Catalytic N-Acylation of Cyclic Amines by Arylglyoxylic Acids via Radical-Radical Cross-Coupling

Article abstract of DOI:10.1002/ejoc.202100381

A methodical mechanistic investigation allowed for the catalytic N-acylation of secondary cyclic amine counterparts by arylglyoxylic acids through radical-radical coupling. The reaction proceeds via a twofold SET-promoted Cu(I)/Cu(II) catalytic cycle under mild conditions. An analogous reaction variant allows for the N-acylation in a one-pot fashion directly starting from a secondary cyclic amine even in the presence of a second amine or hydroxy group.

Full text of DOI:10.1002/ejoc.202100381

Communications
doi.org/10.1002/ejoc.202100381
Catalytic N-Acylation of Cyclic Amines by Arylglyoxylic
Acids via Radical-Radical Cross-Coupling
Ajijur Rahaman,[a] Anupam Kumar Singh,[a] Aniket Gupta,[a] and Sukalyan Bhadra*[a]
with α-oxo acids was realized via the formation of a stable
aniline radical cation with high loading of Ag-mediator or by
photoredox catalysis using oxygen as the electron mediator/
acceptor (Scheme 1).[8] Thus, to expedite the acyl–heteroatom
bond formation, the use of an external electron acceptor
(oxidant) appeared mandatory to complete the oxidative
catalytic cycle.
In the present study, we have demonstrated through
various mechanistic investigations that a copper(II) catalyst
triggered a single electron transfer (SET) process of arylglyoxylic
acids, which further promoted the formation of an N-based
radical species from secondary cyclic amines, essentially com-
pleting the Cu(I)/Cu(II)-based catalytic cycle under mild con-
dition.
A methodical mechanistic investigation allowed for the catalytic
N-acylation of secondary cyclic amine counterparts by aryl-
glyoxylic acids through radical-radical coupling. The reaction
proceeds via a twofold SET-promoted Cu(I)/Cu(II) catalytic cycle
under mild conditions. An analogous reaction variant allows for
the N-acylation in a one-pot fashion directly starting from a
secondary cyclic amine even in the presence of a second amine
or hydroxy group.
Catalytic production of two dissimilar free radicals and the
subsequent radical-radical cross-coupling arguably represents
an attractive method to construct a new CÀ C and/or C-
heteroatom bond and has recently emerged as a perceptible
approach.[1] Whereas the activation energy of radical-radical
coupling reactions is virtually zero, the selectivity of the cross-
coupling of two unlike radicals is hindered by the unavoidable
homocoupling events of either of the two radicals. Nonetheless,
the desired selectivity of a radical-radical cross-coupling reac-
tion is often achieved by engaging a persistent radical and a
transient radical species.[2] We reasoned to apply this strategy
for building amide bonds, given the abundance of those
linkages in numerous natural and synthetic compounds. We
envisioned that the amidation of a specific class of amine, even
in the presence of another comparably reactive amine group,
might be accomplished by this approach through carefully
generating the appropriate amine radical. In this paper, we are
pleased to disclose a copper-catalyzed N-acylation of secondary
cyclic amines by α-oxo acids via radical-radical cross-coupling
reaction (Scheme 1).
