Crystal structures of 10α-gon-5-enes from the synthetic pathway to desogestrel

Article abstract of DOI:10.1016/j.steroids.2005.03.006

X-ray crystallographic studies performed on the product of the ketalization reaction of 13β-ethyl-11α-hydroxy-gon-5-ene-3,17-dione have lead to the unequivocal assignment of the 10α stereochemistry to C10, showing that an inversion of configuration occurred during formation of the 3,17-diketal. From the Swern oxidation of this compound, 11α-(methylthio)methoxy- 10α-gonene was obtained as the major product instead of the desired 11-ketone. Modeling studies showed that the configurational instability at C10 is determined by the presence of the 11α-hydroxyl group.

Full text of DOI:10.1016/j.steroids.2005.03.006

Steroids 70 (2005) 660–666
Crystal structures of 10␣-gon-5-enes from the synthetic
pathway to desogestrel
Emilia Modicaa, Diego Colomboa, Fiamma Ronchettia,∗, Antonio Scalab,
Gabriella Bombieric, Nicoletta Marchinic, Lucio Tomad
a
Dipartimento di Chimica, Biochimica e Biotecnologie per la Medicina, Universita` di Milano, Via Saldini 50, 20133 Milano, Italy
b
Facolta` di Medicina e Chirurgia, Universita` Vita-Salute San Raffaele, Via Olgettina 58, 20132 Milano, Italy
c
Istituto di Chimica Farmaceutica e Tossicologica, Universita` di Milano, Viale Abruzzi 42, 20131 Milano, Italy
d
Dipartimento di Chimica Organica, Universita` di Pavia, Via Taramelli 10, 27100 Pavia, Italy
Received 13 January 2005; received in revised form 22 March 2005; accepted 28 March 2005
Available online 4 May 2005
Abstract
X-ray crystallographic studies performed on the product of the ketalization reaction of 13␤-ethyl-11␣-hydroxy-gon-5-ene-3,17-dione have
lead to the unequivocal assignment of the 10␣ stereochemistry to C10, showing that an inversion of configuration occurred during formation
of the 3,17-diketal. From the Swern oxidation of this compound, 11␣-(methylthio)methoxy-10␣-gonene was obtained as the major product
instead of the desired 11-ketone. Modeling studies showed that the configurational instability at C10 is determined by the presence of the
11␣-hydroxyl group.
© 2005 Elsevier Inc. All rights reserved.
Keywords: Crystal structures; 10␣-Steroids; Ketalization; Desogestrel; Swern oxidation
1. Introduction
Moreover, in order to readdress the synthetic approach
towards the synthesis of the 10␣-analogue of desogestrel [8],
Swern oxidation [9], a mild oxidative method widely used in
the case of steroidal compounds [10,11], was applied to the
ketal 4. Under Swern conditions, instead of the expected 11-
keto compound, the major product of the reaction was found
to be the (methylthio)methyl ether 5.
In this paper, we report the structures (Scheme 1) of the
10␣-gonenes 4 and 5, together with that of compound 3,
which has the “natural” 10␤-stereochemistry at C10, and we
describe modeling studies to explain the configurational in-
stability at C10.
During our studies [1,2] on the synthesis of 13-ethyl-11-
methylene-18,19-dinor-17␣-pregn-4-en-20-yn-17-ol, deso-
gestrel (1), the most prescribed third generation oral contra-
ceptive [3–5], the 13␤-ethyl-11␣-hydroxy-gon-4-ene-3,17-
dione intermediate (2) [6] was submitted to a ketalization
reaction in order to protect the 3- and 17-keto functions as
described by Gao et al. [7], who reported the formation of the
bisdiethyleneketal 3, easily recovered in high yield from the
reaction mixture.
In our hands, after crystallization, the acetalization af-
forded 13␤-ethyl-11␣-hydroxy-10␣-gon-5-ene-3,17-dione-
3,17-diethyleneketal (4), in which the 10-hydrogen atom was
on the ␣-side of the molecule. The “unnatural” stereochem-
istry at C10 of 4 was unambiguously assigned by X-ray
diffraction.
2. Experimental
2.1. General
¨
Uncorrected melting points were determined on a Buchi
∗
apparatus. Optical rotations were determined on a Perkin-
Corresponding author. Tel.: +39 02 50316037; fax: +39 02 50316036.
