Highly enantioselective carbene insertion into N–H bonds of aliphatic amines

Article abstract of DOI:10.1126/science.aaw9939

Aliphatic amines strongly coordinate, and therefore easily inhibit, the activity of transition-metal catalysts, posing a marked challenge to nitrogen-hydrogen (N–H) insertion reactions. Here, we report highly enantioselective carbene insertion into N–H bo

Full text of DOI:10.1126/science.aaw9939

RESEARCH
respect to the diazo reactant in the N–H
insertion of morpholine was investigated
next (Fig. 2C). Diazo esters with linear or
branched a-alkyl chains afforded the desired
products in good to high yields (66 to 99%)
with excellent enantioselectivities (94 to
ORGANIC CHEMISTRY
Highly enantioselective carbene insertion
into N–H bonds of aliphatic amines
Mao-Lin Li, Jin-Han Yu, Yi-Hao Li, Shou-Fei Zhu*, Qi-Lin Zhou*
96% ee) (27 to 31). Various functional groups
(
alkenyl, ester, ether, amide) appended to
the alkyl chain were tolerated, giving high
yields (86 to 99%) and enantioselectivities
Aliphatic amines strongly coordinate, and therefore easily inhibit, the activity of
transition-metal catalysts, posing a marked challenge to nitrogen-hydrogen (N–H) insertion
reactions. Here, we report highly enantioselective carbene insertion into N–H bonds of
aliphatic amines using two catalysts in tandem: an achiral copper complex and chiral
amino-thiourea. Coordination by a homoscorpionate ligand protects the copper center
that activates the carbene precursor. The chiral amino-thiourea catalyst then promotes
enantioselective proton transfer to generate the stereocenter of the insertion product.
This reaction couples a wide variety of diazo esters and amines to produce chiral a-alkyl
a–amino acid derivatives.
(
87 to 96% ee) (32 to 38). Furthermore, a-aryl
diazoacetates also afforded the correspond-
ing arylglycine derivatives (39 to 44) in high
yields (>95%) with good enantioselectivities
(72 to 89% ee).
To demonstrate the further synthetic utility
of the N−H insertion reaction, several trans-
formations of the insertion products were per-
formed. The product (R)-3 was reduced by
LiAlH4 to afford (R)-2-benzylamino-butanol
[(R)-45], an intermediate for the synthesis of
g-secretase inhibitors (30) and PDK1 inhibi-
tors (31) (Fig. 3A). The product 15, which could
be prepared at gram scale from N−H insertion
of morpholine and diazo ester 2, was hydro-
lyzed to acid 46, an intermediate for the syn-
thesis of hyperproliferative disorder (HPD)
treatment agents (Fig. 3B) (32). The (R)-2-
morpholinopropanoic acid (47), which is
prepared by hydrolysis of the product 27,
is a key intermediate for the synthesis of the
phosphatidylinositol 3-kinase d (PI3Kd) inhib-
itors (33), as well as DNA-dependent protein
kinase (DNA-PK) inhibitors (34) (Fig. 3C).
To gain deeper insight into the mechanism
of the N−H insertion reaction, we performed
kinetic analyses by using in situ infrared (IR)
spectroscopy. To accelerate the kinetics, the
initial rates of the reaction were measured at
various concentrations of the components at
40°C (figs. S1 to S5). The rate showed a first-
order dependence on concentrations of Tp*Cu
and diazo compound 2 (Fig. 4A), which indi-
cates that the formation of metal carbenoid
through Tp*Cu-catalyzed decomposition of
diazo ester 2 is the likely rate-limiting step.
