Thermal spin transition between high spin (HS) and low spin (LS) states is known to occur in coordination compounds if the difference between the lowest vibronic states of HS and LS is on the order of thermal energy, kBT. The phenomenon was first seen in iron(III) dithiocarbamates by Cambi and coworkers around 1930. More than thirty years later the first examples for thermal spin transition in iron(II) complexes became known. We started work in this field around 1970 with the primary goal of learning more about the mechanism and the driving force of thermal spin transition, also called „spin crossover” (SCO), in solids. We first concentrated on the various chemical influences on the spin transition behaviour. From studies of the ligand substitution effect /29, 33, 34, 40) in a number of substituted phenanthroline complexes we learned that the spin transition behaviour is influenced essentially in two ways, through electronic push/pull effects via the donating nitrogen atom, and through steric hindrance by the substituent in 2- and or 9-position via the (1/R5) dependence of the cubic crystal field parameter. The experimental findings could be supported by CNDO calculations. Later attempts to „fine-tune” the ligand-field strength suitable for thermal SCO behaviour to occur were carried out with pyrazolyl-borate complexes /219/.
Another chemical influence on the SCO behaviour which we found to be very important is the isomorphic replacement of the iron(II) by a different transition metal, preferentially the diamagnetic zinc(II). The study of the ”metal dilution effectâ in the mixed crystal system [FexZn1-x(2-pic)3]Cl2.C2H5OH /37,46/ (2-pic = 2-picolylamine) was the first of this kind in the field. The essential findings, which were also observed in later similar metal dilution studies in other SCO mixed crystals /53, 59,69, 70, 87, 221, 231/ is that the HS state becomes stabilized with decreasing iron(II) concentration, which is reflected in the spin transition curve, γHS(T), which is shifted to lower temperatures and becomes more gradual with decreasing iron concentration. In other words, as the distance between the spin state changing centers increases with decreasing iron concentration, the cooperative interactions become weaker until they are practically lost like in a liquid solution; then the spin transition curve becomes smooth and can be described as a simple Boltzmann distribution over all HS and LS spin state levels.
These results from the metal dilution studies, together with the observation by X-ray dif-fraction on single crystals of SCO compounds first in a japanese group and also in our laboratory /102/ that the metal-ligand bond distance in the HS state is ca. 10 % longer than in the LS state, were fundamental for the development of the model of „elastic interactions and lattice expansion” in our laboratory /69,87, 101, 147, 154/. This model has generally been considered the most successful theoretical approach among the many attempts that have been reported so far.
We have chosen the SCO system [Fe(2-pic)3]X2.Sol for detailed investigations of the influence of all possible chemical alterations on the SCO behaviour. Except for the effect of metal dilution described above we have observed, using different physical techniques, that the nature of the non-coordinated anion X as well as the crystal solvent molecules (water, ethanol, methanol etc.) /38, 276/ greatly influence the SCO behaviour. Moreover, with great effort we have synthesized this complex with picolylamine labelled with deuterium and 15N in various positions and found that the spin transition behaviour is significantly influenced only in those cases where the labelling isotopes take part in the ”communication chainâ from one metal center to another /54, 122/. This supports the existence of cooperative interactions via electron-phonon coupling. Lattice deformation parameters could be derived from temperature-dependent single-crystal X-ray studies /102, 128, 154/. EPR measurements on a single crystal of the pic complex doped with ca. 1% of Mn2+ (the first study of this kind in SCO research) enabled us to probe the local structural changes via the ZFS parameters D and E /61/. Most surprising was the observation in a high-resolution Mössbauer effect study of the pic system, that the spin transition takes place in two steps with a plateau near the transition temperature of ca. 120 K/71/, which was later confirmed by heat capacity measurements with a home-made specially designed and constructed micro-calorimeter which allows the measurements on single crystals of only 10 mg in weight /172,177/. We have derived the mixing entropy from the heat capacity measurements and found that there is a considerable lack as compared to the expected mixing entropy calculated under the assumption that HS and LS molecules are distributed statistically in the transition region. The missing mixing entropy points actually at a certain ordering phenomenon as a result of short-range interactions between the metal centres. We have performed Monte-Carlo calculations and found that apparently there is preferential HS-LS-HS-LS... ordering like in a chess plate inside domains and statistical distribution outside. The existence of the two-step transition as well as its gradual disappearance upon metal dilution and under pressure could be well reproduced /260, 294/. Another theoretical approach employing the cluster variation method /265/ supported the importance of the short-range interactions leading to special features in spin transition behaviour such as a step. Later, H. Bürgi et al. have confirmed by temperature-dependent single crystal structure determination that the stepwise spin transition is connected with the appearance of an intermediate phase.
