Taming Heavier Group 14 Imine Analogues: Accessing Tin Nitrogen [Sn=N] Double Bonds and their Cycloaddition/Metathesis Chemistry

Abstract

A systematic study to access stable stannaimines is reported, by combining different heteroleptic stannylenes with a range of organic azides. The reactions of terphenyl‐/hypersilyl‐substituted stannylenes yield the putative tin nitrogen double bond, but is directly followed by 1,2‐silyl migration to give Sn^II systems featuring bulky silylamido ligands. By contrast, the transition from a two σ donor ligand set to a mixed σ‐donor/π‐donor scaffold allows access to three new stannaimines which can be handled at room temperature. The reactivity profile of these Sn=N bonded species is crucially dependent on the substituent at the nitrogen atom. As such, the Sn=NMes (Mes=2,4,6‐Me₃C₆H₂) system is capable of activating a broad range of substrates under ambient conditions via 1,2‐addition reactions, [2+2] and [4+2] cycloaddition reactions. Most interestingly, very rare examples of main group multiple bond metathesis reactions are also found to be viable.

A systematic study on taming stannaimines (R₂Sn=NR) is reported by identifying a suitable supporting ligand framework for stannylenes in combination with organic azides. The nitrogen substituent of these compounds was crucial for allowing a broad range of subsequent reactivity ranging from cycloadditions to multiple bond metathesis reactions, thus highlighting the necessary balance of rendering challenging systems isolable but still reactive.

Introduction

Recent years have seen rapid development in the chemistry of heavier main group compounds containing multiple bonds, with investigations continuously uncovering new reactivity patterns, and even entry into catalytic applications. ^[1] Despite these developments, there are still significant gaps when it comes to isolable compounds, with new strategies being required to generate ′bottleable′ species, or at least to allow for clean/facile generation for investigation of subsequent chemistry. We have lately become interested in tin pnictogen multiple bonds and have reported on phosphinidene transfer to Sn^II, which allows for the isolation and characterization of a crystalline stannaphosphene. ^[2] Lighter homologues which possess multiple bonds between tin and nitrogen, often referred to as either stannaimines or tin‐imido complexes (drawing comparisons to tin as a metalloid), are also rare. To date, there is one crystallographically characterized example (I), and even this compound is stable only below −30 °C, undergoing facile intramolecular C−H activation at higher temperatures to give stannacycle II (Scheme 1, middle). ^[3]

Scheme 1.

Top: synthetic strategy to access stannaimines; middle: example of the only structurally characterized stannaimine I which undergoes intramolecular C−H activation above −30 °C to give II; Bottom: Other observed products of reactions of stannylenes with azides III–VI.

Potential routes to stannaimines involve the combination of low‐valent stannylenes and organoazides, with the latter acting as masked nitrene sources via extrusion of dinitrogen (Scheme 1, top). To date however, all literature examples following this synthetic protocol lead either to C−H activation, formal dimerization (cf. distannadiazane III), azadistanniridine formation (cf. IV), reactions with a second equivalent of the organoazide to yield a tetraazastannole (cf. V) or rearrangement to give a tin azide (cf. VI; Scheme 1, bottom).[ ³ , ⁴ ]

These observations point to the high reactivity of putative tin nitrogen double bonds, evidenced by these “self‐activation” processes or reactions with further azide. As a consequence, only in the case of compound I has any subsequent reactivity been reported, focussing on primary amines and carbonyl compounds.

Reflecting the balancing act between obtaining tractable species and systems which might still be expected to show interesting patterns of chemical reactivity, we set out to investigate Sn=N double bonds derived from the reactions between two electronically distinct stannylenes and five organoazides. This approach allows for systematic variation of the σ‐ and π‐donor capabilities of the stannylene supporting framework, and of the steric profile of the putative imine function. As such, it has enabled us to characterize stannaimines which can be handled at room temperature, thereby allowing for studies of reactivity focussing on 1,2‐addition chemistry and cycloaddition processes leading to Sn=N/Sn=E metathesis.

