Mn4Al11: Spin-Selective Magnetic Conductor

• Mn4Al11 is a stoichiometric binary intermetallic compound with triclinic P̅1 symmetry, characterized by a one-dimensional Mn zigzag chain and spin-selective ferrimagnetism.
• Combined single-crystal experiments and spin-polarized DFT reveal a magnetic transition near 68 K, low carrier density, and anomalous optical and transport behavior.
• Contrasting DFT studies classify it as either a half-semimetal candidate or a ferrimagnetic half-metal, highlighting its potential for spintronic applications and further investigation.

Mn$_4$Al$_{11}$ is a stoichiometric binary intermetallic compound of Mn and Al that crystallizes in triclinic $P\bar{1}$ symmetry and has emerged in 2025 as a rare spin-selective magnetic conductor. Combined single-crystal experiments and spin-polarized electronic-structure calculations identify a magnetic phase transition near $68\ \mathrm{K}$, a very low-carrier-density metallic ground state, and a ferrimagnetic state with anomalous optical and transport behavior. In the contemporary literature, it is described both as a half-semimetal candidate and as a ferrimagnetic half-metal, reflecting different emphases in experimental interpretation and first-principles classification (Han et al., 16 Aug 2025, Datta et al., 7 Jul 2025).

1. Crystal chemistry and structural motifs

Single-crystal X-ray diffraction at $92\ \mathrm{K}$ determines Mn$_4$Al$_{11}$ to be centrosymmetric triclinic $P\bar{1}$, with two symmetry-inequivalent Mn sites and six inequivalent Al sites. The refined crystallographic information file was deposited with CCDC under accession number 2414038. High-quality single crystals were grown by a Sn-flux method from starting materials Mn (99.99%), Al (99.99%), and Sn (99.99%) with molar ratio Mn:Al:Sn $= 3:10:20$, heated to $1000\ ^\circ\mathrm{C}$ with a 15 h ramp, held for 50 h, and cooled to $_{11}$0 at $_{11}$1 before centrifugation; the resulting crystals are shiny and plate-like. The compound is also listed as Materials Project mp-2856, and the experimental study does not report decomposition or phase instability over 2–300 K (Han et al., 16 Aug 2025).

The characteristic structural motif is a one-dimensional Mn zigzag chain running along the $_{11}$2 axis. Two Mn(1) and two Mn(2) atoms overlap along $_{11}$3, with reported nearest-neighbor distances Mn(1)–Mn(1) $_{11}$4, Mn(1)–Mn(2) $_{11}$5, and Mn(2)–Mn(2) $_{11}$6. The spacing between adjacent chains is comparatively large, implying weak interchain interactions. A separate first-principles study gives optimized lattice parameters $_{11}$7, $_{11}$8, $_{11}$9, $P\bar{1}$0, $P\bar{1}$1, and $P\bar{1}$2, compared with experimental values $P\bar{1}$3, $P\bar{1}$4, $P\bar{1}$5, $P\bar{1}$6, $P\bar{1}$7, and $P\bar{1}$8. In that description, both Mn1 and Mn2 occupy bicapped square antiprismatic environments with ten Al neighbors and Mn–Al nearest-neighbor distances spanning 2.37–2.75 $P\bar{1}$9 (Datta et al., 7 Jul 2025).

These structural details are central because the magnetic and electronic anomalies are tied to site inequivalence, Mn–Mn chain geometry, and Mn–Al hybridization. A plausible implication is that the one-dimensional zigzag-chain motif provides the geometric setting for the antiferromagnetic-type correlations and anisotropic transport scenarios discussed for the compound.

