Spin-state splittings of iron(II) complexes with trispyrazolyl ligands
Graphical abstract
What determines the spin-state splittings of iron(II) complexes with trispyrazolyl ligands ? Is it steric interaction between substituents on the pyrazolyl-rings, direct covalent metal–ligand bonding, or the presence of counter-ion and solvent molecules? Here, we have studied these intriguing systems with density functional methods at the OPBE/TZP level.
Introduction
In 1966, a versatile new class of ligands appeared, which combines some features of the cyclopentadienyl (Cp) and betadiketonate ligands [1]. These new complexes are the polypyrazolylborate anions with the general structure [RnB(pz)4−n]−, where n can be 0, 1 or 2, pz is a pyrazol-1-yl group and R usually is H, an alkyl or aryl group (see Scheme 1). With n = 1 and R = H, this gives a tridentate ligand [HB(pz)3]−, which had been given the abbreviation Tp (for tripyrazolyl) and its substituents are then often denoted by superscripts (e.g. Tp3Me for the 3-methyl analogue) [2].
Since the systematic name poly(pyrazol-1-yl)borates does not convey the mode of coordination of these ligands, these ligands are also called scorpionates [2] to provide an idea of how the ligand binds to metal ions. In almost all cases there are (at least) two pyrazolyl groups coordinated to the metal ion (see Scheme 2) and the resulting six-membered ring is in a deep boat form that brings the third group arching over the metal ion, like the tail of a scorpion. In the case of homoscorpionates this third group consists of a third pyrazolyl group, but it may also be formed by other groups (e.g. H, OR, SR, NR2), i.e. the heteroscorpionates. However, the coordination of the metal with a third ligating group is not mandatory, since the Tp ligand can sometimes be only bidentate, as in complexes of Rh(I) and Pd(II) [3].
Several transition-metal complexes have been observed where a metal is ligated by two of these Tpx ligands (di-scorpionates, see Fig. 1) with an octahedral (homoleptic) metal coordination. In case of Fe(II), many of these have been classified as spin-crossover complexes (see Table 1), which are interesting complexes for technology and medicine, as they are able to switch between different states depending on the temperature, pressure, etc. However, not all of the di-scorpionate complexes show spin-crossover behavior, as is for instance the case for Fe(Tp3,4,5-Me3)2 that remains high-spin even at low temperatures, or the parent complex Fe(Tp)2 that remains low-spin up until ca. 420 K.
The reason for this different behavior is often attributed to steric intra-molecular interligand interactions between the 3-methyl groups, but electronic substituent effects can not be disregarded altogether and may even play a decisive role. Moreover, spin-crossover properties are generally believed to result from cooperative properties of a large number of these “molecules” within the solid phase, and which involve long-range intermolecular interactions, even though spin-crossover (SCO) has also been observed with complexes in solution [4], [5], [6]. Therefore, the origin for the differences in the spin-crossover behavior that are observed for different substituted di-scorpionate complexes (if any) remains unknown.
In principle, computational studies should be able to clarify (some parts of) this complex situation, as one could perform separate studies on the isolated complexes, and on the solid phase. By comparing these results, one could obtain detailed information about the parameters that govern the spin-crossover (SCO) phenomenon. However, there are a number of pitfalls that hamper this straightforward application of theory. The most important one regards the correct description of relative spin-state energies (spin-state splittings) of isolated transition-metal complexes, which has shown to be a very complicated endeavour. There has been a plethora of papers describing the failure of Density Functional Theory (DFT) [7] functionals for this property [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], especially regarding that of B3LYP. For example, Trautwein and co-workers showed in 2001 that B3LYP is unable to correctly predict the low-spin ground-state (at 0 K) for a number of SCO complexes [11]. Reiher and co-workers subsequently proposed to lower the amount of Hartree–Fock exchange (to 15% to give the B3LYP* functional) [13], [14], which seemed to perform better in many instances, but still failed dramatically for many other complexes. Other functionals had been proposed such as XLYP/X3LYP, [28] or more recently the M06 suite [29], but these were also inaccurate for the spin-state splittings [9].
