Coordination chemistry of cyclic disilylated stannylenes and plumbylenes to group 4 metallocenes.

Reduction of group 4 metallocene dichlorides with magnesium in the presence of cyclic disilylated stannylene or plumbylene phosphine adducts yielded the respective metallocene tetrylene phosphine complexes. Under the same conditions the use of the respective dimerized stannylene or plumbylene gave metallocene ditetrylene complexes. A computational analysis of these reactions revealed for all investigated compounds multiple-bonded character for the M-E(II) linkage, which can be rationalized in the case of the monotetrylene complex with the classical σ-donor/π-acceptor interaction. The strength of the M-E(II) bond increases descending group 4 and decreases going from Sn to its heavier congener Pb. The weakness of the Ti-E(II) bonds is caused by the significantly reduced ability of the titanium atom for d-p π-back-bonding.

Introduction

The chemistry of heavier carbene analogues has attracted the attention of both experimentally and theoretically oriented chemists over the last decades.[1,2] The fundamental differences in electronic ground states, reactivities, and structures between carbenes and their heavier counterparts are certainly a major reason for this attraction. Heavy tetrylenes usually possess a singlet ground state with an increasing singlet–triplet gap with higher atomic number.[3] Responsible for this ground-state preference is a progressing reluctance to form hybrid orbitals. The s-electrons of the thus preferred (ns)2(np)2 configuration remain paired. As a consequence of this, dimerization of divalent species is not necessarily a favored process. Most heavier tetrylene dimers do not feature π-bonds, as common for olefins, but rather exist as dimeric donor–acceptor adducts. Attachment of electropositive substituents to the divalent group 14 atoms forces some mixing of their s- and p-orbitals. This way the singlet–triplet gap can significantly be diminished. Sekiguchi’s distannene (Bu2MeSi)2Sn=Sn(SiMeBu2)2, which, despite rather bulky groups on the tin atoms, does not dissociate into monomers in solution, is a good example for this behavior.[4] Descending group 14 further to lead, the reluctance to form dimeric compounds becomes even more pronounced. This is nicely illustrated by the difference between bis[tris(trimethylsilyl)silyl]tin and the analogous lead compound. While both compounds exist as monomers in solution, the stannylene crystallizes as a distannene, while the plumbylene retains its monomeric structure in the solid state.[5]

Recently, we could show that the bidentate tetrakis(trimethylsilyl)tetramethyltetrasilanylene ligand can be employed to stabilize divalent tin and lead compounds.[6,7] These can be isolated either as the respective base adducts (1 and 2) or as dimers (3 and 4). The dimeric stannylene and plumbylene compounds 3 and 4 exhibit considerable structural differences. While the tin compound 3 is the result of a dimerization–rearrangement process of a disilylated stannylene and exists as an endocyclic bicyclic distannene,[6] the plumbylene appears as a monomer in solution but crystallizes as a weak donor–acceptor adduct (4).[7]

In the present contribution we report on an extension of these studies to explore the coordination chemistry of disilylated stannylenes and plumbylenes with early transition metals and in particular with group 4 metallocenes. Only few examples of related compounds are known so far. In a seminal study Piers and co-workers[8,9] could demonstrate complexation of Lappert’s stannylene[10] Sn[CH(SiMe3)2]2 to zirconocene derivatives. More recently reactions of C,N-chelated tin(II) and lead(II) compounds to zirconocene were reported by Ruzicka and co-workers.[11] In the latter case the employed plumbylene decomposed to elemental lead and free ligand during the reaction, thus no compound with a Zr–Pb bond was observed.[11] In this context also the formation of a hafnocene silylene complex, reported recently by Sekiguchi and co-workers, should be noted.[12] That the heavier tetrylenes can display quite different bonding motifs to transition metals has been shown recently by Hahn and co-workers.[13]

Results and Discussion

Synthesis

For the synthesis of group 4 metallocene plumbylene and stannylene complexes a new general approach was sought. In the mentioned examples, the zirconocene stannylene complexes were prepared by warming Negishi zirconocene Cp2ZrCl2–2BuLi[14] in the presence of 2 equiv of stannylene from −80 °C to room temperature, thus limiting the scope of the reaction to zirconium.[8,9,11] Alternatively, we found that reductions of group 4 metallocene dichlorides with magnesium[15,16] in the presence of stannylene or plumbylene phosphine adducts (1 and 2) provided smooth conversion to the desired complexes (5–10) (Scheme 1). Isolated yields after crystallization from pentane of these highly colored compounds were above 80%. The plumbylene complexes of titanocene (8), zirconocene (9), and hafnocene (10) represent the first examples of compounds with group 4 metal–lead bonds. In general it should be noted that the number of known plumbylene transition-metal complexes is quite small.[17]

Scheme 1

Formation of Group 4 Metallocene Stannylene and Plumbylene Complexes

By reaction of 1,2-dipotassiodisilanes and -digermanes with metallocene dichlorides, we recently succeeded in the synthesis of group 4 disilene and digermene complexes.[18] When this reaction, however, was carried out with the respective 1,2-dipotassiodistannane, instead of the expected distannene complexes, metallacyclotetrastannanes were isolated as the only tin containing products.[18] This behavior of tin was attributed to its greater preference for the divalent state compared to its lighter congeneres.[18] As the distannene unit of 3 (Scheme 2) is held together by two bridging tetrasilanylene ligands, we reasoned that the dissociation to stannylenes might be blocked, and thus an equimolar amount of 3 could serve as a precursor for a distannene complex. However, when Cp2HfCl2 was reduced with magnesium in the presence of an equimolar amount of 3, the only pentane soluble product was the hafnocene bis(stannylene) complex 11 (Scheme 2). Its formation is apparently facilitated by the reversible rearrangement of 3 to the exocyclic distannene isomer,[6] which is the direct dimerization product of the monomeric stannylene 12 (Table 2).[6] Coordination of the exocyclic distannene to the hafnocene would then favor the distannylene over the distannene complex as observed previously.[18]

Scheme 2

Formation of Hafnocene and Titanocene Bis(stannylene) Complexes 11 and 13

Table 2

Selected Experimental and Calculated [in parentheses, at M06-2X/SDD (M,E), 6-31G(d) (Si,C,H)] Structural Parameter, WBI, Molecular Orbital Energy Differences ΔE and Experimental Element NMR Chemical Shifts for Tetrylenes 12, 18, Monotetrylene Complexes 5–10, and Bis-tetrylene Complexes of Group 4 Metallocenes 11, 13, 15, 19–21a

