The Design of Molecular Magnets

• Ashlea Hughes

1. Polyoxometalates

Polyoxometalates (POMs) are clusters consisting of linked metal oxide polyhedral. They are commonly anionic, although some cationic species have been reported.¹ Due to this charge, counter ions are often associated with POMs. The metal ions which form the POMs often belong to group 5 or group 6 in the periodic table and possess a high oxidation state causing an electron configuration of either d⁰ or d¹. The nuclearity of POMs differs widely from single monomeric species to high nuclearity (i.e. over 100 metal ions) species. The three major transition metals that form POMs are: vanadium, molybdenum and tungsten.

1.1 Structures

The building blocks of polyoxometalates often correspond to tetrahedral, tetragonal pyramids, and octahedra, sharing corners and edges which results in a variety of caged, ribboned, wheel or basket like structures e.g. [P₆Mo₁₈O₇₃]^11-.² It is this variety in potential structures of POMs that causes a number of complex molecular spin arrays to be available.

Figure 1: Five Baker-Figgis isomers adapted from Zhou et al. ³

The keggin structure, named after J. J. Keggin who determined its structure, is a very common structure found in POMs.⁴ It consists of a central tetrahedrally coordinated atom, caged by 12 octahedrally coordinated atoms. Five rotational isomers, known as Baker-Figgis isomers, of the keggin structure exist, and are shown in Figure 1, as do defect structures known as lacunary structures.³ Due to the conformation of these structures they have high stability and so aid in the exchange coupling between other keggin molecules. The POMs comprising of the transition metal, tungsten, all favour this structure and larger tungsten POMs can be considered to be made up of keggin subunits. The Dawson structure and Anderson structure are similar structures that have also been reported many times within literature.

The wheel structure is yet again another commonly found structure and is desired for POM formation. The cavity/cavities the wheel contains are accessible and allow the coordination of other clusters, and form chains in a step by step growth, forming very large POMs.

1.2 Traditional synthesis

POM based clusters are traditionally made via a one-pot synthesis. The aggregation and condensation of the reagents is controlled via experimental variables.⁵ These variables can include: concentration, pH, temperature, ligand effects, counter ions and various other experimental variables. The synthesis usually begins with the acidification of the metal salt, sequentially followed by condensation of the molecules forming a variety of architectural structures.

1.3 Uses

Polyoxometalates possess a wide variety of properties. Their diversity means that they are used for many applications. They are used as catalysts in a range of organic reactions. The acidity and solubility of POMs make them ideal candidates to use as homogeneous catalysts, as well as their redox abilities. They have many uses in biological systems, in particular the POM ferritin, which is the protein responsible for iron storage and is a cofactor in the photosynthesis process within plants. This property in particular has inspired research into many energy and photosensitive related devices using POMs. In addition there is also research being done into medicinal applications as potential uses in antiretroviral drugs for diseases such as AIDS. ⁶ Aside from all these applications, and many others, certain POMs also bear magnetic properties that I shall discuss herein.⁵

2. Magnetism

Magnetic properties arise from the spin and orbital angular momentum of unpaired electrons. Paramagnetism is a property of compounds which contain unpaired electrons. These compounds have a relative permeability greater than one and consequently attract magnetic fields. When a field is applied the spins of electrons start to align parallel to each other. The magnetisation of paramagnetic materials is directly proportional to the magnetic field that is applied; this, however, does vary with temperature. This relationship between temperature and magnetization is described by Curies Law: Where M is the resulting magnetisation, B is the applied magnetic field, T is the temperature and C is the Curie constant which is specific for each individual material.

Figure 2: Hysteresis Curve ⁷

Hysteresis is a property of paramagnetic materials. If a magnetic material is magnetised in regard to a direction and the material does not relax back to zero magnetisation then the material is exhibiting hysteresis. If an alternating field is applied a hysteresis loop is formed as shown in Figure 2. Bulk magnets exhibit hysteresis, as do some polyoxometalates.⁸ This provides many advances into information storage as exhibiting hysteresis means the material has magnetic bistability, meaning the material can present two stable phases and can change between the two with response to an external stimulus, a vital property for memory elements.

2.1 Magnetic POMs

POMs commonly contain transition metal centres, some of which are paramagnetic with respect to their spin states. The paramagnetic transition metal centres therefore correspond to the d¹ electron configuration mentioned above, e.g. polyoxovanadates (IV). When the paramagnetic metal ions are brought together they interact and can give rise to a wide variety of magnetic properties. The magnetic interaction between the ions is defined by the Heisenburg Hamiltonian. These POMs which possess magnetic properties are known as single molecule magnets. Polyoxovanadates are the most common magnetic POM. [V^IV₁₅As₆O₄₂(H₂O)]^6- is a relatively large oxovanadium (IV) POM whose magnetic properties have been widely investigated.⁷

