The Tethered Silanoxymercuration of Allylic Alcohols

Abstract

We present the first examples of tethered olefin functionalization reactions using a silanol auxiliary. A range of allylic alcohols are readily condensed with di-tert-butylsilyl bis(trifluoromethanesulfonate) to form allylic silanols. When treated with Hg(OTf)₂ and NaHCO₃, these silanols facilely transform into cyclic silanediol organomercurial compounds. In most cases, the reactions are exquisitely diastereoselective. The scale can be increased more than ten-fold without loss of yield and selectivity. We demonstrate that the product silanediols are versatile synthons for a variety of further reactions.

Graphical Abstract

The family of organomercurials has earned an unfortunate reputation, largely due to the actions of a minority of its members.^{1, 2} Nevertheless, the majority of higher order organomercurial compounds are crystalline, air-stable, and, when treated with appropriate respect, no more dangerous than most chemicals encountered in the organic laboratory.³ The C–Hg linkage is essentially covalent, making it a most unique and versatile organometallic bond. ^4–6 Our laboratory is deeply invested in the field of tethered olefin functionalization reactions.^7–9 In such reactions, a nucleophilic auxiliary (the “tether”) is appended to an alcohol or amine and then cyclized onto a pendant alkene. Here, we disclose our most recent contribution to this area, a cyclization reaction of allylic silanols into cyclic silanediol mercury chloride compounds. Allylic alcohols can be readily transformed into allylic silanols using a combination of di-tert-butylsilyl bis(trifluoromethanesulfonate)¹⁰ and imidazole; the free OH of the silanol tether serves as a convenient nucleophile for olefin attack.

The resulting cyclic silanediol compounds, reminiscent of intermediates en route to polyol natural products (Figure 1),^11–17 are synthons for a variety of further transformations.

Figure 1.

Many antibiotics are polyhydroxylated compounds.

Silanol auxiliaries have been popularized by Gevorgyan and co-workers as directing groups in C–H functionalization reactions.^{18, 19} There is one example from the Lee laboratory of the use of silanols for gold catalyzed alkyne functionalization²⁰, and an example of a silylperoxidation reaction of homoallylic alcohols from the Woerpel laboratory.²¹ Nevertheless, our survey of the literature revealed no examples of the silanol tether employed in alkene functionalization. Our work has been directly inspired by the pioneering contributions of Overman^{22, 23} and Leighton^24–28, who demonstrated that acetal tethers could be used for the mercuric-salt mediated functionalization of olefins (Scheme 1).^{29, 30}

Scheme 1.

A strategy for the conversion of alkenyl alcohols into diol synthons.

Optimization of our silanol-tethered alkene functionalization reaction (Table 1) was performed with (E)-(but-2-en-1-yloxy)di-tert-butylsilanol which was synthesized in one step from crotyl alcohol and di-tert-butylsilyl bis(trifluoromethanesulfonate) (See Supporting Information for full experimental procedures). With 1 equivalent of Hg(OCOCF₃)₂, we were pleased to observe formation of 22% of desired cyclic silanediol (Table 1, Entry 1), giving us hope that our conceived reaction was viable. There was little improvement with increase of temperature (Table 1, Entry 2), equivalents of Hg(OCOCF₃)₂ (Table 1, Entry 3), or time (Table 1, Entry 4). We hypothesized that adventitious trifluoroacetic acid, formed as a byproduct of cyclization, was deleterious to reaction progress. Accordingly, we tested K₂CO₃ and NaHCO₃ in varying equivalents (Table 1, Entries 5-7) as base additives. Nevertheless, we observed little change in reaction performance. Reducing the temperature to −40 °C was markedly deleterious (Table 1, Entry 8). To our amazement and excitement, at this temperature, simply switching to Nishizawa’s salt (Hg(OTf)₂),³¹ in combination with NaHCO₃ (1 equivalent) afforded clean and high-yielding conversion to the desired product (Table 1, Entry 9). It should be noted that the use of 1 equivalent of NaHCO₃ was critical for reaction performance! In its absence, a complex mixture of decomposition products was observed (Table 1, Entry 10).

Table 1.

