New synthetic approaches to polycyclic aromatic hydrocarbons and their carcinogenic oxidized metabolites: Derivatives of benzo[s]picene, benzo[rst]pentaphene, and dibenzo[b,def]chrysene

Article abstract of DOI:10.1021/jo9918044

A new synthetic approach to polycyclic aromatic compounds is described that entails in the key steps double Suzuki coupling of PAH bisboronic acid derivatives with o-bromoaryl aldehydes to furnish aryl dialdehydes that are converted to larger polycyclic aromatic ring systems by either (a) conversion to diolefins by Wittig reaction followed by photocyclization or (b) reductive cyclization with triflic acid and 1,3-propanediol. This synthetic method provides convenient access to as many as three different polycyclic aromatic ring systems from a single Suzuki coupled intermediate. It was utilized to synthesize substituted derivatives of benzo[s]picene, benzo[rst]pentaphene, dibenzo-[b,def]chrysene, and 13,14-dihydro-benz[g]indeno[2,1-a]fluorene, as well as the putative carcinogenic bisdihydrodiol metabolites of benzo[s]picene, benzo[rst]pentaphene, and dibenzo[b,def]chrysene.

Full text of DOI:10.1021/jo9918044

3952
J . Org. Chem. 2000, 65, 3952-3960
New Syn th etic Ap p r oa ch es to P olycyclic Ar om a tic Hyd r oca r bon s
a n d Th eir Ca r cin ogen ic Oxid ized Meta bolites: Der iva tives of
Ben zo[s]p icen e, Ben zo[r st]p en ta p h en e, a n d
Diben zo[b,d ef]ch r ysen e
Fang-J ie Zhang, Cecilia Cortez, and Ronald G. Harvey*
Ben May Institute for Cancer Research, University of Chicago, Chicago, Illinois 60637
Received November 22, 1999
A new synthetic approach to polycyclic aromatic compounds is described that entails in the key
steps double Suzuki coupling of PAH bisboronic acid derivatives with o-bromoaryl aldehydes to
furnish aryl dialdehydes that are converted to larger polycyclic aromatic ring systems by either (a)
conversion to diolefins by Wittig reaction followed by photocyclization or (b) reductive cyclization
with triflic acid and 1,3-propanediol. This synthetic method provides convenient access to as many
as three different polycyclic aromatic ring systems from a single Suzuki coupled intermediate. It
was utilized to synthesize substituted derivatives of benzo[s]picene, benzo[rst]pentaphene, dibenzo-
[b,def]chrysene, and 13,14-dihydro-benz[g]indeno[2,1-a]fluorene, as well as the putative carcinogenic
bisdihydrodiol metabolites of benzo[s]picene, benzo[rst]pentaphene, and dibenzo[b,def]chrysene.
Sch em e 1
Polycyclic aromatic hydrocarbons (PAHs) are wide-
spread environmental pollutants produced in the com-
bustion of organic matter.1,2 Some PAHs, such as benzo-
[a]pyrene, are potent carcinogens that are implicated in
the causation of lung cancer in cigarette smokers, as well
as other cancers.1a,3
The mechanism of PAH carcinogenesis involves meta-
bolic activation by P450 mono-oxygenase enzymes to PAH
dihydrodiol intermediates that are transformed to PAH
diol epoxides in which the epoxide ring resides in a
sterically crowded bay or fjord molecular region. The
latter serves to protect the reactive epoxide function from
further enzymatic degradation. The PAH diol epoxides
react with DNA to form covalent adducts that lead to
mutations ultimately resulting in the induction of
tumors.1a,4 There is evidence that as many as three
additional mechanistic pathways may also play a role.1a
These include (1) one-electron oxidation to radical-cation
intermediates that react with DNA, resulting in depu-
rination;5,6 (2) dihydrodiol dehydrogenase catalyzed de-
hydrogenation of dihydrodiols to quinones that combine
with DNA or enter into a redox cycle with O2 to generate
reactive oxygen species that attack DNA;7 and (3) forma-
tion of benzylic alcohol derivatives (by oxidation of methyl
substituents or by biomethylation)8 that are in turn
converted by sulfotransferase enzymes to benzylic sulfate
esters that attack DNA.9 There is also evidence that more
polar PAH metabolites, such as bisdihydrodiols, contrib-
ute to the carcinogenicity of PAHs that possess two or
more sterically crowded bay or fjord regions in the
molecule (e.g., dibenz[a,j]anthracene).10
Investigations of the mechanisms of PAH carcinogen-
esis at the molecular-genetic level have been hampered
by a deficiency of efficient methods for synthesis of PAHs
(1) (a) Harvey, R. G. Polycyclic Aromatic Hydrocarbons: Chemistry
and Carcinogenesis; Cambridge University Press: Cambridge,
England, 1991. (b) Harvey, R. G. In The Handbook of Environmental
Chemistry, Volume 3, Part I: PAHs and Related Compounds; Hut-
zinger, O., Editor-in-Chief, Neilson, A., Volume Ed., Springer: Berlin,
Heidelberg, 1997; Chapter 1, pp 1-54.
