Angewandte
Communications
Chemie
Table 1: Ligand screening for asymmetric allene synthesis.[a]
[a] Using 0.2 mmol of alkyne 3a, 0.02 mmol of CuI, 0.03 mmol of ligand and 0.4 mmol of Et3N at RT in 2 mL 1,4-dioxane, with 0.4 mmol of hydrazone
1a (with DIPEA as buffer) flowed through activated MnO2 then a back pressure regulator (BPR). All reactions proceeded in >95% conversion.
[b] Allene/alkyne ratio determined by 1H NMR analysis of the crude reaction mixture. [c] ee determined by chiral HPLC.
Fortunately, switching the flanking oxazoline units to
imidazolines (PyBIMs) proved fruitful, allowing an improve-
ment in the enantioselectivity (96% ee, entry 4). In particular,
para-substituted electron-withdrawing groups on the N-aryl
substituent proved essential to providing allene 4a as the
major product, with the 4-SF5 group providing the best result
(entry 9). Further increases in the electron deficiency (4-nitro
and 4-triflyl, entries 10 and 11) of the aromatic ring provided
moderate enantioselectivities, presumably due to poorer
binding to the copper center. We believe that the ambidentate
oxazoline units of the more commonly used PyBOX ligands
are the origin of this differing efficiency in enantioselectivity,
as coordination to the oxygen atom could lead to poorer
enantioselection (see SI for a detailed discussion). In contrast,
PyBIM ligands were superior as they offer only a single
binding site on the imidazoline units. Furthermore, modu-
lation of electronic effects was critical to achieve correct
coupling selectivity. To the best of our knowledge, this is the
first rationalization of the performance of PyBOX versus
PyBIM ligands,[12] which could be important for future
development of new catalytic asymmetric reactions.
With the new ligand in hand, we then tested a variety of
propargylated amine derivatives and flow-generated diazo
compounds for the asymmetric allenylation reaction
(Table 2). Overall although yields in general are moderate
(30–57%), the enantioselectivity of the reaction is excellent
(89–98% de/ee) and proceeds rapidly under mild conditions.
The remaining mass balance was mainly due to the alkyne
cross-coupled product, which could generally be removed by
careful column chromatography. A myriad of sensitive func-
tional groups were found to be compatible with the reaction
conditions, including ketones (4c), aldehydes (4d), epoxides
(4e), unprotected alcohols (4i), esters (4l, 4u) and terminal/
internal olefins (4s, 4t and 4v), all with excellent enantiose-
lectivities. Notably, several heterocyclic examples including
thiophenes (4 f), furans (4j), indoles (4ac) and pyrroles (4ad)
were also tolerated despite their potential to interfere with
the catalytic cycle, all with good enantioselectivity. Variation
in the electronic properties of the diazo compound had
a small effect on the yield and enantioselectivity, with more
electron-withdrawing substituents such as 4-fluoro (4o) and
3-cyano (4q) leading to slightly lower ee values (91% ee and
89% ee, respectively, compared to 94–96% ee for 4m, 4n, 4p
and 4r). It was also easy to scale up the asymmetric
allenylation to 5 mmol of propargylamide for allene 4w in
our flow system without the need to store excessive quantities
of diazo compound, which provided 0.89 g of chiral material
in similar yield and ee to the smaller scale run.
To further probe the functional group tolerance of this
procedure, we were able to conduct the late-stage asymmetric
allenylation of various propargylated amine derivatives of
seven drug molecules/natural products (4x–4ad), including
(S)-ibuprofen, aspirin, penicillin G, topiramate, pregabalin,
indomethacin and atorvastatin, again all proceeding with high
de/ee values. Surprisingly, the presence of the thioether
functionality on penicillin G was not detrimental to the
stereoselectivity.
We were able to demonstrate the utility of the chiral
allenes generated in a medicinal chemistry context by
conducting the silver-mediated cyclization[13] of allene 4a to
its corresponding 3-pyrroline 5 (Scheme 2), with good chir-
ality transfer and excellent yield (97% ee to 95% ee). The
process can therefore overall be regarded as a formal
enantioselective sp2–sp3 coupling, with a functional handle
(internal olefin) for potential further derivatization.
We anticipate that the reaction mechanism proceeds as
depicted below (Scheme 3). In the presence of base and the
copper catalyst, the initially generated ligand–copper acety-
2
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!