Subscribe to RSS
DOI: 10.1055/a-2188-1842
Skeletal Rearrangements Involving Cyclopropyl- and Alkene-Stabilized Silylium Ions
The work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany’s Excellence Strategy (EXC 2008/1-390540038). P.-W.L. thanks the China Scholarship Council for a predoctoral fellowship (2019–2023). M.O. is indebted to the Einstein Foundation Berlin for an endowed professorship.
Abstract
This Account summarizes the fascinating chemistry of cyclopropyl-stabilized silylium ions, which are readily available from vinylcyclopropanes (VCPs). Depending on the nucleophilic partner, these reactive intermediates undergo direct ring opening or ring expansion to nonclassical alkene-stabilized silylium ions. The latter can also be accessed by gold as well as proton electrophiles from silicon compounds containing unsaturated C–C bonds. All these reaction cascades can be terminated by C–H or C–C as well as Si–O bond formation. From this, a clearer picture of the versatility of these rather complex chemistries emerges.
1 Introduction
2 Skeletal Rearrangements of Vinylcyclopropanes Involving Cyclopropyl-Stabilized Silylium Ions
2.1 Termination by C–H Bond Formation
2.2 Termination by C–C Bond Formation
3 Related Bond Reorganizations Involving Alkene-Stabilized Silylium Ions
3.1 Initiation by Cationic Gold(I) Complexes
3.2 Initiation by Brønsted Acids
4 Conclusion
#
Keywords
β-silicon effect - bond reorganization - carbenium ions - ring expansion - silylium ions - skeletal rearrangements - vinylcyclopropanesBiographical Sketches


Peng-Wei Long (born in 1994 in Xi’an, China) obtained his bachelor’s degree in July 2016 from Shangluo University, and completed his master’s studies in July 2019 under the supervision of Professor Li-Wen Xu at Hangzhou Normal University. Supported by a predoctoral fellowship from the China Scholarship Council, he started his Ph.D. research under the guidance of Martin Oestreich at the Technische Universität Berlin in October 2019. His work focuses on the application of strong Lewis acids such as silylium-ion-like species and B(C6F5)3 for the synthesis of organosilicon compounds.


Tao He (born in 1990 in Zhangye, China) received his bachelor’s degree from the University of Science and Technology of China (Hefei/China, 2014) under the supervision of Professor Yun-He Xu. He then moved to Tsinghua University and obtained his doctoral degree under the supervision of Professor Wei He (Beijing/China, 2019). His graduate research involved the synthesis of novel silicon-containing compounds under rhodium catalysis. In 2019, he moved to Berlin and joined the group of Martin Oestreich at the Technische Universität Berlin for his postdoctoral training funded by the Cluster of Excellence UniSysCat. His research interests focus on the application of superelectrophiles in organocatalysis.


Hendrik F. T. Klare (born in 1981 in Ankum, Germany) studied chemistry at the Westfälische Wilhelms-Universität Münster, where he received his diploma in 2007 and his Ph.D. in 2011 under the supervision of Martin Oestreich. His graduate research involved the exploitation of stabilized silylium ions for catalysis. As part of a DFG predoctoral fellowship, he spent six months as a visiting scholar in the group of Professors Kazuyuki Tatsumi and Yasuhiro Ohki at Nagoya University in Japan, investigating cooperative Si–H bond activation modes. After postdoctoral studies with Professor Gerhard Erker in Münster focusing on frustrated Lewis pair chemistry, he joined the laboratory of Professor Brian M. Stoltz at the California Institute of Technology as a DAAD postdoctoral fellow, developing novel methodologies for natural product synthesis. In 2013, Hendrik moved to the Technische Universität Berlin, where he is currently working as a senior scientist and lecturer in the Oestreich group.


Martin Oestreich (born in 1971 in Pforzheim, Germany) is a Professor of Organic Chemistry at the Technische Universität Berlin. He received his diploma degree with Professor Paul Knochel (Marburg, 1996) and his doctoral degree with Professor Dieter Hoppe (Münster, 1999). After a two-year postdoctoral stint with Professor Larry E. Overman (Irvine, 1999–2001), he completed his habilitation with Professor Reinhard Brückner (Freiburg, 2001–2005) and was appointed as Professor of Organic Chemistry at the Westfalische Wilhelms-Universität Münster (2006–2011). He also held visiting positions at Cardiff University in Wales (2005), at The Australian National University in Canberra (2010), and at Kyoto University in Japan (2018).
Introduction
Stabilization of the carbenium ion is key to the success of the SN1 reaction as well as to bond reorganizations involving these reactive intermediates. The type of stabilization can be classified as positive inductive (+I) and mesomeric (+M) effects. The inductive effect is transmitted through the σ-framework of a molecule and emerges from a local change in electron density due to electron-withdrawing or -donating groups in the molecule; it is determined by the electronegativity of the substituents. Conversely, mesomeric effects can either engage a nonbonding electron pair (n) of an α-heteroatom directly attached to the carbon atom with the electron sextet[1] or an adjacent π-system (Scheme [1a]). Aside from these effects, hyperconjugation,[2] which is an interaction of the empty orbital at the cationic carbon atom with an adjacent filled σ-orbital of a C(sp3)–C(sp3) or C(sp3)–H bond, can also lend significant stabilization to the carbenium ion (Scheme [1b]). This orbital interaction can be further enhanced with a more electropositive substituent in the β-position. For example, the third-row element silicon features a larger atomic radius than carbon and it is also more electropositive. As a consequence, a vicinal C(sp3)–Si bond exerts a more pronounced hyperconjugative stabilization of the vacant p-orbital of the carbenium-ion center. Based on this effect, commonly referred to as the β-silicon effect (Scheme [1b]),[3] [4] [5] [6] various synthetically useful transformations with vinyl, allyl, propargyl, allenyl, and other silanes have been established in both Lewis acid and transition-metal catalysis.[3,4] Of note, there is another, far less exploited role of C(sp3)–Si bonds that helps stabilize carbenium ions: the so-called γ-silicon effect.[7] [8] More precisely, the back lobe of a C(sp3)–Si bond can interact with the empty p-orbital by percaudal homohyperconjugation (Scheme [1c]). In intramolecular systems, the bonding situation is even more complex, and stabilization of the positive charge by delocalization from contributions of neighboring C–C and C–H bonds generally leads to the formation of bridged structures featuring a three-center, two-electron (3c2e) bond.[9] One prominent example of such a nonclassical pentacoordinate carbonium ion is the 2-norbornyl cation,[10] [11] and the synthesis and isolation of the corresponding silicon analog have also been accomplished (Scheme [1d], top).[12] [13] These cations can be considered either as classical β-silyl-substituted cyclic carbenium ions or, as a more realistic approximation, as nonclassical alkene-stabilized silylium ions (Scheme [1d], bottom). Inspired by these unusual bonding motifs, we were curious to learn about the generation of cyclopropyl-stabilized silylium ions and their reactivity in bond reorganization processes (Scheme [1d], gray box). Due to its ring strain, the cyclopropyl group features partial C(sp2) hybridization and thus C=C double-bond character.[14]


