Mg 3 Sb 2 –Mg 3 Bi 2 alloys have been heavily studied as a competitive alternative to the state-of-the-art n-type Bi 2 (Te,Se) 3 thermoelectric alloys. Using Mg 3 As 2 alloying, we examine another dimension of exploration in Mg 3 Sb 2 –Mg 3 Bi 2 alloys and the possibility of further improvement of thermoelectric performance was investigated. While the crystal structure of pure Mg 3 As 2 is different from Mg 3 Sb 2 and Mg 3 Bi 2 , at least 15% arsenic solubility on the anion site (Mg 3 ((Sb 0.5 Bi 0.5 ) 1−x As x ) 2 : x = 0.15) was confirmed. Density functional theory calculations showed the possibility of band convergence by alloying Mg 3 Sb 2 –Mg 3 Bi 2 with Mg 3 As 2 . Because of only a small detrimental effect on the charge carrier mobility compared to cation site substitution, the As 5% alloyed sample showed zT = 0.6–1.0 from 350 K to 600 K. This study shows that there is an even larger composition space to examine for the optimization of material properties by considering arsenic introduction into the Mg 3 Sb 2 –Mg 3 Bimore »
The importance of the Mg–Mg interaction in Mg 3 Sb 2 –Mg 3 Bi 2 shown through cation site alloying
Herein we study the effect alloying Yb onto the octahedral cite of Te doped Mg 3 Sb 1.5 Bi 0.5 has on transport and the material's high temperature stability. We show that the reduction in mobility can be well explained with an alloy scattering argument due to disrupting the Mg octahedral –Mg tetrahedral interaction that is important for placing the conduction band minimum at a location with high valley degeneracy. We note this interaction likely dominates the conducting states across n-type Mg 3 Sb 2 –Mg 3 Bi 2 solid solutions and explains why alloying on the anion site with Bi isn't detrimental to Mg 3 Sb 2 's mobility. In addition to disrupting this Mg–Mg interaction, we find that alloying Yb into the Mg 3 Sb 2 structure reduces its n-type dopability, likely originating from a change in the octahedral site's vacancy formation energy. We conclude showing that while the material's figure of merit is reduced with the addition of Yb alloying, its high temperature stability is greatly improved. This study demonstrates a site-specific alloying effect that will be important in other complex thermoelectric semiconductors such as Zintl phases.
- Award ID(s):
- Publication Date:
- NSF-PAR ID:
- Journal Name:
- Journal of Materials Chemistry A
- Page Range or eLocation-ID:
- 2033 to 2038
- Sponsoring Org:
- National Science Foundation
More Like this
The Zintl compound Eu 2 ZnSb 2 was recently shown to have a promising thermoelectric figure of merit, zT ∼ 1 at 823 K, due to its low lattice thermal conductivity and high electronic mobility. In the current study, we show that further increases to the electronic mobility and simultaneous reductions to the lattice thermal conductivity can be achieved by isovalent alloying with Bi on the Sb site in the Eu 2 ZnSb 2−x Bi x series ( x = 0, 0.25, 1, 2). Upon alloying with Bi, the effective mass decreases and the mobility linearly increases, showing no signs of reduction due to alloy scattering. Analysis of the pair distribution functions obtained from synchrotron X-ray diffraction revealed significant local structural distortions caused by the half-occupied Zn site in this structure type. It is all the more surprising, therefore, to find that Eu 2 ZnBi 2 possesses high electronic mobility (∼100 cm 2 V −1 s −1 ) comparable to that of AM 2 X 2 Zintl compounds. The enormous degree of disorder in this series gives rise to exceptionally low lattice thermal conductivity, which is further reduced by Bi substitution due to the decreased speed of sound. Increasing themore »
Improved stability and high thermoelectric performance through cation site doping in n-type La-doped Mg 3 Sb 1.5 Bi 0.5n-Type conduction in a Mg 3 Sb 1.5 Bi 0.5 system is achieved with La-doping at cation sites with a peak zT > 1. La-doped samples exhibit much higher doping efficiency and dopability compared to other chalcogen-doped samples. This allows greater tunability of the electronic properties. La-doping also significantly improves the thermal stability of n-type Mg 3 Sb 1.5 Bi 0.5 measured via a long-term Hall carrier concentration measurement.
Recent, and somewhat surprising, successful n-type doping of Mg 3 Sb 2 was the key to realizing high thermoelectric performance in this material. Herein, we use first-principles defect calculations to investigate different extrinsic n-type doping strategies for Mg 3 Sb 2 and to reveal general chemical trends in terms of dopant solubilities and maximal achievable electron concentrations. In agreement with experiments, we find that Sb substitution is an effective doping strategy, with Se and Te doping predicted to yield up to ∼8 × 10 19 cm −3 electrons. However, we also find that Mg substitution with trivalent (or higher) cations can be even more effective; in particular, the predicted highest achievable electron concentration (∼5 × 10 20 cm −3 ) with La as an extrinsic dopant exceeds that of Se and Te doping. Interstitial doping (Li, Zn, Cu, Be) is found to be largely ineffective either due to self-compensation (Li) or high formation energy (Zn, Cu, Be). Our results offer La as an alternative dopant to Te and Se and reinforce the need for careful phase boundary mapping in achieving high electron concentrations in Mg 3 Sb 2 .
A M 2 X 2 compounds that crystallize in the CaAl 2 Si 2 structure type have emerged as a promising class of n- and p-type thermoelectric materials. Alloying on the cation (A) site is a frequently used approach to optimize the thermoelectric transport properties of A M 2 X 2 compounds, and complete solid solubility has been reported for many combinations of cations. In the present study, we investigate the phase stability of the AMg 2 Sb 2 system with mixed occupancy of Mg, Ca, Sr, or Ba on the cation (A) site. We show that the small ionic radius of Mg 2 + leads to limited solubility when alloyed with larger cations such as Sr or Ba. Phase separation observed in such cases indicates a eutectic-like phase diagram. By combining these results with prior alloying studies, we establish an upper limit for cation radius mismatch in A M 2 X 2 alloys to provide general guidance for future alloying and doping studies.