Oxidation of Δ4- and δ5-steroids with hydrogen peroxide catalyzed by porphyrin complexes of MnIII and FeIII

Article abstract of DOI:10.1002/ejoc.200400461

In this paper we describe a new environmentally friendly method to promote the stereoselective epoxidation of Δ⁴- and Δ⁵- steroids. Metalloporphyrins efficiently catalyze the epoxidation reactions of 17β-acetoxy-4-androstene (1), 4-cholestene (2) and 3β-acetoxy-5- cholestene (3) in the presence of H₂O₂ as oxygen donor. Modeling the molecular structure of the porphyrin as well as the central metal allows the control of the preferential formation of α- or β-epoxides. Porphyrins with bulky, electron-withdrawing groups in the ortho positions of the meso phenyls and with Mn^III as the central metal ion, such as [Mn(TDCPP)Cl], gave preferentially the β-epoxide of Δ⁴- and Δ⁵-steroids. [Fe(TPFPP)Cl] catalyzes preferentially the α-epoxidation of Δ⁴-steroids and also increases the stereoselectivity for the α-epoxide in Δ⁵-steroids, similar to the results obtained with m-CPBA (m-chloroperbenzoic acid) as oxidant. The substrate structure strongly influences the chemoselectivity of the reactions. The X-ray structures of two main products were determined, and two-dimensional NMR techniques allowed the full assignment of ¹H and ¹³C NMR resonances as well as the stereochemistry of these products. A mechanistic proposal involving oxo species for the β-approach and peroxy species for the α-approach is proposed. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200400461

FULL PAPER
Oxidation of ∆4- and ∆5-Steroids with Hydrogen Peroxide Catalyzed by
Porphyrin Complexes of MnIII and FeIII
[a]
[a]
[a]
´
˜
Susana L. H. Rebelo, Mario M. Q. Simoes, M. Graca P. M. S. Neves,
¸
[a]
[a]
[b]
[b]
´
Artur M. S. Silva, Jose A. S. Cavaleiro,* Andreia F. Peixoto, Mariette M. Pereira,
[c]
[c]
[c]
´
˜
Manuela R. Silva, Jose A. Paixao, and Ana M. Beja
Keywords: Catalysis / Epoxidation / Iron / Manganese / Porphyrins / Steroids
In this paper we describe a new environmentally friendly
tially the α-epoxidation of ∆4-steroids and also increases the
stereoselectivity for the α-epoxide in ∆5-steroids, similar to
method to promote the stereoselective epoxidation of ∆4- and
∆5-steroids. Metalloporphyrins efficiently catalyze the epox- the results obtained with m-CPBA (m-chloroperbenzoic acid)
idation reactions of 17β-acetoxy-4-androstene (1), 4-choles-
as oxidant. The substrate structure strongly influences the
tene (2) and 3β-acetoxy-5-cholestene (3) in the presence of chemoselectivity of the reactions. The X-ray structures of two
H2O2 as oxygen donor. Modeling the molecular structure of
the porphyrin as well as the central metal allows the control
of the preferential formation of α- or β-epoxides. Porphyrins
main products were determined, and two-dimensional NMR
techniques allowed the full assignment of 1H and 13C NMR
resonances as well as the stereochemistry of these products.
with bulky, electron-withdrawing groups in the ortho posi- A mechanistic proposal involving oxo species for the β-ap-
tions of the meso phenyls and with MnIII as the central metal
proach and peroxy species for the α-approach is proposed.
ion, such as [Mn(TDCPP)Cl], gave preferentially the β-epox- (© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
ide of ∆4- and ∆5-steroids. [Fe(TPFPP)Cl] catalyzes preferen- Germany, 2004)
Introduction
for the functionalization of the steroid backbone, mainly in
positions 4 to 6 are now challenging goals.[5]
An important methodology is the epoxidation of unsatu-
The chemical modification of steroids can bring signifi-
rated ∆4- and ∆5-steroids, which has been performed either
cant biological effects.[1] This is obviously related to the ap-
directly or in a catalytic way, leading to different stereose-
plication of specific steroids as valuable therapeutic agents
lected products. Furthermore, the careful choice of an agent
in the treatment of several diseases, more recently in the
for the oxirane-ring opening allows the introduction of sev-
treatment of breast cancer.[2] Breast-cancer growth is pre-
eral functionalities into specific positions.[6Ϫ10]
sumably related to the presence of estrogens;[2] these are
biosynthesized in vivo from androgens in three successive
Direct epoxidation of both ∆4- and ∆5-steroids with or-
ganic peroxy acids affords mainly α-epoxides,[11] except in
oxidative steps, mediated by aromatase, a complex enzyme
the particular case when a β-hydroxy or an analogous
group is in the allylic position; in such cases syn stereodi-
of the cytochrome P450 family.[3,4] Therefore, inhibitors of
this enzyme have been developed in the last few years as
an approach to a treatment strategy. Compounds like 4-
hydroxyandrostan-3,17-dione and 6-methyleneandrosta-
1,4-diene-3,17-dione were found to be effective as inhibitors
recting effects have been observed, leading mainly to the
β-epoxides.[12,13] The catalytic epoxidation of ∆4- and ∆5-
steroids leads mainly to β-epoxidation.[14Ϫ16]
Metalloporphyrins have been found to be efficient cata-
of aromatase by blocking the estrogen biosynthesis and
lysts for the epoxidation of alkenes with several oxidants.[17]
Among the oxidants used, hydrogen peroxide allows a clean
chemistry, since water is the only byproduct obtained;[18Ϫ20]
such environmentally friendly procedures are now impera-
tive in chemical synthesis. The application of Ru porphyrins
in the oxidation of ∆5-steroids of the cholestane series by
O2 has been described; β-epoxides were obtained and the
conversion and stereoselectivity of the reaction were found
to be strongly dependent on the type of substituent in the
3-position.[21Ϫ24] During our studies[25] the oxidation of 5-
thereby causing tumor regression.[2] In order to improve the
activity of this type of compounds, stereoselective strategies
[a]
Department of Chemistry, University of Aveiro,
3810-193 Aveiro, Portugal
Fax: (internat.) ϩ 351-234-370-084
E-mail address: jcavaleiro@dq.ua.pt
[b]
Department of Chemistry, University of Coimbra,
3004-535 Coimbra, Portugal
[c]
CEMDRX, Department of Physics, University of Coimbra,
3004-535 Coimbra, Portugal
4778
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200400461
Eur. J. Org. Chem. 2004, 4778Ϫ4787

Catalytic Oxidation of ∆4- and ∆5-Steroids with Hydrogen Peroxide
FULL PAPER
cholestene derivatives by PhIO with a supported electron- Oxidation of 17β-Acetoxy-4-androstene (1)
donating MnIII porphyrin was also found to afford β-epox-
Oxidation of 1 was performed with H2O2 in the presence
ides.[26]
of the catalysts IϪIV and also with m-CPBA alone for com-
parative purposes (Scheme 1 and Table 1). The products
were isolated from the reaction mixtures by preparative
TLC on silica gel.
We report herein the oxidation of 17β-acetoxy-4-andros-
tene (1), 4-cholestene (2) and 3β-acetoxy-5-cholestene (3)
with H2O2, in the presence of several MnIII and FeIII por-
phyrins (Figure 1). The results are discussed on the basis
of a comparative study using m-CPBA (m-chloroperbenzoic
acid) as a standard oxidant.
