Unravelling the mechanism of cobalt-catalysed remote C-H nitration of 8-aminoquinolinamides and expansion of substrate scope towards 1-naphthylpicolinamide

Article abstract of DOI:10.1039/c9sc05076k

Previously, an unexpected Co-catalysed remote C-H nitration of 8-aminoquinolinamide compounds was developed. This report provided a novel reactivity for Co which was assumed to proceed through the mechanistic pathway already known for analogous Cu-catalysed remote couplings of the same substrates. In order to shed light into this intriguing, and previously unobserved reactivity for Co, a thorough computational study has now been performed, which has allowed for a full understanding of the operative mechanism. This study demonstrates that the Co-catalysed remote coupling does not occur through the previously proposed Single Electron Transfer (SET) mechanism, but actually operates through a high-spin induced remote radical coupling mechanism, through a key intermediate with significant proportion of spin density at the 5- and 7-positions of the aminoquinoline ring. Additionally, new experimental data provides expansion of the synthetic utility of the original nitration procedure towards 1-naphthylpicolinamide which unexpectedly appears to operate via a subtly different mechanism despite having a similar chelate environment.

Full text of DOI:10.1039/c9sc05076k

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
Unravelling the mechanism of cobalt-catalysed
remote C–H nitration of 8-aminoquinolinamides
and expansion of substrate scope towards 1-
naphthylpicolinamide†
Cite this: Chem. Sci., 2020, 11, 534
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
b
b
a
Melody Chu,a Oriol Planas,b Anna Company,
Xavi Ribas,
Alex Hamilton
*
a
*
and Christopher J. Whiteoak
Previously, an unexpected Co-catalysed remote C–H nitration of 8-aminoquinolinamide compounds was
developed. This report provided a novel reactivity for Co which was assumed to proceed through the
mechanistic pathway already known for analogous Cu-catalysed remote couplings of the same
substrates. In order to shed light into this intriguing, and previously unobserved reactivity for Co,
a thorough computational study has now been performed, which has allowed for a full understanding of
the operative mechanism. This study demonstrates that the Co-catalysed remote coupling does not
occur through the previously proposed Single Electron Transfer (SET) mechanism, but actually operates
through a high-spin induced remote radical coupling mechanism, through a key intermediate with
significant proportion of spin density at the 5- and 7-positions of the aminoquinoline ring. Additionally,
new experimental data provides expansion of the synthetic utility of the original nitration procedure
Received 8th October 2019
Accepted 16th November 2019
DOI: 10.1039/c9sc05076k
towards 1-naphthylpicolinamide which unexpectedly appears to operate via
a subtly different
rsc.li/chemical-science
mechanism despite having a similar chelate environment.
approaches are very common when employing second and
third-row transition metals in C–H functionalisation. In
contrast, the chemistries of rst-row transition metals are more
diverse than second- and third-row analogues, which provides
an exciting potential opportunity to discover and develop novel
protocols outside of those using traditional CMD and OA steps
(e.g. radical coupling approaches).5 In the case of late stage C–H
bond functionalisation, these unique reactivities could be
applied to provide access to previously inaccessible analogues
of biologically active compounds through new innovative C–C
and C–X bond forming protocols.
Several rst-row transition metals have been shown to
provide reactivities amenable for application in C–H function-
alisation protocols. However, the already known rich mecha-
nistic diversity of Co in the eld of C–H functionalisation makes
this metal stand out as an exciting candidate for the basis of
novel protocols.6,7
Introduction
The controlled and selective functionalisation of C–H bonds is
currently a topic which is attracting a signicant amount of
interest from the chemical community.1 When successfully
applied, C–H functionalisation protocols are generally used to
either reduce the length of complex multi-step syntheses or to
provide rapid access to new derivatives of biologically active
molecules as a late stage functionalisation tool.2 The challenge
in the latter application is to regiochemically functionalise
a complex compound in a predictable and controlled fashion,
whereby there are oen a number of other reactive functional
groups present. Currently, a majority of the developed C–H
functionalisation protocols utilise a chelation directing group
strategy for control of product selectivity3 and well established/
understood Concerted Metalation Deprotonation (CMD) or
Oxidative Addition (OA) steps.4 These key mechanistic
Functionalisation of quinolines beyond traditional C2
modication has recently become a topic of increasing
interest8 as quinoline is an important heterocycle found in
many biologically active compounds. For example, the 8-
aminoquinoline scaffold is the basis of a number of anti-
malarial drugs9 and facile modication as a late stage func-
tionalisation tool may provide access to new potent drugs
which could have signicant impact on one of the world's
major causes of death.10
aDepartment of Biosciences and Chemistry, Sheffield Hallam University, Sheffield, S1
1WB, UK. E-mail: c.whiteoak@shu.ac.uk; a.hamilton@shu.ac.uk
b
´
´
`
Departament de Quımica, Grup de Quımica Bioinspirada, Supramolecular i Catalisi
´
`
(QBIS-CAT), Institut de Quımica Computacional i Catalisi (IQCC), Universitat de
Girona, Campus de Montilivi, 17071 Girona, Catalonia, Spain
† Electronic supplementary information (ESI) available: Experimental;
experimental protocols, characterization data, and NMR spectra of all new
compounds. DFT studies; details of methods used and coordinates for all
complexes. See DOI: 10.1039/c9sc05076k
534 | Chem. Sci., 2020, 11, 534–542
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
One novel approach to the functionalisation of 8-amino- and d).17 The lack of understanding of this new Co-catalysed
quinolinamides has been the remote (non-proximate) C–H approach has now inspired us to examine the mechanism in
bond chlorination catalysed by Cu at the C7 position, which was more detail utilising the open shell DLPNO-CCSD(T) (Domain-
originally reported by Stahl/Ertem and co-workers in 2013 based Local Pair-Natural Orbital Coupled Cluster approxima-
(Scheme 1a).11 Non-proximate C–H bond functionalisation, tions) method, in order that the unique reactivity of Co can be
therefore moving beyond directing group based protocols, is fully understood and as a result be further rationally exploited
a major challenge in synthetic chemistry with only a limited, but by others.
