Morphology of the active layer of an OPV device
Laura Ciammaruchi
(laura.ciammaruchi@rochester.edu)
Opt 407 SEM Practicum - Spring 2012
Introduction
Fabrication & Characterization Techniques
Conclusions & Acknowledgements
1.1 The OPV Technology
Organic photovoltaic (OPV) cells are mostly based on a bulk hetero-junction (BHJ) structure, i.e. an active layer of mixed electron donor and acceptor. The acceptor component is usually a fullerene-based material such as C60 or PCBM, while a variety of materials can be useful as the donor component. It has been proven that it is possible to achieve a large open-circuit voltage with almost any donor, provided that it is present in a small concentration.
1.2. OPV fabrication
The cell fabrication is carried out through “vacuum thermal evaporation” in a vacuum chamber, where the entire device deposition sequence is completed at a pressure of the order of 10-7 Torr, without any vacuum break.
In order to guarantee both an efficient exciton diffusion and an effective charge collection, a bulk hetero-junction with percolated donor-acceptor phase is preferred. The width of the percolated donor-acceptor network phases should be on the order of tens of nanometers so that all the excitons, regardless of their generation location, can be collected. This structure also allows efficient charge transport due to the existence of pure donor and acceptor phases. However, creation of such an ideal nanostructured bulk hetero-junction is challenging.
2. Fabrication and Characterization techniques
2.1. Sample Preparation
A layer of C60:TAPC, as thin as about 150 um was thermally deposited by co-evaporation on top of an ITO coated glass. The respective deposition rates were 1.9 Å /s and 0.1 Å /s. To be as close as possible to the real BHJ structure, the relative concentrations were kept at 95% for C60 and 5% for TAPC. This turned out in a quasi-total presence of C60 in detriment to the donor material.
The as prepared samples were sputter-coated with a thin layer of gold (about 60 Å) to improve their conductivity for the electron microscopy. This process allows to reduce charging on the sample surface and to improve signal and quality of the images. Sample sizing consisted in cutting a cross-section of the layer of the organic material, and finally making a simple scratch on the surface with a regular blade.
On the other hand, atomic force microscopy was performed on the pristine layer top surface without any additional sample coating.
2.2 Electron Microscopy and AFM
Due to its thickness, the film was mainly inspected under SEM (Scanning Electron Microscope) to get the information of its morphology and cross-section.
When imaging the sample through SEM, three detectors were tried: InLens detector, SE2 detector and BSD detector. However, since the structure mainly consisted of Carbon Nitrogen and Oxygen, whose atomic weights are quite similar, the BSD images were not as worthy as the ones taken with the InLens and SE2 detectors.
Also, the sample turned out to build up a charge very quickly despite the coating precaution, so some images taken at higher resolution resulted to be slightly blurry.
Some of the micrographs have been colorized for a more enjoyable visualization.
We notice how the higher the magnification is, the clearer the smooth-edged structures appear, which are almost hidden when imaging the layer at lower scales. These “bubbles” not only are present on the surface, but also in the bulk of the layer itself. Those bubbles could be the very way C60 grows inside the coater during the deposition process, or better moisture inclusions due to the exposition of the sample to air.
The AFM analysis showed that the average height of the bubbles can be estimated in the range of a few nm.
The micrograph on the right shows an attempt to visualize a highly magnified surface (245Kx c.ca). The result is somehow blurred due to the extreme quickness the sample built up charge.
2.3. Electron Flight Simulation and Elemental Map Collection
Electron flight simulation technique was used to simulate the depth of penetration of the beam and the interaction volume at different accelerating voltages: 3, 5 and 20KV. For each simulated voltage, also the actual elemental spectrum through the EDAX X-ray detector was acquired.
Low accelerating voltages – 3 and 5 KV – are able to excite electrons coming from outer layers of the sample, as the actual elemental spectrum also displays (Carbon, Oxygen and Gold peaks are mostly present). At high accelerating voltages also elements relative to the glass itself and the coating come up – like Si, Ca, In - in detriment of the outer ones that are easily reabsorbed.
3. Conclusions and Acknowledgements
A 150 um thick layer of organic material was deposited on top of an ITO coated glass.
What appeared as a smooth and plane surface at relatively low resolutions, started to show features like “bubbles”, or bulges, at higher ones. The origin of these bulges is unclear and other morphological analyses on both a pristine and a aged layer should be done in order to investigate better its nature (by now I am just happy to have discovered them!)
My deep thanks go to Brian McIntyre and Sergey Korjenevski, as well as to Ihab Mardini, Guy Mongelli and all the thoughtful people who helped me out in all sort of needs while I was discovering Microscopy.
- M. Zhang et al., Bulk hetero-junction photovoltaic cells with low donor concentration, Advanced Materials (2011), 23(42), 4960-4964
- Y. Zheng, Nanostructured thin films for organic photovoltaic cells and organic light emitting diodes
- SEM Practicum course notes, Brian L. McIntyre