Synthesis and isomerization of arene oxide metabolites of phenanthrene, triphenylene, dibenz[a,c]anthracene and dibenz[a,h]anthracene

Article abstract of DOI:10.1039/b009712h

Dibenz[a,h]anthracene 3,4-oxide 5A_RS, synthesised from the enantiopure dibromoMTPA precursor 9A_RRS*, was found to have totally racemized and to be accompanied by benz[5,6]anthra[1,2-b]oxepine 11A. Phenanthrene 3,4-oxide 5B_RS

Full text of DOI:10.1039/b009712h

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis and isomerization of arene oxide metabolites of
phenanthrene, triphenylene, dibenz[a,c]anthracene and
dibenz[a,h]anthracene
1
Suresh K. Balani, I. N. Brannigan, Derek R. Boyd,* Narain D. Sharma, Francis Hempenstall
and Allison Smith
School of Chemistry, The Queen’s University of Belfast, Belfast, UK BT9 5AG
Received (in Cambridge, UK) 4th December 2000, Accepted 21st February 2001
First published as an Advance Article on the web 22nd March 2001
Dibenz[a,h]anthracene 3,4-oxide 5ARS, synthesised from the enantiopure dibromoMTPA precursor 9ARRS*, was found
to have totally racemized and to be accompanied by benz[5,6]anthra[1,2-b]oxepine 11A. Phenanthrene 3,4-oxide
5BRS, obtained from the enantiopure bacterial metabolite cis-3,4-dihydroxy-3,4-dihydrophenanthrene 12B by a
modified synthetic approach involving the chlorohydrin ester 16B, was observed to racemize spontaneously at
ambient temperature. Dibenz[a,h]anthracene 3,4-oxide 5ARS/5ASR, phenanthrene 3,4-oxide 5BRS/5BSR, triphenylene
1,2-oxide 5CRS/5CSR, and dibenz[a,c]anthracene 1,2-oxide 5DRS/5DSR, obtained from the corresponding racemic
cis-tetrahydrodiol precursors 14A–14D by the new method, were obtained without any evidence of the formation
of benz[5,6]anthra[1,2-b]oxepine 11A, naphth[1,2-b]oxepine 11B, phenanthro[10,9-b]oxepine 11C, or benz[3,4]-
anthra[1,2-b]oxepine 11D isomers respectively. The total racemization of arene oxide 5ARS and formation of oxepine
11A from the bromoMTPA precursor 8ARRS* are in accord with earlier PMO predictions based on resonance energy
considerations. Photoisomerization of arene oxides 5ARS/5ASR, 5CRS/5CSR, and 5DRS/5DSR was found to yield the
corresponding oxepines 11A, 11C, and 11D.
Introduction
Polycyclic aromatic hydrocarbons (PAHs) occur widely in the
environment and can be cytotoxic, mutagenic and carcinogenic
when activated via monooxygenase-catalysed arene and alkene
epoxidations. The initially formed epoxide (arene oxide) inter-
mediates are generally very difficult to detect due to their pro-
Scheme 1
pensity to aromatize forming the corresponding phenols. The
benzene oxide metabolite 1, resulting from enzyme-catalysed
epoxidation of 2-trifluoromethyl methylbenzoate in the fungus
Phellinus ribis, was thus unusual, being isolable as a result of its
remarkable stability.1 By contrast, naphthalene 1,2-oxide 2, the
first reported example of a polycyclic arene oxide metabolite,
was much less stable and could only be detected by using radio-
chemical tracer methods.2
and that spontaneous racemization of some polycyclic arene
oxides could occur through an electrocyclic rearrangement to
yield undetected oxepine valence tautomers 6 (Scheme 2).3,4,8
The formation of an arene oxide during metabolism could
in principle involve facial stereoselectivity yielding either enan-
tiomer.3,4 Since the rates of further enzyme-catalysed reactions
and biological activity are often dependent upon absolute
configuration, it is important to determine if an arene oxide
enantiomer is configurationally stable. In earlier reports we
have shown that benzene oxide enantiomers e.g. chlorobenzene
2,3-oxide, 3, spontaneously racemize via the corresponding
oxepine valence tautomers e.g. oxepine, 45 (Scheme 1). Con-
versely the configurational stability of naphthalene 1,2-oxide 2
was established by the synthesis of either enantiomer without
evidence of spontaneous racemization.6,7 Synthetic studies on
polycyclic arene oxides 5 derived from single enantiomer
precursors showed that configurational stability was variable
Scheme 2
Perturbational molecular orbital (PMO) calculations of the
arene oxide–oxepine equilibrium, and the associated loss of
resonance energies, allowed the relative ease of racemization
of particular arene oxide structures to be predicted.8 Previous
synthetic studies have provided experimental confirmation
of the PMO-based prediction that monooxygenase-catalysed
epoxidation of both tricyclic, e.g. phenanthrene (3,4-oxide
5B),8–10 and tetracyclic arenes, e.g. triphenylene (1,2-oxide 5C),11
could yield arene oxides which are prone to spontaneous
DOI: 10.1039/b009712h
J. Chem. Soc., Perkin Trans. 1, 2001, 1091–1097
This journal is © The Royal Society of Chemistry 2001
1091

View Article Online
racemisation. The same PMO method8 also led to the predic-
tion that several pentacyclic arene oxides, e.g. 5A–5D, would
spontaneously racemize.
In this report we present (i) the synthesis and spontaneous
total racemization of a pentacyclic arene oxide (dibenz[a,h]-
anthracene 1,2-oxide 5ARS) and the anticipated simultaneous
formation of the corresponding oxepine isomer 11A from
an enantiopure bromoester precursor (8RRS*) using the con-
ventional synthetic route (Scheme 3). (ii) The synthesis and
observed spontaneous racemisation of a tricyclic arene oxide
(phenanthrene 3,4-oxide 5BRS/5BSR), derived from an enantio-
pure cis-dihydrodiol bacterial metabolite (phenanthrene cis-
3S,4R-dihydrodiol 12B), using an alternative synthetic pathway
(Scheme 4). (iii) The synthesis of pure samples of a tetracyclic
Prior to the preliminary report of this work,12 the preferred
method of synthesis of enantiopure polycyclic arene oxides has
involved the formation and resolution of bromo-α-methoxy-
α-(trifluoromethyl)phenylacetate (bromoMTPA) esters 8 and
conversion of the resulting dibromoester intermediates 9 to
arene oxides 5 via an intramolecular SN2 mechanism as shown
in Scheme 3.10 However, when tetra- or penta-cyclic arenes
Scheme 4 i, Pd/C, H2, MeOH; ii, MeC(OMe)3; iii Me3SiCl, CH2Cl2,
Et3N; iv NBS, AIBN, CCl4; v NaOMe, THF; vi OsO4, NMO, CH2Cl2 .
arene oxide (triphenylene 1,2-oxide 5CRS/5CSR) and two penta-
cyclic arene oxides (dibenz[a,h]anthracene 3,4-oxide 5ARS
/
5ASR, and dibenz[a,c]anthracene 1,2-oxide 5DRS/5DSR) from
racemic cis-tetrahydrodiol precursors (13C, 13A, and 13D)
without formation of the corresponding stable oxepines (11C,
11A, and 11D) using a new method (Scheme 4) and (iv) the
photoisomerization of arene oxides 5ARS/5ASR, 5CRS/5CSR and
5DRS/5DSR to yield the corresponding oxepines 11A, 11C, and
11D.
Results and discussion
The established synthetic method, used earlier for the synthesis
of enantiopure polycyclic arene oxides,10 developed in these
laboratories, was itself a modification of the method previously
used by Yagi and Jerina14 in the synthesis of racemic samples
using resolved bromohydrin ester precursors (Scheme 3). To
date this method10 has been successfully applied to the syn-
thesis of single enantiomers of three categories of arene oxides.