Results and Discussion

Reactions of Organoazides with a Silyl‐Stannylene

Initial investigations focussed on the heteroleptic terphenyl‐/hypersilyl‐substituted stannylene ^MesTerSn(Si(SiMe₃)₃) (Sn2) which can be accessed by reacting the previously reported amido/terphenyl analogue ^MesTerSn(N(SiMe₃)₂) ^[2] (Sn1) with (thf)₂K(Si(SiMe₃)₃) (^MesTer=−C₆H₃‐2,6(C₆H₂‐2,4,6‐Me₃)₂; Scheme 2, top).

Scheme 2.

Top: Synthesis of the mixed terphenyl‐/hypersilyl‐substituted stannylene Sn2 from the mixed terphenyl‐/amido‐substituted stannylene Sn1; bottom: set of azides A1–A5 used in this study.

We hypothesized that the kinetically stabilizing features associated with the steric bulk of the terphenyl ligand class ^[5] allied to the presence of a second strong σ‐donor substituent −(Si(SiMe₃)₃) ^[6] should give access to a stannylene with a wide C−Sn−Si angle and a small HOMO–LUMO gap. It was thought that this would likely render Sn2 highly reactive towards organoazides while also offering sufficient shielding towards dimerization of the product stannaimine. Sn2 was obtained as a blue crystalline solid in 87 % isolated yield and characterized by multinuclear NMR, UV/Vis spectroscopy, elemental microanalysis and single crystal X‐ray diffraction. The intense blue colour of Sn2 is consistent with the measured absorption at λ _max=638 nm (Figure S3 ^[7] ). The molecular structure of Sn2 shows a wide C1−Sn1−Si1 bond angle of 109.75(6)° which is similar to that of the recently reported system ^MesTerSn(Si( ^t Bu)₃) (113.50(14)°), ^[8] and characteristic of both significant p‐character in the HOMO and a small HOMO–LUMO gap (Figure 1).

Figure 1.

Molecular structure of ^MesTerSnSi(SiMe₃)₃ (Sn2) in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−Si1 2.6407(7), Sn1−C1 2.189(3), C1−Sn1−Si1 109.75(6).

A range of organoazides was identified for formal nitrene transfer chemistry including commercially available N₃SiMe₃ (A1), aryl‐substituted azides of different steric bulk (N₃Mes (A2), N₃Dipp (A3), and N₃ ^MesTer (A4)) and N₃Quin (A5) which features an additional nitrogen donor functionality designed to stabilize the putative stannaimine through dative Sn−N bond formation (Scheme 2, bottom). Accordingly, the reactivity of Sn2 with each of A1–A5 was scoped (Scheme 3).

Scheme 3.

Reactions of Sn2 with A1–A5.

The reaction of Sn2 with N₃SiMe₃ (A1) is accompanied by immediate N₂ evolution and the formation of two products according to ¹H NMR spectroscopy, even when the reaction is initiated at low temperatures (Figure S5). ^[7] Colourless crystals of the minor product suitable for single crystal X‐ray diffraction could be isolated, which revealed the formation of a product derived from intramolecular C(sp³)−H activation (Sn4a) in which one of the flanking mesityl methyl groups has been activated, resulting in the formation of N−H and Sn−CH₂ functionalities (Figure 2).

Figure 2.

Molecular structure of the C(sp³)−H activation product Sn4a in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms except for H1, H24A and H24B have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−N1 2.0445(14), Sn1−C1 2.1754(16), Sn1−C24 2.1715(16), Sn1−Si1 2.5915(4), C21−C24 1.500(2), C1−Sn1−Si1 125.04(4), C1−Sn1−N1 106.37(6).