Scheme 1
The structure and stereochemistry of epoxides 17β-ace-
toxy-4,5β-epoxyandrostane (1a) and 17β-acetoxy-4,5α-ep-
oxyandrostane (1b) were identified by GC-MS, NMR spec-
troscopy and X-ray crystallography. It is worth pointing out
that the characterization of this type of compounds has
been poorly described in the literature.[15,16]
Figure 1. Metalloporphyrin complexes used in these studies
The full assignment of the 1H and 13C NMR signals
(Table 2) was achieved by two-dimensional techniques, na-
mely gCOSY (1H/1H), gHSQC (1H/13C) and gHMBC (1H/
13C-long range) experiments, as well as by DEPT tech-
niques. The stereochemistry of the α- and β-epoxides was
unambiguously identified by X-ray crystallography (Fig-
ure 2 and 3).
Results and Discussion
Oxidation of Substrates
The oxidation reactions were carried out at room tem-
Compound 1a crystallizes in the chiral space group P21
perature, with progressive addition of H2O2, in the presence with two symmetry-related molecules (formula unit:
of the selected metalloporphyrin (Figure 1). The reactions C21H32O3) in the unit cell. In the chiral molecule all the
were followed by GC every 30 min and the addition of rings are trans-fused, with the exception of A/B, which have
H2O2 was stopped when the proportion of the compounds a cis-junction [torsion angles C(1)ϪC(10)ϪC(5)ϪC(6):
remained constant after two successive GC analysis. The 174.90(19)°; C(19)ϪC(10)ϪC(5)ϪC(4): 138.0(2)°]. Ring A
final results were based on the NMR spectra of the total has a distorted sofa conformation with the substituent O(1)
reaction mixtures (see Exp. Sect.).
atom bonded to C(4) and C(5) in a nearly bisectional posi-
˚
Oxidation reactions carried out with the manganese(iii) tion. The OϪC bond lengths are 1.444(4) and 1.443(3) A
porphyrins were performed in a CH3CN/CH2Cl2 mixture, for C(4) and C(5), respectively, and the angles around C(5)
with ammonium acetate as the co-catalyst, which is essen- are 115.10(19)° [O(1)ϪC(5)ϪC(10)] and 118.1(2)°
tial for the heterolytic cleavage of H2O2.[27] When the [C(4)ϪC(5)ϪC(6)]. Rings B and C have normal, slightly
iron(iii) catalyst was used the reactions were carried out in flattened chair conformations, and in the five-membered
CH3OH/CH2Cl2 and no co-catalyst was added.[28,29]
ring O(2) bonds to C(17) in an equatorial position.
Eur. J. Org. Chem. 2004, 4778Ϫ4787
www.eurjoc.org
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4779

J. A. S. Cavaleiro et al.
FULL PAPER
Table 1. Oxidation of 17β-acetoxy-4-androstene (1) with H2O2 in the presence of different porphyrin catalysts
Time
(h)
Conversion
(%)
Yield (%)
ES[a]
β/(αϩβ)
Entry
Catalyst
1a (β)
1b (α)
1c
1d
1e
1
2
3
4
5
6
7
8
9
[Mn(TDCPP)Cl] (I)[b]
[Mn(TDCPP)Cl] (I)[b]
[Mn(βNO2TDCPP)Cl] (II)[b]
[Mn(βNO2TDCPP)Cl] (II)[b]
[Mn(TPFPP)Cl] (III)[b]
[Fe(TPFPP)Cl] (IV)[c]
[Fe(TPFPP)Cl] (IV)[c]
m-CPBA[d]
1
2.5
1
2
2
0.5
1
1
88
98
74
96
79
81
100
99
0
49
43
25
31
34
24
33
33
Ϫ
18
15
8
10
29
47
57
51
Ϫ
3
0
13
3
3
0
0
5
Ϫ
5
6
10
18
4
0
0
8
32
8
25
8
0
0
73
74
76
76
54
34
37
39
Ϫ
0
Ϫ
0
Ϫ
no catalyst
4
[a]
[b]
Epoxidation selectivity, measured by the proportion of the β-epoxide (1a) with respect to the total amount of epoxides 1a and 1b.
Reaction conditions: substrate (0.03 mmol), catalyst (0.2 µmol) and ammonium acetate (0.08 mmol) were dissolved in 2 mL of CH3CN/
CH2Cl2 (1:1) and stirred at room temperature. Aqueous H2O2 (30% w/w) diluted with CH3CN (1:20) was added to the reaction mixture
[c]
in 30 µL aliquots every 15 min. Reaction conditions: as in footnote b but without ammonium acetate and the reaction medium was
CH2Cl2/CH3OH (1:3). Reaction carried out without catalyst and with m-CPBA as oxidant instead of H2O2.
[d]
1
Table 2. H and 13C NMR spectroscopic data (at 500.13 and 125.77 MHz, respectively) of epoxides 1a and 1b[a]
17β-Acetoxy-4,5β-epoxiandrostane (1a)
17β-Acetoxy-4,5α-epoxiandrostane (1b)
C
13C (δ)
1H (δ)
13C (δ)
1H (δ)
C-1
C-2
C-3
C-4
C-5
C-6
C-7
C-8
29.6
15.3
23.6
61.4
65.5
31.3
30.0
34.9
46.4
36.4
20.7
36.8
42.6
50.5
23.6
27.5
82.7
12.0
19.2
171.2
21.2
1.25Ϫ1.30 (m, 2 H)
1.23Ϫ1.30 (m, 2 H)
31.2
15.9
22.2
60.8
65.9
30.1
27.9
35.3
50.1[b]
35.6
20.2
36.8
42.6
50.3[b]
23.5
27.5
82.8
12.1
17.4
171.2
21.2
1.18Ϫ1.24 (m, 1 H), 1.29Ϫ1.33 (m, 1 H)
1.35Ϫ1.42 (m, 2 H)
1.83Ϫ1.88 (m, 2 H)
2.93 (d, J ϭ 3.9 Hz, 1 H)
Ϫ
0.94Ϫ0.98 (m, 1 H), 2.00Ϫ2.06 (m, 1 H)
1.26Ϫ1.28 (m, 1 H), 1.60Ϫ1.62 (m, 1 H)
1.56Ϫ1.60 (m, 1 H)
1.13Ϫ1.18 (m, 1 H)
Ϫ
1.23Ϫ1.27 (m, 1 H), 1.48Ϫ1.50 (m, 1 H)
1.18Ϫ1.22 (m, 1 H), 1.72Ϫ1.75 (m, 1 H)
Ϫ
1.82Ϫ1.87 (m, 1 H), 1.89Ϫ1.95 (m, 1 H)
2.90 (d, J ϭ 4.6 Hz, 1 H)
Ϫ
0.99Ϫ1.07 (m, 1 H), 2.07Ϫ2.15 (m, 1 H)
0.99Ϫ1.05 (m, 1 H), 1.75Ϫ1.79 (m, 1 H)
1.48Ϫ1.53 (m, 1 H)
1.20Ϫ1.24 (m, 1 H)
Ϫ
1.32Ϫ1.35 (m, 1 H), 1.46Ϫ1.50 (m, 1 H)
1.13Ϫ1.20 (m, 1 H), 1.73Ϫ1.76 (m, 1 H)
Ϫ
C-9
C-10
C-11
C-12
C-13
C-14
C-15
C-16
C-17
C-18
C-19
CO2CH3
CO2CH3
1.08Ϫ1.13 (m, 1 H)
1.13Ϫ1.18 (m, 1 H)
1.31Ϫ1.36 (m, 1 H), 1.63Ϫ1.69 (m, 1 H)
1.49Ϫ1.53 (m, 1 H), 2.16Ϫ2.20 (m, 1 H)
4.60 (dd, J ϭ 0.9 and 17.0 Hz, 1 H)
0.81 (s, 3 H)
1.01 (s, 3 H)
Ϫ
1.29Ϫ1.34 (m, 1 H), 1.62Ϫ1.69 (m, 1 H)
1.47Ϫ1.51 (m, 1 H), 2.15Ϫ2.20 (m, 1 H)
4.60 (dd, J ϭ 1.3 and 17.0 Hz, 1 H)
0.81 (s, 3 H)
1.07 (s, 3 H)
Ϫ
2.05 (s, 3 H)
2.04 (s, 3 H)
[a]
[b]
Chemical shifts are expressed in ppm and the coupling constants (J) in Hz. The assignments of these two carbons could be inter-
changeable.
