Cobalt-Catalyzed ortho-C?H Functionalization/Alkyne Annulation of Benzylamine Derivatives: Access to Dihydroisoquinolines

Article abstract of DOI:10.1002/chem.201702283

A practical picolinamide-directed C?H functionalization/alkyne annulation of benzylamine derivatives enabling access to the previously elusive 1,4-dihydroisoquinoline skeleton was developed using molecular O₂ as the sole oxidant and Co(OAc)₂ as precatalyst. The method is compatible with both internal and terminal alkynes and shows high versatility and functional-group tolerance. Furthermore, full preservation of enantiopurity is observed when using non-racemic α-substituted benzylamine derivatives. Kinetic analysis of the reagents and catalyst, labeling experiments, and the isolation and identification of catalytically competent Co-complexes revealed important insights about the mechanism.

Full text of DOI:10.1002/chem.201702283

A Journal of
Accepted Article
Title: Cobalt-Catalyzed ortho-C-H Functionalization/Alkyne Annulation
of Benzylamine Derivatives: Access to Dihydroisoquinolines
Authors: Angel Manu Martínez, Nuria Rodríguez, Ramon Gomez-
Arrayas, and Juan C Carretero
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201702283
Link to VoR: http://dx.doi.org/10.1002/chem.201702283
Supported by

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
Cobalt-Catalyzed ortho-CH Functionalization/Alkyne Annulation
of Benzylamine Derivatives: Access to Dihydroisoquinolines
Ángel Manu Martínez[a], Nuria Rodríguez,*[a,b] Ramón Gómez-Arrayás,*[a,b] and Juan C. Carretero*[a,b]
Abstract:
A
practical
picolinamide-directed
CH
CH functionalization in concert with alkyne annulation is
typically enabled by a bidentate chelating auxiliary,[7] which is
attached to the arene system by a carbonyl or a heteroatom
linkage, most often benzamide and related functionalities.[8] In
contrast, to the best of our knowledge, there are only a handful
of procedures for the use of benzylamine derivatives in Co-
mediated alkyne annulation. Within a general protocol for cobalt-
catalyzed, aminoquinoline-directed C(sp2)H annulation of
benzamide derivatives with alkynes, Daugulis disclosed a single
functionalization/alkyne annulation of benzylamine derivatives
enabling the access to the previously elusive 1,4-dihydroisoquinoline
skeleton was developed using molecular O2 as the sole oxidant and
Co(OAc)2 as precatalyst. The method is compatible with both
internal and terminal alkynes and shows high versatility and
functional-group tolerance. Furthermore, full preservation of
enantiopurity is observed when using non-racemic -substituted
benzylamine derivatives. Kinetic analysis of the reagents and
catalyst, labeling experiments and the isolation and identification of
catalytically competent Co-complexes revealed important insights
about the mechanism.
isolated example of annulation of
a -methyl-substituted
benzylamine-derived picolinamide substrate with 2-butyne in the
presence of a stoichiometric amount of cobalt in combination
with 2 equiv of Mn(OAc)2 as cooxidant and O2 (air) as the
terminal oxidant to afford the corresponding dihydroisoquinoline
derivative in 44% yield.[5a] This lack of precedents of catalytic
protocols is striking considering that the resulting
dihydroisoquinoline (DHIQ) skeleton is an interesting motif
present in natural products and a versatile scaffold for alkaloid
Introduction
Alkyne annulation of nitrogen containing arenes through
CH/NH functionalization, in which vicinal CN and CC bonds
across an alkyne are installed in a single transformation, has
proven to be a powerful tactic for the step-economical assembly
of various heterocycles with activities of relevance to medicinal
chemistry and biology.[1,2] While impressive progress has been
made using second and third row transition metals such as Pd,
Rh or Ru,[2] homogeneous cobalt catalysis is currently drawing
much attention as a more abundant, less toxic and cheaper
viable alternative.[3] This increasing demand to move toward
more environmentally benign approaches makes also highly
desirable to replace the stoichiometric oxidants generally
required to maintain the catalytic cycle (most often silver or
copper salts) with molecular oxygen, in which no wastes are
formed except for water.[4] However, yet despite the significant
progress in this field, reports relying on oxygen as the terminal
oxidant remain limited, particularly in the context of cobalt-
catalyzed heterocycle synthesis via arene CH annulation with
-systems. In particular, the first broadly applicable Co-catalyzed
aerobic alkyne annulations have appeared only recently.[5,6]
Consequently, catalyst systems combining simple cobalt salts
and O2 as the sole oxidant are being actively sought.
,10]
syntheses.[9
The scarcity of studies involving benzylamine derivatives in
Co-catalyzed CH functionalization[11] may be attributed to their
propensity to undergo dehydrogenation at the benzylic position
under oxidative conditions, potentially leading to imine-type
intermediates and/or aromatization of the final products.[12] We
speculated that the combination of Co-catalytst with oxygen as
smooth oxidant could enable an efficient access to DHIQ
derivatives via annulation of benzylamine derivatives with
alkynes, thus expanding the synthetic utility of this reaction while
moving towards sustainable development. During the
preparation of this manuscript, Cui reported a Co(OAc)2-
catalyzed picolinamide-directed annulation of benzylamine
derivatives with alkynes using a 50 mol% of catalyst loading and
O2 as terminal oxidant.[13] In this report, dehydrogenation at the
benzylic position occurs concomitantly to the annulation process,
resulting in the formation of isoquinolines. Herein we present an
operationally simple and structurally flexible procedure for the
synthesis of a variety of DHIQ, including chiral, non-racemic
derivatives. Important mechanistic insights including isolation
and interception of some presumed Co-species involved in the
catalytic cycle are also provided.
