Unusual electrochemical oxidation of cholesterol

Article abstract of DOI:10.1016/j.steroids.2008.01.014

It has been found that cholesterol undergoes direct electrochemical oxidation on platinum electrode in dichloromethane. Voltammetric measurements show that the process is controlled by the rate of electron transfer and the height of the oxidation peak is linear vs. concentration of cholesterol. Preparative electrolysis with separated cathodic and anodic compartments afforded dicholesteryl ether in a relatively high material yield. Depending on electrolysis conditions (composition of supporting electrolyte and electrolytic cell construction) various by-products with a 3β-chloro, 3β-acetoxy, or 3β-acetylamino group were obtained.

Full text of DOI:10.1016/j.steroids.2008.01.014

steroids 7 3 ( 2 0 0 8 ) 543–548
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/steroids
Unusual electrochemical oxidation of cholesterol
Jan Kowalskia, Zenon Łotowskib, Jacek W. Morzyckib,∗, Jolanta Płoszyn´ skaa,
Andrzej Sobkowiaka,∗∗, Agnieszka Z. Wilczewskab
a
´
´
Faculty of Chemistry, Rzeszow University of Technology, P.O. Box 85, 35-959 Rzeszow, Poland
b
Institute of Chemistry, University of Białystok, Piłsudskiego 11/4, 15-443 Białystok, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
It has been found that cholesterol undergoes direct electrochemical oxidation on platinum
electrode in dichloromethane. Voltammetric measurements show that the process is con-
trolled by the rate of electron transfer and the height of the oxidation peak is linear vs.
concentration of cholesterol. Preparative electrolysis with separated cathodic and anodic
compartments afforded dicholesteryl ether in a relatively high material yield. Depending
on electrolysis conditions (composition of supporting electrolyte and electrolytic cell con-
struction) various by-products with a 3␤-chloro, 3␤-acetoxy, or 3␤-acetylamino group were
obtained.
Received 19 October 2007
Received in revised form
28 December 2007
Accepted 9 January 2008
Published on line 20 January 2008
Keywords:
© 2008 Elsevier Inc. All rights reserved.
Cholesterol
Electrochemical oxidation
Dicholesteryl ether
Steroid
1.
Introduction
oxidation sites of cholesterol are the C5–C6 double bond,
the allylic C-7 position, and the 3␤-OH group. The sys-
Cholesterol oxidation products (COPs) have been recently
widely studied since they are involved in the pathogenesis
of atherosclerosis [1]. Several COPs show various undesirable
biological activities [2,3]. Due to the presence of the dou-
ble bond cholesterol is susceptible to autooxidation in the
presence of dioxygen, light, and metal ions via free radical
reactions [4]. COPs are also formed during food process-
ing and therefore numerous methods for COPs analysis in
food were elaborated [5,6]. A number of papers dealing with
electrochemical oxidation of cholesterol has also appeared.
Most reports concern indirect oxidation with mediators, such
as Mn(III)/IV–porphyrin–O2, Ti(II)/(III)–hematoporphyrin–O2,
or Br−/BrO− [7–9]. The first two systems caused the selec-
tive radical oxidation of the cholesterol side chain at C-25
to the corresponding tertiary alcohol. However, the preferred
tem with hypobromide mediation afforded several products
oxidized at the double bond (cholestan-3␤,5␰,6␰-triol, 5␰,6␰-
epoxycholestan-3␤-ol) or at C-7 (7␰-hydroxycholesterol and
7-oxocholesterol). Recently described electrochemical system
for oxidation of cholesterol acetate induced by dioxygen and
iron picolinate complexes afforded mostly 7-hydroxylated
products [10]. Direct electrochemical oxidation of cholesterol
on platinum electrode in glacial acetic acid led to formation of
7␣-acetoxycholesterol and 7␤-acetoxycholesterol in the ratio
10:3 [11]. The previous studies on cholesterol electrooxidation
were carried out in polar nucleophilic solvents. The proposed
mechanism usually includes formation of a cation-radical in
the first step. However, it is known that cation-radicals are
more stable in dichloromethane, which do not exhibit nucle-
ophilic properties, than in other solvents usually employed in
∗
∗∗
Corresponding author. Tel.: +48 85 7457585; fax: +48 85 7457581.
Corresponding author. Tel.: +48 17 8651656; fax: +48 17 8549830.
E-mail addresses: morzycki@uwb.edu.pl (J.W. Morzycki), asobkow@prz.rzeszow.pl (A. Sobkowiak).