(7) Smithgall, T. E.; Harvey, R. G.; Penning, T. M. Cancer. Res. 1988,
48, 1227. Flowers-Geary, L.; Harvey, R. G.; Penning, T. M. Chem. Res.
Toxicol. 1993, 6, 252. Penning, T. M.; Ohnishi, S. T.; Ohnishi, T.;
Harvey, R. G. Chem. Res. Toxicol. 1996, 9, 84. Burczynski, M. E.;
Harvey, R. G.; Penning, T. M. Biochemistry 1998, 37, 6781.
(8) Flesher, J .; Myers, S. R.; Blake, J . W. In Polynuclear Aromatic
Hydrocarbons; Cooke, M., Dennis, A. J ., Eds.; Battelle Press: Colum-
bus, OH, 1986; pp 271-284.
(2) Harvey, R. G. Polycyclic Aromatic Hydrocarbons; Wiley-VCH:
New York, 1997.
(3) Denissenko, M. F.; Pao, A.; Tang, M.; Pfeifer, G. P. Science 1996,
274, 430. Pfeifer, G. P.; Denissenko, M. F. Environ. Mol. Carcin. 1998,
31, 197. Denissenko, M. F.; Chen, J . X.; Tang, M.; Pfeifer, G. P. Proc.
Natl. Acad. Sci. U.S.A. 1997, 94, 3893. Chen, J . X.; Zheng, Y.; West,
M.; Tang, M. Cancer Res. 1998, 58, 2170.
(4) Conney, A. H. Cancer Res. 1982, 42, 4875. Dipple, A.; Moschel,
R. C.; Bigger, C. A. H. In Chemical Carcinogens, 2nd ed.; Searle, C.
E., Ed.; ACS monograph no. 182, American Chemical Society: Wash-
ington, DC, 1984; pp 63-84. Harvey, R. G.; Geacintov, N. E. Acc. Chem.
Res. 1988, 21, 66. Harvey, R. G. Polycyclic Hydrocarbons and Car-
cinogenesis; ACS Symposium Series monograph 283,; American Chemi-
cal Society: Washington, DC, 1985.
(5) Cavalieri, E. L.; Rogan, E. G. Xenobiotica 1995, 25, 677.
(6) Recent evidence suggests that the depurination pathway may
be of relatively minor importance in the mechanism of PAH carcino-
genesis: Melendez-Colon, V.; Luch, A.; Seidel, A.; Baird, W. Carcino-
genesis 1999, 20, 1885.
(9) Glatt, H.; Pauly, K.; Frank, H.; Seidel, A.; Oesch, F.; Harvey, R.
G.; Werle-Schneider, G. Carcinogenesis 1994, 15, 2605. Surh, Y.; Liem,
A.; Miller, E. C.; Miller, J . Carcinogenesis 1989, 10, 1519.
