En Route to a Molecular Terminal Tin Oxide

In the pursuit of terminal tin chalcogenides, heteroleptic stannylenes bearing terphenyl- and hexamethyldisilazide ligands were reacted with carbodiimides to yield the respective guanidinato complexes. Further supported by quantum chemical calculations, this revealed that the iso-propyl-substituted derivative provides the maximum steric protection achievable. Oxidation with elemental selenium produced monomeric terminal tin selenides with four-coordinate tin centers. In reactions with N2O as oxygen transfer reagent, silyl migration toward putative terminal tin oxide intermediates gave rise to tin complexes with terminal —OSiMe3 functionality. To prevent silyl migration, the silyl groups were substituted with cyclohexyl moieties. This analogue exhibited distinctively different reactivities toward selenium and N2O, yielding a 1,2,3,4,5-tetraselenastannolane and chalcogenide-bridged dimeric compounds, respectively.


■ INTRODUCTION
Compounds featuring the carbonyl functionality (R 2 C�O) such as aldehydes, ketones, and amides are fundamental components in organic chemistry.Despite being thermodynamically robust, these functional groups are straightforward to functionalize, given the polarity of the C�O motif.Due to electronegativity differences, their heavier tetrel analogues R 2 E( 14)�O (E(14) = silicon, germanium, tin, and lead) exhibit even greater charge separation, which increases down Group 14. 1 Furthermore, significantly weaker π overlap between oxygen and the heavier Group 14 elements results in terminal E(14)�O double bond fragments which are thermodynamically unstable, and often adopt a polarized/ylidic form (E(14) + − O − ).The inherent charge disparity in heavier Group 14 carbonyl compounds cannot be quenched effectively by π bond formation, resulting in high reactivity.This is frequently manifested in self-quenching through di-, oligo-, and polymerization reactions; as well as inter and intramolecular C−H activation processes. 2Consequently, heavier Group 14 carbonyl compounds have been elusive species in the past, leaving ample room for the further development of their chemistry.
Over a century ago, Kipping aimed to synthesize the lightest heavier carbonyls, known as silanones (R 2 Si�O).However, the material produced was later identified to be a polysiloxane, a now omnipresent class of polymers and illustrative of one of the typical self-quenching reactivities, vide supra. 3Despite being detected in low-temperature matrices in the 1980s, 4 it was not until 2007 that the first stable silacarbonyl compounds were reported, utilizing external Lewis acid and/or Lewis base stabilization (Figure 1, I). 5 This strategy paved the way for the isolation of main group carbonyl species across the p-block elements.
Interestingly, the first heavier Group 14 analogue of a ketone, devoid of any acid−base stabilization, was reported for germanium instead of silicon.Tamao, Matsuo, and co-workers achieved this milestone in 2012 with the terminal monomeric germanone, 6a having paved the way for further examples featuring the terminal Ge�O moiety (Figure 1, II). 6 Within the next seven years, stable compounds featuring the "free" Si� O functionality were reported by Filippou, 7 Kato, 8 and Inoue 9 (Figure 1, III−V).Iwamoto and co-workers were finally able to tame a cyclic dialkylsilanone, bearing a three-coordinate silicon center with an unperturbed Si�O double bond, utilizing a kinetic stabilization strategy just four years ago (Figure 1, VI). 10 Such achievements have shifted the perception of these compounds from "laboratory curiosities" to versatile tools for exploring classical carbonyl chemistry with the heavier Group 14 elements and uncovering novel reactivity patterns and applications in oxide ion transfer chemistry. 11ne intriguing question that remains is whether "true" stannanones/terminal tin oxides are synthetically accessible. 11he most closely related isolable complexes in literature involve formal "SnO" and "PbO" units trapped by multiple Lewis acid and Lewis base sites. 12erein, we report on our current progress in isolating terminal tin chalcogenides en route to the isolation of a terminal tin oxide.

■ RESULTS AND DISCUSSION
Heteroleptic stannylenes, comprising one terphenyl and one hexamethyldisilazido ligand of the general type Ar TerSn{N-(SiMe 3 ) 2 } [1a: Aryl(Ar) = Mes (2,4,6-Me 3 C 6 H 2 ), 1b: Ar = Dipp (2,6-i Pr 2 C 6 H 3 )], have recently been found to facilitate the isolation of rare instances of terminal stannaphosphenes and stannaimines. 13However, when 1a,b are subjected to typical oxygen transfer reagents, e.g., N 2 O or Me 3 NO, they yield complex reaction mixtures or undergo decomposition.We postulated that a modified ligand set, featuring a threecoordinate tin atom supported by an intramolecular Lewis base, might provide the necessary electronic and steric characteristics to enable the formation of a heteroleptic stannylene capable of generating terminal tin chalcogenides.In this context, upon inspection of the Frontier Kohn−Sham molecular orbitals of 1, it becomes evident that 1 can act as an ambiphile, capable to react nucleophilically either at the tin or nitrogen lone pair observed in the HOMO, and electrophilically at the tin p-orbital observed in the LUMO, which is the major contributor (77%) to the molecular orbital (Scheme 1, A).This consideration, in conjunction with the well-known behavior of carbodiimides, which tend to formally insert into tetrel−amido bonds due to their propensity to act as nucleophiles at nitrogen and electrophiles at carbon, 14 further lays the foundation of our rationale.
