Bin
Wang
a,
Derek Yiren
Ong
a,
Yihang
Li
a,
Jia Hao
Pang
a,
Kohei
Watanabe
b,
Ryo
Takita
*b and
Shunsuke
Chiba
*a
aDivision of Chemistry and Biological Chemistry, School of Physical and Mathematical Sciences, Nanyang Technological University, 637371, Singapore. E-mail: shunsuke@ntu.edu.sg
bGraduate School of Pharmaceutical Sciences, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. E-mail: takita@mol.f.u-tokyo.ac.jp
First published on 4th May 2020
A concise protocol for anti-hydromagnesiation of aryl alkynes was established using 1:1 molar combination of sodium hydride (NaH) and magnesium iodide (MgI2) without the aid of any transition metal catalysts. The resulting alkenylmagnesium intermediates could be trapped with a series of electrophiles, thus providing facile accesses to stereochemically well-defined functionalized alkenes. Mechanistic studies by experimental and theoretical approaches imply that polar hydride addition from magnesium hydride (MgH2) is responsible for the process.
We embarked on our investigation of chemical reactivity of various main group metal hydrides, derived from NaH and the corresponding metal halides, toward reduction diphenylacetylene (1a). Among main group metal iodides examined for the optimization (see the ESI† for details), we found that a combination of NaH and MgI2 shows a promising reactivity toward semi-hydrogenation of 1a to trans-stilbene (2a) (Table 1).15 Use of NaH and MgI2 in 1:1 molar ratio resulted in the best outcome for the formation of 2a (entries 1–3). Full conversion of 1a was attained at 100 °C as the reaction temperature, providing stilbene 2a in 96% yield with high trans-selectivity (trans:cis = 94:6) (entry 4). Interestingly, the reactions with MgBr2 and MgCl2 gave comparable results (entries 5 and 6),16 implying that a common reactive magnesium hydride species is generated and responsible for the present reduction of 1a (vide infra). It should be noted that the reactions with other alkaline earth metal iodides based on Ca, Sr, and Ba were not optimal for this transformation.
Entry | MgX2 (equiv.) | Temp [°C] | Conv.b [%] | Yieldb [%] | |
---|---|---|---|---|---|
2a | 2a′ | ||||
a All the reactions were conducted using 0.5 mmol of 1a in THF (2.5 mL). b GC yields with n-dodecane as an internal standard. c Isolated yield was 96% as a 94:6 trans/cis-mixture. d Isolated yield was 93% as a 94:6 trans/cis-mixture. | |||||
1 | MgI2 (1.5) | 80 | 60 | 38 | 1 |
2 | MgI2 (2) | 80 | 62 | 40 | 2 |
3 | MgI2 (3) | 80 | 73 | 68 | 4 |
4 | MgI2 (3) | 100 | >99 | 93c | 6 |
5 | MgBr2 (3) | 100 | 99 | 93d | 5 |
6 | MgCl2 (3) | 100 | 94 | 89 | 4 |
Having optimized the reaction conditions in hands (Table 1, entry 4), we next investigated the substrate scope of alkynes for their trans-semi-reduction (Scheme 2A). Various diarylalkynes 1b–1g could be converted selectively into the corresponding trans-alkenes 2b–2g in good yields in general. Stereoselectivity was slightly dropped when a sterically bulky aryl group is installed onto the substrates (for 2c and 2d), while the electronic nature of the aryl substituents does not affect much onto the stereoselectivity (2e–2gvs.2h–2i). Alkynes having heteroaryl motifs such as furan, thiophene, and indole could also be reduced in efficient manners (for 2j–2l). Reduction of 1-phenyl-2-t-butylacetylene (1m) afforded the corresponding trans-alkene 2m in good yield, while that of 1-phenyl-2-phenyldimethylsilylacetylene (1n) resulted in formation of alkenyl silane 2n in only moderate yield. However, this method was found not applicable for reduction of dialkyl alkynes such as 5-decyne (1o). Deuterium labeling experiments using D2O (for quenching) and NaD unambiguously suggested the presence of alkenylmagnesium species A as an intermediate, that is formed via anti-hydromagnesiation (Scheme 2B).
