Cooperative Conformational Regulation in N-Heterocyclic Fluorohydrins

Article abstract of DOI:10.1071/CH14256

Seven-membered N-heterocycles are flexible ring structures, and their conformational control is important to their bioactivity. Our prior work shows that stereoselective monofluorination, if installed diastereoselectively, can bias a seven-membered, subst

Full text of DOI:10.1071/CH14256

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem. 2015, 68, 50–56
Full Paper
http://dx.doi.org/10.1071/CH14256
Cooperative Conformational Regulation in
N-Heterocyclic Fluorohydrins
A
A B
,
Alpesh Ramanlal Patel and Fei Liu
A
Department of Chemistry and Biomolecular Sciences, Macquarie University,
Sydney, NSW 2109, Australia.
B
Corresponding author. Email: fei.liu@mq.edu.au
Seven-membered N-heterocycles are flexible ring structures, and their conformational control is important to their
bioactivity. Our prior work shows that stereoselective monofluorination, if installed diastereoselectively, can bias a seven-
membered, substituted azepane ring to one major conformation. However, multiple fluorination may not provide as much
conformational bias due to conflicting effects. Here we show in our model azepane system that fluorohydrins can confer
strong conformational bias if the relative configuration of the fluorine and hydroxy substitutent is appropriate to enable
cooperative conformational control.
Manuscript received: 24 April 2014.
Manuscript accepted: 16 May 2014.
Published online: 18 July 2014.
Introduction
conformational bias, due to conflicting conformational controls
from multiple fluorine atoms.[14c] Here, we describe the findings
of conformational control from concurrent fluorine and hydroxy
substitution at C5 and C6 in this azepane system (Fig. 1). Three
fluorohydrin azepanes, 3–5, were prepared to investigate the
conformational outcomes of this azepane ring system. The aim
was to investigate the stereochemical requirements for achie-
ving cooperative conformational control for highly substituted
fluoroazepanes. This is the first study of conformational regula-
tion of seven-membered azepane fluorohydrins, in which we
show how cooperative conformational control can be achieved.
In these tetra-substituted azepanes (Fig. 1), multiple effects
exist to influence the conformation of the azepane ring. Notably,
the C3/C4 pseudo-diequatorial v. pseudo-diaxial preference, the
C3 azido gauche effect, and the fluorine/hydroxy gauche effect
are significant contributors.[16,17] These preferences in the
order of 1–3 kcal molꢀ1 are comparable in strength, unlike our
earlier cases in which strong fluorine/Nþ gauche effects around
6 kcal molꢀ1 from C6 fluorine substitution may dominate in
controlling the ring conformation. In the azepanes here, these
relatively mild conformational effects may act cooperatively or
in conflict, thereby presenting an interesting opportunity for
investigating the conformational characteristics that may
emerge from cooperative interaction of multiple conformational
effects in these highly substituted azepanes.
Chemical diversity manifests in many ways, the most commonly
explored being compositional and configurational as seen
frequently in diversity-oriented synthesis (DOS).[1–3] Confor-
mational diversity is another source of chemical diversity and
important to receptor–ligand interactions and molecular recog-
nition.[4] Conformation dynamics in ligand–receptor inter-
actions is not easily predictable, and experimental approaches
that allow measured conformational regulation are necessary for
conferring specific ligand–receptor interactions.[5–7] Existing
approaches of conformational control, such as steric directing,
electronic attraction and repulsion, stereoelectronic control, and
hydrogen bonding, have seen great success in tuning ligand
conformation.[8,9] Selective fluorination is a recent addition to
this repertoire in conformation tuning, due to the strong and
unusual stereoelectronic effects resulting from the C–F bond,
which include dipole–dipole interactions, charge–dipole inter-
actions, and hyperconjugation effects.[10–13]
We have recently reported the regulatory effects of stereo-
selective fluorination using a model N-heterocyclic azepane
system.[14] Substituted azepanes are well recognised bioactive
epitopes in medicinal and natural products chemistry, and our
introduction of fluorine-based conformational control repre-
sents an additional avenue for accessing bioactive azepanes.[15]
In our model system, a disubstituted azepane, with an azido
substituent at C3 and an OBn substituent at C4 in trans
relationship, was able to undergo selective monofluorination
at the C6 position.[14a] Our earlier work of conformational
control by monofluorination in this substituted azepane system
shows that stereoselective monofluorination at the C6 position
confers significant conformational bias only if such substitution
occurs in the R configuration.[14b] Installation of an additional
fluorine atom at the same C6 carbon is not conducive to further
Results and Discussion
Synthesis of Fluorohydrin Azepanes
The synthesis of the targeted fluorohydrin azepanes 3–5
followed the deoxyfluorination strategy starting from dihydro-
xyazepanes. Dihydroxyazepanes 6 and 10 were synthesised
according to the procedure reported previously.[14a] The
Journal compilation Ó CSIRO 2015
www.publish.csiro.au/journals/ajc

N-Heterocyclic Fluorohydrins
51
synthesis was initiated with the asymmetric epoxidation of
divinylcarbinol and involved epoxide ring opening with allyl
amine, ring closing metathesis, and dihydroxylation as key steps
to afford 6 and 10 in 49 % yield over seven steps. The fluoro-
hydrin azepane 9 was synthesised using the deoxyfluorination
procedure with inversion of stereochemistry at C5.[14] Sodium
borohydride reduction of azepane 12 furnished a mixture of
diastereomers 13 and 11 in a ,9 : 1 ratio. Boc-deprotection
using triflouroacetic acid (TFA) afforded the TFA salt of aze-
panes 1–5 from 6, 10, 9, 11, and 13, respectively (Scheme 1).