The negative first-order dependence on CAT
is consistent with a pre-equilibrium forma-
tion of a resting-state complex between the
thiourea catalyst CAT and Tp*Cu, which would
suppress the copper-catalyzed decomposition
of the diazo ester. However, benzylamine, gen-
erally coordinating with the metal catalyst and
suppressing the formation of metal carbe-
noid, showed a zero-order kinetic effect in the
reaction, which suggests that the coordina-
tion of CAT to Tp*Cu is much stronger than
that of benzylamine, and the inhibition by
benzylamine can be ignored (fig. S6). We
posit that the negative Tp* ligand renders the
copper catalyst a softer Lewis acid that favors
interaction with a soft base like sulfur. Fur-
ther evidence for the stronger interaction be-
tween CAT and Tp*Cu includes observations
hiral amines are ubiquitous in natural
products, pharmaceuticals, and agro-
chemicals. Approximately 43% of the top
Tp*Cu [Tp* is hydrotris(3,5-dimethylpyrazolyl)
borate] (23–25) with a chiral amino-thiourea
(CAT) bearing a pyrrolidine motif (26–29)
(Fig. 1C, lower). The reaction provides effi-
cient, highly enantioselective access to chiral
a-alkyl a–amino acid derivatives bearing sec-
ondary and tertiary amino substituents, which
are difficult to prepare by other methods.
Under the optimal reaction conditions,
a broad range of aliphatic amines was
then investigated for N–H insertion with 2-
phenylpropan-2-yl a-diazobutyrate 2 (Fig. 2A).
The benzylic primary amines underwent the
N–H insertion reaction smoothly to afford the
corresponding a-aminobutanoic acid deriva-
tives (3 to 9) in high yields (81 to 95%) with
high enantioselectivities [88 to 92% enanti-
omeric excess (ee)], though 2-phenethylamine
and n-butylamine gave moderate enantiose-
lectivities (10 and 11). Secondary amines were
also suitable substrates for the reaction but re-
quired longer reaction times and excess diazo
compounds for satisfactory outcomes. Piper-
idine derivatives generally exhibited high
enantioselectivities, and the introduction of
2
00 prescription medicines in 2016 con-
tain an aliphatic amine moiety (Fig. 1A)
1, 2). The development of highly enantiose-
C
(
lective transition-metal–catalyzed reactions
that form C–N bonds is thus of long-standing
interest in synthetic chemistry (3–5). Transition-
metal–catalyzed carbenoid insertion into N–H
bonds has proven a straightforward method in
this respect, benefitting from mild reaction
conditions, good functional group tolerance,
and readily available reactants (6, 7). Recently,
chiral transition-metal catalysts have been
successfully applied to enantioselective N–H
insertion reactions in the synthesis of natural
or unnatural chiral a–amino acid derivatives
8, 9). However, these reactions have been re-
stricted to aromatic amines (10–15) or amides
16–18) (Fig. 1B). Aliphatic amines are compara-
tively stronger Lewis bases and thus poison the
metal catalysts by strong coordination, inter-
fering with generation of the metal carbenoid
(
(
(
19, 20). Moreover, excess aliphatic amines
can displace the ylide from metal-ylide inter-
mediates, leading to racemic product forma-
tion from the free ylide (Fig. 1C, upper). We
envisioned that a combination of two catalysts
2
electron-withdrawing groups (CO Me and
CN) at the 4-position led to higher yields (71
and 86%, respectively) and better enantiose-
lectivities (90 and 94% ee, respectively) (12
to 14). Morpholine, substituted piperazines,
and thiomorpholine also underwent the N–
H insertion and gave the desired products
(15 to 19) in high yield with 87 to 97% ee.
Fused heterocyclic amines also afforded N–
H insertion products in satisfactory yields
and enantioselectivities (20 and 21). How-
ever, azepane and N-methyl-1-phenylme-
thanamine gave lower enantioselectivities
(73 and 77% ee, respectively, 22 and 23).
The N–H insertion reactions with chiral drugs,
such as amoxapine, trimetazidine, and vorti-
oxetine, proceeded smoothly to afford cor-
responding a-aminobutanoic acid derivatives
(
21, 22) might address these challenges: An
achiral transition-metal catalyst compatible
with aliphatic amines would generate the ylide
intermediate, and a separate chiral catalyst
would then promote enantioselective proton
transfer. After exploring various transition-
metal catalysts and chiral H-bonding catalysts in
the N–H insertion reaction of a-diazobutanoates
with benzylamine (tables S1 to S5), we report
here the success of this approach, pairing the
homoscorpionate-coordinated copper complex
State Key Laboratory and Institute of Elemento-Organic
Chemistry, College of Chemistry, Nankai University, Tianjin
00071, China.
