[1,2]- and [1,4]-Wittig Rearrangements of α-Alkoxysilanes: Effect of Substitutions at both the Migrating Benzylic Carbon and the Terminal sp² Carbon of the Allyl Moiety

Abstract

Substituted α-alkoxysilanes can be deprotonated by alkyllithium bases and made to undergo Wittig rearrangements to afford the #x0005B;1,4]- and [1,2]-rearranged products in varying ratios. Substitution at the benzylic migrating carbon and/or at the allylic carbon of the allyl moiety impacts the rearrangement reaction, influencing the reactivity as well as the [1,4]-/[1,2]-selectivity. Diastereomeric α-alkoxysilanes show different reactivities with the syn diastereomer being the more reactive isomer.

1. Introduction

The isomerizations of ethers under basic conditions to the corresponding alcohols are known as Wittig rearrangements and can proceed via two main pathways: A stepwise [1,2]-migration involving a radical/radical-anion species¹ or via a concerted symmetry-allowed [2,3]-shift.² In addition, during the reorganization of allylic ethers [1,4]-Wittig rearrangements can compete with the [1,2]-migration and give rise to carbonyl compounds.³ Despite its potential to construct adjacent stereocenters, transfer chirality, or obtain stereodefined enolates that could be used in subsequent transformations, very few examples where [1,4]-Wittig rearrangements predominate over [1,2]-pathways exist in the literature.⁴ Furthermore, although some experiments support a stepwise [1,4]-Wittig process, the involvement of a concerted mechanism has not been discarded.⁵

In an earlier communication,^6a we showed that unsubstituted α-alkoxysilane 1 (R₁ = R₂ = H) rearranged exclusively via the [1,4]-pathway at low temperatures to give acylsilane 2 in 80% yield (Scheme 1). We observed that the [1,2]-pathway became competitive with increasing temperature, leading to a gradual erosion in [1,4]-selectivity and producing mixtures of 2 and the isomeric [1,2]-Wittig product 4, with the [1,4]-/[1,2]-ratio reaching 2:1 at room temperature.^6,7 Following this report, we questioned whether a high [1,4]-/[1,2]-selectivity could be retained if we made structural changes in our model substrate 1 (e.g. where R₁ and/or R₂ ≠ H).

Scheme 1.

[1,4]-/[1,2]Wittig rearrangements of α-alkoxysilanes.

Ref 6a: R₁ = R₂ = H; conditions: sec-BuLi (1.5 equiv), THF, −78 °C, 30 min; yield of 2: 80%, 3 and 4 not observed

Nakai and co-workers previously showed that the scope and limitations of the [1,2]-Wittig rearrangement are determined principally by the migratory aptitude of the alkyl group (primary < secondary < tertiary = allyl < benzyl) thus following increasing radical stability.¹ In cases of limited radical stability of the migrating group, carbanion-stabilizing groups similarly facilitate [1,2]-migrations.¹ In contrast, the yield of [1,4]-Wittig products has been reported to be relatively insensitive to substitution at the α- or γ-position of the allylic moiety,³ although Schlosser observed that [1,4]-/[1,2]-selectivity is diminished with increasing alkyl substitution about the migrating carbon.⁸ In the context of these previous findings, we set out to systematically investigate the introduction of alkyl substitution at the benzylic carbon and/or terminal allylic carbon of α-alkoxysilanes so as to gain insight into the steric and stereochemical factors that might control the course of Wittig rearrangements of α-alkoxysilanes.⁹

2. Results and discussion

We initially studied the effect of alkyl substituents at the terminal sp² carbon of the allyl moiety (R₁ = alkyl, R₂ = H in 1, Scheme 1). Compounds 5, 8 and 11, possessing a methyl, n-hexyl and t-butyl at the terminal olefin carbon, respectively, were prepared from the corresponding α-(trimethylsilyl)-γ-alkyl allyl alcohols by acid-catalyzed O-alkylation with benzyl trichloroacetimidate.¹⁰

Subjection of methyl-substituted compound 5 to our previously developed conditions (sec-BuLi, THF, −78 °C) ⁶ afforded in 74% overall yield a 4:1 ratio of acylsilane 6 and alcohol 7 resulting from [1,4]- and [1,2]-Wittig rearrangements, respectively (Scheme 2). The effect on the rate of deprotonation/rearrangement was negligible, consistent with the fact that the site of deprotonation is at a relatively remote position with respect to the alkyl substituent. The observed erosion of the [1,4]-/[1,2]-selectivity with substitution at the terminal allylic position is in apparent contrast with literature reports.³ A longer alkyl chain at the γ-position in compound 8 led to lower ratio (2:1) of [1,4]-/[1,2]- Wittig products 9 and 10 in 78% overall yield. This trend reached a maximum point in the case of compound 11 in which the bulky t-butyl group completely inhibited the [1,4]-Wittig shift (Scheme 2). Allylic deprotonation of 11 preceeded [1,2]-Wittig rearrangement to give alcohol 12, in addition, benzylic deprotonation was competitive and allowed a [2,3]-sigmatropic shift to afford alcohol 13 as a single diastereomer. Thus, bulkier alkyl groups at the γ-carbon appear to lower the reactivity towards deprotonation at the allylic position, allowing competitive benzylic deprotonation. In fact, 11 underwent incomplete conversion in the same period of time that compounds 5 and 8 were completely consumed. In addition, increasing steric hindrance increases the selectivity in favor of the [1,2]-Wittig product. This could be interpreted as an increasingly prohibitive [1,4]- recombination of the radical and radical-anion fragments, but also as the difficulty of the α-(benzyloxy)allyllithium species to adopt an optimum conformation for a concerted [1,4]-shift. Although the inverse relationship between steric demand of the γ-alkyl group and the [1,4]-/[1,2]- selectivity is consistent with both an stepwise and concerted mechanisms for the [1,4]-migration, at this point it is difficult to predict which mechanism would be more susceptible to such steric effects.

Scheme 2.

Effect of alkyl substitution at the terminal sp² carbon of the allyl moiety.

We continued our studies by analyzing the influence of substituents at the migrating (benzylic) carbon, a structural change that inherently led to diastereomeric substrates. Our simplest models, α-(trimethylsilyl)allyl ethers syn–14 and anti–14 bearing a methyl group at the benzylic position, were not separable by column chromatograph on silica gel, thus a 1:1.4 mixture of syn–14/anti–14 was employed (Scheme 4). Employing such a mixture did not constitute a problem since it was possible to monitor the reaction by ¹H NMR. Under our standard reaction conditions we found that the most reactive diastereomer (syn–14) was consumed within 8 hours whereas the ‘less reactive’ diastereomer anti-14 was recovered in 43% (Scheme 3). The [1,4]- and isomeric [1,2]-products (15 and 16, respectively) were obtained in a ratio of 1.5:1 and in a combined 35% yield. It was observed that allowing the reaction to proceed overnight resulted in an increased overall yield of 15 and 16 (46%) and basically the same [1,4]-/[1,2]- ratio (1.7:1) with a corresponding decrease in the recovery of the ‘less reactive’ anti-14 (26%). The near complete erosion of the [1,4]-/[1,2]- selectivity (>99:1 in compound 1)⁶ is in agreement with the detrimental effect of increasing substituents at the migrating carbon in the [1,4]-/[1,2]-selectivity observed by Schlosser.^8b

Scheme 4.

Wittig rearrangements of α-alkoxysilanes 14, 17, and syn-20 bearing a substituent (methyl, 2-propenyl and iso-propyl, respectively) at the benzylic carbon.

Scheme 3.

Plausible mechanism for the isomerization of the [1,2] Wittig alkoxide i to enolate iii.

The marked difference in the reactivity of diastereomers syn–14 and anti–14 points to the determinant role of relative stereochemistry, and more specifically the steric environment around the allylic proton (α to silicon) in allowing the key deprotonation step to take place prior to rearrangement.¹² The relative ability of the syn and anti diastereomers to adopt an optimum conformation for deprotonation might be responsible for the observed difference in reactivity. We propose that the allylic C-H bond should be perpendicular to the olefin and therefore aligned with the π system. At the same time, antiperiplanar alignment of the allylic C-H bond to the cleaving C-O would allow weakening of the C-H bond. The phenyl group would take the less crowded and furthest position, maximizing conjugation by aligning with the C-O bond and thus leading to the pseudo-eclipsed conformers shown in Figure 1. These proposed conformational requirements pose a more severe steric interaction in anti-14, the less reactive diastereomer, in which the pseudo-eclipsing Methyl and TMS groups collide. On the other hand, in syn-14 the TMS group is pseudo-eclipsed with H_b and a less unfavorable steric interaction between the benzylic methyl and vinyl groups is possible. Alternatively, positioning the benzylic H_b proton in an “eclipsed” alignment with the TMS groups in both anti-14 and syn-14 would lead to an more unfavorable steric interaction in anti-14 (Ph vs vinyl) than in syn-14 (Me vs vinyl).

Figure 1.

Proposed relevant conformers for allylic deprotonation in syn-14 and anti-14.

Our conformational analysis is consistent with the fact that increasing the steric demand of the substituent at the benzylic position (methyl in 14) dramatically reduced deprotonation rate. The diastereomeric 2-propenyl (17) and the isopropyl (20) analogues were unreactive under standard reaction conditions (sec-BuLi, THF, −78 °C, 24 h). In these two cases the use of a less bulky base (n-BuLi) was necessary to effect a reaction. A mixture of syn-17/anti-17 (2.6:1) required 30 h for complete reaction at −30 °C, yielding acylsilane 18 and ketone 19 in a 4.3:1 ratio. The seemingly higher [1,4]-/[1,2]-selectivity is clouded by the reaction also affording a complex mixture of alkylated and otherwise unidentified byproducts. A temperature of 0 °C was necessary for the isopropyl substituted syn-20 to undergo deprotonation and rearrangement to give the [1,4]- and isomeric [1,2]-products 21 and 22, in 23% yield (1.8:1 ratio), along with 27% of unreacted syn-20.

We also studied the behavior of diastereomeric ethers 23, containing a para methyl group on the aromatic ring (Scheme 5).⁹ This weakly electron-donating group barely influenced the [1,4]-/[1,2]- selectivity, relative to the unsubstituted analogues 14 (2:1 vs 1.5:1, respectively), providing [1,4]-product 24 and [1,2]-product 25 in 40% overall yield. However, a significant amount of diastereomeric dibenzyl dimer 26 was isolated in significant amounts from the reaction mixture, suggesting that 1) the stepwise mechanism is major pathway in the rearrangements of 23, and 2) that the expected benzyl radical stabilization by the para methyl group prevents radical/radical-anion recombination within the solvent cage,¹³ allowing benzyl radicals to escape and dimerize.

