Julia Volk, Bruno A. Correia Bicho, Clemens Bruhn and Ulrich Siemeling

N-Heterocyclic germylenes and stannylenes of the type [Fe{(η5-C5H4)NR}2E] with bulky alkyl substituents

Accessible
De Gruyter | Published online: September 23, 2017

Abstract

The 1,1′-diaminoferrocene derivatives [Fe(η5-C5H4–NHAd)2] (Ad=2-adamantyl) and [Fe(η5-C5H4–NHtBu)2] were investigated in terms of their suitability for the synthesis of N-heterocyclic tetrylenes of the type [{Fe(η5-C5H4–NR)2}E] (E=Ge, Sn). The synthesis of these target compounds was easily achieved with R=tBu, but failed with R=Ad. In the latter case, the stannylene was not sufficiently stable for isolation and decomposed to the aminoiminoferrocene derivative [Fe(η5-C5H4–NHAd)(η5-C5H4–N=Ad′)] (Ad′=adamant-2-ylidene). Attempts to synthesise [{Fe(η5-C5H4–NAd)2}Ge] afforded intractable material, from which the unusual compound [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] was obtained by serendipity. It contains GeO, stabilised by adduct formation with GeCl2 and the target germylene. [Fe(η5-C5H4–NHtBu)2], [Fe(η5-C5H4–NHAd)2], [Fe(η5-C5H4–NHAd)(η5-C5H4–N=Ad′)], [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] and [{Fe(η5-C5H4–NtBu)2}E] (E=Ge, Sn) were structurally characterised by single-crystal X-ray diffraction.

1 Introduction

N-heterocyclic tetrylenes (NHTs) are the heavier analogues of the highly popular N-heterocyclic carbenes (NHCs) [1]. The lone pair of the divalent carbon atom of an NHC can be viewed as residing in an sp2-hybridised orbital, whereas the corresponding donor orbital of the heavier congeners has significantly higher s character due to the increasing differences in size and energy between the orbitals of the ns and np subshells, which renders hybridisation increasingly unfavourable for the heavier elements [2], [3]. Consequently, NHTs are stronger π acceptors and weaker σ donors than NHCs. Although the chemistry of NHTs has developed remarkably during the past decades [3], [4], [5], [6], [7], it is still far away from the maturity that has been reached in the field of NHCs. For example, the extensive work on NHCs equipped with additional functionalities [8], [9], [10], [11], [12], [13], [14], including photo- and redox-active units for reversible switching behaviour useful in catalysis [15], is not yet mirrored in the area of NHTs. In this context we recently established germanium and tin analogues [16] of the stable ferrocene-based NHCs of the type [{Fe(η5-C5H4–NR)2}C] [17], [18], [19], [20], which exhibit a highly unconventional reactivity [17], [21] similar to that of (alkyl)(amino)carbenes [22]. We described five stable NHTs of the type [{Fe(η5-C5H4–NR)2}E] (E=Ge, R=CH2tBu, Mes, TMS; E=Sn, R=Mes, TMS), which constitute the first examples of redox-functionalised NHTs in general. Independently from us, the germylene [{Fe(η5-C5H4–NMes)2}Ge] was also prepared and studied by Breher and co-workers [23]. While the neopentyl-substituted germylene [{Fe(η5-C5H4–NCH2tBu)2}Ge] was easily obtained in essentially quantitative yield, either by a transamination reaction of the diamine [Fe(η5-C5H4–NHCH2tBu)2] with the stable acyclic germylene [(Me3Si)2N]2Ge [24] or by the reaction of the lithium diamide [Fe(η5-C5H4–NLiCH2tBu)2] with [GeCl2(1,4-dioxane)] [25], analogous attempts to synthesise the corresponding stannylene [{Fe(η5-C5H4–NCH2tBu)2}Sn] were not successful. This compound appears to be not sufficiently stable for isolation due to a decomposition pathway which leads to the aminoiminoferrocene derivative [Fe(η5-C5H4–NHCH2tBu)(η5-C5H4–N=CHtBu)] and plausibly involves an initial β-hydrogen elimination step. The neopentyl substituent is a bulky primary alkyl group. The aim of the work presented here was to investigate whether the use of bulky secondary or tertiary alkyl substituents allows the synthesis and isolation of the corresponding germylenes and stannylenes. The 2-adamantyl (Ad) and the tert-butyl substituent were selected for this purpose. The choice of the former substituent was motivated by our earlier carbene work, where [{Fe(η5-C5H4–NAd)2}C] was the first stable NHC with a ferrocene-based backbone [20]. The tert-butyl substituent was chosen for two reasons. First, it is the simplest tertiary alkyl group. Second, it is closely related to the trimethylsilyl group, which allowed the synthesis and isolation of the corresponding NHTs [{Fe(η5-C5H4–NSiMe3)2}E] (E=Ge, Sn). Note that CMe3 and SiMe3 do not contain β-hydrogen atoms and are therefore not compatible with the decomposition pathway involving a β-hydrogen elimination step suggested above for the neopentyl-substituted stannylene [{Fe(η5-C5H4–NCH2tBu)2}Sn]. The steric effect of a tert-butyl group is known to be larger than that of a trimethylsilyl group, although the former occupies a smaller volume. This is due to the fact that a silicon atom is larger than a carbon atom, leading to comparatively longer bonds (in this case to nitrogen) [26].

2 Results and discussion

The synthesis and isolation of the tert-butyl substituted target NHTs [{Fe(η5-C5H4–NtBu)2}E] (E=Ge, Sn) turned out to be straightforward (Scheme 1).

Scheme 1: Synthesis of the tert-butyl substituted NHTs [{Fe(η5-C5H4–NtBu)2}E] (E=Ge, Sn).