0039-128X/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2008.01.014

544
steroids 7 3 ( 2 0 0 8 ) 543–548
electrochemistry [12]. Therefore, this solvent was used in a
present study on direct electrochemical oxidation of choles-
terol.
[14]. Cholesterol and cholesta-3,5-diene were purchased from
Sigma and was used without further purification. Other sol-
vents were purchased from POCh (Gliwice, Poland), dried and
distilled before use.
2.
Experimental
2.3.
Electrooxidation of cholesterol—procedures and
analysis of products
2.1.
Equipment
In all electrochemical measurements, the supporting elec-
trolyte was the solution containing 0.1 M tetrabutylammo-
nium tetrafluoroborate in dichloromethane.
Cyclic-voltammograms were recorded with iR compensation
at 25 ◦C using a three-electrode potentiostat (Bioanalytical
Systems Model CV-50W). The experiments were conducted
in a 15-mL electrochemical cell with an argon-purge system.
The working electrode was a Bioanalytical Systems platinum
inlay (area, 0.03 cm2), the auxiliary electrode a platinum mesh
(contained in a glass tube with a medium porosity glass-frit)
and the reference electrode Ag/0.1 M AgNO3 in acetonitrile.
The latter was contained in a Pyrex tube with a cracked soft-
glass tip, which was placed inside a Luggin capillary. Before
each experiment the working electrode was polished using a
Buehler Micropolish Alumina Gamma 3B and Buehler Micro-
cloth polishing cloth and rinsed with glacial acetic acid. The
accuracy of potential measurement was equal to 10 mV. All
reported values of potentials and currents are means of three
independent measurements.
The preparative electrolyses were performed with a poten-
tiostat (Princeton Applied Research Model 273 A) by passing
a constant current (10 mA) during 3 h. The working electrode
was a platinum plate (area, 7.5 cm2), the auxiliary electrode a
platinum mesh, and the same reference electrode as described
above. The low current density applied equal to 1.23 mA cm−2
prevented the decomposition of supporting electrolyte and
also the occurrence of consecutive reactions. The electroly-
ses were carried out either in an electrolytic cell, in which the
cathode was placed in a glass tube with a medium porosity
glass frit (G3) or anode and cathode compartments were con-
nected with an electrolytic bridge filled with 0.4 M solution
of tetrabutylammonium tetrafluoroborate (TBABF4) in ace-
tonitrile/dichloromethane (50:50, v/v). All measurements were
performed at room temperature. High-purity argon gas was
used to deaerate the solutions applied in all electrochemical
measurements.
Electrochemical oxidation of cholesterol (3× 200 mg) was
initially carried out in an electrolytic cell, in which the cathode
was placed in a glass tube with a medium porosity glass frit,
with TBABF4 in dichloromethane at 0 ◦C for 3 h. The progress of
reaction was monitored by TLC on silica gel plates developed
with hexane–ethyl acetate 8:2. TLC analysis showed forma-
tion of a complex mixture of compounds. The dry residue
obtained by evaporation of the solvent in vacuo from the
combined reaction mixtures was subjected to the silica gel
column chromatography. Five major fractions were collected;
one of them being the unreacted cholesterol 1 (200 mg), eluted
with hexane–ethyl acetate 95:5 (fraction 4). Other fractions
were subjected to further chromatographic separation. Frac-
tion 1 contained a mixture of nonpolar products (23 mg),
mostly isomeric dienes, which could not be separated, and
a trace amount of dicholesteryl ether (6). Repeated chro-
matography of fraction 2 (elution with hexane–ethyl acetate
98:2) afforded 5␤,6␣-dichlorocholestan-3␤-ol (2; 30 mg, yield
9%). From fraction 3 (eluted with hexane–ethyl acetate 97:3)
5␣,6␤-dichlorocholestan-3␤-ol (3; 50 mg, yield 14%) was iso-
lated [15,16]. Further chromatographic purification of fraction
5 (elution with hexane–ethyl acetate 90:10) afforded 6␣-
chlorocholestane-3␤,5␤-diol (4; 20 mg, yield 6%) [17].
In the case of prolonged cholesterol electrooxidation, 3␤-
chlorocholest-5-ene (5) was additionally isolated. The reaction
carried out in dichloromethane with TBABF4 at 0 ◦C for 7 h
afforded 4% of compound 5 (eluted with hexane), in addition
to other products.
