Divalent metallocenes of the lanthanides – a guideline to properties and reactivity

Abstract

Since the discovery in the early 1980s, the soluble divalent metallocenes of lanthanides have become a steadily growing field in organometallic chemistry. The predominant part of the investigation has been performed with samarium, europium, and ytterbium, whereas only a few reports dealing with other rare earth elements were disclosed. Reactions of these metallocenes can be divided into two major categories: (1) formation of Lewis acid–base complexes, in which the oxidation state remains +II; and (2) single electron transfer (SET) reductions with the ultimate formation of Ln(III) complexes. Due to the increasing reducing character from Eu(II) over Yb(II) to Sm(II), the plethora of literature concerning redox reactions revolves around the metallocenes of Sm and Yb. In addition, a few reactivity studies on Nd(II), Dy(II) and mainly Tm(II) metallocenes were published. These compounds are even stronger reducing agents but significantly more difficult to handle. In most cases, the metals are ligated by the versatile pentamethylcyclopentadienyl ligand: (C₅Me₅). Other cyclopentadienyl ligands are fully covered but only discussed in detail, if the ligand causes differences in synthesis or reactivity. Thus, the focus lays on three compounds: [(C₅Me₅)₂Sm], [(C₅Me₅)₂Eu] and [(C₅Me₅)₂Yb] and their solvates. We discuss the synthesis and physical properties of divalent lanthanide metallocenes first, followed by an overview of the reactivity rendering the full potential of these versatile reactants.

---

Sebastian Schäfer

Sebastian Schäfer graduated from the Technical University of Kaiserslautern in 2012. He joined the group of Peter W. Roesky at KIT for his doctoral studies, focusing on low valent silicon and coinage metal chemistry. After receiving his degree in 2016 and a brief postdoctoral stay at KIT, he joined the chemical industry for a two-year post doc position working on the development of heterogeneous catalysts. Since 2019, he has been working as a head of laboratory and project lead in the field of inorganic materials in the chemical industry.

Sebastian Kaufmann graduated in 2020, receiving his doctoral degree from the Karlsruhe Institute of Technology (KIT). After a short postdoctoral stay at the KIT as the head of a student's laboratory practice, he moved on to the Baden-Württemberg Cooperative State University Karlsruhe. Where he was responsible for the laboratories in the Department of Safety, Health and Environment and assisted in the development of the new study course “sustainable science and technology”. In 2022 he went on to the chemical industry, joining a GLP laboratory.

Esther S. Rösch

Esther S. Rösch graduated from the University of Karlsruhe in 2007 and received her doctoral degree from the Karlsruhe Institute of Technology (KIT) in 2009. After a short postdoctoral stay at the KIT, she moved to the pharmaceutical industry. In 2013, she was appointed professor for Bioanalytics at the Pforzheim University of Applied Sciences. In 2018, she joined the Baden-Württemberg Cooperative State University Karlsruhe, where she headed the Department of Safety, Health and Environment from 2020–2022. Her current research focuses on magnetic nanoparticles and their biomedical applications, as well as on the utilisation of organic and inorganic luminophores, especially lanthanide-based ones.

Peter W. Roesky

Peter W. Roesky obtained his diploma in chemistry in 1992 from the University of Würzburg and his doctoral degree from the Technical University of Munich in 1994. After postdoctoral work at Northwestern University, USA (1995–1996), he completed his habilitation at the University of Karlsruhe in 1999. He was appointed a full professor at the Freie Universität Berlin in 2001, during which he joined the faculty of chemistry and biochemistry. In 2008, he became a Full Professor of inorganic functional materials at the Karlsruhe Institute of Technology (KIT). From 2013 to 2015, he served as Dean of the Faculty of Chemistry and Biosciences at KIT. His current research interest revolves around the synthetic inorganic and organometallic chemistry of s-block metals, silicon, phosphorus, gold, and lanthanides.

---

1. Introduction

The revolution of organometallic chemistry in the mid of the last century, starting from the landmark (but rather serendipitous) isolation of ferrocene^1,2 and the groundbreaking recognition of its sandwich structure^3,4 set off an avalanche of cyclopentadienyl chemistry that did not directly capture the lanthanides. In that time associated with a somewhat meek redox chemistry limited to the oxidation state +III, trivalent cyclopentadienyl compounds were isolated but displayed no noteworthy chemistry besides the formation of Lewis acid–base adducts. Also, cyclopentadienyl compounds of the rare earths are typically pretty insoluble in ethereal and hydrocarbon solvents as they tend to form polymeric structures.⁵ This proved especially true for divalent lanthanide compounds. Classically, three lanthanide elements have a well accessible divalent oxidation state: europium, samarium and ytterbium. Reactions of the corresponding Ln(II) halide precursors LnX₂ with cyclopentadienyl salts or metal/ammonia systems with cyclopentadiene resulted usually in insoluble organometallic polymers that did allow only a very few further reactivity studies.⁶ This changed with the emergence of one of the most prominent ligands in organometallic chemistry, pentamethylcyclopentadienyl. With its enhanced sterical bulk and higher electron density compared to the parent cyclopentadienyl, it allowed the isolation of unusual and reactive species. In combination with its excellent crystallization properties, it played an immense role in the history of organometallics. It also allowed the isolation of the first well-soluble, divalent metallocenes of Sm, Eu and Yb, which resulted in the investigation of the fascinating chemistry of these reactive molecules. Today, we consider these compounds as the “classic” divalent metallocenes. In addition, a few divalent metallocenes of the other rare elements ligated by even more bulky cyclopentadienyl ligands are known. Although divalent metallocenes of Tm,⁷ Dy,⁸ and Nd⁹ were established some years ago, metallocenes of most of the other rare earth metallocenes were published only very recently.^10,11

These newly accessible divalent lanthanides are mostly considered as the “non-classical” divalent compounds. This review focuses mainly on the “classical” divalent lanthanide metallocenes of Sm, Eu and Yb ligated by the permethylated cyclopentadienyl ligand C₅Me₅. Their rich and colourful chemistry is still being actively researched all over the world even after 40 years of discovery and has not ceased to surprise to this day. Other substitution patterns of the cyclopentadienyl ligand are also known. However, they are only discussed herein if the influence of the substituents strongly deviates from the corresponding permethylated cyclopentadienyl ligand derivatives. The first part of this review will deal with the synthesis, structure and physical properties of the divalent metallocenes of the lanthanides. The second part covers the reactivity, divided into reactivity as a Lewis acid, yielding Ln(II) complexes, and reactivity in redox reactions resulting in Ln(III) complexes.

2. Synthesis

[(C₅Me₅)₂Yb] was the first divalent decamethylmetallocene. Its synthesis was published independently by Watson and Andrews in 1980.^12,13 By reacting YbBr₂ with KC₅Me₅ or NaC₅Me₅ with YbCl₂ in ethereal solvents like tetrahydrofuran (THF), diethylether (Et₂O) or 1,2-dimethoxyethane (DME) the corresponding metallocene was obtained as its respective ether adduct. The THF solvate of [(C₅Me₅)₂Sm] was first reported by Evans in 1981.¹⁴ By diffusing samarium vapour into a hexane solution of HC₅Me₅, the diene is reduced with concomitant hydrogen evolution. A dark mixture was obtained that was filtered. Extraction of the remaining dark solid with THF yields [(C₅Me₅)₂Sm(THF)₂] alongside hydride-containing side products. A few years later, a salt metathesis route to the metallocene starting from SmI₂ and KC₅Me₅ was reported (Scheme 1).¹⁵ In contrast to [(C₅Me₅)₂Sm(solvent)_n] and [(C₅Me₅)₂Yb(solvent)_n], [(C₅Me₅)₂Eu(solvent)_n] could be prepared from EuCl₃ and three equivalents of Na(C₅Me₅) in THF and crystallised as etherate.¹³ Here, one equivalent of pentamethylcyclopentadienyl acts as a reducing agent by transferring one electron to the Eu(III) centre with the release of a formal (C₅Me₅) radical that combines to a dimer. The observed reactivity is a result of the redox potential of the Eu(III)/Eu(II) couple.

Scheme 1 Synthesis of the title compounds via salt metathesis.¹⁵

Today, the most established and convenient route to THF-solvated [(C₅Me₅)₂Ln(solvent)_n] and similar metallocenes is the reaction of two equivalents of the cyclopentadienyl derivative with the corresponding divalent iodides in THF.^16–19 The divalent iodides can be cleanly prepared using the respective elemental metals and 1,2-diiodoethane.²⁰ Metallocenes with other substitution patterns on the five-membered ring including larger entities such as indenyl, fluorenyl and ansa-metallocenes are obtained similarly.^11,21–35 With a few exceptions in which the substituents strongly influence the structure, these compounds are not discussed here in detail. Thus, salt metathesis of other cyclopentadienyl derivatives such as pentabenzylcyclopentadienyl (Cp^Bn5) likewise led to the corresponding metallocenes [(Cp^Bn5)₂Ln] (Ln = Sm, Eu, Yb) (Scheme 2).³⁶ Due to the sterically demanding ligand as well as the saturated coordination sphere of the lanthanides by π-interaction with one phenyl ring per ligand, these compounds are present without any coordinating solvent at the metal centre.

Scheme 2 Synthesis [(Cp^Bn5)₂Ln] (Ln = Sm, Eu, Yb) via salt metathesis.³⁶

The synthesis of [TmI₂(DME)₃] in 1997 paved the way for thulium in organometallic chemistry.³⁷ However, metallocenes of thulium have not been synthesised with the pentamethylcyclopentadienyl ligand since bulkier ligands are needed to stabilise the highly reactive metal ions in the divalent oxidation state. In general, thulium metallocenes were synthesised similarly by salt metathesis from the corresponding alkali metal cyclopentadienyl derivatives and [TmI₂(THF)₃]. Thus, [(1,3-C₅H₃(SiMe₃)₂)₂Tm(THF)],³⁸ [(1,2,4-C₅H₂(SiMe₃)₂)₃Tm(THF)],³⁹ [(1,3-C₅H₃^tBu₂)₂Tm],⁴⁰ [(1,2,4-C₅H₂^tBu₃)₂Tm],⁴¹ and [(1,2,4-C₅H₂^tBu₃)₂Tm(THF)]⁷ were obtained by this route.

In contrast, all solution-based synthetic routes yield ethereal adducts of [(C₅Me₅)₂Ln(solvent)_n] (Ln = Sm, Eu, Yb). However, for some applications, it is necessary to start with the unsolvated, more reactive metallocenes. To obtain the solvent-free metallocenes of Sm and Eu, sublimation of the solvated compounds can be carried out. Evans and co-workers studied the synthesis of solvent-free samarocene.⁴² Upon heating [(C₅Me₅)₂Sm(THF)₂] to 85 °C at a pressure of 1 × 10⁻⁵ mbar, the compound desolvates readily and solvent-free samarocene was obtained in 75% yield as a sublimate. The desolvation is a two-stage process. The intermediate monosolvate [(C₅Me₅)₂Sm(THF)] is formed first and was successfully isolated and confirmed by X-ray crystallography.⁴³ In contrast, [(C₅Me₅)₂Eu(THF)] does not release THF that easily and three consecutive sublimations at 85 °C and 1 × 10⁻⁵ mbar are necessary to obtain a THF-free product.⁴⁴ In comparison, the solvent-free ytterbium compound cannot be obtained from the sublimation of either [(C₅Me₅)₂Yb(THF)] or [(C₅Me₅)₂Yb(OEt₂)]. The base-free ytterbocene was obtained by vacuum removal of the solvent from a strong boiling toluene solution of [(C₅Me₅)₂Yb(OEt₂)].¹⁹ The desolvation of the THF-solvate is not possible with this method, indicating the strong bond between the oxygen and ytterbium that appears to be stronger than in the related Eu and Sm metallocenes. When bulkier substituted cyclopentadienyl ligands were used (e.g., [(1,3-C₅H₃(SiMe₃)₂)], the removal of the solvent is facilitated.⁴⁵ In the case of very crowded cyclopentadienyl ligands (e.g. 1,2,4-C₅H₂^b₃, 1,2,4-C₅H₂(SiMe₃)₃, C₅Ph₅, Cp^Bz5 (pentabenzylcyclopentadienyl) the metallocenes are obtained solvent free.^{29,36,46–49}

It is also possible to synthesise the decamethylmetallocenes of Eu and Yb directly from the elements. Solutions of Yb or Eu metal in liquid ammonia reacted with HC₅Me₅ to first yield ammonia solvates of europocene and ytterbocene. Further extraction with THF or Et₂O subsequently resulted in the formation of [(C₅Me₅)₂Eu(THF)] or the mixed solvate, [(C₅Me₅)₂Yb(NH₃)(Et₂O)] (Scheme 3).⁵⁰ An analogous reaction with samarium metal has not been reported in the literature. [(C₅Me₅)₂Eu(THF)₂] was also synthesised from the metal and the plumbocene derivative [(C₅Me₅)₂Pb].⁵¹

Scheme 3 Syntheses of title compounds starting from metals Eu and Yb in liquid ammonia.⁵⁰

The synthesis of metallocenes with very bulky cyclopentadienyl ligands is different in some cases due to sterical reasons. In some cases, the desired metallocenes were directly obtained from the metal. Thus, the reaction of Yb metal with two equivalents of the bulky pentaphenylcyclopentadiene ligand and one equivalent of diphenylmercury in THF at room temperature did not yield the sandwich complex [(C₅Ph₅)₂Yb]. Instead, the ionic species [Yb(THF)₆][C₅Ph₅]₂ was obtained. However, the addition of non-coordinating solvents led to the desired sandwich complex [(C₅Ph₅)₂Yb], which remained insoluble in any nonpolar solvent.⁴⁷ This redox-transmetalation/protolysis (RTP) reaction was later extended to Sm and Eu.⁵² Thus, these metals react with one equivalent of HgPh₂, and two equivalents of C₅Ph₅H to give at 40 °C for several days the decaphenyllanthanocenes [Ln(C₅Ph₅)₂] (Ln = Sm, Eu). When using Hg(C₆F₅)₂ in place of HgPh₂ the reaction was performed at room temperature. The octaphenyllanthanocenes [Sm(C₅Ph₄H)₂(THF)] and [Eu(C₅Ph₄H)₂(DME)] were obtained likewise from HgPh₂ and two equivalents of C₅Ph₄H₂. A similar oxidative approach starting from the metal was used for the synthesis of the phosphine functionalised metallocenes [(η5-C^₅H₄PPh₂)₂Eu(DIME)] and [(η⁵-C₅H₄PPh₂)₂Yb(DIME)] (DIME = diethylene glycol dimethyl ether). Here, [Tl(C₅H₄PPh₂)] was reacted with metallic europium or ytterbium powder in THF in the presence of mercury, followed by crystallization from a solvent mixture of DME and DIME.⁵³ In a similar way bis(tris(trimethylsilyl)cyclopentadienyl)europium was synthesised from europium powder and Tl(1,2,4-C₅H₂(SiMe₃)₃).⁴⁹

A surprising alternative to this route to form tetra- and penta-phenylcyclopentadienyldiphenylphosphines was disclosed by Junk, Deacon, and co-workers. They underwent selective C–P bond cleavage with Eu, Sm, or Yb metal in the presence of catalytic amounts of I₂ to give [(C₅Ph₄H)₂Ln(solvent)] or [(C₅Ph₅)₂Ln] (Scheme 4).⁵⁴

Scheme 4 Syntheses of [(C₅Ph₄H)₂Ln] or [(C₅Ph₅)₂Ln].⁵⁴

Also ansa-metallocenes were obtained directly from the metals, e.g., acenaphthylene reacted directly with activated Sm and Yb metal by samarium or ytterbium in THF to yield the respective C₂-symmetric trans-rac-ansa-lanthanocene complexes [(η⁵-C₁₂H₈)₂Ln(THF)₂] (Ln = Sm, Yb) (Scheme 5).⁵⁵

Scheme 5 Syntheses of [(η⁵-C₁₂H₈)₂Ln(THF)₂] (Ln = Sm, Yb).⁵⁵

Besides salt metathesis and oxidation of lanthanide metals, a reductive approach is also known. Thus, [(C₅H₄Me)₂Yb(DME)] was obtained by reducing [(C₅H₄Me)₂YbCl] with metallic sodium,⁵⁶ while reduction of [(C₅H₄SiMe₃)₂YbCl]₂ with Na/Hg resulted in [(C₅H₄SiMe₃)₂Yb].⁵⁷ Even the highly reactive thulium compound [(1,2,4-C₅H₂^tBu₃)₂Tm(THF)] is accessible from [(1,2,4-C₅H₂^tBu₃)₂TmI] with KC₈ in toluene.⁷ Reduction of the corresponding Nd complex [(1,2,4-C₅H₂^tBu₃)₂NdI] with KC₈ in the presence of [18]crown-6 led to the divalent ate-complex [(1,2,4-C₅H₂^tBu₃)₂Nd(μ-I)(K[18]crown-6)].⁹ This compound reacts with a methylene group from the (1,2,4-C₅H₂^tBu₃) ligand forming a “tuck-in” complex. Metallate complexes of dysprosium [(1,2,4-C₅H₂^tBu₃)₂Dy(μ-X)(K[18]crown-6)] (X = BH₄, Br, I) were obtained in similar way by reduction of the corresponding trivalent precursors (Scheme 6).⁸

Scheme 6 Synthesis of [(1,2,4-C₅H₂^tBu₃)₂Dy(μ-X)(K[18]crown-6)] (X = BH₄, Br, I).⁸

By using the bulkier ligand pentaisopropylcyclopentadienyl (Cp^iPr5) linear divalent metallocenes of almost all rare earth elements [(Cp^iPr5)₂Ln] (Ln = Y, La, Ce, Pr, Nd, Gd, Ho, Er) can be synthesised by potassium graphite reduction of the corresponding iodine precursors [(Cp^iPr5)₂ LnI] (Scheme 7).^10,11 The Tm and Lu derivatives were obtained by in situ reduction. Thus, LnI₃ was reacted with 2.5 equiv. of NaCp^iPr5 first, followed by reduction with KC₈ in benzene.¹¹

Scheme 7 Synthesis of [(Cp^iPr5)₂Ln] (Ln = Y, La, Ce, Pr, Nd, Gd, Ho, Er).¹¹

Another common reductive approach to access donor functionalised metallocenes starts from the trivalent amido complexes [{(Me₃Si)₂N}₃Ln(III)(μ-Cl)Li(THF)₃] (Ln = Yb, Eu). These were mostly reacted with functionalised indenes or related ligands, which resulted in deprotonation of the ligand and concurrent reduction of the metal.^58–61 As byproduct {(Me₃Si)₂N}₂ is formed. Due to the higher redox potential, this reaction pathway does not work with samarium compounds.

A remarkable reductive approach was reported by Harder and co-workers.⁶² They reacted the benzyl complexes [(2-Me₂N-benzyl)₃Ln] (Ln = Sm, Yb) with the perarylated cyclopentadiene (4-nBu-C₆H₄)₅C₅H (Cp^BIGH) to obtain the divalent complexes [(Cp^BIG)₂M] (M = Yb, Sm). The steric bulk of the ligand seems to be a driving force for the reduction process. Although the reaction mechanism could not be fully deduced, the formation of a 2-Me₂N-benzyl radical was anticipated since 1,2-di(2-Me₂N-phenyl)ethane was found as a major side product in the mother liquors. The ethyl and isopropyl derivatives [{(4-EtC₆H₄)₅C₅}₂Sm] and [{(4-ⁱPrC₆H₄)₅C₅}₂Sm] were synthesised in a similar manner from the benzyl compounds [(DMAT)₂Sm(THF)₂] (DMAT = 2-Me₂N-α-Me₃Si-benzyl).⁶³

The corresponding europium compound [(Cp^BIG)₂Eu] was prepared in a simple protonation reaction from [Eu(DMAT)₂(THF)₂] with two equivalents of Cp^BIGH.⁶⁴

3. Properties

3.1. Solid state structures of the base free metallocenes

[(C₅Me₅)₂Ln] (Ln = Sm, Eu, Yb) have been structurally characterised in their unsolvated form in the solid state by single crystal X-ray crystallography.^19,42,44 Instead of forming a coplanar sandwich structure, a “bent-metallocene” structure like in the corresponding alkaline earth metallocenes is preferred (Fig. 1). This bent structure is also observed in gas phase.^65–67 The carbon–Ln distances in [(C₅Me₅)₂Ln] decrease from Sm to Yb in agreement with the lanthanide contraction. However, due to the similarities of the ionic radii in the neighbouring elements Sm and Eu, the structural parameters of [(C₅Me₅)₂Eu] and [(C₅Me₅)₂Sm] are almost identical. For both metallocenes, the average Ln–C distance is 2.79(1) Å. [(C₅Me₅)₂Yb] is obtained in two structural modifications, depending on the crystallization method. Only slight differences in the molecular bond lengths are observed for the two modifications. The corresponding Yb–C average distance is 2.66 Å, around 0.12 Å shorter than in the Sm- or Eu-metallocene. The bend angle centroid–Ln–centroid for [(C₅Me₅)₂Sm] is 140.1° and for [(C₅Me₅)₂Eu] 140.3°.^42,44 With 145/146° the bend angle is slightly wider in the ytterbium compound.¹⁹

Fig. 1 Molecular structure of [(C₅Me₅)₂Yb] in the solid state featuring the bent structure of the metalloecene (Reproduced from the CIF file CCDC: 1242530).¹⁹

Note, that the structures of the pentamethylcyclopentadienyl compounds are closely related to the corresponding metallocenes of calcium and strontium, which also exhibit the bent structural motif. In the [(C₅Me₅)₂Ca], the bending angle is 147° and the Ca–C distances are 2.609(6) Å, which is comparable to the corresponding angles and distances in the Yb-compound.⁶⁸ [(C₅Me₅)₂Sm] and [(C₅Me₅)₂Eu] most likely exhibit structural resemblances to [(C₅Me₅)₂Sr]. However, no solid state structural data is available for the Sr decamethylmetallocene.

The reason why the bent structures are preferred over the linear geometries both in the lanthanide decamethyl metallocenes and the corresponding alkaline earth metal compounds has been intensely discussed in the literature and investigated by theoretical methods. In principle, the following reasons have been discussed in the literature:

Electrostatic reasons. The spherical symmetry of the metal ions is disturbed by the negatively charged pentamethylcyclopentadienyl rings. By adopting a bent structure, the electron shell of the metal is deformed to a half-dumbbell-like form, having a positively and a negatively polarised region. The ligands interact with the positively polarised region, minimising the repulsion of the two negative charges of the ligands.⁴⁴

(n − 1)d-Orbital contribution. Hartree–Fock–Slater calculations on YbCl₂ showed bending, which was explained by an increased contribution of the inner (n − 1)d orbitals also observed in heavier s-block elements. This leads to the observed bent structures. The reason for this contribution is the smaller energetic separation between the ns/(n − 1)d orbitals due to relativistic effects, which results in the preference of ns/(n − 1)d over ns/np hybridisation.^69,70

van-der-Waals interactions between the rings. Force field calculations including electron correlation attributed the bending to van der Waals interactions between the C₅Me₅ rings and were able to reproduce the structures. As the rings approach each other, attractive dispersion forces over-compensate repulsion at a certain angle, resulting in a shallow energy well. By placing an optimised structure in the environment of a crystal, it could be shown that the methyl hydrogen atoms of adjacent molecules interact with the metal centre, thus satisfying the coordination sphere and influencing the bending. However, the impact of the interactions in a crystal on the bending is rather small. Interestingly, by applying the calculation to a (C₅Me₅)₂ unit without a metal ion, a bending was also observed.⁷¹

More recent relativistic, gradient-corrected density functional (DFT) calculations performed on [(C₅Me₅)₂Yb] contradict earlier results gained from molecular mechanics force field calculations.^72,73 Electrostatic and orbital interactions between the metal and the ring are identified as the main reason for the driving force away from linearity. Using this method, no bending was found for (C₅Me₅)₂ unit with a central dummy atom.

In contrast to [(C₅Me₅)₂Ln] compounds, metallocenes with bulkier substituents such as (C₅Ph₅) tend to form linear structures due to a steric clash of the substituents. Thus, single-crystal X-ray diffraction studies of [(C₅Ph₅)₂Yb] exhibit a highly symmetric structure with two parallel cyclopentadienyl ligands in a staggered conformation.⁴⁷ Detailed studies concerning the structure of [(Cp^BIG)₂M] were performed. The propeller-like ligands have opposite chirality and interlock with each other. The ligands are located in a parallel fashion, however, the metals are bound in a nearly perfect η⁵,η⁵-fashion (Fig. 2). However, unusually high displacement factors of the metal atoms parallel to the ring planes were observed, resulting in bent Cp_centre-M-Cp_centre units.⁶² A linear structure with a pseudo-D_5d symmetry was also observed for the pentaisopropylcyclopentadienyl compounds [(Cp^iPr5)₂Ln] (Ln = Y, La, Ce, Pr, Nd, Eu, Sm, Gd, Ho, Er, Tm, Yb, Lu).^10,11

Fig. 2 Molecular structure of [(Cp^BIG)₂Yb] in the solid state showing the propeller-like ligand arrangement (Reproduced from the CIF file CCDC 665304).⁶²

3.2. Bonding

For a long time, no disagreement was found on the bonding in [(C₅Me₅)₂Ln]. As for the metallocenes of the alkaline earth metals, the bonding was supposed to be mainly ionic. No contribution of the f-electrons was observed. Relativistic DFT calculations on [(C₅Me₅)₂Yb] revealed large charge separations between the ligands and the metal indicating significant ionicity in the compounds. The electron population in the f-orbitals were found to be close to 14, confirming no contribution of these orbitals to the binding.^73,74 However, a more recent study using quantum chemical methods on the DFT level comparing [(C₅Me₅)₂Sm] with [(C₅Me₅)₂Sr] showed a substantial covalent interaction in the Sm compound. The authors further elaborate that the ligand–metal bond in lanthanide(II) complexes is in general partially covalent.⁷⁵

3.3. Electronic structure

The electronic structures of [(CpiPr5)2Ln] (Ln = Y, La, Ce, Pr, Nd, Eu, Sm, Gd, Ho, Er, Tm, Yb, Lu) were investigated, e.g., by ultraviolet–visible (UV-Vis) spectroscopy.10,11 The results support the expected 4fn+1 electron configuration for Ln(II) = Sm, Eu, Tm, Yb and a 4fn5dz21 configuration for the other rare earth compounds ([Kr]4dz21 for Y(II)). EPR spectroscopy showed a significant s–d orbital mixing in the highest occupied molecular orbital and hyperfine coupling constants. Magnetic susceptibilities measured at room temperature suggests that the more pronounced 6s–5d mixing may be associated with weaker 4f–5d spin coupling.

The oxidation state of the divalent complexes [(Cp^BIG)₂M] (M = Eu, Yb, Sm) was confirmed by Harder and co-workers using various physical methods.⁶⁴ Temperature-dependent magnetic susceptibility data of [Yb(Cp^BIG)₂] confirmed the divalent oxidation state by showing diamagnetism. Temperature-dependent ¹⁵¹Eu Mössbauer investigations of [(Cp^BIG)₂Eu] ranging from 93K to 215 K showed an agreement with other Eu^II species. Detailed analysis of the Mössbauer spectra provided information about the dynamics of the Eu^II ion within the sandwich complex. X-ray absorbance near edge spectroscopy (XANES) also confirmed the divalent oxidation state of [(Cp^BIG)₂M] (M = Eu, Yb, Sm). Furthermore, [(Cp^BIG)₂Eu] showed extremely bright orange emission under UV excitation at room temperature. The photoluminescence spectra of this compound also confirm the divalent oxidation state.

3.4. Other properties

An interesting feature of [(C₅Me₅)₂Eu] is as mentioned above that Mössbauer spectra of ¹⁵¹Eu can be recorded with a ¹⁵¹Sm source.⁷⁶ At 4.2 K and 77 K broad absorptions at 13 mm s⁻¹ were observed. The broadness of the signals is explained by spherical paramagnetic relaxation that takes place at a rate comparable to the Mössbauer time scale of 9.7 × 10⁻⁹ s.