Reaction Optimization

Entrya	Hg (II)(equiv.)	Base	Temp/Time	P/RSM
1	Hg(OCOCF3)2(1)	None	23 °C, 1h	22/45
2	Hg(OCOCF3)2(1)	None	40 °C, 1h	25/28
3	Hg(OCOCF3)2 (2)	None	23 °C, 1h	29/30
4	Hg(OCOCF3)2 (1)	None	23 °C, 16h	31/13
5	Hg(OCOCF3)2 (1)	K2CO3 (1)	23 °C, 16h	34/18
6	Hg(OCOCF3)2 (1)	NaHCO3 (1)	23 °C, 16h	34/18
7	Hg(OCOCF3)2 (1)	NaHCO3 (2)	23 °C, 16h	38/8
8	Hg(OCOCF3)2 (1)	NaHCO3 (1)	−40 °C, 16h	6/63
9	Hg(OTf)2 (1)	NaHCO3 (1)	−40 °C, 16h	67/<5
10	Hg(OTf)2 (1)	None	−40 °C, 16h	—b

^aYield estimated with methyl phenyl sulfone as a ¹H NMR internal standard.

^bComplex mixture of products.

Our optimized protocol utilizing Hg(OTf)₂ (1 equivalent) and NaHCO₃ (1 equivalent) at −40 °C in THF was compatible with a wide array of alkenyl silanol substrates (Scheme 2). The silanol auxiliary could be appended to both primary allylic alcohols as well as secondary ones (Scheme 2, Entry 9). In all but two instances (Scheme 2, Entry 10), the cyclization reaction proceeded with excellent diastereoselectivity (>20:1), with the pendant alkyl or aryl group and mercury chloride substituent in a trans relationship. A crystal structure of 31 (Scheme 2, Entry 6) unambiguously established this stereochemistry. It is remarkable to note that even with products containing 3 stereocenters (Scheme 2, Entry 9), only a single diastereomer was observed. We hypothesize that a chair-like transition state during silanoxy-mercuration is responsible for this excellent diastereoselectivity (Scheme 3). The reaction was tolerant of a wide variety of alkyl and aryl substituents attached to the olefin. Ethers (Scheme 2, Entries 2, 5, 6, 7), halogens (Scheme 2, Entry 6), and several heterocycles (Scheme 2, Entries 7 and 8) were all compatible. Di-substituted trans-olefins were the most competent substrates in this cyclization. While di-substituted cis-olefins also reacted (Scheme 2, Entry 11), the yield of cyclized product dropped. Furthermore, an endo mode of cyclization was strongly preferred; exo cyclization (Scheme 2, Entry 12) proceeded in only low yield.

Scheme 2.

Substrate Scope

^areactions conducted on a 0.2 mmol scale and relative stereochemistry is shown in all cases. ^bIsolated yields. ^cCrystallographic information deposited in the Cambridge Database (CCDC), Number 2032765 ^ddr = 1:1.

Scheme 3.

A chair-like transition state likely underlies the high diastereoselectivity of cyclization.

With tri-substituted alkenes, the cyclization reaction proceeded only partially, if at all (Scheme 4). Nevertheless, in these reactions, alcohol products were isolated, which themselves may be valuable intermediates for further derivatization.³

Scheme 4.

Some substrates do not fully cyclize but still form valuable diol products.

We were pleased to find that our cyclization reaction scaled greater than 10-fold with no loss in yield or selectivity (Scheme 5).

Scheme 5.

Cyclization scales greater than 10-fold with no loss of yield and selectivity.

Furthermore, as we had envisioned at the conception of this project, the cyclic silanediol organo-mercury products could be used as starting materials for a variety of further transformations, including hydroxylation,^32–34 demercuration, ^35–38 and iodination^39–41 (Scheme 6A–C).

Scheme 6.

The C–Hg organometallic bond is highly versatile.

In summary, we present the first tethered olefin functionalization using a silanol auxiliary. The silanol tether could be conveniently appended to a variety of primary and secondary alcohols via condensation with di-tert-butylsilyl bis(trifluoromethanesulfonate). Hg(OTf)₂ mediated cyclization proceeded with high diastereoselectivity and afforded cyclic silanediol mercury chloride products. The reaction was scalable greater than ten-fold and the products were readily amenable for further demercurative transformations. We expect this reaction to find much application in the pursuit of important polyhydroxylated compounds.