2. Magnetic order, thermodynamics, and anisotropy

DC magnetic susceptibility $68\ \mathrm{K}$0 was measured with $68\ \mathrm{K}$1 applied along the $68\ \mathrm{K}$2 axis under zero-field-cooled warming from 2 to 300 K, and isothermal magnetization $68\ \mathrm{K}$3 was measured at 10 K for $68\ \mathrm{K}$4 from $68\ \mathrm{K}$5 to $68\ \mathrm{K}$6. A magnetic phase transition occurs at $68\ \mathrm{K}$7, visible in both $68\ \mathrm{K}$8 and transport. Below $68\ \mathrm{K}$9, the ground state is weakly ferrimagnetic, supported experimentally by the appearance of a hysteresis loop at 10 K. Above about 100 K, $92\ \mathrm{K}$0 increases upon cooling and then decreases down to about 50 K, which was attributed to enhanced antiferromagnetic-type correlations within the Mn(1)–Mn(2) zigzag chains. A Curie–Weiss fit of $92\ \mathrm{K}$1 above 200 K using $92\ \mathrm{K}$2 yields $92\ \mathrm{K}$3, $92\ \mathrm{K}$4, and $92\ \mathrm{K}$5 per formula unit, establishing predominantly antiferromagnetic interactions without resolving the detailed spin configuration (Han et al., 16 Aug 2025).

Specific heat measurements provide a thermodynamic signature of the same transition. $92\ \mathrm{K}$6 shows a weak anomaly at about 68 K. At low temperature, $92\ \mathrm{K}$7 versus $92\ \mathrm{K}$8 is linear, and a Sommerfeld fit,

$92\ \mathrm{K}$9

gives $_4$0 and $_4$1. The relatively large $_4$2 indicates a finite electronic density of states at $_4$3 and suggests correlation effects from Mn $_4$4 electrons (Han et al., 16 Aug 2025).

First-principles calculations describe the ferrimagnetic state in more microscopic terms. One study reports local moments Mn1 $_4$5, Mn2 $_4$6, and total magnetization $_4$7 per unit cell, with Al sites essentially non-magnetic. Including SOC for magnetic anisotropy yields a maximum MAE of $_4$8 at polar angles $_4$9, $_{11}$0, and the magnetic easy axis at $_{11}$1, $_{11}$2, nearly perpendicular to a slightly buckled Mn layer. The same calculation reports no significant SOC effects and no indication of non-collinear magnetism (Datta et al., 7 Jul 2025).

Experimentally, quantitative net moment per formula unit, coercivity, and anisotropic comparison were not provided, because only $_{11}$3 was measured. This leaves the detailed magnetic structure unresolved; neutron diffraction or spin-resolved probes were explicitly identified as necessary.

3. Transport and optical response

In-plane dc resistivity $_{11}$4 was measured in a PPMS with a standard four-probe configuration, with current flowing in the $_{11}$5 plane. From 300 K to low temperature, $_{11}$6 rises with cooling, reaches a peak around $_{11}$7, and then drops to a minimum near $_{11}$8. The derivative $_{11}$9 shows a sharp feature at $P\bar{1}$0, coincident with the magnetic anomaly. Above about 100 K, the negative temperature coefficient resembles semiconducting behavior, but the combined optical and specific-heat data establish metallicity with very low carrier density and strong scattering rather than a hard semiconductor gap (Han et al., 16 Aug 2025).

Optical spectroscopy was performed by measuring in-plane near-normal-incidence reflectance $P\bar{1}$1 from 40 to $P\bar{1}$2 using a Bruker 80V FTIR, with in situ gold overcoating at low frequency and aluminum reference above $P\bar{1}$3. Kramers–Kronig transformations were used to derive $P\bar{1}$4, with Hagen–Rubens low-frequency extrapolation and x-ray atomic scattering functions at high frequency. At 300 K, $P\bar{1}$5 shows a low-frequency upturn indicative of metallicity, but its absolute value is much smaller than in conventional metals, consistent with a low carrier density. Upon cooling, the low-frequency reflectance decreases, while the response above $P\bar{1}$6 is nearly temperature-independent. Numerous phonon absorptions appear below about $P\bar{1}$7, and they grow in strength at low temperature due to weak electronic screening (Han et al., 16 Aug 2025).