The generation of Sn4a can be hypothesized as proceeding via the formation of the putative stannaimine ^MesTerSn(Si(SiMe₃)₃)=NSiMe₃ followed by a C−H activation step comparable to the previously reported formation of II starting from I (Scheme 1). ^[3]

Structurally, the central tin atom in Sn4a is four‐coordinate, and in a distorted tetrahedral coordination environment. Both, the Sn1−N1 and Sn1−C24 bond lengths (2.0445(14) Å and 2.1715(16) Å) are diagnostic of the respective single bonds (Sn−C 2.15 Å, Sn−N 2.11 Å), ^[9] and are in good accordance with the only comparable structurally characterized C−H activation product, derived from the reaction of (Mes*)₂Sn with N₃Dipp (cf. 2.084(3) Å and 2.137(4) Å). ^[4b] In C₆D₆ solution the ¹H NMR spectrum shows the expected diastereotopic doublet signals located at δ=2.59 and 2.88 ppm (² J _H,H=11.2 Hz) associated with the methylene group. The NH hydrogen atom gives rise to a resonance at δ=−0.99 ppm. The nuclei close to tin show the expected characteristic tin satellites, with these being seen for both the hydrogen atoms of both the methylene group and the NH moiety (e.g. ² J ${}_{{}^{119}\mathrm{Sn},\mathrm{H}}$ =93.8 Hz, ² J ${}_{{}^{117}\mathrm{Sn},\mathrm{H}}$ =71.7 Hz for one of the methylene group signals; Figure S9). ^[7]

Although we were not able to isolate the major product of the reaction in a pure bulk state due to omnipresent co‐crystallization of Sn4a, colorless needles of this compound were identified (and separated by crystal picking) which were suitable for single crystal X‐ray diffraction. These were shown to contain the amido/terphenyl stannylene ^MesTerSn{N(SiMe₃)Si(SiMe₃)₃} (Sn3a) formed by insertion of the N(SiMe₃) function into the Sn−Si bond of Sn2 (Scheme 3 and Figure S11). In this context, the reactions of Sn2 with the aryl‐substituted azides N₃Mes (A2) and N₃Dipp (A3) also result in the clean formation of the formal nitrene insertion products ^MesTerSnN(Aryl)Si(SiMe₃)₃ (Aryl=Mes (Sn3b), Dipp (Sn3c)) (Scheme 3). Both Sn3b and Sn3c could be characterized by single crystal X‐ray diffraction (Figure 3 and Figure S18). Structurally, the Sn−N bond lengths of (on average) 2.09 Å for Sn3a–c are typical of single bonds and the transition from Sn−Si bonding in the starting material Sn2 to Sn−N bonding in Sn3a–c is accompanied by a widening of the C^MesTer−Sn−N bond angle from 109.75(6)° to 112.03(8)° (Sn3c). Since substitution of a strong σ‐donor substituent by a weaker one is expected to lead to more bent carbenoid systems, steric effects are considered to be the primary reason for this observation.

Figure 3.

Molecular structure of ^MesTerSnN(Mes)Si(SiMe₃)₃ (Sn3b) in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−N1 2.079(2), Sn1−C27 2.234(3), C27−Sn1−N1 109.49(9), C1−N1−Sn1 126.02(17), C1−N1−Si1 118.34(17), Si1−N1−Sn1 115.64(11).

Of note mechanistically is the fact that each of Sn3a–c features an aryl(silyl)amido ligand, formed via silyl group migration to the putative Sn=N functionality ‐ highlighting a novel reaction pathway between stannylenes and azides. The formation of Sn3a–c is hypothesized to involve initial formation of the target stannaimine followed by facile silyl migration. Computationally, we find that this pathway is both kinetically and thermodynamically viable (Figure 4), with an activation barrier of 21 kcal mol⁻¹ (for Sn3b) and an overall free energy change (ΔG=−41.3 kcal mol⁻¹) that is highly exergonic.

Figure 4.

Reaction coordinate for the formation of Sn3b from the presumed stannaimine (all energies, G, in kcal mol⁻¹).