Given the selectivity for the formation of 2a,b on the nature of the carbodiimide substitution pattern, the overall Gibbs free energy of reactions between 1 and carbodiimides of varying R (R = i Pr, t Bu, SiMe 3 ) were explored computationally by density functional theory at the BP86-D3BJ/def2-TZVP/Benzene-(PCM) level of theory (Figure S60, Table S12).This found unanimously that only reactions of R = i Pr were exergonic (ΔG 298 : a, −14.5 kcal mol −1 ; b, −14.2 kcal mol −1 ), presumably due to the increased steric hindrance producing thermodynamically unfavorable products with strained conformations.The endergonic energies for reactions with carbodiimides featuring R = t Bu, SiMe 3 also correlate well to the experimental findings, which showed no conversion to the respective tin guanidinates. 15a and 2b were characterized by multinuclear nuclear magnetic resonance (NMR) spectroscopy, bulk purity verified by elemental microanalysis, and, in the case of 2a, single crystal X-ray diffraction (Scheme 1, C).
The molecular structure of 2a shows the three-coordinate tin atom whose coordination environment is distorted trigonal pyramidal [largest bond angle, 114.73( 7 In order to assess the suitability of the selected ligand framework for stabilizing terminal chalcogenides in a broader context, 2a,b were reacted with stoichiometric amounts of elemental selenium (Scheme 2, A).Although no reactions could be observed at room temperature, heating of the reaction mixtures to 70 °C for several hours results in consumption of both starting materials, color changes to a more intense yellow, and main formation of single products according to 1 H NMR spectroscopy.It is worth noting that 2a does not react with elemental tellurium in benzene or tetrahydrofuran neither at room temperature nor elevated temperatures of up to 100 °C.
Crystals of 3b suitable for single crystal X-ray diffraction analysis were obtained from a saturated n-hexane solution at −30 °C, confirming the formation of a terminal tin selenide with a four-coordinate tin center, whose coordination environment is best described as distorted tetrahedral (τ 4 = 0.79 18 ) (Scheme 2, B).The terminal tin−selenium bond length of 2.3818(6) Å is in good agreement with the double bond covalent radii of the respective elements (Σ cov Sn−Se 2.56 Å, Σ cov Sn = Se 2.37 Å) and is on the shorter end of the tin−selenium bonds reported to date (c.f. Figure S57 and Tbt(Ditp)SnSe 2.373(3) Å; 19d Tbt = 2,4,6tris[bis(trimethylsilyl)methyl]phenyl, Ditp = 2,2′-diisopropylm-terphenyl-2′-yl).Generally, 3a,b account for the first monomeric and neutral terminal tin selenides with fourcoordinate tin centers, with most literature-known derivatives bearing five-coordinate tin centers.14c, 19,20 The aforementioned example with tin in a trigonal planar coordination environment is not obtained directly from the reaction of the respective stannylene precursor with selenium due to the initial formation of a 1,2,3,4,5-tetraselenastannolane and has to be further reacted with three equivalents of triphenylphosphine.19d Although the structural parameters of 3b are indicative of pronounced double bond character of the Sn−Se bond and are usually the preferred way to describe these complexes in the literature, 14c,19 the bonding of 3a was investigated by computational methods.A Wiberg bond index of 1.35 and natural charges of +1.61 (Sn)  and −0.78 (Se) were found, indicating a polarized interaction with a formal order between a single and double bond (Scheme 2, C).Furthermore, natural bond orbital (NBO) analysis was employed and found only two NBOs to describe the Sn−Se interaction with a total of 2.05 electrons, both of which polarized toward Se (60.8%), indicating a zwitterionic single bond.This is consistent with the Sn−N interaction in our previously reported stannaimine systems, taking into account the difference in electronegativity between N and Se.13b There are also three NBOs describing lone pairs at Se, accounting for six electrons, one of which is delocalized (86.7% localization on Se).Natural localized molecular orbital analysis shows that a significant amount (12.3%) of the delocalization tail resides in a p-orbital overlap with Sn, explaining the increased Sn−Se bond order above what would be expected for a single bond.
