High-throughput screening and mechanistic insights into solid acid proton conductors

Abstract

Proton-conducting solid acids could enable water-free operation of high-temperature fuel cells. However, systematic materials screening has, hitherto, been computationally prohibitive. Here, we introduce a two-stage high-throughput screening strategy that directly computes proton diffusion coefficients, enabled by machine-learned interatomic potentials fine-tuned to ab initio data. Starting from more than six million materials, our screening -- based on structural motifs rather than empirical descriptors -- identifies $27$ high-performing proton conductors, including over ten previously unexplored compounds. These include sustainable and commercially available materials, candidates that have not yet been synthesized, organic systems that fall outside conventional design rules, and known proton conductors that validate our approach. Importantly, our findings reveal a universal oxygen--oxygen distance of approximately $2.5$~Å at the moment of proton transfer across diverse chemistries, providing mechanistic insight and showing that macroscopic proton conductivity emerges from the interplay between anion rotational dynamics, hydrogen-bond network connectivity, and proton-transfer probability.

Figures (4)

• Figure 1: Structural solid acid motifs used for screening materials databases. Schematic representation of the two complementary proton coordination motifs derived from the solid acids a CsH$_2$PO$_4$ and b Cs$_7$(H$_4$PO$_4$)(H$_2$PO$_4$)$_8$struct_cdpstruct_cpp. $N_\mathrm{A}(\mathrm{B};r_c)$ denotes the number of atoms of type A around an atom of type B within a radial cutoff $r_c$.
• Figure 2: Overview of the high-throughput screening results obtained using the MatterSim foundation model. Parameters for the MD simulations are given in the Methods section. a Elemental composition of the filtered dataset. Colors encode how often each element appears among the filtered structures, with the exact counts printed below the corresponding element symbols. b--d Dynamical descriptors extracted from $100$ ps MD simulations (see Methods). The proton diffusion coefficients $D_\mathrm{H}$ (b), the X--O bond vector autocorrelation values $C_\mathrm{XO}$ at $\tau = 100$ ps (c), and the number of proton transfer events per hydrogen atom and per picosecond (d) are shown for each oxoanion center atom type. Note that proton jumps were evaluated only every $100$th frame to reduce storage, so these values differ from those computed from the trajectories generated in the second stage of this investigation.
• Figure 3: Proton transport analysis based on long-time MD simulations using fine-tuned MLIPs. From top to bottom: proton diffusion coefficients, oxoanion rotation dynamics, frequency of proton transfer events per hydrogen atom and per picosecond, and XO$_\mathrm{y}$ oscillation periods. Proton diffusion, oxoanion rotation, and proton-transfer frequencies are obtained from $400$ K $3$ ns MD trajectories generated with MLIPs fine-tuned to $40$ ps AIMD data for each material. XO$_\mathrm{y}$ oscillation periods are calculated from the $40$ ps AIMD trajectories, see Eq. (\ref{['eq:xoy_acf']}) below.
• Figure 4: Structural origin of proton transfer.a Oxygen--oxygen radial distribution functions (RDF, top) and distribution of proton transfer events as a function of instantaneous oxygen--oxygen distance (bottom) obtained from the $40$ ps AIMD trajectories. b Comparison of the proton transfer distance distributions obtained from AIMD and PIMD simulations for the outlier materials containing the (TeO$_6$)-group: Rb$_2$H$_6$(SO$_4$)(TeO$_6$) (top) and Tl$_2$H$_6$(SeO$_4$)(TeO$_6$) (bottom).