Community Research and Development Information Service - CORDIS

FP7

GLOBASOL Report Summary

Project ID: 309194
Funded under: FP7-ENERGY
Country: Italy

Final Report Summary - GLOBASOL (Global solar spectrum harvesting through highly efficient photovoltaic and thermoelectric integrated cells)

Executive Summary:
The GLOBASOL project aimed at developing new concepts, materials and devices for the efficient conversion of the solar energy to electrical power. The basic idea is to put in synergy new and advanced concepts of light management and new optically active materials in order to harvest the global solar radiation and convert it to electricity, through photovoltaic/thermoelectric integrated devices. Within this approach, the idea is the use of the medium-to-high energy part of the solar radiation, i.e. up to wavelengths of 1100 nm and its conversion by sensitized mesoscopic solar cells (SMSC), where the sensitization will be performed by organometallic complexes, organic dyes, semiconductor quantum dots (QD), and/or suitable combinations of the above materials to reach near quantitative conversion of incident photons-to-electric current across the ultraviolet and visible part of the spectrum.
During the Globasol project new molecular and nanostructured materials suitable for assembling future generation SMSC and TE cells were prepared. A detailed physico-chemical characterisation of the materials, fundamental for the comprehension of interfacial phenomena for enhancing/optimising light absorption, charge transfer processes, and heat/current flow have been also carried out.
For PV applications attention was paid to: i) optimisation of dye adsorption; ii) preparation and characterisation of quantum dot nanocrystals suitable to act as IR absorbers in PV devices; iii) synthesis and characterisation of inorganic and hybrid organic-inorganic materials suitable for the preparation of quasi solid electrolytes.
Beside materials already planned for SMSC, organic-inorganic perovskite for solid-state solar cells were developed, taking into account the strong interest and performances that these materials showed in the last years. The main objectives achieved so far concerned the optimisation of methods allowing a better control of perovskite morphology and of TiO2/ perovskite composite materials. A certified efficiency record of 21 % was indeed reached (see Annex 1).
The long wavelength part of the spectrum (above 1100 nm) is exploited by thermoelectric (TE) devices, specially designed to work with the maximum efficiency in combination with the SMSC.
The identification and selection of material building blocks and architectures suitable for highly efficient optical light management covering a broad spectral range between 400 nm up to 1200 nm, have been also carried out.
The project included highly sophisticated experimental studies, supported by advanced theoretical modeling, to define innovative materials both for the PV (organometallics, organic dyes, QD, electrolytes) and the TE (bulk and nano-sized semiconductor couples) cells: this research led to a substantial increase of the efficiency of the two separate cell types with respect to the state of the art. The newly developed materials have been assembled in high efficiency SMSC and TE cells: the integration of the various parts has been designed carefully, in order to maintain the high performances of the separated cells also in the integrated device. Overall system conversion efficiencies of 28 % have been achieved.
The project is strongly multi-disciplinary, involving well-known research groups in the fields of organic and inorganic chemistry, physical and theoretical chemistry, physics, engineering and energetics: the research management will ensure at any stage the feedback loop between the scientific and technological activities (developing both new SMSC and TE modules and integrated devices). Five Universities, one Research Institution, and one High-Tech SME guarantee a scientific and technological multidisciplinary research activity.

Project Context and Objectives:

The GLOBASOL project aims at developing new concepts, materials and devices for the efficient conversion of the solar energy to electrical power. The basic idea is to put in synergy new and advanced concepts of light management and new optically active materials in order to harvest the global solar radiation and convert it to electricity, through photovoltaic/thermoelectric integrated devices.
Within this approach, the medium-to-high energy part of the solar radiation, i.e. up to wavelengths of 1100 nm, will be harvested and converted by sensitized mesoscopic solar cells (SMSC), where the sensitization will be performed by organometallic complexes, organic dyes, semiconductor quantum dots (QD), and/or suitable combinations of the above materials to reach near quantitative conversion of incident photons-to-electric current across the ultraviolet and visible part of the spectrum. Such unconventional photo-electric converters achieve photon containment - and hence enhanced light harvesting - without sacrificing the transparency to lower energy photons: these systems can therefore be used as top cells in tandem arrangements to harvest a large fraction of solar emission.
The high wavelength part of the spectrum (above 1100 nm) is exploited by thermoelectric (TE) devices, specially designed to work with the maximum efficiency in combination with the SMSC. A key concept is the development of new materials with high ZT (figure of merit) in the temperature range 500 K - 700 K, either based on n- and p-doped nanostructured bulk semiconductor couples, or on assembled quantum dots (QD) semiconductor arrangements.
The efficient separation of the solar spectrum could be accomplished either by tandem architectures, or by suitable optical devices, possibly including concentration lens or mirrors for the infrared fraction to enhance the performances of the TE generation. Figure 1 illustrates the general light management proposed in GLOBASOL.
The integration of the various parts was designed carefully, in order to maintain the high performances of the separated cells also in the integrated device.

One of the objectives of the Globasol project is to produce highly efficient solar cells and to collect information on their life time. A benchmarking analysis and identification of most suitable hole transporting materials for SMSC application has been carried out, as well as the identification of photonic crystals to enhance the performances of SMSC and of optical apparatus for the light splitting.
Thus, particular attention has been focused to the screening of the electronic properties of light absorbers, organic sensitizers and QDs already available from the project partners, or easily achievable on the basis of their experience.
It has to be pointed out that over the last period, we have witnessed an unexpected breakthrough and rapid evolution in the field of emerging photovoltaics, with the realization of highly efficient solid-state hybrid solar cells based on organometal trihalide perovskite absorbers (Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith, H. J. Efficient Hybrid Solar Cells Based on Meso-Superstructured Organometal Halide Perovskites. Science 2012, 338 (6107), 643-647; Burschka, J.; Pellet, N.; Moon, S. J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Grätzel, M. Sequential Deposition as a Route to High-Performance Perovskite-Sensitized Solar Cells. Nature 2013, 499, 316-319).
The perovskites are semiconductors with a direct band gap, large absorption coefficient and high carrier mobility which make these solids an attractive class of materials as light harvesters in heterojunction solar cells (D. Bi et al. J. Phys. Chem. Lett. 2013, 4, 1532−1536). The use of these materials as light harvesters in solar cells led to arrive at values of conversion efficiency of 16-17% (M. Liu et al. Nature 2013, 501, 395-398; T. Moehla, J. Hyeok Im, Y. Hui Lee, K. Domanski, F. Giordano, S. M. Zakeeruddin, I.M. Dar, L-P. Heiniger, M. K. Nazeeruddin, N-G. Park and M. Grätzel, Strong Photocurrent Amplification in Perovskite Solar Cells with a Porous TiO2 Blocking Layer under Reverse Bias, J. Phys. Chem. Letters, 2014, DOI: 10.1021/jz502039k). On the basis of the interesting results reported in the literature, inorganic-organic perovskites were considered as light harvester in PCE solar cells. This type of materials was thus added to the other solids already planned in the Project because of their peculiar performances for PV applications.

As for materials needed for TE applications, PbS, PbSe and Bi2Te2 are the most important materials, nevertheless one of the objective of the Project has been the testing of other solids as well aiming to reach ZT values of around 1.5 in the 300-500 K temperature range.
The selection/definition of most promising materials for both SMSC and TE applications have been driven also by theoretical and computational modelling aiming both at reproducing the effects of specific chemical modifications in the device and at modelling the properties of materials as a function of the physical and chemical nature of the system. To reach this goal, a computational protocol for the multiscale modeling of the Globasol materials, validated on the available experimental data, has been developed.
Efforts are also expected in the design, preparation and characterization of PV and TE devices by exploiting materials developed in the frame of the Globasol Project.
Particular attention is given to the definition of device architectures and light management systems (by exploiting photonic crystals). In the first 18 months of the Project, attention has been paid to the definition of the architecture of the integration of both PV and TE devices.
As a key-step, the main synthetic routes that will be used for the preparation of both SMSC and TE components have been validated in the first 18 months of the Project. This allowed to provide the amount of materials needed for the testing of both PV and TE devices.
Another objective of the first 18 Months of the Project is the development of a proof of concept of SMSC and TE modules incorporating novel materials produced in the frame of Globasol project and of the optical architecture for the management of the solar spectrum by splitting the different components, minimizing losses and dispersions.
At the end of 18 month, the objective was to prepare single devices with the following performances:
- SMSC absorbing below 750 nm (prepared by using Globasol sensitizers and hole conductors): at least 12%;
- SMSC absorbing between 750 and 1100 nm (prepared by using Globasol QDs and hole conductors): at least 8%;
- TE modules (with Globasol QD or nanowires): at least 4%.

