Synthesis of Silylated Cyclobutanone and Cyclobutene Derivatives Involving 1,4‐Addition of Zinc‐Based Silicon Nucleophiles [original]

Synthesis of Silylated Cyclobutanone and Cyclobutene

Derivatives Involving 1,4-Addition of Zinc-Based Silicon

Nucleophiles

Ming Cui[a] and Martin Oestreich*[a]

Abstract: A copper-catalyzed conjugate silylation of various

cyclobutenone derivatives with Me2PhSiZnCl·2LiCl or

(Me2PhSi)2Zn·xLiCl (x�4) to generate β-silylated cyclobuta-

nones is reported. Trapping the intermediate enolate with

ClP(O)(OPh)2affords silylated enol phosphates that can be

further engaged in Kumada cross-coupling reactions to

yield silylated cyclobutene derivatives.

Conjugate addition of silicon nucleophiles to α,β-unsaturated

carbonyl compounds is one of the standard processes for the

formation of C(sp3)Si bonds.[1] The resulting β-silylated

carbonyl compounds[2] can be converted into the correspond-

ing aldols by oxidative degradation of that C(sp3)Si bond.[3] As

to cyclic acceptors, the vast majority of protocols are for

cyclopentenone and -hexenone derivatives.[4,5] Murakami and

co-workers reported the 1,4-addition to cyclobutenone deriva-

tives using Fleming’s (Me2PhSi)2CuLi·LiCN[4a,b] to access func-

tionalized 1,3-dienes after trapping of the enolate intermediate

and electrocyclic ring-opening (Scheme 1, top).[6] Aside from

this isolated example, there are no further methods known,

neither stoichiometric nor catalytic in copper.

Almost 20 years ago, our laboratory introduced copper-

catalyzed and even copper-free protocols for conjugate

silylation employing bis(triorganosilyl)zinc and tris(triorgano-

silyl)zincate reagents.[7–9] We also found copper salts to accel-

erate these reactions and to be essential for hindered and β,β-

disubstituted acceptors, respectively.[8] Zinc-based silicon nucle-

ophiles such as (Me2PhSi)2Zn·4LiCl and also Me2PhSiZnCl·2LiCl

are in fact highly useful. Their functional-group tolerance is

substantially improved over that of the corresponding more

reactive lithium compounds from which the zinc reagents are

typically prepared by transmetalation. To date, none of these

protocols have been applied to cyclobutenones. Moreover, the

synthesis of cyclobutyl-substituted silanes is limited to a few

examples. In 2010, Ito and co-workers reported a copper-

catalyzed borylation of silyl-substituted homoallylic sulfonates,

and cyclobutylsilane derivatives were obtained by insertion of

the CC double bond into an in situ formed CuB bond

followed by an intramolecular SN2 reaction.[10] The Fu group[11]

and our group[12] reported single examples of the synthesis of

cyclobutylsilanes by metal-catalyzed radical cross-coupling of a

tertiary and a secondary cyclobutyl bromide with zinc- and

magnesium-based silicon reagents, respectively. In this work,

we describe copper-catalyzed conjugate silylations of highly

substituted cyclobutenone derivatives with zinc-based silicon

reagents (Scheme 1, bottom). The intermediate metal enolates

can either be hydrolyzed to afford 3-silyl-substituted cyclo-

butanones or captured with ClP(O)(OPh)2as an electrophile to

furnish cyclobutenyl phosphates. Subsequent Kumada cross-

coupling yields silicon-containing cyclobutene derivatives.

Our study commenced with the conjugate silylation of

cyclobutenone 1a with 2.0 equiv. of Me2PhSiZnCl·2LiCl in

THF[13] (Table 1). Using Cu(CH3CN)4PF6as the catalyst in THF at

room temperature, β-silylated β-phenylcyclobutanone 2a was

obtained in 95% yield after hydrolysis (entry 1). Yields were

slightly lower with less silicon nucleophile, for example 91%

yield with 1.5 equiv. of Me2PhSiZnCl·2LiCl. Given the possibility

of a copper-free 1,4-addition,[8] we compared different β-

[a] M. Cui, Prof. Dr. M. Oestreich

Institut für Chemie

Technische Universität Berlin

Strasse des 17. Juni 115, 10623 Berlin (Germany)

E-mail: [email protected]

Supporting information for this article is available on the WWW under

https://doi.org/10.1002/chem.202102993

VCH GmbH. This is an open access article under the terms of the Creative

Commons Attribution License, which permits use, distribution and re-

production in any medium, provided the original work is properly cited.