The pioneering discovery of redox-neutral decarboxylative
acylation by Goossen et al. in 2008 has opened up new avenues
to the synthetic community for utilizing α-oxo acids as an acyl
surrogate primarily for C-based coupling partners.[3–5] However,
the decarboxylative acylation of heteroatom systems is limited
to only a handful of protocols, largely involving acyl radical
intermediates formed from α-oxo acids.[6] For instance, the
decarboxylative CÀ S coupling of α-oxo acids with disulfides or
thiols to give thioesters proceeded in the presence of a strong
oxidant including peroxysulfates.[7] The acylation of anilines
Based on the literature precedence, we anticipated that
arylglyoxylic acids might act as a single electron donor to
reduce the copper(II) catalyst to copper (I), resulting in
arylglyoxylate radicals.[3a,9] To validate this postulate, when a 1:1
mixture of phenylglyoxylic acid (1a) and Cu(OAc)2 in DMF was
°
°
heated at 100 C and monitored by EPR at À 160 C, the intensity
of the X-band EPR signal at gavg =2.95 (gk =2.35, g =2.07) of
⊥
Cu(II) decreased steadily with time (Figure 1 and Figure S1,
Supporting Information (SI)). A gradual decrease in the peak
intensity of Cu(II) was also observed when the reaction was
monitored by UV-Vis spectroscopy (see SI, Figures S2 and S3).[10]
These findings clearly indicate the reduction of Cu(II) to Cu(I) via
SET from 1a. Notably, when the reaction was conducted in the
presence of
a radical scavenger, 2,6-di-tert-butyl-4-meth-
ylphenol (BHT) (2.0 equiv.), the formation of BHT-(1a) adduct
was identified by GC-MS analysis suggesting the involvement of
glyoxylate radicals (see SI, Figure S4).[11] We further hypothe-
sized that the resultant glyoxylate radical, via decarboxylation,
would initiate the required N-acylation process in the presence
of an N-based radical (Scheme 2).[6b,8b] The copper-mediated
[a] A. Rahaman, A. Kumar Singh, A. Gupta, Dr. S. Bhadra
Inorganic Materials and Catalysis Division
CSIR-Central Salt and Marine Chemicals Research Institute
G.B. Marg, Bhavnagar 364002, Gujarat, India
E-mail: sbhadra@csmcri.res.in
https://sukalyanbhadracsmcri.wixsite.com/home
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/ejoc.202100381
Scheme 1. N-Acylation approaches of amines by α-oxo acids.
Eur. J. Org. Chem. 2021, 2198–2202
2198
© 2021 Wiley-VCH GmbH

Communications
doi.org/10.1002/ejoc.202100381
Table 1. Identification of the amine equivalent and optimization of
reaction conditions.[a]
Entry
Z
Deviation from the above condition
5a
[%]
1
2
3
4
5
6
7
8
2a, H none
3, Cl none
4,OBz none
4,OBz no Cu(OAc)2
4,OBz Cu(OAc)2 (1 equiv.)
4,OBz AgNO3 (1 equiv.) instead of Cu(OAc)2
4,OBz Cu(OTf)2 instead of Cu(OAc)2
4,OBz CuI instead of Cu(OAc)2
4,OBz DMAc instead of DMF at RT for 16 h
4,OBz K2CO3 (2.0 equiv.), DMAc instead of DMF
4,OBz Et3N (2.0 equiv.), DMAc instead of DMF
4,OBz Et3N (2.0 equiv.), DMAc instead of DMF at 100 oC
for 16 h
00
00
50
00
50
00
15
24
68
18
69
87
9
10
11
12
Figure 1. X-Band EPR spectra for the conversion of Cu(II) to Cu(I) by
°
phenylglyoxylic acid-mediated SET process recorded at À 160 C.
[a] Reaction conditions: 1a (0.5 mmol), 2–4 (0.5 mmol), Cu(OAc)2 (10 mol%),
°
DMF (3 mL), N2 atm., 145 C, 8 h. DMF=N,N’-Dimethylformamide; DMAc=
N,N’-Dimethylacetamide
(5a–v). A significant feature of the N-acylation is that it allows
transformation of N-OBz bonds in a highly regiospecific fashion
keeping other sensitive substituents and common N-protecting
groups intact (5g, 5l–q). The acylation condition was also
tolerant to a range of functional groups present in the α-oxo
acid fragment (5r–v). No self-coupling product via decarbox-
ylation of α-oxo acids was formed across the CÀ halogen bond
(5r–s, 5v). Notably, the N-acylation was easily scalable, as
exemplified by the multigram synthesis of 5a in comparable
yields (Scheme 3). However, non-cyclic secondary amines,
primary amines, and alkylglyoxylic acids remained unreactive
under the reaction conditions.
Scheme 2. Mechanistic hypothesis of Cu(II)-catalyzed N-acylation by aryl-
glyoxylic acids.
process could be turned into catalytic via oxidation of the
resultant Cu(I)-species into the Cu(II) active catalyst either in the
presence of an oxidant or by means of a second SET process.