(10) Carcinogenic PAHs with more than one bay or fjord molecular
region include dibenzo[a,c]anthracene, dibenzo[a,h]anthracene, diben-
zo[a,j]anthracene, dibenzo[b,def]chrysene, dibenzo[a,e]aceanthrylene,
dibenzo[def,p]chrysene, and benzo[rst]pentaphene. For references to
the role of bisdihydrodiols in the mechanism of carcinogenesis of these
PAHs, see: Harvey, R. G.; Dai, W.; Zhang, J .-T.; Cortez, C. J . Org.
Chem. 1998, 63, 8118. Zhang, J .-T.; Dai, W.; Harvey, R. G. J . Org.
Chem. 1998, 63, 8125. Zhang, J .-T.; Harvey, R. G. Tetrahedron 1999,
55, 625.
10.1021/jo9918044 CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/31/2000

New Synthetic Approaches to PAHs
J . Org. Chem., Vol. 65, No. 13, 2000 3953
Sch em e 2
and their oxidized metabolites.1,2 We now report a new
synthetic approach that involves in the key steps double
Suzuki coupling and triflic acid catalyzed reductive
cyclization. This method was used to synthesize putative
carcinogenic metabolites of benzo[s]picene (1), benzo[rst]-
pentaphene (2), and dibenzo[b,def]chrysene (obsolete
name, dibenzo[a,h]pyrene) (3).
the starting compounds for the preparation of 12, was
itself synthesized from 1,4-dibromonaphthalene (6a ) via
reaction with Mg and B(OMe)3 followed by acidic hy-
drolysis.11 The other starting compound 6-bromo-2,3-
dimethoxybenzaldehyde (8c) was synthesized from 3-meth-
oxysalicylaldehyde (7a ) (Scheme 1) via acetylation of the
phenolic hydroxyl group followed by selective bromination
of the product (7b) with Br2 and KBr in the 6-position to
yield (8a ).12 Compound 8a was converted to its methyl
ether derivative (8c) by hydrolysis with 10% KOH to the
phenol (8b) followed by reaction of the latter with
dimethyl sulfate.13 Direct synthesis of 8c through bro-
mination of 2,3-dimethoxybenzaldehyde was less satis-
factory, affording a mixture of 5-bromo-2,3-dimethoxy-
benzaldehyde and 8c (1.7:1).
Resu lts
The general synthetic approach entails double Suzuki
coupling of a PAH bisboronic acid derivative with a
suitably substituted aryl bromide. This is illustrated by
its application to the synthesis of the bisdihydrodiol (4)
and diol epoxide (5) derivatives of benzo[s]picene. The
Double Suzuki coupling of the aryl bromide 8c with
naphthalene 1,4-bisboronic acid (6c) was carried out in
the presence of Pd(PPh3)4 in DME by the published
procedure.14 The yield of the dialdehyde product (9a )
obtained by this procedure was moderate (20%). How-
ever, it was dramatically improved (to >90%) by a
modified procedure that entailed use of sodium carbonate
as a base and water as a cosolvent.15 The principal
1
product was shown by TLC and H NMR analysis to be
a mixture of stereoisomers of 9a in the ratio of 35:65.
Formation of isomers is apparently due to steric restric-
tion of rotation between the two outer substituted aryl
rings and the central naphthalene ring of 9a . An ad-
ditional product identified as the palladium complex 10a
key intermediate in this synthesis is 3,4,9,10-tetrameth-
oxybenzo[s]picene (12) (Scheme 2). The 1,4-bisboronic
acid derivative of naphthalene (6c), required as one of
1
on the basis of its elemental analysis and its H and 13C
spectra was also isolated. This complex evidently arises
from reaction between 8c and Pd(PPh3)4. It is surpris-
ingly stable in that it can be purified by column chro-
matography on silica gel and is resistant to hydrolysis
by heating in aqueous solution.
Analogous double Suzuki coupling of 2-bromobenzal-
dehyde with 6c in the presence of Pd(PPh3)4 in DME by
the modified procedure afforded the corresponding dial-
dehyde product lacking the methoxy substituents (9b)
in good yield (91%), but the analogous palladium complex
(11) Allen, L. M.; Roscoe, C. W. J . Pharm. Sci. 1969, 58, 368.