These objectives required an accurate optimization of core materials useful for the assembly of the device. The integration of the various parts has been designed carefully, in order to maintain the high performances of the separated cells also in the integrated device.
The final goal of Globasol research is to improve the performance of SMSC tandem cell by absorbing visible and near IR (from below 750 to 1100 nm) light and TE device absorbing IR radiation (above to 1100 nm). For this reason, during the second part of the Project, efforts are carried out for the preparation and characterization of PV and TE devices by exploiting materials developed in the frame of the Project.
Particular attention is given to the definition of device architectures and light management systems by exploiting photonic crystals.
As a final objective, of a proof of concept of SMSC and TE modules incorporating novel materials produced in the frame of Globasol project and of the optical architecture for the management of the solar spectrum by splitting the different components, minimizing losses and dispersions has to be produced. At the end of the Project, the proof of concept of the hybrid device able to co-generate electrical power with efficiencies above 30% with a rational and efficient exploitation of the whole solar spectrum is expected.

Project Results:
The main scientific and technological results obtained during the three-years project are reported in this section in relation to the different type of materials/activities that the Consortium developed in the frame of the Project. In addition, attention was paid to decrease cost production of the materials (e.g. by upscaling material preparation) and to the study of life time and possibility to recycle of the devices prepared in the frame of the Project.
In the sections below, the main results obtained in the frame of the Project are reported following the work plan of the GLOBASOL Project (i.e. Work package organisation).

1.1.3.1 Benchmarking, materials definition and theoretical modelling (activity related to WP1)
The activities carried out in the frame of WP1 have three main goals: i) selection of the best candidates for the core materials of SMSC and TE cells, taking into account the state-of-the-art results in the literature and the expertise of the partners in the various fields, ii) the generation and application of a Life Cycle Assessment (LCA) approach both as a practical decision making tool through the project R&D activities (and the related innovative materials developed) and as a reference tool to guarantee the sustainability of SMSC and TE cells production; iii) the theoretical and computational modelling of new materials and devices.
The LCA activity started at month 6 and a specific questionnaire for data collection related to the materials studied in Globasol Project has been implemented and discussed with the partner. In addition, a literature survey has been also carried out in order to have information on LCA analysis of different photovoltaic cells and energy payback of such systems.

- Materials definition and specification requirements
Selection of materials for developing high performance TE and PV devices has been done, with particular emphasis on:
i) organic and inorganic-organic hybrid sensitizers for SMSC;
ii) quantum dots;
iii) hole transport materials;
iv) nanostructured materials for TE applications;
v) photonic materials.

Concerning organic sensitizers, different porphyrin sensitizers have been studied in the frame of the Project especially by UAM and EPFL partner, unveiling a world record dye sensitized solar cells efficiency of 12%. The porphyrin chromophore possess intrinsically strong light absorption in both the Soret and Q-bands, however lacks absorption between 500-600 nm. To overcome the lack of absorption in the green part, co-sensitization was applied to reach 12% power conversion efficiency. Within the GLOBASOL project the EPFL partner molecularly engineered a novel sensitizer, with bis(2',4'-bis(hexyloxy)-[1,1'-biphenyl]-4-yl)amine donor group and a new anchoring group with the pro-quinoidal 2,1,3-benzothiadiazole (BTD) acceptor resulting panchromatic response. This work has been published in 2014 in Nature Chemistry journal.
Moreover, the new concept of perovskite-based solar cells is taken as a benchmark. The general chemical formula for perovskite compounds is AMX3, where 'A' and 'M' are two cations of very different sizes, and X is an anion that bonds to both. The ideal cubic-symmetry structure has the M cation in 6-fold coordination, surrounded by an octahedron of anions (MX6), and the A cation in 12-fold cuboctahedral coordination.
By virtue of the their ‘customizable’ low band gap, the perovskite absorbers showed superior advantages over traditional dyes by much more strongly absorbing over a broader range, enabling complete light absorption in films as thin as 500 nm. This is specifically favorable for the solid-state cells, where thickness limitations of around 2 µm have historically limited the light absorption and photocurrent generation.
PbS and PbSe Quantum Dots have been also considered as attractive candidates for SMSC devices as they can absorb near IR photons. They have already been demonstrated to show high efficiencies and possible multi-exciton generation (MEG) which potentially enables a significant increase of the conversion efficiency. Due to the device requirements, the activity has been focussed on QDs with an absorption threshold at around 1000 nm (900-1100 nm) for capturing the near IR photons.
Extending the variety of possible light absorbing materials the new synthesis of nanocrystalline Bi2S3 will be investigated as a less toxic alternative.
On the basis of the literature survey and partners knowledge, different materials (both inorganic and hybrid organic-inorganic) have been selected as filler for the preparation of non-liquid or quasi-solid electrolytes (hole transport materials). Particular attention has been devoted to the study of the effect of porous and layered materials with different structural properties when used as additives for electrolytes based on ionic liquids. On the basis of the collected results, the preparation method of these materials (with particular emphasis on monodispersed silica particles) has been optimized also to allow the introduction of organic species on the silica surface of groups suitable to increase the chemical affinity with the electrolyte (aiming at a good dispersion of the additive in the electrolyte) and to positively modify the electronic properties of TiO2 (for instance by introduction of basic groups like –NH2 species). Moreover, the dimension of materials has been varied in order to promote light scattering phenomena that can be useful to improve the final device efficiency. Beside porous materials, cationic clays, such as saponite and hectorite, have been also used as additives for the preparation of quasi-solid electrolytes. These samples already demonstrated a positive effect when used for non-liquid electrolyte preparation (D. Costenaro, C. Bisio, F. Carniato, G. Gatti, F. Oswald, T. B. Meyer, L. Marchese, Solar Energy Mater & Solar Cells, 117 (2013) 9-14).
An analysis of solar irradiation on PV/TE system indicated that the temperature of available heat to the TE system is likely to reach only around 450 K with a maximum possible temperature up to 500 K for a concentration ratio of 100. Clearly, the selection of candidate materials for this application will be limited to this temperature range. Bi2Te3 is the only material that has a good ZT value over 300K-400K. PbTe based alloys also show the highest ZT values over 400K-500K. Therefore, crystalline and PbTe are employed as the benchmarking materials and the thermoelectric properties of crystalline Bi2Te3 have been characterised.

Finally, the identification and selection of material building blocks and architectures suitable for highly efficient optical light management covering a broad spectral range between 400 nm up to 1200 nm, have been finalized. Inks based on selected material building blocks and procedures for deposition and processing ink layers have been developed. One- dimensional photonic crystals (1DPC’s) optically active in the 700 - 750 nm range and the 1000 - 1200 nm range can now be manufactured. Nanoparticles of TiO2 and SiO2 have been selected as building blocks to constitute the alternating periodic one dimensional photonic crystal architecture. Alternating, nanoporous layers of nanoparticles of TiO2 and SiO2 have been used for the manufacturing of the light management system splitting the sun spectrum in separate components.