As a conventional cyclopropane derivative and a highly attractive research target, vinylcyclopropanes (VCPs) have received considerable interest in the past several decades. Transition-metal catalysis[15] [16] [17] unleashed an exceptionally rich variety of products starting from VCPs. Aside from these stimulating developments, the reactivity of VCPs toward Brønsted acids also opened the door to several bond reorganizations.[16,18] Usually perceived as ‘fat protons’,[19] silylium ions[20] were recently found by us to react with VCPs, resulting in various selective rearrangement cascades. These intriguing transformations share a common intermediate, namely an intramolecularly cyclopropyl-stabilized silylium ion in two diastereomeric forms. In all these reactions, the β-silicon effect is a key feature. We describe in this Account how deceptively minor and, as such, subtle changes to the VCP substrate and the reaction initiation can result in distinctly different reaction channels and hence product outcomes. These discoveries are put into perspective and related bond reorganizations that also rely on the intermediacy of β-silicon-stabilized cyclic carbenium ions and alkene-stabilized silylium ions, respectively, will also be discussed.
# 2
Skeletal Rearrangements of Vinylcyclopropanes Involving Cyclopropyl-Stabilized Silylium Ions
Starting from the monosubstituted VCP 1, cyclopropyl-stabilized silylium ions 4 + can be accessed either via a [1,3]-hydride shift of a β-silicon-stabilized carbenium ion 2 + formed by the addition of a donor-stabilized silylium ion to the alkene in VCP 1, or via an intermolecular Lewis acid-mediated abstraction of a leaving group from the silicon atom in hydrosilylation product 5 (Scheme [2]). Depending on the substituents and the donor groups involved, those cyclopropyl-stabilized silylium ions 4 + are the starting point for diverse rearrangements involving C–C bond reorganization. Termination of these reaction cascades can occur by intermolecular hydride transfer from another hydrosilane molecule or a borohydride. Aside from C–H bond formation, even C–C bond formation by allyl transfer from an allylsilane is feasible. Alternatively, intramolecular association with an arene can lead to Friedel–Crafts-type electrophilic aromatic substitution (SEAr). Hence, this section is organized by the type of donor molecule and donor group, respectively.




Termination by C–H Bond Formation
The hydricity of the hydride source can determine the fate of these cyclopropyl-stabilized silylium ions. In 2020, our group disclosed a silylium-ion-promoted formal (5+1) cycloaddition of aryl-substituted VCPs 1 and hydrosilanes 6 involving aryl migration (Scheme [3]).[21] This transformation is initiated by catalytic amounts of trityl borate Ph3C+[B(C6F5)4]– and subsequently maintained by self-regeneration of the silylium ion. Silacyclohexanes 7, in which the aryl group had changed its connectivity, were found as the major product along with minor amounts of silacyclohexane 8 and silacyclopentane 9, both with the aryl group attached to the original carbon atom. The solvent had a substantial effect on the product distribution as the formation of 8 and 9 can be strongly suppressed in electron-deficient, yet highly polar arenes such as 1,2-dichlorobenzene and chlorobenzene. In addition, the electronic properties of the substituents had a marked influence on the yields of the major product 7. Aside from dihydrosilane 6a, tertiary hydrosilanes 6b–d also participated in this reaction. This is attributed to substituent exchange between a quaternary silane and a silylium ion.[22] [23] [24] [25] Such dealkylation of quaternary silanes has been independently studied by us.[26]