Scheme 1:

Synthesis of the tert-butyl substituted NHTs [{Fe(η5-C5H4–NtBu)2}E] (E=Ge, Sn).

The diamine [Fe(η5-C5H4–NHtBu)2] [27] was converted to the corresponding lithium diamide [Fe(η5-C5H4–NLitBu)2] with nBuLi in hexane. The diamide was obtained as a very air- and moisture-sensitive red precipitate. Its subsequent reaction with [GeCl2(1,4-dioxane)] [25] in THF or with SnCl2 in benzene/THF (6:1), respectively afforded [{Fe(η5-C5H4–NtBu)2}Ge] and [{Fe(η5-C5H4–NtBu)2}Sn] in excellent yields. The identity of these NHTs is strongly supported by solution NMR spectroscopic data (C6D6), which are also fully in accord with the molecular structures in the solid state determined by X-ray crystallography (see below). The 1H NMR spectrum exhibits a singlet signal due to the tert-butyl groups at δ=1.39 and 1.31 ppm for the germylene and stannylene, respectively. The cyclopentadienyl rings give rise to an ill-resolved multiplet at δ=3.94 ppm in each case, which almost appears as a broad singlet. For comparison, the signal due to the tert-butyl groups of the diamine [Fe(η5-C5H4–NHtBu)2] is located at δ=1.08 ppm and the cyclopentadienyl protons of the diamine give rise to two singlets at δ=3.92 and 4.01 ppm. The 119Sn NMR signal of [{Fe(η5-C5H4–NtBu)2}Sn] is located at δ=544 ppm, which compares well with the chemical shift of 589 ppm found for the trimethylsilyl congener [{Fe(η5-C5H4–NSiMe3)2}Sn] [16].

The chemistry was more difficult with [Fe(η5-C5H4–NHAd)2]. Its lithiation with nBuLi in toluene/hexane (1:1) afforded the corresponding diamide as an orange precipitate, which was considerably more sensitive than the tert-butyl congener. It was best handled in a glove-box (<1 ppm O2 and H2O) and used immediately after its preparation. The compound showed only rather limited solubility in C6D6 and the 1H NMR spectrum was too complex for a meaningful interpretation. Similar behaviour was previously found for the mesityl congener [Fe(η5-C5H4–NLiMes)2] and explained by aggregate formation [16]. The identity of [Fe(η5-C5H4–NLiAd)2] was confirmed in an NMR experiment by reaction with D2O, which cleanly afforded [Fe(η5-C5H4–NDAd)2], the 1H NMR spectrum of which was identical to that of [Fe(η5-C5H4–NHAd)2], except for the missing NH signal located at δ=2.39 ppm in C6D6 [20].

Attempts to react freshly prepared [Fe(η5-C5H4–NLiAd)2] with [GeCl2(1,4-dioxane)] in a 1:1 molar ratio under various conditions invariably furnished intractable material which resisted further analysis. However, when two equivalents of [GeCl2(1,4-dioxane)] were used in THF, a light brown solid was obtained, which was fairly soluble in C6D6 and subjected to 1H NMR spectroscopic analysis. Two singlets of equal integral at δ=3.95 and 4.00 ppm indicated the formation of a new ferrocene derivative. For comparison, the cyclopentadienyl signals of the diamine [Fe(η5-C5H4–NHAd)2] are located at δ=3.83 and 3.92 ppm in C6D6. Not surprisingly, signals due to small amounts of THF and 1,4-dioxane as well as the diamine [Fe(η5-C5H4–NHAd)2] (very likely formed by trace amounts of adventitious moisture) were also present in the spectrum. These 1H NMR spectroscopic results are in accord with, but do not prove, the formation of the target germylene [{Fe(η5-C5H4–NAd)2}Ge]. After several days at room temperature a small number of light brown crystals formed from the C6D6 solution. A single-crystal X-ray diffraction study revealed that they were composed of the unusual germanium(II) compound [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] (Fig. 1; see Fig. 6 for the molecular structure in the crystal), which from a formal point of view contains GeO, stabilised by adduct formation with GeCl2 (acting as a Lewis acid) and the target germylene [{Fe(η5-C5H4–NAd)2}Ge] (acting as a Lewis base). GeO belongs to the family of neutral molecular tetrel monochalcogenides. These are the heavier analogues of carbon monoxide and are usually investigated at high temperatures in the gas phase [28] or at low temperatures in inert solid matrices [29]. First examples were recently stabilised by adduct formation, which proved possible so far for GeSe [30], GeTe [30], SnO [31], PbO [31] and PbSe [32]. We note in this context that SnO and PbO were respectively generated in situ by hydrolysis of a diaminostannylene and diaminoplumbylene [31]. It is likely that in our case an analogous reaction with trace amounts of adventitious moisture gave rise to GeO.

Fig. 1: The germanium(II) compound [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] obtained by serendipity from the reaction of [Fe(η5-C5H4–NLiAd)2] with two equivalents of [GeCl2(1,4-dioxane)] in THF.

Fig. 1:

The germanium(II) compound [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] obtained by serendipity from the reaction of [Fe(η5-C5H4–NLiAd)2] with two equivalents of [GeCl2(1,4-dioxane)] in THF.

