Reinvestigation of Reactions of HgPh2 with Eu and Yb Metal and the Synthesis of a Europium(II) Bis(tetraphenylborate)

Europium bis(tetraphenylborate) [Eu(thf)7][BPh4]2⋅thf containing a fully solvated [Eu(thf)7]2+ cation, was synthesized by protolysis of “EuPh2” (from Eu and HgPh2) with Et3NHBPh4, and the structure was determined by single-crystal X-ray diffraction. Efforts to characterize the putative “Ph2Ln” (Ln = Eu, Yb) reagents led to the synthesis of a mixed-valence complex, [(thf)3YbII(μ-Ph)3YbIII(Ph)2(thf)]⋅2thf, resulting from the reaction of Yb metal with HgPh2 at a low temperature. This mixed-valence YbII/YbIII compound was studied by 171Yb-NMR spectroscopy and single-crystal X-ray diffraction, and the oxidation states of the Yb atoms were assigned.


Introduction
Rare earth metal compounds containing a carbon-to-metal σ-bond have been of considerable interest for many years [1][2][3][4][5][6][7][8][9][10]. The uniquely high reactivity of this mainly ionic bond not only poses a synthetic challenge from the fundamental research viewpoint but also affords an opportunity for widespread synthetic applications [11][12][13][14][15]. Brönstedt and Lewis basic alkyl or aryl ligands can easily be displaced to afford ionic species [9,16,17], offering access to more reactive systems, in particular for the catalysis of olefin polymerization reactions [18][19][20][21][22][23][24]. In this paper, a weakly coordinating tetraphenylborate BPh 4 anion is used to increase the electropositive character of the central metal and to offer available sites for substrate coordination. In addition, complexes with BPh 4 − offer an interesting reactivity for metalorganic synthesis [25][26][27][28]. In recent years, our group has reported on the synthesis of a number of compounds with BPh 4 − anions [29]. The ionic perfluoroaryllanthanoid(II) tetraphenylborates [Eu(C 6 F 5 )(thf) 6 ]BPh 4 and [Yb(C 6 F 5 )(thf) 5 ]BPh 4 were obtained by treating Yb or Eu metal with HgPh(C 6 F 5 ) in the presence of Me 3 NHBPh 4 [30]. In the course of the present study, we found that the BPh 4 − anion has a stabilizing effect, increasing the thermal robustness of the ionic complexes relative to [Eu(C 6 F 5 ) 2 (thf) 5 ] and [Yb(C 6 F 5 ) 2 (thf) 4 ]. Motivated by these results, we were interested in the possible synthesis of analogous Ln II (Ln = Eu, Yb) complexes with an unsubstituted phenyl ligand.
[Eu(thf) 7 ][BPh 4 ] 2 ·thf (1) crystallized in the monoclinic space group P2 1 /n. Because the coordination chemistry of rare earth metals is primarily governed by ionic interactions and Sm II and Eu II have virtually the same ionic radius [34], the structures are very similar. The inner coordination sphere of the Eu-centered cation resembles a distorted pentagonal bipyramid. This [Eu(thf) 7 ] 2+ core is surrounded by a total of eight BPh 4 anions, forming a cube-like shape around the cation (Figure 1 right) [34].
We were unable to directly characterize the putative "[LnPh 2 (thf) x ]" products from the redox transmetallation reaction (Scheme 1, Equation (2)). Considering the temperature sensitivity of lanthanoid aryls and alkyls and the possibility of ether cleavage reactions [35][36][37][38], the redox transmetallation reaction was repeated at −25 • C. At this temperature, a reaction of HgPh 2 with Eu or Yb was only observed at a concentration of at least~0.5 mol/L. Storage of this slurry mixture at −25 • C for two weeks under daily sonication at r.t. for~1 min. resulted in the formation of large red crystals of the mixed valent species [Yb 2 Ph 5 (thf) 4 ]·2thf (3) (Scheme 2), a complex previously isolated by Bochkarev et al. without THF of crystallization, after synthesis by a reaction of ytterbium naphthalenide with HgPh 2 [39]. Our new facile synthesis route offers a unique opportunity for future selective syntheses of +2/+3 mixed-valence Yb compounds by protolytic ligand exchange reactions of [Yb 2 Ph 5 (thf) 4 ]. No Eu compound similar to 3 or any solely divalent species was isolated from an analogous synthesis. ordination chemistry of rare earth metals is primarily governed by ionic interactions and Sm II and Eu II have virtually the same ionic radius [34], the structures are very similar. The inner coordination sphere of the Eu-centered cation resembles a distorted pentagonal bipyramid. This [Eu(thf)7] 2+ core is surrounded by a total of eight BPh4 -anions, forming a cube-like shape around the cation (Figure 1 right)   We were unable to directly characterize the putative "[LnPh2(thf)x]" p the redox transmetallation reaction (Scheme 1, Equation (2)). Considering th sensitivity of lanthanoid aryls and alkyls and the possibility of ether cleav [35][36][37][38], the redox transmetallation reaction was repeated at -25 °C. At this t reaction of HgPh2 with Eu or Yb was only observed at a concentration o mol/L. Storage of this slurry mixture at -25 °C for two weeks under daily so for ~1 min. resulted in the formation of large red crystals of the mixed [Yb2Ph5(thf)4]2thf (3) (Scheme 2), a complex previously isolated by Boc without THF of crystallization, after synthesis by a reaction of ytterbium n with HgPh2 [39]. Our new facile synthesis route offers a unique opportun selective syntheses of +2/+3 mixed-valence Yb compounds by protolytic lig reactions of [Yb2Ph5(thf)4]. No Eu compound similar to 3 or any solely di was isolated from an analogous synthesis.

