Pd-Catalyzed C(sp2)?H Alkoxycarbonylation of Phenethyl- and Benzylamines with Chloroformates as CO Surrogates

Article abstract of DOI:10.1002/chem.202005338

The site-selective functionalization of C?H bonds within a complex molecule remains a challenging task of capital synthetic importance. Herein, an unprecedented Pd-catalyzed C(sp²)?H alkoxycarbonylation of phenylalanine derivatives and other amines featuring picolinamide as the directing group (DG) is reported. This oxidative coupling is distinguished by its scalability, operational simplicity, and avoids the use of toxic carbon monoxide as the C₁ source. Remarkably, the easy cleavage of the DG enables the efficient assembly of isoindolinone compounds. Density Functional Theory calculations support a Pd^II/Pd^IV catalytic cycle.

Full text of DOI:10.1002/chem.202005338

Full Paper
doi.org/10.1002/chem.202005338
Chemistry—A European Journal
&
CÀH Functionalization
Pd-Catalyzed C(sp2)ÀH Alkoxycarbonylation of Phenethyl- and
Benzylamines with Chloroformates as CO Surrogates
Paula Andrade-Sampedro,[a, c] Jon M. Matxain,[b, c] and Arkaitz Correa*[a]
Abstract: The site-selective functionalization of CÀH bonds
within a complex molecule remains a challenging task of
capital synthetic importance. Herein, an unprecedented Pd-
catalyzed C(sp2)ÀH alkoxycarbonylation of phenylalanine de-
rivatives and other amines featuring picolinamide as the di-
recting group (DG) is reported. This oxidative coupling is dis-
tinguished by its scalability, operational simplicity, and
avoids the use of toxic carbon monoxide as the C1 source.
Remarkably, the easy cleavage of the DG enables the effi-
cient assembly of isoindolinone compounds. Density Func-
tional Theory calculations support a PdII/PdIV catalytic cycle.
Introduction
Carbonylation reactions are of paramount importance in both
academic and industrial environments and represent a crucial
technology for the production of bulk and fine chemicals
worldwide.[1] One of the most relevant carbonylation processes
is the palladium-catalyzed alkoxycarbonylation of aryl halides
featuring the combination of CO (gas) as the C1 source along
with an alcohol for the introduction of the ester unit
(Scheme 1, route a).[1] Despite its high abundance and low
price, the use of carbon monoxide poses severe downsides
such as high risk in handling and storage as well as toxicity
and flammability, among others. As a result, a myriad of CO
surrogates have been investigated in the last several years to
perform carbonylation reactions in a safer and sustainable
fashion.
Scheme 1. Pd-catalyzed alkoxycarbonylations.
streamlined and atom-efficient approach to chemical synthe-
sis.[3] This tactic has allowed for the challenging site-selective
appendance of a variety of CO surrogates at the ortho position
of non-functionalized arenes such as carbon dioxide,[4] a-keto
esters,[5] azodicarboxylates,[6] DMF,[7] chloroform,[8] or alkyl chlor-
oformates,[9] among others (Scheme 1, route b). Although the
latter methods have clearly expanded the toolkit of available
carbonylations, the selective CÀH alkoxycarbonylation[10] occur-
ring at C sites remotely positioned from a given DG still re-
mains an unmet challenge of prime synthetic significance.
The recent years have witnessed a tremendous interest in
the chemical modification of amino acids and peptides derived
thereof.[11] In this regard, transition-metal catalysis has un-
locked new paradigms for the site-selective labeling of a vast
array of amino acids and fueled the development of innovative
bond disconnections upon CÀH functionalization processes.[12]
In 2016, the group of Shi designed an efficient Pd-catalyzed
C(sp3)ÀH alkoxycarbonylation with alkyl chloroformates as the
practical C1 source for the modification of a broad range of ali-
phatic carboxamides bearing 8-aminoquinoline (AQ) as the
DG.[13a] Remarkably, a wide variety of phenylalanine (Phe) resi-
dues housing the DG at the C-terminal position smoothly un-
derwent the selective alkoxycarbonylation at the benzylic site
Compared with classical syntheses occurring in pre-function-
alized substrates, CÀH functionalization has changed the land-
scape of modern chemistry by enabling the direct conversion
of traditionally unreactive hydrocarbon moieties into valuable
functionalized compounds.[2] In particular, the chelation assis-
tance approach based on the installation of a Lewis basic
motif commonly named directing group (DG) offers a more
[a] P. Andrade-Sampedro, Dr. A. Correa
Department of Organic Chemistry I, Joxe Mari Korta R&D Center
University of the Basque Country (UPV/EHU)
Avda. Tolosa 72, 20018 Donostia-San Sebastiµn (Spain)
E-mail: arkaitz.correa@ehu.eus
[b] Dr. J. M. Matxain
Polimero eta Material Aurreratuak: Fisika, Kimika eta Teknologia Saila,
Kimika Fakultatea, Euskal Herriko Unibertsitatea (UPV/EHU)
Paseo Manuel de Lardizabal 3, 20018 Donostia-San Sebastiµn (Spain)
[c] P. Andrade-Sampedro, Dr. J. M. Matxain
Donostia International Physics Center (DIPC)
Paseo Manuel de Lardizabal 4, 20018 Donostia-San Sebastiµn (Spain)
Supporting information and the ORCID identification number(s) for the au-
thor(s) of this article can be found under:
https://doi.org/10.1002/chem.202005338.
