Reactions of m -Terphenyl Stabilized Germylene and Stannylene with Water and Methanol: Oxidative Addition versus Arene Elimination and Different Reaction Pathways for Alkyl and Aryl Substituted Species

: Reactions of the divalent germylene Ge(Ar Me6 ) 2 (Ar Me6 = C 6 H 3 -2,6-{C 6 H 2 -2,4,6-(CH 3 ) 3 } 2 ) with water or methanol gave the Ge(IV) insertion products (Ar Me6 ) 2 Ge(H)OH ( 1 ) or (Ar Me6 ) 2 Ge(H)OMe ( 2 ), respectively. In contrast, its stannylene congener Sn(Ar Me6 ) 2 reacted with water or methanol to produce the Sn(II) species {Ar Me6 Sn(μ-OH)} 2 ( 3 ) or {Ar Me6 Sn(μ-OMe)} 2 ( 4 ), respectively, with elimination of Ar Me6 H. Compounds 1 – 4 were characterized by IR and NMR spectroscopy as well as by X-ray crystallography. Density functional theory calculations yielded mechanistic insight into the formation of (Ar Me6 ) 2 Ge(H)OH and {Ar Me6 Sn( μ - OH)} 2 . The insertion of an m -terphenyl stabilized germylene into the O–H bond was found to be catalytic, aided by a second molecule of water. The lowest energy pathway for the elimination of arene from the corresponding stannylene involved sigma-bond metathesis rather than separate oxidative addition and reductive elimination steps. The reactivity of Sn(Ar Me6 ) 2 with water or methanol contrasts with that of Sn{(CH(SiMe 3 ) 2 } 2 which affords the Sn(IV) insertion products {(Me 3 Si) 2 CH} 2 Sn(H)OH or {(Me 3 Si) 2 CH} 2 Sn(H)OMe. The differences were tentatively ascribed to the Lewis basicity of the employed solvent (Et 2 O vs. THF) and the use of molar vs. millimolar concentrations of the substrate.


INTRODUCTION
Stable divalent group 14 carbene analogues (tetrylenes) have attracted much interest over the last four decades. 1 Some of the more recent advances in this area include the synthesis of twocoordinate acyclic silylenes, 2 the donor-acceptor stabilization of heavy group 14 hydrides GeH2 and SnH2, 3 and the characterization of two-coordinate 1,2-bis-metalylenes, 4 a heavier analogue of 1,2-dicarbenes. The interest in stable tetrylenes is in part due to their facile and in some cases reversible reactivity with fundamentally important small molecules such as CO, 5 H2, 2a,4,6 NH3, 6a,7 and C2H4 8 under mild conditions. Of particular significance have been insertion reactions of which topical examples include the insertion of a silylene into a P-P bond of P4, 9 the insertion of a cationic metallogermylene into X-H bonds (X = H, B, or Si), 10 and the insertion of a boryl substituted stannylene into the C≡C bond in alkynes. 11 In recent years, we have carried out detailed investigations of the reactivity of m-terphenyl stabilized heavier tetrylenes with small molecules. 12 In this context, we have reported the reactions of E(Ar Me 6 )2 (E = Ge, Sn; Ar Me 6 = C6H3-2,6-{C6H2-2,4,6-(CH3)3}2) 13 with H2, 6 NH3, 6a CO, 5 isocyanides, 14 hydrazines, 15 inorganic acids HX (X = CN − , N3 − , F − , SO3CF3 − , and BF4 − ), 16 and AlMe3 and GaMe3. 17 A logical extension of this chemistry is the investigation of the reactions of E(Ar Me 6 )2 with hydroxyl compounds and of water and methanol in particular. Although the insertion of transient tetrylenes such as EH2, EMe2, or EPh2 (E = Si, Ge, or Sn) into O-H bonds has been studied both experimentally and theoretically, 18  LGeGeL with H2 in the presence of adventitious moisture. 4 In addition, insertion reactions of Ge{CH(SiMe3)2}2 into the O-H bond of the enolic forms of several ketones have been reported, although none of the products has been structurally characterized. 21 We now describe the reaction of E(Ar Me 6 )2 (E = Ge, Sn) with water or methanol which give the insertion products (Ar Me 6 )2Ge(H)OR (1, R = H; 2, R = Me) or the dimers {Ar Me 6 Sn(μ-OR)}2 (3, R = H; 4, R = Me) depending on the group 14 element E (Scheme 1). The products 1-4 were characterized by spectroscopic methods and single crystal X-ray crystallography. The observed reactivity is similar to that seen for E(Ar Me 6 )2 with NH3 6a and alkyl-amino tetrylenes [LEEt] (L = N(C6H2-4-Si i Pr3-2,6-{C(H)Ph2}2 i Pr)) with protic reagents, 22 but differs from that of E(Ar Me 6 )2 with HBF4, which resulted in the formation of an insertion product (Ar Me 6 )2E(H)F irrespective of the element E. 15 To shed light on the mechanistic details of the observed transformations, the reactions of stable germylenes and stannylenes with water were also probed computationally. The acquired experimental and computational results were compared with those reported for E{CH(SiMe3)2}2 as well as with the data available for transient tetrylenes EMe2 and EPh2. Scheme 1. Reaction of E(Ar Me 6 )2 (E = Ge, Sn) with water or methanol to form 1-4.

