The Design of Intrinsically Conductive Metal-Organic Frameworks for Thermoelectric Materials

Concludes that intrinsic conductive MOFs remain compositionally narrow and TE-limited by low S, sparse k/zT data, crystallinity challenges, and computational disagreement.

Relevance: Core · p. 29 · 4. Conclusion

Summarises the narrow set of conductive MOFs with reported TE parameters, centred mainly on 2D HXTP/HXB frameworks, and discusses Cu3(HTB)2 as the strongest but still limited benchmark.

Relevance: Core · p. 14 · 2.4. Current MOFs with Reported Thermoelectric Properties · Table 1

Introduces design strategies and computational-screening challenges, arguing that chemical band engineering and better models are central to improving TE MOFs.

Relevance: Core · p. 16 · 3. Design Strategies

Reviews electronic transport mechanisms and conductive pathways in MOFs, highlighting uncertainty about intrinsic mechanisms and the need for single-crystal measurements and better band-structure modelling.

Relevance: Core · p. 10 · 2.2.1. Transport Mechanisms

Frames thermoelectrics as waste-heat conversion materials and positions MOFs as a potential phonon-glass electron-crystal platform combining crystallinity, porosity, and tunable hybrid chemistry.

Relevance: Core · p. 2 · 1. Introduction

Discusses heteroatom choice, redox-active chelating linkers, crystallinity, linker size, and interlayer distance as competing levers for conductivity, Seebeck coefficient, and thermal conductivity.

Relevance: Core · p. 21 · 3.2. Linker Selection

States that TE-specific MOF ML is not yet established, but conductive-MOF models can already indicate descriptors such as ligand rigidity, HOMO, metal electronegativity, HBD, and pKa.

Relevance: Supporting · p. 28 · 3.4. Machine Learning · Figure 28

Covers post-synthetic alignment/compositing and morphology engineering as routes to manage anisotropic charge transport and decouple electrical and thermal conductivity.

Relevance: Supporting · p. 26 · 3.3.1. Compositing · Figures 25-26

Organises metal effects by topology, metal size, mixed valency, and bimetallic substitution, with emphasis on orbital symmetry, interlayer stacking, and environmental sensitivity.

Relevance: Core · p. 18 · 3.1. Metal Substitution

Explains why 2D planar MOFs have moved MOFs from electrical-insulator assumptions toward electronic-device and TE relevance, while warning that zT remains far below inorganic TE materials.

Relevance: Core · p. 2 · 2. Metal-Organic Frameworks as Thermoelectric Materials

Connects Seebeck coefficient to band structure, carrier concentration, and zT, and identifies low S values as a core limitation for conductive MOF thermoelectrics.

Relevance: Core · p. 12 · 2.3. Seebeck Coefficients of MOFs

Synthesises pore size, porosity, pore shape, adsorbates, density, and metal-linker acoustic mismatch as drivers of low lattice thermal conductivity in MOFs.

Relevance: Core · p. 3 · 2.1. Thermal Conductivity of Porous Materials

Charge transport mechanisms

Band-like transport is associated with delocalised carriers in continuous bands; hopping is thermally activated between localised sites.

Categories: band-like transport · hopping transport

p. 10 · 2.2.1. Transport Mechanisms · Figure 12

Intrinsically conductive pathways in MOFs

The review adopts three major design strategies for conductive MOFs and also recognises redox-mediated hopping pathways in the broader conductive-MOF literature.

Categories: through-bond · extended conjugation · through-space pi-pi stacking · redox hopping · guest-promoted redox hopping

p. 14 · 2.2.2. Electrically Conductive Pathways in MOFs · Figure 12

Current TE-tested intrinsic MOF families

The current experimentally assessed TE-MOF landscape is dominated by 2D square-planar HXTP/HXB frameworks, with very few topological or linker exceptions.

Categories: 2D planar HXTP/HXB frameworks · 3D Zn-HIB exception · PTC large-core exception

p. 14 · 2.4. Current MOFs with Reported Thermoelectric Properties · Figure 15

Design strategy classes for TE MOFs

The review's Section 3 organises optimisation around metal nodes, organic linkers, post-synthetic/morphological engineering, and data-driven screening.

Categories: metal substitution · linker selection · materials engineering · machine-learning screening

p. 16 · 3. Design Strategies

Linker selection variables

Linker design is treated as a central modular lever, but with trade-offs between conductivity, crystallinity, metallicity, Seebeck coefficient, and thermal conductivity.

Categories: heteroatom identity · redox-active chelating group · crystallinity effect · aromatic core size · interlayer-distance control

p. 22 · 3.2.3. Ligand Size

Measurement and sample-state caveats

The review repeatedly separates intrinsic transport from sample-form artefacts, including grain boundaries, contact resistance, atmosphere, and anisotropic alignment.