(12) Wriede, U.; Fernandez, M.; West, K. F.; Harcourt, D.; Moore,
H. W. J . Org. Chem. 1987, 52, 4485.
(13) Kametani, T.; Honda, T.; Inoue, H.; Fukumoto, K. J . Chem. Soc.,
Perkin Trans. 1 1976, 1221.
(14) Wright, S. W.; Hageman, D. L.; McClure, L. D. J . Org. Chem.
1994, 59, 6095.
(15) Sharp, M. J .; Cheng, W.; Snieckus, V. Tetrahedron Lett. 1987,
28, 5093.
(16) 1,2-Epoxybutane serves to block competing secondary reactions
by scavenging the HI produced: Liu, L. B.; Yang, B. W.; Katz, T. J .;
Poindexter, M. K. J . Org. Chem. 1991, 56, 3769.

3954 J . Org. Chem., Vol. 65, No. 13, 2000
Zhang et al.
Sch em e 3
10b could not be isolated. To test the role of the
palladium complex in the Suzuki coupling reaction with
6c, 10a was reacted with 6c in the presence of Na2CO3.
The dialdehyde coupling product 9a was obtained in
about 50% yield, confirming that 10a is a likely inter-
mediate in the Suzuki coupling.
the fully aromatic parent compound 13a accompanied by
the PAH compound arising from the alternative mode of
cyclization, 1,2,11,12-tetramethoxy-13,14-dihydrobenz[g]-
indeno[2,1-a]fluorene (15). The 1H NMR spectrum of 13a
was relatively simple as a result of its symmetry. It
exhibited characteristic singlets at δ 9.07, 8.59, and 7.81,
which were assigned to the meso region H5,8 aromatic
protons, the K-region H6,7 protons, and the bay region
H13,14 protons, respectively, and additional peaks consis-
tent with this assignment. The product 15 presumably
arises from reductive cyclization of the aldehyde functions
of 9a to the adjacent positions of the naphthalene ring
Wittig reaction of 9a with the phosphonium salt
prepared from methyltriphenylphosphonium bromide
and n-butyllithium took place smoothly to furnish the
diolefin 11 in 91% yield. Photochemical oxidative cyclo-
dehydrogenation of 11 in dilute benzene solution (1.8 ×
10-3 M) in the presence of iodine and 1,2-epoxybutane16
gave 3,4,9,10-tetramethoxybenzo[s]picene (12) as the sole
cyclized product in 81% yield.17,18 Compound 12 prepared
by this route was identical in its physical and spectral
properties, including its 1H NMR spectrum, with an
authentic sample prepared by a synthetic route not
involving Suzuki coupling.18b Conversion of tetramethox-
ybenzo[s]picene (12) to the bisdihydrodiol (4) and bis-anti-
diol epoxide (5) derivatives of benzo[s]picene was reported
previously.19
In principle, cyclodehydration of the dialdehyde inter-
mediate (9a ) might be expected to provide synthetic
access to 3,4,9,10-tetramethoxybenzo[rst]pentaphene (13a)
(Scheme 3), a convenient synthetic precursor of the
unknown 3,4,9,10-bisdihydrodiol of benzo[rst]pentaphene.
However, cyclodehydration of polycyclic aromatic alde-
hydes is seldom useful synthetically because of the low
yields obtained. Consistent with this experience, acid-
catalyzed cyclodehydration of 9a , even under relatively
dilute conditions, afforded mainly polymeric products. In
view of this difficulty, we investigated a potential alter-
native approach involving reductive Friedel-Crafts cy-
clization of 9a with trifluoromethanesulfonic acid (triflic
acid) and 1,3-propanediol by the procedure of Fukuzawa
et al.20 The product of this reaction was not the expected
5,8-dihydro derivative (14). Instead, there was obtained
1
system. The H NMR spectrum of 15 showed a singlet
at δ 4.06 for four methylene protons, consistent with
either the assigned structure 15 or the alternative
structure 14. However, structure 14 could be ruled out
by the chemical shift pattern of the remaining aromatic
protons, which showed a doublet of doublets at δ 8.81
and a doublet at δ 8.06 assigned to the sterically hindered
fjord region H4,5 and H8,9 protons, respectively, and an
additional doublet of doublets at δ 7.65 and a doublet at
δ 7.06 assigned to the H6,7 and H3,10 protons, respectively,
as well as two singlets at δ 3.98 and 4.04 for the methyl
protons. Moreover, attempted dehydrogenation of this
compound over a palladium catalyst failed to yield the
fully aromatic PAH compound 13a , affording only recov-
ered 15.