The first group included the enantiopure arene oxides, 5, of
naphthalene (1,2-),6 anthracene (1,2-),6 benz[a]anthracene (8,9-,
10,11-),15,16 and benzo[a]pyrene (7,8-)17 which were found to be
configurationally stable at ambient temperature. The second
type included the arene oxides, 5, of phenanthrene (1,2- and
3,4-),9,10 chrysene (1,2- and 3,4-),18,19 and benzo[c]phenanthrene
Scheme 3 i, (Ϫ)-MTPA Cl, pyridine; ii NBS–CCl4; iii NaOMe–THF.
are synthesised from the dibromoester intermediate 9 the latter
approach has often resulted in the concomitant formation of an
isolable oxepine isomer 11. The reaction is assumed to proceed
via a competing SN2Ј mechanism involving the unstable angular
arene oxide intermediate 10 which remained undetected
(Scheme 3).8,13 It was also found that the formation and
isolation of these relatively stable oxepine intermediates, 11,
could be predicted using the PMO method.8,13
1092
J. Chem. Soc., Perkin Trans. 1, 2001, 1091–1097

View Article Online
The absolute configurations were assigned using the reported
NMR method based upon the 1H chemical shift values of
the H-3 signal, the 19F chemical shift values of CF3, and the
coupling constants J3,4.
10 Thus the less polar bromoMTPA ester
(8ARRS*) having the smaller coupling constant (J3,4), the smaller
1H chemical shift (δH-3) and the smaller negative 19F chemical
shift value for CF3 (δF) was assigned the 3R,4R configuration.
Benzylic bromination of the 8ARRS* diastereoisomer using NBS
yielded a relatively unstable mixture of dibromoMTPA esters,
9ARRS*. Treatment of this mixture with sodium methoxide in
THF yielded dibenz[a,h]anthracene 3,4-oxide 5ARS/5ASR as the
major component (ca: 65%) and the relatively stable oxepine
11A as the minor component (35%). While the sample of arene
oxide 5ARS/5ASR could not be chromatographically separated
from oxepine 11A without decomposition, the optical rotation
value of zero from the crude mixture was consistent with the
predicted initial formation of the 5ARS enantiomer followed by
total spontaneous racemization (proceeding via the undetected
unstable oxepine 6). When the synthesis was repeated using a
racemic sample of dibromoacetate, derived from bromohydrin
7, a racemic mixture of arene oxide 5ARS/5ASR (ca. 40%) and
stable oxepine 11A (ca. 60%) was again formed. The arene
oxide enantiomers 5ARS/5ASR decomposed to phenols during
chromatography; a pure sample of the more stable oxepine 11A
was isolated. Further evidence of decomposition was found
when the mixture of arene oxide 5ARS/5ASR and oxepine 11A
was dissolved in CDCl3 solution containing a trace of benzoic
acid. The arene oxide isomerized to the corresponding phenols
leaving the residual oxepine 11A unchanged.
(1,2-)20 which were initially obtained in enantiopure form but
were observed to racemize slowly at ambient temperature via
undetected oxepine isomers, 6. The final category included arene
oxides, 5, of triphenylene (1,2-),21 benzo[e]pyrene (9,10-),22
dibenz[a,j]anthracene (3,4-),23 and benzo[g]chrysene (5,6-,
7,8-),24 benz[a]anthracene (1,2-, 3,4),21 which were all assumed
to have been synthesised as single enantiomers, were found to
have totally racemized (via unstable oxepine valence tautomers
6) and were also accompanied by isolable oxepine isomers 11.
The loss of resonance energy associated with the isomerization
of the third type of arene oxides 5 to the unstable oxepines 6
was calculated8 to be much lower than for the first group and
slightly lower than for the second category. The simultaneous
formation of arene oxide derivatives, 5, of dibenz[a,c]anthra-
cene (1,2- and 3,4-) and the corresponding stable oxepine
isomers, 11, from racemic dibromoacetate precursors has also
been reported.25,26
No evidence for interconversion of the arene oxide 5ARS
/
5ASR to oxepine 11A was obtained when a CDCl3 solution of
the mixture was heated at 50 ЊC or maintained at ambient
temperature for an extended period (>24 h). However, UV
irradiation (>300 nm) of the mixture of arene oxide 5ARS/5ASR
and oxepine 11A at ambient temperature for a short period
(0.5 h), showed a concomitant increase in the proportion of
oxepine (ca. 50%) and decrease in the amount of arene oxide. A
similar type of photochemical circumambulatory rearrange-
ment or “oxygen walk” process was observed for other arene
oxide derivatives among tetra-and penta-cyclic members of the
PAH series.21–25 This observation clearly indicates that oxepine
11A is formed during the synthesis of arene oxide 5ARS/5ASR
from either dibromoacetate or dibromoMTPA (9) precursors.
It was assumed that oxepine 11A was formed via an SN2Ј
mechanism while the arene oxide 5A resulted from the normal
SN2 pathway (Scheme 3). Thus, despite the earlier report on the
synthesis of uncontaminated arene oxide 5A,27 the formation
of oxepine 11A in the present study has allowed our earlier
prediction13 about the concomitant formation of arene oxides
and oxepines of this type, based on PMO calculations, to be
confirmed. In view of the current results it appears that oxepine
11A had escaped detection in the earlier study.27
Resulting from the importance of arene oxides and their
derivatives in mechanism studies of chemically induced carcino-
genesis caused by larger members of the PAH series,38,39 and
the difficulties experienced earlier in obtaining pure samples
of arene oxides 5 of the tetracyclic series (triphenylene,21 benz-
[a]anthracene21), and pentacyclic (benzo[e]pyrene,22 dibenz[a,j]-
anthracene,23 dibenz[a,c]anthracene,25 and benzo[g]chrysene24)
without the accompanying oxepines 11, an alternative to the
bromoester method shown in Scheme 3 was explored.
All of the earlier PMO predictions8,13 have to date proved to
be in accord with experimental observations on the relative
ease of racemization of the arene oxides 5 and the concomitant
formation of stable oxepine isomers 11 from these laboratories.
Thus, it was surprising to find a report27 that the 1,2- and 3,4-
arene oxides of dibenz[a,h]anthracene, synthesised using the
standard dibromoacetate route,14 were unaccompanied by the
corresponding stable oxepines, 11. This apparent exception to
our predictions13 has been reinvestigated as part of our current
programme to utilize arene cis-dihydrodiol metabolites in the
synthesis of arene oxides.5,7 A wide range of cis-dihydrodiol
metabolites of the PAHs including naphthalene,28,29 anthra-
cene,30,31 phenanthrene,30,32 benz[a]anthracene,33,34 chrysene,35
and benzo[a]pyrene33 and heteroPAHs36 are available from
bacterial biotransformations.
PAHs constitute the first class of compounds to be confirmed
as chemical carcinogens,37 and dibenz[a,h]anthracene (DBA)
was the first carcinogenic member to be identified.38 The arene
oxide metabolite 5ARS/5ASR has been proposed as one of three
arene oxide intermediates formed during liver microsomal
metabolism of DBA.27 Enzyme-catalysed hydrolysis of this
arene oxide has in turn been shown to yield the corresponding
trans-dihydrodiol (trans-3,4-dihydroxy-3,4-dihydrodibenz[a,h]-
anthracene), which has been identified as a precursor of the
ultimate mutagenic and carcinogenic diol epoxide and tetraol
epoxide metabolites of DBA.39
The synthetic sequence shown in Scheme 4 is based upon the
availability of cis-tetrahydrodiol precursors 14. The tricyclic
tetrahydrodiol (3S,4R)-14B was derived by catalytic hydrogen-
ation of the (3S,4R)-enantiopure cis-dihydrodiol metabolite
12B, obtained from biotransformation of the corresponding
PAH (phenanthrene) using mutant strains of the bacteria
Sphingomonas yanoikuyae (B8/36)32 and Pseudomonas putida
(NCIB 9816/11).40 Osmylation of the dihydroarenes 13A, 13C,
and 13D, synthesised by literature methods,11,26,27 gave the
The first synthetic route to arene oxide 5ARS/5ASR was based
on the use of bromohydrin 7ARR/7ASS as a precursor (Scheme
3). Compound 7A, obtained from the corresponding bromo-
acetate,27 was treated with (Ϫ)-α-methoxy-α-(trifluoromethyl)-
phenylacetyl chloride (MTPA-Cl derived from S-MTPA and
identified by S*) to yield a mixture of bromoMTPA diastereo-
isomers, 8ARRS* and 8ASSS*. This mixture was separated by PLC
into the less polar (8ARRS*) and more polar (8ASSS*) fractions.