through the formation of
a
five-membered palladacycle
Chem. Eur. J. 2021, 27, 5782 –5789
5782
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005338
Chemistry—A European Journal
Results and Discussion
Since the seminal work by Daugulis on the use of picolinamide
(PA) as a removable, efficient DG,[16] it has demonstrated supe-
rior directing abilities to assist a variety of transformations in
the realm of CÀH activation.[17] Encouraged by these results,
we began our studies by selecting the alkoxycarbonylation of
PA-Phe-OMe (1a) with commercially available ethyl chlorofor-
mate (2a) as the model reaction. Whereas the formation of the
indoline derivative through an intramolecular d-amination was
never detected,[14] initial exploratory screening preferentially af-
forded the undesired N-functionalized product. Control experi-
ments in the absence of Pd(OAc)2 ruled out the formation of
the latter compound through a classical base-assisted substitu-
tion reaction and supported a CÀN bond-forming reductive
elimination of the putative PdIV species (see below). However,
careful screening of all the reaction parameters revealed that
the latter reaction pathway could be minimized and achieved
the desired d-alkoxycarbonylation instead.[18]
Scheme 2. Alkoxycarbonylation of Phe derivatives.
(Scheme 2, top). More recently, they have achieved the assem-
bly of a number of phthalic acid derivatives through a Pd-cata-
lyzed AQ-directed C(sp2)ÀH alkoxycarbonylation process.[13b]
Inspired by these excellent results, we envisioned a comple-
mentary amino acid tagging technique featuring the installa-
tion of picolinamide (PA) as bidentate auxiliary at the N-termi-
nus of the Phe residue, thus enabling the remote d-functionali-
zation upon the intermediacy of a challenging six-membered
palladacycle (Scheme 2, bottom). Assuming an analogous
mechanism to that proposed by Shi involving a PdII/PdIV
regime,[13] we anticipated that the judicious choice of the reac-
tion parameters would be crucial for achieving high positional
selectivity. In fact, careful analysis of the existing literature
clearly verified that site-selectivity issues may hamper the tar-
geted d-functionalization as the transient PdIV intermediate
could undergo competitive reductive elimination processes to
deliver either the N-functionalized product or the correspond-
ing indoline compound upon an intramolecular CÀH amination
reaction.[14] To the best of our knowledge, the C(sp2)ÀH alkoxy-
carbonylation of b-arylethylamines remains unexplored and, if
successful, we could unlock its full synthetic potential toward
the diversification of other arylamines beyond phenylalanine
derivatives. As part of our interest in CÀH functionalization,[15]
herein we disclose a Pd-catalyzed site-selective C(sp2)ÀH alkox-
ycarbonylation of picolinamide-containing phenethyl and
benzyl amines with chloroformates. The salient features of our
method include the broad group tolerance, scalability, reten-
tion of the native chirality, and facile removal of the required
DG, thus streamlining the assembly of biologically relevant iso-
indolinone framework in the absence of carbon monoxide.
Likewise, Density Functional Theory (DFT) studies unraveled a
PdII/PdIV catalytic manifold and rationalized the common use of
tert-amyl alcohol as a non-innocent solvent in CÀH functionali-
zation reactions.
After considerable experimentation, we found that the com-
bination of Pd(OAc)2 (10 mol%), Na2CO3, Ag2CO3, LiI as additive
in a mixture of t-amyl alcohol and PhCl at 1258C under air pro-
vided the best results, giving rise to 3a in 75% yield as a mix-
ture of mono- and di-alkoxycarbonylated products (1:3 ratio;
Table 1, entry 1). Control experiments proved instructive in un-
derstanding the requirements of the process: whereas the Pd
catalyst and silver carbonate had a crucial role as not even
traces of 3a were detected in their absence (entries 2 and 3,
respectively), the addition of Na2CO3 and LiI was found to be
beneficial and resulted in higher yields of 3a (entries 4 and 5,
respectively). Other iodide sources afforded 3a in lower yields
(entries 6 and 7). Likewise, the use of a mixture of PhCl and
Table 1. Pd-catalyzed d-C(sp2)ÀH alkoxycarbonylation of PA-Phe-OMe
with ethyl chloroformate.[a]
Entry
Change from standard conditions
3a [%][b]
1
2
3
4
5
6
7
8
none
75 (1:3)[c]
0
without Pd(OAc)2
without Ag2CO3
without Na2CO3
without LiI
TBAI instead of LiI
KI instead of LiI
Na2CO3 (1.0 equiv)
Ag2CO3 (1.0 equiv)
PhCl as solvent
t-amylOH as solvent
0
42 (3:1)[c]
30 (1:3)[c]
0
51 (1:3)[c]
64 (1:3)[c]
58 (1:3)[c]
54 (3:1)[c]
46 (3:1)[c]
9
10
11
[a] Reaction conditions: 1a (0.25 mmol), 2a (0.75 mmol), Pd(OAc)2
(10 mol%), Ag2CO3 (2.0 equiv), Na2CO3 (2.0 equiv), LiI (50 mol%) in a mix-
ture of t-amylOH/PhCl (1:1; 2 mL) at 1258C for 16 h under air. [b] Yield of
isolated product after column chromatography. [c] Ratio of mono- and di-
functionalized product 3a/3a’.