[a]
Mr, A. M. Martínez, Dr. N. Rodríguez, Dr. R. Gómez Arrayás, Prof.
Dr. Juan C. Carretero
Departamento de Química Orgánica, Facultad de Ciencias
Universidad Autónoma de Madrid (UAM)
Cantoblanco 28049 Madrid (Spain)
E-mail: n.rodriguez@uam.es; ramon.gomez@uam.es;
juancarlos.carretero@uam.es
Dr. N. Rodríguez, Dr. R. Gómez Arrayás, Prof. Dr. Juan C.
Carretero
Institute for Advanced Research in Chemical Sciences (IAdChem)
UAM, 28049 Madrid, Spain
Results and Discussion
Optimization studies. The model reaction between the parent
N-benzylpicolinamide (1) and 4-octyne was chosen for
optimization studies (Table 1). In accordance with the above
mentioned precedents,[8,11] poor outcome obtained in the initial
experiment using Co(OAc)2 (20 mol %) in conjunction with
Mn(OAc)2 (2 equiv) as cooxidant and NaOAc (4 equiv) as a base
under oxygen atmosphere, affording product 2 in 13% GC yield
[b]
Supporting information for this article is given via a link at the end of
the document
This article is protected by copyright. All rights reserved.

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
[c] Isolated yield upon chromatographic purification. [d] Degassed solvent.
[e] In the absence of Co-salt. [f] 4-Octyne (0.22 mmol, 1.50 equiv),
Co(OAc)2 (15 mol%). [g] 80 °C. [h] Scale-up to 1.5 mmol of 1.
(entry 1). Nevertheless, the reaction was very clean, with no
other byproducts detected. In the absence of NaOAc the
catalytic activity was virtually suppressed (5% GC yield, entry 2),
suggesting that an extra-source of acetate ions is necessary for
the reaction to proceed. To our satisfaction, the conversion was
dramatically improved in the absence of any oxidant other than
oxygen, allowing the isolation of 2 in 77% yield (entry 3).
Furthermore, oxygen as terminal oxidant is crucial for this
reaction, given that only traces of 2 could be detected in the
absence of both Mn(OAc)2 and oxygen (degassed solvent, entry
4). As expected, no product was detected when the reaction was
performed in the absence of cobalt catalyst (entry 5). A solvent
screening (entries 6-11) revealed that ethanol, 1,4-dioxane and
DCE are suitable for this transformation (80-84% GC yield),
while HFIP failed to provide 2 in useful conversion (16% GC
yield, entry 8).
Using EtOH as solvent, we sought to enable lower catalyst
loading and decrease the excess of both alkyne and base. After
minor adjustments, it was found that product 2 can be obtained
in 85% isolated yield within just 2.5 hours with 15 mol% of Co-
catalyst in the presence of 1.5 equiv of alkyne and 1.5 equiv of
NaOAc (entry 12), which were set up as the optimized
conditions. The essential role of acetate for catalyst turnover
was verified by lowering NaOAc loading to 1.0 and 0.5 equiv and
observing a significant decrease in the yield of 2 (79% and 22%
GC yield, entries 13 and 14, respectively). We also found a
small decrease in reactivity when the reaction temperature was
reduced to 80 °C (entry 15). Finally, this method allows for
scale-up to 1.5 mmol with no appreciable loss of efficiency,
which is important for practical convenience (entry 16).
Table 1. Optimization studies in the model reaction of 1 with 4-octyne[a]
Chemoselective deprotection and removal of the auxilliary
COPy group. In accordance with our desired goal of accessing
DHIQ, conditions for the efficient removal of the picolinamide
auxiliary without aromatization of the heterocyclic ring were
needed. Under the commonly employed basic conditions (KOH
in EtOH at 125 °C), deprotection/aromatization took place
efficiently to provide the isoquinoline derivative, as illustrated in
the transformation of 2 into 3 (93% yield, Scheme 1). Instead, N-
deprotection without aromatization proved to be feasible by
changing to reductive acidic conditions.[14] In the presence of
Zn/AcOH, the resulting unprotected enamine undergoes a facile
benzylic oxidation by atmospheric O2 to afford the 4-hydroxy-
1,4-dihydroisoquinoline 4. This oxidation of 1,2-DHIQ derivatives
had been documented in the literature.[15]
Entry
Cooxidant
NaOAc
(equiv)
Solvent
Time
(h)
2 (%)[b]
1
Mn(OAc)2
4.0
0
TFE
12
12
12
12
12
4
13
2
Mn(OAc)2
TFE
5
3














4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
1.5
1.0
0.5
1.5
1.5
TFE
85 (77)[c]
4[d]
5[e]
6
TFE
<5
TFE
<5
TFE
85
7
EtOH
HFIP
4
83
Scheme 1. N-Deprotection of 2.