To ascertain a suitable amine counterpart, we studied the
Cu(II) catalyzed N-acylation of various amine equivalents under
inert conditions (Table 1 and Table S1, SI). Soon it turned out
that free amine was not an operational substrate (entry 1).
Gratifyingly, the desired N-acylated compound 5a was formed
when an amine equivalent 4a bearing an electrophilic nitrogen
atom was employed among others (entries 2–3). However, the
reaction did not proceed without copper or with the use of
silver salt (entries 4–6). The use of other Cu(II) and Cu(I) salts
also gave inferior results (entries 7–8). Pleasingly, replacing DMF
with DMAc provided an improved yield of 5a even at room
temperature (entry 9). Further enhancement of yield was
observed when the reaction was conducted in the presence of
To gain insight into the N-acylation process, we next
conducted a sequence of experiments. As expected, employing
a radical quencher, such as 2,2,6,6-tetramethylpiperidine-N-oxyl
(TEMPO) or 2,6-di-tert-butyl-4-methylphenol (BHT) (2.0 equiv.
each), completely shut down the reaction. Both BHT-(1a) and
BHT-(piperidine) adducts were identified by GC-MS and HRMS,
respectively, in the copper-catalyzed N-acylation of 4a with 1a
in the presence of BHT, signifying the formation of arylglyoxyl-
and piperidinyl-radicals (see SI, Figures S5 and S6).
When a mixture of 1a and 4e was added to the solution of
Cu(I)OAc in DMAc, the peak of Cu(II) at gavg =3.01 (gk =2.45,
g =2.06) in X-band EPR spectra appeared almost instantly and
⊥
the signal for Cu(II) remained nearly unchanged with time,
confirming the participation of a rapidly transformable Cu(I)/
Cu(II) catalytic cycle (Figure 2 and Figure S7, SI).[12] While the
initial Cu(II) to Cu(I) transition might be due to SET from 1a, the
transition of Cu(I) to Cu(II) to complete the catalytic cycle is
conceivably attributed to a second SET, promoted by the NÀ O
bond cleavage of 4e.[13]
°
an appropriate base such as Et3N and at 100 C (entries 10–12).
After recognizing the practical reaction system, we contin-
ued to investigate the scope of the N-acylation reaction
(Scheme 3). Numerous cyclic amines with varied ring sizes were
acylated with arylglyoxylic acids in good to excellent yields
Our proposed mechanism, as depicted in Scheme 4, is in
accord with these experimental findings and literature
Eur. J. Org. Chem. 2021, 2198–2202
www.eurjoc.org
2199
© 2021 Wiley-VCH GmbH

Communications
doi.org/10.1002/ejoc.202100381
Figure 2. X-Band EPR spectra of Cu(II) formed from Cu(I) during the N-
°
acylation process recorded at À 160 C.
Scheme 4. Plausible mechanism.
ylic acids were added to the freshly prepared crude O-benzoyl
hydroxylamines into the same reaction vessel and the mixture
°
was heated at 100 C to obtain the desired N-acylated
compound in practical yields (Scheme 5). The N-acylation
approach was applied for late-stage amidation of complex
molecules containing both secondary cyclic amine sites and an
additional cross-coupling sensitive functional group including
aromatic amine, non-cyclic secondary amines, phenolic-OH etc.
(5w–y). Gratifyingly, the amidation took place preferentially at
the secondary cyclic amine site, possibly due to the optimal
reactivity and appropriate stabilization energies of the corre-
sponding N-centered radicals formed via the facile cleavage of
NÀ O bonds.[14]
Scheme 3. Scope of the decarboxylative N-acylation[a] [a] Reaction condi-
tions: 1 (0.5 mmol), 4 (0.5 mmol), Cu(OAc)2 (10 mol%), DMAc (3 mL), 16 h.
°
Condition A: room temp. Condition B: Et3N (1 mmol), 100 C. [b] 1a (1.05 g,
7 mmol), 4a (1.44 g, 7 mmol) using optimized conditions of A and B (see SI).