To investigate the migration of the aryl group, a series of control experiments was performed (Scheme [4]).[21] α-Substituted styrene 10, a possible intermediate resulting from opening of the cyclopropyl ring, was independently prepared and subjected to the standard procedure. The cyclization product 8aa, without aryl migration, was the only obtained product (Scheme [4a], top). Another experiment starting from skipped diene 11 as a plausible intermediate, also exclusively afforded silacyclohexane 8aa in excellent yield (Scheme [4a], bottom). These results convincingly excluded the chemoselective ring opening of the cyclopropane as an entry into the skeletal reorganization. Experiments with cyclopropyl-containing silanes 12, preformed by hydrosilylation of the alkene unit in VCP 1a,[27] yielded the aryl-migrated silacyclohexane 7aa almost quantitatively, either providing the hydride source intramolecularly, as in 12a, or intermolecularly, as for 12b, with Et3SiH (Scheme [4b]). These results indicated that ring opening/hydrosilylation of the cyclopropyl group is directly linked to the aryl migration.
Based on deuterium-labeling experiments (not shown) and density-functional theory (DFT) calculations at the M062X/cc-PVTZ//M062X/6–311G(d,p) level,[28] we proposed a plausible mechanism for the silylium-ion-promoted (5+1) cycloaddition (Scheme [5]).[21] An initial hydride abstraction[25] followed by an exergonic association of the thus-generated silylium ion [Et2HSi(benzene)]+ to the C=C double bond in VCP 1a forms the β-silicon-stabilized carbenium ion 13 +. A subsequent intramolecular [1,3]-hydride shift from the silicon atom to the benzylic carbon atom generates the benzene-stabilized silylium ion 14 +. Replacement of the benzene donor with the cyclopropyl group eventually leads to the intramolecularly cyclopropyl-stabilized silylium ion 15 +, which exists as two diastereomers with cis- and trans-configuration, likely in equilibrium with each other. The diastereomer cis-15 + allows for the lowest transition state for the thermodynamically driven [1,2]-phenyl shift/ring expansion under cleavage of the proximal C4–C6 bond, thereby converting the highly strained bicyclo[3.1.0]hex-2-silyl cation cis-15 + into the alkene-stabilized silylium ion 16 +. An intermolecular hydride transfer from the external dihydrosilane 6a to 16 + affords the aryl-migrated product 7aa. The silylium ion stabilized by C6H6 is regenerated in this step. The reaction outcome with 7aa as the major product and 8aa/9aa as side products [after an additional cyclopropane-to-cyclopropane (CP-to-CP) rearrangement; not shown] is consistent with the free-energy difference of those two barriers, making this (5+1) cycloaddition a kinetically controlled process.
Compared to hydrosilanes 6, the borohydride [HB(C6F5)3]–, generated by hydride abstraction from a hydrosilane with B(C6F5)3, is a more competent hydride donor.[29] Hence, its higher reactivity towards the involved carbenium-ion intermediates was expected to have an impact on the reaction outcome. In 2021, we reported a B(C6F5)3-catalyzed diastereoselective (4+1) cycloaddition of alkyl-substituted VCPs 18 and Et2SiH2 (6a) to afford trans-3,4-disubstituted silolanes 19 (Scheme [6]).[30] This reaction bears similarities to the above (5+1) cycloaddition in that a β-silicon-stabilized carbenium ion 13 + with [HB(C6F5)3]– as the counteranion is generated. Unlike the scenario with the [B(C6F5)4]– counteranion, initial reduction occurs to furnish the hydrosilylation product.[27] Another Si–H bond activation with B(C6F5)3 then leads to a cyclopropyl-stabilized silylium ion with [HB(C6F5)3]– as the counteranion. It is again at this key intermediate stage that the reaction outcome is decided: In this case, cleavage of the distal C5–C6 bond of the cyclopropyl group in the trans-configured diastereomer occurs with concomitant intermolecular hydride transfer from the borohydride (Scheme [7], top). The corresponding transition state TS20→19 for the reaction of the aforementioned hydrosilylation product 20 [27] helps to explain the trans-configuration of the silanes 19. Complete loss of stereocontrol was seen when probing this transformation with trans-1,2-disubstituted VCP 21, obtaining trans-22 and cis-22 as a 50:50 mixture of diastereomers (Scheme [7], bottom).




# 2.2
Termination by C–C Bond Formation
Other than the hydride transfer from either a hydrosilane or an in-situ-generated borohydride, the cyclopropyl-stabilized silylium ion can also engage in the reaction with π-basic donors such as an arene or the C=C double bond of an allyl group. In 2021, we disclosed a 6/6/5-fused tricyclic ring construction starting from benzyl-substituted VCPs 23 and hydrosilanes 6 via an intramolecular Friedel–Crafts alkylation (Scheme [8]).[31] The functional-group tolerance is modest, but one example stands out. The silyl-substituted VCP 23f did not yield the corresponding product 24fa but, instead, the parent 24aa was observed as the sole product. Strikingly, the yield of 24aa was significantly higher with 23f than that with 23a as the cyclization precursor (80% versus 54%). Hence, the quantitative protodesilylation as well as the improved yield shed light on the reaction mechanism (vide infra). Dealkylation and dearylation at the silicon atom were involved when tertiary hydrosilanes 6b–d were used instead of secondary hydrosilane 6a, again arising from substituent redistribution.[22] [23] [24] [25] [26]


VCP 27, bearing a phenethyl group instead of the benzyl substituent, afforded another unknown 6/6/6-fused silicon-containing tricycle (Scheme [9]).


Control experiments with different precursors were conducted to probe the reaction mechanism (Scheme [10]).[31] The predominant formation of an endo-cyclization product 30 rather than tricyclic product 24aa from 2-substituted allyl benzene 29 again excluded a preceding ring-opening hydrosilylation of the cyclopropyl group in VCPs 23. Potential intermediate 31, arising from hydrosilylation of VCPs 23 and Et2SiH2 (6a), was cleanly converted into the 6/6/5-fused tricycle 24aa upon treatment with trityl borate Ph3C+[B(C6F5)4]– in the absence of an external hydrosilane 6. These results indicate that the ring opening of the cyclopropyl group is intimately connected to an intramolecular Friedel–Crafts-type bond formation at the tethered aryl group and is downstream of the alkene hydrosilylation.




