Small Molecule Activation by Two‐Coordinate Acyclic Silylenes

In recent decades, the chemistry of stable silylenes (R2Si:) has evolved significantly. The first major development in this chemistry was the isolation of a silicocene which is stabilized by the Cp* (Cp* = η5‐C5Me5) ligand in 1986 and subsequently the isolation of a first N‐heterocyclic silylene (NHSi:) in 1994. Since the groundbreaking discoveries, a large number of isolable cyclic silylenes and higher coordinated silylenes, i.e. Si(II) compounds with coordination number greater than two, have been prepared and the properties investigated. However, the first isolable two‐coordinate acyclic silylene was finally reported in 2012. The achievements in the synthesis of acyclic silylenes have allowed for the utilization of silylenes in small molecule activation including inert H2 activation, a process previously exclusive to transition metals. This minireview highlights the developments in silylene chemistry, specifically two‐coordinate acyclic silylenes, including experimental and computational studies which investigate the extremely high reactivity of the acyclic silylenes.


Introduction
The cleavage of rigid σ-bonds, such as H-H bond, is a key step in a lot of important catalytic processes, conventionally the domain of transition metals. [1] The ability of transition metals to bind reversibly with various functional groups enables transition-metal complexes to perform as effective catalysts. In recent decades, the field of main-group compounds has grown signifi-the first isolable two-coordinate acyclic silylene was finally reported in 2012. The achievements in the synthesis of acyclic silylenes have allowed for the utilization of silylenes in small molecule activation including inert H 2 activation, a process previously exclusive to transition metals. This minireview highlights the developments in silylene chemistry, specifically two-coordinate acyclic silylenes, including experimental and computational studies which investigate the extremely high reactivity of the acyclic silylenes.
cantly and a variety of low-valent main-group compounds which show interesting reactivity have been reported. [2] Silicon, the second most abundant element in the Earth's crust, is especially of interest due to its high natural abundance and lowtoxicity. Bond activation reactions using low-valent silicon species in place of transition-metal complexes are of importance, as the former are more environmentally friendly and cost-effective than the latter. Over the past four decades, many low-valent silicon compounds have been prepared using sterically demanding ligands (kinetic stabilization) and/or electronically stabilizing ligands based on heteroatom substituents (thermodynamic stabilization). [3] Silylenes (:SiR 2 ), the silicon analogues of carbenes (:CR 2 ), have gained much attention due to their propensity to selectively activate small molecules. [2e] While the ground electronic state of carbenes (singlet or triplet) depends on the nature of the pendent substituents, silylenes generally exhibit a singlet ground state. [4] The frontier molecular orbitals of singlet-state silylenes consist of a high energy lone pair (HOMO) and an available vacant p-orbital (LUMO). This dual donor/acceptor character (ambiphilicity) mimics the frontier d-orbitals found in transition metals. [2a] As such, the activation of inert molecules (such as H 2 ) using silylenes has been shown to be possible, a process previously exclusive to transition metals.