Structures.
Single crystal X-ray diffraction studies revealed a similar structural arrangement in 1 and 2 (Figures 1 and 2; selected structural data are given in Table 1). Compound 1 crystallizes in the orthorhombic Fdd2 space group, while the space group of the methoxy derivative 2 is monoclinic P21/c. In 1, the germanium atom and the hydrogen and hydroxide groups bound to it are disordered over two sites with 50% occupancy. A similar distortion is seen in the structure of 2, in which case the crystals were also found to contain co-crystallized 1 (18%) due to the presence of adventitious water in the reaction mixture.  In both 1 and 2, the germanium atom has tetrahedral coordination due to bonding to two terphenyl ligands, an oxygen atom from the hydroxyl/methoxy group, and a hydrogen atom. The germanium-bound hydrogen atom could not be clearly identified from the electron density difference map for either 1 or 2, but its presence in the structures was confirmed with IR and 1 H NMR spectroscopies.
Compounds 3 and 4 crystallize in monoclinic space groups C2/m and P21/n, respectively. Their structures revealed a dimeric arrangement with two SnAr Me 6 moieties bridged by two hydroxide (3) or methoxide (4) groups (Figures 3 and 4; selected structural data are given in Table 1). In contrast to compounds 1 and 2, no structural disorder was observed either for 3 or 4. However, crystals of the arene Ar Me 6 H, 25 a side product in the synthesis of both 3 and 4, were obtained together with 3 and 4 in all crystallizations.
The structures of 3 and 4 are similar with comparable structural parameters. Notable differences are only observed in the orientation of the Ar Me 6 ligands in relation to the four-membered Sn2O2 ring that is influenced by the steric requirements of the   Table 1. Selected bond lengths (Å) and angles (°) in products 1-4.
bridging hydroxide/methoxide group. In both 3 and 4, the coordination around tin atoms is trigonal pyramidal with near 90° interligand C-Sn-O angles. We note that the structure of 3 is very similar to that of the related dimer {Ar iPr 4 Sn(μ-OH)}2 (Ar iPr 4 = C6H3-2,6-{C6H2-2,4-[CH(CH3)2]2}2) that was synthesized previously by us via reaction of a distannyne with TEMPO or pyridine-N-oxide. 26 Spectroscopy. The 1 H NMR spectra of 1-4 display the characteristic signals of the Ar Me 6 group with methyl and aromatic proton resonances appearing from 1.8 to 2.3 ppm and from 6.6 to 7.1 ppm, respectively (see Supporting Information).
The methoxy proton in 2 appears as a singlet at 2.74 ppm in the 1 H NMR spectrum, whereas the germanium-bound proton in 1 and 2 shows a resonance at 6.06 and 6.05 ppm, respectively. The Ge-OH proton chemical shift (0.47 ppm) was confirmed by COSY NMR spectroscopy. It is further upfield than that reported for the O-H proton (Eind)2Ge(H)OH (5.21 ppm), possibly as a result of shielding by the flanking rings. The Ge-H proton resonance was found to be a doublet with a coupling constant to the O-H proton of 1.4 Hz, confirming, albeit indirectly, the existence of the hydroxyl proton in the structure. The presence of the O-H proton is also supported by IR spectroscopy and the synthesis of the deuterium analogue of 1, Products 3 and 4 could not be readily separated from the Ar Me 6 H byproduct to provide completely clean NMR spectra. This is due to the low solubility of 3 and 4 and the relatively high solubility of Ar Me 6 H to common solvents. Because of this, the NMR spectra of 3 and 4 were deconvoluted through comparison of the data to a spectrum of pure Ar Me 6 H (see Supporting Information). The methoxy proton in 4 appears as a singlet at 2.74 ppm in the 1 H NMR spectrum, while the O-H proton of 3 shows a resonance at 0.00 ppm. The latter assignment could be confirmed via synthesis of {Ar Me 6 Sn(μ-OD)}2 (3′), the deuterium analogue of 3 (see Supporting Information). Even though the O-H stretch of 3 could not be seen in the IR spectrum at the expected region (ca. 3500 cm −1 ), 24 the O-D stretch of 3′ is clearly visible at 2620 cm −1 .