A plausible reaction mechanism supported by DFT calculations at the M062X/cc-PVTZ//M062X/6-31G(d,p) level[28] is outlined in Scheme [11] (top).[31] Again, the cyclopropyl-stabilized silylium ions cis-33 + and trans-33 + marked the crossroads of this reaction sequence. Their formation is the same as before (cf. Scheme [5]). However, unlike for aryl-substituted VCPs 1, the sequence of [1,2]-benzyl shift/ring expansion is kinetically less favorable for benzyl-substituted VCPs 23, consistent with the unobserved formation of silacyclohexane 36aa (Scheme [11], right cycle). Instead, the trans-diastereomer 33 +, which is lower in energy, undergoes intramolecular nucleophilic attack of the phenyl group at C5 with concomitant cleavage of the distal C5–C6 (Scheme [11], left cycle), resembling the hydride transfer in the (4+1) cycloaddition (cf. Scheme [7]). The formation of a bond between C6 and the silicon atom results in a 6-endo-tet ring closure, leading to the tricyclic Wheland intermediate 34 +. This Brønsted acidic intermediate then reacts with dihydrosilane 6a by dehydrogenative protolysis[32] [33] to arrive at the experimentally obtained product 24aa. However, 34 + is also a proton source that can be the origin of side reactions (Scheme [11], bottom). This Brønsted acid can transfer a proton to the C=C double bond of another molecule of VCP 23a, releasing the product 24aa under the formation of the cyclopropylcarbinyl cation 37 +. The barrier of this protonation step is kinetically competitive. Subsequent hydride transfer from Et2SiH2 (6a) to 37 + gives the γ-silicon-stabilized carbenium ion 38 +. An intermolecular hydride transfer from dihydrosilane 6a eventually yields the open-chain product 26 along with regeneration of the donor-stabilized silylium ion. Notably, the protodesilylation seen in the transformation of VCP 23f, with a silyl group attached to the aryl group, into parent 24aa greatly suppresses that undesired reaction channel by sequestering protons and turning them into silylium ions.[26] , [33] [34] [35] [36] As a consequence of this, the yield of tricyclic product 24aa is higher than directly starting from 23a (see Scheme [8], bottom).
In addition to the intramolecular interception with a tethered arene to afford a tricyclic ring system, the alkene-stabilized silylium ion emerging from the rearrangement of the cyclopropyl-stabilized silylium ion can also be trapped intermolecularly by an allyl group, again terminating the reaction sequence with a C–C bond formation. In 2022, we accomplished a skeletal reorganization of β-silylated cyclopropanes 39, bearing an allyl group at the silicon atom (Scheme [12]).[37] From this, silacyclohexanes 40 containing a quaternary carbon atom are formed. Trityl borate Ph3C+[B(C6F5)4]– proved to be an optimal initiator, and diminished yields were obtained with a silylium ion, the benzenium ion, or other Lewis acids. Moderate to good yields were achieved, and functional groups such as silyl (40f) and halides (40g–j) were tolerated. A four-carbon-atom tether in VCPs 40 secured excellent chemoselectivity, as the potentially competing Friedel–Crafts cyclization is no longer an option. This situation changed when the carbon chain was shortened by one carbon (41a and 42a). A substrate with the ortho-positions occupied again rearranged selectively to the allylated silacyclohexane 41b, offering strong experimental evidence for the Friedel–Crafts cyclization being a side reaction.