Chem. Eur. J. 2021, 27, 5782 –5789
www.chemeurj.org
5783
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005338
Chemistry—A European Journal
tert-amyl alcohol led to the best results (entries 10 and 11). To
overcome the persistent problem of regioselectivity between
the mono- and di-functionalization reaction, the evaluation of
supporting ligands, equivalents of 2a and other parameters
were carefully analyzed. Unfortunately, higher selectivity
toward the mono-alkoxycarbonylated product 3a was only
achieved at the expense of having much lower overall yields.
Importantly, different DGs were evaluated under the optimized
conditions and PA showed a superior coordinating ability as a
bidentate DG (Scheme 3). In this regard, benzoyl- and tosyl-
protected Phe derivatives devoid of an additional nitrogen-
chelating atom remained unreactive as well as the parent de-
rivative bearing a 3-pyridine unit, thereby supporting the bi-
dentate nature of PA. Likewise, a related carboxamide housing
a 1,2,3-triazole unit could be also employed as an efficient bi-
dentate DG, albeit with lower efficiency.
Scheme 3. Influence of the DG.
went the target alkoxycarbonylation with a wide variety of
electronically diverse chloroformates. Not only simple alkyl
We next investigated the preparative scope of the d-C(sp2)À chloroformates such as ethyl (2a), benzyl (2d), hexadecyl (2 f),
H alkoxycarbonylation protocol to assemble a new family of
decorated Phe compounds in a simple fashion (Table 2). Grati-
fyingly, the model substrate PA-Phe-OMe (1a) smoothly under-
and 9-fluorenylmethyl (2h) derivatives but also structurally
complex menthyl chloroformate (2g) furnished the corre-
sponding products 3 as variable mixtures of mono- and di-al-
Table 2. Pd-catalyzed d-C(sp2)ÀH alkoxycarbonylation of phenylalanine derivatives.[a,b]
[a] As for Table 1, entry 1. [b] Yield of isolated product or product mixture after column chromatography, average of at least two independent runs.
[c] Ratio of mono- and diacylated product (3/3’). [d] Gram-scale experiment. [e] t-amylOH/PhCl (1:1; 4 mL).
Chem. Eur. J. 2021, 27, 5782 –5789
www.chemeurj.org
5784
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005338
Chemistry—A European Journal
koxycarbonylated compounds, which were separated by
column chromatography. Notably, the process was tolerant of
chloroformates bearing methoxy and chloro groups within the
alkyl chain, thereby affording the corresponding Phe com-
pounds 3b and 3c, respectively, in high yields. Whereas benzyl
chloroformate preferentially delivered the mono-carbonylated
product 3d in 49% yield, hexadecyl and 9-fluorenylmethyl de-
rivatives resulted in the exclusive formation of difunctionalized
Phe compounds 3 f’ and 3g’ in 56 and 71% yield, respectively.
Importantly, aryl chloroformates such as 2e could be also
used, albeit with lower efficiency. The use of unnatural Phe de-
rivatives accommodating different substitution patterns within
the aromatic ring led to the exclusive formation of the mono-
alkoxycarbonylated compounds. In this regard, Phe residues
bearing ortho or meta substituents, which blocked the difunc-
tionalization process, resulted in 3i–k in good yields. Converse-
ly, para-substituted Phe residues resulted in the preferential
(3l) or exclusive formation (3m’) of the difunctionalized prod-
uct. Importantly, HPLC analysis of 3 f’ verified that no racemiza-
tion occurred along the oxidative process,[18] and crystallo-
graphic analysis of 3m’ confirmed that the absolute stereo-
chemistry was identical to that of the starting Phe residue. No-
tably, the method boded well with Phe derivatives bearing al-
kylnitriles (3n) as well as biologically relevant cyclic ethers (3o
and 3p). Furthermore, the process could be performed on the
gram-scale with a remarkable 65% yield, thus highlighting the
synthetic utility and robustness of our d-C(sp2)ÀH alkoxycarbo-
nylation manifold.
