Olefin Cyclopropanation by Radical Carbene Transfer Reactions Promoted by Cobalt(II)/Porphyrinates: Active-Metal-Template Synthesis of [2]Rotaxanes

Article abstract of DOI:10.1002/anie.201803934

A Co^II/porphyrinate-based macrocycle in the presence of a 3,5-diphenylpyridine axial ligand functions as an endotopic ligand to direct the assembly of [2]rotaxanes from diazo and styrene half-threads, by radical-carbene-transfer reactions, in excellent 95 % yield. The method reported herein applies the active-metal-template strategy to include radical-type activation of ligands by the metal-template ion during the organometallic process which ultimately yields the mechanical bond. A careful quantitative analysis of the product distribution afforded from the rotaxane self-assembly reaction shows that the Co^II/porphyrinate subunit is still active after formation of the mechanical bond and, upon coordination of an additional diazo half-thread derivative, promotes a novel intercomponent C?H insertion reaction to yield a new rotaxane-like species. This unexpected intercomponent C?H insertion illustrates the distinct reactivity brought to the Co^II/porphyrinate catalyst by the mechanical bond.

Full text of DOI:10.1002/anie.201803934

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Olefin Cyclopropanation via Radical Carbene Transfer Reaction
Promoted by Co(II)porphyrinates for the Active-Metal-Template
Synthesis of [2]Rotaxanes
Authors: Arthur Alcantara, Liniquer Fontana, Vitor Rigolin, Yuri
Andrade, Marcos Ribeiro, Wdeson Barros, Catia Ornelas, and
Jackson D Megiatto
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: Angew. Chem. Int. Ed. 10.1002/anie.201803934
Angew. Chem. 10.1002/ange.201803934
Link to VoR: http://dx.doi.org/10.1002/anie.201803934
http://dx.doi.org/10.1002/ange.201803934

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
10.1002/anie.201803934
Angewandte Chemie International Edition
COMMUNICATION
Olefin Cyclopropanation via Radical Carbene Transfer Reaction
Promoted by Co(II)porphyrinates for the Active-Metal-Template
Synthesis of [2]Rotaxanes
Arthur F. P. Alcântara, Liniquer A. Fontana, Vitor H. Rigolin, Yuri F. S. Andrade, Marcos A. Ribeiro,
Wdeson P. Barros, Catia Ornelas and Jackson D. Megiatto Jr*
We reasoned that introduction of a molecular strap to a
Co(II)porphyrinate would yield a porphyrin-based macrocycle
with a well-defined cavity and with the two axial positions of the
Co(II)porphyrinate available for further coordination of
monodentate ligands and/or activation of substrates. However,
the distinct size of the macrocycle’s cavity would create an
asymmetry on the two axial positions of the Co(II)porphyrinate
due to steric hindrance. Therefore, selective coordination of
strong-field and kinetically robust bulky ligands would occur only
outside the cavity of the macrocycle, while leaving the internal
axial site free for the activation of non-bulky substrates.
Accordingly, the Co(II)porphyrinate-based macrocycle with a
bulky axial ligand would function as an endotopic catalyst able to
mediate radical-carbene transfer reactions between diazo and
styrene half-thread derivatives to afford cyclopropane-linked
rotaxanes by the active-metal-template approach.
Abstract: A Co(II)porphyrinate-based macrocycle in the presence of
3,5-diphenylpyridine axial ligand functions as an endotopic ligand to
direct the assembly of [2]rotaxanes from diazo and styrene half-
threads via radical carbene transfer reactions in excellent 95% yield.
The method reported herein applies the active-metal template
strategy to include radical-type activation of ligands by the metal-
template ion during the organometallic process that ultimately yield
the mechanical bond. A careful quantitative analysis of the product
distribution afforded from the rotaxane self-assembly reaction shows
that the Co(II)porphyrinate subunit is still active after formation of the
mechanical bond and, upon coordination of an additional diazo half-
thread derivative, promotes a novel intercomponent C–H insertion
reaction to yield a new rotaxane-like species. This unexpected
intercomponent C–H insertion illustrates the distinct reactivity
brought to the Co(II)porphyrinate catalyst by the mechanical bond.
We have designed semi-rigid macrocycle 1 (Scheme 1) for
our synthetic strategy. The crystal structure of 1 (CCDC number
1823132, grown from slow evaporation of a DMSO/CH2Cl2
solution under N2 atmosphere, Figure 1) reveals that the pair of
ortho-biphenyl moieties attached to the meso-positions of the
porphyrin core and connected to each other by the flexible
aliphatic ester linker create an aperture with cavity width of 10.3
Å, which is sufficient large to fit a residual DMSO molecule
coordinate to the Co(II) ion. This large cavity prompts us to use
the bulky 3,5-diphenylpyridine moiety as axial ligand in our
studies.