Scheme 2. Synthesis of complex 3.
Compound 3 crystallized in the monoclinic space group C2/c wi molecule in the asymmetric unit ( Figure 2), whereas the structure reported was solved and refined in monoclinic space group P21 with two whole mo asymmetric unit (one had two disordered THF ligands) [39]. In the struc coordinate Yb atoms are bridged by three phenyl ligands, each through a s Yb1 has two terminal phenyl ligands and one THF donor, whereas Yb2 h terminal THF ligands. The simplest explanation for the stoichiometry is tha is Yb II and the other is Yb III , and these should be distinguishable based on th difference of 0.15 Å between six-coordinate Yb 2+ and Yb 3+ [34]. However, the complex. Relevant bond lengths are listed in Table 1, together with com reported by Bochkarev for molecule Yb3,4 of [Yb2Ph5(thf)4] [39]. Both Yb1 terminal THF ligands, where Yb1-O1 is 2.400(6) Å, whereas <Yb-O2,3,4> is is a difference much less than that expected if Yb1 is trivalent and Yb2 is corresponding difference for the reported complex is much larger, at 0.14 Å difference is not significant according to the 3 e.s.d. criterion. On the ot terminal Yb1-C1,7 bond lengths of 3 (2.446(8), 2.445 (7)) correspond closely  Scheme 2. Synthesis of complex 3.
Compound 3 crystallized in the monoclinic space group C2/c with one whole molecule in the asymmetric unit (Figure 2), whereas the structure reported by Bochkarev was solved and refined in monoclinic space group P2 1 with two whole molecules in the asymmetric unit (one had two disordered THF ligands) [39]. In the structure, two six-coordinate Yb atoms are bridged by three phenyl ligands, each through a single C atom. Yb1 has two terminal phenyl ligands and one THF donor, whereas Yb2 has only three terminal THF ligands. The simplest explanation for the stoichiometry is that one Yb atom is Yb II and the other is Yb III , and these should be distinguishable based on the ionic radius difference of 0.15 Å between six-coordinate Yb 2+ and Yb 3+ [34]. However, the reality is more complex. Relevant bond lengths are listed in Table 1, together with comparable data reported by Bochkarev for molecule Yb3,4 of [Yb 2 Ph 5 (thf) 4 ] [39]. Both Yb1 and Yb2 have terminal THF ligands, where Yb1-O1 is 2.400(6) Å, whereas <Yb-O2,3,4> is 2.45 Å, which is a difference much less than that expected if Yb1 is trivalent and Yb2 is divalent. The corresponding difference for the reported complex is much larger, at 0.14 Å [39], but the difference is not significant according to the 3 e.s.d. criterion. On the other hand, the terminal Yb1-C1,7 bond lengths of complex [Yb(C 6 F 5 ) 2 (thf) 4 ], Yb-C is considerably longer, at 2.649(3) Å [42]. Accordingly, the Yb-C bond lengths provide evidence that Yb1 is +III; therefore, Yb2 is in the +II oxidation state. The reported data support the view that Yb3 (Table 1) is trivalent, despite the high e.s.d. values [39]. The small difference in the terminal Yb-O bond lengths despite the differing oxidation states proposed above is consistent with the reported data; Yb-O is only marginally shorter in [YbPh 3 bridging phenyl groups lie outside of the plane generated by C(13), C (19), and C(25) by 10, 17, and 9, respectively, and all twist in the same direction towards Yb(2) in a propellorlike fashion. In summary, π and possible C-H-Yb agostic interactions reinforce the bridging C-Yb2 interaction, leading to a closer approach of the bridging arene rings to divalent Yb2. If the C13 Ph group, which has, by far the greatest difference between the Yb1-C and Yb2-C bond lengths, is allocated to Yb1, then the structure would be an association of [YbPh3(thf)] and [YbPh2(thf)3], an outcome consistent with the proposal of Bochkarev that the complex is formed by an association of YbPh3 and YbPh2 [39]. Applying the 3 e.s.d. criterion to the reported [Yb2Ph5(thf)4] structure, the separations of Yb3 and Yb4 from the same bridging C atom [39] are essentially indistinguishable. In the present structural refinement, the average Yb2-C bond distance is 2.64 Å (2.66 Å for Bochkarev's structure [39]), whereas the analogous Yb1-C average is 2.51 Å (2.48 Å for that reported in [39]), and the average Yb-O distances are 2.45 and 2.400(6) Å for Yb2 and Yb1, respectively (2.44 and 2.30 Å for those reported in [39]).   