Based on the insights from our earlier studies as well as DFT computations at the M06-2X/cc-PVTZ//M06-2X/6-31G(d,p) level,[28] the mechanistic picture depicted in Scheme [13] evolved.[37] The reaction cascade is initiated by trityl-cation-mediated deallylation to preferentially generate the energetically favored cyclopropyl-stabilized silylium ion trans-43 +, which undergoes skeletal rearrangement by a [1,2]-hydride shift/ring expansion through transition state trans-44 +‡, thereby forming an energetically more stable tertiary carbocation stabilized by the β-silicon effect; i.e., alkene-stabilized silylium ion 45 + (upper cycle). Subsequent transfer of an allyl group from another molecule of 39 to the carbocation center in 45 + leads to the allylation product 40 along with the regeneration of trans-43 +. The competing Friedel–Crafts cyclization (lower cycle) traces back to a CP-to-CP rearrangement to furnish another cyclopropyl-stabilized silylium ion trans-47 + through transition state trans-46 +‡. Subsequent intramolecular nucleophilic attack of the phenyl ring results in a 7-exo-tet ring closure under cleavage of the proximal C2–C3 bond to give Wheland intermediate 48 +. The proton released from 48 + is captured by the allylsilane 39 under the release of propylene, thereby regenerating trans-43 +, which can enter the next catalytic cycle. While silacyclopentane 49a was not detected experimentally starting from 39a with n = 4, the Friedel–Crafts cyclization becomes a viable pathway for substrates with a tether that is shortened by one carbon atom (n = 3; cf. 42a in Scheme [12], gray box).
#
# 3
Related Bond Reorganizations Involving Alkene-Stabilized Silylium Ions
The bond reorganizations discussed in Section 2 reveal that cyclopropyl-stabilized silylium ions derived from vinylcyclopropanes (VCPs) tend to undergo strain-release ring opening to alkene-stabilized silylium ions, eventually arriving at silacyclohexanes (cf. Scheme [5] and Scheme [13]). In fact, these β-silicon-stabilized cyclic carbenium ion-like intermediates can also be accessed starting from different precursors. Various cleverly designed silicon-containing hydrocarbon skeletons were employed for this purpose,[38] [39] [40] [41] [42] [43] leading to thermodynamically favored silacyclohexane or -cyclopentane derivatives. It is noteworthy that all of these reactions proceed intramolecularly under either gold(I) or Brønsted acid initiation with preinstalled π-donors at the silicon atom, including allyl,[38–40] , [43] alkenyl,[41,42] (het)aryl,[41] and alkynyl[38] [39] substituents. Remarkable examples have been disclosed using phenols as nucleophiles to intermolecularly trap the formed alkene-stabilized silylium ion intermediate, affording the corresponding acyclic silicon-stereogenic silyl ethers, even in an enantioselective manner.[43]
3.1Initiation by Cationic Gold(I) Complexes
In 2006, the groups of Toste[39] (Scheme [14], top) and Lee[38] (Scheme [15], top) independently reported a formal sila-Cope rearrangement of (alkynyl)allylsilanes 50 catalyzed by cationic gold(I) complexes generated in situ from LAuCl and AgSbF6 or AgBF4. Activation of the alkynyl substituent by the carbophilic catalyst induces a 6-endo-tet cyclization through nucleophilic attack of the pendant allyl group to give cationic six-membered ring intermediate 56 + and its resonance form, carbenoid 57 + (Scheme [14], bottom). The β-silicon-stabilized carbenium ion 56 + can undergo either direct trapping by an oxygen-based nucleophile or β-silyl fragmentation to form intramolecularly alkene-stabilized silylium ion 58 +. In Toste’s work, acyclic silyl ether 53 and cyclic vinyl silane 54 can be chemoselectively accessed depending on the alcohol nucleophile used: 53 with phenol (51a) and 54 with methanol (52a). The different reaction outcomes were attributed to a faster rate of β-silyl fragmentation than cation capture so that the more nucleophilic aliphatic alcohol 52a (i.e., MeOH) leads to the formation of cyclic product 54 while less nucleophilic aromatic alcohol 51a (i.e., PhOH) mainly affords the acyclic silyl ether 53. The reaction is rather sensitive to the steric situation since α-substituted allylsilanes exclusively yielded acyclic product 53 even when employing methanol (52a) as the nucleophile. Conversely, cyclic product 54 was not obtained in Lee’s work (Scheme [15], top), which can presumably be explained by the slightly modified protocol using less electron-rich Ph3P instead of t-Bu3P as the ligand, resulting in less back-bonding from gold(I) (cf. 57 + in Scheme [14], bottom) and faster β-silyl fragmentation of 56 +. Direct alcoholysis of the allyl moiety in 50 was observed when replacing the allyl (R3 = H) with a methallyl group (R3 = Me), a consequence of preferential activation of the more electron-rich methallyl group over the alkyne by the catalyst to give alcoholysis product 59 (Scheme [15], bottom).


# 3.2
Initiation by Brønsted Acids
Allylsilanes have been extensively utilized as silylium-ion precursors[20] because of their reactivity toward various electrophiles such as protons[26] [36] and carbenium ions[44] (see also Scheme [12] and Scheme [13]). Their nucleophilicity is increased in methallyl-substituted derivatives. On this basis, List and co-workers developed a protocol for the asymmetric synthesis of silicon-stereogenic silyl ethers 66 from bis(methallyl)silanes 60 and phenols 51 induced by catalytic amounts of an imidodiphosphorimidate (IDPi) 61 as Brønsted acid (Scheme [16]).[43] The authors’ initial plan was to convert the bismethallylated quaternary silane 60 into silicon-stereogenic silyl ether 63 by protonation of 60 with the chiral IDPi acid catalyst 61 followed by capture of the silylium ion generated after C–Si bond cleavage in the assumed β-silicon-stabilized carbenium ion 62 + with a phenol nucleophile 51 (Scheme [16a]). However, direct alcoholysis under the liberation of isobutene was not observed. Instead, intermediate 62+ underwent a 6-endo-trig ring closure with the other methallyl substituent at the silicon atom to afford another β-silicon-stabilized carbenium ion 64 + (Scheme [16b]). Subsequent C–Si bond cleavage leads to the formation of the intramolecularly alkene-stabilized silylium ion 65+ paired with the chiral counteranion of the IDPi acid. Its reaction with phenol derivative 51 eventually furnished the enantioenriched silyl ether 66.




This desymmetrization protocol features good functional-group compatibility, and, aside from benzyl-substituted derivatives 60a–e, aryl-substituted congeners 60f–i with both electron-donating and -withdrawing groups also produced comparable results (Scheme [17]).[43] A diminished enantioselectivity was obtained with silane 60j, bearing three methallyl groups. The enantioselectivity was highly sensitive regarding the alkyl substituent at the silicon atom; even replacing the methyl with an ethyl group resulted in significantly decreased enantioinduction, as seen for 66k.