The description of a Sn δ+ −Se δ− single bond with partial charges is in agreement with a weak π-acceptor character of the tin atom and is usually reflected by an upfield shift in the 77 Se NMR spectrum (shielded selenium).19k The observed 77  Having shown that the chosen supporting ligand set at tin is capable of stabilizing terminal tin selenides, terminal tin oxide complexes were targeted next.By pressurizing a C 6 D 6 solution of 2a,b with 1 bar of nitrous oxide at room temperature, and following the reaction by 1 H NMR spectroscopy, clean formations of single species over the course of approximately 5 h are observed (Figures S26 and S30). 15Subsequent workup led to the isolation of colorless solids and liquid injection field desorption ionization mass spectrometry (LIFDI-MS) of the Mes Ter-substituted derivative is in agreement with the envisioned net oxygen transfer to precursors 2a,b. 15e 119 Sn NMR chemical shifts of the obtained compounds are upfield shifted [δ 119 Sn = −209.9(4a) and −203.7 (4b) ppm] when compared to the starting material [δ 119 Sn = 90.5 (2a) and 95.2 (2b) ppm] and are in the same range as observed for terminal selenides 3a,b (vide supra).Although the solution NMR and LIFDI-MS data generally support the formation of terminal tin oxides, the 29 Si{ 1 H} NMR data indicate different product formation.For the starting material 2a,b, as well as the terminal tin selenides 3a,b, the 29 Si{ 1 H} NMR spectra each exhibit two signals in close proximity, as expected when both trimethylsilyl groups are located at nitrogen [δ 29 Si{ 1 H} = 3.9 and 7.8 ppm (2a), 4.7 and 7.8 ppm (2b), 4.7 and 10.1 ppm (3a), 5.5 and 10.9 (3b) ppm].In contrast, the 29  The formal 1,4-silyl migration observed in this study is suggested to occur through a putative terminal tin oxide intermediate, a hypothesis supported by comparable silyl migrations observed in the context of terminal silanones and in our recently reported stannaimine study.9a,13b,22 Given this reaction behavior, our focus shifted to a heteroleptic terphenyl-/guanidinato-tin system devoid of silyl groups.Initially, we synthesized the heteroleptic terphenyl-/ dicyclohexylamido-stannylene 6 through a salt metathesis reaction between Mes TerSnCl (5) 23 and freshly prepared LiNCy 2 (Scheme 4, A). 15 Characterization of 6 was carried Inorganic Chemistry out in solution using NMR spectroscopy and in the solid state by single crystal X-ray diffraction. 15dditionally, 6 was found to react with N,N′-diisopropylcarbodiimide, yielding the corresponding guanidinato complex 7. Notably, this compound can be conveniently synthesized in a one-pot procedure starting from compound 5 (Scheme 4, A).The analytical data for 7 show only marginal differences from those of 2a,b, so that a detailed discussion is omitted at this stage (cf.Scheme 4, B for the structural data). 17nterestingly, when 7 was reacted with equimolar amounts of elemental selenium at elevated temperatures (reaction did not initiate at room temperature), most of the compound remained unreacted as confirmed by 1 H NMR spectroscopy.Simultaneously, the elemental selenium was entirely consumed, as demonstrated by the absence of any remaining gray precipitate in the reaction mixture.Accordingly, 7 was reacted with an excess of elemental selenium until 7 was completely consumed (Figure S47).From the respective crude 1 H NMR spectrum, it was already evident that small amounts of i PrN�C�N i Pr were liberated.After multiple crystallization attempts, we eventually succeeded in growing both yellow and orange crystals suitable for single crystal X-ray diffraction.The orange crystalline material revealed the formation of 1,2,3,4,5-tetraselenastannolane 8 (Scheme 4, A,C).The structural data within the SnSe 4 linkage is in good agreement with the also structurally characterized 1,2,3,4,5tetraselenastannolane Tbf(Mes)SnSe 4 (Sn−Se av 2.58 Å, Se− Se av 2.31 Å). 19d,24 The coordination environment at tin is best described as square pyramidal, according to the structural parameter τ 5 (0.04). 25 The identity of the yellow crystalline material explains why free i PrN�C�N i Pr was detected in the crude NMR spectra of the reaction and is linked to the formation of the selenium-bridged dimer (1,3,2,4-diselenadistannetane) Mes TerSn(NCy 2 )(μ-Se 2 )Sn{N( i Pr)C(NCy 2 )N( i Pr)} Mes Ter (9) with the terphenyl substituents in a cis configuration (Scheme 4, A,D).To the best of our knowledge, the release of carbodiimides from guanidinato ligands upon addition of another substrate has not been observed so far.Although 8 and 9 invariably cocrystallized in our hands, small amounts of 8 could be separated and further analyzed by elemental microanalysis, 1 H and 119 Sn{ 1 H} NMR spectroscopy (δ119Sn{ 1 H} = −252.9ppm) (Figures S46, S48 and S49). 15he reactivity of 2a,b and 7 toward elemental selenium differs significantly despite comparatively small differences in backbone substitution patterns.
