Designing new Zintl phases SrBaX (X = Si, Ge, Sn) for thermoelectric applications using \textit{ab initio} techniques

TL;DR

This work addresses the design of new thermoelectric materials by proposing SrBaX (X = Si, Ge, Sn) Zintl phases and evaluating their stability, bonding, phonons, and transport from first principles. It employs DFT (VASP with PAW, PBE/mBJ), lattice-dynamics calculations (Phonopy, ShengBTE), electron-phonon transport (Amset), COHP analysis (Lobster), and AIMD to map structural and transport properties. The results show ultralow lattice thermal conductivities ($\kappa_L<1$ W m$^{-1}$ K$^{-1}$) driven by weak bonding and strong anharmonic scattering, with SrBaGe achieving $ZT$ up to about $2.0$ at 700 K, while SrBaSn exhibits bipolar conduction that limits performance. Guided by these insights, the study discusses dopant and bandgap engineering strategies and motivates experimental synthesis of the SrBaX series for practical device applications.

Abstract

Slack's phonon-glass and electron-crystal concept has been the guiding paradigm for designing new thermoelectric materials. Zintl phases, in principle, have been shown as great contenders of the concept and thereby good thermoelectric candidates. With this as motivation, we design new Zintl phases SrBaX (X = Si, Ge, Sn) using state-of-the-art computational methods. Herein, we use first-principles simulations to provide key theoretical insights to thermal and electrical transport properties. Some of the key findings of our work feature remarkably low lattice thermal conductivities ($<$~1~W~m$^{-1}$~K$^{-1}$), putting proposed materials among the well-known thermoelectric materials such as SnSe and other contemporary Zintl phases. We ascribe such low values to antibonding states induced weak bonding in the lattice and intrinsically weak phonon transport, resulting in low phonon velocities, short lifetimes, and considerable anharmonic scattering phase spaces. Besides, our results on electronic structure and transport properties reveal tremendous performance of SrBaGe ($ZT\sim$ 2.0 at 700~K), highlighting the relevance among state-of-the-art materials such as SnSe. Further, the similar performances for both $p$- and $n$-type dopings render these materials attractive from device fabrication perspective. We believe that our study would invite experimental investigations for realizing the true thermoelectric potential of SrBaX series.

Figures (11)

• Figure 1: Energy versus volume curves for SrBaSi (dark shade), SrBaGe (medium shade), and SrBaSn (light shade) in different space groups. The inset shows the enlarged view for orthorhombic and cubic space groups. Note that the values overlap for P$6_3$/mmc and P$6_3$mc space groups.
• Figure 2: Phase diagram for SrBaSn, where colorbar shows the formation energy (eV/atom), and orange diamonds represent phases above convex hull.
• Figure 3: Energy evolution for $ab$$initio$ molecular dynamics simulations as a function of steps for SrBaX (X = Si, Ge, Sn) at 300 and 700 K.
• Figure 4: Phonon dispersion curves and density of states (in arb. units) for (a) SrBaSi, (b) SrBaGe, and (c) SrBaSn.
• Figure 5: (a) Crystal structure of SrBaX (X = Si, Ge, Sn) in orthorhombic $Pnma$ space group and (b) sideview of the structure showing puckered hexagonal rings of Ba and X atoms along $bc$-plane, while Sr atoms are stacked along $a$-axis, (c)-(e) crystal orbital Hamilton population of SrBaSi, SrBaGe, and SrBaSn, respectively. The numerical values represent integrated crystal orbital Hamilton populations for different interaction pairs.