- Life Cicle Analysis of materials and PV/TE components
A specific questionnaire for data collection related to the materials studied has been implemented and the acquisition of LCA information is started in July 2014. A thorough literature review was also conducted in relation to LCA analysis of different photovoltaic cells and energy payback of such systems.
The objective of this task is to generate and apply a Life Cycle Assessment (LCA) approach either as a practical decision making tool through the project R&D activities and the related innovative materials developed, and as a reference tool to guarantee the sustainability of SMSC and TE cells production. From month 18 to month 36, a specific questionnaire for data collection related to the devices has been sent to partners for the acquisition of LCA information and data has been elaborated to obtain environmental impacts for the final device produced within the Globasol project.
A detailed literature review was also conducted in relation to LCA analysis of different photovoltaic cells and energy payback of such systems.
A cradle to gate analysis of the final device developed in the project has been performed considering: i)SMSC device from Fraunhofer ISE; ii) TE device from CU and iii) Integrated device from EXEGER.
The devices analysis was possible thanks to the results obtained by the “materials for energy” analysis provided in the first part of the project considering i) photosensitizer from UAM, ii) gel electrolytes from UNIPMN, iii) Photosensitizer and TE materials from TUD, iv) TE Materials from CU and v) photonic materials from EXEGER.
- Development of specific hybrid computational procedures and modelling of novel materials and interfaces
In the frame of this activity, new multiscale computational techniques, involving and combining different theoretical approaches, for the modelisation of materials and devices developed in the frame of the Project have been developed. Besides optimizing the theoretical methods for the various layers, on the basis of the literature data and experimental results, efficient multiscale procedure have been developed to perform calculations on composite materials.
The computational modeling has been performed at different levels of theory, to account for the various scales of complexity of the studied systems.
Several aspects of the calculation procedure have been set up, in order to optimize the results for the systems of interest.
The theoretical approaches developed and validated during the first step of the project have been employed to model composite materials, which are the core of Globasol devices. The complex materials have been modelled exploiting the multiscale procedures: several computational approaches have been applied to the various part of the system and the results integrated by suitable programs which have allowed to describe the geometrical features of the interfaces (electrode-sensitisers, electrolyte-sensitisers..) as well as the energy and the electron transfer processes across the interfaces, and help designing the best inorganic and organic materials and the most efficient device architecture.

- Modeling of hybrid devices
Potential system architectures for a hybrid device consisting of a photovoltaic (PV) cell and a thermoelectric generator (TEG) have been modelled in order to estimate the maximum efficiency of the system. Two designs have been considered, a planar architecture and a parabolic design.
Whereas the planar design does not seem to bring any benefit for the device’s efficiency, the architecture including optical as well as thermal concentration can lead to combine efficiencies in the range of 19%. This corresponds to an increase of the efficiency of about one percentage point. For the calculations, it has been assumed that the SMSC is cooled, i.e. the efficiency of the solar cell stays constant. In reality, a sole SMSC would heat up and its efficiency would drop significantly. If the temperature of the SMSC would increase from e.g. 20°C to 50°C, the efficiency would drop by 5% (www.dyesol.com). In the considered Globasol hybrid setups, the IR part of the solar spectrum, which mainly causes the heating up, does not hit the solar cell. Thus, the solar cell is not affected by such an efficiency decrease compared to the sole solar cell. In conclusion, that means that for the parabolic architecture the efficiency increase for the complete hybrid setup can potentially reach six percentage points compared to the sole SMSC cell.

1.1.3.2 Synthesis and characterisation of new materials for photovoltaic and thermoelectric applications (activity related to WP2)

a) Highly absorbing panchromatic sensitizers
In the frame of the GLOBASOL Project, UAM developed the synthesis of novel polypyridine ruthenium complexes- and phthalocyanine-based sensitizers for SMSC capable of harvesting light and injecting electrons efficiently into the semiconductor metal-oxide conduction band.
-Polypyridine ruthenium complexes
Increased conjugation on the ancillary bipyridine ligands by the incorporation of appropriate subunits based on two or more conjugated thiophene units were attained by UAM researchers. In addition, bulky groups were also added in the sensitizer molecular structure in order to reduce aggregation.
The photovoltaic performances of novel Ru(II)-bipyridine heteroleptic complexes (M. Sánchez Carballo, M. Urbani, A. Kumar Chandiran, D. González-Rodríguez, P. Vázquez, M. Grätzel, M.K. Nazeeruddin and T. Torres, Dalton Trans. 2014, M. Urbani, M. Medel,S. Amit Kumar, M. Ince, A. N. Bhaskarwar, D. Gonzalez-Rodriguez, M. Graetzel, M. K. Nazeeruddin, T. Torres, Chem. Eur. J. 2015, M. Urbani, A. Abate, M. Grätzel, S. Ahmad, T. Torres, M. K. Nazeeruddin, Dalton Trans., 2015, 44, 10847–10851), incorporating branched and bulkier alkyl chains compared to their linear analogues, were studied. In both series, it was found that dyes containing 2-methyl-hex-2-yl substitution gave better performances than 1,1-dipropylbutyl. The best overall performances over the four dyes were obtained for TT207 (CYC-B11 analogue) that contain 2-methylhex-2-yl type substitution, achieving an overall PCE of 8.5%. Further optimization of dye DSSCs with respect to the dye-uptake solvent, and electrolyte composition, led to a maximum PCE of 9.1% under AM1.5G standard conditions.
The use of bulky substitutions in a cyclopenta(2,1-b;3,4-bA)dithiophene (CDT) Ru (II) bipyridyl metal complex, to obtain high open-circuit potential in the dye sensitized solar cell was also studied (Adapting Ruthenium Sensitizers to Cobalt Electrolyte Systems”, S. Amit Kumar, M. Urbani, M. Medel, M. Ince, D. Gonzalez Rodriguez, A. Kumar Chandiran, A. N. Bhaskarwar, T. Torres, Md. K. Nazeeruddin, M. Grätzel, J. Phys. Chem. Lett., 2014, 5, 501–505).
Ruthenium sensitizers TT-230 and TT-232 based on cyclopenta[2,1-b:3,4-b']dithiophene (CDT) which were tested in DSSCs using a conventional iodine-based liquid electrolyte containing the iodine/triiodine redox couple, and their performances compared to the benchmark Z907 were also prepared (M. Urbani, M. Medel, S. Amit Kumar, D. González-Rodríguez, M. Grätzel, Md. K. Nazeeruddin and T. Torres, Polyhedron 2014).