The alkenylmagnesium A could be further functionalized by subsequent treatment with various electrophiles (Scheme 3A).17 Formylation with N,N-dimethylformamide (DMF) proceeded smoothly to form 2,3-diphenylacrylaldehyde (3a) in 70% yield. Addition of 1.2 equiv. of 4-anisaldehyde afforded allylalcohol 4a in 87% yield, whereas use of 2.1 equiv. of 4-anisaldehyde resulted in Oppenauer-type oxidation18 to form α,β-unsaturated ketone 5a in 76% yield. Allylation with allylbromide was facilitated by a catalytic amount (10 mol%) of CuCN·2LiCl19 to afford skipped diene 6a in 94% yield. Similarly, installation of a benzyloxymethyl (BOM) unit was achieved for synthesis of 7a using BOMCl. Use of 1,2-dibromo-1,1,2,2-tetrachloroethane allowed for smooth bromination of A to form alkenyl bromide 8a. Borylation of A could be achieved with pinacolborane by following the Singaram's protocol,20 affording alkenylboronic ester 9a in 91% yield.
We next questioned how the steric and electronic bias could affect the regioselectivity on the hydromagnesiation of unsymmetrical aryl alkynes (Scheme 3B). The reaction of 1c having a sterically bulkier 1-naphthyl group underwent selective hydromagnesiation to form alkenylmagnesium B with installation of a hydride at the less hindered distal carbon (marked in blue) as a major form, that could be trapped by subsequent allylation to provide 6c in 66% yield.21 On the other hand, hydromagnesiation of 1e having an electron-rich 4-dimethylaminophenyl group resulted in installation of hydride on the proximal carbon (marked in red) to generate alkenylmagnesium C, that was functionalized with BOMCl to provide 7e in 66% yield. Hydromagnesiation of 1-phenyl-2-t-butylacetylene (1m) occurred in regioselective manner, where hydride attack was observed at the β-carbon (marked in red) to the phenyl group, while the side was sterically shielded by the t-Bu group. The resulting alkenylmagnesium D could be further transformed into BOM adduct 7m in 77% yield.
To elucidate the active hydride species responsible for the present anti-hydromagnesiation of alkynes, we conducted several control experiments (Scheme 4 and see the ESI†). Ashby reported preparation of magnesium hydride MgH2 in bulk state by treatment of MgBr2 with 2 equiv. of NaH in THF at room temperature (in quantitative yield together with the formation of inert sodium bromide).22,23 We observed that the reaction of alkyne 1a with bulk MgH2, prepared from 3 equiv. of NaH and 1.5 equiv. of MgBr2 by following the Ashby's protocol, resulted in selective formation of trans-stilbene (2a) despite poor conversion of 1a (Scheme 4A). On the other hand, treatment of 1a with bulk MgH2 in the presence of MgI2 (1.5 equiv.) greatly enhanced the conversion of 1a, providing almost the same outcome with that in the optimized reaction conditions (Scheme 4B). The IR spectrum of the mixture obtained from the reaction of NaH and MgI2 in 1:1 molar ratio showed the presence of only MgH2 and MgI2 (NaI is transparent in the IR window) (Scheme 4C). We reasoned that while bulk MgH2 itself is less hydridic due to its polymerized structure with high lattice energy,24 synergistic cooperation between bulk MgH2 and MgI2 through counter ion metathesis allows for freshly generating more hydridic MgH2 probably of smaller units (Scheme 4D), that is the key for the success in use of NaH and MgI2 in 1:1 molar ratio for the efficient anti-hydromagnesiation.