In the case of the bridging phenyl groups, C13, 19, and 25 are bound more closely to Yb1 than Yb2, by 0.176, 0.068, and 0.019 Å, respectively (Table 1, Figure 3), the last difference not being significant; however, overall, these data are consistent with the above oxidation state assignments for Yb1 and Yb2. In addition an ortho carbon (C18, C24, and C30), of the phenyl rings of C13, 19, and 25 approach Yb2 at 2.993(4)-3.130 Å (Table 1), whereas the other ortho carbon (C14, 20, and 26) of the same rings is more distant from Yb1, despite the higher oxidation state of Yb1. The Yb2-ortho-C contacts are in the range for intramolecular Yb II -π-C(Ph) interactions [44][45][46][47][48][49], whereas the Yb1-C distances are too long for credible Yb III -C π interactions [44]. There are also possible agostic interactions with the ortho C-H groups on the phenyl rings of C (18,24,30), as the Yb(2)-H distances ( Table 1) are just within the sum of the van der Waals radius of hydrogen (1.20 Å) [50] and the metallic radius (pseudo-van der Waals radius) of ytterbium (1.94 Å) [51]. The three bridging phenyl groups lie outside of the plane generated by C(13), C (19), and C(25) by 10, 17, and 9 • , respectively, and all twist in the same direction towards Yb(2) in a propellor-like fashion. In summary, π and possible C-H-Yb agostic interactions reinforce the bridging C-Yb2 interaction, leading to a closer approach of the bridging arene rings to divalent Yb2. If the C13 Ph group, which has, by far the greatest difference between the Yb1-C and Yb2-C bond lengths, is allocated to Yb1, then the structure would be an association of [YbPh 3 (thf)] and [YbPh 2 (thf) 3 ], an outcome consistent with the proposal of Bochkarev that the complex is formed by an association of YbPh 3 and YbPh 2 [39]. Applying the 3 e.s.d. criterion to the reported [Yb 2 Ph 5 (thf) 4 ] structure, the separations of Yb3 and Yb4 from the same bridging C atom [39] are essentially indistinguishable. In the present structural refinement, the average Yb2-C bond distance is 2.64 Å (2.66 Å for Bochkarev's structure [39]), whereas the analogous Yb1-C average is 2.51 Å (2.48 Å for that reported in [39]), and the average Yb-O distances are 2.45 and 2.400(6) Å for Yb2 and Yb1, respectively (2.44 and 2.30 Å for those reported in [39]).   The 171 Yb NMR spectrum suggests that complex 3 dissociates into two neutral species of differing oxidation states, namely YbPh 2 and YbPh 3 . As paramagnetic Yb III complexes have not been observed in 171 Yb NMR spectra, it is unlikely that a mixed-valence Yb 2 Ph 5 species would be observed, owing to bridging of Yb II through phenyl groups to Yb III . In the 171 Yb NMR spectrum of the compound (Figure 4), a single resonance is observed at δ = 694 ppm. This value is similar to a resonance (686 ppm) in the 171 Yb NMR spectrum of the reaction mixture from Yb metal and iodobenzene in THF at -78 • C, a system that reacts as "PhYbI(thf) n " [40]. Because the spectrum also has a resonance at 388 ppm attributable to YbI 2 [40,41,52], the resonance at 686 ppm can be assigned to [YbPh 2 (thf) n ], the product with YbI 2 of the Schlenk equilibrium (Scheme 3). This, in turn, leads to the assignment of 694 ppm of 3 to [YbPh 2 (thf) n ], with coproduct YbPh 3 silent in the 171 Yb NMR spectrum, owing to paramagnetism. The chemical shift lies between those of four-coordinate [Yb(2,6-Ph 2 C 6 H 3 ) 2 (thf) 2 ] reported by Niemeyer (927 ppm) [52], and that of six-coordinate [Yb(C 6 F 5 ) 2 (thf) 4 ] [ (463 ppm) [30,42]. Because coordination number is a key determinant of the chemical shift in the 171 Yb NMR spectra within a series of related compounds [53,54] and [YbPh 2 (thf) n ] can be expected to have a coordination number more similar to that of [Yb(C 6 F 5 ) 2 (thf) 4 ] than the sterically crowded [Yb(2,6-Ph 2 C 6 H 3 ) 2 (thf) 2 ], the observed value is probably reasonable for n = 3 (5 coordination), given the differing electron-withdrawing properties of C 6 F 5 and Ph, or even n = 4.