After subtracting phonon contributions with Lorentzian fits, the electronic background remains finite at low frequency, demonstrating that bands cross $P\bar{1}$8. Two features are particularly important: the low-frequency $P\bar{1}$9 is nearly flat rather than Drude-like, and its magnitude decreases upon cooling from 300 K to about 50 K. The spectral-weight analysis uses

$= 3:10:20$0

and the reduction of low-frequency spectral weight with decreasing temperature was interpreted as evidence for a decreasing carrier density and substantial carrier scattering. The likely scattering source proposed in the experimental study is spin-disorder scattering from localized Mn $= 3:10:20$1 moments along the zigzag chains; when moments order at $= 3:10:20$2, spin-disorder diminishes and $= 3:10:20$3 drops sharply below the transition (Han et al., 16 Aug 2025).

The transport characterization remains incomplete in several standard respects. Magnetoresistance and Hall effect were not reported, so explicit $= 3:10:20$4, $= 3:10:20$5, and $= 3:10:20$6 are not available. The paper notes the single-band relations

$= 3:10:20$7

but also states that strong phonon modes and flat low-frequency $= 3:10:20$8 prevent reliable extraction of $= 3:10:20$9 and the scattering rate from a clean Drude analysis.

4. Electronic structure and the half-semimetal versus half-metal classification

Spin-polarized DFT calculations in the combined experimental-theoretical study were carried out with VASP, PAW pseudopotentials, the PBE-GGA functional, a plane-wave cutoff of 500 eV, and a $1000\ ^\circ\mathrm{C}$0 Monkhorst–Pack mesh, with structures relaxed to $1000\ ^\circ\mathrm{C}$1 force per atom and without SOC or Hubbard $1000\ ^\circ\mathrm{C}$2. That study reports a spin-resolved electronic structure in which the spin-up channel is metallic, with both electron and hole pockets crossing $1000\ ^\circ\mathrm{C}$3, while the spin-down channel is semiconducting near $1000\ ^\circ\mathrm{C}$4. The total DOS at $1000\ ^\circ\mathrm{C}$5 is low, and the states near $1000\ ^\circ\mathrm{C}$6 are dominated by Mn $1000\ ^\circ\mathrm{C}$7-derived bands hybridized with Al $1000\ ^\circ\mathrm{C}$8 states. On that basis, together with the small and nearly frequency-independent far-infrared conductivity and the metallic but low-carrier-density ground state, Mn$1000\ ^\circ\mathrm{C}$9Al$_{11}$00 is proposed as a half-semimetal candidate (Han et al., 16 Aug 2025).

A separate first-principles study uses spin-polarized GGA-PBEsol with PAW in Quantum ESPRESSO, a $_{11}$01 $_{11}$02-mesh, Marzari–Vanderbilt cold smearing of width 0.02 Ry, plane-wave cutoffs of 55 Ry and 400 Ry, and no Hubbard $_{11}$03. In that description, pristine Mn$_{11}$04Al$_{11}$05 is a ferrimagnetic half-metal: the down-spin channel is metallic, with a single isolated band crossing $_{11}$06, while the up-spin channel hosts a gap of about $_{11}$07 at $_{11}$08. The metallic down-spin band is primarily Mn2-$_{11}$09 in character; in the conduction-band region it is separated from other bands by about $_{11}$10, more specifically 0.33 eV at $_{11}$11 and 0.41 eV at $_{11}$12 relative to $_{11}$13, and an electron pocket is present along $_{11}$14–$_{11}$15 near $_{11}$16. Using

$_{11}$17

the reported values $_{11}$18 and $_{11}$19 states/cell give $_{11}$20, corresponding to 100% spin polarization in the metallic down-spin channel (Datta et al., 7 Jul 2025).