increasing, number of approaches currently available.12 This
Herein, we describe our ndings and also demonstrate that
initial work by Stahl/Ertem and co-workers demonstrated the the protocol can also be successfully transferred to the remote
rst example of a metal-catalysed remote C–H functionalisation nitration of 1-naphthylpicolinamide analogues which replace
of 8-aminoquinolinamides and Density Functional Theory the aminoquinolinamide chelate environment with a picolina-
(DFT) calculations were employed to elucidate the mechanism, mide, thus further expanding the applicability of the previously
which was found to be based on a key Single Electron Transfer discovered protocol.
(SET) step. This seminal work has been the inspiration for the
development of an ever-expanding number of practical Cu-
Results and discussion
catalysed remote C–H functionalisation protocols of the same
substrate.13 Meanwhile, other metals have also been employed
The initial approach in this study was to identify the nature of
as catalysts for remote C–H functionalisation of these 8-ami-
the starting “Co(II) source” in the mechanism. Given the use of
noquinolinamides, although to a signicantly reduced extent.14
AcOH as solvent, the experimental conditions would allow for
Resulting from the rich aforementioned chemistry of Co, in
extensive counter-ion ligand exchange, leading to a number of
2016 some of us reported on the unexpected remote nitration of
potential Co(II) species. Relative free energies for ligand
8-aminoquinolinamide scaffold using tert-butyl nitrite (TBN) as
exchange (Table 1) identied Co(OAc)2 as the most thermody-
nitro source (Scheme 1b).15,16 This was the rst time that Co had
namically stable Co species. This compound was therefore used
been observed to display this type of reactivity. Since this report,
as the basis for the active catalyst species for the remainder of
others have also been attracted to the potential of this novel
potential for Co in the eld of C–H functionalisation (Scheme 1c
the reaction mechanism. It should be noted at this point that
the observed benet of using Co(NO3)2 over Co(OAc)2 in the
Scheme 1 (a) Cu-catalysed remote chlorination reported by Stahl. (b) First Co-catalysed remote nitration reported by Whiteoak and Ribas. (c)
Co-catalysed remote perfluoroalkylation reported by Niu and Song. (d) Co-catalysed remote sulfonylation reported by Xia.
This journal is © The Royal Society of Chemistry 2020
Chem. Sci., 2020, 11, 534–542 | 535

View Article Online
Chemical Science
Edge Article
original protocol15 is likely a result of the background nitration assignment of 1Int 1$L1 and 3Int 1$L1 being Co(III) whereas 5Int
of the substrate through traditional nitration utilising in situ 1$L1 being high spin Co(II) with an associated radical ligand,
formed HNO3 in an acidic environment.
based on the spin density analysis.
Aer this initial study on catalyst species we embarked on
The transition states for the remote NO2 radical coupling at
full elucidation of the reaction mechanism. As a result of the the 5- and 7-positions starting from 5Int 1$L1 lie at 4.7 and
moderate to large size of the catalytic system being studied, DFT 5.2 kcal molꢀ1 respectively (Fig. 2). This low barrier height is in
calculations would traditionally be the method of choice. good agreement with the protocol being operative under
However, due to the potential for different spin states to be ambient conditions. Additionally, the nitration step is also the
involved in the mechanism, which pose a signicant challenge regioselective transition state with a DDG‡298 K of 0.5 kcal molꢀ1
for DFT based methods, the recently implemented open shell in favour of the observed 5-substituted product. This difference
DLPNO-CCSD(T) method was utilised.18,19 This near linear results in a calculated regioselectivity of 60%, which is in
scaling ab initio electronic structure method has the potential reasonably good agreement with the experimentally observed
for coupled cluster accuracy for large, synthetically relevant 3 : 1 product ratio. Calculation of the low spin nitration reac-
catalysts. Recent work by Chen and co-worker on the same tion, via 2TS 1–2$LA/B
, showed it to be approximately
1
catalyst, but focusing on a different C–H activation protocol, 17 kcal molꢀ1 less favourable than the proposed high spin
highlighted the importance of these methods due to the mechanism shown in Fig. 2. This substantial stabilisation of the
complicated multi-state reactivity which has been previously high-spin mechanism further strengthens the proposed remote
reported.20 In this report Multi-State Reactivity (MSR) was found radical coupling mechanism via the quintet state. There has
to be an intriguing feature of the chemistry of high-spin Co(III), been discussion in the literature regarding the potential over-
where this report revealed the highly complex mechanisms stabilisation of high spin complexes with DLPNO-CCSD(T)
operative in Co-catalyzed C–H activation processes.