Scheme 5.

Wittig rearrangements of diastereomeric 23 having an electron-rich migrating (benzylic) group.

It is important to mention that we have consistently observed that syn diastereomers (general structure 1, R₂ ≠ H, Scheme 1) are more reactive than the corresponding anti isomers in all cases. The relative stereochemistry of diastereomers syn-14/anti-14 and syn-17/anti-17 was determined as shown in Scheme 5.¹⁴ Derivatization of syn-14/anti-14 to the 3,5-dinitrobenzoyl esters syn-28/anti-28 and crystallization of syn-28 allowed the determination of its crystal structure.¹⁵ On the other hand, ring-closing metathesis of syn-17 and anti-17 followed by NOE studies of the corresponding products trans-29 and cis-29 led to the assignment of relative stereochemistry in syn-17 and anti-17 (Scheme 6).

Scheme 6.

Determination of relative stereochemistry of syn-14/anti-14 and syn-17/anti-17.

Finally, we studied the behavior of substrates bearing substitution at both the migrating carbon and the terminal allylic carbon. These experiments gave us the opportunity to evaluate the effect of olefin geometry not only on the reactivity and selectivity of the rearrangements, but also on the stereochemistry of the bond reorganization.

Compounds 30 were synthesized as geometrically pure Z or E isomers, however, while the Z diastereomers (syn Z-30 and anti Z-30)¹⁶ could be largely separated by column chromatography (dr >19:1), the E diastereomers proved very difficult to separate and therefore were used as a diastereomeric mixture (E-30).

In theory, clean deprotonation of 30 (E or Z) followed by rearrangement should afford pairs of diastereomeric [1,4]- and [1,2]-products (31 and 32 respectively). As described above, further isomerization of the [1,2]-products via silyl migration could also lead to another pair of diastereomeric ketones (33). In practice, syn and anti Z-30 were very unreactive when treated with n-BuLi or other bases¹⁷ at low temperature, and even at room temperature these diastereomers reacted sluggishly. Reaction of syn Z-30 (Scheme 7) with n-BuLi led to almost 50% conversion and ~20% yield of a complex mixture of products. Careful examination and separation of these mixtures revealed that compound 33 was accompanied by [2,3]-Wittig (34),¹⁸ diastereomeric [1,2]- and [1,4]-Wittig products lacking the trimethylsilyl group (35 and 36, respectively) and alkylated products (not shown). The actual [1,2]-Wittig product (32, Scheme 8) was not observed.

Scheme 7.

Substitution at the migrating carbon and terminal sp² carbon, Z isomers.

Scheme 8.

Substitution at the migrating carbon and terminal sp² carbon, E isomers.

Although it was not possible to obtain exact ratios of products or diastereomers from either ¹H NMR or HPLC of crude reaction mixtures, the products could be partially purified allowing their approximate ratios to be determined.¹⁹ The [2,3]-Wittig rearrangement of syn Z-30 proceeded through a standard transition structure^2a to give syn-34 as a single diastereomer,¹⁸ on the other hand all other products from [1,4]- and [1,2]-migrations were obtained in low diastereomeric ratios (ranging from 1:1 to ~3:1). Isolation of the [2,3]-Wittig product is diagnostic of competitive deprotonation at the benzylic position, rather than α to silicon, likely as a consequence of the relatively high temperature required for the desired reaction to occur. The elevated temperature may allow these substrates to overcome an unfavorable conformation for deprotonation of the allylic hydrogen due to the minimization of A(1,3)-strain.²⁰

As expected, anti Z-30 was less reactive under the same reaction conditions. Here, the starting material was recovered in 77% and only a total ~10% yield of products 34–36 was obtained (Scheme 7). The absence of compounds 31–33 suggests that deprotonation α to silicon is inhibited due to severe steric crowding. [2,3]-Wittig rearrangement of anti Z-30 proceeded stereospecifically^2a to give only anti-34.²¹ Compounds 35 and 36, likely to be formed via a silicon/lithium exchange⁷ followed by [1,2]- and [1,4]-Wittig rearrangement, respectively, were obtained again in low diastereomeric ratios. Interestingly, the [1,2]-product 35 showed an inverse diastereoselection in comparison to that observed in the rearrangement of syn Z-30.

Changing the geometry of the olefin had a pronounced effect (Scheme 8). In line with our previous discussion, the reactivity towards initial deprotonation was dominated by the relative configuration at the α and α′ positions of the ethers, as illustrated by the rearrangement of E-30 (syn/anti = 1:1.5). For example, syn E-30 was completely consumed by n-BuLi at low temperature, while its diastereomer anti E-30 was mostly recovered (Scheme 7). Quenching the reaction at −30 °C led to the isolation the [1,4]-Wittig product (31) in 23% yield and with low diastereoselectivity,²² accompanied by the isomeric [1,2]-product (33) also in low yield. Interestingly, quenching the reaction at lower temperature allowed the isolation of the direct [1,2]-Wittig product 32,²³ which in our previous room temperature experiments (Scheme 6) had undergone silicon migration and rearrangement to 33. This was evidenced in an experiment run at 0 °C for 52 hours and quenched at −78 °C, which gave the [1,4]- and [1,2]-Wittig products 31 and 32 in 30% yield (1:1 ratio) with only traces of the isomeric [1,2]-product 33. Thus, in certain cases, quenching the reaction at low temperature significantly reduces silyl migration.

The relative stereochemistry of diastereomeric acylsilane 31 was determined by oxidation to the corresponding carboxylic acid.^{22, 23} The relative stereochemistry of diastereomeric alcohol 32 was not determined. Attempts to derivatize 32 to the corresponding 3,5-dinitrobenzoyl ester failed presumably due to the congested nature of the tertiary alcohol. The relative stereochemistry in isomeric [1,2]-Wittig product (33) was determined as shown in Scheme 9. Compound 33 (dr = 1.4:1) was reduced to the corresponding alcohol 39 as a mixture of only 3 diastereomers. Partial separation of the diastereomers of 39 led to a mixture with a ratio 10:2:1. Benzoylation of this mixture with 3,5-dinitrobenzoyl chloride in pyridine gave 40 as a mixture o diastereomers. Recrystallization from CH₂Cl₂/hexanes gave a single diastereomer and its relative stereochemistry was determined by X-ray. Hydrolysis (NaOH) and oxidation (DMP) of this single isomer (40) gave anti-33, which matched spectroscopically with the major diastereomer in the initial diastereomeric mixture of 33.

Scheme 9.

Determination of the relative stereochemistry of 33 by derivatization to the crystalline ester 40 (major diastereomer) followed by the reverse transformations to anti-33.

3. Conclusions

In conclusion, we have shown that substitution at the migrating carbon impacts the Wittig rearrangement of α-alkoxysilanes, decreasing reactivity towards deprotonation and eroding the [1,4]-/[1,2]-selectivity. Similarly, substitutions at the terminal carbon of the allyl moiety alone or in combination with substitution at the migrating carbon also lowers the [1,4]-/[1,2]-selectivity, especially where substitution comes in the form of Z-olefins. Increasingly bulkier substituents at the terminal allylic position gradually decrease the [1,4]-/[1,2]-selectivity, leading to complete inhibition of the [1,4]-pathway in the bulkier case (t-butyl). The reactivity in these diastereomeric substrates heavily depends on their relative stereochemistry, syn or anti, the former being more reactive in all cases. Taken together these results show the beneficial effect of silyl groups on reactivity (by lowering the acidity of allylic hydrogens) is countered by the steric congestion afforded upon substitution at the benzylic or allylic (or both) positions, which presumably prevents the access of optimal conformation for the key deprotonation step.

4. Experimental section

4.1 General considerations

All reactions were run under nitrogen in previously flame-dried flasks or disposable vials. THF was freshly distilled from sodium benzophenone ketyl. Dichloromethane was distilled from calcium hydride. Cyclohexane and hexane were used without further purification. Ultrapure silica gel 60 (230–400 mesh ASTM) from Silicycle was used for flash column chromatography. Melting points are not corrected.

Nuclear Magnetic Resonance spectra were recorded on 300 MHz, 500 MHz and 600 MHz Varian instruments using CDCl₃ as solvent and referenced to 7.24 ppm (residual proton). IR was recorded on a Perkin-Elmer instrument. High Resolution Mass Spectrometry spectra were recorded at the Michigan State University Mass Spectrometry Facility using a Waters Qtof Ultima (ESI) and JEOL AX 505H (EI/CI) instruments.

4.2 Preparation of α-alkoxysilanes – General procedure A

Trichloroacetimidate of the appropriate alcohol (prepared according to literature procedure)¹⁰ (2.0 equiv) was added to a stirred solution of the requisite α-(trimethylsilyl)allyl alcohol^11d (1.0 equiv) in cyclohexane or hexane (0.2 M) at room temperature. A solution of TMSOTf (0.055 equiv) in cyclohexane or hexane (usually 0.1 mL/1.0 mL cyclohexane) or, alternatively, BF₃•OEt₂ in dry diethyl ether, was then added dropwise. White precipitate formed upon addition of the Lewis acid. The reaction mixture was stirred at room temperature until completion as judged by ¹H NMR (typically overnight) and filtered through a plug of celite. The precipitate was then washed with pentane or hexane (precipitate is soluble in ether) and the filtrate diluted with ether. The diluted filtrate was subsequently washed with NaHCO₃ (sat. aq.) (twice), 1M HCl (twice), and lastly with brine (twice). The organic phase was dried over anhydrous MgSO₄, filtered, and concentrated to furnish the crude product. Purification by column chromatography on silica gel (0–2% EtOAc in hexane gradient) afforded the pure product.

4.2.1 Preparation of 5

Applying general procedure A to 6.75 g (46.86 mmol) of (E)-1-(trimethylsilyl)but-2-en-1-ol, 17.75 g (70.29 mmol) and the trichloroacetimidate of benzyl alcohol and BF₃•OEt₂ (0.65 mL, 5.15 mmol) in cyclohexane afforded 2.11 g (34%) of 5 as a colorless oil. ¹H NMR (300 MHz, CDCl₃) δ 7.35 (m, 5 H), 5.55–5.39 (m, 2 H), 4.69–4.65 (d, J = 12.4 Hz, 1 H), 4.37–4.28 (d, J = 12.4 Hz, 1 H), 3.52–3.50 (d, J = 7.1 Hz, 1 H), 1.74–1.72 (d, J = 4.7 Hz, 3 H), 0.01 (s, 9 H). ¹³C NMR (75 MHz, CDCl₃) δ 139.4, 129.7, 128.0 (2C), 127.5 (2C), 127.0, 125.1, 75.1, 71.3, 18.0, −3.7. HRMS (CI) m/z 252.1775 [(M+NH₄)⁺; calcd for C₁₄H₂₂OSi, 252.1784].