[c] CCDC no. for compounds 5m, 5n and 5r are 2033263, 2033255 and
2033256 respectively (see SI).
reports.[3a,7b,13] In a catalytic cycle, the Cu(II) catalyst is initially
reduced to Cu(I) species via a SET from 1. In the presence of 4,
the resulting Cu(I) species forms Cu(II) catalyst, thereby
generating amine radicals that undergo decarboxylative N-
acylation with arylglyoxylate radicals via radical-radical cross-
coupling to produce 5. The role of a base, such as Et3N, is to
accelerate the deprotonation of 1 thereby increasing the
reaction rate.
Further, when a mixture of a secondary cyclic amine and a
non-cyclic amine (or aniline) was subjected to the N-benzoyla-
tion with phenylglyoxylic acid (1 equivalent) under our
optimized condition as in Table 2, the secondary cyclic amine
was benzoylated exclusively. This finding might be found
beneficial in the future to benzoylate specifically the secondary
cyclic amine in an amine mixture.[15]
Subsequently, we studied an analogous N-acylation reaction
that is operational in a one-pot manner directly starting from
an amine and arylglyoxylic acid. Thus, Cu(OAc)2 and arylglyox-
In conclusion, we have disclosed a copper-catalyzed N-
acylation of O-benzoyl hydroxylamines as the secondary cyclic
Eur. J. Org. Chem. 2021, 2198–2202
www.eurjoc.org
2200
© 2021 Wiley-VCH GmbH

Communications
doi.org/10.1002/ejoc.202100381
arylglyoxylic acid. As predicted, the one-pot N-acylation took
place favorably at the secondary cyclic amine site over other
amine/hydroxy groups or in a mixture of two different amines.
We envisage that this unique mode of action of copper catalysis
will find numerous applications in radical chemistry in the
future.
Deposition Numbers 2033263 (for 5 m), 2033255 (for 5n)
and 2033256 (for 5r) contain the supplementary crystallo-
graphic data for this paper. These data are provided free of
charge by the joint Cambridge Crystallographic Data Centre
and Fachinformationszentrum Karlsruhe Access Structures serv-
ice www.ccdc.cam.ac.uk/structures.
Acknowledgements
We sincerely thank DST (grant no. DST/INSPIRE/04/2015/002248),
CSIR (CSMCRI project no. MLP 0028, and a fellowship to AR), Dr.
Eringathodi Suresh of CSIR-CSMCRI for X-ray crystallographic
analysis, Dr. Pabitra B. Chatterjee of CSIR-CSMCRI for helpful
discussion on EPR analysis and “Analytical and Environmental
Science Division and Centralized Instrument Facility” of CSIR-
CSMCRI for providing instrument facilities. CSIR-CSMCRI communi-
cation no. 193/2020.
Scheme 5. One-pot N-acylation of cyclic amines[a] [a] Reaction conditions: 2
(0.5 mmol), (BzO)2 (0.6 mmol), K2HPO4 (0.5 mmol), DMAc (3 mL) at rt for 10 h;
subsequent acylation with ArCOCO2H (0.5 mmol) and Cu(OAc)2 (10 mol%) at
°
100 C for 16 h. [b] The acylation step was conducted at rt. [c] 1 mmol
K2HPO4. [d] The acylation step was run for 8 h.
Conflict of Interest
Table 2. Comparison of the one-pot N-benzoylation method with the
conventional method.
The authors declare no conflict of interest.
Entry
Current N-benzoylation[a]
Conventional N-benzoylation[b]
Keywords: Acylation · Amides · Copper · Radical reactions ·
Synthetic methods
1
[1] For selected examples on radical-radical coupling, see: a) T. Ishii, K.
Nagao, H. Ohmiya, Chem. Sci. 2020, 11, 5630–5636; b) F. Zhou, L. Li, K.
Lin, F. Zhang, G.-J. Deng, H. Gong, Chem. Eur. J. 2020, 26, 4246–4250;
c) D. Kong, P. J. Moon, R. J. Lundgren, Nature Catalysis 2019, 2, 473–476;
d) Q. Wang, Y. Qu, H. Tian, Y. Liu, H. Song, Q. Wang, Chem. Eur. J. 2019,
25, 8686–8690; e) C. Zheng, F. Lu, H. Lu, J. Xin, Y. Deng, D. Yang, S.