J. Chem. Soc., Perkin Trans. 1, 2001, 1091–1097
1093

View Article Online
corresponding racemic cis-tetrahydrodiol precursors 14A, 14C,
and 14D in variable yields (15–80%) (Scheme 4).
13A, 13C, and 13D. An earlier observation using the enantio-
pure trans-bromoMTPA ester 8C,10 allied to results obtained in
the present study using trans-bromoMTPA ester 8A, has
allowed the predicted8,13 spontaneous racemization of the
derived arene oxides 5ARS/5ASR and 5DRS/5DSR to be confirmed
(in the presence of the corresponding oxepines isomers 11A and
11C).
Reaction of the tetrahydrodiols, 14A–14D, with trimethyl-
orthoacetate and a catalytic amount of benzoic acid pro-
vided the corresponding 1,3-dioxole derivatives 15A–15D, as
diastereoisomeric mixtures (63–82% yield) whose structures
1
were confirmed by H NMR and MS analysis. Separation of
the diastereoisomeric pairs 15A–15D was considered unneces-
sary as both gave the same chloroacetate products 16A–16D
in the next stage of the synthesis. The isomeric mixtures were
considered to be of sufficient purity to be used without
chromatography.
Photoisomerization of the pure arene oxides 5ARS/5ASR
,
5CRS/5CSR 5DRS/5DSR yielded the corresponding isolable
,
oxepines 11A, 11C and 11D which were in turn found to yield
phenolic products upon prolonged exposure to UV light. No
evidence of oxepine 5B was obtained from UV irradiation of
phenanthrene 3,4-oxide 5BRS/5BSR under similar conditions.
Reaction of the dioxolane diastereoisomers 15A–15D with
chlorotrimethylsilane in the presence of a trace of triethylamine
yielded the corresponding trans-chloroacetates 16A–16D.
This reaction was earlier applied to the PAH bond of highest
electron density (the 5,6- K-region band). The K-region arene
cis-dihydrodiols were used for the synthesis of enantiopure
trans-chloroacetate precursors of the K-region arene oxides
of chrysene, 7,12-dimethylbenz[a]anthracene and benzo[c]-
phenanthrene.41 A qualitative comparison of the relative stabili-
ties of the trans-halohydrin ester types exemplified by com-
pounds 8 (containing a non-benzylic halogen, Scheme 3) and
16A–16D (containing a benzylic halogen, Scheme 4) showed
that the latter category was less stable. The trans-chloroacetates
16A–16D, in common with other compounds bearing a halogen
atom at a benzylic position e.g. compounds 9A, 17A–17D, were
found to decompose during attempted purification during
silica-gel chromatography. Fortunately these compounds were
obtained in good yields (ca. 70%) and were of sufficient purity
(1H NMR analysis) for using directly in the next stage.
Treatment of the trans-chloroacetates 16A–16D with N-
bromosuccinimide in tetrachloromethane yielded the corre-
sponding bromo trans-chloroacetates 17A–17D as isomeric
mixtures. Compounds 17A–17D appeared to be relatively stable
during their synthesis and isolation (1H NMR spectral
analysis). However, in order to avoid decomposition during
chromatography they were converted directly to the corre-
sponding arene oxides 5ARS/5ASR–5DRS/5DSR by treatment with
sodium methoxide without chromatographic separation. The
Conclusion
Arene oxide enantiomers 5ARS/5ASR and oxepine 11A have been
simultaneously synthesised via the trans-bromoMTPA ester
intermediate 8ARRS* (Scheme 3) in contrast to an earlier report.
An alternative synthetic route to arene oxides 5ARS/5ASR–5DRS
/
5DSR via the corresponding cis-tetrahydrodiols, 14A–14D, and
trans-chloroacetates, 16A–16D involves exclusively an SN2
pathway without formation of oxepine by-products 11A–11D
(Scheme 4). This is, in our view, the preferred method for the
synthesis of pure arene oxide metabolites of PAHs. The roles of
two distinct types of oxepines, 6 and 11, have been established
in the context of arene oxide chemistry of PAHs. Thus, spon-
taneous racemization via the unstable and unobserved oxepines,
6A and 6B, has been found to give arene oxides 5ARS/5ASR and
5BRS/5BSR while photoisomerization of arene oxides 5ARS/5ASR
,
5CRS/5CSR, and 5DRS/5DSR was found to yield the stable and
isolable oxepines 11A, 11C and 11D.
Experimental
1H NMR spectra were recorded at 250 MHz (Bruker WH250),
300 MHz (Bruker Avance DPX-500) and at 500 MHz (Bruker
Avance DRX-500) in CDCl3 solvent unless stated otherwise.
Chemical shifts (δ) are reported in ppm relative to SiMe4
and coupling constants (J) are given in Hz. Mass spectra
were recorded at 70 eV on a VG Autospec Mass Spectrometer,
using a heated inlet system. Accurate molecular weights were
determined by the peak matching method with perfluoro-
kerosene as standard. Elemental microanalyses were obtained
on a Perkin-Elmer 2400 CHN microanalyser. Optical rotations
were recorded at ambient temperature and are given in 10Ϫ1 deg
synthesis of the unstable arene oxides 5BRS/5BSR, 5CRS/5CSR
,
and 5DRS/5DSR from the corresponding cis-tetrahydrodiol
precursors 14B, 14C, 14D using the route shown in Scheme 4
eliminates the need for separation from the stable oxepines 11B,
11C,20 11D25,26 which are by-products from the method shown
in Scheme 3. The unstable oxepine tautomers 6A–6C, while
not detected, were assumed to be present in solution and to be
responsible for the spontaneous racemization of arene oxides
5A, 5B, and 5C derived from enantiopure trans-bromoMTPA
ester precursors 8A, 8B,9,10 and 8C.21
Since the tetrahydrodiol precursor 14B ([α]D Ϫ62, CHCl3)
was obtained by hydrogenation of the enantiopure cis-dihydro-
diol bacterial metabolite of phenanthrene 12B ([α]D +35,
MeOH), the product phenanthrene 3,4-oxide 5BRS must be
a single enantiomer initially. However, based on the PMO
calculations8 and its earlier synthesis from an enantiopure
trans-bromoMTPA precursor,9,10 a degree of racemization was
expected. The sample of phenanthrene 3,4-oxide obtained after
recrystallization at Ϫ70 ЊC initially showed a significant optical
rotation ([α]D +165, CDCl3) which decreased progressively to
zero over 24 h at ambient temperature. Simultaneous 1H NMR
analysis of this sample of phenanthrene 3,4-oxide 5B over the
same period showed no decomposition and hence it was con-
cluded that spontaneous racemization (yielding enantiomers
5BRS/5BSR) was solely responsible for the decrease in [α]D value.
The question of spontaneous racemization of arene oxides
cm2
g
Ϫ1. PLC was carried out on glass plates (20 × 20 cm)
coated with Merck Kieselgel PF254+356
.
The dihydroPAHs 1,2-dihydrodibenz[a,h]anthracene 13A,27
3,4-dihydrotriphenylene 13C,11 and 3,4-dihydrobenzo[a,c]-
anthracene 13D26 and ( )-trans-1-acetoxy-2-bromo-1,2,3,4-
tetrahydrodibenz[a,h]anthracene27 were obtained using the
literature methods.
