Session C11: Defects in Semiconductors -- Energy Materials

Bright triplet excitons in cesium lead halide perovskites

The observation of a ground optically forbidden “dark” exciton state in semiconductor nanocrystals was first reported in the seminal paper of Nirmal et al. in 1995. ¹ Later research in nanowires, nanorods, and nanoplatelets has shown that the ground exciton state in all these semiconductor structures is a dark exciton, leading us to believe that the ground exciton must be dark. Because dark excitons release photons slowly, hindering emission, semiconductor nanostructures that disobey this rule have been sought. Three years ago however cesium lead halide perovskite (CsPbX3, with X = Cl, Br or I ) nanocrystals were grown, which demonstrated very bright photoluminescence (PL) with quantum yield 50-90% at room temperature. This bright emission was traced to a very short radiative decay time. The nanocrystals emit light about 20 and 1,000 times faster than any other semiconductor nanocrystal at room and cryogenic temperatures, respectively. The increase of the decay time with temperature is inconsistent with a dark ground state exciton suggesting that in these nanocrystals the ground exciton state is bright. We use an effective-mass model and group theory to demonstrate the possibility of such a ground bright state existing, which can occur when the strong spin–orbit coupling in the conduction band of perovskites is combined with the Rashba effect.² We then apply our model to CsPbX3 nanocrystals, and measure size- and composition- dependent fluorescence at the single-nanocrystal level. The bright triplet character of the lowest exciton explains the anomalous photon-emission rates of these materials. The existence of this bright triplet exciton is further confirmed by analysis of the fine structure in low-temperature fluorescence spectra.
¹Nirmal, M. et al. “Observation of the “dark exciton” in CdSe quantum dots.” Phys. Rev. Lett. 75, 3728–3731 (1995).
² Becker, M. A. “Bright triplet excitons in caesium lead halide perovskites,” Nature, 553, 189-193 (2018)   

Formation of DY Defect Centers in Bi-Doped Hybrid Halide Perovskites

The DX center is a major killer defect that limits the n-type doping in some four-fold coordinated semiconductors. It is a deep negatively charged defect complex converted from a nominal shallow donor defect, which can serve as a trap center of electrons, thus is detrimental to the performance of optoelectronic devices. Similar to the DX center, we find that a donor-yielded complex center (DY center) also exists in six-fold coordination semiconducting materials. For example, Bi is commonly used as n-type dopant in perovskite APbX₃. However, our first-principles calculations show that the DY centers are formed in Bi doped MAPbBr₃ when the Fermi level is high in the gap, but, interestingly, it does not form in MAPbI₃. The reason that the DY center is formed in MAPbBr₃ instead of MAPbI₃ is attributed to the high conduction band minimum (CBM) of MAPbBr₃. Our results are able to explain recent puzzling experiment observations and the thorough discussions of the formation and the properties of the DY center in perovskites provide enlightening insights to the defect study in six-fold coordinated semiconductors.
  

Significantly Different Solubility of Li, Na and K Dopants in Cu(In,Ga)Se2 and Cu2ZnSn(S,Se)4 Solar Cells

The doping of alkaline elements such as Na, K, Rb and Cs in Cu(In,Ga)Se₂ (CIGS) and Cu₂ZnSn(S,Se)₄ (CZTSSe) thin film solar cells had been intensively studied for decades and was shown to be critical for achieving high photovoltaic efficiency. However, the doping of the light alkaline element Li has been much less studied. Using the first-principles calculations, we show here that the formation energy of Li dopants is very low in CIGS and CZTSSe, while the formation energies of Na and K dopants are much higher, so the doping concentration (solubility) of Li can be very high in the CIGS and CZTS lattices at room temperature, much higher than those of Na and other heavier alkaline elements. Na and K may be doped into the CIGS and CZTSSe lattices with a high solubility at high temperature, but the dopants will diffuse out of the grains and prefer staying on the grain boundaries at room temperature because of the low solubility in the lattice. The concentration increase of the hole carriers after the alkaline doping will be discussed based on the calculated solubility.   

