Fluorination of steroid estrogens with Selectfluor??: Elucidation of regio- and stereoselectivity

Article abstract of DOI:10.1016/j.jfluchem.2014.09.030

Two steroid estrogens, natural estradiol and 8?±-estradiol, were fluorinated using Selectfluor?? in ionic liquid. Unexpectedly, mostly products of fluorination at position C-10 were isolated instead of corresponding fluorophenols, as it was reported previ

Full text of DOI:10.1016/j.jfluchem.2014.09.030

Journal of Fluorine Chemistry 168 (2014) 218–222
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
1
Fluorination of steroid estrogens with Selectfluor : Elucidation of
regio- and stereoselectivity
Roman P. Bogautdinov, Alan F. Fidarov *, Svetlana N. Morozkina, Andrei A. Zolotarev,
Galina L. Starova, Stanislav I. Selivanov, Alexander G. Shavva
Department of Natural Products Chemistry, Institute of Chemistry, Saint-Petersburg State University, Universitetsky pr. 26, Petrodvorets 198504, Russia
A R T I C L E I N F O
A B S T R A C T
1
Two steroid estrogens, natural estradiol and 8a-estradiol, were fluorinated using Selectfluor in ionic
Article history:
Received 5 September 2014
Received in revised form 25 September 2014
Accepted 27 September 2014
Available online 6 October 2014
liquid. Unexpectedly, mostly products of fluorination at position C-10 were isolated instead of
corresponding fluorophenols, as it was reported previously. The reaction proceeds stereoselectively,
yielding different stereoisomers for different types of the steroid skeleton. Stereoselectivity was proved
by 1D/2D NMR experiments and X-ray analysis of obtained products. The obtained compounds are
suitable intermediates for synthesis of various classes of estrogens.
Keywords:
ß 2014 Elsevier B.V. All rights reserved.
1
Selectfluor
8
1
a
-Estradiol
0-Fluorosteroids
Stereoselectivity of fluorination
NMR spectroscopy methods
1
. Introduction
Since steroidal estrogens used in hormonal replacement
hormonal activity and can be less vulnerable to metabolic
oxidation as compared to natural substrates. Unexpectedly, mostly
isolated was the product with fluorination at position C-10, 3b,
along with minute amount of 2- and 4-fluoro-8 -estradiol 1b and
therapy are proved to be carcinogenic [1], there is an urgent need
for corresponding new safe medications. Liehr in his work has
shown that carcinogenicity and estrogenicity can be separated [2];
the obvious example is 2-fluoroestradiol 1a (Fig. 1), an active
estrogen that possesses no carcinogenic properties. Unfortunately,
this substance is rather hard to synthesize. The known methods
include various types of fluorination using decomposition of
diazonium salts [3], reactions with acetyl hypofluorite [4], cesium
fluorosulfate [5] etc. All of these methods provide yields and
regioselectivities only from poor to mediocre and require harsh
conditions or highly active instable reagents.
a
4b. Further, we have checked the original method on natural
estradiol and got analogous results, that are, in turn, completely
different from those reported by Heravi [6]. Structure elucidation
and observations on the stereoselectivity of fluorination are
reported in this paper. Such compounds can be used in synthesis
of antraquinone-estrogen hybrids [11], 3-aminosteroids [12] and
1-hydroxy-4-fluorosteroids [13].
2. Experimental
1
Heravi reported that under treatment by Selectfluor in ionic
NMR spectroscopic data were recorded with Bruker DPX-300
1
liquid estradiol 2a was converted into 2-fluoroestradiol 1a with
excellent yield [6]. That was the best result of such fluorination so
far, although it contradicted with results of some other electro-
philic fluorinations using different N–F derivatives [7,8].
and Avance 400 spectrometers (300 and 400 MHz for H, 75 and
1
3
125 MHz for C, respectively) in CDCl3 and DMSO-d6 and are
referenced to the residual solvent proton (dH = 7.26 and 2.50 ppm,
respectively) and carbon signals (dC = 77.00 and 39.52 ppm,
respectively). High-resolution mass spectra were recorded with
a Bruker Maxis HRMS-ESI-Q-TOF spectrometer (electrospray
ionization mode). X-ray diffraction analysis was performed with
an Agilent Technologies Xcalibur diffractometer. Melting points
were recorded on a Buchi melting point apparatus and are
uncorrected. Fine chemicals were purchased from the Sigma-
Aldrich Co. LLC. Chromatographic purification was performed
using Acros Organics silica gel 60, 35–70 mesh.