8
4
16
9
DCE
4
80
Evaluation of the substitution at the N-benzylpicolinamide
substrate. Encouraged by these results, we next evaluated the
scope of the CH/NH annulation of various benzylamine
derivatives with 4-octyne (Scheme 2). A broad range of p-
substituted derivatives underwent annulation reaction in
acceptable to good yields (products 22-29, 57-96%) with no
remarkable sensitivity to the electronic characteristics of the
substituent (OMe, SMe, Me, Cl, F, CF3, CN, CO2Me). Eventually,
the crude reaction mixture can be subjected to the N-
deprotection procedure under Zn/HCl accompanied by
spontaneous oxidation by air to afford the corresponding 4-
hydroxy-1,4-DHIQ derivatives in useful overall yields (products
24-27, 57-92%). It is also worth mentioning that those substrates
bearing a OMe or a SMe substituents led to the in situ formation
of the corresponding fully aromatic isoquinoline derivatives in the
10
11
12[f]
13
14
15[g]
16[h]
1,4-Dioxane
p-Xyle[c]ne
EtOH
EtOH
EtOH
EtOH
EtOH
4
84
4
4
2.5
2.5
2.5
2.5
12
92 (85)[c]
79
22
70
80 (70)[c]
[a] 1 (0.15 mmol, 1.00 equiv), 4-octyne (0.37 mmol, 2.50 equiv), Co(OAc)2
(20 mol%), cooxidant (2.00 equiv), NaOAc (x equiv), solvent (1 mL), O2
(1 atm), 100 °C, t (h). [b] GC yields (n-hexadecane as internal standard).
This article is protected by copyright. All rights reserved.

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
Co-catalyzed annulation step (28 and 29, respectively). The
reason behind this different behavior is not clear at the present
time.
Co-catalyzed CH alkenylation/annulation of benzylamine 1
with internal alkynes. To study the alkyne substrate scope,
several types of internal alkynes were surveyed in the reaction
with N-benzylpicolinamide 1 (Scheme 3). Unfortunately, a diaryl-
substituted alkyne such as diphenylacetylene was found to be
unsuitable for this set of conditions (36, <5% GC yield).
Interestingly, this reactivity was opposite to that observed in our
previously reported Rh-catalyzed CH functionalization of
picolinamides with alkynes, in which diarylacetylenes displayed
excellent reactivity whereas dialkyl-alkynes resulted in a total
lack of reactivity.[16] This, along with the different reaction
outcome observed, highlights the complementarity between Rh-
catalysis and Co-catalysis.
Unsymmetrical internal alkynes represents a greater challenge
due to their inherent difficulty to achieve useful regiocontrol in
the 1,2-migratory insertion step. We were pleased to find that
the reaction with conjugated 1-aryl-1-alkynes such as 1-phenyl-
1-propyne occurred smoothly to afford the corresponding
isoquinoline 37 as a single regioisomer, in which the aryl
substituent is regioselectively placed proximal to the nitrogen
atom (Scheme 3, 94% yield). Other conjugated alkynes with
electronically dissimilar acetylenic substituents did also
participate in the coupling reaction maintaining the level and
sense of regioselectivity, to provide the corresponding
isoquinoline derivatives (products 38 and 39, 52% and 74%,
respectively). In these cases, the intermediate DHIQ product
could not be isolated, which seems to indicate that the use of
conjugated alkynes as the coupling partner favors the
aromatization/N-deprotection of the heterocyclic ring leading to
the isoquinoline product rather than the DHIQ. Accordingly, a
mixture of DHIQ regioisomers was observed with a sterically
biased internal dialkyl-alkynes such as isopropyl methyl
acetylene, albeit with virtually no regiocontrol (40a+40b, 84%).
This result suggests that the regioselectivity of insertion appears
to be dictated mainly by electronic factors rather than sterics
factors.
In contrast, the electronic effects of the meta-substituents
appear to have more influence over the reactivity than do the
substituents at para-position. For example, whereas a meta-Me
benzylamine derivative provides the corresponding DHIQ
product 30 in high yield (87%), the presence of a strong
electron-withdrawing CF3 substituent at meta-position decreased
the yield significantly (31, 4% GC yield). The functionalization
took place at the more sterically accessible ortho-CH bond of
the meta-substituted benzylamine, the DHIQ being obtained as
a single regioisomer (see SI for structure determination). In
contrast to many reported methods in which an ortho-substituent
retards or shuts down the reaction completely, this
transformation showed no appreciable sensitivity to steric
hindrance in the aryl ortho-position (32-34, 56-95%). Slightly
reduced reactivity was observed in the case of an electron
withdrawing ortho-F substituent, yet the corresponding DHIQ 34
was obtained in a synthetically useful 56% yield.
Overall, it is important to stress the considerable
chemoselectivity offered by this catalyst system, tolerating
substrates bearing a range of useful functionalities, including
halides (F, Cl and Br) and the potentially coordinating SMe and
CN moieties. Even a twofold CH functionalization/alkyne
annulation was feasible in the case of using
a 1,4-
phenylenedimethanamine derivative, resulting in the direct
construction of a polyheteroaromatic pyrido[3,4-g]isoquinoline
skeleton (35, 68%) under the reaction conditions.
Scheme 2. Evaluation of the substitution at the N-benzylpicolinamide
substrate. [a] Overall yield for the annulation / N-deprotection steps. [b]
Co(OAc)2 (20 mol%), 12 h. [c] 1,4-dioxane as solvent and heating at 110 °C
for 12 h. [d] Detected by GC.
Scheme 3. Co-catalyzed CH alkenylation/annulation of benzylamine 1 with
internal alkynes. [a] Determined by GC. [b] Regiochemistry determined by
NOESY experiments. [c] Using 1,4-dioxane as solvent.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
Co-catalyzed CH annulation with terminal alkynes. A
limitation frequently encountered in rhodium(III)- and
ruthenium(II)-catalyzed oxidative synthesis of nitrogen-
containing heterocycles is the inability to incorporate terminal
alkynes due to the problematic alkyne dimerization.[17] It is,
therefore, noteworthy that a variety of terminal alkynes with both
alkyl and aryl substitution reacted smoothly under these Co-
catalyzed conditions, affording in good yields as single
regioisomers the corresponding 1,2-DHIQ in the case of alkyl-
alkynes (Scheme 4, 41-46, 62-91%) or the corresponding
isoquinolines in the case of aryl-alkynes (47-51, 68-94%). In all
cases studied the alkyl or aryl acetylenic substituent ends up
distal to the nitrogen atom (at C4) in the DHIQ skeleton.
