Synthesis and characterisation of medium-sized ring systems by oxidative cleavage. Part 2: Insights from the study of ring expanded analogues

Article abstract of DOI:10.1016/j.tet.2010.10.043

Variable temperature NMR analysis and computational methods have been used to develop a detailed understanding of the ¹H NMR spectra of a family of medium-sized ring containing compounds. The family consists of analogues containing 10-, 11- and

Full text of DOI:10.1016/j.tet.2010.10.043

Tetrahedron 66 (2010) 9694e9702
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Synthesis and characterisation of medium-sized ring systems by oxidative
cleavage. Part 2: Insights from the study of ring expanded analogues
*
*
Tomas Lebl , Magali M. Lorion, Alan M. Jones, Douglas Philp, Nicholas J. Westwood
School of Chemistry and Biomedical Sciences Research Complex, University of St Andrews, St Andrews, Fife KY16 9ST, Scotland, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 July 2010
Received in revised form 28 September 2010
Accepted 18 October 2010
Available online 11 November 2010
Variable temperature NMR analysis and computational methods have been used to develop a detailed
understanding of the 1H NMR spectra of a family of medium-sized ring containing compounds. The
family consists of analogues containing 10-, 11- and 12-membered rings and in all cases the NMR spectra
at room temperature showed a series of diastereotopic methylene signals despite the lack of a stereo-
genic centre in these systems. On repeating the NMR analysis at higher temperatures, all the signals
coalesced for the 12-membered ring system consistent with full interconversion of ring conformers. This
was not observed in the analogous 10- and 11-membered ring systems with the interchange of con-
formers remaining slow on the NMR timescale. However, 1D gs-NOESY/EXSY NMR experiments showed
that in the smaller ring systems interconversion of diastereotopic protons did occur. Computational
studies suggest that the dynamic process observed by NMR for the 10- and 11-membered rings systems
is different from that observed in the 12-membered ring containing compound.
Keywords:
Heterocycle
Oxidative fragmentation
Medium-sized ring
VT NMR analysis
Mechanism
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
makes a contribution to the complexity of the 1H NMR spectrum of
1a at room temperature, variable temperature NMR studies showed
There remains significant interest in the search for chemical
scaffolds that exhibit restricted conformational freedom and hence
display functional groups in well-defined orientations.1 Medium-
sized ring systems are considered useful in this context due to the
often strong conformational preferences they exhibit.2,3 Importantly,
very few examples of medium-sized ring containing compounds are
presentinthecommercialcompoundcollectionsthatdominatemany
high-throughput screening programmes. For this reason and others,
synthetic routes to access medium-sized rings are of continued im-
portance. To date, our focus has been on the use of oxidative frag-
mentation reactions as a means of accessing these types of systems.
In the previous paper,4 we revisited the synthesis of the me-
dium-sized ring containing compound 1a by oxidative fragmenta-
tion of the known anti-psoriasis drug 2a (Fig. 1A).5 Compound 1a
had a relatively complex 1H NMR spectrum containing a series of
diastereotopic protons associated with the methylene groups
present in the 10-membered ring. This observation implied that 1a
contains at least one non-classical stereogenic axis. One possible
explanation was that this stereogenic axis resulted from restricted
rotation about the N12eC10-bond in 1a and whilst this clearly
that the situation is more complex as the C7eC10 protons remained
diastereotopic at temperatures well above that required for co-
alescence of the aromatic signals in 1a. In addition, the small
molecule X-ray crystal structure of 1a showed that in the unit cell
four different conformers were represented.6 Two of these struc-
tures ((i) and (ii), Fig. 1B) are non-superimposable mirror images.
The other two ((iii) and (iv)) correspond to a second enantiomeric
pair in which the chlorine atom is on the opposite face of the
molecule relative to (i) and (ii), respectively. The relative orienta-
tion of the three carbonyl groups is the same in all four conformers.
Whilst the observation of restricted bond rotation in medium-
sized ring systems is relatively common,2,7 we became interested in
exploring this system further as a function of ring size. The results
of these studies are presented here including a detailed discussion
of variations in the ability of the amide group to reorient itself as
the size of the ring varies.
2. Results and discussion
Our previously reported4 variable temperature NMR studies on 1a
showed a considerable broadening of the two signals associated with
the diastereotopic C7 protons as a function of temperature (Fig. 2(a)
and (b), Fig. 1A for numbering and Ref. 4). Broadening of these reso-
nances at higher temperatures is probably consistent with a re-
laxation of restrictions to bond rotation in the 10-membered ring of
* Corresponding authors. E-mail addresses: tl21@st-andrews.ac.uk (T. Lebl),
njw3@st-andrews.ac.uk (N.J. Westwood).
0040-4020/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2010.10.043

T. Lebl et al. / Tetrahedron 66 (2010) 9694e9702
9695
Fig. 1. A. Synthesis of the medium-sized ring containing compounds 1aec. aA small increase in yield was observed when 4.0 equiv of m-CPBA was used. B. The four independent
conformers that make up the unit cell in the X-ray crystal structure analysis of 1a.
1a. In order to characterise this further, it was decided to repeat the
NMR analysis on analogues of 1a that contained larger ring systems
(11- or 12-membered ring in 1b and 1c, respectively, Fig. 1A).
2.1. Synthesis of ring expanded analogues of 1a
The required analogues 1b and 1c were synthesised in an
analogous manner to 1a (Fig. 1A). The enamines 3b and 3c were
prepared according to a literature procedure from cycloheptanone
(4b) and cyclooctanone (4c), respectively, and were converted to
the desired pyridones 2b and 2c.8 Oxidative cleavage of 2b and 2c
was achieved on reaction with m-CPBA, with the 11-membered
macrocycle 1b being afforded in a similar yield (72%) to that
obtained for the 10-membered macrocycle 1a (65%). However,
synthesis of the 12-membered ring macrocycle 1c proved much
more difficult, apparently due to an overall decrease in the re-
activity of 2c with m-CPBA. When 2 equiv of m-CPBA was used, 1c
was isolated in 24% yield and a second product was obtained in 30%
yield. 1H NMR analysis indicated that this new product contained
an aldehyde functional group. Further NMR studies led to the as-
signment of this product as 5 and a plausible mechanism for for-
mation of 5 from 2c is shown in Scheme 1. It seems likely that when
n¼1 (2a) the corresponding iminium ion 6 is rapidly intercepted by
a second equivalent of m-CPBA to give 7 (probably as a single syn-
diastereomer9). However, if in the case of the eight-membered ring,
this attack is slowed due to increased steric hindrance to attack of
the nucleophile, conversion of iminium ion 8 to enamine 9 may
compete. Subsequent oxidative cleavage of the enamine double
bond can then occur to give 5. A small increase in the yield of 1c
from 2c was observed when 4 equiv of m-CPBA was used.