- Zinc-phthalocyanines
UAM prepared a variety of phthalocyanines with functional groups to enhance absorption in the visible region, and with bulky substituents to reduce aggregation for both DSSCs and HTM in perovskites-based solar cells.
UAM have also synthesized push-pull phthalocyanine systems (related to the “champion” porphyrin sensitizers that have shown efficiencies higher than 13%). Moreover, in collaboration with EPFL they have synthesized and characterized novel dyes capable of absorbing in the red region and beyond, and exhibiting particular photophysical properties.
One breakthrough in the phthalocyanine-sensitized DSSC area was the preparation of the dye TT1 (J.-J. Cid, J.-H. Yum, S.-R. Jang, M. K. Nazeeruddin, E. Martínez-Ferrero, E. Palomares, J. Ko, M. Grätzel, T. Torres, Angew. Chem. Int. Ed., 2007, 46, 8358; M. E. Ragoussi, J. J. Cid, J. H. Yum, G. de la Torre, D. Di Censo, M. Grätzel, M. K. Nazeeruddin, T. Torres, Angew. Chem. Int. Ed. 2012, 51, 4375), which hold three tert-butyl groups in the periphery of the macrocycle in order to suppress the stacking of the dye on the TiO2 surface.
In this case, it was demonstrated that the suppression of aggregation was successful in a way that no addition of co-adsorbent chenodeoxycholic acid (CDCA) was necessary.
Moreover, new Pc TT58 along with its use as a sensitizer in DSSCs, in an attempt to shed light on the actual influence of the steric congestion around the anchoring group. Moreover, UAM and EPFL have revisited the preparation of TT40 based devices in order to optimize their efficiency (J.-H. Yum, A. Kumar Chandiran, M. Ince, G. de la Torre, M. Grätzel, M. K. Nazeeruddin and T. Torres, ChemPhysChem 2014, 15, 1033 – 1036). The cells prepared with this dye yielded a PCE of 5.5% under 100 mWcm-2 (1 sun irradiation) and 6.1% under 9.5 mWcm-2.
The photovoltaic performance of SMSCs based on Pcs dyes have been greatly improved to date, but the spectral tuning of the Pc is limited in some extent because of chromophore absorption. Therefore, UAM have also synthesized a series of novel near IR absorbing zinc phthalocyanines bearing donor-chromophore-acceptor/anchoring groups and have investigated their influence on the solar cell performance (Molecular Engineering of Phthalocyanine Sensitizers for Dye-Sensitized Solar Cells”, M. Ince, J.-H. Yum, Y. Kim, S. Mathew, M.l Grätzel, T. Torres, and Md. K. Nazeeruddin, J. Phys. Chem. C 2014, 118, 17166−17170).
UAM also prepared a series of zinc-phthalocyanines bearing bulky aryl groups at non-peripheral positions for improving the device efficiency and stability (L. Tejerina, M. V. Martinez, T. Torres, M. K. Nazeeruddin, Chem. Eur. J. 2016, 22, 4369-4373). A new design of non-aggregated zinc(II) carboxyphthalocyanines for dye-sensitized solar cells have been proposed. UAM have also synthesized a series of zinc-phthalocyanines bearing bulky aryl groups at the periphery for completing the series described before (Lara Tejerina, M. V. Martínez-Díaz, M.K. Nazeeruddin, T. Torres, J. Porphy. Phthalocy. 2016, in press). Thus, for example carboxyphthalocyanine TT65 has an extended conjugation and a highly lipophilic periphery. The synthesis has been carried out following an efficient convergent route in which the bulky aryl groups were introduced by multiple Suzuki-Miyaura cross-coupling reaction on a preformed tri-iodophthalocyanine Pc2 - prepared from easily accesible Pc1 - to give Pc3, which was further oxidized to Pc4 (TT65, Fig. 2).

With the advent of organohalide perovskite structures as efficient, solution processable light harvesters, major shifts towards the further development, understanding and controlling of such materials have been undertaken. Although research on novel sensitizers hasn’t been fully halted, more focus has been given to the development of high-efficiency CH3NH3PbI3 solar cells that were eventually used as the top cell in the presented tandem.

b) Colloidal nanocrystals for UV/Vis/NIR absorption

TUD group activity has been focused on establishing the synthetic protocols to provide the materials with the specific properties required for the implementation of the project as well as the improvement of the colloidal synthesis of quite well known IV-VI semiconductor materials, such as PbS and PbSe nanocrystals (NCs). These materials have been designed to act as IR absorbers in PV devices instead of the organic dyes currently used in the present generation of dye sensitized solar cells. They have been demonstrated to show high efficiencies and possible multi-exciton generation (MEG). QD materials had been synthesized in nonpolar media and their surface ligands had been exchanged to facilitate incorporation into the PV devices and to achieve better charge transfer from the QDs to the conducting oxide layer.
The most promising materials to achieve the objectives of GLOBASOL project are nanocrystals based on the IV-VI class of semiconductor materials, especially those of PbS and PbSe. Therefore the main efforts of the hot-injection synthesis were concentrated on the improvement of the material preparation and the tailoring of the optical properties to suit the designs of the project.
PbS and PbSe quantum dots of different sizes were synthesized using a modified hot injection method employing oleic acid (OlA) and trioctylphosphine (TOP) to control both the growth and resulting sizes of the particles. In Fig. 3 exemplary absorption spectra of the resulting quantum dots with different sizes, including very small ones, recorded in tetrachlorethylene (TCE) are presented. The very small PbS and PbSe QDs have the highest available conduction band positions and therefore should be energetically favourable for the electron transfer from the nanoparticle conduction band to the conducting oxide layer. The very small PbS NCs were provided to the EPFL partner for their evaluation as absorbers and testing in solar cell devices.

The as-synthesized nanoparticles are coated by a layer of organic ligands (OlA) and are therefore soluble in non-polar solvents. The particles, once dissolved are very stable with respect to their solubility and no aggregation has been observed.
The synthesis parameters for the up-scaled synthesis were adjusted to achieve a higher amount of QDs with the same optical properties.

Lead chalcogenide nanostructures are sensitive to oxygen, and this is a huge disadvantage for the further use in the device. Therefore the synthesis of core-shell structures is a good possibility to increase the stability of the QDs. Especially the exposure of PbSe QD solutions to air under ambient conditions leads to rapid oxidation of the QDs such that up to 50 % of their volume is transformed into PbO, SeO2 or PbSeO3 within 24 h. This leads to a significant blue shift of the absorption features. To increase the stability of the PbSe QDs core-shell structures had been synthesized.

For the synthesis of the core-shell structures the successive ion layer absorption and reaction (SILAR) technique had been successfully used.
Moreover, bismuth sulfide nanoparticles were synthesized (Figure 4). The nanoparticles have irregular elongated shape having a mean width of 15 nm and length of 110 nm.

To facilitate the incorporation of the QDs into PV devices, as well as to improve their charge transfer to the conducting oxide layer, surface modifications of the QDs have been performed. 1,6-Hexanedithiol (HDT) has been utilized as a possible short chained dithiol ligand for surface modification. This offered the possibility to vary the functional group bound to the oxide layer as well as introduce a shorter distance between the QDs and the oxide layer compared to the long chained initial oleic acid ligand.

As stated before and largely indicated in the recent literature, organometal halide perovskites have been shown as an excellent visible-light absorbing materials useful for photovoltaic applications. Therefore QDs should be soluble in polar media like the perovskite solutions. An increased coupling in between the QDs and the perovskite to achieve a very high electron transfer rate is assumed if there is a direct contact in between the perovskite and the QDs. Therefore first investigations of the surface modification of PbS QDs with methylammonium lead iodide had been performed.
The success of the surface modification is revealed by FTIR spectroscopy, TEM investigations as well as by the visible phase transfer from the nonpolar octane phase to the polar DMF phase.
The activities in this task have been focused on establishing the synthetic protocols, synthesis upscale and the tuning of desired properties of materials with the specific properties required for the implementation of the project. After interim 18 M report, the perovskites were chosen for the development in the framework of this project, as potentially the most promising IR light absorber/sensitizer. The experience of TUD partner in the synthesis of colloidal quantum dots was extended to synthesize colloidal perovskite nanoparticles. Perovskite nanoparticles had been designed to facilitate their incorporation into the PV devices and to achieve better charge transfer to the conducting oxide layer.
Figure 5 summarizes the results of the band gap tuning of CsPbX3 perovskites determined by the halide anion (i.e. ratio between Br and I, used for the synthesis).

By adjusting the initial composition of the lead halide precursors, the emission and absorption spectra are tuneable over the entire visible spectral region between 410-700 nm.
By tuning the injection temperature (e.g. CsPbBr3 at 150 °C and 200 °C) the band gap can be also tuned because of quantum size-effects. Moreover, we introduced cleaning procedure consisting of two centrifugation steps for the removal of smaller and bigger particles and by-products which allows the stabilization of CsPbBr3 for several months.

TUD is able to synthesize high quality perovskite nanoparticles with different compositions and by this with tuneable optical and electrical properties. We introduced the gentle cleaning procedure, which allows stabilizing the perovskite nanoparticles for months. However, achieving stability of the iodide-based nanoparticles is still a challenging task, which should be addressed in further research, as these nanoparticles are most promising for the absorption in the near-IR region.

