Ytterbium sensitization in KY₃F₁₀: Pr³⁺, Yb³⁺ for silicon solar cells efficiency enhancement

Abstract

The codoping of KY₃F₁₀ with Pr³⁺ and Yb³⁺ ions is investigated as a possible quantum cutting system to enhance solar cells efficiency. For one visible photon absorbed by Pr ions, two ytterbium ions are expected to be excited by two consecutive energy transfers. The subsequent emission of two infrared photons reduces thus the thermalization losses usually observed in Si solar cells. Emission spectra and lifetime decays in KY₃F₁₀ doped with 0.5% Pr³⁺ and codoped with 0%, 1%, 10% and 20% Yb³⁺ show an increase of the energy transfer efficiency from Pr³⁺ to Yb³⁺ with the Yb³⁺ concentration. For the first Pr³⁺ to Yb³⁺ energy transfer, an efficiency close to 100% is achieved in KY₃F₁₀: 0.5%Pr³⁺, 20%Yb³⁺. However, this promising result faces challenging issues since an increase in Yb concentration induces energy migration between Yb³⁺ ions which impairs the Yb³⁺ luminescence.

Introduction

Solar cells present an intrinsic limitation for the efficient conversion of sunlight radiation into electricity as only the photons with energy close to the semiconductor’s bandgap energy can be profitably absorbed by the solar cell. Low energy photons go through the cell without being absorbed (sub-bandgap transparency) therefore without generating electron–hole pairs. For the high energy part of the solar spectrum, supra-bandgap photons are absorbed, but the excited electrons undergo a rapid thermalization losing part of their energy as heat. As a consequence, the conversion efficiency is limited to 30% [1] in classical silicon solar cells.

Thermalization losses can be reduced with a luminescent converter by absorbing UV or visible photons and splitting them into two or more near-infrared (NIR) photons [2]. This downconversion process is usually referred to as quantum cutting (QC). In this study, we investigated KY₃F₁₀ codoped with Pr³⁺ and Yb³⁺ ions as a promising quantum cutting material.

Trivalent ytterbium has a single excited state around 10,000 cm⁻¹. Therefore, the Yb³⁺ emission can be efficiently absorbed by silicon solar cells without significant thermalization losses. An efficient quantum cutting scheme using Yb³⁺ ions as acceptors entails donor ions able to transfer efficiently their energy to Yb³⁺ ions so as to obtain two excited Yb ions for one excited donor ion. It implies that the donor ion has an excited energy level around twice the energy of the Yb ²F_5/2 energy level, as well as an intermediate level at approximately the same energy as the ²F_5/2 energy level in order to obtain two resonant energy transfers. Praseodymium appears to be an ideal candidate since ³P₀ is located at 20,000 cm⁻¹ approximately and ¹G₄ around 10,000 cm⁻¹ [3]. Once excited, the depopulation of the ³P₀ level can take place through two consecutive energy transfers, the first one can be written schematically as Pr³⁺(³P₀ → ¹G₄); Yb³⁺(²F_7/2 → ²F_5/2) and the second one as Pr³⁺(¹G₄ → ³H₄); Yb³⁺(²F_7/2 → ²F_5/2) as displayed in Fig. 1. As a result, the absorption of a blue photon in the ³P_J levels leads, in an ideal situation, to the excitation of two Yb³⁺ ions and to the emission of two NIR photons.

Energy transfer processes between Pr³⁺ and Yb³⁺ have been studied for upconversion and downconversion applications in host materials such as fibers [4], chlorides [5] and fluorides [6]. An efficient quantum cutting mechanism for solar cell applications was reported in SrF₂: Pr³⁺, Yb³⁺ [6]. The possibility of a cooperative mechanism between Pr³⁺ and Yb³⁺ has also been proposed [7], [8]. In a cooperative process, two Yb³⁺ ions would be simultaneously excited by energy transfer from one excited Pr³⁺ ion. This result is somewhat surprising since a cooperative mechanism is a second-order process which typically is four or five orders of magnitude less effective than two-steps first-order processes [9]. Unless a quantitative analysis describing the respective contributions of first-order and second-order processes is performed, first-order processes such as the one described in Fig. 1 are expected to dominate.

The doping of KY₃F₁₀ with rare-earth ions has been reported in few publications [10], [11], [12], [13], [14], [15], [16], [17]. As a fluoride material, its relatively low phonon energy (420 cm⁻¹) makes it highly suitable for applications involving energy transfers. A low phonon energy host lattice reduces non-radiative relaxations via multiphonon emission. Therefore, it favours high fluorescence quantum yields and efficient energy transfers. The codoping of KY₃F₁₀ with Pr³⁺ and Yb³⁺ ions was previously investigated as a promising upconversion material. A strong fluorescence signal from ³P₀ under IR excitation was reported in KY₃F₁₀ codoped Pr³⁺ and Yb³⁺ [15]. To our knowledge, the study presented in this paper is the first investigation of Yb sensitization in KY₃F₁₀: Pr³⁺, Yb³⁺ for solar cells efficiency enhancement.