In general, silylenes are of high reactivity, have short lifetimes and tend to undergo facile dimerization, oligomerization or polymerization. For example, silylenes bearing sterically bulky substituents such as Mes (Mes = 2,4,6-Me 3 C 6 H 2 ), dimerize to form the corresponding disilenes (R 2 Si=SiR 2 ). [5] Therefore, kinetic and/or thermodynamic stabilization is required to isolate such silylenes as stable compounds. One of important developments in main-group chemistry was the isolation of a disilene Mes 2 Si= SiMes 2 which was formed through the dimerization of the transient divalent silylene :SiMes 2 at 77 K. [6] Since the groundbreaking discovery, a variety of isolable cyclic silylenes and functionalized silylenes with a higher coordinated silicon(II) atom have been reported to date (Figure 1). One of the significant developments in the chemistry of stable silylenes is the isolation of the dodecamethylsilicocene :SiCp* 2 (1) (Cp* = η 5 -C 5 Me 5 ) as a Si(II) compound with higher coordination number by Jutzi and co-workers in 1986. [7] Silicocene 1 is stabilized using the thermodynamic stabilization effect of the Cp* ligands. In 1994, Denk and co-workers reported the first stable two-coordinate N-heterocyclic silylene (NHSi) 2, [8] which is the silicon analogue of the stable N-heterocyclic carbene (NHC) isolated by Arduengo and co-workers in 1991. [9] Subsequent to this, the groups of Lappert and Gehrhus succeeded in isolating the benzo-fused silylenes 3. [10] In 2006, the six-membered NHSi 4 was described by Driess and co-workers. [11] Roesky and coworkers reported the first base-free bis-silylene 5 in 2011. [12] A large number of examples of NHSi's, which are stabilized by the effect of cyclic systems along with the interaction from the lone pairs on the directly bonded nitrogen atoms to the vacant 3p orbital on the silicon atom, were reported. [13] In 1999, Kira and co-workers succeeded in the synthesis of the first isolable cyclic dialkylsilylene 6 using the kinetic stabilization effect by the sterically bulky dialkyl based helmet-type ligand. [14] Furthermore, Driess and co-workers reported the synthesis of carbocyclic silylenes 7 bearing two phosphonium ylides which exhibit comparable aromatic character. [15] Apart from these cyclic silylenes, a number of Lewis base stabilized silylenes have been prepared. [16] In this decade, many studies on three coordinate Si(II) compounds have been demonstrated and their interesting electronic features and fascinating reactivity have been revealed. A significant discovery in this field was the synthesis of three coordinate silylenes bearing halogens [:SiX 2 (L)] (X = halogen) which are widely used as precursors to synthesize novel silicon compounds. In 2006, the first example of isolable monomeric chlorosilylene (:SiCl[PhC(N t Bu) 2 ]) 8 stabilized by an amidinate ligand was reported by Roesky and co-workers. [17] NHC-stabilized dihalosilylenes [:SiX 2 (NHC) (X = halogen)] 9 are also indispensable building blocks in synthetic chemistry. [16c,18] Another remarkable recent achievement in such three coordinate systems are the preparation of hydrosilylenes [:Si(H)R] which are attractive compounds for applications in catalytic transformations such as the hydrosilylation of alkenes, alkynes and carbonyl com- pounds. There are only a few examples of isolated hydrosilylenes without using Lewis acid stabilization. [19] Kato and Baceiredo reported the isolation of a three coordinate Si(II) hydride 10a which is stabilized by an intramolecular phosphine coordination. [19a] Similarly, the phenyl-substituted silylene 10b stabilized by a similar phosphine based ligand was isolated. [20] In addition, four coordinate silicon(II) compounds, e.g. tetraphosphorus coordinated Si(II) compound (11), have been reported. [21] While many isolable cyclic silylenes [13,22] and Lewis base stabilized silylenes [16] have been described to date, only a few examples of simple dicoordinate acyclic silylenes are known, because the isolation of such silylenes as stable compounds is synthetically challenging due to their highly reactive nature.
In a previous theoretical study, Wang and Ma investigated the small molecule activation, specifically H 2 , to the variety of cyclic and acyclic silylenes. [23] Some of the key factors which influence the reactivity of silylenes towards H 2 activation are the HOMO-LUMO and singlet-triplet gaps. In the case of NHSi's which exhibit large HOMO-LUMO and singlet-triplet energy gaps, high activation energies are required to reach the corresponding product. Another important factor in the reaction behavior of silylenes is the geometry around the silicon center, especially the angle at the silicon atom. When the ring strain is large, a higher barrier is required in breaking a H 2 molecule. For instance, the activation energy for the three-membered silylene ring, silacyclopropenylidene (53.15 kcal/mol), is much higher compared with that of the acyclic dimethylsilylene (:SiMe 2 ) (13.31 kcal/mol) due to the ring strain along with the 2π-electrons-delocalization on the C-C-Si ring ( Figure 2). Similarly, in the case of nitrogen-substituted systems, the higher activation barrier (63.46 kcal/mol) for the N-heterocyclic silylene is required to reach the product than that (45.59 kcal/mol) of the diaminosilylene [:Si(NH 2 ) 2 ]. [23] It is also found that the splitting of H 2 with acyclic silylenes is more exothermic (-50.89 kcal/mol for dimethylsilylene, -23.34 kcal/mol for diaminosilylene) than that of cyclic silylenes (-18.65 kcal/mol for silacyclopropenylidene, -6.65 kcal/mol for N-heterocyclic silylene). In a recent computational study, Kuriakose and Vanka investigated the single site small molecule activation by acyclic silylenes and the undesired side reaction. [24] In these systems during H 2 splitting the undesired side reaction, which leads to decomposition of the silylenes forming the products :Si(H)R′ and HR (decomposition reaction), would be competitive to the desired reaction in which both hydrogen atoms bind to the same atom to form the tetravalent RSi(R′)H 2 product (single site reaction) (R/R′ = thiolato/thiolato, boryl/amido, or silyl/amido). The study indicated that the angle at the silicon center also affects the preference of silylenes for the single site or the decomposition pathway. When the angle becomes even smaller, the dissociation pathway is favored over the single site pathway significantly. Figure 2. The activation energy and the exothermic energy for the H 2 insertion reaction with the acyclic silylene (black line) and the cyclic silylene (light green line), calculated at the B3LYP/6-311+G** level of theory. [23] Transition state (TS), addition product (Pr).