- Quasi solid electrolytes for p-type conductors in SMSC
This task deals with the development of quasi-solid electrolyte for SMSC devices.
Novel quasi-solid electrolytes are produced aiming at eliminating the liquid solvents currently used in SMSC cells as hole transporting media. Following the results obtained by UNIPMN researchers (L. Etgar et al., J. Mater. Chem. A, 2013, 1, 10142–10147), silica-based materials with different structural properties (passing from amorphous silica to mesoporous or layered solids) were prepared, fully characterised and used as additives for the preparation of quasi-solid electrolytes. Materials were also properly modified by introducing –NH2, ¬¬–COOH, –NH–CH2–CH2–NH, –SH and phenyl groups in order to optimize processes at the interfaces. It has been pointed out indeed that the presence of these groups has a positive effect on electrochemical parameters of SMSC cells.

Two different approaches were used in the frame of the project for the preparation of quasi-solid electrolytes. From one side, silica-based materials with different structure (amorphous silica, monodispersed silica particles and ordered mesostructured silicas, Fig. 6) were used as host materials for I-/I3- redox species for the preparation of quasi-solid electrolytes. The siliceous materials were used to minimize the need of volatile solvent (i.e. acetonitrile) and to limit possible evaporation and associated safety risks.

Particular attention was given to the modification of particle size of prepared solids, aiming at obtaining good suspensions in liquid electrolytes and to optimize the interaction with TiO2 semiconductor. From the other side, the surface of silica solids were properly modified with amino groups, because of their positive effect on the final performances of the SMSC devices (L. Etgar et al., J. Mater. Chem. A, 2013, 1, 10142–10147).

Besides silicas, special interest was also devoted to the preparation of synthetic clays (i.e. saponite, Fig. 7) with controlled morphology and chemical compositions to be used as additives for DSSC purposes.
Cationic saponite clays, especially when in the form of nanosized particles, appeared particularly effective for improving the efficiency of liquid electrolytes. In this last period, the activity was focussed to the optimisation of a synthesis method to reduce as much as possible the particle dimensions of inorganic saponites samples.
As reported previously, inorganic saponite clays were prepared by hydrothermal synthesis by properly adapting the procedures reported in the literature (Kloprogge, J. T.; Breukelaar, J.; Jansen, J. B. H.; Geus, J. W. Clays Clay Miner. 1993, 41, 103). It was pointed out that the H2O/Si ratio used for the preparation of the synthesis gel influences the morphology and the aspect ratio of the produced materials passing from 200 nm particle size for solids prepared by using H2O/Si ratio of 20 to ca. 50 nm for saponite obtained by using H2O/Si ratio of 110. Aiming to verify if it is possible to prepare a nanosized saponite by further increasing the synthesis gel dilution, the synthesis procedure was conducted by using a H2O/Si ratio of 150.

The adopted hydrothermal synthesis procedure affords in general high yield of layered product, with essentially low amounts of amorphous by-products.
The dilution of the gel mixture has a clear effect on the average size of the lamellae, being larger than 200 nm for SAP20 and less than 50 nm for SAP150. The latter sample is formed of very thin particles.

As an alternative redox mediator to the conventional (Co(II)/Co(III)) and (I-/I3-), copper complexes (Cu(I)/Cu(II)) have been studied both in liquid state (Yu Bai, 2011) (Shigeki Hattori, 2005) (Michele Brugnati, 2007), and solid state (Edwin C. Constable, 2009) DSCs. Hattori et al. obtained the maximum photon to current efficiency (PCE) as 1.4% with bis(2,9-dimethyl-1,10-phenantroline)copper(I)/(II) (Cu(dmp)2) which has a distorted tetragonal shape providing lower re-organization energy. In the frame of the Project, we show here that for the Cu(dmp)2 complex, the conductivity of the complex can actually be increased up to 50 fold, when measured by AC voltammetry inside a gold-lined conductivity cell.
Similarly, we found that a similar increase of diffusion coefficient appears for the classical Co(bpy)3II/III electrolyte, where an increase of the diffusion coefficient up to 10-fold could be witnessed These results highlight the potential of soggy sands electrolytes for use in dye-sensitized cells not just to improve the rheological properties of the electrolytes, but also to decrease their diffusion coefficient, increase their mass diffusion current and therefore mitigate diffusion losses that are commonly attributed to slow diffusion of the Co(II/III) electrolyte.

- Synthesis and characterisation of new TE materials
The key activities associated with this task include demonstration of: 1) the capability of producing thermoelectric nanoparticles, 2) the capability of fabricating nanostructured thermoelectric bulk materials, and 3) the potential improvement through appropriate design of nanostructures. In the initial phase, the efforts have focused on the production of TE nanoparticles following two approaches: 1) a bottom-up approach by TUD based on hot-injection method and 2) a top-down approach by CU based on milling method. Following on the successful effort of production of thermoelectric nanoparticles, the efforts focused on developing fabrication processes and suitable conditions for preparing nanostructured bulk materials. The density and thermoelectric properties of the prepared samples were characterised to evaluate the potential of the proposed strategy, approaches and materials.
The thermal conductivity of first samples of PbS nanowires has been characterized at Fraunhofer IPM.
The TUD group activities have focused on the development of synthetic strategies for the ligand-assisted preparation of novel nanosized thermoelectric (TE) materials as well as the development of assembly approaches and surface chemistry alterations for the improvement of the properties of the materials with respect to their end use. Therefore the synthesis of PbS was optimized to achieve sufficient amounts of PbS quantum dots. The ligands used in the synthesis of the QDs unfortunately act as an insulating layer resulting in QD solids with very high resistivity. For this reason different surface modification strategies were introduced to increase the conductivity of the resulting quantum dot solids. One very promising method was the exchange of the initial ligands by shorter ones that can be destroyed at lower temperatures. By such heat treatments the resistivity could be decreased by up to six orders of magnitude. For the pressing of the quantum dots two different compaction methods (SPS and hydraulic pressing) were compared. Samples that had been surface modified with MPA and subsequently thermally treated show the best results with respect to their thermopower and resistivities. Next to the PbS QDs, also a synthesis that enables the formation of PbS nanowires of different diameters and that is easy up-scalable was developed. The resulting nanowires were used as building blocks for film formation on glass substrates by an easily implemented method that requires no special equipment. Surface modifications of the films were performed to improve the charge transfer in the films and the Seebeck coefficients as well as the thermal conductivity of the resulting films were measured. Due to the fewer grain boundaries present in the films composed of nanowires as compared to the QD assemblies the conductivity is significantly higher.
MgAgSb has been identified as alternative for “BiTe” compounds in the project. So far, only p-type material has been published with maximum ZT values around 1.2 to 1.4. Within the project, the reproducibility of the published values have been checked. Therefore, several annealing and doping strategies have been tested.
MgAgSb compounds have been synthesized at Fraunhofer IPM by mechanical alloying. By careful adjustment of the Ni content, a significant improvement of the ZT value has been achieved. In order get a closer reproduction of the literature values, it is suggested to analyse the grain size distribution of the samples and, if applicable, to improve the nanostructuring of the samples.
At the beginning of month 19, TUD research groups was able to create different kinds of Bi2S3 nanoparticles, possessing shapes of dots, rods and more complex nanostructures. We also proved their high crystallinity. However, already at those early stages of research, we have found, that the creation of highly conducting layers of these materials was very challenging task. This motivate us to look for other promising material systems. Thus, the nanosheets from Bi2Se3 and Bi2Te3 were successfully synthesized, thorough characterized and utilized for the formation of thin layers and pellets with very attractive thermoelectric properties.
The so named polyol synthesis performed in ethyleneglycole (EG) in presence of polyvnylpyrrolidone (PVP) as capping ligand was found to be the most promising procedure resulting in Bi2Se3 and Bi2Te3 nanosheets (Figure 8).