We also examined the use of sterically more demanding organoazide reagents and systems featuring additional pendant functionality. However, with Sn2 and N₃ ^MesTer (A4), for example, no reaction could be observed even after prolonged heating of the reaction mixture to 80 °C (Figure S19). On the other hand, the reaction of Sn2 with N₃Quin(A5) in a 1 : 1 stoichiometry resulted in significant amounts of unreacted Sn2. The reaction was therefore repeated with an excess of A5, resulting in further gas evolution and the formation of a clear orange solution. After work‐up, a small quantity of orange crystals suitable for single crystal X‐ray diffraction were obtained from a saturated n‐hexane solution at −30 °C. Crystallographic analysis reveals that a total of three quinoline moieties are incorporated in the product ^MesTerSn(NHQuin)(QuinNSi(SiMe₃)₂NQuin) (Sn5; Figure 5). As can be seen from Figure 5, the reaction between Sn2 and A5 results in deconstruction of the hypersilyl functionality in Sn2 to generate a Si(SiMe₃)₂ moiety which bridges two azide‐derived N atoms to form a chelating tridentate N,N,N‐ligand in the product Sn5. The third coordination site is the result of one quinoline nitrogen atom additionally coordinating to tin (2.3256(15) Å). The other two tin nitrogen linkages have separations typical of single bonds (2.0808(13) Å and 2.1162(12) Å). The third equivalent of quinoline azide is coordinated as the corresponding amido ligand, so that formally a Me₂Si=CH₂ unit is the leaving group in this reaction. Together with the remaining terphenyl ligand, the tin center is in a distorted tetragonal pyramidal coordination environment.

Figure 5.

Molecular structure of ^MesTerSn(NHQuin)(QuinNSi(SiMe₃)₂NQuin) (Sn5) in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−N2 2.0719(14), Sn1−N3 2.3256(15), Sn1−N4 2.0808(13), Sn1−N6 2.1162(12), Sn1−C1 2.1765(15), N4−Si1 1.7965(14), N6−Si1 1.7685(14), N2−Sn1−N6 101.39(5), N6−Sn1−N4 72.06(5), N4−Sn1−N3 74.31(6), C1−Sn1−N2 105.44(6), C1−Sn1−N6 117.90(5), C1−Sn1−N4 108.10(6), C1−Sn1−N3 110.20(5).

Synthesis of Stannaimines from an Amido‐Stannylene

None of the above reactions using (silyl)stannylene Sn2 allowed for the isolation of a stannaimine, with formal nitrene insertion into the Sn−Si bond being commonly encountered. In order to circumvent this problem, we therefore considered the use of stannylene precursor bearing a pendant amido (rather than silyl) substituent. Although likely to be less reactive kinetically, we perceived (from a thermodynamic perspective) that the π donor characteristics of the amido group would complement the strongly σ‐donating terphenyl moiety to stabilize the electron‐deficient stannaimine Sn centre through both inductive and resonance effects. The combination of strong σ and strong π donor substituents has recently been employed to great effect in the stabilization of silanone compounds, R₂Si=O. ^[10]

Systems of type Sn3—bearing sterically very demanding amido substituents—are unreactive towards a second equivalent of the azide substrate, so we targeted sterically less demanding analogues. The stoichiometric reaction of Sn1 with N₃SiMe₃ (A1) results in the retention of half an equivalent of unreacted Sn1. Addition of a second equivalent of A1 gives clean conversion to a single product which was unambiguously identified as the azido/bis(amido) Sn^IV compound ^MesTerSn{N(SiMe₃)₂}₂N₃ (Sn6) by single crystal X‐ray diffraction (Scheme 4 and Figure S25). N−Si bond cleavage of a second molecule of N₃SiMe₃ at a putative stannaimine is assumed to be the reaction path, in accordance with the previously reported reaction of Sn{N(SiMe₃)₂}₂ with N₃SiMe₃ which yields Sn{N(SiMe₃)₂}₃N₃. ^[4c] The structural parameters for Sn6 are in good agreement with the latter system, and a detailed discussion is not warranted at this point.

Scheme 4.

Reactions of Sn1 with A1–A5.