compd	M–E	d(M–E) [pm]	d(E–Si) [pm]	α(E,M,P) [°]	α(Si,E,Si) [°]	WBI (ME)	ΔE(dxz/π)b [eV]	ΔE(px/π*)b [eV]	ΔEpara b [eV]	δ ISOE expt	BDE (ME) [kJ mol–1]	BDEB3LYP (ME) [kJ mol–1]	BDENCI (ME) [kJ mol–1]
5	Ti–Sn	269.4 (265.3)	264.7 (267.5)	89.5 (91.1)	99.1 (98.3)	1.23	0.27	1.21	5.20	1635c	151	70	81
6	Zr–Sn	279.4 (278.0)	263.6 (265.9)	90.8 (92.0)	99.3 (98.8)	1.52	0.60	1.42	5.45	1263c	249	162	87
7	Hf–Sn	275.9 (278.3)	262.4 (265.2)	90.5 (92.0)	99.4 (99.1)	1.58	0.66	1.55	5.57	1080c	270	180	90
8	Ti–Pb	272.7 (275.4)	271.1 (274.1)	89.2 (90.6)	97.8 (95.9)	1.07	0.34	0.96	5.04	5299d	118	58	60
9	Zr–Pb	282.1 (284.3)	270.8 (272.4)	90.1 (92.2)	98.0 (97.5)	1.38	0.62	1.25	5.40	4165d	215	145	70
10	Hf–Pb	279.3 (284.2)	269.7 (271.7)	90.1 (92.2)	98.2 (97.2)	1.44	0.67	1.40	5.83	3462d	234	161	73
12	Sn		(267.4)		(91.7)
18	Pb		(275.1)		(90.5)					19516d
13	Ti–Sn	271.4 (269.3)	264.8 (267.1)	89.0 (86.1)	98.4 (97.6)	1.07				2172c	153e	60e	93
19	Zr–Sn	(285.3)	(267.3)	(90.1)	(97.6)	1.19					222e	133e	89
11	Hf–Sn	(286.1)	(267.1)	(90.2)	(97.7)	1.22				1785c	239e	151e	88
20	Ti–Pb	(280.0)	(274.6)	(87.5)	(95.9)	0.93					120e	47e	73
21	Zr–Pb	(291.6)	(273.8)	(89.1)	(96.5)	1.15					193e	116e	77
15	Hf–Pb	(291.6)	(273.5)	(88.7)	(96.6)	1.15					209e	132e	77

BDE of the ME bonds calculated using the M06-2X functional are given as BDE(ME). For comparison, the BDE computed applying the B3LYP functional, here denoted as BDEB3LYP(ME), is summarized as well. Finally, the noncovalent contributions to the BDE, BDENCI(ME), calculated from the difference between BDE(ME) and BDEB3LYP(ME) are listed.

For definition see Figure 5.

119Sn NMR

207Pb NMR.

Mean value computed from the dissociation of the complex into a Cp2M fragment and two tetrylenes.

Complex 11 was isolated in about 60% yield as thin almost black needles. Piers et al. briefly mentioned the possibility of replacing just one stannylene ligand in their zirconocene bis(stannylene) complex with PMe3, but no detailed information on this reaction was provided.[9] Reaction of 11 with 2 equiv of PEt3 in fact led to selective formation of 7 and 1 (Scheme 2). Reaction of 3 with Cp2TiCl2 and magnesium yielded the titanocene bis(stannylene) complex 13 (Scheme 2), but in this case, starting material 3 was contaminated with KN(SiMe3)2 leading in addition to the formation of the trivalent titanium compound Cp2TiN(SiMe3)2 (14). The eventually obtained crystals contained 13 and 14 in a 1:1 ratio. An interesting aspect of these reactions is that attempts to react 3 with excess metallocene dichloride did not lead to monostannylene complexes but selectively to the distannylene compounds.

Under very similar conditions, employing magnesium reduction of hafnocene dichloride in the presence of 2 equiv of 4 even the hafnocene bis(plumbylene) complex 15 could be obtained in moderate yield (Scheme 3). Again, addition of PEt3 to 15 led to the selective formation of plumbylene phosphine adduct 2 and hafnocene complex 10. Due to the very high solubility in pentane of both 11 and 15, they were difficult to isolate, and unfortunately only rather low-quality crystals could be obtained and used for X-ray diffraction experiments.

Scheme 3

Synthesis of Hafnocene Bis(plumbylene) Complex 15

A reaction of Cp2ZrCl2 and magnesium with a substoichiometric amount of 4 (Scheme 4) was carried out to check whether the preference for the formation of ditetrylene complexes extends to lead. The formation of the zirconocene plumbylene complex 16, where the vacant coordination site at zirconium is occupied by a THF molecule, showed this not to be the case.

Scheme 4

Synthesis of zirconocene plumbylene THF complex 16

In order to assess the general applicability of this simple access to group 4 metal tetrylenes, we reacted TaCl5 as a group 5 compound with magnesium in the presence of stannylene adduct 1 (Scheme 5). The successful formation of a stannylene complex of TaCl3, 17, was confirmed by NMR spectroscopy, and a low-quality crystal structure, which revealed a distorted octahedral coordination geometry around tantalum with two trans-oriented PEt3 ligands.

Scheme 5

Synthesis of tantalum stannylene complex 17

NMR Spectroscopy

Distannene 3 exhibits a typical distannene 119Sn NMR shift of 545 ppm, strongly indicating the persistence of the Sn=Sn double bond in solution.[6] Further proof for the retained Sn–Sn bond in solution is the presence of 117Sn satellites in the 119Sn NMR spectrum of 3. The coupling constant of 1J( = 1240 Hz is relatively small compared to a value of 2930 Hz found for Masamune’s tetraaryldistannene,[19] suggesting only weak bonding of donor–acceptor type. By coordination of a phosphine ligand to the stannylene (1), electron octet configuration is achieved. Therefore, chemical shifts far upfield from those expected for the free stannylene 12 are observed. In 1 the 119Sn resonance was found at −224 ppm,[6] consistent with the very recent report by Escudié and co-workers for the NHC adduct of bis[tris(trimethylsilyl)silyl]tin (δ = −197 ppm).[20] The 119Sn NMR signal of 1 is split into a doublet by the adjacent 31P nucleus with a coupling constant of 2220 Hz. The group 4 metallocene stannylene complexes 5–7 display downfield shifted 119Sn resonances compared to 1. The chemical shift of titanocene complex 5 was found to be 1635 ppm with the signal being a doublet with 2JSnP = 276 Hz coupling to phosphorus. The respective zirconocene and hafnocene complexes 6 and 7 exhibit their 119Sn signals at considerably higher field at 1263 ppm (6, 2JPSn = 132 Hz) and 1080 ppm (7, 2JPSn = 92 Hz) with smaller coupling constants. The same trend can also be observed in the 31P spectra of these complexes, as the chemical shift ranges from 50.2 ppm in the titanocene complex (5) down to 38.5 ppm for zirconium (6) to finally 34.8 ppm in the hafnium case (7) (Table 1). NMR spectroscopy thus clearly suggests an increasing degree of π-back-donation when descending group 4.

Table 1

Selected Spectroscopic and Structural Features of Group 4 Metallocene Tetrylene Complexes 5–10

compound	5	6	7	8	9	10
distance (Å) M–E	Ti–Sn 2.69	Zr–Sn 2.79	Hf–Sn 2.76	Ti–Pb 2.73	Zr–Pb 2.82	Hf–Pb 2.79
sum of covalent radii (Å)	2.99/2.76a	3.14/2.94a	3.14/2.92a	3.06/2.80a	3.21/2.98a	3.21/2.96a
range of known M–Sn single bond lengths (Å)b	2.81 – 2.89	2.93 - 3.09	2.91 - 3.06
sum of angles around X (°)	359.6	359.5	359.6	359.6	359.6	359.7
	X = Sn	X = Sn	X = Sn	X = Pb	X = Pb	X = Pb
NMR (ppm)	119Sn: 1635	119Sn: 1263	119Sn: 1080	207Pb: 5299	207Pb: 4165	207Pb: 3462
	31P: 50.2	31P: 38.5	31P: 34.8	31P: 57.5	31P: 46.3	31P: 45.5
2Jx (Hz)	PSn: 276	PSn: 132	PSn: 92

Differences in the published values of tabulated covalent radii[22,23] of elements lead to different sums.