Corresponding author. Email: sfzhu@nankai.edu.cn (S.-F.Z.);
qlzhou@nankai.edu.cn (Q.-L.Z.)
3
*
(
24 to 26) in high yields with good enan-
tioselectivities (Fig. 2B). The scope with
Li et al., Science 366, 990–994 (2019)
22 November 2019
1 of 5

RESEARCH
| REPORT
of the changes in nuclear magnetic resonance
NMR) and ultraviolet (UV)–visible spectra af-
ter adding CAT into a mixture of Tp*Cu and
benzylamine (figs. S7 to S11). Taken together,
the kinetic, NMR, and UV studies are con-
sistent with the Tp*Cu•CAT complex, rather
of the Tp*Cu•CAT complex using DFT revealed
a minimum energy structure in which bonding
of the Tp*Cu to the sulfur atom of the thiourea
enhances Brønsted acidity (Fig. 4B).
pyrazol rings were also evaluated under the
standard reaction conditions (Fig. 4D). De-
spite a large fluctuation in the yield, the in situ
IR studies showed that all tested Tp ligands
promoted high conversions and that the major
by-product was 2-phenylpropan-2-yl but-2-
enoate, resulting from the b-H migration
of the metal carbenoid (figs. S15 to S19).
Modifying the Tp ligands also influenced
the enantioselectivity when the same chiral
thiourea catalyst was used, indicating involve-
ment of the copper catalyst in the enantio-
determining step. By contrast, upon tuning
the electronic properties of the arene ring of
the chiral thiourea catalyst, the enantiose-
lectivity decreased precipitously, whereas the
yield remained almost unchanged (table S7).
We again hypothesize that copper coordina-
tion enhances the Brønsted acidity of the
thiourea catalyst while minimally influencing
the distant site of enantioinduction (Fig. 4C).
On the basis of the aforementioned mech-
anistic studies, a catalytic cycle is proposed in
Fig. 4E. The Tp*Cu•CAT complex serves as a
resting state of the catalyst and dissociates to
release Tp*Cu, which catalyzes transformation
of the diazo ester into the metal carbenoid in
the rate-determining step. Nucleophilic attack
on the metal carbenoid by the aliphatic amine
(
The enantio-determining proton-transfer
step was studied computationally by using
DFT with the Tp*Cu•CAT complex as a cata-
lyst. The structures corresponding to the lowest-
energy transition states for the major and minor
enantiomers of product are presented in Fig.
4C. The optimal transition state TSRaCu-I is
only 2.8 kcal/mol higher in free energy than the
free enol, implying extremely efficient proton-
transfer catalysis. In accord with experimental
observations, the calculated energy of transi-
tion state TSRaCu-I was 4.9 kcal/mol lower
than that of the TSSaCu-I transition state lead-
ing to the unfavored (S)-product of N−H in-
sertion. Besides the different hydrogen-bonding
interactions in the transition states TSRaCu-I
and TSSaCu-I, the S−Cu bond in TSRaCu-I is
markedly shorter than that in TSSaCu-I. The
shorter S−Cu bond indicates stronger coor-
dination of copper by thiourea and likely
higher Brønsted acidity of the thiourea cata-
lyst, which would promote proton transfer to
the substrate.
2
than the Tp*Cu•BnNH complex, as the rest-
ing state of the catalyst in the reaction. Al-
though the Tp*Cu•CAT complex is the main
resting state of the copper, free Tp*Cu is still
evident under the reaction conditions (fig. S10)
and can react with the diazo compound.
Density functional theory (DFT) calculations
indicated that the copper catalyst in the inter-
mediate Cu-ylide could be recoordinated by
thiourea to release a free ylide or its more
stable tautomer free enol (Fig. 4B). Even with-
out the added chiral catalyst, the proton trans-
fer of these intermediates is still a very rapid
process and can be mediated by trace water,
enol intermediates, and even substrates them-
a
selves (18, 35). The similar pK values (where
K
a
is the acid dissociation constant) (figs. S12
to S14 and table S8) of the Brønsted acidic site
and protonated basic site of the thiourea pro-
mote concerted proton transfer, whereby the
thiourea protonates the newly formed stereo-
genic center, and the amino group deproton-
ates the free enol. Computational modeling
Several other tris(pyrazolyl)borate (Tp) lig-
ands bearing different substituents on the
Fig. 1. Strategy for enantiocontrol of
N–H insertion reactions of aliphatic
amines with carbenes. (A) Representa-
tive drugs demonstrating the ubiquity of
chiral aliphatic amines in bioactive
molecules. (B) Amine sources
reported for enantioselective
N–H insertion reactions.