The reactions of Sn1 with the aryl‐substituted azides N₃Mes (A2) or N₃Dipp (A3) occur readily at room temperature, with immediate gas evolution being observed, accompanied by a color change from bright orange/red to dark red. NMR analysis in each case reveals clean formation of a single product. However, on standing at room temperature overnight, significant numbers of additional signals were observed to grow in, accompanied by decolorization of the solution. The initially formed products could, however, be obtained as dark red crystalline materials by crystallization from aliphatic hydrocarbons at −30 °C. In case of the reaction of Sn1 with A2, these crystals were suitable for single crystal X‐ray diffraction, and crystallographic studies reveal the formation of the target stannaimine Sn7a, clearly evident by the short Sn1−N1 bond of 1.9354(16) Å and the sum of covalent radii around the tin center of 360° (Figure 6).

Figure 6.

Top: Molecular structure of ^MesTerSn(N(SiMe₃)₂)=NMes (Sn7a) in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−N1 1.9354(16), Sn1−N2 2.0455(16), Sn1−C1 2.1481(18), N1−C25 1.401(2), C1−Sn1−N1 116.46(7), N1−Sn1−N2 125.17(7), C1−Sn1−N2 118.34(7); bottom: Excerpt of the molecular structure of Sn7a in the crystal showing the planarity of the central unit.

To gain further insight into the bonding situation in stannaimine systems, we carried out quantum chemical calculations on the less computationally demanding derivative Sn7a. The HOMO‐1 and LUMO are best described as Sn=N π‐bonding and π*‐antibonding orbitals, respectively and the Kohn–Sham molecular orbitals of Sn7a reveal that the Sn−N interaction is largely polarized towards nitrogen (Figure 7, top). This is further demonstrated by Natural Bonding Orbital (NBO) analysis. Two NBOs contribute to the Sn−N bonding in Sn7a, both of which are highly polarized (78 % each) towards nitrogen and account for a total of 2.06 electrons between tin and nitrogen (Table S8). This finding is consistent with the calculated Sn−N Wiberg bond index, which was found to be only slightly greater than unity (1.06) (Figure 7, bottom). Taken together these data suggest that the bonding between tin and nitrogen is highly polarized and possesses a high degree of ionic character, consistent with the significant electronegativity difference between the two atoms. This polarity can also be seen in the NBO charges calculated for Sn7a (Sn, +2.26; N, −1.17; Figure 7, bottom). Furthermore, the electron density at the bond critical point (BCP) located between Sn and N using Atoms in Molecules (QTAIM) analysis is relatively low [ρ(r)=0.1338 e Å⁻³] (Figure S82). This finding, along with the positive Laplacian of the electron density at this point [∇² ρ(r)=0.5142 e Å⁻⁵] are indicative of a highly electrostatic interaction. Nonetheless, the non‐spherical nature of the BCP (ϵ=0.122) is consistent with a degree of π delocalization between these two atoms.

Figure 7.

Top: Selected molecular orbitals of the optimized structure of Sn7a; bottom: Optimized structure of Sn7a with the tin‐nitrogen Wiberg bond index (shown in black and selected natural atomic charges (shown in red).

The observation of additional signals in the ¹H NMR spectra of Sn7a and Sn7b over time is due to intramolecular C(sp³)−H activation across the Sn=N bond. This benzylic C−H activation (effectively deprotonation) by the basic imido group within the polar Sn=N unit results in the formation of the corresponding amido functionalities, and coordination of a methylene group to the tin center to give Sn4b,c (Scheme 4). Sn4b,c could both be characterized by single crystal X‐ray diffraction and the molecular structures are shown in Figures S32 and S38. The structural parameters of these six‐membered stannacycles are consistent with those measured for Sn4a (Figure 2). In contrast to the only other isolated stannaimine [Sn(N(SiMe₃)₂)=NDipp], ^[4c] Sn7a,b can be handled at room temperature for prolonged periods allowing for systematic investigation of their reaction chemistry. In this context, storage of a benzene solution of Sn7a for four hours at room temperature yields only 10 % of Sn4b.