Other special features of spin transition behaviour and highlights which we have learned from our detailed examination of SCO complexes are briefly described in the following:
- The crystal quality (single crystal vs. powder, mechanical treatment) influences strongly the SCO behaviour, e.g. a hysteresis becomes broader and more gradual /73, 85/.
- Following the spin transition curve with hysteresis in steps such that smaller loops appear inside the hysteresis and using Everettâs theorems leads to special information about cluster formation, depending on shape and area of the inner loops /77/.
- Applying a magnetic field has marginal influence, it decreases the spin transition temperature /80/.
- From relaxation studies using Mössbauer spectroscopy one finds rates of fluctuation between HS and LS states on the order of 107 s-1 /120, 140, 143, 197/.
- Applying pressure stabilizes in general the LS state, which is expected from the fact that the metal-ligand bond distance shrinks by up to 10 % upon HS to LS conversion, and yield information about cooperative interactions /74, 127, 141, 142, 276, 277, 281, 308, 312, 379, 419, 439/. Substituted triazole complexes /276, 277, 281, 308, 312/ turn out to be suitable SCO complexes for pressure sensors for measurements in remote positions. The very rare case of thermal spin crossover in an chromium(II) compound was also studied under pressure /379/. Our pressure effect studies in molecular magnetism, particularly on spin crossover systems have been summarized in several comprehensive review articles /378, 389, 400, 419/.
- Although by far most of the iron(II) compounds exhibiting thermal spin transition possess a FeN6 core, we have documented that SCO also takes place in complexes with FeP4Cl2 core containing phosphine ligands /267/.
- Polynuclear SCO complexes of iron(II) can be synthesized with Fe2+ ions in different lattice sites exhibiting different SCO behaviour, e.g. complexes with substituted tetrazole and triazole ligands /126, 236, 243, 259, 268/.
- Positron annihilation measurements have proven to be a suitable technique for local structure changes in spin transition processes / 109, 203, 211, 233/.
- Following temperature dependent HS and LS fractions with Mössbauer spectroscopy requires the knowledge of the Debye-Waller factors (Lamb-Mössbauer factors, LMF) of the two spin states. We have worked out the differences in LMF between HS and LS in neat crystals and in metal-diluted SCO materials, and pointed at the eminent importance of the nature (acoustic modes) of the host lattice and of the intramolecular vibrations, respectively /105, 108, 154, 231, 237, 240/.
- Co(III) complexes with d6 electron configuration like Fe(II) are known to be mostly diamagnetic. Only [CoF6]3- is paramagnetic with S=2 ground state. A few Co(III) complexes, which were synthesized by W. Kläui and characterized in our laboratory, exhibit thermal spin transition /60, 119).
We have devoted particular attention to the class of dinuclear iron(II) complexes, work mostly in collaboration with the University of Valencia. An interesting interplay between antiferromagnetic coupling and thermal spin crossover has been observed in bipyrimidine-bridged compounds and extensively studied by magnetic and Mössbauer measurements, also under pressure /332, 345, 362, 383, 396/. We have developed a method, using Mössbauer spectroscopy in applied magnetic field, for direct monitoring the spin state in these systems, where the thermal spin transition in most cases takes place from HS-HS pairs via HS-LS intermediates to LS-LS pairs. A plateau at 50 % transition may occur, which is clearly identified as consisting of HS-LS pairs /339, 351, 380/. Recently we have studied a dinuclear SCO complex of iron(II) with a very rigid bridge between the two iron centers as an exceptional case where the spin transition takes place at only one iron(II) center leading from HS-HS to HS-LS pairs /434/. It could be demonstrated by temperature-dependent Mössbauer spectroscopy without magnetic field that the spin transition in one center is felt by the other one, still remaining in the HS state, but giving rise to a new quadrupole doublet due to a slight molecular distortion.
