Lanthanide Ions for Solar Cells

Andries Meijerink

Condensed Matter and Interfaces, Debye Institute for NanoMaterials Science
Science Utrecht University, Princetonplein 5
3584 CC Utrecht, The Netherlands

Introduction

Global energy consumption is on the rise and is projected to double by 2050, compared with worldwide energy consumption rates in 2001.¹ Sustainable energy production based on the direct conversion of energy radiated from the sun into heat or electricity is expected to gain importance, as it may be the only renewable source capable of generating sufficient energy to meet the long-term worldwide energy demand.^1,2 The capacity of photovoltaic cells to convert sunlight into electricity makes them prime candidates for effective large-scale capture and conversion of solar energy, but at present the contribution of photovoltaic energy is limited, due to its relatively high cost per kilowatt-hour. A reduction in price may be achieved by either lowering production costs or increasing conversion efficiency.

Crystalline Si (c-Si) solar cells dominate the photovoltaic market and have energy efficiencies around 15 %. In a solar cell, a single electronhole pair is generated in a semiconductor upon absorbing a photon above the energy bandgap. The most significant loss mechanisms are due to relaxation of "hot" charge carriers that are created upon absorption of a high-energy photon and transmission of photons with energies below the bandgap of the semiconductor material.³ The excess energy of high energy photons is rapidly lost as heat by thermalization. In a detailed balance model developed by Shockley and Queisser,⁴ the theoretical efficiency limit for a single-junction solar cell with Eg equal to 1.1 eV can be determined to be 30%. A promising approach to raise the theoretical efficiency beyond the Shockley-Queisser limit is to adapt the solar spectrum through upconversion or downconversion. In upconversion, two low-energy photons are "added up" to give one higherenergy photon,⁵ thus converting sub-bandgap photons, which are otherwise lost, into supra-bandgap photons, which can be absorbed. Downconversion, or "quantum cutting," is the opposite process, whereby one high-energy photon is "cut" into two lower-energy photons that are both absorbed. Figure 1 shows the standard terrestrial solar spectrum and (in green) the fraction of the energy that can be harvested by a c-Si solar cell, assuming there are no other losses besides spectral mismatch losses.

Figure 1. Potential efficiency gain of down- and 2-photon upconversion for a c-Si solar cell. The green part gives the part of the energy from the solar spectrum that can be converted into electrical energy by a c-Si solar cell, assuming no other losses than spectral mismatch losses. The yellow area gives the energy gain that can be obtained with an ideal upconverter, while the red part gives the maximum gain with an ideal downconverter.

The gain that can be achieved by using a perfect upconverter (in yellow) and perfect downconverter (in red) is about 20 % for each process for c-Si solar cells. In this review, the potential of using lanthanides as downconverters and upconverters will be outlined, after a brief introduction to the unique optical properties of lanthanides.

Luminescent Lanthanides

Lanthanides (Ln) are a group of elements that can be found at the bottom of the periodic table. The 4f inner shell of lanthanides is partially filled with electrons. They are mostly stable in the trivalent state, and the Ln³⁺ ions have the electronic configuration 4 fⁿ 5 s ² 5 p⁶ , where n varies from 0 to 14. The partially filled 4f inner shell is responsible for the characteristic optical and magnetic properties of the lanthanides. The number of configurations for n electrons divided over the 14 4f orbitals is large (14/n), and all of the configurations can have different energies. In Figure 2, the energy level diagrams for the trivalent lanthanide ions, from Ce to Yb, are shown.

Figure 2. Free ion energy levels of the trivalent lanthanide ions from Ce (4f¹) to Yb³⁺ (4fy¹³). Levels are labelled by term symbols or, for some higher levels, capital letters.

The horizontal black lines represent energy levels, labeled by term symbols ^2S+1LJ. This diagram, often called the "Dieke diagram,"⁶ exhibits the rich energy level structure of the free ions. It is also representative of the 4f energy level structure of these ions when doped into crystalline or glassy materials, because the optically-active 4f electrons are shielded from the host environment by the outer filled 5s and 5p orbitals. The shielding also results in sharp, atomic-like lines in the optical spectra. Quantum efficiency, defined as the number of photons emitted divided by the number of photons absorbed, can be very high, typically above 90%. The high efficiencies are the reason for the widespread application of lanthanides in light-emitting devices. In Figure 3, efficient luminescence is demonstrated for two applications.

Figure 3. (top) Luminescence of blue (Eu²⁺), green (Tb³⁺) and red (Eu³⁺) phosphors for (compact) fluorescent tubes under 254 nm UV irradiation. The white luminescence is obtained for a mixture of the blue, green and red phosphors. (bottom) Euro banknote under 360 nm UV irradiation. The red, green and blue luminescence from the stars and fibers is generated by Eu²⁺ (blue and green) and Eu³⁺ (red).