4.2.2 Preparation of compound 8

Compound 8 was prepared by alkylation of (E)-1-(trimethylsilyl)non-2-en-1-ol. Preparation of (E)-1-(trimethylsilyl)non-2-en-1-ol: To a solution of (E)-2-nonen-1-ol (800 mg, 5.62 mmol, 1 equiv) in THF (15 mL) at −78 °C was added n-BuLi (1.6 M in hexanes, 3.9 mL, 6.19 mmol, 1.1 equiv) slowly. After 30 minutes, TMSCl (0.79 mL, 6.19 mmol, 1.1 equiv) was added and the mixture stirred at room temperature for 1 hour. The reaction mixture was cooled down at −78 °C and t-BuLi (3.97 mL, 6.74 mmol, 1.2 equiv) was added dropwise (within 15 minutes). The reaction was monitored by ¹H NMR. After 8 hours the reaction was quenched by quickly adding NH₄Cl_(sat) (7 mL). The mixture was diluted with Et₂O (20 mL) and the aqueous phase was extracted with Et₂O (3 × 10 mL). Combined organic extracts were washed with brine, dried over MgSO₄ and concentrated to give 1.1 (91%, crude yield) of almost pure (E)-1-(trimethylsilyl)non-2-en-1-ol as a yellowish oil. ¹H NMR (500 MHz, CDCl₃) δ 5.56 (ddt, J = 1.0, 6.5, 15.0 Hz, 1 H), 5.46 (m, 1 H), 3.88 (dt, J = 1.5, 7.0 Hz, 1 H), 2.01 (q, J = 7.0 Hz, 2 H), 1.35–1.25 (m, 8 H), 0.86 (t, J = 7.0 Hz, 3 H), 0.02 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 131.2, 127.8, 68.4, 32.5, 31.7, 29.7, 28.8, 22.7, 14.1, 1.3, −4.2. IR (neat) 3416, 2928, 1248, 841 cm⁻¹. HRMS (EI) m/z 196.1647 [(M-H₂O)⁺; calcd for C₁₂H₂₄Si, 196.1647].

Alkylation of (E)-1-(trimethylsilyl)non-2-en-1-ol: Applying general procedure A to (E)-1-(trimethylsilyl)non-2-en-1-ol (300 mg, 1.4 mmol, 1 equiv) and trichloroacetimidate of benzyl alcohol (671 mg, 2.5 mmol, 1.8 equiv) in 10:1 hexane/CH₂Cl₂ (8 mL) with TMSOTf (38 μ, 0.21 mmol, 0.15 equiv) for 2.5 hours afforded 134 mg (31%) of 8 as a colorless oil. ¹H NMR (500 MHz, CDCl₃) δ 7.30 (m, 4 H), 7.24 (m, 1 H), 5.43 (m, 2 H), 4.66 (d, J = 12.5 Hz, 1 H), 4.30 (d, J = 12.0 Hz, 1 H), 3.50 (d, J = 8.0 Hz, 1 H), 2.04 (q, J = 7.0 Hz, 2 H), 1.38–1.24 (m, 8 H), 0.88 (t, J = 8.0 Hz, 3 H). ¹³C NMR (126 MHz, CDCl₃) δ 139.5, 130.9, 128.6, 128.1, 127.6, 127.1, 75.1, 71.3, 32.5, 31.7, 29.6, 28.8, 22.7, 14.1, −3.9. IR (film) 3030, 2957, 1246, 841 cm⁻¹. HRMS (EI) m/z 289.1985 [(M-CH₃)⁺; calcd for C₁₈H₂₉OSi, 289.1988].

4.2.3 Preparation of compound 11

Compound 11 was prepared by alkylation of (E)-4,4-dimethyl-1-(trimethylsilyl)pent-2-en-1-ol. Preparation of (E)-4,4-dimethyl-1-(trimethylsilyl)pent-2-en-1-ol: To a solution of (E)-4,4-dimethyl-2-penten-1-ol (700 mg, 6.13 mmol, 1 equiv) in THF (17.5 mL) at −78 °C was added n-BuLi (1.6 M in hexanes, 2.7 mL, 6.74 mmol, 1.1 equiv) slowly. After 30 minutes, TMSCl (0.86 mL, 6.74 mmol, 1.1 equiv) was added and the mixture stirred at room temperature for 1 hour. The reaction mixture was cooled down at −78 °C and t-BuLi (9 mL, 15.3 mmol, 2.5 equiv) was added dropwise (within 25 minutes). The reaction was transferred to a cold bath at −30 °C and monitored by ¹H NMR. After 6 hours the reaction was quenched by quickly adding NH₄Cl_(sat) (10 mL). The mixture was diluted with Et₂O (20 mL) and the aqueous phase was extracted with Et₂O (3 × 10 mL). Combined organic extracts were washed with brine, dried over MgSO₄ and concentrated. The crude product was purified by column chromatography (5% EtOAc in hexanes) to give 790 (69%) of 11 as a yellowish oil. ¹H NMR (500 MHz, CDCl₃) δ 5.49 (m, 2 H), 3.90 (dd, J = 1.0, 3.5 Hz, 1 H), 1.0 (s, 9 H), 0.01 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 138.6, 125.9, 68.4, 32.8, 29.8, −4.2. IR (film) 3406, 2959, 1248, 843 cm⁻¹. HRMS (EI) m/z 168.1334 [(M-H₂O)⁺; calcd for C₁₀H₂₀Si, 168.1334].

Alkylation of (E)-4,4-dimethyl-1-(trimethylsilyl)pent-2-en-1-ol: Applying general procedure A to (E)-4,4-dimethyl-1-(trimethylsilyl)pent-2-en-1-ol (350 mg, 1.88 mmol, 1 equiv) and trichloroacetimidate of benzyl alcohol (900 mg, 3.38 mmol, 1.8 equiv) in 10:1 hexane/CH₂Cl₂ (10 mL) with TMSOTf (51 μ, 0.282 mmol, 0.15 equiv) for 2.5 hours afforded 245 mg (49%) of 11 as a colorless oil. ¹H NMR (500 MHz, CDCl₃) δ 7.30 (m, 4 H), 7.24 (m, 1 H), 5.49 (dd, J = 1.0, 15.5 Hz, 1 H), 5.31 (dd, J = 6.5, 16.0 Hz, 1 H), 4.64 (d, J = 12.5 Hz, 1 H), 4.30 (d, J = 12.5 Hz, 1 H), 3.51 (dd, J = 1.0, 8.5 Hz, 1 H), 1.01 (s, 9 H), 0.00 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 141.7, 139.5, 128.1, 127.7, 127.1, 123.2, 75.2, 71.3, 29.8, −3.9. IR (film) 3026, 2959, 1246, 841 cm⁻¹. HRMS (EI) m/z 185.1353 [(M+Bn)⁺; calcd for C₁₀H₂₁OSi, 185.1362].

4.2.4 Preparation of compounds 14

Applying general procedure A to 4.01 g (30.82 mmol) of α-hydroxysilane 1-(trimethylsilyl)-prop-2-en-1-ol, 17.25 g (58.57 mmol) of the trichloroacetimidate of 2-methyl-1-phenylpropan-1-ol, and 0.38 g (1.70 mmol) of TMSOTf, and stirring the reaction overnight afforded 5.7 g (79%) of 14 as a 1:1 mixture of diastereomers. IR (neat) 2972, 2928, 2899, 1628, 1493, 1452, 1248 cm⁻¹. Mixture of syn-14/anti-14: ¹H NMR (300 MHz, CDCl₃) δ 7.36–7.10 (m, 10 H), 5.83–5.68 (m, 2 H), 5.06–4.87 (m, 4 H), 4.56–4.46 (m, 2 H), 3.82–3.80 (dt, J = 6.9, 1.4 Hz, 1 H), 3.43–3.41 (dt, J = 6.9, 1.4 Hz, 1 H), 1.39 (d, J = 6.6 Hz, 3 H), 1.35 (d, J = 6.6 Hz, 3 H), 0.06 (s, 9 H), 0.02 (s, 9 H). ¹³C NMR (75 MHz, CDCl₃) δ 145.3, 144.2, 137.6, 128.4, 128.0, 127.9, 127.1, 126.7, 126.6, 125.8, 112.1, 111.7, 76.0, 75.6, 74.1, 73.2, 24.8, 22.3, −3.7, −3.8. HRMS (EI) m/z 234.1434 [(M)⁺; calcd for C₁₄H₂₂OSi, 234.1440]. anti-14: ¹H NMR (500 MHz, CDCl₃) δ 7.35–7.21 (m, 5 H), 5.82–5.70 (m, 1 H), 5.05–4.95 (m, 2 H), 4.56–4.49 (q, J = 6.6, Hz, 1 H), 3.43–3.40 (dt, J = 7.1, 1.3 Hz, 1 H), 1.39 (d, J = 6.6 Hz, 3 H), 0.00 (s, 9 H). ¹³C NMR (125 MHz, CDCl3) δ 144.4, 137.8, 128.2, 127.2, 126.8, 112.2, 75.7, 73.3, 24.6, −3.9. HRMS (EI) m/z 234.1428 [(M)⁺; calcd for C₁₄H₂₂OSi, 234.1440].

4.2.5 Preparation of compound 20

Applying general procedure A to 0.88 g of 1-(trimethylsilyl)-prop-2-en-1-ol (6.73 mmol), 3.96 g of the trichloroacetimidate of 1-phenylbutan-1-ol (13.45 mmol, 2 equiv) and 0.07 mL of TMSOTf (0.4 mmol, 0.055 equiv) overnight afforded 1.32 g of 20 (75%) as a 1:1 mixture of diastereomers after column chromatography (0–2% EtOAc gradient). IR (neat) 1628 cm⁻¹.

Anti-20: ¹H NMR (300 MHz, CDCl₃) δ 7.31–7.19 (m, 5 H), 5.77–5.70 (m, 1 H), 5.03–4.92 (dd, J = 10.6, 17.2 Hz, 2 H), 4.07–4.06 (d, J = 7.5 Hz, 1 H), 3.39–3.37 (d, J = 8.0 Hz, 1 H), 1.91–1.85 (m, 1 H), 1.01–1.00 (d, J = 6.6 Hz, 3 H), 0.72–0.70 (d, J = 7.1 Hz, 3 H), −0.01 (s, 9 H). ¹³C NMR (75 MHz, CDCl₃) δ 141.8, 138.0, 128.1 (2C), 127.8, 127.1 (2C), 113.1, 84.7, 72.6, 35.0, 19.2, 19.0, −4.0. HRMS (APCI) m/z 263.1821 [(M+H)⁺; calcd for C₁₆H₂₇OSi, 263.1831].