We have made similar observations when studying by Mössbauer spectroscopy (with Prof. J. Reedijkâs group in Leiden) a trinuclear iron(II) complex, where only the central metal ion with FeN6 core undergoes thermal spin transition, but the two outer iron(II) ions with FeN3O3 core (from capping water molecules) remain in the HS state throughout /259/. The outer two iron(II) ions ”feelâ the thermally induced spin transition at the central metal through a slight molecular distortion causing a new quadrupole doublet with area fraction of 2:1 in relation to that of the new LS resonance signal at the center. These are exceptionally well suited examples for teaching Mössbauer spectroscopy.
We have extended the series of oligomeric iron(II) complexes exhibiting thermal spin transition by investigating tetranuclear iron(II) complexes, so-called grid complexes, prepared by J.M. Lehnâs group. All four iron(II) ions are in the HS state at sufficiently high temperatures and change spin state successively on cooling /357, 393, 397/. The iron centres are weakly ”mechanicallyâ coupled and as a consequence the thermal spin transition is very gradual, which is indicative of weak cooperativity. The spin transition is also influenced greatly by application of pressure. These grid-like systems show the LIESST effect.
A pentanuclear complex containing five iron(II) ions, one in the center with FeN4O2 core (from two coordinated water molecules) remaining in the HS state and four iron(II) ions with FeN6 core was prepared and structurally characterized in the group of J. Reedijk and studied with magnetic and Mössbauer measurements in our laboratory /268/. The system shows rather sharp thermal spin transition at the four outer metal ions.
One of the most significant discovery in our research laboratory was the observation that spin state switching in SCO substances can also be achieved with light, from LS to HS using green and from HS to LS using red light. The first (accidental) observation was in 1984 on [Fe(ptz)6]BF4 (ptz = propyl-tetrazole) /84/. We have called this phenomenon „Light-Induced Excited Spin State Trapping (LIESST)”. The mechanism has been fully elucidated on ligand field theoretical grounds /95/. More examples exhibiting this fascinating photophysical phenomenon have been found shortly after the discovery /94, 110, 126, 220, 236, 253, 254/, and LIESST was observed even with SCO compounds embedded in polymer matrices /137/. The dynamics with particular emphasis on cooperative interactions has been investigated using optical and Mössbauer techniques /212, 218, 257/. Unequivocal evidence for the occurrence of cooperative interactions during LIESST state relaxation could be derived from the shape of the relaxation curves, which are sigmoidal in a neat (undiluted) crystal, but single-exponential in metal-diluted crystals /95, 212, 257/. Interestingly, the LIESST phenomenon was also observed for iron(II) sites with HS ground state, which was converted into the long-lived LS state with red light /149, 171/. Very recently, we have discovered new photophysical effects: An existing thermal hysteresis of a SCO material is shifted to the low temperature region using green light /302/ and, as a preliminary result, to higher temperatures with red light. Absolutely surprising was the recent observation of the LIESST effect in a strong-field compound with spin transition temperature far above 300 K /300/, contrary to what would be expected from the „inverse energy gap law”, which predicts in this case much shorter lifetimes for the LIESST state by ca. 15 orders of magnitude. These photophysical effects may well be suited for fast optical switches and displays. Photoswitchable coordination compounds based on the LIESST effect and other phenomena have been discussed in several review articles /182, 191, 228, 272, 280, 305, 322, 327, 331, 368/.
