Concise total synthesis and stereochemical analysis of tetraponerines T3 and T4

Article abstract of DOI:10.1021/jo302045y

An efficient stereocontrolled preparation of tetraponerines T3 and T4 is detailed. The sequence takes advantage of two consecutive stereoselective aminoallylations of appropriate aldehydes with chiral tert-butanesulfinamide and in situ generated allyl indium species. The absolute configuration of the carbon stereogenic center at the aminal moiety is thermodynamically controlled. This was ascertained on the basis of an exhaustive DFT configurational study of tetraponerines, which fulfils the lack of detailed structural information for these systems. It was found that the trans-transoid-configuration of the AB rings is the most stable geometry for T3 and T4. However, the C ring prefers a cis-configuration in T3 (ttc-T3) and a trans-fusion in T4 (ttt-T4). Regarding their dynamic behavior, low activation barriers were found by DFT calculations for the inversion of the nitrogen at the indolizidine framework, allowing rapid equilibration between the major configurations (ttc and ttt) in T3 and T4.

Full text of DOI:10.1021/jo302045y

Article
pubs.acs.org/joc
Concise Total Synthesis and Stereochemical Analysis of
Tetraponerines T3 and T4
Irene Bosque, Jose C. Gonzalez-Gomez,* Albert Guijarro,* Francisco Foubelo, and Miguel Yus
́
́
́
Departamento de Química Organ
99, 03080 Alicante, Spain
́ ́
ica, Facultad de Ciencias and Instituto de Síntesis Organica (ISO), Universidad de Alicante, Apdo.
S
* Supporting Information
ABSTRACT: An efficient stereocontrolled preparation of
tetraponerines T3 and T4 is detailed. The sequence takes
advantage of two consecutive stereoselective aminoallylations
of appropriate aldehydes with chiral tert-butanesulfinamide and
in situ generated allyl indium species. The absolute
configuration of the carbon stereogenic center at the aminal
moiety is thermodynamically controlled. This was ascertained
on the basis of an exhaustive DFT configurational study of
tetraponerines, which fulfils the lack of detailed structural
information for these systems. It was found that the trans-
transoid-configuration of the AB rings is the most stable geometry for T3 and T4. However, the C ring prefers a cis-configuration
in T3 (ttc-T3) and a trans-fusion in T4 (ttt-T4). Regarding their dynamic behavior, low activation barriers were found by DFT
calculations for the inversion of the nitrogen at the indolizidine framework, allowing rapid equilibration between the major
configurations (ttc and ttt) in T3 and T4.
Tetraponerines act as paralyzing venoms that are used as
chemical warfare by the host species, and consequently it was
earlier found that they exhibit insecticidal activity.1 Recent
studies categorize them as efficient inhibitors of a range of
nAChRs (nicotinic acetylcholine receptors).7 More recently, it
has also been found that these molecules and some analogues
possess cytotoxic activity.8 The unprecedented tricyclic
skeleton along with the interest in their biological activities
have made them attractive targets for total synthesis.9 Despite
these elegant and creative synthetic approaches, efficient
syntheses of this family of alkaloids that can be easily modified
to prepare biologically active analogues are still desirable.
Moreover, insight into the configurational spectrum of
tetraponerines could be important to understand their mode
of action in any activity displayed as well as to perform a
rationale search of structure−activity for tetraponerine derived
compounds. In this context, we present herein a concise
divergent synthesis of tetraponerines T3 and T4, as well as their
configurational and conformational analysis using DFT
calculations.
INTRODUCTION
■
In 1987 Braekman et al. reported the isolation and the structure
determination of tetraponerines, a family of alkaloids that
Pseudomyrmecine ants of the genus Tetraponera segregate
against their enemies.1 These tricyclic alkaloids, named
tetraponerines T1−T8, possess a singular aminal moiety and
can be divided into two structural families according to the size
of the ring A (Figure 1). Interestingly, some experimental
Figure 1. Structures of tetraponerines T1−T8.
evidence has been found to support two different pathways for
the biosynthesis of the two tetraponerine skeletons.2,3 In each
family, the compounds differ from each other by the length of
the alkyl chain (propyl or pentyl) and/or by the configuration
at C5. The structure and relative configuration of the major
component of the mixture of these eight toxic alkaloids,
(+)-T8, was determined by X-ray diffraction analysis,1 and its
absolute configuration was confirmed by enantioselective
synthesis.4 Although the original structures proposed for T3,
T5, T6, and T7 were wrong, extensive spectroscopic and
circular dichroism studies allowed the correction of these
structures.5,6
RESULTS AND DISCUSSION
■
Upon biosynthetic studies, T8 was subjected to chemical
degradation, and it was found that N11−C11a could be
selectively cleaved by catalytic hydrogenation in acidic media,
most likely through the corresponding iminium ion.2 Notably,
the absolute stereochemistry at the aminal center (C11a) of all
Received: September 18, 2012
Published: October 23, 2012
© 2012 American Chemical Society
10340
dx.doi.org/10.1021/jo302045y | J. Org. Chem. 2012, 77, 10340−10346

The Journal of Organic Chemistry
Article
natural tetraponerines is the same regardless the configuration
(R/S) at C5. With this in mind, we speculated that the
configuration at the aminal center of natural tetraponerines
could be thermodynamically controlled by equilibration
between the most stable aminals and the corresponding open
iminium form. On the basis of this hypothesis we anticipated
that rings B and C of tetraponerines T3 and T4 could be
formed, with the right stereochemistry at C11a, by reaction of 4-
bromobutanal10 with the corresponding diamines. Relying on
our indium stereoselective aminoallylation of aldehydes,11 we
traced back the synthesis of diamine derivatives to 5-
bromopentanal, chiral tert-butanesulfinamide, and in situ
generated allyl indium species (Scheme 1). Hence we
dr, where 14% is the sum of other isomers). More importantly,
compound 3a was isolated in 72% yield as a single isomer after
column purification. When aldehyde 2 was submitted to the
same reaction conditions, but using (RS)-tert-butanesulfina-
mide, better diasteroselection was observed for compound 3b
(96:4 dr), which was isolated as a single isomer in 80% yield
after column purification. Remarkably, even for the mismatched
case (3a), the stereoinduction of the chiral sulfinyl group
exceeds the effect of the stereocenter at the piperidine ring,
which was crucial to access both required configurations at the
homoallylic stereocenter. Moreover, the iterative use of the
stereoselective aminoallylation procedure (initially to prepare
1a in 94:6 dr) resulted in an enantiomeric enhancement that
allowed the isolation of 3a or 3b in more than 99:1 er.