In this context, we finally investigated the reactivity of 7 toward N 2 O.
The reaction is overall clean and results in the formation of a single product according to 1 H NMR spectroscopy (Figure S52).Removal of all volatile components and recrystallization from n-hexane yields a colorless microcrystalline solid which was first analyzed by LIFDI mass spectrometry and is in agreement with the formation of the oxygen bridged dimer Mes TerSn{N-( i Pr)N(Cy 2 )C( i Pr)}(μ-O 2 )Sn{N( i Pr)-N(Cy 2 )C( i Pr)} Mes Ter (1,3,2,4-dioxadistannetane) (10) (Scheme 5, A). 15 The cis configuration of both, the terphenyl and guanidinato ligands, could further be verified by single crystal X-ray diffraction with crystals obtained from a saturated n-pentane solution of 10 stored at −30 °C (Scheme 5, B).
Computational investigation found that the dimerization of the proposed terminal oxide intermediate to To prevent silyl migration, tin compound 7 with an aliphatic cyclohexyl substitution pattern instead of SiMe 3 groups was successfully synthesized.Reacting 7 with elemental selenium does not lead to the formation of a terminal tin selenide and gives rise to both the 1,2,3,4,5-tetraselenostannolane 8 and 1,3,2,4-diselenadistannetane 9, the formation of which is accompanied by the release of i PrN�C�N i Pr.
The reaction of 7 with N 2 O also deviates significantly from those of 2a,b, leading to the clean formation of the 1,3,2,4dioxadistannetane 10, showing that comparatively small changes in substitution have a significant influence on the reaction outcome and further emphasize the difficulties in stabilizing a compound with a terminal tin−oxygen bond.The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.inorgchem.4c00598.

Scheme 3 .
Scheme 3. Reaction of 2a,b with N 2 O to Give 4a,b and Overall Reaction Free Energy from 2a,b
Scheme 4. (A) Two-step and One-Pot Synthesis of the Heteroleptic Stannylene 7 and Its Reactivity with Elemental Selenium to Give 8; (B−D) Molecular Structures of Mes TerSn{N( i Pr)N(Cy 2 )C( i Pr)} (7), Mes TerSn(Se 4 ){N( i Pr)C(NCy 2 )N i Pr} (8), and Mes TerSn(NCy 2 )(μ-Se 2 )Sn{N( i Pr)C(NCy 2 )N( i Pr)} Mes Ter (9) in the Crystal a dimer 10 is exergonic by ΔG 298 = −26.2kcal mol −1 .The observed cis configuration is, albeit only slightly, thermodynamically favored over its trans configuration by ΔG 298 = −1.4kcal mol −1 .■ CONCLUSIONS We present the reactions of heteroleptic terphenyl-/amidosubstituted stannylenes 1a,b with carbodiimides.Investigated through combined experimental and computational studies, 1a,b react with the iso-propyl-substituted derivative, yielding the corresponding guanidinato complexes 2a,b.Sterically more demanding carbodiimides are unable to undergo a comparable metathesis-type reaction.Consequently, 2a,b offers maximum steric protection, which should ultimately facilitate the targeted synthesis of terminal tin chalcogenides.Compounds 2a,b react cleanly with elemental selenium to give respective terminal tin selenides 3a,b.By contrast, in reactions with N 2 O as an oxygen transfer reagent, instead of yielding a terminal tin oxide, silyl migration of the guanidinato ligand to the putative tin−oxygen moiety occurs, yielding the corresponding tin complexes 4a,b, bearing the Sn−OSiMe 3 functionality.
The obtained compounds have been comprehensively characterized in solution and in the solid state, including single crystal X-ray diffraction of one compound of each accessed class.The bonding situation in the first examples of four-coordinate terminal tin selenide 4a,b was further analyzed by quantum chemical calculations.■ ASSOCIATED CONTENT * sı Supporting Information

a
Scheme 5. (A) Reaction of 7 with N 2 O to Give 10; (B) Molecular Structure of Mes TerSn{N( i Pr)C(NCy 2 )N i Pr}(μ-O 2 )Sn{N( i Pr)C(NCy 2 )N i Pr} Mes Ter (10) in the Crystal a