Mechanistic control experiments were performed to elucidate the assumed order of bond-forming events (Scheme [18]).[43] The isolated side product 63f was reacted with starting material 60f and phenol (51a) but did not give any silyl ether 66f (Scheme [18a]). A shorter reaction time with lower catalyst loading led to the conversion of silane 60a into silacyclohexane 67a in nearly racemic form (Scheme [18b]). Subjecting racemic silane 67a as substrate under otherwise standard reaction conditions afforded silyl ether 66a with comparable enantioselectivity (Scheme [18c]). These reaction outcomes indicated that the Si–O bond formation is the enantioselectivity-determining step (rather than the formation of cyclic 67). Moreover, the substantial enantiomeric ratio obtained for 66j when starting from trismethallyl-substituted silane 60j showed that the enantioselectivity was unlikely to be determined by protonation of one of the alkenes. Based on these experimental results, a mechanism was proposed that commences with the protonation of prochiral bis(methallyl)silane 60 by the IDPi acid 61 (Scheme [18d]). Ion pair 62+ X* – is formed with the carbocation stabilized by hyperconjugation; i.e., the β-silicon effect. A cationic π-cyclization then gives the ion pair 64 +X*–; loss of a proton and reprotonation put this intermediate in equilibrium with cyclic silane 67. Alternatively, ion pair 64 +X*– converts into silylium-based ion pair 65 +X*– by C–Si bond cleavage. Nucleophilic attack of phenol 51b at the silicon atom of 65 +X*– delivers product 66 and regenerates the catalyst. Steric effects on the transition states of that enantiodetermining step were also investigated by DFT calculations.[43]