• 5␤,6␣-Dichlorocholestan-3␤-ol (2):
m.p. 137–138 ◦C (methanol). 1H NMR ı (ppm): 4.49 (dd,
J = 12.4, 5.3, 1H, 6␤-H); 4.15 (m, 1H, 3␣-H); 1.15 (s, 3H, 19-H);
0.90 (d, J = 6.6 Hz, 3H, 21-H), 0.87 (d, J = 6.7 Hz, 6H, 26,27-H);
0.66 (s, 3H, 18-H). 13C NMR ı (ppm): 83.0 (C), 67.9 (CH), 66.0
(CH), 56.0 (CH), 56.0 (CH), 44.4 (C), 42.5 (C), 42.4 (CH), 40.2
(CH2), 39.6 (CH2), 39.4 (CH2), 36.1 (CH2), 35.7 (CH), 35.1 (CH),
34.1 (CH2), 28.1 (CH2), 28.0 (CH), 27.0 (CH2), 26.7 (CH2), 24.0
(CH2), 23.8 (CH2), 22.8 (CH3), 22.5 (CH3), 22.0 (CH2), 19.1 (CH3),
18.6 (CH3), 11.8 (CH3). IR ꢀ (cm−1): 1467, 1283, 1086, 1061,
1023, 1000, 964. MS m/z (%): 458 (16), 456 (23, M+), 423 (38),
421 (100), 386 (24), 384 (62), 368 (18), 366 (26).
NMR spectra were measured at 200 MHz with a Brucker AC
200F spectrometer using CDCl3 solutions with TMS as internal
standard. Column chromatography was performed on Merck
silica gel (70–230 mesh). Merck Silica Gel 60, F 256 TLC alu-
minum sheets were applied for thin layer chromatographic
analysis. For visualization of products, a 5% solution of phos-
phomolybdic acid in ethanol [13] was used. IR spectra were
recorded on a Nicolet series II Magna-IR 550 FT-IR spectrom-
eter as chloroform solutions. Mass spectra were obtained at
70 eV with an ADM-604 spectrometer. Melting points were
determined on a Kofler apparatus of the Boetius type.
• 5␣,6␤-Dichlorocholestan-3␤-ol (3):
m.p. 135–139 ◦C (methanol), literature [16] m.p. 143–144 ◦C
(ethanol). 1H NMR ı (ppm): 4.35 (dd, J = 4.2, 1.8 Hz, 1H, 6␣-H);
4.31 (m, 1H, 3␣-H); 2.48 (dd, J = 10.4, 3.3 Hz, 1H, 4␣-H), 2.33
(m, 1H, 7␤-H); 1.37 (s, 3H, 19-H); 0.90 (d, J = 7.9 Hz, 3H, 21-H),
0.87 (d, J = 6.6 Hz, 6H, 26,27-H); 0.71 (s, 3H, 18-H). 13C NMR
ı (ppm): 85.4 (C), 67.9 (CH), 63.8 (CH), 56.1 (CH), 55.2 (CH),
45.9 (CH), 43.0 (CH2), 42.7 (C), 40.7 (C), 39.7 (CH2), 39.5 (CH2),
2.2.
Chemicals and reagents
The reagents for the investigations and syntheses were the
commercially available highest purity. The solvent for all
electrochemical experiments was dichloromethane (Aldrich),
which was purified by the procedure described elsewhere

steroids 7 3 ( 2 0 0 8 ) 543–548
545
36.1 (CH2), 35.7 (CH), 35.4 (CH2), 34.5 (CH2), 30.3 (CH), 30.3
(CH2), 28.2 (CH2), 28.0 (CH), 24.0 (CH2), 23.8 (CH2), 22.8 (CH3),
22.5 (CH3), 21.2 (CH2), 19.7 (CH3), 18.6 (CH3), 12.2 (CH3). IR ꢀ
(cm−1): 1467, 1382, 1154, 1085, 1023, 1000, 964. MS m/z (%):
458 (12), 456 (20, M+), 386 (28), 384 (25), 368 (21), 301 (21).
• 6␣-Chlorocholestane-3␤,5␤-diol (4):
6-H); 3.29 (m, 1H, 3␣-H); 1.00 (s, 3H, 19-H); 0.92 (d, J = 6.5 Hz,
3H, 21-H), 0.87 (d, J = 6.6 Hz, 6H, 26,27-H); 0.68 (s, 3H, 18-H).