(C) Enantioselective transition-metal–
catalyzed N–H insertion reactions
with aliphatic amines: challenges and
solutions. Optimal reaction conditions:
The reaction of 1 (0.2 mmol),
2
(0.22 mmol), Tp*Cu (5 mole %),
and CAT (6 mol %) was carried out in
3
2
ml of methyl tert-butyl ether (MTBE) at
5°C for 20 hours. BnNH , benzylamine;
, tert-butyl carbamate; CbzNH
2
BocNH
2
2
,
benzyl carbamate; Me, methyl; Et, ethyl;
Ph, phenyl; M, metal; ref., reference.
Li et al., Science 366, 990–994 (2019)
22 November 2019
2 of 5

RESEARCH
| REPORT
Fig. 2. Scope of aliphatic amines
and a-diazo esters in the
enantioselective N–H insertion
reaction. Reaction conditions:
amines (0.2 mmol), a-diazo esters
(
0.22 mmol), Tp*Cu (5 mol %),
CAT (6 mol %), 3 ml MTBE,
5°C, 20 hours. Isolated yields
A
B
2
are given. The ee values were
determined by high-performance
liquid chromatography. (A) Scope of
aliphatic amines. (B) Application
to enantioselective late-stage
functionalization of pharmaceuticals.
(
*
3
C) Scope of a-diazo esters.
Diazo esters (0.3 mmol),
6 hours. †Diazo esters (0.3 mmol),
MTBE:CH
2
Cl
2
= 10:1, 36 hours.
‡
Diazo esters (0.3 mmol), 40°C,
0 hours. Bu, tert-butyl; Pr, iso-propyl.
t
i
2
C
Fig. 3. Synthetic transformations
of the N−H insertion products.
A
(A) Transformation of 3 to (R)-2-
benzylamino-butanol [(R)-45],
a key intermediate for the
synthesis of bioactive molecules.
THF, tetrahydrofuran; rt, room
temperature. (B) Formal synthesis
of HPD treatment agents with
the N−H insertion as key step.
B
C
(C) Transformation of 27 to
4
7, a key intermediate for the
synthesis of bioactive molecules.
Li et al., Science 366, 990–994 (2019)
22 November 2019
3 of 5

RESEARCH
| REPORT
Fig. 4. Mechanistic studies.
A) Kinetic profiles of
Cu-catalyzed N−H insertion
reaction of 2 and BnNH
B) Calculated Gibbs free
(
2
.
(
energy (DG) of Cu-ylide,
free ylide, and free enol.
Lowest-energy ground-state
structure of the Tp*Cu•CAT
complex. Structures of
alternative higher-energy
complexes are provided in
fig. S20. (C) DFT-optimized
lowest-energy transition
structures for R and
S products. Calculations
were performed at the
m062x-D3/def2tzvpp//m062x-
D3/def2svp level. Structures
of alternative higher-energy
complexes are provided
in figs. S21 to S25. (D) Influ-
ence of different Tp ligands.
(E) Proposed catalytic cycle for
the enantioselective carbene
insertion into N–H bonds
of aliphatic amines. conv.,
conversion; DMSO, dimethyl
sulfoxide; equiv, equivalents.
generates a metal ylide. The catalyst CAT dis-
places the ylide from the metal-ylide in-
termediate to generate free enol and the
Tp*Cu•CAT complex. The Tp*Cu•CAT com-
plex then promotes proton transfer in the free
enol through a push-pull mechanism: The
amino moiety accepts a proton from the hy-
droxy group of the enol while the thiourea
moiety donates a proton to the b-carbon of
the enol.
