The intramolecular β-fluorine...ammonium interaction in 4- and 8-membered rings

Article abstract of DOI:10.1039/b606334a

The structures of 3-fluoroazetidinium hydrochloride and 3-fluoro-1,5-diazacyclooctane hydrobromide are explored both by X-ray diffraction analysis and DFT calculations, and the conformations of these molecules are shown to be significantly influenced by t

Full text of DOI:10.1039/b606334a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
…
The intramolecular b-fluorine ammonium interaction in 4- and
8-membered rings
Natalie E. J. Gooseman,a David O’Hagan,*a Alexandra M. Z. Slawin,a Andrew M. Teale,b David J. Tozer*b
and Robert J. Youngc
Received (in Cambridge, UK) 4th May 2006, Accepted 26th May 2006
First published as an Advance Article on the web 21st June 2006
DOI: 10.1039/b606334a
vicinal difluorosystems, most classically in 1,2-difluoroethane,5–7
The structures of 3-fluoroazetidinium hydrochloride and
but extends to systems such as amides of b-fluoroethylamine8 and
3-fluoro-1,5-diazacyclooctane hydrobromide are explored both
esters of fluoroethanol.9 Generally the differences in anti–gauche
by X-ray diffraction analysis and DFT calculations, and the
energiesintheseneutralsystemsareinthe range 1.0–2.0kcalmol21
,
conformations of these molecules are shown to be significantly
+
…
influenced by the through space C–F N interaction.
whereas the energy differences between equatorial and axial
conformational preferences for C–F in the 3-fluoropyrimidiniums
1–3 are much larger, and thus the stabilising effect is considerably
larger (y4.0–5.0 kcal mol21). We have investigated the conforma-
tion of b-fluoroethylamine hydrochloride 4.10 The X-ray derived
structure of 4 shows a clear gauche conformational preference as in
4a and density functional theory (DFT) calculations indicated a
gauche 4a over anti 4b preference of y5.8 kcal mol21. Inallofthese
cases the influence of intramolecular hydrogen bonding is not
dominating. Both theoretical and X-ray derived structures do not
Snyder, Lankin et al. have reported1 that 3-fluoropiperidinium 1
had a strong preference for fluorine in the axial 1a over the
equatorial 1b conformation (see Fig. 1). Their studies2–4 have
extended to exploring the conformational preference of cis-3,5-
difluoropiperidine 2 and again they conclude a very clear
preference for both of the fluorines to lie axial 2a rather than
equatorial 2b.
+
…
They have proposed a C–F N charge dipole orientating effect
to account for this, rather than an intrinsic stereoelectronic gauche
effect. The stereoelectronic gauche effect is observed in neutral
…
revealparticularlyshortH Fcontactsinthesecases.Alsotheeffect
is retained in quaternary ammonium systems such as 3, although in
+
…
thesecasessomeN –CH3 F–Chydrogenbondingisimplicatedas
the positive charge density from nitrogen extends to these
hydrogens. None-the-less organic fluorine forms only weak
hydrogen bonding interactions11–13 and such stabilisation could
onlyaccountforuptoafew(2.0–3.0)kcalmol21 ineachcase. Inour
study10 on the fluoroethylammonium ion 4, we found that the
directionality of the hydrogen atoms towards fluorine did not
significantly influence the stabilisation energy. These observations
reinforce the charge–dipole interaction proposed by Snyder and
Landkin as the dominant interaction in these systems.
To explore the consequences of the interaction further, this
Communication evaluates its influence in the 4-membered 3-fluoro-
azetidinium14 5 ring system, the smallest b-fluoroammonium ring
system that can be constructed, and also the larger 8-membered
1,5-diaza-ammoniumcyclooctane 6 ring system (see Fig. 2).