Scheme 1. Conjugate silylation of cyclobutenone derivatives and follow-up

chemistry of the in situ-formed enolates.

Chemistry—A European Journal

Communication

doi.org/10.1002/chem.202102993

Wiley VCH Donnerstag, 11.11.2021

2165 / 221299 [S. 16103/16106] 1

substituted and α,β-disubstituted cyclobutenones in reactions

with and without the copper catalyst. The silylation of 1a in the

absence of Cu(CH3CN)4PF6did lead to 2a yet with a substantial

decrease in yield (entry 1). Other cyclobutenones such as β-

butyl-substituted 1g and α,β-disubstituted 1o and 1p were

tested, and the low yields of the copper-free protocol confirmed

the importance of a copper catalyst (entries 2–4).

We further tested the substrate scope of this conjugate

silylation (Scheme 2). β-Aryl-substituted cyclobutenones were

generally suitable substrates, affording the corresponding β-

silylated cyclobutanones in good to excellent yields (1a–f!2a–

f). Electron-donating groups at the aryl ring such as methyl and

methoxy led to higher yields than halogenated derivatives.

Likewise, cyclobutenones bearing a primary alkyl substituent in

the β-position furnished the corresponding products in equally

high yields (1g–k!2g–k); the yield was lowest for 1 k

containing a C(sp3)Cl bond. With sterically more demanding

secondary alkyl groups such as cyclopropyl and cyclohexyl,

yields were still good (1l,m!2l,m). A silyl group in the β-

position was also compatible (1n!2n). The reactions of α,β-

disubstituted cyclobutenones 1o and 1p proceeded equally

well. Product 2o was obtained with high diastereoselectivity

while 2p formed with a poor diastereomeric ratio. We believe

that the diastereoselectivity is mainly controlled by steric factors

in the protolysis of the enolate intermediate.

Next, we tried to capture the enolate intermediate as an

enol phosphate,[14] that is cyclobutenyl phosphates 3, to allow

for subsequent cross-coupling reactions.[15] The brief survey

outlined in Table 2 shows that copper-catalyzed 1,4-addition of

either Me2PhSiZnCl· 2LiCl or (Me2PhSi)2Zn·xLiCl (x�4) to 1 a

followed by enolate trapping with ClP(O)(OPh)2furnishes the

enol phosphate 3a in moderate yields (entries 1 and 2).

Relevant to an enantioselective variant, no uncatalyzed back-

ground reaction was seen with an almost salt-free stock

solution of (Me2PhSi)2Zn· xLiCl in Et2O[16] (entry 2). In the light of

our recent work about an enantioselective conjugate silylation

with a zinc-based silicon nucleophile,[17] we decided to inves-

tigate the asymmetric version. The yield increased in the

presence of the chiral phosphoramidite ligand (S,R,R)-L1 but

enantioinduction was low, even at 78 °C (entries 3 and 4). A

Table 1. Comparison of copper-catalyzed and copper-free protocols with

Me2PhSiZnCl·2LiCl.[a]

Entry Acceptor Product Yield of 2[%][b]

w/ Cu(CH3CN)4PF6

Yield of 2[%][c]

w/ o Cu(CH3CN)4PF6

11a 2a 95 71

21g 2g quant. 24

31o 2o 95 30

41p 2p 95 0

[a] All reactions were performed on a 0.2 mmol scale for 2 h. [b] Isolated

yield after flash chromatography on silica gel. [c] Determined by 1H NMR

spectroscopy by using CH2Br2as the internal standard.

Scheme 2. Synthesis of β-silylated cyclobutanones by conjugate addition of

Me2PhSiZnCl·2LiCl. Unless otherwise noted, all reactions were performed on

a 0.2 mmol scale for 2 h. Yields are of analytically pure product obtained

after flash chromatography on silica gel. The relative configuration was

assigned by 1H NMR spectroscopic analysis prior to purification (see the

Supporting Information for details). [a] Value in parentheses for the reaction

on a 1.0 mmol scale.