The success of the overall transformation
relies on the combined properties of the achiral
copper catalyst and chiral organocatalyst. This
study not only solves a long-standing challenge
in enantioselective carbene insertion reactions
Li et al., Science 366, 990–994 (2019)
22 November 2019
4 of 5

RESEARCH
| REPORT
but also provides a potentially general strat-
egy for transition-metal–catalyzed asymmetric
transformations involving strongly coordinat-
ing substrates.
16. E. C. Lee, G. C. Fu, J. Am. Chem. Soc. 129, 12066–12067
2007).
B. T. Golding, R. J. Griffin, I. R. Hardcastle, S. Rodriguez-Aristegui,
WO patent 2010136778 (2010).
35. Y. Liang, H. Zhou, Z.-X. Yu, J. Am. Chem. Soc. 131, 17783–17785
(2009).
(
1
7. B. Xu, S.-F. Zhu, X.-L. Xie, J.-J. Shen, Q.-L. Zhou, Angew.
Chem. Int. Ed. 50, 11483–11486 (2011).
18. Y.-Y. Ren, S.-F. Zhu, Q.-L. Zhou, Org. Biomol. Chem. 16,
087–3094 (2018).
3
ACKNOWLEDGMENTS
1
9. S. R. Hansen, J. E. Spangler, J. H. Hansen, H. M. L. Davies,
Org. Lett. 14, 4626–4629 (2012).
0. I. D. Jurberg, H. M. L. Davies, Chem. Sci. 9, 5112–5118 (2018).
1. A. E. Allen, D. W. C. Macmillan, Chem. Sci. 2012, 633–658
(
We thank X.-S. Xue for discussions of the DFT calculations
and M. P. Doyle for comments and suggestions on the
preparation of the manuscript. Funding: We thank the
National Natural Science Foundation of China (21625204,
21790332, 21532003), the “111” project (B06005) of the
Ministry of Education of China, and the National Program
for Special Support of Eminent Professionals for financial
support. Author contributions: Q.-L.Z. and S.-F.Z.
conceived the study; M.-L.L., S.-F.Z., and Q.-L.Z. designed
the experiments and analyzed the data; M.-L.L. performed
the reactions and the mechanistic and DFT studies;
J.-H.Y. and Y.-H.L. made some of the diazo substrates;
and M.-L.L., S.-F.Z. and Q.-L.Z. wrote the manuscript.
Competing interests: The authors declare no competing
interests. Data and materials availability: Additional
optimization and mechanistic data are provided in the
supplementary materials.
REFERENCES AND NOTES
1
.
N. A. McGrath, M. Brichacek, J. T. Njardarson, J. Chem. Educ.
7, 1348–1349 (2010).
2
2
8
2
.
.
These pharmaceutical posters are freely accessible
to anyone as PDF files at https://njardarson.lab.
arizona.edu.
T. C. Nugent, Chiral Amine Synthesis: Methods, Developments
and Applications (Wiley VCH, 2010).
2012).
2. D.-F. Chen, Z.-Y. Han, X.-L. Zhou, L.-Z. Gong, Acc. Chem. Res.
7, 2365–2377 (2014).
23. S. E. Bromberg et al., Science 278, 260–263 (1997).
2
4
3
2
4. S. Trofimenko, Scorpionates: Polypyrazolylborate Ligands
and Their Coordination Chemistry (Imperial College Press,
1999).
4
.
Y. Yang, S.-L. Shi, D. Niu, P. Liu, S. L. Buchwald, Science 349,
6
2–66 (2015).
S.-L. Shi, Z. L. Wong, S. L. Buchwald, Nature 532, 353–356
2016).
5
.
25. M. E. Morilla et al., Chem. Commun. 2002, 2998–2999
(2002).
26. M. S. Sigman, E. N. Jacobsen, J. Am. Chem. Soc. 120,
4901–4902 (1998).
(
6
.
M. P. Doyle, M. A. McKervey, T. Ye, Modern Catalytic
Methods for Organic Synthesis with Diazo Compounds
(
Wiley, 1998).
2
7. T. Okino, Y. Hoashi, Y. Takemoto, J. Am. Chem. Soc. 125,
12672–12673 (2003).
7.
A. Ford et al., Chem. Rev. 115, 9981–10080 (2015).
8
.