In this work we made an attempt to perform this reaction with
another substrate, 8
Scheme 1); various 2-fluoro-8
osteo- and cardioprotective properties [9,10] while retaining
a
-estradiol 2b, under the same conditions
(
a-estrogens are known to possess
*
Corresponding author. Tel.: +7 9117504828.
E-mail address: f-alan@mail.ru (A.F. Fidarov).
http://dx.doi.org/10.1016/j.jfluchem.2014.09.030
022-1139/ß 2014 Elsevier B.V. All rights reserved.
0

R.P. Bogautdinov et al. / Journal of Fluorine Chemistry 168 (2014) 218–222
219
+
+
2
HRMS (ESI-TOF), Calc. for [M + H] , [E18=24F? ] : 291.1755,
found 291.1751.
2.2. X-ray diffraction studies
Single crystals of 3a were grown from methanol, single crystals
of 3b—from Et O. A suitable crystal was selected and analyzed on a
2
Fig. 1. Structures of 2-fluoroestradiol 1a and estradiol 2a.
SuperNova, Dual, Cu at zero, Atlas diffractometer. The crystal was
kept at 100(2) K during data collection. Using Olex2 [14], the
structure was solved with the SHELXS structure solution program
using Direct Methods and refined with the SHELXL refinement
package using Least Squares minimization [15].
2
.1. General procedure for the reaction of estrogen analogues with
1
Selectfluor
Crystallographic data form compounds have been deposited
with the Cambridge Crystallographic Data Center as supplemen-
tary material numbers CCDC 1014305 (3a) and 1014306 (3b).
1
A suspension of steroid 2a, b (0.2 mmol) and Selectfluor
0.21 mmol) in the 1:1 mixture of bmimBF (0.2 g) and CH OH
0.2 g) was stirred for 1 h at 20 8C. The mixture was then poured in
(
(
4
3
water (50 ml) and extracted with ethyl acetate (3 Â 25 ml). The
3. Results and discussion
combined organic layers were washed with brine, dried over
anhydrous MgSO
chromatography of the residue (petroleum ether/ethyl acetate
:1) afforded 3a, b and the mixture of 1a, b, 4a, b and starting
material 2a, b.
-Fluoroestra-1,4-diene-17
51–153 8C (lit. mp 152–154 8C [13]).
4
and the solvent was evaporated. Flash
3.1. Synthesis of 10-fluorosteroids
4
The initial task in this paper was reproduction of previous
results [6] with unnatural estrogen analogue, 8
a-estradiol 2b,
1
0
b
b-ol-3-one 3a. Yield 63%. mp
under the same conditions. However, treatment of this compound
1
1
4
with Selectfluor in 1:1 mixture of ionic liquid (bmimBF ) and
1
H NMR (CDCl
3
,
d): 0.86 s (3H), 0.96–1.56 m (8H), 1.60–1.69 m
methanol for 24 h resulted only in 83% conversion, yielding 10-
fluorosteroid 3b (69%) and an inseparable mixture of 2- and 4-
fluoro-8alpha-estradiol 1b and 4b (4%, 1:1, determined by NMR)
(Scheme 1). Such results were completely unexpected, since the
author of original method reported only the products of fluorina-
tion at positions 2 and 4 with complete conversion of substrate and
corresponding 88%/11% yields, using natural estradiol under the
same conditions. We assumed that molecular geometry can affect
the regioselectivity of fluorination and conducted a control
experiment with natural estradiol as described by Heravi.