Conjugated aryl-alkynes worked effectively in the absence of
NaOAc as additive.
that the reaction may proceed through an irreversible
cyclometalation pathway.[18]
Scheme 5. Tolerance towards substitution at the benzylic position and
preservation of stereochemical integrity.
Scheme 6. D-Labelling experiments.
Kinetic studies. Next, we focused on the reaction rate
dependences on the catalyst and the reagents concentrations in
the model coupling between 1 and 4-octyne under optimized
conditions (see SI for details).[19] The reaction exhibited a first
order kinetic dependence with respect to cobalt concentration,
suggesting a mononuclear species as active catalyst. Reaction
rate versus time curve revealed a significant induction period of
25 min.[5a] Nonetheless, during the portion of the reaction that
exhibits steady state behavior, the rate of the annulation was
found to be zeroth-order with respect to the alkyne, suggesting
that the alkyne is not involved in the rate-determining step. The
reaction showed a partial negative order in the benzylamine
derivative concentration, which can be plausibly ascribed to off-
cycle nonproductive binding interactions between this species
and the catalyst prior to the CꢀH cleavage, thereby decreasing
the effective concentration of catalyst. A similar behavior of the
negative-order rate dependence on substrate has also been
observed in recent related literature on Co-catalysis.[20]
Isolation of a Co-complex and study of its competency. To
gain insight about possible intermediates during the reaction
course, we explored the reaction of equimolar amounts of 1 and
Co(OAc)2 in the presence of NaOAc (1.5 equiv) in EtOH under
O2 atmosphere at 100 °C for 12 hours (Scheme 7a). In this
reaction we could isolate the air-stable CoIII-complex A (20%),
which was characterized through ESI-HRMS and NMR analysis
upon chromatographic purification (see SI). Unfortunately, no
suitable crystals for X-ray diffraction analysis could be obtained
for this complex. In CoIII-complex A, three units of 1 are
coordinated to the metal center as monoanionic bidentate N,N-
donor ligands. It may be noted that cobalt undergoes a one-
electron oxidation (CoII → CoIII) during the complex formation in
which oxygen appears to have served as the oxidant.[21]
Scheme 4. Co-catalyzed CH annulation with terminal alkynes. [a] Co(OAc)2
(10 mol%), NaOAc (40 mol%). [b] Regiochemistry determined by NOESY
experiments. [c] 20 mol% NaOAc. [d] In absence of NaOAc.
Tolerance towards substitution at the benzylic position and
preservation of stereochemical integrity. To test whether this
annulation protocol would be sensitive to substitution at the
benzylic position and whether the stereochemical integrity of a
stereogenic center at such position would be preserved, the
enantiopure (S)--methylbenzylamine derivative (S)-52 was
subjected to the standard conditions and subsequent N-
deprotection with Zn/HCl (Scheme 5). Pleasingly, the resulting
4-hydroxy-1,4-DHIQ ()-53 was obtained in good yield (75%)
and no appreciable loss of enantiopurity (>98% ee, major
diastereoisomer), yet with moderate diastereoselectivity (dr =
65:35). The stereochemical assignment of 53 was tentatively
established by NOESY experiments (see SI for details).
Mechanistic insights. Deuterium labelling studies. We next
performed a series of experiments to gain some insights into the
mechanism (Scheme 6). First, reactions with isotopically labeled
co-solvents (D2O or CD3OD) in both the presence and absence
of the alkyne coupling partner, showed no deuterium
incorporation at either the recovered unreacted starting material
or at the DHIQ product 2. This lack of H/D scrambling suggests
This article is protected by copyright. All rights reserved.

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
Interestingly, Co-species A smoothly reacted with 4-octyne to
afford the DHIQ derivative 2 as the only product in good yield
(89%, Scheme 7b). Furthermore, complex A was found to be
catalytically competent in the reaction of 1 with 4-octyne,
providing 2 in good yield (Scheme 7c), suggesting that this
complex could be an active catalyst precursor for this reaction.
ESI-HRMS experiment. ESI-HRMS experiment shed some
more light on the nature of the presumed Co-complexes that
were formed in the reaction of 1 with 4-octyne under the
optimized catalytic conditions. The mass spectrum of the crude
reaction mixture upon 45 min (Figure 1) was quite clean, with a
dominant molecular ion peak of DHIQ 2 (m/z=343.1778).
Interestingly, four peaks could be assigned to Co-complexes.
The peak at m/z=693.1991 corresponds to CoIII-complex A
previously isolated (calculated value: m/z=693.2019). The peak
at m/z=775.1922 was assigned to CoIII-complex B (theoretical
value: m/z=775.2050, M+Na+), which keeps the three units of
picolinamide ligand in the coordination sphere of Co, but one of
which is now coordinated as a monodentate neutral ligand
(rather than in a bidentate monoanionic fashion), with an acetate
occupying the vacant site. The peak at m/z=563.1093 was
assigned to CoIII-complex C (theoretical value: m/z=563.1099,
M+Na+), holding two units of 1 as ligands and an acetate bonded
to the metal (the latter likely in a bidentate fashion). Presumably,
complex C results from complex A through ligand displacement
via complex B. Interestingly, a minor peak D at m/z=481.0991
was also observed, whose mass would correspond to the five
membered cobaltacycle presumably formed from C upon ortho-
CH functionalization (CoIII-complex D, consistent with the
theoretical value m/z=481.1069).[22a]
Scheme 7. Isolation of a preformed Co-complex and study of its competency.