methods,23 however the degree of over-stabilisation is in the
The rst step of the reaction is hydrogen abstraction by the in order of 5 kcal molꢀ1 at most and therefore as a result we are
t
situ formed BuOc radical leading to the radical of the 8-ami- condent this reaction proceeds via the proposed high spin
noquinolinamide ligand, L1. Coupling L1 to the [Co(AcO)2] pre- pathway.
catalyst forms the initial catalytic species Int1$L1. This species
To ensure the previously proposed intramolecular Singlet
can exist in three possible spin states; low spin singlet (1Int 1, S Electron Transfers (SET) mechanism was not involved in the
¼ 0), medium spin triplet (3Int 1, S ¼ 1) and high spin quintet reaction, LR-TD-DFT (Linear Response Time Dependent-
(5Int 1, S ¼ 2). With the DLPNO-CCSD(T) method the high spin Density Functional Theory) calculations were performed on
quintet state (5Int 1$L1) proved to be the ground-state species, 1Int 1$L1. The rst excited state transition, corresponding to the
which is in full agreement with previous work from Chen and excitation of an electron from the 8-aminoquinolinamide
co-worker.20 Examination of the electronic structure of the three ligand to the Co(III) centre, is found to be in the range of 37–
potential spin state complexes leads to the conclusion that both 43 kcal molꢀ1, depending on the choice of DFT functional and
3
5
1Int 1$L1 and Int 1$L1 are Co(III) complexes whereas the Int amount of HF exchange included (see ESI for full details†). The
1$L1 is actually a high-spin quartet Co(II) species with the radical transition energy is similar to that reported by Stahl/Ertem and
still residing on the quinoline moiety of the ligand. Cumulative co-workers for the related Cu(II) complex.11 Optimisation of the
spin densities on the aminoquinoline ligand, calculated using corresponding excited state structure, 1Int 1ES$L1, with the
1
3
5
DLPNO-CCSD/def2-tzvp,21 for Int 1$L1, Int 1$L1 and Int 1$L1 hybrid functional PBE0 provides a value of 24.4 kcal molꢀ1
are 0, 0.18 and 0.72, respectively. Fig. 1 highlights the increased above Int 1$L1. Although not directly comparable due to the
1
electron spin density residing at 5- and 7-positions, showing the differing methodologies used, both the transition energy and
5
potential sites for radical addition to Int 1$L1. A similar spin the optimised excited state complex is signicantly higher in
population distribution was previously observed by Chen and
co-worker leading to the description as a quartet Co(II) ferro-
magnetically coupled with a ligand radical.20
Topological analysis using Quantum Theory of Atom in
Molecules (QTAIM)22 and different bond order descriptors
(Table 2) for the three spin state structures highlight a signi-
cant decrease in bonding/interaction between the Co centre and
the 8-aminoquinolinamide ligand on progression from singlet
to quintet state. This is in further agreement with the
Table 1 Relative free energy, DG298 kcal molꢀ1, for Co(II) species
Co(II) species
Relative free energy, DG298 (kcal molꢀ1
)
Fig. 1 (Left) Molecular structure with calculated spin densities for 5Int
1$L1 obtained using DLPNO-CCSD/def2-tzvp highlighting the spin
accumulated at the 5- and 7-positions on the 8-aminoquinolinamide
ligand. (Right) QTAIM wavefunction analysis of spin density gradients
and bond critical points, focusing on the 8-aminoquinoline moiety.