Obtained by searching for M–Sn single bond using CCDC's ConQuest 1.13.

For the hafnium bis(stannylene) complex 11 the 119Sn NMR resonance was found at a much lower field at 1785 ppm. This is in good agreement with the value of 1677 ppm observed by Piers and co-workers for their zirconocene bis(stannylene) complex.[9] The zirconocene bis(stannylene) complex published by Ruzicka and co-workers on the other hand resonates considerably upfield at 923 ppm.[11] However, the different behavior of the latter can be explained by intramolecular donation of electron density into the empty tin p-orbital from the attached amino groups of the parent stannylene. The difference between the 119Sn NMR chemical shifts of 7 and 11 can be rationalized on the basis of π-back-donation. In 7 there is one phosphine and one stannylene ligand present, of which the first can be considered to be mainly a σ-donor. For this reason π-back-donation to tin is enhanced in this system, whereas in 11 the situation can be described as competition of two σ-donor π-acceptor ligands for the same electron pair. Therefore, the electronic environment approaches that of the free stannylene 12 and exhibits the pronounced downfield shift. While it is not completely clear how valid a comparison of the 119Sn NMR shift of the tantalum compound 17 with the group 4 metallocene complexes of the same stannylene is, it seems fair to state that the chemical shift of δ = 1985 ppm is in the same region. The associated downfield shift compared to δ = 1080 ppm found for 7, which contains also a third row transition metal and a phosphine ligand, is likely caused by a combination of several factors. Together with the fact that compound 17 is a 14 electron complex, the electronegative chloride substituents further diminish electron density at the metal. This certainly affects the ability of the tantalum atom to engage in back-donation.

Four signals in the 29Si NMR spectrum of complex 11 indicate the equivalence of both five-membered rings in the complex. However, the two faces of each ring are not equivalent as two signals for the four SiMe3 groups were observed. The same molecular symmetry was also derived from the respective 1H and 13C NMR spectra. This face differentiation suggests hindered rotation around the Zr–Sn bond in compound 11.

The 207Pb NMR spectroscopic results for the plumbylene complexes 8–10 show a similar trend as observed for their stannylene counterparts (5–7). A steady decrease in the chemical shift value descending group 4 was found: titanocene plumbylene complex 8 resonates at δ = 5299 ppm, zirconocene complex 9 at δ = 4165 ppm, and finally hafnocene plumbylene 10 at δ = 3462 ppm. The expected splitting into doublets caused by coupling to the 31P nucleus of the phosphine ligand was not observed. The weaker bonds between Pb and the respective metal, compared to the analogous stannylene complexes, in addition to the fact that the 207Pb signals are comparably broad [values for full width at half-maximum (fwhm) range from 130 to 220 Hz] seem to impede the observation of the 2JPbP coupling. 31P NMR spectroscopic results for compounds 8–10 parallel the behavior observed for their lighter tin congeners. The 31P NMR shifts are δ = 57.6, 46.3, and 45.5 ppm for 8–10, respectively. The observed trend points again to a higher degree of back-bonding between the group 4 metal and Pb in Zr and Hf compounds 9 and 10 compared to titanocene complex 8. The dependence of the M–E back-bonding from the additional phosphine ligands can be estimated from a comparison of the 207Pb NMR chemical shifts of the plumbylene zirconocene 9(δ = 4165 ppm) with the respective THF adduct 16 (δ = 5770). A downfield shift of ca. Δδ 207Pb = 1600 ppm illustrates the superior electron-donating ability of the phosphine compared to THF.

Compared to a 207Pb NMR chemical shift of δ = 3587 ppm reported for [Fe(CO)4]4Pb,[21] the strongly downfield shifted resonances between δ = 3462–5770 ppm for compounds 8–10 and 16 seem to indicate a marked plumbylene character. However, these chemical shifts have to be seen in the context of the free plumbylene, 18, for which an extremely downfield shifted 207Pb resonance at δ = 19516 ppm was found.[7] Compared to this, the plumbylene character of 8–10 and 16 seems to be not so pronounced.

The increasing electronic saturation of the Pb atom also can be observed in the 29Si NMR spectra. In 8 the central silicon atoms resonate at δ = −16 ppm. With stronger π-back-bonding from the d2 transition metal to the plumbylene, this resonance shifts upfield to δ = −42 and −54 ppm for 9 and 10, respectively. Consistent with this argumentation, the comparison of 9 and 16 revealed a downfield shift of the 29Si resonance of the silicon attached to Pb from δ = −42 ppm for 9 to δ = −37 ppm for 16. The 29Si signals for the more remote silicon atoms in 8–10 are found at almost identical positions, suggesting a very similar chemical environment. For 8–10 two different resonances for the trimethylsilyl groups were observed indicating hindered rotation around the metal–Pb bond and also no dissociation of phosphines. In contrast to this, only one SiMe3 signal was observed for 16, which suggests dissociation of THF, which can also be concluded from broadened signals in the respective 1H spectrum. For hafnocene bis(plumbylene) complex 15 only 1H, 13C, and 29Si NMR data could be obtained. The spectra for these nuclei are very similar to those of 11 and consistent with the proposed structure.

X-ray Crystallography

As all six phosphine tetrylene complexes (5–10) are isostructural, only one example of each group is shown [Figure 1: 7 and Figure 2: 10; depictions of the others (5, 6, 8, and 9) are available in the Supporting Information (Figures S1–S4)]. Compounds 5–10 feature planar geometries around the tetrel(II) center and short transition-metal tetrel(II) bond lengths. The sums of bond angles around tin or lead in all six examples match 360° almost exactly with a maximum deviation of 0.5°. The group 4–E (E = Sn or Pb) bond lengths are significantly shorter than the sum of tabulated covalent radii[22,23] or respective single bonds in the case of tin (Table 1). For lead this comparison cannot be made, as compounds containing Ti, Zr, or Hf bonds to Pb have not been reported so far. The X-ray crystallographic study strongly supports the conclusions drawn from the NMR observations with respect to the extent of bond order between the group 4 transition metals and tin or lead. Interesting in this respect is also the P transition-metal–E (E = Sn or Pb) angle of 90° in all compounds that ensures non-disturbance of the transition-metal interaction by the orthogonally coordinated phosphine ligand.

Figure 1

Molecular structure of 7 (thermal ellipsoid plot drawn at the 30% probability level). Hydrogen atoms omitted for clarity (bond lengths in Å, angles in °). Hf(1)–P(1) 2.663(3), Hf(1)–Sn(1) 2.7585(11), P(1)–C(13) 1.826(9), Si(1)–Si(2) 2.356(4), Si(1)–Sn(1) 2.626(2), Si(2)–C(17) 1.889(10), Si(4)–Sn(1) 2.622(3), P(1)–Hf(1)–Sn(1) 90.53(7), Si(4)–Sn(1)–Si(1) 99.41(8), Si(4)–Sn(1)–Hf(1) 129.95(6), Si(1)–Sn(1)–Hf(1) 130.25(6).