Syn-20: ¹H NMR (300 MHz, CDCl₃) δ 7.35–7.17 (m, 5 H), 5.64–5.53 (m, 1 H), 4.89–4.72 (dd, J = 10.7, 16.7 Hz, 2 H), 4.01–3.99 (d, J = 6.9 Hz, 1 H), 3.72–3.68 (d, J = 7.4 Hz, 1 H), 1.99–1.89 (m, 1 H), 0.93–0.91 (d, J = 6.9 Hz, 3 H), 0.77–0.74 (d, J = 6.9 Hz, 3 H), 0.05 (s, 9 H), 0.08 (s, 9 H). ¹³C NMR (75 MHz, CDCl₃) δ 142.6, 138.2, 127.4 (2C), 127.3 (2C), 126.5, 111.4, 87.5, 76.6, 34.5, 18.7 (2C), −3.5. HRMS (APCI) m/z 263.1821 [(M)⁺; calcd for C₁₆H₂₇OSi, 263.1831].

4.2.6 Preparation of 23

Applying general procedure A to 1 g (7.68 mmol) of 1-(trimethylsilyl)-prop-2-en-1-ol, 3 g (10.75 mmol) of the trichloroacetimidate of 4-methylbenzyl alcohol and TMSOTf (35 μL, 0.19 mmol) in hexane afforded 1.7 g (95%) of diastereomeric 23 as a colorless oil. Mixture of syn-23/anti-23 (1:1) δ ¹H NMR (500 MHz, CDCl₃) δ 7.20 (d, J = 8.5 Hz, 2 H), 7.15–7.09 (m, 6 H), 5.73 (m, 2 H), 5.01–4.94 (m, 3 H), 4.88 (dt, J = 11.0 Hz, 1 H), 4.48 (q, J = 6.5 Hz, 1 H), 4.45 (d, J = 6.5 Hz, 1 H), 3.78 (dt, J = 1.5, 6.5 Hz, 1 H), 3.40 (dt, J = 1.5, 7.0 Hz, 1 H), 2.33 (s, 3 H), 2.31 (s, 3 H), 1.35 (d, J = 6.5 Hz, 3 H), 1.32 (d, J = 6.5 Hz, 3 H), 0.03 (s, 9 H), −0.05 (s, 9 H). ¹³C NMR (151 MHz, CDCl₃) δ 142.5, 141.3, 137.84, 137.81, 136.7, 136.3, 128.8 (2 C), 128.7 (2 C), 126.7 (2 C), 125.9 (2 C), 112.1, 111.6, 75.9, 75.4, 74.1, 73.1, 24.7, 22.2, 21.13, 21.09, −3.0, −4.0. IR (film) 3050, 2972, 1248, 841 cm⁻¹. HRMS (EI) m/z 248.1597 [(M⁺); calcd for C₁₅H₂₄OSi, 248.1596].

4.2.7 Preparation of compound E-30

Applying general procedure A to 3.6 g of (E)-1-(trimethylsilyl)but-2-en-1-ol (24.98 mmol),^11d 13.32 g of the trichloroacetimidate of phenethyl alcohol (49.5 mmol, 2 equiv) and 0.47 mL of BF₃•OEt₂ (3.74 mmol, 0.15 equiv) afforded 3.07 g of E-30 (39%).

E-30 (mixture of diastereomers anti/syn 0.58:0.42): ¹H NMR (500 MHz, CDCl₃) δ 7.33–7.19 (m, 5 H), 5.42–5.31 (m, 2 H), 4.52 (q, J = 6.5 Hz, 0.58 H), 4.49 (q, J = 6.0 Hz, 0.42 H), 3.69 (d, J = 7.0 Hz, 0.42 H), 3.29 (d, J = 7.0 Hz, 0.58 H), 1.71 (d, J = 5.5 Hz, 1.74 H), 1.61 (d, J = 6.0 Hz, 1.26 H), 1.35 (d, J = 6.5 Hz, 1.74 H), 1.32 (d, J = 6.0 Hz, 1.26 H), 0.03 (s, 3.78 H), −0.05 (s, 5.22 H). ¹³C NMR (125 MHz, CDCl₃) anti E-30: δ 144.6, 130.2, 128.1 (2C), 127.0, 126.8 (2C), 124.6, 75.0, 72.4, 24.7, 17.9, −3.9. syn E-30: δ 145.8, 130.4, 127.9 (2C), 126.6, 125.9 (2C), 124.0, 75.2, 73.2, 22.0, 17.8, −3.8. HRMS (CI) m/z 248.1591 [(M)⁺; calcd for C₁₅H₂₄OSi, 248.1596].

4.2.8 Preparation of compound Z-30

Following the general procedure A to 3.49 g of 1-(trimethylsilyl)but-2-yn-1-ol (24.53 mmol) and 13.1 g of the trichloroacetimidate of sec-phenethyl alcohol (49.06 mmol, 2 equiv) in cyclohexane (140 mL) at 0 °C was added 0.46 mL of BF₃•OEt₂ (3.68 mmol, 0.15 equiv). After 1 hour the reaction was stopped, worked up according to the general procedure A followed by column chromatography (8% DCM in hexanes) to afford 5 g (83%) of diastereomeric alkyne 37 as a colorless oil (diastereomers partially separated). Alkyne reduction: To a solution of 1.435 g of 37 (5.82 mmol, dr = 1:1) in hexanes (210 mL) was added Et₃N (2.6 mL, 2.5 mL/mmol 37) and Lindlar’s catalyst (38.8 mg, 37.5 mg/mmol 37). The flask was flushed with hydrogen and a hydrogen balloon attached. The mixture was vigorously stirred and the reaction monitored by NMR (about 4 h). The reaction mixture was partially concentrated, filtered through a plug of celite and fully concentrated. Column chromatography (10% DCM in hexanes) afforded 900 mg (62%) of Z-30. Note: Pure alkyne 37 decomposes relatively quickly after isolation and its decomposition products appear to poison the catalyst and hamper reduction thus requiring addition of more catalyst. Samples of 37 stored at −20 °C slowly decomposed turning yellow, such samples in hexanes were filtered through a short silica gel plug and rinsed with more hexanes. After concentration clean 37 was immediately submitted to the reduction reaction. IR (neat) 2963, 2203, 1248, 1082, 843 cm⁻¹. HRMS (EI) m/z 246.1444 [(M)⁺; calcd for C₁₅H₂₂OSi, 246.1440].

anti-37: ¹H NMR (500 MHz, CDCl₃) δ 7.37–7.23 (m, 5 H), 4.79 (q, J = 6.5 Hz, 1 H), 3.43 (q, J = 2.5 Hz, 1 H), 1.88 (d, J = 2.5 Hz, 3 H), 1.40 (d, J = 7.0 Hz, 3 H), 0.06 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 143.8, 128.3 (2C), 127.3, 126.8 (2C), 82.7, 77.6, 76.3, 60.5, 24.4, 3.9, −4.0.

syn-37: ¹H NMR (500 MHz, CDCl₃) δ 7.35 (d, J = 7.5 Hz, 2 H), 7.30 (m, 2 H), 7.21 (tt, J = 1.5, 7.5 Hz, 1 H), 4.71 (q, J = 6.5 Hz, 1 H), 3.88 (q, J = 2.5 Hz, 1 H), 1.78 (d, J = 2.5 Hz, 3 H), 1.36 (d, J = 6.5 Hz, 3 H), 0.12 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 145.0, 128.0 (2C), 126.8, 126.1 (2C), 83.1, 77.6, 76.1, 61.0, 21.4, 3.8, −3.8.

anti Z-30 ¹H NMR (500 MHz, CDCl₃) δ 7.29 (t, J = 7.0 Hz, 2 H), 7.23 (m, 3 H), 5.50 (ddq, J = 1.0, 7.0, 11.0 Hz, 1 H), 5.39 (ddq, J = 1.5, 10.5, 11.0 Hz, 1 H), 4.44 (q, J = 6.5 Hz, 1 H), 3.67 (d, J = 10.5 Hz, 1 H), 1.35 (ddd, J = 0.5, 2.0, 7.0 Hz, 3 H), 1.34 (d, J = 6.5 Hz, 3 H), −0.03 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 144.6, 130.5, 128.1 (2C), 127.1, 126.9 (2C), 124.4, 75.6, 67.4, 24.6, 13.4, −3.9. HRMS (ESI) m/z 249.1663 [(M+H)⁺; calcd for C₁₅H₂₅OSi, 249.1675].

syn Z-30 ¹H NMR (500 MHz, CDCl₃) δ 7.32–7.26 (m, 4 H), 7.20 (m, 1 H), 5.43–5.32 (m, 2 H), 4.45 (q, J = 6.5 Hz, 1 H), 4.13 (d, J = 9.5 Hz, 1 H), 1.51 (m, 3 H), 1.34 (d, J = 6.5 Hz, 3 H), 0.04 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 145.6, 130.9, 127.9 (2C), 126.7, 126.0 (2C), 123.2, 75.8, 68.4, 22.0, 13.5, −3.8. HRMS (ESI) m/z 249.1675 [(M+H)⁺; calcd for C₁₅H₂₅OSi, 249.1675].