Finally, to probe whether an internal tethered donor unit could be employed to further stabilize a stannaimine, the reaction of Sn1 with N₃Quin (A5) was investigated. On combining the two compounds in benzene, an immediate colour change to dark green was observed, with ¹H NMR analysis confirming clean formation of a single product. This species was shown by X‐ray crystallography to be the corresponding stannaimine Sn8, as expected stabilized by dative coordination of the pendant quinoline nitrogen atom (Scheme 4, Figure 8). As a result of the dative coordination of the quinoline nitrogen atom, the tin center in this case is four‐coordinate and in a distorted tetrahedral coordination environment. The Sn1−N1 bond length of 2.022(4) Å is significantly elongated when compared to stannaimine Sn7a (1.9354(16) Å). However, despite the tin center being datively stabilized and four‐coordinate, similar intramolecular C(sp³)−H activation is observed at room temperature, as seen for Sn7a,b, yielding the corresponding complex Sn4d which was also characterized by multinuclear NMR spectroscopy and single crystal X‐ray diffraction.[ ⁷ , ¹¹ ]

Figure 8.

Top: Molecular structure of ^MesTerSn(N(SiMe₃)₂)=NQuin (Sn8) in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−N1 2.022(4), Sn1−N2 2.198(3), Sn1−N3 2.032(3), Sn1−C1 2.158(4), N1−Sn1−N2 93.09(13), N1−Sn1−N3 114.71(14); bottom: Excerpt of the molecular structure of Sn8 in the crystal showing the coordination environment at the tin center.

Reactivity Studies of Stannaimines

For the only previously reported example of a stannaimine (I), 1,2‐addition reactions with primary amines and cycloaddition reactions with carbonyl compounds were investigated, but no structural data for cycloaddition products were reported. ^[3b] With this in mind, we perceived that the more stable stannaimines Sn7a,b and Sn8 might provide a convenient basis for systematically investigating the reactivity of this substance class. In a first exploration of reactivity, Sn7a,b and Sn8 were reacted with phenylacetylene in attempts to generate the corresponding 1,2‐addition products.

Interestingly, although the Dipp‐ and Quin‐substituted derivatives Sn7b and Sn8 react with phenylacetylene, these reactions are slow and intramolecular C−H activation competes, so that only mixtures of products could be obtained (Figure S46). This emphasizes the fine balance needed to tune the system to render it isolable, but at the same time sufficiently reactive towards external substrates. By contrast, Mes‐substituted derivative Sn7a reacts rapidly and cleanly with phenylacetylene to give the corresponding 1,2‐addition product ^MesTerSn{N(SiMe₃)₂}(CCPh)(NHMes) (Sn10a) (Scheme 5), implying that this specific substitution pattern is an ideal compromise, allowing for both the isolation of the reactive tin nitrogen double bond and clean reactivity towards external substrates. The reduced steric profile of the Mes substituent (cf. Dipp) presumably allows faster kinetics in intermolecular reactivity. A similar result was obtained by employing an aliphatic terminal alkyne to give the corresponding C−H activation product ^MesTerSn(N(SiMe₃)₂)(CC₄H₈Cl)(NHMes) (Sn10b). Sn10a and Sn10b were both characterized by single crystal X‐ray diffraction (Figure 9 (top) and Figure S51). The regiochemistry of C−H addition in each case is consistent with the polarity of the Sn=N bond revealed computationally, and with (effectively) deprotonation of the terminal alkyne by the strongly basic imide nitrogen.

Scheme 5.

Reactivity of stannaimine Sn7a.

Figure 9.

