Stereospecific Cu(I)-Catalyzed C–O Cross-Coupling Synthesis of Acyclic 1,2-Di- and Trisubstituted Vinylic Ethers from Alcohols and Vinylic Halides

CuI and trans-N,N′-dimethylcyclohexyldiamine catalyze the single-step C–O bond cross-coupling between 1,2-di- and trisubstituted vinylic halides with functionalized alcohols, producing acyclic vinylic ethers. This stereospecific transformation selectively gives each of the (E)- and (Z)-vinylic ether products from the corresponding vinyl halide precursors. This method is compatible with carbohydrate-derived primary and secondary alcohols and several other functional groups. The conditions are mild enough to reliably generate vinylic allylic ethers without promoting Claisen rearrangements.

Correcting conditions from a published literature example 1 In 2003, Nordmann and Buchwald reported C-O cross-couplings of several primary alcohols with the 1,2-disubstituted (E)-vinylic iodide 5, using catalytic CuI and 3,4,7,8-tetramethyl-1,10phenanthroline (L1) with stoichiometric Cs2CO3 in toluene or o-xylene solvent, at 80 o C, with reaction times in the range 12 -36 h. 1 For these examples (in ref. 1, Table 1, entries 4-5, and also in Table 3, entries 1 -3), the publication states that air atmosphere was used. The "General procedure A" in the supporting information describing the experimental protocol for these examples did not indicate that inert atmosphere or air-free technique was required.
In contrast, the examples that combined CuI-L1-catalyzed cross-couplings of allylic alcohols + vinylic iodides with Claisen rearrangement of the allylic vinylic ether intermediates were conducted at 120 o C using argon atmosphere. The "General procedure B" in the supporting information for these examples also states that argon atmosphere was used. However, this difference implies that argon atmosphere was required only at the higher temperature of 120 o C.
After initially encountering difficulties extending the Nordmann -Buchwald method to other alcohols, we repeated an example in the 2003 paper to validate our technique, preparing (E)vinylic ether S1 from (E)-vinylic iodide 5 with benzyl alcohol: Table 3, entry 3: Our deviations from the published experimental procedure (ref. 1) are described in bold. We did not use dodecane as an internal standard, but relied on the isolated mass of purified product S1 and calculated the isolated yield relative to the limiting reactant, (E)-1-iodo-1-decene (5).
1) An early validation experiment was conducted on 30% of the scale reported in ref. 1, with 0.2 equiv CuI and 0.4 equiv L1, to ensure that enough catalyst and ligand was present. The reactants were weighed in air, and the reaction was performed in a sealed vial under air. Specifically: A 5 mL conical vial was charged with benzyl alcohol (65 mg, 0.60 mmol, 2.0 equiv), CuI (11.4 mg, 0.06 mmol, 0.2 equiv relative to vinylic iodide 5), L1 (28.4 mg, 0.12 mmol, 0.4 equiv relative to vinylic iodide 5), (E)-vinylic iodide 5 (80 mg, 0.30 mmol, 1.0 equiv), Cs2CO3 (146.9 mg, 0.45 mmol, 1.5 equiv relative to vinylic iodide 5) and toluene (0.2 mL, 1.5 M based on vinylic iodide 5). The vial was sealed with a screw cap. The reaction mixture was vigorously stirred at 80 ºC for 17 h. The resulting suspension was cooled to room temperature and filtered through a 1 x 2 cm pad of silica gel eluting with diethyl ether (100 mL).The filtrate was concentrated by rotary evaporation, and the residue was purified by flash chromatography on silica gel (eluent = 100% hexanes) to afford pure product S1 (25.6 mg, 35% yield). 2) A subsequent validation experiment was conducted on the same scale, using 1 mmol of (E)vinylic iodide 5 as the limiting reactant, which contained 8% 1-decene. Specifically: A 5 mL conical vial was charged with benzyl alcohol (216 mg, 2.0 mmol, 2.0 equiv), CuI (19.0 mg, 0.1 mmol, 0.1 equiv), L1 (47.3 mg, 0.2 equiv), (E)-vinylic iodide 5 (266 mg, 1.0 mmol, 1.0 equiv), Cs2CO3 (489 mg, 1.5 mmol, 1.5 equiv) and toluene (0.5 mL, 2 M). The vial was topped with a water-cooled condenser, topped with a septum, connected by a needle to an argon line. (The reaction vessel was not purged.) The reaction mixture was vigorously stirred at 80 ºC for 17 h. The resulting suspension was cooled to room temperature and filtered through a 1 x 2 cm pad of silica gel eluting with diethyl ether (100 mL).The filtrate was concentrated by rotary evaporation, and the residue was purified by flash chromatography on silica gel (eluent = 100% hexanes) to afford pure product S1 in (118 mg, 53% yield when accounting for the 1-decene impurity in the sample of vinylic iodide 5).