The stereochemistry of azepanes 1–13 was secured using
HSQC, HMBC, COSY, and NOESY 2D NMR spectroscopy as
described previously.[14] The experimental J values were
extracted via 1H NMR simulation with the use of the program
Bruker TopSpin 3.2 (Table 1). The conformational analysis of
azepanes 1–5 was based on these experimental J values to
identify the contributing conformers.[18] Conformation searches
were performed as implemented in the program MOE, and
conformers within 3–5 kcal molꢀ1 for each azepane were clus-
tered to identify unique ring conformations.[19] Each of the
unique ring structures was subjected to density functional theory
(DFT) geometry optimisations (Turbomole),[20] followed by
general J calculations as described previously.[18] The experi-
mental J values were then compared with the calculated values
to identify the matching conformers.[21]
Conformational Analysis of Fluorohydrin Azepanes
3–5 in CDCl3
Comparison of Fluorohydrin Azepane 3
with Hydroxyazepane 14
In our earlier work,[14c] two major conformers for hydro-
xyazepane 14 were identified (14a and 14b in a 3 : 2 ratio,
Fig. 2). In both conformers, the C3/C4 pseudo-diequatorial
preference is maintained. However, the azido gauche preference
was only maintained in the major conformer 14a but not in 14b.
Fluorohydrin azepane 3 differs from the hydroxyazepane 14 by
only the fluorine substitution at C5 trans to the C6 hydroxy
group. The addition of this fluorine substituent at C5 does not
alter the conformational preference of 14, since the major
conformer 3a is similar to 14a. In both conformers, the benzy-
loxy and azido groups are pseudo-diequatorial as expected, and
the azido gauche and C6-oxygen/C7-nitrogen gauche prefer-
ences are also satisfied. The minor conformer 3b is also quite
similar to 3a with the difference mainly due to the puckering of
the nitrogen. Interestingly, the addition of fluorine in this case
improves the conformational bias only slightly to a 2 : 1 ratio. In
the case here, the fluorine substitution at C5 does not appreci-
ably reduce conformational disorder nor change the major
conformational preferences.
Comparison of Fluorohydrin Azepanes 4
with Hydroxyazepane 15
OBn
4
X
Hydroxyazepane 15 differs from hydroxyazepane 14 only in
the configuration of the C6 hydroxy group (Fig. 2). Yet for 15
the conformational disorder was high enough to prevent the
identification of major contributors from eight clusters of
conformers found by a conformational search.[14c] Similar to
15, azepane 2, with an additional C5 hydroxy substitution in the
R configuration (cis-diol), is also too conformationally disor-
dered to permit identification of major conformers out of a
dozen distinct ring conformation clusters. However, the addition
of one fluorine trans to the C6 hydroxy group provided the
fluorohydrin 4 with clear conformational bias to one conformer
4a (Fig. 3). In this conformation, the benzyloxy and azido
groups are pseudo-diequatorial as expected, and similar to 3a,
the C5-fluorine/C6-oxygen relationship is anti rather than
gauche. This shows that the cooperativity between the fluorine
1: X ϭ H, Y ϭ OH, M ϭ H, N ϭ OH
2: X ϭ OH, Y ϭ H, M ϭ OH, N ϭ H
3: X ϭ F, Y ϭ H, M ϭ H, N ϭ OH
4: X ϭ H, Y ϭ F, M ϭ OH, N ϭ H
5: X ϭ H, Y ϭ F, M ϭ H, N ϭ OH
5
Y
N3
6
M
3
2
ϩ
N
7
Ϫ
N
H21 TFA
Conformational influences
H
H
H
H
H
C
C
XO
F
H
H
H
H
OBn
RH2N
NH2R
N3
ϩ
ϩ
N3
H
OH
C3/C4 pseudo-
azido gauche
hydroxy gauche
fluorine-oxygen
diequatorial preference
effect
effect
gauche effect
Fig. 1. Structures of azepanes 1–5 and the expected contributors of
stereoelectronic influence on their conformations.
OBn
OBn
F
X
OBn
HO
Y
N3
N3
N3
d
b
HO
HO
TBSO
67 %
77 %
N
R
N
R
N
Boc
7
8
X ϭ H; Y ϭ OH
X ϭ F; Y ϭ H
6
1
R ϭ Boc
R ϭ H
9
3
R ϭ Boc
R ϭ H
c
65 %
a
a
97 %
96 %
OBn
OBn
F
OBn
OBn
HO
F
F
Ref. 14c
N3 b,c,d
N3
N3
N3
e
f
HO
HO
O
HO
71 %
92 %
36 % over
3 steps
N
R
N
R
N
N
R
Boc
11 R ϭ Boc
13 R ϭ Boc
12
10 R ϭ Boc
a
a
a
97 %
96 %
96 %
4
R ϭ H
5
R ϭ H
2
R ϭ H
Scheme 1. Synthesis of fluorohydrin azepanes 1–5. Reagents and conditions: (a) trifluoroacetic acid (TFA), neat,
5 min; (b) Imidazole, tert-butyldimethylsilyl chloride (TBSCl), dry DMF, 258C, 10 h; (c) Deoxofluor, dry dichloro-
methane (DCM), 08C, 8 h; (d) tetrabutylammonium fluoride (TBAF), dry THF, 45 min; (e) pyridinium chlorochromate
˚
(PCC) (1.1 equiv.), MS 4 A, dry DCM, 258C, 1.5 h; (f) NaBH4, EtOH, 258C, 1.5 h.

52
A. R. Patel and F. Liu
Table 1. Experimental 3JHH and 3JHF values for azepanes 1–5 in CDCl3
Experimental 3JHH and 3JHF (Hz)A
Azepanes
2a–3
2.5
2b–3
3–4
4–5a
4–5b
5a–6a
5a–6b
5b–6a
5b–6b
6a–7a
6a–7b
6b–7a
n.a.