While 1 and 2 are both colorless, products 3 and 4 are pale yellow and absorb in the near UV and violet region with λmax of 365 and 302 nm for 3 and 4 (See Supporting Information).
Discussion. The observation of the dimeric products 3 and 4 parallels the reactivity of Sn(Ar Me 6 )2 with NH3. 6a However, it differs from the results obtained by Pörschke and coworkers who found that the tin species Sn{CH(SiMe3)2}2 undergoes simple insertion into both water and methanol, 20 forming {(Me3Si)2CH}2Sn(H)OH and {(Me3Si)2CH}2Sn(H)OMe, respectively. The germanium congener Ge{C(H)(SiMe3)2}2 is also known to form insertion products {(Me3Si)2CH}2Ge(H)R (R = H, Me) analogous to 1 and 2. 19 Hence, in order to understand the reactivity observed for E(Ar Me 6 )2 with water or methanol, mechanistic investigations of possible reaction pathways were conducted computationally using density functional theory (DFT). The calculations were performed for model compounds E(Ar Me 4 )2 (Ar Me 4 = C6H3-2,6-{C6H3-2,6-(CH3)2}2) as the p-methyl substituent in Ar Me 6 is not expected to influence the reaction mechanism and can therefore be omitted to lower the computational cost.
Computational Results. Insertion into O-H bonds is one of the most studied reactions of transient silylenes and germylenes. 18 It has been shown both experimentally and theoretically that the reaction proceeds by formation of a Lewis acid-base complex, R′2E′-OHR (E′ = Si, Ge; R′ = H, Me, Ph), followed by a proton transfer from oxygen to the group 14 element to give the O-H insertion product R'2E'(H)OR. The latter step in the mechanism is not unimolecular but rather involves catalytic proton transfer by a second molecule of water or alcohol. An alternative reaction channel involving elimination of H2 from water/alcohol complexes of parent tetrylenes, H2E'-OHR, to yield HE'OR has also been identified computationally and experimentally verified by Leigh et al. 18a Calculations (PBE1PBE-D3BJ/def-TZVP) performed for the reaction of E(Ar Me 4 )2 with water show that the formation of complexes (Ar Me 4 )2E-OH2 is slightly endergonic or energy neutral for E = Ge and Sn, respectively. The insertion into the O-H bond was found to take place catalytically, aided by a second molecule of water ( Figure 5, left). This parallels the reactivity of transient tetrylenes with water/alcohols 18 or amines 28 as well as the mechanism calculated for the insertion of Ge(Ar Me 6 )2 into the N-H bond in NH3. 6a The Gibbs energy of activation for the formation of (Ar Me 4 )2E(H)OH was found to be 46 and 73 kJ mol -1 for E = Ge and Sn, respectively. We note that the mechanism for a unimolecular proton transfer was found to have a considerably greater barrier, well above 100 kJ mol -1 , representing an improbable reaction pathway on the potential energy surface.
The arene elimination pathway ( Figure 5, right) was calculated to proceed via one-step sigma bond metathesis 27 that has a Gibbs energy of activation of 64 and 40 kJ mol -1 for E = Ge and Sn, respectively. The reaction was found to proceed via a cyclic transition state in which a hydrogen from the coordinated water molecule in (Ar Me 4 )2E-OH2 is transferred to the aryl substituent, yielding Ar Me 4 EOH and Ar Me 4 H. This is reminiscent of the mechanism calculated for the reaction of Sn(Ar Me 6 )2 with NH3 to give Ar Me 6 SnNH2. 6a The hydroxytetrylene Ar Me 4 EOH can undergo a subsequent barrierless dimerization to form the hydroxide-bridged dimer {Ar Me 4 E(μ-OH)}2 as the final product. We also tested the possibility that arene elimination would involve the product of the oxidative addition pathway. However, the energy barrier for elimination of Ar Me 4 H from (Ar Me 4 )2E(H)OH was found to be considerably higher than that calculated for the sigma bond metathesis.