( )-trans-3-Bromo-4-hydroxy-1,2,3,4-tetrahydrodibenz[a,h]-
anthracene (7A)
To a stirring solution of ( )-trans-1-acetoxy-2-bromo-1,2,3,4-
tetrahydrodibenz[a,h]anthracene (0.4 g, 0.72 mmol), in dry
THF (4 cm3) under N2, was added dropwise diborane–THF
solution (8 cm3, 1.0 M solution) and the reaction mixture was
stirred for a further 20 h. Water (25 cm3) was added and the
quenched reaction mixture extracted with chloroform (2 × 50
cm3). PLC purification (CHCl3–MeOH; 98 : 2) of the crude
product, obtained after removal of chloroform, yielded bromo-
hydrin 7A (0.26 g, 70%), mp 175 ЊC (from CH2Cl2) (Found: M+,
from the larger PAHs, dibenz[a,h]anthracene 3,4-oxide 5ARS
/
1
5ASR, triphenylene 1,2-oxide 5CRS/5CSR, and dibenz[a,c]anthra-
cene 1,2-oxide 5DRS/5DSR could not be addressed as these were
synthesised from the corresponding racemic tetrahydrodiols
376.0465. C22H17BrO requires M+, 376.0463); H NMR (250
MHz, THF-d8–CDCl3, 1 : 1) δ 2.50 (2 H, m, H-2), 3.30 (2 H, m,
H-1), 4.49 (1 H, m, H-3), 4.88 (1 H, d, J3,4 4.7, H-4), 7.3–8.0 (7
1094
J. Chem. Soc., Perkin Trans. 1, 2001, 1091–1097

View Article Online
H, m, Ar-H), 8.5 (1 H, s, Ar-H), 8.82 (1 H, d, J 4.7, Ar-H), 9.18
(14B). Typical procedure: a solution of (+)-(3S,4R)-3,4-di-
(1 H, s, Ar-H).
hydroxy-3,4-dihydrophenanthrene (12B), (0.2 g, 0.93 mmol;
[α]D +35, c = 0.4 in MeOH), isolated from the metabolism of
phenanthrene using a mutant strain (B 8/36) of the bacterium
Sphingomonas yanoikuyae,28 in methanol (15 cm3), was stirred
(24 h) with 10% Pd/C catalyst (0.02 g) under hydrogen at
atmospheric pressure. The filtrate, on concentration under
reduced pressure yielded (Ϫ)-(3S,4R)-3,4-dihydroxy-1,2,3,4-
tetrahydrophenanthrene (14B) as a white crystalline solid
(0.19 g, 94%); mp 190–193 ЊC (from CHCl3); [α]D Ϫ62 (c = 0.5
in CHCl3) (Found: C, 78.1; H, 6.4. C14H18O2 requires C, 78.5;
(Ϫ)-(3R,4R) and (؉)-(3S,4S)-3-Bromo-1,2,3,4-tetrahydrodi-
benz[a,h]anthracen-4-yl 3,3,3-trifluoro-2-methoxy-2-phenyl-
propionate (8ARRS* and 8ASSS*
)
Bromohydrin 7A (0.4 g, 1.03 mmol) dissolved in pyridine
(1 cm3) was converted into the corresponding bromoMTPA
esters 8ARRS* and 8ASSS* by treatment with (Ϫ)-MTPA chloride
(0.4 g, 1.5 mmol); the product was extracted using chloroform
and purified by column chromatography (Found: C, 64.6: H,
4.2. C32H24BrF3O3 requires C, 64.8; H, 4.1%). Separation of
the MTPA diastereoisomers by multiple elution PLC (EtOAc–
hexane; 8 : 92) yielded the high Rf diastereoisomer 8ARRS* (0.26
g, 41%), mp 173–174 ЊC (from CHCl3–hexane); [α]D Ϫ34
(c = 0.6 in CHCl3); 1H NMR (250 MHz, CDCl3) δ 2.49 (2 H, m,
H-2), 3.45 (2 H, m, H-1), 3.54 (3 H, s, OMe), 4.59 (1 H, m, H-3),
6.55 (1 H, d, J 4, H-4), 7.3–8.0 (12 H, m, Ar-H), 8.49 (1 H, s,
Ar-H), 8.83 (1 H, d, J 8, Ar-H), 9.13 (1 H, s, Ar-H); δF (94.18
MHz, CDCl3) Ϫ8.62 (3 F, s, CF3). The low Rf diastereoisomer
8ASSS* (0.24 g, 38%), mp 177–178 ЊC (from CHCl3–hexane);
1
H, 6.6%); H NMR (250 MHz, CDCl3) δ 2.05 (2 H, m, H-2),
3.03 (2 H, m, H-1), 3.91–4.08 (1 H, m, H-3), 5.39 (1 H, d,
J4,3 3.8, H-4), 7.23 (1 H, d, J9,10 8.5, H-9), 7.47 (1 H, dd, J7,6 7.0,
J7,8 8.0, H-7), 7.57 (1 H, ddd, J6,7 7.0, J6,8 1.0, J6,5 8.5, H-6), 7.74
(1 H, d, J9,10 8.5, H-10), 7.81 (1 H, dd, J8,7 7.0, J8,6 1.0, H-8), 8.25
(1 H, d, J5,6 8.5, H-5).
( )-cis-3,4-Dihydroxy-1,2,3,4-tetrahydrophenanthrene (14B).
Typical procedure: 4-methylmorpholine N-oxide (NMO, 0.5 g)
and osmium tetraoxide (0.01 g) were added to a stirred solution
of alkene 13B (1.0 g, 5.6 mmol) in CH2Cl2 (15 cm3). On com-
pletion of the reaction (24 h) a saturated solution of Na2SO3
was added and stirring was continued (1 h). The solution was
filtered and the filtrate washed with water. The racemic diol 14B
(0.93 g, 78%), obtained after removal of the solvent, was found
to be spectrally identical with a sample of compound (Ϫ)-14B.
1
[α]D +1 (c = 0.5 in CHCl3); H NMR (250 MHz, CDCl3) δ 2.6
(2 H, m, H-2), 3.45 (2 H, m, H-1), 3.56 (3 H, s, OMe), 4.67 (1 H,
m, H-3), 6.55 (1 H, d, J 4, H-4), 7.4–7.9 (12 H, m, Ar-H), 8.49
(1 H, s, Ar-H), 8.82 (1 H, d, J 8, Ar-H), 9.10 (1 H, s, Ar-H);
δF (94.18 MHz, CDCl3) Ϫ8.64 (3 F, s, CF3).
(3R, 4R)-1,3-Dibromo-1,2,3,4-tetrahydrodibenz[a,h]anthracen-
2-Methoxy-2-methyl-3a,4,5,11c-tetrahydrophenanthro[3,4-d]-
[1,3]dioxole (15B). A solution of cis-tetrahydrodiol 14B (0.2 g,
0.92 mmol), trimethylorthoacetate (0.5 cm3) and benzoic acid
(0.005 g) was refluxed (2 h); the cooled reaction mixture was
filtered, dried (K2CO3) and concentrated to yield a mixture
(70:30) of diastereoisomers 15B (0.22 g, 79%) which could not
be separated by PLC without partial decomposition. (Found:
M+, 270.1252. C17H18O3 requires M+, 270.1255); 1H NMR (250
MHz, CDCl3) δ 1.40 (3 H, s, Me), 1.61 (3 H, s, MeЈ), 1.84–1.96
(2 H, m, H-4, HЈ-4), 2.0–2.3 (2 H, m, H-4, HЈ-4), 2.58–2.76
(2 H, m, H-5, HЈ-5), 2.84–2.99 (1 H, m, HЈ-5), 3.00–3.18 (1 H,
m, H-5), 3.42 (3 H, s, OMeЈ), 4.66 (1 H, m, H-3a), 4.75 (1 H, d,
HЈ-3a), 5.62 (1 H, d, J11c,3a 6.7, H-11c), 5.81 (1 H, d, J11c,3a 6.7,
HЈ-11c), 7.14–8.17 (12 H, m, Ar-H, Ar-HЈ).