E-mail address: fiamma.ronchetti@unimi.it (F. Ronchetti).
Elmer 241 polarimeter in a 1 dm cell at 20 ◦C using chloro-
0039-128X/$ – see front matter © 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2005.03.006

E. Modica et al. / Steroids 70 (2005) 660–666
661
(0.7 g, 3.7 mmol) were added to a solution of 13␤-ethyl-
11␣-hydroxy-gon-5-ene-3,17-dione (2) (5 g, 16.6 mmol) in
CH2Cl2 (90 ml), and the mixture was refluxed for 6 h under
argon. The reaction was worked up by washing with aqueous
NaHCO3, and the organic layer was washed with brine, dried
over anhydrous sodium sulfate, and evaporated to dryness.
From the reaction mixture, pure compound 4 (3.5 g) was ob-
tained after crystallization from methanol. Mp 158–160 ◦C;
1
[␣]D −204.0◦ (c 1); selected H NMR signals: δ 1.00 (3H,
t, J = 8 Hz, CH2CH3), 3.70 (1H, ddd, J = 10, 10, and 4 Hz,
11-H), 3.82–3.98 (8H, broad, 3- and 17-ethyleneketal), 5.42
(1H, dd, J = 6 and 1 Hz, 6-H); 13C NMR: δ 9.5, 21.2, 22.3,
25.7, 31.1, 31.8, 35.5, 36.0, 37.1, 38.0, 46.2, 48.9, 49.6, 51.9,
65.0, 65.1, 65.2, 66.0, 68.6, 110.2, 120.5, 121.0, 138.1. EI-
MS m/z 390 [M]+. C23H34O5 (390.50): calcd. C 70.74, H
8.78; found C 71.10, H 8.84.
2.2.2. 13β-Ethyl-11α-hydroxy-gon-5-ene-3,17-
dione-3,17-diethylene ketal (3)
Scheme 1.
An analytical sample of pure compound 3 was obtained,
from the mother liquors of the reaction mixture described
above, using reversed-phase HPLC on a 25 cm × 4 mm col-
umn packed with 5 ␮m LiChrosorb RP-18. Initial conditions
for the chromatography were 65% solvent A (water) and 35%
solvent B (CH3CN). A gradient was initiated after 5 min; the
concentration of solvent B was increased linearly from 35 to
70% over a period of 20 min and these conditions were main-
tained for an additional 10 min. The flow-rate was 1.2 ml/min,
T = 30 ◦C, UV detector 215 nm. The retention times were
19.03 min for compound 3 and 16.34 min for compound 4.
3: Mp 186–188 ◦C (from methanol); [␣]D −23.0◦ (c 1);
selected 1H NMR signals: δ 1.0 (3H, t, J = 8 Hz, CH2CH3),
2.68 (1H, m, 10-H), 3.56 (1H, m, 11-H), 3.80–4.00 (8H,
broad, 3- and 17-ethylene ketal), 5.50 (1H, dd, J = 6 and 1 Hz,
6-H); 13C NMR: δ 9.7, 21.4, 22.3, 30.2, 32.6, 35.6, 35.7, 36.0,
38.5, 43.9, 45.7, 49.1, 51.1, 52.7, 65.0, 65.1, 65.2, 66.0, 75.7,
110.3, 120.0, 120.4, 138.4. EI-MS m/z 390 [M]+. C23H34O5
(390.50): calcd. C 70.74, H 8.78; found C 71.27, H 8.86.
form solutions. Dry solvents and liquid reagents were dis-
tilled prior to use; triethylamine and pyridine were distilled
from calcium hydride, and DMSO was dried over 4 A molec-
˚
ular sieves. All reagents were purchased from Aldrich. All re-
actions were monitored by TLC on Silica Gel 60 F-254 plates
(Merck) with detection by spraying with 50% H2SO4 solu-
tion and heating to 110 ◦C. Flash column chromatography
was performed on Silica Gel 60 (230–400 mesh, Merck).
Mass spectrometry was performed on a particle beam
quadrupolar mass spectrometer Hewlett-Packard HP 5988A
equipped with an interface PB 5998A and a low pressure
HPLC HP 1050. The mass spectrometric analyses were per-
formed in electron impact (EI-MS) ionization mode at an
electron energy of 70 eV with a source temperature of 250 ◦C.