Synthesis of (E)-1-bromo-1-decene (S7):
In an oven-dried 25mL round bottom flask was charged with Cp2ZrCl2 (987 mg, 1.7 equiv.) and 1decyne (0.36 mL, 1.0 equiv) in THF (5.5 mL, 0.35 M). To this solution at room temperature was added a 1.0 M THF solution of LiHBEt3 (2.6 mL, 1.3 equiv). After 10 min of stirring, Nbromosuccinimide (NBS, 471 mg, 1.2 equiv) was added. After additional stirring for 30 min, the reaction mixture was quenched with a 1N HCl solution (50 mL) and then extracted with diethyl ether (50 mL× 3). The organic layer was washed with aqueous saturated NaHCO3 (50 mL) and brine (50 mL), dried over MgSO4, and then filtered on filter paper. The solvent was concentrated on rotary evaporator. The crude was dissolved in hexanes, and white solids crashed out. The solution was filtered through a short pad of Celite ® /silica gel and washed with hexanes to yield (E)-1-bromo-1-decene as a colorless oil (0.391 g, 89% yield
Step 1: (Z)-1-iodo-1-undecene (crude 11, 50:1 Z/E, containing 15% 1,1-diiodo-1-undecene S8): In a 250 mL oven-dried round bottom flask charged with a stirbar and wrapped in aluminum foil, (iodomethyl)triphenylphosphonium iodide (15.0 g, 28.5 mmol, 1 equiv) was added and placed under inert atmosphere by purging with a constant stream of argon gas. Dry THF (71 mL, 0.4 M based on phosphonium salt, freshly distilled) was added to the reaction flask, followed by the dropwise addition of sodium hexamethyldisilazide (NaHMDS, 1 M in THF, 29 mL, 1 equiv) over 5 minutes in the dark at room temperature to give an orange liquid. After the addition, the mixture was stirred at room temperature for 20-30 minutes, resulting in a dark red liquid with some precipitate. The reaction flask was then cooled to -78 °C in a dry ice/acetone bath, and hexamethylphosphoramide (HMPA, 9 mL) was added. The mixture was stirred for approximately 5 minutes before adding decanal (4.2 mL, 22.8 mmol, 0.8 equiv) to the reaction mixture. For optimal Z/E ratio, the reaction was stirred vigorously at -78 °C in the dark under argon for 3 h to give a thick dark orange mixture. The reaction mixture warmed to about 10-15 °C, then saturated NaHCO3 solution (~36 mL) was added to quench the reaction, and hexanes were added to dilute the reaction mixture, which gave a solution of orange liquid and a yellow-white precipitate. The mixture was filtered through a Celite ® plug on a 100 mL fritted funnel, using hexanes and water to rinse. The organic layer was separated from the water layer in a 500 mL separatory funnel, and the aqueous layer was extracted twice with hexanes. The organic layer was filtered twice through a silica plug, using 100 mL fritted funnel. The silica plug was washed with additional hexanes, and the total organic layer was washed with saturated Na2S2O3 solution, dried with anhydrous Na2SO4, filtered, and concentrated to give a yellow crude oil of vinylic iodide (4.2 g). The crude mixture was analyzed by 1 H NMR, to determine the Z/E ratio of the vinylic iodide 11 (50:1 Z/E) and the content of 1,1-diiodo-1-undecene (S8, approximately 15%) by integrating the following alkenyl proton resonances: • 6.94 ppm (t, J = 7.0 Hz, 1H) for 1,1-diiodo-1-undecene (S8) • 6.54 ppm (dt, J = 14.3, 7.2 Hz, 1H) for (E)-1-iodo-1-undecene • 6.22-6.11 ppm (m, 2H) for (Z)-1-iodo-1-undecene (11).
The crude mixture was subsequently subjected to Zn-Cu-mediated stereoselective reduction without further purification. 10