6b–7b
n.a.
OH
OBn
H
Ha
6
HO
Ha
5
4
H
N3
3
7
2
7.5
5.5
5.0
n.a.
1.8
n.a.
n.a.
n.a.
3.4
7.5
1
Hb
Ha
N
ϩ
Ha
H2
Hb
1
Hb
HO
OBn
H
Hb
HO
5
4
6
H
3
2
N3
7
2.0
3.5
2.0
1.9
6.0
4.0
8.9
8.2
8.0
7.1
8.7
7.0
n.a.
24.2
2.0
4.0
n.a.
11.0
n.a.
2.2
n.a.
n.a.
5.9
n.a.
2.0
n.a.
7.0
n.a.
n.a.
3.0
n.a.
1.7
n.a.
3.0
n.a.
7.5
4.0
n.a.
6.2
n.a.
4.0
n.a.
1.4
n.a.
1
Hb
Ha
N
ϩ
Ha
H2
Hb
2
Hb
F
OBn
H
HO
Ha
5
4
6
H
N3
3
7
2
1.7
6.0
1
Hb
Ha
N
Ha
ϩ
H2
Hb
3
F
OBn
H
Ha
Hb
HO
5
4
6
H
N3
3
7
2
21.4
n.a.
1
N
H2
Hb
Ha
ϩ
Ha
Hb
4
F
OBn
H
Ha
HO
Ha
5
4
6
H
3
N3
5.8
13.1
n.a.
19.4
7
2
1
Hb
Ha
N
Ha
ϩ
H2
Hb
5
AThe 1H NMR spectrum of azepane 2 in CDCl3 was too broad to allow accurate extraction of specific J values therefore the J values reported for azepane
2 should be regarded as approximate. n.a.: not applicable.
ϩ
ϩ
3
4
3a
3b
(2:1)
14a
14b
(3:2)
14
15
ϩ
ϩ
4a
ϩ
ϩ
5
5a
5b
(4:1)
1a
1b
1c (5:4:1)
1
Fig. 2. Major conformers of azepanes 1, 3, 4, and 5. Azepane 2 was too disordered to allow identification of major conformers. For comparison
purposes, the conformers of hydroxyazepanes 14 and 15 are included.[14c] All carbon atoms are shown in black and heteroatoms grey. All hydrogen
atoms are omitted for clarity.

N-Heterocyclic Fluorohydrins
53
12
disorder if not more. In other words, having an additional
hydroxy substituent at C5 does not appreciably confer further
conformational bias nor alter the conformational preferences
in this azepane. Conformers 14a and 1a are in fact quite
similar.
Experimental J values
Conformer 4a J values
10
8
6
In the case of azepane 5, substitution with fluorine at C5 in
the same configuration, however, is able to confer conforma-
tional control cooperatively with three other substituents at C3,
C4, and C6. Not only is the ring conformation biased to one
major conformer but also the conformational preferences in the
major conformer are changed. Interestingly, this level of con-
formational control is absent in fluorohydrin azepane 3, in which
the fluorine substituent is in the opposite configuration. This
again reinforces the observation that, given the appropriate
configuration in this highly substituted azepane system, fluorine
is a stronger director than the hydroxy group. It is also note-
worthy that in all three major conformations of the fluorohydrins
3a, 4a, and 5a, the C3/C4 pseudo-diequatorial preference and
hydroxy gauche effect are satisfied, while the azido gauche
effect group is consistently lost. In the case of the dihydroxy
azepane 1 with two major conformers 1a and 1b, the C3/C4
pseudo-diequatorial preference and hydroxy gauche effect are
either satisfied (1a) or lost (1b), while the azido gauche effect
group is consistently satisfied in both 1a and 1b. This also
suggests that fluorine at C5 is able to control the relative
influences of the other conformational factors better than the
hydroxy group. However, fluorine alone is not able to achieve
conformational regulation at the most optimal level. The best-
case scenario of conformational control is found in fluorohydrin
azepane 4, where the C5 fluorine remains in the S configuration
with the C6 hydroxy group trans to the fluorine. In this azepane
the cooperativity between the four substituents in conformational
control is at the highest level given the appropriate relative
configuration of the substituents, resulting in bias to one
predominant conformer 4a.
4
2
0
1.3
2.3
3.4
4.6
6.7
7.8
7.9
Spins
Fig. 3. The well matched experimental and calculated J values of
conformer 4a in CDCl3.
and hydroxy substituents in conferring strong conformational
bias can be achieved only if the relative stereochemistry of the
two groups is appropriate.
Comparison of Fluorohydrin Azepane 5 with
Hydroxyazepane 1
The comparison between the conformational populations of
azepane 5 and 1 reveals interesting effects of the fluorine
substitution at C5 (Fig. 2). Both azepanes are tetra-substituted
in identical configurations; for azepane 5, the C5 substitution is a
fluorine atom rather than a hydroxy group. Hydroxyazepane 1
has two major conformers 1a and 1b, plus one minor conformer
1c. Conformers 1a and 1b are distinct, with different dihedral
angles at C3, C4, C5, and C6. Conformer 1a adopts a chair–boat
like geometry, and the C3/C4 pseudo-diequatorial preference is
only partly satisfied; however in 1b, the C3/C4 substituents are
in the pseudo-diaxial position. The minor conformer 1c
resembles 1a, except that the C3/C4 substituents are pseudo-
diequatorial and that the azido preference in 1c is anti rather than
gauche.
Changing the C5 substituent from OH to fluorine appears to
significantly alter the conformational landscape of the azepane.
