 60 (1988) 2379
[4.c]S.Holdcroft, B.L.Funt, J.Electroanal.Chem. 240 (1989) 89
[5.a]Z.Deng, W.H.Smyre, H.S.White, J.Electrochem.Soc. 136 (1989) 2152
[5.b]G.Tourillon, F.Garnier, J.Phys.Chem. 88 (1984) 5281
[6]R.Kötz, S.Stucki, Electrochim.Acta 31 (1986) 1311
[7] V.Kölle, M.Grätzel, Angew.Chem. 99 (1987) 564
[8.a] P.Bäuerle, F.Würthner, S.Heid, Angew.Chem. 102 (1990) 414
[8.b] P.Bäuerle, S.Scheib, Adv.Mater. 5 (1993) 848
[9]A.Bandi, H.-M.Kühne, J.Schefold, M.Specht, M.Wendt, DECHEMA-Monogr. 117 (1989)

Back to list of projects

Project Title: Modification and characterisation of fuel generating photo electrodes with new organic and anorganic surface layers		Ref.No.: 25
Project Type and Category:	Hydrogen production
Project Duration:	1989 – 1998
Project Participants:	Institute f. organ. Chemie (IOC), Prof. Dr. F. Effenberger (A4.2) Inst. f. physikal. Elektronik (IPE), Prof. Dr. J.H Werner (A4.1) Other participants of sfb 270: Project Manager:Prof. Dr. Werner, Prof. Dr. Effenberger
Sponsor:	DFG: 294,800 DM/y since 1989
Project Budget and Funding:	Approx.. 2,6 Mio DM
Project Description and Objectives:	In a photoelectrochemical cell hydrogen is generated at the surface of a semiconductor electrode which is illuminated and dips in an electrolyte. Early approaches with TiO2-electrodes succeeded in principle [1], but, because of its high band-gap, this material turns out to be unfavourable for collecting of solar energy. In the other hand silicon with a band-gap of 1,1 eV has a much lower threshold of absorption but is thermodynamically instable towards photoanodic but also cathodic processes of corrosion [2]. Besides thin layers of metal or metal oxides organic polymers provide a useful means for coating and hence protection of the electrodes [3]. As thin films of conducting polymers like e.g. polyacetylene, polyaniline, polypyrrole, and polythiophene [4,5] are useful materials for passivation of photoanodes, electronmediators must be incorporated in the surface layer of the cathodes to guarantee an efficient charge transfer. Moreover from a kinetic point of view it is necessary to link catalysts like e.g. RuO2, IrO2, Pt or Pd [6] or molecular systems [7] to the surface to accelerate such multiple electron processes like that in water electrolysis and to reduce the overpotential. Due to the stability of polythiophene in aqueous solution [5b] this polymer seems to be appropriate for protection of electrodes. Consequently this polymer, provided with redox-active, catalytic sidegroups, should match the expected features. According to their redox potentials, 4,4'-bipyridine salts as well as an organometallic rhodium-complex, a real homogeneous molecular catalyst, were considered convenient for the reductive part of water electrolysis and ferrocene for the oxidative part (Scheme 1). The polymers should be built up via electropolymerization of functionalized mono- or oligothiophenes. In the first part of this programm a convenient synthetic route to thiophenes and oligo-thiophenes with 3-(-haloalkyl)-substituents had to be developed in order to bind the redox catalysts. The second part should include the electrooxidative polymerization of the monomeric building blocks as well as investigations refering to polymer and film characteristics like redox properties, conductivity and absorption. To probe the protective and catalytic properties of these materials under working conditions films on silicon wafers were prepared by spin coating of polymer solutions followed by evaporation of the solvent.