An important breakthrough was made in 2004, when one of us reported for the first time the application of a new GGA functional (OPBE) to spin-state splittings [8]. This new functional combines the non-empirical PBEc correlation functional by Perdew and co-workers with Handy and Cohen’s empirical OPTX exchange functional [30], [31], [32]. The spin-state splittings as predicted by OPBE give in almost all cases the correct spin ground-state [19], [22], [24], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43], with as only exception the iron–porphyrin with an axial histidine ligand [44]. For this latter system OPBE predicts, like all DFT functionals, a triplet ground-state while CCSD(T) predicts a quintet ground-state, which would be more inline with experimental data. For all other systems studied so far, OPBE correctly predicts the spin ground-state, including for difficult systems like the Fe(phen)2(NCS)2 spin-crossover complex, and iron with (amp, dpa) pyridylmethylamine ligands. For these complexes, it was shown very recently that OPBE is the only DFT functional able to correctly predict a high-spin ground-state for Fe(amp)2Cl2 and a low-spin for the related Fe(dpa)22+ complex [9]. Moreover, also for NMR chemical shifts and reaction barriers does OPBE seem to perform significantly better than other DFT functionals [35], [36], [45], [46].
It should be noted also that not only the DFT functional is important, but also the basis set that is being used [47]. We recently reported a systematic study into the effect of the basis set, and found that Slater-type orbital (STO) basis sets performed well, and also very large Gaussian-type orbital (GTO) basis sets [47]. However, neither small/medium GTOs nor basis sets containing effective core potentials (ECPs) were reliable. Moreover, for a series of Pople basis sets, it was observed that the largest basis set (6-311 + G**) actually gave the largest deviation from the reference STO/very-large-GTO data. Here, we use both a reliable functional (OPBE) and reliable (STO) basis sets so that we can confide in the results obtained. In particular, we have studied the spin-state splittings for isolated Fe(Tpx)2 complexes to see how these are influenced by substituent effects on the pyrazolyl-ring.
We have studied also the neutral analogues of the anionic Tpx ligands, i.e. the trispyrazolylmethane ligands [48]. Note that in order to distinguish the methane ligand from the borate ligand, from here on we will refer to these as Tc and Tb respectively. Despite the fact that trispyrazolylmethane was first reported in 1937 [49], its chemistry is underdeveloped in comparison with the boron counterpart. However, recent breakthroughs in the synthesis of ring-substituted trispyrazolylmethanes [50] offer the opportunity for the development of this promising class of ligand. They are formally derived by replacing the apical (BH)− anionic moiety with the isoelectronic CH group [48]. Because the Tcx ligand is neutral, the iron(II) di-scorpionate with two Tcx ligands has a total charge of +2, and hence in experimental studies is accompanied by one or more counter-ions. The presence of the counter-ion(s), and even which one is present, was shown to have a large influence on the stability and properties of the complex [51], [52]. Nowadays, complexes with a wide range of metals supported by Tcx ligands have been synthesized [52], [53], and many studies about spin-transitions in the Fe(II) complexes have been reported [54], [55], [56], [57], [58], [59], [60], [61]. Consequently, we decided to carry out a study on a wide range of iron(II) complexes with Tbx and Tcx ligands, with particular emphasis on how these are influenced by substitution patterns [51].
Section snippets
Computational
The calculations using the unrestricted formalism have been performed with the Amsterdam Density Functional (ADF) suite of program [62]. MOs were expanded in an uncontracted set of Slater type orbitals (STOs) of triple-ζ quality containing diffuse functions and one (TZP) sets of polarization functions. Energies and gradients were calculated using the local density approximation (LDA; Slater exchange and VWN correlation) with gradient-corrections (GGA) for exchange (OPTX) [31] and correlation
Results and discussion
We have studied several iron(II) complexes with the trispyrazolylborate Tbx and trispyrazolylmethane Tcx ligands, where substitutions have been made on the 3-, 4- and 5-position of the pyrazolyl groups (see Scheme 3 and Fig. 1). The total charge on the ligands (−1 for borate ligands, zero for methane ligands) has a substantial effect on the molecular orbitals. The highest occupied molecular orbitals of the borate ligand are found at elevated energies compared to those of the methane ligand (see
Conclusions
We have studied the spin-state preferences of iron(II) complexes with trispyrazolylborate and trispyrazolylmethane ligands, and determined at the OPBE/TZP level how these for an individual iron(II) complex in the gas-phase are influenced by substitution patterns at the pyrazolyl-ring. By using an energy decomposition analysis of the metal–ligand and ligand–ligand interactions, we have shown that there is strong resemblance between the covalent metal–ligand interactions for both ligands.