13C NMR ı (ppm): 141,4 (C), 121.53 (CH), 76.4 (CH), 56.8 (CH),
56.2 (CH), 50.3 (CH), 42.3 (C), 40.6 (CH2), 39.8 (CH2), 39.5 (CH2),
37.4 (CH2), 36.9 (C), 36.2 (CH2), 35.8 (CH); 32.0 (CH2), 31.9 (CH),
29.4 (CH2), 28.2 (CH2), 28.0 (CH), 24.3 (CH2), 23.8 (CH2), 22.8
(CH3), 22.6 (CH3), 21.1 (CH2), 19.4 (CH3), 18.7 (CH3), 11.9 (CH3).
IR ꢀ (cm−1): 1467, 1381, 1366, 1079. MS m/z (%): 754 (2, M+),
739 (1) 384 (31), 369 (100), 368 (70), 355 (11). Elemental anal-
ysis calcd for C54H90O: C, 85.87; H, 12.01. Found: C, 85.60; H,
11.93.
m.p. 139–142 ◦C (dichloromethane), literature [17] m.p.
140–142.5 ◦C. 1H NMR ı (ppm): 4.31 (m, 1H, 6␤-H); 4.17 (m,
1H, 3␣-H); 0.99 (s, 3H, 19-H); 0.90 (d, J = 6.9 Hz, 3H, 21-H), 0.87
(d, J = 6.8 Hz, 6H, 26,27-H); 0.65 (s, 3H, 18-H). 13C NMR ı (ppm):
77.2 (C), 67.8 (CH), 67.3 (CH), 56.1 (CH), 56.1 (CH), 42.8 (C), 42.8
(CH), 42.7 (C), 39.7 (CH2), 39.5 (CH2), 38.4 (CH2), 36.1 (CH2),
35.7 (CH), 35.7 (CH); 31.5 (CH2), 28.1 (CH2), 28.0 (CH), 27.6
(CH2), 25.7 (CH2), 24.0 (CH2), 23.8 (CH2), 22.8 (CH3), 22.6 (CH3),
21.4 (CH2), 18.6 (CH3), 16.7 (CH3), 12.0 (CH3). IR ꢀ (cm−1): 1467,
1264, 1094. MS m/z (%): 440 (2), 438 (6, M+), 402 (24), 384 (29),
368 (36), 366 (100), 351 (4).
• Cholest-5-en-3␤-yl acetate (cholesteryl acetate; 7):
1H NMR ı (ppm): 5.38 (m, 1H, 6-H); 4.61 (m, 1H, 3(-H); 2.04 (s,
3H, Ac), 1.03 (s, 3H, 19-H); 0.94 (d, J = 6.5 Hz, 3H, 21-H), 0.87
(d, J = 6.5 Hz, 6H, 26, 27-H); 0.68 (s, 3H, 18-H).
• N-Acetyl-cholest-5-en-3␤-amine (8):
m.p. 238–240 ◦C (dichloromethane), literature [19] m.p.
239–241 ◦C (methanol). 1H NMR ı (ppm): 5.37 (m, 1H, 6-H);
5.29 (bd, J = 7.4Hz, 1H, NH), 3.70 (m, 1H, 3␣-H); 1.98 (s, 3H,
Ac), 1.00 (s, 3H, 19-H); 0.92 (d, J = 6.5 Hz, 3H, 21-H), 0.87 (d,
J = 6.5 Hz, 6H, 26,27-H); 0.68 (s, 3H, 18-H). 13C NMR ı (ppm):
169.1 (C), 140.1 (C), 122.0 (CH), 56.7 (CH), 56.1 (CH), 50.1 (CH),
49.7 (CH), 42.3 (C), 39.7 (CH2), 39.5 (CH2), 39.3 (CH2), 37.8
(CH2), 36.5 (C), 36.2 (CH2), 35.8 (CH); 31.8 (CH2), 31.8 (CH), 29.2
(CH2), 28.2 (CH2), 28.0 (CH), 24.3 (CH2), 23.8 (CH2), 23.6 (CH3),
22.8 (CH3), 22.5 (CH3), 21.0 (CH2), 19.3 (CH3), 18.7 (CH3), 11.8
(CH3). IR ꢀ (cm−1): 1637, 1560, 1466, 1383, 1368, 1112, 1040.
MS m/z (%): 384 (1), 369 (31), 368 (100); MS ESI m/z: 450.4 (100,
MNa+).