In the last three decades, the chemistry of stable silylenes has grown significantly and has been subject to many recent reviews regarding isolable cyclic silylenes, [13,22] functionalized silylenes with higher coordinate silicon(II) centers, [16] and their application towards small molecule activation and catalytic reactions. [2c,25] Despite recent progress, the activation of inert molecules such as H 2 using silylenes remains scarce. The computational studies implied that acyclic silylenes with a highly obtuse angle at the silicon center would have a low-lying triplet excited state and allow for the activation of inert molecules. Therefore, the study of such silylenes may open new doors to the reactivity of main-group compounds as transition metal mimics. The focus of this minireview is the properties and key reactivity highlights from simple dicoordinate acyclic silylenes.

Properties
Several examples of stable two-coordinate acyclic silylenes have been structurally characterized using single-crystal X-ray diffraction analysis. The information can help to explain the compound's reactivity and stability. Selected structural parameters of these acyclic silylenes are shown in Table 1 [8,[10][11][12][13] Silylenes bearing electronegative siloxy (21) and boryloxy (24) substituents and the vinyl(silyl)silylene (27)  Additionally, the geometry of silylenes has an effect on the HOMO-LUMO energy, as silylenes which have wider E-Si-E′ angles tend to exhibit smaller HOMO-LUMO energy gaps (Table 1). Previous computational studies imply silylenes which exhibit small HOMO-LUMO gaps and coordinative flexibility should be ideal for selective activation of relatively unreactive small molecules. Further information regarding the chemical bonding in silylenes is gained by the 29 Si NMR spectrum ( Table 1). The 29 Si NMR chemical shifts for the two-coordinate silicon center of the acyclic bis(amido)silylene 19 (+204.6 ppm) is downfield shifted relative to those in NHSi's (δ = 78-119 ppm). [43] In the case of NHSi's, the lone pairs on the nitrogen atoms are parallel to the empty 3p orbital on the silicon atom leading to an efficient πoverlap and increased shielding of the silylene resonance. Similarly, bis(arylthiolato)silylenes 17a-c and imino(silyl)silylene 37a show a downfield shift in the 29 Si NMR spectrum [+285.5 (17a), +270.4 (17b), +270.9 (17c), +300.0 (37a) ppm]. These results suggested much less π-donation from the sulfur or nitrogen atoms to the vacant p orbital on the silicon atom relative to that in NHSi's. Furthermore, amido(boryl)silylene 14, amido-  (14), +438.2, +467.5 (15), +432.9 (27)] in the 29 Si NMR spectrum. The results indicate a significantly large electrophilicity of the divalent silicon center which is reminiscent of that observed in the dialkyl-substituted cyclic silylene 6 (567.4 ppm). [14] The 29 Si NMR spectrum of 21 and 24 feature a significantly highfield resonance [+58.9 (21), +35.5 (24) ppm] compared to other dicoordinate acyclic silylenes, which suggests additional π-donation by the siloxy or boryloxy ligands.