Modeling of disorder in II-IV-V2 semiconductors for tuning of novel properties

The customization of multinary semiconductors has attracted great interest in the science community for a diverse set of novel applications. The properties of these materials can be fine-tuned by controlling composition and atomic ordering. However, an understanding of the structure-synthesis-property relationship is essential for a rational design. In this contribution, we use two II-IV-V₂ materials: ZnSnN₂ and ZnGeN₂ as examples to illustrate the ordering effects. Model Hamiltonian based Monte Carlo simulations were used to create structures with different degrees of disorder. The energies and electronic structures were estimated from first principles calculations. We find that energies of ZnSnN₂ can be well approximated by a Motif Hamiltonian which incorporates only short-range ordering. We demonstrate that, only with the consideration of disorder and oxygen contamination, the net doping level in ZnSnN₂ can be lowered to agree with experiments (10¹⁷ cm^-3). However, for ZnGeN₂, long-range ordering effects step in, and thus, we use the Cluster Expansion Hamiltonian to create disordered structures. The thermodynamics and optoelectronic properties of disordered ZnGeN₂ will be discussed.   

Sulfur Vacancies as the Origin of n-type Doping in Unintentionally Doped Pyrite FeS2 Single Crystals

Pyrite FeS₂ has long been considered an ideal semiconductor for low-cost solar cells as it is composed of earth-abundant, non-toxic, inexpensive elements, has a suitable band gap (0.95 eV), and absorbs sunlight strongly. Disappointing power conversion efficiencies, however, have plagued pyrite solar cells. This failing is now widely acknowledged to be due to problems with pyrite’s surface, and a lack of doping control. An important unanswered question is the origin of the n-type behavior seen in unintentionally-doped pyrite single crystals and thin films. Here, we present the first substantial electronic transport evidence that sulfur vacancies are this n-type dopant by varying sulfur vapor pressure (P_S) during single crystal growth. Crystals grown under high P_S exhibit semiconducting behavior, with transport activation energies of 225 meV and 300 K electron densities (n_300K) of 10¹⁶ cm^-3. Decreasing P_S increases n_300K to >10¹⁷ cm^-3 and decreases the activation energy ten-fold, evidencing an evolution towards an insulator-metal transition. These trends are independent of metal impurity concentrations and, moreover, n_300K is too large to be explained by these impurities.   

The nature of band gap of Co3O4 – a revisit from first-principles

Cobalt Oxide (Co₃O₄) has emerged as a highly promising material for a wide variety of energy technologies, including hydrogen generation through solar-water-splitting and Li ion batteries. Yet, a detailed understanding of the electronic properties of this material is largely lacking. For example, contradicting experimental results have been reported for the optical gap, leading to two commonly reported values of 0.8 eV and 1.6 eV. Here we have employed hybrid functional calculations compliant with the generalized Koopmans' theorem, to demonstrate that the intrinsic band gap of Co₃O₄ is ~1.6 eV. Meanwhile, the ~0.8 eV transition found experimentally is due to the presence of polaron or defect states. In particular, our calculations predict the spontaneous formation of electron and hole polarons, that in turn exhibit significant contribution to the absorption spectra of the material and are responsible for the optical excitation at 0.8 eV. Finally, we resolve the nature of the stable spin states of electron and hole polarons, and we discuss how the interaction between polarons with n-type dopants (carbon) could improve the electrical conductivity of this intrinsic p-type material.   

Grain-size dependent carrier compensation in Cu2O

Cuprous oxide (Cu₂O) is an absorber material for low cost solar cells and various fundamental phenomena have been explained using this material. Phase pure Cu₂O samples with different grain sizes, in the range of 400 µm to 3 mm, were obtained by thermal oxidation of a Cu sheet. The temperature dependent carrier concentration of all samples follows the compensated semiconductor model. In this model, two independent monovalent acceptors (N_A1, N_A2) are assumed with one compensating donor level (N_D) and the carrier concentration is obtained from charge neutrality condition. It is observed that the difference in energy levels between two acceptors is ~80 meV, pertaining to a normal and a split Cu-vacancy in Cu₂O as acceptor defects. Interestingly, the concentration of donor defects increases with increasing grain boundary cross section (Λ_GB), from 5.8×10¹³ cm^-3 for a sample with Λ_GB=0.22x10^-3 µm^-1 to 3.0×10¹⁴ cm^-3 for a sample with Λ_GB=0.45x10^-3 µm^-1. It indicates that the grain boundaries act as source of compensating donor defects in Cu₂O. This study suggests that the increment in grain size of Cu₂O can improve the performance of electronic devices based on it.   