Co(NO3)2
Co(NO3) (OAc)
Co(OAc)2
0.00
ꢀ20.21
ꢀ48.34
536 | Chem. Sci., 2020, 11, 534–542
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
Table 2 Bonding and QTAIM parameters for different spin-states of Int1$L1a
1Int1$L1
Co–N(1)
3Int1$L1
5Int1$L1
Co–N(1)
Bonding parameter
Co–N(2)
Co–N(1)
Co–N(2)
Co–N(2)
r
r
1.948
0.104
0.408
ꢀ0.030
1.053
0.736
0.369
1.894
0.113
0.563
ꢀ0.031
1.010
0.651
0.515
1.878
0.125
0.441
ꢀ0.045
1.183
0.777
0.450
1.925
0.108
0.474
ꢀ0.032
0.982
0.626
0.454
1.985
0.087
0.477
ꢀ0.020
0.901
0.434
0.343
2.073
0.072
0.356
ꢀ0.014
0.784
0.399
0.268
V2r
H(r)
FBO
MBO
LBO
a
2
˚
Bonding parameters: Co–N(1)amide Co–N(2)pyrrole: distance r in A, QTAIM parameters in a.u.; r electron density, V r laplacian of electron density,
H(r) local energy density. FBO ¼ Fuzzy Bond Order,24 MBO ¼ Mayer Bond Order,25 LBO ¼ Laplacian Bond Order.26
energy than the transition state on the proposed high-spin 3.8 kcal molꢀ1 more stable than the corresponding 7-
mechanism starting from the quintet ground state, Int 1$L1. substituted product (4Int 2$LB1), and 16.6 kcal molꢀ1 more
5
As a result of the requirement for a coupling transition state stable than the equivalent low spin doublet mechanism (2Int
aer 1Int 1ES$L1, this precludes this mechanism on the grounds 2$LA1). A number of different pathways for the deprotonation of
of it being energetically unfeasible.
4Int 2$L1 at the 5- and 7-positions were explored. Addition of
Continuing with the high-spin mechanism, the nitration of [Co(AcO)3]ꢀ to 4Int 2$L1 forms 5Int 3$L1, this step is exergonic by
5Int 1$L1 results in the formation of the intermediate complex, 9.7 kcal molꢀ1. Cobalt assisted deprotonation by acetate anion
4Int 2$L1. This intermediate is 8.0 kcal molꢀ1 lower in energy has a barrier of 9.8 kcal molꢀ1, leading to the nal intermediate
than the preceding starting complex. The lower energy pathway [4Int 4$L1]ꢀ, which undergoes proto-demetallation to form the
leading to nitration at the 5-position forms Int 2$LA1 which is observed product and regenerates the catalyst, completing the
4
Fig. 2 Solvated free energy surface (DG298) for the nitration of 8-aminoquinolinamide calculated with DLPNO-CCSD(T)/def2-TZVP. Black (A)
and blue (B) lines represent high spin nitration at the 5 and 7 positions respectively. Dark grey represents first excited state optimised geometry of
Int 1$L1. Light grey represents the corresponding low spin pathway. Calculated geometries shown are for the observed 5-position nitration via the
high spin mechanism.
This journal is © The Royal Society of Chemistry 2020
Chem. Sci., 2020, 11, 534–542 | 537

View Article Online
Chemical Science
Edge Article
catalytic cycle. Attempts to optimise the transition state for abstraction from the amide by the tBuOc radical generated from
deprotonation by an isolated acetate anion led directly to [4Int in situ decomposition of the TBN. This results in a high spin
4$L1]ꢀ, suggesting this pathway could also be occurring, but no Co(II) species with partially delocalised radical character on the
energetic barrier is reported. Due to the reaction conditions 5- and 7-positions of the 8-aminoquinolinamide moiety which
t
allowing for the formation of BuOc (eqn (1)), a hydrogen atom then reacts with the nitrogen dioxide (NO2) formed from the
abstraction pathway was also explored. Formation of the adduct decomposition of TBN (eqn (1) and (2)).27 Deprotonation of the
4
complex Int 2$LA1–tBuOc is endergonic by 2.9 kcal molꢀ1 and intermediate coupling product with the in situ formed
5
1
therefore 12.2 kcal molꢀ1 less favourable than In t3$LA and [Co(OAc)3]ꢀ furnishes the chelate species of the nal product,
2.4 kcal molꢀ1 less favourable than 5TS 3–4$L1. Based on these which can then undergo proto-demetallation by acetic acid to
results we propose cobalt assisted acetate deprotonation is the reform the initial Co(OAc)2 catalyst, releasing the nal nitrated
active pathway (Fig. 2). Nitration at the 7-position (route B) is 8-aminoquinolinamide product.
consistently between 3 and 8 kcal molꢀ1 less favourable
throughout the entire mechanism, and in good agreement with to expanding the scope of this novel reaction mechanism.
With this understanding in hand, we turned our attention
the experimental observations for regioselectivity.