2.2 Quantum or classical behaviour?

POMs are relatively large, organically bridged, molecular magnets, and are thought to potentially be the bridge between the quantum properties of the smallest atoms and the classical properties possessed by larger magnets that we know in our day-to-day lives. Events in classical physics are either allowed or forbidden in contrast to the events in quantum physics which all have a various probability of occurring. It has been thought that the mesoscopic size of the POMs would cause the classical properties seen in bulk magnets to also be observed in these intermediate sized magnets. Some magnetic POMs are thought to behave quasi-classically as they can exhibit both quantum as well as classical magnetic properties.⁸ This 15 vanadium centred POM was however, found not to behave in a classical way. Though, it has been reported that the quantum/classical behaviour is not just based on the size of the magnet, but also the individual spins of the metal ions. If the individual ions within the POM have a small spin (S= 1/2) then quantum models approximate properties of the material the best, however when reaching relatively high spins similar results to the quantum models are shown by using classical models. This has been confirmed in smaller clusters which contain larger spins.⁷ It is interesting that smaller clusters with higher spins possess more classical magnetic properties, opposed to larger clusters which possess small spin ions. Electron paramagnetic resonance (EPR) is a technique used to study materials with unpaired electrons, i.e. paramagnets. Line widths present in EPR spectra can dictate whether a material is expressing quantum or classical behaviour. ⁹

2.3 Spin frustration

Figure 3. Spin fustrastion within [V^IV₁₅As₆O₄₂(H₂O)]^{6- 7}

To achieve a large spin ground state is a difficult challenge for scientists. However, spin frustration effects can lead to isolated systems that contain this property. The layered structure of [V^IV₁₅As₆O₄₂(H₂O)]^6- comprising of three layers, with the central layer only having three interacting centres, then the spin cannot be resolved by Hund’s rule and is said to be frustrated, as shown in Figure 3.

Much research has been done into the use of large POMs as magnets, however with current technology allowing the miniaturisation of most devices, the question posed is who can make the smallest POM in which both quantum and classical effects coexist? For this to be answered the synthesis of POMs needs to be carefully controlled and fine-tuned.

2.4 Mixed valence clusters

The magnetic properties of mixed valence clusters are often more difficult to interpret due to their electrons which are delocalised over the structure as opposed to the simpler localised valence species, described above. The mixed valence cluster [Mn₁₂O₁₂(O₂CCH₃)16(H₂O)₄].2CH₃COOH.4H₂O, comprising of Mn₄^IV and Mn₈^III has been found to exhibit both quantum and classical properties. It has been established that this cluster behaves like a small bulk magnet when subjected to low temperatures.

2.5 Single molecule magnets

Single molecule magnets (SMMs) are usually based on first row transition metals. They require a high spin ground state along with a negative uniaxial anisotropy. A relatively high blocking temperature (the temperature at which relaxation of magnetisation is slow) is also desirable.⁹ Flexibility in the structure of SMMs provides difficulties in the regulation of intermolecular exchange coupling, however, due to their rigid conformation, POMs have an environment which is ideal for magnetic structures. To produce SMMs the intermolecular magnetic interactions between molecules must be negligible.

2.6 POMs as ligands

Although POMs themselves can have magnetic properties, due to their variety of shapes if they possess defects, as in the lacunary structure, and the cavities the wheel structure often has, they can also be used as multidentate ligands.¹⁰ They can bind to paramagnetic coordination clusters, causing encapsulation of small clusters of magnetic ions.¹¹ By doing this the POM ligand effectively dilutes the smaller, encapsulated molecule and therefore reduces the dipolar interactions that are undesirable for SMMs. The magnetic molecule encapsulated could possess both antiferromagnetic and ferromagnetic properties; however, which property is expressed could potentially be determined by the POM surrounding it. The wheel structure, due to the nature of its shape also possesses a cavity, which could also be used to bind smaller clusters, increasing the distance between the magnetic molecules and in doing so reducing the interactions between them. ¹²

3. Designing POMs

The design of molecular clusters with magnetic properties, or as ligands for small clusters, that possess magnetic properties is challenging. The traditional one-pot synthesis controlled by numerous variables does not give the desired purity of species or the fine tuning abilities required to design SMMs. The degree of functionality required for these systems relies upon directed assembly of the building blocks into disciplined architectures.¹³

3.1 Solvothermal and ionothermal synthesis

Solvothermal synthesis reduces the limitations of the experimental variable of temperature used in the regular synthesis. Water and common organic solvents used in the regular synthesis limit the temperature the system can reach; however, autoclaves used in the solvothermal process reduce this limitation and create the opportunity for higher temperatures and pressures of the reaction mixture. Reproducibility of these reactions is very difficult as it requires tight control of experimental conditions. Ionothermal synthesis adapts this solvothermal synthesis by using an ionic liquid as a solvent. These ionic liquids influence the assembly of the POMs by functioning as templates and can potentially create selective architectures.¹⁴

3.2 Linking clusters

Lacunary structures have been used as precursor molecules with other metal ions as linkers to produce high nuclearity POMs.^11,15 This is often known as the ‘building block strategy’.^{16} It opens up opportunities to expand the POM size beyond what has currently been synthesised. However, difficulties have been found due to the architecture of the potential building blocks rearranging and reorganizing due to not being stable enough in solution. Due to the stability of tungsten POMs this methodology is widely used within tungsten chemistry; however, these limitations cause problems for other complexes such as molybdenum. Stabilising the initial building blocks is currently a task being researched widely by several research groups.^11,16