In azepane 5, the conformational disorder is clearly reduced to
one major conformer 5a with one minor conformer 5b in a 4 : 1
ratio. The conformational preferences are also considerably
changed, in that 5a resembles the minor conformer 1c, with
only slight dihedral differences at C2 and C3. The minor
conformer 5b does not resemble any of the conformational
contributors of azepane 1. For both conformer 5a and 5b, the
C3/C4 preference is pseudo-diequatorial, and the azido prefer-
ence is anti rather than gauche. Both of the C4/C5 and C5/C6
fluorine/OH gauche preferences are met in 5a but not in 5b. This
comparison demonstrates that C5 fluorine substitution is more
effective at reducing conformational disorder compared with the
hydroxy substituent.
Solvent Effects
The coupling constants of azepanes 1–5 in polar aprotic sol-
vents, such as (CD3)2CO, CD3CN, (CD3)2SO, and protic solvent
CD3OD were also investigated. For bishydroxyazepane 1, the
proton chemical shifts were within 0.5 ppm in different solvents.
Small coupling constant deviations of 2–3 Hz appeared in the
C3–C5 region; however, the rest of the J values were identical in
different solvents, suggesting no significant changes in the
conformer populations in different solvents. For bishydrox-
yazepane 2, the solvent effects were more pronounced. The
broad proton signals became more defined with more than one
1
conformer population visible. H NMR spectra in (CD3)2CO
and CD3CN displayed at least three different conformer popu-
lations; while the 1H NMR spectra in CD3OD and (CD3)2SO had
one major conformer population with a minor population. It was
possible to extract the coupling constants of the major popula-
tion from the CD3OD spectrum. Conformational analysis was
able to identify two major conformers 2a and 2b in a 3 : 2 ratio
(Fig. 4). These two conformers adopt similar conformations at
C3, C4, and C5; however, the nitrogen puckerings are in
opposite directions (up in 2a but down in 2b). In the major
conformer 2a, the C6 oxygen prefers to be gauche to the ring
nitrogen, although the azido preference is anti rather than
gauche.
Cooperative Conformational Regulation
in Fluorohydrin Azepane 4
The tetra-substituted fluorohydrin azepane 4 exhibits
strong conformational bias, and this control is the result of
all four substituents in an appropriate relative configuration
on the azepane ring. One possibility might be that perhaps
tetra-substitution generally favours a more ordered ring con-
formation. However, the conformational properties of aze-
pane 1 are not supportive of such a generalisation. Compared
with the mono-hydroxylated azepane 14, the bis-hydroxylated
azepane 1 exhibits a comparable level of conformational
Azepane 4 and 5 had no significant changes in conformer
populations in the solvents studied, and lowering the

54
A. R. Patel and F. Liu
ϩ
ϩ
2a
2b
3c
3d
3b
3 (CD3CN)
2 (CD3OD)
(3:2)
(6:3:1)
Fig. 4. Major conformers of azepanes 2 in CD3OD and 3 in CD3CN. All carbon atoms are shown in black and heteroatoms grey.
All hydrogen atoms are omitted for clarity.
temperature did not lead to apparent changes in conformational
populations.[21,22] However, azepane 3 in acetonitrile exihibits
very different conformational populations (Fig. 4). Only one set
was amenable to J-based analysis. This population consists of
two major conformers 3c and 3d, with 3b being a minor
contributor. Interestingly, 3d is only slightly different from 3a
by virtue of the azido anti rather than gauche to the ring nitrogen.
However, conformer 3c is a distinct new conformer with the C5
fluorine and C6 hydroxy groups gauche and the C3/C4 sub-
stituents in pseudo-diaxial positions. These cases demonstrate
that some of these highly substituted azepane systems, while
conformationally biased, still retain significant conformational
mobility and are able to sample very different conformations in
different solvents. For seven-membered N-heterocycles that are
conformationally very diverse, such conformational mobility
could not have been predicted prior but only observed
experimentally.
Environment 2011.10; Chemical Computing Group), and
Turbomole (version 6.3).
Synthesis of Azepanes 1–13
Azepanes 6, 10–12 were synthesised according to the procedure
described previously.[14]
(3R,4S,5S,6S)-3-Azido-4-benzyloxy-5,6-
dihydroxyazepane (1)
The azepane 6 (13.60 mg, 35.9 mmol) was dissolved in TFA
(500 mL) at 258C. The solution was allowed to stir for 5 min
before TFA was evaporated under N2 flow. The reaction flask
was kept under high vacuum (0.005 torr, 258C) for 3 h to remove
traces of TFA and the colourless oily residue obtained was
characterised as 1 (9.70 mg, 97 %).
[a]2D0 ¼ þ32.3 (c 0.8, CH2Cl2). nmax (film)/cmꢀ1 3581, 2349,
1690, 1673, 1646, 1551, 1167, 1130. dH (600 MHz, CDCl3)
7.39–7.28 (m, 5H), 4.65 (d, J 10.10, 1H), 4.57 (d, J 10.10, 1H),
4.34–4.28 (m, 1H), 4.07–3.95 (m, 2H), 3.62–3.53 (m, 1H), 3.38–
3.13 (m, 3H), 2.83–2.71 (m, 1H). dC (150 MHz, CDCl3) 136.8,
128.9, 128.6, 128.4, 80.6, 74.6, 73.9, 66.1, 59.9, 46.3, 45.1. m/z
(HRMS ESI) 279.1451; [M þ H]þ C13H19N4O3 requires
279.1457.
Conclusions
In conclusion, a series of fluorohydrin azepanes were examined
for their conformational properties. One of their dihydroxy-
substituted counterparts was also investigated as a comparison.
In such highly substituted seven-membered N-heterocyles,
positive cooperativity in conformational control was achieved
with a C5-S fluorine and a C6-R hydroxy substituent. Deviation
from this configurational profile leads to increased conforma-
tional disorder. Replacing the C5-S fluorine with a C5-S
hydroxy substituent also causes loss of conformational bias,
suggesting that the C5 fluorine exhibits stronger effects than the
hydroxy group in conformational control of this azepane system.