The solids prepared from the sheets (both drop-casting and spray coating were utilized for this purposes) possess limited conductivity, as the coupling between individual nanoentities is limited due to the grain boundaries and the capping agent. Several methods were utilized to avoid this disadvantage of the new material, including capping agent exchange to sulphide ions, vacuum drying and annealing at various temperatures. Among them, the annealing was found to be the most efficient.

Thermoelectrical characterizations of the thin films samples were performed by Cardiff partner, the pellets were measured in Dresden (Dr. I. Veremchuk, MPI CPS). Excellent thermoelectric characteristics of the samples of Bi2Te3 and mixed Bi2Se3/Bi2Te3 have been obtained. The ZT value of the mixed sample approaches 1.4 at 418 K. To the best of our knowledge, such high values were not reported by other groups for colloidally prepared samples up to date.
Thermoelectric bulk samples based on Skutterudite compounds have been prepared at Fraunhofer IPM. They exhibit high figures of merit, i.e. ZT = 1.0 at 697 K for n-type and ZT = 0.9 at 690 K for p-type skutterudites.

- Synthesis and testing of photonic materials for light management

The identification and selection of material building blocks and architectures suitable for highly efficient optical light management covering a broad spectral range between 400 nm up to 1200 nm, have been finalized. Inks based on selected material building blocks and procedures for deposition and processing ink layers have been developed. One- dimensional photonic crystals (1DPC’s) optically active in the 700 nm - 750 nm range and the 1000 nm - 1200 nm range can now be manufactured.
Within the Project, nanoparticles of TiO2 and SiO2 have been chosen as building blocks constituting the alternating periodic one dimensional photonic crystal architecture. Alternating, nanoporous layers of nanoparticles of TiO2 and SiO2 will be used for the manufacturing of the light management system splitting the sun spectrum in separate components.
A crucial requisite for achieving the goals within Globasol is the ability to manufacture one dimensional photonic crystals with high specular reflectance in selected wavelength regions that suit the needs imposed by the partners in the Globasol project.
Regarding the specular reflectance in the targeted NIR wavelength region between 700 nm and 750 nm, EXEGER achieved promising results with reflectances well above 90% (where 100% reflectance is ideal). By optimizing the processing conditions further we expect to reach an reflectance above 97%. In the IR region between 1000 nm – 1200 nm further optimization work is needed. Currently we can achieve reflectance values above 70%. By tuning the process parameters further we expect to reach at least 90% reflection.

EXEGER achieved to manufacture large area photonic crystals by dip coating. The size is 10cm x 10cm which is four times larger than previously (5cm x 5cm). The new large photonic crystal (Fig. 9) exhibit very promising characteristics:
• Reflection properties can be controlled (resulting in different colours);
• Reflectance close to 90% can be achieved;
• Size can be made larger than 10cm x 10cm;
• Multilayers are deposited using a robot assisted dip coating system for maximum thickness precision.

By tuning the thicknesses of the deposited layers it is possible to change the colour of reflection, as can be seen in the figure. Additionally, very high reflectances are achieved. Consequently, reflection properties can be controlled to different wavelengths. Reflectance close to 90% can be achieved.
High reflectance values (over 97%) and high transmittance values (over 80%) can be achieved on small substrates using spin coating.
The identification and selection of material building blocks and architectures suitable for highly efficient optical light management covering a broad spectral range between 400 nm up to 1200 nm, have been finalized. Inks based on selected material building blocks and procedures for deposition and processing ink layers have been developed. One-dimensional photonic crystals (1DPC’s) optically active in the 700 nm - 750 nm range and the 1000 nm - 1200 nm range can now be manufactured.

1.1.3.2 Design, preparation and characterisation of innovative PV and TE cells (activity related to WP3)

- Integration of new materials in SMSC devices
This WP aims at the preparation of new devices exploiting the materials prepared during the Project. The activity of this WP started at month 6. Since in the last years, an every growing interest was paid in the literature to the possibility to use perovskite solids for the preparation of high performance solar device, major attention was focused on this type of technology also in the frame of the Globasol Project.
Spiro-MeOTAD based PSCs have been fabricated following the 1-step procedure reported in literature. A toluene anti-solvent process has been used to facilitate the crystallization formation and produce more uniform layers. The cell structure used is reported in the figure 10, where 3 cells per substrate can be obtained.

Good efficiencies approaching 16% with a good reproducibility have been obtained. Similar cell fabrication processes have been performed at EPFL where record efficiencies of PSCs were achieved (20.8%). This high efficient cells have been used in Globasol for the efficiency demonstration of the whole device. The second step was the concept and the fabrication of small PSCs based on graphite counter electrodes. After having tested spin coating and spray pyrolysis techniques, the process has been successfully converted into screen-printed layers, as reported in the following picture. The cell structure follows the same structure of the front-electrode which is under development for in-situ cells. The cells are screen-printed on a 10 x 10 cm2 glass plate, in order to have a high statistic, good reproducibility and a fast process. A maximum efficiency of 4.35% has been obtained. Though the cell structure need further optimization in the layer thickness, paste uniformity and perovskite infiltration into the porous structure, the values obtained for the electrical parameters give confirmation of the right strategy pursued for this architecture.
Analysis of the possible recycling of perovskite solar cells was performed at EPFL.
Perovskite based solar cells based on CH3NH3PbI3 and related materials have reached impressive efficiencies that, on a lab scale, can compete with established solar cell technologies, at least in short-term observations. Despite frequently voiced concerns about the solubility of the lead salts that make up the absorber material, several life cycle analyses have come to overall positive conclusions regarding environmental impact of perovskite solar cells (PSC) production. Their particularly short energy payback time (EBPT) in comparison to other established PV technologies makes them truly competitive. Several studies have identified valuable components such as FTO, gold and high temperature processes as the most significant contributors to the environmental impact of PSCs. Considering these findings, EPFL have developed a rapid dismantling process allowing recovery of all major components, saving both raw materials, energy and production time in the fabrication of recycled PSCs.
A detailed work was also done in collaboration between UNIPMN and FRAUNHOFER concerning the testing of quasi-solid electrolytes in SMSC devices prepared adding silicas with different structure and particle size to ionic liquid based electrolyte. All materials have been also functionalized by introducing 3-aminopropyltriethoxysilane (APTS) in order to increase the affinity with the liquid electrolyte. From photovoltaic measurements it was noticed that the introduction of solids in the electrolyte leads to an increase of Jsc with respect to the reference cell. Moreover, the introduction of samples functionalized with –NH2 species led to an increase of the Voc value and this is especially evident in the case of silica particles prepared by water in oil method (Voc passing from 0.653 in the case of reference cell to 0.670 V). Both of these parameters contribute to improve the efficiency of the cell, which increase up to 14% compared to the reference cell. The electrochemical behavior and the spatial photocurrent distribution were then carefully studied (C. Vittoni, V. Sacchetto, D. Costenaro, S. Mastroianni, A. Hinsch, L. Marchese and C. Bisio “Gelation of solvent-free electrolyte using siliceous materials with different size and porosity for applications in dye sensitized solar cells”, Solar Energy, vol. 124, pp. 101-113, 2016).

- Introduction of photonic materials in SMSC devices
The finalization of this task consisted primarily in a fine tuning the specific reflection and transmission wavelengths of the photonic crystals to match precisely the agreed reflection wavelength of 750 nm and to maximize the reflection and transmission of light at this specific wavelength. Care was also taken in preparing substrates of correct geometrical dimensions to seamlessly be integrated into the opto-electromechanical light management system at hand.
EXEGER have achieved to deposit fairly homogeneous one-dimensional crystals designed for the custom made mechanical holder in the light management system. The photonic crystal exhibited excellent transmission characteristics above 750 nm with transmission values greater than 80 % in the entire wavelength range from 750 nm up to above 2000 nm. On the other hand the reflection characteristics of the Globasol photonic crystal could not compete with the commrecially available photonic crystals and the reflection below 750 nm was around 50 % between 400 nm and 750 nm and peaking at 750 nm (65% reflectance) and 550 nm (around 70 % reflectance), Fig. 11.