4.2.9 Preparation of compound 17

17 was prepared following a procedure reported in literature.²⁴ Allyltrimethylsilane 1-(trimethylsilyl)prop-2-en-1-ol (1.26 g, 11.0 mmol, 1.75 mL), benzaldehyde (1.67 g, 11.0 mmol, 1.12 mL), and TMSOTf (0.36 mL, 2.0 mmol, 0.44 g) were successively added to a stirred cold (−78 °C) solution of α-(trimethylsilyl)allyl trimethysilyl ether (2.0 g, 10.0 mmol) in CH₂Cl₂ (100 mL). The reaction was stirred for 70 min and then quenched with NaHCO₃ (aq. sat.). The aqueous phase was extracted with CH₂Cl₂ (100 mL × 4), and the combined organic layers were washed with NaHCO₃ (100 mL × 2), brine (100 mL × 2), and then dried (MgSO₄). Filtration and concentration afforded the crude product as a 1:2.56 mixture of diastereomers. After silica gel chromatography 1.96 g (7.58 mmol) of the pure products were obtained in a combined yield of 77%. The pair of diastereomers is separable by column chromatography on silica gel (5% and 10% CH₂Cl₂ in hexanes).

anti-17: ¹H NMR (300 MHz, CDCl₃) δ 7.35–7.22 (m, 5 H), 5.88–5.69 (m, 2 H), 5.05–4.95 (m, 4 H), 4.46–4.42 (dd, J = 7.7, 5.8 Hz, 1 H), 3.44–3.42 (d, J = 7.4 Hz, 1 H), 2.59–2.49 (quint, J = 7.7 Hz, 1 H), 2.39–2.30 (quint, J = 6.86, 1 H), −0.01 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 142.5, 137.7, 135.4, 128.1 (2C), 127.4 (3C), 116.3, 112.9, 79.3, 73.0, 43.03, −4.0. HRMS (CI) m/z 261.1664 [(M+H)⁺; calcd for C₁₆H₂₄OSi, 261.1675.

syn-17: ¹H NMR (300 MHz, CDCl₃) δ 7.30–7.20 (m, 5 H), 5.79–5.60 (m, 2 H), 5.01–4.80 (m, 4 H), 4.39–4.35 (t, J = 6.2 Hz, 1 H), 3.82–3.78 (dt, J = 7.1, 1.3 Hz, 1 H), 2.54–2.40 (m, 2 H), 0.05 (s, 9 H). ¹³C NMR (500 MHz, CDCl₃) δ 143.6, 137.9, 134.9, 127.8 (2C), 126.9, 126.6 (2C), 116.8, 111.9, 81.1, 75.8, 41.5, −3.7. HRMS (CI) m/z 261.1681 [(M+H)⁺; calcd for C₁₆H₂₄OSi, 261.1675].

4.3 Wittig rearrangements of α-alkoxysilanes – General procedure B

A solution of α-alkoxysilane (1.0 equiv) in freshly distilled THF (0.06–0.07 M) was cooled to the desired temperature under nitrogen. The required amount of s-BuLi (1.5–4.0 equiv, 1.3 M in cyclohexane) or n-BuLi (1.6 M in hexanes) was added dropwise via syringe. The reaction mixture was stirred at the reaction temperature for the desired length of time, then quenched with saturated aqueous NH₄Cl and diluted with ether. Phases separated and the organic phase was washed with water and brine. The organic phase was dried over anhydrous MgSO₄ and concentrated. Silical gel chromatography (0 to 2% EtOAc in hexane gradient) afforded the rearranged products usually as light oils.

4.3.1 Wittig rearrangements of compound 5

Applying the General procedure B to 141 mg (0.60 mmol) of 5 and 0.69 mL (0.90 mmol) of s-BuLi (1.3 M in cyclohexane) at −78 °C for for 30 min, after purification by column chromatography on silica gel, afforded 106 mg (75%) of a 4:1 mixture of both [1,4]- and [1,2]-rearrangement products 6 (a light yellow oil) and 7 as a colorless oil.

Compound 6: ¹H NMR (300 MHz, CDCl₃) δ 7.32–7.17 (m, 5 H), 2.61–2.50 (m, 2 H), 2.47–2.30 (m, 3 H), 0.84–0.81 (d, J = 6.6 Hz, 3 H), −0.13 (s, 9 H). ¹³C NMR (125 MHz, CDCl₃) δ 248.6, 140.6, 129.2 (2C), 128.2 (2C), 125.9, 54.9, 43.3, 29.6, 19.9, −3.3. IR (neat) 1709 cm⁻¹. HRMS (EI) m/z 233.1355 [(M–H)⁺; calcd for C₁₄H₂₁OSi, 233.1362].

Compound 7: ¹H NMR (500 MHz, CDCl₃) δ 7.26–7.10 (m, 5 H), 5.60–5.56 (dq, J = 15.4, 1.6 Hz, 1 H), 5.19–5.12 (apparent dq, J = 15.4, 6.6 Hz, 1H), 2.86 (d, J = 7.7 Hz, 1 H), 2.81 (d, J = 7.7 Hz, 1 H), 1.64–1.62 (dd, J = 6.6, 1.6 Hz, 3 H), 0.05 (s, 9 H). ¹³C NMR (125 MHz, CDCl₃) δ 136.2, 135.4, 130.6 (2C), 127.9 (2C), 126.3, 121.62, 70.4, 43.1, 17.8, −4.2. IR (neat) 3432 cm⁻¹. HRMS (EI) m/z 234.1435 [(M)⁺; calcd for C₁₄H₂₂OSi, 234.1440].

4.3.2 Wittig rearrangements of compound 8

Applying the General procedure B to compound 8 (62 mg, 0.204 mmol, 1 equiv) and sec-BuLi (1.4 M in cyclohexane, 0.22 mL, 1.5 equiv) at −78 °C for 40 min, after purification by silica gel column chromatography (3% EtOAc in hexanes), afforded 48.4 mg (78%) of a 2:1 mixture of both [1,4]- and [1,2]-rearrangement products 9 and 10 as a colorless oils. An analytically pure sample of 9 could be obtained by subsequent silica gel column chromatography (3% EtOAc in hexanes).

Compound 9: ¹H NMR (500 MHz, CDCl₃) δ 7.24 (m, 2 H), 7.15 (m, 1 H), 7.11 (m, 2 H), 2.58 (dd, A of ABX system, J = 6.5, 13.5 Hz, 1 H), 2.49 (d, J = 6.5 Hz, 2 H), 2.42 (dd, B of ABX system, J = 7.0, 13.5 Hz, 1 H), 2.29 (m, 1 H), 1.23 (m, 10 H), 0.85 (t, J = 7.5 Hz, 3 H), 0.12 (s, 9 H). ¹³C NMR (125 MHz, CDCl₃) δ 248.7, 140.7, 129.3 (2 C), 128.2 (2 C), 125.8, 52.5, 40.5, 34.1, 33.9, 31.8, 29.5, 26.8, 22.6, 14.1, −3.2. IR (film) 3026, 2926, 1643, 1456, 1250, 844 cm⁻¹. HRMS (EE) m/z 304.2242 [(M)⁺; calcd for C₁₉H₃₂OSi, 304.2222].

Mixture of 9 and 10 (1.8:1) ¹H NMR (500 MHz, CDCl₃) δ 7.23–7.11 (m, 14 H), 5.55 (dt, J = 1.0, 15.5 Hz, 1 H), 5.13 (dt, J = 7.0, 15.5 Hz, 1 H), 2.87 (d, J = 13.0 Hz, 1 H), 2.81 (d, J = 13.5 Hz, 1 H), 2.58 (dd, J = 7.0, 13.5 Hz, 1.8 H), 2.49 (d, J = 6.5 Hz, 3.6 H), 2.42 (dd, J = 7.0, 13.5 Hz, 1.8 H), 2.29 (m, 1.8 H), 1.97 (m, 2 H), 1.21 (m, 26 H), 0.85 (m, 8.4 H), 0.12 (s, 16.2 H), 0.06 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) signals corresponding to 9 δ 248.7, 140.7, 129.3 (2 C), 128.2 (2 C), 125.8, 52.5, 40.5, 34.1, 33.9, 31.8, 29.5, 26.8, 22.6, 14.1, −3.2. signals corresponding to 10 δ 136.2, 134.3, 130.6 (2 C), 127.8 (2 C), 127.3, 126.3, 70.3, 43.1, 32.5, 31.7, 29.7, 28.7, 22.6, 14.0, −4.2. IR (film) 3534, 3028, 2957, 1745, 1643, 1248, 843 cm⁻¹. HRMS for 10 (EI) m/z 286.2150 [(M-H₂O)⁺; calcd for C₁₉H₃₀Si, 286.2117].

4.3.3 Wittig rearrangements of compound 11

Applying the General procedure B to compound 11 (91 mg, 0.329 mmol, 1 equiv) and sec-BuLi (1.4 M in cyclohexane, 0.35 mL, 1.5 equiv) at −78 °C for 40 min, after purification by silica gel column chromatography (3% EtOAc in hexanes), afforded 56.2 mg (62%) of [1,2]-product 12, 3.7 mg (4%) of [2,3]-product 13 and 14 mg of unreacted 11 (15%) as colorless oils.

Compound 12: ¹H NMR (500 MHz, CDCl₃) δ 7.23 (m, 2 H), 7.17 (m, 1 H), 7.10 (m, 2 H), 5.46 (d, J = 16.0 Hz, 1 H), 5.14 (d, J = 16.0 Hz, 1 H), 2.88 (d, J = 13.5 Hz, 1 H), 2.83 (d, J = 13.0 Hz, 1 H), 0.92 (s, 9 H), 0.05 (s, 9 H). ¹³C NMR (125 MHz, CDCl₃) δ 137.9, 136.1, 130.6 (2 C), 129.2, 127.7 (2 C), 126.3, 70.0, 43.1, 32.8, 29.9, −4.2. IR (film) 3534, 3028, 2959, 1248, 843 cm⁻¹. HRMS (EI) m/z 258.1820 [(M-H₂O)⁺; calcd for C₁₇H₂₆Si, 258.1804].

Compound 13: ¹H NMR (500 MHz, CDCl₃) δ 7.25 (m, 2 H), 7.19 (m, 3 H), 6.11 (dd, J = 10.0, 18.5 Hz, 1 H), 5.25 (dd, J = 0.5, 18.0 Hz, 1 H), 5.03 (t, J = 3.0 Hz, 1 H), 1.93 (dd, J = 2.5, 9.0 Hz, 1 H), 1.70 (d, J = 3.5 Hz, 1 H), 0.97 (s, 9 H), 0.00 (s, 9 H). ¹³C NMR (125 MHz, CDCl₃) δ 145.3, 142.1, 136.3, 127.8 (2 C), 126.9 (2 C), 126.2, 73.6, 65.0, 33.0, 28.9, −1.2. IR (film) 3422, 2955, 1246, 837 cm⁻¹. HRMS (EI) m/z 258.1791 [(M-H₂O)⁺; calcd for C₁₇H₂₆Si, 258.1804].

4.3.4 Wittig rearrangements of compound 14

Applying General procedure B to 360 mg (1.53 mmol) of 14 and 1.8 mL (2.30 mmol) of s-BuLi (1.3 M in cyclohexane) at −78 °C overnight, afforded 162 mg (46%) of a 1.68:1 mixture of 15 and 16 as a colorless oil.

Compound 15: ¹H NMR (300 MHz, CDCl₃) δ 7.33–7.12 (m, 5 H), 2.67–2.57 (m, 1 H), 2.54–2.41 (m, 1 H), 1.89–1.67 (m, 2 H), 1.24–1.21 (d, J = 7.1 Hz, 3 H), 0.11 (s, 9 H). ¹³C NMR (125 MHz, CDCl₃) δ 248.2, 146.6, 128.4, 127.0, 126.0, 46.4, 39.3, 30.2, 22.4, −3.2. IR (neat) 1643 cm⁻¹. HRMS (EI) m/z 233.1358 [(M–H)⁺; calcd for C₁₄H₂₁OSi, 233.1362].