Rather than focus on addition of different coupling partners at
the 5- and 7-positions of the 8-aminoquinolinamide substrate,
something which has already been exemplied by both Xia
and co-workers17a and Niu and Song and co-workers,17b we
choose to focus on selective nitration of substrates that have
proven more challenging by more traditional synthetic
methods. In an effort to direct experiments a new substrate
was rationally proposed which would facilitate selective 1-
aminonaphthyl nitration while maintaining the key bi-dentate
nitrogen chelation. 1-Naphthylpicolinamide, L2, (Scheme 3a)
was chosen to ensure formation of the proposed required Co
t
tBuONO / BuOc + NO
(1)
(2)
1
NO þ O2/NO2
2
As a result of this new in-depth study, in adjustment to our
previously proposed SET mechanism for the Co-catalysed
remote nitration of 8-aminoquinolamide substrates, we now
propose the mechanism shown in Scheme 2. In the rst step,
the Co(II) coordination to substrate is promoted by Hc
Scheme 2 Proposed mechanism for the remote nitration of 8-aminoquinolinamide substrates catalysed by Co(OAc)2 at the 5-position based on
the computational studies in from this work. Note: the NO2 is formed in situ from the decomposition of TBN (see eqn (1) and (2)).
538 | Chem. Sci., 2020, 11, 534–542
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
Scheme 3 (a) 1-Napththylpicolinamide substrates used in the experimental work; nitration at either the 2- or 4-position for the 1-napth-
thylpicolinamide. (b) Products isolated from the Co-catalysed remote nitration of 1-napththylpicolinamide substrates; reaction conditions:
0.50 mmol 1-napththylpicolinamide substrate, 20 mol% Co(NO3)2$6H2O, 4.0 equivalents TBN, 3.5 mL AcOH, r.t., 18 h. a6.0 equivalents of TBN.
bOnly the 4-substituted product could be successfully isolated, and the yield shown is of this regioisomer. (c) Examples of product upgrading of
the nitrated compound obtained from 1-naphthylpicolinamide. (d) Examples of product upgrading of the nitrated compound obtained from 4-
chloro-substituted 1-naphthylpicolinamide substrate.
chelate intermediate. Indeed, Cu examples of remote func- equivalents.15 The lower reactivity of electron-withdrawing
tionalisation of these substrates are known and have previ- substituted substrates was previously observed with the 8-ami-
ously been reported.28
noquinolinamide
As with the 8-aminoquinolinamide substrate, it was possible thylpicolinamide substrate with no substituents could also be
to convert number of differently substituted 1-naph- nitrated, although despite our best efforts it was not possible to
substrates.15
The
parent
1-naph-
a
thylpicolinamide substrates to the corresponding nitrated successfully isolate the 2-nitration product. As a result of this,
products as shown in Scheme 3b. The electron-withdrawing 4- the reported yield therefore corresponds to the 4-nitration
chloro-substituted 1-naphthylpicolinamide, provided clean product only, which was successfully obtained as an analytically
nitration at the 2-position. It should be noted that in order to pure sample. Finally, the electron-donating methoxy-
get an almost quantitative yield (97%), it was necessary to substituted substrate could also be cleanly converted to the
increase the TBN loading from 4.0 equivalents to 6.0 corresponding 2-nitrated product quantitatively.
This journal is © The Royal Society of Chemistry 2020
Chem. Sci., 2020, 11, 534–542 | 539

View Article Online
Chemical Science
Edge Article
Fig. 4 (Left) Molecular structure with calculated spin densities for 5Int
1$L2 obtained using DLPNO-CCSD/def2-tzvp highlighting the spin
accumulated at the 2- and 4-positions on the 1-aminonaphthyl
moiety. (Right) QTAIM wavefunction analysis of spin density gradients,
focusing on the 1-aminonaphthyl moiety.
can then proceed via either a low spin (S ¼ 1/2) or high spin (S ¼
3/2) pathway. As with the 8-aminoquiolinamide mechanism the
high spin pathway, through 4TS 1-2$L2, is signicantly more
favourable. This suggests a Multi-State Reactivity (MSR) mech-
anism with crossing between potential energy surfaces,29 from
Fig. 3 Solvated free energy surface (DG298) for the nitration of 1-
naphthylpicolinamide calculated with DLPNO-CCSD(T)/def2-TZVP.
Black (A) and blue (B) lines represent high spin nitration at the 4 and 2
positions respectively. Dark grey represents first excited state opti-
mised geometry of Intermediate 1. Light grey represents the corre-
sponding low spin pathways. Calculated geometries shown are for the low to high spin, as the nitration step proceeds in order to
observed 4-position nitration via the high spin mechanism, as well as
the intermediate for the 2-position nitration.
facilitate the remote radical coupling. The high spin products of
the nitration step, 4Int 2$LA/B2, are approximately 25 kcal molꢀ1
more energetically stable compared to the equivalent low spin
intermediate, Int 2$LA/B2. Interestingly the observed substitu-
2
In order to exemplify the potential for these compounds for
tion at the 4-position is kinetically favoured, while the 2-posi-
further utility, the 1-naphthylpicolinamide with nitration at the
tion is the thermodynamic intermediate. This can be accounted
4-position was initially reduced to the corresponding amine
using Pd/C and H2 (Scheme 3c). Additionally, the picolinamide
bond could be hydrolysed under acid conditions to furnish 1-
amino-4-nitronaphthalene in good yield (85%) and thereaer,
the 1,4-diaminonaphthalene was prepared in good yield
through reduction of this nitro-product using Pd/C and
hydrogen (78%; Scheme 3c). Finally, a highly substituted
product containing synthetically useful chloro and amine
functionalities again obtained through a facile Pd/C and H2
reduction of the nitration product of the 4-chloro-substituted 1-
naphthylpicolinamide could be easily prepared (Scheme 3d).
