ACS Catalysis
Letter
Nitrogen protecting groups such as CBz and Boc (7) are
tolerated; however, a N-nosylate-protected substrate did not
afford product (not shown). Simple carbon-substituted
benzene rings afforded products 8, 9, 10, and 11 all with
good selectivity. The reaction was successful with function-
alized boronic acids as well. Using p-OMe and p-CF3-
substituted phenylboronic acids yielded 12 and 13. Organo-
boron nucleophiles bearing functional group handles such as
halides or nitriles, allowing for further elaboration, were also
tolerated in the reaction to afford products 14, 15, 16, 17, and
18. Heterocyclic boronic acids also performed well in the
reaction to afford compounds 19, 20, 21, and 22, although
using quinolinyl nucleophiles resulted in lower yields and
products with lower ee (23 and 24).
a large array of chiral ligands that identified a highly efficient
and selective catalyst, and the optimized conditions were
general, achieving good to excellent ee and drs for a variety of
substrates. This transformation was suitable for the mod-
ification of a variety of dehydroamino ester derivatives,
showcasing the robustness of the reaction and its potential
for late-stage functionalization of complex biological targets,
which is currently under further exploration. This method can
be used as a reliable synthetic tool for the synthesis of
analogues of both L- and D-amino acids and other substituted
pyrrolidine scaffolds to enable the exploration of novel
chemical matter.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
■
sı
Under standard reaction conditions, however, reactivity was
not observed with pyridyl organoboron reagents. Because these
are important structural motifs in the context of pharmaceuti-
cally relevant chemical matter, we felt it was important to
attempt to engage these nucleophiles in the reaction. We found
more forcing conditions (dioxane as solvent, heating to 100
°C) were necessary for product formation. Increased yield was
observed with increasing the boronic acid loadings to 3 equiv,
using Cs2CO3 as base, and lowering the H2O loading to 10
equiv. With these modifications to the reaction protocol, we
were able to obtain products 25−28 in synthetically useful
yields. In certain cases, with the 4-pyridyl substituted
nucleophiles, low levels of diastereo- and enantioinduction
were observed (25, 26, and 27).28 Despite the diminished
selectivity, these products are challenging to prepare otherwise.
Sterically encumbered substrates (29 and 30) gave low
yields under our standard reaction conditions because of
consumption of the organoboron reagent in an undesired
protodeboronation pathway.14a Whereas background reactions
can cause protodeboronation, a competing protonation of a
Rh(I)-aryl intermediate after transmetalation may contribute
to the composition as well. With simple hydrocarbon boronic
acid nucleophiles, we observed that protodeboronation only
occurred in the presence of the Ph-BPE-Rh catalyst. This Rh-
catalyzed protodeboration pathway becomes more competitive
with sterically bulky substrates, likely because of the substrate
(1) being unable to bind to a more hindered Rh-center. Higher
yields of racemic mixtures of product were observed with these
substrates with Rh(nbd)2BF4 to give the results shown. The
reaction with cyclohexenyl boronic acid to give 31 was also low
yielding under the optimized reaction conditions, but good
yield of the racemate was obtained with the achiral catalyst.29
To gain a better understanding of the high diastereose-
lectivities observed in the reaction, several additional experi-
ments were performed. When single stereoisomer cis-2 is
subjected to the standard reaction conditions, trans-2 is not
detected, indicating that epimerization is not responsible for
the observed diastereomeric ratios. DFT calculations show the
difference between the ground state energies of trans-2 and cis-
2 (ΔG) is approximately 0.18 kcal/mol, equating to a 1.3:1 dr
single stereoisomer trans-2 is subjected to basic conditions,
erosion of the diastereomeric ratio is observed, indicating that
the diastereoselectivity seen in the catalytic transformation is
kinetically driven.30
Instrumental information, optimization details, detailed
experimental procedures, absolute stereochemistry de-
termination, and product characterization (PDF)
AUTHOR INFORMATION
Corresponding Author
■
Emma K. Edelstein − Department of Process Research and
Development, Merck & Co., Inc., Boston, Massachusetts
Authors
Danica A. Rankic − Department of Process Research and
Development, Merck & Co., Inc., Boston, Massachusetts
Caroline C. Dudley − Department of Process Research and
Development, Merck & Co., Inc., Boston, Massachusetts
02115, United States
Spencer E. McMinn − Department of Discovery Chemistry,
Merck & Co., Inc., Boston, Massachusetts 02115, United
States
Donovon A. Adpressa − Department of Process Research and
Development, Merck & Co., Inc., Boston, Massachusetts
02115, United States
Complete contact information is available at:
Author Contributions
All authors have given approval to the final version of the
manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Shaoguang Zhang for assistance with the high-
throughput chiral ligand screen. We thank Aaron Sather,
Michael Ardolino, Thomas Lyons, Emily Corcoran, Kelsey
Poremba, Matthew Maddess, and L.-C. Campeau for helpful
conversation, advice, and guidance during the preparation of
this manuscript. We would also like to acknowledge the
internship program which C.C.D. participated in.
We have developed a highly stereoselective Rh-ACA for the
modular synthesis of trans-3-aryl proline derivatives, a chemical
space which is difficult to access efficiently with current
methods. High-throughput experimentation was used to screen
REFERENCES
■
747
ACS Catal. 2021, 11, 743−749