Categories: polycrystalline pellets · thin films · single crystals · contact resistance · grain boundaries · ambient adsorbates · crystallite alignment

p. 10 · 2.2.1. Transport Mechanisms

Metal substitution variables

Metal choice is interpreted through square-planar topology, d-orbital character, metal-linker orbital overlap, possible mixed valency, and bimetallic interpolation.

Categories: metal-node topology · metal size · mixed valency · bimetallic composition

p. 18 · 3.1. Metal Substitution

Thermoelectric performance parameters

The review uses zT = S2 sigma T / kappa to organise the TE design problem and emphasises that increasing one parameter can compromise another.

Categories: Seebeck coefficient · electrical conductivity · thermal conductivity · temperature · power factor

p. 1 · 1. Introduction · Equation 1

MOF lattice thermal-conductivity design levers

Thermal transport is organised around pore scattering, density, adsorbate effects, and interfacial thermal bottlenecks between organic linker and inorganic node.

Categories: porosity · density · pore size · pore shape · adsorbates · metal-linker acoustic mismatch

p. 9 · 2.1.4. Effect of Metal Node and Linker Acoustic Mismatch

Copper-sulfur benzene-based 2D conductive MOFs reported with the highest intrinsic MOF conductivities and the largest review-reported zT.

Conduction: Very high sigma but low S and relatively high kappa, consistent with metallic domains and sample crystallinity effects.

Representative materials: Cu3(HTB)2 · Cu3(HIB)2 as named in review text for Cu-BHT benchmark

Nodes / linkers: Cu · HTB · HXB sulfur analogue

p. 15 · 2.4. Current MOFs · Figure 16

Copper THQ framework modified by BPY pillar insertion to increase interlayer distance and reduce pi-pi stacking.

Conduction: Pillar insertion decreases electrical conductivity and increases activation barrier by disrupting through-space pi-pi interactions.

Representative materials: Cu3(THQ)2 · Cu3(THQ)2-BPY

Nodes / linkers: Cu · THQ · BPY pillar

p. 24 · 3.2.4. Interlayer Distance · Figure 24

Ni, Pd, and Pt HITP-based MOFs used to examine metal size and d-orbital dispersion effects on S and sigma.

Conduction: Larger metals are interpreted as improving metal-linker orbital overlap and increasing S magnitude, but available samples are not all high-quality crystalline materials.

Representative materials: Ni3(HITP)2 · Pd3(HITP)2 · Pt3(HITP)2

Nodes / linkers: Ni · Pd · Pt · HITP

p. 19 · 3.1.2. Metal Size

Modified Ni-HITP-like 2D MOFs with alkyl side chains used to tune interlayer distance and trade conductivity against Seebeck coefficient.

Conduction: Longer side chains increase interlayer distance, weakening through-space conduction but increasing S enough to create an optimum PF.

Representative materials: Ni3(HATI_C1)2 · Ni3(HATI_C3)2 · Ni3(HATI_C4)2

Nodes / linkers: Ni · HATI_Cx · HITP-derived linker

p. 23 · 3.2.4. Interlayer Distance · Figure 23

Electron-withdrawing modified HHTP/HHTT linkers used to improve acidity and crystallinity for single-crystal conductivity measurements.

Conduction: Improved crystallinity reduces grain-boundary artefacts and improves sigma relative to polycrystalline O-linked analogues.

Representative materials: Ni3(HHTT)2 · Cu3(HHTT)2

Nodes / linkers: Ni · Cu · HHTT · oxygen heteroatom linker

p. 22 · 3.2.2. Heteroatoms and Effect on Crystallinity

HIB-based Co, Ni, Cu, and Zn MOFs used to compare metal-node orbital effects, atmosphere sensitivity, and 3D versus 2D topology.

Conduction: Seebeck and conductivity vary strongly with metal and atmosphere; Zn-HIB is much less conductive due to the 3D structure and weaker delocalisation.

Representative materials: Co3(HIB)2 · Ni3(HIB)2 · Cu3(HIB)2 · Zn3(HIB)2

Nodes / linkers: Co · Ni · Cu · Zn · HIB · HXB nitrogen analogue

p. 18 · 3.1.1. Metal Node Topology · Figure 19

Layered square-planar MOFs based on triphenylene or benzene cores with O, S, or NH chelating heteroatoms.

Conduction: Conductivity arises from through-bond metal-linker interactions, extended conjugation in the ab plane, and through-space pi-pi pathways between layers.