In Part 1,4 it was shown that whilst it was possible to prepare 1a
from 2a in a catalytic ruthenium tetroxide-mediated oxidative
cleavage reaction, the major product formed in this reaction was
often the syn-diol 10 (Scheme 1). Under conditions that gave only
10 from 2a, it was found that 2c was converted to 1c in 16% yield
(73% based on recovered starting material 2c) with no indication of
the corresponding diol 11 being present. Whilst an increase in the
catalyst loading to 5 mol %, resulted in an increase in the isolated
Fig. 2. Comparison of the aliphatic region of the 1H NMR (300 MHz) spectra for 1a, 1b
and 1c recorded in toluene-d8 at 293 and 363 K showing broadening or coalescence of
the C7-diastereotopic proton resonances.

9696
T. Lebl et al. / Tetrahedron 66 (2010) 9694e9702
Table 1
Chemical shift differences of diastereotopic protons in ppm derived from 1H,13C-
HSQC of 1c at 233 K and calculated for the ground state of 1c at the B3LYP/6-31G**
level of theory
Experimental
B3LYP/6-31G**
C7
C8
C9
C10
C11
C12
1.97
0.84
0.21
0.45
1.16
0.16
1.51
0.76
0.66
0.68
0.88
0.48
Raising the temperature at which the 1H NMR analysis was
carried out to 363 K further emphasised the difference in behaviour
between the 10- and 11-membered systems, 1a and 1b, as com-
pared to the 12-membered system, 1c. In the case of 1c, a 1H,13C-
HSQC spectrum recorded at 363 K showed that the signals corre-
sponding to all the diastereotopic protons in 1c coalesced on
heating the sample (see Figs. 2e and S211). Therefore the dynamics
of the 12-membered ring are such that full interconversion of
conformers occurs. On the other hand, the signals in the aliphatic
region of the spectrum of 1b remained separate and broadened as
previously described4 for 1a (Fig. 2c vs Fig. 2a).
Since broadening of lines in the NMR spectrum as a function of
temperature did not allow for unambiguous conclusions to be made
concerning the dynamics of the 10- and 11-membered ring com-
pounds, 1a and 1b, further studies using a 1D gs-NOESY/EXSY ex-
periment were carried out at various temperatures (Fig. 3). For both
1a and 1b qualitatively similar results were obtained. At room
temperature antiphase peaks due to a positive NOE were observed
(Fig. 3a for 1a, data for 1b not shown). However, at higher temper-
atures the resonances associated with H7a and H7b appeared in
phase indicating that chemical exchange of those two diastereotopic
protons was occurring (Fig. 3b). As expected, the intensity of the
Scheme 1. A possible mechanism for the formation of 5.
yield of 1c to 37% isolated yield (66% based on recovered starting
material 2c), 11 was still not observed. The subtle difference in
behaviour of 2a and 2c in this reaction is currently rationalised
either based on differences in the rate of cleavage of the corre-
sponding diols 10 and 11 by sodium periodate (which is used as
a co-oxidant) or by a change in mechanism with, for example, direct
cleavage10 of the double bond in 2c occurring.
2.2. Variable temperature NMR study of the 10, 11 and 12-
membered ring containing systems, 1aec
A comparison of the aliphatic region of the 1H NMR spectra
obtained for the three medium-sized ring systems in 1a, 1b and 1c
as a function of temperature is shown in Fig. 2.
The results at 293 K (Fig. 2(b), (d) and (f)) indicate that the dy-
namics of this system vary as a function of ring size with a signifi-
cant transition taking place as the ring size is increased from 11 to
12 (cf. Fig. 2d and f). In 1c (12-membered ring) the aliphatic signals
are very broad with the signals assigned to the C7-diastereotopic
protons (labelled 7a and 7b) being barely visible, whereas the
analogous signals in 1b (11-membered ring) are similar to those
observed for 1a. When the 1H,13C-HSQC spectrum of 1c was
recorded at lower temperature (233 K, see Fig. S1)11 six sets of
diastereotopic aliphatic protons were observed with the spectrum
being further complicated by the presence of two rotamers (2:1
ratio) resulting from restricted rotation about the N(14)eC(10) bond
at this temperature (see Fig. 1A for numbering system). The largest
and smallest differences in chemical shifts (1.97 and 0.16 ppm)
were associated with the C7 and C12 diastereotopic protons, re-
spectively. Therefore, a preliminary explanation for this observa-
tion was that the diastereotopic nature of these protons stems from
restricted rotation about the C5eC6 carbon bond of the dicarbonyl
functionality in 1c (Fig. 1A). However, further assignment of the 13C
resonances of the other methylene carbons in 1c using a 1H,13C-
H2BC experiment revealed that the diastereotopic pair with the
second largest chemical shift difference belonged to C11.11 The
situation was clearly more complex than the initial explanation.
Therefore, further studies using NMR and computational methods
were carried out and are described below. Interestingly, the
chemical shift differences of the diastereotopic protons calculated
for the ground state of 1c at the B3LYP/6-31G** level of theory
showed a reasonable agreement with the observed trend (Table 1)
providing evidence that this level of theory is relevant to the study
of these systems, as expected.
Fig. 3. 1D gs-NOESY/EXSY spectra recorded for 1a following selective irradiation of the
H7a resonance. (a) At 293 K the H7b peak has the opposite phase to the H7a peak as
a result of a positive NOE (mixing time¼0.4 s); (b) At 310 K the H7a and H7b peaks appear
in the spectrum in phase because magnetisation transfer due to chemical exchange
outweighs the NOE effect (mixing time¼0.4 s); (c) At 314.5 K a faster rate of exchange and
longer mixing time (0.5 s) results in a stronger enhancement of the H7b peak.

T. Lebl et al. / Tetrahedron 66 (2010) 9694e9702
9697
EXSYpeaks provedtobe dependenton both the temperatureand the
mixing time (Fig. 3c).
Unfortunately, a rather unsatisfactory agreement between experi-
mental and theoretical data was obtained for the entropy of acti-
vation
D
Sz even when all vibrational entropy contributions from
frequencies below 300 cmꢁ1 were removed (cf. Table 2, column 4 vs
column 7).
2.3. Rationalisation of the trends observed in the variable
temperature NMR studies
The studies described above indicated that the dynamics of this
system are very dependent on the size of the ring. The next step in
providing a full rationalisation was to calculate the activation pa-
rameters for the dynamic processes associated with 1aec and to
assess possible pathways to explain these using calculations at the
B3LYP/6-31G** level of theory.
2.5. Dynamic processes associated with the medium-sized
ring
2.5.1. Identification of the activation parameters. The resonances
associated with the C7-diastereotopic protons in the 1H NMR
spectra of 1a, 1b and 1c recorded between 278 and 373 K were
subjected to total lineshape analysis. No interference caused by
restricted rotation about the N(12/13/14)eC(10) bond occurred as
this process was already in the fast exchange limit in this temper-
2.4. Restricted rotation about the N(12/13/14)eC(10) bond in
1a, 1b and 1c
ature range. The resulting activation parameters
D
Gz
, D
Hz298 and
298
D
Sz
are presented in Table 3. In the case of 1a and 1b, the tem-
The activation parameters for restricted rotation about the N(13/
14)eC(10) bond in 1b and 1c were determined by total lineshape
analysis of the C20 and/or C60 proton resonances and calculated at
the B3LYP/6-31G** level of theory to be compared with the values
already obtained for 1a (Table 2).4,11
298
perature range over which broadening of resonances was suitable
for lineshape analysis was narrow and poor R2 values resulted from
the Eyring plots.11 Therefore, quantitative EXSY spectra were also
used for the calculation of the activation parameters. Very good
agreement between the two methods was obtained for
D
Hz
298
(Table 3, column 3 vs column 6) and whilst values showed
D
Sz
298
Table 2
Activation parameters
restricted rotation about the N(12/13/14)eC(10) bond determined experimentally
using total lineshape analysis of 1H NMR resonances of aromatic protons C20 and/or
C60 and calculated at the B3LYP/6-31G** level of theory
considerable discrepancies, the errors resulting from the EXSY data
were significantly smaller (Table 3, columns 4 and 7). All activation
parameters clearly confirmed the difference in behaviour between
the 12-membered system, 1c, as compared to the 10- and 11-
membered systems, 1a and 1b.