At the top, the emission of light by three lanthanide doped luminescent materials, or "phosphors," for application in fluorescent tubes is shown. At the bottom, the luminescence of a Euro banknote under UV-irradiation is shown as it is observed under the blacklight at a cash register. The red, green, and blue lights originate from Eu³⁺ and Eu²⁺ ions. The choice of safeguarding the Euro banknotes with the element europium has been deliberate.⁷ For up- and downconversion for solar cells, lanthanide ions are ideal candidates. The rich energy level structure allows many pathways to add or split the energy of incoming photons without significant energy losses. The process can take place within a single type of lanthanide ion, or it can involve energy transfer between two or more types of ions codoped within the same host material.

Upconversion

Upconversion with lanthanides was discovered in the early 1960s.⁵ For the lanthanide ion couple Yb³⁺-Er³⁺, it was shown that upon infrared excitation of Yb³⁺ around 1000 nm, green and red luminescence from Er³⁺ is observed. The steps involved are illustrated in Figure 4.

Figure 4. Schematic energy level scheme for the Yb/Er couple. The Yb³⁺ ion absorbs around 980 nm and transfers the energy from the ²F_5/2 level to the ⁴I_11/2 level of Er³⁺. Subsequent energy transfer from a second excited Yb³⁺ ion to Er³⁺ (⁴I_11/2), excites Er³⁺ ion to to the ⁴F_7/2 excited state. After multi-phonon relaxation to the lower lying ⁴S_3/2 and ⁴F_9/2 states, green and red emission are observed, as indicated.

The process was originally called "addition de photons par transfer d′energie" (APTE), but it is now generally known as energy transfer upconversion (ETU), and it is the most efficient mechanism for upconversion with lanthanide ions.⁵ Different types of energy transfer mechanisms are possible, but non-radiative energy transfer via dipoledipole interaction is mainly the dominant mechanism. Efficient energy transfer requires the ions to be in close proximity; thus, ETU requires high concentration of the dopants. One of the most efficient upconverter materials for NIR to VIS upconversion is NaYF₄: Er³⁺, Yb³⁺. Lanthanide upconverters have been applied in solar cells. Upconversion efficiency is low, and most demonstrations merely serve as proof of the principle showing that an increase in efficiency can be realized by applying an upconversion layer. In general, the upconverter is applied at the back of a cell as an electrically isolated layer. A back reflector reflects all emitted photons back into the solar cell. The first experiment was conducted on GaAs (Prod. No. 329010) solar cells combined with a vitroceramic material doped with Yb³⁺ and Er³⁺ under extremely high excitation densities. An efficiency of 2.5 % of the solar cell was obtained. In 2005, Shalav et al.⁸ showed upconversion under lower excitation density of 2.4 W/cm² reaching 3.4 % quantum efficiency at 1,523 nm in a crystalline silicon solar cell with NaYF₄ doped with Er³⁺ as upconverter. Since c-Si has a rather small bandgap (1.12 eV), transmission losses are not as high as for wider bandgap solar cells. Therefore, the efficiency gain for larger bandgap solar cells can be higher.

Presently, upconversion efficiencies are still rather low; only at high excitation densities, exceeding the solar power density, are efficiencies above 1 % reached. Concentration of solar light is needed. In addition, it is crucial to broaden the absorption spectrum, as the absorption lines for lanthanide upconverters are narrow, and absorption strengths are low for the parity forbidden transitions within the 4fⁿ configuration. Broadening of the absorption spectrum for lanthanide upconverters can be achieved by using a sensitizer with a broad absorption band and a narrow emission line resonant with the lanthanide absorption line. Sensitization can be achieved by an external sensitizer (e.g., quantum dots), or an internal sensitizer (e.g., transition metal ions). Even more challenging are options to enhance upconversion efficiencies by manipulating emission and excitation processes through plasmonic coupling.⁹ The use of plasmonic effects with upconverter materials is a new and emerging field, with many possibilities and challenges.

