Final Report Summary - HIGH-VOLTAGE PV (New materials for high voltage solar cells used as building blocks for third generation photovoltaics)
Final report:
Within this project methods were investigated to develop single junction photovoltaic cells with a high photovoltage as building blocks for multi-junction tandem devices. Increasing the light to electric power conversion efficiency of single junction solar cells beyond the thermodynamic limit of 33 % to divide the solar spectrum into spectral windows which are converted by photovoltaic cells optimised for the respective spectral range. Especially for the visible spectral range photovoltaic cells that generated a voltage above 1V are missing.
For a photovoltaic cell one needs an absorber material, a meta-stable excited state, a driving force for electron and holes transport and contacts which are selective to either electrons or holes. To increase the photovoltage the energy levels have to be optimised in a way that potential losses during charge generation, separation and transport are minimised. Furthermore, recombination has to be suppressed. Over the 2 years of the project two different types of photovoltaic cells were investigated:
a) Nano-composite mesoporous films that are sensitised with a dye monolayer or a thin quantum dot (QD) layer and
b) Thin film all-oxide photovoltaic cells based on semiconductors with a bandgap in the visible range.
In nano-composite solar cells the following methods to increase the photovoltage were investigated:
a1) Minimising potential losses by optimisation of the conduction band level of the electron conducting phase. To achieve this, inorganic coatings were deposited generating an interface dipole between the electron and hole conducting phase. It was shown that such an interface dipole shifts the conduction band level of mesoporous TiO2 film with respect to the redox potential of a liquid electrolyte or to the valence band level of a solid state hole conductor. It was successfully demonstrated in this project that a Lanthanum surface treatment can increase the photovoltage of dye-sensitised solar cells by more than 100 mV, however on the expense of the photocurrent. Open circuit potential decay and charge extraction measurements were used to understand the mechanism of the La coating and it was concluded that basic deposition conditions lead to an upward shift of the TiO2 conduction band but lead simultaneously to a reduced charge injection efficiency from the dye into the TiO2. This work was published in J. Phys. Chem. C (see Publication No.
4) and Adv. Funct. Mater. (see Publication No. 5).
a2) Minimising potential losses between the absorber and the electron conducting material by deposition of organic dipoles onto CdS QDs that were synthesized onto a mesoporous TiO2 film. Here it was possible to show that the excited QD state shifts with respect to the TiO2 conduction band as a function of the molecular dipole moment, even though the dipole was present at the QD/electrolyte interface and not at the QD/TiO2 interface. Thus molecular dipoles proved to be a powerful tool to align the excited state energy level of the absorber with respect to the electron and hole conducting phase. This work was published in J. Am. Chem. Soc. (see Publication No.
3)
a3) Optimising the energy difference between the electron and hole conducting phase using alternative electron and hole conducting materials. ZrO2 was investigated as an alternative wide bandgap oxide with a conduction band edge closer to the vacuum level compared to TiO2. In dye-sensitised solar cells it is known that electron injection from the photo-excited dye into ZrO2 is not possible because the exited states are located within the bandgap of ZrO2. Unexpectedly it was found that CdSe QDs are able to inject electrons from the excited into the ZrO2 conduction band, however the photovoltage was not enhanced, which was also surprising. We proposed that specific adsorption of ions from the polysulfide electrolyte to the TiO2 and ZrO2 surface creates interface dipoles of different strength such that the location of band edges is strongly dependent on the electrolyte composition. Thus in aqueous polysulfide electrolyte electron injection was possible while dyes in conjunction with organic poly-iodine electrolyte cannot inject. This work was published in Phys. Chem. Chem. Phys. (see Publication No. 2). Furthermore, effective counter electrodes were developed for aqueous polysulfide, published in J. Phys. Chem. C (see Publication No. 6).
a4) Suppressing recombination at the conducting transparent front contact in contact with the redox electrolyte. While this recombination path is insignificant in dye-sensitised solar cells it was found that it is significant in QD sensitised solar cells that use an aqueous polysulfide electrolyte. It was demonstrated that suppressing the recombination by the deposition of a compact oxide layer enhances the fill factor and the photovoltage and leads to an overall efficiency improvement of CdSe QD sensitised solar cells. For this project a computer controlled spray pyrolysis scanner was build and the results have been submitted for publication (J. Phys. Chem. C).
a5) A general overview of QD sensitised solar cells with an extended discussion on energy level alignment was presented in a review article in ChemPhysChem (see Publication No. 1).
b1) Development of thin film all-oxide photovoltaic cell with a high photovoltage. The spray pyrolysis system was furthermore used to synthesize all-oxide photovoltaic cells based on a compact wide bandgap oxide layer (TiO2, ZnO, ZrO2 or SnO2) and a metal oxide semiconductor with a bandgap in the visible such as Cu2O or Fe2O3. First all-oxide prototypes were demonstrated such as TiO2/Cu2O and SnO2/Fe2O3 heterojunction cells with photocurrents in the range of hundreds of mA/cm2 and photovoltages of hundreds of mV.
During the period of the fellowship it was possible to increase the light to electric power conversion efficiency of QD sensitised solar cells beyond 4 %, shifting it into a range where it starts to become attractive for further commercial development.