Representative materials: Ni3(HITP)2 · Cu3(HTB)2 · Cu3(HHTP)2 · Ni3(HIB)2

Nodes / linkers: Ni · Cu · Co · Pd · Pt · HXTP · HXB · HITP · HHTP · HTB · HIB

p. 14 · 2.4. Current MOFs · Figure 15

Mixed-metal CoxNi3-x(HITP)2, CoxCu3-x(HITP)2, and CuxNi3-x(HITP)2 frameworks used to tune conductivity by composition.

Conduction: Conductivity changes almost linearly with higher-conductivity metal content rather than exceeding the best monometallic endpoint.

Representative materials: CoxNi3-x(HITP)2 · CoxCu3-x(HITP)2 · CuxNi3-x(HITP)2

Nodes / linkers: Co · Ni · Cu · HITP

p. 21 · 3.1.4. Bi-Metallic MOFs · Figure 20

Canonical nickel-HITP 2D conductive MOF used as a benchmark for high conductivity, band-structure disputes, defects, and single-crystal versus polycrystalline behaviour.

Conduction: Reported behaviour ranges from metallic-like single-crystal evidence to semiconducting polycrystalline film behaviour; defects and stacking strongly influence bands.

Representative materials: Ni3(HITP)2

Nodes / linkers: Ni · HITP · HXTP nitrogen analogue

p. 10 · 2.2.1. Transport Mechanisms

Zeolitic imidazolate frameworks and their glasses used as secondary evidence that crystalline porosity can lower kappa versus amorphous analogues.

Conduction: Used for thermal conductivity rather than electronic TE performance; crystalline ZIFs have lower kappa than denser glasses.

Representative materials: ZIF-4 · ZIF-8 · ZIF-62

Nodes / linkers: Zn · imidazolate

p. 3 · 2.1.1. Effect of Porosity and MOF Density · Figures 2-3

Post-synthetic alignment in composites

Align anisotropic conductive MOF crystallites in a polymer matrix with an applied field, then fix alignment by photopolymerisation.

Claimed effects: Improves both parallel and perpendicular composite conductivities by more than three orders of magnitude without doping.

Controlling variables: electric field · MOF loading · crystal alignment · polymer matrix

Representative materials: Cu3(HHTP)2 in PEGDA

Caveat: Higher MOF loading can hinder alignment; composite conductivity remains influenced by matrix pathways and loading.

p. 26 · 3.3.1. Compositing · Figures 25-26

Bimetallic composition tuning

Mix metal nodes in conductive HITP frameworks to interpolate electronic and structural properties across monometallic endpoints.

Claimed effects: Can tune conductivity continuously; potential TE benefit remains possible if S also varies favourably.

Controlling variables: metal ratio · metal distribution · synthetic selectivity · interlayer distance

Representative materials: CoxNi3-x(HITP)2 · CuxNi3-x(HITP)2

Caveat: Reported studies did not investigate S, and conductivity did not exceed the best endpoint material.

p. 21 · 3.1.4. Bi-Metallic MOFs · Figure 20

Oxidise redox-active linkers to charge-neutral conductive frameworks

Exploit deprotonation and subsequent ligand oxidation in HXTP/HXB linkers to create mixed-valence/radical linkers and remove counterions.

Claimed effects: Complete oxidation can strongly increase electrical conductivity and may convert otherwise insulating frameworks into conductive materials.

Controlling variables: linker redox state · counterion content · post-synthetic oxidant · organic radical formation

Representative materials: Ni3(HTB)2 · NiTAA-MOF

Caveat: Oxidation improves sigma but the impact on S, stability, and TE balance is not yet systematically established.

p. 21 · 3.2.1. Heteroatoms · Figure 21

Design linkers for improved crystallinity and intrinsic measurements

Use ligand functionalisation, such as electron-withdrawing groups, to improve reversible crystallisation and enable single-crystal conductivity measurements.

Claimed effects: Can greatly increase measured sigma by reducing grain boundaries and better revealing intrinsic transport.

Controlling variables: linker acidity · bond reversibility · single-crystal growth · grain-boundary resistance

Representative materials: Ni3(HHTT)2 · Cu3(HHTT)2

Caveat: More crystalline O-linked MOFs may still lag behind softer-heteroatom systems in absolute conductivity.

p. 22 · 3.2.2. Heteroatoms and Effect on Crystallinity

Tune interlayer distance to balance sigma and S

Modify side chains, stacking, or pillar linkers to control pi-pi through-space interactions and the degree of metallicity.

Claimed effects: Short distances favour sigma; expanded distances can increase S and create an optimum PF rather than maximum conductivity.

