Primary studyCore evidenceTheory Transport

Metallic Metal-Organic Frameworks Predicted by the Combination of Machine Learning Methods and Ab Initio Calculations

He Y., Cubuk E.D., Allendorf M.D. et al. · Journal of Physical Chemistry Letters · 2018 · 4562-4569

10materials
10samples
0synthesis routes
13measurements
137results
6claims and caveats

Evidence map

Open a family to keep every result attached to its sample, method and conditions.

Author interpretations and caveats

Paraphrased for this database from the authors’ stated interpretations — never quoted verbatim — and kept separate from reported measurements.

CaveatSupport assessment: Medium

Five of the six DFT-confirmed metallic MOF structures had been synthesised in prior literature, while CdC4 was present in CoRE-MOF but the authors could not find a synthesis/characterisation report.

Caveat: The article cites prior literature but gives no experimental recipes; no synthesis_routes are emitted for this computational paper.

4-5 · Results and discussion · Text discussion · Linked to 6 structured results

CaveatSupport assessment: High

The paper predicts candidate conductive MOFs computationally but does not report first-hand electrical conductivity, thermoelectric, porosity, electrochemical, or device measurements.

Caveat: Conductivity is inferred from calculated band structures/gaps only.

2 · Introduction/results narrative · Methods/results narrative · Linked to 6 structured results

OtherSupport assessment: Medium

Feature-space overlap between inorganic materials and MOFs supports transfer learning for some regions of MOF space, but not necessarily all regions.

Caveat: Overlap assessment is qualitative/visual in the main figure, and the authors explicitly state applicability is expected only for some regions.

3-4 · Feature-space analysis · Figure 3 · Linked to 2 structured results

Phase AssignmentSupport assessment: High

The combination of ML multivoting and DFT+PBE confirms six intrinsic metallic MOF model systems with zero reported band gap.

Caveat: Metallicity is at the semilocal DFT band-theory level; no first-hand electrical transport measurements were performed.

6 · Summary · Summary paragraph · Linked to 6 structured results

Structure Property LinkSupport assessment: Medium

Cd2C8 and Mn4Re12Te16C12N12 are suggested as the most likely candidates for gapless electronic transport because of their relatively large band dispersions.

Caveat: This is a theoretical inference from band dispersion, not a measured transport result.

5 · Results and discussion · Figure 4 · Linked to 3 structured results

Transport MechanismSupport assessment: Medium

All six reported Mott constants C are below 0.25, so physics beyond DFT+PBE could open electronic band gaps or produce Mott-insulating behaviour.

Caveat: The authors describe C as a very crude estimate and approximate carrier density by number of metals per unit cell.

5 · Results and discussion · Table 2 · Linked to 6 structured results

Material identities

Names and aliases are kept exactly within the paper’s own identity model.

MaterialCompositionStructure contextSource
CdC4 (Cd2C8 model)Cd2 C8Cd nodes in a carbon-only CoRE-MOF CdC4 model · carbon-only framework connectivity as represented in CoRE-MOF entry JUTCUWunknown · Model SystemDFT-optimised face-centred cubic CdC4 model from CoRE-MOF4 · Results and discussion · Table 2
CdC4 (Cd8C32 model)Cd8 C32Cd nodes in a distinct carbon-only CoRE-MOF CdC4 model · carbon-only framework connectivity as represented in CoRE-MOF entry LEJBUXunknown · Model SystemDFT-optimised simple monoclinic CdC4 model; not DFT-confirmed metallic4 · Results and discussion · Table 2
Hg[SCN]4Co[NCS]4Co4 Hg4 C16 S16 N16Hg and Co thiocyanate/isothiocyanate coordination framework · SCN/NCS thiocyanate ligandsunknown · Model SystemDFT-optimised simple orthorhombic Co-Hg thiocyanate framework model from CoRE-MOF4 · Results and discussion · Table 2
KNd[Re4Te4(CN)12]K4 Nd4 Re16 Te16 C48 N48K, Nd and Re-Te cyanide cluster framework model · cyanide (CN)unknown · Model SystemDFT-optimised base-centred orthorhombic ML-predicted candidate; not DFT-confirmed metallic4 · Results and discussion · Table 2
Mn[Re3Te4(CN)3]Mn4 Re12 Te16 C12 N12Mn centres and Re-Te cyanide cluster units · cyanide (CN)3D · Model SystemDFT-optimised simple monoclinic metal cyanide framework model related to Prussian Blue chemistry4 · Results and discussion · Table 2
Mn2[Re6S8(CN)6]4Mn8 Re24 S32 C24 N24Mn centres coordinated by hexarhenium sulfide cyanide cluster anions · cyanide (CN)3D · Model SystemDFT-optimised simple cubic metal cyanide 3D coordination network related to Prussian Blue4 · Results and discussion · Table 2
Mn2[Re6Se8(CN)6]4Mn8 Re24 Se32 C24 N24Mn centres coordinated by hexarhenium selenide cyanide cluster anions · cyanide (CN)3D · Model SystemDFT-optimised simple orthorhombic metal cyanide 3D coordination network related to Prussian Blue4 · Results and discussion · Table 2
Mn2[Re6Te8(CN)6]4Mn8 Re24 Te32 C24 N24Mn centres coordinated by hexarhenium telluride cyanide cluster anions · cyanide (CN)3D · Model SystemDFT-optimised simple orthorhombic metal cyanide 3D coordination network related to Prussian Blue4 · Results and discussion · Table 2
Na13Fe4Sb2W18(C4O43)2Na13 Fe4 Sb2 W18 C8 O86Na, Fe, Sb and W-containing framework model · C4O43-containing formula fragment as reportedunknown · Model SystemDFT-optimised triclinic ML-predicted candidate; not DFT-confirmed metallic4 · Results and discussion · Table 2
Transfer-learning MOF screening workflowNot specifiedunknown · Model SystemPaper-level computational workflow using OQMD inorganic training data, CoRE-MOF screening structures, four ML classifiers, statistical multivoting, and DFT validation.2 · Methods/results narrative · Figure 1

Sample register

Sample form, processing state and composition status define the context for measurements.

Show 10 sample records
SampleForm and roleProcessing and geometrySource
DFT model of CdC4 (Cd2C8 model)research_0682__mat__mat_cdc4_cubicModel · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
DFT model of CdC4 (Cd8C32 model)research_0682__mat__mat_cdc4_monoModel · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
DFT model of Hg[SCN]4Co[NCS]4research_0682__mat__mat_hg_co_scnModel · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
DFT model of KNd[Re4Te4(CN)12]research_0682__mat__mat_k_nd_re_teModel · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
DFT model of Mn[Re3Te4(CN)3]research_0682__mat__mat_mn_re3te4Model · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
DFT model of Mn2[Re6S8(CN)6]4research_0682__mat__mat_mn_re_sModel · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
DFT model of Mn2[Re6Se8(CN)6]4research_0682__mat__mat_mn_re_seModel · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
DFT model of Mn2[Re6Te8(CN)6]4research_0682__mat__mat_mn_re_teModel · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
DFT model of Na13Fe4Sb2W18(C4O43)2research_0682__mat__mat_na_fe_sb_wModel · Model System · ModelCrystal structure relaxed computationally for DFT+PBE electronic-structure calculations.10 · Ab Initio Calculations · Table SIV
Paper-level ML/DFT screening workflowresearch_0682__mat__mat_screening_workflowModel · Paper Level Unspecified · ModelComputational screening and validation workflow, not a physical sample.2 · Methods/results narrative · Figure 1