The reaction of freshly prepared [Fe(η5-C5H4–NLiAd)2] with SnCl2 in a 1:1 molar ratio was investigated under various conditions. Frequently, a completely insoluble brownish solid was obtained (very likely elemental tin) together with a mixture of soluble compounds, containing, inter alia, the aminoiminoferrocene derivative [Fe(η5-C5H4–NHAd)(η5-C5H4–N=Ad′)] (Ad′=adamant-2-ylidene). This compound was subsequently prepared by reacting [(Me3Si)2N]2Pb [24] with [Fe(η5-C5H4–NHAd)2]. The initially formed plumbylene undergoes a rapid and clean decomposition to elemental lead and [Fe(η5-C5H4–NHAd)(η5-C5H4–N=Ad′)], in analogy to the formation of [Fe(η5-C5H4–NHCH2tBu)(η5-C5H4–N=CHtBu)] which was previously described by us [16]. When the reaction of [Fe(η5-C5H4–NLiAd)2] with SnCl2 was performed in a toluene/THF mixture (9:1) at −20°C, followed by rapid work-up at this temperature after reaction times of only a few minutes, a dark brown solid was obtained, which was largely soluble in cold C6D6 and gave rise to three signals in the 119Sn NMR spectrum, located at δ=−400, 172 and 425 ppm, the latter one being compatible with the target stannylene [{Fe(η5-C5H4–NAd)2}Sn]. For comparison, the 119Sn NMR signals of the stable tert-butyl (see above), trimethylsilyl and mesityl congeners [16] were observed at δ=540, 589 and 403 ppm, respectively.

2.1 Crystal structures

Surprisingly, only two examples of structurally characterised 1,1′-di(alkylamino)ferrocenes have been reported to date, namely the benzyl-substituted derivatives [Fe(η5-C5H4–NHCH2-p-C6H4-NMe2)2] and [Fe(η5-C5H4–NHCH2-p-C6H4–OMe)2] [33]. Single-crystal X-ray diffraction studies were therefore performed for [Fe(η5-C5H4–NHAd)2] (Fig. 2) and [Fe(η5-C5H4–NHtBu)2] (Fig. 3).

Fig. 2: Molecular structure of [Fe(η5-C5H4–NHAd)2] in the crystal.

Fig. 2:

Molecular structure of [Fe(η5-C5H4–NHAd)2] in the crystal.

Fig. 3: Molecular structure of [Fe(η5-C5H4–NHtBu)2] in the crystal.

Fig. 3:

Molecular structure of [Fe(η5-C5H4–NHtBu)2] in the crystal.

The cyclopentadienyl rings are almost coplanar in each case (see below). Their conformation is synperiplanar in the case of [Fe(η5-C5H4–NHtBu)2] (torsion angle N–Cipso–Cipso–N 7.2°) and synclinal in the case of [Fe(η5-C5H4–NHAd)2] (torsion angle N–Cipso–Cipso–N 72.9°). This differs markedly from the antiperiplanar arrangement observed for [Fe(η5-C5H4–NHCH2-p-C6H4X)2] (X=NMe2, OMe) [33]. The almost fully eclipsed conformation of [Fe(η5-C5H4–NHtBu)2] coincides with intramolecular hydrogen bonding between the two NH units (not shown in Fig. 3). Due to this attractive interaction, the cyclopentadienyl rings are tilted by 2.7° in that direction which brings the two Cipso atoms closer together. On top of that the Cipso–N bond vectors are inclined slightly towards the Fe atom by −3.1 and −2.2° (cg–Cipso–N 176.9 and 177.8°, cg=cyclopentadienyl ring centroid). This allows an N1···N2 distance of 3.046(3) Å, ca. 0.30 Å shorter than the interplanar distance between the Cp rings in ferrocene. The average N–H···N angle is ca. 112°, which is at the low end of the values commonly observed for such hydrogen bonds. These structural data point to a weak intramolecular interaction [34], [35], [36], which is not at all unusual for diamines [37]. Hydrogen bonding is absent in the structure of [Fe(η5-C5H4–NHAd)2]. In this case, the cyclopentadienyl ring tilt angle is merely 1.5°. However, in contrast to the tert-butyl congener the wedge opens up in the direction of the Cipso atoms and the Cipso–N bond vectors are pointing away from the Fe atom by 4.7 and 3.3°.

We have also determined the structure of the closely related aminoiminoferrocene derivative [Fe(η5-C5H4–NHAd)(η5-C5H4–N=Ad′)] (Fig. 4). The molecules are aggregated as centrosymmetric dimers by intermolecular N–H···N hydrogen bonds, whose metrical parameters [N1···N2 2.989(5) Å, N–H···N 161(3)°] indicate a significantly stronger interaction than that discussed above. The cyclopentadienyl ring tilt angle is 3.9° in this case. Similar to the diamine [Fe(η5-C5H4–NHAd)2], the wedge opens up in the direction of the Cipso atoms and the Cipso–N bond vectors are pointing away from the Fe atom by 3.6 and 1.8°. The Nimine atom exhibits two substantially different C–N bond lengths, viz. 1.415(5) and 1.289(5) Å, in accord with a single and a double bond between sp2 hybridised carbon and nitrogen atoms. The corresponding interatomic distances of the Namine atom are 1.380(5) and 1.459(5) Å for the bond to the sp2 and sp3 hybridised carbon atom, respectively, which compares well to the values determined for the diamines [Fe(η5-C5H4–NHAd)2] [1.388(8) and 1.399(8) Å for N–C(sp2) vs. 1.458(8) and 1.449(8) Å for N–C(sp3)] and [Fe(η5-C5H4–NHtBu)2] [1.411(4) and 1.414(3) Å for N–C(sp2) vs. 1.474(3) and 1.463(4) Å for N–C(sp3)]. The C–Namine–C and C–Nimine–C angle is 120.7(3) and 117.8(3)°, respectively. This similarity of the bond angles at the Namine and Nimine atoms is not uncommon. For example, the C–Namine–C angles of [Fe(η5-C5H4–NHCH2-p-C6H4–NMe2)2] are 117.9(4)°. This value lies in the narrow range of the C–Nimine–C angles [116.4(2)–120.7(2)°, average value 118.5°] found for the two independent molecules of [Fe(η5-C5H4–N=CH-p-C6H4–NMe2)2] present in the crystal [33].

