Full Paper
Chemistry—A European Journal
doi.org/10.1002/chem.202100446
[15]
ols have been intensely investigated. However, the low solu-
bility of steroids in aqueous systems severely limits product
17-ketosteroid 2a (Table 1, entry 1). Next, a screen of the opti-
mal reaction conditions was carried out. MeCN/H O 40:1
2
[15]
yield and productivity.
indeed proved to be the most suitable solvent mixture. Metha-
The classical chemical methods for the removal of the C17
side chain outlined above typically require a quench or neu-
tralization of the excess of reagents, in addition to work-up
and purification steps to isolate the corresponding 17-ketoste-
roid before it can be further derivatized to the target com-
nol (entry 2) and THF/H O (entry 3) provided lower conversion
2
or selectivity. Other solvent systems (EtOH, acetone/water) and
other proportions of MeCN/H O resulted in lower conversions
2
as well (see Table S1 in the Supporting Information). When
sodium and lithium perchlorate were used as the supporting
electrolyte instead of the tetraalkylammonium salt, lower con-
versions and a significantly lower selectivity was observed (en-
tries 4 and 5). Interestingly, other anode materials such as
glassy carbon, reticulated vitreous carbon (RVC) or platinum
performed very poorly (entries 6–8), with very low conversions
of the substrate (<5% in all cases and rather poor selectivity).
Modification of other reaction parameters, including higher
current densities and different substrate or supporting electro-
lyte concentrations (Table S1) did not improve the results.
Once the optimal reaction conditions had been established
[
7,16]
pound (Figure 1a).
We hypothesized that cleavage of the
a-hydroxyketone C17-C20 bond might be possible via anodic
oxidation in a catalyst- and reagent-free fashion (Figure 1b).
Electrochemistry has been shown as a safe and green method-
[17,18]
ology to induce organic redox transformations.
Such
methodology for the generation of 17-ketosteroids, in the ab-
sence of external reagents, would permit direct derivatization
of the intermediate without additional work-up or purification
steps, resulting in a convenient and sustainable one-pot proce-
dure for the multistep preparation of essential C19 steroids.
(
0.2m substrate concentration in MeCN/H O 40:1, 5 mA con-
2
stant current with graphite as the anode and stainless steel as
the cathode material), the amount of charge was gradually in-
creased (Table 1, entries 9 and 10). To our delight, full conver-
sion of the starting material and excellent selectivity towards
Results and Discussion
To test our hypothesis, the electrolysis of hydrocortisone 1a
was initially investigated as model reaction (Table 1). In a typi-
cal experiment, the corticosteroid and a supporting electrolyte
were placed in an undivided electrochemical cell (5 mL IKA
ElectraSyn vial) and, after addition of a solvent, the mixture
was electrolyzed under constant current at room temperature.
Gratifyingly, a first attempt in acetonitrile/water 40:1 with
À1
the target 17-ketosteroid 2a was achieved after 4 Fmol of
charge had been passed. The excellent selectivity of the reac-
tion enabled a very simple work-up procedure, which consist-
ed of evaporation of the solvent, dilution of the crude mixture
with a saturated aqueous solution of NaHCO3 (or distilled
water) and extraction of the product with DCM. Evaporation of
the organic solvent provided a quantitative yield of 2a.
Et NBF4 as the supporting electrolyte and graphite as the
4
anode, provided a promising 79% conversion of the starting
material and an excellent selectivity towards the corresponding
The scope and applicability of this electrochemical C17 side
chain cleavage protocol was subsequently investigated. Thus,
several key 17-ketosteroids 2 were prepared from their corre-
sponding corticosteroid precursors (Figure 2). As mentioned
above, 11b-hydroxyandrostenedione 2a was obtained in quan-
titative yield (Figure 2a). When cortisone and cortexolone were
used as starting materials, Reichstein’s substance G (adrenos-
terone) 2b and androstenedione 2c were obtained in quanti-
tative and excellent yield, respectively. Anodic side chain cleav-
age of the more sensitive prednisone and prednisolone also
provided satisfactory results, and 17-ketosteroids 2d and 2e
were obtained in excellent yields. Importantly, it could also be
shown that this reagent-free electrochemical procedure does
not require the presence of a hydroxyketone side chain, a
rather common limitation of some conventional chemical
cleavage methods. Thus, 17a-hydroxyprogesterone 3, which
contains a methylketone side chain, could be cleanly electro-
lyzed to the corresponding 17-ketosteroid 2c in quantitative
yield (Figure 2b). In contrast, the presence of the 17-hydroxyl
group proved to be essential for the side chain cleavage. Thus,
when the standard electrolysis conditions were applied to cor-
ticosterone, 17-ketosteroid formation was not observed.
Table 1. Optimization of the reaction conditions for the anodic C17 side
chain cleavage of hydrocortisone 1a.
[a]
[b]
Entry Conditions
Conv /Select
[
c]
(
%)
1
2
3
4
5
6
7
8
9
1
MeCN:H
MeOH
2
O (40:1) Et
Et
4
4
4
NBF
NBF
NBF
4
4
4
(+)G/Fe(À)
(+)G/Fe(À)
(+)G/Fe(À)
(+)G/Fe(À)
(+)G/Fe(À)
2 F/mol 79/99
2 F/mol 65/92
2 F/mol 78/91
2 F/mol 47/73
2 F/mol 35/73
THF:H
2
O (10:1)
Et
MeCN:H
MeCN:H
MeCN:H
MeCN:H
MeCN:H
MeCN:H
2
2
2
2
2
2
O (40:1) NaClO
O (40:1) LiClO
4
4
O (40:1) Et
O (40:1) Et
O (40:1) Et
O (40:1) Et
4
4
4
4
NBF
NBF
NBF
NBF
4
4
4
4
(+)GC/Fe(À) 2 F/mol <1/–
(+)RVC/Fe(À) 2 F/mol 9/67
(+)Pt/Fe(À) 2 F/mol 11/75
(+)G/Fe(À)
(+)G/Fe(À)
3 F/mol 84/99
0
MeCN:H
2
O (40:1) Et
4
NBF
4
4 F/mol 99/98
[
1
a] Undivided cell (Electrasyn 2.0, 5 mL vial), 0.6 mmol 1a, 3 mL solvent,
000 rpm stirring speed, constant current electrolysis at rt. [b] Deter-
As mentioned above, the preparation of many C19 andro-
gen steroids require several reaction steps. Side chain cleavage
utilizing conventional reagents complicate these multistep syn-
thetic routes, as quench and workup procedures for the purifi-
cation of the 17-ketosteroid intermediates are required. The
mined by HPLC peak area percent (254 nm). [c] Determined by HPLC
peak area percent (254 nm) as the percentage of product with respect to
all other peaks except the substrate. G: graphite. GC: glassy carbon. RVC:
reticulated vitreous carbon. Fe: stainless steel.
Chem. Eur. J. 2021, 27, 6044 – 6049
6045
ꢀ 2021 Wiley-VCH GmbH