Figure 1. Key area of the ESI-HRMS spectrum of the crude reaction mixture between 1 and 4-octyne after 45 min.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
Plausible mechanistic hypothesis. Taking together all these
experimental observations and current literature,[7,22] a plausible
simplified mechanism is depicted in Scheme 8. Initial
enantiomeric purity when starting from chiral non-racemic
substrates, thereby allowing the preparation of chiral, non-
racemic DHIQ derivatives. Mechanistic studies including kinetic
analysis, labelling studies, Co-complex isolation and HRMS-
interception provided valuable insights about the reaction course
and the structure of some presumed Co-species involved in the
catalytic cycle.
coordination of
1 to Co(OAc)2 species with concomitant
oxidation of cobalt(II) by oxygen (dissolved in solution)21 would
generate the presumed active CoIII species C, in equilibrium with
CoIII complex A via complex B. In fact, complex A may function
as a reversible off-cycle reservoir that could explain the partial
negative order for the substrate 1 previously observed in kinetic
studies. Acetate-assisted ortho-CH bond functionalization
would lead to the coordinatively unsaturated 16-electron CoIII
complex D. Alkynyl coordination (E) and subsequent 1,2-
migratory insertion of the cobalt-carbon bond across the alkyne
would result in the formation of the seven-membered alkenyl-
cobalt intermediate F. Subsequent reductive elimination step
releases the 1,2-DHIQ product while the concomitantly formed
CoI species is reoxidized by O2 to regenerate the CoII and/or
CoIII species, thus closing the catalytic cycle. The relative ease
of forming the off-cycle tris-picolinamide chelate CoIII complexes
of type A, which should be less favorable in the case of using
the more-hindered 8-quinolinamide substrates, could explain the
higher reactivity of the latter substrates over picolinamide
derivatives[8] and the use of high catalyst loadings[8,13] described
in previous reports. Interestingly, the excess of acetate ions
seems to counteract the formation of tris-picolinamide chelate
CoIII complexes of type A by promoting ligand displacement to
form complexes of type C.
Experimental Section
Representative procedure for Co-catalyzed CH alkenylation-
annulation:
The conversion of N-benzylpicolinamide (1) into [(3,4-
dipropylisoquinolin-2(1H)-yl)(pyridin-2-yl)methanone] (2): An oven-
dried, nitrogen-flushed 10 mL vessel was charged with N-
benzylpicolinamide (1) (31.8 mg, 0.15 mmol, 1.00 equiv), cobalt(II)
acetate (3.98 mg, 0.023 mmol, 0.15 equiv) and sodium acetate (18.5 mg,
0.23 mmol, 1.50 equiv). The reaction vessel was sealed with a Teflon
lined cap, then evacuated and flushed with oxygen. Under the
atmosphere of oxygen, ethanol (1.00 mL) and 4-octyne (33.0 µL,
0.23 mmol, 1.50 equiv) were added via syringe. The resulting mixture
was then stirred at 100 °C for 2.5 h before it was quenched with a
saturated solution of potassium tartrate (5.00 mL) and extracted with
EtOAc (2 x 5.00 mL). The organic layer was washed with brine (5.00 mL),
dried over Na2SO4, filtered and concentrated in vacuo. The residue was
purified by column chromatography (CH2Cl2-Et2O 20:1) to afford 2 as a
colorless oil (42.7 mg; 89%; see the Supporting Information for
spectroscopic data).
Representative procedure for the deprotection of the picolinamide
group with KOH:
The conversion of [(3,4-dipropylisoquinolin-2(1H)-yl)(pyridin-2-
yl)methanone] (2) into 3,4-dipropylisoquinoline (3): An oven-dried,
nitrogen-flushed 20 mL vessel was charged with 2 (48.0 mg, 0.15 mmol,
1.00 equiv) and KOH (504 mg, 9.00 mmol, 60.0 equiv). The reaction
vessel was sealed with a Teflon lined cap, and ethanol (3.00 mL) was
added via syringe. The resulting mixture was stirred at 125 °C for 24 h
before it was cooled down to room temperature, diluted with ethyl acetate
and washed with water. The organic phase was separated, dried over
Na2SO4 and concentrated in vacuo. The residue was purified by column
chromatography (n-hexane-EtOAc-CH2Cl2 5:1:1), to give 3 as a yellow oil
(29.7 mg; 93%; see the Supporting Information for spectroscopical data).
Representative procedure for the deprotection of the picolinamide
group with Zn/HCl:
Scheme 8. Proposed catalytic cycle.
The conversion of [(3,4-dipropylisoquinolin-2(1H)-yl)(pyridin-2-
yl)methanone] (2) into 3,4-dipropyl-1,4-dihydroisoquinolin-4-ol (4):
To a suspension of 2 (48.0 mg, 0.12 mmol, 1.00 equiv) in a mixture of
H2O (3 mL) and THF (1 mL) was added 12 M HCl (0.30 mL, 30.0 equiv)
and the mixture was stirred for 5 minutes at room temperature. Zinc dust
(118 mg, 1.80 mmol, 15.0 equiv) was then added in three portions and
the mixture was stirred at room temperature until consumption of 2. Then,
the reaction was mixed with ca. 200 mg of sand and filtered through a
celite plug. The filtrate was transferred to a separatory funnel with 2M
NaOH (2 × 20 mL) and extracted twice with dichloromethane (2 × 20 mL).