General Considerations
All manipulations were performed with the rigorous exclusion of oxygen and moisture in flame-dried glassware with greaseless stopcocks, either on a dual-manifold Schlenk line or in an argon-filled Saffron glovebox. All solvents were predried over Na wire, stored over Na/K benzophenone ketyl, and distilled prior to use. "Ph2Ln" (Ln = Eu, Yb) and "PhYbI(thf)n" were prepared according to the procedure described in [32,33,40,41]. NMR spectra were obtained using a Bruker DPX 300 MHz spectrometer. 171 Yb-NMR spectra are referenced to 0.15 M [Yb(C5Me5)2(thf)] in THF/10 % C6D6 [53].

Synthesis of [Eu(thf)7][BPh4]2⋅thf (1)
"Ph2Eu" (60 mg) was dissolved in THF (3 mL) in a glass tube. The solution was slowly topped with a solution of Et3NHBPh4 (100 mg, 0.24 mmol) in 7 mL of THF. The solutions were allowed to mix over a period of 3 days. The remaining dark solution was decanted, evaporated and the resulting off-white crystalline solid was washed with 2 mL of THF, affording 71 mg (55 % based on Et3NHBPh4) of compound 1. Single crystals suitable for X-ray diffraction were found in this solid. IR (Nujol, cm

Synthesis of [Yb(thf)6][BPh4]2 (2)
"Ph2Yb" (60 mg) was dissolved in THF (3 mL) in a glass tube. The solution was slowly topped with a solution of Et3NHBPh4 (100 mg, 0.24 mmol) in 7 mL of THF. The solutions were allowed to mix over a period of 3 days. The remaining dark solution was decanted, evaporated and the resulting yellow crystalline solid was washed with 2 mL of THF, [Yb(C6F5)2(thf)4] [ (463 ppm) [30,42]. Because coordination number is a key determinant of the chemical shift in the 171 Yb NMR spectra within a series of related compounds [53,54] and [YbPh2(thf)n] can be expected to have a coordination number more similar to that of [Yb(C6F5)2(thf)4] than the sterically crowded [Yb(2,6-Ph2C6H3)2(thf)2], the observed value is probably reasonable for n = 3 (5 coordination), given the differing electron-withdrawing properties of C6F5 and Ph, or even n = 4.

General Considerations
All manipulations were performed with the rigorous exclusion of oxygen and moisture in flame-dried glassware with greaseless stopcocks, either on a dual-manifold Schlenk line or in an argon-filled Saffron glovebox. All solvents were predried over Na wire, stored over Na/K benzophenone ketyl, and distilled prior to use. "Ph2Ln" (Ln = Eu, Yb) and "PhYbI(thf)n" were prepared according to the procedure described in [32,33,40,41]. NMR spectra were obtained using a Bruker DPX 300 MHz spectrometer. 171 Yb-NMR spectra are referenced to 0.15 M [Yb(C5Me5)2(thf)] in THF/10 % C6D6 [53].

Synthesis of [Eu(thf)7][BPh4]2thf (1)
"Ph2Eu" (60 mg) was dissolved in THF (3 mL) in a glass tube. The solution was slowly topped with a solution of Et3NHBPh4 (100 mg, 0.24 mmol) in 7 mL of THF. The solutions were allowed to mix over a period of 3 days. The remaining dark solution was decanted, evaporated and the resulting off-white crystalline solid was washed with 2 mL of THF, affording 71 mg (55 % based on Et3NHBPh4) of compound 1. Single crystals suitable for X-ray diffraction were found in this solid. IR (Nujol, cm

Synthesis of [Yb(thf)6][BPh4]2 (2)
"Ph2Yb" (60 mg) was dissolved in THF (3 mL) in a glass tube. The solution was slowly topped with a solution of Et3NHBPh4 (100 mg, 0.24 mmol) in 7 mL of THF. The solutions were allowed to mix over a period of 3 days. The remaining dark solution was decanted, evaporated and the resulting yellow crystalline solid was washed with 2 mL of THF, Scheme 3. Schlenk equilibrium in the PhYbI system (coordinated solvent not shown) [41].