In order to quantify and investigate the intramolecular effect,
calculations were performed on isolated systems (i.e. non periodic),
using Kohn–Sham DFT with the B97-2 hybrid exchange–
correlation energy functional.15 We have confirmed that qualita-
tively similar results are obtained with the widely used B3LYP
DFT functional and with the computationally demanding MP2
correlated wavefunction-based method. All calculations were
performed using the TZ2P basis set,16 augmented with an
additional s and p diffuse function on the non-H atoms, as in
ref. 10. Molecular structures were optimised and analytic harmonic
vibrational frequencies were calculated in order to confirm that the
located stationary points are minima on the potential energy
surface. Quoted energy differences include zero-point vibrational
corrections determined using these harmonic frequencies. All
calculations were performed using the Gaussian 03 program.17
+
…
Fig. 1 Vicinal C–F
N
axial/gauche conformations are significantly
favoured over the corresponding equatorial/trans relationships.
aSchool of Chemistry and Centre for Biomolecular Sciences, University
of St Andrews, North Haugh, St Andrews, Fife, UK KY16 9ST.
E-mail: do1@st-and.ac.uk; Fax: +44 1334 463808; Tel: +44 1334 467176
bDepartment of Chemistry, University of Durham, South Road, Durham,
UK DH1 3LE. E-mail: d.j.tozer@durham.ac.uk; Fax: +44 191 3844737;
Tel: +44 191 3342111
cHigh Throughput Medicinal Chemistry, Discovery Research, GSK,
Stevenage, UK. E-mail: rob.j.young@gsk.com
3190 | Chem. Commun., 2006, 3190–3192
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Fig. 2 X-Ray crystal structures of the 5 (Cl2) and 6ax (2Br2).{
We commenced by considering the 4-membered ring, 5. In our
Fig. 4 X-ray and DFT optimized molecular conformations.
+
previous study,10 we quantified the influence of the C–F
N
…
the positive charge, we then took the optimised structure of cation
5, added an extra electron to the LUMO (which occupies the now
HOMO molecular orbital with significant s character on the N
and opposite-phase s character on the two bonded H atoms,
together with p character on the bonded C atoms) and then re-
optimised the structure of the resulting chemically non-intuitive
neutral molecule 5a. During the course of the optimisation, the
ring pucker inverted to give the optimised structure of 5a shown in
Fig. 4, which closely resembles the amine structure 7. The
calculations clearly demonstrate that the positive charge has a
pronounced effect on the structure—the observations are fully
interaction in the cation 4 by comparing the calculated electronic
energies of the gauche and anti conformations 4a and 4b. Such an
approach cannot be used for cation 5 due to the ring constraints
where it is not possible to establish gauche and anti conformations.
We have therefore taken an alternative strategy, which suppresses
the interaction by removing the positive charge and/or the fluorine
atom. We have also considered the influence of including the
counterion in the calculations.
To investigate the influence of the net positive charge, we first
compared the structure of the b-fluoroazetidinium cation 5 with
the corresponding neutral amine 7 shown in Fig. 3. For the amine,
stationary points were determined, commencing from a range of
conformations. Three stationary points were located and the
lowest energy structure is presented in Fig. 4. The C–F and N–H
+
…
consistent with a favourable C–F N interaction.
To assess the influence of the F atom, we have performed
calculations on the corresponding non-fluorinated, azetidinium
ring 8, obtained by replacing the F atom in 5 with a hydrogen
atom. Fig. 4 shows the optimised structure of 8. The dihedral
angle, which was 100.0u in 5, is now slightly wider, at 102.3u in 8,
…
groups are spatially well separated, with F H and F
…
N
˚
distances of 3.96 A and 3.28 A, respectively. The N–C–C–F
˚
torsion angle is 137.2u.
+
…
N
which is consistent with the removal of the favourable C–F
Next, we took the optimised structure of amine 7 and added a
proton to obtain an approximate geometry for the b-fluoroaze-
tidinium cation 5. A full geometry optimisation was then
performed, commencing from this structure. During the course
of the optimisation the ring pucker inverted, relative to the F atom,
causing the C–F and +NH groups to approach one another more
closely. The optimised structure is presented in Fig 4, together with
interaction.