The ¹⁷¹Yb isotope has a nuclear spin of 1/2 and a natural abundance of 14.27%. The gyroscopic moment is 4.712 × 10⁷ rad T⁻¹ s⁻¹. Due to the full 14 f shell, Yb(II) ions are diamagnetic and therefore organometallic Yb(II) species can be well studied by NMR spectroscopy.⁷⁷ For [(C₅Me₅)₂Yb(Et₂O)] a shift of 36 ppm with ω_1/2 = 90 Hz was reported. The corresponding THF solvate gives a resonance at 87 ppm with ω_1/2 = 24 Hz.

The metal-bond disruption energies were determined for [(C₅Me₅)₂Sm] and [(C₅Me₅)₂Sm(THF)₂], which are 69.4(2.4) and 72.7(2.9) kcal mol⁻¹, respectively.⁷⁸

3.5. Influence of the substituents of the cyclopentadienyl ligand

As already pointed out, most metallocenes were synthesised using the pentamethylcyclopentadienyl ligand. On the one hand, bulkier substituents such as (1,2,4-C₅H₂^tBu₃) are needed to stabilise the highly reactive divalent thulium and dysprosium complexes. Bulky substituents therefore also reduce the reactivity. In a detailed study, the reactivity of [{(4-EtC₆H₄)₅C₅}₂Sm] and [{(4-ⁱPrC₆H₄)₅C₅}₂Sm] was compared with [(C₅Me₅)₂Sm] and [(C₅Me₅)₂Sm(THF)_n]. Not too surprisingly, the strongly shielded Sm(II) compounds showed significantly lower reactivity.⁶³ By using pentaisopropylcyclopentadienyl linear divalent metallocenes of almost all rare earth elements [(Cp^iPr5)₂Ln] (Ln = Y, La, Ce, Pr, Nd, Eu, Sm, Gd, Ho, Er, Tm, Yb, Lu) are accessible.¹¹

In general, the ansa-samarocenes feature structures with a stronger bending of the ligands and hence a wider biting angle. As a result, they showed higher activity and selectivity as catalysts for the polymerization of ethylene and 1-olefines.³²

4. Lewis acid–base complexes

Due to the large size of the Sm(II), Eu(II) and Yb(II) ions in the (C₅Me₅)₂Ln moieties, all three compounds act as Lewis acids. The reactivity in this case is determined by the ionic radii and is similar for all three compounds. Like their relatives [(C₅Me₅)₂Ca] and [(C₅Me₅)₂Sr], [(C₅Me₅)₂Ln] qualify as hard Lewis acids. The synthesis of Lewis acid–base complexes can be generalised (Scheme 8). Usually, the simple combination of the desired ligand and [(C₅Me₅)₂Ln(solvent)_n] in a suitable solvent leads to the formation of the anticipated complex. However, in some cases it is necessary to use the solvent-free reactants, e.g. for the preparation of complexes with ligands containing rather soft donor atoms that form weaker bonds with the hard lanthanide atoms. As for the geometry of the Lewis base adducts, either one or two additional neutral ligands can coordinate to the [(C₅Me₅)₂Ln] fragment, depending on the steric demand of the Lewis bases and the size of the metal. Besides Lewis acid–base complexes also some protolysis and substitution reactions are discussed in the chapter.

Scheme 8 General synthesis routes to Lewis acid–base adducts of the divalent lanthanocenes.

4.1. Oxygen donor ligands

As hard Lewis's acids, the metallocenes form stable complexes with oxygen donors, e.g. ethers. All three pentametyhlcyclopentadienyl derivatives were first reported as solvates of THF or diethylether.^13,14 These solvates are the most commonly used reactants, not only because of their lower reactivity, but also because they can be easily obtained by the aforementioned reactions in the respective ethers. The crystal structure of [(C₅Me₅)₂Sm(DME)] solvated with DME was also reported.⁷⁹ Selected (C₅Me₅)₂Sm-ether complexes are depicted in Scheme 9 with the corresponding Sm–O bond lengths.

Scheme 9 Ethereal adducts of samarocene.^14,79,80

In order to gain more information about the reactivity of samarium(II), a bis(dihydropyran) and a mono(tetrahydropyran) complex were synthesised.⁸⁰ [(C₅Me₅)₂Sm(OC₅H₁₀)] was obtained by dissolving [(C₅Me₅)₂Sm(THF)] in tetrahydropyran (OC₅H₁₀). After evaporation, the obtained maroon solid was dissolved in toluene and evaporated again to yield the dark brown solid of [(C₅Me₅)₂Sm(OC₅H₁₀)]. [(C₅Me₅)₂Sm(OC₅H₈)₂] (OC₅H₈ = 3,4-dihydro-2H-pyran) was synthesised by dissolving [(C₅Me₅)₂Sm(THF)₂] in toluene and evaporating the solvent until the solid monosolvate [(C₅Me₅)₂Sm(THF)] was present. The solid was dissolved in 3,4-dihydro-2H-pyran. After evaporation, [(C₅Me₅)₂Sm(OC₅H₈)₂] was obtained quantitatively. Investigation of the physical properties of both complexes indicated the presence of samarium(II). The molecular structure determined by X-ray methods revealed Sm–O bond lengths of 2.655(6) and 2.699(7) Å for [(C₅Me₅)₂Sm(OC₅H₈)₂] and 2.630(6) and 2.770(9) Å for [(C₅Me₅)₂Sm(OC₅H₁₀)]. Divalent complexes have Sm–O bond lengths in the range of 2.62(2)–2.699(7) Å, whereas trivalent complexes range from 2.44(2) to 2.511(4) Å.

Phosphine oxides were also employed as oxygen donors as ligands. [(C₅Me₅)₂Sm(OPPh₃)(THF)] was reported by Evans.⁸¹ It was prepared by combining a solution of Ph₃PO and [(C₅Me₅)₂Sm(THF)₂] and isolated as black crystals. The compound was identified by ¹H NMR and IR spectroscopy, magnetic moment measurements and elemental analysis. No solid state structure is available. The syntheses of the phosphine oxide complexes [(C₅Me₅)₂Yb(OPMe₃)] (yellow orange) and [(C₅Me₅)₂Yb(OPMe₃)₂] (orange) were reported as well.⁸² They were prepared by adding either one or two equivalents of OPMe₃ to a toluene solution of [(C₅Me₅)₂Yb]. The solution behaviour was studied by variable temperature ¹H and ³¹P NMR spectroscopy, showing slow intermolecular exchange at room temperature. THF is competitive with phosphine oxide ligands, while diethylether is not. Addition of diethylether to a toluene solution of the monophosphine oxide complex results in precipitation of [(C₅Me₅)₂Yb(OPMe₃)₂] and concomitant formation of [(C₅Me₅)₂Yb(OEt₂)]. The authors explain this by facilitated molecular exchange via intermediate [(C₅Me₅)₂Yb(OPMe₃)(OEt₂)]. Despite isolated as crystalline solids, no X-ray crystallographic studies are available.

Reactions of [(C₅Me₅)₂Sm(THF)₂] with the anionic O-donors KOAr (OAr = OC₆H₂^tBu₂-2,6-Me-4, OC₆H₃ⁱPr₂-2,6) in THF resulted in the ionic Sm(II) phenolate complexes [(C₅Me₅)Sm(THF)₂(OAr)(μ-C₅Me₅)K(THF)₂]_n, in which a polymeric structure is formed (Scheme 10).⁸³ A similar reaction was seen by using the thiolate SC₆H₂ⁱPr₃-2,4,6 as reagent. The obtained polymeric structural motif, which is also seen by using carbanions as well as N- and P-donors, is frequently observed (see Schemes 14 and 23). Within these polymeric structures “intermolecular” interactions between the potassium atom and the C₅Me₅ ligands are observed.

Scheme 10 Polymeric chain in [(C₅Me₅)Sm(THF)₂(OR)(μ-C₅Me₅)K(THF)₂]_n.⁸³

Dimeric, divalent lanthanide complexes [(C₅Me₅)Sm(μ-OAr)]₂ (Ar = C₆H₃^tBu₂-2,6, C₆H₂^tBu₂-2,6-Me-4 and C₆H₂^tBu₃-2,4,6)) can be obtained by diffusion of ArOH into a toluene solution of [(C₅Me₅)₂Sm(THF)₂] (Scheme 11).⁸⁴ Interestingly, the samarium atom is not oxidised but one of the C₅Me₅ ligands is protonated. Single crystal X-ray analysis of [(C₅Me₅)Sm(μ-OAr)]₂ with Ar = C₆H₂^tBu₃-2,4,6 showed that it is an unsolvated dimeric samarium(II) complex with mixed ArO and C₅Me₅ ligands with an Sm(μ-O)₂Sm unit, which is exactly planar. However, the μ-OAr bridges are unsymmetric. The bond distance of the Sm(1)–O(1) bond (2.425(5) Å) is significantly shorter than that of the Sm(1)–O(1′) bond (2.512(6) Å). The solubility of these three complexes in toluene follows the order C₆H₂^tBu₃-2,4,6 > C₆H₂^tBu₂-2,6-Me-4 > C₆H₃^tBu₂-2,6. There are two other synthetic routes that yield [(C₅Me₅)Sm(μ-OAr)]₂. In addition to the above-mentioned route, the reaction of [(C₅Me₅)₂Sm(THF)₂] and [Sm(OAr)₂(THF)₃] leads to the same complexes (Scheme 11). Divalent lanthanide complexes with both ArO and C₅Me₅ ligands are extremely rare. When [(C₅Me₅)Sm(μ-OAr)]₂ (Ar = C₆H₂^tBu₃-2,4,6) is exposed to a trace amount of air, the trivalent samarium complex [(C₅Me₅)₂Sm(OAr)] is obtained.

Scheme 11 Synthesis of [(C₅Me₅)Sm(μ-OAr)]₂.⁸⁴

In addition to the reaction with alcohols, a rather unusual dinuclear samarium compound was reported from a protolysis reaction of [(C₅Me₅)₂Sm(THF)₂] with the bulky silanole (^tBuO)₃SiOH.⁸⁵ By reacting both compounds in a 2 : 3 stoichiometric ratio in toluene, green crystals of the dinuclear siloxide-bridged species [(C₅Me₅)Sm{μ-OSi(O^tBu)₃}₃Sm] were isolated in 93% yield (Scheme 12). The two Sm(II) atoms are bridged by three -OSi(O^tBu)₃ ligands. During the reaction, one pentamethylcyclopentadienyl ligand is cleaved and one samarocene is completely protolysed. Thus, one of the Sm centres is coordinated by a C₅Me₅ ligand and three bridging OSi(O^tBu)₃ units, while the second Sm(II) atom is surrounded only by these bridging silanols. The latter Sm(II) atom is additionally coordinated by three of the Si–O–C oxygen atoms of the silanols, creating a six-fold coordination sphere. The bridging Sm–O distances are in the range of 2.432(3)–2.523(3) Å. The reactivity of this complex was further investigated.⁸⁵

Scheme 12 Reaction of [(C₅Me₅)₂Sm(THF)₂] with (^tBuO)₃SiOH.⁸⁵

4.2. Monodentate nitrogen ligands

Complexes are also formed with various nitrogen donor containing ligands. However, amine complexes of the metallocenes are rarely reported in the literature.

When [(C₅Me₅)₂Yb] is synthesised in ammonia (see above), a mixed ammonia–THF adduct was formed after work up in THF.⁵⁰ The small ammonia ligand seems capable of coordinating to the small Lewis acidic Yb ion. Interestingly, the analogous reaction with [(C₅Me₅)₂Eu] yields the mono-THF solvate despite the larger ionic radius. Here, the more pronounced Lewis acidic character of the smaller Yb(II) could be the reason for the additional ammonia ligand (Scheme 3).

Green [(C₅Me₅)₂Sm(^tBuNH₂)] was prepared from [(C₅Me₅)₂Sm] and tert-butylamine in toluene, but not from (C₅Me₅)₂Sm–THF solvates (Scheme 13).⁸⁶ The amine in this case is a weaker donor than THF. When the reaction was carried out in the amine as solvent, a purple solid was obtained that was most likely [(C₅Me₅)₂Sm(^tBuNH₂)₂], but no conclusive data were collected. The solid state structure of [(C₅Me₅)₂Sm(^tBuNH₂)] was determined by single crystal X-ray crystallography, which revealed Sm–N distances of 2.804(10) and 2.737(7) Å for the two independent molecules in the asymmetric unit. In the same study the reaction with N-methyl imidazole (N-MeIm) with [(C₅Me₅)₂Sm(THF)₂] is reported. This ligand can replace THF by forming the purple double substituted complex [(C₅Me₅)₂Sm(N-MeIm)₂] (Scheme 13). The solid state structure reveals Sm–N distances of 2.618(10) and 2.673(10) Å. The authors note the resemblance to the structure of [(C₅Me₅)₂Sm(THF)₂].

Scheme 13 Reaction of tert-butyl amine and N-methyl imidazole with samarocene.⁸⁶

Although the reaction of samarocene with nitriles usually leads to redox reactions yielding Sm(III) products (see below), it is possible to isolate Lewis acid–base complexes from Sm(II) and tert-butyl nitrile.⁸⁷ Combining [(C₅Me₅)₂Sm(THF)₂] and Me₃CCN in THF and subsequent cooling yields brown crystalline [(C₅Me₅)₂Sm(NCCMe₃)₂(THF)] (Fig. 3). Structural data was obtained by single crystal X-ray crystallography to reveal a rather unusual coordination of three additional neutral ligands to the (C₅Me₅)₂Sm fragment. The N or O donor atoms of THF and Me₃CCN create a plane perpendicular to the C₅Me₅–Sm–C₅Me₅ axis. The N–Sm distances are 2.735(3) Å and the Sm–O bond length is 2.716(3) Å. If the analogous reaction is carried out in toluene as solvent, green crystals of the THF-free compound [(C₅Me₅)₂Sm(NCCMe₃)₂] were obtained in addition to other unstable products. Though the connectivity of the atoms was unambiguously determined, the structural data was too poor to discuss metric parameters.

Fig. 3 Molecular structure of [(C₅Me₅)₂Sm(NCCMe₃)₂(THF)] in the solid state emphasizing the unusual coordination of three additional ligands (Reproduced from the CIF file CCDC: 637938).⁸⁷

Lewis-acid–base interaction is also possible with phosphine imines; in particular, the interaction of the phosphine imine Et₃PNH with [(C₅Me₅)₂Yb] was investigated.⁸² The ligand behaves similar to the isoelectronic phosphine oxide ligands (see above). The 1 : 1 complex is formed during the reaction of Et₃PNH and [(C₅Me₅)₂Yb] in toluene. The solution behaviour of [(C₅Me₅)₂Yb(Et₃PNH)] was studied by variable temperature, multinuclear NMR spectroscopy. The ligand shows slow intermolecular exchange at room temperature. If a second equivalent of Et₃PNH is added, the 2 : 1 complex [(C₅Me₅)₂Yb(Et₃PNH)₂] precipitates as an orange solid from solution at −80 °C. No X-ray crystallographic data is available.

The first pyridine adduct of the metallocenes is [(C₅Me₅)₂Yb(NC₅H₆)₂].⁸⁸ It is prepared by reaction of excess pyridine with [(C₅Me₅)₂Yb(OEt₂)] in toluene. The complex was isolated as dark green prisms that were investigated with X-ray crystallography. The Yb–N bond lengths are 2.585 and 2.544 Å.

Nocton and co-workers disclosed a number of N-donor adducts of [(1,3-C₅H₃^tBu₂)₂Sm]. The N-aromatic heterocycles, pyridine, picoline, 4-tert-butyl-pyridine, isoquinoline and quinolone were coordinated to the metal atom via the N-atom.⁸⁹ As magnetic measurements show, the electronic structure of Sm(II) of these compounds is f⁶. Thus, the simple coordination adducts were formed and no redox reaction took place.

Reactions of [(C₅Me₅)₂Sm(THF)₂] with the anionic N-donors K(NRR′) (NRR′ = NHC₆H₂^tBu₃-2,4,6 and N(SiMe₃)₂) in THF resulted in the ionic Sm(II) complexes [(C₅Me₅)Sm(THF)_x(NRR′)(μ-C₅Me₅)K(THF)₂]_n, in which a polymeric structure is formed (Scheme 14).⁸³ In this structure, a potassium cation bridges the complexes by coordinating with one of each of the neighbouring C₅Me₅ ligands. In a similar reaction of [(C₅Me₅)₂Yb(THF)₂] with KN(SiMe₃)₂ the corresponding Yb(II) complex [(C₅Me₅)Yb(N(SiMe₃)₂)(μ-C₅Me₅)K(THF)₂]_n was obtained.

Scheme 14 Polymeric chain in [(C₅Me₅)Ln(N(SiMe₃)₂)(μ-C₅Me₅)K(THF)₂]_n (Ln = Sm, Yb).⁸³

4.3. Bipyridine and terpyridine, and related systems

Complexes of 2,2’-bipyridine (bipy) with all three classical metallocenes were reported in literature.^36,90,91 They are prepared by the combination of the ether containing precursors, i.e. [(C₅Me₅)₂Ln(OEt₂)], in a suitable solvent. The Sm and Yb (red-brown) complexes have been structurally characterised. The nitrogen-Yb bond lengths are 2.324 and 2.318(5) Å, while for the larger Sm ion the two distances are 2.427(2) and 2.436(2) Å, respectively. It is worth noting that due to their strong reducing properties, Yb and Sm are able to transfer electron density to the bipyridine, thereby reducing it to the radical ligand. This property is more pronounced in the corresponding samarium complexes, while for ytterbium bipy the concept of intermediate valence tautomerism was coined by Andersen, meaning that the ground state consists of a [(C₅Me₅)₂Yb(II)(bipy)] and a [(C₅Me₅)₂Yb(III)(bipy)˙⁻] in different ratios.⁹² The amount of electron density transferred to the corresponding ligand is also determined by the ligand, e.g. phenanthroline is reduced to the corresponding radical ligand,⁹³ while the oxidation state of ytterbium in the bipyridine-complex is between two and three. Bipy-complexes and complexes with bipy-related ligands such as terpyridine⁹⁴ or phenanthroline⁹³ have been extensively investigated, but a full review would go beyond the scope of this chapter. The novel charge-transfer from Yb(II) to nitrogen-containing aromatic ligands was also studied in 2 : 1 metal-to-ligand adducts of the type [Yb(II)–(ligand)–Yb(II)(C₅Me₅)₂] (ligand = tetra(2-pyridyl)pyrazine (tppz), 6′,6′′-bis(2-pyridyl)-2,2′:4′,4′′:2′′,2′′′-quaterpyridine and 1,4-di(terpyridyl)-benzene (dtb)).⁹⁵

Unique complexes were obtained by reaction of different tris(2-pyridyl)stannate derivatives with the Ln(II) metallocenes. Thus, treatment of [LiSn(2-C₅H₃N-5-Me)₃(THF)] with [(C₅Me₅)₂Ln(OEt₂)] (Ln = Eu, Yb) resulted in Ln^II sandwich complexes [Ln{Sn(2-C₅H₃N-5-Me)₃}₂] comprising the anionic tris(pyridyl)-stannate as a κ³N-coordinating ligand (Scheme 15).⁹⁶ Further reaction of [Yb{Sn(2-C₅H₃N-3-Me)₃}₂] with another equivalent of [(C₅Me)₂Yb(OEt₂)] resulted in a ligand rearrangement and the mixed complex [(C₅Me₅)Yb{Sn(2-C₅H₃N-3-Me)₃}] (Scheme 15).⁹⁷

Scheme 15 Reaction of a tris(2-pyridyl)stannate derivatives with [(C₅Me₅)₂Ln(OEt₂)] (Ln = Eu, Yb).^96,97

The divalent tin atom in the ligand backbone is suitable to coordinate to other Lewis-acid complexes, such as [(C₅H₅)₃La].⁹⁸ The resulting ligand [(C₅H₅)₃La{Sn(2-py^5Me)₃Li(THF)}] (py^5Me = C₅H₃N-5-Me) gave in the presence of [(C₅Me₅)₂Yb(OEt₂)] via substitution of the (C₅Me₅)-rings the pentametallic complex [Yb{Sn(2-py^5Me)₃La(C₅H₅)₃}₂], in which the two tris(2-pyridyl)stannate units and two unsupported Sn–La bonds in terminal positions encapsulate the Yb²⁺ cation (Scheme 16).⁹⁷ Similarly, the group 13 adducts [{Li(THF)Sn(2-py^5Me)₃}MEt₃] (M = Ga, In), which exhibit long Sn–M bonds, afforded with [(C₅Me₅)₂Eu(OEt₂)] the corresponding compounds [Eu{Sn(2-py^5Me)₃MEt₃}₂] (Scheme 16).⁹⁹ A complete different reactivity was seen by using [(C₅Me₅)₂Sm(OEt₂)] as precursor. Instead of a (C₅Me₅)-ring substitution, a redox reaction, which alters the tin ligand was seen and [(C₅Me₅)₂Sm{MEt₂(2-py^5Me)₂}] (M = Ga, In) were formed as products.⁹⁷

Scheme 16 Reaction of a tris(2-pyridyl)stannate adduct with (C₅Me₅)₂Ln(OEt₂)] (Ln = Eu, Yb).^97,99

4.4. Ytterbium – transition metal complexes bridged by redox non-innocent N-donor ligands

An elegant method for studying the communication between a divalent organo-lanthanide fragment and a transition metal complex is the use of redox non-innocent N-donor ligands as bridge. Thus, in a related synthetic method as described in the previous chapter [(C₅Me₅)₂Yb(OEt₂)] was reacted with palladium complexes, [(bipym)PdMe₂] (bipym = bipyrimidine) and [(taphen)PdMe₂] (taphen = 4,5,9,10-tetraazaphenanthrene) to give the heterobimetallic complexes [(C₅Me₅)₂Yb(bipym)PdMe₂] and [(C₅Me₅)₂Yb(taphen)PdMe₂], respectively (Scheme 17). In these complexes, an electron is transferred from the ytterbocene fragment to the ligand, increasing the electron-donating properties of the ligands. As shown by oxidative addition of MeI, this strongly influences the reactivity of the Pd species.¹⁰⁰

Scheme 17 Synthesis of the heterobimetallic complexes [(C₅Me₅)₂Yb (bipym)PdMe₂] and [(C₅Me₅)₂Yb(taphen)PdMe₂].¹⁰⁰

In addition to the Pd complex, also the corresponding Ni complex [(bipym)NiMe₂] was treated with [(C₅Me₅)₂Yb(OEt₂)]. This resulted in the analogue heterobimetallic complex [(C₅Me₅)₂Yb(bipym)NiMe₂], which was further reacted with CO to give [(C₅Me₅)₂Yb(bipym)Ni(CO)₂] and acetone. By comparison with the reactivity of the parent [(bipym)NiMe₂] complex, it was shown that the divalent lanthanide fragment has a strong influence on the reaction kinetics.¹⁰¹ [(C₅Me₅)₂Yb(bipym)NiMe₂] was also used as very efficient catalyst for alkene isomerization in the presence of catecholborane.¹⁰²

In another study, deprotonated 2-pyrimidin-2-yl-1H-benzimidazole (Hbimpm) was coordinated with a NiMe₂ fragment. The corresponding ionic complex [K(bimpm)NiMe₂] was subsequently reacted with [(C₅Me₅)₂Yb(OEt₂)]. Instead of a simple addition product, the bimpm underwent a coupling reaction, which is a result of the single electron transfer process from the {(C₅Me₅)₂Yb} fragment.¹⁰³

Intramolecular electron transfer was also observed in a complex in which a nickel atom and a {(C₅Me₅)₂Yb} fragment are bridged by a macrocyclic biquinazoline ligand (Mabiq). The mixed Yb–Ni complex, [(C₅Me₅)₂Yb(Mabiq)Ni]B(C₆F₅)₄, was synthesised upon reaction of [Ni(II)(Mabiq)]B(C₆F₅)₄ with [(C₅Me₅)₂Yb(OEt₂)] (Scheme 18). As supported by spectroscopic studies the complex is best described as [(C₅Me₅)₂Yb(III)(Mabiq˙)Ni(II)]⁺ with a ligand-centred radical delocalised over both the diketiminate and bipyrimidine units of the Mabiq ligand.¹⁰⁴

Scheme 18 Synthesis of [(C₅Me₅)₂Yb(Mabiq)Ni(II)]⁺.¹⁰⁴

Electronic and magnetic communication was also investigated between trimetallic mixed actinide–lanthanide molecular complexes. The target compounds [(C₅Me₅)₂An{N=C(CH₂C₆H₅)(tpyYb(C₅Me₄R)₂)}]₂ (An = Th, U; R = Me, Et; tpy = terpyridine) were obtained by reacting uranium(IV)- and thorium(IV)-bis(ketimide) complexes with [(C₅Me₅)₂Yb(OEt₂)]¹⁰⁵ and [(C₅Me₄Et)₂Yb(OEt₂)] (Scheme 19).¹⁰⁶ As linker between the metals terpyridyl-functionalised ketimides were employed. The Yb-ions show a valence equilibria between the divalent and the trivalent oxidation state and exhibit rich electrochemical behaviour consistent with electronic coupling between the actinide and Yb(II/III)tpy˙⁻ moieties. Magnetic studies of the uranium complex indicate a coupled magnetic state between the U(IV) and Yb(III)tpy˙⁻ groups at low temperatures.

Scheme 19 Synthesis of [(C₅Me₅)₂An{N=C(CH₂C₆H₅)(tpyYb(C₅Me₄R)₂)}]₂.^105,106

In a follow-up study, the corresponding Sm complexes, which were prepared in a similar fashion as the Yb complexes from [(C₅Me₄Et)₂Sm(OEt₂)], were synthesised. As seen for the Yb complex, a strong electronic coupling between the metals was observed in the case of the uranium compound [(C₅Me₅)₂U{N=C(CH₂C₆H₅)(tpySm(C₅Me₄Et)₂)}]₂.¹⁰⁶

4.5. Carbon donors

Labile carbon monoxide adducts of the solvent free metallocenes are formed under CO pressure in toluene or methylcyclohexane.^107,108 Ethereal solvents are not suitable because the strong oxygen donors prevent carbon monoxide coordination. The formed labile complexes were studied by IR spectroscopy and other spectroscopic methods. The CO stretching frequencies of [(C₅Me₅)₂Sm(CO)] (2153 cm⁻¹) and [(C₅Me₅)₂Eu(CO)] (2150 cm⁻¹) are greater than those of free CO (2134 cm⁻¹), indicating a complete absence of back bonding. In the respective ytterbocene compound, in contrast, somewhat lower CO stretching frequencies were observed. Ytterbocene also forms a CO complex with two CO ligands that dominates at higher pressures. A DFT study carried out in 2002 revealed that the lower back bonding frequency of the Yb bound CO is due to isocarbonyl formation.¹⁰⁹

The coordination in all carbon monoxide complexes is reversible and CO can be removed in vacuo. Due to their lability, no X-ray crystallographic studies could be carried out on carbon monoxide complexes.

Isocyanides are ligand systems that are very closely related to carbon monoxide. The first 1 : 2 complex of [(C₅Me₅)₂Yb] and other ytterbocenes with the isocyanide 2,6-Me₂C₆H₃NC was reported by Andersen and co-workers.¹⁰⁷ It was synthesised by combining [(C₅Me₅)₂Yb(OEt₂)] and 2,6-Me₂C₆H₃NC in toluene. The mean Yb–C bond distance in [(C₅Me₅)₂Yb(2,6-Me₂C₆H₃NC)₂] is 2.538 Å. No isocyanide complexes are reported for [(C₅Me₅)₂Eu], whereas the reaction of [(C₅Me₅)₂Sm] with an isocyanide resulted in reductive cleavage of the ligand (see below). However, with 1,2,4-C₅H₂^tBu₃ as the ligand of the samarium atom, the crystal structure of an isocyanide complex could be successfully resolved.¹¹⁰

Andersen and co-workers also reported the synthesis of two phosphine–ylidene ytterbium complexes, namely [(C₅Me₅)₂Yb(Me₂PhPCHSiMe₃)] and [(C₅Me₅)₂Yb(Me₂PhPCH₂)] (Scheme 20).⁸² They were prepared by combining the respective ylidene with [(C₅Me₅)₂Yb] in toluene. Extensive, multinuclear, temperature-variable NMR spectroscopic investigations were carried out to elucidate the solution behaviour of the phosphine ylidene complexes as well as the interaction of the carbon with the ytterbium atom. Based on this, a mechanism for the fluxional processes in solution was suggested (please refer to the literature for further details).⁸² Briefly, the compounds underwent fast intermolecular exchange at room temperature. Dark green crystals of [(C₅Me₅)₂Yb(Me₂PhPCHSiMe₃)] were investigated by X-ray crystallography to reveal a Yb–C bond distance of 2.69(2) Å.