General Considerations
All manipulations were performed with the rigorous exclusion of oxygen and moisture in flame-dried glassware with greaseless stopcocks, either on a dual-manifold Schlenk line or in an argon-filled Saffron glovebox. All solvents were predried over Na wire, stored over Na/K benzophenone ketyl, and distilled prior to use. "Ph 2 Ln" (Ln = Eu, Yb) and "PhYbI(thf) n " were prepared according to the procedure described in [32,33,40,41]. NMR spectra were obtained using a Bruker DPX 300 MHz spectrometer. 171 Yb-NMR spectra are referenced to 0.15 M [Yb(C 5 Me 5 ) 2 (thf)] in THF/10 % C 6 D 6 [53].

Synthesis of [Eu(thf) 7 ][BPh 4 ] 2 ·thf (1)
"Ph 2 Eu" (60 mg) was dissolved in THF (3 mL) in a glass tube. The solution was slowly topped with a solution of Et 3 NHBPh 4 (100 mg, 0.24 mmol) in 7 mL of THF. The solutions were allowed to mix over a period of 3 days. The remaining dark solution was decanted, evaporated and the resulting off-white crystalline solid was washed with 2 mL of THF, affording 71 mg (55 % based on Et 3 NHBPh 4 ) of compound 1. Single crystals suitable for X-ray diffraction were found in this solid. IR (Nujol, cm

Synthesis of [Yb(thf) 6 ][BPh 4 ] 2 (2)
"Ph 2 Yb" (60 mg) was dissolved in THF (3 mL) in a glass tube. The solution was slowly topped with a solution of Et 3 NHBPh 4 (100 mg, 0.24 mmol) in 7 mL of THF. The solutions were allowed to mix over a period of 3 days. The remaining dark solution was decanted, evaporated and the resulting yellow crystalline solid was washed with 2 mL of THF, affording compound 2. Single crystals suitable for X-ray diffraction were found in this solid, and 2 was identified by unit cell [31].

Synthesis of [Ph 5 Yb 2 (thf) 4 ]·2thf (3)
Ytterbium metal (260 mg, 1.5 mmol) and HgPh 2 (355 mg, 1.0 mmol) were suspended in 2 mL of THF. Upon sonication at r.t. for 2 h, the mixture developed a red color and was thereafter cooled to −25 • C. Daily sonication for 1 min at r.t. and storage at −25 • C resulted in the formation of mercury metal and a dark red crystalline solid. The remaining solution was decanted, and the residue was treated with 10 mL of THF and decanted from the mercury metal. Concentration of the solution to 5 mL resulted in the formation of large red crystals of 3·(thf) 2 suitable for X-ray diffraction. Isolated yield: 209 mg (41 %). 171 Yb-NMR (thf, 52.6 MHz, −30 • C): δ = 694 ppm.

X-ray Crystallography
Complexes 1 and 3 were measured on a Bruker APEX-II CCD diffractometer equipped with graphite-monochromated Mo-Kα radiation (λ = 0.71073 Å) at 123 K and mounted on a fiber loop in crystallography oil. Absorption corrections were completed using the Apex II program suite with SADABS [55].

Conclusions
In conclusion, we prepared the ionic Eu(II) complex [Eu(thf) 7 ][BPh 4 ] 2 ·thf. In an attempt to characterize the previously described "Ph 2 Ln" compounds, which can be used to prepare [Ln(thf) n ][BPh 4 ] 2 (Ln = Eu, Yb), an improved synthesis of the Yb II /Yb III compound [Yb 2 Ph 5 (thf) 4 ] was developed. The Eu complex 1 (and the Yb complex 2) should be a source of homoleptic complexes via displacement of THF by neutral ligands in non-polar solvents, whereas 3 may yield other mixed-oxidation-state Yb complexes via protolysis reactions, whereby the phenyl groups of 3 are replaced by ligands (L) (e.g., L = OAr, OR, pyrazolate, amidinate, etc.) by reaction with the corresponding proligand (HL).
Author Contributions: The complexes were synthesized and characterized by M.W. and Z.G; the original draft of the manuscript was written by M.W. and subsequently reviewed and edited by Z.G., G.B.D. and P.C.J.; supervision, project administration and funding acquisition, G.B.D. and P.C.J. All authors have read and agreed to the published version of the manuscript.