With an efficient method in hand to prepare diamine
derivatives 3a and 3b with high enantiomeric purity, the free
diamines were prepared by aqueous acidic removal of the chiral
auxiliary followed by catalytic hydrogenation of the olefin
moiety with concomitant removal of the carbamate group
(Scheme 3). The crude diamines were treated with 4-
Scheme 1. Retrosynthetic Analysis of Tetraponerines T3 and
T4
Scheme 3. Endgame in the Synthesis of T3 and T4
envisioned a protocol where a protected 2-allylpiperidine is
formed from 5-bromopentanal, transformed into the corre-
sponding aldehyde, and submitted to a second aminoallylation
to afford the desired diamines. Importantly, both required
diastereomeric diamines could be available by using either (SS)-
tert-butanesulfinamide or its enantiomer.
As depicted in Scheme 2, our synthesis commenced with an
efficient preparation of the 2-allylpiperidine derivative 1a.12
Scheme 2. Divergent Stereoselective Synthesis of Diamine
Derivatives 3a and 3b
bromobutanal in the presence of K2CO3 to obtain tetraponer-
ines T3 and T4, respectively, in good overall yield. Spectral and
1
physical data for these compounds (e.g., H and 13C NMR, IR,
optical rotation) were in good agreement with those reported
in the literature for natural5 and synthetic samples.9b
Importantly, when the biological activity of tetraponerine
analogues was recently evaluated, the long-chain derivatives at
C5 systematically displayed the lowest IC50 values.7,8 In this
context, the synthesis described here could be easily modified
to enlarge the chain at C5 of T3 or T4 by cross-metathesis of
intermediates 3a or 3b with the corresponding terminal
alkenes.
As we mentioned before, our synthetic approach to
tetraponerines T3 and T4 relied on the hypothesis that the
natural alkaloids are the most stable aminals possible. We have
not found any comprehensive study on the conformational and
One-pot transformation of 1a into 1b was followed by oxidative
cleavage of the olefinic double bond using a modified Johnson-
Lemieux oxidation.13 Aminoallylation of aldehyde 2 using our
indium-mediated protocol with (SS)-tert-butanesulfinamide11
took place smoothly with moderate diasteroselectivity (86:14
10341
dx.doi.org/10.1021/jo302045y | J. Org. Chem. 2012, 77, 10340−10346

The Journal of Organic Chemistry
Article
configurational analysis of tetraponerines, so in order to fulfill
the lack of detailed structural information, we did an exhaustive
computational study of all of the different stereoisomeric
configurations that, in principle, are conceivable using this as
well as other synthetic methodologies. For that purpose, a few
remarks on the stereochemical nomenclature and notation used
are pertinent. Concerning the nitrogen atoms, the relevant
configurational descriptors of the pyramidal nitrogen were
specifically defined, in coherence with the calculation inputs.
Until now, this task has been always overlooked, probably on
the basis of the expected low nitrogen inversion barriers. Thus,
we speak of configuration of the pyramidal nitrogen regardless
of its inversion barrier. Both T3 and T4 share a bicyclic
quinolizidine (rings AB, Figure 1) as well as an indolizidine
(rings BC, Figure 1) fragment. We used as a starting anchor
point the configuration of C6a, which is invariably R in these
natural products. The configuration of the nearest nitrogen,
N11, defines the type of ring fusion of the AB fragment, either
cis- or trans-. Next, the relative stereochemistry of the
quinolizidine and indolizidine frameworks (i.e., the rings AB
and BC) was defined as either cisoid or transoid. Finally the
indolizidine ring fusion BC is again defined as either cis or trans.
Figure 2 exemplifies it for some important configurations of T3
and T4.
the conformational and configurational space of this family of
compounds.
The results of this DFT study are represented in Charts 1
and 2.15,16 The hypothesis of a thermodynamic control of the
reaction is sound. For both T3 and T4, the most stable
geometry corresponds to the trans-transoid-configurations of
the AB rings, while the C ring prefers a cis-configuration in T3
(ttc-T3, ΔE = −0.38 kcal mol−1, population 66%), closely
followed by the opposite C ring fusion (ttt-T3, ΔE = 0 kcal
mol−1, population 33%). The third configuration in order of
stability is a T3 diastereomer with an inverted configuration at
C
11a and an expected hypothetical population of less than 1% in
the mixture (i.e., the ctt2-T3, ΔE = 2.34 kcal mol−1, population
0.47%). For T4, the configurational spectrum is dominated by
the trans-fusion at the C ring (ttt-T4, ΔE = 0 kcal mol−1,
population 95%), with only a minor population of <5% of the
cis-isomer at the C ring (i.e., the ttc-T4, ΔE = 1.81 kcal mol−1,
population 4.5%).
The reason for the reported order of stabilities are a
combination of strain, steric, and to a lesser extent, electronic
effects. Among the last ones, the anomeric effect in the T3-4
system may be described as a stabilizing effect originated by
hyperconjugation of one nitrogen lone pair with the empty
C11a−N σ* bond of the second nitrogen, which may operate
along with an alternative mechanism of dipole cancellation,
expected to be less important in these systems. Depending on
the relative geometric orientation of the two nitrogen atoms
around C11a, the possibility of none (ccc2-, ctt2-, tcc- and ttt-T3-
4), one, or two (ccc-T3-4) anomeric interactions stabilizing the
structure may arise. The effect is strongly dependent on the
overlap of these two orbitals and is maximum with an
antiparallel alignment between them. In general, the anomeric
effect was found to be less important, and its additive effect to
the overall energy is easily masked by other stronger
interactions. For instance, two out of the three T3
configurations with lower energy as well as the lowest energy
configuration of T4 have zero anomeric effects (ttt-T3-4 and
ctt2-T3). The most stable T3 configuration (the ttc-T3) has
one clear anomeric effect, but so have most of the remaining
configurations. The ccc-T3 configuration is the only one having
two simultaneous anomeric interactions; however, it lies 6.28
kcal mol−1 above the minimum and is therefore not
representative in practice of the T3.17 The same is true for
the ccc-T4, which lies 8.29 kcal mol−1 above the minimum.