S-10
Synthesis of (S,Z)-4-(2-iodovinyl)-2,2-dimethyl-1,3-dioxolane (S10): 12 In a 250 mL oven-dried round bottom flask charged with a stirbar and wrapped in aluminum foil, (iodomethyl)triphenylphosphonium iodide (4.5 g, 8.5 mmol) was added (with minimal exposure to light) and placed under inert atmosphere via purging with a constant stream of argon gas. A portion of dry THF (10 mL, freshly distilled) was added to the reaction flask, followed by the dropwise addition of NaHMDS (1 M in THF, 8.5 mL) over 5 -10 minutes in the dark at room temperature (wrapped in aluminum foil) to give an orange liquid. After the addition, the mixture was stirred at room temperature for 20-30 minutes, resulting in a dark red liquid with some precipitate. While waiting for ylide formation, D-glyceraldehyde acetonide (S3, 50 wt% in CH2Cl2, 1.77 g, 6.8 mmol) was weighed by pouring directly into a separate 25 mL round bottom flask. The aldehyde was placed under argon atmosphere and dissolved in dry THF (11 mL); this required constant swirling over 10 minutes to dissolve the aldehyde. The reaction flask for ylide generation was cooled to -78 °C with a dry ice/acetone bath, and HMPA (2.6 mL) was added. The aldehyde/THF solution was added to the reaction flask, and the reaction mixture was stirred in the dark, at -78 °C to -35 °C, over 3 hours. After completion of the reaction, saturated NaHCO3 (12 mL) was added to quench and ethyl acetate (EtOAc) was used to dilute the reaction. The mixture was filtered through a Celite ® plug (on a 100 mL fritted funnel) using EtOAc and water to rinse. Then, the organic layer was separated from the aqueous layer, and the organic components were extracted two more times from the aqueous layer before being washed with saturated Na2S2O3 solution and brine, dried with anhydrous Na2SO4, filtered, and concentrated to give a yellow crude oil. The crude mixture was dissolved in 10:1 hexanes/EtOAc (100 mL), which caused precipitation to crash out, and was filtered through a silica plug using 200 mL of the same eluent to rinse the crude flask. The filtrate was concentrated to furnish the title compound S10 as a yellow oil (1.17 g, 68% yield, 10: 1 Z/E containing <2% of 1,1-diiodoalkene). The mixture was analyzed by 1 H NMR for the Z/E ratio and the content of 1,1-diiodoalkene using the integrations of the following alkenyl protons: • 7.15 (d, J = 7.3 Hz, 1H) for 1,1-diiodoalkene • 6.59 -6.46 (m, 2H) for E-isomer S6 • 6.46 -6.35 ppm (m, 2H) for Z-isomer S10.
Synthesis of 100% (Z)-1-iodo-1-decene (S12) via cis-hydrogenation of 1-iodo-1-decyne (S11): 1-iodo-1-decyne (S11): To a solution of 1-decyne (2 g) in acetone (72 mL, purged with argon balloon) under argon atmosphere in the dark, N-iodosuccinimide (NIS, 3.9 g) was added followed by AgNO3 (296 mg). The reaction was stirred overnight at room temperature under argon in the dark to give a pale clear liquid with yellow precipitate. The reaction was diluted with deionized H2O and hexanes to give an orange mixture, which was then transferred into a separatory funnel. The aqueous layer was extracted with hexanes three times, and the combined organic layer was washed in sequence with deionized H2O, saturated aqueous Na2S2O3, and brine. The combined organic extracts were filtered through a silica plug, and rinsed with hexanes. The filtrate was concentrated at medium vacuum with the water bath at 20 °C to avoid product evaporation, to give a pink oil. 1 H NMR signals matched the literature report. 13 The crude product S11 (3.2 g) was used without further purification.
Characterization data for 1-iodo-1-decyne (S11): Synthesis of ortho-nitrobenzenesulfonylhydrazide (NBSH): To a 50 mL round bottom flask charged with a stir bar, ortho-nitrobenzenesulfonyl chloride (4 g) was added, purged with argon atmosphere, and dissolved with dry THF (18 mL). The reaction flask was cooled to -30 °C using a carefully monitored cold bath of dry ice and acetone. Hydrazine hydrate (N2H4-H2O, 2.2 mL) was carefully added to the mixture, and the reaction was stirred at -30 °C over 30 minutes. After stirring was complete, 40 mL of ethyl acetate was added to dilute the reaction, and the organic layer was washed with 10% aqueous NaCl solution (w/v) (five times with 30 mL portions). The organic layer was dried with anhydrous Na2SO4 and filtered slowly into ~210 mL of hexanes in a separate flask over 5 minutes, which caused a precipitate to crash out of solution. The flask was stored in a freezer overnight. The solid precipitate was filtered and washed with cold hexanes (~100 mL total). The solid was dried overnight at high vacuum, to give NBSH as a pale yellow flaky solid (

