What kind of solar cells are suitable for thermoelectric hybridization?

One might ask if all kind of solar cells are suitable for thermoelectric hybridization. If we consider only the case of thermal coupling (namely when the solar cell is placed in thermal contact with the thermoelectric part, which is the case taken into account in this project) the answer is “absolutely no.” The reason is the temperature sensitivity of solar cells versus temperature. Here I show you why.

Actually most solar cells lose efficiency when heated up. Namely their efficiency decrease linearly (at least for most single-junction solar cells) versus temperature (see here an explanation about the physics of this behavior). Thus it is possible to define the solar cell efficiency (η_pv) as

η_pv = η_pv⁰ [1 + β(T_h - T_a)]

where η_pv⁰ is the solar cell efficiency at room temperature, β is the slope of the linear decrease of the solar cell efficiency (normally is negative), and T_h the solar cell actual temperature.

The efficiency of the thermoelectric part has instead the opposite trend (namely increasing with temperature), and depends on the materials characteristics through the so-called figure of merit (take a look to the following links for simple introduction to thermoelectrics, and their efficiency).

Since essentially the hybrid efficiency it is the sum of the efficiencies of the two parts it turns out that is necessary to find a trade-off between the behavior of the two parts (as shown in the figure).

In the figure, η_htepvg , η_pv, and η_teg are respectively the hybrid, the photovoltaic, and the thermoelectric efficiency. The last important parameter is the opto-thermal efficiency (η_ot) which is the efficiency in which the systems converts the solar power into heat actually flowing through the thermoelectric part. η_ot deals with the capability of the system to avoid heat losses, and is related to the encapsulation strategy.

What is important to understand here is that the temperature sensitivity of the solar cell plays a crucial role on the final efficiency of the hybrid device. However not all the solar cells behaves in the same way. In particular every kind of material defines the slope of the solar cell temperature sensitivity β.

Analyzing the physics of photon absorption, it is possible to understand the higher the energy gap, the smaller the temperature sensitivity. This is made clear by the following graph, where the value of β are reported as a function of the energy gap, and the material quality.

Thus, it turns out that materials with small energy gap (such as Silicon, CIGS, and GaAs) are not suitable for thermoelectric hybridization. Actually it can be found that for this materials the energy gains due to the addition of the thermoelectric part are negative. Other materials with wider energy gaps (such as GaInP, amorphous silicon, and perovskites) can lead to positive energy gains due to thermoelectric hybridization.

In the following figure a summary of this finding is shown. In particular most of the available photovoltaic materials are reported as a function of their efficiency and theri energy gap. A line is drawn at approximately 1.5 eV, where the switch from negative to positive efficiency gains was found to happen.

More details about this evidence are reported in one of my paper titled "Conditions for beneficial coupling of thermoelectric and photovoltaic devices".