Small Molecule Activation with Acyclic Silylenes
an energetically accessible LUMO can mimic the reactivity of transition metal complexes. The ability of transition metals to facilitate the activation of small molecules (H 2 , CO, alkenes etc.) has enabled the widespread development of homogeneous transition metal catalysis. Recently, key catalytic reaction steps (oxidative addition, insertion reactions, reductive elimination) have been reported for the low-valent main-group compounds. [2] However, the cleavage of rigid σ-bonds such as H-H by main-group compounds remain scarce. Silylenes are of extremely high reactivity due to the high-energy lone-pair on silicon (HOMO) and the low-lying vacant p orbital (LUMO) that enable to activate these small molecules. Acyclic silylenes are expected to exhibit high reactivity relative to their cyclic counterparts due to their wide E-Si-E′ angles and small HOMO-LUMO gaps. In this section, the small molecule activation by two-coordinate acyclic silylenes is outlined.

Activation of H 2
The cleavage of dihydrogen is a key step in numerous homogeneous catalytic processes such as the hydrogenation of unsaturated organic compounds and hydroformylation reactions. [44] Additionally, the adsorption/regeneration of H 2 is important processes in potential hydrogen storage materials. [45] This desirable reactivity towards H 2 is generally mediated by transition metals, however it has recently been demonstrated that reduced main-group centers exhibit this reactivity as well.

Activation of NH 3
Although many examples of H 2 activation by transition metal complexes have been reported, N-H bond activation is more challenging as Werner-type complexes are readily formed in the reaction with Lewis basic amines. [47] The activation of the N-H bonds of ammonia has attracted attention for applications in catalytic transformations such as hydroamination. While very few of examples of N-H bond activation by transition metal complexes are known, it has been revealed that many lowvalent main-group compounds undergo such activation processes. [47b] Bis(amido)silylene 19 reacts with NH 3 to yield triaminosilane 50 together with the secondary amine TBoNH (48) (Scheme 13). [32] The plausible mechanism involves the formation of diaminosilylene 49 via a σ-bond metathesis reaction between 19 and NH 3 , followed by the oxidation addition to NH 3 to afford the triaminosilane 50. [48] This observation is in agreement with a previously computed σ-bond metathesis H 2 activation pathway mediated by silylene. [24] Recently, our group reported the reactivity of the imino(siloxy)silylene 21. [49] The reaction of silylene 21 with 1 equivalent of NH 3 affords the hydroamination product 51. It is of note that compound 51 even Scheme 13. Activation of NH 3 by acyclic silylenes. reacts with excessive amounts of NH 3 to yield an unidentified mixture. In the reaction, IPrNH and (H 2 N)( t Bu 3 SiO)Si(H)(NH 2 ) were formed, similar σ-bond metathesis reaction to that observed for the bis(amido)silylene 19.

C-O Bond Activation
Carbon dioxide is a potent greenhouse gas in the atmosphere and a versatile feedstock for chemical or material production. [50] To date, many studies on carbon capture and storage (CCS) of CO 2 along with its chemical activation and utilization as a C1 source have been demonstrated. While transition metals have been utilized in most CO 2 activation, the development of transition-metal free and eco-friendly systems have been underexplored. Currently, some low-valent silicon compounds which undergo CO 2 activation have been reported. [51] Our group found that silepin 38a, which behaves as dormant form of imino(silyl)silylene 37a, rapidly reacts with CO 2 under mild conditions (1 atm, r.t., within 1h) to afford the corresponding silicon carbonate 53 (Scheme 14). [40] Comparable to mechanisms described in literature, [51e,52] it is plausible that the transient silanone (O=Si(IPrN){Si(SiMe 3 ) 3 }) was formed by the oxidative addition of CO 2 and extrusion of CO, followed by the cycloaddition