Scheme 20 Reaction of a phosphine ylide with [(C₅Me₅)₂Yb].⁸²

The first two lanthanide N-heterocyclic carbene (NHC) complexes reported in literature were [(C₅Me₅)₂Sm(NHC)] and the corresponding 1 : 2 complex [(C₅Me₅)₂Sm(NHC)₂] (NHC = 1,3,4,5-tetramethylimidazol-2-ylidene) (Scheme 21).¹¹¹ Despite the rather soft carbon donor atom, [(C₅Me₅)₂Sm(NHC)] was prepared by combining [(C₅Me₅)₂Sm(THF)] with one equivalent of NHC in toluene, the resulting [(C₅Me₅)₂Sm(NHC)] being a dark green, high melting compound. By adding a second equivalent of NHC, the bis-carbene-complex [(C₅Me₅)₂Sm(NHC)₂] was obtained. It was possible to structurally investigate [(C₅Me₅)₂Sm(NHC)₂] by single crystal X-ray crystallography. The Sm–NHC distances are 2.837(7) and 2.845(7) Å, respectively.

Scheme 21 Successive reaction of [(C₅Me₅)₂Sm(THF)₂] with two equivalents of 1,3,4,5-tetramethylimidazol-2-ylidene.¹¹²

Furthermore, [(C₅Me₅)₂Sm(C₃N₂Me₂ⁱPr₂)] was reported.¹¹² The 1,3-diisopropyl-4,5-dimethylimidazoline-2-ylidene reacts with [(C₅Me₅)₂Sm(THF)₂] which results in the carbene-complex. The complex was characterised by single crystal X-ray structure analysis. The Cp–Sm distances are 2.5823(16) and 2.5957(15) Å, and the Sm–C(carbene) distance is 2.782(3) Å, which is considerably shorter than in the prior described compound. Another Sm–NHC complex ligated by 1,3-diisopropyl-4,5-dimethylimidazoline-2-ylidene is similar.¹¹²

In addition to the Sm–NHC complexes, Yb–NHC compounds of composition [(C₅Me₄Et)₂Yb(NHC)] (NHC = 1,3,4,5-tetramethylimidazol-2-ylidene and 1,3-diisopropyl-4,5-dimethylimidazol-2-ylidene) were reported. Due to the smaller ion radius of Yb compared to Sm, only one NHC ligand is bound to the metal atom. The Yb–C bond length for the tetramethyl derivative is 2.669(4) Å.¹¹³

Compounds in which the triple bond of alkynes act as an electron donor for a Lewis acidic metal are common for transition metals. Even after thorough literature research, however, we are only aware of one η²-coordinated alkyne complex of the lanthanide metallocenes.¹¹⁴ Combining desolvated [(C₅Me₅)₂Yb] and 2-butyne in pentane, [(C₅Me₅)₂Yb(η²-(H₃–C≡C–CH₃)] was obtained as dark purple/red needles (Scheme 22). The complex was investigated with IR, ¹H and ¹³C NMR spectroscopy and its solid state structure was elucidated by X-ray crystallography. Based on the investigations carried out, the complex fragments only weakly influence each other and little or no backbonding is observed. The average Yb carbon distance is 2.850 Å.

Scheme 22 Reaction of [(C₅Me₅)₂Yb] with 2-butyne and [(η²-C₂H₄)Pt(PPh₃)₂].¹¹⁴

[(C₅Me₅)₂Yb] can initiate the polymerization of ethylene, though no complex with an olefin as a ligand was isolated. However, the reaction of base-free [(C₅Me₅)₂Yb] with the platinum ethylene complex [(η²-C₂H₄)Pt(PPh₃)₂] in toluene yielded the ethylene bridged, heterobimetallic complex [(C₅Me₅)₂Yb(μ-C₂H₄)Pt(PPh₃)₂] as red crystals (Scheme 22).¹¹⁵ IR and NMR spectroscopy indicate an exchanging system in solution. With respect to the solid structure, the Yb–ethylene–carbon distances are 2.770 and 2.793 Å, respectively. The hydrogen atoms of the ethylene moiety were refined isotropically and are slightly tilted towards the ytterbium atom, resulting in four distinct Yb–H distances of 2.58(5), 2.64(3), 3.09(4) and 3.15 (3) Å. The authors conclude that the interaction between the olefin and the ytterbium centre is rather weak.

Another complex, which was synthesised as polymerization catalyst, is [(C₅Me₅)Ln{CH(SiMe₃)₂}(C₅Me₅)K(THF)₂]_n (Ln = Sm, Eu, Yb).¹¹⁶ The reaction of [(C₅Me₅)₂Ln(THF)₂] (Ln = Sm, Eu, Yb) with KCH(SiMe₃)₂ in THF resulted in the Ln(II) alkyl complexes in 90–92% isolated yields (Scheme 23). The Sm–C₅Me₅ average bond distances in the Sm-complex are 2.85(2) and 2.86(2) Å and are therefore in the 2.84(2)–2.97(2) Å bond distance range of those found in the analogous Sm(II) complexes with a heteroatom-containing monodentate anionic ligand. The Sm–CH(SiMe₃)₂ bond distance is 2.64(1) Å and is hence between the Sm–C bond lengths of Sm(II) (2.787(5) and 2.845(5) Å) and Sm(III) alkyl complexes (2.48(1) Å). The Eu–C₅Me₅ average bond distances of the Eu-complex are 2.83(2) and 2.91(2) Å and are comparable with those of the Sm–C₅Me₅ bonds in the Sm-complex due to the similar ion sizes of Eu(III) and Sm(II). The Eu–CH(SiMe₃)₂ bond distance is 2.65(1) Å. The Yb–C₅Me₅ average bond distances are 2.74(3) and 2.79(3) Å and are in the range of those found in comparable Yb(II) complexes.

Scheme 23 Polymeric chain in [(C₅Me₅)Ln{CH(SiMe₃)₂}(C₅Me₅)K(THF)₂]_n (Ln = Eu, Sm, Yb).¹¹⁶

Reaction of unsolvated [(C₅Me₅)₂Yb] and the beryllium piano stool complex [(CH₃)Be(C₅Me₅)] in pentane yields the uncommon methyl bridged complex [(C₅Me₅)₂Yb(μ-CH₃)Be(C₅Me₅)] as dark orange crystals (Scheme 24).¹¹⁷ The interaction in solution proved to be rather weak and a fast exchange process was observed by NMR spectroscopy. The solid state structure reveals a Yb–C distance of 2.766 Å, which is similar to the distance in the bridging olefin complex of platinum and shorter than the Yb–C distance in the Yb-alkyne complex (see above). The hydrogen atoms derived from the Fourier difference map show an average distance of 2.59 ± 0.08 Å to the ytterbium atom. The authors emphasise the complex as a model for methane complexation by ytterbium.

Scheme 24 Structure of [(C₅Me₅)₂Yb(μ-CH₃)Be(C₅Me₅)].⁵¹

A related complex is formed by the reaction of [(C₅Me₅)₂Yb(THF)] and AlEt₃ in toluene. The ethyl bridged, heterobimetallic complex [(C₅Me₅)₂Yb(μ-Et)AlEt₂(THF)] was isolated as green crystals.¹¹⁸ AlMe₃ and Al(ⁱC₄H₉)₃ give similar structures. The structure of the ethyl derivative was determined. The Al and Yb centres are linked by the ethyl moiety attached to the aluminium ion. The latter is tetrahedrally coordinated by three ethyl units and one THF molecule. Both carbon atoms of the bridging C₂H₅ unit bind to the ytterbium ion. The Yb–C distances are 2.854(18) and 2.939(12) Å. The bond is weak and therefore broken by the addition of THF or other donors. The use of the ytterbocene THF solvate as a precursor is still possible because the organoaluminum compound is a stronger Lewis acid and therefore removes THF from the Yb atom.

The solvent free metallocenes [(C₅Me₅)₂Ln] (Ln = Sm, Eu, Yb) react with [Et₃NH][BPh₄] to form the divalent π-coordinated tetraphenyl borates [(C₅Me₅)Ln(μ-η⁶:η¹-Ph)₂BPh₂] (Ln = Sm, Eu, Yb). Two of the phenyl rings of the tetraphenylborate counteranion coordinate η⁶ to the lanthanide atoms (Scheme 25).^119,120 [(1,3-C₅H₃^tBu₂)₂Eu] reacts in a similar way to give [(1,3-C₅H₃^tBu₂)Eu(μ-η⁶:η¹-Ph)₂BPh₂].¹²¹

Scheme 25 Synthesis of [(C₅Me₅)Ln(μ-η⁶:η¹-Ph)₂BPh₂] (Ln = Sm, Eu, Yb).^119,120

4.6. Silicon, phosphorus, and sulphur donors

Complexes containing ligands with soft donor atoms of the third row period of the periodic table are comparatively less common in combination with divalent metallocenes of the lanthanides. Ligands with sulphur, phosphorus or silicon are easily replaced by oxygen containing ligands like THF. Therefore, the solvent free metallocenes or at least the diethylether complexes have to be used as precursor.

Recently, lanthanide(II) silyl complexes were reported.¹¹⁶ [(C₅Me₅)₂Ln(THF)₂] reacted with KH and H₃SiPh yielding [(C₅Me₅)Ln(SiH₃)(THF)(C₅Me₅)K(THF)]_n (Ln = Yb(II), Sm(II), Eu(II)) (Scheme 26). Due to poor crystal quality, the Sm complex could not be characterised, whereas the Eu and Yb-analogues were investigated by crystallographic studies. The Ln–(C₅Me₅) average bond distances are for Ln = Eu 2.85(1) and 2.86(1) Å, and for the Yb-complex 2.76(1) for both C₅Me₅ units. The Eu–SiH₃ and Yb–SiH₃ average bond distances are 3.239(3) and 3.091(3) Å, respectively.

Scheme 26 Polymeric chain in [(C₅Me₅)Ln(SiH₃)(THF)(C₅Me₅)K(THF)]_n (Ln = Eu, Sm, Yb).¹¹⁶

The first [(C₅Me₅)₂Ln-silylene] complex was reported in 2003 by West and Evans.¹²² [(C₅Me₅)₂Sm-(NHSi)] was prepared by combining solvent-free samarocene and NHSi (NHSi = 1,3-di-tert-butyl-2,3-dihydro-1H-1,3,2-diazasilol-2-ylidene) in toluene. As expected, the bond between the Si(II)-centre and Sm is rather weak. The silylene ligand is immediately replaced by THF if present in solution. The solid structure shows a Sm–Si bond length of 3.1903(10) Å. The silylene coordinates asymmetrically and is slightly tilted to one side, allowing a long-range interaction with one tert-butyl methyl group of the silylene. In the course of renewed interest in low-valent main group donors, two new silylene complexes of decamethylsamarocene, [(C₅Me₅)₂Sm(Si(OC₆H₄-2-^tBu)){(N^tBu)₂CPh}] and [(C₅Me₅)₂Sm(Si(O^tBu)){(N^tBu)₂CPh}] were reported in 2015.¹²³ They were obtained from the reaction of [(C₅Me₅)₂Sm(OEt₂)] and the respective four-membered N-heterocyclic silylenes {PhC(N^tBu)₂}SiO^tBu and {PhC(N^tBu)₂}SiO(2-^tBu-C₆H₄) in toluene and isolated as emerald green solids. The solid state structures of both complexes were determined, whereby long Sm–Si distances of 3.4396(15) Å ({PhC(N^tBu)₂}SiO^tBu) and 3.3142(18) Å in {PhC(N^tBu)₂}SiO(2-^tBu-C₆H₄) indicate weak interactions. The compounds were extensively characterised and investigated by quantum chemical calculations on the DFT level confirming the weak bond without covalent character.

Among the phosphorus compounds, the reactions of the etherates [(C₅Me₅)₂Ln(OEt₂)] (Ln = Eu, Yb) and the bidentate phosphines Me₂PCH₂CH₂PMe₂ and Me₂PCH₂PMe₂ were investigated.¹²⁴ In the first case, insoluble coordination polymers are formed with Me₂PCH₂CH₂PMe₂. In the second case, the sterically less flexible Me₂PCH₂PMe₂ yields the soluble complexes [(C₅Me₅)₂Ln(Me₂PCH₂PMe₂)] (Eu: red, Yb: green). According to NMR investigations of the Yb compound, the phosphine coordinates in a bidentate fashion. No X-ray data is available for these complexes.

Complexes of monodentate phosphines were reported for ytterbocene. Due to the reduced donor strength of the phosphines, the solvent-free complex [(C₅Me₅)₂Yb] was used as precusor.⁸² Upon reaction with PMe₃ in toluene, the green 1 : 2 complex [(C₅Me₅)₂Yb(PMe₃)₂] was formed and isolated from toluene at −80 °C. A similar protocol with PEt₃ yielded the blue 1 : 1 complex [(C₅Me₅)₂Yb(PEt₃)]. The solution behaviour of the complexes was investigated and NMR spectroscopy shows that the interaction between phosphine and ytterbium is very weak in solution. Although isolated as crystals, no X-ray crystallographic data is available.⁸²

As with the lanthanide(II) silyl complexes, a phosphide complex was also synthesised (Scheme 27). Reaction of [(C₅Me₅)₂Sm(THF)₂] with KPH(C₆H₂^tBu₃-2,4,6) resulted in decomplexion of one C₅Me₅ ligand from the Sm(II) atom and the formation of the polymeric species [(C₅Me₅)Sm(THF)(μ-PHAr)K(C₅Me₅)(THF)]_∞ (Ar = C₆H₂^tBu₃-2,4,6) (Scheme 27).⁸³ Within this compound a “C₅Me₅K” unit is bonded to the phosphide site with its K atom. The Sm–P bond (3.234(2) Å) is significantly longer than those found in samarium(II) bis(phosphide) complexes. A similar reaction was seen by using the thiolate SC₆H₂ⁱPr₃-2,4,6 as reagent.⁸³

Scheme 27 Synthesis of [(C₅Me₅)Sm(THF)(μ-PHAr)K(C₅Me₅)(THF)]_∞.⁸³

More recently, the reactions between metallocenes of ytterbium or samarium with tetramethylbiphosphinine (tmbp), a bipyridine analogue with phosphorus donors, were reported.¹²⁵ In this case, the diethyl solvates were employed for the synthesis of [(C₅Me₅)₂Ln(tmbp)] (Ln = Sm, Yb) in toluene as dark brown crystals. The solid state structures were determined by X-ray diffraction (Fig. 4). The Ln–P bond lengths in the solid state are 2.909(2) and 2.972(2) Å for the Sm compound, and 2.872(2) Å and 2.983(2) Å for Yb one respectively. The complexes were extensively investigated by NMR spectroscopy and theoretical calculations to assess the electron density transferred from the ligand to the lanthanide ion. The authors aimed to investigate whether the metals are in a divalent oxidation state or in a trivalent state with a reduced tmbp ligand. In the case of Yb, the oxidation state +II was found to be the most accurate, while in the samarocene complex the Sm ion is best described as trivalent.

Fig. 4 Molecular structures of tetramethylbiphosphinine comlexes of ytterbocene and samarocenene in the solid state (Reproduced from the CIF file CCDC 1452181).¹²⁵

It should be noted that, to the best of our knowledge, no Lewis acid–base adduct with neutral sulphur donor ligands has been reported in the literature, probably due to the very weak coordination abilities of the soft sulphur donors with regard to the hard character of the lanthanide ions.

4.7. Aluminium and gallium compounds

The low valent aluminium compound [(C₅Me₅)Al]₄ can act as a Lewis base due to its free electron pair. Heating [(C₅Me₅)₂Eu] or [(C₅Me₅)₂Yb] and [(C₅Me₅)Al]₄ in an evacuated ampoule resulted in red (Eu) or green (Yb) crystals of [(η⁵-C₅Me₅)₂Ln-Al(η⁵-C₅Me₅)], respectively.¹²⁶ These compounds were the first compounds with an aluminium-4f metal bond. Both compounds were structurally investigated by single crystal X-ray crystallography (Fig. 5). The Eu–Al bond is 3.3652(10) Å long, whereas the Yb–Al bond is slightly shorter with 3.1981(11) Å. The compounds even decompose in hydrocarbon solvents. However, they were thoroughly characterised using standard analytical methods and special care was taken to ensure that there was no hydride species between aluminium and the lanthanide that would lead to an Ln(III) complex. In addition, extensive quantum chemical calculations were carried out. According to these calculations, the interaction between the lanthanide and the aluminium atom is mainly electrostatic, resulting in very low binding energies, which is consistent with the low stability of both compounds in solution.

Fig. 5 Molecular structure of [(η⁵-C₅Me₅)₂Ln-Al(η⁵-C₅Me₅)] (Ln = Eu, Yb) in the solid state (based on the CIF file CCDC 296933).¹²⁶

Complexes of [(C₅Me₅)₂Eu] and [(C₅Me₅)₂Yb] and the heavier Ga congener of [(C₅Me₅)Al]₄, namely [(C₅Me₅)Ga], have also been prepared.¹²⁷ The reaction of solvent free [(C₅Me₅)₂Eu] and THF deficient [(C₅Me₅)₂Yb(THF)_1−n] with [(C₅Me₅)Ga] in toluene in the correct stoichiometric ratio yielded [(η⁵-C₅Me₅)₂Eu–Ga(η⁵-C₅Me₅)₂)] and [(η⁵-C₅Me₅)₂(THF)Yb-Ga(η⁵-C₅Me₅)₂], respectively. In contrast to the Al compounds, the two Ga complexes can be prepared and are stable in solution. Both compounds were characterised by standard analytical methods including X-ray crystallography. The Eu–Ga contacts in [(η⁵-C₅Me₅)₂Eu–Ga(η⁵-C₅Me₅)₂] are 3.2499(6) Å and 3.3907(6) Å. In [(η⁵-C₅Me₅)₂(THF)Yb–Ga(η⁵-C₅Me₅)₂], the Yb–Ga distance is 3.2872(4) Å. The Yb–O distance from the THF–oxygen atom is 2.418(2) Å. The most striking difference between [(η⁵-C₅Me₅)₂Eu–Ga(η⁵-C₅Me₅)₂] and [(η⁵-C₅Me₅)₂Eu–Al(η⁵-C₅Me₅)] is that two Ga(C₅Me₅) fragments coordinate to the europium atom while in the analogue Al compound only one Al(C₅Me₅) donor coordinates to the europium centre. This is rationalised by the longer Eu–Ga bonds compared to the Eu–Al bonds, allowing the coordination of two ligands.

5. Reactivity as a reducing agent

As early as 1984, Deacon et al. reported that [(C₅H₅)₂Yb(DME)] can be oxidised with thallous, mercuric, argentic and cuprous salts to give [(C₅H₅)₂Yb-X] (X = O₂CMe, O₂CC₆F₅, O₂CC₅H₄N, Cl, Br, I, C≡CPh, C₆F₅, (MeCO)₂CH, (PhCO)₂CH).⁶ In recent years, the focus has been on substituted metallocenes, mainly the decamethyl derivatives. In general, [(C₅Me₅)₂Ln] (Ln = Sm, Eu, Yb) or its solvates can act as reducing agents according to the reaction shown in Scheme 28 (i.e. Eu(III)/Eu(II) = −0.35 V, Yb(III)/Yb(II) = −1.1 V, Sm(III)/Sm(II) = −1.5 V vs. normal hydrogen electrode).¹²⁸ While europocene is a weak reductant and quite stable in the oxidation state +II, ytterbocene is more strongly reducing.

Scheme 28 Redox reactivity of the three pentametyhlcyclopentadienyl derivatives (Ln = Eu, Sm, Yb).

Samarocene is the strongest reducing agent in this series being able to react with dinitrogen (see below, Scheme 72). Its reactivity has been compared with the reactivity of alkaline metals, which is why most of the work on redox reaction involving the title compounds revolves around Sm. The reduction strength can thus be ranked in the order (C₅Me₅)₂Eu < (C₅Me₅)₂Yb < (C₅Me₅)₂Sm.¹²⁹ In addition, divalent thulium metallocenes, which have a very strong reduction potential, are known. However, due to their high reactivity these compounds are difficult to handle and only a few reactions were reported.⁴¹ Note, [(C₅Me₅)₂Tm(solvate)_n] was not isolated. Only derivatives with bulkier substituents are known. The following section provides an overview of the reduction reactivity patterns of the divalent metallocenes.

5.1. Reactivity towards organic reagents

5.1.1. Reactivity towards unsaturated hydrocarbons. The two more reactive metallocenes of Yb and Sm, especially samarocene, readily react with carbon–carbon multiple bonds of alkenes and alkynes. In particular, the reactivity of [(C₅Me₅)₂Sm] and its solvates towards organic substrates has been extensively studied by Evans et al. In general, the reactivity is based on the transfer of one electron from the lanthanide(II) centre to the multiple bonds of the organic substrate.

As early as in the first publication of the synthesis of [(C₅Me₅)₂Sm(THF)₂], Evans mentioned the ability of the molecule to enhance the hydrogenation of 3-hexyne to cis-3-hexene.¹⁴ The ability of the samarocene to reduce carbon–carbon triple bonds was later demonstrated by the reaction of Ph–C≡C–Ph and [(C₅Me₅)₂Sm(THF)₂] (Scheme 29).^130,131 According to elemental analysis the isolated black material has the empirical formula [(C₅Me₅)₂SmCC₆H₅]. By further spectroscopic analysis, the structure of a bis-samarium complex with a bridging enediyl-ligand [(C₅Me₅)₂Sm(C₆H₅)C=C(C₆H₅)Sm(C₅Me₅)₂] was identified, and the oxidation state of +III was confirmed. Hydrolysis yielded pure trans-stilbene. Interestingly, by adding THF to the complex, divalent [(C₅Me₅)₂Sm(THF)₂] was formed again. Reaction of the enediyl complex with molecular hydrogen yields the hydride-bridged Sm(III) complex [(C₅Me₅)₂Sm(μ-H)]₂.

Scheme 29 Reaction of [(C₅Me₅)₂Sm(THF)₂] with diphenyl acetylene.^130,131

The chemistry of [(C₅Me₅)₂Sm(solvate)_n] and terminal alkynes as well as those of the products obtained has been widely studied.^132,133 Samarocene reacts with terminal alkynes HC≡C–R to form trivalent products of the type [(C₅Me₅)₂Sm(C≡C–R)(THF)] (R = –Ph, –CH₂CH₂Ph, –CH₂NEt₂, –CH₂CH₂CHMe₂, –CHMe₂, –CMe₃) when the THF solvate of the metallocene was used as reactant (Scheme 30).^132,133 Similar reactivity was observed in the reaction of [(1,3-C₅H₃^tBu₂)₂Sm] with phenylacetylene.¹³⁴ The complex [(C₅Me₅)₂Sm(C≡C–Ph)(THF)] was structurally characterised, however from the reaction of trivalent [(C₅Me₅)₂Sm(THF)₂(BPh₄)] and the corresponding potassium alkynide.¹³³ The carbon Sm–C(alkyne) bond length is 2.50(2) Å.

Scheme 30 Reactions of [(C₅Me₅)₂Sm(THF)₂] with various alkynes.^132,133

The reactivity of these alkynide complexes, regarding the coupling of the alkynide moieties was investigated in depth.¹³² If the reaction between solvent-free [(C₅Me₅)₂Sm] and terminal alkynes was carried out in the absence of THF, coupling of two alkynide moieties occurs to yield structurally diverse Sm complexes (Scheme 31). The structure of the resulting bridging ligand strongly depends on the nature of the electronic and steric factors of the substituents at the alkynide ligand. Coupling does not occur for the bulky C≡C-CMe₃ moiety (Scheme 31). For example, [(C₅Me₅)₂Sm(C≡C–Ph)] yields the coupled trienediyl complex [{(C₅Me₅)₂Sm}₂(μ-η²:η²-Ph-C=C=C=C–Ph)], which is also obtained from the reaction of [(C₅Me₅)₂Sm(THF)₂] and the butadiyne Ph–C≡C–C≡C–Ph (Scheme 30).^135,136 It is also possible to synthesise the coupled complex by thermolysis of the solvated THF complex [(C₅Me₅)₂Sm(C≡C–Ph)(THF)] or from the divalent [(C₅Me₅)₂Sm] and PhC≡CH (Scheme 31).¹³⁶ The use of unsolvated [(C₅Me₅)₂Sm] also yields the coupled products for RC≡CH (R = (CH₂)₂Ph, CH₂NEt₂, (CH₂)₂CHMe₂ and CHMe₂). It should be noted that this synthesis route does not work for any other substituent on the alkynide unit. The authors point out that the reactivity of these compounds has to be evaluated on a case-to-case basis, considering the properties of the solvent and the steric and electronic nature of the alkynide ligands.

Scheme 31 Reactions of unsolvated [(C₅Me₅)₂Sm] with various alkynes.¹³⁶

In contrast, decamethylytterbocene reacts with PhC≡CH to form the red trinuclear, mixed valence compound [(C₅Me₅)₂Yb(μ-C≡CPh)₂Yb(μ-C≡CPh)₂Yb(C₅Me₅)₂] (Scheme 32).¹³⁷ In this case, a redox reaction and a concomitant acid–base elimination of two HC₅Me₅ molecules formally take place. However, no closer elucidation of the mechanism is reported. The two outer Yb atoms, which are coordinated by two C₅Me₅-ligands and two bridging phenylacetylene moieties, are in the oxidation state +III as confirmed by the averaged Yb–C pentamethylcyclopentadienyl-distances of 2.61(2) Å. The central Yb atom remains in the divalent oxidation state and is surrounded by four bridging phenylacetylide ligands in a distorted tetrahedral fashion (average Yb–C distance 2.52 Å). Furthermore, four close distances to the C–Ph atoms of the respective ligands are found in the solid state structure, saturating the coordination sphere of the divalent Yb atom. An electron exchange between the di- and the trivalent Yb centres was excluded by magnetic susceptibility studies.

Scheme 32 Reaction of with phenylacetylene.¹³⁷

Interestingly, [(C₅Me₅)₂Sm(THF)₂] reacts in THF with acetylene to yield the binuclear acetylide-bridged complex [{(C₅Me₅)₂Sm(THF)}₂(μ-η¹:η¹-C₂)] with concomitant H₂ evolution (Scheme 33).¹³⁸ When the reaction was carried out with desolvated samarocene, no clear product was isolated. The increased reactivity of [(C₅Me₅)₂Sm] resulted in an intractable mixture of products. A comparable reactivity was observed by carrying out the reaction of [(C₅Me₅)₂Sm(THF)₂] and acetylene in toluene, which is a weakly coordinating solvent (Scheme 33). The formation of by-products was, to a lesser extent, also observed with the use of THF as a solvent in NMR studies. The X-ray crystallographic analysis of [{(C₅Me₅)₂Sm(THF)}₂(μ-η¹:η¹-C₂)] reveals a rare bridging motive of the carbon ligand and an almost linear arrangement of the metals and the carbon atoms. In addition to the acetylide ligand, a THF molecule is bound to each of the samarium atoms. The Sm–C bond lengths are 2.438(7) and 2.448(8) Å. Note that the C–C bond length is 1.21(1) Å, meaning only a slight deviation from acetylene. Hence, the alkyne is not activated.