D
Gz
[kJ molꢁ1], Hz [kJ molꢁ1] and Sz [kJ Kꢁ1 molꢁ1] for
D D
298
Experimental
Gz
B3LYP/6-31G**
Gz Hz
D
D
Hz
D
Sz
D
D
D
Sza
298
298
2.5.2. Detailed assessment of 1c. With the experimentally de-
termined activation parameters in hand, it was decided to in-
vestigate the dynamic processes associated with the medium-sized
ring using computational techniques. As a starting point, it was
assumed that the dynamic process being observed in the 1H VT
NMR experiment for 1c corresponded to the interconversion of
enantiomeric conformers of 1c (e.g., between conformers (i) and
(ii), Fig. 4). The main step in this process was proposed to involve
the C13eN14 amide passing through the 12-membered ring al-
though flipping of the C5eC6 dicarbonyl unit is also required. Initial
attempts to identify a reasonable pathway for the movement of the
amide by driving the C14aeN14eC13eO13 torsion angle (amide
bond rotation through the ring) in conformer (i) of 1c were dis-
couraging (see structure in Fig. 1A for numbering). However,
progress was made when an initial flip of the C5eC6 dicarbonyl
functionality was attempted. This dicarbonyl flip was briefly de-
scribed in Part 14 for the 10-membered ring in 1a and in the case of
1c was calculated by driving the O5eC5eC6eO6 torsion angle us-
ing the semi-empirical RM1 method and refined at the B3LYP/6-
31G** level of theory. The calculations confirmed that conversion of
conformer (i) to intermediate conformer (I1) via transition state
(TS1) for 1c (Fig. 4) was an energetically uphill but feasible process,
as was the case for 1a.4 Importantly, starting at conformer (I1), it
was now possible to identify a reasonable path by which the
1a
1b
1c
55.6ꢀ0.1
50.7ꢀ0.1
51.9ꢀ0.9
56.1ꢀ0.7
55.5ꢀ1.0
54.8ꢀ1.4
1.7ꢀ2.8
16.3ꢀ3.9
9.6ꢀ5.8
60.8
59.0
55.7
61.0
60.8
54.3
0.8
6.0
ꢁ4.7
a
All vibrational modes with frequencies below 300 cmꢁ1 were ignored in cal-
culating activation entropy.
The trend in the Gibbs free energy of activation (D
Gz298) shows
that the number of methylene groups has only a small effect on the
restricted rotation about the N(12/13/14)eC(10) bond (Table 2,
column 2). The calculated values of activation enthalpy D
Hz (Table
2, column 6) are in reasonably good agreement with the experi-
mental values (Table 2, column 3), providing further support for the
use of these computational methods in the context of this system.
Surprisingly, however, the activation entropies for the restricted
rotation about the N(12/13/14)eC(10) bond ( Sz, column 4) were all
D
positive. One possible explanation for this comes from comparison
of the calculated ground and transition state geometries of 1aec
(Fig. S6 and Table S2).11 When the transition states of 1aec are
reached, the average increase in the N(12/13/14)eC(11/12/13) and
N(12/13/14)eC(12a/13a/14a) bond lengths within the medium-
ꢀ
sized ring was calculated to be 0.0150ꢀ0.0032 and 0.0185ꢀ0.003 A,
respectively. This may in turn result in a higher conformational
flexibility of the ring and consequently to higher entropy.
Table 3
Activation parameters
D
Gz
[kJ molꢁ1], Hz [kJ molꢁ1] and Sz [J.Kꢁ1.molꢁ1] for the dynamic process associated with the medium-sized ring in 1aec were determined
D D
298
experimentally using total lineshape analysis and the intensities of the EXSY crosspeaks between the 1H NMR resonances associated with the C7-diastereotopic protons.
Calculated values were generated at the B3LYP/6-31G** level of theory and are given for the corresponding rate determining step (RDS), which varies for each particular
analogue (see main text)
Total lineshape analysis
Gz Hz
EXSY
Gz
B3LYP/6-31G**
RDS
D
D
D
Sz
D
D
Hz
D
Sz
D
Gz
298
D
Hz
D
Sz
298
298
1a
78.4ꢀ0.1
76.8ꢀ1.7
ꢁ5.6ꢀ4.9
76.5ꢀ0.1
79.8ꢀ0.3
11.1ꢀ1.0
(i)/(TS1)
72.5
71.7
79.8
61.8
70.6
66.9
73.5
53.9
ꢁ7.0
ꢁ15.9
ꢁ21.0
ꢁ26.3
(I1)/(TS20)
(I1)/(TS20)
(I1)/(TS2)
1b
1c
80.2ꢀ0.4
58.8ꢀ0.1
77.7ꢀ4.9
49.5ꢀ0.6
ꢁ8.5ꢀ13.6
ꢁ31.1ꢀ2.1
80.5ꢀ0.1
78.4ꢀ1.2
ꢁ7.0ꢀ3.6
d
d
d

9698
T. Lebl et al. / Tetrahedron 66 (2010) 9694e9702
Fig. 4. The fixed amide rotation pathway. Ball and stick representation of ground and transition state structures involved in this pathway for 1c localised using semi-empirical RM1
method and refined at B3LYP/6-31G** level of theory. The Gibbs free energy profile calculated for this pathway at the same level of theory.
C13eN14 amide could pass through the 12-membered ring by
progressive driving of the C4aeC14aeC13eO13 torsion angle
fromþ30ꢂ to ꢁ70ꢂ. The geometry of the transition state (TS2) for
this process was refined at the B3LYP/6-31G** level of theory and
shows that during this process the C13N14-amide in 1c travels
through the medium-sized ring (the C4aeC14aeC13eO13 torsion
angle in (TS2)¼ꢁ0.9ꢂ) without distortion of the amide away from
planarity (C14aeN14eC13eO13¼ꢁ4.7ꢂ, Fig. 4). For this reason, this
pathway is subsequently referred to as the fixed amide rotation
pathway. The calculated Gibbs free energy profile for this overall
process is depicted in Fig. 4 and suggests that transition from
conformer (I1) to (ii) is the rate determining step for the in-
terconversion of the enantiomeric conformers of 1c. As the half-life
(t1/2) for interconversion of (i) to (ii) for 1c was calculated to be
2.3 ms at 298 K then 1c does not exist as atropisomers (as defined
by Oki12). The corresponding activation parameters calculated at
the B3LYP/6-31G** level of theory for the fixed amide rotation
pathway are in very good agreement with the experimentally de-
termined values (Table 3).
of both (TS1) and (I1) with respect to the original global minimum
ground state (i) are shown as a function of ring size in Table 4) al-
though it is energetically feasible for all three compounds 1aec.