Figure 2

Molecular structure of 10 (thermal ellipsoid plot drawn at the 30% probability level). Hydrogen atoms omitted for clarity (bond lengths in Å, angles in °). Hf(1)–P(1) 2.670(5), Hf(1)–Pb(1) 2.7927(12), P(1)–C(11) 1.830(19), Pb(1)–Si(4) 2.690(5), Pb(1)–Si(1) 2.705(4), Si(1)–Si(2) 2.362(7), Si(2)–C(17) 1.889(17), C(15)–C(16) 1.55(3), P(1)–Hf(1)–Pb(1) 90.13(12), Si(4)–Pb(1)–Si(1) 98.24(14), Si(4)–Pb(1)–Hf(1) 130.62(10), Si(1)–Pb(1)–Hf(1) 130.81(10).

The crystals obtained of hafnocene bis(stannylene) 11 and hafnocene bis(plumbylene) 15 were not of high enough quality to permit a detailed structural discussion, therefore no metrical data can be given. But these structures (Figures S5 and S6, Supporting Information) still serve as proof for the atom connectivity and therefore validate the assignments made based on NMR spectroscopy. The crystals containing titanocene bis(stannylene) 13 (Figure 3) together with Cp2TiN(SiMe3)2 (14, Figure 4) in a 1:1 ratio were of much better quality. The bond lengths of Ti–Sn in 13, with 2.71 and 2.72 Å, reflect the diminished degree of back-bonding compared to 5 (2.69 Å). The structure of compound 14 is also interesting as it represents a rare example of a structurally characterized titanocene amide with Ti in the oxidation state +3.[24,25]

Figure 3

Molecular structure of 13 (thermal ellipsoid plot drawn at the 30% probability level). Hydrogen atoms omitted for clarity (bond lengths in Å, angles in °). Ti(1)–Sn(1) 2.7122(13), Ti(1)–Sn(2) 2.7154(14), Sn(1)–Si(4) 2.6399(19), Sn(1)–Si(1) 2.6570(19), Sn(2)–Si(9) 2.639(2), Sn(2)–Si(12) 2.6577(19), Sn(1)–Ti(1)–Sn(2) 88.98(4), Si(4)–Sn(1)–Ti(1) 132.35(5), Si(1)–Sn(1)–Ti(1) 126.06(5), Si(9)–Sn(2)–Si(12) 98.36(6), Si(9)–Sn(2)–Ti(1) 132.67(5), Si(12)–Sn(2)–Ti(1) 126.22(5).

Figure 4

Molecular structure of 14 (thermal ellipsoid plot drawn at the 30% probability level). Hydrogen atoms omitted for clarity (bond lengths in Å, angles in °). Ti(2)–N(1) 2.020(6), N(1)–Si(18) 1.710(6), N(1)–Si(17) 1.722(6), Si(18)–N(1)–Si(17) 122.4(3), Si(18)–N(1)–Ti(2) 117.8(3), Si(17)–N(1)–Ti(2) 119.8(3).

Theoretical Studies.[26]

Optimizations of the molecular structures of the free tetrylenes 12 and 18, the group 4 element mono(tetrylene) complexes 5–10, and the metallocenes bis(tetrylene) complexes 11, 13, 15, and 19–21 at the density functional M06-2X/SDD (Sn, Pb, Ti, Zr, Hf) and 6-31G(d) (P, Si, C, H) level of theory result in structural parameters which are very close to those found by X-ray diffraction methods for some of these compounds. Data which are important for the discussion are summarized in Table 2. For the tetrylenes 12 and 18 for which no experimental structural data are available, half-chair conformations of the metallacyclopentasilane rings were predicted with the heavy group 14 element and the two neighboring silicon atoms spanning the central plane.[27] A common feature of all optimized molecular structures of the mono(tetrylene) (5–10) and of the metallocene bis(tetrylene) complexes 11, 13, 15, and 19–21 are trigonal planar coordinated Sn or Pb atoms (sum of the bond angles α around the element atom, Σα(E) = 358–360°) embedded in a half-chair metallacyclopentasilane ring of local C2 symmetry. The computational results indicate no significant influence of the complexation on the molecular structure of the tetrylene. The most obvious structural modification is a widening of the endocyclic SiESi bond angle α(SiESi) by 5.4–7.4° (see Table 2). In accordance with the available experimental structures, the results of the computation predict that the tetrylene units are oriented mostly perpendicular to the central E–M–P plane in tetrylene complexes 5–10 (dihedral angle β = 82–84°),[28] while in the metallocene bis(tetrylene) complexes 11, 13, 15, and 19–21, the equivalent dihedral angle is somewhat smaller (β = 65–74°).[28] This specific arrangement allows for an efficient back-bonding from metal d-orbitals to the formally empty p-orbital at the tetrel atom. The calculated M–E(II) bond lengths, which are summarized in Table 2, show the expected trends. That is, for a given tetrylene the E–M distances increase along the series E–Ti < E–Zr ∼ E–Hf, and the Sn–M separations are always smaller than the corresponding Pb–M distances. The calculated M–E(II) bond lengths as well as those determined experimentally (see Table 2) are all smaller than standard values for E–M single bonds (Sn–M: 276 pm (Ti), 294 pm (Zr), 292 pm (Hf); Pb–M: 280 pm (Ti), 298 pm (Zr), 296 pm (Hf)),[23] in no case, however, the values predicted for σ2π2 E=M double bonds are reached (Sn=M: 247 pm (Ti), 257 pm (Zr), 258 pm (Hf); Pb=M: 252 pm (Ti), 262 pm (Zr), 263 pm (Hf)).[23] In agreement with these structural criteria also, the results of a natural bond orbital (NBO) analysis indicate the multiple-bond character for the M–E(II) linkage in mono- and bis(tetrylene) complexes 5–10, 11, 13, 15, 19–21 (see Table 2). In detail, all calculated Wiberg bond indices (WBIs) are significantly larger than computed for the respective M–E(IV) single bond in the corresponding metallocene–ditetryl compound (Cp2M(EMe3)2)[E = Sn: 0.83 (Ti), 0.93 (Zr), 0.94 (Hf); E = Pb: 0.79 (Ti), 0.91 (Zr), 0.92 (Hf)]. As it is expected, the WBIs for the M–E(II) bond for stannylene complexes are always larger than computed for the corresponding plumbylene complex, and the calculated bond orders for the M–E(II) bond, as expressed by the WBIs, increase for a given tetrel element in the order Ti < Zr ≤ Hf. In addition the calculated bond order is for each M–E pair larger for the mono(tetrylene) complex than for the metallocene bis(tetrylene). These trends are also reflected by the computed bond dissociation energies for the M–E(II) bond BDE(ME) for the tetrylene complexes (see Table 2). The E–Ti bonds are significantly less stable than the E–Zr bonds [by 97 (5/6) and 98 kJ mol–1 (8/9) in the case of the mono(tetrylene) complexes and by 68 (13/19) and 72 kJ mol–1 (20/21) for the bis(tetrylene) complexes], and there is a second although smaller increase predicted for the BDE of the E–Hf bonds (by 17–21 kJ mol–1). The Pb–M bonds are for all calculated metallocene complexes 29–36 kJ mol–1 weaker than the corresponding Sn–M linkages. The BDE values for the Ti–E bonds in mono(tetrylene) (5, 8) and bis(tetrylene) complexes 13, 20 are very similar (see Table 2). The situation differs, however, for the hafnocene and zirconocene complexes for which the computed BDEs of the M–E(II) bond are smaller in bis(tetrylene) complexes than in their mono(tetrylene) counterparts by 18–31 kJ mol–1.