Scheme 33 Reaction of [(C₅Me₅)₂Sm(THF)₂] and acetylene.¹³⁸

Like the triple bonds of alkynes, the double bonds of alkenes can also be reduced by [(C₅Me₅)₂Sm]. The simplest alkene ethylene is polymerised to polyethylene by the solvated as well as the desolvated complex.¹³⁹ With substituted alkenes like propylene, trans-butene and allylbenzene Sm(III) allyl-complexes of composition [(C₅Me₅)₂Sm(allyl)] are formed. In some cases, the THF solvate shows no reactivity, pronouncing once more the ability of THF to block the reactive site of the [(C₅Me₅)₂Sm] fragment. The allyl ligand is in all cases in a η³-coordination mode in the solid state. With 1,3-butadiene, a coupling reaction of two allyl ligands occurred to give a bis-samarocene complex with a bridging bis-allyl ligand consisting of eight CH_n-moieties. Depending on the substituents, the three different Sm–C^allyl distances vary in the range of 2.551(17)–2.730(17) Å. Furthermore, [(C₅Me₅)₂Sm] isomerises cis-stilbene to the trans isomer.¹³⁹ The dark maroon-purple, dinuclear complex [{(C₅Me₅)₂Sm}₂(μ-η²:η⁴-PhCHCHPh)] was isolated from the stoichiometric reaction of decamethylsamarocene and cis-stilbene (Scheme 34). Here, two [(C₅Me₅)₂Sm] fragments are bridged by the double bond of stilbene. In addition, both stilbene phenyl rings exhibit close contacts of the ortho carbon atom with one of the samarium centres. This results in the asymmetric tilt of the phenyl moieties towards one samarium centre. A similar dinuclear product, [{(C₅Me₅)₂Sm}₂(μ-η²:η⁴-PhCHCH₂)], was obtained from the reaction of decamethylsamarocene with styrene (Scheme 34). The [(C₅Me₅)₂Sm] fragments coordinate on opposite sides to the styrene double bond. Both the ipso as well as the ortho carbon of the phenyl ring coordinate to one of the Sm ions which results in asymmetric η⁴:η² coordination and unequal bond Sm–C double bond distances. Distances in the range of 2.537(15) to 2.732(15) Å were determined. Shorter distances belong to the η²-bound samarium fragment. The phenyl-C-Sm contacts are 2.85(2) and 2.77(2) Å, respectively.

Scheme 34 Reactions of [(C₅Me₅)₂Sm] with styrene and stilbene.¹³⁶

In contrast to the chemistry described above, the reaction of [(C₅Me₅)₂Sm] with the diene monomers isoprene (C₅H₈) and myrcene (C₁₀H₁₆) resulted in weakly coordinated hydrocarbon complexes.¹³¹ The isoprene complex [{(C₅Me₅)₂Sm}₂(μ-η²:η⁴-CH₂CHC(Me)CH₂)] was formed at −36 °C in toluene (Scheme 35). The addition of THF to this compound quantitatively resulted in the divalent species [(C₅Me₅)₂Sm(THF)₂]. The reaction of myrcene with [(C₅Me₅)₂Sm] was analogous to that of isoprene. The olefin complex [{(C₅Me₅)₂Sm}₂(μ-η²:η⁴-CH₂CHC(CH₂)CH₂CH₂CHC(Me)₂] was obtained as product. The addition of THF reversed the reaction and myrcene as well as [(C₅Me₅)₂Sm(THF)₂] was formed.¹³¹

Scheme 35 Reactions of [(C₅Me₅)₂Sm] with isoprene.¹³¹

The reaction of samarocene and cyclopentadiene results in trivalent [(C₅Me₅)₂Sm(C₅H₅)] with the concomitant formation of molecular hydrogen.¹⁴⁰ The authors emphasise the similarity to the reactivity of the alkaline metals. All three cyclopentadienyl ligands coordinate in a η⁵ mode. Interestingly, the addition of a further equivalent of solvent-free [(C₅Me₅)₂Sm] yields the mixed valent, dinuclear complex [{(C₅Me₅)₂Sm(III)}(μ-C₅H₅)Sm(II)(C₅Me₅)₂]. The addition of THF to a toluene solution of this dimer resulted in the reformation of [(C₅Me₅)₂Sm(C₅H₅)] and [(C₅Me₅)₂Sm(THF)₂].

Various substituted samarocenes and ansa-samarocenes have been used as catalysts for the polymerisation of ethylene and 1-olefines. In general, ansa-samarocenes showed higher activity and selectivity.³² A highly regio- and diastereoselective cross-coupling of allyl/propargyl ethers and δ-ketoesters to functionalised δ-lactones was observed by using various samarocenes as reagents. In dependence on the nature of the cyclopentadienyl ligand, high regio and diastereocontrol were reported in some cases. In contrast, SmI₂ gave unsatisfactory results in the transformation indicating the influence of the cyclopentadienyl ligand.¹⁴¹

Another parallel between the chemistry of samarocene and that of alkaline metals is the ability to transfer electrons to the aromatic system of polycyclic aromatic hydrocarbons (PAH).¹⁴² The intensely coloured reaction products of solvent-free [(C₅Me₅)₂Sm] and PAH such as anthracene, pyrene, acenaphtylene, 2,3-benzanthracene and 9-methylanthracene were investigated. No reaction occurred with naphthalene and coronene. In all other cases, 2 : 1 complexes of the general formula [{(C₅Me₅)₂Sm}₂(μ-PAH)] were isolated (Scheme 36). The solid structures of the anthracene, pyrene and 2,3-benzanthracene complexes were determined by single crystal X-ray crystallography. In general, samarium atoms coordinate to three PAH C-atoms resulting in a η³-coordination mode on opposing sides of the ligand, one above the ligand plane and one below. Slight differences are observed in terms of the positions of the (C₅Me₅)₂Sm fragments relative to each other. While for the anthracene ligand both metallocenes coordinate at the opposite middle carbons but at the opposite outer benzannulated rings, in the complex with the 2,3-benzanthracene ligand the samarium moieties are located directly opposite each other. In the case of pyrene, the metallocenes are located on the same end of the molecule. In contrast to PAH complexes of alkaline metals the PAH retains its planarity upon reduction in [(C₅Me₅)₂Sm–PAH] complexes. The Sm–C(PAH) bond lengths are in the range of 2.595(4) to 2.840(4) Å and are comparable to the allyl complexes described above. No single crystals of 6-methylanthracene and acenaphthylene for structural studies were obtained, however, the 2 : 1 stoichiometry was confirmed via elemental analyses. No reaction of samarocene with corocene was observed, which was explained by its low solubility. The interactions of [(C₅Me₅)₂Sm] and the PAH appears to be rather weak, as THF addition yields [(C₅Me₅)₂Sm(THF)₂] and the free hydrocarbon.

Scheme 36 Reaction products of samarocene with the polycyclic aromatic hydrocarbons anthracene, 2,3-benzanthracene and pyrene.¹⁴²

The reaction with cyclooctatetraene (COT) proceeded differently and yielded after a reaction time of 12 hours the Sm(III) complexes [(C₅Me₅)Sm(COT)] and [(C₅Me₅)₃Sm] via rearrangement and ligand exchange.^143,144 The corresponding diindenyl complex [(C₉H₇)₂Sm(THF)_x] shows a similar reactivity.¹⁸ With shorter reaction times of 15 min, [(C₅Me₅)₂Sm(THF)₂] and COT not only yielded [(C₅Me₅)Sm(COT)(THF)] and the oxidatively coupled dimer [(C₅Me₅)₂] with toluene as solvent, but also [(C₅Me₅)₂Sm{O(CH₂)₄(C₅Me₅)}(THF)] and [(C₅Me₅)₃Sm].¹⁴⁵ The complex [(C₅Me₅)₂Eu] shows no reactivity towards these substrates.

5.1.2. Reactions with carbon halides. [(C₅Me₅)₂Ln] or their solvates react with aryl and alkyl halides, though only [(C₅Me₅)₂Sm] and [(C₅Me₅)₂Yb] undergo oxidative addition reactions yielding Ln(III) compounds.

The C₅Me₅ ligands in decamethyleuropocene undergo coupling reactions with the organic moieties of the respective alkyl or aryl halides such as chlorides and bromides, yielding C₅Me₅R and EuX₂.¹⁴⁶

The reactions of the ytterbocene and samarocene with various organic chlorides, bromides and iodides were broadly investigated in terms of kinetics and the reaction mechanism. Organic substrates such as benzylchloride, –bromide and fluoride, butylchlorides, isopropylchloride, methyl iodide, phenyliodide and 1,2-diiodoethane were used.^{129,146–148} The corresponding diindenyl complex [(C₉H₇)₂Sm(THF)_x] shows a similar reactivity.¹⁸ The products observed in these reactions with [(C₅Me₅)₂Sm(solvent)_n] and [(C₅Me₅)₂Yb(solvent)_n] are [(C₅Me₅)₂LnX], [(C₅Me₅)₂LnR], [(C₅Me₅)LnX₂] together with radical coupling products of R (Scheme 37). The reactivity of RX follows the order I > Br > Cl ≫ F and benzyl ≈ tertiary > secondary > primary > phenyl corresponding to an atomic abstraction and concomitant generation of an organic radical. As a result of the reduction potential, samarocene reacts more readily than ytterbocene.

Scheme 37 General reactivity of the metallocenes towards halogenated hydrocarbons (Ln = Sm, Yb; X = F, Cl, Br, I). Not all products shown are formed in each case. Sometimes other byproducts were found.^{129,146–148}

Even unreactive carbon–fluorine bonds of certain perfluoro-olefins were reductively cleaved by the ethereal adducts of all three metallocenes.¹⁴⁹ In general, these reactions yielded [(C₅Me₅)₂LnF] with a terminal Ln–F bond as well as complexes with [(C₅Me₅)LnF₂] fragments. For example, [Yb₅(C₅Me₅)₆(μ₄-F)(μ₃-F)₂(μ-F)₆] was crystallised as a minor reaction product from the reaction of [(C₅Me₅)₂Yb(OEt₂)] and a perfluoro olefin. Structural data for diethylether and THF adducts of [(C₅Me₅)₂LnF] for Sm, Eu and Yb is available.

As expected, the reactivity of the metallocenes with perfluoro-olefins also follows the order of the reduction potential, i.e. [(C₅Me₅)₂Sm] > [(C₅Me₅)₂Yb] > [(C₅Me₅)₂Eu].¹⁴⁹ The reaction proceeds fast and rather clean if the substrate can form an allylic perfluoro-radical as an intermediate, which was shown by the reaction with perfluoro-2,4-dimethyl-3-ethylpent-2-ene and perfluoro-2,3-dimethylpent-2-ene. Up to four fluorine atoms were abstracted to ultimately yield perfluorotrienes. Perfluorocyclohexene reacts much slower by initial abstraction of an olefinic fluorine atom. In this case, hydrogen containing products were observed that most likely form via hydrogen abstraction from a C₅Me₅ ligand or a solvent. Successive defluoronization yields perfluorobenzene. Interestingly, visible light enhances the reaction rate. C₆F₁₀ does not react with [(C₅Me₅)₂Eu] in the dark, but when illuminated with a tungsten lamp, conversions have been observed in the NMR spectra. It should also be noted that reactions in toluene proceed faster than in diethylether, which is in accordance with the blocking of the reactive, inner sphere reaction site at the metal centre.

Finally, hexafluorobenzene was reported to react with [(C₅Me₅)₂Yb] in hexane to yield the mixed valent complexes [(C₅Me₅)₆Yb₄(μ-F)₄] and [(C₅Me₅)₄Yb₂(μ-F)] (Scheme 38).¹⁵⁰ The latter one can be considered as a Lewis acid–base complex between the Yb(III) complex [(C₅Me₅)₂YbF], in which the fluorine ligand is acting as the electron donor and the divalent [(C₅Me₅)₂Yb], which acts as Lewis-base. This is reflected by the unequal Yb-F bond lengths of 2.317(2) Å (Yb(II)–F) and 2.084(2) Å (Yb(III)–F) in its crystal structure. The organometallic [(C₅Me₅)₂Yb(C₆F₅)] compound is a by-product of the reaction. The mixed valent complex can also be obtained by reaction with various fluorinated olefins and aryls, namely CFHCH₂, CF₂CH₂, C₂F₄, PhF and PhCF₃. No reaction occurred with the fluorinated alkyls like C₂F₆ or 1,1,1-CF₃CH₃.¹⁴⁹ According to the authors, the strength of the carbon–fluorine bond is not the decisive factor for the reaction of ytterbocene with fluorinated hydrocarbons, since the C–F bond in C₆F₆ is stronger than in C₂F₆ (154 kJ mol⁻¹vs. 127 kJ mol⁻¹), but rather a polarisable group on the fluorine atom together with a free metal coordination site appear to be the necessary requirements for C–F activation.

Scheme 38 Reaction of [(C₅Me₅)₂Yb] with C₆F₆.¹⁵⁰

In contrast, to these methods, the organolanthanide fluoride [(η⁵-C₅H₄^tBu)₂Sm(μ-F)]₃ was prepared by oxidation of [(η⁵-C₅H₄^tBu)₂Sm(THF)₂] with Me₃SnF.¹⁵¹

5.1.3. Reactions with nitrogen-containing organic molecules. Especially the strongly reducing [(C₅Me₅)₂Sm] and its solvates react with nitrogen-containing functional groups of organic molecules. In the reaction of cyclohexyl isocyanide or tert-butyl isocyanide, the R–NC bond was cleaved to form the trimeric compounds of composition [(C₅Me₅)₂Sm(CNR)(μ-CN)]₃ (R: cyclohexyl, tert-butyl) with bridging cyanide ligands.¹⁵² An interacting isocyanide ligand coordinates with each of the samarium(III) centres. The cyclohexyl-containing trimer was structurally characterised by X-ray crystallography (Fig. 6). The three CN-bridged Sm atoms are almost planar, forming a triangle. Similar reactivity was observed in the reactions of samarocene with nitriles. Reaction in THF with variously substituted nitriles yielded the insoluble polymeric products [(C₅Me₅)₂Sm)(CN)]_n, which were also formed when the trimeric isocyanide compounds were desolvated. The trimeric tert-butyl nitrile adduct [(C₅Me₅)₂Sm(NC^tBu)(μ-CN)]₃ was isolated and structurally characterised and resembles the aforementioned isocyanide compounds (Fig. 6).⁸⁷ The reaction proceeded by initial coordination of nitrile ligands to the divalent precursor as proven by the isolation of tert-butyl nitrile adducts (see above, Fig. 3).

Fig. 6 Molecular structure of trimeric [(C₅Me₅)₂Sm(NC^tBu)(μ-CN)]₃ in the solid state (the rear C₅Me₅ and the hydrogen atoms are omitted for clarity) (Reproduced from the CIF file CCDC 638959).⁸⁷

The reactivity of the two cyano-group bearing molecules 7,7,8,8-tetracyanoquinodimethane (TCNQ) and 1,2,4,5-tetracyanobenzene (TCNB) towards [(C₅Me₅)₂Yb(THF)₂] was investigated by Trifonov et al.¹⁵³ The reaction with an equimolar amount of TCNQ in acetonitrile gave the dinuclear symmetric complex [(C₅Me₅)₂Yb(CH₃CN){(μ-CN)₂C(C₆H₄)C(CN)₂(C₅Me₅)}]₂ (Fig. 7). The dark red-brown complex was investigated by single crystal X-ray crystallography. During the reaction, a C₅Me₅-moiety of another [(C₅Me₅)₂Yb] molecule adds to one methylidene moiety, creating a monoanionic bridging ligand. The bond lengths observed in the solid structure point to a Yb(III) compound, as do magnetic measurements. When [(C₅Me₅)₂Yb(THF)₂] was reacted with TCNB in DMF, the green complex [(C₅Me₅)₂Yb{(μ-CN)₂(C₆H₄)(CN)₂}]₂ was formed. In this case, the TCNB moiety stays intact. Each of the dianionic ligands coordinates with both Yb-atoms with opposing cyano groups. The Yb–N bond lengths are in the range of 2.338(5) to 2.363(4) Å. The two aromatic rings of the bridging ligands show a parallel, almost eclipsed orientation indicating an interaction of their π-systems.

Fig. 7 Molecular structure of dinuclear [(C₅Me₅)₂Yb(CH₃CN){(μ-CN)₂C(C₆H₄)C(CN)₂(C₅Me₅)}]₂ in the solid state (Reproduced from the CIF file CCDC: 707131).¹⁵³

For a comparison with the activation of dinitrogen (see below, Scheme 72), the reactivity of samarocene towards azobenzenes and hydrazines was extensively studied. One molar equivalent of [(C₅Me₅)₂Sm(THF)₂] reacted with azobenzene in toluene to form the green complex [(C₅Me₅)₂Sm(N₂Ph₂)(THF)] (Scheme 39).^154,155 X-Ray crystallographic data shows the coordination or the ligand in a η²-fashion. The N–N distances of the crystallographically independent molecules in the asymmetric units are 2.390(10) and 2.450(10) Å, meaning significant lengthening upon reduction. The reverse addition of azobenzene to two equivalents samarocene in toluene yielded the blue complex [{(C₅Me₅)₂Sm}₂(N₂Ph₂)] including the azobenzene as a bridging ligand (Scheme 39). The solid state structure shows that each of the samarium atoms coordinates only one of the nitrogen atoms in contrast to the η²-mode in the 1 : 1 compound. The N–N bond however does not change significantly upon coordination. Instead, a lengthening of N–C bonds was observed. This allows the samarium metal to interact with one of the ortho hydrogen atoms of the phenyl ring. The azobenzene ligand in this molecule can be considered the PhNNPh²⁻-anion.

Scheme 39 Reactivity of [(C₅Me₅)₂Sm(THF)₂] towards different amounts of PhN=NPh.^154,155

The reaction of [(C₅Me₅)₂Sm(THF)₂] with the analogous hydrazine compound PhHNNHPh in hexane leads to the formation of yellow [(C₅Me₅)₂Sm(NHPh)]_xvia N–N bond cleavage.³⁵ The precipitate was recrystallised from THF to give [(C₅Me₅)₂Sm(NHPh)(THF)]. The same product was also observed when [(C₅Me₅)₂Sm(THF)₂] was reacted with aniline or phenylhydrazine (Scheme 40). However, unidentified by-products occur and yields are low. The solid state structure of [(C₅Me₅)₂Sm(NHPh)(THF)] reveals a terminal coordinated NHPh-amide. A different reaction pathway was observed when desolvated [(C₅Me₅)₂Sm] was reacted with diphenylhydrazine in toluene. As product [(C₅Me₅)₂Sm(PhNHNPh)] was obtained, which forms the solvate [(C₅Me₅)₂Sm(PhNHNPh)(THF)] after crystallization from THF. In this compound single deprotonated hydrazine acts as a ligand, which coordinates in a η²-fashion to the samarium atom. Further treatment of [(C₅Me₅)₂Sm(PhNHNPh)] with [(C₅Me₅)₂Sm(THF)₂] resulted again in [(C₅Me₅)₂Sm(NHPh)(THF)] (Scheme 40).

Scheme 40 Reactivity of [(C₅Me₅)₂Sm(THF)₂] towards aniline, diphenyl hydrazine and phenyl hydrazine and of [(C₅Me₅)₂Sm] towards phenylhydrazine.³⁵

The parent molecule hydrazine (H₂NNH₂) reacted with solvent-free samarocene to form the doubly deprotonated dinuclear product [(C₅Me₅)₂Sm(μ-η²:η²-HNNH)Sm(C₅Me₅)₂] as red crystals (Scheme 41). The solid state structure revealed an unsymmetrically bridging mode of the ligand. The hydrazido anion in [(C₅Me₅)₂Sm(μ-η²:η²-HNNH)Sm(C₅Me₅)₂] shows a lengthening in the nitrogen–nitrogen bond compared to hydrazine.

Scheme 41 Reaction of [(C₅Me₅)₂Sm] and hydrazine.³⁵

When desolvated samarocene was reacted with excess hydrazine in benzene, the yellow tetranuclear compound [(C₅Me₅)₄Sm₄(N₂H₂)₂(N₂H₃)₄(NH₃)₂] was isolated.¹⁵⁶ The Sm atoms are arranged in a tetrahedral fashion and are bridged by deprotonated hydrazine molecules. During the reaction, one pentamethylcyclopentadienyl ligand of each samarocene unit was protonated. Both HNNH and HNNH₂ are present in the molecule as ligands.

A deprotonation was also observed by reacting the neutral formazan ligand L²H (L² = {PhNNC(4-^tBuPh)NNPh} with [(C₅Me₅)₂Sm(THF)₂] (Scheme 42). As shown by single crystal X-ray diffraction, the bonding in the six-membered core is delocalised.¹⁵⁷

Scheme 42 Reaction of [(C₅Me₅)₂Sm(THF)₂] with a formazan.¹⁵⁷

The reaction of [(C₅Me₅)₂Sm(THF)₂] with benzaldehyde azine (PhHC=NN=CHPh) as a further compound with a nitrogen-nitrogen bond resulted in the dinuclear coupling product [{(C₅Me₅)₂Sm}₂{μ,η⁴-(PhCH=NNCHPh)₂}] (Scheme 43).⁹¹ The nitrogen moieties bind in a η²-fashion to the {(C₅Me₅)₂Sm} units. In contrast to the previously described reactions containing C–N bonds, no reduction of the multiple bonds occurs but instead a reductive coupling of the substrate to the bridging ligand. Most likely, the coupling involves some radical intermediates.

Scheme 43 Reaction of [(C₅Me₅)₂Sm(THF)₂] with PhCNNCPh.⁹¹

The reaction of either one or two equivalents of benzophenone imine with [(C₅Me₅)₂Sm(THF)₂] gave orange red-crystals of [(C₅Me₅)₂Sm(N=CPh₂)(NH₂CHPh₂)] (Scheme 44).¹⁵⁸ The mechanism was investigated by deuteration experiments. Most likely the reaction proceeds via an electron transfer to an imine, which deprotonates a second molecule present in the reaction mixture.

Scheme 44 Reactions of [(C₅Me₅)₂Sm(THF)₂] with different amounts of benzophenone imine.¹⁵⁸

The reduction of [(1,2,4-C₅H₂^tBu₃)₂Tm] with pyridine resulted in a coupling of the pyridine in para-position to give [{(1,2,4-C₅H₂^tBu₃)₂Tm}₂{μ-(NC₅H₅-C₅H₅N)}], in which the metal atoms are bridged by a 1,1′-bis(1,4-dihydropyridylamide) ligand (Scheme 45).³⁹

Scheme 45 Reactions of [(1,2,4-C₅H₂^tBu₃)₂Tm] with pyridine.³⁹

A similar reaction of [(C₅Me₅)₂Sm(THF)₂] with pyridazine in toluene resulted in the transfer of one electron and coupling of two complexes to give yellow-orange [{(C₅Me₅)₂Sm(THF)}₂{μ-(N₂C₄H₄)₂}] (Scheme 46).⁹¹ The product was studied by single crystal X-ray crystallography. The neighbouring N atoms are coordinating the samarium atoms in a η² coordination mode. Coupling occurs via one of the carbon atoms not directly adjacent to the nitrogen atoms.

Scheme 46 Reaction of [(C₅Me₅)₂Sm(THF)₂] and pyridazine.⁹¹

Typically, [(C₅Me₅)₂Sm(THF)₂] reacts with multiple bonds to give either [{(C₅Me₅)₂Sm}₂(substrate)] with a doubly-reduced substrate, or reductive coupling occurs to form [{(C₅Me₅)₂Sm}₂(substrate–substrate)] species with a dianionic, coupled-substrate moiety.¹⁵⁹ When [(C₅Me₅)₂Sm(THF)₂] reacts with 1,2-bis(2-pyridyl)ethene (py), both types of trivalent Sm-complexes are formed at the same time to yield [{(C₅Me₅)₂Sm}₂(μ-η²:η²-pyCHCHpy)] and [{(C₅Me₅)₂Sm}₂(μ-η³:η³-1,2,3,4-(py)₄C₄H₄)]. Spectroscopic data was consistent with a trivalent samarium complex. However, the crystallographic data was not sufficient for a detailed bond analysis.

As mentioned above, the reaction of [(C₅Me₅)₂Yb(OEt₂)] with 1,10-phenanthroline leads to oxidation of the metal and a reduced radical ligand,⁹³ which is in a monomer/dimer equilibrium. A similar coupling of 1,10-phenanthroline was observed by using the stronger reducing agents [(C₅Me₅)₂Sm(OEt₂)] or the tert-butyl substituted metallocenes [(1,3-C₅H₃^tBu₂)₂Sm], [(1,2,4-C₅H₂^tBu₃)₂Sm], and [(1,2,4-C₅H₂^tBu₃)₂Tm].⁴⁶ In all cases a C–C bond formation on the 4-position of the phenanthroline ring was observed in the solid state. Upon analysis of the solution structure, the authors suggest a thermally reversible C–C coupling in all cases.

An interesting behaviour is found in the reaction of [(C₅Me₅)₂Yb(OEt₂)] with 4,5-diazafluorene (Scheme 47).¹⁶⁰ The reaction in toluene gave the green intermediate-valent species [(C₅Me₅)₂Yb(N₂C₁₁H₈)], which could not be structurally characterised. The green compound transformed into the red trivalent diazafluorenyl compound [(C₅Me₅)₂Yb(N₂C₁₁H₇)] via hydrogen abstraction. This product was structurally characterised with X-ray crystallography. The Yb–N distances are 2.372 ± 0.001 Å, which are in the range of ionic, trivalent [(C₅Me₅)₂Yb(bipy)] complex. Extensive mechanical investigations on the formation of the complex as well as kinetic measurements were carried out. It was shown that [(C₅Me₅)₂Yb(N₂C₁₁H₇)] is formed via an isolable dimeric complex, formed by the coupling of two diazafluorene ligands and concomitant H₂ abstraction.

Scheme 47 Reaction of [(C₅Me₅)₂Yb(OEt₂)] with diazafluorene.¹⁶⁰

No C–C coupling occurs by using redox-active ligands such as diazabutadienes. Decamethylytterbocene reacted with N,N-di-tert-butyl-1,4-diazabutadiene to give the oxidised complex [(C₅Me₅)₂Yb(N₂(^tBu)₂C₂H₂)] via one electron reduction of the diene (Scheme 48).^161,162 Similarly, [(C₅Me₅)₂Yb(THF)₂] reacts with the diazabutadiene 2-MeC₆H₄N=C(Me)C(Me)=NC₆H₄Me-2 to the corresponding complex [(C₅Me₅)₂Yb(2-MeC₆H₄N=C(Me)C(Me)=NC₆H₄Me-2)] ligated by a diazabutadiene anion.¹⁶³ Similar studies were performed with other ytterbocenes or other diazabutadienes.^162–164 Reaction with samarocene gave the analogous Sm(III) complex.¹⁴³ The redox potential of [(C₅Me₅)₂Eu(Et₂O)] is too low to reduce the N,N-di-tert-butyl-1,4-diazabutadiene and the respective Lewis-acid base complex was formed.¹⁶⁵ In contrast, when N,N-di-pentafluorophenyl-1,4-diazabutadiene derivative was reacted with a [(C₅Me₅)Eu(Et₂O)], the corresponding Eu(III) complex [(C₅Me₅)₂Eu(N₂(C₆F₅)₂(C(Me))₂)] with the reduced ligand as a radical anion was obtained (Scheme 48). The pentafluorophenyl groups induce a higher electron affinity of this ligand.

Scheme 48 Reactions of ytterbocene and europocene with differently substituted diazabutadienes.^161,165

Another redox-active ligand is 1,2-bis(imino)acenaphthene (BIAN), which consists of both a naphthalene ring and a 1,4-diaza-1,3-butadiene moiety. The reaction of [(C₅Me₅)₂Ln(OEt₂)] (Ln = Sm, Eu) with an equimolar quantity of the corresponding R-BIAN ligand (R = mesityl, tert-butyl, p-methoxyphenyl) in toluene solution at ambient temperature resulted in [(C₅Me₅)₂Sm(mes-BIAN)], [(C₅Me₅)₂Eu(^tBu-BIAN)] and [(C₅Me₅)₂Eu(p-MeO-BIAN)]. Magnetic measurements indicate a trivalent oxidation state of the lanthanide metal (Scheme 49).¹⁶⁶ Thus, the BIAN ligand is reduced by one electron. In contrast, the reaction of [(C₅Me₅)₂Sm(OEt₂)] with one equivalent of the sterically encumbered ligand, dpp-BIAN (dpp = 2,6-diisopropylphenyl) in THF solution resulted in the loss of a C₅Me₅ group and formation of [(C₅Me₅)Sm(dpp-BIAN)(THF)].¹⁶⁶ Here a two-electron reduction of the dpp-BIAN ligand has taken place, which is also supported by the corresponding bonding parameters.