Nevertheless, our results here were also totally different—
incomplete conversion and 10-fluorosteroid 3a as a main product
(Scheme 1).
(
(
(
1H), 1.76–1.85 m (1H), 1.84–2.04 m (4H), 2.04–2.15, 2.40–2.47 m
1H), 2.63–2.75 m (1H), 3.66 t (1H, J = 8.5 Hz), 6.04 brs (1H), 6.24 d
1H, J = 10.2 Hz), 7.09 dd (1H, J = 10.4 Hz, 7.6 Hz).
13
3
C NMR (CDCl , d): 10.9, 22.6, 23.5, 30.3, 31.8, 32.6, 35.8, 36.1,
4
1
1
3.1, 49.7, 54.2 d (JC,F = 25.0 Hz), 81.4, 89.2 d (JC,F = 168.0 Hz),
23.6 d (JC,F = 4.4 Hz), 129.4 d (JC,F = 8.1 Hz), 145.1 d (JC,F = 24.2 Hz),
60.3 d (JC,F = 19.1 Hz), 185.1 d (JC,F = 4.4 Hz).
+
+
HRMS (ESI–TOF), Calc. for [M + H] , [E18
found 291.1743.
-Fluoro-8
mp 121–123 8C.
2
=24F? ] : 291.1755,
1
0
a
a-estra-1,4-diene-17b-ol-3-one 3b. Yield 69%.
1
H NMR (DMSO-d
m, 2H), 2.07–2.34 (m, 2H), 2.35–2.45 (m, 2H), 3.27–3.46 (m, 1H),
.53 (d, 1H, J = 4.5 Hz), 6.14 (s, 1H), 6.23 (d, 1H, J = 10.0 Hz), 6.99
dd, 1H, J = 10.0 Hz, 6.7 Hz).
6
): d 0.69 (s, 3H), 0.87–1.69 (m, 9H), 1.73–1.95
(
4
(
3.2. Elucidation of fluorination stereoselectivity
13
C NMR (DMSO-d
6
):
d
13.77, 18.01 (JC,F = 7.1 Hz), 23.01, 25.13,
The structure of the first analogue 3b obtained from 8
a-
3
8
0.14, 31.51, 35.01, 36.96, 42.63, 47.09, 49.05 (JC,F = 25.5 Hz),
0.69, 90.81 (JC,F = 160.3 Hz), 125.64, 130.25 (JC,F = 8.4 Hz), 147.35
C,F = 21.4 Hz), 158.34 (JC,F = 18.6 Hz), 186.02.
estradiol was examined by NMR spectroscopy methods. First, a full
1
13
signal-atom assignment was performed for H and C NMR
1
9
(
J
spectra (Table 1). A F spectrum was also recorded and displayed
Scheme 1. Fluorination of estradiol 2a and 8a-estradiol 2b.

2
20
R.P. Bogautdinov et al. / Journal of Fluorine Chemistry 168 (2014) 218–222
Table 1
one double doublet (6.5 and 12.5 Hz) because of scalar interactions
with H1 and H9 protons correspondingly. These interactions
Signal-atom assignment for steroid 3b.
a
1
13
H assignment
C assignment
were established by analysis of multiplet structure of correspond-
1
1
n
13
n
ing signals in H spectrum as well as cross-peaks in several homo-
Atom no
d
( =), ppm ( J=–F
)
Atom no
d( E), ppm ( JC–F)
1
2
4
6.99 (6.5 Hz)
6.23 (1.6 Hz)
6.14 (1.8 Hz)
2.40
1
2
3
4
146.42 (21.4 Hz)
129.31 (8.4 Hz)
185.12 (4.6 Hz)
124.66 (4.8 Hz)
and heteronuclear correlation spectra (COSY-DQF, J-COSY, HSQC
and NOESY) (Fig. 2).
3
However, the obtained value of vicinal constant
J
F10–
6
6
7
7
8
9
1
1
1
1
1
1
1
1
1
1
1
1
a
b
a
b
a
a
1
1
2
2
4
5
5
6
6
7
7
8
H9
a
= 12.5 Hz cannot serve as unambiguous proof of their mutual
2.40
cis- or trans-orientation, and, therefore, cannot be used as a clue for
determination of fluorine spatial position.