The metal-template synthesis is a powerful methodology to
prepare interlocked molecules.[1] In the so-called active-metal-
template approach,[1b] the metal template ions are coordinate to
endotopic macrocyclic ligands to afford active complexes that
promote formation of covalent bonds between molecular
fragments only through the cavity of the macrocycle to yield
interlocked molecules.[2] The metal complexes explored so far in
the active-template method are formal transition metal species in
which the redox processes occurring during catalysis are
primarily metal centered. However, further developments in
transition metal complexes have opened the possibility for metal
ions to share redox information with the ligands during the
organometallic cycle.[3] For example, some synthetic complexes
can combine 1e– redox change at the ligands with a 1e– redox
change at the metal ion to enable first-row transition metal
complexes to undergo the 2e– redox processes usually observed
only for noble metals or multinuclear species. The
cobalt(II)porphyrinate-mediated radical carbene transfer reaction
are particularly noteworthy as this methodology allows formation
of more nucleophilic radical carbene intermediates that easily
add to olefins affording cyclopropane derivatives instead of
usual coupling reactions.[4]
[*]
Prof. J. D. Megiatto Jr, A. F. P. Alcântara, L. A. Fontana, V. H.
Rigolin, Y. F. S. Ferreira, Dr. M. A. Ribeiro, Prof. W. P. Barros, Prof.
C. Ornelas
Institute of Chemistry, University of Campinas (UNICAMP)
POBox 6154, 13083-970, Campinas, SP, Brazil
E-mail: jackson@iqm.unicamp.br
A. F. P. Alcântara
Instituto Federal do Sertão Pernambucano
Estrada do Tamboril, 56200-000, Ouricuri, PE, Brazil
Scheme 1. Proposed mechanism for the active-metal-template synthesis of
rotaxanes using the radical-carbene transfer reaction to olefins promoted by
Co(II)porphyrinates.[4,5]
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.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
10.1002/anie.201803934
Angewandte Chemie International Edition
COMMUNICATION
MALDI-TOF mass spectrum (Figure S2, S.I.) of the most polar
byproduct shows m/z signals for complexes 5 and 7 along with
that for macrocycle 1. Complexes 5 and 7 are relatively stable
alkyl-Co(III)porphyrinates, which are formed from radical
carbene intermediates[5] 4 and 6 via radical hydrogen abstraction
reactions; a well-known side-reaction that leads to deactivation
of the Co(II)porphyrinate catalyst in the cyclopropanation
process.[4] The presence of m/z peak for macrocycle 1 in the
mass spectrum of the porphyrin byproduct should come from
fragmentation of complexes 5 and/or 7 as Co–C bonds in alkyl-
Co(III)porphyrinates are photosensitive[6] and can break upon
the MALDI-TOF laser ionization process.
The MALDI-TOF mass spectrum of the other isolated
porphyrin byproduct (Figure S3, S.I.) shows a m/z peak at
2059.72, along with that for rotaxane 9 and its characteristic
stepwise fragmentation pattern. The mass spectrum of this
byproduct corresponds to a rotaxane-like species that is formed
when one extra half-thread 2 reacts with rotaxane 9 to yield a
carbenoid species. The coordination of the extra half-thread 2
can occur outside or inside the cavity of the macrocycle
component of rotaxane 9. If coordination of 2 to 9 occurs endo-
to the macrocyclic cavity, cyclopropanation reaction must yield a
[3]rotaxane, which is never observed in our investigation. In
contrast, transfer of the carbenoid intermediate to half-thread 3
exo- to the macrocyclic cavity must regenerate rotaxane 9 while
forming thread 8 and no porphyrin byproduct should have been
observed. Therefore, cyclopropanation reactions do not yield
this rotaxane byproduct. Radical hydrogen abstraction reactions
are also ruled out as those reactions should yield alkyl-
Co(III)porphyrinate-based rotaxanes and the UV-Vis absorption
spectrum of this porphyrin byproduct is very similar to that of
Figure 1. Crystal structure of macrocycle 1-DMSO. Carbon atoms are shown
in gray, nitrogen in light blue, cobalt in dark blue, oxygen in red and sulfur in
yellow. Hydrogen atoms are omitted for clarity purposes. Ellipsoids are drawn
at 50% probability levels.