Compound 16: ¹H NMR (300 MHz, CDCl₃) δ 7.33–7.19 (m, 5 H), 3.82–3.75 (q, J = 6.9 Hz, 1 H), 2.34–2.28 (m, 2 H), 1.38 (d, J = 6.9 Hz, 3 H), 0.77–0.55 (m, 2 H), −0.11 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 211.8, 140.8, 128.8 (2C), 127.8 (2C), 127.0, 52.3, 35.6, 17.7, 10.3, −1.9. IR (neat) 1717, 1601 cm⁻¹. HRMS (CI) m/z 234.1466 [(M)⁺; calcd for C₁₄H₂₂OSi, 234.1440].

4.3.5 Wittig rearrangements of compound 17

Applying General procedure B to 165 mg (0.638 mmol) of 17 and 1.6 mL of n-BuLi (2.55 mmol, 4 equiv, 1.6 M in hexanes) at −78 °C, allowing the reaction to warm to −30 °C and stirring at this temperature for about 48 h, after purification by column chromatography on silica gel afforded 45 mg (32%) of a 4.53:1 mixture of 18 and 19 as light yellow oils. Note: the reported yield is based on 2.64:1 diastereomeric ratio of anti/syn 17.

Compound 18: ¹H NMR (500 MHz, CDCl₃) δ 7.34–7.16 (m, 5 H), 5.78–5.60 (m, 1 H), 5.00–4.89 (m, 2 H), 2.58–2.52 (m, 1 H), 2.50–2.45 (m, 1 H), 2.39–2.31 (m, 3 H), 2.00–1.93 (m, 1 H), 1.72–1.64 (m, 1 H), 0.09 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 248.1, 144.4, 136.8, 128.4 (2C), 127.7 (2C), 126.2, 116.0, 46.1, 45.1, 41.4, 27.9, −3.2. IR (neat) 1717, 1643 cm⁻¹. HRMS (EI) m/z 260.1595 [(M)⁺; calcd for C₁₆H₂₄OSi, 260.1596].

Compound 19: ¹H NMR (500 MHz, CDCl₃) δ 7.31–7.08 (m, 5 H), 5.68–5.60 (m, 1 H), 5.00–4.91 (m, 2 H), 3.72 (t, J = 7.4 Hz, 1 H), 2.81–2.75 (m, 1 H), 2.45–2.39 (m, 1 H), 2.31 (m, 2 H), 0.74–0.68 (m, 1 H), 0.62–0.56 (m, 1 H), −0.11 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 210.6, 138.6, 135.9, 128.8 (2C), 128.2 (2C), 127.2, 116.6, 58.1, 36.7, 36.5, 10.1, −1.9. IR (neat) 1716, 1643 cm⁻¹. HRMS (EI) m/z 260.1593 [(M)⁺; calcd for C₁₆H₂₄OSi, 260.1596].

4.3.6 Wittig rearrangements of compound 20

Applying General procedure B to 69.5 mg of syn-20 (0.265 mmol) and 0.33 mL of n-BuLi (0.5296 mmol, 2 equiv) in THF (3.3 mL) at −78 °C and then at 0 °C for 17 H afforded a mixture (15.8 mg) of 21 and 22 in a combined 23% yield as colorless oil along with 18.6 mg of unreacted syn-20. Column chromatography was performed with 3% EtOAc in hexanes.

Compound 21: ¹H NMR (500 MHz, CDCl₃) δ 7.26–7.03 (m, 5 H), 2.39–2.35 (m, 1 H), 2.30–2.24 (m, 1 H), 2.20–2.15 (m, 1 H), 2.08–2.02 (m, 1 H), 1.80–1.73 (m, 1 H), 1.71–1.63 (m, 1 H), 0.94 (d, J = 6.8 Hz, 3 H), 0.68 (d, J = 6.8 Hz, 3 H), 0.06 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 248.6, 143.8, 128.4 (2C), 128.1 (2C), 126.0, 52.4, 33.7, 25.2, 20.9, 15.3, −3.3. IR (neat) 1719, 1643 cm⁻¹. HRMS (APCI) m/z 263.1840 [(M+H)⁺; calcd for C₁₆H₂₇OSi, 263.1831].

Compound 22: ¹H NMR (500 MHz, CDCl₃) δ 7.29–7.20 (m, 5 H), 3.30 (d, J = 10.2 Hz, 1 H), 2.42–2.25 (m, 3 H), 0.94 (d, J = 6.3 Hz, 3 H), 0.74–0.67 (m, 1 H), 0.63 (d, J = 6.8 Hz, 3 H), 0.61–0.54 (m, 1 H), −0.11 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 211.7, 138.4, 128.7 (2C), 128.6 (2C), 127.0, 66.4, 37.8, 30.7, 21.7, 20.4, 9.9, −1.9. IR (neat) 1715 cm⁻¹. HRMS (CI) m/z 262.1755 [(M)⁺; calcd for C₁₆H₂₆OSi, 262.1753].

4.3.7 Wittig rearrangements of compound 23

Applying the General procedure B to compound 23 (anti/syn = 1:1) (66 mg, 0.282 mmol, 1 equiv) and n-BuLi (1.6 M in cyclohexane, 0.53 mL, 3 equiv) at −78 °C and then at −30 °C for 3 hours, after purification by silica gel column chromatography (3% EtOAc in hexanes), 16.3 mg (26%) of 24, 10.2 mg (14%) of 25 as colorless oils, and 7.5 mg of a mixture of anti-23 and 26 (1:9 ratio). An analytical sample of 26 was obtained by column chromatography eluting with hexanes.

Compound 24: ¹H NMR (600 MHz, CDCl₃) δ 7.08 (d, J = 7.8 Hz, 2 H), 7.02 (d, J = 8.4 Hz, 2 H), 2.59 (m, 1 H), 2.51 (ddd, A of ABX system, J = 6.0, 9.0, 16.8 Hz, 1 H), 2.41 (ddd, B of ABX system, J = 6.0, 9.0, 17.4 Hz, 1 H), 2.30 (s, 3 H), 1.80 (m, 1 H), 1.72 (m, 1 H), 1.20 (d, J = 6.6 Hz, 3 H), 0.12 (s, 9 H). ¹³C NMR (151 MHz, CDCl₃) δ 248.2, 143.6, 135.5, 129.1 (2 C), 126.9 (2 C), 46.5, 38.9, 30.3, 22.5, 21.0, −3.2. IR (film) 2959, 1643, 1250, 844 cm⁻¹. HRMS (EI) m/z 248.1595 [(M)⁺; calcd for C₁₅H₂₄OSi, 248.1596].

Compound 25: ¹H NMR (500 MHz, CDCl₃) δ 7.11 (d, J = 8.0 Hz, 2 H), 7.08 (d, J = 8.5 Hz, 2 H), 3.74 (q, J = 7.0 Hz, 1 H), 2.30 (m, 6 H), 1.35 (d, J = 7.0 Hz, 3 H), 0.71 (ddd, A of ABX system, J = 6.5, 10.0, 15.0 Hz, 1 H), 0.60 (ddd, B of ABX system, J = 6.5, 9.0, 14.0 Hz, 1 H), 0.11 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 212.0, 137.9, 136.7, 129.5 (2 C), 127.7 (2 C), 51.9, 35.5, 21.0, 17.7, 10.4, −1.9. IR (film) 2955, 1716, 1456, 1250, 837 cm⁻¹. HRMS (EI) m/z 233.1363 [(M-CH₃)⁺; calcd for C₁₄H₂₁OSi, 233.1362].

Compound 26: (1:0.7 mixture of diastereomers) ¹H NMR (500 MHz, CDCl₃) δ 7.09 (m, 8 H), 6.97 (d, J = 8.0 Hz, 2.8 H), 6.91 (d, J = 8.5 Hz, 2.8 H), 2.90 (m, 1.4 H), 2.72 (m, 2 H), 2.32 (s, 6 H), 2.25 (s, 4.2 H), 1.20 (m, 4.2 H), 0.98 (m, 6 H). ¹³C NMR (126 MHz, CDCl₃) δ 143.6, 142.8, 135.4, 135.0, 128.9, 128.5, 127.7, 127.5, 46.8, 45.8, 21.2, 21.0, 20.9, 17.8. IR (film)3021, 2961, 1514, 1452, 817 cm⁻¹. HRMS (EI) m/z 238.1724 [(M)⁺; calcd for C₁₈H₂₂, 238.1722].

4.3.8 Wittig rearrangements of anti/syn E-30

Applying representative procedure B to 235 mg (0.946 mmol) of E-30 (anti/syn = 1.5:1) and 2.36 mL of n-BuLi (3.78 mmol, 4 equiv, 1.6 M in hexanes) in THF (12 mL) at −30 °C for 44 h. After purification by column chromatography on silica gel (30% CH₂Cl₂ in hexanes) 101 mg of anti E-30 (43%), 10.1 mg of 32 (16%) and 72 mg of a mixture of 31 (23%, anti/syn = 1.9:1) and 33 (6%, anti/syn = 1.44:1) were obtained. Analytical samples of 31 and 33 were obtained by subsequent column chromatography of the mixture (30% CH₂Cl₂ in hexanes).

Compound 31 (mixture of diastereomers anti-31/syn-31, 0.65:0.35 ratio): ¹H NMR (600 MHz, CDCl₃) δ 7.28–7.24 (m, 2 H), 7.18–7.12 (m, 3 H), 2.64 (m, 0.65 H), 2.61 (m, 0.65 H), 2.52 (m, 0.35 H), 2.46 (m, 0.35 H), 2.38–2.26 (m, 2 H), 1.21 (d, J = 6.6 Hz, 1.05 H), 1.20 (d, J = 7.2 Hz, 1.95 H), 0.84 (d, J = 6.6 Hz, 1.05 H), 0.70 (d, J = 6.6 Hz, 1.95 H), 0.14 (s, 5.85 H), 0.08 (s, 3.15 H). ¹³C NMR (151 MHz, CDCl₃) major diastereomer (anti-31): δ 248.6, 145.4, 128.1 (2C), 127.9 (2C), 126.0, 53.0, 44.6, 33.5, 18.2, 18.04, −3.16. Minor diastereomer (syn-31): δ 248.7, 146.3, 128.2 (2C), 127.6 (2C), 126.0, 54.0, 45.0, 33.8, 18.05, 17.4, −3.25. IR (neat) 1643 cm⁻¹. HRMS (ESI) m/z 249.1665 [(M+H⁺); calcd for C₁₅H₂₅OSi, 249.1675].