Therefore, additional NOE experiments were conducted for two
types of spatial interactions in the molecule—H–H and H–F.
1.81
5
157.42 (18.6 Hz)
1.58
2.29
6
7
8
30.57
24.19
34.08
2.17 (12.5 Hz)
1.20
a
Firstly, in case of homonuclear experiment,
a-orientation was
b
a
b
a
a
b
a
b
a
b
1.06
0.96
9
49.05 (25.5 Hz)
assigned for fluorine atom, based on estimated ratio of distances
1.60
between protons r=1–=9 = 2.58 A and r=1–=11 = 2.37 A. The same
˚
˚
a
a
1.38
10
11
90.81 (160.3 Hz)
18.01 (7.1 Hz)
conclusion can be made considering results of heteronuclear
1.38
1
experiment: integral intensities of H9
a
, H8
a
and H6
a
in
H
1.38
spectrum rise upon selective saturation of fluorine signal (Fig. 3).
Thus, the results of homo- and heteronuclear experiments on
measuring NOE in this molecule represent independent evidences
1.83
12
36.02
1.26
3.37
13
14
15
41.69
46.15
22.07
29.20
79.75
12.82
-OH
2
<H O>
for
a
-orientation of fluorine atom at position 10.
Lately, an X-ray crystallographic analysis was performed for
both analogues, also confirming the result of NMR studies for 8
compound 3b (Table 2) (Fig. 4). Interestingly, the stereoselectivity
0.69
1
1
1
6
7
8
a
-
Fig. 2. Fragments of compound’s 3b EOSY, NOESY, J-COSY and HSQC spectra used for signal-atom assignment and for determination of scalar constants.

R.P. Bogautdinov et al. / Journal of Fluorine Chemistry 168 (2014) 218–222
221
1
19
Fig. 3. The results of heteronuclear H{ F} NOE-difference experiment (
t
m
= 1.0 s) for compound 3b.
Table 2
X-ray crystal structure parameters of compounds 3a and 3b.
Crystal data and structure refinements
Compound
3a
3b
Empirical formula
Formula weight
Temperature (K)
Crystal system
C
18
H
23
O
2
F
18 23 2
C H O F
290.36
100 (2)
290.36
100 (2)
Orthorhombic
P2
Monoclinic
Space group
1
2
1
2
1
Cc
a ( A˚ )
b ( A˚ )
c ( A˚ )
6.7680 (2)
12.0219 (3)
18.4639 (6)
90.00
15.5496 (2)
15.6885 (3)
24.7217 (4)
a
b
g
(8)
(8)
(8)
90.00
90.00
94.2025 (16)
90.00
90.00
Volume (A˚
3
)
1502.30 (8)
4
6014.61 (17)
Z
16
3
r
calc (mg/mm )
1.284
1.283
À1)
m (mm
F (000)
0.729
0.729
624.0
2496.0
3
Crystal size (mm )
Range for data collection
0.22 Â 0.18 Â 0.10
7.18 to 144.948
À16 ꢀ h ꢀ 19, À19 ꢀ kꢀ 19, À30 ꢀ l ꢀ 30
34080
2Q
8.78 to 1458
Index ranges
À8 ꢀ h ꢀ 8, À14 ꢀ k ꢀ 5, À18 ꢀ l ꢀ 22
Reflections collected
Independent reflections
Data/restraints/parameters
7203
2890[R(int) = 0.0300]
2890/0/192
1.048
9222[R(int) = 0.0469]
9222/2/765
2
Goodness-of-fit on F
1.038
Final R indexes [I > = 2
s
(I)]
R
R
1
= 0.0331, wR
= 0.0373, wR
2
2
= 0.0811
= 0.0838
R
R
1
= 0.0423, wR
= 0.0466, wR
2
2
= 0.1075
= 0.1114
Final R indexes [all data]
Largest diff. peak/hole (e A˚
Flack parameter
1
1
À3
)
0.20/À0.15
À0.06 (15)
0.26/À0.19
À0.08 (11)

2
22
R.P. Bogautdinov et al. / Journal of Fluorine Chemistry 168 (2014) 218–222
Fig. 4. Molecular structures of 3a (left) and 3b (right) (X-Ray data).