Samples were dissolved in methanol and introduced into the
mass spectrometer via the particle beam LC–MS interface
directly connected with the HPLC system. Elution was per-
formed with methanol (0.4 ml/min).
The HPLC system consisted of an L-6200 Single
pump (Merck-Hitachi), an L-4250 UV–vis detector (Merck-
Hitachi), and a T-6300 Column Thermostat (Merck). The
recorder was a Shimadzu C-R6A Chromatopac.
All NMR spectra were recorded at 303 K with a Bruker
FT-NMR Avance DRX500 spectrometer in CDCl3 solution;
chemical shifts are reported as δ (ppm) relative to CHCl3
fixed at 7.24 ppm for 1H NMR spectra and relative to CDCl3
fixed at 77.00 ppm for 13C NMR spectra.
2.2.3. 13β-Ethyl-11α- (methylthio)methoxy-10α-gon-5-
ene-3,17-dione-3,17-diethylene ketal
(5)
Triethylamine (8.6 ml, 61 mmol) and pyridine (0.5 ml,
5 mmol) were added to a solution of 4 (1 g, 2.6 mmol) dis-
solved in dimethyl sulfoxide (5.7 ml, 80.5 mmol) under ar-
gon.The mixture was cooled to 10 ◦C, and sulfur trioxide
pyridine complex (2 g, 12.6 mmol) was added over a period of
4 h. The mixture was stirred for another 3 h at 10 ◦C and then
diluted with water (25 ml). The solution was extracted with
toluene (3 × 20 ml), and the combined organic phases were
washed with brine, dried over sodium sulphate, and evapo-
rated to dryness. The resulting crude reaction mixture was
purified by flash chromatography hexane/ethyl acetate (8:2)
affording 5 (702 mg, 60%). Mp 126–128 ◦C (from methanol);
[␣]D −285.5◦ (c 1); selected 1H NMR signals: δ 1.0 (3H, t,
J = 8 Hz, CH2CH3), 3.75 (1H, ddd, J = 10, 10, and 4 Hz,
2.2. Chemistry
2.2.1. 13β-Ethyl-11α-hydroxy-10α-gon-5-ene-3,17-
dione-3,17-diethylene ketal
(4)
Ethylene glycol (40 ml, 717 mmol), triethyl orthofor-
mate (20 ml, 120 mmol) and p-toluenesulfonic acid hydrate

662
E. Modica et al. / Steroids 70 (2005) 660–666
Table 1
Crystal data and structure refinement for compounds 3, 4, and 5
Compound 3
Compound 4
23H34O5
Compound 5
Empirical formula
Formula weight
Temperature (K)
C
23H34O5
C
C25H38O5S
450.61
390.50
390.50
293(2)
293(2)
293(2)
˚
Wavelength (A)
0.71073
0.71073
0.71073
Crystal system
Space group
Unit cell parameters (A)
Orthorhombic
P212121(19)
a = 7.302(1)
b = 11.818(3)
c = 23.745(5)
Monoclinic
P21(4)
a = 10.413(3)
b = 7.696(1)
c = 13.423(5)
β = 111.44(3)◦
1001.3(5)
Orthorhombic
P212121(19)
a = 7.972(2)
b = 12.331(2)
c = 24.437(8)
˚
3
˚
Volume (A )
2049.1(7)
2402.2(11)
Z
4
2
4
Calculated density (Mg/m3)
Absolute coefficient (mm−1
F(0 0 0)
1.266
0.087
848
1.292
0.089
424
1.246
0.167
976
)
Crystal size (mm3)
Theta range (◦)
0.7 × 0.9 × 0.6
0.8 × 0.9 × 0.5
0.5 × 0.8 × 0.7
2.43–22.94
2.10–24.96
2.35–22.99
Limit. indexes
−1 ≤ h ≤ 8, −1 ≤ k ≤ 12, −1 ≤ l ≤ 23
2199/2030 [Rint = 0.0563]
22.94, 98.2%
−12 ≤ h ≤ 11, −1 ≤ k ≤ 9, −1 ≤ l ≤ 15
2474/2180 [Rint = 0.0302]
24.96, 99.9%
−1 ≤ h ≤ 8, −1 ≤ k ≤ 13, 0 ≤ l ≤ 26
2461/2336 [Rint = 0.0365]
22.99, 99.7%
Full-matrix least-squares on F2
2336/0/362
Refl coll./unique
Compl. to theta
Refinement method
Data/rest./param.