Figure 5

FMO interaction scheme for monotetrylene complex 6, derived from M06-2X/SDD (Zr, Sn), 6-31G(d) (Si, P, C, H) calculations. This MO scheme is qualitatively valid for all investigated monotetrylene complexes 5–10.

The bonding between the zirconium and the tin atoms in the mono(tetrylene) complex 6 is rationalized by the orbital interaction diagram shown in Figure 5. This orbital interaction diagram is also valid qualitatively for all investigated metallocene mono(tetrylene) complexes 5–10. Plots of the surface diagrams for frontier molecular orbitals (FMOs) of compound 6 can be found in the Supporting Information. The M–E(II) bond in complexes 5–10 is best described by the conventional σ-bonding/(d/p) π-back-bonding scheme for carbene complexes. In the framework of perturbation theory, the relative extent of back-bonding in the metallocene tetrylene complexes can be estimated by the evaluation of the calculated orbital stabilization energy ΔE(d/π) and the corresponding destabilization energy ΔE(π*/p) (see Figure 5 and Table 2). Both energy differences increase for both kinds of metallocene mono(tetrylene) complexes along the series Ti < Zr < Hf. This suggests that the d → p π-back-bonding is smallest for the titanium complexes (5, 8) and largest for the hafnium compounds (7, 10).

The analysis also indicates that for each group 4 metallocene, the lowering of the π-orbital as expressed by ΔE(d/π) is slightly smaller in the stannylene complexes (5–7) than in the corresponding plumbylene complexes (8–10, by 0.07 eV (Ti), 0.02 eV (Zr), 0.01 eV (Hf)). On the other hand, the effect of the tetrylene on the π* level is more significant, as the destabilization energy ΔE(π*/p) is markedly larger for the stannylene complexes (5–7) than for the corresponding plumbylene complexes (8–10, by 0.25 eV (Ti), 0.17 eV (Zr), 0.15 eV (Hf)).

The poor ability of the titanocene to engage in π-bonding is mostly due to the poor spatial and energetic match between the 3d orbital of titanium and the 5p orbital of the tin (ΔE(d/p) = 2.65 vs 2.08 eV for Zr or 1.88 eV for Hf) or 6p of lead atom ((ΔE(d/p) = 2.80 vs 2.22 eV for Zr or 2.03 eV for Hf).

NMR chemical shifts of group 14 carbene analogs are always extremely large due to a dominant paramagnetic contribution, which arises from the efficient interaction of the applied magnetic field with the filled sp-type MO and the orthogonal empty p-type orbital at the dicoordinated tetrel element.[29] The paramagnetic shift is very large for small energy differences ΔEpara between these two magnetically active orbitals.[29] In the mono(tetrylene) complexes the sp orbital is transformed to the σ-orbital of the M–E(II) bond, and the p orbital can be associated with the π*-orbital (Figure 5). Therefore, the 119Sn NMR chemical shift of the stannylene complexes 5–7 and the 207Pb NMR chemical shift of the metallocene plumbylene complexes 8–10 are determined mainly by the energy difference between these two molecular orbitals, ΔEpara (see Figure 5). While the energy of the σ-orbital remains nearly constant for a given tetrel element along the series of group 4 metals (E(σ) = −6.09 eV (5), −6.13 eV (6), −6.12 eV (7)), the π*-level is significantly altered by the different extent of back-bonding (E(π*) = −0.90 eV (5), −0.68 eV (6), −0.55 eV (7)).[30] Consequently, the changes in ΔEpara, summarized in Table 2, are mostly due to the different extent of back-bonding in that sense as the larger ΔEpara, the higher the multiple-bond character of the M–E(II) bond. Therefore the experimental 119Sn NMR chemical shift of stannylene compounds 5–7 and the experimental 207Pb NMR shift of the metallocene plumbylene complexes 8–10 can be used as a tool to estimate the degree of multiple bonding in these complexes. Increasing multiple-bond character of the M–E(II) bond leads to increasing energy differences ΔEpara, which becomes manifest in a upfield shift of the element resonance in NMR spectroscopy. These relations become obvious by analyzing the data summarized in Table 2, and although only a very limited set of data is used, by the correlations between the reciprocal calculated ΔEpara and the experimental chemical shifts δ119Sn (5–7) and δ207Pb (8–10) shown in Figure 6.

Figure 6

Plot of the experimental δ 119Sn (●) and δ 207Pb NMR (▲) chemical shifts versus the reciprocal ΔEpara for monostannylene complexes 5–7 and monoplumbylene metallocenes 8–10.

The bonding in bis(stannylene) zirconocenes was treated principally already by Piers.[9] In agreement with that earlier investigation we found that multicenter interactions are important for the understanding of the bonding in bis(tetrylene) complexes 11, 13, 15, and 19–21. The FMO interaction diagram of the bis(stannylene) zirconocene 19 is shown in Figure 7. Qualitatively, it is also valid for the investigated bis(tetrylene) complexes 11, 13, 15, 20, and 21. The analysis reveals that the ME(II)2 group is built up by two σ-type orbitals and the π-type HOMO, each delocalized across all three constituent atoms (see Figures 7 and 8 for surface diagrams of the respective MOs). The delocalization of the π-type HOMO across all three atoms helps to rationalize the smaller bond order and the weaker M–E(II) bond in group 4 metallocene bis(tetrylene) complexes compared to their mono(tetrylene) counterparts. In addition, inspection of the HOMO of compound 19 indicates some degree of bonding interaction between the two distant Sn atoms, although their separation, d(SnSn), approaches the sum of the van der Waals radii, ΣvdWR (d(SnSn) 403.8 pm, ΣvdWR 434 pm).[31] The computed WBI index between this pair of atoms differs significantly from zero; it is however only 22% of the bond index computed at the same theoretical level for the central Sn–Sn single bond in (H3Si)6Sn2 (19: WBI(SnSn) = 0.20; (H3Si)6Sn2: WBI(SnSn) = 0.92, d(SnSn) = 283.7 pm). Similar small WBIs were computed for the bis-(tetrylene) complexes 11, 13, 15, 20, and 21 (WBIs range from 0.15 (PbPb in complex 20) to 0.23 (SnSn in compound 11)). These computational data suggest that in the continuum of possible bonding modes for group 4 metallocene bis(tetrylene) complexes, beginning with the ditetren complex A, passing the metallacyclopropane B and ending at the delocalized bis(tetrylene) structure C, the here investigated complexes 11, 13, 15 and , 19–21 are best described by canonical structure C with only minor contribution from structure B (Scheme 6).

Figure 7

FMO interaction scheme for bis(tetrylene) complex 19 in C2 symmetry, derived from M06-2X/SDD(Zr,Sn), 6-31G(d) (Si,C,H) calculations. The FMO diagram for Cp2Zr is derived from that of Cp2Zr(PEt3) (see Figure 5) by removal of the phosphane ligand (see the Supporting Information for further details). This MO scheme is qualitatively valid for all investigated bis-tetrylene complexes 11, 13, 15, and 19–21.