Despite this, the anomeric effect exists and is patent in the
calculated geometries. It can be identified by a small but regular
shortening of the donor N−C11a bond and a lengthening of the
acceptor N−C11a bond, which is statistically consistent for both
the T3 and T4 configurations. The C11a−N average bond
devoid of anomeric effects is 1.461 Å (σ = 0.006 Å). The
average N−C11a bond with anomeric shortening is 1.454 Å (σ =
0.005 Å), while the average bond with anomeric lengthening
is 1.483 Å (σ = 0.008 Å).18 We have quantitatively evaluated
the magnitude of the anomeric effect by means of second order
perturbation theory, estimating the donor−acceptor (bond−
antibond) interactions on the NBO (Natural Bond Orbital)
basis.19 Typical stabilization energies in the 10−11 kcal mol−1
range were actually found for the nitrogen lone pair (donor N
LP) interacting with an antiparallel carbon−nitrogen sigma
antibond (acceptor C11a−N σ*). Although these are fairly
typical values for an anomeric stabilization, we found that the N
LP was also capable to stabilize a range of acceptor antibonds
provided that they were antiperiplanar to the N LP, with
Figure 2. Stereochemical descriptors of T3, T4 stereoisomers. As an
example, the trans-transoid-trans-T3 and -T4 (ttt-T3 and ttt-T4) and
the corresponding trans-transoid-cis-T3 and -T4 (ttc-T3 and ttc-T4) are
represented.
When the AB ring fusion is cis-, the quinolizidine framework
exists as an equilibrium mixture of two well-defined conformers
(just as in cis-decalin). These have been named here as series 1
and 2, which interconvert by simple ring flip and are illustrated
in Figure 3 in the case of cis-cisoid-cis-T4 (ccc-T4 ⇆ ccc2-T4). In
general, other local conformations (e.g., the envelope in the
pyrrolidine ring) were imposed by the rest of the structure.
Finally, the n-propyl side chain was always found to
accommodate in an extended conformation gauche to the
C6.14 Altogether, these stereochemical motifs defined precisely
Figure 3. Conformations of the cis- series 1 and 2 exchanging by ring
flip. By definition, H6a (boldface) is axial in ring B in series 1 and
equatorial in series 2.
10342
dx.doi.org/10.1021/jo302045y | J. Org. Chem. 2012, 77, 10340−10346

The Journal of Organic Chemistry
Article
a
Chart 1. DFT Energy Profiles (ΔE Including ZPE Corrections in kcal mol−1) of Local Minima for All T3 Configurations
a
The structures of the six configurations lower in energy are drawn.
a
Chart 2. DFT Energy Profiles (ΔE Including ZPE Corrections in kcal mol−1) of Local Minima for All T4 Configurations
a
The structures of the six configurations lower in energies are drawn. During geometry optimization, the ccc2-T4 became cct2-T4 and the ctt-T4
became ctc-T4. Both are consistent with a spontaneous release of strain through nitrogen inversion at the indolizidine framework.
surprisingly high stabilization energies, that altogether may
compensate or even overcome the anomeric effect in the
inverted configuration (i.e., that lacking anomeric effect).20
Among these interactions, the N LP-C−Hax σ* is historically
relevant since it gives rise to the so-called Bohlmann bands in
IR spectroscopy,21 a specific C−H stretching of the axial
hydrogens adjacent to the nitrogen atom.
Pr group, as well as no anomeric effects. For T4, the ttt-T4 has
a trans-quinolizidine type of fusion also and no anomeric
stabilization. There is a small gauche interaction between C3
and Pr that is not relieved now by nitrogen inversion at N4,
which makes it the more stable configuration. The inverted ttc-
T4 has still that gauche interaction, as well as an additional 1,3-
diaxial methylene interaction with only 1 hydrogen (Figure 2
and Chart 2), which seems not to be compensated by the
appearance of one anomeric stabilization similar to the one
described before (calculated C11a−N4 and C11a−N11 bond
lengths are 1.454 and 1.489 Å, respectively).
Finally, although the T3 and T4 do not interconvert
spontaneously, a comparison of the intrinsic stability of both
tetraponerines was done. Overall, for the theoretical
interconversion T3 ⇆ T4, ΔG° is −2.52 kcal mol−1, meaning
that the tetraponerine T4 would account for 98.6% and the T3
only for the remaining 1.4% of a hypothetical equilibrating
mixture.23 This issue proved to be crucial in a previous
stereoselective synthesis of T4, the final step of which relied on
a thermodynamic equilibration of the Pr group attached to a
fast inverting radical at C5.9a
We next focused our attention on the inversion barriers of
the N11 (quinolizidine framework) and the N4 (indolizidine
framework). Starting from the major components of the T3
and T4, the nitrogen inversion at the quinolizidine part of the
molecule (i.e., at N11) affords substantially more unstable
configurations: ttc-T3 (66%) ⇆ ccc-T3 (ca. 0.0013%) and ttt-
T3 (33%) ⇆ cct-T3 (ca. 0.020%), as well as ttc-T4 (4.5%) ⇆
ccc-T4 (3.7 × 10−5 %) and ttt-T4 ⇆ cct-T4 (ca. 0.15%). These
inverted configurations are unstable enough to prevent (or at
least to complicate very much) experimental determination of
Concerning tensional and steric effects, two important
patterns of destabilization are easily identified. The methyl-
ene-methylene (or alkyl-alkyl) 1,3-diaxial interactions, as
occurring in ccc-T3-4, ccc2-T3, ctc-T3, ctc2-T4, ctt2-T4, and
tcc-T3, is one of them. The second, even more severe, is a ring
B deformation from chair into other distorted conformations
such as twisted boat, which occurs when a tentative trans-diaxial
fusion with ring C is imposed, e.g., cct2-T3-4, ctt-T3, tct-T3-4.15
None of these destabilizing elements are found in the more
stable configurations of T3 and T4. The main stereochemical
elements of these configurations follow.22 For T3, ttt-T3 has a
trans-quinolizidine type of fusion and lacks anomeric
stabilization. There is a small gauche interaction between C3
and Pr. This interaction can be avoided by nitrogen inversion at
N4, becoming then the more stable configuration, ttc-T3, at the
expense of an additional 1,3-diaxial methylene interaction with
only one hydrogen atom (Figure 2 and Chart 1). In addition,
ttc-T3 takes advantage of one anomeric stabilization between
the N4 lone pair and the empty C11a−N11 σ* (the calculated
C11a−N4 and C11a−N11 bond lengths are 1.455 and 1.489 Å,
respectively). In the less stable ctt2-T3, there is a cis-
quinolizidine type of fusion, but the Pr group is equatorial.