Wang, Z. Huang, M. Gao, A. Lei, Chem. Commun. 2018, 54, 5574–5577;
f) H. Yi, G. Zhang, H. Wang, Z. Huang, J. Wang, A. K. Singh, A. Lei, Chem.
Rev. 2017, 117, 9016–9085; g) Z. Huang, D. Zhang, X. Qi, Z. Yan, M.
Wang, H. Yan, A. Lei, Org. Lett. 2016, 18, 2351–2354; h) A. Studer, D. P.
Curran, Angew. Chem. 2016, 128, 58–106; Angew. Chem. Int. Ed. 2016,
55, 58–102.
2
3
[2] a) H. Fischer, Chem. Rev. 2001, 101, 3581–3610; b) D. Leifert, A. Studer,
Angew. Chem. 2020, 132, 74–110; Angew. Chem. Int. Ed. 2020, 59, 74–
108.
[3] For reviews, see: a) L.-N. Guo, H. Wang, X.-H. Duan, Org. Biomol. Chem.
2016, 14, 7380–7391; b) W. I. Dzik, P. P. Lange, L. J. Goossen, Chem. Sci.
2012, 3, 2671–2678; c) R. Shang, L. Liu, Sci. China Chem. 2011, 54, 1670–
1687; d) N. Rodríguez, L. J. Goossen, Chem. Soc. Rev. 2011, 40, 5030–
5048.
[a] Cyclic amine (0.5 mmol), non-cyclic amine (0.5 mmol), (BzO)2 (0.6 mmol),
K2HPO4 (0.5 mmol), DMAc (3 mL), rt, 10 h; subsequent addition of
°
PhCOCO2H (0.5 mmol) and Cu(OAc)2 (10 mol%), 100 C, 16 h. [b] Cyclic
amine (0.5 mmol), non-cyclic amine (0.5 mmol), PhCOCl (0.5 mmol), Et3N
°
(0.75 mmol), DCM (3 mL), 0 C–rt, 16 h. [c] 88% Bn2NH recovery. [d] 94%
PhNH2 recovery. [e] 95% PhNH2 recovery.
[4] a) L. J. Goossen, F. Rudolphi, C. Oppel, N. Rodríguez, Angew. Chem.
2008, 120, 3085–3088; Angew. Chem. Int. Ed. 2008, 47, 3043–3045;
b) L. J. Goossen, B. Zimmermann, T. Knauber, Angew. Chem. 2008, 120,
7211–7214; Angew. Chem. Int. Ed. 2008, 47, 7103–7106.
[5] For selected examples, see: a) P. Das, P. Biswas, J. Guin, Chem. Asian J.
2020, 15, 920–925; b) G.-Z. Wang, R. Shang, W.-M. Cheng, Y. Fu, Org.
Lett. 2015, 17, 4830–4833; c) Z. He, X. Qi, S. Li, Y. Zhao, G. Gao, Y. Lan, Y.
Wu, J. Lan, J. You, Angew. Chem. 2015, 127, 869–873; Angew. Chem. Int.
Ed. 2015, 54, 855–859; d) L. Yu, P. Li, L. Wang, Chem. Commun. 2013, 49,
2368–2370; e) J. Park, M. Kim, S. Sharma, E. Park, A. Kim, S. H. Lee, J. H.
amine equivalents by arylglyoxylic acids via radical-radical
cross-coupling strategy. As evidenced by EPR, UV-Vis, and mass
spectrometric analysis, the required CÀ N bond formation
proceeds via a Cu(I)/Cu(II) based catalytic cycle involving a
twofold SET process. The catalytic N-acylation is also opera-
tional in a one-pot fashion starting from a free amine and
Eur. J. Org. Chem. 2021, 2198–2202
www.eurjoc.org
2201
© 2021 Wiley-VCH GmbH

Communications
doi.org/10.1002/ejoc.202100381
Kwak, Y. H. Jung, I. S. Kim, Chem. Commun. 2013, 49, 1654–1656; f) Z.-Y.