Figure 8

Calculated surface diagrams for pertinent molecular orbitals of bis(stannylene) complex 19 in C2 symmetry, derived from M06-2X/SDD(Zr,Sn), 6-31G(d) (Si,C,H) calculations (isodensity value: 0.05). (a) 1a orbital (σ-EME bonding (HOMO-4)); (b) 1b orbital (σ EME bonding (HOMO-1)); (c) 2a orbital (π EME bonding (HOMO)); (d) 2b orbital (π EME nonbonding (LUMO)) (Color code: light blue: Zr, greenish gray: Sn; blue gray: Si; and gray: carbon).

Scheme 6

Possible Structural Arrangements Group 4 Metallocene Bis(tetrylene) Complexes (M = Ti–Hf; E = Sn, Pb).

The large and polarizable substituents which are present in all investigated metallocene complexes suggest that attractive dispersion energy contributions to the overall binding energy of the complexes might be a decisive factor. The here applied M06-2X functional[32] properly accounts for dispersion forces, while the most prominent deficit of the popular B3LYP functional is the nearly complete negligence of noncovalent van der Waals interactions. Therefore, the difference in the calculated bond dissociation energies (BDEs) using these two functionals allows estimating the contribution of noncovalent bonding in metallocene complexes 5–11, 13, 15, and 19–21.[33−35] The contribution of noncovalent interactions, BDENCI, to the overall BDE, which is calculated as the difference between the BDE obtained at the M06-2X level and the reduced BDEB3LYP obtained at the B3LYP level (see Table 2), is substantial in all cases. In the case of the titanium complexes it accounts for 60% of the overall BDE and even in the hafnium complexes it amounts to 30%.

Conclusion

The present study demonstrates that stannylene and plumbylene complexes of all three group 4 metallocenes are synthetically accessible by magnesium reduction of the corresponding metallocene dichlorides in the presence of the phosphine-stabilized tetrylenes 1 or 2. Significant π back-bonding from the transition metal to the heavy main group atom was shown by NMR spectroscopy and confirmed by X-ray structure analyses of all six complexes 5–10. The stannylene complexes 5 and 7 are the first stannylene complexes of titanium and hafnium to be reported, whereas 8–10 are the first compounds to feature group 4–lead bonds at all. Using the base-free compounds 3 and 4 hafnocene bis(tetrylene) complexes, 11 and 15, could be prepared. In these complexes the extent of π-back-bonding is decreased because of competition of two π-acceptor ligands for only one electron pair.

The theoretical analysis of the bonding in metallocene mono- and bis(tetrylene) complexes 5–10, 11, 13, 15, and 19–21 revealed for all investigated compounds multiple-bonded character for the M–E(II) linkage in agreement with the interpretation of the experimental data. The bonding between the group 4 metal and the group 14 element atom can be rationalized in the case of the mono(tetrylene) complexes with the classical σ-donor−π-acceptor interaction. The strength of the M–E(II) linkage increases descending the group 4 metals and decreases going from Sn to its heavier congener Pb. As a consequence, the weakest M–E(II) bonds are found between Ti and Pb atoms, while the strongest are predicted by the computation for the pair Hf–Sn. The reason for the weakness of the Ti–E(II) bonds is the significantly reduced ability of the titanium atom for d–p π-back-bonding. The theoretical analysis of the bonding in mono(tetrylene) complexes 5–10 supports the presumption based on experimental data that the NMR chemical shift of the tetrel atom in the complex is a qualitative measure for the extent of multiple-bonding between the metal and the tetrel atom in these complexes. Increased d–p back-bonding results in significant shielding of the tetrel atom. The calculated WBIs and BDEs indicate that the individual M–E(II) bond is weaker in the bis(tetrylene) zirconocene and hafnocene complexes 11, 15, 19, and 21 than in the corresponding mono(tetrylene) complexes 6, 7, 9, and 10. Clearly, this is a result of the reduced multiple-bond character of the M–E(II) linkage, due to the competition between two acceptor π-type orbitals for one metal d-orbital in the metallocene bis(tetrylene) complexes. This situation is best described by a multicenter bonding which involves the two tetrel atoms and the central metal atom. The degree of Ti–E π-back-bonding in the titanium complexes 5 and 8 is already small and is not further reduced by the addition of a second tetrylene unit in complexes 13 and 20. Consequently, similar BDEs for the Ti–E(II) bonds in compounds 5, 8, 13, and 20 are predicted by the calculations.

Experimental Section

General Remarks

All reactions involving air-sensitive compounds were carried out under an atmosphere of dry nitrogen or argon using either Schlenk techniques or a glovebox. All solvents were dried using column-based solvent purification system.[36] Chemicals were obtained from different suppliers and used without further purification.

1H (300 MHz), 13C (75.4 MHz), 29Si (59.3 MHz), 31P (124.4 MHz), 119Sn (111.8 MHz), and 207Pb (62.8 MHz) NMR spectra were recorded on a Varian INOVA 300 spectrometer. If not noted otherwise for all samples, C6D6 was used as solvent. To compensate for the low isotopic abundance of 29Si, the INEPT pulse sequence[37,38] was used for the amplification of the signal. Elemental analyses were carried out using a Heraeus VARIO ELEMENTAR instrument. For the plumbylene complexes, attempts to obtain elemental analysis data gave consistently too low values for C and H.

X-ray Structure Determination

For X-ray structure analyses the crystals were mounted onto the tip of glass fibers, and data collection was performed with a BRUKER-AXS SMART APEX CCD diffractometer using graphite-monochromated Mo Kα radiation (0.71073 Å). The data were reduced to Fo2 and corrected for absorption effects with SAINT[39] and SADABS,[40,41] respectively. The structures were solved by direct methods and refined by full-matrix least-squares method (SHELXL97).[42] If not noted otherwise, all non-hydrogen atoms were refined with anisotropic displacement parameter. All hydrogen atoms were located in calculated positions to correspond to standard bond lengths and angles. All diagrams were drawn with 30% probability thermal ellipsoids, and all hydrogen atoms were omitted for clarity. Unfortunately the obtained crystal quality of some substances was poor. This fact is reflected by quite high R and low θ values.

Crystallographic data (excluding structure factors) for the structures of compounds 5–11 and 13–15 reported in this paper have been deposited with the Cambridge Crystallographic Data Center as supplementary publication no. CCDC-831742 (5), 831743 (6), 831744 (7), 831751 (8), 831748 (9), 831745 (10), 855936 (11), 855938 (13 and 14), and 855937 (15). Copies of data can be obtained free of charge at: http://www.ccdc.cam.ac.uk/products/csd/request/.

Compounds 1,[6]2,[7]3,[6] and 4(7) were prepared according to published procedures.

General Procedure

For group 4 metallocene monotetrylene complexes 5–10: An equimolar (0.5 mmol each) mixture of group 4 metallocene dichloride, magnesium turnings, and 1 or 2 was stirred in THF (5 mL) for 3 h. During this time deeply colored solutions formed. The THF was removed under reduced pressure, and the remaining solid was extracted with pentane (3×, 5 mL each). The filtrate was concentrated to 6 mL and stored at −60 °C for 16 h. Crystals could be isolated by decantation. Typical yield: about 80%.