Downconversion

The idea to obtain quantum efficiencies above 100 % by creating multiple photons through "cutting" a single photon into two lowerenergy photons was first proposed by Dexter in 1957.¹⁰ The mechanism involved the simultaneous energy transfer from a donor to two acceptors, each accepting half the energy of the excited donor. It was not until 1974 that experimental evidence for quantum yields above 100 % was obtained, for YF₃:Pr³⁺.^11,12 The mechanism was not the one proposed by Dexter, but it involved two sequential emission steps from the high-energy ¹S₀. Later, quantum cutting via two sequential energy transfer steps in the Gd³⁺-Eu³⁺ couple was discovered and, based on the analogy of the two-step energy transfer process in upconversion, it was called "downconversion."¹³ The aim in all this work was to achieve the emission of two visible photons from a single UV photon in order to boost the efficiency of light-emitting devices. The potential of downconversion for increasing the efficiency of solar cells was realized soon afterwards.² The first experimental demonstration of downconversion for solar cells involved the Tb³⁺-Yb³⁺ couple, where quantum cutting was achieved through cooperative energy transfer from Tb³⁺ to two Yb³⁺ ions, the very mechanism suggested almost 50 years earlier by Dexter.¹⁵ It is evident from the Dieke diagram (Figure 2) that the energy level structure of Yb³⁺ is ideally suited for use in downconversion for c-Si solar cells. The Yb³⁺ ion has a single excited state approximately 10,000 cm^-1 above the ground state, corresponding to an emission of around 1,000 nm. The absence of other energy levels allows Yb³⁺ to exclusively "pick up" energy packages of 10,000 cm^-1 from other codoped lanthanide ions and emit photons at ~1,000 nm, which can be absorbed by c-Si. Efficient downconversion using Yb³⁺ via resonant energy transfer requires donor ions with an energy level at about 20,000 cm^-1 and an intermediate energy level at approximately 10,000 cm^-1. Examination of Figure 2 reveals potential ion couples, such as E^r3+ - Yb³⁺, Nd³⁺ - Yb³⁺, Ho³⁺ - Yb³⁺, and Pr³⁺ - Yb³⁺. As an example, Figure 5 shows the energy transfer processes leading downconversion for the Pr³⁺-Yb³⁺ couple.

Figure 5. Energy levels and quantum cutting mechanism for the Pr³⁺ - Yb³⁺ couple. A two-step energy transfer occurs upon excitation into the ³P_J (J=0, 1, 2) and ¹I₆ levels of Pr³⁺. A single visible photon absorbed by these levels is thereby converted into two ~1,000 nm photons. Solid, dotted, and curved arrows represent optical transitions, nonradiative energy transfer processes, and nonradiative relaxation, respectively.

After excitation into one of the ³P_J levels between 450 and 490 nm, efficient two-step energy transfer to two neighboring Yb³⁺ ions occurs with internal quantum efficiencies close to 200 %. Both Yb³⁺ ions can emit a 1,000 nm photon that can be absorbed by a c-Si solar cell.

The promising results of downconversion with lanthanides does not mean that one can expect implementation of downconversion materials in solar cells in the near future. On the contrary, serious challenges need to be addressed before downconversion materials can be applied. A major limitation, just as for upconversion, is the weak absorption for 4fⁿ transitions. A solution may be a sensitizer that absorbs efficiently in the UV/VIS and transfers the energy to the downconversion couple. Work on the sensitization is only just starting. A second issue is concentration quenching. High Yb³⁺-concentrations are needed to achieve complete energy transfer to the Yb³⁺ acceptor ions. At these high concentrations, quenching of the emission occurs through energy migration over the Yb³⁺-sublattice (concentration quenching) and lowers efficiency. Finally, if an efficient downconversion couple is developed, the material needs to be incorporated in a transparent layer on top of the solar cell. To prevent losses due to isotropic emission, an anti-reflective coating for the 1,000-nm Yb³⁺ emission is required on top of the downconversion layer. Clearly, the road toward implementation is long.

Conclusions

Spectral conversion for solar cells is an emerging concept in the field of photovoltaics, and it has the potential to increase significantly the efficiency of solar cells. Lanthanide ions are ideal candidates for spectral conversion, due to their high luminescence efficiencies and rich energy level structure that allows for great flexibility in the upconversion and downconversion of photons in a wide spectral region (NIR-VIS-UV). Proof-of-concept experiments have been reported for upconversion, demonstrating an increase in efficiency for sub-bandgap illumination for different types of solar cells by application of upconverter materials. The challenges lie in improving upconversion efficiency, especially for the relatively low excitation densities that are typical for solar illumination.

Quantum cutting through downconversion has only been studied in the last decade. It offers great potential for efficiency enhancement for narrow bandgap solar cells, such as c-Si solar cells. Efficient downconversion has been reported for several lanthanide couples. In these couples, Yb³⁺ serves as an ideal acceptor with a single excited state, just above the bandgap of c-Si. High internal quantum yields (close to 200 %) have been demonstrated and are independent of the incident power. Before implementation in solar cell systems can be realized, some serious issues need to be addressed: sensitization, concentration quenching, and a transparent downconversion layer on top of the solar cell. Research on solving the various issues is in progress, and new results can be expected in the coming years. Based on these results, it will become clear if photon management using lanthanides is indeed a viable option to realize more efficient solar cells.