The methylene C7 is in 1,3-diaxial interaction but only with 2
hydrogens, and there is a small gauche interaction C3 with the
10343
dx.doi.org/10.1021/jo302045y | J. Org. Chem. 2012, 77, 10340−10346

The Journal of Organic Chemistry
Article
°C and concentrations (c) are given in g/100 mL. Infrared analysis was
performed with a spectrophotometer equipped with an ATR
component; wavenumbers are given in cm−1. Mass spectra (EI)
were obtained at 70 eV, and fragment ions are in m/z with relative
intensities (%) in parentheses. HRMS analyses were also carried out in
the electron impact mode (EI) at 70 eV using a quadrupole mass
analyzer. GC analyses were obtained with an HP-5 column (30 m ×
0.25 mm, i.d. × 0.25 μm) and an EI (70 EV) detector. The
temperature program was as follows: hold at 60 °C for 3 min, ramp
from 60 to 270 °C at 15 °C/min, hold at 270 °C for 10 min.
the equilibria and have little practical relevance. However, this
is not the case for the indolizidine fragment. The inversion at
the N4 exchanges two substantially populated configurations,
particularly in the case of T3: ttc-T3 (66%) ⇆ ttt-T3 (33%)
and ttt-T4 (95%) ⇆ ttc-T4 (4.5%). The calculated activation
barriers for these two processes are reported in Figure 4 and
1
1H NMR spectra were recorded at 300 or 400 MHz for H NMR
and 75 or 100 MHz for 13C NMR, using CDCl3 or C6D6 as solvents
and TMS as internal standard (0.00 ppm). The data is reported as (s =
singlet, d = doublet, t = triplet, m = multiplet or unresolved, br s =
broad signal, coupling constant(s) in Hz, integration). 13C NMR
spectra were recorded with 1H-decoupling at 100 MHz and referenced
to CDCl3 at 77.16 ppm or to C6D6 at 128.06 ppm. DEPT-135
experiments were performed to assign CH, CH2 and CH3.
(R)-2-Allyl-1-(benzyloxicarbonyl)piperidine (1b). An aqueous
6 M solution of HCl (1.5 mL, 9.00 mmol) was added dropwise to a
solution of 1a12 (687 mg, 3.00 mmol) in dry THF (3 mL) at 0 °C
under Ar. The solution was allowed to reach 23 °C and was stirred for
1.5 h. After the mixture cooled to 0 °C, an aqueous 2 M solution of
NaOH (14 mL, 27.00 mmol) was added dropwise, and the resulting
mixture was stirred at the same temperature. After 5 min, a solution of
benzyloxycarbonyl chloride (515 μL, 3.60 mmol) in CH2Cl2 (11 mL)
was added to the stirred solution. The resulting mixture was then
allowed to reach 23 °C and was stirred for 3 h. The reaction mixture
was extracted with EtOAc (3 × 20 mL), and the combined organic
layers were washed with H2O, dried over MgSO4, filtered, and
concentrated under reduced pressure. The residue was purified by
column chromatograpy (95:5 hexane/EtOAc) to provide the desired
product as a colorless oil (715 mg, 92%). Rf1= 0.26 (9:1 hexane/
Figure 4. Calculated activation barriers for N4 inversion in T3 (ΔE =
1.90 kcal mol−1) and T4 (ΔE = 4.92 kcal mol−1) at the B3LYP/6-
311+G (2d,p) level. Notice the trigonal planar geometry at the
inverting nitrogen; ν = 126.27 i cm−1 for the transition state TS-T3
(top left) and ν = 163.64 i cm−1 for the transition state TS-T4 (top
right).
correspond to the processes drawn in Figure 2. In both cases, at
the transition state the nitrogen displays a trigonal planar
geometry. Visual examination of the imaginary frequency mode
passing through the transition state confirmed the correct
nature of the calculated saddle points TS-T3-4. Since the
barriers encountered are low, a rapid interconversion between
these two configurations for each tetraponerine is expected.24
EtOAc); [α]20 = +44 (c 0.498 in CHCl3); H NMR (300 MHz,
D
CDCl3) δ 7.37−7.28 (m, 5H), 5.71 (td, J = 16.5, 7.3 Hz, 1H), 5.19−
4.95 (m, 4H), 4.37 (br s, 1H), 4.05 (d, J = 12.3 Hz, 1H), 2.86 (td, J =
12.8, 2.5 Hz, 1H), 2.43 (dddd, J = 10.5, 7.4, 2.4, 1.2 Hz, 1H), 2.25 (dt,
J = 14.0, 7.1 Hz, 1H), 1.73 −1.54 (m, 5H), 1.54 −1.31 (m, 1H); 13C
NMR (101 MHz, CDCl3) δ 155.7 (C), 137.2 (C), 135.4 (CH), 128.6
(CH), 128.0 (CH), 127.9 (CH), 117.0 (CH2), 67.0 (CH2), 50.5
(CH), 39.4 (CH2), 34.6 (CH2), 27.7 (CH2), 25.6 (CH2), 18.9 (CH2);
IR (ATR) ν 3035, 2937, 2859, 1691, 1641, 1445, 1421, 1258, 1240
cm−1; LRMS (EI) m/z (%) 259 (M+, 0.1), 219 (19), 218 (100), 175
(32), 174 (100), 92 (40), 91 (100), 65 (35), 55 (20); HRMS (EI) m/z
calcd for C16H21NO2 259.1572, found 259.1601.