• 3␤-Chlorocholest-5-ene (5):
m.p. 89–91 ◦C (dichloromethane), literature (Steraloids cat-
alogue) m.p. 95–96 ◦C. 1H NMR ı (ppm): 5.38 (m, 1H, 6-H);
3.78 (m, 1H, 3␣-H); 1.04 (s, 3H, 19-H); 0.92 (d, J = 6.5 Hz, 3H,
21-H), 0.87 (d, J = 7.5Hz, 6H, 26,27-H); 0.69 (s, 3H, 18-H). 13C
NMR ı (ppm): 140.8 (C), 122.5 (CH), 60.3 (CH), 56.7 (CH), 56.1
(CH), 50.1 (CH), 43.4 (CH2), 42.3 (C), 39.7 (CH2), 39.5 (CH2), 39.1
(CH2), 36.4 (C), 36.2 (CH2), 35.8 (CH); 31.8 (CH2), 31.8 (CH), 31.1
(CH2), 28.2 (CH2), 28.0 (CH), 24.3 (CH2), 23.8 (CH2), 22.8 (CH3),
22.6 (CH3), 21.0 (CH2), 19.2 (CH3), 18.7 (CH3), 11.8 (CH3). IR ꢀ
(cm−1): 1467, 1377, 869. MS m/z (%): 406 (35), 404 (100, M+),
389 (42), 368 (14), 353 (9).
In the typical preparative electrolysis with separated
electrodes, 15 cm3 of the supporting electrolyte (0.1 M
tetrabutylammonium tetrafluoroborate in dichloromethane)
containing 900 mg (2.33 mmol) of cholesterol was placed
into the anodic compartment of electrolyser. The cathodic
compartment contained the supporting electrolyte with
approximately 2% of acetic acid. Anode and cathode com-
partments were connected with an electrolytic bridge filled
with 0.4 M solution of TBABF4 in 50:50 (v/v) acetonitrile and
dichloromethane. The total volume of the solution in the
electrolytic bridge was equal to 4 cm2. The approximate con-
centration (due to a volume contraction effect) of acetonitrile
in the bridge was about 9 M.
The dry residue obtained by evaporation in vacuo of the
solvent from the crude reaction mixture was subjected to sil-
ica gel column chromatography. The reaction products were
eluted consecutively with hexane–ethyl acetate mixtures.
The non-polar 3␤-chlorocholest-5-ene (5; 25 mg, yield 3%)
was eluted with hexane. Further elution with hexane–ethyl
acetate (1–5%) afforded dicholesteryl ether (6; 247 mg, yield
28%) followed by cholesterol acetate (7; 35 mg, yield 4%)
and an unidentified cholestenone (M 384; 7 mg). The starting
cholesterol was then eluted with 5–6% ethyl acetate–hexane
(257 mg, 29%). The most polar product, N-acetyl-cholest-5-en-
3␤-amine (8; 38 mg, yield 4%), was eluted with hexane–ethyl
acetate (7:3).
3.
Results and discussion
3.1.
Voltammetric measurements
Cyclic-voltammogram of cholesterol in dichloromethane
(Fig. 1, curve a) shows in the first anodic scan the broad
anodic oxidation peak in the region from +1.55 to +1.85 V
(vs. Ag/0.1 M AgNO3 in acetonitrile) and at scan rate 0.1 V s−1
the peak potential was estimated as +1.75 0.02 V. In the
subsequent cathodic scan three reduction peaks appear at
+0.50, −0.25, and −0.53 V. In the second anodic scan an oxida-
tion peak at +0.15 V in addition to cholesterol oxidation peak
(+1.75 V) is present. The height of the peak of cholesterol oxi-
dation depends linearly on cholesterol concentration within
the investigated range 0.1–5 mM. This peak is rather broad and
therefore a half-peak potential (Ep/2) was used to analyze its
dependence on logarithm of scan rate. The plot is presented in
Fig. 2. The half peak potential is shifted towards more positive
values during the increase of scan rate and the dependence is
linear. This indicates that the process is controlled by electron
transfer and value of ˛n˛ equal to 0.48 04 was calculated [20].
Fig. 1 presents also a cyclic-voltammogram of cholesta-
3,5-diene (curve b). Cholesta-3,5-diene is oxidized at less
positive potential (+1.1 V) than cholesterol (+1.75 V) but in the
subsequent cathodic scan and the following second anodic
scan similar peaks as on cyclic voltammogram of choles-
terol are present. A cyclic voltammogram registered for a
mixture of cholesterol and cholesta-3,5-diene (curve 3) con-
firms the observation. The resemblance of the curves may
• Dicholest-5-en-3␤-yl ether (dicholesteryl ether; 6):
m.p. 202–203 ◦C (dichloromethane), literature [18] m.p.