With successful transfer of the original nitration reaction to
the 1-naphthylpicolinamide substrates we wondered if it fol-
lowed the same reaction mechanism. Initially it was assumed
that the 1-naphthylpicolinamide substrate would follow
a similar reaction path to the 8-aminoquinolinamide example.
In contrast to the 8-aminoquinalineamide substrate, the
DLPNO-CCSD(T) calculations suggest the ground state of the 1-
naphthylpicolinamide complex is the singlet (S ¼ 0) with the
quintet (S ¼ 2) and triplet (S ¼ 1) states being 2.2 and
3.7 kcal molꢀ1 higher in energy respectively (Fig. 3). TDDFT
calculations on 1Int 1$L2 gives the rst excited state transition,
corresponding to the SET mechanism, of 37.7 kcal molꢀ1. An
˚
for by the additional long range hydrogen bonding (2.45 A)
between the hydrogen at the 2-position of the nitrated naphthyl
moiety and the Co bound acetate. The calculated higher energy
barriers for 1-naphthylpicolinamide compared to 8-amino-
quinolinamide agrees with the experimental requirement for
more forcing reaction conditions (increase in number of
equivalents of TBN in the case of 4-chloro-substituted 1-naph-
thylpicolinamide). Increased spin density on the Co and
decreased spin density on the 1-aminonaphthyl moiety further
support this decrease in reactivity (Fig. 4).
Conclusions
In summary, we have studied the mechanism of the remote
nitration of 8-aminoquinolinamides catalysed by Co using the
robust DLPNO-CCSD(T) methodology and found there to be
a signicant difference in mechanism to the well-studied
analogous Cu-catalysed remote functionalisation protocols of
8-aminoquinolinamides. The results reported within this work
identify an unusual mechanism which does not pass through
the initially proposed SET process, but instead, the reaction
takes place via a High-Spin Remote Induced Radical Coupling,
providing an alternative route for the NO2 to couple compared
with a traditional SET process. The high spin reactivity has been
successfully transferred to 1-naphthylpicolinamide substrates,
further enhancing the applicability of the original protocol,
where the computational study has demonstrated that a subtly
different mechanism is operative (Multi-State Reactivity).
Overall, the distinct mechanistic pathways shown in this work
1
equivalent to the barrier to this was seen for Int 1$L1.
Thereaer spin density analysis of 5Int 1$L2 showed
a signicant proportion of density at the observed reactive 2-
and 4-positions of the 1-aminonaphthyl moiety (Fig. 4), again
highlighting the potential for remote radical coupling. Nitration
at the 2- and 4-position of the 1-naphthylpicolinamide substrate
540 | Chem. Sci., 2020, 11, 534–542
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
provide a complimentary tool for the remote C–H functionali-
sation of both 8-aminoquinolinamides and 1-naph-
thylpicolinamides. This new mechanistic understanding
9 (a) S. Vandekerckhove and M. D'Hooghe, Bioorg. Med. Chem.,
2015, 23, 5098–5119; (b) X. Nqoro, N. Tobeka and
B. A. Aderibigbe, Molecules, 2017, 22, 2268.
should now inspire others to develop new synthetic protocols 10 R. Sharma, S. Patil and P. Maurya, SAR QSAR Environ. Res.,
for remote functionalisation using Co-catalysis.
2013, 25, 189–203.
11 A. M. Suess, M. Z. Ertem, C. J. Cramer and S. S. Stahl, J. Am.
Chem. Soc., 2013, 135, 9797–9804.
12 For some innovative examples see: (a) Z. Zhang, K. Tanaka
and J.-Q. Yu, Nature, 2017, 543, 538–542; (b) J. Luo,
S. Preciado and I. Larrosa, J. Am. Chem. Soc., 2014, 136,
4109–4112; (c) O. Saidi, J. Marae, A. E. W. Ledger,
Conflicts of interest
There are no conicts to declare.