Scheme 49 Reactions of samarocene and europocene with differently substituted 1,2-bis(imino)acenaphthenes (BIAN).¹⁶⁶

A metal-induced oxidation of the metallocenes was observed by reacting [(C₅Me₅)₂Yb] with thallium compounds. In these cases, the thallium atom is reduced upon reaction and the corresponding anion is transferred to the lanthanide metal. Following this strategy, [(C₅Me₅)₂Yb] reacts with [Tl(Ph₂pz)] and [Tl(azin)] (Ph₂pz = 3,5-diphenylpyrazolate, azin = 7-azaindolate) to give [(C₅Me₅)₂Yb(Ph₂pz)] and [(C₅Me₅)₂Yb(azin)]. In both cases, the N-donor ligands coordinate with both nitrogen atoms to the Yb-atom.¹⁶⁷

5.1.4. Reactions with organic substrates containing carbon oxygen bonds. It was shown that epoxides react with decamethylsamarocene to yield the oxo-bridged dimer [{(C₅Me₅)₂Sm}₂(μ-O)],⁸¹ which is often obtained as an undesired side product with oxygen, moisture or oxygen-containing molecules that can transfer oxygen atoms (Scheme 50).^81,168 The crystal structure reveals a centrosymmetric molecule with the centre of symmetry residing at the oxygen atom. The Sm–O–Sm atoms are arranged linearly, with the centroid–Sm–centroid axes of the two (C₅Me₅)₂Sm units standing perpendicular to each other. The Sm–O distance is 2.094 Å. An alternative rational access to [{(C₅Me₅)₂Sm}₂(μ-O)] is the oxidation of [(C₅Me₅)₂Sm(THF)₂] with pyridine N-oxide (Scheme 50).

Scheme 50 Oxidation of [(C₅Me₅)₂Sm(THF)₂] with epoxides and pyridine N-oxide.⁸¹

The cleavage of allylic-, propargylic and vinylic ethers by [(C₅Me₅)₂Sm(THF)_n] was investigated via NMR.^169–171 Allylic ethers are cleaved by [(C₅Me₅)₂Sm(THF)₂], as evidenced by various examples, e.g. the reaction of desolvated [(C₅Me₅)₂Sm] and allyl benzyl ether gave the allyl compound [(C₅Me₅)₂Sm(C₃H₅)] and the alkoxide compound [(C₅Me₅)₂Sm(OBn)] (Scheme 51).

Scheme 51 Reactivity of [(C₅Me₅)₂Sm(THF)₂] towards allylic ethers to yield allyl complexes.¹⁶⁹

Alkoxide formation was seen upon the reaction of [(C₅Me₅)₂Sm(THF)₂] with 2,3,5,6-tetramethylphenol in toluene (Scheme 52).¹⁷² Concomitant with the release of hydrogen, an orange crystalline solid was isolated. It was identified as the phenolate complex [(C₅Me₅)₂Sm(OC₆HMe₄)] by X-ray crystallography. As expected, the phenolate ligand binds via the oxygen atom to the Sm centre with a Sm–O bond of 2.13(1) Å. The observed results are in contrast to the reaction of [(C₅Me₅)₂Sm(THF)₂] with very bulky phenols (see above, Scheme 11).⁸⁴ In these reactions the samarium atom is not oxidised but one of the C₅Me₅ ligands is protonated.

Scheme 52 Reaction of [(C₅Me₅)₂Sm(THF)₂] with 2,3,5,6-tetramethylphenol.¹⁷²

Trivalent alkoxide complexes were obtained by the reaction with diorganoperoxides of the type R-OO-Ar.¹⁷³ The reaction of [(C₅Me₅)₂Yb(NH₃)₂] and R-OO-Ar in toluene (R = CMe₃, SiMe₃) yields products of the composition [(C₅Me₅)₂Yb(OR)(NH₃)] as orange solids (Scheme 53). No structural data exist; however, the products were characterised by NMR spectroscopy, elemental analysis, IR spectroscopy and mass spectrometry. The corresponding chalcogenide compounds of S, Se and Te are similarly accessible (see below, Scheme 59).

Scheme 53 Reactivity of decamethylytterbocenes towards aryl substituted organic peroxides.¹⁷³

Note, the reaction of [(C₅Me₅)₂Sm(THF)₂] with bulky phenols does not give Sm(III) compounds, but dimeric, divalent complexes [(C₅Me₅)Sm(μ-OAr)]₂ (Ar = C₆H₃^tBu₂-2,6, C₆H₂^tBu₂-2,6-Me-4 and C₆H₂^tBu₃-2,4,6)) (see above, Scheme 11).⁸⁴

Benzoate complexes are accessible by the oxidation of divalent Yb with thallium reagents.^6,174 Thus, [(C₅H₄Me)₂Yb(O₂CPh)]₂ was obtained by reaction of [(C₅H₄Me)₂Yb] with Tl(O₂CPh) in THF.¹⁷⁴

Multinuclear complexes are observed if the divalent metallocenes of the lanthanides were reacted with 3,6-di-tert-butyl-o-benzoquinone (3,6-dbbq).¹⁷⁵ For [(C₅Me₅)₂Sm(THF)₂] and [(C₅Me₅)₂Yb(THF)] the reaction with 3,6-dbbq in hexane resulted in the dinuclear compounds [{(C₅Me₅)Ln}₂(dbcat)₂] (Fig. 8), in which the former quinone ligand has been reduced to its catecholate (dbcat) form. During the reaction, each lanthanide ion loses one pentamethylcyclopentadienyl ligand. The crystal structure was determined for both compounds. The structure is centrosymmetric, with one of the oxygen atoms of each catecholate ligand bridging the two lanthanide ions. The bond length for the non-bridging oxygen atoms are 2.3127(18) and 2.2269(17) Å (Sm) and 2.2264(19) and 2.1285(18) Å for the Yb complex. The Ln–O bond to the neighbouring Ln ion is slightly longer (0.09 Å for Sm, 0.04 Å for Yb). The lesser reducing europocene yields a different product. The trinuclear mixed-valent compound [(C₅Me₅Eu)(Eu(THF))₂(dbcat)₃] is obtained (Fig. 8), in which only one Eu ion is in +III oxidation state. One europium ion has retained one cyclopentadienyl ligand, while the two others have lost both cyclopentadienyl ligands. The structure is rather complex, and the Eu ions have different ligand environments. Every oxygen-catechol atom has a bridging function. The different catecholate bonds are in the wide range from 2.280(2) to 2.504(2) Å, due to the different oxidation states of the three Eu ions.

Fig. 8 Molecular structures of [{(C₅Me₅)Sm}₂(dbcat)₂] (top) and [(C₅Me₅Eu)(Eu(THF))₂(dbcat)₃] (bottom) in the solid state (Reproduced from the CIF file CCDC: 1409389, 1409391).¹⁷⁵

As an example for a ketone, the reactivity of fluorenone was investigated in the reaction with ytterbocene and samarocene.¹⁷⁶ In both cases, the THF solvates reacted in THF with one equivalent of fluorenone to yield the corresponding radical anion ketyl complexes, [(C₅Me₅)₂Ln(III)(˙OC₁₃H₈)(THF)] (Scheme 54). Further reaction with hexamethylphosphoramide (HMPA) led to the replacement of the THF molecule, giving rise to [(C₅Me₅)₂Ln(III)(˙OC₁₃H₈)(HMPA)]. The structures of [(C₅Me₅)₂Sm(˙OC₁₃H₈)(THF)] and [(C₅Me₅)₂Yb(˙OC₁₃H₈)(HMPA)] were determined. Ln–O distances of 2.234(7) Å (Sm) and 2.108(7) Å (Yb) were measured.

Scheme 54 Reaction of decamethylsamarocene and decamethylytterbocene with fluorenone.¹⁷⁶

A dihydroindenoindene diolate complex was obtained by reacting [(C₅Me₅)₂Sm(THF)₂] successively with C₆H₅C≡CC₆H₅ and CO to yield the trivalent seven-coordinate samarium complex [{(C₅Me₅)₂Sm}₂(OC₁₆H₁₀)] (Scheme 55).¹⁷⁷ [{(C₅Me₅)₂Sm}₂(OC₁₆H₁₀)] can be recrystallised from THF which results in the eight coordinated THF-solvate [{(C₅Me₅)₂(THF)Sm}₂(OC₁₆H₁₀)]. The average Sm–C(ring) length is in the seven-coordinate complex 2.70(3) Å and thus shorter than those of the eight-coordinate complex with a bond length of 2.75(2) Å. The Sm–O(O₂C₁₆H₁₀) bond lengths are 2.08(2) and 2.099(9) Å, respectively. As expected, the bond lengths of the seven-coordinate complex are shorter than of the eight-coordinate complex.

Scheme 55 Reaction of [(C₅Me₅)₂Sm(THF)] with diphenylacetylene and CO.¹⁷⁷

5.1.5. Organo-pnictogen compound. The redox reactivity of [(C₅Me₅)₂Sm] and its solvates towards organo-pnictogen compounds of the type ER₃ and R₄E₂ was the focus of several studies. When PPh₃ was reacted with solvent-free samarocene, coordination occurs in the solution while no reaction was observed with the corresponding arsenic compound AsPh₃.¹⁷⁸ The samarocene THF adduct is less reactive towards phosphines as they cannot substitute the strongly coordinating oxygen donor. However, solvent-free samarocene reacted with SbPh₃via a reductive Sb–C bond cleavage to form [(C₅Me₅)₂SmPh] and a product mixture as monitored by NMR spectroscopy. This reaction does not occur with the THF solvate. By treating the more reactive BiPh₃ with either [(C₅Me₅)₂Sm] or [(C₅Me₅)₂Sm(THF)₂] in toluene or cyclohexane the remarkable binuclear compound [(C₅Me₅)₂Sm(μ-η²:η²-Bi₂)Sm(C₅Me₅)₂] and a complicated mixture of side products were obtained. These can be avoided by using a 4 : 1 stoichiometry of Sm to Bi (Scheme 56).¹⁷⁹ The crystal structure of dark red [(C₅Me₅)₂Sm(μ-η²:η²-Bi₂)Sm(C₅Me₅)₂] was determined revealing a bridging Bi₂²⁻ moiety side-on coordinated with the (C₅Me₅)₂Sm units. The Bi-Bi distance is 2.851(1) Å and thus shorter than typical Bi-Bi single bond distances, which range from 2.990(2) to 3.092(2) Å.^180–184 For the Sm-Bi distances values ranging from 3.265(1) Å to 3.311(1) Å were determined. The C₅Me₅–Sm–C₅Me₅ axes in [(C₅Me₅)₂Sm(μ-η²:η²-Bi₂)Sm(C₅Me₅)₂] are reminiscent of the prominent N₂ complexes, but are arranged coplanar and not perpendicular as in the corresponding N₂ complex (Scheme 72).²⁰²

Scheme 56 Reaction of [(C₅Me₅)₂Sm(THF)₂] with Ph₃Bi.¹⁷⁹

Though no isolable products were obtained by treatment of [(C₅Me₅)₂Sm] with SbPh₃, the reaction with the aliphatic stibine SbⁿBu₃ resulted in a complex mixture, containing the isolable product [{(C₅Me₅)₂Sm}₃(μ-η²:η²:η¹-Sb₃)].¹⁸⁵ In contrast to the dinuclear pnictogenic complexes of N and Bi, a trinuclear Zintl type anion Sb₃³⁻ was formed. X-Ray crystallographic studies showed Sb–Sb distances of 2.689(1) and 2.686(1) Å. Two of the (C₅Me₅)₂Sm moieties are bound to two Sb anions of the triangular anion, while one moiety binds only to one and has an additional THF molecule bound to satisfy the coordination sphere.

When [(C₅Me₅)₂Sm] was reacted with Ph₂PPPh₂ or Ph₂AsAsPh₂, reductive cleavage of the E–E bond occurred and the trivalent pnictogenide compounds [(C₅Me₅)₂Sm(EPh₂)] were formed (Scheme 57).^178,186 When THF was present either as a solvent or in the reactant, the corresponding THF-complexes [(C₅Me₅)₂Sm(EPh₂)(THF)] were isolated. However, these compounds were unstable and ether cleavage of the coordinated THF took place, yielding a diphenylpnicgtogenyl-functionalised butoxide ligand. However, the phosphide [(1,3-C₅H₃^tBu₂)₂Sm(PPh₂)], which was prepared from [(1,3-C₅H₃^tBu₂)₂Sm] and Ph₂PPPh₂, is not prone to ring-opening owing to insufficient space in the Sm coordination sphere for a THF ligand.¹⁸⁶

Scheme 57 Reactivity of decamethylsamarocene towards diphosphides and diarsenides.^180,188

The crystal structure of [(C₅Me₅)₂Sm(AsPh₂)] shows an unsymmetrically coordinated diphenylarsenide ligand that is tilted in a way to enable an η²-interaction of a phenyl ring with the samarium atom, emphasizing the high Lewis acidity of the large Sm(III) ion. The geometry around the arsenic atom strongly deviates from pyramidal or planar geometry. The Sm–As bond lengths of the two crystallographically independent molecules are 2.973(3) and 2.966(3) Å. In contrast, the AsPh₂ ligand in [(C₅Me₅)₂Sm(AsPh₂)(THF)] is considerably less distorted and features an Sm–As longer bond (3.049(3) Å).

For the corresponding phosphide, only the ring-opened THF product was structurally characterised. In contrast to the arsenic compound, the complex crystallises either as a dimer [{(C₅Me₅)₂Sm}(μ-OC₄H₈PPh₂)]₂ or as a polymer [{(C₅Me₅)₂Sm}(μ-OC₄H₈PPh₂)]_∞. The Sm atoms are linked by the butoxide ligands via oxo functions and neutral phosphorus donors.

Using a 4 : 1 stoichiometry of desolvated [(C₅Me₅)₂Sm] to Ph₂EEPh₂ in non-polar solvents the reaction resulted in mixed valent complexes of the type [(C₅Me₅)₂Sm(μ-EPh₂)Sm(C₅Me₅)₂]. The arsenic compound [(C₅Me₅)₂Sm(μ-AsPh₂)Sm(C₅Me₅)₂] was only detected by NMR spectroscopy as it quickly decomposes. The phosphorus compound [(C₅Me₅)₂Sm(μ-PPh₂)Sm(C₅Me₅)₂] was structurally characterised; however, X-ray data was not of sufficient quality to determine bond distances. The connectivity of all atoms was confirmed, showing a bridging phosphide between the samarium atoms. The reactions of [(C₅Me₅)₂Sm] or [(C₅Me₅)₂Sm(THF)₂] with either Ph₂SbSbPh₂ or Ph₂BiBiPh₂ did not yield any products with Sm–E bonds. Instead [(C₅Me₅)₂Sm(Ph)] was isolated.

In a different series, phospholyl and arsolyl ligands, such as [(C₄Me₄P)₂], [(C₄H₂Me₂P)₂], [(C₄-^tBu₂H₂P)₂], and [(C₄H₂Me₂As)₂] were reacted with [(C₅Me₅)₂Sm(Et₂O)] to obtain the samarium(III) complexes [{(C₅Me₅)₂Sm}(C₄Me₄P)], [{(C₅Me₅)₂Sm}(C₄H₂Me₂P)], [{(C₅Me₅)₂Sm}(C₄-^tBu₂H₂P)] and [{(C₅Me₅)₂Sm}(C₄H₂Me₂As)], respectively.¹⁸⁷ The reaction of [(C₅Me₅)₂Sm(Et₂O)] with [Tl(C₄H₄P)] yielded [{(C₅Me₅)₂Sm}(C₄H₄P)].

The reactivity of 2-tert-butyl-1-phosphaethyne ^tBuC≡P towards [(C₅Me₅)₂Sm(THF)₂] in toluene was investigated (Scheme 58).¹⁸⁸ Bright red crystals of [{(C₅Me₅)₂Sm}₂(μ-^tBuC=P–P=C-^tBu)] were isolated. The reductive coupling of two molecules phospha-alkyne took place to form a P–P bond concomitant with the formation of double bonds between the phosphorus and the carbon atoms. The newly formed ligand bridges two (C₅Me₅)₂Sm(III) moieties on opposing sides of the ligand via one P and one C atom. The Sm–P distances are 2.952(2) and 2.945(2) Å and the Sm–C bond lengths are almost equidistant with 2.557(6) and 2.556(6) Å.

Scheme 58 Reaction of [(C₅Me₅)₂Sm(THF)₂] with ^tBuC≡P.¹⁸⁸

5.1.6. Chalcogen containing organics. Similar to the above-mentioned diorganoperoxides (Scheme 53), the corresponding organic disulfides, diselenides and ditellurides (ArEEAr; E = S, Se, Te) are reported to react with either [(C₅Me₅)₂Sm(solvate)_n] or [(C₅Me₅)₂Yb(solvate)_n] to yield the respective chalcogenide compounds of the type [(C₅Me₅)₂Ln(III)(EAr)(L)] (Scheme 59).^173,189 For example, the reaction of [(C₅Me₅)₂Sm(THF)₂] with Mes-E-E-Mes (Mes = Mesityl; E = S, Se, Te) in toluene gave compounds of the composition [(C₅Me₅)₂Sm(III)(E-Mes)(THF)] in every case. The complexes were isolated as orange crystals. The crystal structures of the selenium and tellurium compounds reveal Sm–E distances of 3.088(2) Å for Te and 2.919(1) Å for Se following decreasing atomic radii.

Scheme 59 Reactivity of decamethylytterbocene towards organic disulfides, diselenides and ditellurides.^173,189

Analogous reactions occurred when either [(C₅Me₅)₂Yb(NH₃)₂] or [(C₅Me₅)₂Yb(Et₂O)] was reacted with organodichalcogenides with different organic moieties such as phenyl or mesityl. Red compounds of the type [(C₅Me₅)₂Yb(ER)(L)] were isolated. The crystal structures of [(C₅Me₅)₂Yb(TePh)(NH₃)] and [(C₅Me₅)₂Yb(SPh)(NH₃)] (R = phenyl or substituted phenyl group) were determined and revealed Yb–Te bonding distance of 3.039(1) Å and the Yb–S bond length of 2.670(3) Å.

A similar reactivity was observed by using metallocenes with bulkier substituents on the five-membered ring. However, no solvent is coordinated with the metal atoms of these products. Thus, the reaction of [(Cp^Bz5)₂Sm] with Ph₂E₂ (E = Se, Te) at room temperature in THF gave the monometallic compounds [(Cp^Bz5)₂Sm(SePh)] and [(Cp^Bz5)₂Sm(TePh)].⁴⁸

Pentavalent phosphine chalcogens readily oxidise samarocene to yield chalcogenide-bridged dinuclear compounds.¹⁹⁰ [(C₅Me₅)₂Sm(THF)₂] was reacted with SPPh₃ or SePPh₃ in THF and [{(C₅Me₅)₂Sm(THF)}₂(μ-S)] and [{(C₅Me₅)₂Sm(THF)}₂(μ-Se)] were obtained as yellow (S) or orange (Se) crystals (Scheme 60). The structures are comparable to the oxygen-bridged product described above. Hence, two samarium atoms are bridged by one E²⁻ ligand. One additional THF molecule is bound to each metal centre to saturate the coordination sphere. The two Sm–E distances are of almost equal length with values of 2.783(1) and 2.779(1) Å for the selenium complex and 2.663(1) and 2.665(1) Å for the sulphur analogue in accordance with the smaller atomic radius of S.

Scheme 60 Reaction of [(C₅Me₅)₂Sm(THF)₂] with phosphine selenides and tellurides.¹⁹⁰

Ytterbocene was also oxidised by SPPh₃ or SePPh₃ to give the corresponding chalcogenide complexes, in which the E²⁻ bridges the two metal centres (Scheme 61).¹⁹¹ However, due to the smaller ionic radius of Yb³⁺, no additional THF ligands are bound to the metal. The crystal structure of purple [{(C₅Me₅)₂Yb}₂(μ-Se)] reveals a centrosymmetric molecule with equal Yb–Se distances of 2.621(1) Å. The tellurium analogue was prepared from [(C₅Me₅)₂Yb(OEt₂)] and tri-n-butylphosphine telluride in hexane (Scheme 61).

Scheme 61 Reactivity of [(C₅Me₅)₂Yb(OEt₂)] towards phopshine chalcogenides.¹⁹¹

The reactivity of tetramethylthiuram disulfide (S₂NCMe₂)₂ towards [(C₅Me₅)₂Sm(THF)₂] was investigated by Edelmann and co-workers.¹⁸⁹ The sulphur–sulphur bond was reductively cleaved resulting in a chelating dithiocarbamate ligand, which was found in the orange product [(C₅Me₅)₂Sm(S₂NCMe₂)].

The related redox active sulphurdiimine [(Me₃SiN=)₂S] was investigated as a ligand for [(C₅Me₅)₂Ln(THF)₂] (Ln = Sm, Eu, Yb).¹⁹² The reaction of [(C₅Me₅)₂Ln(THF)₂] with [(Me₃SiN=)₂S] in Et₂O yielded in all three cases [(C₅Me₅)₂Ln(Me₃SiN=)₂S] (Scheme 62). The compounds were investigated with single-crystal X-ray diffraction and revealed symmetrical coordination of the diimide ligand. The N–Ln bond distances decrease from Sm to Yb in accordance with the decrease of the ionic radii (Sm–N: 2.458(2)/2.456(2) Å; Eu–N 2.449(2)/2.437(2) Å; Yb–N 2.352(8)/2.361(8) Å). The S–N distances are between those for single and double S–N bonds and are best described as a radical anion. The products were extensively characterised and investigated with DFT calculations with regard to their electronic nature.

Scheme 62 Synthesis of [(C₅Me₅)₂Ln(Me₃SiN=)₂S].¹⁹²

5.2. Reactivity towards inorganic molecules

The reactions of [(C₅Me₅)₂Sm(THF)₂] with strongly oxidizing nitric oxide NO or nitrous oxide N₂O resulted in a mixture of oxidised products, accompanied by a colour change from purple to orange.⁸¹ Identification of all the products was not possible. One of the products was samarocene oxide [{(C₅Me₅)₂Sm(THF)₂}₂(μ-O)]. The corresponding diindenyl complex [(C₉H₇)₂Sm(THF)_x] shows a similar reactivity.¹⁸

Evans investigated the reactivity of [(C₅Me₅)₂Sm(THF)₂] towards H₂O under controlled conditions, a generally undesired reaction in organosamarium chemistry.¹⁹³ By exposing a samarocene THF solution to a nitrogen atmosphere containing water vapour, brown crystals of the hexanuclear samarium cluster [{(C₅Me₅)Sm}₆(O₉H₆)] were formed over a longer period. Two types of reaction take place. First, hydrolysis of a cyclopentadienyl ligand and second, water reduction by Sm(II). The six samarium atoms build an axially elongated octahedron having an oxygen atom in the centre. The remaining oxygen atoms bridge three Sm atoms of each plane of the octahedral Sm₆ core. Due to the distorted geometry of the octahedron, the Sm–O distances vary between 2.499(4) and 2.504(4) Å, depending on the position of the Sm atom in the octahedron.

In contrast, no clusters but the organolanthanide(III) hydroxide complexes [{1,3-C₅H₃(SiMe₃)₂}₂Sm(μ-OH)]₂ and [{C₅H₄(SiMe₃)}₂Yb(μ-OH)]₂ were obtained, when the corresponding metallocene precursors [{1,3-C₅H₃(SiMe₃)₂}₂Sm(THF)] and [{C₅H₄(SiMe₃)}₂Yb(OEt₂)] were reacted with water in an ethereal solution. Both complexes are dimeric in the solid state with bridging hydroxide groups, while NMR data indicate that these structures persist in aprotic media.¹⁹⁴

Evans et al. reported that [(C₅Me₅)₂Sm(THF)₂] is able to reduce CO in THF under a pressure of 6 bar to form the dark orange, tetranuclear complex [({C₅Me₅)₂Sm}₂(μ-O₂CCCO)(THF)]₂ in low yield.¹⁹⁵ The bridging ligands can be regarded as ketenecarboxylates and are formally constituted of three carbon monoxide molecules. A schematic drawing of the structure can be found in Scheme 63. Each of the ligands binds one Sm atom at each end of the ligand. Additionally, a kind of dimeric structure is formed, in which the oxygen atoms of the carboxylic acid units are bridging both subunits. However, studies by Andersen et al. on carbon monoxide complexes did not show any redox reactivity of desolvated decamethyl samarocene under a CO atmosphere (see above).¹⁰⁸

Scheme 63 Schematic drawing of [{(C₅Me₅)₂Sm}₂(μ-O₂CCCO)(THF)]₂.¹⁹⁵

The divalent thulium complex [(1,2,4-C₅H₂^tBu₃)₂Tm] reacted with CO to give selective CO reductive dimerization and trimerization in dependence of the stoichiometric ratio (Scheme 64).⁴¹ By using an equimolar ratio, the ethynediolate complex [{(1,2,4-C₅H₂^tBu₃)₂Tm}₂(μ-κ(O):κ(O′)-C₂O₂)] was formed. During the reaction, a dimerization of CO to a ethynediolate (C₂O₂)²⁻ fragment bridging the two Tm(III) ions was observed. In contrast, treatment of [(1,2,4-C₅H₂^tBu₃)₂Tm] with an excess of CO led to [{(1,2,4-C₅H₂^tBu₃)₂Tm}₂(μ-O₂CCCO)]. This reaction is similar to the one observed of [(C₅Me₅)₂Sm(THF)₂] with CO (Scheme 63).¹⁹⁵

Scheme 64 Reaction of [(1,2,4-C₅H₂^tBu₃)₂Tm] with CO.⁴¹

CO₂ can be reduced by [(C₅Me₅)₂Sm(THF)₂] and related samarocenes (Scheme 65).^23,196 When [(C₅Me₅)₂Sm(THF)₂] is exposed to an atmosphere of carbon dioxide in toluene or hexane, an intractable mixture of products was obtained according to NMR spectroscopy. Even at −78 °C, no single product was isolated. However, when carried out in THF, [(C₅Me₅)₂Sm(μ-η²:η²-O₂C₂O₂)Sm(C₅Me₅)₂] is cleanly formed and isolated in >90% yield. As shown by X-ray crystallography, reductive coupling of two carbon dioxides gave an oxalate ligand, which bridges two (C₅Me₅)₂Sm moieties. Reductive coupling is a typical reaction pattern of decamethylsamarocene. The structural data was not of sufficient quality for the determination of bond distances and angles. However, the structure was unequivocally established. In contrast to reductive coupling, the related COS undergoes disproportionation when reacted with [(C₅Me₅)₂Sm(THF)₂] to give [(C₅Me₅)₂Sm(μ-η²:η¹-S₂CO)Sm(C₅Me₅)₂(THF)] (Scheme 65). However, according to NMR spectroscopy, another unidentified product is present in this reaction. By adjusting reaction conditions, the complex can be obtained in 90% yield. COS is transformed into a dithiocarbonate ligand, that bridges one samarocene fragment in a bidentate way with both sulphur donors, and the other with the single oxygen atom. Orange crystals were investigated with X-ray crystallography. The dithiocarbonate is planar, the Sm–S bonds are similar with 2.773(2) and 2.821(2) Å while the Sm–O bond is expectedly shorter (2.270(5) Å). Mechanisms for the formation of both CO₂ and COS-derived compounds were discussed by the authors.¹⁹⁶ Interestingly, though no experimental details are given, [(C₅Me₅)₂Yb(THF)₂] forms the sulphur bridged dimer [(C₅Me₅)₂YbSYb(C₅Me₅)₂].¹⁹¹

Scheme 65 Reactivity of [(C₅Me₅)₂Sm(THF)₂] towards CO₂ and COS.¹⁹⁶

A different reactivity of CO₂ was seen by using the base free and bulky substituted samarocenes [(1,3-C₅H₃^tBu₂)₂Sm] and [(1,2,4-C₅H₂^tBu₃)₂Sm].³¹ A clean formation of bridged carbonate samarium dimers [{(1,3-C₅H₃^tBu₂)₂Sm}₂(μ-CO₃)] and [{(1,2,4-C₅H₂^tBu₃)₂Sm}₂(μ-CO₃)] was seen. Apparently, a reductive disproportionation of CO₂ with the release of CO must have taken place. This contrasts with the formation of the oxalate-bridged samarium dimer shown in Scheme 65. The structures of both compounds are slightly different. While in [{(1,3-C₅H₃^tBu₂)₂Sm}₂(μ-CO₃)] a μ-η²:η²-coordination mode of the carbonate ligand was observed, a μ-η¹:η²-coordination mode was seen in [{(1,2,4-C₅H₂^tBu₃)₂Sm}₂(μ-CO₃)].³¹

Recently, the reactivity of SO₂ towards some of the divalent metallocenes of the lanthanides was extensively investigated.^197,198 Each compound was dissolved in THF. SO₂ was subsequently condensed onto the solution at −50 °C. In the case of [(C₅Me₅)₂Eu(THF)₂], the dinuclear complex [{(C₅Me₅)Eu}₂(μ,κO,κO′-C₅Me₅SO₂)₄] was isolated as the only product (Scheme 66). A sulfinate ligand was formed via nucleophilic attack of one Eu(C₅Me₅) ligand to SO₂ In total, four C₅Me₅SO₂ ligands bridge two Eu(C₅Me₅)-fragments. Thereby, a cage-like structure was created. The O-Eu distances are in the range of 2.309(3) to 2.354(2) Å.