CONCLUSION
■
The iterative use of stereoselective aminoallylation of aldehydes
with chiral tert-butanesulfinamide and in situ generated allyl
indium species has been successfully applied to the synthesis of
tetraponerines T3 and T4. The formation of rings B and C was
achieved by reaction of the corresponding enantioenriched free
diamines with 4-bromobutanal, obtaining the desired aminals
with the right stereochemistry at the aminal center (C11a). An
exhaustive configurational analysis of T3 and T4 by DFT
calculations revealed that they are mainly populated (>99%) by
isomers with the trans-transoid-configuration at the AB rings,
which clearly supports the hypothesis of a thermodynamic
control of the stereochemistry at the aminal center. This study
also shows that the C ring prefers a cis-configuration in T3 (ttc-
T3) and a trans-fusion in T4 (ttt-T4), although both are in
rapid equilibration with their respective second most populated
configurations (i.e., ttt-T3 and ttc-T4) given the low activation
barriers calculated for the nitrogen inversion at the indolizidine
framework.
(R)-1-(Benzyloxicarbonyl)-2-(2-oxoethyl)piperidine (2). To a
solution of 1b (684 mg, 2.64 mmol) in 1,4-dioxane/H2O (3:1, 27 mL)
were successively added 2,6-lutidine (615 μL, 5.28 mmol), NaIO4
(2.30 g, 10.56 mmol), and a solution of OsO4 in t-BuOH (2.5% w/w
in t-BuOH, 260 μL). The mixture was stirred at 23 °C for 1.5 h before
being quenched with water (20 mL). The mixture was extracted with
EtOAc (3 × 20 mL), and the collected organic layers were washed
with H2O, dried over MgSO4, filtered, and concentrated under
reduced pressure. The residue was purified by column chromatograpy
(4:1 hexane/EtOAc) to provide the pure product as a colorless oil
(610 mg, 88%). Rf = 0.30 (7:3 hexane/EtOAc); [α]20 = +42 (c 0.34
D
1
in CHCl3); H NMR (300 MHz, CDCl3) δ 9.71 (s, 1H), 7.42−7.29
(m, 5H), 5.12 (s, 2H), 4.93 (dd, J = 13.0, 6.9 Hz, 1H), 4.08 (d, J =
13.1 Hz, 1H), 2.85 (t, J = 13.0 Hz, 1H), 2.75 (ddd, J = 15.6, 8.1, 3.0
Hz, 1H), 2.60 (ddd, J = 15.7, 6.9, 2.0 Hz, 1H), 1.81 −1.32 (m, 6H);
13C NMR (101 MHz, CDCl3) δ 200.7 (CH), 155.4 (C), 136.7 (C),
EXPERIMENTAL SECTION
128.6 (CH), 128.2 (CH), 128.0 (CH), 67.4 (CH2), 46.3 (CH), 44.6
(CH2), 39.8 (CH2), 28.8 (CH2), 25.3 (CH2), 19.0 (CH2); IR (ATR) ν
3026, 2939, 2860, 2731, 1721, 1687, 1445, 1418, 1257, 1238 cm−1;
LRMS (EI) m/z (%) 261 (M+, 3), 233 (27), 218 (20), 175 (13), 174
(100), 170 (18), 126 (27), 108 (20), 92 (29), 91 (100), 65 (30), 55
(13); HRMS (EI) m/z calcd for C15H19NO3 261.1365, found
261.1351.
■
General Remarks. (RS)-tert-Butanesulfinamide and its enantiomer
were a gift of Medalchemy (>99% ee by chiral HPLC on a Chiracel AS
column, 90:10 hexane/i-PrOH, 1.2 mL/min, λ = 222 nm). TLC was
performed on silica gel 60 F254 using aluminum plates and visualized
with phosphomolybdic acid (PMA) stain. Flash chromatography was
carried out on handpacked columns of silica gel 60 (230−400 mesh).
Melting points are uncorrected. Optical rotations were measured using
a polarimeter with a thermally jacketted 5 cm cell at approximately 20
(2R,2′S,SS)-1-(Benzyloxicarbonyl)-2-[2′-tert-(butylsulfina-
mide)-4′-pentenyl]piperidine (3a). To a dry flask were added (SS)-
10344
dx.doi.org/10.1021/jo302045y | J. Org. Chem. 2012, 77, 10340−10346

The Journal of Organic Chemistry
Article
tert-butanesulfinamide (220 mg, 1.82 mmol) and indium powder (259
mg, 2.27 mmol) under Ar. Then was added a solution of compound 2
(496 mg, 1.90 mmol) in dry THF (4.5 mL) followed by Ti(OEt)4
(818 μL, 3.64 mmol), and the reaction mixture was stirred under Ar
for 1 h at 23 °C. At this time allyl bromide (236 μL, 2.73 mmol) was
added to the mixture, and it was heated to 60 °C for 5 h. The mixture
was allowed to reach 23 °C and was carefully added over a stirring
mixture of 4:1 EtOAc/brine (15 mL). The resulting white suspension
was filtered through a short pad of Celite, washed with EtOAc, and
organics were concentrated in vacuo. According to HPLC analysis of
the crude product, 86% of the diastereomeric mixture corresponds to
the major diastereoisomer. After column chromatography (3:2
hexane/EtOAc), the major isomer was isolated pure (>99:1 according
to HPLC) as a white solid (532 mg, 72%). Rf = 0.26 (1:1 hexane/
EtOAc); mp 31.6−32.9 °C; [α]20D = +76 (c 0.62 in CHCl3); 1H NMR
(300 MHz, CDCl3) δ 7.42−7.30 (m, 5H), 5.72 (dt, J = 16.6, 8.4 Hz,
1H), 5.26−5.01 (m, 4H), 4.88 (br s, 1H), 4.65 (br s, 1H), 4.03 (m,
1H), 2.94 (br s, 1H), 2.77 (t, J = 12.7 Hz, 1H), 2.53 (br s, 1H), 2.34
(dt, J = 13.9, 6.9 Hz, 1H), 2.02 (t, J = 12.7 Hz, 1H), 1.84−1.36 (m,
7H), 1.31−0.98 (2 x br s, 9H); 13C NMR (101 MHz, CDCl3) δ 156.0
(C), 137.0 (C), 135.4 (CH), 128.6 (CH), 128.1 (CH), 127.9 (CH),
117.6 (CH2), 67.3 (CH2), 55.8 (C), 53.3 (CH), 47.6 (CH), 39.4
(CH2), 39.1 (CH2), 36.4 (CH2), 29.4 (CH2), 25.6 (CH2), 22.9 (CH3),
19.3 (CH2); IR (ATR) ν 3242, 3033, 2936, 1674, 1641, 1421, 1260,
1064 cm−1; LRMS (EI) m/z (%) 350 (M+ − C4H8, 3), 218 (10), 174
(29), 91 (100), 84 (19); HRMS (EI) m/z calcd for C22H34N2O3S −
C4H8 350.1664, found 350.1665; HPLC (Chiralcel AD-H column 25
cm × 0.46 cm, 95:5 hexane/i-PrOH, 1.0 mL min−1, λ = 217 nm) tR for
major isomer 15.93 min, tR other diastereoisomers 20.76−23.39 min.