The microscopic explanation for this spin selectivity also differs in emphasis. The experimental paper discusses $_{11}$21–$_{11}$22 hybridization and a pseudogap tendency that depletes Mn $_{11}$23 DOS at $_{11}$24 in Al–Mn systems. The theoretical tuning study identifies spin-asymmetric Mn2–Al hybridization from COHP analysis as the origin of the half-metallic gap: Mn1–Al bonding near $_{11}$25 is almost spin-independent, whereas Mn2–Al is bonding in the up-spin channel and strongly anti-bonding in the down-spin channel near $_{11}$26, suppressing up-spin states at $_{11}$27 while permitting metallic down-spin conduction (Han et al., 16 Aug 2025, Datta et al., 7 Jul 2025).

The coexistence of these two classifications is one of the main interpretive issues in the 2025 literature. One work defines Mn$_{11}$28Al$_{11}$29 as a half-semimetal candidate because the metallic spin channel has very low carrier density and semimetal-like spectral weight behavior, while the other treats pristine Mn$_{11}$30Al$_{11}$31 as a ferrimagnetic half-metal with a finite half-metallic gap. This suggests that the low-DOS, spin-selective state in Mn$_{11}$32Al$_{11}$33 lies close to the boundary between half-metallic and half-semimetallic descriptions.

5. Strain, pressure, and site-selective Ge substitution

The most extensive tuning study examines uniaxial strain along the $_{11}$34 axis, isotropic pressure, and Ge substitution at all six inequivalent Al sites. Under uniaxial strain $_{11}$35 from roughly $_{11}$36 to $_{11}$37, the half-metallic band gap collapses for compressive strain beyond $_{11}$38. The down-spin band structure progressively develops additional crossings at $_{11}$39, destroying half-metallicity while retaining ferrimagnetism. Magnetization varies approximately parabolically with $_{11}$40, reaching a minimum at $_{11}$41 and a maximum at $_{11}$42. Under hydrostatic compression, the half-metal-to-metal transition occurs at $_{11}$43: at 10 GPa, a down-spin band at $_{11}$44 crosses $_{11}$45, and at higher pressures additional crossings appear at $_{11}$46, $_{11}$47, and along $_{11}$48–$_{11}$49. The total magnetization decreases from $_{11}$50 at 0 GPa to $_{11}$51 at 15 GPa, but ferrimagnetic order persists to at least 15 GPa (Datta et al., 7 Jul 2025).

Ge substitution is strongly site selective. To preserve inversion symmetry, off-center substitutions at Al1–Al5 are introduced in inversion pairs, producing Mn$_{11}$52Al$_{11}$53Ge$_{11}$54-1 through Mn$_{11}$55Al$_{11}$56Ge$_{11}$57-5, while substitution at the inversion-center site Al6 $_{11}$58 yields Mn$_{11}$59Al$_{11}$60Ge. The nominal Ge concentrations are $_{11}$61 for Mn$_{11}$62Al$_{11}$63Ge$_{11}$64-$_{11}$65 and $_{11}$66 for Mn$_{11}$67Al$_{11}$68Ge. Formation-energy analysis indicates that Ge-substituted structures are thermodynamically more favorable than analogous Si-doped ones, a trend attributed to larger Ge size and stronger Mn–Ge hybridization (Datta et al., 7 Jul 2025).

All five off-center Ge substitutions destroy the pristine half-metallicity and produce metallic states. In Mn$_{11}$69Al$_{11}$70Ge$_{11}$71-1, -2, -3, -4, and -5, the band formerly isolated at $_{11}$72 shifts into the valence manifold and loses its isolation; DOS at $_{11}$73 increase, with sharper bands indicating increased localization. The reported spin polarizations at $_{11}$74 are $_{11}$75, 0.46, 0.08, 0.40, and 0.66, respectively. Their total magnetizations are 0.22, 0.42, 0.08, 0.49, and $_{11}$76cell, with local moments remaining site differentiated and Mn2 $_{11}$77 Mn1. Among them, the -5 structure shows the largest magnetization and a sizable MAE of about $_{11}$78 at $_{11}$79, $_{11}$80 (Datta et al., 7 Jul 2025).