Within this project methods were investigated to develop single junction photovoltaic cells with a high photovoltage as building blocks for multi-junction tandem devices. Increasing the light to electric power conversion efficiency of single junction solar cells beyond the thermodynamic limit of 33 % to divide the solar spectrum into spectral windows which are converted by photovoltaic cells optimised for the respective spectral range. Especially for the visible spectral range photovoltaic cells that generated a voltage above 1V are missing.
For a photovoltaic cell one needs an absorber material, a meta-stable excited state, a driving force for electron and holes transport and contacts which are selective to either electrons or holes. To increase the photovoltage the energy levels have to be optimised in a way that potential losses during charge generation, separation and transport are minimised. Furthermore, recombination has to be suppressed. Over the 2 years of the project two different types of photovoltaic cells were investigated:
a) Nano-composite mesoporous films that are sensitised with a dye monolayer or a thin quantum dot (QD) layer and
b) Thin film all-oxide photovoltaic cells based on semiconductors with a bandgap in the visible range.
In nano-composite solar cells the following methods to increase the photovoltage were investigated:
a1) Minimising potential losses by optimisation of the conduction band level of the electron conducting phase. To achieve this, inorganic coatings were deposited generating an interface dipole between the electron and hole conducting phase. It was shown that such an interface dipole shifts the conduction band level of mesoporous TiO2 film with respect to the redox potential of a liquid electrolyte or to the valence band level of a solid state hole conductor. It was successfully demonstrated in this project that a Lanthanum surface treatment can increase the photovoltage of dye-sensitised solar cells by more than 100 mV, however on the expense of the photocurrent. Open circuit potential decay and charge extraction measurements were used to understand the mechanism of the La coating and it was concluded that basic deposition conditions lead to an upward shift of the TiO2 conduction band but lead simultaneously to a reduced charge injection efficiency from the dye into the TiO2. This work was published in J. Phys. Chem. C (see Publication No.
4) and Adv. Funct. Mater. (see Publication No. 5).
a2) Minimising potential losses between the absorber and the electron conducting material by deposition of organic dipoles onto CdS QDs that were synthesized onto a mesoporous TiO2 film. Here it was possible to show that the excited QD state shifts with respect to the TiO2 conduction band as a function of the molecular dipole moment, even though the dipole was present at the QD/electrolyte interface and not at the QD/TiO2 interface. Thus molecular dipoles proved to be a powerful tool to align the excited state energy level of the absorber with respect to the electron and hole conducting phase. This work was published in J. Am. Chem. Soc. (see Publication No.
3)
a3) Optimising the energy difference between the electron and hole conducting phase using alternative electron and hole conducting materials. ZrO2 was investigated as an alternative wide bandgap oxide with a conduction band edge closer to the vacuum level compared to TiO2. In dye-sensitised solar cells it is known that electron injection from the photo-excited dye into ZrO2 is not possible because the exited states are located within the bandgap of ZrO2. Unexpectedly it was found that CdSe QDs are able to inject electrons from the excited into the ZrO2 conduction band, however the photovoltage was not enhanced, which was also surprising. We proposed that specific adsorption of ions from the polysulfide electrolyte to the TiO2 and ZrO2 surface creates interface dipoles of different strength such that the location of band edges is strongly dependent on the electrolyte composition. Thus in aqueous polysulfide electrolyte electron injection was possible while dyes in conjunction with organic poly-iodine electrolyte cannot inject. This work was published in Phys. Chem. Chem. Phys. (see Publication No. 2). Furthermore, effective counter electrodes were developed for aqueous polysulfide, published in J. Phys. Chem. C (see Publication No. 6).
a4) Suppressing recombination at the conducting transparent front contact in contact with the redox electrolyte. While this recombination path is insignificant in dye-sensitised solar cells it was found that it is significant in QD sensitised solar cells that use an aqueous polysulfide electrolyte. It was demonstrated that suppressing the recombination by the deposition of a compact oxide layer enhances the fill factor and the photovoltage and leads to an overall efficiency improvement of CdSe QD sensitised solar cells. For this project a computer controlled spray pyrolysis scanner was build and the results have been submitted for publication (J. Phys. Chem. C).
a5) A general overview of QD sensitised solar cells with an extended discussion on energy level alignment was presented in a review article in ChemPhysChem (see Publication No. 1).
b1) Development of thin film all-oxide photovoltaic cell with a high photovoltage. The spray pyrolysis system was furthermore used to synthesize all-oxide photovoltaic cells based on a compact wide bandgap oxide layer (TiO2, ZnO, ZrO2 or SnO2) and a metal oxide semiconductor with a bandgap in the visible such as Cu2O or Fe2O3. First all-oxide prototypes were demonstrated such as TiO2/Cu2O and SnO2/Fe2O3 heterojunction cells with photocurrents in the range of hundreds of mA/cm2 and photovoltages of hundreds of mV.
During the period of the fellowship it was possible to increase the light to electric power conversion efficiency of QD sensitised solar cells beyond 4 %, shifting it into a range where it starts to become attractive for further commercial development.