(2R,2′R,RS)-1-(Benzyloxicarbonyl)-2-[2′-tert-(butylsulfina-
mide)-4′-pentenyl]piperidine (3b). This compound was prepared
from (RS)-tert-butanesulfinamide (282 mg, 2.33 mmol) and compound
2 (610 mg, 2.34 mmol) following the same procedure described above
for compound 3a. According to HPLC analysis of the crude product,
95% of the diasteromeric mixture corresponds to the major
diastereoisomer. After column chromatography (3:2 hexane/EtOAc)
the major isomer was isolated pure (>99:1 dr according to HPLC) as a
white solid (757 mg, 80%): Rf = 0.21 (1:1 hexane/EtOAc); mp 88.9−
(96:4:0.05 CH2Cl2/MeOH/20% NH4OH) to provide the desired
product as an oil (151 mg, 54% from 3a). Rf = 0.40 (9:1 CH2Cl2/
MeOH); [α]20D = +35 (c 0.49 in CHCl3) {lit.5 [α]20D + 27 (c 0.07 in
CHCl3), lit.9b [α]20D = +35 (c 0.51 in CHCl3)}; 1H NMR (300 MHz,
C6D6) δ 3.28 (dd, J = 4.8, 2.2 Hz, 1H), 3.20 (dd, J = 14.6, 7.4 Hz, 1H),
2.85−2.69 (m, 3H), 2.08−1.98 (m, 1H), 1.97−1.87 (m, 1H), 1.86−
1.50 (m, 8H), 1.50−1.01 (m, 8H), 0.93 (t, J = 7.1 Hz, 3H); 13C NMR
(101 MHz, C6D6): δ = 75.7 (CH), 56.8 (CH), 52.8 (CH), 50.8
(CH2), 50.7 (CH2), 33.9 (CH2), 33.2 (CH2), 32.0 (CH2), 30.4 (CH2),
26.2 (CH2), 25.1 (CH2), 22.2 (CH2), 20.7 (CH2), 14.5 (CH3); IR
(ATR) ν 2952, 2927, 2869, 2803, 1455, 1390, 1354, 1157, 1129, 1113
cm−1; LRMS (EI) m/z (%) 222 (M+, 49), 221 (64), 194 (14), 193
(100), 180 (8), 179 (14), 152 (44), 138 (19), 137 (14), 124 (12), 110
(12), 97 (10), 96 (51), 84 (19); HRMS (EI) m/z calcd for C14H26N2
222.2096, found 222.2076; GC major peak (>98%) at 12.40 min.
Tetraponerine T4. It was prepared from 3b (199 mg, 0.49 mmol)
and 4-bromobutanal (111 mg, 0.74 mmol), following the same
procedure described above for tetraponerine T3. The expected
product was obtained as an oil (61 mg, 56% from 3b). Rf = 0.43
(9:1 CH2Cl2/MeOH); [α]20D = +102 (c 0.34 in CHCl3) {lit.5 [α]20
=
D
+94 (c 0.2 in CHCl3), lit.9b [α]20 = +107 (c 1.16 in CHCl3)}; H
1
D
NMR (400 MHz, C6D6) δ 3.16 (td, J = 8.2, 2.2 Hz, 1H), 2.83 (d, J =
8.1 Hz, 1H), 2.35 (t, J = 6.3 Hz, 1H), 2.13 (ddd, J = 10.7, 7.1, 3.4 Hz,
1H), 2.04 (dd, J = 15.9, 8.5 Hz, 1H), 1.86−1.55 (m, 7H), 1.55−1.05
(m, 11H), 0.88 (t, J = 7.1 Hz, 3H); 13C NMR (101 MHz, C6D6) δ
86.0 (CH), 63.3 (CH), 61.4 (CH), 51.9 (CH2), 49.2 (CH2), 37.9
(CH2), 37.3 (CH2), 33.2 (CH2), 30.0 (CH2), 26.5 (CH2), 25.5 (CH2),
20.8 (CH2), 19.3 (CH2), 15.3 (CH3); IR (ATR) ν 2930, 2870, 2789,
1646, 1454, 1377, 1337, 1190, 1157, 1025 cm−1; LRMS (EI) m/z (%)
222 (M+, 49), 221 (100), 194 (14), 193 (98), 180 (15), 179 (15), 152
(30), 151 (20), 138 (22), 137 (14), 124 (14), 110 (13), 96 (35), 84
(19); HRMS (EI) m/z calcd for C14H26N2 222.2096, found 222.2064;
GC major peak (>98%) at 12.74 min.