Technical Goals:	Modification and optimization of electrode surfaces
Project Status	Final Phase
Preliminary or Final Results:	The synthesis of a series of mono- and oligothiophenes 1 - 4 with various alkyl chain length and substitution pattern was realized via nickel catalyzed Grignard coupling reaction [8]. As outlined in Scheme 2, the redox catalysts were linked by nucleophilic substitution of the terminal bromide. The 2,2'-bipyridine derivative 6 was converted to the desired rhodium complex 8 (x=2,3; n=5) in cooperation with project A5. Apart from compounds, in which the oxidation potential of the thiophene is unfavourable with respect to its catalytic moiety, the oligothiophenes in 5, 7 und 8 were electropolymerized. In the cyclic voltammogram of a platinum electrode coated with polythiophene 5 (x=1, n=8) redox curves of the incorporated viologen (Eo[V2+/+]= -0,74 V; Eo[V+/o]= -1,2 V; gg. Ag/AgCl) are clearly visible (Figure 1). Figur1:Cyclic volCyclic voltammogram of a P5 (n=8) film at a scan rate of 50 mV/s. Inset: electrooxidative polymerization of 5 (n=8) [5x10-2 M] in acetonitrile (0.1 M TBAHPF) in a potentiodynamic experiment.tammogram of a P5 (n=8) film at a scan rate of 50 mV/s. Inset: electrooxidative polymerization of 5 (n=8) [5x10-2 M] in acetonitrile (0.1 M TBAHPF) in a potentiodynamic experiment. Films prepared from 5 and 8, respectively, either on silicon or glassy carbon electrodes, on which hydrogen reveals a considerable overpotential, demonstrate the suitability of these polymers by both passivating the surface of the photoelectrode and activating the generation of hydrogen catalytically. The polymers obtained by electrooxidation of 5 or 8 are soluble in polar organic solvents. While p-Si is applicable only in the reductive but not in the oxidative potential range polymer layers with thicknesses between 0.05 and 5 µm were obtained by spin coating from those solutions. As charge transfer through the layer is assumed to take place via viologen units additional platinum particle are necessary to release hydrogen. Figure 2 presents the possible variations in distribution investigated. Type 1: precipitation of Pt directly on the semiconductor surface by electron-beam evaporation providing a homogeneous Pt-layer (A) or by galvanostatic reduction of H2PtCl6 solutions yielding island structures (B) with subsequent polymer film precipitation; type 2: homogeneous distribution of Pt particles achieved by casting with a suspension of finest Pt powder in the polymer solution; type 3: precipitation of Pt of the polymer film by sputtering (A) or by galvanostatic reduction of H2PtCl6 solutions (B). Figure 2: Possible distribution of platinum in the system p-silicon/polymer/ platinum. The photoelectrodes were characterized via I-U-relationships as shown in Figure 3a. From comparative measurements the following results can be summerized. p-Si-electrodes, both coated and uncoated with polymer resemble in the I-U-diagram (curves 2, 3) and reveal no photopotential in absence of a catalyst. Layers of poly(3-dodecylthiophene) (P3DDT), having no viologen side groups, behave as an insulator (curve 1). Due to individual polymer absorption lamdamax = 480-500 nm) electrodes with catalytically active layers work with a gain in energy only if the layer thickness is <1µm. Thin films are preferred in long-term measurements since thick layers often peel off after short operating time. Coated and catalytically active photoelectrodes provide a greater energy gain in strong acidic medium rather than in neutral environment according to literature findings [9]. In thin films the distribution of Pt (type 1-3) hardly influences the observed photopotential (100-280 mV), but type 1 electrodes result in a higher long-term stability than type 3. Electrodes of type 1 (A) show an ohmic contact between Si and Pt and are therefore less suitable. On the contrary, the polymer in type 3 (A) acts as a Schottky barrier. Photocathodes of type 1 (B) (curve 5) provide the best result with an efficiency of 1.2 % and a high long-term stability. In a test, broken off 120 h later (Figure 3b) in which the electrode worked at the maximum power, the photocurrent practically remains constant after a small initial reduction but slightly improved photovoltage (curve a). Comparatively the photocurrent of an uncoated p-Si/Pt-electrode decreases significantly within 24 h due to photocorrosion (curve b). Within this long testing time about 0.35 l hydrogen evolved at an electrode area of 1 cm² without degradation of the polymer layer and the silicon surface, respectively. Figur3: I-U-relationships of p-Si/polymer(P3DDT)- (1), p-Si/polymer(P5)- (2), p-Si- (3), p-Si/ Pt- (4) and p-Si/Pt/polymer(P5)-photocathode at pH 0 and 20 mV/s. Photocurrent-time-relationship of a p-Si/Pt/polymer(P5)-photocathode (a) working at maximum power at pH 0 within 120 h; (b) p-Si/Pt-photoelectrode without polymer. Rhodium complex functionalized polythiophene P8 (x=2, n=5), polymerized on glassy carbon
Related Reference Papers and Other Publications:	Internet-Site: www.uni-stuttgart/UNIuser/sfb270/A4_2_E.htm Wasserstoff als Energieträger, ISBN 3-00-001 796-8 [1]A.Fujishima, K.Honda, Bull.Chem.Soc.Jpn. 44 (1971) 1148 [2]H.Gerischer, Pure Appl.Chem. 52 (1980) 2649 [3]M.S.Wrighton, Science 231 (1986) 32 [4.a] R.A.Simon, M.S.Wrighton, Appl.Phys.Lett. 44 (1984) 930 [4.b]K.M.Kost, D.E.Bartak, B.Kazee, T.Kuwana, Anal.Chem. 60 (1988) 2379 [4.c]S.Holdcroft, B.L.Funt, J.Electroanal.Chem. 240 (1989) 89 [5.a]Z.Deng, W.H.Smyre, H.S.White, J.Electrochem.Soc. 136 (1989) 2152 [5.b]G.Tourillon, F.Garnier, J.Phys.Chem. 88 (1984) 5281 [6]R.Kötz, S.Stucki, Electrochim.Acta 31 (1986) 1311 [7] V.Kölle, M.Grätzel, Angew.Chem. 99 (1987) 564 [8.a] P.Bäuerle, F.Würthner, S.Heid, Angew.Chem. 102 (1990) 414 [8.b] P.Bäuerle, S.Scheib, Adv.Mater. 5 (1993) 848 [9]A.Bandi, H.-M.Kühne, J.Schefold, M.Specht, M.Wendt, DECHEMA-Monogr. 117 (1989)