A Taxonomy for Three-Terminal Tandem Solar Cells

Abstract

Tandem and multijunction solar cells offer the only demonstrated path to terrestrial 1-sun solar cell efficiency over 30%. Three-terminal tandem (3TT) solar cells can overcome some of the limitations of two-terminal and four-terminal tandem solar cell designs. However, the coupled nature of the cells adds a degree of complexity to the devices themselves and the ways that their performance can be measured and reported. While many different configurations of 3TT devices have been proposed, there is no standard taxonomy to discuss the device structure or loading topology. This Perspective proposes a taxonomy for 3TT solar cells to enable a common nomenclature for discussing these devices and their performance. It also provides a brief history of three-terminal devices in the literature and demonstrates that many different 3TT devices can work at efficiencies above 30% if properly designed.

Figure 1

Mapping the wide variety of three-terminal tandem configurations. In all schematics, n-type materials are red and p-type materials are blue. “Top” is used as a representative top cell, and in a real device would be replaced by the name of the material, e.g. “perovskite” or “GaInP”. The naming terminology above each schematic is explained in detail in the main text. The purple letters (T, F, R, and Z) correspond to the names of the nodes used for different loading configurations.

Figure 2

Node naming and loading conventions for (a) a standalone 3T IBC subcell and (b) a generic s-connected 3TT device. The node names are independent of the cell doping, so neutral colors are used. Each cell can be loaded in three different ways, indicated by which contact is common between the loads (CR, CZ, C(F/T)) as shown schematically below the larger diagrams. The voltage difference or current flowing between two nodes is indicated using the two subscripts of the respective nodes (e.g V_RT). The current through each node is indicated with the subscript of the node (e.g., J_Z).

Figure 3

(a) Current density vs voltage and power density vs voltage plots for a 4TT GaAs//Si experimental device measured under AM1.5G illumination. (b) The same 4TT data, plotted as power contours in V_GaAs vs V_Si space.

Figure 4

Simulated 3TT GaAs/s/nuIBC-Si device power contours under AM1.5G illumination. (a) Schematic of CZ loading; (b) schematic of CR loading; (c) CZ P–V–V plot; (d) CR P–V–V plot; (e) CZ P–J–J plot; (f) CR P–J–J plot. In the CZ case, both loads are producing power at the max power point, but in the CR case, the load across the R and Z nodes is injecting power into the device to achieve the same operating state (hence negative currents for J_ZR). (This simulation neglects luminescent coupling between the cells.)

Figure 5

Simulated CZ P–V–V contours for multiple configurations of a 3TT device based on GaInP and Si (under AM1.5G illumination): (a) GaInP/s/nuIBC, (b) GaInP/r/nuIBC, (c) GaInP/s/pbIBC, and (d) GaInP/r/pbIBC.

Figure 6

Simulated CZ P–V–V contours for a subset of configurations of a 3TT device based on GaAs and Si (under AM1.5G illumination): (a) GaAs/r/nuIBC and (b) GaAs/r/pbIBC.