Computational Details. All of the 24 structures of Charts 1 and 2
were built as explained in the text and were initially optimized using a
MM2 force field before being submitted to density functional theory
calculations (DFT). DFT calculations of the structures,26 energies, and
harmonic vibrational analysis were carried out using the Becke−Lee−
Yang−Parr (B3LYP) exchange-correlation functional.27 We relied on
the widely used B3LYP functional, the performance of which was
reviewed recently with a collection of molecules of biological relevance
on the basis of the reported errors in barrier height energy (for singlet
transition states) and conformational energies,28 and also specifically
for this functional with a larger set of neutral, closed-shell organic
molecules containing C, H, N, and O atoms, on the basis of the
isomerization energies for nitrogen-containing molecules.29 Important
correlation energy corrections due to noncovalent, medium-range
interactions are not expected.30 The geometries of the isolated species
have been fully optimized in the gas phase using the split valence
triple-ζ 6-311+G (2d,p) basis set.31 When both polarization and
diffuse functions are used, an improvement of the isomerization
energies of amines was reported for this functional.29 In addition a
double set of polarization functions was used to get a better
description of the inversion barriers at the nitrogen. A preliminary
study with a typical double-ζ 6-31G (d) basis set was also carried out
to test the level of convergence of the energies with the size of the
base, which was considered adequate for our purposes. Analytic second
derivative calculations, which yield the harmonic frequencies, were
performed on the optimized geometries at the same level of theory to
ensure that the optimized geometries were true minima and to provide
corrections for the zero-point energy (ZPE) effects. The Hessian
matrices of the optimized geometries had only positive eigenvalues.
The activation barriers were located using the synchronous transit-
guided quasi-Newton (STQN) method,32 requested both with the
QST2 (two input structures) and QST3 (three input structures)
formalism. Frequency analysis was carried out subsequently to make
sure that true first order saddle points were located, giving rise to a one
negative eigenvalue. The calculations were carried out with the
GAUSSIAN 09 suite of programs.33
89.5 °C; [α]20 = −26 (c 0.80 in CHCl3); 1H NMR (300 MHz,
D
CDCl3) δ 7.48−7.29 (m, 5H), 5.73 (dt, J = 16.6, 8.4 Hz, 1H), 5.21−
5.03 (m, 5H), 4.48 (br s, 1H), 4.07 (d, J = 13.3 Hz, 1H), 3.27 (dd, J =
12.8, 6.3 Hz, 1H), 2.89 (t, J = 12.5 Hz, 1H), 2.54−2.28 (m, 2H),
1.88−1.32 (m, 8H), 1.17 (br s, 9H); 13C NMR (101 MHz, CDCl3) δ
155.5 (C), 136.9 (C), 134.0 (CH), 128.6 (CH), 128.1 (CH), 128.0
(CH), 119.2 (CH2), 67.2 (CH2), 56.0 (C), 53.0 (CH), 47.8 (CH),
40.7 (CH2), 39.6 (CH2), 34.7 (CH2), 27.8 (CH2), 25.6 (CH2), 22.7
(CH3), 18.9 (CH2); IR (ATR) ν 3219, 3032, 2947, 1664, 1642, 1438,
1267, 1056 cm−1; LRMS (EI) m/z (%) 350 (M+ − C4H8, 7), 218 (12),
174 (30), 91 (100), 84 (20); HRMS (EI) m/z calcd for C22H34N2O3S
− C4H8 350.1664, found 350.1680; HPLC (same conditions described
for 3a) tR for major diastereoisomer 19.25 min, tR for the minor
diastereoisomer 17.88 min.
Tetraponerine T3. To a solution of compound 3a (510 mg, 1.26
mmol) in THF (3.2 mL) was added dropwise aqueous 6 M HCl (628
μL, 3.77 mmol) at 0 °C under Ar. The reaction mixture was stirred for
1 h while reaching 23 °C. Aqueous 2 M NaOH (5 mL) was added to
the mixture, and the free amine was extracted with EtOAc (3 × 10
mL) and washed with brine (1 × 10 mL). Organics were dried over
MgSO4, filtered, and concentrated under reduced pressure. The
residue was then disolved in dry MeOH (24 mL), and Pd/C 10% (420
mg) was added to the mixture. The suspension was shacked under
hydrogen atmosphere (4 atm) for 12 h at 23 °C and filtered though
Celite, and the obtained solution was concentrated under reduced
pressure. The residue (free diamine) was then dissolved in dry CH2Cl2
(13 mL), and K2CO3 (520 mg, 3.77 mmol) was added, followed by 4-
bromobutanal25 (285 mg, 1.89 mmol). The mixture was stirred at 23
°C for 4 h, after which time inorganic salts were removed by filtration.
The filtrate was washed with aqueous NaHCO3, followed by brine, and
then dried over MgSO4. Organics were concentrated under reduced
pressure, and the residue was purified by column chromatography
10345
dx.doi.org/10.1021/jo302045y | J. Org. Chem. 2012, 77, 10340−10346

The Journal of Organic Chemistry
Article
(15) The actual optimized geometries, absolute energies, zero-point
energies, enthalpies, and free energies under standard conditions (1
atm, 298.15 K) are available for all of the structures in the Supporting
Information.
(16) The accuracy of these results has been estimated by comparison
with a preliminary study at the B3LYP/6-31G (d) level. Both
calculations have identical profiles, with an average absolute deviation
of the energies in the whole set of 0.4 and 0.7 kcal mol−1 for T3
and T4 respectively.
(17) Analogues of this configuration have been apparently used to
represent the structure of some tetraponerines in the past. See: Kim, J.
T.; Gevorgyan, V. Org. Lett. 2002, 4, 4697−4699 (Supporting
Information).
(18) For a sample of 16 structures from Charts 1 and 2, only those
with nondeformed chair geometries in rings A and B were included.
(19) (a) NBO Version 3.1; Glendening, E. D.; Reed, A. E.; Carpenter,
J. E.; Weinhold, F. (b) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J.
Chem. Phys. 1985, 83, 735−46. (c) Reed, A. E.; Weinhold, F. J. Chem.
Phys. 1985, 83, 1736−40.