Acknowledgements
¨
P. M. Liu, M. F. Mahon, G. Kociok-Kohn, M. K. Whittlesey
A. H., C. J. W. and M. C. would like to thank the Department of
Biosciences and Chemistry/Biomolecular Sciences Research
Centre (BMRC) at Sheffield Hallam University for funding. X. R.
and A. C. thank the Spanish MICINN (CTQ2016-77989-P), and
and C. G. Frost, J. Am. Chem. Soc., 2011, 133, 19298–19301;
(d) N. Hofmann and L. Ackermann, J. Am. Chem. Soc.,
2013, 135, 5877–5884; (e) C. Yuan, L. Zhu, R. Zeng, Y. Lan
and Y. Zhao, Angew. Chem., Int. Ed., 2018, 57, 1277–1281.
the Catalan DIUE of the Generalitat de Catalunya (2017SGR264). 13 For some representative examples see: (a) X.-X. Liu, Z.-Y. Wu,
`
X. R. thanks ICREA for an ICREA Academia award. C. J. W. and
X.-L. Luo, Y.-Q. He, X.-Q. Zhou, Y.-X. Fan and G.-S. Huang,
RSC Adv., 2016, 6, 71485–71488; (b) J. Xu, C. Shen, X. Zhu,
P. Zhang, M. J. Ajitha, K.-W. Huang, Z. An and X. Liu,
Chem.–Asian J., 2016, 11, 882–892; (c) C. Xia, K. Wang,
J. Xu, Z. Wei, C. Shen, G. Duan, Q. Zhu and P. Zhang, RSC
Adv., 2016, 6, 37173–37179; (d) X. Zhu, L. Qiao, P. Ye,
B. Ying, J. Xu, C. Shen and P. Zhang, RSC Adv., 2016, 6,
89979–89983; (e) Y. Dou, Z. Xie, Z. Sun, H. Fang, C. Shen,
P. Zhang and Q. Zhu, ChemCatChem, 2016, 8, 3570–3574;
(f) C. Xia, K. Wang, J. Xu, C. Shen, D. Sun, H. Li, G. Wang
and P. Zhang, Org. Biomol. Chem., 2017, 15, 531–535; (g)
L. Zhu, R. Qiu, X. Cao, S. Xiao, X. Xu, C.-T. Au and
S.-F. Yin, Org. Lett., 2015, 17, 5528–5531; (h) H. Saloo,
A. Mandal, J. Selvakumarm and M. Baidya, Eur. J. Org.
Chem., 2016, 4321–4327; (i) H. Li, H. Chen, C. Bian,
Z. Tang, A. K. Singh, X. Qi, X. Yue, Y. Lan, J.-F. Lee and
A. Lei, Chem. Commun., 2017, 53, 6736–6739; (j) S. Han,
A. Liang, X. Ren, X. Gao, J. Li, D. Zou, Y. Wu and Y. Wu,
Tetrahedron Lett., 2017, 58, 4859–4863.
X. R. also thank COST Action CHAOS (CA15106). O. P. thanks
the Spanish MECD for a pre-doctoral fellowship (FPU13-04099).
We thank IQCC for granting access to their computing cluster.
Notes and references
1 J. J. Li, C–H Bond Activation in Organic Synthesis, CRC Press,
Boca Raton, 1st edn, 2017.
2 For an overview see: (a) D. C. Blakemore, L. Castro,
I. Churcher, D. C. Rees, A. W. Thomas, D. M. Wilson and
A. Wood, Nat. Chem., 2018, 10, 383–394; (b) J. Wencel-
Delord and F. Glorius, Nat. Chem., 2013, 5, 369–375.
3 For an overview of directing groups in metal-catalysed C–H
functionalisations, see: (a) C. Sambiagio, D. Schonbauer,
R. Blieck, T. Dao-Huy, G. Pototschnig, P. Schaaf,
T. Wiesinger, M. F. Zia, J. Wencel-Delord, T. Besset,
¨
¨
B. U. W. Maes and M. Schnurch, Chem. Soc. Rev., 2018, 47,
6603–6743; (b) M. Zhang, Y. Zhang, X. Jie, H. Zhao, G. Li
and W. Su, Org. Chem. Front., 2014, 1, 843–895; (c) 14 For a comprehensive overview of recent advances in metal
M. Corbet and F. De Campo, Angew. Chem., Int. Ed., 2013,
52, 9896–9898.
4 For an overview of mechanisms of C–H bond activation see:
catalysed remote C–H functionalisation's of 8-
aminoquinolinamides see: B. Khan, H. S. Dutta and
D. Koley, Asian J. Org. Chem., 2018, 7, 1270–1297.
A. E. Shilov and G. B. Shul'pin, Chem. Rev., 1997, 98, 2879– 15 C. J. Whiteoak, O. Planas, A. Company and X. Ribas, Adv.
2932.
Synth. Catal., 2016, 358, 1679–1688.
5 H. Yi, G. Zhang, H. Wang, Z. Huang, J. Wang, A. K. Singh and 16 tert-Butyl nitrite has found a variety of uses in organic
A. Lei, Chem. Rev., 2017, 117, 9016–9085.
6 For reviews of Co-catalysed C–H functionalisation protocols
see: (a) G. Pototschnig, N. Maulide and M. Schnurch,
chemistry and has recently been well reviewed: A. Dahiya,
A. K. Sahoo, T. Alam and B. K. Patel, Chem.–Asian J., DOI:
10.1002/asia.201901072.