S-12
(Z)-1-iodo-1-decene (S12): 1-Iodo-1-decyne (S11, 869 mg, 3.3 mmol) was added to a 250 mL round bottom flask charged with a stir bar, and then purged with argon atmosphere. Compound S10 was dissolved with i-PrOH (41.5 mL) and dry THF (41.5 mL). The reaction mixture was cooled to 0 °C in an ice bath and wrapped with aluminum foil. Freshly prepared NBSH (1.5 g, 6.9 mmol, 2.1 equiv) was added under argon, followed by triethylamine (1.6 mL, 12 mmol, 3.5 equiv). The reaction mixture slowly warmed to room temperature overnight. After 19 hours, the reaction mixture was dark clear yellow, which was concentrated by rotary evaporation to give an orange residue with pale yellow oil. The crude compound was dissolved in a minimum amount of CH2Cl2 and loaded onto 4 g of Celite ® , and purified by silica gel chromatography, eluting with 100% hexanes. The desired product was collected (note: S12 was UV-active and the first band to elute), and concentrated to give a clear colorless oil (450 mg, 51% yield, containing 5% iododecane). The 1 H NMR spectrum of the product S12 matched previous literature. 13 Characterization data for (Z)-1-iodo-1-decene (S12):

Synthesis of trisubstituted vinylic iodide
(E)-1-iodo-2-methyldec-1-ene (S13): The reaction protocol was modified based on reported procedures. [15][16][17] To an argon-charged 100 mL oven-dried round bottom flask, Cp2ZrCl2 (292 mg, 1 mmol) was added to dry CH2Cl2 (20 mL), followed by slow addition of Al(CH3)3 (2M in heptane, 7.5 mL) at room temperature (CAUTION: Smoke was observed). The reaction mixture was cooled to 0 o C over 10 min, and deionized water (0.13 mL) was added dropwise. The reaction mixture was then warmed to room temperature and stirred for 20 min, followed by the addition of 1-decyne (0.9 mL, 5 mmol). After stirring for 30 min, an argon-charged solution of iodine (1.42 g, 6 mmol) in dry THF (7.5 mL) was added to the reaction mixture, which was then stirred for an additional 30 min. The reaction was quenched by careful pouring into saturated aqueous K2CO3 (50 mL) (CAUTION: vigorous bubbling occurs) and was filtered through a Celite ® pad using CH2Cl2 to rinse the reaction vessel. The organic solution was washed with saturated aqueous Na2S2O3 and brine, dried with anhydrous Na2SO4, filtered and concentrated to give a clear light-yellow crude oil (730 mg). The crude mixture was purified by column chromatography using silica gel in 100% pentane to furnish the title compound (495 mg, 95:5 regioisomer ratio S13:S14, also containing ~15% of 2-methyl-1-decene, determined by 1 H NMR analysis
K2CO3-treated deuterated CDCl3 was used for all NMR spectra, both before and after the purification process.

S-20
The initial solvent and ligand screening (Table S1) revealed that the anionic ligand L2 in tetraglyme solvent facilitated the C-O cross-coupling reaction, using a 1 : 1 ratio of CuI : L2.
Reducing the temperature from 125-130 °C to 80 °C only slightly decreased the yield of (E)-vinylic ether 9, and more substantially decreased the yield of enyne side-product 10. With anionic ligand L2 in tetraglyme solvent and increasing catalyst loading to 0.2 equiv CuI and 0.2 equiv L2 with 3 equiv of Cs2CO3, we achieved full conversion of vinylic iodide 5 with equimolar amount of alcohol 8, resulting in excellent yield of (E)-vinylic ether 9. Anionic ligands L5, L6, and L7 gave inferior results, as did o-xylene and DMF solvent with L2. Upon additional screening of polyether solvents, 1,2-dimethoxyethane (DME) gave excellent results for C-O cross-couplings with the neutral ligand L4. With neutral ligand L4 in DME solvent and increasing catalyst loading to 0.2 equiv CuI and 0.4 equiv L4 with 3 equiv of Cs2CO3, we achieved full conversion of vinylic iodide 5 with equimolar amount of alcohol 8, resulting in excellent yield of (E)-vinylic ether 9. The anionic ligand L8 was also incompatible with the (Z)-selective cross-coupling reaction. The additional basic carboxylate moiety of both amino acid ligands L2 and L8 accelerates the enyneforming side reaction, outcompeting C-O bond formation. The increased steric bulk of the neutral diamine ligand L9 adversely affected the outcome of the copper-catalyzed reaction. Iminopyridine ligand L10, were previously developed for C-O and C-N cross-couplings with aryl halides, providing different electronic environments for the oxidative addition step (electron-rich pyridine nitrogen) and reductive elimination step (electron-poor imine nitrogen). However, the only product observed was a trace of enyne 13, arising from elimination of (Z)-vinylic iodide 11, which is a pathway that was not available for the substrates described in that literature. 32  Reaction optimization was initially performed with L3 after its initial success with the (Z)-selective C-O cross-coupling reaction. To suppress formation of enyne 13, we screened several weaker bases such as K2CO3 and CsHCO3. However, the lower basicity also retarded the rate of alcohol deprotonation, resulting in inferior yields of vinylic ether 12 and incomplete conversion of vinylic iodide 11. Cs2CO3 was found to be the most effective base.