(3R,4S,5R,6S)-3-Azido-4-benzyloxy-6-(tert-
butyldimethylsilyl)oxy-5-hydroxyazepane-1-
carboxylic Acid tert-Butyl Ester (7)
A solution of tert-butyldimethylsilyl chloride (96.5 mg,
0.641 mmol) in DMF (3.25 mL) was added dropwise via a
syringe to a well stirred solution of 6 (220 mg, 0.583 mmol)
and imidazole (43.6 mg, 0.641 mmol) in DMF (3.25 mL) at 258C
under N2 atmosphere. The reaction mixture was allowed to stir
at the same temperature for 10 h. The reaction mixture was
quenched by addition of water (5 mL) and was extracted with
EtOAc (10 mL ꢁ 3). The combined organic layer was washed
with water (5 mL ꢁ 2) and brine (5 mL) to remove DMF residues
and dried (MgSO4) before evaporation under reduced pressure
to obtain the crude product, which was subjected to flash
chromatography (petroleum ether/EtOAc, 9/1) to give 7
(192.4 mg, 67 %, Rf 0.33) as a colourless oil. It is noteworthy
that 7 was obtained exclusively and unreacted 6 was recovered.
[a]2D0 ¼ þ15.1 (c 1.0, CH2Cl2). nmax (film)/cmꢀ1 3600–3130
(br), 2997, 2933, 2365, 2114, 1680, 1411, 1157, 1088, 1045,
990. dH (600 MHz, CDCl3) 7.39–7.33 (m, 5H), 4.93 (d, J 11.11,
1H), 4.60 (d, J 11.11, 1H), 4.28–4.25 (m, 1H), 4.02 (dd, J 14.28,
5.32, 1H), 3.82–3.70 (m, 2H), 3.62 (br s, 1H), 3.58–3.53 (m, 1H),
3.31 (dd, J 14.17, 10.88, 1H), 3.01 (ddd, J 13.70, 4.69, 2.86, 1H),
2.60 (br s, 1H), 1.47 (s, 9H), 0.92 (s, 9H), 0.1 (s, 6H). dC
(150 MHz, CDCl3) 154.7, 137.6, 128.9, 128.4, 128.3, 81.7, 80.7,
75.4, 73.4, 71.4, 64.4, 50.5, 46.0, 28.5, 26.0, ꢀ4.53, ꢀ4.85. m/z
(HRMS ESI) 493.2844; [M þ H]þ C24H41N4O5Si requires
493.2846.
Experimental
General
All reactions were conducted under N2 atmosphere. Unless
otherwise specified, all reagents were purchased from Sigma–
Aldrich and used without further purification. Deoxofluor was
obtained from Matrix Scientific and used without further
purification. CH2Cl2 was obtained from a solvent purification
system (Innovative Technology SPS400) and stored over MS
˚
4 A beads. Ethyl acetate and petroleum ether were distilled
before use and the later refers to the fraction collected between
60 and 808C. THF was distilled over Na–benzophenone and
˚
stored over MS 4 A beads. Anhydrous DMF was obtained
from Sigma–Aldrich and used without further purification. 1H
NMR spectra were recorded at 258C on either a Bruker
DRX600K or DPX400 NMR spectrometer and are reported in
parts per million (ppm) using the specified solvent as the
internal standard (CDCl3 at 7.26 ppm). 13C NMR spectra are
reported in ppm using the specified solvent as the internal
standard (CDCl3 at 77.16 ppm). Computational investigations
were performed using programs MOE (Molecular Operating

N-Heterocyclic Fluorohydrins
55
(3R,4S,5S,6S)-3-Azido-4-benzyloxy-6-(tert-
butyldimethylsilyl)oxy-5-fluoroazepane-1-carboxylic
Acid tert-Butyl Ester (8)
136.5, 128.9, 128.6, 128.3, 93.0 (d, 1JCF. 178.86), 78.0 (d, 2JCF.
19.98), 74.0, 66.6 (d, 2JCF. 26.97), 57.6 (d, 3JCF. 9.21), 50.1 (d,
3JCF. 5.58), 48.2. m/z (HRMS ESI) 281.1410; [M þ H]þ
C13H18FN4O2 requires 281.1408.
A solution of bis(2-methoxyethyl)aminosulfur trifluoride
(Deoxofluor) (89.8 mg, 0.406 mmol) in dry dichloromethane
(DCM) (5.5 mL) was added dropwise via a syringe to a well
stirred solution of 7 (200.0 mg, 0.406 mmol) in dry DCM
(5.5 mL) under N2 atmosphere at 08C. The reaction mixture
was allowed to stir at the same temperature for 8 h. The reaction
mixture was quenched by ice-cooled water before the crude was
extracted with EtOAc and washed with water and brine. The
organic phase was evaporated under reduced pressure after
drying over MgSO4 to obtain the crude product, which was
subjected to flash chromatography (petroleum ether/EtOAc,
9/1) to give 8 (130.5 mg, 65 %, Rf 0.52) as a colourless oil.