Top: Molecular structure of ^MesTerSn(N(SiMe₃)₂)(NHMes)CCC₄H₈Cl (Sn10b) in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms (except H1) have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−N1 2.057(2), Sn1−N2 2.051(2), Sn1−C1 2.181(3), Sn1−C40 2.087(3), C40−C41 1.189(4), C41−C42 1.468(4), N1−Sn1−N2 106.37(9), N1−Sn1−C40 104.17(10); bottom: Molecular structure of the [2+2] cycloaddition product Sn11a in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−N1 2.021(2), Sn1−N2 2.0976(19), Sn1−O1 2.0361(16), Sn1−C1 2.140(2), O1−C30 1.355(3), N2−C30 1.390(3), O2−C30 1.214(3), N1−Sn1−N2 112.46(8), O1−Sn1−N2 65.10(7).

In the context of cycloaddition chemistry, a range of heteroallenes (CO₂, ⁱ PrNCN ⁱ Pr, CS₂, SCNR (R= ^t Bu, Xyl)) was reacted with Sn7a. The addition of 1 bar of CO₂ or one equivalent of N,N′‐diisopropylcarbodiimide results in an immediate color change of the solution to bright yellow. In each case ¹H NMR analyses confirm the formation of a single product, namely the [2+2] cycloaddition products Sn11a,b, whose connectivity could be unambiguously verified by single crystal X‐ray diffraction (Figure 9 (bottom)). These complexes are the first examples of structurally characterized [2+2] cycloaddition products derived from stannaimines. Comparing their structures with those of related compounds derived from germaimines show that they are in good accordance (near planarity of the central four‐membered ring, exocyclic double bond character, and behavior as a dianionic ligand to the tetrel center). ^[12] The general scalabilty of the synthesis of stannaimine Sn7a and compounds derived therefrom was demonstrated starting with 500 mg of Sn1. Since Sn1—although comparatively slowly—undergoes intramolecular C−H activation, we recommend to always synthesize it in situ, determine purity by ¹H NMR spectroscopy of an aliquot and directly react it with the substrate. Accordingly, Sn1 was converted to Sn7a and further reacted exemplarily with iPrNCNiPr to give 628 mg of Sn11b (which corresponds to an isolated yield of 87 %). ^[7]

Interestingly, the reaction of Sn7a with an excess of the heavier (but usually more reactive) CO₂ analogue carbon disulfide is significantly slower, and therefore significant amounts of the competing C−H activation product Sn4b can be observed by in situ ¹H NMR monitoring (Figure S66). Despite the formation of Sn4b, however, we succeeded in crystallizing the formal [4+2] cycloaddition product of Sn7a with CS₂ (Sn12; Scheme 5). The molecular structure of Sn12 reveals that cycloaddition occurs between one of the C=S double bonds of CS₂ and the heterodiene motif (Sn=N−C(sp²)=C(sp²)), comprised of the tin‐nitrogen double bond and the adjacent (localized) carbon‐carbon double bond of the mesityl substituent. Comparable reactivity has been observed in reactions of a germanimine with carbonyl compounds and alkenes, respectively. ^[12] Notably, although both the tin center and the ortho carbon atom of the dearomatized mesityl group are stereogenic centers, Sn12 is formed diastereomerically pure as a mixture of the R,S and S,R enantiomers (as determined by single crystal X‐ray diffraction). The molecular structure of Sn12 is shown in Figure 10 (top).

Figure 10.

Top: Molecular structure of the [4+2] cycloaddition product Sn12 in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1A−N1A 2.036(11), Sn1A−N2A 2.058(9), Sn1A−C1A 2.152(11), Sn1A−S1A 2.438(3), S1A−C56A 1.775(13), S2A−C56A 1.649(14), N1A−C50A 1.269(17), C50A−C51A 1.559(19), C51A−C52A 1.502(18), C52A−C53A 1.31(2), C53A−C54A 1.48(2), C54A−C55A 1.37(2), C50A−C55A 1.483(18), C51A−C56A 1.503(19), N1A−Sn1A−S1A 98.9(3), N1A−Sn1A−C1A 110.7(4); bottom: Molecular structure of (^MesTerSn(N(SiMe₃)₂)S)₂ (Sn13) in the crystal. Thermal ellipsoids are drawn at the 50 % probability level (hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [deg]: Sn1−S1 2.4218(5), Sn1−S2 2.4218(5), Sn2−S1 2.4225(5), Sn2−S2 2.4204(4), Sn1−N1 2.0627(16), Sn1−C7 2.2010(18), N1−Sn1−C7 115.68(7), Sn1−S1−Sn2 87.965(15), S1−Sn1−S2 91.990(16).