4-yl 3,3,3-trifluoro-2-methoxy-2-phenylpropionate (9ARRS*
)
A solution of bromoMTPA ester 8ARRS* (0.16 g, 0.27 mmol)
in CCl4 (15 cm3), containing N-bromosuccinimide (0.58 g,
0.3 mmol) and α,αЈ-azoisobutyronitrile (0.002 g), was refluxed
(0.5 h). The reaction mixture was filtered and the filtrate con-
centrated to yield the dibromo product 9ARRS* (0.16 g, 88%
1
yield). H NMR (250 MHz, CDCl3) δ 2.8–3.2 (2 H, m, H-2),
3.62 (3 H, s, OMe), 5.04 (1 H, m, H-3), 6.08 (1 H, t, J 3,
H-1), 6.84 (1 H, d, J 9, H-4), 7.0–8.6 (12 H, m, Ar-H), 8.76 (1 H,
s, Ar-H), 8.78 (1 H, d, J 7, Ar-H), 9.10 (1 H, s, Ar-H).
Attempted PLC purification of compound 9ARRS* resulted in
its decomposition.
( )-3,4-Dihydro-3,4-epoxydibenz[a,h]anthracene (dibenz[a,h]-
anthracene 3,4-oxide) 5ARS/5ASR and benz[5,6]anthra[1,2-b]-
oxepine (11A)
( )-cis-1,2-Dihydroxy-1,2,3,4-tetrahydrotriphenylene (14C).
Alkene 13C (0.2 g, 0.87 mmol) on dihydroxylation (OsO4)
yielded tetrahydrodiol 14C (0.160 g, 70%); mp 220 ЊC (from
MeOH–CHCl3) (lit.11 mp 220 ЊC); 1H NMR (250 MHz, CDCl3)
δ 1.26 (1 H, br s, OH), 2.16 (1 H, m, H-3), 2.26 (1 H, m, H-3),
3.20 (1 H, m, H-4), 3.46 (1 H, m, H-4), 4.01 (1 H, m, H-2), 5.35
(1 H, d, J1,2 4.4, H-1), 7.64 (4 H, m, Ar-H), 8.10 (1 H, m, Ar-H),
8.38 (1 H, m, Ar-H), 8.71 (2 H, m, Ar-H).
Sodium methoxide (0.2 g, 4 mmol) was added to a solution of
dibromoMTPA ester 9ARRS* (0.15 g, 0.22 mmol) in THF (4 cm3)
and the reaction mixture stirred at 0 ЊC (16 h). Attempted
PLC separation of the product mixture, arene oxide 5A–
oxepine 11A (2 : 3), resulted in the isolation of oxepine 11A and
decomposition of arene oxide 5A and hence the two com-
pounds were characterized without separation (0.03 g, 45%)
(Found: M+, 294.10462. C22H14O requires M+, 294.10446).
12-Methoxy-12-methyl-9,10,10a,13a-tetrahydrotriphenyleno-
[1,2-d][1,3]dioxole (15C). Yield 82% (Found: M+, 320.1412.
C21H20O3 requires M+, 320.1414); 1H NMR (250 MHz, CDCl3)
δ 1.45 (3 H, s, MeЈ), 1.71 (3 H, s, Me), 2.07 (1 H, m, H-10), 2.12
(1 H, m, HЈ-10), 2.40 (2 H, m, H-10, HЈ-10), 2.70 (3 H, s, OMe),
3.27 (4 H, m, H-9, HЈ-9), 3.52 (3 H, s, OMeЈ), 4.78 (1 H, m,
H-10a), 4.84 (1 H, m, HЈ-10a), 5.79 (1 H, d, J 6.7, H-13a), 5.92
(1 H, d, J 6.2, HЈ-13a), 7.66 (8 H, m, Ar-H), 8.14 (2 H, m, Ar-H,
Ar-HЈ), 8.32 (2 H, m, Ar-H, Ar-HЈ), 8.72 (4 H, m, Ar-H,
Ar-HЈ).
( )-3,4-Dihydro-3,4-epoxydibenz[a,h]anthracene (dibenz[a,h]-
1
anthracene 3,4-oxide) (5ARS/5ASR). [α]D 0 (CHCl3); H NMR
(250 MHz, CDCl3) δ 4.37 (1 H, m, H-3), 4.74 (1 H, d, J 4, H-4),
6.74 (1 H, dd, J 4, J 10, H-2), 7.61–9.15 (11 H, m, H-1 and
1
Ar-H). H NMR spectrum for compound 5A was very similar
to the literature report.27
Benz[5,6]anthra[1,2-b]oxepine (11A). Mp 180 ЊC (decomp.);
1H NMR (250 MHz, CDCl3) δ 5.75 (1 H, m, H-3), 6.41 (1 H, d,
J 5.5, H-2), 6.53 (1 H, dd, J 4, J 11.4, H-4), 7.61–9.15 (11 H, m,
H-5 and Ar-H).
( )-cis-1,2-Dihydroxy-1,2,3,4-tetrahydrobenzo[b]triphenylene
(14D). Alkene 13D (0.05 g. 0.18 mmol) yielded tetrahydrodiol
14D (0.03 g, 53%), mp 242 ЊC (from MeOH–CHCl3) (Found: C,
83.6; H, 5.7. C22H18O2 requires C, 84.0; H, 5.8%); 1H NMR (300
MHz, CDCl3–C5D5N) δ 1.29 (2 H, br s, OH), 2.10 (1 H, m, H-
3), 2.54 (1 H, m, H-3), 3.00 (1 H, m, H-4), 3.28 (1 H, dd, J4,4Ј 1.7,
J4,3 7.6, H-4), 4.21 (1 H, d, J2,112.4, H-2), 5.73 (1 H, s, H-1),
Synthesis of cis-tetrahydrodiols (14A–14D) and the corre-
sponding 2-methoxy-2-methyl-1,3-dioxole derivatives (15A–15D)
(Ϫ)-(3S,4R)-3,4-Dihydroxy-1,2,3,4-tetrahydrophenanthrene
J. Chem. Soc., Perkin Trans. 1, 2001, 1091–1097
1095

View Article Online
7.49–7.65 (4 H, m, Ar-H), 8.00–8.15 (3 H, m, Ar-H), 8.89 (1 H,
d, J8,9 8.0, H-8), 9.18 (1 H, s, H-14), 9.29 (1 H, s, H-9).
(2 H, m, H-3), 5.57 (1 H, d, J4,3 2.7, H-4), 5.67 (1 H, m, H-1),
7.14–8.10 (6 H, m, Ar-H).
( )-trans-2-Acetoxy-1-chloro-1,2,3,4-tetrahydrotriphenylene
(16C). 1,3-Dioxolane 15C (0.100 g, 0.031 mmol) gave the crude
chloroacetate 16C (0.080 g, 79%) (Found: M+, 324.0899.
C20H17O2Cl requires M+, 324.0917); 1H NMR (300 MHz,
CDCl3) δ 1.95 (3 H, s, OCOMe), 2.39 (1 H, m, H-3), 2.70 (1 H,
m, H-3Ј), 3.32 (1 H, m, H-4Ј), 5.64 (1 H, m, H-2), 5.66 (1 H, s,
H-1), 7.67 (4 H, m, H-8, H-9, H-14, H-15), 8.15 (1 H, d, J7,8, 9.1,
H-7), 8.24 (1 H, d, J15,16 9.3, H-16), 8.73 (2 H, d, J 7.3, Ar-H).
2-Methoxy-2-methyl-3a,4,5,15c-tetrahydrobenzo[10,11]tri-
phenyleno[1,2-d][1,3]dioxole (15D). Yield 73%; H NMR (300
1
MHz, CDCl3) δ 1.47 (3 H, s, Me), 1.77 (3 H, s, MeЈ), 2.11 (2 H,
m, H-4), 2.35 (1 H, m, H-5), 2.62 (1 H, m, HЈ-4), 2.75 (3 H, s,
OMeЈ), 3.14 (2 H, m, H-5, HЈ-4), 4.92 (2 H, m, H-3a), 5.90 (1 H,
d, J 6.2, HЈ-15c), 6.03 (1 H, d, J 6.2, H-15c), 7.58 (4 H, m, Ar-H,
Ar-HЈ), 7.71 (4 H, m, Ar-H, Ar-HЈ), 8.12 (6 H, m, Ar-H,
Ar-HЈ), 8.79 (1 H, s, HЈ-14), 8.82 (1 H, s, H-14), 8.87 (1 H, s,
H-8), 8.90 (1 H, s, HЈ-8), 9.20 (2 H, s, H-9, HЈ-9).