Table 4
Relative Gibbs free energies of transition states and intermediate conformers with
respect to the global minimum ground state (i) calculated at the B3LYP/6-31G** level
of theory for analogues 1aec
DG298 [kJ mol]
(i)
(TS1)
(I1)
(TS2)
(ii)
(TS20)
(I20)
(TS30)
1a
1b
1c
0.0
0.0
0.0
72.5
58.8
44.2
31.0
24.6
12.1
N/A
110.1
73.9
0.0
0.0
0.0
102.7
104.4
96.8
44.5
29.8
31.1
101.6
97.8
109.1
Next, attempts were made to force the amide group through the
11-membered ring in 1b. In this case, calculations showed that the
fixed amide rotation pathway described for 1c is plausible for 1b
although there is
a significantly higher activation barrier
Having rationalised the activation parameters associated with
1c, an analogous explanation (initial dicarbonyl flip followed by
passing of the amide through the medium-sized ring) was com-
putationally assessed for both 1a and 1b. As expected, the initially
required dicarbonyl flip was energetically feasible for 1b and 1a
(see Part 14). However, the dicarbonyl flip was calculated to be more
difficult as the ring size decreased (the relative Gibbs free energies
(
D
Gz ¼85.5 kJ molꢁ1, Fig. 6). This value is reasonably consistent
298
with the
D
Gz
value observed in the VT experiments
298
(
D
Gz ¼80.5 kJ molꢁ1, Table 3, column 5). By contrast, attempts to
298
locate the analogous transition state for the 10-membered ring
system 1a using semi-empirical RM1 methods resulted in un-
realistic structures with relative energies above 200 kJ molꢁ1. As
EXSY experiments had clearly shown that the C7-diastereotopic

T. Lebl et al. / Tetrahedron 66 (2010) 9694e9702
9699
protons in 1a could interconvert, an alternative rationalisation to
the fixed amide rotation pathway was required for 1a and possibly
1b.
transecisetrans amide bond rotation pathway are very close in
value (
D
Gz ¼72.5 and 71.7 kJ molꢁ1, respectively) and are sig-
298
nificantly larger than for the final step in the pathway
(
D
Gz ¼57.1 kJ molꢁ1). Although the similarity in
D
Gz
for the
298
298
2.5.3. Detailed assessment of 1a. As conformer (I1) had proved
a more productive starting point in the case of 1c (Fig. 4), a detailed
search of the conformational behaviour of the analogous I1 con-
former of 1a using semi-empirical RM1 method was carried out.
Interestingly, it was found that the mirror image of the 1a ground
state geometry could be accessed from conformer (I1) by pro-
gressive driving of O11eC11eN12eC10 dihedral angle from ꢁ170ꢂ
to 170ꢂ (see Fig. 1A for numbering). This process can be viewed as
a ‘transecisetrans’ amide bond rotation which flips both the
C11N12-amide carbonyl and the N12-chlorophenyl ring in 1a. The
amide moiety in both the global minimum ground state (i) and
two steps does not allow for an unambiguous determination of the
rate determining step, the
D
Hz
and
D
Sz
values calculated for
298
298
the first step are in better agreement with the experimental values.
2.5.4. Reassessment of 1c. To date two alternative pathways, the
fixed amide and the transecisetrans amide bond rotation pathways,
have been identified by which enantiomeric conformers in this
series can interconvert. The most clear situation is for the 10-
membered ring containing 1a, where the only energetically feasible
pathway is the transecisetrans amide bond rotation pathway de-
scribed in Fig. 5. Reanalysis of 1c showed that this pathway was also
feasible for 1c (see Table 4) but that the activation barriers for this
pathway are significantly higher for 1c than those calculated for the
fixed amide rotation pathway (Fig. 4). It was therefore concluded
that, whilst both 1a and 1c are both capable of interconverting
enantiomeric conformers, they do so by different pathways. Next
the focus turned to 1b, which contains an 11-membered ring in-
termediate in size between the rings in 1a and 1c.
the intermediate conformer (I1) adopts
a trans-orientation
(O11eC11eN12eC10¼ꢁ166.2ꢂ and ꢁ167.9ꢂ, respectively). The
transformation is calculated to be achieved through two transition
states (TS20) and (TS30) and the intermediate conformer (I20), which
has a cis-configured amide (O11eC11eN12eC10¼4.8ꢂ, Fig. 5). The
relative Gibbs free energies with respect to the global minimum
ground state (i) obtained after refinement at the B3LYP/6-31G**
level of theory are summarised for 1a in Table 4. The overall Gibbs
free energy profile for interconversion of the enantiomeric con-
formers of 1a calculated at the B3LYP/6-31G** level of theory shows
three activation barriers (Fig. 5). In this process, flipping of the
C5eC6 dicarbonyl group also starts the interconversion process as
the activation barrier for rotation about the C11eN12 amide bond
in conformer (i) is some 30 kJ molꢁ1 higher in energy than that
from conformer (I1) for 1a (data not shown). The calculated Gibbs
free energy of activation for the first and second steps in the
2.5.5. Detailed assessment of 1b. Conformational intermediates
and transition states for both amide rotation pathways were cal-
culated for 1b. The resulting relative energies with respect to the
corresponding global minimum ground state (i) are summarised in
Table 4 and are depicted in Fig. 6. Unlike the situation for transition
state (TS1) and conformational intermediate (I1), the relative en-
ergies of both the transition states (TS20) and (TS30) are only slightly
affected by the ring size (Table 4). Consequently the Gibbs free
Fig. 5. transecisetrans amide bond rotation pathway: Ball and stick representation of ground and transition state structures for amide rotation of 1a localised using semi-empirical
RM1 method and refined at the B3LYP/6-31G** level of theory. The key torsion angle
terconversion of mirror images of 1a calculated at the B3LYP/6-31G** level of theory is also shown.
a
¼O11eC11eN12eC10 is highlighted by red dashed line. The free energy profile for in-

9700
T. Lebl et al. / Tetrahedron 66 (2010) 9694e9702
provided the amide bond in 1a undergoes trans- to cis- to trans-
isomerisation after a flip of the C5eC6 dicarbonyl group has oc-
curred. By this pathway (the transecisetrans amide bond rotation
pathway), which was shown to be energetically feasible for 1a, the
two enantiomeric conformers of 1a can interconvert without the
amide group passing through the 10-membered ring. Whilst 1a and
1c use two different pathways to interconvert enantiomeric con-
formers, the situation for the 11-membered ring containing ana-
logue 11b is less certain as both processes are calculated to be
energetically feasible. Our current studies suggest that 11b prefer-
entially uses the transecisetrans amide bond rotation pathway.