Compound 1b crystallizes in the chiral space group
There are another three isolated products, identified by
1
P212121 with four symmetry-related molecules (formula GC-MS and H and 13C NMR spectroscopy, as 17β-ace-
unit: C21H32O3) in the unit cell. In the chiral molecule all toxy-4-androsten-3-ol (1c), 17β-acetoxy-4-androsten-3-one
the rings are trans-fused. The torsion angles around the (1d) and 17β-acetoxy-4,5-epoxyandrostan-3-ol (1e).
A/B junction are 178.3(10)° [C(1)ϪC(10)ϪC(5)ϪC(6)] and
The catalytic results are shown in Table 1 and these data
94.7(12)° [C(19)ϪC(10)ϪC(5)ϪC(4)]. Ring A has a dis- clearly indicate that the oxidation efficiency is dependent
torted sofa conformation with the substituent O(1) atom on the catalyst used and on the reaction time. Maximum
bonded to C(4) and C(5) in a nearly bisectional position. conversions of 98, 96, 79 and 100% were obtained with
˚
The OϪC bond lengths are 1.473(13) and 1.442(13) A for [Mn(TDCPP)Cl], [Mn(βNO2TDCPP)Cl], [Mn(TPFPP)Cl]
C(4) and C(5), respectively, and the angles around C(5) and [Fe(TPFPP)Cl], respectively. With manganese porphy-
are 114.3(9)° [O(1)ϪC(5)ϪC(10)] and 120.4(11)° rins, the fastest reaction was obtained with
[C(4)ϪC(5)ϪC(6)]. Rings B and C have normal slightly [Mn(TDCPP)Cl], which gave rise to 88% conversion after
flattened chair conformations, and in the five-membered one hour; however, in an equal time period, [Fe(TPFPP)Cl]
ring O(2) the bonds to C(17) are in the equatorial positions. gave complete conversion of substrate.
4780
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4778Ϫ4787

Catalytic Oxidation of ∆4- and ∆5-Steroids with Hydrogen Peroxide
FULL PAPER
Figure 2. ORTEP[44] diagram of compound 1a drawn at the 30% probability level
Figure 3. ORTEP[44] diagram of compound 1b drawn at the 30% probability level
The maximum yield of the β-epoxide (1a) formation was
achieved with [Mn(TDCPP)Cl] after 1 h of reaction (49%,
Table 1, entry 1). With this catalyst we also observed an
increase in the yield of compound 1e from 8 to 32% when
the reaction time was increased from 1 h to 2.5 h (entry 2).
Despite the high β-stereoselectivity observed in the reac-
tions catalyzed by [Mn(βNO2TDCPP)Cl], lower yields of
epoxides were found since this catalyst seems to promote
allylic oxidation, producing the highest yields of alcohol 1c
(13%) and ketone 1d (18%) after 1 h and 2 h of reaction,
respectively (entries 3 and 4). Also in this case, after 2 h of
reaction, compound 1e was obtained in 25% yield. Moder-
ate yields for the epoxides were observed with
[Mn(TPFPP)Cl], but nearly equal amounts of β- and α-
epoxides were obtained (entry 5).
Figure 4. Oxidation of 17β-acetoxy-4-androstene (1) with H2O2 in
the presence of catalysts IϪIV, after one hour reaction time; the
results are compared with those from stoichiometric m-CPBA oxi-
dation
A change in stereoselectivity was observed in the pres-
ence of [Fe(TPFPP)Cl], since this reaction took place with
higher selectivity for the α-epoxide 1b. With this catalyst,
and after 1 h of reaction, total conversion and 57% yield of
1b were observed (Table 1, entry 7). This catalyst shows an
oxidation pattern similar to that due to m-CPBA, where the
α-epoxide was obtained in 51% yield (entry 8). The epoxid-
Oxidation of 4-Cholestene (2)
The oxidation of ∆4-steroid 2 was also performed with
ation stereoselectivity distribution can be more clearly ob- H2O2 in the presence of the catalysts IϪIV, as well as with
served in Figure 4.
m-CPBA for comparative purposes (Table 3).
Eur. J. Org. Chem. 2004, 4778Ϫ4787
www.eurjoc.org
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4781

J. A. S. Cavaleiro et al.
FULL PAPER
Table 3. Oxidation of 4-cholestene 2 with H2O2 in the presence of different porphyrin catalysts
Time
(h)
Conversion
(%)
Yield (%)
ES[a]
β/(βϩα)
Entry
Catalyst
2a (β)
2b (α)
2c
2d
2e
1
2
3
4
5
6
7
8
9
[Mn(TDCPP)Cl] (I)[b]
[Mn(TDCPP)Cl] (I)[b]
[Mn(βNO2TDCPP)Cl] (II)[b]
[Mn(βNO2TDCPP)Cl] (II)[b]
[Mn(TPFPP)Cl] (III)[b]
[Fe(TPFPP)Cl] (IV)[c]
[Fe(TPFPP)Cl] (IV)[c]
m-CPBA[d]
1
2.5
1
2
2
0.5
1
1
84
96
72
95
79
80
98
94
0
42
41
27
27
34
25
33
36
Ϫ
17
16
11
11
29
44
54
50
Ϫ
6
2
11
5
4
0
0
4
Ϫ
5
8
9
21
3
0
0
1
Ϫ
9
25
8
24
6
0
0
71
72
71
71
54
36
38
42
Ϫ
0
Ϫ
no catalyst
4
[a]
[b]
Epoxidation selectivity, measured by the proportion of the β-epoxide (2a) with respect to the total amount of epoxides 2a and 2b.
Reaction conditions: substrate (0.03 mmol), catalyst (0.2 µmol) and ammonium acetate (0.08 mmol) were dissolved in 2 mL of CH3CN/
CH2Cl2 (1:1) and stirred at room temperature. Aqueous H2O2 (30% w/w) diluted with CH3CN (1:20) was added to the reaction mixture
[c]
in 30 µL aliquots every 15 min. Reaction conditions: as in footnote b but without ammonium acetate and the reaction medium was
CH2Cl2/CH3OH (1:3). Reaction carried out without catalyst and with m-CPBA as oxidant instead of H2O2.
[d]
The product mixtures resulting from substrate 2 oxi- two-dimensional techniques, namely gCOSY (1H/1H),
dation reactions were fractionated by preparative TLC on gHSQC (1H/13C) and gHMBC (1H/13C-long range) experi-
silica gel; the isolated products (Scheme 1) were 4,5β-epoxy- ments, as well as by DEPT techniques. The results are col-
cholestane (2a), 4,5α-epoxycholestane (2b), 4-cholesten-3-ol lected in Table 4.
