Abstract
© 2014 AIP Publishing LLC. Colloidal quantum dots (CQDs) are semiconductor nanocrystals synthesized with, processed in, and deposited from the solution phase, potentially enabling low-cost, facile manufacture of solar cells. Unfortunately, CQD solar cell reports, until now, have only explored batch-processing methods - such as spin-coating - that offer limited capacity for scaling. Spray-coating could offer a means of producing uniform colloidal quantum dot films that yield high-quality devices. Here, we explore the versatility of the spray-coating method by producing CQD solar cells in a variety of previously unexplored substrate arrangements. The potential transferability of the spray-coating method to a roll-to-roll manufacturing process was tested by spray-coating the CQD active layer onto six substrates mounted on a rapidly rotating drum, yielding devices with an average power conversion efficiency of 6.7%. We further tested the manufacturability of the process by endeavoring to spray onto flexible substrates, only to find that spraying while the substrate was flexed was crucial to achieving champion performance of 7.2% without compromise to open-circuit voltage. Having deposited onto a substrate with one axis of curvature, we then built our CQD solar cells onto a spherical lens substrate having two axes of curvature resulting in a 5% efficient device. These results show that CQDs deposited using our spraying method can be integrated to large-area manufacturing processes and can be used to make solar cells on unconventional shapes.
Original language | English (US) |
---|---|
Pages (from-to) | 163902 |
Journal | Applied Physics Letters |
Volume | 105 |
Issue number | 16 |
DOIs | |
State | Published - Oct 20 2014 |
Externally published | Yes |
Bibliographical note
KAUST Repository Item: Exported on 2020-10-01Acknowledged KAUST grant number(s): KUS-11-009-21
Acknowledgements: The authors would like to thank Ikeuchi USA for useful discussions on the appropriate nozzles to use for our materials and solvents. This research was supported in part by IBM Canada Research and Development Center. This publication is based in part on work supported by Award KUS-11-009-21, made by King Abdullah University of Science and Technology (KAUST).
This publication acknowledges KAUST support, but has no KAUST affiliated authors.