Fig. 4: Molecular structure and aggregation of [Fe(η5-C5H4–NHAd)(η5-C5H4–N=Ad′)] in the crystal. Hydrogen bonds are indicated by broken lines.

Fig. 4:

Molecular structure and aggregation of [Fe(η5-C5H4–NHAd)(η5-C5H4–N=Ad′)] in the crystal. Hydrogen bonds are indicated by broken lines.

We next come to the two structurally characterised germanium(II) compounds of this study, viz. [{Fe(η5-C5H4–NtBu)2}Ge] (Fig. 5) and [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] (Fig. 6).

Fig. 5: Molecular structure of [{Fe(η5-C5H4–NtBu)2}Ge] in the crystal.

Fig. 5:

Molecular structure of [{Fe(η5-C5H4–NtBu)2}Ge] in the crystal.

Fig. 6: Molecular structure of [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] in the crystal. Selected bond angles (deg): Ge1–O1–Ge2 89.75(12), Ge1–O1–Ge3 133.41(15), Ge2–O1–Ge3 136.82(15), N1–Ge1–N2 83.19(13), N1–Ge1–O1 78.35(12), N2–Ge1–O1 79.06(12), N1–Ge2–N2 82.40(13), N1–Ge2–O1 77.74(12), N2–Ge2–O1 78.64(13), O1–Ge3–Cl1 94.66(9), O1–Ge3–Cl2 94.29(9), Cl1–Ge3–Cl2 94.84(5).

Fig. 6:

Molecular structure of [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] in the crystal. Selected bond angles (deg): Ge1–O1–Ge2 89.75(12), Ge1–O1–Ge3 133.41(15), Ge2–O1–Ge3 136.82(15), N1–Ge1–N2 83.19(13), N1–Ge1–O1 78.35(12), N2–Ge1–O1 79.06(12), N1–Ge2–N2 82.40(13), N1–Ge2–O1 77.74(12), N2–Ge2–O1 78.64(13), O1–Ge3–Cl1 94.66(9), O1–Ge3–Cl2 94.29(9), Cl1–Ge3–Cl2 94.84(5).

The molecular structure of the N-heterocyclic germylene [{Fe(η5-C5H4–NtBu)2}Ge] is similar to that of the neopentyl and trimethylsilyl congeners recently described by us [16]. Pertinent metric parameters are collected in Table 1. For conciseness, the data for the corresponding stannylenes are also included.

Table 1:

Pertinent metric parameters (bond lengths in Å, angles in deg) for [{Fe(η5-C5H4–NtBu)2}E] (E=Ge, Sn) and their neopentyl and/or trimethylsilyl congeners.

N–E N–E–N Σ∠N Tilt angle N–Cipso–Cipso–N
[{Fe(η5-C5H4–NtBu)2}Ge]a 1.861(4)

1.854(4)
107.44(16) 359.2(3)

359.9(3)
6.5 4.8
[{Fe(η5-C5H4–NCH2tBu)2}Ge]a,b 1.852(5)

1.834(6)
106.7(2) 359.0(4)

358.7(4)
6.5 3.1
[{Fe(η5-C5H4–NSiMe3)2}Ge]a,b 1.845(2) 106.48(13) 359.55(15) 8.3 9.5
[{Fe(η5-C5H4–NtBu)2}Sn]c 2.066(4) 103.2(2) 359.0(4) 3.0 10.9
[{Fe(η5-C5H4–NSiMe3)2}Sn]b,c 2.058(6) 103.1(3) 359.3(5) 5.1 10.4

    aE=Ge; bRef. [16]; cE=Sn.

The N atoms are trigonal planar in each case. The cyclopentadienyl rings are tilted only slightly, the angle being close to the value of 7.4° reported for the almost unstrained [3]ferrocenophane in each case [38]. The germanium bond angles of the ferrocene-based germylenes in Table 1 are very similar to the values reported for Lappert’s iconic acyclic diaminogermylene [(Me3Si)2N]2Ge (ca. 107°) [39] and the closely related compound [(iPrMe2Si)2N]2Ge (ca. 105°) [40]. This is strongly reminiscent of the fact that the N–C–N bond angles of ferrocene-based N-heterocyclic carbenes of the type [{Fe(η5-C5H4–NR)2}C] (ca. 120°) [19], [20] are essentially identical with the values determined for acyclic diaminocarbenes [41], [42], [43].