( )-4-Bromo-2-acetoxy-1-chloro-1,2,3,4-tetrahydrotriphenyl-
ene (17C). The trans-chloroacetate 16C (0.07 g, 0.2 mmol) on
bromination with N-bromosuccinimide gave the crude bromo-
( )-cis-3,4-Dihydroxy-1,2,3,4-tetrahydrodibenz[a,h]anthracene
(14A). Phenylboronic acid (0.35 g) was added to the solution of
alkene 13A (0.15 g, 1.8 mmol) in CH2Cl2 during the osmylation
procedure. The intermediate phenyl boronate (0.11 g, 15%)
obtained was treated with propane-1,3-diol (2 cm3) and the
reaction mixture stirred (3 h) at room temperature to liberate
the cis-tetrahydrodiol product 14A (0.05 g, 10%); mp 250–
252 ЊC (Found: M+, 314.1312. C22H18O2 requires M+,
1
trans-chloroacetates 17C (0.68 g, 80%); H NMR (300 MHz,
CDCl3) δ 2.10 (3 H, s, OCOCH3), 2.39 (1 H, m, H-3), 2.71 (1 H,
m, H-3Ј), 5.75 (1 H, d, J1,2 2.6, H-2), 5.83 (1 H, m, H-1), 6.11
(1 H, m, H-4), 7.61–8.82 (8 H, m, Ar-H).
( )-trans-2-Acetoxy-1-chloro-1,2,3,4-tetrahydrobenzo[b]tri-
phenylene (16D). 1,3-Dioxolane 15D (0.030 g, 0.081 mmol) gave
1
314.1307); H NMR (300 MHz, THF-d8) δ 2.08–2.32 (2 H, m,
1
the crude chloroacetate 16D (0.023 g, 79%); H NMR (300
H-2), 3.07–3.23 (2 H, m, H-1), 3.96 (1 H, m, H-3), 4.02 (1 H,
d, J 5.8, H-4), 7.62–7.81 (5 H, m, Ar-H), 7.94–8.02 (2 H, m,
Ar-H), 8.68 (1 H, s, Ar-H), 9.04 (1 H, d, J 8.1, Ar-H), 9.38 (1 H,
s, Ar-H).
MHz, CDCl3) δ 1.96 (3 H, s, OCOMe), 2.38 (1 H, m, H-3Ј), 2.68
(1 H, m, H-3), 3.34 (2 H, m, H-4), 5.69 (1 H, m, H-2), 5.81 (1 H,
m, H-1), 7.42 (1 H, m, Ar-H), 7.56–7.64 (4 H, m, Ar-H), 8.07
(2 H, dd, J12,13 7, J12,13 7.5, H-13, H-10), 8.67 (1 H, s, H-14), 8.85
(1 H, d, J8,7 8.0, H-8), 9.17 (1 H, s, H-9).
2-Methoxy-2-methyl-3a,14,15,15a-tetrahydronaphtho-
[2Ј,1Ј:6,7]phenanthro[1,2-d][1,3]dioxole (15A). Yield 70%; 1H
NMR (250 MHz, CDCl3) δ 1.47 (3 H, s, Me), 1.68 (3 H, s, MeЈ),
2.11 (2 H, m, H-15), 2.24 (2 H, m, H-15), 2.75 (3 H, s, OMeЈ),
3.14 (2 H, m, HЈ-14), 3.30 (2 H, m, H-14, HЈ-14), 3.50 (3 H,
s, OMe), 4.78 (1 H, m, HЈ-15a), 4.88 (1 H, m, H-15a), 5.40 (1 H,
d, J4,3 6.1, HЈ-3a), 5.52 (1 H, d, J4,3 6.2, H-3a), 7.58 (4 H, m,
Ar-H, Ar-HЈ), 7.70 (6 H, m, Ar-H, Ar-HЈ), 7.82 (4 H, m, Ar-H,
Ar-HЈ), 8.08 (2 H, m, Ar-H, Ar-HЈ), 8.54 (2 H, m, Ar-H), 8.80
(2 H, m, Ar-H), 9.25 (2 H, m, Ar-H).
( )-1-Bromo-3-acetoxy-4-chloro-1,2,3,4-tetrahydrobenzo[b]-
triphenylene (17D). The trans-chloroacetate 16D (0.025 g,
0.067 mmol) was converted into the crude bromo-trans-chloro-
acetates 17D (0.025 g, 83%); 1H NMR (300 MHz, CDCl3)
δ 2.10 (3 H, s, OCOCH3), 3.22 (2 H, m, H-3), 5.76 (1 H, d,
J3,2 2.6, H-2), 5.94 (1 H, m, H-1), 6.06 (1 H, d, J4,3 4.4, H-4),
7.60–7.71 (4 H, m, Ar-H), 8.09 (2 H, m, Ar-H), 8.33 (1 H, m,
Ar-H), 8.74 (1 H, s, H-14), 8.85 (1 H, m, Ar-H), 9.16 (1 H, s,
H-9).
Typical procedure for the synthesis of trans-chloroacetates
(16A–16D) and bromo-trans-chloroacetates (17A–17D)
( )-trans-3-Acetoxy-4-chloro-1,2,3,4-tetrahydrodibenz[a,h]-
anthracene (16A). 1,3-Dioxolane 15A (0.030 g, 0.081 mmol)
gave the crude chloroacetate 16A (0.025 g, 81%); 1H NMR (300
MHz, CDCl3) δ 1.98 (3 H, s, OCOMe), 2.35 (1 H, m, H-2), 2.64
(1 H, m, H-2), 3.46 (2 H, m, H-1), 5.35 (1 H, m, H-3), 5.50 (1 H,
m, 4-H), 7.38–7.88 (5 H, m, Ar-H), 8.10 (2 H, m, Ar-H), 8.5
(1 H, s, Ar-H), 8.88 (1-H, d, J 8.0, Ar-H), 9.20 (1-H, s, Ar-H).
(Ϫ)-(3S,4S)-3-Acetoxy-4-chloro-1,2,3,4-tetrahydrophenan-
threne (16B). A mixture of chlorotrimethylsilane (0.2 cm3,
1.5 mmol) and triethylamine (0.1 cm3) in CH2Cl2 (5 cm3) was
stirred at 0 ЊC under nitrogen. A solution of 1,3-dioxolane 15B
(0.15 g, 0.5 mmol) in CH2Cl2 (5 cm3) was then added and
the reaction mixture stirred at 0 ЊC (2 h). The crude trans-
chloroacetate 16B, obtained as an oil, was used without purifi-
cation as its attempted PLC purification resulted in partial
decomposition (0.025 g, 81%); [α]D Ϫ22 (c 0.5 in CHCl3)
(Found: M+, 274.0782. C16H15O2Cl requires M+, 274.0776);
1H NMR (300 MHz, CDCl3) δ 1.80 (3 H, s, OCOCH3), 2.03
(1 H, m, H-2Ј), 2.43 (1 H, m, H-2), 2.79 (1 H, dd, J1,1Ј 11.6,
J1Ј,2 5.5, H-1Ј), 3.02 (1 H, m, H-1), 5.50 (1 H, m, H-3), 5.55 (1 H,
m, H-4), 7.06 (1 H, d, J9,10 8.5, H-10), 7.32 (1 H, dd, J6,7 7.2,
J6,5 8.2, H-6), 7.59 (1 H, d, J9,10 8.4, H-9), 7.66 (1 H, d, J5,6 8.4,
H-5).