205–207 ◦C (benzene–ethanol). 1H NMR ı (ppm): 5.34 (m, 1H,

546
steroids 7 3 ( 2 0 0 8 ) 543–548
suggest that cholesta-3,5-diene is formed during reaction in
dichloromethane as an intermediate and its further reaction
in the presence of cholesterol may be partly responsible for
dicholesteryl ether (6) formation. However, it is not very likely
that it is a dominating pathway leading to compound 6. Anal-
ysis of the reaction products strongly support an alternative
mechanism of compound 6 formation (vide infra). An elec-
trochemical oxidation of cholesta-3,5-diene is being under
investigation.
3.2.
Preparative electrolyses
The initial experiments were performed using a divided cell, in
which the cathode was placed in a glass tube with a glass frit.
The reaction products were separated by column chromatog-
raphy and carefully analyzed (Scheme 1). The major products
appeared to be cholesterol dichlorides 2 and 3 differing in con-
figurations at chiral centers. The major product 3 showed in
its 1H NMR spectrum a broad (w/2 = 23.5 Hz) multiplet at ı
4.31 ppm coming from an axial proton at C-3. This proved that
rings A and B in this compound are trans fused and the chlo-
Fig. 2 – Dependence of the half-peak potential (Ep/2) of
cholesterol (1 mM) oxidation process vs. log(v) (v is the scan
rate).
rine atom at ring junction occupies an ␣ position. Contrary
to that in the 1H NMR spectrum of compound 2 a narrow
signal of 3␣-proton (w/2 = 15.6 Hz) was present at 4.15 indi-
cating a trans fusion of the ring system. The configurations at
C-6 in dichlorides 2 and 3 were established by analyzing cou-
pling constants and NOE effects of a proton at C-6. Another
electrooxidation product was chlorohydrine 4. The structure
of compound 4 was established by spectroscopic methods
and confirmed by its acetylation with Ac2O in pyridine. 3-
Monoacetate was obtained proving that the second hydroxyl
group is placed at the tertiary carbon atom (C-5). In the case
of prolonged cholesterol electrooxidation, 3␤-chlorocholest-
5-ene (5) was additionally isolated. The nonpolar fraction of
products was not abounded and appeared to be a mixture of
hydrocarbons (dienes).
The results obtained were similar to those described
by Japanese authors [8]. They have also obtained products
of the double bond chlorination by electrolysis of choles-
terol in dichloromethane with metal chloride additives (e.g.
FeCl3). The reaction was then stereoselective and only 5␣,6␤-
dichloride 3 was formed in addition to chlorohydrine 4. It
is clear that dichloromethane was a source of chlorine in
our experiments since no additives were used. Its cathodic
Fig. 1 – Cyclic-voltammograms registered in 0.1 M
tetrabutylammonium tetrafluoroborate in dichloromethane
on platinum electrode (area, 0.03 cm2) of (a) 0.5 mM
cholesterol, (b) 0.25 mM cholesta-3,5-diene, and (c) the
mixture of 0.5 mM cholesterol with 0.25 mM
cholesta-3,5-diene. Scan rate 0.5 V s−1, potentials measured
vs. Ag/0.1 M AgNO3 electrode in acetonitrile.
Scheme 1

steroids 7 3 ( 2 0 0 8 ) 543–548
547
Scheme 2
reduction to chloride ions, their diffusion to anodic compart-
ment and electrooxidation in presence of cholesterol causes
the formation of chloro-derivatives of cholesterol [12]. To
avoid the process, cathodic and anodic compartments were
separated and connected with electrolytic bridge, which con-
tained 0.4 M solution of TBABF4 in the mixture of acetonitrile
and dichloromethane (50:50, v/v) to increase the conduc-
tance. Moreover, glacial acetic acid was added to cathodic
compartment of the electrolyser to prevent the electrochem-
ical reduction of dichloromethane [12]. This however, as
shown below, resulted in the appearance of 3␤-acetylamino
and 3␤-acetyl derivatives of cholesterol among the prod-
ucts.
The main product of preparative electrolysis in this con-
ditions was dicholesteryl ether 6 (Scheme 2), with material
yield equal to about 37%. This naturally occurring compound
is formed from cholesteryl sulfates or sulfonates [18]. The for-
mation of 6 as an unwanted side product was observed during
synthesis of steroid glycosides by the orthoester method [21]
or other methods [22]. The TLC analysis of nonpolar reaction
products showed presence of an intensive UV active spot (pre-
sumably from cholestadienes) but separation of electrolysis
products by column chromatography proved that cholestadi-
enes are formed only in trace amounts.