A comparison of the two energy diagrams in Figure 5 shows that the oxidative insertion pathway has a lower barrier when E = Ge, whereas for E = Sn, the arene elimination pathway is kinetically favored. However, the dimer {Ar Me 4 E(μ-OH)}2 is the thermodynamically preferred product independent of the identity of the group 14 element. As a whole, the computational results predict different reactivity for Ge(Ar Me 4 )2 and Sn(Ar Me 4 )2 with water, which is in good agreement with experimental observations. The lower energy barrier for Ge(Ar Me 4 )2 in oxidative insertion can be rationalized with its greater Lewis basicity as compared to Sn(Ar Me 4 )2. In a similar fashion, the barriers for reductive elimination parallel the strength of the E-C bond with the weaker bond corresponding to a lower barrier.
The computational results allow a straightforward rationalization as to why Sn(Ar Me 6 )2 reacts with HBF4 by simple insertion. 15 The elimination of Ar Me 6 H from (Ar Me 6 )2Sn(H)F would most likely proceed via high energy transition state, rendering such transformation unlikely. It is, however, less obvious why Sn{CH(SiMe3)2}2 reacts with water or alcohol by oxidative insertion, 19 while Sn(Ar Me 6 )2 undergoes arene elimination and subsequent dimerization to yield {Ar Me 4 Sn(μ-OR)}2. For this reason, we also examined the reaction of E{CH(SiMe3)2}2 with water using computational methods.
The results from DFT calculations (PBE1PBE-D3BJ/def-TZVP) suggested that the reaction of E{CH(SiMe3)2}2 with water proceeds qualitatively similarly as observed for E(Ar Me 4 )2 (see Supporting Information). Specifically, Ge{CH(SiMe3)2}2 was predicted to react via oxidative insertion, whereas Sn{CH(SiMe3)2}2 should eliminate CH2(SiMe3)2 to give {(Me3Si)2HC}SnOH that would readily dimerize to [{(Me3Si)2HC}Sn(μ-OR)]2. The dimer [{(Me3Si)2HC}Sn(μ-OR)]2 was found to be both the thermodynamically and kinetically favored product with a difference of 16 kJ mol -1 in activation energy for the two investigated pathways. We note that the energy barriers for the elimination of the alkane CH2(SiMe3)2 were calculated to be approximately twice as high as those determined for the arene Ar Me 4 H, as expected based on the differences in their electronic structures.
At this point we can only speculate as to why the experimental and computational results for the reactivity of Sn{CH(SiMe3)2}2 with water differ. We note that the reactions of Sn{CH(SiMe3)2}2 have used an excess (14 equivalents) of water at molar concentrations, 20  It should also be pointed out that the experiments reported by Pörschke et al. employ THF as the solvent, 20 while the analogous reactions with Ge{CH(SiMe3)2}2, 19 as well as those involving E(Ar Me 6 )2, were performed in Et2O and in the presence of millimolar quantities of the substrate. The properties of strong electron pair donor solvents are known to have significant influence to the reactivity of GeH2 and SiMe2, as the reacting species in such cases is not the free tetrylene but its Lewis acid-base adduct. 18a, 29 In a similar fashion, the product distribution from the reaction of GePh2 with CCl4 has been observed to depend on the employed solvent, with the insertion product Ph2Ge(Cl)CCl3 being favored in neat THF solutions or in hexanes containing catalytic amounts of THF; 30 with an even stronger donor, NEt3, the insertion product was obtained exclusively. These data strongly suggest that the reactivity of Sn{CH(SiMe3)2}2 with water or methanol could be vastly different if the reactions were carried out in Et2O due to its less coordinating nature as compared to THF. Solvent modified reactivity of stable acyclic tetrylenes are currently under further computational and experimental investigations in our laboratories.

CONCLUSION
The m-terphenyl stabilized tetrylenes Ge(Ar Me 6 )2 and Sn(Ar Me 6 )2 were found to have differing reactivity towards both water and methanol. While the germylene Ge(Ar Me 6 )2 inserts into the O-H bond to form the Ge(IV) products (Ar Me 6 )2Ge(H)OH (1) and (Ar Me 6 )2Ge(H)OMe (2), the corresponding stannylene Sn(Ar Me 6 )2 undergoes arene elimination followed by dimerization to yield the Sn(II) species {Ar Me 6 Sn(μ-OH)}2 (3) and {Ar Me 6 Sn(μ-OMe)}2 (4). Computational analyses at the DFT level reproduced the experimental results and gave mechanistic insight into the formation of (Ar Me 6 )2Ge(H)OH and {Ar Me 6 Sn(μ-OH)}2. The insertion of an Ar Me 4 stabilized germylene into the O-H bond was found to be catalytic, aided by a second molecule of water. The lowest energy pathway for the elimination of arene from the corresponding stannylene involved sigma-bond metathesis rather than separate oxidative addition and reductive elimination steps. DFT calculations found no difference in the preferred reactivity of Sn(Ar Me 4 )2 and Sn{CH(SiMe3)2}2 even though the latter is known to undergo oxidative insertion into O-H bonds. A plausible explanation for the experimental behavior of the alkyl stannylene is the use of reaction conditions that favour insertion (molar concentrations of water and THF as the solvent).