Scheme 66 Reaction of [(C₅Me₅)₂Eu(THF)₂] with SO₂.¹⁹⁷

For [(C₅Me₅)₂Yb(THF)₂], three different products were crystallised from the reaction with SO₂. By reductive coupling of two SO₂ molecules, the dinuclear Yb(III) complex [(C₅Me₅)₂Yb(μ-η²:η²-O₂S₂O₂)Yb(C₅Me₅)₂] was formed as a minor product (Scheme 67).¹⁹⁷

Scheme 67 Reaction of [(C₅Me₅)₂Yb(THF)₂] with SO₂.¹⁹⁷

The coupling results in a bridging dithionite ligand, in which the sulphur atoms adopt a pyramidal structure. The Yb fragments are coordinated by the oxygen atoms. The Yb–O bond lengths range from 2.256(7) to 2.273(8) Å. Since this compound is only a minor product full characterization was not feasible. The main product of the reaction between SO₂ and [(C₅Me₅)₂Yb(THF)₂] is the tetranuclear, red-violet complex [{(C₅Me₅)₂Yb(S₂O₄)}₂{(C₅Me₅)Yb(C₅Me₅SO₂)}₂], wherein both the reductive coupling of two SO₂ molecules and the nucleophilic attack of C₅Me₅ anion on an SO₂ occurred to form the sulfonate ligands C₅Me₅SO₂. Structural investigations also show a cage-like structure, in which two (C₅Me₅)Yb units are alternatively bridged by sulfonate and dithionite ligands. The two dithionite ligands additionally chelate one more (C₅Me₅)₂Yb fragment each, ultimately resulting in the tetranuclear compound. The (C₅Me₅)Yb–O bonds to the sulfinate ligands are 2.222(4) and 2.252(4) Å, the (C₅Me₅)Yb–O dithionite bonds are 2.356(4) and 2.225(4) Å. The outer (C₅Me₅)₂Yb fragments show Yb–O distances of 2.223(4) and 2.255(4) Å. The third product, which was obtained upon heating, is the structural analogue of the dinuclear tetrasulfinate europium complex described above (Scheme 66).

Upon treatment of the more reactive metallocene [(C₅Me₅)₂Sm(THF)₂] with SO₂, four isolable compounds were obtained (Scheme 68).¹⁹⁸ As major products [{(C₅Me₅)₂Sm(S₂O₄)}₂{(C₅Me₅)Sm(C₅Me₅SO₂)}₂], which is an analogue to the corresponding Yb compound, and [{(C₅Me₅)₂Sm(C₅Me₅SO₂)}₂] were formed. The latter is a dinuclear sulfinate complex, in which the two C₅Me₅SO ligands were created by the nucleophilic attack of a C₅Me₅ anion to SO₂. The resulting mono-(C₅Me₅)Sm units were also found in the other products. The two Sm centres in [{(C₅Me₅)₂Sm(C₅Me₅SO₂)}₂] are bridged by the oxygen atoms of the sulfinate, creating an eight-membered ring. The Sm–O distances determined in the structure are 2.339(2) and 2.355(2) Å, respectively. The other two minor products are [(C₅Me₅)₂Sm(μ-η²:η²-O₂S₂O₂)Sm(C₅Me₅)₂], also observed for Yb, and the dinuclear compound [(C₅Me₅)Sm}₂(μ,κO,κO’-C₅Me₅SO₂)₄]. All structures were investigated by single crystal X-ray diffraction and revealed no unexpected features when compared with the corresponding Yb and Eu complexes.

Scheme 68 Products from the reaction of [(C₅Me₅)₂Sm(THF)₂] with SO₂.¹⁹⁸

The influence of the steric demand of the ligands was seen in the reaction of various metallocenes with the mineral realgar (As₄S₄). The reaction of [(C₅Me₅)₂Ln(THF)₂] (Ln = Sm, Yb) with As₄S₄ resulted the tetrametallic cage complex [{(C₅Me₅)₂Sm}((C₅Me₅)Sm)₃AsS₃{(C₅Me₅)AsS₂}₂(THF)₃] or the trimetallic cage compound [{(C₅Me₅)Yb}₃As₂S₄((C₅Me₅)AsS₂)(THF)₂] (Scheme 69).¹⁹⁹ The reduction of realgar leads to the formation of the unprecedented (C₅Me₅)AsS₂²⁻ anion in both compounds. Moreover, thioarsenate (AsS₃³⁻) is found in the Sm complex, while an As₂S₄⁴⁻ anion is seen in the Yb compound.

Scheme 69 Reaction of [(C₅Me₅)₂Ln(THF)₂] with the mineral realgar (As₄S₄).¹⁹⁹

Closed cage compounds are formed by either using bulkier ligands or a different As/S cage. The reaction of the bulky substituted samarocene [(1,2,4-C₅H₂^tBu₃)₂Sm] with As₄S₄ and the reaction of [(C₅Me₅)₂Yb(THF)₂] with dimorphite (As₄S₃) gave the closed eleven-vertex cage clusters [{(1,2,4-C₅H₂^tBu₃)Sm}₃(AsS₃)₂] and [{(C₅Me₅)Yb}₃(AsS₃)₂], which have similar structures (Scheme 70).¹⁹⁹ Both cages consist of two AsS₃³⁻ anions and three lanthanide cations.

Scheme 70 Synthesis of the closed cage compounds [(1,2,4-C₅H₂^tBu₃)₃(AsS₃)₂] and [{(C₅Me₅)Yb}₃(AsS₃)₂].¹⁹⁹

Reaction of [(MeO)₂P(S)S]₂ with [(C₅Me₅)₂Sm(THF)₂] gave the unusual trivalent dithiophosphate mono-pentamethylcyclopentadienyl-complex [(C₅Me₅)Sm{S₂P(OMe)₂}₂]₂.²⁰⁰ In the dinuclear complex, the samarium atom possesses a high coordination number, which is rather atypical for Sm²⁺ and more often found in Sm³⁺ ions. This effect would explain this unusual bridging between the Sm atoms.

Heterometallic complexes of samarium and tungsten or molybdenum were prepared from the reaction of [(C₅Me₅)₂Sm(THF)₂] and the tetrathiometalates [(PPh₄)₂MoS₄] and [(PPh₄)₂WS₄], respectively (Scheme 71).²⁰¹ By using the molybdenum compound with samarocene in THF a red compound was obtained, which was proven to be the trimetallic complex [{(C₅Me₅)₂Sm}₂Mo(μ-S)₄][PPh₄]. The central metal atom of the thiometalate was reduced during the reaction. The structure was established by single crystal X-ray diffraction. The central molybdenum atom is surrounded tetrahedrally by four sulphur ions, which are also bridging the Sm ions of the (C₅Me₅)₂Sm units. The Sm–S distances are 2.791(2) and 2.796(2) Å. In contrast to the molybdenum compound, the reaction of [(C₅Me₅)₂Sm(THF)₂] with [(PPh₄)₂WS₄] resulted in the bimetallic complex [(C₅Me₅)₂Sm(μ-S)₂WS₂][PPh₄]. In this case, only the reduction of one PPh₄ anion occurs. The sulphur atoms are tetrahedrally arranged around the tungsten atom. Only two of them are also binding to the Sm(C₅Me₅)₂ unit, while the remaining two sulphur atoms are arranged terminally. The Sm–S bond distances are 2.817(8) and 2.841(7) Å. The authors explain the different reactivity of group 6 thiometalates with the different redox potentials of Mo(VI) and W(VI).

Scheme 71 Reactions of [(C₅Me₅)₂Sm(THF)₂] with [(PPh₄)₂MoS₄] and [(PPh₄)₂WS₄].²⁰¹

5.3. Reactivity towards elements

Especially the two stronger reducing metallocenes ytterbocene and samarocene are well known to reduce elements to anionic ligands.

5.3.1. Group 15. One of the landmark discoveries of [(C₅Me₅)₂Sm] in terms of reactivity was the activation of molecular dinitrogen (Scheme 72).²⁰² When a toluene solution of desolvated decamethylsamarocene was exposed to a nitrogen atmosphere, brown-reddish crystals of [(C₅Me₅)₂Sm(μ-η²:η²-N₂)Sm(C₅Me₅)₂] crystallised during four weeks. The reaction occurred also to some extent when solid [(C₅Me₅)₂Sm] was stored under a nitrogen atmosphere. It was the first example of a planar side on the coordination of an N₂ molecule between two metal centres. The Sm–N distances are between 2.3 and 2.4 Å, which is typical for Sm(III)–N bonds. However, the N–N bond is rather short, indicating only a weak activation (Fig. 9). The samarocene units in the complex are orientated perpendicular to each other such that the centroids of the pentamethylcyclopentadienyl ligands form a tetrahedron. The dinitrogen coordination in [(C₅Me₅)₂Sm(μ-η²:η²-N₂)Sm(C₅Me₅)₂] is reversible. When the compound was redissolved in toluene, nitrogen was released. It is also possible to remove the coordinated nitrogen by exposing the solid compound to vacuum.

Scheme 72 Reactivity of unsolvated decamethylsamarocene towards molecular nitrogen.²⁰²

Fig. 9 Molecular structure of [(C₅Me₅)₂Sm(μ-η²:η²-N₂)Sm(C₅Me₅)₂] in the solid state (Reproduced from the CIF file CCDC: 1181490).²⁰²

NMR studies revealed a temperature dependent equilibrium between samarocene and the corresponding dinitrogen complex in solution.

A similar dinitrogen activation was seen when [(1,2,4-C₅H₂^tBu₃)₂Nd(μ-I)(K[18]crown-6)] was treated with N₂.⁹ As result the dimer [(1,2,4-C₅H₂^tBu₃)₂Nd(μ-η²:η²-N₂)Nd(1,2,4-C₅H₂^tBu₃)₂] was isolated.

A cooperative dinitrogen activation was observed by using a mixture of [(C₅Me₅)₂Sm(THF)₂] and 9,10-Me₂-9,10-diboraanthracene (Scheme 73).²⁰³ As result the salt [(C₅Me₅)₂Sm(THF)₂][(C₅Me₅)₂Sm(η²-N₂B₂C₁₄H₁₄)] was obtained, in which the N₂²⁻ anion is stabilised between a [(C₅Me₅)₂Sm(THF)₂]⁺ cation and the diboraanthracene Lewis acid.

Scheme 73 Synthesis of [(C₅Me₅)₂Sm(THF)₂][(C₅Me₅)₂Sm(η²-N₂B₂C₁₄H₁₄)].²⁰³

The reactions of heavier congeners of dinitrogen with the samarocene were investigated by Roesky and co-workers.^126,204,205 A toluene solution of desolvated [(C₅Me₅)₂Sm] activates white phosphorus (P₄) vapour to yield a molecular polyphosphide [{(C₅Me₅)₂Sm}₄P₈] (Scheme 74).²⁰⁶ By diffusing P₄ vapour in the solution, pyrophoric red crystals were formed. Structural studies show a tetranuclear Sm compound with a realgar-type P₈ cage in the centre of the molecule. The four {(C₅Me₅)₂Sm} moieties are residing on the corners of a square, each one binding to two phosphorous atoms (Sm–P 2.997(2)–3.100(2) Å). The authors note the analogy of the central P₈⁴⁻ anion to the well-known Zintl anion P₇³⁻ and theoretical studies were carried out. The formation of the Zintl type P₈⁴⁻ anion by samarium pronounces once more the parallels to the reductive strength of the alkaline metals.

Scheme 74 Reactions of [(C₅Me₅)₂Sm] with P₄ (Reproduced from the CIF file CCDC 740249).²⁰⁶

The heavier congeners of P₄ are either highly reactive (As) or non-existing under ambient conditions (Sb, Bi). Yellow arsenic, As₄, is inconvenient to synthesise, extremely photosensitive and highly prone towards decomposition into its thermodynamically stable modification, grey arsenic. Thus, the reaction of freshly prepared solution of As₄ in toluene at room temperature with unsolvated decamethylsamarocene [(C₅Me₅)₂Sm] as divalent lanthanide source resulted besides unidentified and inseparable side products in [{(C₅Me₅)₂Sm}₂(μ-η²:η²-As₂)].²⁰⁵ [{(C₅Me₅)₂Sm}₂(μ-η²:η²-As₂)] (Scheme 75) is similar to Evans et al. dinitrogen complex [{(C₅Me₅)₂Sm}₂(μ-η²:η²-N₂)] (Scheme 72).²⁰² The As–As bond length is 2.278(2) Å, which corresponds to a double bond rather than a single bond, since {As=As}²⁻ moieties have bond lengths of 2.2 to 2.3 Å in transition metal complexes.^207,208

Scheme 75 Reactions of [(C₅Me₅)₂Sm] with As₄ and As⁰_nano.²⁰⁵

A more straightforward approach to polyarsenides starts from elemental As⁰ nanoparticles (As⁰_nano, d = 7.2 ± 1.8 nm), which were obtained by the reduction of AsI₃ with a freshly prepared solution of lithium naphthalenide.²⁰⁵ Treatment of the as-prepared As nanoparticles with [(C₅Me₅)₂Sm] at 60 °C for short reaction times (<1 day) resulted again in [{(C₅Me₅)₂Sm}₂(μ-η²:η²-As₂)] in low yields. Also here, by-products formed during the reaction accompanied the product.

In contrast, the samarium polyarsenide [{(C₅Me₅)₂Sm}₄As₈], which is the heavier congener of the realgar-type [{(C₅Me₅)₂Sm}₄P₈] polyphosphide, was formed after prolonged heating of As⁰_nano and [(C₅Me₅)₂Sm] at 120 °C. This polyarsenide is the most important congener of the realgar type [{(C₅Me₅)₂Sm}₄P₈]. All Sm–As bond distances are similar and are between 3.0814(10) Å and 3.1734(10) Å. Within the {As₈}⁴⁻ tetraanion, all As–As bonds are similar (2.4044(12) Å and 2.5003(12) Å) and in the range of single bonds (Fig. 10).

Fig. 10 Molecular structure of [{(C₅Me₅)₂Sm}₄As₈] in the solid (Reproduced from the CIF file CCDC 1880716).²⁰⁵

As already mentioned, there is no reactive elemental modification for antimony. To increase the reactivity, Sb/Hg alloy or Sb⁰ nanoparticles were used as an activated antimony source.²⁰⁴ Reaction of [(C₅Me₅)₂Sm] with Sb/Hg stirred at room temperature for 48 hours resulted in the polystibide complexes [{(C₅Me₅)₂Sm}₂(μ-η²:η²-Sb₂)] and surprisingly, a mercury-containing compound [{{(C₅Me₅)₂Sm}₂Sb}₂(μ-Hg)] (Scheme 76).^204,205 The authors could not separate the complexes on a preparative scale. The compound [{(C₅Me₅)₂Sm}₂(μ-η²:η²-Sb₂)] is isostructural to other group 15 compounds such as [{(C₅Me₅)₂Sm}₂(μ-η²:η²-E₂)] (E = N, As, Bi).^179,202,205 The Sm–Sb bond lengths of 3.2141(9) Å are in the range of those in [{(C₅Me₅)₂Sm}₃(μ-η²:η²:η¹-Sb₃)(THF)] (Sm–Sb 3.162(1) Å - 3.205(1) Å)¹⁸⁵ and the short Sb–Sb bond length of 2.6593(15) Å is similar to other Sb–Sb double bonds (Sb–Sb’ 2.642(1) Å) in the distibene [TbtSb=SbTbt] (Tbt = 2,4,6-tris[bis(trimethylsilyl)-methyl]phenyl).²⁰⁹ [{{(C₅Me₅)₂Sm}₂Sb}₂(μ-Hg)] is the first molecular Sb–Hg bond ever reported with d(Hg–Sb) = 2.6400(6) Å.²⁰⁴

Scheme 76 Isolated intermediates from the reaction of [(C₅Me₅)₂Sm] with Sb/Hg alloy.²⁰⁴

Performing the reaction with an identical reaction mixture at 60–70 °C for two days resulted in [{{(C₅Me₅)₂Sm}₃(μ⁴,η^1:2:2:2-Sb₄)}₂Hg] as minor product. The elevated reaction temperature thus led to compounds with a higher Sb and a lower Hg ratio. As shown by X-ray analysis, two [{(C₅Me₅)₂Sm}₃Sb₄]⁻-moieties are linked by Hg²⁺ (Scheme 76). In [{{(C₅Me₅)₂Sm}₃(μ⁴,η^1:2:2:2-Sb₄)}₂Hg] three samarocene fragments are connected by a [Sb₄]⁴⁻ anion, in which the Sb atoms clearly adopt different formal oxidation states with an average of +1.²⁰⁴

Finally, the authors heated the reaction to 120 °C for two days in order to obtain the f-element realgar-type polystibide complex [{(C₅Me₅)₂Sm}₄Sb₈] (Scheme 77).²⁰⁴ [{(C₅Me₅)₂Sm}₄Sb₈] is isostructural to its lighter analogue [{(C₅Me₅)₂Sm}₄E₈] (E = P, As),^205,206 demonstrating the accessibility of also heavier group 15 polyanions in f-element chemistry. Four {(C₅Me₅)₂Sm}⁺-fragments bridge the corners of a [Sb₈]⁴⁻ Zintl anion.

Scheme 77 Reaction of Sb/Hg alloy with [(C₅Me₅)₂Sm] to give different {(C₅Me₅)₂Sm}/Sb/Hg species, which further react to give [{(C₅Me₅)₂Sm}₄Sb₈] (left). Reactions of Sb⁰ nanoparticles with [(C₅Me₅)₂Sm] yield [{(C₅Me₅)₂Sm}₄Sb₈] (right).²⁰⁴

As an alternative approach, Sb⁰ nanoparticles synthesised under salt-free and inert conditions from SbCl₃ and 2,3,5,6-tetramethyl-1,4-bis(trimethylsilyl)-1,4-diaza-2,5-cyclo-hexadiene²¹⁰ were employed as reactive antimony source.²⁰⁴ The nanoparticles were subsequently reacted with [(C₅Me₅)₂Sm] to give the polystibide complexes [{(C₅Me₅)₂Sm}₂(μ-η²:η²-Sb₂)] again. After prolonged heating of the reaction mixture, the Zintl complex [{(C₅Me₅)₂Sm}₄Sb₈] was isolated as the sole product demonstrating that the formation of this compound is the thermodynamically favoured pathway (Scheme 77).

5.3.2. Group 16. Regarding the chalcogens, [(C₅Me₅)₂Yb(OEt₂)] does not react with S₈, while the more reactive samarocene reacts with an excess of elemental sulphur in THF in a 2 : 3 ratio to give [(C₅Me₅)₂Sm(THF)(μ-η¹:η³-S₃)Sm(C₅Me₅)₂] (Scheme 78).¹⁹⁰

Scheme 78 Reactivity of [(C₅Me₅)₂Sm(THF)₂] towards sulphur, selenium and tellurium.¹⁹⁰

The reaction is so rapid that the S²⁻ containing complex (see above, Scheme 59) cannot be isolated from this reaction. However, [(C₅Me₅)₂Yb(μ-S)Yb(C₅Me₅)₂] was isolated from the reaction mixture of [(C₅Me₅)₂Yb(THF)₂] and [(V^MeCp)₂(μ-S)(μ,η^1:1-S₂)(μ,η^2:2-S₂)] (^MeCp = C₅H₄Me).²¹¹

The more reactive thulium reagent [(1,2,4-C₅H₂^tBu₃)₂Tm] reacts with S₈ to give a [(1,2,4-C₅H₂^tBu₃)₂Tm₂(μ-η²:η²-S₂)], in which the S₂²⁻ unit is side-on coordinated with both Tm ions.²¹²

By treating [(C₅Me₅)₂Sm(THF)₂] with excess of selenium, the selenium analogue [(C₅Me₅)₂Sm(THF)(μ-η¹:η³-Se₃)Sm(C₅Me₅)₂] was isolated as dark red crystals. In the case of solid tellurium, the corresponding tellurium complexes are formed; however, the reaction with tellurium is slower, owing to the lower oxidative reactivity of this element. [(C₅Me₅)₂Sm(THF)(μ-η¹:η³-Se₃)Sm(C₅Me₅)₂] was characterised by single crystal X-ray crystallography. The structure reveals a dinuclear Sm compound, in which the two metal centres are bridged by a Se₃²⁻ anion. One of the Sm atoms has an additional THF ligand bound to it. The Se₃²⁻ anion is asymmetrically coordinated, in the sense that one (C₅Me₅)₂Sm unit coordinates to all three selenium atoms in an η³-fashion, while the other (C₅Me₅)₂Sm moiety coordinates only to one Se atom from the opposing side of the triangular anion (Scheme 79). [(C₅Me₅)₂Sm(THF)(μ-η¹:η³-Se₃)Sm(C₅Me₅)₂] slowly transformed to the interesting hexanuclear Se-Sm cluster [(C₅Me₅)Sm]₆(Se₁₁)].²¹³

Scheme 79 Structure of [(C₅Me₅)₂Sm(THF)(μ-η¹:η³-Se₃)Sm(C₅Me₅)₂].¹⁹⁰

[(C₅Me₅)₂Yb(OEt₂)] reacted with selenium and tellurium to give the anticipated chalcogenide bridged compounds [{(C₅Me₅)₂Yb}₂(μ-Se)] and [{(C₅Me₅)₂Yb}₂(μ-Te)] (also accessible from the reaction of phosphine chalcogenide compounds, see above, Scheme 59).¹⁹¹ However, when [(C₅Me₅)₂Yb(OEt₂)] is stirred with a “large excess” of tellurium in hexane over a longer time, black crystals of [(C₅Me₅)₂Yb(μ-η²:η²-Te₂)Yb(C₅Me₅)₂] were isolated (Scheme 80).²¹⁴ The ditelluride anion is located between the two (C₅Me₅)₂Yb units. The Te–Te bond is perpendicular to the Yb–Yb axis. The Yb–Te bond lengths are 3.153(9) and 3.1598(7) Å.

Scheme 80 Structure of [(C₅Me₅)₂Yb(μ-η²:η²-Te₂)Yb(C₅Me₅)₂].²¹⁴

Similar compounds can also be obtained by the reaction of polychalcogenide complexes of vanadium with [(C₅Me₅)₂Sm(THF)₂] and [(C₅Me₅)₂Yb(THF)₂]. Thus, a mixture of the mono-[(C₅Me₅)₂Yb(μ-Se)Yb(C₅Me₅)₂] and the diselenides, [(C₅Me₅)₂Yb(μ-η²:η²-Se₂)Yb(C₅Me₅)₂] was isolated from the reaction of [(C₅Me₅)₂Yb(THF)₂] and [(V(C₅H₄Me)₂(μ-Se)(μ,η^1:1-Se₂)(μ,η^2:2-Se₂)].²¹¹ [(C₅Me₅)₂Sm(μ-Te)Sm(C₅Me₅)₂] was obtained from [(C₅Me₅)₂Sm(THF)₂] and [{V(nacnac)}₂(μ-Te)₂] ((nacnac) = (HC(C(Me)NC₆H₃-ⁱPr₂)₂)).²¹¹

The reaction of the metallocenes with organometallic reagents leads us to the next chapter.

5.4. Reaction with coordination compounds and organometallic reagents

Especially in the early 1980s, Andersen investigated the reactivity of [(C₅Me₅)₂Yb] towards transition metal–carbonyl complexes.^215–217 In general, reactions of [(C₅Me₅)₂Ln(OEt₂)] (Ln = Sm, Yb) with [Co₂(CO)₈],^215,218 [Fe₂(CO)₉],²¹⁶ [Mn₂(CO)₁₀]^217,219 and [Re₂(CO)₁₀]^217,219 were studied. In every case reduction of the carbonyl complexes occurs, yielding compounds with bridging carbonyl ligands. Isocarbonyl complexes with oxygen atoms binding to the metal centres were formed in all cases. For [Co₂(CO)₈], the heterobimetallic complex [(C₅Me₅)₂Yb(OC)Co(CO)₃(THF)] was obtained (Scheme 81).

Scheme 81 Reaction of [(C₅Me₅)₂Yb(OEt₂)] and [Co₂(CO)₈].²¹⁵

The stronger reducing agent [(C₅Me₅)₂Sm(THF)₂] compared to the above mentioned [(C₅Me₅)₂Yb(THF)₂] also reacts with [Co₂(CO)₈] and forms the structurally related product [(C₅Me₅)₂Sm(OC)Co(CO)₃(THF)].²¹⁸ The infinite chain [{(C₅Me₅)₂Sm(THF)}{(μ-CO)₂(CO)₃Mn}]_∞ was obtained upon reaction of [(C₅Me₅)₂Sm(THF)₂] with [Mn₂(CO)₁₀]. Here, two CO ligands are coordinating with each samarocene (Scheme 82).²¹⁹

Scheme 82 Reaction of [(C₅Me₅)₂Sm(THF)₂] with [Mn₂(CO)₁₀].²¹⁹

In contrast, the reaction with [Re₂(CO)₁₀] led to the formation of [{(C₅Me₅)₂Sm}₃{(O₄C₄)(μ₃-CO)₂(μ-CO)(CO)₅Re₂}Sm(C₅Me₅)₂(THF)] (Scheme 83).²¹⁹ The central [(O₄C₄)(μ₃-CO)₂(μ-CO)(CO)₅Re₂]⁴⁻ core has been formed by a four-fold reduction process and the [μ-O₄C₄] unit can formally be considered as tetra-anionic where each oxygen atom is negatively charged. The Re–Re bond is slightly shorter than the Re–Re single bond in Re₂(CO)₁₀²²⁰ (2.934(3) vs. 3.041(11) Å, respectively) but theoretical investigations indicate only a weak interaction.