of another molecule of CO 2 . While such compounds tend to dimerize, [51] our group demonstrated that the isolation of the first four-coordinate, monomeric silicon carbonate 53 in high yields. The reaction of 14 with CO 2 under mild conditions (1 atm, r.t.) resulted in the formation of the (trimethylsiloxy)iminosilane {(HCDippN) 2 B}Si(NDipp)(OSiMe 3 ) (52). [53] It is plausible that the in situ generation of the silanone [O=Si{B(NDippCH) 2 }{N(SiMe 3 )-Dipp}], [54] followed by silyl group migration than the bimolecular reaction with CO 2 yields the carbonate. [55] Silylene 14 also reacts with CO at ambient temperature to yield 54 which was characterized by standard spectroscopic techniques and X-ray crystallographic analysis (Scheme 15). [53] Compound 54 contains two Si(IV) centers, which bind to one carbon and two oxygen atoms derived from CO, along with the amide ligand. Although the activation of carbon monoxide and formation of the stable carbonyl complexes under mild condition is well known for transition-metal complexes, such reaction is virtually unknown for main-group compounds. Recently, the groups of Schulz and Schreiner reported the isolation of the silylene carbonyl complex [L(Br)Ga] 2 Si:-CO (55) (L = HC[C(Me)N(2,6-i Pr 2 C 6 H 3 )] 2 ). The reaction of GaL with SiBr 4 under a CO atmosphere, generates the silylene [L(Br)Ga] 2 Si: (46) in situ, subsequently affording the silylene carbonyl complex 55. [46] Compound 55 is remarkably stable both in the solid state (T d = 176-177°C) and in solution, no decomposition was observed in toluene solution up to 80°C. Furthermore, silylene carbonyl complex 55 acts as a masked silylene and reacts with H 2 to give the dihydrosilane H 2 Si[Ga(Br)L] 2 (47).

C=C and C≡C Bonds Activation
The activation of small organic molecules and the formation of C-C bonds is a fundamentally important process for transfor- mation of simple molecules into essential chemical compounds in both academia and industry. [56] For this purpose, transition metal catalysts have been utilized and a large number of useful catalysts have been developed. On the one hand, the catalytic bond activation and C-C bond formation are challenging for main group compounds as their oxidation states vary in a much narrower range. Recently, some main group element compounds have shown the activation of C-C bonds in neutral organic molecules such as alkenes alkynes as well as its dynamic equilibrium that is a key step in catalytic processes. Furthermore, it was demonstrated that catalytic activation of alkynes, followed by the formation of C-C bonds by utilizing low-valent main-group compounds is also possible. [57] Silylenes are well known to undergo cycloaddition reactions with unsaturated C-C bonds. [13,43] Similarly, silylenes 21, 37a, and 40 reacted with ethylene under mild conditions (1 atm, r.t.) to form the corresponding cycloaddition products 56, 59, and 60 (Scheme 16). [33,40,41] In the case of silylene 15, the reaction with ethylene at ambient temperatures gave the silirane product Si{CH 2 -CH 2 }{NDipp(SiMe 3 )}{Si(SiMe 3 ) 3 } (57) in high yields. Furthermore, when compound 57 was heated to 60°C under an ethylene atmosphere, an exceptional insertion of ethylene into Si-Si bond occurred to yield the modified silirane Si{CH 2 -CH 2 }{NDipp(SiMe 3 )}{CH 2 -CH 2 -Si(SiMe 3 ) 3 } (58). [58] A NMR experiment with deuterated ethylene indicated that the reaction proceeds via migratory insertion of the coordinated ethylene into the Si-Si bond, followed by the formation of the silirane with a C 2 D 4 molecule. The groups of Power and Tuononen found that silylenes 17a and 17b also react with ethylene or alkynes to afford the [1+2] cycloaddition products 61a, 61b, and 62b (Scheme 17). [59] Interestingly, the ethylene addition products 61a and 61b were found to undergo reversible reactions with ethylene under ambient conditions. Notably, while many maingroup compounds can react with ethylene under mild conditions, the reversible reaction, which is a key step in catalytic cycles, remains rare.