EXPERIMENTAL SECTION
General Procedures. All manipulations were carried out in either an inert atmosphere glovebox or by using modified Schlenk techniques to maintain strictly anaerobic and anhydrous conditions. Ge(Ar Me 6) 2 and Sn(Ar Me 6) 2 were prepared by literature procedures. 13 Solvents were dried using a Grubbs-style purification system 31 and stored over NaK. Water was collected from a Thermo Scientific Barnstead Nanopure and degassed under reduced pressure (< 1 torr) for 2 h. Methanol was dried with CaH2 and distilled onto 3 Å molecular sieves. 1 H and 13 C{ 1 H} spectra were recorded on a Bruker 500 MHz spectrometer and referenced to solvent signals. No 119 Sn signals could be observed in the region between +2400 and -1000 ppm. Melting points were determined on a Mel-Temp II apparatus using capillary tubes sealed with vacuum grease under an inert atmosphere. IR spectra were recorded as Nujol mulls between CsI plates on a Perkin-Elmer 1430 ratio recording infrared spectrometer. UV-Visible spectra were recorded from dilute solutions in toluene using 3 mL quartz cuvette and an Olis 17 modernized Cary 14 UV/Vis/NIR spectrophotometer. (Ar Me 6 )2Ge(H)OMe (2). MeOH (15.9 μL, 0.3924 mmol) was dissolved in Et2O (20 mL) and this solution was added dropwise to a slurry of Ge(Ar Me 6 )2 (0.2745 g 0.3924 mmol) in Et2O (35 mL) at −78 °C. The resulting mixture was stirred and allowed to warm slowly to room temperature overnight, giving a pale purple solution. After 48 hours of stirring, a pale yellow solution was obtained. Reduction of the solvent to ca. 4 mL gave colorless X-ray quality crystals that contained 2 (82 %) co-crystallized with 1 (18 % (15 mL) and this solution was added dropwise to a slurry of Sn(Ar Me 6 )2 (0.1300 g, 0.174 mmol) in Et2O (40 mL) at −78 °C. The resulting mixture was stirred and allowed to warm slowly to room temperature overnight, which resulted in a slightly cloudy pale yellow solution. The solvent volume was reduced to ca. 5 mL after which the solution was decanted. The resulting solid was dried under reduced pressure giving a yellow powder. X-ray quality single crystals of both 3 and Ar Me 6 H were grown from a saturated Et2O solution. Total mass yield: 90% (0.1175 g). Upon heating, crystals of 3 become visibly orange at ca. 160 °C but no melting is observed < 250 °C. 1  Computational Details. All calculations were performed with Gaussian09. 32 The structures of the studied systems were optimized using the PBE1PBE hybrid exchange-correlation functional 33 in conjunction with the def-TZVP basis sets. 34 For tin, a def2-TZVP basis with an effective core potential (ECP) was used to treat scalar relativistic effects. 35 Dispersion interactions were modelled by applying Grimme's empirical dispersion correction with Becke-Johnson damping (D3BJ). 36 The choice of a particular functional-basis set combination was motivated by our recent theoretical-experimental studies of the chemistry of metallylenes as well as computational efficiency. 2a, 8, 14 The correction for dispersion effects was considered crucial in order to obtain accurate energetics for systems employing bulky terphenyl substituents.
Calculations were performed only for the reactions of heavier tetrylenes with water. Model compounds based on the Ar Me 4 ligand (Ar Me 4 = C6H3-2,6-{C6H3-2,6-(CH3)2}2) were used to reduce the computational cost. The nature of stationary points found (minimum or transition state) was assessed with calculation of full Hessian matrices using analytic or numerical gradients.

ASSOCIATED CONTENT Supporting Information
Crystallographic information files for 1-4, synthetic details of deuterium analogues 1′ and 3′, and additional spectroscopic (NMR, IR, and UV/Vis) and computational data.

Notes
The authors declare no competing financial interest.