The DFT calculations for the reactions of diphenylacetylene (1a) with a MgH2 dimer as the model of activated (MgH2)n species were thus carried out at the ωB97XD/6-311++G**/SMD(THF)//ωB97XD/6-31++G** level of theory (Scheme 5). From INT1anti, the reduction of diphenylacetylene smoothly proceeded viaTSanti (ΔG‡ +23.6 kcal mol−1), in the manner of anti-hydromagnesiation to afford alkenylmagnesium species INT2anti. In this event, the hydride in the same plane as one of the benzene rings attacks on the proximal alkyne carbon center and, the magnesium cation successively shifts to the distal alkyne carbon, indicating the polar hydride transfer mechanism. Thus, the steric and electronic bias of the aromatic ring should result in profound effects on the regioselectivity (i.e.Scheme 3B). Further investigations led to find the syn-reduction pathway. In TSsyn, diphenylacetylene is distorted, where two phenyl rings are almost perpendicular (the dihedral angle is ca. 84°). Accordingly, TSsyn was located 2.6 kcal mol−1 higher than TSanti. These computed results well corroborated the reducing ability of the NaH–MgI2 system for the anti-selective hydromagnesiation.25
We observed that the reaction of propargyl alcohol 10a with 3 equiv. of NaH26 and 2 equiv. of MgI2 proceeded smoothly at 60 °C, affording trans-alkene 11a as a single isomer in 87% yield (Scheme 6A). Based on the deuterium labelling experiments using NaD and D2O (Scheme 6A for use of NaD in the reduction of 10a. See the ESI† for details), we proposed that the process is triggered by the formation of alkoxymagnesium hydride E, that mediates hydromagnesiation to afford 5-membered magnesiocycle F. This hydroxy-guided approach27 allowed for trans-semi-reduction of various propargylic alcohols 10b–10g into the corresponding allylic alcohols 11b–11g (Scheme 6B). Heteroaromatic motifs such as 2-thienyl and 2-pyridyl groups were compatible with the process (for 11b and 11c). Sterically more hindered substrates based on adamantane (for 10d) and derived from (R)-carvone (for 10e) generated homoallylic alcohols 11d and 11e in 56% and 77% yields, respectively. Secondary propargylic alcohols 10f and 10g could also be smoothly reduced. Formation of 11e and 11g kept alkenyl moieties intact, suggesting that the present protocol is selective in the hydromagnesiation of alkynes. Phenol and aniline moieties were also capable in guiding trans-semi-reduction of alkynes for providing the corresponding 11h–11k in good to moderate yields.
Scheme 6 Guided reduction. a The reactions were conducted using 0.5 mmol of 10 with NaH (3 equiv.) and MgI2 (2 equiv.) THF (2.5 mL) at 60 °C. b The reaction with NaD installed a deuterium on the β-styryl moiety in 93% incorporation rate (see the ESI† for details). cTrans:cis = 99:1. d11g was isolated as its TBS ether. See the ESI† for details. e The alkyne reduction was conducted at 80 °C. f The alkyne reduction was conducted using NaH (6 equiv.) and MgI2 (4 equiv.) at 100 °C. g CuCN·2LiCl (10 mol%), allylbromide (4 equiv.), 0 °C, 2 h. h Diethyl carbonate (4 equiv.), 40 °C, 15 h. i HBpin (1.1 equiv.), 0–24 °C, 0.5 h then treatment with Si gel. |
The 5-membered magnesiocycle intermediate F could be further functionalized by CuCN-catalyzed allylation to form skipped diene 12a in good yield (Scheme 6C). Treatment of F with diethyl carbonate allowed for construction of lactone 13a. Furthermore, the reaction with pinacolborane followed by workup with Si gel resulted in formation of 1,5-dihydro-1,2-oxaborole 14a. On the other hand, we found that the reaction of 2-(phenylethynyl)phenol (10h) likely involves a mixture of 6-membered magnesiocycle G and 5-membered one H, which could be trapped under CuCN-catalyzed allylation reaction conditions to afford skipped dienes 12h (41% yield) and 12h′ (37% yield), respectively.
Footnote |
† Electronic supplementary information (ESI) available: Experimental details, including procedures, syntheses and characterization of new compounds; 1H and 13C NMR spectra. CCDC 1987693 and 1987694. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/d0sc01773f |
This journal is © The Royal Society of Chemistry 2020 |