(2c), 4-cholesten-3-one (2d) and 4,5-epoxycholestan-3-ol
The stereochemistry of the epoxides 2a and 2b was estab-
(2e), which were fully characterized by GC-MS and 1H and lished by comparison of their NMR spectra with those de-
13C NMR spectroscopy.
scribed for the 4,5-epoxides of the androstane series (1a and
The assignment of the resonances in the 1H and 13C 1b). Previous literature data reported the partial charac-
NMR spectra of compounds 2a and 2b was achieved using terization of these compounds,[7,10,11] but from these data
1
Table 4. 13C and H NMR spectroscopic data (at 75.47 and 300.13 MHz, respectively) of epoxides 2a and 2b[a]
4,5β-Epoxycholestane (2a)
4,5α-Epoxycholestane (2b)
C
13C (δ)
1H (δ)
13C (δ)
1H (δ)
C-1
C-2
C-3
C-4
C-5
C-6
C-7
C-8
29.6
15.3
23.7
61.5
65.7
31.5
30.6
35.1
46.4
36.2
23.8
39.8
42.6
56.2
21.3
24.3
56.2
12.0
19.2
35.8
18.6
36.2
28.2
39.5
28.0
22.6
22.8
1.25Ϫ1.32 (m, 2 H)
1.20Ϫ1.33 (m, 2 H)
1.82Ϫ1.94 (m, 2 H)
2.89 (d, J ϭ 4.3 Hz, 1 H)
Ϫ
0.97Ϫ1.02 (m, 1 H), 2.09 (dt, J ϭ 4.1 and 14.2 Hz, 1 H)
0.96Ϫ1.05 (m, 1 H), 1.71Ϫ1.80 (m, 1 H)
1.41Ϫ1.47 (m, 1 H)
1.18Ϫ1.25 (m, 1 H)
Ϫ
1.11Ϫ1.14 (m, 1 H), 1.32Ϫ1.40 (m, 1 H)
1.99 (dt, J ϭ 3.2 and 12.5 Hz, 2 H)
Ϫ
1.11Ϫ1.14 (m, 1 H)
1.32Ϫ1.47 (m, 2 H)
1.60Ϫ1.64 (m, 2 H)
0.97Ϫ1.09 (m, 1 H)
0.68 (s, 3 H)
1.00 (s, 3 H)
0.96Ϫ1.05 (m, 1 H)
0.91 (d, J ϭ 6.5 Hz, 3 H)
1.32Ϫ1.40 (m, 2 H)
1.81Ϫ1.88 (m, 2 H)
1.10Ϫ1.18 (m, 2 H)
1.47Ϫ1.54 (m, 1 H)
0.863 (d, J ϭ 6.6 Hz, 3 H)
0.867 (d, J ϭ 6.6 Hz, 3 H)
31.1
15.9
22.3
60.7
66.1
30.2
28.4
35.5
50.2
35.5
23.8
39.8
42.6
55.7
20.7
24.1
56.1
12.0
17.4
36.1
18.6
36.2
28.2
39.5
28.0
22.6[d]
22.8[d]
1.19Ϫ1.24 (m, 2 H)
1.35Ϫ1.41 (m, 2 H)
1.85Ϫ1.88 (m, 2 H)
2.92 (d, J ϭ 3.0 Hz, 1 H)
Ϫ
0.94Ϫ0.99 (m, 1 H), 2.00Ϫ2.06 (m, 1 H)
1.25Ϫ1.30 (m,1 H), 1.60Ϫ1.62 (m, 1 H)
1.58Ϫ1.60 (m,1 H)
1.13Ϫ1.19 (m, 1 H)
Ϫ
1.11Ϫ1.16 (m, 1 H), 1.30Ϫ1.38 (m, 1 H)
1.98Ϫ2.01 (m, 2 H)
Ϫ
1.11Ϫ1.14 (m, 1 H)
1.30Ϫ1.41 (m, 2 H)
1.58Ϫ1.62 (m, 2 H)
1.00Ϫ1.09 (m, 1 H)
0.68 (s, 3 H)
C-9
C-10
C-11
C-12
C-13
C-14
C-15[b]
C-16[b]
C-17
C-18
C-19
C-20
C-21
C-22
C-23
C-24
C-25
C-26[c]
C-27[c]
1.05 (s, 3 H)
0.96Ϫ1.05 (m, 1 H)
0.90 (d, J ϭ 6.5 Hz, 3 H)
1.32Ϫ1.40 (m, 2 H)
1.85Ϫ1.87 (m, 2 H)
1.10Ϫ1.16 (m, 2 H)
1.47Ϫ1.54 (m, 1 H)
0.87 (d, J ϭ 6.6 Hz, 3 H)
0.85 (d, J ϭ 6.6 Hz, 3 H)
[a]
[b]
Chemical shifts are expressed in ppm and the coupling constants (J) in Hz.
The assignments of these two carbons could be
[c]
[d]
interchangeable.
The assignments of these two carbons could be interchangeable.
The assignments of these two carbons could
be interchangeable.
4782
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4778Ϫ4787

Catalytic Oxidation of ∆4- and ∆5-Steroids with Hydrogen Peroxide
FULL PAPER
we were unable to distinguish between the two epoxides due Oxidation of 3β-Acetoxy-5-cholestene (3)
to incoherent NMR results.[10,11] With the techniques now
The two epoxides resulting from the oxidation of sub-
available (NMR spectroscopy and X-ray crystallography)
we concluded that several previous NMR assignments are
not correct. The chemical shift of H-4 in 4,5β-epoxycholes-
tane (2a) was unequivocally assigned at δ ϭ 2.89 ppm in-
stead of δ ϭ 2.98 ppm.[11] The assignment of the carbon
resonances of 4,5α-epoxycholestane (2b)[10] was corrected
for C-2 (δ ϭ 15.9 instead of 22.5 ppm), C-3 (δ ϭ 22.3 in-
stead of 28.5 ppm) and C-19 (δ ϭ 17.4 instead of 16.0 ppm).
The assignment of the carbon resonances of 4,5β-epoxycho-
lestane (2a)[10] was corrected for C-2 (δ ϭ 15.3 instead of
17.4 ppm) and for C-10 (δ ϭ 36.2 instead of 35.2 ppm). The
unequivocal assignment of all resonances of compounds 2a
and 2b is given in Table 4.
strate 3 (Scheme 2) were isolated from the reaction mixture
by column chromatography and identified by comparing
1
their H NMR spectra with literature data.[23] The signal
of the 6β proton of the 5,6α-epoxycholestane appears as a
doublet at δ ϭ 2.9 ppm (J ϭ 4 Hz), and is clearly dis-
tinguishable from the doublet at δ ϭ 3.1 ppm (J ϭ 2 Hz)
for the 6α proton of the 5,6β-epoxide. The mass spectra of
these compounds show that a loss of acetic acid occurs in
the injector, as the highest m/z values observed correspond
to [Mϩ· Ϫ 60] ions.
The behavior towards oxidation of 4-cholestene by H2O2,
catalyzed by metalloporphyrins, is similar to that observed
with steroid 1. The catalytic results collected in Table 3
show that the oxidation efficiency depends on the catalyst
used and on the reaction time. Maximum conversions of
96, 95, 79 and 98% were obtained with [Mn(TDCPP)Cl],
[Mn(βNO2TDCPP)Cl], [Mn(TPFPP)Cl] and [Fe(TPFPP)-
Cl], respectively. The fastest reaction occurred in the pres-
ence of catalyst IV, followed by the reaction catalyzed by
catalyst I, which reached 98 and 84% conversion, respec-
tively, after one hour of reaction.