Goodness-of-fit
Fin. R ind[I > 2σ(I)]
Full-matrix least-squares on F2
2030/0/365
Full-matrix least-squares on F2
2180/1/389
1.109
R1 = 0.0529
1.104
R1 = 0.0359
1.140
R1 = 0.0598
wR2 = 0.1524
0.316, −0.327
wR2 = 0.0927
0.176, −0.190
wR2 = 0.1504
0.604, −0.437
−3
˚
Larg. diff. peak and hole (A
)
11-H), 3.84–3.98 (8H, broad, 3- and 17-ethylene ketal), 4.63
(1H, d, J = 12 Hz, S–CHa), 4.71 (1H, d, J = 12 Hz, S–CHb),
5.40 (1H, dd, J = 6 and 1 Hz, 6-H). 13C NMR: δ 9.7, 15.0,
21.0, 22.4, 25.8, 30.9, 31.2, 32.2, 35.5, 36.3, 38.3, 46.3, 47.2,
48.4, 51.7, 64.9, 65.0, 65.2, 66.0, 72.1, 72.8, 110.3, 120.0,
120.4, 138.4. EI-MS m/z 450 [M]+. C25H38O5S (450.61):
calcd. C 66.63, H 8.50; found C 66.90, H 8.57.
and angles are reported in Table 2. The supplementary crys-
tallographic data have been deposited with the Cambridge
Crystallographic Data Centre (CCDC deposition numbers
254438, 254439, 254440). Copies can be obtained free of
charge from: CCDC, 12 Union Road, Cambridge CB2 1EZ,
UK;fax:+441223336033;e-mail:deposit@ccdc.cam.ac.uk.
2.4. Theoretical calculations
2.3. X-raycrystal studies of 4, 5 and 3
All calculations were carried out using the Gaussian03
[17] program package. The conformational space of com-
pounds 3, 4, 6, and 7 was explored through optimization of
all the possible starting geometries, which were optimized at
the B3LYP level with the 6-31G* basis set.
The crystals of C23H34O5 (3), C23H34O5 (4), and
C25H38O5S (5) were obtained by recrystallization from
methanol as colorless prisms. A summary of the crystal
data, data collection, and structure refinement is presented
in Table 1.
The intensity data were collected with a CAD4 diffrac-
tometer with graphite monochromated Mo K␣ radiation (λ
0.71073 A). The cell parameters were determined and refined
3. Results and discussion
˚
by least-squares fit of 20 high angle reflections. The structures
were solved by direct methods [15] and conventional Fourier
synthesis [16]. Refinement of the structures was made by
full matrix least-squares on F2. All non-H-atoms were re-
fined anisotropically. In compound 4, the H-atom positions
were detected in a difference Fourier synthesis and refined
with isotropic thermal factors, while in compounds 3 and 5,
some of the H-atoms were introduced at calculated positions
in their described geometries and allowed to ride on the at-
tached carbon atom with fixed isotropic thermal parameters
(1.2Ueq of the parent carbon atom). Selected bond distances
3.1. Chemistry
Oneofthekeystepsinthesynthesisofdesogestrel(1)isin-
troduction of the 11-methylene moiety, which can be attained
by oxidation of an 11-hydroxyl function [6,11,12], followed
either by conversion into the 11-methyl-11-hydroxyl deriva-
tive, which is then submitted to dehydration [6] or by a Wittig
procedure [11,12]. When 11-hydroxylsterols, such as 2, are
the starting compounds, it seems convenient to protect the 3-
and 17-keto groups before the subsequent oxidative step to
the 11-ketone [7,12,13].

E. Modica et al. / Steroids 70 (2005) 660–666
663
Table 2
Selected bond lengths (A) and angles ( ) for compounds 3, 4, and 5
Small amounts of compound 3 were detected in the ke-
talization reaction mixture. An analytical sample was ob-
tained by HPLC purification followed by crystallization. The
X-ray diffraction study was also performed in this case, al-
lowing us to confirm that compound 3 had the natural 10␤-
configuration.