[60] Products 61a and 61b were characterized using NMR spectroscopy (61a and 61b) and X-ray crystallographic analysis (61b). Van't Hoff analysis of the association of ethylene with 17b, as determined by variable-temperature 1 H NMR spectroscopy, revealed a small value of Gibbs free energy (ΔG assn = -24.9 kJ/mol at 300 K), which is comparably more favorable compared with that for the reaction of the phosphine supported Si(II) complex reported by the groups of Kato and Baceiredo (-3.0 kJ/mol). [60b] Similarly, the rare reversibility between Si(II) and Si(IV) compounds was found in silylenes 37a and 37b which undergo an intermolecular insertion reaction into the C=C bond of the aromatic ligand framework to give silepins 38a and 38b. The equilibrium between 37a and 38a was revealed by experimental and computational studies.

Activation of P 4
White phosphorus (P 4 ), which is easily obtained by the reduction of phosphate rock, is widely used as a starting material to synthesize organophosphorus compounds. In industrial processes, phosphorus chloride (PCl n ) and phosphoryl chloride (POCl 3 ) are precursors to organo-phosphorus products are prepared by the chlorination or oxychlorination of white phosphorus. [61] The eco-friendly and atom efficient method of direct conversion of white phosphorus to phosphine containing products has also been considered. [62] Currently, it was demonstrated that direct catalytic transformation of P 4 into organophosphorus compounds using a transition metal complex is possible. [63] However, the stoichiometric and catalytic reaction of white phosphorus under mild conditions is challenging for both transition metal and main group compounds.
While several cyclic silylenes can react with P 4 , in most cases, the oxidative addition of a single P-P bond at the silicon center is observed, [64] and the controlled reaction of P 4 by main-group compounds remains scarce. [61a,65] The treatment of the vinyl(silyl)silylene 27 with P 4 resulted in the formation of ( Me IPrCH)Si(P 4 ){Si(SiMe 3 ) 3 } (63) (Scheme 18). [35] It is plausible that compound 63 was formed by the oxidative addition of a P-P bond of P 4 to silylene 27 with subsequent 1,2-silyl migration. In this reaction, the cleavage of two P-P bonds of P 4 and the regioselective formation of four new Si-P bonds were observed. In addition, imino(siloxy)silylene 21 was found to occur via the oxidative addition of 1 equivalent of P 4 to give

Summary and Outlook
Since the discovery of Jutzi's silicocene and Denk's NHSi, various cyclic silylenes and Lewis base stabilized silylenes have been reported. However, isolable two-coordinate acyclic silylenes had been considered to be transient and non-isolable compounds for a long time. The recent synthesis of stable acyclic silylenes has unlocked a new avenue in silicon chemistry, enabling metallomimetic behavior of this Earth-abundant element. In this minireview, we mainly focused on the two-coordinate acyclic silylene chemistry containing synthesis, properties, and application for small molecule activation. Acyclic silylenes bearing wide E-Si-E′ angles and small HOMO-LUMO gaps are of extremely high reactivity which lead to the activation of important small molecules such as H 2 , NH 3 , ethylene, and CO 2 . Interestingly, while bis(silyl)silylene 40 exhibits a relatively large HOMO-LUMO gap (4.18 eV), 40 was found to occur the activation of H 2 . The singlet-triplet gap of 40 is comparatively small (10.5 kJ/mol). In addition, it was revealed that bis(arylthiolato)silylenes 17a and 17b show equilibrium reactions with ethylene at room temperature which is a key step in catalytic processes and is rare for silicon due to the unfavorable reduction of Si(IV) to Si(II).
These results for acyclic silylenes indicate the potential of main group compounds for future applications in the realms of catalytic and materials science. This study is inspiring for the molecular design of new silicon compounds which enable small molecule activation and the unprecedented cleavage of N-N bond in dinitrogen, N 2 . Acyclic silylenes are of extremely high reactivity due to the high-energy lone-pair on silicon (HOMO) and a low-lying vacant p-orbital (LUMO) which may interact with an empty π* orbital and an n-orbital (a lone pair) on N 2 leading to a weakening of the N-N bond. Furthermore, the hydrogenation and transfer hydrogenation of unsaturated molecules such as alkenes and alkynes mediated by acyclic silylenes is expected to be feasible. These studies imply that the steric and electronic tuning of the substituents enable the control of hydrogenation reactions. The accessibility of both Si(II) and Si(IV) oxidation states may lead to the utilization of silicon compounds in catalysts.