[11] As detected by GC-MS, the partial formation of benzaldehyde is
possibly due to incomplete decarboxylation of PhCOCO2H under Cu-
catalysis (see SI).
[12] The conversion of Cu(I) to Cu(II) took place almost immediately, as
monitored by EPR analysis. For the X-band EPR spectra, see Supporting
Information, Figure S7.
[13] The cleavage of NÀ OBz bond may lead to transform Cu(I) to Cu(II)
through SET mechanism. For details see: a) P. Wang, B. Zhao, Y. Yuan, Z.
Shi, Chem. Commun. 2019, 55, 1971–1974; b) Z. Wang, M.-Y. Han, P. Li,
L. Wang, Eur. J. Org. Chem. 2018, 5954–5960.
Li, D.-D. Li, G.-W. Wang, J. Org. Chem. 2013, 78, 10414–10420; g) H.
Wang, L.-N. Guo, X.-H. Duan, Adv. Synth. Catal. 2013, 355, 2222–2226;
h) H. Wang, L.-N. Guo, X.-H. Duan, Org. Lett. 2012, 14, 4358–4361; i) P.
Fang, M. Li, H. Ge, J. Am. Chem. Soc. 2010, 132, 11898–11899.
[6] For important early works on radical decarboxylation of carboxylic
acids, see: a) J. M. Anderson, J. K. Kochi, J. Am. Chem. Soc. 1970, 92,
1651–1659; b) J. M. Anderson, J. K. Kochi, J. Org. Chem. 1970, 35, 986–
989; c) F. Fontana, F. Minisci, M. C. N. Barbosa, E. Vismara, J. Org. Chem.
1991, 56, 2866–2869.
[7] a) K. Yan, D. Yang, W. Wei, J. Zhao, Y. Shuai, L. Tian, H. Wang, Org.
Biomol. Chem. 2015, 13, 7323–7330; b) G. Rong, J. Mao, D. Liu, H. Yan, Y.
Zheng, J. Chen, RSC Adv. 2015, 5, 26461–26464.
[8] a) W.-T. Xu, B. Huang, J.-J. Dai, J. Xu, H.-J. Xu, Org. Lett. 2016, 18, 3114–
3117; b) X.-L. Xu, W.-T. Xu, J.-W. Wu, J.-B. He, H.-J. Xu, Org. Biomol. Chem.
2016, 14, 9970–9973; c) J. Liu, Q. Liu, H. Yi, C. Qin, R. Bai, X. Qi, Y. Lan, A.
Lei, Angew. Chem. 2014, 126, 512–516; Angew. Chem. Int. Ed. 2014, 53,
502–506.
[14] J. Hioe, D. Šakić, V. Vrček, H. Zipse, Org. Biomol. Chem. 2015, 13, 157–
169.
[15] P. G. M. Wuts, T. W. Greene, Protection for the Amino Group. In Greene’s
Protective Groups in Organic Synthesis, 4th Edn. (Wiley, USA, 2006, pp.
696–926).
[9] a) S. Senaweera, K. C. Cartwright, J. A. Tunge, J. Org. Chem. 2019, 84,
12553–12561; b) A. Gupta, A. Rahaman, S. Bhadra, Org. Lett. 2019, 21,
6164–6168; c) Y. Wu, Z. Huang, Y. Luo, D. Liu, Y. Deng, H. Yi, J.-F. Lee, C.-
W. Pao, J.-L. Chen, A. Lei, Org. Lett. 2017, 19, 2330–2333.
[10] A gradual decrease in slope was observed in the room temperature
time-dependent OD measurement of 1a and Cu(II) mixture in DMAc at
Manuscript received: March 27, 2021
Revised manuscript received: April 1, 2021
Accepted manuscript online: April 2, 2021
λmax =770 nm. For details, see Supporting Information, Figure S3.
Eur. J. Org. Chem. 2021, 2198–2202
www.eurjoc.org
2202
© 2021 Wiley-VCH GmbH