A combination of synthesis, NMR and computational methods
has led to an improved understanding of the conformational
equilibria associated with a set of medium-sized ring containing
analogues. Future reports will focus on the study of related systems
where interconversion of enantiomeric conformers is not possible.
Fig. 6. Gibbs free energy profile for the interconversion of the mirror images of 1b
calculated at the B3LYP/6-31G** level of theory.
energy for the transition from (I1) to (TS20) becomes larger for 1b
compared to the situation for 1a but the calculated barrier for 1b is
still smaller than the barrier calculated for the fixed amide rotation
pathway via transition state (TS2) for 1b (Fig. 6). Thus the most
favoured pathway for 1b is calculated to be the transecisetrans
amide bond rotation pathway seen for 1a, although a key difference
exist as the rate determining step is clearly the conversion of (I1) to
(I20) for 1b. Gratifyingly, the experimental and calculated values of
the activation parameters associated with 1b (Table 3), assuming
that the process observed in the VT NMR experiments is the con-
version of (I1) to (I20), are in good agreement.
4. Experimental11
4.1. General procedure A8
A solution of triethylamine (4.81 mL, 34.5 mmol, 1.10 equiv) and
the enamine (3a (13.6 g, 34.5 mmol, 1.10 equiv) or 3b (14.9 g,
34.5 mmol, 1.10 equiv) or 3c (16.1 g, 34.5 mmol, 1.10 equiv)) in tol-
uene (18.0 mL) was added dropwise over 10 min to a stirred, cooled
(iceeacetone, ꢁ10 ꢂC bath) suspension of crude 2-((3-chlorophenyl)
amino)-3-pyridinecarbonyl chloride (8.40 g, 31.4 mmol, 1.00 equiv)
in toluene (140 mL). As the solid dissolved, the dark red reaction
mixture was kept between ꢁ5 ꢂC and þ5 ꢂC for 2 h, then allowed to
warm up to room temperature over 30 min and heated at 80 ꢂC for
4 h. After cooling down, the reaction mixture was stirred at room
temperature for 12 h, after which the toluene was removed under
reduced pressure. The residue was dissolved in dichloromethane
(150 mL) and the solution was washed with H2O (100 mL), 1.0 M
HCl(aq) (100 mL), 1.0 M Na2CO3(aq) (100 mL) and with H2O (100 mL).
Aqueous extracts were back extracted with dichloromethane
(2ꢃ150 mL), and the combined organic portion was dried over
Na2SO4, filtered and concentrated in vacuo. The residue was purified
by column chromatography over silica gel (3:2, ethyl acetate/hex-
ane). An analytically pure sample was obtained by recrystallisation
from acetonitrile to afford the title compound.
3. Conclusion
The attempted synthesis of the 10-membered ring containing 1a
from 2a led to a compound that exhibited a relatively complex 1H
NMR at room temperature. In addition, X-ray crystallographic
analysis of 1a showed that two enantiomeric pairs of conformers
were present in the unit cell. These observations raised the possi-
bility that 1a existed as enantiomeric atropisomers, which in theory
could be isolated. It was initially proposed that this might result from
the fact that the amide carbonyl group in 1a was unable to pass
through the 10-membered ring, hence rendering interconversion of
enantiomeric conformers of 1a impossible. This naive view led to
a proposal that investigation of the ring-expanded versions of 1a,
compounds 1b (11-membered ring) and 1c (12-membered ring)
may cast further light on the system. Indeed it was found through
variable temperature NMR studies using 1c that interconversion of
the enantiomeric conformers of 1c was relatively fast with a t1/2 for
this process of approximately 2 ms. Furthermore it was shown by
computational methods that the interconversion of enantiomeric
conformers of 1c very likely occurred by passage of the amide bond
in 1c through the 12-membered ring, oreceded by a C5eC6 dicar-
bonyl flip. As the calculations suggest that the amide bond in 1c
remains in a trans-orientation throughout this process it has been
referred to here as the fixed amide rotation pathway. Intriguingly,
when the analogous process was explored computationally for 1a it
was shown that the fixed amide rotation pathway was impossible in
the 10-membered ring series.
4.2. General procedure B
To a solution of 2 (2a (311 mg, 1.00 mmol, 1.00 equiv) or 2b
(325 mg, 1.00 mmol, 1.00 equiv) or 2c (339 mg, 1.00 mmol,
1.00 equiv)) in anhydrous dichloromethane (16.0 mL) was added
portion-wise m-CPBA (345 mg, 2.00 mmol, 2.00 equiv) and the
reaction mixture was stirred at room temperature for 40 h before
being washed with 1.0 M Na2CO3(aq) (10.0 mL) and H2O (10.0 mL).
The aqueous washes were back extracted with dichloromethane
(2ꢃ30.0 mL). The combined organic layers were dried (MgSO4),
filtered and concentrated in vacuo. The residue was purified by
column chromatography over silica gel (1:1, ethyl acetate/hexane).
An analytically pure sample was obtained by recrystallisation from
ethanol.
Further light was cast on the 10-membered system 1a when
a
1D gs-NOESY/EXSY experiment demonstrated that in-
4.3. General procedure C
terconversion of the diastereotopic protons in 1a was occurring on
the NMR timescale. This observation was also consistent with the
half-life calculated from the activation parameters determined for
1a using VT NMR techniques (t1/2 (1a)¼4.3 s at 298 K). Whilst this
information argues that 1a cannot exists as atropisomers (as de-
fined by Oki12), the question of how the enantiomeric conformers
of 1a interconvert remained. Further computational studies led to
the identification of a pathway that could explain this observation
A solution of the cyclic ketone (cycloheptanone 4b (10.1 g,
90.0 mmol, 1.00 equiv) or cyclooctanone 4c (11.4 g, 90.0 mmol,
1.00 equiv)), pyrrolidine (8.26 mL, 99.0 mmol, 1.10 equiv) and
p-toluenesulfonic acid (catalytic 0.01 g) in anhydrous toluene
(200 mL) was refluxed under argon using a Dean and Stark appa-
ratus. Refluxing whilst stirring was continued until separation of
water had ceased (3b, 5 h and 3c, 20 h). The reaction mixture was

T. Lebl et al. / Tetrahedron 66 (2010) 9694e9702
9701
then cooled and the solution concentrated in vacuo. The residue was
purified by distillation.
procedure A. Mp 168e169 ꢂC; 1H NMR (300 MHz, CDCl3):
d
¼8.73
(dd, 3J¼7.9 Hz, 4J¼2.0 Hz, 1H, C4-H), 8.51 (dd, 3J¼4.5 Hz, 4J¼2.0 Hz,
1H, C2-H), 7.53e7.50 (m, 2H, AreH), 7.28e7.24 (m, 2H, AreH, C3-H),
7.17e7.13 (m, 1H, AreH), 3.08e2.93 (m, 2H, C6-H2), 2.64e2.59 (m,
2H, C10-H2), 1.87e1.79 (m, 2H, C8-H2), 1.67e1.60 (m, 2H, C7-H2),
4.3.1. 13-(3-Chlorophenyl)-8,9,10,11-tetrahydro-5H-pyrido[2,3-b][1]
azacycloundecine-5,6,12-(7H,13H)-trione (1b). The title compound
1b (257 mg, 0.720 mmol, 72%) was afforded as a yellow solid using
general procedure B.