The unusual compound [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] contains three GeII atoms, which are each three-coordinate and exhibit a trigonal pyramidal configuration (sum of angles: 240.6, 238.8 and 283.8° for Ge1, Ge2 and Ge3, respectively). The oxygen atom is in a trigonal-planar bonding environment (sum of angles 360.0°). The germanium atoms Ge1 and Ge2 are each bridged by the oxygen atom and both nitrogen atoms, which results in rather acute germanium bond angles close to 80°. The bond angles of the third germanium atom are ca. 95°. The molecule has approximate Cs symmetry (the Fe atom lies only 0.1 Å out of the plane of the three Ge atoms). It can be formally viewed as GeO stabilised by adduct formation with dichlorogermylene (acting as a Lewis acid) and the diaminogermylene [{Fe(η5-C5H4–NAd)2}Ge] (acting as a Lewis base). The Ge–N bond lengths lie in the narrow range from 2.074(3) to 2.098(3) Å [average value 2.084(3) Å] and are thus ca. 0.2 Å longer than those of [{Fe(η5-C5H4–NtBu)2}Ge], whose GeII atom is only two-coordinate. They are similar to those of the Ge2N2 core [average value 2.074(4) Å] of the aggregational diaminogermylene dimer [(CH2NBz)2Ge]2 (Bz=benzyl), whose GeII atoms are also in a tricoordinate bonding environment [44]. The Ge–O bond lengths are 1.928(3) and 1.933(3) Å for the two nitrogen-bonded germanium atoms. These values agree well with those reported for the OSnII3 core of the bis(diaminostannylene)-stabilised SnO (ca. 2.08–2.13 Å) [31], considering the difference of the covalent radii of germanium (1.20 Å) and tin (1.39 Å) [45]. The corresponding Ge–O distance of 1.897(3) Å determined for the chlorine-bonded germanium atom is slightly, but significantly, shorter. For comparison, the Ge–O bond lengths of the polymeric dichlorogermylene adduct [GeCl2(μ2-1,4-dioxane)], which contains tetracoordinate GeII atoms, are 2.3989(12) Å [46], whereas the digermylene [Ar(SiMe3)NGe]2O [Ar=C6H2(CHPh2)2Me-2,6,4], whose oxygen and germanium atoms are exclusively dicoordinate, has Ge–O bond lengths of 1.8088(16) and 1.8154(15) Å and Ge–N bond lengths of 1.875(2) and 1.878(2) Å [47]. The Ge2N2 unit of [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] is not planar. The folding is due to the fact that both germanium atoms are also bonded to the oxygen atom. The bridging oxygen atom causes a rather short distance of these two germanium atoms of 2.7240(6) Å, only ca. 13% longer than the sum of the covalent radii (2.40 Å) [45] and much less than the sum of the van der Waals radii (4.58 Å) [36].

Finally, we come to the structure of the stannylene [{Fe(η5-C5H4–NtBu)2}Sn] (Fig. 7), which is very similar to that of the trimethylsilyl congener (see Table 1).

Fig. 7: Molecular structure of [{Fe(η5-C5H4–NtBu)2}Sn] in the crystal.

Fig. 7:

Molecular structure of [{Fe(η5-C5H4–NtBu)2}Sn] in the crystal.

The tin bond angles of these ferrocene-based stannylenes (ca. 103°) are very close to those reported for the acyclic diaminostannylenes [(Me3Si)2N]2Sn (ca. 105°) [48] and [(PhMe2Si)2N]2Sn (ca. 102°) [49]. A comparison of the bond parameters of the germylenes of this study with their tin homologues (Table 1) reveals two trends which are well-known in this area of chemistry [5]. First, the Ge–N bonds are shorter than the Sn–N bonds by ca. 0.2 Å, which is in accord with the difference of the covalent radii of germanium (1.20 Å) and tin (1.39 Å) [47]. Second, the bond angles of germanium are wider than those of tin (by ca. 3°), which is in accord with Bent’s rule [50], [51].

3 Conclusion

The facile synthesis of the tert-butyl substituted NHTs [{Fe(η5-C5H4–NtBu)2}E] (E=Ge, Sn) indicates that steric hindrance is not a major issue for the synthesis of such compounds. Consequently, we surmise that the problems encountered in this chemistry with the 2-adamantyl substituent are predominantly due to other reasons. The outcome of our attempts to synthesise and isolate the stannylene [{Fe(η5-C5H4–NAd)2}Sn] is strongly reminiscent of what we observed before in the case of the neopentyl-substituted congener [{Fe(η5-C5H4–NCH2tBu)2}Sn] [16]. This further supports the notion that stannylenes of this type which bear primary or secondary N-alkyl substituents are prone to decomposition by a pathway involving a β-hydrogen elimination step [52]. The structurally characterised germanium(II) compound [μ2-{Fe(η5-C5H4–NAd)2}Ge2(μ3-O)(GeCl2)] adds GeO to the limited number of neutral molecular tetrel monochalcogenides stabilised by adduct formation.

4 Experimental section

All reactions involving air-sensitive compounds were performed in an inert atmosphere (argon or dinitrogen) by using standard Schlenk techniques or a conventional glovebox. Starting materials were procured from standard commercial sources and used as received. [Fe(η5-C5H4–NHAd)2] [20], [GeCl2(1,4-dioxane)] [25] and [(Me3Si)2N]2Pb [24], [53] were synthesised by following adapted versions of the published procedures. NMR spectra were recorded at ambient temperature with Varian NMRS-500 and MR-400 spectrometers operating at 500 and 400 MHz, respectively, for 1H. 119Sn NMR spectra were recorded with a Varian NMRS-500 spectrometer.

4.1 [{Fe(η5-C5H4–NtBu)2}Ge]

nBuLi (0.50 mL of a 1.60 M solution in hexane, 0.80 mmol, 2.6 equivalents) was added dropwise to a stirred solution of [Fe(η5-C5H4–NHtBu)2] [27] (100 mg, 0.30 mmol) in hexane (5 mL) cooled to 0°C, leading to the formation of [Fe(η5-C5H4–NLitBu)2] as a red precipitate. After 15 min the solid was isolated by filtration, washed with hexane (3×5 mL) and dried under vacuum. Yield 76 mg (73%). [Fe(η5-C5H4–NLitBu)2] (76 mg, 0.22 mmol) was dissolved in THF (5 mL) and [GeCl2(1,4-dioxane)] (52 mg, 0.22 mmol) was added to the stirred solution. After 20 h volatile components were removed under vacuum. The solid residue was extracted with hexane (10 mL) and the extract reduced to dryness under vacuum, leaving the product as a yellow solid. Yield 85 mg (95%). – 1H NMR (C6D6): δ=1.39 (s, 18 H, tBu), 3.94 (m, 8 H, C5H4). – 13C{1H} NMR (C6D6): δ=34.5 (CMe3), 58.8 (CMe3), 67.5, 70.5 (2×CH), 109.1 (Cipso). – C18H26N2FeGe (398.9): calcd. C 54.20, H 6.57, N 7.02; found C 54.31, H 6.67, N 6.79.