Despite these differences, epoxidation is the main path-
way observed in all reactions. For synthetic purposes
[Mn(TDCPP)Cl] is the best catalyst to obtain the 4,5β-
epoxide (2a): a 42% yield of this compound was achieved
after 1 h of reaction time (entry 1). With the same catalyst
a longer reaction time (2.5 h) afforded a 25% yield of the
Scheme 2
Oxidation of steroid 3 in the presence of catalysts IϪIV
epoxy alcohol (2e). 4,5α-Epoxycholestane (2b) was ob- proceeds with 100% chemoselectivity for epoxidation
tained as the major product with [Fe(TPFPP)Cl] (entry 7); (Table 5). High conversions are obtained with
these results are similar to the ones obtained with m-CPBA [Mn(TDCPP)Cl] and [Mn(βNO2TDCPP)Cl] (79 and 80%,
(entry 8), although the iron porphyrin complex is slightly respectively) and these reactions lead to high β-stereoselec-
more selective for the epoxidation reaction. Although high tivity. [Mn(TDCPP)Cl] is the best catalyst for β-epoxid-
conversions are achieved with [Mn(βNO2TDCPP)Cl], this ation, affording compound 3a in 71% yield and with 90%
is not a catalyst of choice for these epoxidations due to the β-selectivity.
high yields of allylic alcohol and ketone obtained. Catalyst
III presents the lowest stereoselectivity observed in the without catalyst gave 100% conversion, but lower stereo-
epoxidation of 4-cholestene. selectivities. In the reaction catalyzed by the iron porphyrin
The reactions with [Fe(TPFPP)Cl] and H2O2 or m-CPBA
Table 5. Oxidation of 3β-acetoxy-5-cholestene (3) with H2O2 in the presence of different porphyrin catalysts
Entry
Catalyst
Time
(h)
Conversion
(%)
Yield (%)
3a (β)
ES[a]
β/(βϩα)
3b (α)
1
2
3
4
5
6
[Mn(TDCPP)Cl] (I)[b]
[Mn(βNO2TDCPP)Cl] (II)[b]
[Mn(TPFPP)Cl] (III)[b]
[Fe(TPFPP)Cl] (IV)[c]
m-CPBA[d]
3
4
3
2
2
5
79
80
43
100
100
0
71
66
31
66
43
Ϫ
8
14
12
34
57
Ϫ
90
82
72
66
43
Ϫ
no catalyst
[a]
[b]
Epoxidation selectivity, measured by the proportion of the β-epoxide (3a) with respect to the total amount of epoxides 3a and 3b.
Reaction conditions: substrate (0.03 mmol), catalyst (0.2 µmol) and ammonium acetate (0.08 mmol) were dissolved in 2 mL of CH3CN/
CH2Cl2 (1:1) and stirred at room temperature. Aqueous H2O2 (30% w/w) diluted with CH3CN (1:20) was added to the reaction mixture
[c]
in 30 µL aliquots every 15 min. Reaction conditions as in footnote b but without ammonium acetate and the reaction medium was
CH2Cl2/CH3OH (1:3). Reaction carried out without catalyst and with m-CPBA as oxidant instead of H2O2.
[d]
Eur. J. Org. Chem. 2004, 4778Ϫ4787
www.eurjoc.org
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4783

J. A. S. Cavaleiro et al.
FULL PAPER
IV a β-selectivity of 66% was achieved. The stoichiometric or it can even act as a metalloperoxy complex as the final
m-CPBA oxidation afforded a selectivity of 43% for the β- oxidant (species D).[32]
epoxide 3b and 57% for the α-isomer (Figure 5).
Although in some cases indirect evidence suggests that
homolytic cleavage can occur with alkyl hydroperoxides, to
generate RO· radicals at low temperature, which can initiate
a radical chain oxidation mechanism,[33,34] many reported
results also show that with H2O2 in the presence of MnIII
porphyrins like [Mn(TDCPP)Cl] and a buffering co-cata-
lyst, a manganese oxo complex is the final active species,
and therefore heterolytic cleavage takes place. Evidence for
that is: (a) the type of reactivity (conversion and selectivity
in several substrates oxidation) demonstrated by the system
(MnTDCPP/H2O2/imidazole) is identical to that observed
with the system (MnTDCPP/PhIO/imidazole) and this
points out to the involvement of the same final active spec-
ies;[19] (b) the change from almost no reactivity to high reac-
tivity when the [Mn(TDCPP)Cl]/H2O2 system is used in the
absence or in the presence of a buffering co-catalyst which
can contribute to the heterolytic cleavage by basic and acid
catalysis;[20,27] or (c) the preferential epoxidation of un-
reactive aromatic hydrocarbons like naphthalene by the
[Mn(TDCPP)Cl]/H2O2/NH4OAc system, which is very
unlikely in a radical-type mechanism.[35]
In the conditions used here for MnIII porphyrin catalysis,
the heterolytic cleavage of hydrogen peroxide probably oc-
curs and a manganese oxo species is the final oxidant; this
species is capable of both double-bond epoxidation and hy-
droxylation of activated allylic CϪH bonds.
In the case of [Fe(TPFPP)Cl], which shows identical reac-
tivity to m-CPBA, we consider it likely that under the reac-
tion conditions used (protic solvent/no buffering co-cata-
lyst) the final oxidation species is a metalloperoxy species
(D in Scheme 3; R ϭ H). This kind of species has been de-
scribed to be inert towards alkanes, but is probably reactive
enough to epoxidise alkenes.[36]
Figure 5. Oxidation of 3-β-acetoxy-5-cholestene (3) with H2O2 in
the presence of catalysts IϪIV; the results are compared with those
from stoichiometric m-CPBA oxidation
Mechanistic Considerations
The stereochemical differences observed in the epoxid-
ation of ∆4-steroids by H2O2 in the presence of different
metalloporphyrins point to the possibility that different
mechanisms and intermediate species could be present. It is
now generally accepted that oxidation reactions catalyzed
by metalloporphyrins with oxidants such as PhIO, m-
CPBA, KHSO5 or NaOCl proceed through a short cycle of
the cytochrome P450-type mechanism, with the formation
of high-valent oxo species, which seem to be the final oxi-
dant species (B in Scheme 3). Many theoretical studies and
the detection of the elusive metalloporphyrin oxo species
by stopped-flow visible spectrometry have corroborated this
hypothesis.[30,31] However, the mechanism of oxidation reac-
tions with H2O2 or alkyl hydroperoxides in the presence of
metalloporphyrins has been more controversial because, in
Further evidence for these two different mechanisms was
the presence of these oxidants, the intermediate species A obtained in the visible spectra of the reaction mixtures:
(Scheme 3) can undergo heterolytic or homolytic cleavage, MnIII porphyrins, with Soret bands near 480 nm, are con-
Scheme 3
4784
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4778Ϫ4787

Catalytic Oxidation of ∆4- and ∆5-Steroids with Hydrogen Peroxide
FULL PAPER
verted, during the course of the reaction, into another spec- the major compounds produced, similar to the stoichio-
ies with a brief lifetime and with Soret bands close to metric oxidation by m-CPBA. In the oxidation of ∆5-ster-
430 nm; the Soret band of [Fe(TPFPP)Cl] is maintained oids, an increase in the α-face attack was observed when
close to 410 nm during the whole reaction.
catalyst IV was used.
The stereochemical hindrance of the substrates and the
These results have been rationalized in terms of the active
different intervening catalytic species referred to above can metalloporphyrin catalytic species together with the stereo-
also justify the approach of the metalloporphyrin complex chemical hindrance of the substrate, and we have proposed
to the substrate from either the α or the β side. In ∆4-ster- the involvement of an oxo species for the β approach and a
oids, the most stable folded structure along the A and B peroxy species for the α approach. Considering recent pub-
ring-fusion bond has previously been invoked to explain the lications on the possible existence of multiple active oxy-
formation of β products.[37] Therefore, the porphyrin oxo genating species in porphyrin-catalyzed oxidations,[43] the
species should approach the double bond preferably from present results give further insight into such a hypothesis
the cis side, producing preferentially the β-epoxides, when hydrogen peroxide is used as the oxidant.
whereas metalloperoxyporphyrin species having a periph-
The main products were fully characterized by single-
eral atom bridge, analogous to m-CPBA, can approach the crystal X-ray diffraction and double resonance NMR tech-
double bond from the trans side, giving rise to α-epoxides niques; the results constitute an important and clarifying
as the main products.
study of this type of compounds.