S.-F. Zhu, Q.-L. Zhou, Acc. Chem. Res. 45, 1365–1377 (2012).
28. S.-Z. Nie et al., Tetrahedron Asymmetry 21, 2055–2059
(2010).
9. D. Gillingham, N. Fei, Chem. Soc. Rev. 42, 4918–4931 (2013).
SUPPLEMENTARY MATERIALS
1
0. B. Liu, S.-F. Zhu, W. Zhang, C. Chen, Q.-L. Zhou,
J. Am. Chem. Soc. 129, 5834–5835 (2007).
29. A. G. Doyle, E. N. Jacobsen, Chem. Rev. 107, 5713–5743
(2007).
science.sciencemag.org/content/366/6468/990/suppl/DC1
Materials and Methods
Supplementary Text
Figs. S1 to S25
Tables S1 to S9
Spectral Data
Calculation Data
References (36–69)
1
1
1. Z. Hou et al., Angew. Chem. Int. Ed. 49, 4763–4766 (2010).
2. S.-F. Zhu, B. Xu, G.-P. Wang, Q.-L. Zhou, J. Am. Chem. Soc.
30. M. Neitzel, J. Marugg, WO patent 2005090296 (2005).
31. J. R. Medina et al., J. Med. Chem. 54, 1871–1895 (2011).
32. K. Thede, W. J. Scott, E. Bender, S. Golz, A. Haegebarth,
P. Lienau, F. Puehler, D. Basting, D. Schneider, M. Moewes,
WO patent 2014147182 (2014).
33. I. Terstiege et al., Bioorg. Med. Chem. Lett. 27, 679–687
(2017).
34. M. Frigerio, M. G. Hummersone, K. A. Menear, M. R. V. Finlay,
E. J. Griffen, L. L. Ruston, J. J. Morris, A. K. T. Ting,
1
34, 436–442 (2012).
3. X. Xu, P. Y. Zavalij, M. P. Doyle, Angew. Chem. Int. Ed. 51,
829–9833 (2012).
4. Y. Zhu et al., Angew. Chem. Int. Ed. 53, 1636–1640 (2014).
1
9
1
1
5. V. Arredondo, S. C. Hiew, E. S. Gutman,
I. D. U. A. Premachandra, D. L. Van Vranken, Angew. Chem.
Int. Ed. 56, 4156–4159 (2017).
20 February 2019; accepted 17 September 2019
10.1126/science.aaw9939
Li et al., Science 366, 990–994 (2019)
22 November 2019
5 of 5

Highly enantioselective carbene insertion into N−H bonds of aliphatic amines
Mao-Lin Li, Jin-Han Yu, Yi-Hao Li, Shou-Fei Zhu and Qi-Lin Zhou
Science 366 (6468), 990-994.
DOI: 10.1126/science.aaw9939
A tag team approach to forming C−N bonds
Many pharmaceutical compounds contain carbon-nitrogen (C−N) bonds in just one of two mirror-image
orientations. Forging these bonds with electron-rich nitrogen reactants is challenging because the nitrogen groups can
coordinate with, and thus interfere with, the catalyst. Li et al. report a cooperative approach to overcoming this obstacle
(
see the Perspective by Ovian and Jacobsen). They used a copper catalyst to activate the carbon reactant and then a
hydrogen-bonding thiourea catalyst to set the product geometry with high selectivity. The reaction is compatible with a
broad range of diazo ester and amine coupling partners.
Science, this issue p. 990; see also p. 948
ARTICLE TOOLS
http://science.sciencemag.org/content/366/6468/990
SUPPLEMENTARY
MATERIALS
http://science.sciencemag.org/content/suppl/2019/11/20/366.6468.990.DC1
REFERENCES
PERMISSIONS
This article cites 65 articles, 3 of which you can access for free
http://science.sciencemag.org/content/366/6468/990#BIBL
http://www.sciencemag.org/help/reprints-and-permissions
Use of this article is subject to the Terms of Service
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement of
Science, 1200 New York Avenue NW, Washington, DC 20005. The title Science is a registered trademark of AAAS.
Copyright © 2019 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of
Science. No claim to original U.S. Government Works