Scheme 83 Reaction of [(C₅Me₅)₂Sm(THF)₂] with [Re₂(CO)₁₀].²¹⁹

The reaction with iron and manganese carbonyls resulted in more sophisticated structures. The pentanuclear Yb–Fe complex [{(C₅Me₅)₂Yb}₂{Fe₃(CO)₇(μ-CO)₄}]²¹⁶ and the polymeric Mn–Yb compound [{(C₅Me₅)₂Yb}Mn(CO)₅]_n²¹⁷ consist of polymeric chains with both [(C₅Me₅)₂Yb(μ-CO)₃Mn(CO)₂] moieties and dimeric (C₅Me₅)₂Yb(μ-CO)₂Mn(CO)₃ subunits. A similar structure was found for the reaction with [Re₂(CO)₁₀].²¹⁷ No significant influence of bulkier cyclopentadienyl ligands is observed on the reactivity. Thus, reaction of [(Cp^Bz5)₂Sm] with [Co₂(CO)₈] or [Mn₂(CO)₁₀] in toluene gave the bridged tetrametallic complexes [{(Cp^Bz5)₂Sm}₂{(μ-OC)₂Co(CO)₂}₂] or [{(Cp^Bz5)₂Sm}₂{(μ-OC)₂Mn(CO)₃}₂].⁴⁸

Besides pure metal carbonyls also organometallic carbonyl derivatives were reacted with [(C₅Me₅)₂Sm(THF)₂]. Thus, the reaction with [(C₅Me₅)Fe(CO)₂]₂ led to the tetranuclear complex [(C₅Me₅)₂Sm(μ-OC)₂Fe(C₅Me₅)]₂, which forms a twelve membered ring.²²¹

The sulphur carbonyl complex [Fe₂(μ-S₂)(CO)₆] was reacted with [(C₅Me₅)₂Ln(THF)₂] (Ln = Eu, Sm, Yb) in toluene (Scheme 84).²²² While for [(C₅Me₅)₂Eu(THF)₂] no reaction was observed, [(C₅Me₅)₂Sm(THF)₂] and [(C₅Me₅)₂Yb(THF)] yielded octanuclear complexes of the type [Fe₆Ln₂(μ₃-S)₆(μ,η²-CO)₄(CO)₈(η⁵-C₅Me₅)₄] as black crystals. The compounds form wheel-type structures, with a dianionic Fe₆(μ₃-S₂)(CO)₁₂ unit in the centre. The (C₅Me₅)₂Ln moieties are bound via a bridging sulphur atom and the oxygen atoms of the carbonyl ligands. The sulphur atoms are connected to the inner iron atoms.

Scheme 84 Reactions of [(C₅Me₅)₂Ln(THF)₂] (Ln = Sm, Yb) with [Fe₂(μ-S₂)(CO)₆].²²²

Solvent-free [(C₅Me₅)₂Sm] is able to reduce the organometallic compound AlEt₃. With an excess of AlEt₃ in toluene, elemental aluminium was deposited and the red complex [(C₅Me₅)₂Sm(μ-Et)₂AlEt₂] was formed.^223,224 Two carbon atoms attached to the aluminium ion form a bridge to the samarium atom with Sm–C distances of 2.662(4) Å. Here, the difference in the reactivity between [(C₅Me₅)₂Sm] and [(C₅Me₅)₂Yb] was emphasised: R₃Al compounds react with ytterbocene with the formation of Lewis acid–base adducts (see above).

Samarocene [(C₅Me₄R)₂Sm(THF)₂] (R = H, Me) also reduces AlMe₃ in toluene to form [(C₅Me₄R)₂Sm{(μ-Me)AlMe₂(μ-Me)}₂Sm(C₅Me₄R)₂] (Fig. 11).^225,226 Each samarocene unit is linked to two tetrahedral (μ-Me)₂AlMe₂ moieties with the Sm(μ-Me)-Al linkages being nearly linear with angles in [(C₅Me₅)₂Sm{(μ-Me)AlMe₂(μ-Me)}₂Sm(C₅Me₅)₂] of 175.2(9)° and 177.8(7)° and a planar arrangement.

Fig. 11 Molecular structure of [(C₅Me₅)₂Sm{(μ-Me)AlMe₂(μ-Me)}₂Sm(C₅Me₅)₂] in the solid state (hydrogen atoms omitted for clarity). The central [Sm–C–Al–C–Sm–C–Al–C–] unit is almost planar (Reproduced from the CIF file CCDC: 1278704).²²⁵

Upon reaction with dimethylzinc [Me₂Zn], [(C₅Me₅)₂Yb] reacts with the methyl-bridged zincate [(C₅Me₅)₂Yb(μ-Me)₂ZnMe], which can be readily prepared in 2–3 g scale.²²⁷ Upon recrystallization from THF the [ZnMe]-fragment is cleaved and [(C₅Me₅)₂YbMe(THF)] is obtained. When [ZnMe₂] is replaced with the more reactive [CuMe], the compounds [(C₅Me₅)₂Yb(μ-Me)Yb(C₅Me₅)₂] and [(C₅Me₅)₂Yb(μ-Me)Yb(C₅Me₅)₂(Me)] were obtained.²²⁷ [(C₅Me₅)₂Yb(μ-Me)Yb(C₅Me₅)₂] is a mixed valent compound containing a Yb(II) and a Yb(III) species. When [(C₅Me₅)₂Yb(Et₂O)_n] (n = 0, 1) is combined with [ZnPh₂] in toluene, [(C₅Me₅)₂Yb(μ-Ph)₂ZnPh] is yielded (Scheme 85). Its structural motive is similar to the one of [(C₅Me₅)₂Yb(μ-Me)₂ZnMe].

Scheme 85 Reaction of [(C₅Me₅)₂Yb(Et₂O)_n] (n = 0, 1) with [ZnPh₂].²²⁷

Since the reaction with diphenyl mercury [HgPh₂] did not lead to any results, [Hg(C₆F₅)₂] was investigated. This led to the compound [(C₅Me₅)₂Yb(C₆F₅)], which is similar to the previously reported reaction of [(C₅H₅)₂Yb] and [Hg(C₆F₅)₂].²²⁸ To investigate more into the formation of [(C₅Me₅)₂YbMe] other organometallic methyl transfer reagents were employed. When [(C₅Me₅)₂VMe] is mixed with [(C₅Me₅)₂Yb] in a 2 : 1 ratio, [(C₅Me₅)₂Yb(μ-Me)Yb(C₅Me₅)₂] is formed and [(C₅Me₅)₂V] can be isolated from the mother liquor. This is also applicable for halides and other anions and [(C₅Me₅)₂Yb(μ-X)Yb(C₅Me₅)₂] (X = F, Cl, Br, I, BH₄, H) can be synthesised in the same manner. When the same reaction was attempted with [(C₅Me₅)₂TiX] (X = Cl, Br, H, Me, BH₄) the bridged compounds [(C₅Me₅)₂Yb(μ-X)Ti(C₅Me₅)₂] were obtained, in which the key feature is the nearly linear Yb–X–Ti bridge.²²⁷

[(C₅Me₅)₂Sm(THF)₂] reduces diphenylmercury to clean elemental mercury and the orange complex [(C₅Me₅)₂Sm(Ph)(THF)].²²⁹ The structure was established by X-ray crystallography.

In 2001, [{(C₅Me₅)₂SmPh}₂] was synthesised again from [(C₅Me₅)₂Sm] with HgPh₂ for mechanistic studies of the samarium-catalyzed redistribution of Ph₃SiH₃ to Ph₂SiH₂ and SiH₄.²³⁰ It was assumed that Sm–Ph would play a key role as an intermediate. The first attempts to synthesise [{(C₅Me₅)₂SmPh}₂] were conducted in toluene, but the reaction led to [(C₅Me₅)₂Sm(CH₂Ph)]. The anticipated product was finally obtained with benzene as solvent. However, [{(C₅Me₅)₂SmPh}₂] is thermally unstable and decomposes even at −35 °C under a nitrogen atmosphere. Thus, the complex was only characterised in solution by NMR.

[(C₅Me₅)₂Yb] reacts with [(C₅H₅)₂M] (M = V, Cr, Mn, Co and Ni) in an exchange reaction to yield [(C₅Me₅)₂Yb(C₅H₅)(THF)_n] (n = 0 or 2, depending on the conditions). It is noteworthy that the reaction proceeds on different timescales. The Ni and Mn compounds react within 24 hours, whereas the other metallocenes need timespans ranging from days to weeks.²³¹

In recent years, the reactivity towards polyphosphide complexes was investigated.^232–234 Pentaphosphaferrocene [(C₅Me₅)Fe(P₅)] and [(C₅Me₄R)₂Sm(THF)₂] (R = Me, ⁿPr) reacted under reductive coupling to [((C₅Me₅)Fe)₂P₁₀{Sm(η⁵-C₅Me₄R)₂}₂] (R = Me, ⁿPr) (Scheme 86).²³² A phosphorous atom of each P₅ ring was coupled, thus forming the bicyclic (P₁₀)⁴⁻ ion. The iron atoms bind to four phosphorous atoms each, while the fifth is connected to the other ring. Each samarocene unit connects to two phosphorus atoms, each on opposing sides of the polyphosphide ion.

Scheme 86 Synthesis of [((C₅Me₅)Fe)₂P₁₀{Sm(η⁵-C₅Me₄R)₂}₂] (R = Me, ⁿPr).²³²

In contrast to pentaphosphaferrocene [(C₅Me₅)Fe(P₅)], the corresponding arsenide species [(C₅Me₅)Fe(As₅)] is less stable due to a weaker As–As bond. Reduction of [(C₅Me₅)Fe(As₅)] with [(1,3-C₅H₃^tBu₂)₂Sm(THF)] led to the arsenic poor [{(1,3-C₅H₃^tBu₂)₂Sm}(μ,η⁴:η⁴-As₄){Fe(C₅Me₅)}] and the arsenic-rich species [{(1,3-C₅H₃^tBu₂)₂Sm}₂(As₇){Fe(C₅Me₅)}] (Scheme 87).²³⁵ These compounds were the first polyarsenides of rare earth metals. Both compounds form parallel. They were separated by crystallization from different solvents. [{(1,3-C₅H₃^tBu₂)₂Sm}(μ,η⁴:η⁴-As₄)[Fe(C₅Me₅)] is a nice example of a d/f-triple decker sandwich complex with a purely inorganic planar middle deck. As supported by DFT calculations the As₄²⁻ unit is a 6π-aromatic system, which is related to the cyclobutadiene dianion (CH)₄²⁻. The bond distances within the As₄²⁻ dianion are equal with an average As–As bond length of 2.410 Å. The arsenic-rich species [{(1,3-C₅H₃^tBu₂)₂Sm}₂(As₇){Fe(C₅Me₅)}] features an As₇³⁻ cage, which has a norbornadiene-like structure with two short As–As bonds in the scaffold.

Scheme 87 Synthesis of the first polyarsenides of rare earth metals: the arsenic poor species [{(1,3-C₅H₃^tBu₂)₂Sm}(μ,η⁴:η⁴-As₄){Fe(C₅Me₅)}] and the arsenic-rich species [{(1,3-C₅H₃^tBu₂)₂Sm}₂(As₇){Fe(C₅Me₅)}].²³⁵

The molecule [(C₅H₂(^tBu)₃Co)₂(μ,η^2:2-P₂)₂] containing two P₂ units was reacted with [(C₅Me₄R)₂Sm(THF)₂] (R = Me, ⁿPr) in heptane.²³⁶ Single crystal X-ray diffraction studies revealed the structure of the product as [(C₅H₂(^tBu)₃Co)₂P₄Sm(C₅Me₄R)₂] (R = Me, ⁿPr) (Scheme 88). Upon reduction, a bond between the two P₂ ligands was formed to establish a P₄ chain ligand. The samarocene unit is located between the C₅H₂(^tBu)₃Co units, bonding to two phosphorus atoms. The Sm–P bond lengths are 2.874(3) and 2.921(3) Å. Quantum chemical calculations were carried out. They showed that the P–P bond formation was not a direct consequence of reduction by samarium, as all the spin density resides on the Co atoms.

Scheme 88 Synthesis of [(C₅H₂(^tBu)₃Co)₂E₄Sm(C₅Me₅)₂] (E = P, As).^236,237

In a similar reaction, the arsenic compound [(C₅H₂(^tBu)₃Co)₂(μ,η^2:2-As₂)₂] featuring two As₂ units was reacted with [(C₅Me₄R)₂Sm(THF)₂] (R = Me, ⁿPr) to give [(C₅H₂(^tBu)₃Co)₂As₄Sm(C₅Me₄R)₂] (R = Me, ⁿPr) (Scheme 88), which is the first structural representative of open chain-like polyarsenides as ligands in the coordination sphere of the lanthanides.²³⁷ The central As₄Co₂ scaffold forms a (non-crystallographic) C₂-symmetric distorted trigonal prism, which is chiral. However, both enantiomers crystallised as racemates. The observed As–As bonds within the chain are in the range of those observed in yellow arsenic (2.44 Å) and thus can be considered single bonds.²³⁸

Treatment of [(C₅Me₅)₂Ln(THF)₂] (Ln = Sm and Yb) with the phosphide complex [{(C₅H₅)Mo(CO)₂}₂(μ,η^2:2-P₂)] resulted in a mixture of phosphide containing complexes, which were separated by fractional crystallization. The major product is the sixteen-membered, bicyclic complexes [{(C₅Me₅)₂Ln}₂P₂{(C₅H₅)Mo(CO)₂}₄] (Scheme 89).²³⁴ In the case of samarocene, two further complexes were isolated as side products, namely the P₄ complex [{(C₅Me₅)₂Sm}₂P₄{(C₅H₅)Mo(CO)₂}₂] and the P₅ complex [{(C₅Me₅)₂Sm}₃P₅{(C₅H₅)Mo(CO)₂}₃] (Scheme 89). Both complexes consist of cyclo-phosphide ligands resulting from reduction by the lanthanide. All compounds were crystallographically characterised. DFT calculations elucidate the electronic structure of the complexes.

Scheme 89 Reactions of [(C₅Me₅)₂Ln(THF)₂] (Ln = Sm, Yb) with [{(C₅H₅)Mo(CO)₂}₂(μ,η^2:2-P₂)].²³⁴

By reacting the heavier group 15 congener [{(C₅H₄^tBu)Mo(CO)₂}₂(μ,η^2:2-As₂)] with solvent-free [(C₅Me₅)₂Sm], it resulted in a straightforward formation of the mixed d/f-metal species [{(C₅Me₅)₂Sm}₂As₂{(C₅H₄^tBu)Mo(CO)₂}₂] (Scheme 90).²³⁹ Due to the two-electron reduction of [{(C₅H₄^tBu)Mo(CO)₂}₂(μ,η^2:2-As₂)] the Mo–Mo bond is cleaved and the [{(C₅H₄^tBu)Mo(CO)₂}₂(μ,η^2:2-As₂)]²⁻ dianion is formed. The central As₂²⁻ unit is side-on coordinated with the Mo atoms. This is in contrast to [{(C₅Me₅)₂Ln}₂P₂{(C₅H₅)Mo(CO)₂}₄] (Scheme 89)²³⁴ in which the P₂²⁻ unit is end-on bound. The As–As-bond length of 2.238(2) Å is in the range of As–As double bonds known in the literature and only slightly shortened in comparison to [{(C₅H₅)Mo(CO)₂}₂(μ,η^2:2-As₂)].^240,241 On the other hand, quantum chemical calculations suggest a weakened double bond.

Scheme 90 Reduction of [{(C₅H₄^tBu)Mo(CO)₂}₂(μ,η^2:2-As₂)] with [(C₅Me₅)₂Sm(THF)₂] resulting in [{(C₅Me₅)₂Sm}₂As₂{(C₅H₄^tBu)Mo(CO)₂}₂].²³⁹

When the analogue reaction of [(C₅Me₅)₂Sm] with [{(C₅H₄^tBu)Mo(CO)₂}₂(μ,η^2:2-Sb₂)] was performed the polystibide [{(C₅Me₅)₂Sm}₂Sb₄{(C₅H₄^tBu)Mo(CO)₂}₂], featuring a planar Sb₄ ring was isolated (Scheme 91).²³⁹ The planar Sb₄ ring, which consists of two short (Sb1–Sb2 2.7313(8) Å) and two longer (Sb1–Sb2′ 2.8608(7) Å) Sb–Sb bonds, is as a unprecedented ligand for an organometallic or coordination compound. This scaffold can be considered a metal-coordinated tetrastibacyclobutadiene unit.

Scheme 91 Synthesis of [{(C₅Me₅)₂Sm}₂Sb₄{(C₅H₄^tBu)Mo(CO)₂}₂].²³⁹

When [(C₅Me₅)₂Ln(THF)₂] (Ln = Sm, Yb) was treated with the molybdenum-polyphosphide sandwich compound [(C₅Me₅)Mo(CO)₂(η³-P₃)], reductive coupling of the triangular phosphide ions occurs.²³⁴ This resulted in a new P–P bond formation via a reductive dimerization. As a result, the 4d/4f hexaphosphide complexes [{(C₅Me₅)₂Ln}₂P₆{(C₅Me₅)Mo(CO)₂}₂] featuring an unusual P₆ ligand in the centre were formed (Scheme 92). The P–P bond length of the newly formed bond is in the range of a P–P single bond. Each of the two molybdenum-containing units and each of the two lanthanocene fragments coordinates with one of the triangular P₃ units.²³⁴

Scheme 92 Synthesis of [{(C₅Me₅)₂Ln}₂P₆{(C₅Me₅)Mo(CO)₂}₂] (Ln = Sm, Yb).²³⁴

6. Conclusions

In this article, we provided a comprehensive overview of the syntheses, properties and reactivities of divalent metallocenes of the lanthanides. Two main reaction pathways of these compounds have become clear: Lewis acid–base complex formation and redox reactions in which the lanthanide atom is oxidised in a one-electron reaction. These interesting reagents are characterised by the rich and diverse structures of the resulting products and their high reactivity. The observed reduction chemistry of the divalent lanthanocenes shows some similarities with the reduction chemistry of alkali metals. In addition to the similarities in reactivity, some products also show structural correlations. However, in contrast to the alkaline metals, the reactivity can be fine-tuned via the steric bulk of the substituents of the cyclopentadienyl rings. Moreover, the electron transfer is in some cases reversible. The discovery of the so-called non-classical divalent lanthanides (divalent lanthanides beyond Sm, Eu, Yb) opens another almost unexplored pathway towards unprecedented reactivity. However, even with the established classical compounds, a large number of new reactions have been published in recent years. Thus, the journey is far from being over. We hope that we have been able to show that divalent metallocenes of the lanthanides have an enriching chemistry, which even after almost 40 years of the first reports is far from being fully investigated.

Author contributions

All authors did the literature research and wrote parts of the manuscript. PWR originated the idea and supervised the work.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors gratefully acknowledge support from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) through the Collaborative Research Centre “4f for Future” (CRC 1573 project number 471424360, project C1).