It was informative to include the Cl2 counter ion in calculations
on 5. We performed both an unconstrained optimisation and a
second optimisation where the positions of the hydrogen on
nitrogen and the Cl2 ion were constrained to correlate with their
positions in the X-ray derived structure. The ring puckers in a
similar manner to that found in the free amine 7 with a similar
N–C–C–F torsion angle (134.1u), reflecting the strong electrostatic
…
the corresponding X-ray structure. The calculated F H and
…
˚ ˚
N distances reduce to 3.10 A and 2.95 A, respectively and the
F
+
…
N
influence of the counterion, which clearly attenuates the C–F
N–C–C–F torsion angle reduces to 100.0u. Geometry optimisa-
tions commencing from alternative conformations also produced
the same optimised structure. To further highlight the influence of
interaction, consistent with an electrostatic interaction (structures
not shown).
Optimised structures determined in the above manner are not
directly comparable with X-ray structures, since they take no
account of intermolecular effects. None-the-less we note that the
X-ray structure of salt 5 in Fig. 2 and 4 is intermediate between the
calculated structures of the free cation 5 and that with its
counterion, with a ring puckering more closely resembling that
determined for the free cation 5.
Fig. 3 The comparitive azetidine structures evaluated by DFT.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 3190–3192 | 3191

View Article Online
Sollis (GSK) for kindly preparing18 and providing a sample of
3-fluoroazetidinium chloride 5 for X-ray analysis.
Notes and references
{ Crystal data for 5: C3H7ClFN, M = 111.55, monoclinic, space group
˚
P21/n, a = 6.583(3), b = 8.278(3), c = 9.052(3) A, b = 97.556(12)u, U =
489.0(3) A , F(000) = 232, Z = 4, Dc = 1.515 Mg m23, m = 0.645
3
˚
mm21(Mo-Ka, l = 0.71073 A). The data were collected at T = 93(2) K,
˚
2553 reflections (3.35 , h , 25.33u) were measured on a Rigaku Mercury
CCD diffractometer yielding 843 unique data (Rmerg = 0.0359).
Conventional R = 0.0887 for 69 reflections with I ¢ 2s(I), GOF =
1.000. Final wR2 = 0.0648 for all data (64 refined parameters). The largest
peak in the residual map is 0.239 e A23. CCDC 606604.Crystal data for 6:
˚
C6H15Br2FN2, M = 294.02, monoclinic, space group P21/c, a = 7.0901(16),
˚
3
˚
b = 12.250(3), c = 12.581(3) A, b = 106.266(6)u, U = 1049.0(4) A , F(000) =
576, Z = 4, Dc = 1.862 Mg m23, m = 7.691 mm21(Mo-Ka, l = 0.71073 A).
˚
The data were collected at T = 93(2) K, 6572 reflections (2.37 , h ,
25.37u) were measured on a Rigaku Mercury CCD diffractometer yielding
1887 unique data (Rmerg = 0.0333). Conventional R = 0.0284 for 1707
reflections with I > 2s(I), GOF = 1.026. Final wR2 = 0.0589 for all
data (177 refined parameters). The largest peak in the residual map is
0.646 e A23. CCDC 606605. For crystallographic data in CIF or other
˚
electronic format see DOI: 10.1039/b606334a
Scheme 1 Reagents i. TsCl, NEt3, DCM, 0 uC, 23 h, 97%; ii. Pd(OH)2/
C, H2, EtOH, 23 uC, 99%; iii. Deoxofluor, DCM, 23 uC, 18 h, 67%; iv. 13,
NaH, DMF, 100 uC, 17 h, 36%; v. HBr (33%), phenol in AcOH, 90 uC,
96 h, 100%.
1 J. P. Snyder, N. S. Chandrakumar, S. N. Rao, D. P. Spangler and
D. C. Lankin, J. Am. Chem. Soc., 1993, 115, 3356–3357.
2 J. P. Snyder, N. S. Chandrakumar, H. Sato and D. C. Lankin, J. Am.
Chem. Soc., 2000, 122, 544–545.
In this study we have also prepared the 3-fluoro-1,5-diazacy-
clooctane HBr salt 6. This allowed an exploration of a
conformationally flexible larger ring system, and in this particular
case, the fluorine has the potential to participate in two
3 A. M. Sum, D. C. Lankin, K. Hardcastle and J. P. Snyder, Chem.–Eur.
J., 2005, 11, 1579–1591.
