Effect of surface termination on ion intercalation selectivity of bilayer Ti₃C₂T₂ (T = F, O and OH) MXene
Summary¶
MXene electrodes and membranes exploit interlayer galleries that can swell or contract when ions and solvents intercalate, so predicting how surface chemistry steers interlayer spacing is central to selectivity and transport modeling. Berdiyorov and Mahmoud study bilayer Ti₃C₂T₂ with T = F, O, or OH using density functional theory for intercalation energetics and ReaxFF molecular dynamics to capture finite-temperature interlayer distance response. They compare pristine bilayers to functionalized bilayers when anions and cations are inserted, focusing on how termination reverses qualitative trends in gallery expansion versus contraction. The abstract frames the problem as intercalation selectivity: functionalization changes whether gallery opening tracks anions similarly to cations, which matters when designing MXene stacks for ion sieving or pseudocapacitive charge storage.
Methods¶
Static QM (DFT). PBE GGA exchange–correlation with Grimme D2-type dispersion on PBE; calculations use Atomistix ToolKit with an 8×8×8 k-mesh, 150 Ry real-space mesh cutoff, DZP numerical orbitals, and convergence thresholds 10⁻⁴ eV (energy) and 0.01 eV/Å (forces). Supercells are bilayer Ti₃C₂T₂ with T = none, F, O, OH, >10 Å vacuum along the layer normal, and relaxed in-plane dimensions on the order of a ≈ 3.145 Å, b ≈ 5.457 Å, c ≈ 20 Å (reported relaxed values). In-plane PBC apply; the stack is isolated from its periodic image by vacuum in the normal direction. Intercalants include Cl⁻, Na⁺, K⁺, Li⁺, Mg²⁺, Ca²⁺, Al³⁺, and As⁵⁺, with high formal charges treated via the article’s compensation-charge scheme.
DFT-based MD runs use the NVT ensemble, Δt = 0.5 fs, and a Nosé–Hoover chain thermostat (Appl. Surf. Sci. §2). Barostat / stress control for those DFT-MD segments is not emphasized on the pages indexed for this note (canonical NVT thermal sampling of d(t) at fixed cell).
MD application (ReaxFF). ReaxFF employs a Ti₃C₂(OH)₂ / water / K⁺-oriented parameterization from prior work (cited in the article). The minimal bilayer gallery places one ion and one water between layers (Fig. 1 schematic in the paper). The abstract reports that finite-temperature ReaxFF MD reproduces the qualitative sign change in interlayer spacing vs intercalant type seen in DFT-MD after surface functionalization. ReaxFF run settings (code, timestep, target T, duration, barostat, E-field, enhanced sampling) appear in Appl. Surf. Sci. §2–3 and are not transcribed on this wiki page.
Force-field training is N/A: an existing ReaxFF set is applied, not refitted here.
MD blueprint honesty. DFT-MD details above include NVT, timestep, and Nosé–Hoover thermostat language from the indexed Appl. Surf. Sci. text. ReaxFF MD trajectories are expected to use PBC bilayer cells analogous to the DFT setup; LAMMPS (or the code named in the paper) should be confirmed in §2–3. Production duration (ps/ns), any NPT/barostat segment for ReaxFF, target pressure, and electric-field/enhanced-sampling knobs are N/A on this page unless copied from the PDF.
Findings¶
Outcomes and mechanism readout (authors’ framing). For pristine bilayer Ti₃C₂, the minimum-energy interlayer spacing trends toward larger d when either anions or cations are intercalated (together with a water molecule in the gallery model of Fig. 1). After surface functionalization (F, O, OH terminations), the response reverses qualitatively: d increases with anion insertion but contracts under cation insertion, with stronger contraction as the cation charge state increases—most pronounced for Ti₃C₂O₂ in the abstract’s summary.
Comparisons. The study contrasts pristine vs functionalized bilayers across the same ion set, and reports that the sign change in d vs intercalant type is reproduced in both DFT-based MD and ReaxFF MD (abstract), i.e. finite-temperature modeling agrees with the static DFT E(d) picture on this qualitative point.
Sensitivity and design levers. Termination chemistry (T) and ion formal charge are the dominant knobs in the reported trends; injecting extra electrons to the layers is also discussed in the article as amplifying contraction effects (see Appl. Surf. Sci. discussion).
Limitations and outlook (as implied by scope). The bilayer + single ion + one water construct is intentionally minimal; multilayer stacks, concentrated electrolytes, and long-time ion transport are outside what the abstract claims. Force-field transferability for highly charged intercalants should be checked against experiment when used for engineering estimates.
Quantitative E(d) curves, d_min tables, and time series d(t) are in the peer-reviewed PDF and figures.
Limitations¶
Solvent-explicit electrochemical environments and long-time diffusion statistics require larger models and may expose force-field limitations outside the fitted ion–surface chemistry.
Relevance to group¶
Pairs DFT with ReaxFF for MXene ion intercalation—useful cross-reference for 2D electrochemical materials modeling.
Citations and evidence anchors¶
- DOI:
10.1016/j.apsusc.2017.04.195.