1.59e1.52 (m, 2H, C9-H2); 13C NMR (75.5 MHz, CDCl3):
d
¼176.5
(C5), 155.4 (C10a), 151.9 (C2), 151.1 (C11a), 140.8 (C10), 136.1 (C4),
135.3 (C30), 130.7 (Ar), 129.8 (Ar), 129.4 (Ar), 127.7 (Ar), 124.9 (C5a),
119.6 (C3), 119.3 (C4a), 32.2 (C10), 32.0 (C8), 26.7 (C7), 25.5 (C9),
24.1 (C6); IR (KBr): nmax¼2932 (m), 1617 (s) (C]O), 1589 (m), 1542
(w), 1475 (m), 1425 (m), 1296 (m), 785 (m), 750 (w) cmꢁ1; LRMS
Mp 115e116 ꢂC; 1H NMR (300 MHz, CDCl3):
d
¼8.29 (dd,
3J¼4.9 Hz, 4J¼2.0 Hz, 1H, C2-H), 8.04 (dd, 3J¼7.7 Hz, 4J¼2.0 Hz, 1H,
C4-H), 7.34e7.30 (m, 1H, AreH), 7.29e7.23 (m, 2H, AreH), 7.22e7.17
(m, 1H, AreH), 7.14e7.09 (m, 1H, C3-H), 3.44e3.30 (m, 1H, C7-H),
2.65e2.54 (m, 1H, C70-H), 2.34e2.21 (m, 1H, C11-H), 2.07e1.98 (m,
1H, C110-H), 1.92e1.78 (m, 1H, C8-H), 1.69e1.46 (m, 4H, C80-H, C10-
H2, C9-H), 1.22e1.08 (m, 1H, C90-H); 13C NMR (75.5 MHz, CDCl3):
(ESþ): m/z (%) 347.13 (100) [M35ClþNa]þ; HRMS (ESþ): m/z calcd for
35
C19
H
ClN2NaO [M35ClþNa]þ: 347.0927; found: 347.0913.
17
d
¼201.5 (C6), 188.8 (C5), 175.5 (C12), 151.6 (C2), 151.4 (C13a), 142.6
4.3.4. 12-(3-Chlorophenyl)-6,7,8,9,10,11-hexahydrocycloocta[b][1,8]
naphthyridin-5(12H)-one (2c). The title compound 2c (5.43 g,
16.0 mmol, 51%) was afforded as a cream solid using general pro-
(C10), 139.2 (C4), 135.1 (C30), 130.5 (Ar), 129.9 (Ar), 128.8 (Ar), 128.0
(Ar), 126.0 (C4a), 121.8 (C3), 36.0 (C11), 34.3 (C7), 25.0 (C9), 22.5
(C10), 21.5 (C8); IR (KBr): nmax 2944 (m), 1710 (s) (C]O), 1686 (s)
(C]O), 1655 (s) (C]O), 1589 (s), 1475 (m), 1423 (m), 1364 (m), 1266
cedure A. Mp 190e191 ꢂC; 1H NMR (300 MHz, CDCl3):
d¼8.74 (dd,
3J¼7.6 Hz, 4J¼1.9 Hz, 1H, C4-H), 8.51 (dd, 3J¼4.6 Hz, 4J¼1.9 Hz, 1H,
C2-H), 7.55e7.50 (m, 2H, AreH), 7.31e7.25 (m, 2H, AreH, C3-H),
7.21e7.18 (m, 1H, AreH), 2.96e2.83 (m, 2H, C6-H2), 2.74e2.69 (m,
2H, C11-H2), 1.81e1.73 (m, 2H, C7-H2), 1.56e1.42 (m, 6H, C10-H2,
(s), 741 (s) cmꢁ1; LRMS (ESþ): m/z (%) 379.18 (100) [M35ClþNa]þ;
HRMS (ESþ): m/z calcd for C19
H
ClN2NaO3 [M35ClþNa]þ:
35
17
379.0835; found: 379.0825.
C8-H2, C9-H2); 13C NMR (75.5 MHz, CDCl3):
d
¼177.0 (C5), 151.9 (C2),
4.3.2. 14-(3-Chlorophenyl)-7,8,9,10,11,12-hexahydropyrido[2,3-b][1]
azacyclododecine-5,6,13-(14H)-trione (1c). Method 1: The title
compound 1c (89 mg, 0.24 mmol, 24%) was afforded as a yellow oil
using general procedure B with an additional 20 h stirring at room
151.7 (C12a), 151.6 (C11a), 140.6 (C10), 136.0 (C4), 135.3 (C30), 130.5
(Ar), 130.1 (Ar), 129.6 (Ar), 128.1 (Ar), 123.4 (C5a), 119.5 (C3), 119.2
(C4a), 29.6 (C8), 29.5 (C7), 29.2 (C11), 27.6 (C10), 25.7 (C9), 25.2
(C6); IR (KBr): nmax¼1617 (s) (C]O), 1588 (m), 1543 (m), 1475 (m),
1426 (m), 1266 (s), 745 (s) cmꢁ1; LRMS (ESþ): m/z (%) 361.11 (100)
temperature. 1H NMR (300 MHz, CDCl3):
d
¼8.41 (dd, 3J¼4.8 Hz,
4J¼2.0 Hz, 1H, C2-H), 8.01 (dd, 3J¼7.7 Hz, 4J¼2.0 Hz, 1H, C4-H),
7.57e7.53 (m, 1H, AreH), 7.46e7.41 (m, 3H, AreH), 7.23 (dd, 3J¼7.7,
4.8 Hz, 1H, C3-H), 3.96e3.41 (m, 1H, C7-H), 2.53e1.39 (m, 11H, C70-
H, C12-H2, C8-H2, C9-H2, C11-H2, C10-H2); 13C NMR (75.5 MHz,
[M35ClþNa]þ,
699.27
(25)
[2M35ClþNa]þ,
363.15
(10)
[[M37ClþNa]þ, 701.28 (10) [M35ClþM37ClþNa]þ; HRMS (ESþ): m/z
35
calcd for C20
H
ClN2O [M35ClþH]þ: 339.1264; found: 339.1268.
20
CDCl3):
d
¼199.6 (C6), 187.8 (C5), 173.7 (C13), 150.8 (C14a), 150.5
4.3.5. (E)-1-Cycloheptenylpyrrolidine13 (3b). The title compound 3b
(C2), 142.0 (C10), 139.2 (C4), 135.0 (C30), 130.4e129.0 (4ꢃAr), 126.5
(C4a), 121.6 (C3), 34.2 (C12), 26.0 (C7), 24.8 (C9), 24.5 (C10), 23.3
(C11), 21.2 (C8); IR (NaCl): nmax¼2932 (m), 1693 (s) (C]O), 1665 (s)
(C]O),1591 (s), 1475 (w), 1420 (m), 1368 (m), 1274 (m) cmꢁ1; LRMS
(9.20 g, 55.7 mmol, 62%) was afforded as a yellow oil after distil-
lation using general procedure C. 1H NMR (300 MHz, CDCl3):
d
¼4.47
(br s, 1H, C2-H), 2.97e2.91 (m, 4H, C20-H2), 2.49e2.45 (m, 4H, C3-
H2, C7-H2), 1.84e1.80 (m, 4H, C30-H2), 1.73e1.62 (m, 6H, C4-H2, C6-
H2, C5-H2); LRMS (ESþ): m/z (%) 166.15 (100) [MþH]þ.