Notes and references

1. T. J. Kealy and P. L. Pauson, Nature, 1951, 168, 1039–1040.
2. G. B. Kauffman, J. Chem. Educ., 1983, 60, 185.
3. E. O. Fischer and W. Pfab, Z Naturforsch B, 1952, 7, 377–379.
4. H. Werner, Angew. Chem., Int. Ed., 2012, 51, 6052–6058.
5. G. W. Watt and E. W. Gillow, J. Am. Chem. Soc., 1969, 91, 775–776.
6. G. B. Deacon, G. D. Fallon, P. I. MacKinnon, R. H. Newnham, G. N. Pain, T. D. Tuong and D. L. Wilkinson, J. Organomet. Chem., 1984, 277, C21–C24.
7. F. Jaroschik, F. Nief and L. Ricard, Chem. Commun., 2006, 426–428.
8. F. Jaroschik, F. Nief, X.-F. Le Goff and L. Ricard, Organometallics, 2007, 26, 1123–1125.
9. F. Jaroschik, A. Momin, F. Nief, X.-F. Le Goff, G. B. Deacon and P. C. Junk, Angew. Chem., Int. Ed., 2009, 48, 1117–1121.
10. C. A. Gould, K. R. McClain, J. M. Yu, T. J. Groshens, F. Furche, B. G. Harvey and J. R. Long, J. Am. Chem. Soc., 2019, 141, 12967–12973.
11. K. R. McClain, C. A. Gould, D. A. Marchiori, H. Kwon, T. T. Nguyen, K. E. Rosenkoetter, D. Kuzmina, F. Tuna, R. D. Britt, J. R. Long and B. G. Harvey, J. Am. Chem. Soc., 2022, 144, 22193–22201.
12. P. L. Watson, J Chem Soc Chem Comm, 1980, 14, 652–653.
13. T. D. Tilley, R. A. Andersen, B. Spencer, H. Ruben, A. Zalkin and D. H. Templeton, Inorg. Chem., 1980, 19, 2999–3003.
14. W. J. Evans, I. Bloom, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1981, 103, 6507–6508.
15. W. J. Evans, J. W. Grate, H. W. Choi, I. Bloom, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1985, 107, 941–946.
16. M. Schultz, C. D. Sofield, M. D. Walter and R. A. Andersen, New J. Chem, 2005, 29, 919–927.
17. C. A. P. Goodwin, D. Reta, F. Ortu, N. F. Chilton and D. P. Mills, J. Am. Chem. Soc., 2017, 139, 18714–18724.
18. W. J. Evans, T. S. Gummersheimer and J. W. Ziller, Appl. Organomet. Chem., 1995, 9, 437–447.
19. M. Schultz, C. J. Burns, D. J. Schwartz and R. A. Andersen, Organometallics, 2000, 19, 781–789.
20. P. Girard, J. L. Namy and H. B. Kagan, J. Am. Chem. Soc., 1980, 102, 2693–2698.
21. W. J. Evans, T. S. Gummersheimer, T. J. Boyle and J. W. Ziller, Organometallics, 1994, 13, 1281–1284.
22. W. J. Evans, K. J. Forrestal and J. W. Ziller, Polyhedron, 1998, 17, 4015–4021.
23. W. J. Evans, J. M. Perotti, J. C. Brady and J. W. Ziller, J. Am. Chem. Soc., 2003, 125, 5204–5212.
24. H. Nakamura, Y. Nakayama, H. Yasuda, T. Maruo, N. Kanehisa and Y. Kai, Organometallics, 2000, 19, 5392–5399.
25. C. Qian, G. Zou, W. Jiang, Y. Chen, J. Sun and N. Li, Organometallics, 2004, 23, 4980–4986.
26. M. Visseaux, D. Barbier-Baudry, O. Blacque, A. Hafid, P. Richard and F. Weber, New J. Chem, 2000, 24, 939–942.
27. V. K. Bel'sky, Y. K. Gunko, B. M. Bulychev, A. I. Sizov and G. L. Soloveichik, J. Organomet. Chem., 1990, 390, 35–44.
28. S. K. Adas and G. J. Balaich, J. Organomet. Chem., 2018, 857, 200–206.
29. F. Weber, M. Schultz, C. D. Sofield and R. A. Andersen, Organometallics, 2002, 21, 3139–3146.
30. S. Zhou, Z. Wu, L. Zhou, S. Wang, L. Zhang, X. Zhu, Y. Wei, J. Zhai and J. Wu, Inorg. Chem., 2013, 52, 6417–6426.
31. M. Xémard, V. Goudy, A. Braun, M. Tricoire, M. Cordier, L. Ricard, L. Castro, E. Louyriac, C. E. Kefalidis, C. Clavaguéra, L. Maron and G. Nocton, Organometallics, 2017, 36, 4660–4668.
32. E. Ihara, M. Nodono, K. Katsura, Y. Adachi, H. Yasuda, M. Yamagashira, H. Hashimoto, N. Kanehisa and Y. Kai, Organometallics, 1998, 17, 3945–3956.
33. C. A. P. Goodwin, J. Su, L. M. Stevens, F. D. White, N. H. Anderson, J. D. Auxier, T. E. Albrecht-Schönzart, E. R. Batista, S. F. Briscoe, J. N. Cross, W. J. Evans, A. N. Gaiser, A. J. Gaunt, M. R. James, M. T. Janicke, T. F. Jenkins, Z. R. Jones, S. A. Kozimor, B. L. Scott, J. M. Sperling, J. C. Wedal, C. J. Windorff, P. Yang and J. W. Ziller, Nature, 2021, 599, 421–424.
34. H. Sitzmann, T. Dezember, O. Schmitt, F. Weber, G. Wolmershäuser and M. Ruck, Z. Anorg. Allg. Chem., 2000, 626, 2241–2244.
35. W. J. Evans, G. Kociok-Köhn, V. S. Leong and J. W. Ziller, Inorg. Chem., 1992, 31, 3592–3600.
36. A. N. Selikhov, T. V. Mahrova, A. V. Cherkasov, G. K. Fukin, J. Larionova, J. Long and A. A. Trifonov, Organometallics, 2015, 34, 1991–1999.
37. M. N. Bochkarev, I. L. Fedushkin, A. A. Fagin, T. V. Petrovskaya, J. W. Ziller, R. N. R. Broomhall-Dillard and W. J. Evans, Angew. Chem., Int. Ed. Engl., 1997, 36, 133–135.
38. W. J. Evans, N. T. Allen and J. W. Ziller, Angew. Chem., Int. Ed., 2002, 41, 359–361.
39. F. Jaroschik, F. Nief, X.-F. Le Goff and L. Ricard, Organometallics, 2007, 26, 3552–3558.
40. F. Nief, B. T. de Borms, L. Ricard and D. Carmichael, Eur. J. Inorg. Chem., 2005, 637–643.
41. T. Simler, K. N. McCabe, L. Maron and G. Nocton, Chem. Sci, 2022, 13, 7449–7461.
42. W. J. Evans, L. A. Hughes and T. P. Hanusa, J. Am. Chem. Soc., 1984, 106, 4270–4272.
43. W. J. Evans, G. Kociok-Köhn, S. E. Foster, J. W. Ziller and R. J. Doedens, J. Organomet. Chem., 1993, 444, 61–66.
44. W. J. Evans, L. A. Hughes and T. P. Hanusa, Organometallics, 1986, 5, 1285–1291.
45. P. B. Hitchcock, J. A. K. Howard, M. F. Lappert and S. Prashar, J. Organomet. Chem., 1992, 437, 177–189.
46. G. Nocton and L. Ricard, Chem. Commun., 2015, 51, 3578–3581.
47. G. B. Deacon, C. M. Forsyth, F. Jaroschik, P. C. Junk, D. L. Kay, T. Maschmeyer, A. F. Masters, J. Wang and L. D. Field, Organometallics, 2008, 27, 4772–4778.
48. N. Reinfandt and P. W. Roesky, Inorganics, 2022, 10, 25.
49. M. H. Kuiper and H. Lueken, Z. Anorg. Allg. Chem., 2007, 633, 1407–1409.
50. A. L. Wayda, J. L. Dye and R. D. Rogers, Organometallics, 1984, 3, 1605–1610.
51. B. M. Wolf, C. Stuhl, C. Maichle-Mössmer and R. Anwander, Chem. – Eur. J., 2018, 24, 15921–15929.
52. R. P. Kelly, T. D. M. Bell, R. P. Cox, D. P. Daniels, G. B. Deacon, F. Jaroschik, P. C. Junk, X. F. Le Goff, G. Lemercier, A. Martinez, J. Wang and D. Werner, Organometallics, 2015, 34, 5624–5636.
53. G. Lin and W.-T. Wong, J. Organomet. Chem., 1996, 523, 93–98.
54. A. C. G. Shephard, D. P. Daniels, G. B. Deacon, Z. Guo, F. Jaroschik and P. C. Junk, Chem. Commun., 2022, 58, 4344–4347.
55. I. L. Fedushkin, S. Dechert and H. Schumann, Angew. Chem., Int. Ed., 2001, 40, 561–563.
56. J. Tao, S. Qi, L. Yonghua and J. Songchun, J. Organomet. Chem., 1993, 450, 121–124.
57. M. F. Lappert, P. I. W. Yarrow, J. L. Atwood, R. Shakir and J. Holton, J. Chem. Soc., Chem. Commun., 1980, 987–988.
58. E. Sheng, S. Zhou, S. Wang, G. Yang, Y. Wu, Y. Feng, L. Mao and Z. Huang, Eur. J. Inorg. Chem., 2004, 2923–2932.
59. K. Zhang, W. Zhang, S. Wang, E. Sheng, G. Yang, M. Xie, S. Zhou, Y. Feng, L. Mao and Z. Huang, Dalton Trans, 2004, 1029–1037.
60. S. Zhou, S. Wang, E. Sheng, L. Zhang, Z. Yu, X. Xi, G. Chen, W. Luo and Y. Li, Eur. J. Inorg. Chem., 2007, 1519–1528.
61. S. Wang, S. Wang, S. Zhou, G. Yang, W. Luo, N. Hu, Z. Zhou and H.-B. Song, J. Organomet. Chem., 2007, 692, 2099–2106.
62. C. Ruspic, J. R. Moss, M. Schürmann and S. Harder, Angew. Chem., Int. Ed., 2008, 47, 2121–2126.
63. N. J. C. van Velzen and S. Harder, Organometallics, 2018, 37, 2263–2271.
64. S. Harder, D. Naglav, C. Ruspic, C. Wickleder, M. Adlung, W. Hermes, M. Eul, R. Pöttgen, D. B. Rego, F. Poineau, K. R. Czerwinski, R. H. Herber and I. Nowik, Chem. – Eur. J., 2013, 19, 12272–12280.
65. R. A. Andersen, J. M. Boncella, C. J. Burns, R. Blom, A. Haaland and H. V. Volden, J. Organomet. Chem., 1986, 312, C49–C52.
66. R. A. Andersen, R. Blom, C. J. Burns and H. V. Volden, J Chem Soc Chem Comm, 1987, 10, 768–769.
67. J. Marçalo, A. Pires de Matos and W. J. Evans, Organometallics, 1996, 15, 345–349.
68. T. P. Hanusa, Chem. Rev., 1993, 93, 1023–1036.
69. C. Elschenbroich, Organometallchemie, Teubner, Wiesbaden, 2008.
70. R. L. De Kock, M. A. Peterson, L. K. Timmer, E. J. Baerends and P. Vernooijs, Polyhedron, 1990, 9, 1919–1934.
71. T. V. Timofeeva, J. H. Lii and N. L. Allinger, J. Am. Chem. Soc., 1995, 117, 7452–7459.
72. T. Ziegler and A. Rauk, Theor. Chim. Acta, 1977, 46, 1–10.
73. N. Kaltsoyannis and M. R. Russo, J. Nucl. Sci. Technol., 2014, 39, 393–399.
74. J. C. Green, D. Hohl and N. Roesch, Organometallics, 1987, 6, 712–720.
75. S. Labouille, C. Clavaguera and F. Nief, Organometallics, 2013, 32, 1265–1271.
76. A. F. Williams, F. Grandjean, G. J. Long, T. A. Ulibarri and W. J. Evans, Inorg. Chem., 1989, 28, 4584–4588.
77. A. G. Avent, M. A. Edelman, M. F. Lappert and G. A. Lawless, J. Am. Chem. Soc., 1989, 111, 3423–3425.
78. S. P. Nolan, D. Stern and T. J. Marks, J. Am. Chem. Soc., 1989, 111, 7844–7853.
79. S. J. Swamy, J. Loebel, J. Pickardt and H. Schumann, J. Organomet. Chem., 1988, 353, 27–34.
80. W. J. Evans and T. A. Ulibarri, Polyhedron, 1989, 8, 1007–1014.
81. W. J. Evans, J. W. Grate, I. Bloom, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1985, 107, 405–409.
82. D. J. Schwartz and R. A. Andersen, Organometallics, 1995, 14, 4308–4318.
83. Z. Hou, Y. Zhang, H. Tezuka, P. Xie, O. Tardif, T.-A. Koizumi, H. Yamazaki and Y. Wakatsuki, J. Am. Chem. Soc., 2000, 122, 10533–10543.
84. Z. M. Hou, Y. G. Zhang, T. Yoshimura and Y. Wakatsuki, Organometallics, 1997, 16, 2963–2970.
85. M. Nishiura, Z. Hou and Y. Wakatsuki, Organometallics, 2004, 23, 1359–1368.
86. W. J. Evans, G. W. Rabe and J. W. Ziller, J. Organomet. Chem., 1994, 483, 39–45.
87. W. J. Evans, E. Montalvo, S. E. Foster, K. A. Harada and J. W. Ziller, Organometallics, 2007, 26, 2904–2910.
88. T. D. Tilley, R. A. Andersen, B. Spencer and A. Zalkin, Inorg. Chem., 1982, 21, 2647–2649.
89. G. Nocton and L. Ricard, Dalton Trans, 2014, 43, 4380–4387.
90. M. Schultz, J. M. Boncella, D. J. Berg, T. D. Tilley and R. A. Andersen, Organometallics, 2002, 21, 460–472.
91. W. J. Evans and D. K. Drummond, J. Am. Chem. Soc., 1989, 111, 3329–3335.
92. C. H. Booth, D. Kazhdan, E. L. Werkema, M. D. Walter, W. W. Lukens, E. D. Bauer, Y. J. Hu, L. Maron, O. Eisenstein, M. Head-Gordon and R. A. Andersen, J. Am. Chem. Soc., 2010, 132, 17537–17549.
93. G. Nocton, W. W. Lukens, C. H. Booth, S. S. Rozenel, S. A. Medling, L. Maron and R. A. Andersen, J. Am. Chem. Soc., 2014, 136, 8626–8641.
94. J. M. Veauthier, E. J. Schelter, C. N. Carlson, B. L. Scott, R. E. Da Re, J. D. Thompson, J. L. Kiplinger, D. E. Morris and K. D. John, Inorg. Chem., 2008, 47, 5841–5849.
95. C. N. Carlson, C. J. Kuehl, R. E. Da Re, J. M. Veauthier, E. J. Schelter, A. E. Milligan, B. L. Scott, E. D. Bauer, J. D. Thompson, D. E. Morris and K. D. John, J. Am. Chem. Soc., 2006, 128, 7230–7241.
96. F. Reichart, M. Kischel and K. Zeckert, Chem. – Eur. J., 2009, 15, 10018–10020.
97. K. Zeckert, Organometallics, 2013, 32, 1387–1393.
98. K. Zeckert, S. Zahn and B. Kirchner, Chem. Commun., 2010, 46, 2638–2640.
99. K. Zeckert, Dalton Trans, 2012, 41, 14101–14106.
100. V. Goudy, A. Jaoul, M. Cordier, C. Clavaguéra and G. Nocton, J. Am. Chem. Soc., 2017, 139, 10633–10636.
101. D. Wang, J. Moutet, M. Tricoire, M. Cordier and G. Nocton, Inorganics, 2019, 7, 58.
102. M. Tricoire, D. Wang, T. Rajeshkumar, L. Maron, G. Danoun and G. Nocton, JACS Au, 2022, 2, 1881–1888.
103. D. Wang, M. Tricoire, V. Cemortan, J. Moutet and G. Nocton, Inorg. Chem. Front, 2021, 8, 647–657.
104. S. A. J. Boyce, J. Moutet, L. Niederegger, T. Simler, G. Nocton and C. R. Hess, Inorg. Chem., 2021, 60, 403–411.
105. E. J. Schelter, J. M. Veauthier, J. D. Thompson, B. L. Scott, K. D. John, D. E. Morris and J. L. Kiplinger, J. Am. Chem. Soc., 2006, 128, 2198–2199.
106. E. J. Schelter, R. Wu, J. M. Veauthier, E. D. Bauer, C. H. Booth, R. K. Thomson, C. R. Graves, K. D. John, B. L. Scott, J. D. Thompson, D. E. Morris and J. L. Kiplinger, Inorg. Chem., 2010, 49, 1995–2007.
107. M. Schultz, C. J. Burns, D. J. Schwartz and R. A. Andersen, Organometallics, 2001, 20, 5690–5699.
108. P. Selg, H. H. Brintzinger, M. Schultz and R. A. Andersen, Organometallics, 2002, 21, 3100–3107.
109. L. Maron, L. Perrin, O. Eisenstein and R. A. Andersen, J. Am. Chem. Soc., 2002, 124, 5614–5615.
110. F. Weber, H. Sitzmann, M. Schultz, C. D. Sofield and R. A. Andersen, Organometallics, 2002, 21, 3139–3146.
111. A. J. Arduengo, M. Tamm, S. J. Mclain, J. C. Calabrese, F. Davidson and W. J. Marshall, J. Am. Chem. Soc., 1994, 116, 7927–7928.
112. M. Glanz, S. Dechert, H. Schumann, D. Wolff and J. Springer, Z. Anorg. Allg. Chem., 2000, 626, 2467–2477.
113. H. Schumann, M. Glanz, J. Winterfeld, H. Hemling, N. Kuhn and T. Kratz, Angew. Chem., Int. Ed. Engl., 1994, 33, 1733–1734.
114. C. J. Burns and R. A. Andersen, J. Am. Chem. Soc., 1987, 109, 941–942.
115. C. J. Burns and R. A. Andersen, J. Am. Chem. Soc., 1987, 109, 915–917.
116. Z. M. Hou, Y. Zhang, M. Nishiura and Y. Wakatsuki, Organometallics, 2003, 22, 129–135.
117. C. J. Burns and R. A. Andersen, J. Am. Chem. Soc., 1987, 109, 5853–5855.
118. H. Yamamoto, H. Yasuda, K. Yokota, A. Nakamura, Y. Kai and N. Kasai, Chem. Lett., 1988, 1963–1966.
119. W. J. Evans, T. M. Champagne and J. W. Ziller, Organometallics, 2007, 26, 1204–1211.
120. W. J. Evans, J. R. Walensky, F. Furche, A. G. DiPasquale and A. L. Rheingold, Organometallics, 2009, 28, 6073–6078.
121. C. T. Palumbo, J. W. Ziller and W. J. Evans, J. Organomet. Chem., 2018, 867, 142–148.
122. W. J. Evans, J. M. Perotti, J. W. Ziller, D. F. Moser and R. West, Organometallics, 2003, 22, 1160–1163.
123. R. Zitz, H. Arp, J. Hlina, M. Walewska, C. Marschner, T. Szilvási, B. Blom and J. Baumgartner, Inorg. Chem., 2015, 54, 3306–3315.
124. T. D. Tilley, R. A. Andersen and A. Zalkin, Inorg. Chem., 1983, 22, 856–859.
125. A. Jaoul, C. Clavaguéra and G. Nocton, New J. Chem, 2016, 40, 6643–6649.
126. M. T. Gamer, P. W. Roesky, S. N. Konchenko, P. Nava and R. Ahlrichs, Angew. Chem., Int. Ed., 2006, 45, 4447–4451.
127. M. Wiecko and P. W. Roesky, Organometallics, 2007, 26, 4846–4848.
128. L. R. Morss, Chem. Rev., 1976, 76, 827–841.
129. R. G. Finke, S. R. Keenan and P. L. Watson, Organometallics, 1989, 8, 263–277.
130. W. J. Evans, I. Bloom, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1983, 105, 1401–1403.
131. W. J. Evans, D. G. Giarikos, C. B. Robledo, V. S. Leong and J. W. Ziller, Organometallics, 2001, 20, 5648–5652.
132. W. J. Evans, R. A. Keyer and J. W. Ziller, Organometallics, 1993, 12, 2618–2633.
133. W. J. Evans, T. A. Ulibarri, L. R. Chamberlain, J. W. Ziller and D. Alvarez, Organometallics, 1990, 9, 2124–2130.
134. V. Goudy, M. Xémard, S. Karleskind, M. Cordier, C. Alvarez Lamsfus, L. Maron and G. Nocton, Inorganics, 2018, 6, 82.
135. W. J. Evans, R. A. Keyer, H. Zhang and J. L. Atwood, J. Chem. Soc., Chem. Commun., 1987, 837–838.
136. W. J. Evans, R. A. Keyer and J. W. Ziller, Organometallics, 1990, 9, 2628–2631.
137. J. M. Boncella, T. D. Tilley and R. A. Andersen, J. Chem. Soc., Chem. Commun., 1984, 710–712.
138. W. J. Evans, G. W. Rabe and J. W. Ziller, J. Organomet. Chem., 1994, 483, 21–25.
139. W. J. Evans, T. A. Ulibarri and J. W. Ziller, J. Am. Chem. Soc., 1990, 112, 2314–2324.
140. W. J. Evans and T. A. Ulibarri, J. Am. Chem. Soc., 1987, 109, 4292–4297.
141. M. P. Plesniak, X. Just-Baringo, F. Ortu, D. P. Mills and D. J. Procter, Chem. Commun., 2016, 52, 13503–13506.
142. W. J. Evans, S. L. Gonzales and J. W. Ziller, J. Am. Chem. Soc., 1994, 116, 2600–2608.
143. A. Recknagel, M. Noltemeyer and F. T. Edelmann, J. Organomet. Chem., 1991, 410, 53–61.
144. W. J. Evans, S. L. Gonzales and J. W. Ziller, J. Am. Chem. Soc., 1991, 113, 7423–7424.
145. W. J. Evans, K. J. Forrestal, J. T. Leman and J. W. Ziller, Organometallics, 1996, 15, 527–531.
146. R. G. Finke, S. R. Keenan, D. A. Schiraldi and P. L. Watson, Organometallics, 1987, 6, 1356–1358.
147. R. G. Finke, S. R. Keenan, D. A. Schiraldi and P. L. Watson, Organometallics, 1986, 5, 598–601.
148. W. J. Evans, J. W. Grate, K. R. Levan, I. Bloom, T. T. Peterson, R. J. Doedens, H. Zhang and J. L. Atwood, Inorg. Chem., 1986, 25, 3614–3619.
149. P. L. Watson, T. H. Tulip and I. Williams, Organometallics, 1990, 9, 1999–2009.
150. C. J. Burns and R. A. Anderson, J. Chem. Soc., Chem. Commun., 1989, 2, 136–137.
151. H. Schumann, M. R. Keitsch, J. Winterfeld and J. Demtschuk, J. Organomet. Chem., 1996, 525, 279–281.
152. W. J. Evans and D. K. Drummond, Organometallics, 1988, 7, 797–802.
153. A. A. Trifonov, I. D. Gudilenkov, G. K. Fukin, A. V. Cherkasov and J. Larionova, Organometallics, 2009, 28, 3421–3425.
154. W. J. Evans, D. K. Drummond, S. G. Bott and J. L. Atwood, Organometallics, 1986, 5, 2389–2391.
155. W. J. Evans, D. K. Drummond, L. R. Chamberlain, R. J. Doedens, S. G. Bott, H. Zhang and J. L. Atwood, J. Am. Chem. Soc., 1988, 110, 4983–4994.
156. K. G. Wang, E. D. Stevens and S. P. Nolan, Organometallics, 1992, 11, 1011–1013.
157. D. Jin, X. Sun, A. Hinz and P. W. Roesky, Dalton Trans, 2022, 51, 5218–5226.
158. Z. Hou, C. Yoda, T.-A. Koizumi, M. Nishiura, Y. Wakatsuki, S.-I. Fukuzawa and J. Takats, Organometallics, 2003, 22, 3586–3592.
159. W. J. Evans, R. A. Keyer, G. W. Rabe, D. K. Drummond and J. W. Ziller, Organometallics, 1993, 12, 4664–4667.
160. G. Nocton, C. H. Booth, L. Maron and R. A. Andersen, Organometallics, 2012, 32, 1150–1158.
161. A. A. Trifonov, Russ. Chem. Bull., 2003, 52, 601–606.
162. M. D. Walter, D. J. Berg and R. A. Andersen, Organometallics, 2007, 26, 2296–2307.
163. A. A. Trifonov, B. G. Shestakov, K. A. Lyssenko, J. Larionova, G. K. Fukin and A. V. Cherkasov, Organometallics, 2011, 30, 4882–4889.
164. A. A. Trifonov, I. A. Borovkov, E. A. Fedorova, G. K. Fukin, J. Larionova, N. O. Druzhkov and V. K. Cherkasov, Chem. – Eur. J., 2007, 13, 4981–4987.
165. J. A. Moore, A. H. Cowley and J. C. Gordon, Organometallics, 2006, 25, 5207–5209.
166. K. Vasudevan and A. H. Cowley, Chem. Commun., 2007, 3464–3466.
167. G. B. Deacon, E. E. Delbridge, G. D. Fallon, C. Jones, D. E. Hibbs, M. B. Hursthouse, B. W. Skelton and A. H. White, Organometallics, 2000, 19, 1713–1721.
168. C. Schoo, S. V. Klementyeva, M. T. Gamer, S. N. Konchenko and P. W. Roesky, Chem. Commun., 2016, 52, 6654–6657.
169. K. Takaki, T. Kusudo, S. Uebori, Y. Makioka, Y. Taniguchi and Y. Fujiwara, Tetrahedron Lett, 1995, 36, 1505–1508.
170. K. Takaki, T. Kusudo, S. Uebori, T. Nishiyama, T. Kamata, M. Yokoyama, K. Takehira, Y. Makioka and Y. Fujiwara, J. Org. Chem., 1998, 63, 4299–4304.
171. K. Takaki, M. Maruo, T. Kamata, Y. Makioka and Y. Fujiwara, J. Org. Chem., 1996, 61, 8332–8334.
172. W. J. Evans, T. P. Hanusa and K. R. Levan, Inorg. Chim. Acta, 1985, 110, 191–195.
173. D. J. Berg, R. A. Andersen and A. Zalkin, Organometallics, 1988, 7, 1858–1863.
174. G. B. Deacon, C. C. Quitmann, K. Müller-Buschbaum and G. Meyer, Z. Anorg. Allg. Chem., 2001, 627, 1431–1432.
175. N. A. Pushkarevsky, M. A. Ogienko, A. I. Smolentsev, I. N. Novozhilov, A. Witt, M. M. Khusniyarov, V. K. Cherkasov and S. N. Konchenko, Dalton Trans, 2016, 45, 1269–1278.
176. Z. M. Hou, A. Fujita, Y. G. Zhang, T. Miyano, H. Yamazaki and Y. Wakatsuki, J. Am. Chem. Soc., 1998, 120, 754–766.
177. W. J. Evans, D. K. Drummond, L. A. Hughes, H. M. Zhang and J. L. Atwood, Polyhedron, 1988, 7, 1693–1703.
178. W. J. Evans, J. T. Leman, J. W. Ziller and S. I. Khan, Inorg. Chem., 1996, 35, 4283–4291.
179. W. J. Evans, S. L. Gonzales and J. W. Ziller, J. Am. Chem. Soc., 1991, 113, 9880–9882.
180. F. Calderazzo, A. Morvillo, G. Pelizzi and R. Poli, J. Chem. Soc., Chem. Commun., 1983, 507–508.
181. E. V. Dikarev and B. Li, Inorg. Chem., 2004, 43, 3461–3466.
182. K. Y. Monakhov, T. Zessin and G. Linti, Eur. J. Inorg. Chem., 2010, 322–332.
183. R. J. Schwamm, M. Lein, M. P. Coles and C. M. Fitchett, Angew. Chem., Int. Ed., 2016, 55, 14798–14801.
184. M. G. Zhao, T. T. Hao, X. Zhang, J. P. Ma, J. H. Su and W. Zheng, Inorg. Chem., 2017, 56, 12678–12681.
185. W. J. Evans, S. L. Gonzales and J. W. Ziller, J Chem Soc Chem Comm, 1992, 16, 1138.
186. N. A. Pushkarevsky, I. Y. Ilyin, P. A. Petrov, D. G. Samsonenko, M. R. Ryzhikov, P. W. Roesky and S. N. Konchenko, Organometallics, 2017, 36, 1287–1295.
187. F. Nief and L. Ricard, Organometallics, 2001, 20, 3884–3890.
188. A. Recknagel, D. Stalke, H. W. Roesky and F. T. Edelmann, Angew. Chem., Int. Ed. Engl., 1989, 28, 445–446.
189. A. Recknagel, M. Noltemeyer, D. Stalke, U. Pieper, H.-G. Schmidt and F. T. Edelmann, J. Organomet. Chem., 1991, 411, 347–356.
190. W. J. Evans, G. W. Rabe, J. W. Ziller and R. J. Doedens, Inorg. Chem., 1994, 33, 2719–2726.
191. D. J. Berg, C. J. Burns, R. A. Andersen and A. Zalkin, Organometallics, 1989, 8, 1865–1870.
192. S. V. Klementyeva, N. P. Gritsan, M. M. Khusniyarov, A. Witt, A. A. Dmitriev, E. A. Suturina, N. D. Hill, T. L. Roemmele, M. T. Gamer, R. T. Boeré, P. W. Roesky, A. V. Zibarev and S. N. Konchenko, Chem. – Eur. J., 2017, 23, 1278–1290.
193. W. J. Evans, N. T. Allen, M. A. Greci and J. W. Ziller, Organometallics, 2001, 20, 2936–2937.
194. P. B. Hitchcock, M. F. Lappert and S. Prashar, J. Organomet. Chem., 1991, 413, 79–90.
195. W. J. Evans, J. W. Grate, L. A. Hughes, H. Zhang and J. L. Atwood, J. Am. Chem. Soc., 1985, 107, 3728–3730.
196. W. J. Evans, C. A. Seibel and J. W. Ziller, Inorg. Chem., 1998, 37, 770–776.
197. S. V. Klementyeva, M. T. Gamer, A. C. Schmidt, K. Meyer, S. N. Konchenko and P. W. Roesky, Chem. – Eur. J., 2014, 20, 13497–13500.
198. S. V. Klementyeva, N. Arleth, K. Meyer, S. N. Konchenko and P. W. Roesky, New J. Chem, 2015, 39, 7589–7594.
199. N. Arleth, S. Bestgen, M. T. Gamer and P. W. Roesky, J. Am. Chem. Soc., 2014, 136, 14023–14026.
200. M. Rieckhoff, M. Noltemeyer, F. T. Edelmann, I. Haiduc and I. Silaghi-Dumitrescu, J. Organomet. Chem., 1994, 469, C19–C21.
201. W. J. Evans, M. A. Ansari, J. W. Ziller and S. I. Khan, Organometallics, 1995, 14, 3–4.
202. W. J. Evans, T. A. Ulibarri and J. W. Ziller, J. Am. Chem. Soc., 1988, 110, 6877–6879.
203. S. Xu, L. A. Essex, J. Q. Nguyen, P. Farias, J. W. Ziller, W. H. Harman and W. J. Evans, Dalton Trans, 2021, 50, 15000–15002.
204. C. Schoo, S. Bestgen, A. Egeberg, S. Klementyeva, C. Feldmann, S. N. Konchenko and P. W. Roesky, Angew. Chem., Int. Ed., 2018, 57, 5912–5916.
205. C. Schoo, S. Bestgen, A. Egeberg, J. Seibert, S. N. Konchenko, C. Feldmann and P. W. Roesky, Angew. Chem., Int. Ed., 2019, 58, 4386–4389.
206. S. N. Konchenko, N. A. Pushkarevsky, M. T. Gamer, R. Köppe, H. Schnockel and P. W. Roesky, J. Am. Chem. Soc., 2009, 131, 5740–5741.
207. D. Fenske, H. Fleischer and C. Persau, Angew. Chem., Int. Ed. Engl., 1989, 28, 1665–1667.
208. H. A. Spinney, N. A. Piro and C. C. Cummins, J. Am. Chem. Soc., 2009, 131, 16233–16243.
209. N. Tokitoh, Y. Arai, T. Sasamori, R. Okazaki, S. Nagase, H. Uekusa and Y. S. Ohashi, J. Am. Chem. Soc., 1998, 120, 433–434.
210. T. Saito, H. Nishiyama, H. Tanahashi, K. Kawakita, H. Tsurugi and K. Mashima, J. Am. Chem. Soc., 2014, 136, 5161–5170.
211. A. Y. Konokhova, M. Y. Afonin, T. S. Sukhikh and S. N. Konchenko, J. Struct. Chem., 2020, 61, 1244–1252.
212. A. G. Demkin, B. Y. Savkov, T. S. Sukhikh and S. N. Konchenko, J. Struct. Chem., 2021, 62, 957–965.
213. W. J. Evans, G. W. Rabe, M. A. Ansari and J. W. Ziller, Angew. Chem., Int. Ed. Engl., 1994, 33, 2110–2111.
214. A. Zalkin and D. J. Berg, Acta Crystallogr C, 1988, 44, 1488–1489.
215. T. D. Tilley and R. A. Andersen, J. Chem. Soc., Chem. Commun., 1981, 19, 985–986.
216. T. D. Tilley and R. A. Andersen, J. Am. Chem. Soc., 1982, 104, 1772–1774.
217. J. M. Boncella and R. A. Andersen, Inorg. Chem., 1984, 23, 432–437.
218. W. J. Evans, I. Bloom, J. W. Grate, L. A. Hughes, W. E. Hunter and J. L. Atwood, Inorg. Chem., 1985, 24, 4620–4623.
219. R. Yadav, T. Simler, M. T. Gamer, R. Köppe and P. W. Roesky, J. Chem. Soc., Chem. Commun., 2019, 55, 5765–5768.
220. M. R. Churchill, K. N. Amoh and H. J. Wasserman, Inorg. Chem., 1981, 20, 1609–1611.
221. A. Recknagel, A. Steiner, S. Brooker, D. Stalke and F. T. Edelmann, Chem. Ber, 1991, 124, 1373–1375.
222. S. N. Konchenko, T. Sanden, N. A. Pushkarevsky, R. Köppe and P. W. Roesky, Chem. – Eur. J., 2010, 16, 14278–14280.
223. W. J. Evans, L. R. Chamberlain and J. W. Ziller, J. Am. Chem. Soc., 1987, 109, 7209–7211.
224. W. J. Evans, J. T. Leman, R. D. Clark and J. W. Ziller, Main Group Met. Chem., 2000, 23, 163–168.
225. W. J. Evans, L. R. Chamberlain, T. A. Ulibarri and J. W. Ziller, J. Am. Chem. Soc., 1988, 110, 6423–6432.
226. M. Glanz, S. Dechert, H. Schumann, D. Wolff, D. Wolff and J. Springer, Z. Anorg. Allg. Chem., 2000, 626, 2467–2477.
227. M. D. Walter, P. T. Matsunaga, C. J. Burns, L. Maron and R. A. Andersen, Organometallics, 2017, 36, 4564–4578.
228. G. B. Deacon and D. L. Wilkinson, Inorg. Chim. Acta, 1988, 142, 155–159.
229. W. J. Evans, I. Bloom, W. E. Hunter and J. L. Atwood, Organometallics, 1985, 4, 112–119.
230. I. Castillo and T. D. Tilley, J. Am. Chem. Soc., 2001, 123, 10526–10534.
231. M. D. Walter, C. J. Burns, P. T. Matsunaga, M. E. Smith and R. A. Andersen, Organometallics, 2016, 35, 3488–3497.
232. T. Li, M. T. Gamer, M. Scheer, S. N. Konchenko and P. W. Roesky, Chem. Commun., 2013, 49, 2183–2185.
233. T. Li, N. Arleth, M. T. Gamer, R. Köppe, T. Augenstein, F. Dielmann, M. Scheer, S. N. Konchenko and P. W. Roesky, Inorg. Chem., 2013, 52, 14231–14236.
234. N. Arleth, M. T. Gamer, R. Köppe, N. A. Pushkarevsky, S. N. Konchenko, M. Fleischmann, M. Bodensteiner, M. Scheer and P. W. Roesky, Chem. Sci, 2015, 6, 7179–7184.
235. N. Arleth, M. T. Gamer, R. Köppe, S. N. Konchenko, M. Fleischmann, M. Scheer and P. W. Roesky, Angew. Chem., Int. Ed., 2016, 55, 1557–1560.
236. T. Li, N. Arleth, M. T. Gamer, R. Köppe, T. Augenstein, F. Dielmann, M. Scheer, S. N. Konchenko and P. W. Roesky, Inorg. Chem., 2013, 52, 14231–14236.
237. C. Schoo, R. Köppe, M. Piesch, M. T. Gamer, S. N. Konchenko, M. Scheer and P. W. Roesky, Chem. – Eur. J., 2018, 24, 7890–7895.
238. L. R. Maxwell, S. B. Hendricks and V. M. Mosley, J. Chem. Phys., 1935, 3, 699–709.
239. N. Reinfandt, C. Schoo, L. Dütsch, R. Köppe, S. N. Konchenko, M. Scheer and P. W. Roesky, Chem. – Eur. J., 2021, 27, 3974–3978.
240. L. Tuscher, C. Helling, C. Wölper, W. Frank, A. S. Nizovtsev and S. Schulz, Chem. – Eur. J., 2018, 24, 3241–3250.
241. P. J. Sullivan and A. L. Rheingold, Organometallics, 1982, 1, 1547–1549.

---