#
# 4
Conclusion
In summary, we have demonstrated that cyclopropyl-stabilized silylium ions, which are readily accessible from VCPs, are versatile reactive intermediates en route to a diverse set of silicon-containing compounds. Deceptively simple variations of the substituent at the VCP and the nucleophilic coupling partner result in dramatically different reaction outcomes. The intramolecular stabilization of the transient silylium ion by the cyclopropyl group can occur with cis- or trans-configuration. These diastereomeric isomers are in equilibrium with each other and are the starting point of different bond reorganizations that can be terminated by either C–H (in the presence of a hydride source) or C–C bond formation (in the presence of a π-bond donor such as an arene or allyl group). Dependent on the nucleophilicity of the reactant and the substituent on the VCP, the intramolecularly stabilized silylium ions undergo ring opening of the cyclopropyl group by cleavage of the distal or proximal C–C bond, resulting in the formation of five- and six-membered silacycles, respectively. This strain-release ring opening of the cyclopropyl group can be accompanied by 1,2-aryl (for R = Ar) or 1,2-hydride migration (for R = Alk), while an intramolecularly accessible aryl group is the entry into Friedel–Crafts-type cyclizations (for R = Bn). The ring expansion leads to a cyclic β-silicon-stabilized carbenium ion that typically undergoes fast β-silyl fragmentation to an alkene-stabilized silylium ion. Other synthetic approaches to arrive at these reactive intermediates by gold(I) or Brønsted acid initiation have been put into context in this Account. Although kinetically less favored, a cyclopropane-to-cyclopropane rearrangement at the stage of the trans-configured cyclopropyl-stabilized silylium ion must be taken into account, opening the door to further reaction channels and potential side products. We find it remarkable that, despite these numerous possibilities, a specific product can be obtained selectively by deliberate choice of the reaction conditions.
#
#
Conflict of Interest
The authors declare no conflict of interest.
Acknowledgment
We thank Vittorio Bonetti for his initial contributions as part of this master thesis, and Dr. Guoqiang Wang for his substantial intellectual input during the quantum-chemical elucidation of several reaction mechanisms.
-
References
- 1 Grützmacher H, Marchand CM. Coord. Chem. Rev. 1997; 163: 287
- 2 Deasy CL. Chem. Rev. 1945; 36: 145
- 3 Roberts DD, McLaughlin MG. Adv. Synth. Catal. 2022; 364: 2307
- 4 Beļaunieks R, Puriņš M, Turks M. Synthesis 2020; 52: 2147
- 5a Siehl H.-U. In Recent Developments in Carbocation and Onium Ion Chemistry . Laali KK. ACS Symposium Series 965; American Chemical Society; Washington: 2007: 1
- 5b Lambert JB, Zhao Y, Emblidge RW, Salvador LA, Liu X, So J.-H, Chelius EC. Acc. Chem. Res. 1999; 32: 183
- 5c Lambert JB. Tetrahedron 1990; 46: 2677
- 6a Ibrahim MR, Jorgensen WL. J. Am. Chem. Soc. 1989; 111: 819
- 6b Lambert JB, Zhao Y. J. Am. Chem. Soc. 1996; 118: 7867
- 6c Lambert JB, Zhao Y, Wu H. J. Org. Chem. 1999; 64: 2729
- 6d Dabbagh HA, Zamani M, Fakhraee S. Res. Chem. Intermed. 2012; 38: 1551
- 6e Dabbagh HA, Zamani M, Fakhraee S. Res. Chem. Intermed. 2013; 39: 2011
- 7a Davidson ER, Shiner Jr. VJ. J. Am. Chem. Soc. 1986; 108: 3135
- 7b Sugawara M, Yoshida J. J. Org. Chem. 2000; 65: 3135
- 7c For more recent work, see: Creary X. Beilstein J. Org. Chem. 2019; 15: 1769 ; and cited references
- 8 Roy A, Bonetti V, Wang G, Wu Q, Klare HF. T, Oestreich M. Org. Lett. 2020; 22: 1213
- 9 Brown HC. The Nonclassical Ion Problem . Plenum Press; New York: 1977
- 10 Scholz F, Himmel D, Heinemann FW, Schleyer P. vR, Meyer K, Krossing I. Science 2013; 341: 62
- 11 Moss RA. J. Phys. Org. Chem. 2014; 27: 374
- 12 Steinberger H.-U, Müller T, Auner N, Maerker C, Schleyer P. vR. Angew. Chem. Int. Ed. Engl. 1997; 36: 626
- 13 Müller T, Bauch C, Ostermeier M, Bolte M, Auner N. J. Am. Chem. Soc. 2003; 125: 2158
- 14 de Meijere A. Angew. Chem. Int. Ed. Engl. 1979; 18: 809
- 15 Wang J, Blaszczyk SA, Li X, Tang W. Chem. Rev. 2021; 121: 110
- 16 Ganesh V, Chandrasekaran S. Synthesis 2016; 48: 4347
- 17 Jiao L, Yu Z.-X. J. Org. Chem. 2013; 78: 6842
- 18 Goldschmidt Z, Crammer B. Chem. Soc. Rev. 1988; 17: 229
- 19 Fleming I. Chem. Soc. Rev. 1981; 10: 83
- 20 Klare HF. T, Albers L, Süsse L, Keess S, Müller T, Oestreich M. Chem. Rev. 2021; 121: 5889
- 21 He T, Wang G, Bonetti V, Klare HF. T, Oestreich M. Angew. Chem. Int. Ed. 2020; 59: 12186
- 22 Schäfer A, Reißmann M, Schäfer A, Saak W, Haase D, Müller T. Angew. Chem. Int. Ed. 2011; 50: 12636
- 23 Schäfer A, Reißmann M, Jung S, Schäfer A, Saak W, Brendler E, Müller T. Organometallics 2013; 32: 4713
- 24 Labbow R, Reiß F, Schulz A, Villinger A. Organometallics 2014; 33: 3223
- 25 Omann L, Pudasaini B, Irran E, Klare HF. T, Baik M.-H, Oestreich M. Chem. Sci. 2018; 9: 5600
- 26 Wu Q, Qu Z.-W, Omann L, Irran E, Klare HF. T, Oestreich M. Angew. Chem. Int. Ed. 2018; 57: 9176
- 27 Long P.-W, He T, Oestreich M. Org. Lett. 2020; 22: 7383
- 28a Zhao Y, Truhlar DG. Theor. Chem. Acc. 2008; 120: 215
- 28b Zhao Y, Truhlar DG. Acc. Chem. Res. 2008; 41: 157
- 28c The solvent effect was taken into consideration using a polarizable continuum model (PCM) for both geometry optimizations and single-point energy calculations, see: Tomasi J, Persico M. Chem. Rev. 1994; 94: 2027
- 29 Heiden ZM, Lathem AP. Organometallics 2015; 34: 1818
- 30 Long P.-W, Oestreich M. Org. Lett. 2021; 23: 4834
- 31 He T, Wang G, Long P.-W, Kemper S, Irran E, Klare HF. T, Oestreich M. Chem. Sci. 2021; 12: 569
- 32 Chen Q.-A, Klare HF. T, Oestreich M. J. Am. Chem. Soc. 2016; 138: 7868