We consider that [Mn(TDCPP)Cl] activates the hydrogen
Further studies on the epoxidation reaction mechanism,
peroxide mainly through the formation of a MnV oxo spec- in order to disclose the effect of the porphyrin structure
ies, leading mainly to β-epoxidation. The differences ob- and the central metal on the catalytically active species, are
served for the performances of [Mn(TDCPP)Cl] and currently being carried out using different catalysts and
[Mn(TPFPP)Cl] could be explained by a more difficult for- substrates. The results will be published in due course.
mation of this oxo species in the [Mn(TPFPP)Cl] case, due
to its higher redox potential. In this case, both oxo and
peroxy species could be active. The differences between the
Experimental Section
General Details: H and 13C NMR spectra were acquired at 22 °C
in CDCl3 solutions at 500.13 MHz or 300.13 MHz and at 125.76
or 75.47 MHz, respectively, using Bruker DRX 500 and 300 spec-
trometers. The chemical shifts are expressed in δ (ppm) values rela-
tive to tetramethylsilane (TMS) as internal reference. GC-MS
analyses were performed using a Finnigan Trace GC-MS (Thermo
Quest CE instruments) using helium as the carrier gas (35 cm/s).
GC-FID analyses were performed using a Varian Star 3400 CX gas
chromatograph and hydrogen as the carrier gas (55 cm/s). In both
cases fused silica Supelco capillary columns SPB-5 (30 m ϫ
0.25 mm i.d.; 0.25 µm film thickness) were used. Gas chromato-
graphic conditions: temperature 250 °C, during 1 min; temperature
rate 30 °C/min; final temperature 290 °C during 12 min; injector
temperature 290 °C; detector temperature 300 °C.
1
two catalysts could also be explained by the smaller sub-
stituents on [Mn(TPFPP)Cl], which lead to a lower stereo-
selectivity. In the oxidation of the ∆5-steroid, the approach
of the catalyst from the trans side to afford the α-epoxide is
even more hindered, and this can justify the lower reaction
conversions and higher β-selectivities reported with all the
metalloporphyrins. On the other hand, m-CPBA is a small
molecule with a peroxy atom bridge, and it can approach
the double bond from both sides.
Conclusions
The epoxidation of ∆4- and ∆5-steroids of the androstane
and cholestane series proceeds efficiently in the presence of
metalloporphyrin catalysts under mild conditions and using
an ecologically safe oxidant such as hydrogen peroxide.
The system H2O2/[Mn(TDCPP)Cl] or [Mn(βNO2-
TDCPP)Cl]/NH4AcO in CH3CN/CH2Cl2 performs the
epoxidation of ∆4- and ∆5-steroids in a β-stereoselective
way. In the reactions of ∆4-steroids 1 and 2 epoxidation was
always the main pathway, but allylic alcohols and ketones
were also observed in minor amounts, especially with
[Mn(βNO2TDCPP)Cl]. Catalyst I afforded the highest
yields of β-epoxides. For conversions above 90%, significant
and increasing quantities of 4,5-epoxy-3-hydroxysteroids
were also observed with catalysts I and II. Catalyst III was
the least stereoselective. In the reaction of the ∆5-steroid 3,
100% chemoselectivity for the epoxidation and higher β-
selectivities were observed.
Preparative thin-layer chromatography (TLC) was carried out on
silica gel plates (Riedel-de Hae¨n silica gel 60 DGF254). Hydrogen
peroxide (30 wt.-% solution in water) and acetonitrile were pur-
chased from Riedel-de Hae¨n. All other chemicals and solvents were
obtained from commercial sources and either used as received or
distilled and dried using standard procedures. Light petroleum was
the fraction with a boiling point of 40Ϫ60 °C.
Preparation of Substrates: The substrates were prepared by pub-
lished procedures. Substrates 1 and 2 were obtained by reduction
of testosterone acetate and 4-cholesten-3-one, respectively.[5] Sub-
strate 3 was prepared by reaction of 5-cholesten-3β-ol with Ac2O
in pyridine.[38] The spectroscopic data obtained for 1, 2 and 3 were
identical to those reported in the literature.
Catalyst Synthesis: The free bases of the metalloporphyrins IϪIV
(Figure 1) were prepared according to described procedures.[39Ϫ41]
Metallation of the free bases leading to the formation of complexes
IϪIV was performed with MnCl2 or FeCl2 according to conven-
tional methods.[40,42]
High conversions were observed with the system H2O2/
[Fe(TPFPP)Cl] in CH3OH/CH2Cl2. In oxidation of ∆4-ster-
Oxidation Reactions with MnIII Porphyrins: In a typical experiment,
oids the stereochemistry was inverted, and α-epoxides were the substrate (0.03 mmol) and ammonium acetate (0.08 mmol)
Eur. J. Org. Chem. 2004, 4778Ϫ4787
www.eurjoc.org © 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 4785

J. A. S. Cavaleiro et al.
FULL PAPER
were dissolved in 2 mL of a stock solution of the catalyst [0.1 µmol
4-Cholesten-3-ol (2c): tR ϭ 3.96 min (dehydration occurs in the in-
of catalyst per mL of CH2Cl2/CH3CN (1:1)] and stirred at room jector). MS (EI): m/z (%) ϭ 368 (100) [Mϩ· Ϫ 18], 353 (50). 1H
temperature. Aqueous hydrogen peroxide (30% w/w) diluted with
acetonitrile (1:20) was added to the reaction mixture in 30 µL ali-
quots every 15 min.
NMR (300.13 MHz): δ ϭ 0.68 (s, 3 H, 18-H), 0.862 and 0.865 (2
d, J ϭ 6.6 Hz, 3 ϫ 2 H, 26-H and 27-H), 0.90 (d, J ϭ 6.1 Hz, 3
H, 21-H), 1.00 (s, 3 H, 19-H), 4.05Ϫ4.11 (m, 1 H, 3-H), 5.46 (dd,
J ϭ 1.5, 4.8 Hz, 1 H, 4-H) ppm. 13C NMR (75.47 MHz): δ ϭ 64.3
(C-3), 120.6 (C-4), 150.5 (C-5) ppm.
Oxidation Reactions with FeIII Porphyrins: In a typical experiment,
the substrate (0.03 mmol) was dissolved in 2 mL of a stock solution
of the catalyst [0.1 µmol of catalyst per mL of CH3OH/CH2Cl2
(3:1)] and stirred at room temperature. Aqueous hydrogen peroxide
(30% w/w) diluted with acetonitrile (1:20) was added to the reaction
mixture in 30 µL aliquots every 15 min.
4-Cholesten-3-one (2d): tR ϭ 6.35 min. MS (EI): m/z (%) ϭ 384 (40)
1
[Mϩ·], 342 (20). H NMR (300.13 MHz): δ ϭ 0.67 (s, 3 H, 18-H),
0.86 (d, J ϭ 6.6 Hz, 6 H, 26-H and 27-H), 0.91 (d, J ϭ 6.4 Hz, 3
H, 21-H), 1.05 (s, 3 H, 19-H), 5.72 (s-broad, 1 H, 4-H) ppm. 13C
NMR (75.47 MHz): δ ϭ 123.7 (C-4), 171.8 (C-5), 119.7 (C-3) ppm.
Reaction Control: Aliquots were withdrawn from the reaction mix-
ture and injected directly into the GC injector. The addition of
H2O2 was stopped when the relative proportion of the compounds
remained constant after two successive GC analyses. Reaction con-
versions and product yields were based on the 1H NMR spectra of
the reaction mixture. This type of approach was possible because
the various components of the reaction mixture exhibit distinct sig-
nals between δ ϭ 2 and 6 ppm. Only the H-4 signals of the two
isomers of 4,5-epoxides (Scheme 1) are not distinguishable in that
region of the 1H NMR spectra; their relative abundance was deter-
3-Hydroxy-4,5-epoxycholestane (2e): tR ϭ 6.30 min (dehydration
occurs in the injector). MS (EI): m/z (%) ϭ 384 (10) [Mϩ· Ϫ 18].
