Doping-induced screening of the built-in-field in organic solar cells: effect on charge transport and recombination

Ardalan Armin, Gytis Juska, Bronson W. Philippa, Paul L. Burn, Paul Meredith, Ronald D. White, and Almantas Pivrikas. Advanced Energy Materials, doi:10.1002/aenm.201200581 (2012).

Non-technical summary

A perfect semiconductor would not contain any charge carriers unless energy was applied. For example, a solar cell would have no freely moving electrons (negative charges) or holes (positive charges) until sunlight creates them. Conversely, a semiconductor is “doped” if there exist charge carriers despite the absence of any excitation. These equilibrium carriers are created by impurities or defects called dopants. Dopants are sometimes introduced deliberately to create certain electrical properties, but other times dopants are unwanted and can have a negative influence on the device.

Doping is unwanted in organic solar cells because it can interfere with the photogenerated charge carriers. For example, charges created by dopants will disturb the electric field inside the device, hindering the extraction of the carriers created by the sunlight. Doping is undesirable, but it is sometimes unavoidable due to impurities in the materials used and/or the introduction of reactive species such as oxygen. The more pure a material must be, the more expensive it is to make, so it’s important to characterise how pure is “pure enough.”

To answer the question of whether a material is sufficiently pure, we need to examine the impact of doping on the performance of actual devices. We need to compare devices with different levels of doping, and see how the performance varies. Unfortunately, this is exceedingly difficult to do in a reliable way. Different devices may not be directly comparable to each other, because the raw materials and the processing conditions may not always be identical. This is especially true for devices made by hand in the lab. Consequently, one could never be sure that the only thing that changed between any two devices was the level of doping. (You could work around this limitation with statistics if you made many devices, but this is time consuming and expensive.)

What we really need to do this doping-vs-performance experiment reliably is a way to adjust the level of doping in a single device. If we could do that, then all the reproducibility problems would disappear. This paper demonstrates how this can be achieved.

In any solar cell (or diode), there is a natural threshold scale for charge concentration at which the physics changes. This threshold is the capacitance of the diode structure (C) multiplied by the effective voltage (U). The result of this multiplication, CU, has units of charge. If the amount of charge present in the device is less than CU, the electric field is largely undisturbed; conversely, if the amount of charge exceeds CU then the electric field becomes “screened” by the high amount of charge that is present. Importantly, CU is under our control because we can adjust the voltage U.

Suppose that the amount of charge created by doping is Q. If Q is less than CU then the amount of doping is small, and will not interfere. On the other hand, if Q is larger than CU, then the device is highly doped, and this might cause problems. In this paper, we show that we can dynamically adjust the voltage U to switch between low and high levels of doping.

Having developed a methodology to “switch on” the dopants, we are now in a position to characterise their impact on a solar cell. To do this we measure two key performance metrics at different levels of doping. These are the charge carrier mobility (how quickly charges move through the device), and the bimolecular recombination coefficient (how quickly charges annihilate in the presence of the opposite type). The conclusion: for this particular blend of materials, the mobility and recombination coefficient do not change as the device is moved from a low to a high level of doping. Consequently, for this material, the level of doping that is already present is not a problem, and there would be negligible benefit to be gained from using more pure materials.