Table 3. Pd-catalyzed C(sp2)ÀH alkoxycarbonylation of benzylamines and
phenethylamines.[a,b]
Deposition Numbers 2063453 (3m’), 2063455 (5 f), and
2063452 (6b) contain the supplementary crystallographic data
for this paper. These data are provided free of charge by the
joint Cambridge Crystallographic Data Centre and Fachinfor-
mationszentrum Karlsruhe Access Structures service.
[a] Reaction conditions: 4 (0.25 mmol), ClCO2iBu (0.75 mmol), Pd(OAc)2
(10 mol%), Na2CO3 (2.0 equiv), AgOAc (2.0 equiv) in a mixture of t-
amylOH/PhCl (1:3; 4 mL) at 1258C for 16 h under air. [b] Yield of isolated
product after column chromatography, average of at least two independ-
ent runs.
Although the method described by Shi was not applied to a
more challenging peptide framework,[13a] we submitted a varie-
ty of picolinamide-protected Phe dipeptides to the developed
reaction conditions. However, PA-Phe-Gly-OMe, PA-Phe-Phe-
OMe, and PA-Phe-Pro-OMe remained unreactive.[18] In stark
contrast, a wide variety of simple benzyl amines bearing the
PA as DG smoothly underwent the mono-functionalization pro-
cess to furnish compounds 5a–h in moderate to good yields
after slight modification of the reaction conditions (Table 3).
The latter would occur through the formation of a kinetically
more favored five-membered palladacycle. This protocol com-
plements the method by Shi for the assembly of phthalic acid
derivatives from benzamides bearing AQ as the DG,[13b] thereby
enabling the g-C(sp2)ÀH alkoxycarbonylation of simple benzyl
amines.
compounds through an intramolecular condensation event in
excellent yields (Scheme 4). X-Ray analysis of compound 6b
verified the formation of the N-unprotected heterocyclic core.
This tandem alkoxycarbonylation/deprotection sequence offers
an attractive alternative to the commonly used Pd-catalyzed
carbonylation techniques of alkylamines, which customarily in-
volve the use of toxic and flammable CO (gas).[20]
To our surprise, other simple b-arylethylamines devoid of the
ester group of the corresponding Phe residue resulted in the
corresponding alkoxycarbonylated products 5i–k in low yields
(Table 3). We hypothesized that the ester motif of the Phe unit
could have a key role in the stabilization of the transient inter-
mediates and a Thorp–Ingold effect could not be discarded.
Notably, the removal of the DG could be easily performed
upon treatment with NaOH[19] with simultaneous hydrolysis of
the ester motif, thus delivering the corresponding benzolactam
Scheme 4. Assembly of isoindolinones upon cleavage of the DG.
Chem. Eur. J. 2021, 27, 5782 –5789
www.chemeurj.org
5785
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005338
Chemistry—A European Journal
To understand the reaction pathway as well as some of the
observed experimental evidences, we further performed DFT
studies for the C(sp2)ÀH alkoxycarbonylation of PA-Phe-OMe
(1a) with ClCO2Et (2a). Assuming a similar reaction pathway to
that of related PA-directed Pd-catalyzed CÀH functionalization
processes,[15d,17] we proposed the mechanism depicted in
Scheme 5 entailing monomeric palladium intermediates. Com-
plexation of 1a with Pd(OAc)2 would initially afford PdII com-
plex In-tA,[15d,17c] which would next undergo a directed ortho-
selective cyclometallation to provide the six-membered palla-
dacycle IntB.[21] The latter would next undergo oxidative addi-
tion with ethyl chloroformate to provide the corresponding
PdIV IntC, which would ultimately deliver the mono-carbonylat-
ed product complex PC-1 through a CÀC bond-forming reduc-
tive elimination. Eventually, PC-1 could either release product
3a, thereby recovering the active catalyst, or undergo a
second functionalization event to yield di-alkoxycarbonylated
derivative 3a’. Notice that in this study, to ensure the continui-
ty of the reaction energy profiles, infinitely separated reactants
and products are considered along with reactant complexes,
transition states, intermediates, and products. Hence, with the
aim of avoiding the unphysical overestimation of entropic ef-
fects owing to the lack of inclusion of explicit solvent mole-
cules, all the energetic discussion will be carried out with en-
thalpies.[22]
inhibited when adding t-amylOH to the reaction medium. Ac-
cordingly, DFT calculations were undertaken both in the ab-
sence and presence of t-amylOH molecules in an implicit
manner (Scheme 6, black and blue reaction pathway, respec-
tively).[18] With the energy values in hand, we can conclude
that the reaction pathway when considering t-amylOH is ener-
getically more favored and the latter showed a high stabilizing
effect through coordination of the alcohol to the metal center
and the formation of hydrogen bonds with the ester group of
the Phe residue.