A distinct outcome occurs for Ge at Al6 $_{11}$81. Mn$_{11}$82Al$_{11}$83Ge undergoes a metal-to-insulator transition and becomes non-magnetic and semiconducting with an indirect gap of $_{11}$84, with VBM at $_{11}$85 and CBM at $_{11}$86. The reported net moment is $_{11}$87cell, with Mn1 $_{11}$88 and Mn2 $_{11}$89, and Bader analysis gives equal charges for Mn1 and Mn2, both $_{11}$90. COHP analysis identifies the Mn1–Ge bond as strongest among the doped structures, with ICOHP at $_{11}$91, while the Mn1–Mn2 direct COHP becomes negligible because Mn1 and Mn2 are bridged by Ge and the dominant interaction is superexchange via Mn–Ge–Mn. This is presented as a symmetry-controlled metal–insulator transition (Datta et al., 7 Jul 2025).

Phonon calculations add a separate qualification about structural stability. Pristine Mn$_{11}$92Al$_{11}$93 shows imaginary modes, interpreted as metastability consistent with experimental reports of mixed or metastable phases. Ge alloying removes these instabilities for Mn$_{11}$94Al$_{11}$95Ge$_{11}$96-1, -2, -3, -4, and Mn$_{11}$97Al$_{11}$98Ge, while Mn$_{11}$99Al$P\bar{1}$00Ge$P\bar{1}$01-5 retains negative modes (Datta et al., 7 Jul 2025).

6. Scientific significance, related systems, and unresolved questions

Mn$P\bar{1}$02Al$P\bar{1}$03 is significant because spin-selective intermetallics with both low density of states at $P\bar{1}$04 and ferrimagnetism are rare. The experimental study explicitly compares its low spectral weight and temperature-dependent optical conductivity to low-carrier-density metals such as 1T-TiSe$P\bar{1}$05 and ZrTe$P\bar{1}$06, and to pseudogap-like behavior in YBa$P\bar{1}$07Cu$P\bar{1}$08O$P\bar{1}$09. Among reported half-semimetals, HgCr$P\bar{1}$10Se$P\bar{1}$11 is cited as a comparator. The combination of ferrimagnetism and spin-selective transport implies possible use in spin filters, spin valves, and sources of highly spin-polarized currents, while the one-dimensional Mn chain motif suggests possible anisotropic spin transport along $P\bar{1}$12, although this has not yet been quantified (Han et al., 16 Aug 2025).

Several limitations remain explicit. No Hall data or magnetoresistance were reported in the experimental study, so $P\bar{1}$13, $P\bar{1}$14, and $P\bar{1}$15 remain undetermined. Neutron diffraction is needed to resolve Mn site moments, sublattice magnetizations, and the detailed ferrimagnetic arrangement. ARPES would test the predicted spin-resolved semimetallicity or half-metallicity, while spin-resolved ARPES or point-contact Andreev reflection could quantify transport spin polarization. Improved optical modeling with more robust phonon subtraction could enable extraction of $P\bar{1}$16 and the scattering rate, and the relatively large Sommerfeld coefficient motivates further study of correlation effects, including the role of Hubbard $P\bar{1}$17, spin–orbit coupling, and low-temperature quantum oscillations (Han et al., 16 Aug 2025).

The main conceptual caution is that pristine Mn$P\bar{1}$18Al$P\bar{1}$19 is not yet a settled case of a single electronic classification. One 2025 study proposes a half-semimetal candidate grounded in transport, optical spectroscopy, and DFT, whereas another reports a ferrimagnetic half-metal with a $P\bar{1}$20 spin gap and $P\bar{1}$21 from first-principles analysis. The available evidence consistently supports spin-selective ferrimagnetic conduction with a low DOS at $P\bar{1}$22, but the precise designation—half-semimetal candidate or half-metal—remains tied to how the low-carrier metallic channel is interpreted and to whether future measurements validate the predicted spin-resolved Fermiology (Han et al., 16 Aug 2025, Datta et al., 7 Jul 2025).