1H NMR (500.13 MHz): δ ϭ 0.68 (s, 3 H, 18-H), 0.868 and 0.872
(2 d, J ϭ 6.6 Hz, 2 ϫ 3 H, 26-H and 27-H), 0.91 (d, J ϭ 6.5 Hz,
3 H, 21-H), 1.00 (s, 3 H, 19-H), 1.03Ϫ1.09 (m, 1 H, 6-H),
1.05Ϫ1.07 (m, 1 H, 22-H), 1.05Ϫ1.12 (2m, 2 ϫ 1 H, 14-H and 17-
H), 1.08Ϫ1.18 (2m, 1H ϩ2 H, 12-H and 24-H), 1.14Ϫ1.18 (m, 1
H, 1-H), 1.28Ϫ1.32 (2m, 2 ϫ 1 H, 2-H and 9-H), 1.34Ϫ1.40 (2m,
2 ϫ 1 H, 20-H and 22-H), 1.36Ϫ1.40 (m, 1 H, 1-H), 1.41Ϫ1.46 (2m,
1H ϩ2 H, 8-H and 11-H), 1.70Ϫ1.73 (m, 1 H, 2-H), 1.74Ϫ1.79 (m,
2 H, 7-H), 2.00 (dt, J ϭ 3.3, 12.8 Hz, 1 H, 12-H), 2.03Ϫ2.13 (m, 1
H, 6-H), 2.85 (s, 1 H, 4-H), 3.97 (t, J ϭ 8.4 Hz, 1 H, 3-H) ppm.
13C NMR (125.76 MHz): δ ϭ 11.9 (C-18), 18.6 (C-21), 18.7 (C-19),
21.1 (C-11), 22.5 and 22.8 (C-26,27), 23.8 and 24.2 (C-15,16), 25.7
(C-2), 26.0 (C-1), 28.0 (C-25), 28.2 (C-23), 30.2 (C-7), 30.9 (C-6),
35.0 (C-8), 35.7 (C-20), 36.1 (C-22), 36.3 (C-10), 39.4 (C-24), 39.7
(C-12), 42.5 (C-13), 46.0 (C-9), 56.0 and 56.2 (C-14,17), 65.8 (C-
4), 66.8 (C-5,3) ppm.
1
mined by the intensity of the H NMR singlets of H-19 (δ ϭ 1.07
ppm for the α-epoxide and δ ϭ 1.01 ppm for the β-epoxide).
Isolation and Characterization of Reaction Products: The reaction
mixture components were separated by preparative thin-layer chro-
matography on silica gel. Oxidation products from 1 were eluted
with a mixture of CH2Cl2/light petroleum (4:1), whereas the oxi-
dation products from 2 were eluted with a mixture of CH2Cl2/light
petroleum (1:1). Crystals of epoxides 1a and 1b, suitable for X-ray
diffraction, were obtained by recrystallization from a mixture of
dichloromethane and methanol. Compounds 3a and 3b were iso-
lated by column chromatography on silica gel eluting with a (7:3)
mixture of CH2Cl2 and light petroleum.
3-β-Acetoxy-5,6-β-epoxycholestane (3a): tR ϭ 7.46 min (loss of
acetic acid occurs in the injector). MS (EI): m/z (%) ϭ 384 (14)
1
[Mϩ Ϫ 60], 356 (8). H NMR (300.13 MHz): δ ϭ 0.64 (s, 3 H, 18-
H), 0.858 and 0.861 (2 d, J ϭ 6.7 Hz, 2 ϫ 3 H, 26-H and 27-H),
0.89 (d, J ϭ 7.7 Hz, 3 H, 21-H), 1.00 (s, 3 H, 19-H), 2.03 (s, 3 H,
3-OAc), 3.08 (d, J ϭ 1.6 Hz, 1 H, 6-H), 4.71Ϫ4.82 (m, 1-H, 3-
H) ppm.
17β-Acetoxy-4,5β-epoxyandrostane (1a): Retention time (tR):
2.64 min. MS (EI): m/z (%) ϭ 332 (15) [Mϩ·], 304 (6). M.p.
156.4Ϫ157.6 °C. NMR spectroscopic data are given in Table 2.
3-β-Acetoxy-5,6-α-epoxycholestane (3b): tR ϭ 8.20 min (loss of
acetic acid occurs in the injector). MS (EI): m/z (%) ϭ 384 (13)
17β-Acetoxy-4,5α-epoxyandrostane (1b): tR ϭ 2.74 min. MS (EI):
m/z (%) ϭ 332 (26) [Mϩ·], 304 (15). M.p. 137.6Ϫ139.4 °C. NMR
spectroscopic data are given in Table 2.
1
[Mϩ· Ϫ 60], 366 (9). H NMR (300.13 MHz): δ ϭ 0.61 (s, 3 H, 18-
H), 0.860 and 0.861 (2 d, J ϭ 6.7 Hz, 2 ϫ 3 H, 26-H and 27-H),
0.89 (d, J ϭ 7.7 Hz, 3 H, 21-H), 1.07 (s, 3 H, 19-H), 2.02 (s, 3 H,
3-OAc), 2.90 (d, J ϭ 4.3 Hz, 1 H, 6-H), 4.89Ϫ5.00 (m, 1-H, 3-
H) ppm.
17β-Acetoxy-4-androsten-3-ol (1c): tR ϭ 2.45 min (dehydration oc-
curs in the injector). MS (EI): m/z (%) ϭ 314 (25) [Mϩ· Ϫ 18], 299
(5). H NMR (300.13 MHz): δ ϭ 0.81 (s, 3 H, 18-H), 0.99 (s, 3 H,
19-H), 2.04 (s, 3 H, 21-H), 4.04Ϫ4.09 (m, 1 H, 3-H), 4.58 (dd, J ϭ
7.9, 9.0 Hz, 1 H, 17-H), 5.47 (dd, J ϭ 1.6, 5.0 Hz, 1 H, 4-H) ppm.
1
Acknowledgments
Thanks are due to Fundac¸a˜o para a Ciencia e a Tecnologia (FCT)
for funding the Project 39593/Qui/2001 and also the Aveiro Organic
Chemistry Research Group. S. L. H. R. also thanks FCT for a
PhD grant.
17β-Acetoxy-4-androsten-3-one (1d): tR ϭ 3.57 min. MS (EI): m/z
ˆ
1
(%) ϭ 330 (10) [Mϩ·], 288 (15). H NMR (300.13 MHz): δ ϭ 0.84
(s, 3 H, 18-H), 0.99 (s, 3 H, 19-H), 2.05 (s, 3 H, 21-H), 4.58 (dd,
J ϭ 7.9, 9.0 Hz, 1 H, 17-H), 5.74 (d, J ϭ 1.1 Hz, 1 H, 4-H) ppm.
17β-Acetoxy-3-hydroxy-4,5-epoxyandrostane (1e): tR ϭ 3.45 min
(dehydration occurs in the injector). MS (EI): m/z (%) ϭ 330 (15)
[1]
P. M. Dewick, in Medicinal Natural Products, a Biosynthetic
1
[Mϩ· Ϫ 18]. H NMR (300.13 MHz): δ ϭ 0.81 (s, 3 H, 18-H), 1.01
Approach, John Wiley & Sons, Chichester, 1997.