Although the whole reaction pathway is energetically more
favored, the stabilization effect is stronger on the oxidative ad-
dition step. In fact, without considering the effect of t-amylOH,
the oxidative addition step would be endothermic. In contrast,
with the aid of the solvent molecules, this step would become
exothermic. As a result, under these reaction conditions the
C(sp2)ÀH alkoxycarbonylation could be favored over other
competitive side reactions, such as the N-alkoxycarbonylation
reaction and the basic hydrolysis of the ester group. As men-
tioned above, the proposed mechanism would involve three
fundamental steps. The first one would consist of the forma-
tion of Int-A through deprotonation and coordination of sub-
strate 1a to the initial catalyst, leading to a reactant complex
stabilized by À22.5 kcalmolÀ1 with respect to the separated
species. The latter would next undergo a CÀH activation event
to afford Int-B through TS1. Although this path could proceed
through different mechanisms, we have assumed a CMD path-
way wherein the CÀH bond activation was assisted by an auxil-
iary carboxylate/carbonate ion acting as a base, which is often
invoked in the directed ortho-palladation of aromatic sub-
strates.[23] The optimized structure of the transition state (TS1)
reveals an elongation of the CÀH bond from 1.08 to 1.38 Š
Owing to the subtleties of our catalyst system including the
use of t-amylOH as solvent as well as Ag2CO3 and LiI as addi-
tives, we anticipated that they could have a crucial role in fa-
voring the formation of some of the putative reaction inter-
mediates. We experimentally observed that the nature of the
solvent had a profound effect on the reaction rate, yield, and
selectivity of the process. Indeed, competitive N-alkoxycarbo-
nylation (resulting from reductive elimination from Int-C) and
hydrolysis of the ester group under basic conditions could be
and the approximation of the O atom to the H atom (dO-H
1.34 Š) coupled with the formation of the PdÀC bond (dPd-C
=
=
Scheme 5. Proposed mechanism for the d-alkoxycarbonylation of 1a with ClCO2Et.
Chem. Eur. J. 2021, 27, 5782 –5789
www.chemeurj.org
5786
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005338
Chemistry—A European Journal
Scheme 6. Energy profile for the d-(mono)alkoxycarbonylation of 1a with ClCO2Et.
2.17 Š). In this case, the CÀH activation step would be the
rate-limiting step of the catalytic cycle with a barrier of
27.4 kcalmolÀ1. Intermediate Int-B would next coordinate with
ethyl chloroformate to deliver Int-C’ with an energy penalty of
5.2 kcalmolÀ1. The optimized structure of this intermediate re-
veals a PdÀC bond length of 2.63 Š. The latter would undergo
oxidative addition through a concerted pathway with a barrier
of 11.8 kcalmolÀ1 to afford thermodynamically favored PdIV
species Int-C through TS2. In this three-membered transition
state, the PdÀCl and PdÀC distances are shortened to 2.42 and
2.23 Š, respectively, whereas the CÀCl distance is lengthened
to 2.02 Š. With the formation of Int-C and dissociation of the
CÀCl bond, the PdÀCl and PdÀC distances are shortened and
maintained over 1.96 and 2.36 Š, respectively. Although this
species has been proposed to exist as both monomeric PdIV
and dimeric PdIII species, we have considered the monomeric
form owing to its relative stability. Finally, this reactant com-
plex could undergo a reductive elimination with a barrier of
12.9 kcalmolÀ1 via a three-membered transition state TS3, in
which the CÀC distance is shortened to 2.03 Š and the PdÀC
distance is lengthened in 0.11 Š, leading to the thermodynami-
cally favored product complex PC-1 with an energy of
À47.4 kcalmolÀ1. At this point, two possible scenarios could
occur: the dissociation of the product complex to provide the
mono-functionalized product 3a, thereby releasing the active
PdII catalyst, or a series of ligand exchange reactions to deliver
Int-D, which could undergo the second CÀH functionalization
process (Scheme 7). It is worth noting that the formation of
Int-D has been simplified and reduced to a simple one-step re-
action. In this regard, despite the fact that the presence of
silver carbonate was found to be indispensable for the process
to occur, its actual role cannot be attributed as a mere halide
scavenger as heterodimeric PdÀAg intermediates[24] could be
also formed within our catalytic cycle.