Structural and electronic properties of amorphous In2O3:Li,Na from first principles

Amorphous oxide semiconductors (AOS) have found wide application in consumer electronic devices due to their unique properties, namely, high carrier mobility and transparency in the visible. Among the AOS materials, indium-based oxides show best overall performance. Additional cations, such as Zn, Ga, or Sn, allow one to tune the structural, electronic, and optical properties of ternary and quaternary AOS over wide ranges [1-2].
Here we present a thorough systematic analysis of the structural, electronic, and optical properties of amorphous In₂O_3-x doped with Li and Na. From the theory of glass formation, it is known that alkali metal oxides such as Li₂O and Na₂O serve as structural modifiers, increasing the materials density. We employ ab-initio molecular dynamics liquid-quench simulations and accurate density-functional calculations to understand how addition of Li or Na affects the local In-O structure as well as medium-range In-In structure that determines the morphology of the amorphous network. The resulting electronic and optical properties of Li and Na doped amorphous indium oxide are investigated as a function of density, quench rates, cation composition, and oxygen stoichiometry.
[1] J.E. Medvedeva, et al., Adv. Electron. Mater., (2017)
[2] R. Khanal, et al., PRB, (2015)   

Probing the Electronic States in GaAsNBi Alloys

Although solar energy is the most abundant renewable energy source, an approach to combine energy harvesting and energy storage is urgently needed. Recently, solar to hydrogen (STH) conversion using photoelectrochemical cells (PEC) has emerged as a promising approach to harvest and store solar energy. Indeed, record STH efficiency is predicted for a tandem PEC which includes a 1.05 eV bandgap junction.¹ Recently, we synthesized a novel alloy, gallium arsenide nitride bismide,² which is nearly lattice-matched to gallium arsenide, with bandgaps tunable from 0.87 to 1.34 eV. To determine the energy bandoffsets and electronic states of GaAsNBi alloys, we are examining GaAs/GaAsNBi/GaAs single quantum well structures with dilute concentrations of N and Bi. Using a combination of photoluminescence spectroscopy, in conjunction with capacitance-voltage measurements and 1D-Schrödinger-Poisson simulations, we determine the band offsets and deep levels in GaAs/GaAs_1-x-yN_xBi_y/GaAs heterostructures.
¹J. L. Young, M.A. Steiner, H. Doscher, R.M. France, J.A. Turner, T.G. Deutsch, Nature Energy, 17028 (2017).
²J. Occena, T. Jen, E.E. Rizzi, T.M. Johnson, J. Horwath, Y.Q. Wang, R.S. Goldman, Appl. Phys. Lett.110, 242102 (2017).   

A theoretical study of the out-of-plane electrical transport of n-type thermoelectric SnS

As a promising thermoelectric material containing non-toxic and abundant elements, SnS has attracted significant research focus in recent years. To further improve the thermoelectric energy conversion efficiency of SnS, it is critical to have a thorough understanding of its electronic structure. In this work, we have calculated the band structure of SnS from first principles using the HSE06 hybrid functional to unveil the detailed valence band features. We have then compared the band features of SnS with those of the analogue SnSe, which was recently shown to have a very high thermoelectric figure of merit. By further investigating the doping effects of n-type impurities on the electronic structure and electrical transport properties of SnS, we identify resonant states near the bottom of the conduction band and find a considerable increase of electron delocalization along the out-of-plane direction of the layer crystal structure. Boltzmann transport calculations show that these effects on the electronic structure may result in enhanced electrical transport properties.   

scanning tunneling microscopy of defects in SnS2

We present our room- temperature scanning tunneling microscopy measurements of SnS₂, a quasi-two dimensional layered semiconductor. SnS₂ is a material of interest due to its high carrier mobility and its potential use in applications including in photonics, electronics, and solar energy conversion. Understanding the surface properties of SnS₂ are particularly important for its incorporation and optimization in applications. Here we present our surface characterization of SnS₂ and detail several surface features including surface/subsurface defects which locally distort the crystal lattice.