Titanocene Stannylene Phosphine Complex (5)

Dark-blue crystals (yield: 81%): 1H NMR (δ in ppm): 5.24 (s, 10H, Cp), 1.05 (m, 9H, P(CH2CH3)3), 0.88 (m, 6H, P(CH2CH3)3), 0.59 (s, 18H, SiMe3), 0.54 (s, 18H, SiMe3), 0.48 (s, 6H, SiMe2), 0.43 (s, 6H, SiMe2). 13C NMR (δ in ppm): 96.3 (Cp), 22.7 (P(CH2CH3)3), 9.4 (P(CH2CH3)3), 5.6 (SiMe3), 5.4 (SiMe3), 0.3 (SiMe2), 0.1 (SiMe2). 29Si NMR (δ in ppm): −4.1 (SiMe3), −5.0 (SiMe3), −20.1 (SiMe2), −101.4 (quart. Si). 31P NMR (δ in ppm): 50.2 (2J117Sn/119Sn= 232 Hz, 266 Hz). 119Sn NMR(δ in ppm): 1635 (d, 1JSnP = 266 Hz). Anal. calcd for C32H73PSi8SnTi (880.16): C, 43.67; H, 8.36. Found: C, 43.22; H, 8.09.

Zirconocene Stannylene Phosphine Complex (6)

Deep-purple crystals (yield: 85%): 1H NMR (δ in ppm): 5.41 (s, 10H, Cp), 1.05 (m, 9H, P(CH2CH3)3), 0.86 (m, 6H, P(CH2CH3)3), 0.60 (s, 18H, SiMe3), 0.50 (s, 18H, SiMe3), 0.49 (s, 6H, SiMe2), 0.36 (s, 6H, SiMe2). 13C NMR (δ in ppm): 96.7 (Cp), 20.2 (d, 2JPC = 15.8 Hz, P(CH2CH3)3), 8.4 (P(CH2CH3)3), 5.1 (SiMe3), 4.7 (SiMe3), 3.4 (SiMe2). 29Si NMR (δ in ppm): −5.0 (d, 4JPSi = 2.1 Hz, SiMe3), −5.9 (d, 4JPSi = 3.2 Hz, SiMe3), −19.4 (SiMe2), −108.2 (d, 3JPSi = 1.6 Hz, quart. Si). 31P NMR (δ in ppm): 37.5(1JPSn = 133 Hz). 119Sn NMR (δ in ppm): 1263 (d, 1JPSn = 133 Hz). Anal. calcd for C32H73PSi8SnZr (923.51): C, 41.62; H, 7.97. Found: C, 40.96; H, 8.05.

Hafnocene Stannylene Phosphine Complex (7)

Deep-purple crystals (yield: 79%): 1H NMR (δ in ppm): 5.29 (s, 10H, Cp), 1.13 (td, 3JHH = 7.6 Hz, 3JPH = 13.0 Hz, 9H, P(CH2CH3)3), 0.88 (m, 6H, P(CH2CH3)3)3, 0.62 (s, 18H, SiMe3), 0.60 (s, 6H, SiMe2), 0.52 (s, 6H, SiMe2), 0.51 (s, 18H, SiMe3). 13C NMR (δ in ppm): 95.5 (Cp), 22.1 (d, 2JPC = 18.9 Hz, P(CH2CH3)3), 9.3 (P(CH2CH3)3), 5.5 (SiMe3), 5.2 (SiMe3), −0.3 (SiMe2), −0.4 (SiMe2). 29Si NMR (δ in ppm): −4.4 (d, 4JPSi = 2.6 Hz, SiMe3), −5.9 (d, 4JPSi = 3.6 Hz, SiMe3), −19.0 (SiMe2), −110.8 (d, 3JPSi = 3.1 Hz, quart. Si). 31P NMR (δ in ppm): 34.8 (no tin satellites observed). 119Sn NMR (δ in ppm): 1079 (d, 2JSnP = 92 Hz). Anal. calcd for C32H73HfPSi8Sn (1010.78): C, 38.02; H, 7.28. Found: C, 37.73; H, 7.25.

Titanocene Plumbylene Phosphine Complex (8)

Dark-green crystals (yield: 80%): 1H NMR (δ in ppm): 5.2 (s, 10H, Cp), 1.08 (m, 9H, P(CH2CH3)3), 0.85 (m, 6H, P(CH2CH3)3), 0.62 (s, 18H, SiMe3), 0.58 (s, 18H, SiMe3), 0.52 (s, 6H, SiMe2), 0.37 (s, 6H, SiMe2). 13C NMR (δ in ppm): 96.2 (Cp), 23.0 (d, 2JPC = 5.4 Hz, P(CH2CH3)3), 6.7 (SiMe3), 6.6 (SiMe3), 4.0 (SiMe2), 3.2 (SiMe2). 29Si NMR (δ in ppm): −2.3 (d, 4JPSi = 2.7 Hz, SiMe3), −3.7 (d, 4JPSi = 3.8 Hz, SiMe3), −6.9 (SiMe2), −15.6 (d, 3JPSi = 2.5 Hz, quart. Si). 31P NMR (δ in ppm): 57.5. 207Pb NMR (δ in ppm): 5299 (fwhm: 217.8 Hz).

Zirconocene Plumbylene Phosphine Complex (9)

Deep-purple crystals (yield: 77%): 1H NMR (δ in ppm): 5.36 (s, 10H, Cp), 1.08 (m, 9H, P(CH2CH3)3), 0.85 (m, 6H, P(CH2CH3)3), 0.65 (s, 18H, SiMe3), 0.62 (s, 6H, SiMe2), 0.55 (s, 18H, SiMe3), 0.53 (s, 6H, SiMe2). 13C NMR (δ in ppm): 96.5 (Cp), 19.9 (d, 2JPC = 15.1 Hz, P(CH2CH3)3), 8.6 (P(CH2CH3)3), 6.3 (SiMe3), 6.2 (SiMe3), 3.2 (SiMe2), 2.4 (SiMe2). 29Si NMR (δ in ppm): −3.7 (d, 4JPSi = 2.0 Hz, SiMe3), −4.1 (d, 4JPSi = 3.1 Hz, SiMe3), −7.1 (SiMe2), −42.0 (d, 3JPSi = 2.4 Hz, quart. Si). 31P NMR (δ in ppm): 46.3. 207Pb NMR (δ in ppm): 4165 (fwhm: 128.5 Hz).

Hafnocene Plumbylene Phosphine Complex (10)

Deep-purple crystals (yield: 84%): 1H NMR (δ in ppm): 5.24 (s, 10H, Cp), 1.15 (td, 3JHH = 7.7 Hz, 3JPH = 13.1 Hz, 9H, P(CH2CH3)3), 0.87 (m, 6H, P(CH2CH3)3), 0.67 (s, 18H, SiMe3), 0.63 (s, 6H, SiMe3), 0.62 (s, 6H, SiMe3), 0.54 (s, 18H, SiMe3). 13C NMR (δ in ppm): 94.6 (Cp), 21.3 (d, 2JPC = 20.2 Hz, P(CH2CH3)3), 9.2 (P(CH2CH3)3), 6.2 (SiMe3), 6.1 (SiMe3), 2.9 (SiMe2), 2.2 (SiMe2). 29Si NMR (δ in ppm): −2.5 (d, 4JPSi = 2.2 Hz, SiMe3), −4.3 (d, 4JPsi = 3.2 Hz, SiMe3), −7.7 (SiMe2), −53.9 (d, 3JPsi = 3.7 Hz, quart. Si). 31P NMR (δ in ppm): 45.5 (no 207Pb satellites observed). 207Pb NMR (δ in ppm): 3462 (fwhm: 172.1 Hz).