[a]D20 þ55.1 (c 0.9, CH2Cl2). nmax (film)/cmꢀ1 2997, 2933, 2365,
2114, 1680, 1521, 1473, 1456, 1411, 1157, 1088, 1045, 990. dH
(600 MHz, CDCl3) 7.44–7.28 (m, 5H), 4.80 (d, J 10.29, 1H),
4.71 (d, J 10.29, 1H), 4.55 (dd, J 46.56 (1JHF), 8.47, 1H), 4.43–
4.36 (m, 1H), 4.11 (dd, J 14.01, 4.75, 1H), 4.01–3.90 (m, 1H),
3.75 (dd, J 14.30, 2.45, 1H), 3.53–3.47 (m, 1H), 3.25 (dd,
J 14.30, 11.33, 1H), 3.02 (dd, J 14.01, 12.36, 1H), 1.45
(s, 9H), 0.93 (s, 1H), 0.13 (d, J 23.66, 6H). dC (150 MHz,
(3R,4S,5R,6R)-3-Azido-4-benzyloxy-5,6-
dihydroxyazepane (2)
The procedure for the synthesis of 1 was followed to give a
colourless oil in 96 % yield from 10; [a]2D0 ꢀ18.2 (c 0.8,
CH2Cl2). nmax (film)/cmꢀ1 3581, 2349, 1690, 1673, 1646,
1551, 1167, 1130. dH (600 MHz, CDCl3) 7.42–7.29 (m, 5H),
4.65 (d, 2H), 4.29–4.23 (m, 1H), 4.06–4.00 (m, 1H), 3.87–3.80
(m, 1H), 3.74–3.69 (m, 1H), 3.48–3.40 (m, 1H), 3.28–3.15 (m,
3H). dC (150 MHz, CDCl3) 136.6, 128.9, 128.7, 128.7, 81.8,
73.1, 72.5, 66.3, 58.9, 47.6, 46.6. m/z (HRMS ESI) 279.1459;
[M þ H]þ C13H19N4O3 279.1457.
(3R,4S,5S,6R)-3-Azido-4-benzyloxy-5-
fluoro-6-hydroxyazepane (4)
The procedure for the synthesis of 1 was followed to give a
colourless oil in 97 % yield from 11; [a]2D0 ꢀ22.1 (c 0.7,
CH2Cl2). nmax (film)/cmꢀ1 3610–3160 (br), 2999, 2939, 2369,
2134, 1482, 1442, 1410, 1143, 1089, 1012. dH (600 MHz,
1
1
CDCl3) 154.6, 137.8, 128.7, 128.6, 128.2, 95.6 (d, JCF.
177.79), 81.1, 80.8 (d, 2JCF. 19.18), 75.8, 71.6 (d, 2JCF. 28.23),
CDCl3) 7.41–7.34 (m, 5H), 4.85 (dd, JHF 44.09, 6.07, 1H),
4.77 (d, J 11.33, 1H), 4.67 (d, J 11.33, 1H), 4.42–4.37 (m, 1H),
4.24 (dd, J 9.01, 8.37, 1H), 3.78 (dd, J 21.34, 8.48, 1H), 3.40 (dd,
J 14.06, 6.03, 1H), 3.36 (d, J 14.08, 1H), 3.14 (d, J 13.58, 1H),
2.85 (dd, J 13.96, 9.50, 1H). dC (150 MHz, CDCl3) 136.2, 128.9,
128.9, 128.7, 92.1 (d, 1JCF. 173.7), 82.7 (d, 2JCF. 21.7), 73.8, 66.4
3
3
63.5 (d, JCF. 10.16), 49.4 (d, JCF. 12.65), 45.3, 28.5, 25.9,
ꢀ4.61, ꢀ4.83. m/z (HRMS ESI) 495.2798; [M þ H]þ
C24H40FN4OSi requires 495.2803.
2
(3R,4S,5R,6S)-3-Azido-4-benzyloxy-5-fluoro-6-
hydroxyazepane-1-carboxylic Acid tert-Butyl Ester (9)
(d, JCF. 28.33), 59.5, 46.3, 46.0. m/z (HRMS ESI) 281.1411;
[M þ H]þ C13H18FN4O2 requires 281.1408.
A solution of tetrabutylammonium fluoride (TBAF, 1 M
solution in THF, 80 mL, 80.1 mmol) was added dropwise via a
syringe to a well stirred solution of 8 (36.0 mg, 72.8 mmol) in dry
THF (1 mL) under N2 atmosphere at 258C. The reaction mixture
was allowed to stir at the same temperature until completion
(45 min, TLC). The reaction mixture was evaporated to dryness
and the crude obtained was subjected to flash chromatography
(petroleum ether/EtOAc, 4/1) to give 9 (21.3 mg, 77 %, Rf 0.24)
as a colourless oil. [a]2D0 þ36.7 (c 0.9, CH2Cl2). nmax (film)/cmꢀ1
3600–3130 (br), 2990, 2963, 2365, 2134, 1684, 1526, 1479,
1466, 1419, 1167, 1081, 1042. dH (600 MHz, CDCl3) 7.39–7.30
(m, 5H), 4.78 (d, J 11.80, 1H), 4.74 (br d, 1JHF 45.70, 1H), 4.62
(d, J 11.80, 1H), 4.16–4.08 (m, 1H), 4.01 (br d, J 15.22, 1H), 3.98
(br d, J 17.69, 1H), 3.95 (dd, J 14.59, 4.09, 1H), 3.78 (br s, 1H),
3.39 (dd, J 15.31, 4.32, 1H), 3.24 (dd, J 15.22, 3.72, 1H), 1.48 (s,
9H). dC (150 MHz, CDCl3) 157.8, 137.3, 128.7, 128.3, 128.1,
(3R,4S,5S,6S)-3-Azido-4-benzyloxy-5-fluoro-
6-hydroxyazepane-1-carboxylic Acid tert-Butyl
Ester (13)
To a solution of azepane 12 (14.6 mg, 38.7 mmol) in EtOH
(0.12 mL) was added sodium borohydride (1.5 mg, 38.7 mmol)
and the solution was allowed to stir at room temperature for 1.5 h
before addition of EtOAc (5 mL) and brine (2 mL). The organic
layer was separated and the aqueous layer was extracted with
EtOAc (5 mL ꢁ 2). The combined organic layer was rotary
evaporated to obtain the crude mixture, which was subjected
to flash column chromatography using 25 % EtOAc/petroleum
ether to obtain 13 (10.4 mg, 71 %, Rf 0.24) and 11 (1.2 mg, 8 %,
Rf 0.25) as colourless oils. 13: [a]2D0 þ15.5 (c 1.1, CH2Cl2). nmax
(film)/cmꢀ1 3600–3130 (br), 2990, 2963, 2365, 2134, 1684,
1526, 1479, 1466, 1419, 1167, 1081, 1042. dH (400 MHz,
CDCl3) 7.44–7.30 (m, 5H), 4.82–4.67 (m, 3H), 4.44–4.26
(m, 1H), 4.04–3.91 (m, 2H), 3.76 (br d, J 13.85, 1H), 3.62–
3.53 (m, 1H), 3.24–3.13 (m, 2H), 2.29 (d, J 4.98, 1H), 1.46
(s, 9H). dC (100 MHz, CDCl3) 154.8, 137.5, 128.7, 128.6, 128.3,
1
2
96.5 (d, JCF. 176.11), 81.8, 79.5 (d, JCF. 22.61), 73.9, 72.6
(d, 2JCF. 24.36), 61.4 (d, 3JCF. 12.05), 52.1 (d, 3JCF. 9.44), 48.8,
28.3. m/z (HRMS ESI) 381.1938; [M þ H]þ C18H26FN4O4
requires 381.1938.