Compound 32 (mixture of diastereomers, 0.88:0.12 ratio, relative stereochemistry not assigned): ¹H NMR (600 MHz, CDCl₃) δ 7.26–7.14 (m, 5 H), 5.72 (dd, J = 1.2, 15.6 Hz, 0.12 H), 5.59 (dd, J = 1.2, 15.6 Hz, 0.88 H), 5.28 (dq, J = 6.6, 15.6 Hz, 0.12 H), 5.12 (dq, J = 6.6, 15.6 Hz, 0.88 H), 3.04 (q, J = 7.2 Hz, 0.88 H), 3.00 (q, J = 7.2 Hz, 0.12 H), 1.71 (dd, J = 1.8, 6.6 Hz, 0.36 H), 1.61 (dd, J = 1.2, 6.6 Hz, 2.64 H), 1.32 (d, J = 7.2 Hz, 2.64 H), 1.27 (d, J = 7.2 Hz, 0.36 H), 1.03 (s, 0.12 H), 1.02 (s, 0.88 H), −0.03 (s, 7.92 H), −0.09 (s, 1.08 H). ¹³C NMR (151 MHz, CDCl₃) major diastereomer: δ 142.8, 135.1, 128.9 (2C), 127.8 (2C), 126.3, 120.8, 73.6, 46.8, 17.8, 16.4, −2.5. Minor diastereomer: δ 143.0 133.5, 129.2 (2C), 127.9 (2C), 126.6, 122.0, 73.4, 46.9, 18.0, 16.5, −3.0. HRMS (CI) m/z 249.1666 [(M+H)⁺; calcd for C₁₅H₂₄OSi, 249.1675].

Compound 33 (mixture of diastereomers, anti-33/syn-33, 0.55:0.45) ¹H NMR (600 MHz, CDCl₃) δ 7.30 (m, 2 H), 7.24–7.20 (m, 3 H), 3.78 (q, J = 7.2 Hz, 0.55 H), 3.71 (q, J = 7.2 Hz, 0.45 H), 2.36 (m, 0.45 H), 2.33 (m, 0.55 H), 2.14 (dd, J = 10.2, 16.8 Hz, 0.45 H), 2.08 (dd, J = 10.8, 16.2 Hz, 0.55 H), 1.38 (d, J = 6.6 Hz, 1.35 H), 1.37 (d, J = 7.2 Hz, 1.65 H), 1.18 (m, 0.55 H), 1.13 (m, 0.45 H), 0.82 (d, J = 7.8 Hz, 1.35 H), 0.69 (d, J = 7.2 Hz, 1.65 H), −0.118 (s, 4.95 H), −0.160 (s, 4.05 H). ¹³C NMR (151 MHz, CDCl₃) major diastereomer (anti-333): δ 210.8, 140.6, 128.8 (2C), 127.91 (2C), 127.05, 52.2, 43.3, 17.6, 15.3, 14.1, −3.55. Minor diastereomer (syn-33): δ 211.4, 128.7 (2C), 127.89 (2C), 127.03, 53.6, 43.5, 17.4, 15.5, 14.7, −3.54. IR (neat) 1718 cm⁻¹. HRMS (ESI) m/z 249.1667 [(M+H⁺); calcd for C₁₅H₂₅OSi, 249.1675].

4.3.9 Wittig rearrangement of compound syn Z-30

Following the general procedure B to 253.8 mg of syn Z-30 (1.02 mmol) in 10.5 mL of THF at −78 °C was added 2.56 mL of n-BuLi (4.086 mmol, 4 equiv, 1.6 M in hexanes), the cold bath was removed and the reaction stirred at room temperature for 48 h. After work up and column chromatography (gradient of 2–10% EtOAc in hexanes, then 50% EtOAc in hexanes) afforded 142.1 mg of syn Z-30 (56%, dr = 18:1), 11.1 mg of a 1:1 mixture of 31 (syn/anti = 1.1:1) and 33 (syn/anti = 1:1.6), 15.6 mg of 34 (6%, single diastereomer), 13 mg of 35 (7%, dr = 3:1) and 5.4 mg of 36 (3%, syn/anti = 1.3:1).

Compound 34 (tentatively assigned as syn-34):^{17 1}H NMR (500 MHz, CDCl₃) δ 7.36 (m, 2 H), 7.31 (m, 2 H), 7.21 (tt, J = 1.5, 7.5 Hz, 1 H), 5.97 (dd, J = 8.0, 19.0 Hz, 1 H), 5.70 (dd, J = 1.0, 19.0 Hz, 1 H), 2.52 (m, 1 H), 1.85 (s, 1 H), 1.50 (s, 3 H), 0.84 (d, J = 6.5 Hz, 1 H), 0.04 (s, 9 H). ¹³C NMR (151 MHz, CDCl₃) δ 147.8, 147.0, 132.4, 127.8 (2C), 126.4, 125.2 (2C), 75.9, 51.4, 28.3, 14.5, −1.2. IR (neat) 3474 cm⁻¹. HRMS (ESI) m/z 249.1680 [(M+H)⁺; calcd for C₁₅H₂₅OSi, 249.1675].

Compound 35 (mixture of diastereomers, 0.63:0.37 ratio): ¹H NMR (500 MHz, CDCl₃) δ 7.34–7.26 (m, 3 H), 7.19 (m, 2 H), 5.68 (m, 0.37 H), 5.49 (m, 0.63 H), 5.40 (m, 0.37 H), 5.31 (m, 0.63 H), 4.51 (m, 1 H), 1.66 (ddd, J = 0.5, 2.0, 7.0 Hz, 1.11 H), 1.53 (ddd, J = 0.5, 1.5, 6.5 Hz, 1.89 H), 1.43 (s, 1 H), 1.33 (dd, J = 0.5, 7.0 Hz, 1.89 H), 1.21 (dd, J = 0.5, 7.5 Hz, 1.11 H). ¹³C NMR (151 MHz, CDCl₃) δ 143.28, 143.26, 131.33, 131.30, 128.6, 128.2, 128.1, 128.1, 127.7, 126.9, 126.7, 126.4, 71.7, 71.5, 46.6, 45.9, 17.5, 16.0, 13.5, 13.2. IR (neat) 3397 cm⁻¹.

Compound 36 (mixture of diastereomers 1:1 ratio) ¹H NMR (500 MHz, CDCl₃) δ 9.70 (m, 1 H) 9.57 (m, 1 H), 7.28 (m, 4 H), 7.18 (m, 2 H), 7.14 (m, 4 H), 2.67 (m, 1 H), 2.56 (m, 1 H), 2.49 (ddd, J = 1.0, 4.5, 16.0 Hz, 1 H), 2.33–2.26 (m, 3 H), 2.17 (ddd, J = 3.0, 9.0, 16.0 Hz, 1 H), 2.07 (ddd, J = 2.5, 8.5, 16.0 Hz, 1 H), 1.26 (d, J = 2.5 Hz, 3 H), 1.25 (d, J = 2.5 Hz, 3 H), 0.99 (d, J = 6.5 Hz, 3 H), 0.85 (d, J = 6.5 Hz, 1 H). ¹³C NMR (151 MHz, CDCl₃) δ 202.8, 202.5, 145.7, 144.8, 128.4, 128.2, 127.8, 127.5, 126.3, 126.2, 49.5, 48.3, 45.1, 44.5, 34.6, 34.4, 18.4, 18.03, 17.98, 17.6. IR (neat) 1724 cm⁻¹. Diastereomers 36 are known compounds and have spectral data in accord with those previously reported.^20b

4.4 Preparation of diastereomeric 27

A 100 mL round-bottomed flask equipped with a magnetic stir bar and a N₂ line was charged with 9-BBN (0.5 M solution in THF, 2.04 mL, 1.02 mmol) and substrate 14 (671 mg, 2.85 mmol) was then added as a THF solution (0.57 M). The reaction was refluxed at an oil bath temperature of 90 °C, for 10 h. The reaction mixture was cooled to 55 to 65 °C and ethanol (2.0 mL), NaOH (6 M, 0.5 mL) and H₂O₂ (30% w/w, 1.0 mL) were added. The reaction was stirred at 55 to 65 °C for 1 h and cooled to room temperature. The aqueous phase was saturated with K₂CO₃, phases were separated, the organic phase was dried over anhydrous magnesium sulfate and concentrated to afford the crude product. Purification by silica gel (EtOAc 0–10% in hexanes) gave 702 mg of pure syn-27/anti-27 in 97% yield. IR (neat) 3370 cm⁻¹.

anti-27: ¹H NMR (500 MHz, CDCl₃) δ 7.40–7.23 (m, 5 H), 4.56 (q, J = 6.5 Hz, 1 H), 3.86 (m, 1 H), 3.78 (m, 1 H), 3.23 (t, J = 5.0 Hz, 1 H), 2.75 (m, 1 H), 2.20–2.13 (m, 1 H), 1.63 (m, 1H), 1.42 (d, J = 6.5 Hz, 3 H), −0.06 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 143.5, 128.2 (2C), 127.5, 126.8 (2C), 76.4, 70.1, 62.4, 31.7, 23.7, −2.8. HRMS (EI) m/z 253.1618 [(M+H)⁺; calcd for C₁₄H₂₅O₂Si, 253.1624].

syn-27: ¹H NMR (500 MHz, CDCl₃) δ 7.34 (m, 4 H), 7.29 (m, 1H), 4.36 (q, J = 6.5 Hz, 1 H), 3.54–3.45 (m, 2 H), 3.26 (dd, J = 4.5, 9.5 Hz, 1 H), 1.81 (m, 1 H), 1.67-1.60 (m, 2 H), 1.42 (d, J = 6.5 Hz, 3 H), 0.09 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 143.6, 128.5 (2C), 127.8, 126.8 (2C), 77.9, 69.1, 61.3, 33.8, 23.6, −2.5. HRMS (EI) m/z 253.1627 [(M+H)⁺; calcd for C₁₄H₂₅O₂Si, 253.1624].