Hafnocene Bis(Stannylene) Complex (11)

A mixture of hafnocene dichloride (20 mg, 0.05 mmol), magnesium turnings (5 mg, 4.8 mmol, excess), and 3 (61 mg, 0.05 mmol) was suspended in THF (5 mL), subjected to ultrasonification, and stirred for 1 h. A black suspension developed. All volatile materials were removed in vacuo, and the residue was extracted with pentane (10 mL). The dark filtrate was concentrated to 3 mL and stored at −60 °C for 60 h. Black needle-shaped crystals of 11 (40 mg, 0.03 mmol, 54%) were isolated upon filtration and dried in vacuo. 1H NMR (δ in ppm): 4.85 (s, 10H, Cp), 0.47 (s, 36H, SiMe3), 0.46 (s, 36H, SiMe3), 0.29 (s, 12H, SiMe2), 0.25 (s, 12H, SiMe2). 13C NMR (δ in ppm): 96.3 (Cp), 13.9 (SiMe2), 13.8 (SiMe2), 4.9 (SiMe3), 4.8 (SiMe3). 29Si NMR (δ in ppm): −6.8 (SiMe3), −10.8 (SiMe3), −20.1 (SiMe2), −85.3 (quart. Si). 119Sn NMR (δ in ppm): 1785. Anal. calcd for C42H106HfSi16Sn2 (1476.57): C, 34.16; H, 7.24. Found: C, 33.71; H, 7.29.

Titanocene Bis(stannylene) Complex (13)

A mixture of titanocene dichloride (25 mg, 0.10 mmol), magnesium turnings (6 mg, 0.25 mmol), and distannene 3 (117 mg, 0.10 mmol) was suspended in THF (4 mL), subjected to ultrasonification for 2 min, and stirred for 1 h. A deep-purple suspension developed. All volatile materials were removed in vacuo, and the blue residue was extracted 3× with pentane (2 mL each). The dark-blue filtrate was concentrated to 2 mL and stored at −60 °C for 16 h. Blue crystals of 13 (87 mg (0.07 mmol, 70%) were isolated by decantation and dried in vacuo. 1H NMR (δ in ppm): 5.36 (s, 10H, Cp), 0.46 (s, 72H, SiMe3), 0.34 (s, 24H, SiMe2). 13C NMR (δ in ppm): 94.6 (Cp), 4.7 (SiMe3), −0.4 (SiMe2). 29Si NMR (δ in ppm): −1.7 (SiMe3), −19.4 (SiMe2), −79.0 (quart. Si). 119Sn NMR (δ in ppm): 2172.

The crystals containing Cp2TiN(SiMe3)2 were presumably formed by reduction of Cp2TiCl2 with magnesium in the presence of a batch of 3 which was contaminated with KN(SiMe3)2 from the synthesis of 3. The mixed crystals are green from the Ti(III)-compound but were suitable for X-ray crystallography, whereas from pure 13, no crystals of good quality could be obtained.

Hafnocene Bis(Plumbylene) Complex (15)

A mixture of hafnocene dichloride (20 mg, 0.05 mmol), magnesium turnings (4 mg, 0.16 mmol), and 4 (67 mg, 0.05 mmol) was suspended in THF (5 mL), subjected to ultrasonification for 2 min, and stirred for 1 h. The suspension developed a red coloration and was evaporated to dryness. Three times extraction with pentane (3 mL each) yielded a black solution which was concentrated to 2 mL and stored at −60 °C for 36 h. Black needle-shaped crystals of 15 (33 mg, 0.02 mmol, 40%) were isolated by decantation and dried in vacuo. All NMR of 15 were measured in pentane with a D2O capillary. 1H NMR (δ in ppm): 6.21 (10H, Cp), 0.81 (24H, SiMe2), 0.74 (72H, SiMe3). 13C NMR (δ in ppm): 96.9 (Cp), 6.9 (SiMe3), 3.6(SiMe2). 29Si NMR(δ in ppm): 1.0 (SiMe3), −4.9 (SiMe2), −16.6 (quart. Si). 207Pb NMR: no signal was observed.

Zirconocene Plumbylene Tetrahydrofurane Complex (16)

Cp2ZrCl2 (100 mg, 0.34 mmol), 4 (114 mg, 0.17 mmol (considered to be monomeric in solution), and magnesium turnings (24 mg, 1.0 mmol, excess) were suspended in THF and subjected to ultrasonification for 5 min. The mixture turned deep red and was stirred for an additional hour. All volatile materials were removed in vacuo, and the black residue was extracted with pentane (3× 3 mL). The red-purple extract was concentrated to 3 mL and stored for 72 h at −60 °C. Purple crystals of 16 (106 mg, 0.11 mmol, 63%) were isolated by decantation and cautiously dried in vacuo. NMR (C6D6, rt, δ in ppm): 1H: 4.82 (s, 10H, Cp), 2.68 (br, 4H, THF), 1.63 (br, 4H, THF), 0.44 (s, 12H, SiMe2), 0.29 (s, 36H, SiMe3). 13C: 95.6 (Cp), 60.6 (THF), 20.1 (THF), 3.5 (SiMe3), 2.0(SiMe2). 29Si: −10.0 (SiMe3), −11.3 (SiMe2), −36.7 (quart. Si). 207Pb: 5770.

Trichlorotantalum Stannylene Diphosphine Complex (17)

A mixture of TaCl5 (75 mg, 0.21 mmol), magnesium turnings (12 mg, 0.5 mmol, excess), PEt3 (50 mg, 0.42 mmol), and 1 (150 mg, 0.21 mmol) were suspended in THF and subjected to ultrasonification for 2 min, during which time the suspension turned purple. The mixture was stirred for 3 h at rt, and then all volatile materials were removed in vacuo. The residue was extracted with pentane (10 mL). The purple extract was concentrated (4 mL) and stored at −60 °C for 48 h. Deep-purple crystals of 17 (107 mg, 0.10 mmol, 46%) were isolated by decantation and dried in vacuo. 1H NMR (δ in ppm): 2.17 (m, 6H, P(CH2CH3)3), 1.05 (m, 9H, P(CH2CH3)3), 0.51 (s, 12H, SiMe2), 0.47 (s, 36H, SiMe3). 13C NMR (δ in ppm): 24.2 (P(CH2CH3)3), 9.0 (P(CH2CH3)3), 4.5 (SiMe3), −0.9 (SiMe2). 29Si NMR (δ in ppm): 6.7 (SiMe3), −18.6 (SiMe2), −105.9 (quart. Si). 31P NMR (δ in ppm): 38.7 (s, 2JPSn = 73 Hz). 119Sn NMR (δ in ppm): 1985 (t, 2JPSn = 73 Hz). Anal. calcd for C28H78Cl3P2Si8SnTa (1107.57): C, 30.36; H, 7.10. Found: C, 30.17; H, 7.025.