1
2
95.6 (d, JCF. 176.57), 81.2, 80.3 (d, JCF. 21.01), 75.3, 70.2
(d, 2JCF. 22.97), 63.6 (d, 3JCF. 7.24), 48.5 (d, 3JCF. 10.28), 46.4,
28.5. m/z (HRMS ESI) 381.1941; [M þ H]þ C18H26FN4O4
requires 381.1938.
(3R,4S,5R,6S)-3-Azido-4-benzyloxy-5-
fluoro-6-hydroxyazepane (3)
The procedure for the synthesis of 1 was followed to give a
colourless oil in 96 % yield from 9; [a]2D0 þ43.2 (c 0.7, CH2Cl2).
nmax (film)/cmꢀ1 3610–3160 (br), 2999, 2939, 2369, 2134,
1482, 1442, 1410, 1143, 1089, 1012. dH (600 MHz, CDCl3)
(3R,4S,5S,6S)-3-Azido-4-benzyloxy-5-fluoro-6-
hydroxyazepane (5)
1
7.40–7.30 (m, 5H), 5.02 (dd, JHF 46.02, 5.69, 1H), 4.75 (d, J
The procedure for the synthesis of 1 was followed to give a
colourless oil in 96 % yield from 13; [a]2D0 þ24.5 (c 0.8,
CH2Cl2). nmax (film)/cmꢀ1 3610–3160 (br), 2999, 2939, 2369,
2134, 1482, 1442, 1410, 1143, 1089, 1012. dH (600 MHz,
11.86, 1H), 4.65 (d, J 11.86, 1H), 4.33–4.26 (m, 1H), 4.16 (dd, J
24.30, 6.78, 1H), 3.99–3.95 (m, 1H), 3.52 (d, J 14.60, 1H), 3.40
(d, J 14.09, 1H), 3.42–3.22 (m, 2H). dC (150 MHz, CDCl3)

56
A. R. Patel and F. Liu
1
CDCl3) 7.42–7.35 (m, 5H), 4.81 (ddd, JHF 44.79, 5.78, 177,
1H), 4.79 (d, J 11.33, 1H), 4.72 (d, J 11.33, 1H), 4.52 (dd,
J 19.41, 7.08, 1H), 4.02 (dt, J 12.77, 5.73, 1H), 3.97 (ddd, J 6.75,
5.73, 2.35, 1H), 3.61 (ddd, J 14.00, 7.54, 2.06, 1H), 3.43–3.39
(m, 2H), 3.34 (d, J 13.49, 1H). dC (150 MHz, CDCl3) 136.5,
128.9, 128.6, 128.3, 93.0 (d, 1JCF. 178.86), 78.0 (d, 2JCF. 19.98),
[14] (a) A. R. Patel, F. Liu, Tetrahedron 2013, 69, 744. doi:10.1016/J.TET.
2012.10.078
(b) A. R. Patel, G. Ball, L. Hunter, F. Liu, Org. Biomol. Chem. 2013,
11, 3781. doi:10.1039/C3OB40391B
(c) A. R. Patel, L. Hunter, M. M. Bhadbhade, F. Liu, Eur. J. Org. Chem.
2014, 2584. doi:10.1002/EJOC.201301811
[15] (a) E. A. Ilardi, E. Vitaku, J. T. Njardarson, J. Med. Chem. 2014, 57,
2832. doi:10.1021/JM401375Q
2
3
74.0, 66.6 (d, JCF. 26.97), 57.6 (d, JCF. 9.21), 50.1 (d,
3JCF. 5.58), 48.2. m/z (HRMS ESI) 281.1410; [M þ H]þ
C13H18FN4O2 requires 281.1408.
(b) T. K. Beng, S. M. Wilkerson-Hill, R. Sarpong, Org. Lett. 2014, 16,
916. doi:10.1021/OL403671S
[16] L.-S. Sonntag, S. Schweizer, C. Ochsenfeld, H. Wennemers, J. Am.
Chem. Soc. 2006, 128, 14697. doi:10.1021/JA0654938
[17] (a) C. R. S. Briggs, M. J. Allen, D. O’Hagan, D. J. Tozer, D. J. A. M. Z.
Slawin, A. E. Goeta, J. A. K. Howard, Org. Biomol. Chem. 2004, 2,
732. doi:10.1039/B312188G
Supplementary Material
Copies of NMR spectra and conformational analysis details are
available on the Journal’s website.