The unusual stereochemistry of the A/B ring fusion of
the 11-hydroxyl compound 4 stimulated our interest in the
synthesis of the 10␣-analogue of desogestrel, with respect to
better understand its biological role. To this aim, the inter-
mediate 4 was submitted to oxidation using the mild Swern
conditions, which have often been used in the steroid field
[10,11] in order to obtain keto-compounds in high yields.
However, after purification, rather than the desired 11-ketone,
the reaction mixture afforded the thio-compound 5 as main
product. This is a common side reaction product, which is of-
ten recovered together with the keto-derivative during Swern
oxidations [14]. The structure of 5 has been elucidated using
both NMR spectroscopy and X-ray diffraction.
◦
˚
Compound 3
Compound 4
Compound 5
O18–C18
O18–C3
O19–C19
O19–C3
O11–C11
O11–C24
O20–C20
O20–C17
O21–C21
O21–C17
C1–C10
C3–C4
1.383(9)
1.436(7)
1.377(9)
1.407(8)
1.439(8)
1.427(4)
1.425(4)
1.424(4)
1.434(4)
1.432(3)
1.394(11)
1.438(8)
1.412(10)
1.430(8)
1.457(7)
1.405(11)
1.350(10)
1.427(8)
1.406(9)
1.421(8)
1.389(10)
1.439(9)
1.397(9)
1.425(8)
1.544(10)
1.509(12)
1.482(9)
1.313(10)
1.477(10)
1.520(9)
1.557(9)
1.523(9)
1.551(12)
1.466(12)
1.473(12)
1.415(5)
1.432(4)
1.378(5)
1.426(4)
1.537(4)
1.519(4)
1.506(4)
1.325(5)
1.501(4)
1.530(4)
1.534(4)
1.534(4)
1.537(5)
1.496(5)
1.480(6)
1.541(10)
1.494(12)
1.493(10)
1.328(10)
1.489(11)
1.521(10)
1.543(9)
1.535(10)
1.527(11)
1.466(12)
1.481(13)
1.774(12)
1.830(13)
C4–C5
C5–C6
C6–C7
C7–C8
C8–C9
C9–C10
C16–C17
C19–C18
C21–C20
S1–C25
S1–C24
3.2. Crystallographic studies
The stereochemistry of compounds 3, 4, and 5 was as-
signed by X-ray crystallography. Their molecular structures
are reported in Fig. 1 as ORTEP [18] views with the rela-
tive atom numbering scheme. Compound 3 had the natural
configuration at all the stereogenic centers, unchanged with
respect to its synthetic precursor 2, while 4 and 5 showed the
inverted 10S configuration at C10 so that they are classified
as 10␣-gonenes.
The main difference among the three compounds was ob-
served in ring A, as a consequence of the different configu-
ration at C10; in fact, this ring had the 3␣,10␤ chair confor-
mation in 3 and the inverted 3␤,10␣ chair conformation in 4
and 5.
C18–O18–C3
C19–O19–C3
C20–O20–C17
C21–O21–C17
O19–C3–O18
O19–C3–C2
O18–C3–C2
O19–C3–C4
O18–C3–C4
O11–C11–C12
O11–C11–C9
C24–O11–C11
O21–C17–O20
O21–C17–C13
O20–C17–C13
O21–C17–C16
O20–C17–C16
O19–C19–C18
O18–C18–C19
O21–C21–C20
O20–C20–C21
C25–S1–C24
O11–C24–S1
108.8(5)
108.9(5)
109.2(6)
107.0(5)
105.2(5)
110.9(6)
110.7(6)
111.1(7)
109.6(6)
107.1(5)
112.9(5)
107.2(2)
108.0(2)
108.0(3)
109.5(3)
106.3(2)
108.7(3)
110.8(3)
110.6(2)
108.5(3)
111.5(2)
111.6(2)
108.1(6)
107.9(5)
110.2(6)
109.0(6)
105.2(5)
109.5(6)
109.9(6)
110.8(6)
108.7(6)
109.6(5)
105.3(5)
116.0(6)
105.7(5)
116.2(6)
110.3(5)
110.7(6)
109.0(7)
107.3(7)
106.0(8)
104.3(7)
107.4(7)
99.0(5)
105.9(6)
113.8(5)
109.9(6)
112.9(7)
108.5(6)
108.2(7)
104.9(6)
106.1(6)
105.7(7)
105.4(2)
113.8(3)
111.6(2)
112.6(3)
109.1(3)
102.5(2)
102.8(3)
106.6(3)
103.1(3)
The other three rings of the steroidal skeleton adopted very
similar conformations in the examined compounds. In fact,
ring B always exhibited a twisted 8␤ envelope conformation
with small differences in the distance of C8 from the best
˚
plane of the remaining five atoms, which was 0.607(3) A in
˚
˚
4, 0.624(4) A in 5, and 0.655(6) A in 3, where the twisting
was more pronounced.