4 J. P. Snyder, G. L. Grunewald, F. A. Romero, I. Y. Oren and
D. C. Lankin, Org. Lett., 2004, 4, 3557–3560.
5 S. Wolfe, Acc. Chem. Res., 1972, 5, 102–111.
6 S. F. Nelsen, Acc. Chem. Res., 1978, 11, 14–20.
7 P. R. Rablen, R. W. Hoffmann, D. A. Horvat and W. T. Borden,
J. Chem. Soc., Perkin Trans. 2, 1999, 1719–1726.
8 D. O’Hagan, C. Bilton, J. A. K. Howard, L. Knight and D. J. Tozer,
J. Chem. Soc., Perkin Trans. 2, 2000, 605–607.
9 C. R. S. Briggs, D. O’Hagan, H. S. Rzepa and A. M. Z. Slawin,
J. Fluorine Chem., 2004, 125, 19–25.
10 C. R. S. Briggs, M. J. Allen, D. O’Hagan, D. J. Tozer, A. M. Z. Slawin,
A. E. Goeta and J. A. K. Howard, Org. Biomol. Chem., 2004, 2,
732–740.
11 J. A. K. Howard, V. J. Hoy, D. O’Hagan and G. T. Smith, Tetrahedron,
1996, 52, 12613–12622.
12 J. D. Dunitz and R. Taylor, Chem.–Eur. J., 1997, 3, 89–98.
13 J. D. Dunitz, ChemBioChem, 2004, 5, 614–621.
+
…
intramolecular C–F
N
interactions. The synthetic route to 6
started from the glycerol ether 9 followed by tosylation, ether
cleavage and then fluorination to generate 12. Reaction of 12 with
the bis-N-tosylamide 13 generated the desired ring system with 14,
and the HBr salt 6 was isolated after HBr treatment of 14. These
transformations are summarised in Scheme 1. It is obvious from
the resultant X-ray structure shown in Fig. 2 that the C–F bond
occupies an axial orientation. There was no evidence of any
disorder in the structure and particularly of any molecules with the
C–F bond lying in an equatorial conformation.
DFT absolute energy calculations comparing the 6ax and 6eq
free cation structures indicate a preference for the axial conforma-
tional isomer of 9.2 kcal mol21. This energy difference is
comparable to that found2 for 2a (8.9 kcal mol21), again reflecting
14 3-Fluoroazetidinium chloride was obtained from the GSK compound
collection.
15 T. W. Bradley, P. J. Wilson and D. J. Tozer, J. Chem. Phys., 2001, 115,
9233–9242.
+
16 T. H. Dunning, J. Chem. Phys., 1971, 55, 716–723.
17 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin,
J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene,
X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. Ochterski, P. Y. Ayala, K. Morokuma,
G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski,
S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui,
A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox,
T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
M. Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen,
M. W. Wong, C. Gonzalez and J. A. Pople, GAUSSIAN 03
(Revision C.02), Gaussian, Inc., Wallingford, CT, 2004.
18 L. Provins, B. J. Van Keulen, J. Surtees, P. Talaga and B. Christophe,
PCT Int. Appl. WO, 087064 A1, 2003, .
…
the presence of two C–F N interactions.
+
…
In summary the intramolecular C–F N interaction has been
explored in a small and a large ring system and in both cases it is
significant in influencing the ring conformation. The magnitude of
+
…
a single C–F N interaction is similar to that of a good hydrogen
bond and it merits consideration in the design of biologically
relevant amine analogues, where the C–F bond can be inserted as a
strategic tool for influencing conformation without dramatically
affecting the steric profile of a given molecule. In view of the fact
that most aliphatic amines are protonated at physiological pH,
fluorine incorporation b to amines will be expected to have a
dramatic influence on the solution conformation of bio-active
amines, through both pKa modulation and the profound
b-fluorine ammonium interaction.
We thank the BBSRC and GSK for a CASE Studentship
(NEJG), the EPSRC for a DTA Studentship (AMT), and Steve
3192 | Chem. Commun., 2006, 3190–3192
This journal is ß The Royal Society of Chemistry 2006