Table 2. Comparison of copper-catalyzed and copper-free protocols with

enolate trapping.[a]

Entry Zinc-based silicon

nucleophile

Yield of 3a [%][b]

w/ Cu(CH3CN)4PF6

Yield of 3a [%][b]

w/ o Cu(CH3CN)4PF6

1 Me2PhSiZnCl·2LiCl

(2.0 equiv.)

56 18

2 (Me2PhSi)2Zn·xLiCl

(1.2 equiv.)

36 0

3 (Me2PhSi)2Zn·xLiCl

(1.2 equiv.)

53 (6% ee)[c]

w/ L1

–

4[d] (Me2PhSi)2Zn·xLiCl

(1.2 equiv.)

55 (15% ee)[c]

w/ L1

–

[a] All reactions were performed on a 0.2 mmol scale for 2 h. [b]

Determined by 1H NMR spectroscopy by using CH2Br2as the internal

standard. [c] Determined by HPLC analysis on a chiral stationary phase. [d]

The 1,4-addition was conducted at 78°C for 16 h prior to the addition of

ClP(O)(OPh)2.

Chemistry—A European Journal

Communication

doi.org/10.1002/chem.202102993

Wiley VCH Donnerstag, 11.11.2021

2165 / 221299 [S. 16104/16106] 1

systematic screening of various chiral ligands was completely

unsuccessful (see the Supporting Information for the details).

However, the yield could be improved to 76% with no

enantioselectivity with (R,R,R)-L2 (see Scheme 3), and we

continued using this ligand for the reaction scope (a racemic

ligand such as rac-binap afforded significantly lower yields;

19% yield). For completion, the corresponding 1,4-addition of

Me2PhSiZnCl·2LiCl in the presence of (S,R,R)-L1 proceeded with

no enantioinduction.

The optimized reaction conditions are 5.0 mol% of Cu-

(CH3CN)4PF6and 6.0 mol% of L2 in THF with 1.2 equiv. of

(Me2PhSi)2Zn·xLiCl as the silicon source and ClP(O)(OPh)2as the

electrophilic trapping reagent (Scheme 3). The reaction scope

was done with the same set of cyclobutenones 1 a–p(cf.

Scheme 2). Yields were good throughout with β-aryl-substituted

cyclobutenones (1a–f!3a–f). Conversely, the β-alkyl-substi-

tuted derivatives were less reactive, and moderate yields were

obtained (1g–m!3g–m). Again, a silyl group as in 1n was

tolerated to give 3n in 52% yield. Both α,β-disubstituted

substrates 1o and 1 p did react in acceptable yields, affording

fully substituted enol phosphates 3o and 3p, respectively.

Enol phosphates can serve as electrophiles in cross-coupling

reactions,[15] and we tested several of the above cyclobutenyl

phosphates in Kumada coupling reactions (3!4, Scheme 4).

These representative reactions proceeded in moderate yields in

the presence of catalytic amounts of (dppe)NiCl2.[18] Arylation

with PhMgBr reliably gave the corresponding silylated cyclo-

butenes. In turn, alkylation with the primary alkyl Grignard

reagent n-HexMgBr was low yielding but an acceptable yield

was restored with secondary CyMgBr.

To summarize, we reported here a copper-catalyzed con-

jugate addition of zinc-based silicon reagents to highly

substituted cyclobutenones, providing a general and efficient

method to access various β-silylated cyclobutanones. Moreover,

the enolate intermediate can be trapped with a phosphorus

electrophile to arrive at silylated enol phosphates, and these

can be converted into the corresponding cyclobutenes by

Kumada cross-coupling.

Acknowledgements

M.C. thanks the China Scholarship Council for a predoctoral

fellowship (2018–2022). M.O. is indebted to the Einstein

Foundation Berlin for an endowed professorship. Open Access

funding enabled and organized by Projekt DEAL.

Conflict of Interest

The authors declare no conflict of interest.

Keywords: conjugate addition ·copper ·silicon ·synthetic

methods ·zinc

[1] a) E. Hartmann, D. J. Vyas, M. Oestreich, Chem. Commun. 2011,47,

7917–7932; for a recent Outlook, see: b) W. Xue, M. Oestreich, ACS Cent.

Sci. 2020,6, 1070–1081.

[2] F. Nahra, O. Riant in Science of Synthesis: Knowledge Updates 2013/2 (Ed.:

M. Oestreich), Thieme, Stuttgart, 2013, pp. 197–214.

[3] G. R. Jones, Y. Landais, Tetrahedron 1996,52, 7599–7662.