- Design and preparation of interdigitated SMSC for applications in concentrated sunlight
During discussions with partners working on integration of TE and PV cells it has turned out that the concept of PV cells are serving as substrates for NIR mirrors will be followed further in the project with high priority.

The in-situ perovskite solar cells have been fabricated through lasering and screen printing following the CAD design in Figure 12.
Cells are printed on the same substrate and sealed through glass frit processes. The figure also shows an image taken by optical microscope where the good alignment between front and counter electrode can be observed.

The analysis of the perovskite layer formation, and the control of crystal growth during cells processing inside the in-situ concept device, was followed and investigated through the photoluminescence (PL) signal.
This still non-optimized situation is seen back in the current-voltage characteristic of the cell under 1 sun illumination where a photocurrent JSC of 5.44 mA/cm2 has been measured. On the other side, an already remarkable photovoltage of 821 mV has been reached. Both JSC and VOC are further affected by the modest fill factor (59%) which is caused in this case by low ohmic shunt resistance. The total active area is 0.98 cm2.

- Assembling of new materials in TE devices
This task includes the fabrication and characterization of thermoelectric modules fabricated employing the bismuth telluride materials prepared in the Project.
Several modules were prepared utilising the original ingot materials (used as reference) and the n- and p-type materials with the best performance (size range 250-106 µm for both n- and p-type materials). These modules were characterised by impedance spectroscopy (IS). This technique can be considered as an advanced Harman method and is able to determine the module ZT [García-Cañadas, J. and G. Min, Impedance spectroscopy models for the complete characterization of thermoelectric materials. Journal of Applied Physics, 2014. 116(17): p. 174510.]. The characteristic impedance response of a thermoelectric module is shown in Figure 13.

The fabrication of high quality thermoelectric modules requires optimisation of the soldering conditions, which involve the temperature and duration of the soldering process. The quality of soldering is monitored by determining the (ZT)module of the modules using impedance spectroscope. The thermoelectric junctions were soldered together using commercial Sn/Pb (60/40) solders. The use of the Sn42Bi58 paste was also explored.
Initially, a number of modules were fabricated using the reference materials (crystalline Bi2Te3 materials obtained from an international supplier) under different soldering conditions.

The best performing bismuth telluride materials developed in the Project were employed to fabricate modules using the above-identified “optimal” conditions.
The highest (ZT)module value obtained among this batch of modules is 0.35, which also has the lowest module resistance. This value is lower than that of the best module fabricated using crystalline materials (0.51). Since the hot-pressed materials possess higher (ZT)material values, it is expected that the (ZT)module of the modules using the hot-pressed materials should be higher. Unexpectedly, this is not the case, which may be related to larger module resistance. Further investigation is needed to identify the reasons behind it.
Thermoelectric modules were fabricated using commercial crystalline materials and the best performing hot-pressed materials prepared in the Project. The performance of these modules was evaluated by the module ZT values determined using thermoelectric impedance spectroscopy. The aim of this work is to demonstrate the improved performance of thermoelectric modules using the materials developed in this research project. However, the result is inconclusive due to unclear influences of the contact resistance in these modules. Further investigation is needed to focus on in-depth characterisation of the contact resistances.
1.1.3.3 Design and fabrication of integrated PV/TE cells for global spectrum harvesting (activity related to WP4)

The table below summarizes the planned and achieved milestones at M18.

The TE efficiency refers to the conversion efficiency of a skutterudite module in case, a temperature difference of 480K between hot side and cold side of the module is achieved.

For single components, the planned milestones were correctly fulfilled.
In the second part of the project, the optimal elements from the Globasol project were selected for the preparation of the hybrid device. The elements where then assembled and tested under standard solar light condition.
The Globasol partners capability of manufacturing high efficiency devices at certain specified sizes and geometries have been mapped. Based on the specific input from all Globasol partners a design for integrating all the Globasol devices have been created.
The integrated design includes Globasol produced i) perovskite solar cells prepared at EPFL, ii) crystalline silicon solar cell prepared at Fraunhofer ISE and iii) photonic crystals prepared at EXEGER.

The optical splitting system is composed by a sample holder which hosts the two mirrors and the PV devices (Fig. 14). The stack on top (1) defines the splitting system for the perovskite cell which absorbs photons in the region 300-800 nm, while the stack no the botton (2) is dedicated to the silicon cell absorbing the remaining spectrum up to 1200 nm. The real absorption spectrum of the two devices is then function of the properties of the two mirrors. The spectrum of the incident light transmitted by the two mirrors is then used for electricity generation by the thermoelectric device.

The two mirror have been characterised define the angle dependence (from 10° to 45°) and the change of its optical properties in the range 380-1950 nm. Figure 3.60 shows transmittance T and reflectance + transmittance R+T curves for the mirror taken into consideration (Edmund Optics), being this last parameter a quality factor of the mirror which reports the absorbance properties. The characterization was performed with s and p polarization of the incident light. The shift of the cutoff wavelength (WLcutoff)due to the angle dependence is from 750 to 795 nm for s polarization and from 720 to 795 nm for the p polarization, measured from 45° to 10°.
To be noted is that the optical splitting device hosts the mirror with a 45° tilted angle, meaning that the curves to be taken into account for the electrical parameters acquisition of the PV devices are an average between the s and p polarization at 45°. Here the WLcutoff is 735 nm, meaning that a large fraction of the incident light with WL<735 nm (R in this range is >98%) is reflected towards the perovskite cell and the rest is transmitted further.
The commercial mirror characterized in so far was compared with a photonic crystal mirror fabricated by Exeger. FAlthough the mirror does not show a clear cutoff wavelength, it can be noted that the absortion values are limited to < 2% Nevertheless the mirror shows a good transmittance (> 80%) in the region beyond 750 nm.
Fig. 15 shows the direct comparison of transmittance T and reflectance R curves for the two mirrors analysed so far measured with 8° tilt angle in the range 320 – 2400 nm.

A similar analysis performed for the mirror 1 is described in this following section for mirror 2, which is dedicated to the Silicon cell. The WLcutoff considering the average between the s and p polarization at the various tilted angles is in the range 1040 – 1125 nm, being 1040 nm at 45°. The absorption in the range of interest is < 2% .
The photo of the perovskite cell fabricated at EPFL (and used for the integrated device) and its electrical parameters are reported in Figs. 16 and 17. The active area given by the golden region in the centre of the cell is masked during the I-V measurement, giving a final active area of 0.09 cm2. The resulting photo-conversion efficiency is 19.22%.

The photo of the silicon solar cell (fabricated at Fraunhofer ISE), its electrical parameters and the external quantum efficiency are reported in the Figs. 18 and 19. The active area of the cell is 4.12 cm2 and the resulting photo-conversion efficiency is 20.9%.

With the aim of testing the highest reachable efficiency by using the PV devices fabricated by Fraunhofer ISE and EPFL, the optical splitting device with 1-mirror stack has been measured. The total efficiency under AM1.5G solar spectrum is 27.3% by using the commercial mirror. As a final efficiency, the PV devices give 27.3%.
The PSC sees a decrease of its efficiency equal to 10% with a small variation between the forward and reverse scans, while the Si cell looses 52% of its initial efficiency measured under standard test conditions.
Replacing the commercial mirror with the mirror fabricated with photonic crystals (from EXEGER), the resulting efficiency is 22.8% (13,73% from the Si cell and 9,08% from the PSC cell).

The optical splitting system with the 2-mirror stack was also tested and a total efficiency under AM1.5G simulated solar spectrum (1000 W/m2) is 24.88%. The PSC sees a decrease of its efficiency equal to 10%, while the Si cell looses 63% of its initial efficiency measured under standard test conditions.
As a final efficiency, the PV devices give 24.88% plus the contribution of the TE device equal to 0.12 - 0.22%, meaning η in the range 25 - 25.1%.