114.8(6)
Ring C had a 8␤,12␣ chair conformation in the three com-
pounds, while the five-membered ring D adopted a 13␤,14␣
half-chair conformation (the distances of C13 from the plane
According to Gao et al. [6], the protection of 13␤-ethyl-
11␣-hydroxy-gon-4-ene-3,17-dione (2) as a diketal should
easily lead to 3 in high yields. In our hands, when compound
2 was treated with ethylene glycol, triethylorthoformate and
p-toluene sulfonic acid, a predominant product could be re-
covered from the reaction mixture by crystallization. How-
ever, its chemical–physical properties and NMR data were
not in agreement with those reported in the literature [6];
thus we decided to carry out a X-ray diffraction study, which
showed that an inversion of configuration had occurred at
C-10, and the 10␣-diketal 4 had been formed instead of the
foreseen10␤-diketal 3.
˚
of the four other atoms were, respectively, 0.723(7) A in 3,
˚
˚
0.704(3) A in 4, and 0.698(4) A in 5).
The dioxolane rings at the two opposite sides of the molec-
ular skeleton assumed different conformations ranging from
an envelope in 4 to more or less twisted conformations in
3 and in 5. In Table 3 the significant torsion angles that de-
fine the differences in the relative conformations of the three
compounds are reported.
The orientation of O11–H in 3 and the crystal pack-
ing determined the sequence of donor and acceptor groups
involved in the pattern O11–H → O18ꢀ (at 1 − x, 0.5 + y,

664
E. Modica et al. / Steroids 70 (2005) 660–666
Fig. 1. ORTEP Stereo views of compounds 3, 4, and 5 drawn at 50% probability level (the H atoms are drawn as spheres of arbitrary radii).
Table 3
formation of molecular chains in the b-axis direction as
Selected torsion angles (◦) for compounds 3, 4, and 5
shown in Fig. 3. Similar interactions commonly referred to
as ␲-hydrogen bonds [19] have also been found in the crys-
tal structure of donazole [20] with analogous values of bond
distances between the hydroxyl H and the alkyne group of
adjacent molecules. In 5, the CH2SCH3 moiety replaced the
hydroxyl hydrogen preventing molecular chain formation.
A superimposition of 3 and 4 through the best fit of rings
B, C, and D (Fig. 4) clearly showed the effects of the C10
configurational change and the different orientation of the
hydrogen atom of the hydroxyl group (O11–H), which in
3 and 4 was involved in different intermolecular hydrogen
bonds.
Compound 3
Compound 4
Compound 5
C1–C10–C9–C8
−152.4(6)
83.7(7)
−53.8(7)
−158.0(5)
68.6(8)
84.5(3)
−42.4(3)
−48.3(3)
−42.0(3)
174.1(2)
−176.3(3)
31.7(3)
−69.7(3)
−36.9(4)
142.1(3)
27.1(4)
−162.5(3)
−120.4(4)
−25.4(5)
−36.2(3)
79.5(7)
−49.5(8)
−60.8(7)
−59.0(6)
173.4(6)
−175.3(6)
20.3(9)
−71.6(7)
−38.5(8)
147.7(6)
−11.0(1)
−166.1(6)
−110.9(7)
18.0(1)
C1–C10–C9–C11
O11–C11–C9–C10
H10–C10–C9–H9
O19–C3–C4–C5
C1–C2–C3–O19
C3–O18–C18–C19
C5–C4–C3–O18
C12–C13–C17–O20
C15–C16–C17–O21
C17–O20–C20–C21
C14–C13–C22–C23
C13–C17–O21–C21
O18–C18–C19–O19
O20–C20–C21–O21
−67.9(8)
17.7(8)
−175.5(6)
−38.8(8)
144.6(6)
7.3(9)
−165.6(6)
141.8(6)
−8.0(9)
−20.3(9)
3.3. Modeling
−8.3(9)
In order to explain the configurational instability at C10,
we undertook a theoretical study of compounds 3 and 4 to
ascertain their relative stability.