( )-3-Acetoxy-1-bromo-4-chloro-1,2,3,4-tetrahydrodibenz-
[a,h]anthracene (17A). The trans-chloroacetate 16A (0.025 g,
0.067 mmol) formed the crude bromo-trans-chloroacetates 17A
1
(0.02 g, 72%); H NMR (300 MHz, CDCl3) δ 2.10 (3 H, s,
OCOCH3), 3.18 (2 H, m, H-2), 5.45 (1H, d, J3,2 3.1, H-3), 5.62
(1 H, m, H-4), 6.09 (1 H, m, H-1), 7.42–7.78 (5 H, m, Ar-H),
8.18 (2 H, m, Ar-H), 8.63 (1 H, s, Ar-H), 8.80 (1 H, d, J 8.0,
Ar-H), 9.18 (1 H, s, Ar-H).
Typical procedure for the synthesis of arene oxides (5ARS/5ASR
–
5DRS/5DSR
)
(Ϫ)-(3S,4S)-1-Bromo-3-acetoxy-4-chloro-1,2,3,4-tetrahydro-
phenanthrene (17B). N-Bromosuccinimide (0.74 g, 0.41 mmol)
and azoisobutyronitrile (0.005 g) was added to a solution of the
trans-chloroacetate 16B (0.1 g, 0.36 mmol) in CCl4 (20 cm3).
The reaction mixture was heated under reflux (0.75 h) cooled,
filtered and concentrated under reduced pressure. (Ϫ)-(3S,4S)-
1-Bromo-3-acetoxy-4-chloro-1,2,3,4-tetrahydrophenanthrene
17B obtained was found to be a mixture of isomers which
decomposed during attempted separation (0.96 g, 76%);
[α]D Ϫ12 (c 0.4 in CHCl3) (Found: M+, 274.0782. C16H14BrClO2
(؉)-(3S,4R)-3,4-Epoxy-3,4-dihydrophenanthrene (5BRS/5BSR).
Sodium methoxide (0.055 g, 1.0 mmol) was added to a solution
of the bromochloroacetate isomers 17B (0.05 g, 0.14 mmol) in
dry THF (40 cm3) under nitrogen and the reaction mixture was
then stirred (12 h) at 0 ЊC in the absence of light. The solvent
was removed under reduced pressure at ice-bath temperature
and the residue extracted with diethyl ether (30 cm3). The ether
layer was washed with KOH solution (1%) followed by cold
water, dried (K2CO3) and concentrated to yield the crude oxide
5B; it was recrystallized from diethyl ether–pentane at Ϫ70 ЊC
(0.025 g, 60% yield); [α]D +165 (c 0.7 in CHCl3) (lit.10 [α]D +271,
CHCl3); 1H NMR (300 MHz, CDCl3) δ 4.18 (1 H, m, H-3), 5.14
1
requires M+, 274.0776); H NMR (300 MHz, CDCl3) δ 2.04
(3 H, s, OCOCH3), 2.90 (1 H, m, H-2), 3.08 (1 H, m, H-2Ј), 5.53
1096
J. Chem. Soc., Perkin Trans. 1, 2001, 1091–1097

View Article Online
4 D. R. Boyd and N. D. Sharma, Chem. Soc. Rev., 1996, 289.
5 D. R. Boyd, N. D. Sharma, H. Dalton and D. A. Clarke, Chem.
Commun., 1996, 45.
6 M. N. Akhtar, D. R. Boyd and J. G. Hamilton, J. Chem. Soc., Perkin
Trans. 1, 1979, 2437.
(1 H, d, J4,3 4.0, H-4), 6.44 (1 H, dd, J2,3 4.0, J2,1 9.5, H-2), 6.78
(1 H, dd, J1,10 1.7, J2,1 9.5, H-1), 7.47 (1 H, d, J9,10 8.4, H-10),
7.59 (1 H, dd, J7,6 7.1, J7,8 8.2, H-7), 7.65 (1 H, dd, J6,7 7.1, J6,5
7.9, H-6), 7.81 (1 H, d, J9,10 8.4, H-9), 8.44 (1 H, d, J8,7 8.2, H-8).
7 D. R. Boyd, N. D. Sharma, R. Agarwal, N. A. Kerley, R. A. S.
McMordie, A. Smith, H. Dalton, A. J. Blacker and G. N. Sheldrake,
J. Chem. Soc., Chem. Commun., 1994, 1693.
8 D. R. Boyd and M. E. Stubbs, J. Am. Chem. Soc., 1983, 105, 2554.
9 D. R. Boyd, R. M. E. Greene, J. D. Neill, M. E. Stubbs, H. Yagi and
D. M. Jerina, J. Chem. Soc., Perkin Trans. 1, 1981, 1447.
10 S. K. Balani, D. R. Boyd, E. S. Cassidy, G. I. Devine, J. F. Malone,
K. M. McCombe, N. D. Sharma and W. B. Jennings, J. Chem. Soc.,
Perkin Trans. 1, 1983, 2751.
11 D. R. Boyd, D. A. Kennedy, J. F. Malone, G. A. O’Kane, D. M.
Jerina, D. R. Thakker and H. Yagi, J. Chem. Soc., Perkin Trans. 1,
1987, 369.
12 D. R. Boyd, N. D. Sharma, R. Agarwal, N. A. Kerley, R. A. S.
McMordie, A. Smith, H. Dalton, A. J. Blacker and G. N. Sheldrake,
J. Chem. Soc., Chem. Commun., 1994, 1693.
13 D. R. Boyd, S. K. Agarwal, S. K. Balani, R. Dunlop, G. A. O’Kane,
N. D. Sharma, W. B. Jennings, H. Yagi and D. M. Jerina, J. Chem.
Soc., Chem. Commun., 1987, 1633.
14 H. Yagi and D. M. Jerina, J. Am. Chem. Soc., 1973, 95, 243; H. Yagi
and D. M. Jerina, J. Am. Chem. Soc., 1975, 97, 3185.
15 D. R. Boyd, K. A. Dawson, G. S. Gadaginamath, J. G. Hamilton,
J. F. Malone and N. D. Sharma, J. Chem. Soc., Perkin Trans. 1, 1981,
94.
1,2-Epoxy-1,2-dihydrotriphenylene (5CRS/5CSR). A mixture of
bromochloroacetate isomers 17C (0.07 g, 0.17 mmol) was con-
verted to 1,2-epoxy-1,2-dihydrotriphenylene 5CRS/5CSR (0.03 g,
70%); mp 165–167 ЊC (from CHCl3) (lit.21mp 167–168 ЊC);
1H NMR (300 MHz, CDCl3) δ 4.34 (1 H, m, H-2), 5.32 (1 H,
d, J1,2 3.8, H-1), 6.73 (1 H, dd, J2,3 3.8, J3,4 9.8, H-3), 7.50–7.67
(5 H, m, H-4, H-6, H-7, H-10, H-11), 8.39 (1 H, m, H-12), 8.49
(1 H, m, H-5), 8.73 (2 H, m, H-8, H-9).
1,2-Epoxy-1,2-dihydrobenzo[b]triphenylene (5DRS/5DSR). The
bromochloroacetate isomeric mixture 17D (0.025 g, 0.06 mmol)
was converted to a relatively unstable product which was
identified as 1,2-epoxy-1,2-dihydrobenzo[b]triphenylene 5DRS
/
5DSR (0.010 g, 68%);1H NMR (300 MHz, CDCl3) δ 4.38 (1 H,
m, H-2), 5.46 (1 H, d, J1,2 4.1, H-1), 6.78 (1 H, dd, J2,3 4.0, J3,4
10.0, H-3), 7.55–7.69 (5 H, m, H-4, Ar-H), 8.08 (3 H, m, Ar-H),
8.73–8.80 (1 H, m, Ar-H), 8.85 (1 H, s, H-14), 9.18 (1H, s, H-9).
3,4-Epoxy-3,4-dihydrodibenzo[a,h]anthracene
(5ARS/5ASR).