As a consequence of the dearomatization of the mesityl substituent, carbon‐carbon double bonds are localized between C52A and C53A and between C54A and C55A, respectively (1.31(2) Å and 1.37(2) Å). The exocyclic N1A−C50A bond length of 1.269(17) Å is typical of a double bond with all the other bond lengths within the stannacycle being characteristic of single bonds. The exocyclic C56A−S2A bond length of 1.649(14) Å is significantly shorter than the C56A−S1A bond length of 1.775(13) Å within the six‐membered ring system, consistent with a description of the former as a double bond.

Due to diverging reactivity of Sn1 towards these heteroallenes—in particular towards ⁱ PrNCN ⁱ Pr and CS₂—the reactivity of the Sn=N double bond towards isothiocyanates (which contain both C=N and C=S functionalities) was viewed as being of significant interest. Accordingly, the reaction of the aryl‐substituted isothiocyanate SCNXyl with Sn1 was investigated under comparable conditions. In this case, the reaction can be shown by a combination of spectroscopic and crystallographic methods to lead to the selective formation of the [2+2] cyclo‐addition product Sn11c (Scheme 5). The regioselectivity of this transformation was unambiguously confirmed by single crystal X‐ray diffraction, showing that cycloaddition has occurred between the Sn=N double bond of the stannaimine and the weaker C=S double bond of the isothiocyanate (Figure S74). Interestingly, the reaction of Sn1 with the aliphatic isothiocyanate SCN ^t Bu at room temperature results in clean formation of the sulfur‐bridged dimer {^MesTerSn(N(SiMe₃)₂)S}₂ (Sn13), as verified by single crystal X‐ray diffraction (Figure 10 (bottom)), together with the carbodiimide MesN=C=N ^t Bu, implying that an overall metathesis process has occurred, with accompanying dimerization of the stannathione. The formation of Sn13 could also be achieved starting from Sn11c by heating a benzene solution of the Sn,N,C,S‐containing metallacycle at 80 °C for one hour, or by prolonged storage at room temperature (cf. Figure S73). In this case trapping of the intermediate Sn11c which links stannaimine Sn7a (plus isothiocyanate) with thiostannone Sn13 (plus carbodiimide) implies that the metathesis process occurs via a similar mechanism to more well‐known d‐block systems.

Conclusions

We have shown that stannaimines can be stabilized systematically by combining an organoazide with a heteroleptic stannylene featuring one strong σ donor (terphenyl) and one π donor substituent (N(SiMe₃)₂). This allows for systems which can (uniquely) be handled under ambient conditions. It was further shown that, the substituent at the stannaimine nitrogen functionality is crucial for allowing subsequent reactivity. This detailed study allowed for the realisation of 1,2‐addition reactions, [2+2] and [4+2] cycloaddition reactions and a very rare example of main group double bond metathesis chemistry at a stannaimine template. By contrast, using heteroleptic stannylenes with two strong σ donor ligands (terphenyl/hypersilyl) in reactions with organic azides seem to give the putative tin nitrogen double bond but subsequent silyl migration is favoured significantly, so that a series of novel heteroleptic stannylenes with terphenyl and bulky amido ligands are formed. ^[13]

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

M. Fischer, M. M. D. Roy, L. L. Wales, M. A. Ellwanger, C. McManus, A. F. Roper, A. Heilmann, S. Aldridge, Angew. Chem. Int. Ed. 2022, 61, e202211616; Angew. Chem. 2022, 134, e202211616.

Data Availability Statement

Data in support of this publication is publicly available through the Oxford Research Archive (ORA).