of fluorination in case of natural estradiol is reversed and only
-fluorosteroid 3a was obtained. This fact may lead to an
assumption that position 10 in 8 -steroids is more hindered for
-face of the molecule. The
reversed situation may be surmised for natural steroid estrogens.
Noteworthy, although analogous 10-chlorosteroid readily and
irreversibly rearranges into 4-chloroestrone and 2-chloroestrone
1
0
b
Appendix A. Supplementary data
a
1
the attack of Selectfluor from
b
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.jfluchem.
2014.09.030.
upon heating in acetonitrile [16], 10
stable and remains intact under these conditions.
b-fluorosteroid 3a is more
References
[1] The Endogenous Hormones and Breast Cancer Collaborative Group, J. Nat. Cancer
Inst. 94 (2002) 606–616.
2] J.G. Liehr, Mol. Pharmacol. 23 (1983) 278–281.
3] T. Utne, R.B. Jobson, R.D. Babson, J. Org. Chem. 33 (1968) 2469–2473.
[4] J.S. Ward, C.D. Jones, US4522758, 1985.
4
. Conclusions
[
[
1
-estradiol with Selectfluor in
The reaction of estradiol and 8
:1 mixture of bmimBF
a
1
4
and methanol leads to 10-fluorosteroids
[5] L.J. Diorazio, D.A. Widdowson, J.M. Clough, J. Chem. Soc., Perkin Trans. 1 (1992)
421–425.
as main products. The fluorination occurs stereoselectively,
yielding 10 -fluorosteroid with natural estradiol as a substrate
and 10 -fluorosteroid with 8 -estradiol in good yields. These
compounds are suitable intermediates for synthesis of various
classes steroids: antraquinone-estrogen hybrids, 3-aminosteroids
and 1-hydroxy-4-fluorosteroids.
[
[
6] M.R.P. Heravi, J. Fluorine Chem. 129 (2008) 217–221.
7] W.T. Pennington, G. Resnati, D.D. DesMarteau, J. Org. Chem. 57 (1992) 1536–
1539.
8] S. Stavber, M. Jereb, M. Zupan, Synlett 1999 (1999) 1375–1378.
9] V.N. Belov, V.Y. Dudkin, E.A. Urusova, G.L. Starova, S.I. Selivanov, S.V. Nikolaev,
N.D. Eshchenko, S.N. Morozkina, A.G. Shavva, Russ. J. Bioorg. Chem. 33 (2007)
293–301.
b
a
a
[
[
[
10] S.N. Morozkina, A.F. Fidarov, A.G. Shavva, J. Steroid Biochem. Mol. Biol. 126S
2011) e180.
[11] P. Pahari, B. Senapati, D. Mal, Tetrahedron Lett. 48 (2007) 2635–2638.
(
Acknowledgements
[
[
[
12] E. Hecker, M. Hopp, Justus Liebigs Ann. Chem. 692 (1966) 174–179.
13] J.S. Mills, J. Barrera, E. Olivares, H. Garc ı´ a, J. Am. Chem. Soc. 82 (1960) 5882–5889.
14] O.V. Dolomanov, L.J. Bourhis, R.J. Gildea, J.A.K. Howard, H. Puschmann, J. Appl.
Crystallogr. 42 (2009) 339–341.
NMR, HRMS and XRD analyses were performed at the Saint
Petersburg State University Center for Magnetic Resonance, Center
for Chemical Analysis and Materials Research and X-ray Diffraction
Center, respectively.
[15] G.M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr. 64 (2008) 112–122.
16] M. Maumy, P. Capdevielle, Tetrahedron 49 (1993) 7455–7462.
[