The formation of the di-alkoxycarbonylated product 3a’
would take place following an analogous catalytic cycle featur-
ing CÀH functionalization, oxidative addition, and reductive
elimination steps.[25] The geometries of the transient species
are similar to those mentioned in the catalytic cycle toward
the mono-alkoxycarbonylated compound 3a. As depicted in
Scheme 7, the CÀH activation event is also the rate-limiting
step with an energy barrier of 35.3 kcalmolÀ1. Moreover, not
only the oxidative addition (TS5) and reductive elimination
(TS6) steps, with energy barriers of 11.4 and 13.5 kcalmolÀ1, re-
spectively, but also all the intermediates described are energet-
ically viable. Therefore, the thermodynamically favored product
complex PC-2 would be easily formed at the optimized reac-
tion conditions involving a reaction temperature of 1258C. As
in the first catalytic cycle, coordination of the transient species
with t-amylOH led to a more favored reaction pathway (blue
vs. black pathway in Scheme 7).
Concerning the key role of LiI within the reaction outcome,
we performed some DFT calculations assuming a ligand ex-
change prior to the reductive elimination step but we did not
obtain any significant energy values.[18] Accordingly, further
studies are required to clarify the role of iodide additives as
they could accelerate other fundamental steps and, likewise,
the intermediacy of iodide-bridged Pd dimers could not be dis-
carded.[26]
Finally, we carried out some calculations to rationalize the
experimentally observed lower or lack of reactivity of aryl
chloroformates and dipeptides devoid of the ester motif within
our alkoxycarbonylation manifold. As shown on Figure S3 (in
the Supporting Information),[18] the kinetic barriers and the
thermodynamic values when using ClCO2Ph are similar to
those obtained with highly reactive ClCO2Et; however, the use
of ClCO2Ph led to the target product 3e in low yields (Table 2).
Accordingly, we hypothesized that its lower reactivity might be
Chem. Eur. J. 2021, 27, 5782 –5789
www.chemeurj.org
5787
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005338
Chemistry—A European Journal
Scheme 7. Energy profile for the d-(di)alkoxycarbonylation of 1a with ClCO2Et.
derived from unproductive reaction pathways. Concerning the
lack of reactivity of related Phe-containing dipeptides, compu-
tational studies with PA-Phe-Gly-OMe as the model substrate
were undertaken. As depicted on Figures S4 and S5 (in the
Supporting Information),[18] the nitrogen atom of the peptide
backbone could also coordinate to the palladium center, there-
by resulting in a distinct reaction intermediate that could even-
tually undergo oxidative addition of ClCO2Et instead of the de-
sired C(sp2)ÀH carbonylation reaction. In agreement with the
experiments, the use of dipeptides could result in an energeti-
cally more favored N-alkoxycarbonylation event.
this Pd-catalyzed oxidative CÀH carboxylation manifold could
become a useful synthetic tool for the rapid diversification of a
virtually unlimited set of b-arylethylamines and benzylamines.
Acknowledgments
We are grateful to MINECO (RTI2018-093721-B-I00) and the
Basque Government (IT1033-16 and IT1254-19) for financial
support. We thank the technical and human support provided
by SGIker of UPV/EHU and European funding (ERDF and ESF).
Dr. I. Rivilla is kindly acknowledged for his support with HPLC
analysis.
Conclusion
We have developed an unprecedented site-selective d-C(sp2)À
H alkoxycarbonylation technique for the modification of a vari-
ety of Phe residues in a simple fashion. This protocol avoids
the use of toxic carbon monoxide as the C1 source and com-
plements the method by Shi for the installation of ester moiet-
ies now at remote sites of Phe derivatives. Notably, this alkoxy-
carbonylation reaction could be also applied in simple benzyla-
mines and upon cleavage of the DG, it results in the assembly
of the privileged isoindolinone core in a straightforward
manner and in the absence of commonly used carbon monox-
ide. Salient features of the protocol are the scalability, the func-
tional group tolerance, and the performance under air, which
represents a practical bonus in terms of operational simplicity.
Computational studies supported a PdII/PdIV catalytic cycle and
provided valuable insights into the reaction mechanism such
as the key role of t-amylOH as co-solvent. We anticipate that
Conflict of interest
The authors declare no conflict of interest.
Keywords: alkoxycarbonylation
functionalization · DFT calculations · phenylalanine
·
benzylamine
·
CÀH
[1] a) J.-B. Peng, H.-Q. Geng, X.-F. Wu, Chem 2019, 5, 526; b) X.-F. Wu, H.
Neumann, M. Beller, Chem. Soc. Rev. 2011, 40, 4986; c) A. Brennführer, H.
Neumann, M. Beller, Angew. Chem. Int. Ed. 2009, 48, 4114; Angew. Chem.
2009, 121, 4176.
[2] For selected reviews, see: a) J. Das, S. Guin, D. Maiti, Chem. Sci. 2020, 11,
10887; b) R. R. Karimov, J. F. Hartwig, Angew. Chem. Int. Ed. 2018, 57,
4234; Angew. Chem. 2018, 130, 4309; c) J. C. K. Chu, T. Rovis, Angew.
Chem. Int. Ed. 2018, 57, 62; Angew. Chem. 2018, 130, 64; d) J. He, M.
Wasa, K. S. L. Chan, Q. Shao, J.-Q. Yu, Chem. Rev. 2017, 117, 8754; e) J.