Optical splitting device with 2-mirror stack combined with the barrel setup
As final step, the indoor characterized optical splitting system has been installed in the outdoor facility available at Fraunhofer IPM and integrated with the thermoelectric setup developed by them (Fig. 20). The data have been described in the related sections and deliverables.

Despite the fact that results obtained in the first 18M were in line with what it was expected (the global efficiency of the final device, calculated on the basis of available single solar cells and TE devices was in the range 23.6-28%), the maximum final efficiency for the hybrid device is 25.1%.
This deviation was especially related to the TE part, because of the challenge of creating a high temp difference between cold and hot side.
It can be envisaged an increase of the efficiency in order to reach a value > 30% by combining the following contributions, that in the last month of the Project were achieved:

1. Use of higher efficient perovskite cell available at EPFL (from 19.22% to 20.8%);
2. Use of higher efficient silicon cell available at ISE (from 20.90% to 24.89%);
3. Higher EQE of the silicon cell in the photon wavelenghts reflected by the second mirror;

Potential Impact:
The research work carried out in the frame of the Globasol project can have a high potential impact on reseach fields related to photovoltaic and thermoelectric applications. Indeed, the specific attention that was paid in the first part of the Project allowed to reach interesting results concerning the implementation of components related to the single SMSC (i.e. light absorbers, quantum dots, perovskite materials, hole conductors) and TE devices. This will pave the way to the possibility to further improvements of materials related to these technologies, also because the research work was conducted considering not solely the optimization of synthesis and characterisation methods aimied to improve specific properties of studied materials, but also developing new and specific hybrid theorical procedures that are useful to model the properties of novel materials and interfaces. The co-presence of experimental and computational approaches was a key-element for the optimisation of novel nanostructured materials developed in the frame of the Globasol project, and the obtained results can be exploited in all field in which materials optimization is required.
As a matter of fact, another important aspect that was considered in the frame of the Project is the LCA analysis of produced materials and devices. It is worthnoting that the increasing demand for sustainable renewable energy sources to reduce the pollution and dependency on fossil energy resources is associated to the need of understanding the sustainability of novel energy sources.
The commercial-scale production of novel energy sources requires careful consideration of several issues that can be categorized as raw material production, technology, by-products, etc. The life cycle assessment (LCA) is a tool that can be used effectively in evaluating various renewable energy sources for their sustainability and can help policy makers choose the best energy source for specific purpose. The LCA analysis that was carried out in the frame of the Globasol project, that was started from the fundamental study of production of specific materials, can be an important point support the optimization of processes, thus highlitingh the applicative target of the single and hybrid devices as well as their reuse prescription. In this respect, a detailed analysis of possibility of recycle of different components of solid state perovskite-based solar cells (that are now collecting a growing interest in the scientific community) has been carried out. Perovskite based solar cells and related materials have reached impressive efficiencies that, on a lab scale, can compete with established solar cell technologies, at least in short-term observations.
Despite frequently voiced concerns about the solubility of the lead salts that make up the absorber material, several life cycle analyses have come to overall positive conclusions regarding environmental impact of perovskite solar cells (PSC) production. Their particularly short energy payback time (EBPT) in comparison to other established PV technologies makes them truly competitive. Several studies have identified valuable components such as FTO, gold and high temperature processes as the most significant contributors to the environmental impact of PSCs. In the frame of the Project, it was demonstrated that the performance of PSC fabricated from recycled substrates can compete with that of devices fabricated from raw materials and this will pave the way to the possibile development of new PV technologies.
In the second half of the Project, attention was paid to the benchmarking analysis of proof-of-concept of single PV and TE devices. This was especially important to develop specific competencies related to the knolewge of phenomena occurring in single devices, alsoaiming to develop sustainable cell concept for in-situ formation of active and contact layers with high potential impact of PV solutions.

Great efforts were paid to the selection of the most useful architectures for the hybrid PV/TE devices starting from Globasol materials. Different architecture were proposed, simulated and then tested.
Performance of hybrid devices has been tested in both outdoor condition and under Solar simulated light in laboratory and this will allow the estimation of Lifetime for the parts in the hybrid device. Additionally evaluation of cost is made with benchmarking with existing technology concentrated solar, which is available on the market.
From accelerated lifetime tests the PSC cell showed 250h of continuous illumination @1sun with 70% initial efficiency. The crystalline Si component will survive several outdoor studies exceeding 20 years in length.
For concentrated solar the cost is about 8000 USD/kW, the Globasol hybrid cost would be estimated to 12000 USD/kW. The higher price is due to the early stage for this type of solar technology, upscaling and further optimisation will bring the costs down further.

Outdoor measurements of the barrel-shaped hybrid device have been performed at Fraunhofer IPM (Freiburg/Germany 48°N, 7°E). The setup has been tested without tracking of the sun, with manual solar tracking as well as with automatic solar tracking via a turning mechanism.
The best performance has been achieved with the automated turning mechanism. Results for clear sky conditions are shown in Figure 21 and they will be important to experiment these type of technology. The temperature fluctuations are flattened out in comparison to the temperature results in the case of manually turning the concentrator. The temperature at the heat pipe (black line) gradually rises till it reaches a value of 130°C at around 13:00. The rate of change of the temperature is very low till it reaches a peak value of 140°C at 16:00. On the other hand, the cooler temperature and the temperature at the upper heat flux meter increase gradually. This results from the increase of the water temperature inside the cooler as a consequence of continuous solar radiation on the cooler surface which leads to an increase of the cooler temperature and the related upper heat flux meter temperature. Although the hot side temperatures of the TEGs increase, the resulting output power (diagram in the middle) remains almost constant at 500 mW and 300 mW for the lower and upper TEG respectively. The reason behind this behavior is that the cold side temperature of both TEGs increases simultaneously along with the hot side temperature.

In order to validate the accuracy of the efficiency measurement done during the outdoor measurements and accordingly assess the quality of thermal coupling between the TEG, the heat flux meters and the heat pipe, it is necessary to compare the efficiency results obtained from the outdoor measurements (with solar tracker) to those of the laboratory measurements under controlled environment.
Figure 22 shows the efficiency values as a function of the temperature difference at the upper and lower heat flux meter for two cases : The laboratory measurements (black and red triangles) and the outdoor measurements (green and blue square).
Figure 22 shows a good agreement between the outdoor and the laboratory results obtained from the upper and lower heat flux meter (at a temperature difference of 100 K, the difference between the upper and lower efficiency is around 0.4% for both cases). Having this result means that the thermal coupling is successful and that the real-time efficiency measurement is accurate.
In order to take into account that the TEG will only exploit a part of the solar spectrum, the power density of incoming irradiation has been reduced to 50% by applying a half transmissive foil above the barrel setup. The efficiency related to 50% of the solar irradiation drops to around 0.4% The solar to electrical efficiency can prospectively be improved by improving the thermal coupling between vacuum tube, heat pipe and heat flux mount. Without the heat flux mount, i.e. without thermal load, temperatures inside the vacuum tube can reach more than 250°C for a solar irradiation of > 800 W/m2 whereas the measured hot side temperature of the thermoelectric module reaches only 140°C at ~900 W/m2. The hot side temperature could be increased by enhancing the heat transport from the vacuum tube to the converter head and by analyzing the heat losses in detail.
During the three-years of activity of the Project, different dissemination activities have been carried out. The table below summarize the efforts that have been done in this respect during the Globasol Project.

N° Type of dissemination activity
43 Papers on International Journals
20 Oral Presentation to a wide public
44 Oral Presentation to a scientific event
18 Posters
12 Organisation of conferences and workshop
17 Thesis/ dissertation
1 Website/ applications

List of Websites:
The address of project website is www.globasol.eu

Related information

Contact

Luciano Pugliese, (Department Manager)
Tel.: +390131360311
Fax: +390131390
E-mail
Record Number: 184983 / Last updated on: 2016-06-23