MOFs that act as magnets, wires, proton conductors and spin switches

Frameworks can be built to carry magnetism, electrons, protons or spin states

Intuition

A MOF is a scaffold you can furnish. Put the right metals and ligands on it and the scaffold becomes a magnet. Put in orbitals that overlap and it becomes a wire. Put acid groups on its walls and protons hop through it. Put in a metal that can flip its spin and a guest molecule can flip the switch.

The lecture lists current and future applications of MOFs: 1) separations; 2) catalysis; 3) storage; 4) sensing; 5) data storage; 6) magnets; 7) molecular switches or computers; 8) mechanical properties; 9) stereoregular polymerisation; 10) molecular hosts; 11) controlled release; 12) optical properties; 13) conductors or wires; 14) light-harvesting. Some materials combine several. This topic covers the magnetic, electronic, proton-conducting and spin-switching ones.

Magnetic order can live in a porous framework

Atoms with unpaired spins in a crystal can have long-range magnetic order, and most solid-state magnet applications use ordered states, not paramagnetic or diamagnetic ones. Ferromagnetic (FM) means all spins parallel. Antiferromagnetic (AFM) means alternating spins. Even a cubic cell with equal parallel spins can be ordered as FM or as A-type, C-type, E-type or G-type AFM.

The slide’s magnetic MOF is $Cu_3(PTMTC)_2(py)_6(EtOH)_2(H_2O)$, where PTMTC is a polychlorinated triphenylmethane tricarboxylate. The ligand is a stable radical with $S = \tfrac12$. The metal is Cu(II), d⁹, with one unpaired electron, so $S = \tfrac12$ as well. Metal and ligand couple antiferromagnetically, but the antiparallel moments do not cancel because of the stoichiometry. The result is a ferrimagnetic material with a net magnetisation. The structure is stacked hexagonal grids with 30 Å pores perpendicular to the layers. The magnetic properties change slightly when guests are removed, so applications include molecular sensors and separating magnetic from non-magnetic materials. The slide plots $\chi T$ against $T$ for empty and solvent-filled forms. The values are read from the plot and the slide gives no equation, so only the trend is stated here.

Conduction needs delocalised electrons, and MOF design can supply them

Intuition

A wire needs electrons that are free to move. In a crystal that means electrons spread over many atoms through overlapping orbitals, in a partly filled band. So the way to make a conducting MOF is to arrange for overlap.

The band lectures set this up. ReO₃ and WO₃ have the same crystal structure but different electronic structure. In ReO₃ the Re(VI) has one 5d electron, which delocalises through the band made from π-type overlap of $t_{2g}$ orbitals on the metal with oxygen p orbitals, so it is metallic. W(VI) has no 5d electrons, so the band is empty and WO₃ is an insulator.

TTF-TCNQ, “the first organic metal” (1973), combines two insulating molecules into 1D chains. Orbital overlap and charge transfer from electron donor to acceptor create a partly filled band and metal-like 1D conduction. The composition is $(TTF)^{\delta+}(TCNQ)^{\delta-}$.

Electron-conducting MOFs. $Ni_3(HITP)_2$ (HITP is hexaiminotriphenylene) has electron delocalisation over large aromatic ligands, with metal and ligand orbitals overlapping through the framework. Uses are molecular sensors (a guest would change the conductivity) and electrocatalysis (redox reactions inside the pores). In $Zn_2(TTFTB)$, the pores are filled with layers of the donor TTF. Charge delocalisation through overlapping bands of neighbouring TTF gives 1D conduction through the structure.

Proton-conducting MOFs. Surface sulfonate groups give high proton conductivity:

$\sigma = 1.27 \times 10^{-1}\ \text{S cm}^{-1} \quad \text{at 100\% RH and 80 °C}$

RH is relative humidity.

A spin switch flips between high spin and low spin

A material is a spin switch if a stimulus such as light, temperature or pressure changes the intrinsic spin of its electrons. The lecture’s example uses Fe(II), d⁶, in an octahedral field. In a weak ligand field ($\Delta < P$, with $\Delta$ the ligand-field splitting and $P$ the spin-pairing energy) the electrons spread out as $(t_{2g})^4(e_g)^2$. This is high spin. In a strong field ($\Delta > P$) all six pair in $t_{2g}$ as $(t_{2g})^6$. This is low spin.

The MOF is $[Fe(azpy)_2(NCS)_2]\cdot\tfrac12(EtOH)$. The Fe centre has four pyridyl N donors and two NCS ligands. The structure changes with whether gas or solvent molecules, such as ethanol, are in the pores. The desorbed material shows no spin switching. Its magnetic moment stays near 5.2 to 5.3 B.M. across 20 to 300 K. Materials with MeOH, EtOH or 1-PrOH absorbed do switch, with behaviour that depends on the guest. For these, the moment sits near 3.7 to 4.1 B.M. at 20 to 30 K and rises to about 5.2 B.M. by 150 K, with different curve shapes for each guest. All these values are read from the plot. Because the response depends on the guest, the material works as a chemical sensor.

Compare

High spin Fe(II)

Ligand field splitting smaller than pairing energy ($\Delta < P$). Configuration $(t_{2g})^4(e_g)^2$.

Low spin Fe(II)

Ligand field splitting larger than pairing energy ($\Delta > P$). Configuration $(t_{2g})^6$.

Exam detail

The learning outcomes ask you to understand how magnetic ordering, electron delocalisation, proton conduction and spin switching can be built into MOFs, and why that matters. One example for each (the Cu radical MOF, Ni₃(HITP)₂ or TTF-TCNQ, the sulfonated MOF, and the Fe(azpy) switch) covers it.

Recall

Why does the desorbed Fe(azpy) MOF not switch, while the one with ethanol does?

The slide says the structure changes depending on whether guest molecules are in the pores, and the structure influences the magnetic properties. With no guest the structure does not support spin switching. With a guest present the Fe(II) can move between spin states, and how it does so depends on which guest it is.