4.2 [{Fe(η5-C5H4–NtBu)2}Sn]

[Fe(η5-C5H4–NLitBu)2] (prepared as described above; 81 mg, 0.24 mmol) was dissolved in a mixture of benzene (12 mL) and THF (2 mL). SnCl2 (45 mg, 0.24 mmol) was added and the mixture stirred for 1 h. Volatile components were removed under vacuum. The solid residue was extracted with hexane (10 mL) and the extract reduced to dryness under vacuum, leaving the product as an orange solid. Yield 102 mg (97%). – 1H NMR (C6D6): δ=1.31 (s, 18 H, tBu), 3.94 (m, 8 H, C5H4). – 13C{1H} NMR (C6D6): δ=34.9 (CMe3), 59.8 (CMe3), 66.6, 71.3 (2×CH), 112.4 (Cipso). – 119Sn NMR (C6D6): δ=544. – C18H26N2FeSn (445.0): calcd. C 48.59, H 5.89, N 6.30; found C 49.00, H 6.08, N 5.65.

4.3 [Fe(η5-C5H4–NHAd)(η5-C5H4–N=Ad′)]

A solution of [(Me3Si)2N]2Pb (545 mg, 1.00 mmol) in benzene (4 mL) was added to a solution of [Fe(η5-C5H4–NHAd)2] (500 mg, 1.03 mmol) in benzene (5 mL). The solution was stirred for 16 h. The grey solid was removed by filtration. The filtrate was reduced to dryness under vacuum. The brown residue was washed with hexane (2 mL) and recrystallised from toluene, which furnished the product as an orange-brown microcrystalline solid. Yield 320 mg (64%). – 1H NMR (C6D6): δ=1.47–2.21 (m, 29 H, Ad/Ad′), 3.15 (s, 1 H, NH), 3.25, 3.40 (2 s, 2×1 H, Ad/Ad′), 4.01 (m, 2 H, C5H4), 4.03 (m, 4 H, C5H4), 4.17 (m, 2 H, C5H4). – 13C{1H} NMR (C6D6): δ=28.2, 32.1, 33.2, 35.3, 36.8, 37.7, 38.3, 38.9, 39.5, 44.0 (10×Ad/Ad′ CH and CH2), 59.7 (cyclopentadienyl CH), 61.0 (CHNH), 63.6, 64.7, 65.5 (3×cyclopentadienyl CH), 105.8, 110.3 (2×Cipso), 181.7 (C=N). – C30H38N2Fe·H2O (500.5): calcd. C 71.99, H 8.06, N 5.60; found C 72.76, H 8.03, N 5.47.

4.4 Crystal structure determinations

For each data collection a single crystal was taken from the mother liquor, mounted on a micro-mount and all geometric and intensity data were taken from this sample. Data collections were carried out using Mo radiation (λ=0.71073 Å) on a Stoe IPDS2 diffractometer equipped with a 2-circle goniometer and an area detector or on a Stoe StadiVari diffractometer equipped with a 200 K Dectris Pilatus detector. The data sets were corrected for absorption (by integration), Lorentz and polarisation effects. The structures were solved by direct methods (Sir 2008) [54] and refined using alternating cycles of full-matrix least-squares refinements against F2 (Shelxl2014/7) [55]. H atoms were included to the models in calculated positions with the 1.2 fold isotropic displacement parameter of their bonding partner. Experimental details for each diffraction experiment are given in Table 2.

Table 2:

Crystallographic data for the structurally characterised compounds of this study.

[Fe(η5-C5H4–NHAd)2] [Fe(η5-C5H4–NHtBu)2] [Fe(η5-C5H4–NHAd)-(η5-C5H4–N=Ad′)] [{Fe(η5-C5H4–NtBu)2}Ge] 2-{Fe(η5-C5H4–NAd)2}-Ge23-O)(GeCl2)]·½C6H6 [{Fe(η5-C5H4–NtBu)2}Sn]
Chemical formula C30H40FeN2 C18H28FeN2 C30H38FeN2 C18H26FeGeN2 C33H41Cl2FeGe3N2O C18H26FeN2Sn
Formula mass 484.49 328.27 482.47 398.85 826.20 444.95
Crystal size, mm3 0.13×0.04×0.02 0.24×0.12×0.08 0.12×0.11×0.09 0.23×0.15×0.05 0.15×0.14×0.05 0.17×0.14×0.06
Tmin/Tmax 0.9592/0.9857 0.8507/0.9554 0.8141/0.8892 0.6197/0.8943 0.8418/0.9297 0.7244/0.8816
T, K 100(2) 298(2) 100(2) 258(2) 100(2) 100(2)
Crystal system Triclinic Triclinic Triclinic Monoclinic Triclinic Monoclinic
Space group (no.) P1̅ (2) P1̅ (2) P1̅ (2) P21/c (14) P1̅ (2) C2/c (15)
a, Å 6.5374(4) 9.5572(11) 10.3411(6) 17.0834(14) 11.7242(5) 18.2986(12)
b, Å 12.7602(10) 10.2170(11) 10.4143(7) 9.4753(5) 11.7143(6) 9.2759(9)
c, Å 14.3209(11) 11.3957(14) 11.5129(6) 11.4053(9) 12.4602(6) 10.9584(8)
α, deg 93.762(6) 68.169(8) 89.682(5) 90 86.994(4) 90
β, deg 97.837(6) 81.251(10) 69.177(4) 104.670(6) 66.684(4) 108.956(5)
γ, deg 101.848(6) 62.438(8) 82.406(5) 90 78.065(4) 90
Unit cell volume, Å3 1152.85(15) 915.4(2) 1147.59(12) 1786.0(2) 1536.75(14) 1759.2(2)
Z 2 2 2 4 2 4
Dcalcd, g cm−1 1.40 1.19 1.40 1.48 1.79 1.68
μ, mm−1 0.7 0.8 0.7 2.5 3.6 2.2
F(000), e 520 352 516 824 834 896
θ range, deg 1.44–25.50 1.93–26.00 1.89–25.50 2.47–27.00 1.78–25.50 2.35–26.00
hkl range −7→7