15 min and was stable for the extended period of time.
This indicates that organic cation-radicals of cholesterol are
formed [23].
It seems that the first step of electrooxidation is one-
electron transfer from cholesterol to anode. The cation-radical
centered at the oxygen atom is formed since the heteroatom
lone pair electrons are easily removable. The next step is
heterolysis of the C3 O bond leading to the homoallylic car-
bocation. The cleavage of this bond is assisted by the ␲
electrons of the neighboring double bond and the mesomeri-
cally stabilized nonclassical carbocation is formed [24]. Since
electrolyses were performed in dichloromethane, which does
not show nucleophilic properties, the only partner for this
carbocation in the reaction medium was the oxygen atom
of the second molecule of cholesterol 1 and dicholesteryl
ether
6 was formed. However, leakage of nucleophilic
ions (Cl−) or molecules (acetic acid, acetonitrile) from the
cathodic compartment or from the electrolytic bridge caused
the unwanted side reactions. The identification of reac-
tion by-products, cholesteryl 3␤-chloride (5), 3␤-acetate (7),
and 3␤-acetamide (8), unequivocally confirmed the reaction
mechanism.
Apart from the main product some minor products were
also formed. One of them was cholesteryl chloride indicat-
ing that the separation of anodic and cathodic compartments
of electrolyser did not fully prevent the migration of chlo-
rine ions. The addition of glacial acetic acid resulted in the
appearance of cholesterol acetate (7) among the products (due
to diffusion of acetic acid to anodic compartment). The pres-
ence of acetonitrile in electrolytic bridge caused the formation
of 3␤-acetylamino derivative of cholesterol (8). These results
indicate that it is very difficult to prevent leakage of unwanted
species to anodic compartment during the electrolysis pro-
cess.
3.3.
Tentative reaction mechanism
The tentative mechanism of electrochemical oxidation of
cholesterol is drawn in Fig. 3. During preparative electrolyses
the color of the analyte turns light-blue after approximately
Fig. 3 – The tentative mechanism of electrochemical
oxidation of cholesterol.

548
4.
steroids 7 3 ( 2 0 0 8 ) 543–548
[9] Wu X, Song C, Cheng F. Studies of the indirect
Conclusion
electrochemical oxidation of cholesterol. J Electroanal Chem
1992;327:321–5.
[10] Okamoto I, Funaki W, Nakaya K, Kotani E, Takeya T. A new
electrochemical system for stereoselective allylic
hydroxylation of cholesteryl acetate with dioxygen induced
by iron picolinate complexes. Chem Pharm Bull
2004;52:756–9.
[11] Kowalski J, Płoszyn´ ska J, Sobkowiak A, Morzycki J,
Wilczewska AZ. Direct electrochemical acetoxylation of
cholesterol at the allylic position. J Electroanal Chem
2005;585:275–80.
[12] Lund H. Practical problems in electrolysis. In: Lund H,
Hammerich O, editors. Organic electrochemistry. 4th ed New
York: Dekker; 2001. p. 223–92.
[13] Stahl E. Du¨ nnschicht-chromatographie. Berlin:
Springer-Verlag; 1967. p. 313.
[14] Perrin DD, Armarego WLF. Purification of laboratory
chemicals. 3rd ed Oxford: Pergamon Press; 1988. p. 145.
[15] Jacquesy R, Levisalles J. Configuration en C-5 du dichlorure
de cholesterol, F = 174–175 ◦C. Bull Soc Chim 1965;5:396–400.
[16] Barton DHR, Miller E. Stereochemistry of the cholesterol
dichlorides. J Am Chem Soc 1950;72:370–4.
[17] Kimura M, Tohma M, Tomita T. Metal ion catalyzed
oxidation of steroids. III. Reactions of cholesterol with
ferrous ions- and titanous ions–hydrogen peroxide systems
in acetonitrile. Chem Pharm Bull 1973;21:2521–8.
[18] Sobel AE, Owades PS, Owades JL. Sulfate esters as
intermediates in the formation of 7-dehydrocholesterol and
dicholesteryl ether. J Am Chem Soc 1949;71:1487–9.
[19] Barton DHR, Magnus PD, Garbarino JA, Young RN.