The allowed conformations of 3 and 4 were determined
within the DFT framework at the B3LYP/6-31G* level as
implemented in Gaussian03 [17]. The main degrees of con-
formational freedom resided in the dioxolane rings, which
could assume two different twist conformations, and in the
hydroxyl group. The preferred conformations of the two
compounds determined through the calculations were very
0.5 − z) of the dioxolane ring of the adjacent molecule
with formation of molecular chains developing in the b-axis
direction via intermolecular hydrogen bonds (O11· · ·O18ꢀ
ꢀ
ꢀ
◦
˚
2.878(6); O11–H· · ·O18 1.78(9) A; O11–H· · ·O18 163(7) )
as shown in Fig. 2. In 4, these interactions were replaced
by short intermolecular contacts between the hydroxyl H
atom and the C5 C6 double bond of the adjacent molecule
(O11· · ·C5ꢀ 3.195(3); O11· · ·C6ꢀ 3.213(4); O11–H· · ·C5ꢀ
ꢀ
˚
2.51(5); O11–H· · ·C6 2.44(5) A; at 1 − x, 0.5 + y, 1 − z) with

E. Modica et al. / Steroids 70 (2005) 660–666
665
Fig. 2. Crystal packing of 3 showing the molecular chains running along the b-axis direction and the intermolecular hydrogen bond connections (only the
hydroxyl hydrogen is reported for sake of clarity).
Fig. 3. Crystal packing of 4 showing the molecular chains running along the b-axis direction and the intermolecular ␲-hydrogen bond connections (only the
hydroxyl hydrogen is reported for sake of clarity).
Fig. 4. Superimposition of the crystal structures of 3 (grey) and 4 (black).

666
E. Modica et al. / Steroids 70 (2005) 660–666
close to those determined through the X-ray analysis, the
main difference being the conformation of the dioxolane
ring at C17; however, for both compounds the calculated
conformations corresponding to the X-ray geometries re-
sulted higher in energy than the global minima by only
0.1 kcal/mol.
[2] Modica E, Colombo D, Compostella F, Scala A, Ronchetti F. Reduc-
tion of aromatic steroidal A rings by lithium in ethyl amine. Steroids
2002;67:145–50.
[3] Weijers MJ. Clinical trial of an oral contraceptive containing deso-
gestrel and ethynyl estradiol. Clin Ther 1982;4:359–66.
[4] Corson SL. Contraceptive efficacy of a monophasic oral contracep-
tive containing desogestrel. Am J Obstet Gynecol 1993;168:1017–20.
[5] Viinikka L, Ylikorkala O, Nummi S, Virkkunen P, Ranta T, Alapiessa
U, et al. Biological effects of a new potent progestagen. A clinical
study. Acta Endocr 1976;83:429–38.
[6] Gao H, Su X, Zhou L, Li Z. Synthesis of 13-ethyl-11-
methylene-18,19-dinor-17␣-pregn-4-en-20-yn-17-ol. Org Prep Proc
Int 1997;29:572–6.
[7] Gao H, Su X, Huang L, Li Z. An improved synthesis of 13-ethyl-
11-methylengon-4-en-3,17-dione. Synth Commun 1997;27:1981–7.
[8] Modica E, Compostella F, Colombo D, Ronchetti F, Scala A, Bovio
B. Crystal structure of steroidal monoenes obtained by reduction of
the aromatic A ring of 13-ethyl-3-ethoxy-gona-1,3,5(10)-triene-11␣-
17␤-diol. Steroids 2003;68:367–72.
[9] Omura K, Swern D. Oxidation of alcohols by “activated” dimethyl
sulfoxide. A preparative steric and mechanistic study. Tetrahedron
1978;34:1651–60.
[10] Marx M, Tidwell TT. Reactivity-selectivity in the Swern oxidation
of alcohols using dimethyl sulfoxide-oxalyl chloride. J Org Chem
1984;49:788–93.