(ESþ): m/z (%) 393.19 (100) [M35ClþNa]þ; HRMS (ESþ): m/z calcd for
35
C20
H
ClN2NaO3 [M35ClþNa]þ: 393.0982; found: 393.0967.
19
When 4.0 equiv of m-CPBA was used, 1c was obtained as a yel-
4.3.6. (E)-1-Cyclooctenylpyrrolidine13 (3c). The title compound 3c
(13.9 g, 77.5 mmol, 86%) was afforded as a yellow oil after distil-
low oil (133 mg, 0.360 mmol, 36%). In addition, aldehyde 5 (116 mg,
0.300 mmol, 30%) and 2c (115 mg) were isolated from this reaction.
Method 2: A. Sodium periodate (321 mg, 1.50 mmol, 3.00 equiv)
was stirred in H2O (0.750 mL) and 2 N sulfuric acid (100
0.100 mmol, 20 mol %). After the solid had dissolved, the solution
was cooled to 0 ꢂC. Ruthenium(III) chloride hydrate (519
g,
lation using general procedure C. 1H NMR (300 MHz, CDCl3):
d¼4.27
(br s,1H, C2-H), 3.15e3.08 (m, 4H, C20-H2), 2.52e2.43 (m, 4H, C3-H2,
C8-H2), 2.28e2.17 (m, 2H, C4-H2), 2.01e1.91 (m, 4H, C30-H2),
1.68e1.54 (m, 6H, C5-H2, C7-H2, C6-H2); LRMS (ESþ): m/z (%) 180.12
(100) [MþH]þ.
mL,
m
0.00300 mmol, 0.5 mol %) was then added and the reaction was
stirred until the colour turned bright yellow. Ethyl acetate (3 mL)
and acetonitrile (3 mL) were added and stirring was continued for
a further 5 min 2c (169 mg, 0.500 mmol, 1.00 equiv) was then
added to the reaction and the slurry was stirred at room temper-
ature for 24 h. The reaction mixture was poured into a mixture of
saturated NaHCO3 solution (10.0 mL) and saturated sodium thio-
sulfate solution (10.0 mL). The phases were separated, and the
aqueous layer was extracted with ethyl acetate (3ꢃ20.0 mL). The
combined organic layers were dried (Na2SO4) and concentrated in
vacuo. The crude product was purified by flash column chroma-
tography using silica gel (4:6, ethyl acetate/hexane) to afford 1c
(30.0 mg, 0.0800 mmol, 16%) as a white solid; B. Method 2A was
repeated using ruthenium(III) chloride hydrate (5.19 mg,
0.0300 mmol, 5.0 mol %) to afford 1c (69.0 mg, 0.190 mmol, 37%) as
a white solid.
4.3.7. 6-(1-(3-Chlorophenyl)-3-hydroxy-2,4-dioxo-1,2,3,4-tetrahy-
dro-1,8-naphthyridin-3-yl)-hexanal (5). The title compound
5
(116 mg, 0.300 mmol, 30%) was afforded as a yellow solid using
general procedure B with an additional 20 h stirring at room
temperature.
Mp 93e94 ꢂC; 1H NMR (400 MHz, CDCl3):
d
¼9.73 (t, 3J¼1.4 Hz,
1H, C12-H), 8.47 (dd, 3J¼4.8 Hz, 4J¼1.9 Hz, 1H, C2-H), 8.29 (dd,
3J¼7.6 Hz, 4J¼1.9 Hz, 1H, C4-H), 7.48e7.47 (m, 2H, C50-H and C40-H),
7.22e7.18 (m, 2H, C60-H and C3-H), 7.11 (br s, 1H, C20-H), 2.41 (td,
3J¼7.2 Hz, 4J¼1.4 Hz, 1H, C11-H), 2.09e1.94 (m, 2H, C7-H), 1.64e1.56
(m, 2H, C10-H2), 1.54e1.40 (m, 2H, C8-H2), 1.35e1.24 (m, 2H, C9-
H2); 13C NMR (100.6 MHz, CDCl3):
d
¼202.5 (C12), 193.6 (C5), 173.1
(C13), 154.8 (C2), 154.4 (C1a), 137.4 (C10), 137.2 (C4), 135.5 (C30),
130.9 (C50), 129.7 (C40 and C3), 127.5 (C20), 120.3 (C60), 115.8 (C4a),
83.4 (C6), 43.9 (C11), 41.1 (C7), 29.0 (C10), 23.1 (C8), 22.0 (C11); IR
(NaCl): nmax¼3409 (OH), 2926 (m), 1722 (s) (C]O), 1688 (s) (C]O),
1587 (s), 1469 (w), 1435 (m), 1346 (m), 1174 (CeN), 773 (s), 732
(s) cmꢁ1; LRMS (ESþ): m/z (%) 408.8 (100) [M35ClþNa]þ, 440.8 (30)
4.3.3. 11-(3-Chlorophenyl)-6,7,8,9,10,11-hexahydro-5H-cyclohepta
[b][1,8]naphthyridin-5-one (2b). The title compound 2b (4.08 g,
12.6 mmol, 40%) was afforded as a cream solid using general

9702
T. Lebl et al. / Tetrahedron 66 (2010) 9694e9702
35
[M35ClþNaþMeOH]; HRMS (ESþ): m/z calcd for C20
[M35ClþNa]þ: 409.0931; found: 409.0942.
H
ClN2NaO3
References and notes
19
1. (a) Dolle, R. E.; Le Bourdonnec, B.; Goodman, A. J.; Morales, G. A.; Thomas, C. J.;
Zhang, W. J. Comb. Chem. 2009, 11, 739e790 and previous articles in this review
series; (b) Kim, Y.; Arai, M. A.; Arai, T.; Lamenzo, J. O.; Dean, E. F., III; Patterson,
N.; Clemons, P. A.; Schreiber, S. L. J. Am. Chem. Soc. 2004, 126, 14740e14745.
2. For a brief discussion see Wessjohann, L. A.; Rivera, D. G.; Vercillo, O. E. Chem.
Rev. 2009, 109, 796e814.
3. For some recent examples of library synthesis involving medium-sized rings
see: (a) Lee, D.; Sello, J. K.; Schreiber, S. L. J. Am. Chem. Soc. 1999, 121,
10648e10649; (b) Spring, D. R.; Krishnan, S.; Blackwell, H. E.; Schreiber, S. L. J.
Am. Chem. Soc. 2002, 124, 1354e1363; (c) Looper, R. E.; Pizzirani, D.; Schreiber,
S. L. Org. Lett. 2006, 8, 2063e2066; (d) Morton, D.; Leach, S.; Cordier, C.;
Warriner, S.; Nelson, A. Angew. Chem., Int. Ed. 2009, 48, 104e109.