Additional optimization data for (Z)-vinylic ether synthesis
We also tested other Cu(I) sources with L3 as alternatives to CuI. CuBr provided comparable yield of vinylic ether 12, but the Z : E ratio was significantly reduced when compared with CuI. The halide-free catalyst Cu(MeCN)4(BF4) 33 gave lower yield of vinylic ether 12 despite the complete consumption of iodide 11. 1 H NMR analysis of the crude mixture revealed that a portion of the vinylic iodide 11 was converted to the terminal alkyne product (1-undecyne) via basepromoted elimination (ca. 10%). Thus, we concluded that CuI and Cs2CO3 are necessary to catalyze and promote the desired Ullmann-type cross-coupling reaction.   Using only 1.2 equiv of alcohol 8 relative to the (Z)-vinylic iodide 11 provided comparable yield and retention of configuration to that of the optimized condition only when 3 equiv of Cs2CO3 were utilized, suggesting that the alcohol deprotonation step is critically important for the catalytic cycle. Utilizing 2 equiv of alcohol 8 for our optimized procedure offered more generality as we subsequently extended substrate scope. The success of the C-O cross-coupling reaction with (Z)-vinylic iodide 11 was highly sensitive to temperature: a slight increase from 75 to 85 o C reduced the yield of (Z)-vinylic ether 12 and increased the yield of (Z)-enyne side product 13.

Scope of (E)-and (Z)-selective cross-coupling products from primary alcohols (Figure 2a)
Representative procedure for the synthesis of 1,2-disubstituted (E)-vinylic ethers (Conditions A, Figure 2): An oven-dried 4 mL vial charged with a stir bar was added alcohol (1 equiv), Cs2CO3 (3 equiv), L4 (0.4 equiv), and E-vinylic iodide 5 (1 equiv) under argon. The reaction vial was purged continuously with argon for 5 min before CuI (0.2 equiv) was added. Anhydrous DME (0.5 mL, 0.7 M based on vinylic iodide) was added, and the reaction mixture was bubbled with argon for 5 minutes. The reaction vial was quickly closed with a solid cap, sealed with Teflon tape and electrical tape, and placed on the heat with internal temperature at 75 °C. The reaction was stirred over 40 -48 hours, cooled to room temperature, and diluted with diethyl ether. The mixture was filtered through a Celite ® pad and rinsed with diethyl ether (100 mL). The filtrate was concentrated by rotary evaporator, and the crude mixture was subjected to flash column chromatography with 2% Et3N-treated silica gel using hexanes/EtOAc as eluent to purify vinylic ethers. Figure 2): Similar protocol to E-vinylic ether synthesis was used with the following stoichiometry: alcohol (2 equiv), Z-vinylic iodide 11 (1 equiv), L4 (1 equiv), CuI (0.5 equiv.), Cs2CO3 (3 equiv), and anhydrous DME (0.5 mL). The work-up was performed using EtOAc as diluting solvent. The crude mixture was subjected to flash column chromatography with either neutral alumina or 2% Et3N-treated silica gel. The Z/E ratio of Z-vinylic ether product was determined by 1 H NMR analysis using alkenyl proton signals at 6.00 -5.95 ppm (J = 6 -8 Hz) for Z-isomer and 6.25 -6.20 ppm (J = 12 -14 Hz) for E-isomer, using K2CO3-treated deuterated chloroform before and after the purification process. Yields and isomer ratios of vinylic ether products are reported after chromatography, as combined yields for inseparable Z-and Eisomers.
We also observed an identical mass peak in the mass spectrum of the (Z)-isomer 40, see below.