3.3 Counter ion effect

As previously mentioned, POMs are charged species, generally anionic, and so must possess a counter ion. Anionic POMs possess cationic counter ions and vice versa. The properties of the counter ion i.e. size, charge etc, can affect the reactivity of the POM. The larger the counter ions, the more stable the POM is against aggregation. Amines are a common example of bulky counter ions which provide this stability. Organic amines have also been found to be capable of directing the self-assembly of small building blocks, enabling the creation of the desired structure.¹⁷

3.4 Top down

A novel “top down” synthesis has been reported.¹⁸ Opposed to all previously mention syntheses this is a truly unique and novel idea. It suggests the idea of a pH controlled decomposition of a very large POM into smaller fragments. The large POM; [P₄W₅₂O₁₇₈]^24- , has confirmed this approach to be successful with the controlled fragmentation of [P₃W₃₉O₁₃₄]^19-, the scheme is shown in Figure 4.⁵ Not only could this technique potentially enable the formation of new architectures, but it could also help to identify the building blocks that made the POM in the first place, revealing new insights into the mechanism of formation.

4. Conclusion

Polyoxometalates are a very versatile class of compounds. Their potential has been subdued in the past due to a lack of understanding and the traditional one-pot synthesis. With the discovery of the importance of POMs in their various applications, and in particular as molecular magnets, there have been many developments within the past decade with regard to their careful and controlled synthesis. However, the mechanism of formation of POMs is still not fully understood. The controlled design of the structure and function of POMs is still being investigated and there is little doubt it will lead to the engineered functionality required for many capabilities POMs are already showing to have.

The debate between quantum and classical mechanics has been going on for centuries. Quantum properties describe those with discrete levels, whereas classical properties have a continuous scale. The idea that POMs can be used as a “stepping stone” between the quantum and classical magnetic behaviour has started to unite both the theories and has started collaborative research between both physicists and chemists. This breakthrough enables the development and use of POMs in quantum computing and miniaturisation of devices, paving the way for new and exciting technologies.

5. Reference List

1. G. Johansson, Acta Chem. Scand., 1960, 14, 771-773.
2. W. G. Klemperer, T. A. Marquart, and O. M. Yaghi, Angew. Chem. Int. Ed., 1992, 31, 49-51.
3. B. Zhou, B. L. Sherriff, F. Taulelle, and G. Wu, Can. Mineral., 2003, 41, 891-903.
4. J. F. Keggin, Proc. R. Soc. London, Ser. A, 1934, 144, 75-100.
5. H. N. Miras, J. Yan, D. L. Long, and L. Cronin, Chem. Soc. Rev., 2012, 41, 7403-7430.
6. W. Rozenbaum, D. Dormont, B. Spire, E. Vilmer, M. Gentilini, C. Griscelli, L. Montagnier, F. Barre-Sinoussi, and J. C. Chermann, Lancet, 1985, 1, 450-451.
7. M. T. Pope, A. Muller, and Editors., Polyoxometalates: From Platonic Solids to Anti-Retroviral Activity. Kluwer, 1994, 411.
8. D. Gatteschi, A. Caneschi, L. Pardi, and R. Sessoli, Science, 1994, 265, 1054-1058.
9. E. C. Yang, C. Kirman, J. Lawrence, L. N. Zakharov, A. L. Rheingold, S. Hill, and D. N. Hendrickson, Inorg. Chem., 2005, 44, 3827-3836.
10. J. Liu, J. Guo, B. Zhao, G. Xu, and M. Li, Transition Met. Chem., 1993, 18, 205-208.
11. H. Abbas, A. L. Pickering, D. L. Long, P. Kogerler, and L. Cronin, Chem. Eur. J., 2005, 11, 1071-1078.
12. J. Lehmann, A. Gaita-Arino, E. Coronado, and D. Loss, Nat. Nanotechnol., 2007, 2, 312-317.
13. A. Muller, F. Peters, M. T. Pope, and D. Gatteschi, Chem. Rev., 1998, 98, 239-271.
14. A. S. Pakhomova and S. V. Krivovichev, Inorg. Chem. Commun., 2010, 13, 1463-1465.
15. D. L. Long, H. Abbas, P. Kogerler, and L. Cronin, J. Am. Chem. Soc., 2004, 126, 13880-13881.
16. J. M. Cameron, J. Gao, L. Vilá-Nadal, D. L. Long, and L. Cronin, Chem. Commun., 2014, 50, 2155-2157.
17. Y. Jun, L. De-Liang, N. M. Haralampos, and L. Cronin, Inorg. Chem., 2010, 49, 1819-1825.
18. C. P. Pradeep, D. L. Long, C. Streb, and L. Cronin, J. Am. Chem. Soc., 2008, 130, 14946-14947.