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Copies of H and 13C NMR spectra for all new or relevant
compounds prepared, as well as HPLC traces of 3a/3b and
GC−MS for T3 and T4, are provided. Optimized geometries
and zero point energies (Ha), as well as thermodynamic data
for all calculated structures by DFT is also given. This material
is available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: josecarlos.gonzalez@ua.es; aguijarro@ua.es.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(20) A detailed account of these specific interactions for the main
configurations (those in Figure 2) can be found in the Supporting
Information S27−S28.
(21) Gribble, G. W.; Nelson, R. B. J. Org. Chem. 1973, 38, 2831−
2834.
We thank the Spanish Ministerio de Ciencia e Innovacion
́
(Grant No. CTQ2007-65218, Consolider Ingenio 2010-CSD-
2007-00006 and CTQ2011-24165), the Generalitat Valenciana
(Grant No. PROMETEO/2009/039 and FEDER), and the
University of Alicante for their financial support. I.B. also
thanks the Generalitat Valenciana for a predoctoral fellowship
(ACIF/2011/159).
(22) We shall limit the description to the structures within 3 kcal
mol−1 from the absolute minimum.
(23) We determined this ΔG° from the calculated standard free
energies taking into consideration that both T3 and T4 are actual
mixtures of conformations:
DEDICATION
■
°
°
i
j
ΔG° = −RT[ln
e−G (T4)/RT − ln
e−G (T3)/RT
]
∑
∑
Dedicated to the memory of Professor Balbino Mancheno.
̃
i
j
(24) Interestingly, the nitrogen inversion for T4 is ca. 3 kcal mol−1
higher than for T3. During the inversion process, the C3 has to
overcome a gauche interaction with the propyl group in T4, which is
missing in T3.
REFERENCES
■
(1) Braekman, J. C.; Daloze, D.; Pasteels, J. M.; Vanhecke, P.;
Declercq, J. P.; Sinnwell, V.; Francke, W. Z. Naturforsch. 1987, 42c,
627−630.
(25) 4-Bromobutanal is commercially available, but we prepared it by
DIBAL-H reduction of the corresponding methyl esther.
(26) (a) Hohenberg, P.; Kohn, W. Phys. Rev. 1964, 136, B864−B871.
(b) Kohn, W.; Sham, L. J. Phys. Rev. 1965, 140, A1133−A1138.
(27) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(b) Stephens, P. J.; Devlin, F. J.; Chablowski, C. F.; Frisch, M. J. J.
Phys. Chem. 1994, 98, 11623−11627.
(2) Renson, B.; Merlin, P.; Daloze, D.; Braekman, J. C.; Roisin, Y.;
Pasteels, J. M. Can. J. Chem. 1994, 72, 105−109.
(3) Devijver, C.; Braekman, J. C.; Daloze, D.; Pasteels, J. M. Chem.
Commun. 1997, 661−662.
(4) Yue, C.; Royer, J.; Husson, H.-P. J. Org. Chem. 1990, 55, 1140−
1141.
(5) Macours, P.; Braekman, J. C.; Daloze, D.; Pasteels, J. M.
Tetrahedron 1995, 51, 1415−1428.
(28) Riley, K. E.; Op’t Holt, B. T.; Merz, K. M., Jr. J. Chem. Theory
Comput. 2007, 3, 407−433.
(6) Structures of T5 and T6 were also confirmed by chemical
synthesis in: Devijver, C.; Macours, P.; Braekman, J. C.; Daloze, D.;
Pasteels, J. M. Tetrahedron 1995, 51, 10913−10922.
(7) Kem, W. R.; Wildeboer, K.; LeFrancois, S.; Raja, M.; Marszalec,
W.; Braekman, J. C. Cell Mol. Neurobiol. 2004, 24, 535−551.
(8) Rouchaud, A.; Braekman, J. C. Eur. J. Org. Chem. 2009, 2666−
2674.
(29) Tirado-Rives, J.; Jorgensen, W. L. J. Chem. Theory Comput. 2008,
4, 297−306.
(30) Zhao, Y.; Truhlar, D. G. J. Chem. Theory Comput. 2007, 3, 286−
300.
(31) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.
1980, 72, 650−654.
(32) Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J. J. Comput.
Chem. 1996, 17, 49−56.
(9) For selected syntheses of enantioenriched tetraponerines, see:
(a) Chongwei, Y.; Gauthier, I.; Royer, J.; Husson, H.-P. J. Org. Chem.
1996, 61, 4949−4954. (b) Takahata, H.; Kubota, M.; Ikota, N. J. Org.
Chem. 1999, 64, 8594−8601. (c) Stragies, R.; Blechert, S. J. Am. Chem.
Soc. 2000, 122, 9584−9591. (d) Airiau, E.; Girard, N.; Pizzeti, M.;
Salvadori, J.; Taddei, M.; Mann, A. J. Org. Chem. 2010, 75, 8670−8673.
(10) The use of 4-bromobutanal was already reported for the
(33) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A.; Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J. Fox, D. J. Gaussian 09,
Revision A.02; Gaussian, Inc.: Wallingford, CT, 2009.
synthesis of ( )-T8 using a different approach: Barluenga, J.; Tomas
M.; Kouznetsov, V.; Rubio, E. J. Org. Chem. 1994, 59, 3699−3700.
(11) (a) Gonzalez-Gomez, J. C.; Medjahdi, M.; Foubelo, F.; Yus, M.
J. Org. Chem. 2010, 75, 6308−6311. (b) Gonzalez-Gomez, J. C.;
Foubelo, F.; Yus, M. Org. Synth. 2012, 89, 88−97.
(12) Bosque, I.; Gonzalez-Gomez, J. C.; Foubelo, F.; Yus, M. J. Org.
́
,
́
́
́
́
́
́
Chem. 2012, 77, 780−784; Corrigenda: J. Org. Chem. 2012, 77, 4190.
(13) Yu, W.; Mei, Y.; Kang, Y.; Hua, Z.; Jin, Z. Org. Lett. 2004, 6,
3217−3219.
(14) This was expected since the dC‑N (C5−N4 bond lengths) are
customarily some 4−9% shorter than the dC‑C (C5−C6 bond lengths).
10346
dx.doi.org/10.1021/jo302045y | J. Org. Chem. 2012, 77, 10340−10346