The bromochloroacetate isomers 17A (0.025 g, 0.06 mmol)
also gave a relatively unstable product 3,4-epoxy-3,4-dihydrodi-
benzo[a,h]anthracene 5ARS/5ASR (0.012 g, 70%); 1H NMR (500
MHz, THF-d8) δ 4.42 (1 H, m, H-3), 4.85 (1 H, d, J3,4 3.9, H-4),
6.94 (1 H, dd, J2,3 4.0, J2,1 10.0, H-2), 7.55–7.99 (5 H, m, H-1,
Ar-H), 8.36 (3 H, m, Ar-H), 8.73 (2 H, m, Ar-H), 9.15 (1 H, m,
Ar-H).
16 D. R. Boyd, G. S. Gadaginamath, N. D. Sharma, A. F. Drake,
S. F. Mason and D. M. Jerina, J. Chem. Soc., Perkin Trans. 1, 1981,
2233.
17 D. R. Boyd, G. S. Gadaginamath, A. Kher, J. F. Malone,
D. M. Jerina and H. Yagi, J. Chem. Soc., Perkin Trans. 1, 1980,
2113.
18 D. R. Boyd, M. G. Burnett and R. M. E. Greene, J. Chem. Soc.,
Chem. Commun., 1981, 838.
19 D. R. Boyd and R. M. E. Greene, J. Chem. Soc., Perkin Trans. 1,
1982, 1535.
20 S. K. Balani, D. R. Boyd, R. M. E. Greene and D. M. Jerina,
J. Chem. Soc., Perkin Trans. 1, 1984, 1781.
21 D. R. Boyd, N. D. Sharma, S. K. Agarwal, G. S. Gadaginamath,
G. A. O’Kane, W. B. Jennings, H. Yagi and D. M. Jerina, J. Chem.
Soc., Perkin Trans. 1, 1993, 423.
22 S. K. Agarwal, D. R. Boyd, R. Dunlop and W. B. Jennings, J. Chem.
Soc., Perkin Trans. 1, 1988, 3013.
23 D. R. Boyd and G. A. O’Kane, J. Chem. Soc., Perkin Trans. 1, 1990,
2079.
24 R. Agarwal and D. R. Boyd, J. Chem. Soc., Perkin Trans. 1, 1993,
2869.
25 D. R. Boyd and G. A. O’Kane, Tetrahedron Lett., 1987, 28, 6395.
26 P. L. Kole and S. Kumar, J. Chem. Soc., Perkin Trans. 1, 1991, 2151.
27 K. L. Platt, H. Frank and F. Oesch, J. Chem. Soc., Perkin Trans. 1,
1989, 2299.
28 A. M. Jeffrey, H. J. C. Yeh, D. M. Jerina, T. R. Patel, J. F. Davey and
D. T. Gibson, Biochemistry, 1975, 14, 575.
29 D. R. Boyd, R. A. S. McMordie, H. P. Porter, H. Dalton,
R. O. Jenkins and O. W. Howarth, J. Chem. Soc., Chem. Commun.,
1987, 1722.
30 D. M. Jerina, H. Selander, H. Yagi, M. C. Wells, J. F. Davey,
V. Mahadevan and D. T. Gibson, J. Am. Chem. Soc., 1976, 98, 5988.
31 M. N. Akhtar, D. R. Boyd, N. J. Thompson, M. Koreeda,
D. T. Gibson, V. Mahadevan and D. M. Jerina, J. Chem. Soc., Perkin
Trans. 1, 1975, 2506.
Typical photoisomerization reactions of arene oxides (5ARS
5ASR, 5CRS–5CSR and 5DRS/5DSR) to yield oxepines (11A, 11C
and 11D)
/
A mixture of benz[5,6]anthra[1,2-b]oxepine 11A and arene
oxide 5ARS/5ASR (0.1 g) was dissolved in CDCl3 (1 cm3) con-
taining triethylamine (0.005 g) in a standard pyrex NMR tube;
the solution was irradiated (0.5 h) under UV light generated
from a medium pressure mercury lamp (>300 nm) at ambient
temperature. The reaction mixture, on separation by PLC, gave
oxepine 11A and a minor phenolic product. Further irradiation
resulted in isomerization of oxepine 11A to phenolic isomers.
Benz[5,6]anthra[1,2-b]oxepine 11A: 1H NMR (300 MHz,
CDCl3) δ 5.80 (1 H, m, H-3), 6.40 (1 H, d, J 5.5, H-2), 6.53 (1 H,
dd, J 5.0, J 11.4, H-4), 7.60–9.15 (11 H, m, H-5 and ArH).
Phenanthro[10,9-b]oxepine 11C21: 1H NMR (300 MHz,
CDCl3) δ 5.83 (1 H, dd, J2,3 = J3,4 5.1, H-3), 6.50 (1 H, d, J2,3 5.1,
H-2), 6.57 (1 H, dd, J3,4 5.1, J4,5 11.3, H-4), 7.50–8.70 (9 H, m,
H-5 and Ar-H).
1
Benz[3,4]anthra[1,2-b]oxepine 11D26: H NMR (300 MHz,
CDCl3) δ 5.84 (1 H, dd, J2,3 = J3,4 5.1, H-3), 6.53 (1 H, dd,
J3,4 5.1, J4,5 11.2, H-4), 6.57 (1 H, d, J2,3 5.1, H-2), 7.50–7.73
(5 H, m, H-5 and Ar-H), 8.02–8.14 (3 H, m, Ar-H), 8.84–8.97
(1 H, m, Ar-H), 9.07 (1 H, s, Ar-H), 9.3 (1 H, s, H-9).
32 M. Koreeda, M. N. Akhtar, D. R. Boyd, J. D. Neill, D. T. Gibson
and D. M. Jerina, J. Org. Chem., 1978, 43, 1023.
33 D. T. Gibson, V. Mahadevan, D. M. Jerina, H. Yagi and H. J. C. Yeh,
Science, 1975, 189, 295.
34 D. M. Jerina, P. J. van Bladeren, H. Yagi, D. T. Gibson, V.
Mahadevan, A. S. Neese, M. Koreeda, N. D. Sharma and D. R.
Boyd, J. Org. Chem., 1984, 49, 3621.
35 D. R. Boyd, N. D. Sharma, R. Agarwal, S. M. Resnick, M. J.
Schoken, D. T. Gibson, J. M. Sayer, H. Yagi and D. M. Jerina,
J. Chem. Soc., Perkin Trans. 1, 1997, 1715.
36 D. R. Boyd and G. N. Sheldrake, Nat. Prod. Rep., 1998, 15, 309.
37 Y. Pommier, G. Kohlhagen, P. Pourquier, J. M. Sayer, H. Kroth and
D. M. Jerina, Proc. Natl. Acad. Sci. U.S.A., 2000, 97, 2040.
38 E. L. Kennaway, Biochem. J., 1930, 24, 497.
39 K. L. Platt and M. Schollmeier, Chem. Res. Toxicol., 1994, 7, 89.
40 R. E. Parales, S. M. Resnick, C. L. Yu, D. R. Boyd, N. D. Sharma
and D. T. Gibson, J. Bacteriol., 2000, 182, 5495.
41 S. K. Balani, P. J. van Bladeren, E. S. Cassidy, D. R. Boyd and
D. M. Jerina, J. Org. Chem., 1987, 52, 137.
Acknowledgements
We thank DENI (I.N.B., A.S.), the Queen’s University of
Belfast (S. K. B., F. H.) and the BBSRC (N.D.S.) for funding.
References
1 D. R. Boyd, N. D. Sharma, J. S. Harrison, J. T. G. Hamilton, D. B.
Harper and W. C. McRoberts, J. Chem. Soc., Chem. Commun., 2000,
1481.
2 D. M. Jerina, J. W. Daly, B. Witkop, P. Zaltzman-Nirenberg and
S. Udenfriend, J. Am. Chem. Soc., 1968, 90, 6525.
3 D. R. Boyd and D. M. Jerina, in Small Ring Heterocycles, Chemistry
of Heterocyclic Compounds, part 3, ed. A. Hassner, vol. 42, Wiley,
New York, 1985, 197.
J. Chem. Soc., Perkin Trans. 1, 2001, 1091–1097
1097