We were pleased to find that just mixing macrocycle 1 with
half-threads 2 and 3 (Scheme 1) in a 1:1:1 molar ratio in toluene
under a nitrogen atmosphere affords rotaxane 9 (steps a and b,
Scheme 1) in reasonable 45% yield after 18 h at room
temperature (entry 1, Table 1). Preliminary evidence supporting
the structure assignment to rotaxane 9 came from MALDI-TOF
analysis (see Figure S1 in Supporting Information, S.I.). The
fragmentation pattern of the mass spectrum is very
characteristic of interlocked structures.[1d-f] The molecular ion
peak for rotaxane 9 is observed at m/z 1683.28, along with the
molecular ion peak for macrocycle 1 (m/z 854.95), indicating
cleavage of the thread component upon the MALDI-TOF
ionization process. Furthermore, a molecular ion peak fragment
at m/z 1348.20 informs that cleavage of the thread probably
occurs via homolytic rupture of the benzyl linkage with
subsequent loss of a tritylphenoxyl moiety and coordination of
the remaining radical to the Co(II)porphyrinate group on the ring
component during the MALDI-TOF process.
rotaxane
9 (Figure S4, S.I.). Accordingly, the porphyrin
byproduct is a Co(II)-based species.
We reason that this porphyrin byproduct is formed by
coordination of half-thread 2 endo- to the macrocyclic cavity of 9
to produce the carbenoid fragment, which further reacts with the
thread component via insertion reactions into the C–H bonds of
the cyclopropane moiety or those on the benzylic position (For
proposed structures for the rotaxane byproduct, see Figure S5,
S.I.).[7] Those C–H insertions are not observed in the
cyclopropanation
reactions
promoted
by
acyclic
Table 1. Experiments to determine the optimal conditions for formation of
rotaxane 9 using the radical-carbene transfer cyclopropanation reaction as a
metal-active-template methodology.[a,b]
Co(II)porphyrinates[4] because the cyclopropane product diffuses
away from the Co(II)porphyrinate catalyst after the carbenoid
transfer reaction occurs. In the case of rotaxanes, the
mechanical bond always keeps the thread component near the
Co(II)-active center, thereby facilitating this follow up
intercomponent reaction with additional half-thread 2 (step e,
Scheme 1). Furthermore, this rotaxane byproduct is always
isolated in 2-3% yield even though the concentration of 2 is
lower in the reaction medium (Table 1, vide infra), corroborating
the hypothesis of intercomponent reaction.
Molar
Ratio
1:2:3
Rotaxane
byproduct
(%)[f]
Thread
Complex
Rotaxane
Macrocycle
1 (%)
Entry
1
8 (%)[e]
5/7 (%)[f]
9 (%)[f]
1:1:1
1:1:5
12
21
20
-
30
17
19
-
45
56
74
95
3
2
3
3
20
23
-
2[c]
Identification of compounds 5 and 7 as well as the rotaxane
byproduct in the mixture by MALDI-TOF provide important
information for optimization of the reaction conditions. The usual
protocols for cyclopropanation reactions promoted by
Co(II)porphyrinates require no slow addition of the diazo
derivative, while the olefin counterpart is the limit reactant.4
However, Co(II)porphyrinates are used in catalytic amounts in
the traditional one-pot methods. In our case, macrocycle 1 is a
stochiometric reactant and its Co(II)porphyrinate subunit is still
active after formation of rotaxane 9. If concentration of half-
thread 2, a potential radical hydrogen source,[4] is high in the
reaction medium, the deleterious radical hydrogen abstraction
reactions and intercomponent carbenoid insertions are favored.
On the other hand, increasing the concentration of half-thread 3
in the medium should kinetically favor the cyclopropanation
process over those side-reactions.[4]
3[c]
1:1.2:5
1:1.2:5
4[c,d]
-
[a] Reactions carried out at 0.053 M concentration with respect to 1 in toluene,
under N2 atmosphere for 18 h at room temperature. [b] Isolated yields. [c] slow
addition of half-thread 2. [d] Addition of 1 equiv. (with respect to 1) of 3,5-
diphenylpyridine. [e] Relative to diazo 2. [f] Relative to macrocycle 1.
Thread 8 is isolated from the reaction in 12% yield,
confirming that the cyclopropanation reaction also occurs exo- to
the macrocyclic cavity (steps c and d, Scheme 1). Congruently,
noninterlocked macrocycle 1 is recovered from the reaction
mixture in 20% yield. Two porphyrin byproducts with higher
polarity on silica are also isolated from the crude mixture. The
This article is protected by copyright. All rights reserved.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
10.1002/anie.201803934
Angewandte Chemie International Edition
COMMUNICATION
Performing the reaction in the presence of 5-fold excess of
half-thread 3 (with respect to 1, entry 2, Table 1) with portion-
wise addition of half-thread 2 (0.20 equiv. every 2h) to lower its
concentration in the reaction medium, improves the yield of
rotaxane 9 from 45% to 56%, while reducing the yield of
byproduct 5+7 from 30% to 17%. Thread 8 is also isolated in
21% yield and informs that the cyclopropanation reactions
occurring exo- to the macrocyclic cavity are consuming
considerable amount of half-thread 2. Therefore, total
HN’ of the thread component and nuclei HG of the macrocycle
subunit, confirming the proposed structure for rotaxane 10.