−15→15

−17→17
−11→10

−12→12

−13→14
−12→12

−10→12

−13→13
−20→21

−11→12

−14→14
−13→14

−14→14

−15→15
−22→22

−11→11

−13→13
No. of refl. measured 8442 6753 8439 14 094 11 287 5853
Unique refl./Rint 4267/0.1091 3574/0.0431 4258/0.0620 3890/0.0355 5688/0.0414 1733/0.0325
Parameters 304 202 301 205 379 104
Final R1/wR2a [I>2 σ (I)] 0.0767/0.1956 0.0429/0.1028 0.0624/0.1731 0.0505/0.1438 0.0477/0.1313 0.0479/0.1171
Final R1/wR2a (all data) 0.1193/0.2241 0.0581/0.1104 0.0763/0.1834 0.0746/0.1596 0.0554/0.1369 0.0584/0.1243
Goodness-of-fit (F2) 1.055 0.973 1.061 1.107 1.059 1.174
Δρfin (max/min), e Å3 0.76/−0.88 0.46/−0.23 0.91/−0.76 0.81/−0.59 1.81/−1.11 0.99/−1.23

    awR2=[Σw(Fo2Fc2)2/w(Fo2)2]1/2; w=[σ2(Fo2)+(AP)2+BP]−1 with P=(Fo2+2Fc2)/3; A and B are constants adjusted by the programme.

CCDC 1546004–1546009 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre viawww.ccdc.cam.ac.uk/data_request/cif.

Acknowledgment

Support by the Fonds der Chemischen Industrie is gratefully acknowledged (stipend for J. V.).

References

[1] For a recent review on NHCs, see: M. N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature2014, 510, 485. Search in Google Scholar

[2] J. Baumgartner, C. Marschner, Rev. Inorg. Chem.2014, 34, 119. Search in Google Scholar

[3] Y. Mizuhata, T. Sasamori, N. Tokitoh, Chem. Rev.2009, 109, 3479. Search in Google Scholar

[4] M. Asay, C. Jones, M. Driess, Chem. Rev.2011, 111, 354. Search in Google Scholar

[5] M. Lappert, A. Protchenko, P. Power, A. Seeber, Metal Amide Chemistry, Wiley, Chichester, 2009, p. 279. Search in Google Scholar

[6] A. V. Zabula, F. E. Hahn, Eur. J. Inorg. Chem.2008, 5165. Search in Google Scholar

[7] S. Nagendran, H. W. Roesky, Organometallics2008, 27, 457. Search in Google Scholar

[8] S. Hameury, P. de Frémont, P. Braunstein, Chem. Soc. Rev.2017, 46, 632. Search in Google Scholar

[9] A. Nasr, A. Winkler, M. Tamm, Coord. Chem. Rev.2016, 316, 68. Search in Google Scholar

[10] Z. Mo, L. Deng, Synlett2014, 25, 1045. Search in Google Scholar

[11] C. Fliedel, P. Braunstein, J. Organomet. Chem.2014, 751, 286. Search in Google Scholar

[12] U. Siemeling, Eur. J. Inorg. Chem.2012, 3523. Search in Google Scholar

[13] B. Royo, E. Peris, Eur. J. Inorg. Chem.2012, 1309. Search in Google Scholar

[14] A. John, P. Ghosh, Dalton Trans.2010, 39, 7183. Search in Google Scholar

[15] E. Peris, Chem. Rev.2017, 53, 3733. Search in Google Scholar

[16] J. Oetzel, N. Weyer, C. Bruhn, M. Leibold, B. Gerke, R. Pöttgen, M. Maier, R. F. Winter, M. C. Holthausen, U. Siemeling, Chem. Eur. J.2017, 23, 1187. Search in Google Scholar

[17] A. R. Petrov, A. Derheim, J. Oetzel, M. Leibold, C. Bruhn, S. Scheerer, S. Oßwald, R. F. Winter, U. Siemeling, Inorg. Chem.2015, 54, 6657. Search in Google Scholar

[18] S. Rittinghaus, C. Färber, C. Bruhn, U. Siemeling, Dalton Trans.2014, 43, 3508. Search in Google Scholar

[19] U. Siemeling, C. Färber, M. Leibold, C. Bruhn, P. Mücke, R. F. Winter, B. Sarkar, M. von Hopffgarten, G. Frenking, Eur. J. Inorg. Chem.2009, 2009, 4607. Search in Google Scholar

[20] U. Siemeling, C. Färber, C. Bruhn, Chem. Commun.2009, 98. Search in Google Scholar

[21] U. Siemeling, C. Färber, C. Bruhn, M. Leibold, D. Selent, W. Baumann, M. von Hopffgarten, C. Goedecke, G. Frenking, Chem. Sci.2010, 1, 697. Search in Google Scholar

[22] D. Martin, M. Soleilhavoup, G. Bertrand, Chem. Sci.2011, 2, 389. Search in Google Scholar

[23] F. Walz, E. Moos, D. Garnier, R. Köppe, C. E. Anson, F. Breher, Chem. Eur. J.2017, 23, 1173. Search in Google Scholar

[24] D. H. Harris, M. F. Lappert, J. Chem. Soc. Chem. Commun.1974, 895. Search in Google Scholar

[25] T. Fjeldberg, A. Haaland, B. E. R. Schilling, M. F. Lappert, A. J. Thorne, J. Chem. Soc., Dalton Trans.1986, 1551. Search in Google Scholar

