Three of CHEOPS’ partners have tackled the task to optimize the material, processes and encapsulation to achieve the desired high-efficiency 5x5 cm² perovskite modules: CHOSE (Università degli Studi di Roma 'Tor Vergata'), CSEM (Centre Suisse d’Électronique et de Microtechnique) and UOXF (University of Oxford). They have all succeeded to produce such modules with different architectures (mesoscopic and planar), with different techniques (e.g. spin and blade coating) and different materials. The table below shows their interim results.
In the coming year, CHEOPS will select the most promising formulations and techniques to further upscale the modules. The goal is to present a 10x10 cm² demonstrator in January 2018 (other project milestones are documented in our online timeline).
In the following sections, we describe advantages and disadvantages of different architectures and the optimized material and manufacturing flow to achieve these results.
Device architectures: mesoscopic vs planar architecture
The 5x5 cm² perovskite modules have been produced with both mesoscopic and planar architectures. The mesoscopic architecture comprises a mesoporous TiO₂ scaffold that facilitates perovskite growth and the collection of the photo-generated electrons. In the planar architecture, the perovskite layer is directly deposited on the charge selective layer without using the scaffold layer. While still giving a higher efficiency, the mesoscopic architecture has a higher thermal budget due to the high temperature sintering step required by the mesoporous TiO₂ scaffold layer. This high temperature step is not compatible with applications in monolithic tandem cells in combination with silicon heterojunctions. Therefore the planar architecture is investigated as well.
Optimization of perovskite material and charge transport layers
In order to realize a highly efficient 5x5 cm² perovskite module, the main challenge was to optimize the active perovskite material and the charge transport layers (ETL (electron transport layer) and HTL (hole transport layer)), realizing reproducible results in terms of uniformity, thickness and morphology. The ETL was chosen to be either TiO₂ or SnO₂. TiO₂ was deposited using easily upscalable processes, namely spray pyrolysis and magnetron sputtering. These processes produce dense, pinhole-free layers over large area, sputtering having the advantage of enabling thinner layers. SnO₂ is deposited by spin coating. The perovskite active layer is typically deposited by spin coating. Using an anti-solvent quenching, it is possible to promote the crystallization of the films, leading to smooth, uniform layers. Finally, the HTM (hole transport material) layer (spiro-OMeTAD) is deposited either by spin coating or blade coating, which has the potential for coating larger areas.
Optimized manufacturing flow
The manufacturing flow has been optimized in order to realize the up-scaling process for each architecture.
The manufacturing flow was investigated varying the n-i-p device architecture (planar and mesoporous). In the following, we describe the manufacturing flow dividing it in several sequential tasks that refer to the constituent layer forming the device.
TCO (transparent conducting oxide): In this task, we included all the processes related to the glass/FTO substrate such as the FTO etching realized by infrared laser, the deposition of the marker for alignment of the screenprinter and finally the cleaning procedure of the etched substrates.
ETL (electron transport layer): The ETL task leads to the patterned deposition of the BL(= blocking layer)-TiO₂ layer realized by spray pyrolysis deposition. The lift-off procedure permits the mask removal leaving a patterned deposition of the BL-TiO₂ that avoids the presence of the BL-TiO₂ on the interconnection areas of the module.
mp-TiO₂: The patterned BL-TiO₂ samples are used for the deposition of the mesoporous TiO₂ layer with a thickness of about 150nm after its sintering procedure at high temperature. The mp-TiO₂ deposition can be realized using spin-coating or blade-coating techniques.
PK (perovskite): The PK task involves the perovskite deposition on the ETL-based substrates realized in air without controlling the environment during the deposition. The PK layers can be varied with respect the deposition techniques including spin coating and blade coating and the realization procedure such as one-step and two-step deposition
HTL (hole transport layer): In this task, we included the realization procedure of the Spiro-OMeTAD formulation and its deposition. Spiro-OMeTAD is a small molecule generally doped with TBP and Li-TFSI in order to increase its conductivity. In order to scale up the process without using the spin-coating technique, we optimized the Spiro-OMeTAD by using blade coating technique in air. The blade coating process is preferred to reduce the material waste respect to the spin-coating process.
P2: The P2 task is the main issue related to the manufacturing flow. In fact, it is mandatory to remove the photo-active layers from the interconnection areas between the series connected cells. For this action, we optimized a laser assisted ablation of the entire stack. The laser parameters used for the P2 ablation depend on the device stack deposited on the interconnection area of the PK modules. For the mesoscopic architecture the P2 process is optimized to ablate the mp-TiO₂/PK/Spiro-OMeTAD stack.