4.5 Preparation of ester syn-28

A mixture of the substrate alcohol (obtained by 9-BBN oxidation of reactive syn-27 (201 mg, 0.80 mmol), and 3,5-dinitrobenzoyl chloride (366 mg, 1.59 mmol) in pyridine as solvent, was heated to reflux for 52–55 h. Then the solvent was removed under reduced pressure and the crude product purified by chromatography on silica gel (hexanes/EtOAc (0–10%) to afford 194 mg, 55% of the expected ester syn-28 as a solid. Recrystallization from a 1:1 EtOH/hexane mixed solvent afforded product as colorless crystals mp 74.5–75.5 °C. IR (neat) 1728, 1630 cm ^{−1. 1}H NMR (300 MHz, CDCl₃) δ 9.17 (t, J = 2.2 Hz, 1 H), 8.88 (d, J = 2.2 Hz, 2 H), 7.29–6.99 (m, 5 H), 4.36–4.29 (m, 2 H), 4.18–4.10 (m, 1 H), 3.25–3.20 (dd, J = 8.0, 6.0 Hz, 1 H), 1.90–1.83 (m, 2 H), 1.41 (d, J = 6.3 Hz, 3 H), 0.15 (s, 9 H). ¹³C NMR (500 MHz, CDCl₃) δ 162.2, 148.4, 143.9, 134.0, 129.2, 128.2, 127.4, 126.7, 122.1, 78.1, 66.7, 64.4, 30.6, 23.6, −2.7. HRMS (ESI) m/z 447.1599 [(M+H)⁺; calcd for C₂₁H₂₇N₂O₇Si, 447.1587].

4.6 Preparation of cis-29 and trans-29

To a solution of unreactive (anti) 17 (167 mg, 0.641 mmol) in CH₂Cl₂ (10 mL, ~0.7 M) was added Grubbs second-generation catalyst (4 mol%, 21.4 mg, 0.025 mmol) and the solution was stirred under nitrogen at room temperature for 3 h. The reaction mixture was concentrated and purified by column chromatography (10% CH₂Cl₂ in hexanes) to afford 144 mg 72 of cis-29 (97%).

cis-29: ¹H NMR (500 MHz, CDCl₃) δ 7.34-7.22 (m, 5 H), 5.82–5.78 (m, 2 H), 4.38 (dd, J = 3.5, 10 Hz, 1 H), 4.17–4.15 (m, 1 H), 2.26–2.12 (m, 2 H), 0.08 (s, 9 H); ¹³C NMR (126 MHz, CDCl₃) δ 144.0, 128.10 (2C), 128.06, 126.9, 125.6 (2C), 121.1, 75.3, 71.6, 34.2, −4.0. HRMS (CI) m/z 261.1681 [(M+H⁺) calcd for C₁₆H₂₄OSi, 261.1675]. The relative stereochemistry of cis-29 was assigned based on positive NOESY signals between protons at 4.15 ppm and 2.26–2.12 ppm.

Following the same procedure for the reactive (syn) 17 (184 mg, 0.707 mmol) and Grubbs second-generation catalyst (4 mol%, 24 mg, 0.028 mmol) in CH₂Cl₂ for 3 h, followed by column chromatography (30% CH₂Cl₂ in hexanes) afforded 151 mg (92%) of trans-29.

trans-29: ¹H NMR (500 MHz, CDCl₃) δ 7.38–7.24 (m, 5 H), 5.83-5.76 (m, 2 H), 4.72 (t, J = 5.5 Hz, 1 H), 4.01 (m, 1 H), 2.41–2.38 (m, 2 H), 0.09 (s, 9 H); ¹³C NMR (126 MHz, CDCl₃) δ 142.3, 128.5 (2C), 128.4, 127.5, 126.9 (2C), 120.3, 72.6, 70.4, 30.4, −2.7. HRMS (CI) m/z 261.1664 [(M+H⁺) calcd for C₁₆H₂₄OSi, 261.1675]. The relative stereochemistry of trans-29 was confirmed based on negative NOESY signals between protons at 4.72 ppm and 4.01 ppm.

4.7 Preparation of compound 38

The [1,4]-Wittig product 31 from the Wittig rearrangement of α-alkoxysilane of E-30 (217 mg, 0.87 mmol) was dissolved in THF (0.24 M, 3.6 mL), and 3N NaOH (0.83 mL/mmol starting material, 0.72 mL) added. The mixture was heated to 35–40 °C, and then oxidized by dropwise addition of 30% H₂O₂ (0.42 mL/mmol starting material, 0.36 mL), while maintaining the reaction temperature below 50 °C for 2 h. The aqueous phase was cooled to 0 °C, and acidified to pH of 1–2 wit h 6 M HCl. The resulting aqueous material was extracted with ether (5 × 20 mL), and the ether solution dried with anhydrous MgSO₄. Filtration and concentration afforded 158 mg (94% yield) of diastereomeric 38 as a thick colorless oil. Purification by column chromatography on silica gel (hexane/EtOAc 0–10%) afforded 387 as a 2.8:1 mixture of diastereomers (ratio by ¹H NMR). IR (neat) 3100–2500 (br), 2967, 1707, 1495, 1452, 1412, 1290 cm⁻¹.

Major, anti-38: ¹H NMR (500 MHz, CDCl₃) δ 11.49 (br s, 1 H), 7.31–7.18 (m, 5 H), 2.72–2.65 (quint, J = 7.1 Hz, 1 H), 2.53–2.48 (dd, J = 14.8, 4.4 Hz, 1 H), 2.28–2.18 (m, 1 H), 2.14–2.09 (dd, J = 15.1, 9.1 Hz, 1 H), 1.28–1.26 (d, J = 7.1 Hz, 3 H), 0.88 (d, J = 6.6 Hz, 3 H). ¹³C NMR (125 MHz, CDCl₃) δ 179.8, 144.8, 128.2, 127.6, 126.2, 44.4, 38.9, 36.2, 18.3, 17.5.

Minor, syn-38: ¹H NMR (500 MHz, CDCl₃) δ 11.49 (br s, 1 H), 7.33–7.18 (m, 5 H), 2.63–2.56 (quint, J = 7.1 Hz, 1 H), 2.35–2.31 (apparent dd, J = 15.4, 4.4 Hz, 1 H), 2.28–2.18 (m, 1 H), 2.04–1.99 (dd, J = 14.8, 9.3 Hz, 1 H), 1.27–1.24 (d, J = 7.1 Hz, 3 H), 1.04 (d, J = 6.6 Hz, 3 H). ¹³C NMR (125 MHz, CDCl₃) δ 179.8, 145.7, 128.2, 127.5, 126.1, 44.7, 39.8, 36.4, 18.3, 17.2. Anti-38 and syn-38 are known compounds and have spectral data in accord with those previously reported.¹⁹

4.8 Assignment of relative stereochemistry for compound 33

4.8.1 Preparation of 39

To a cold (−78 °C) solution of 33 (50 mg, 0.20 mmol, dr = 1.4:1) in 1:1 CH₂Cl₂/EtOH (3 mL) was added a suspension of NaBH₄ (15.2 mg, 2 equiv) in EtOH (0.8 mL). After 1 hour the temperature was slowly raised to rt and the reaction stirred overnight. The reaction mixture was then treated with H₂O (2 mL) and diluted with diethyl ether. The aqueous phase was washed with diethyl ether (x2). Combined organic extracts were washed with brine, dried over MgSO₄ and concentrated. Column chromatography (5% EtOAc in hexanes) gave 28.5 mg (57%) of 39 in two fractions of different diastereomeric ratio along with 5.2 mg (11%) of unreacted 33.

Compounds 39 (major diastereomer): ¹H NMR (600 MHz, CDCl₃) δ 7.29 (m, 2 H), 7.20 (m, 3 H), 3.80 (m, 1 H), 2.72 (quint, J = 6.6 Hz, 1 H), 1.42 (m, 1 H), 1.31 (d, J = 7.2 Hz, 3 H), 1.29 (d, J = 4.8 Hz, 1 H), 1.07 (m, 1 H), 0.84 (s, 3 H), 0.82 (m, 1 H), −0.10 (m, 9 H). ¹³C NMR (151 MHz, CDCl₃) δ 144.8, 128.4 (2C), 127.7 (2C), 126.3, 73.0, 46.4, 36.3, 15.9, 15.0, 13.1, −3.6. IR (neat) 3423, 2955, 1456, 1248, 839 cm⁻¹. HRMS (ESI) m/z 232.1651 [(M–OH)⁺; calcd for C₁₅H₂₄Si, 232.1647].

4.8.2 Preparation of 40

One fraction of 39 (15.6 mg, 0.063 mmol, dr = 10:2:1) was dissolved in pyridine (1 mL) and 3,5-dinitrobenzoyl chloride (flakes were crushed prior to addition) was added in one portion. After 48 hours the mixture was diluted with diethyl ether (15 mL) and washed with 1M HCl (2 mL × 3), H₂O, brine, dried over MgSO₄ and concentrated. Partial separation of the diastereomers (2 fractions) by column chromatography (4% EtOAc in hexanes) gave 24.4 mg (56%) of 40 as a solid. Recrystallization of one fraction from CH₂Cl₂/hexanes gave a single diastereomer of 40 whose relative stereochemistry was determined by x-ray crystallography.

Compound 40 (major diastereomer): ¹H NMR (500 MHz, CDCl₃) δ 9.22 (t, J = 2.0 Hz, 1 H), 9.13 (d, J = 2.0 Hz, 2 H), 7.30 (t, J = 7.5 Hz, 2 H), 7.22 (m, 3 H), 5.86 (m, 1 H), 3.09 (m, 1 H), 1.75 (ddd, J = 2.5, 10.0, 13.0 Hz, 1 H), 1.32 (d, J = 7.0 Hz, 3 H), 1.26 (ddd, J = 2.5, 12.0, 14.5 Hz, 1 H), 0.87 (d, J = 7.5 Hz, 3 H), 0.47 (m, 1 H), −0.14 (s, 9 H). ¹³C NMR (126 MHz, CDCl₃) δ 162.4, 148.8 (2C), 142.7, 134.3, 129.4 (2C), 128.7 (2C), 127.7 (2C), 127.0, 122.3, 79.4, 44.5, 34.2, 17.7, 15.4, 13.5, −3.7. mp 139–140 °C. IR (neat) 3107, 2957, 1726, 1545, 1348, 1278, 1170, cm⁻¹.

4.8.3 Preparation of anti-33 from recrystallized 40

To a solution of 40 (4.4 mg, 0.010 mmol, dr > 95:5) in THF (1 mL) was added 3M NaOH (0.5 mL) and the mixture stirred for 2 hours in an oil bath at 45 °C. Then, the reaction mixture was diluted with diethyl ether (10 mL). The aqueous phase was washed with diethyl ether (2 mL × 2). Combined organic extracts were washed with H₂O, brine, dried over MgSO₄ and concentrated. Pasteur pipette chromatography (5% EtOAc in hexanes) gave 1.5 mg (61%) of 39 as a single diastereomer. This alcohol (1.5 mg) was dissolved in dry CH₂Cl₂ (0.5 mL) and DMP (0.3 M in CH₂Cl₂, 0.25 mL, excess) was added at room temperature. After 1 hour the reaction mixture was concentrated, suspended in 5% EtOAc in hexanes and filtered through a plug of silica to give ~1.5 mg (ca. 100%) of anti-33.