The combined organic phase was dried over MgSO4 and concentrated in
vacuo. The residue was purified by column chromatography (n-hexane-
EtOAc-CH2Cl2 5:1:1), to give 4 as a white oil (31.1 mg; 96%; see the
Supporting Information for spectroscopical data).
Conclusions
In conclusion, we have developed
a
sustainable and
operationally simple Co-catalyzed CH alkenylation/annulation
of benzylamine derivatives with both terminal and internal
alkynes that enables the access DHIQ structures. This method
relies on simple Co(OA)2 as precatalyst and O2 as the sole
terminal oxidant and features high selectivity and broad
functional group tolerance. The reaction is tolerant of
substitution at the benzylic position, with no appreciable loss of
This article is protected by copyright. All rights reserved.

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
Front. 2014, 1, 843; f) R. K. Rit, M. R. Yadav, K. Ghosh, A. K. Sahoo,
Tetrahedron 2015, 71, 4450; g) M. R. Yadav, R. K. Rit, M. Shankar, A.
K. Sahoo, Asian J. Org. Chem. 2015, 4, 846.
Acknowledgements
We thank the Spanish Ministerio de Economía, Industria y
Competitividad (MINECO, Grants CTQ2012-35790 and
CTQ2015-66954-P, MINECO/FEDER, UE) for financial support.
N.R. thanks the MICINN for a Ramón y Cajal contract and the
European Commission for a for a Marie Curie Career Integration
Grant (CIG: CHAAS-304085). We also thank Mario Martínez
Mingo for experimental help in the final stage of the project.
[8]
For pioneering work on low valent Co-catalyzed CH functionalization,
see: a) K. Gao, N. Yoshikai, J. Am. Chem. Soc. 2011, 133, 400; b) L.
Ilies, Q. Chen, X. Zeng, E. Nakamura, J. Am. Chem. Soc. 2011, 133,
5221; c) K. Gao, N. Yoshikai, Angew. Chem., Int. Ed. 2011, 50, 6888.
For pioneering work on high valent CoIII-catalyzed CH
functionalization: d) T. Yoshino, H. Ikemoto, S. Matsunaga, M. Nakai,
Angew. Chem. Int. Ed. 2013, 52, 2207; e) H. Ikemoto, T. Yoshino, K.
Sakata, S. Matsunaga, M. Kanai, J. Am. Chem. Soc. 2014, 136, 5424.
For selected recent examples on Co-catalyzed annulation with alkynes:
f) T. Yamakawa, N. Yoshikai, Org. Lett. 2013, 15, 196; g) B. Sun, T.
Yoshino, M. Kanai, S. Matsunaga, Angew. Chem., Int. Ed. 2015, 54,
12968; h) H. Wang, J. Koeller, W. Liu, L. Ackermann, Chem. Eur. J.
2015, 21, 15525; i) G. Sivakumar, A. Vijeta, M. Jeganmohan, Chem.
Eur. J. 2016, 22, 5899; (j) X.-Q. Hao, C. Du, X. Zhu, P.-X. Li, J.-H.
Zhang, J.-L. Niu, M.-P. Song, Org. Lett. 2016, 18, 3610; k) T. T.
Nguyen, L. Grigorjeva, O. Daugulis, ACS Catal. 2016, 6, 551; l) K.
Muralirajan, R. Kuppusamy, S. Prakash, C.-H. Cheng, Adv. Synth.
Catal. 2016, 358, 774; m) W. Yu, W. Zhang, Z. Liua, Y. Zhang, Chem.
Commun. 2016, 52, 6837; n) Z.-Z. Zhang, B. Liu, J.-W. Xu, S.-Y. Yan,
B.-F. Shi, Org. Lett. 2016, 18, 1776; o) Q. Lu, S. Vásquez-Céspedes, T.
Gensch, F. Glorius, ACS Catal. 2016, 6, 2352; p) L. Kong, S. Yu, X.
Zhou, X. Li, Org. Lett. 2016, 18, 588; q) H. Wang, M. Moselage, M. J.
González, L. Ackermann, ACS Catal. 2016, 6, 2705; r) S. Prakash, K.
Muralirajan, C.-H. Cheng, Angew. Chem. Int. Ed. 2016, 55, 1844; s) O.
Planas, C. J. Whiteoak, A. Company, X. Ribas, Adv. Synth. Catal. 2015,
357, 4003; t) A. B. Pawar, D. Agarwal, D. M. Lade, J. Org. Chem. 2016,
81, 11409; u) A. Lerchen, S. Vasquez-Céspedes, F. Glorius, Angew.
Chem., Int. Ed. 2016, 55, 3208; v) D. Kalsi, B. Sundararaju, Org. Lett.
2015, 17, 6118; w) S. Zhou, M. Wang, L. Wang, K. Chen, J. Wang, C.
Song, J. Zhu, Org. Lett. 2016, 18, 5632; x) L.-B. Zhang, X.-Q. Hao, Z.-J.
Liu, X.-X. Zheng, S.-K. Zhang, J.-L. Niu, M.-P. Song, Angew. Chem. Int.
Ed. 2015, 54, 10012; y) Y. Liu, C. Wang, N. Lv, Z. Liu, Y. Zhanga, Adv.
Synth. Catal. 2017, DOI: 10.1002/adsc.201600834. For selected recent
examples on Co-catalyzed annulation with alkenes: z) L. Grigorjeva, O.