A key question in our studies of the light-induced spin state conversion leading to long-lived metastable states has been the behaviour of the crystal lattice during the photophysical processes. We have installed a single-crystal X-ray diffractometer equipped with low temperature facility (down to ca. 8 K), CCD detector and fibre optics and investigated the structure of the LIESST state in the iron(II) methyl-tetrazole complex /329, 335/, as well as the superstructure of the LIESST state in the iron(II) picolylamine complex under continuous irradiation /401/. In the case of the iron(II) propyl-tetrazole complex we were able to perform structural studies on five different phases, in making use of the LIESST phenomenon and a super-cooling technique /390/. The same five phases have also been characterized by nuclear inelastic scattering with synchrotron radiation (NIS) at ESRF in Grenoble /413/. This newly developed technique uses the sharp 14.4 keV synchrotron radiation of the Mössbauer nuclear transition in Fe-57, inelastic scattering yield phonon spectra which characterize the local vibrational properties and densities of states. NIS was also employed to explore the local vibrational properties and density of states of the ”classicalâ iron(II) SCO complex [Fe(phen)2(NCS)2] in comparison to results from IR and Raman spectroscopy and DFT calculations /414/.
Another recently developed tool for solid state research is muon spin relaxation (MuSR). We have used the facilities at ISIS (England) to investigate successfully for the first time dynamic lattice processes in spin crossover systems /348, 370, 394, 395, 405/.
One of the objectives of current spin crossover research is to synthesize compounds which exhibit thermal spin crossover and at least one other physical property such as liquid crystalline behaviour, nonlinear optical behaviour, electrical conductivity, to name a few. In recent years we have directed our activities in spin crossover research mainly towards systems that combine thermal spin transition with liquid crystalline properties. A first report describing an iron(III) system with thermal spin transition below room temperature and liquid crystal characteristics above room temperature appeared in 2003 in Angew. Chem. /337/. Soon afterwards we have succeeded to prepare iron(II) spincrossover compounds with long alkoxy chains in 4-position of the 1,2,4-triazole bridging ligands showing thermal spin transition and liquid crystal properties in the same temperature region, i.e. coexisting phenomena, even at room temperature. The intriguing feature is that the LC phase transition is fully retained and, in addition, the system changes colour from pink (LS state) to white (HS state) upon spin transition on heating. These properties bear considerable potential for technical applications. So far we have succeeded to prepare and characterize mononuclear and polynuclear chain-like systems exhibiting coexisting SCO and LC phase transitions /402, 411, 416, 430, 433/. The spin transition characteristics in such systems can be controlled by varying the chain length containing the LC properties on one hand and by varying the distance between the discs in discotic LC systems on the other hand.
Another effort was undertaken recently to prepare iron(II) SCO complexes with dentritic ligands (in collaboration with D. Schlüter/Zürich). Despite the enormous difficulties in synthesizing and characterizing such systems we could show /437/ that thermal spin transition does occur, in a very gradual manner due to the softness of the material and the resulting weak cooperative interactions.
Our work on thermal, optical, pressure- and nuclear-decay-induced spin transition has been reviewed in many review articles /50, 55, 58, 92, 93, 98, 135, 182, 206, 238, 251, 272, 280, 304, 313, 378, 383, 386, 387, 388, 389, 390, 400/.
In 1998 we have started extensive work together with 9 other European research laboratories of the European Community TMR Network „Thermal and Optical Switching of Molecular Spin states (TOSS)”, which I had been coordinating and administering. The main goals were the synthesis and physical characterization of new SCO materials suitable for technical applications.
Together with H.A. Goodwin I have edited the first comprehensive book on ”Spin Crossover in Transition Metal Compoundsâ which appeared in 2004 in the Springer Series ”Topics in Current Chemistryâ /385/. The three volumes (No. 233, 234, 235) with close to 700 pages comprise 30 review articles from well reputed research groups in the field and is presently considered the richest literature source in spin crossover research. Six articles have been contributed from our own group: ”Spin Crossover - An Overall Perspectiveâ /386/, ”Thermal spin crossover in Mn(II), Mn(III), Cr(II) and Co(III) coordination compoundsâ /387/, ”Nuclear-decay-induced excited spin trapping (NIESST)â /388/, ”Pressure effect studies on spin crossover and valence tautomeric systemsâ /389/, ”Structural investigations of tetrazole complexes of iron(II)â /390/, ”Bipyrimidine-bridged dinuclear iron(II) spin crossover compounds, in Spin Crossover in Transition Metal Compoundsâ /383/.