1
Under the conditions investigated, H NMR investigation informs
that macrocycle 1 selectively yields both thread 8 and rotaxane
10 as the trans-diastereoisomer.[4]
interlocking of macrocycle
1
seems impossible with
stoichiometric amounts of 2. Running the reaction with a small
excess of 2 (1.20 equiv. with respect to 1) while keeping the
slow addition (0.20 equiv. every 2h) and the 5-fold excess of 3
allows isolation of rotaxane 9 in good 74% yield (entry 3, Table
1) along with 20% of thread 8, and total conversion of
macrocycle 1.
Addition of 1.00 equiv. of 3,5-diphenylpyridine to the reaction
medium (entry 4, Table 1) under the optimized conditions boosts
the yield of rotaxane 9 to impressive 95% (with respect to 1)
along with 3% of the rotaxane byproduct. This unequivocally
shows that coordination of the 3,5-diphenylpyridine axial ligand
Figure 2. Conversion of paramagnetic rotaxane 9 to the diamagnetic rotaxane
10 by oxidation with AgBF4 in the presence of imidazole as axial ligand for
NMR investigation.
to macrocycle
1 mitigates formation of byproduct 7 by
accelerating the cyclopropanation reaction,[4a,c] while protecting
the external axial coordination site that promotes the deleterious
formation of thread 8 and byproduct 5. However, UV-Vis
investigation on the rotaxane isolated from the experiment with
addition of 3,5-diphenylpyridine reveals that the axial ligand is
lost during the purification process and that rotaxane 9 is the
isolated product (see Figure S6 and its caption for further
discussion, S.I.).
(A)
B
A
H
G
D,F
E
C
H,L,M,P
L
B
To investigate the structure of the rotaxane by NMR
spectroscopy, we use AgBF4 as oxidizing agent in the presence
of imidazole as axial ligand to convert the paramagnetic
G,N
L
D,E
F
*
A
M’
U
C
N’
O
Q
NHIM
L,M,Q
M
Co(II)porphyrinate-based rotaxane
9
into the diamagnetic
Co(III)porphyrinate-based 10 (Figure 2).[8] Comparison of the H
NMR spectra of rotaxane 10, thread 8 and that of the free-base
porphyrin macrocycle precursor of 1[9] (Figure 3) provide
unequivocal evidence for the interlocked architecture. The
pyrrolic nuclei (HB) resonances split in several signals, while the
1
O
N’
P
N
(B)
meso-ones (HA) and those of the ortho-biphenyl rings (HC-H
)
K
J
I
NHINNER
appear duplicate in 10 because of the unsymmetrical threaded
axle that renders the porphyrinate protons magnetically
inequivalent. Imidazole axial coordination is confirmed by the
large shielding observed for protons HU, HV and HV’ as they
experience the strong ring-current effect of the porphyrin
aromatic system. Resonances for protons associated with the
cyclopropane moiety (HR, HS and HT) are significantly shielded in
the rotaxane compared to noninterlocked thread 8, with HT
dramatically shifting upfield ( = 6.21 ppm). This is indicative of
the cyclopropane ring on the axle being held very close to the
Co(III)porphyrinate subunit. On the other hand, the signals for
HO, HP and HQ show much smaller shielding ( = 0.49, 0.37
and 0.74 ppm, respectively) than those observed for the
cyclopropane moiety. Therefore, the macrocycle component
seems to preferentially stay over the half-thread bearing the
carbonyl group. Congruently, the protons on the trityl-stoppering
groups (HL, HM, HN and HN’) experience distinct chemical
environments in 10 and split into multiple signals, with HN’
shifting upfield ( = 3.47 ppm), informing that the macrocycle
component partially encapsulates the phenoxy-ester ring on the
trityl-stopper group. Two dimensional NOESY NMR analysis
(Figure S7, S.I.) reveals NOEs between protons HR, HS, HT and
TMS
**
K
J
I
V
V’
**
**
R
S
S’
T
*
TMS
**
*
S
S’
R
T
Figure 3. 1H NMR spectra (400 MHz, CDCl3, 298 K) of the free-base porphyrin
macrocycle precursor of 1,[9] rotaxane 10 and thread 8. (A) aromatic region;
(B) aliphatic region. Proton assignments were based on 2D-NMR
spectroscopy (S.I.) and they correspond to labelling shown in Figure 2. TMS =
Tetramethylsilane. *residual chloroform (7.26 ppm), dichloromethane (5.29
ppm) and water (1.56 ppm) molecules; ** aliphatic impurities.
The strong shielding experienced by HT and HN’ suggests
that the Co(III)porphyrinate should be axially coordinate to the
carbonyl group on the thread to form
a
heteroleptic
hexacoordinate complex.[10] 13C NMR investigation confirms this
This article is protected by copyright. All rights reserved.