Scheme 3. Synthesis of silylated cyclobutenyl phosphates by sequential

conjugate addition of (Me2PhSi)2Zn·xLiCl (x�4) and enolate trapping. Unless

otherwise noted, all reactions were performed on a 0.2 mmol scale. Yields

are of analytically pure product obtained after flash chromatography on

silica gel. [a] Value in parentheses for the reaction on a 1.5 mmol scale.

Scheme 4. Nickel-catalyzed Kumada cross-coupling of silylated cyclobutenyl

phosphates and Grignard reagents. Unless otherweise noted, all reactions

were performed on a 0.10 mmol scale. Yields are of analytically pure product

obtained after flash chromatography on silica gel. [a] Value in parentheses

for the reaction on a 1.0 mmol scale. [b] Performed on a 0.065 mmol scale.

Chemistry—A European Journal

Communication

doi.org/10.1002/chem.202102993

Wiley VCH Donnerstag, 11.11.2021

2165 / 221299 [S. 16105/16106] 1

[4] a) D. J. Ager, I. Fleming, J. Chem. Soc. Chem. Commun. 1978, 177–178;

b) D. J. Ager, I. Fleming, S. K. Patel, J. Chem. Soc. Perkin Trans. 1 1981,

2520–2526; c) B. H. Lipshutz, J. A. Sclafani, T. Takanami, J. Am. Chem.

Soc. 1998,120, 4021–4022.

[5] a) C. Walter, G. Auer, M. Oestreich, Angew. Chem. Int. Ed. 2006,45, 5675–

5677; Angew. Chem. 2006,118, 5803–5805; b) C. Walter, R. Fröhlich, M.

Oestreich, Tetrahedron 2009,65, 5513–5520; c) K.-s. Lee, A. H. Hoveyda,

J. Am. Chem. Soc. 2010,132, 2898–2900.

[6] M. Murakami, Y. Miyamoto, Y. Ito, J. Am. Chem. Soc. 2001,123, 6441–

6442.

[7] M. Oestreich, B. Weiner, Synlett 2004, 2139–2142.

[8] G. Auer, B. Weiner, M. Oestreich, Synthesis 2006, 2113–2116.

[9] A. Weickgenannt, M. Oestreich, Chem. Eur. J. 2010,16, 402–412.

[10] H. Ito, T. Toyoda, M. Sawamura, J. Am. Chem. Soc. 2010,132, 5990–5992.

[11] C. K. Chu, Y. Liang, G. C. Fu, J. Am. Chem. Soc. 2016,138, 6404–6407.

[12] W. Xue, R. Shishido, M. Oestreich, Angew. Chem. Int. Ed. 2018,57,

12141–12145; Angew. Chem. 2018,130, 12318–12322.

[13] L. Zhang, M. Oestreich, Org. Lett. 2018,20, 8061–8063.

[14] C. Zhong, Y. Huang, H. Zhang, Q. Zhou, Y. Liu, P. Lu, Angew. Chem. Int.

Ed. 2020,59, 2750–2754; Angew. Chem. 2020,132, 2772–2776.

[15] Y. Nassar, F. Rodier, V. Ferey, J. Cossy, ACS Catal. 2021,11, 5736–5761.

[16] A solvent change from THF to Et2O was done to precipitate and remove

as much LiCl as possible. As we did not quantify the residual LiCl, we

decided to continue with Me2PhS2Zn·xLiCl (x�4). C. Fopp, E. Romain, K.

Isaac, F. Chemla, F. Ferreira, O. Jackowski, M. Oestreich, A. Perez-Luna,

Org. Lett. 2016,18, 2054–2057.

[17] L. Zhang, M. Oestreich, ACS Catal. 2021,11, 3516–3522.

[18] A. S. E. Karlström, K. Itami, J.-E. Bäckvall, J. Org. Chem. 1999,64, 1745–

1749.

Manuscript received: August 16, 2021

Accepted manuscript online: September 7, 2021

Version of record online: October 7, 2021

Chemistry—A European Journal

Communication

doi.org/10.1002/chem.202102993

Wiley VCH Donnerstag, 11.11.2021

2165 / 221299 [S. 16106/16106] 1

Loading more pages...