Daugulis, Org. Lett. 2014, 16, 4684; aa) W. Ma, L. Ackermann, ACS
Catal. 2015, 5, 2822; ab) P. Gandeepan, P. Rajamalli, C.-H. Cheng,
Angew. Chem., Int. Ed. 2016, 55, 4308. For selected recent examples
on Co-catalyzed annulation with allenes: ac) R. Kuppusamy, K.
Muralirajan, C.-H. Cheng, ACS Catal. 2016, 6, 3909.
Keywords: cobalt • alkenylation/annulation • alkyne •
picolinamide • dihydroisoquinoline
[1]
[2]
Selected general reviews on CH functionalization: a) D. A. Colby, R. G.
Bergman, J. A. Ellman, Chem. Rev. 2010, 110, 624; b) J. Wencel-
Delord, T. Dröge, F. Liu, F. Glorius, Chem. Soc. Rev. 2011, 40, 4740;
c) B. Li, P. H. Dixneuf, Chem. Soc. Rev. 2013, 42, 5744.
Reviews on metal-catalyzed annulation of arenes with alkynes through
directed C–H functionalization: a) Y. Yang, K. Li, Y. Cheng, D. Wan, M.
Li, J. You, Chem. Commun. 2016, 52, 2872; b) L. Ackermann, Acc.
Chem. Res. 2014, 47, 281; c) P. B. Arockiam, C. Bruneau, P. H.
Dixneuf, Chem. Rev. 2012, 112, 5879; d) G. Song, F. Wang, X. Li,
Chem. Soc. Rev. 2012, 41, 3651; e) T. Satoh, M. Miura, Chem. Eur. J.
2010, 16, 11212; f) P. Thansandote, M. Lautens, Chem. Eur. J. 2009,
15, 5874.
[3]
Selected reviews on Co-catalyzed CH functionalization: a) N. Yoshikai,
Synlett 2011, 1047; b) K. Gao, N. Yoshikai, Acc. Chem. Res. 2014, 47,
1208; c) N. Yoshikai, Bull. Chem. Soc. Jpn. 2014, 87, 843; d) L.
Ackermann, J. Org. Chem. 2014, 79, 8948; e) M. Moselage, J. Li, L.
Ackermann, ACS Catal. 2016, 6, 498; f) T. K. Hyster, Catal. Lett. 2015,
145, 458; g) N. Yoshikai, ChemCatChem, 2015, 7, 732; h) D. Wei, X.
Zhu, L. Niu, M.-P. Song, ChemCatChem 2016, 8, 1242. For selected
reviews on first row transition metal catalyzed CH bond
functionalization, see: i) A. A. Kulkarni, O. Daugulis, Synthesis 2009,
4087; j) S. Z. Tasker, E. A. Standley, T. F. Jamison, Nature 2014, 509,
299; k) I. Bauer, H. Knölker, Chem. Rev. 2015, 115, 3170.
[4]
Selected reviews: a) J. Piera, J.-E. Bäckvall, Angew. Chem., Int. Ed.
2008, 47, 3506; b) S. Ueda, H. Nagasawa, Angew. Chem., Int. Ed.
2008, 47, 6411; c) K. M. Gligorich, M. S. Sigman, Chem. Commun.
2009, 3854; d) J.-S. Tian, T.-P. Loh, Angew. Chem., Int. Ed. 2010, 49,
8417; e) S. Guo, B. Qian, Y. Xie, C. Xia, H. Huang, Org. Lett. 2011, 13,
522; f) A. N. Campbell, S. S. Stahl, Acc. Chem. Res. 2012, 45, 851; g)
W. Wu, H. Jiang, Acc. Chem. Res. 2012, 45, 1736; h) Z. Shi, C. Zhang,
C. Tang, N. Jiao, Chem. Soc. Rev. 2012, 41, 3381; i) Y. Xie, B. Qian, P.
Xie, H. Huang, Adv. Synth. Catal. 2013, 355, 1315.
[9]
a) K. W. Bentley, The Isoquinoline Alkaloids. 1st Ed.; Pergamon:
London, 1965; b) J. P. Michael, Nat. Prod. Rep. 2002, 19, 742.
[10] For biological activity of DHIQ derivatives, see: a) M. Abdel-Aziz, G. E.-
D. A. A. Abuo-Rahma, H. A. Hassan, H. H. Farag, Arch. Pharm. Chem.
Life Sci. 2010, 343, 54; b) H. Lou, S. Q. Ye, J. F. Zhang, J. Wu,
Tetrahedron 2011, 67, 2060.
[5]
For a general Co-catalyzed alkyne annulation using O2 as the terminal
oxidant and Mn(OAc)2 as cooxidant: a) L. Grigorjeva, O. Daugulis,
Angew. Chem. Int. Ed. 2014, 53, 10209. For a general Co-catalyzed
alkyne annulation using O2 as the sole oxidant: b) R. Mei, H. Wang, S.
Warratz, S. A. Macgregor, L. Ackermann, Chem. Eur. J. 2016, 22, 6759.
For the use of the internal oxidant strategy (oxidizing directing group),
see: b) G. Sivakumar, A. Vijeta, M. Jeganmohan, Chem. Eur. J. 2016,
22, 5899; c) M. Sen, D. Kalsi, B. Sundararaju, Chem. Eur. J. 2015, 21,
15529.
[11] For the use of benzylamine derivatives in Co-catalyzed ortho CH
alkynylation: G. Landge, S. P. Midya, J. Rana, D. R. Shinde, E.
Balaraman, Org. Lett. 2016, 18, 5252.
[12] a) K. Morimoto, K. Hirano, T. Satoh, M. Miura, Chem. Lett. 2011, 40,
600; b) D.-S. Kim, J.-W. Park, C.-H. Jun, Adv. Synth. Catal. 2013, 355,
2667.