[26] R. V. Hoffman, Organic Chemistry, 2nd edition. Wiley, Hoboken, NJ, 2004, p. 167. Search in Google Scholar

[27] The preparation of this compound will be reported elsewhere. It was inspired by the synthesis of N-tert-butyl-3,5-dimethylaniline; see: C. E. Laplaza, W. M. Davis, C. C. Cummins, Organometallics1995, 14, 577. Search in Google Scholar

[28] For a gas-phase IR study of GeO, see: G. A. Thompson, A. G. Maki, A. Weber, J. Mol. Spectrosc.1986, 116, 136. Search in Google Scholar

[29] For a vibrational spectroscopy study of GeO in solid argon, see: A. Zumbusch, H. Schnöckel, J. Chem. Phys.1998, 108, 8092. Search in Google Scholar

[30] Y. Xiong, S. Yao, M. Karni, A. Kostenko, A. Burchert, Y. Apeloig, M. Driess, Chem. Sci.2016, 7, 5462. Search in Google Scholar

[31] A. V. Zabula, T. Pape, A. Hepp, F. M. Schappacher, U. C. Rodewald, R. Pöttgen, F. E. Hahn, J. Am. Chem. Soc.2008, 130, 5648. Search in Google Scholar

[32] G. Thiele, Y. Franzke, F. Weigend, S. Dehnen, Angew. Chem. Int. Ed.2015, 54, 11283. Search in Google Scholar

[33] C. Thie, C. Bruhn, U. Siemeling, Eur. J. Inorg. Chem.2015, 5457. Search in Google Scholar

[34] G. A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press, Oxford, 1997. Search in Google Scholar

[35] E. Arunan, G. R. Desiraju, R. A. Klein, J. Sadlej, S. Schreiner, I. Alkorta, D. C. Clary, R. H. Crabtree, J. J. Dannenberg, P. Hobza, H. G. Kjaergaard, A. C. Legon, B. Menucci, D. J. Nesbitt, Pure Appl. Chem.2011, 83, 1619. Search in Google Scholar

[36] S. Alvarez, Dalton Trans.2013, 42, 8617. Search in Google Scholar

[37] M. A. Czarnecki, Vib. Spectrosc.2012, 62, 207. Search in Google Scholar

[38] M. Hillman, J. D. Austin, Organometallics1987, 6, 1737. Search in Google Scholar

[39] R. W. Chorley, P. B. Hitchcock, M. F. Lappert, W.-P. Leung, P. P. Power, M. M. Olmstead, Inorg. Chim. Acta1992, 198–200, 203. Search in Google Scholar

[40] A. Schnepf, Z. Anorg. Allg. Chem.2006, 632, 935. Search in Google Scholar

[41] T. Schulz, D. Weismann, L. Wallbaum, R. Guthardt, C. Thie, M. Leibold, C. Bruhn, U. Siemeling, Chem. Eur. J.2015, 21, 14107. Search in Google Scholar

[42] M. S. Collins, E. L. Rosen, V. M. Lynch, C. W. Bielawski, Organometallics2010, 29, 3047. Search in Google Scholar

[43] R. W. Alder, P. R. Allen, M. Murray, A. G. Orpen, Angew. Chem., Int. Ed. Engl.1996, 35, 1121. Search in Google Scholar

[44] T. J. Boyle, L. J. Tribby, L. A. M. Ottley, S. M. Han, Eur. J. Inorg. Chem.2009, 5550. Search in Google Scholar

[45] B. Cordero, V. Gómez, A. E. Platero-Prats, M. Revés, J. Echeverría, E. Cremades, F. Barragán, S. Alvarez, Dalton Trans.2008, 2832. Search in Google Scholar

[46] M. K. Denk, M. Khan, A. J. Lough, K. Shuchi, Acta Crystallogr.1998, C54, 1830. Search in Google Scholar

[47] J. Li, M. Hermann, G. Frenking, C. Jones, Angew. Chem. Int. Ed.2012, 51, 8611. Search in Google Scholar

[48] T. Fjeldberg, H. Hope, M. F. Lappert, P. P. Power, A. J. Thorne, J. Chem. Soc. Chem. Commun.1983, 639. Search in Google Scholar

[49] J. R. Babcock, L. Liable-Sands, A. L. Rheingold, L. R. Sita, Organometallics1999, 18, 4437. Search in Google Scholar

[50] H. A. Bent, Chem. Rev.1961, 61, 275. Search in Google Scholar

[51] The electronegativity of Ge is higher than that of Sn; see, for example: J. E. Huheey, E. A. Keiter, R. L. Keiter, Anorganische Chemie, 5th edition (Ed.: R. Steudel), de Gruyter, Berlin, 2014, p. 203. Search in Google Scholar

[52] T. J. Hadlington, M. Hermann, G. Frenking, C. Jones, Chem. Sci.2015, 6, 7249. Search in Google Scholar

[53] M. J. S. Gynane, D. H. Harris, M. F. Lappert, P. P. Power, P. Rivière, M. Rivière-Baudet, J. Chem. Soc. Dalton Trans.1977, 2004. Search in Google Scholar

[54] M. C. Burla, R. Caliandro, M. Camalli, B. Carrozzini, G. L. Cascarano, L. De Caro, C. Giacovazzo, G. Polidori, R. Spagna, J. Appl. Crystallogr. 2007, 40, 609. Search in Google Scholar

[55] G. M. Sheldrick, Acta Crystallogr.2008, A64, 112. Search in Google Scholar

Received: 2017-5-29
Accepted: 2017-7-21
Published Online: 2017-9-23
Published in Print: 2017-11-27

©2017 Walter de Gruyter GmbH, Berlin/Boston