Controlling variables: interlayer distance · lateral displacement · side-chain length · pillar-linker ratio

Representative materials: Ni3(HATI_C1)2 · Ni3(HATI_C3)2 · Ni3(HATI_C4)2 · Cu3(THQ)2-BPY

Caveat: Excessive expansion can sharply reduce conductivity; not every interlayer modification changes optical bandgap measurably.

p. 26 · 3.2.4. Interlayer Distance · Figures 23-24

Use larger second- or third-row metals to improve orbital overlap

Select larger metal cations whose more diffuse d orbitals can improve metal-linker overlap and band dispersion.

Claimed effects: May increase both S and sigma and potentially reduce kappa through heavier metal-linker mass mismatch.

Controlling variables: metal row · d-orbital radial extent · metal-linker orbital overlap · crystallinity

Representative materials: Ni3(HITP)2 · Pd3(HITP)2 · Pt3(HITP)2

Caveat: Available Pd/Pt examples include amorphous or lower-quality samples; synthesis quality remains a major uncertainty.

p. 19 · 3.1.2. Metal Size

Tune metal-node topology and d-orbital symmetry

Use square-planar metal nodes and favourable d-orbital z-character to preserve 2D conjugation and through-space interactions.

Claimed effects: Can substantially alter charge-carrier pathways and electrical conductivity without nominally changing the framework topology.

Controlling variables: d-electron count · square-planar distortion · Jahn-Teller effects · interlayer displacement

Representative materials: Co3(HIB)2 · Ni3(HIB)2 · Cu3(HIB)2 · M3(HITP)2

Caveat: Metal substitution can appear isostructural while still changing buckling, stacking, atmosphere response, and S.

p. 18 · 3.1.1. Metal Node Topology · Figure 18

Use ML descriptors for conductive-MOF screening

Train models on metal, ligand, and metal-ligand overlap descriptors to prioritise conductive MOF candidates and identify design features.

Claimed effects: Can accelerate identification of conductive MOFs and provide interpretable descriptors, but TE-specific ML is not yet established.

Controlling variables: metal descriptor · ligand descriptor · overlap descriptor · database size · single-crystal experimental data

Representative materials: CuTTPD MOF

Caveat: Current datasets are small and sparse in thermoelectric labels; predictions may deviate by orders of magnitude.

p. 28 · 3.4. Machine Learning · Figure 28

Use softer linker heteroatoms for covalent metal-linker coupling

Replace hard oxygen-donor linkers with softer nitrogen or sulfur analogues to improve orbital energy matching and charge delocalisation.

Claimed effects: Raises electrical conductivity but can reduce band gap and Seebeck coefficient and may worsen crystallinity.

Controlling variables: heteroatom identity · metal-linker covalency · band gap · framework oxidation state

Representative materials: Cu3(HHTP)2 · Cu3(HITP)2 · Cu3(HTB)2

Caveat: The review warns that sulfur may maximise sigma but is not necessarily optimum for TE because S and crystallinity can suffer.

p. 21 · 3.2.1. Heteroatoms

Adsorbate effects on thermal and electronic properties need systematic evaluation before atmosphere/guest loading can be used as a TE design principle.

p. 5 · 2.1.3. Effect of Atmosphere on Thermal Conductivity

Computational models disagree on metallic versus semiconducting behaviour and need accurate crystal structures, defects, and disorder.

p. 10 · 2.2.1. Transport Mechanisms

Thermal-conductivity screening has not yet been carried out specifically for intrinsically conductive MOFs, and existing models often omit electronic thermal conductivity.

p. 9 · 2.1.4. Effect of Metal Node and Linker Acoustic Mismatch

TE-tested intrinsic conductive MOFs remain concentrated in a narrow set of 2D HXTP/HXB topologies and Co/Cu/Ni metal-linker combinations.

p. 29 · 4. Conclusion

Intrinsic electrical transport mechanisms remain poorly understood because most measurements are on polycrystalline pellets or thin films.

p. 10 · 2.2.1. Transport Mechanisms

Most intrinsically conductive MOFs have low S values, so strategies must shift from maximum sigma alone toward band structures resembling heavily doped semiconductors.

p. 13 · 2.3. Seebeck Coefficients of MOFs

There is no systematic experimental map of optimal metal-linker pairings for both high sigma and favourable S.

p. 29 · 4. Conclusion

Mixed valency and bimetallic strategies have not yet been evaluated for their impact on S and PF in a way that would establish TE benefit.

p. 21 · 3.1.4. Bi-Metallic MOFs

TE-MOF machine learning is blocked by sparse experimental data, especially single-crystal structural and intrinsic electronic measurements.

p. 28 · 3.4. Machine Learning

Very few TE-tested MOFs have thermal conductivity and zT reported, limiting comparison to established TE materials.

p. 13 · 2.4. Current MOFs · Table 1