[6]
[7]
[13] C. Kuai, L. Wang, B. Li, Z. Yang, X. Cui, Org. Lett. 2017, 19, 2102.
[14] The use of Zn/NH4Cl or Mg/MeOH for N-deprotection led to similar
results in terms of product yield, yet after a longer period of time. For
the use of Zn/HCl for picolinamide deprotection, see: D. H. O’ Donovan,
C. De Fusco, D. R. Spring, Tetrahedron 2016, 57, 2962.
Daugulis’s group pioneered the use of 2-picolinamide and 8-
quinolinamide as directing groups in metal-catalyzed functionalization
of inert C−H bonds: a) V. G. Zaitsev, D. Shabashov, O. Daugulis, J. Am.
Chem. Soc. 2005, 127, 13154. For reviews on directing group strategy
in CH functionalization: b) G. Rousseau, B. Breit, Angew. Chem. Int.
Ed. 2011, 50, 2450; c) G. Rouquet, N. Chatani, Angew. Chem. Int. Ed.
2013, 52, 11726; d) F. Zhang, D. R. Spring, Chem. Soc. Rev. 2014, 43,
6906; e) M. Zhang, Y. Zhang, X. Jie, H. Zhao, G. Li, W. Su, Org. Chem.
[15] a) J. R. Brooks, D. N. Harcoutt, R. D.Waigh, J. Chem. Soc. Perkin
Trans.
1
1973, 2588. See also: b) D. A. Claremon, J.
Hirshfield, P. K. Lumma, D. E. Mcclure, J. P. Springer Synthesis
1986, 144.
[16] A. M. Martínez, J. Echavarren, I. Alonso, N. Rodríguez, R. Gómez-
Arrayás, J. C. Carretero, Chem. Sci. 2015, 6, 5802.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
[17] See, for example: a) D. R. Stuart, M. Bertrand-Laperle, K. M. N.
Burgess, K. Fagnou, J. Am. Chem. Soc. 2008, 130, 16474; b) N.
Guimond, K. Fagnou, J. Am. Chem. Soc. 2009, 131, 12050; c) L.
Ackermann, A. V. Lygin, N. Hofmann, Angew. Chem. Int. Ed. 2011, 50,
6379. For exceptions to this observation, see: d) N. Guimond, S. I.
Gorelsky, K. Fagnou, J. Am. Chem. Soc. 2011, 133, 6449; e) H. Wang,
C. Grohmann, C. Nimphius, F. Glorius, J. Am. Chem. Soc. 2012, 134,
19592.
order in substrate in Pd-catalysis: c) A. C. Ferretti, C. Brennan, D. G.
Blackmond, Inorg. Chim. Acta 2011, 369, 292; d) R. D. Baxter, D. Sale,
K. M. Engle, J.-Q. Yu, D. G. Blackmond, J. Am. Chem. Soc. 2012, 134,
4600.
[21] For the oxidation of cobalt(II) by air (dissolved in solution) during the
formation of related tris-chelate complexes of cobalt(III), see: N.
Gunasekaran, P. Jerome, S. W. Ng, E. R. T. Tiekink, R. Karvembu, J.
Mol. Catal. A 2012, 353–354, 156. See also ref. 22a.
[18] A similar outcome was also observed using CD3CO2D as solvent (see
SI for further details).
[22] a) S. Maity, R. Kancherla, U. Dhawa, E. Hoque, S. Pimparkar, D. Maiti,
ACS Catal. 2016, 6, 5493; b) J. Zhang, H. Chen, C. Lin, Z. Liu, C.
Wang, Y. Zhang, J. Am. Chem. Soc. 2015, 137, 12990; c) O. Planas, C.
J. Whiteoak, V. Martin-Diaconescu, I. Gamba, J. M. Luis, T. Parella, A.
Company, X. Ribas, J. Am. Chem. Soc. 2016, 138, 14388; d) X.-K Guo,
L.-B. Zhang, D. Wei, J.-L. Niu, Chem. Sci. 2015, 6, 7059.
[19] These kinetic studies were performed in accordance with the protocols
reported by Blackmond and Burés: a) D. G. Blackmond, Angew. Chem.
Int. Ed. 2005, 44, 4302; b) D. G. Blackmond, J. Am. Chem. Soc. 2015,
137, 10866; c) J. Burés, Angew. Chem. Int. Ed. 2016, 55, 2028.
[20] a) S. Maity, R. Kancherla, U. Dhawa, E. Hoque, S. Pimparkar, D. Maiti,
ACS Catal. 2016, 6, 5493; b) S. Nakanowatari, R. Mei, M. Feldt,
L .Ackermann, ACS Catal. 2017, 7, 2511. For examples on negative
This article is protected by copyright. All rights reserved.

10.1002/chem.201702283
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Ángel Manu Martínez, Nuria Rodríguez,*
Ramón Gómez-Arrayás,* and Juan C.
Carretero*
Page No. – Page No.
Cobalt-Catalyzed ortho-CH
Functionalization/Alkyne Annulation
of Benzylamine Derivatives: Access
to Dihydroisoquinolines
Text for Table of Contents
Molecular O2 as the sole oxidant and simple Co(OAc)2 as precatalyst are key features of a practical
procedure for the CH alkenylation/annulation of benzylamine derivatives with both internal and
terminal alkynes, thereby enabling the access to dihydroisoquinolines, including chiral, non racemic
derivatives. Valuable insights about the reaction course and presumed Co-species involved in the
catalytic cycle are provided.
This article is protected by copyright. All rights reserved.