[2]
(s, 3 H, 19-H), 2.05 (s, 3 H, 21-H), 2.85 (s, 1 H, 4-H), 3.98 (t, J ϭ
5.6 Hz, 1 H, 3-H), 4.60 (t, J ϭ 8.4 Hz, 1 H, 17-H) ppm.
A. M. H. Brodie, V. C. O. Njar, Steroids 2000, 65, 171.
[3]
J. Mc Lain, J. Lee, J. T. Groves, in Biomimetic Oxidations Cata-
lysed by Transition Metal Complexes (Ed.: B. Meunier), Im-
perial College Press, London, 2000, p. 127.
4,5β-Epoxycholestane (2a): tR ϭ 4.47 min. MS (EI): m/z (%) ϭ 386
[4]
(100) [Mϩ·], 358 (35). NMR spectroscopic data are given in Table 4.
M. Numazawa, K. Yamada, Steroids 1999, 64, 320.
[5]
´
Z. Freixa, M. M. Pereira, J. C. Bayon, A. M. S. Silva, J. A. R.
4,5α-Epoxycholestane (2b): tR ϭ 4.84 min. MS (EI): m/z (%) ϭ 386
Salvador, A. M. Beja, J. A. Paixa˜o, M. Ramos, Tetrahedron:
Asymmetry 2001, 12, 1083.
(100) [Mϩ·], 358 (35). NMR spectroscopic data are given in Table 4.
4786
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4778Ϫ4787

Catalytic Oxidation of ∆4- and ∆5-Steroids with Hydrogen Peroxide
FULL PAPER
C.-P. Du, Z.-K. Li, X.-M. Wen, J. Wu, X.-Q. Yu, M. Yang, R.-
G. Xie, J. Mol. Catal. A: Chem. 2004, 216, 7.
S. L. H. Rebelo, M. M. Q. Simo˜es, M. G. P. M. S. Neves, J. A.
S. Cavaleiro, J. Mol. Catal. A: Chem. 2003, 201, 9.
A. M. A. R. Gonsalves, R. A. W. Johnstone, M. M. Pereira, J.
Shaw, J. Chem. Soc., Perkin Trans. 1 1991, 645.
W. Nam, H. J. Han, S.-Y. Oh, Y. J. Lee, M.-H. Choi, S.-Y.
Han, C. Kim, S. K. Woo, W. Shin, J. Am. Chem. Soc. 2000,
122, 8677.
N. Jin, J. T. Groves, J. Am. Chem. Soc. 1999, 121, 2923.
F. Ogliaro, S. P. Visser, J. T. Groves, S. Shaik, Angew. Chem.
Int. Ed. 2001, 40, 2874.
G. Strukul, in Catalytic Oxidations with Hydrogen Peroxide as
Oxidant (Ed.: G. Strukul), Kluwer Academic Press, Dordrecht,
1992, p. 178.
P. A. MacFaul, I. W. C. E. Arends, K. U. Ingold, D. D. M.
Wayner, J. Chem. Soc., Perkin Trans. 2 1997, 135.
P. A. MacFaul, K. U. Ingold, D. D. M. Wayner, L. Que Jr., J.
Am. Chem. Soc. 1997, 119, 10594.
[6]
[26]
[27]
[28]
[29]
J. A. Boynton, J. R. Hanson, C. Uyanik, J. Chem. Res. (S)
1995, 334.
[7]
H. L. Holland, Jahangir, Can. J. Chem. 1983, 61, 2165.
R. J. Ryan, S. Julia, Tetrahedron 1973, 29, 3649.
D. H. R. Barton, Y. Houminer, J. Chem. Soc., Chem. Commun.
[8]
[9]
1973, 839.
[10]
B. N. Blackett, J. M. Coxon, M. P. Hartshorn, K. E. Richards,
Tetrahedron 1969, 25, 4999.
Y. Houminer, J. Chem. Soc., Perkin Trans. 1 1975, 1663.
[11]
[30]
[31]
[12]
´
´
P. Kocovshy, I. Stary, J. Org. Chem. 1990, 55, 3236.
[13]
[14]
[15]
[16]
[17]
D. K. Fukushima, M. Smulowitz, J. S. Liang, G. Lukacs, J.
Org. Chem. 1969, 34, 2702.
M. Tohma, T. Tomita, M. Kimura, Tetrahedron Lett. 1973,
44, 4359.
J. R. Hanson, P. B. Hichcock, M. D. Liman, S. Nagaratnam,
R. Manickavasagar, J. Chem. Res. (S) 1995, 220.
J. R. Hanson, S. Nagaratnam, J. Stevens, J. Chem. Res. (S)
1996, 102.
B. Meunier, A. Robert, G. Pratviel, J. Bernadou, in The Porphy-
rin Handbook, vol. 4, Academic Press, New York, 2000, p. 119.
W. R. Sanderson, Pure Appl. Chem. 2000, 72, 1289.
P. Battioni, J. P. Renaud, J. F. Bartoli, M. Reina-Artiles, M.
[32]
[33]
[34]
[35]
[36]
S. L. H. Rebelo, M. M. Q. Simo˜es, M. G. P. M. S. Neves, A.
M. S. Silva, J. A. S. Cavaleiro, Chem. Commun. 2004, 608.
A. M. Khenkin, A. A. Shteinman, J. Chem. Soc., Chem. Com-
mun. 1984, 1219.
[18]
[19]
[37]
[38]
J. R. Bull, J. Floor, J. Chem. Soc., Perkin Trans. 1 1977, 724.
´
M. M. Pereira, A. Peixoto, C. Foca, J. C. Bayon, A. M. S. Silva,
Fort, D. Mansuy, J. Am. Chem. Soc. 1988, 110, 8462.
A. Thellend, P. Battioni, D. Mansuy, J. Chem. Soc., Chem.
Commun. 1994, 1035.
T. Shingaki, K. Miura, T. Higuchi, M. Hirobe, T. Nagano,
Chem. Commun. 1997, 861.
[20]
[21]
[22]
M. J. Moreno, unpublished results.
[39]
[40]
A. M. A. R. Gonsalves, J. M. T. B. Vareja˜o, M. M. Pereira, J.
Heterocycl. Chem. 1991, 28, 635.
J. W. Buchler, in The Porphyrins (Ed.: D. Dolphin), Academic
Press, New York, 1978, vol. 1, p. 398 and references cited ther-
ein.
`
M. Tavares, R. Ramasseul, J.-C. Marchon, B. Bachet, C.
Brassy, J.-P. Mornon, J. Chem. Soc., Perkin Trans. 2 1992, 1321.
J. C. Marchon, R. Ramasseul, Synthesis 1989, 389.
[41]
[42]
[43]
[44]
M. Autret, Z. Ou, A. Antonini, T. Boschi, P. Tagliatesta, K.
M. Kadish, J. Chem. Soc., Dalton Trans. 1996, 2793.
A. D. Adler, F. R. Longo, F. Kampes, J. J. Kim, Inorg. Nucl.
Chem. 1970, 32, 2443.
[23]
[24]
J.-C. Marchon, R. Ramasseul, J. Chem. Soc., Chem. Commun.
1988, 298.
S. L. H. Rebelo, M. M. Pereira, M. M. Q. Simo˜es, M. G. P. M.
S. Neves, A.-M. S. Silva, J. A. S. Cavaleiro, ‘‘Metalloporphyrins
in steroids’ catalytic oxidation’’, 13th International Symposium
on Homogeneous Catalysis, Tarragona (Spain), 3Ϫ7th Sep-
tember 2002, p. 234.
[25]
´
J. P. Collman, L. Zeng, R. A. Decreau, Chem. Commun. 2003,
2974 and references cited therein.
C. K. Johnson, ORTEP-II, Report ORNL-5138, Oak Ridge
National Laboratory, Tennessee, USA, 1976.
Received July 2, 2004
Eur. J. Org. Chem. 2004, 4778Ϫ4787
www.eurjoc.org
© 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4787