¨
Chem.–Eur. J., 2017, 23, 9206–9232; (b) W. Liu and 17 (a) K. Wang, G. Wang, G. Duan and C. Xia, RSC Adv., 2017, 7,
L. Ackermann, ACS Catal., 2016, 6, 3743–3752; (c)
P. G. Chirila and C. J. Whiteoak, Dalton Trans., 2017, 46,
9721; (d) T. Yoshino and S. Matsunaga, Adv. Synth. Catal.,
51313–51317; (b) J.-F. Suo, X.-M. Zhao, K.-X. Zhang,
S.-L. Zhou, J.-L. Niu and M.-P. Song, Synthesis, 2017, 49,
3916–3924.
2017, 359, 1245; (e) S. Wang, S.-Y. Chen and X.-Q. Yu, 18 M. Sparta and F. Neese, Review application of closed shell
Chem. Commun., 2017, 53, 3165.
DLPNO-CCSD(T) methods, Chem. Soc. Rev., 2014, 43, 5032–
7 For an overview of some of the many different mechanisms
5041.
reported for Co-catalysed C–H functionalisation protocols 19 (a) C. Riplinger, P. Pinski, U. Becker, E. F. Valeev and
see: O. Planas, P. G. Chirila, C. J. Whiteoak and X. Ribas,
Adv. Organomet. Chem., 2018, 69, 209–282.
8 T. Iwai and M. Sawamura, ACS Catal., 2015, 5, 5031–5040.
F. Neese, J. Chem. Phys., 2016, 144, 024109; (b) M. Saitow,
U. Becker, C. Riplinger, E. F. Valeev and F. Neese, J. Chem.
Phys., 2017, 146, 164105.
This journal is © The Royal Society of Chemistry 2020
Chem. Sci., 2020, 11, 534–542 | 541

View Article Online
Chemical Science
Edge Article
20 P. Ma and H. Chen, ACS Catal., 2019, 9, 1962–1972.
21 L. Goerigk and S. Grimme, J. Chem. Theory Comput., 2011, 7,
291–309.
22 T. Lu and F. Chen, J. Comput. Chem., 2012, 33, 580–592.
23 Q. M. Phung, C. Martin-Fernandez, J. N. Harvey and
M. Feldt, J. Chem. Theory Comput., 2019, 15, 4297–4304.
24 I. Mayer and P. Salvador, Chem. Phys. Lett., 2004, 383, 368–
375.
B. Yin, P. Zhang and Q. Zhu, Eur. J. Org. Chem., 2018,
4571–4576; (c) G. You, K. Wang, X. Wang, G. Wang, J. Sun,
G. Duan and C. Xia, Org. Lett., 2018, 20, 4005–4009; (d)
H. Zhu, S. Sun, H. Qiao, F. Yang, Y. Wu and Y. Wu, Org.
Lett., 2018, 20, 620–623; (e) J.-M. Li, Y.-H. Wang, Y. Yu,
R.-B. Wu, J. Weng and G. Lu, ACS Catal., 2017, 7, 2661; (f)
P. Bai, S. Sun, Z. Li, H. Qiao, X. Su, Y. Wu and Y. Wu, J.
Org. Chem., 2017, 82, 12119–12127.
25 A. J. Bridgeman, G. Cavigliasso, L. R. Ireland and J. Rothery, 29 For selected examples of multi-state reactivity in other
J. Chem. Soc., Dalton Trans., 2001, 2095–2108.
systems, see: (a) N. Agenet, V. Gandon, K. P. C. Vollhardt,
M. Malacria and C. Aubert, J. Am. Chem. Soc., 2007, 129,
26 T. Lu and F. Chen, J. Phys. Chem. A, 2013, 117, 3100–3108.
27 P. G. Wang, T. B. Cai and N. Taniguchi in Nitric Oxide Donors:
for Pharmaceutical and Biological Application, ed. P. G. Wang,
T. B. Cai and N. Taniguchi, Wiley-VCH, Weinheim, 2005.
28 For some recent examples of the remote functionalisation of
naphthalylamides see: (a) J.-N. Han, C. Du, X. Zhu,
Z.-L. Wang, Y. Zhu, Z.-Y. Chu, J.-L. Niu and M.-P. Song,
Beilstein J. Org. Chem., 2018, 14, 2090–2097; (b) Y. Dong,
¨
8860–8871; (b) D. Schroder, S. Shaik and H. Schwarz, Acc.
Chem. Res., 2000, 33, 139–145; (c) Y. Sun, H. Tang,
K. Chen, L. Hu, J. Yao, S. Shaik and H. Chen, J. Am. Chem.
Soc., 2016, 138, 3715–3730; (d) L. Hu and H. Chen, J. Am.
Chem. Soc., 2017, 139, 15564–15567; (e) Ref. 20 of this
manuscript.
542 | Chem. Sci., 2020, 11, 534–542
This journal is © The Royal Society of Chemistry 2020