- 33 Wu Q, Irran E, Müller R, Kaupp M, Klare HF. T, Oestreich M. Science 2019; 365: 168
- 34 Allemann O, Duttwyler S, Romanato P, Baldridge KK, Siegel JS. Science 2011; 332: 574
- 35 He T, Klare HF. T, Oestreich M. ACS Catal. 2021; 11: 12186
- 36 Klare HF. T, Oestreich M. J. Am. Chem. Soc. 2021; 143: 15490
- 37 Long P.-W, Wang G, Klare HF. T, Oestreich M. ACS Catal. 2022; 12: 12310
- 38 Park S, Lee D. J. Am. Chem. Soc. 2006; 128: 10664
- 39 Horino Y, Luzung MR, Toste FD. J. Am. Chem. Soc. 2006; 128: 11364
- 40 Matsuda T, Kadowaki S, Yamaguchi Y, Murakami M. Chem. Commun. 2008; 2744
- 41 Matsuda T, Yamaguchi Y, Shigeno M, Sato S, Murakami M. Chem. Commun. 2011; 47: 8697
- 42 Zhou T, Xu L, Xia Y. Org. Lett. 2013; 15: 6074
- 43 Zhou H, Han JT, Nöthling N, Lindner MM, Jenniches J, Kühn C, Tsuji N, Zhang L, List B. J. Am. Chem. Soc. 2022; 144: 10156
- 44 Lambert JB, Zhao Y, Wu H, Tse WC, Kuhlmann B. J. Am. Chem. Soc. 1999; 121: 5001
For early computational and experimental data on β-silyl carbocations, see:
For more recent computational analyses, see:
Corresponding Author
Publication History
Received: 07 September 2023
Accepted after revision: 09 October 2023
Accepted Manuscript online:
09 October 2023
Article published online:
23 November 2023
© 2023. Thieme. All rights reserved
Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany
-
References
- 1 Grützmacher H, Marchand CM. Coord. Chem. Rev. 1997; 163: 287
- 2 Deasy CL. Chem. Rev. 1945; 36: 145
- 3 Roberts DD, McLaughlin MG. Adv. Synth. Catal. 2022; 364: 2307
- 4 Beļaunieks R, Puriņš M, Turks M. Synthesis 2020; 52: 2147
- 5a Siehl H.-U. In Recent Developments in Carbocation and Onium Ion Chemistry . Laali KK. ACS Symposium Series 965; American Chemical Society; Washington: 2007: 1
- 5b Lambert JB, Zhao Y, Emblidge RW, Salvador LA, Liu X, So J.-H, Chelius EC. Acc. Chem. Res. 1999; 32: 183
- 5c Lambert JB. Tetrahedron 1990; 46: 2677
- 6a Ibrahim MR, Jorgensen WL. J. Am. Chem. Soc. 1989; 111: 819
- 6b Lambert JB, Zhao Y. J. Am. Chem. Soc. 1996; 118: 7867
- 6c Lambert JB, Zhao Y, Wu H. J. Org. Chem. 1999; 64: 2729
- 6d Dabbagh HA, Zamani M, Fakhraee S. Res. Chem. Intermed. 2012; 38: 1551
- 6e Dabbagh HA, Zamani M, Fakhraee S. Res. Chem. Intermed. 2013; 39: 2011
- 7a Davidson ER, Shiner Jr. VJ. J. Am. Chem. Soc. 1986; 108: 3135
- 7b Sugawara M, Yoshida J. J. Org. Chem. 2000; 65: 3135
- 7c For more recent work, see: Creary X. Beilstein J. Org. Chem. 2019; 15: 1769 ; and cited references
- 8 Roy A, Bonetti V, Wang G, Wu Q, Klare HF. T, Oestreich M. Org. Lett. 2020; 22: 1213
- 9 Brown HC. The Nonclassical Ion Problem . Plenum Press; New York: 1977
- 10 Scholz F, Himmel D, Heinemann FW, Schleyer P. vR, Meyer K, Krossing I. Science 2013; 341: 62
- 11 Moss RA. J. Phys. Org. Chem. 2014; 27: 374
- 12 Steinberger H.-U, Müller T, Auner N, Maerker C, Schleyer P. vR. Angew. Chem. Int. Ed. Engl. 1997; 36: 626
- 13 Müller T, Bauch C, Ostermeier M, Bolte M, Auner N. J. Am. Chem. Soc. 2003; 125: 2158
- 14 de Meijere A. Angew. Chem. Int. Ed. Engl. 1979; 18: 809
- 15 Wang J, Blaszczyk SA, Li X, Tang W. Chem. Rev. 2021; 121: 110
- 16 Ganesh V, Chandrasekaran S. Synthesis 2016; 48: 4347
- 17 Jiao L, Yu Z.-X. J. Org. Chem. 2013; 78: 6842
- 18 Goldschmidt Z, Crammer B. Chem. Soc. Rev. 1988; 17: 229
- 19 Fleming I. Chem. Soc. Rev. 1981; 10: 83
- 20 Klare HF. T, Albers L, Süsse L, Keess S, Müller T, Oestreich M. Chem. Rev. 2021; 121: 5889
- 21 He T, Wang G, Bonetti V, Klare HF. T, Oestreich M. Angew. Chem. Int. Ed. 2020; 59: 12186
- 22 Schäfer A, Reißmann M, Schäfer A, Saak W, Haase D, Müller T. Angew. Chem. Int. Ed. 2011; 50: 12636
- 23 Schäfer A, Reißmann M, Jung S, Schäfer A, Saak W, Brendler E, Müller T. Organometallics 2013; 32: 4713
- 24 Labbow R, Reiß F, Schulz A, Villinger A. Organometallics 2014; 33: 3223
- 25 Omann L, Pudasaini B, Irran E, Klare HF. T, Baik M.-H, Oestreich M. Chem. Sci. 2018; 9: 5600
- 26 Wu Q, Qu Z.-W, Omann L, Irran E, Klare HF. T, Oestreich M. Angew. Chem. Int. Ed. 2018; 57: 9176
- 27 Long P.-W, He T, Oestreich M. Org. Lett. 2020; 22: 7383
- 28a Zhao Y, Truhlar DG. Theor. Chem. Acc. 2008; 120: 215
- 28b Zhao Y, Truhlar DG. Acc. Chem. Res. 2008; 41: 157
- 28c The solvent effect was taken into consideration using a polarizable continuum model (PCM) for both geometry optimizations and single-point energy calculations, see: Tomasi J, Persico M. Chem. Rev. 1994; 94: 2027
- 29 Heiden ZM, Lathem AP. Organometallics 2015; 34: 1818
- 30 Long P.-W, Oestreich M. Org. Lett. 2021; 23: 4834
- 31 He T, Wang G, Long P.-W, Kemper S, Irran E, Klare HF. T, Oestreich M. Chem. Sci. 2021; 12: 569
- 32 Chen Q.-A, Klare HF. T, Oestreich M. J. Am. Chem. Soc. 2016; 138: 7868
- 33 Wu Q, Irran E, Müller R, Kaupp M, Klare HF. T, Oestreich M. Science 2019; 365: 168
- 34 Allemann O, Duttwyler S, Romanato P, Baldridge KK, Siegel JS. Science 2011; 332: 574
- 35 He T, Klare HF. T, Oestreich M. ACS Catal. 2021; 11: 12186
- 36 Klare HF. T, Oestreich M. J. Am. Chem. Soc. 2021; 143: 15490
- 37 Long P.-W, Wang G, Klare HF. T, Oestreich M. ACS Catal. 2022; 12: 12310
- 38 Park S, Lee D. J. Am. Chem. Soc. 2006; 128: 10664
- 39 Horino Y, Luzung MR, Toste FD. J. Am. Chem. Soc. 2006; 128: 11364
- 40 Matsuda T, Kadowaki S, Yamaguchi Y, Murakami M. Chem. Commun. 2008; 2744
- 41 Matsuda T, Yamaguchi Y, Shigeno M, Sato S, Murakami M. Chem. Commun. 2011; 47: 8697
- 42 Zhou T, Xu L, Xia Y. Org. Lett. 2013; 15: 6074
- 43 Zhou H, Han JT, Nöthling N, Lindner MM, Jenniches J, Kühn C, Tsuji N, Zhang L, List B. J. Am. Chem. Soc. 2022; 144: 10156
- 44 Lambert JB, Zhao Y, Wu H, Tse WC, Kuhlmann B. J. Am. Chem. Soc. 1999; 121: 5001
For early computational and experimental data on β-silyl carbocations, see:
For more recent computational analyses, see:











