4. Jones, A.M.; Lorion, M.M.; Lebl, T.; Slawin, A.M.Z.; Philp, D.; Westwood, N.J.
Tetrahedron, 2010.
5. Friary, R. J.; Schwerdt, J. H. Tetrahedron 1991, 47, 9981e9984.
6. Crystallographic data (excluding structure factors) for 1a has been deposited
with the Cambridge Crystallographic Data Centre as supplementary publication
no. CCDC 741362. Copies of the data can be obtained, free of charge, on ap-
plication to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK, (fax: þ44(0)1223
336033 or e-mail: deposit@ccdc.cam.ac.uk/).
4.4. Computational methods
Semi-empirical electronic structure calculations were carried
out using the RM1 method14 as implemented in Spartan ’08 (Ver-
sion 1.1.2, Build 131)15 running on Windows XP computer equipped
with a Intel Core 2 Quad processor at 2.67 GHz and 4 GB of memory.
DFT calculations were carried using the same software and hard-
ware. The transition states were located by generation of an initial
guess using the results of the RM1 calculations. This model tran-
sition state was refined at the B3LYP/6-31G** level of theory to
a transition state structure possessing single imaginary vibration
that corresponded to the coordinate under study.
7. For some recent example of atropisomerism in medium-sized rings see: (a)
Gibson, K. R.; Hitzel, L.; Mortshire-Smith, R. J.; Gerhard, U.; Jelley, R. A.; Reeve,
A. J.; Rowley, M.; Nadin, A.; Owens, A. P. J. Org. Chem. 2002, 67, 9354e9360; (b)
Lee, S.; Kamide, T.; Tababta, H.; Takahashi, H.; Shiro, M.; Natsugari, H. Biorg.
Med. Chem. 2008, 16, 9519e9523; (c) King, F. D.; Aliev, A. E.; Caddick, S.; Tocher,
D. A.; Coutier-Murias, D. Org. Biomol. Chem. 2009, 7, 167e177.
8. (a) Friary, R. J.; Seidl, V.; Schwerdt, J. H.; Chan, T.-M.; Cohen, M. P.; Conklin, E. R.;
Duelfer, T.; Hou, D.; Nafissi, M.; Runkle, R. L.; Tahbaz, P.; Tiberi, R. L.; McPhail, A.
T. Tetrahedron 1993, 49, 7179e7192; (b) Friary, R. J.; Seidl, V.; Schwerdt, J. H.;
Cohen, M. P.; Hou, D.; Nafissi, M. Tetrahedron 1993, 49, 7169e7178.
9. (a) Davies, S. G.; Key, M.-S.; Rodriquez-Solla, H.; Sanganee, H. J.; Savory, E. D.;
Smith, A. D. Synlett 2003, 1659e1662; (b) Aciro, A.; Davies, S. G.; Garner, A. C.;
Ishii, Y.; Key, M.-S.; Ling, K. B.; Prasad, R. S.; Roberts, P. M.; Rodriquez-Solla, H.;
O’Leary-Steele, C.; Russell, A. J.; Sanganee, H. J.; Savory, E. D.; Smith, A. D.;
Thomson, J. E. Tetrahedron 2008, 64, 9320e9344.
10. Frunzke, J.; Loschen, C.; Frenking, G. J. Am. Chem. Soc. 2004, 126, 3642e3652.
11. For a more information see Supplementary data.
12. Oki, M. Recent Advances in Atropisomerism, 1st ed.; John Wiley: New York, NY,
1983.
13. Kuehne, M. E. J. Am. Chem. Soc. 1959, 81, 5400e5404.
14. Rocha, G. B.; Freire, R. O.; Simas, A. M.; Stewart, J. J. P. J. Comput. Chem. 2006, 27,
1101e1111.
15. Shao, Y.; Molnar, L. F.; Jung, Y.; Kussmann, J.; Ochsenfeld, C.; Brown, S. T.; Gil-
bert, A. T. B.; Slipchenko, L. V.; Levchenko, S. V.; O’Neill, D. P.; DiStasio, R. A., Jr.;
Lochan, R. C.; Wang, T.; Beran, G. J. O.; Besley, N. A.; Herbert, J. M.; Lin, C. Y.; Van
Voorhis, T.; Chien, S. H.; Sodt, A.; Steele, R. P.; Rassolov, V. A.; Maslen, P. E.;
Korambath, P. P.; Adamson, R. D.; Austin, B.; Baker, J.; Byrd, E. F. C.; Dachsel, H.;
Doerksen, R. J.; Dreuw, A.; Dunietz, B. D.; Dutoi, A. D.; Furlani, T. R.; Gwaltney, S.
R.; Heyden, A.; Hirata, S.; Hsu, C.-P.; Kedziora, G.; Khalliulin, R. Z.; Klunzinger, P.;
Lee, A. M.; Lee, M. S.; Liang, W. Z.; Lotan, I.; Nair, N.; Peters, B.; Proynov, E. I.;
Pieniazek, P. A.; Rhee, Y. M.; Ritchie, J.; Rosta, E.; Sherrill, C. D.; Simmonett, A. C.;
Subotnik, J. E.; Woodcock, H. L., III; Zhang, W.; Bell, A. T.; Chakraborty, A. K.;
Chipman, D. M.; Keil, F. J.; Warshel, A.; Hehre, W. J.; Schaefer, H. F.; Kong, J.;
Krylov, A. I.; Gill, P. M. W.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2006, 8,
3172e3191.
4.5. Dynamic NMR spectroscopy
All dynamic NMR experiments were carried out using a Bruker
AVANCE 500 spectrometer equipped with inverse tuneable BBI
probe. The temperature, controlled by a Bruker BVT unit, was
measured before each experiment using either 4% MeOH in
methanol-d4 or 80% 1,2-ethanediol in DMSO-d6 samples.16 The rate
constants for exchange processes were determined by complete
lineshape analysis using the Bruker TOPSPIN DNMR module and/or
quantitative EXSY experiment.17 1D NOESY/EXSY experiments were
recorded using standard selnogp sequence with 50 ms 180ꢂ
Gaussian pulse for selective inversion of H7 resonances and mixing
times ranging from 0 to 0.4 s.18 The resultant intensities of EXSY
crosspeaks were heightened by the addition of NOE crosspeaks
intensities obtained from the same experiment recorded at room
temperature and these were evaluated using EXSYCalc software
(Mestrelab Research). The temperature dependence of rate con-
stants was used to determine the activation parameters by fitting
the experimental data to the Eyring equation.
Acknowledgements
The authors thank BBSRC (A.M.J.), the Wellcome Trust (M.M.L.)
and the Royal Society (Research Fellowship to N.J.W.) for funding.
16. Berger, S.; Braun, S. 200 and More NMR Experiments; Wiley-VCH: Weinheim,
2004.
Supplementary data
17. Perrin, C. L.; Dwyer, T. J. Chem. Rev. 1990, 90, 935e967.
18. Stonehouse, J.; Adell, P.; Keeler, J.; Shaka, A. J. J. Am. Chem. Soc. 1994, 116,
Supplementary data associated with this article can be found in
online version at doi:10.1016/j.tet.2010.10.043.
6037e6038.