(b) K. B. Wiberg, M. A. Murcko, K. E. Laidig, P. J. MacDougall,
J. Phys. Chem. 1990, 94, 6956. doi:10.1021/J100381A008
(c) C. R. S. Briggs, D. O’Hagan, H. S. Rzepa, A. M. Z. Slawin, J. Fluor.
Chem. 2004, 125, 19. doi:10.1016/J.JFLUCHEM.2003.08.011
(d) D. O’Hagan, C. Bilton, J. A. K. Howard, L. Knight, D. J. Tozer,
J. Chem. Soc., Perkin Trans. 2 2000, 605. doi:10.1039/B000205O
[18] (a) C. A. G. Haasnoot, F. A. A. M. DeLeeuw, C. Altona, Tetrahedron
1980, 36, 2783. doi:10.1016/0040-4020(80)80155-4
Acknowledgement
The authors thank Dr Luke Hunter for helpful suggestions.
References
[1] S. L. Schreiber, Science 2000, 287, 1964. doi:10.1126/SCIENCE.287.
5460.1964
[2] D. S. Tan, Nat. Chem. Biol. 2005, 1, 74. doi:10.1038/NCHEM
BIO0705-74
[3] W. R. Galloway, A. Isidro-Llobet, D. R. Spring, Nat. Commun. 2010, 1.
doi:10.1038/NCOMMS1081
[4] S. Llabre´s, J. Jua´rez, F. Forti, R. Pouplana, F. J. Luque, in Physico–
Chemical and Computational Approaches to Drug Discovery
(Eds J. Luque, X. Barril) 2012, Ch. 1, pp. 1–22 (The Royal Society
of Chemistry: London).
Three-bond H–H and H–F coupling constants obey Karplus relation-
ships with the corresponding H–C–C–H and H–C–C–F dihedral
3
angles: ‘typical’ JHH values are 1–3 Hz for gauche alignments and
6–9 Hz for anti alignments; see (b) A. M. Ihrig, S. L. Smith, J. Am.
Chem. Soc. 1970, 92, 759. doi:10.1021/JA00707A003
‘Typical’ 3JHF values are 5–15 Hz for gauche alignments and 25–35 Hz
for anti alignments;see (c)C. Thibaudeau, J. Plavec, J. Chattopadhyaya,
J. Org. Chem. 1998, 63, 4967. doi:10.1021/JO980144K
For an example of conformational analysis by NMRand DFT methods,
see (d) P. Ta¨htinen, A. Bagno, K. D. Klika, K. Pihlaja, J. Am. Chem.
Soc. 2003, 125, 4609. doi:10.1021/JA021237T
[5] L. Skjaerven, N. Reuter, A. Martinez, Future Med. Chem. 2011, 3,
2079. doi:10.4155/FMC.11.159
[6] K. K. Frederick, M. S. Marlow, K. G. Valentine, A. J. Wand, Nature
2007, 448, 325. doi:10.1038/NATURE05959
[7] B. Honarparvar, T. Govender, G. E. Maguire, M. E. Soliman, H. G.
Kruger, Chem. Rev. 2014, 114, 493. doi:10.1021/CR300314Q
[8] J. Rebek, Jr, Science 1987, 235, 1478. doi:10.1126/SCIENCE.3823899
[9] E. A. Meyer, R. K. Castellano, F. Diederich, Angew. Chem. Int. Ed.
2003, 42, 1210. doi:10.1002/ANIE.200390319
[19] MOE (Molecular Operating Environment software) 2013 (Chemical
Computing Group Inc.: Montreal). Available at http://www.
chemcomp.com
[20] (a) R. Ahlrichs, M. Ba¨r, M. Ha¨ser, H. Horn, C. Ko¨lmel, Chem. Phys.
Lett. 1989, 162, 165. doi:10.1016/0009-2614(89)85118-8
(b) A. Scha¨fer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571.
doi:10.1063/1.463096
[10] (a) X.-G. Hu, L. Hunter, Beilstein J. Org. Chem. 2013, 9, 2696.
doi:10.3762/BJOC.9.306
(c) O. Treutler, R. Ahlrichs, J. Chem. Phys. 1995, 102, 346.
doi:10.1063/1.469408
(d) A. Klamt, G. Schu¨u¨rmann, J. Chem. Soc. Perkin Trans. 2 1993,
799. doi:10.1039/P29930000799
(b) L. Hunter, Beilstein J. Org. Chem. 2010, 6, 38. doi:10.3762/
BJOC.6.38
(c) D. O’Hagan, Chem. Soc. Rev. 2008, 37, 308. doi:10.1039/
B711844A
(e) M. von Arnim, R. Ahlrichs, J. Chem. Phys. 1999, 111, 9183.
doi:10.1063/1.479510
[21] See Supplementary Material.
[11] (a) R. Filler, R. Saha, Future Med. Chem. 2009, 1, 777. doi:10.4155/
FMC.09.65
(b) W. K. Hagmann, J. Med. Chem. 2008, 51, 4359. doi:10.1021/
JM800219F
[12] N. E. Gooseman, D. O’Hagan, M. J. G. Peach, A. M. Slawin, A. M.
Teale, D. J. Tozer, R. J. Young, Angew. Chem. Int. Ed. 2007, 46, 5904.
doi:10.1002/ANIE.200700714
[22] Consistent with earlier observations (Ref. [14c]), none of the fluoro-
hydrin azepane 1H NMR spectra revealed new conformational
equilibria at different temperatures suggesting low barriers for the
conformational interchange.
[13] H. Amii, K. Uneyama, Chem. Rev. 2009, 109, 2119. doi:10.1021/
CR800388C