Wencel-Delord, F. Glorius, Nat. Chem. 2013, 5, 369.
Chem. Eur. J. 2021, 27, 5782 –5789
www.chemeurj.org
5788
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005338
Chemistry—A European Journal
[3] For selected reviews, see: a) S. Rej, Y. Ano, N. Chatani, Chem. Rev. 2020,
120, 1788; b) Q. Zhang, B.-F. Shi, Chin. J. Chem. 2019, 37, 647; c) I. Guer-
rero, A. Correa, Eur. J. Org. Chem. 2018, 6034; d) C. Sambiagio, D. Schçn-
bauer, R. Blieck, T. Dao-Huy, G. Pototschnig, P. Schaaf, T. Wiesinger, M. F.
Zia, J. Wencel-Delord, T. Besset, B. U. W. Maes, M. Schnürch, Chem. Soc.
Rev. 2018, 47, 6603; e) W.-H. Rao, B.-F. Shi, Org. Chem. Front. 2016, 3,
1028; f) L. Ackermann, Chem. Rev. 2011, 111, 1315; g) T. W. Lyons, M. S.
Sanford, Chem. Rev. 2010, 110, 1147.
[17] For selected examples, see: a) B. S. Schreib, M. Fadel, E. Carreira, Angew.
Chem. Int. Ed. 2020, 59, 7818; Angew. Chem. 2020, 132, 7892; b) B. Han,
B. Li, L. Qi, P. Yang, G. He, G. Chen, Org. Lett. 2020, 22, 6879; c) S. Guin,
P. Dolui, X. Zhang, S. Paul, V. K. Singh, S. Pradhan, H. B. Chandrashekar,
S. S. Anjana, R. S. Paton, D. Maiti, Angew. Chem. Int. Ed. 2019, 58, 5633;
Angew. Chem. 2019, 131, 5689; d) B.-B. Zhan, J. Fan, L. Jin, B.-F. Shi, ACS
Catal. 2019, 9, 3298; e) B.-B. Zhan, Y. Li, J.-W. Xu, X.-L. Nie, J. Fan, L. Jin,
B.-F. Shi, Angew. Chem. Int. Ed. 2018, 57, 5858; Angew. Chem. 2018, 130,
5960; f) S.-Y. Zhang, G. He, W. A. Nack, Y. Zhao, Q. Li, G. Chen, J. Am.
Chem. Soc. 2013, 135, 2124; g) G. He, G. Chen, Angew. Chem. Int. Ed.
2011, 50, 5192; Angew. Chem. 2011, 123, 5298.
[4] L. Wang, W. Sun, C. Liu, Chin. J. Chem. 2018, 36, 353.
[5] W. Zhou, P. Li, Y. Zhang, L. Wang, Adv. Synth. Catal. 2013, 355, 2343.
[6] W.-Y. Yu, W. N. Sit, K.-M. Lai, Z. Zhou, A. S. C. Chan, J. Am. Chem. Soc.
2008, 130, 3304.
[18] For more details, see the Supporting Information.
[7] J. Chen, J.-B. Feng, K. Natte, X.-F. Wu, Chem. Eur. J. 2015, 21, 16370.
[8] K. Mondal, P. Halder, G. Gopalan, P. Sasikumar, K. Radhakrishnan, V. P.
Das, Org. Biomol. Chem. 2019, 17, 5212.
[9] T. Kochi, S. Urano, H. Seki, E. Mizushima, M. Sato, F. Kakiuchi, J. Am.
Chem. Soc. 2009, 131, 2792.
[10] For selected reviews, see: a) C. Zhu, J. Liu, M.-B. Li, J.-E. Bäckvall, Chem.
Soc. Rev. 2020, 49, 341; b) B. Liu, F. Hu, B.-F. Shi, ACS Catal. 2015, 5,
1863; c) X.-F. Wu, H. Neumann, M. Beller, ChemSusChem 2013, 6, 229.
[11] E. Lenci, A. Trabocchi, Chem. Soc. Rev. 2020, 49, 3262.
[19] S. Kumar, S. Pradhan, S. Roy, P. B. De, P. Punniyamurthy, J. Org. Chem.
2019, 84, 10481.
[20] For example, see: a) C. Zhang, Y. Ding, Y. Gao, S. Li, G. Li, Org. Lett.
2018, 20, 2595; b) F. Ling, C. Ai, Y. Lv, W. Zhong, Adv. Synth. Catal. 2017,
359, 3707; c) B. Haffemayer, M. Gulias, M. J. Gaunt, Chem. Sci. 2011, 2,
312; d) K. Orito, A. Horibata, T. Nakamura, H. Ushito, H. Nagasaki, M. Yu-
guchi, S. Yamashita, M. Tokuda, J. Am. Chem. Soc. 2004, 126, 14342.