Acknowledgments
This study was financially supported by the Spanish research projects CTQ2008-03077/BQU and CTQ2008-06532/BQU and the DURSI project nr. 2005SGR-00238. MG thanks MEC for a research grant. We thank W. Han and L. Noodleman for providing the utility programs to compute the Mössbauer parameters of the iron complexes.
References (103)
- et al.
Inorg. Chim. Acta
(1995) - et al.
Curr. Opin. Chem. Biol.
(2003) - et al.
Theor. Chem. Acc.
(2008) - et al.
Chem. Phys. Lett.
(2006) Inorg. Chim. Acta
(2007)- et al.
J. Organomet. Chem.
(2000) - et al.
Coord. Chem. Rev.
(2005) - et al.
Polyhedron
(2004) - et al.
Chem. Phys. Lett.
(1997) - et al.
Inorg. Chim. Acta
(2008)
J. Phys. Chem. Solids
Polyhedron
Polyhedron
Inorg. Chim. Acta
Tetrahedron Lett.
Solid State Sci.
Inorg. Nucl. Chem. Lett.
Polyhedron
J. Am. Chem. Soc.
Chem. Rev.
Scorpionates. The Coordination Chemistry of Polypyrazolylborate Ligands
J. Am. Chem. Soc.
J. Phys. Chem. B
A Chemist’s Guide to Density Functional Theory
J. Phys. Chem. A
J. Chem. Theory Comput.
J. Phys. Chem. A
Inorg. Chem.
Inorg. Chem.
Theor. Chem. Acc.
Inorg. Chem.
J. Biol. Inorg. Chem.
Inorg. Chem.
J. Am. Chem. Soc.
Inorg. Chem.
J. Biol. Inorg. Chem.
J. Phys. Chem. B
J. Phys. Chem. B
J. Chem. Theory Comput.
J. Chem. Phys.
J. Chem. Phys.
J. Chem. Phys.
ChemPhysChem
Proc. Natl. Acad. Sci. USA
Mol. Phys.
Mol. Phys.
Phys. Rev. Lett.
J. Chem. Theory Comput.
J. Phys. Chem. A
J. Comput. Chem.
Cited by (39)
Iron complexes of tris(pyrazolyl)ethane ligands methylated in the 3-, 4-, and 5-positions
2014, Inorganica Chimica ActaAllyl molybdenum(II) and tungsten(II) compounds bearing bidentate and tridentate pyrazolylmethane ligands
2014, PolyhedronCitation Excerpt :Neutral tris(pyrazolyl)methanes are typical tripodal ligands but their coordination behavior is much less understood than in case of more pronounced isoelectronic anionic borate analogues probably owing to poor synthetic routes giving low yields that leads to a much smaller family of accessible modifications on the ligand framework [18–22]. In last decade, the attention in tris(pyrazolyl)methanes reappeared [23–29] mainly due to the development of new synthetic routes giving ligands modified on the bridging carbon atom [30–33] and the higher stability of the ligand framework in aqueous media when compared to the borate analogues [33–36]. These developments led us to the cationic allyl molybdenum species [(η3-C3H5)Mo(CO)2(κ3-tpm)]+ that was firstly described by Trofimenko in his pioneering work [3].
Computational insight into Wilkinson's complex catalyzed [2 + 2 + 2] cycloaddition mechanism leading to pyridine formation
2014, Journal of Organometallic ChemistryVibrational properties of the mononuclear Fe[HBpz<inf>3</inf>]<inf>2</inf> spin crossover complex
2022, Hyperfine Interactions