Conversion of alcohols into amides by
chlorodiphenylmethylium hexachloroantimonate in nitrile
solvents; some further reactions of the triphenylmethyl
cation. J Chem Soc, Perkin Trans 1974;1:2101–7.
[20] Nicholson RS, Shain I. Theory of stationary electrode
polarography. Single scan and cyclic methods applied to
reversible, irreversible, and kinetic systems. Anal Chem
1964;36:706–23.
[21] Uvarova NI, Oshitok GI, Elyakov GB. Synthesis of steroid and
triterpenoid glycosides by the orthoester method.
Carbohydrate Res 1973;27:79–87.
In the previous studies on cholesterol electrooxidation vari-
ous products were formed oxidized at the C5 C6 double bond
or at the allylic C-7 position. Electrooxidation of cholesterol
in dichloromethane with tetrabutylammonium tetrafluorobo-
rate as a supporting electrolyte led to the nonstereoselective
chlorination of the C5 C6 double bond. However, the reaction
carried out in separated cathodic and anodic compartments
unexpectedly afforded dicholesteryl ether (6) in a relatively
high yield. This reaction is rather surprising since secondary
alcohols are usually oxidized to ketones [25–27]. The most
likely reason for such unusual course of reaction is presence
of the C5 C6 double bond. It seems to be clear that during
electrooxidation a cation-radical centered at the cholesterol
oxygen atom is initially formed. The heterolytic cleavage of
the C3 O bond assisted by the ␲ electrons of the double bond
afforded the homoallylic cation which reacted with the sec-
ond molecule of cholesterol (or any other nucleophile present
in the reaction mixture) yielding the corresponding product.
Acknowledgement
Financial support from the Polish Ministry of Science and
Higher Education is gratefully acknowledged.
r e f e r e n c e s
[1] Brown AJ, Jessup W. Oxysterols and atherosclerosis.
Atherosclerosis 1999;142:1–28.
[2] Leonarduzzi G, Sottero B, Poli G. Oxidized products of
cholesterol: dietary and metabolic origin, and
proatherosclerotic effects. J Nutr Biochem 2002;13:700–10
[Review].
[3] Guardiola F, Codony R, Addis PB, Rafecas M, Boatella J.
Biological effects of oxysterols: current status. Food Chem
Toxic 1996;34:193–211.
[4] Chang YH, Abdalla DSP, Sevanian A. Characterization of
cholesterol oxidation products formed by oxidative
modification of low density lipoprotein. Free Radic Biol Med
1997;23:202–14.
[5] Guardiola F, Bou R, Boatella J, Codony R. Analysis of sterol
oxidation products in foods. J AOAC Int 2004;87:441–6.
[6] Ubhayasekera SJKA, Verleyen T, Dutta PC. Evaluation of GC
and GC–MS methods for the analysis of cholesterol
oxidation products. Food Chem 2004;84:149–57.
[7] Groves JT, Neumann R. Enzymic regioselectivity in the
hydroxylation of cholesterol catalyzed by a
membrane-spanning metalloporphyrin. J Org Chem
1988;53:3891–3.
[8] Maki S, Konno K, Ohba S, Takayama H. Stereoselective
chlorination of steroidal 5,6-olefin by an electrochemical
method; a convenient synthesis of blattelastanoside B.
Tetrahedron Lett 1998;39:3541–2.
¨
[22] Wulff G, Schmidt W. Uber die orthoesterbildung als
konkurrenzreaktion zur glykosylierung. Carbohydrate Res
1977;53:33–46.
[23] Kurasawa Y, Takada A, Ueda T. Color reaction of cholesterol
with trichloroacetic acid and antimony trichloride. On the
reaction mechanism. Steroids 1978;31:163–74.
[24] Kirk DN, Hartshorn MP. Steroid reaction mechanisms.
Amsterdam/London/New York: Elsevier Publishing
Company; 1968. p. 236–48.
[25] Morrow GW. Anodic oxidation of alcohols and derivatives.
In: Lund H, Hammerich O, editors. Organic electrochemistry.
4th ed New York: Dekker; 2001. p. 611–3.
[26] Amatore C, Cammoun C, Jurand A.
Palladium/benzoquinone-catalyzed electrochemical
oxidation of alcohols under anaerobic conditions. Synlett
2007;14:2173–8.
[27] Rodrigues IA, Nart FC. 2-Propanol oxidation on platinum
and platinum–rhodium electrodeposits. J Electroanal Chem
2006;590:145–51.