[11] Schwarz S, Ring S, Weber G, Teichmuller G, Palme HJ, Pfeiffer C, et
al. Synthesis of 13-Ethyl-11-methylene-18,19-dinor-17␣-pregn-4-en-
20-yn-ol (Desogestrel) and its main metabolite 3-Oxo Desogestrel.
Tetrahedron 1994;50:10709–20.
[12] Van den Heuvel MJ, Van Bokhoven CW, De Jongh HP, Zeelen FJ. A
partial synthesis of 13-ethyl-11-methylene-18,19-dinor-17␣-pregn-4-
en-20-yn-17-ol (desogestrel) based upon intramolecular oxidation of
an 11␤-hydroxyl-19-norsteroid to the 18→11␤-lactone. Recl Trav
Chim Pays-Bas 1988;107:331–4.
It is worthy to note that the energies of the preferred con-
formations of 3 and 4 were comparable, the former being
more stable than the latter by only 0.2 kcal/mol. The stabil-
ity of 4, which was close to that of 3 in spite of the “un-
natural” junction between the A and B rings, might seem
surprising. Actually, it is known [11] that the presence of
the 11␣-hydroxyl group gives rise to an intrinsic instability
at C10, which may lead to a partial inversion of this cen-
ter whenever 10␤-H gets the opportunity for abstraction. If
this holds true, the 11-deoxy-derivatives of 3 and 4 should
differ in their stability, such that the compound with the “nat-
ural” skeleton is preferred over the other one. In fact, when
the preferred conformations and the relative energies of the
11-deoxy-derivatives 6 and 7 were determined, compound
7 was less stable than 6 by 1.60 kcal/mol. The reason why
compound 4 was as stable as 3 is due to the presence of the
11␣-hydroxyl group, which destabilized 3 through a steric
˚
interaction with the 1␤-hydrogen atom (2.28 and 2.42 A in
the calculated and in the X-ray geometry, respectively). Con-
versely, in 4, the 11␣-hydroxyl group was at a greater dis-
˚
tance from the 1␣ or 1␤ hydrogen atoms (2.73 and 3.54 A
˚
in the calculated and 2.61 and 3.53 A in the X-ray geome-
try, respectively). The instability produced by the interaction
present in 3 compromised the higher stability of its steroidal
skeletonmakingitenergeticallycomparableto4. Thus, acon-
figurational inversion at C10 became possible on energetic
grounds.
[13] Van den Broek AJ, Van Bokhoven C, Hobbelen PMJ, Leemhuis J.
11-Alkylidene steroids in the 19-nor series. Recl Trav Chim Pays-
Bas 1975;94:35–9.
[14] Albright JD, Goldman L. Dimethyl sulfoxide-acid anhydride
mixtures for the oxidation of alcohols.
1967;89:2416–23.
J Am Chem Soc
[15] Altomare A, Burla MC, Camalli M, Cascarano G, Giacovazzo C,
Guagliardi A, et al. SIR92 - a program for automatic solution of crys-
tal structures by direct methods. J Appl Crystallogr 1994;27:435–6.
[16] Sheldrick GM. SHELX-97. University of Go¨ttingen, Germany.
[17] Gaussian 03, Revision B.02, Frisch MJ, Trucks GW, Schlegel HB,
Scuseria GE, Robb MA, Cheeseman JR, et al., Pittsburgh, PA: Gaus-
sian Inc., 2003.
Acknowledgments
This work was supported by the Ministero dell’Istruzione,
`
dell’Universita e della Ricerca (MIUR, FIRST ex-60%) and
by the Universities of Milan and Pavia.
[18] Johnson CK. ORTEP 11, Report ORNL-5138. Oak Ridge National
Laboratory, TN; 1976.
[19] Atwood JL, Hamada F, Robinson KD, Orr GW, Vincent RL. X-
ray diffraction evidence for aromatic ␲ hydrogen bonding to water.
Nature 1991;349:683–4.
References
[1] Compostella F, Colombo D, Modica E, Scala A, Toma L, Bovio B,
et al. Base promoted oxidation of 13␤-ethyl-11-methylenegon-4-en-
17-one. Steroids 2002;67:111–7.
[20] Viswamitra MA, Radhakrishnan R, Bandekar J, Desiraju GR. Evi-
dence for O H. . .C and N H. . .C hydrogen bonding in crystalline
alkynes, alkenes, and aromatics. J Am Chem Soc 1993;115:4868–9.