[21] a) M. J. Terrey, C. C. Perry, W. B. Cross, Org. Lett. 2019, 21, 104; b) Y.
Zheng, W. Song, Org. Lett. 2019, 21, 3257.
[12] For selected reviews, see: a) I. Guerrero, A. Correa, Asian J. Org. Chem.
2020, 9, 898; b) B.-B. Zhan, M.-X. Jiang, B.-F. Shi, Chem. Commun. 2020,
56, 13950; c) C. Bottecchia, T. Nol, Chem. Eur. J. 2019, 25, 26; d) H. Y.
Chow, Y. Zhang, E. Matheson, X. Li, Chem. Rev. 2019, 119, 9971; e) W.
Wang, M. M. Lorion, J. Shah, A. R. Kapdi, L. Ackermann, Angew. Chem.
Int. Ed. 2018, 57, 14700; Angew. Chem. 2018, 130, 14912; f) M. San Se-
gundo, A. Correa, Synthesis 2018, 50, 2853; g) J. N. deGruyter, L. R.
Malins, P. S. Baran, Biochemistry 2017, 56, 3863; h) G. He, B. Wang, W. A.
Nack, G. Chen, Acc. Chem. Res. 2016, 49, 635; i) A. F. M. Noisier, M. A.
Brimble, Chem. Rev. 2014, 114, 8775.
[22] J.-I. Mujika, J. Uranga, J.-M. Matxain, Chem. Eur. J. 2013, 19, 6862.
[23] a) D. Lapointe, K. Fagnou, Chem. Lett. 2010, 39, 1118; b) D. García-Cua-
drado, A. A. C. Braga, F. Maseras, A. M. Echavarren, J. Am. Chem. Soc.
2006, 128, 1066.
[24] The mode of action of Ag2CO3 as a non-innocent additive within the re-
action pathway still remains unclear. A dual role has been proposed in
related metal-catalyzed CÀH functionalizations acting as an oxidant to
turn over the catalytic cycle and to form heterodimeric PdÀAg com-
plexes, thus lowering the activation barriers of some elemental steps:
a) B. Bhaskararao, S. Singh, M. Anand, P. Verma, P. Prakash, S. A. C. Mala-
kar, H. F. Schaefer, R. B. Sunoj, Chem. Sci. 2020, 11, 208; b) M. Anand,
R. B. Sunoj, H. F. Schaefer III, J. Am. Chem. Soc. 2014, 136, 5535; c) Y.-F.
Yang, G.-J. Cheng, P. Liu, D. Leow, T.-Y. Sun, P. Chen, X. Zhang, J.-Q. Yu,
Y.-D. Wu, K. N. Houk, J. Am. Chem. Soc. 2014, 136, 344.
[25] For the full reaction mechanism and energy profile, please see Figur-
es S2 and S3, respectively, within the Supporting Information including
computational details.
[26] Y. Zhang, Z.-N. Chen, X. Zhang, X. Deng, W. Zhuang, W. Su, Commun.
Chem. 2020, 3, 41.
[13] a) G. Liao, X.-S. Yin, K. Chen, Q. Zhang, S.-Q. Zhang, B.-F. Shi, Nat.
Commun. 2016, 7, 12901; b) G. Liao, H.-M. Chen, B.-F. Shi, Chem.
Commun. 2018, 54, 10859.
[14] a) G. He, G. Lu, Z. Guo, P. Liu, G. Chen, Nat. Chem. 2016, 8, 1131; b) G.
He, Y. Zhao, S. Zhang, C. Lu, G. Chen, J. Am. Chem. Soc. 2012, 134, 3;
c) G. He, C. Lu, Y. Zhao, W. A. Nack, G. Chen, Org. Lett. 2012, 14, 2944.
[15] a) M. San Segundo, A. Correa, Chem. Sci. 2020, 11, 11531; b) P. Andrade-
Sampedro, A. Correa, J. M. Matxain, J. Org. Chem. 2020, 85, 13133; c) I.
Guerrero, A. Correa, Org. Lett. 2020, 22, 1574; d) M. San Segundo, A.
Correa, Chem. Sci. 2019, 10, 8872; e) M. San Segundo, A. Correa, Chem-
SusChem 2018, 11, 3893; f) I. Guerrero, M. San Segundo, A. Correa,
Chem. Commun. 2018, 54, 1627; g) A. Goitia, E. Gómez-Bengoa, A.
Correa, Org. Lett. 2017, 19, 962.
Manuscript received: December 15, 2020
Accepted manuscript online: January 12, 2021
Version of record online: March 1, 2021
[16] V. G. Zaitsev, D. Shabashov, O. Daugulis, J. Am. Chem. Soc. 2005, 127,
13154.
Chem. Eur. J. 2021, 27, 5782 –5789
www.chemeurj.org
5789
ꢀ 2021 Wiley-VCH GmbH