The yields of 13a and 15 obtained from reductive
cyclization of 9a with triflic acid and 1,3-propanediol in
1,2-dichloroethane by the original procedure of Fukuzawa
et al.20 (Method A) were modest (15% and 13%, respec-
tively). Substitution of other acids (CF3CO2H, H2SO4,
Sc(OTf)3, or BF3‚Et2O) for triflic acid proved even less
satisfactory, affording only polymeric products. The use
of CH2Cl2 or CH3NO2 as solvents in place of 1,2-dichlo-
roethane resulted in formation of complex mixtures of
polar products of uncertain identity. However, use of
dioxane in place of dichloroethane as the solvent (Method
B) doubled the yield of 13a (31%), and instead of 15 there
was obtained a new compound (25%) identified as 16, a
cyclic diether derivative of 15. Compound 16 was con-
verted to 15 by heating with 1,3-propanediol and a
catalytic amount of triflic acid in 1,2-dichloroethane. This
reaction failed to take place in dioxane. These findings
(17) Mallory, F. B.; Mallory, C. W. Org. React. 1984, 30, 1.
(18) (a) Zhang, F.-J .; Harvey, R. G. J . Org. Chem. 1998, 63, 2773.
(b) Zhang, F.-J .; Harvey, R. G. J . Org. Chem. 1998, 63, 1168.
(19) The bisdihydrodiol (1) was obtained as a mixture of diastere-
omers consisting of a meso isomer and a racemic mixture of enanti-
omers.
(20) Fukuzawa, S.; Tsuhimoto, T.; Hiyama, T. J . Org. Chem. 1997,
62, 151.

New Synthetic Approaches to PAHs
J . Org. Chem., Vol. 65, No. 13, 2000 3955
Sch em e 4
tion of 13c to the bisdihydrodiol by treatment with a large
excess of NaBH4 in EtOH with O2 bubbling through the
solution (Scheme 5) by the usual procedure21,22 furnished
trans-3,4-trans-9,10-tetrahydroxy-3,4,9,10-tetrahydroben-
zo[rst]pentaphene (17) as a mixture of isomers (17a -c)
along with the overreduced products 18 and 19. The ratio
of 18 to 19 was found to be dependent upon both the
amount of NaBH4 and reaction time. Optimum yield of
the bisdihydrodiol was favored by a low ratio of NaBH4
and relatively short reaction time, but some degree of
over- or under-reduction could not be avoided. However,
17 was obtained in good yield free of 18 and 19 by a
modified procedure that entailed monitoring the reaction
by HPLC to ensure complete conversion of 13c to
products, followed by acetylation of the crude product
mixture, and dehydrogenation with DDQ in refluxing
benzene to yield the tetraacetate derivative of 17. Hy-
drolysis gave the pure bisdihydrodiol 17, which exhibited
a tendency to undergo autoxidation in air and light.
suggest that 16 is an intermediate in the formation of
15. Although 16 is stable in dioxane, it readily undergoes
reduction to 15 in 1,2-dichloroethane. In confirmation of
this hypothesis, when reductive cyclization of 9a in 1,2-
dichloroethane was monitored by TLC, compound 16 was
detected as an initially formed product, which disap-
peared over time and was replaced by 15.
Two other compounds that might conceivably partici-
pate as intermediates in the reductive cyclization of the
dialdehyde 9a are its diacetal (9c) and the related
dialcohol 1,4-bis(2-hydroxymethyl-3,4-dimethoxyphenyl)-
naphthalene (9d ) (Scheme 3). The diacetal 9c was
synthesized from 9a by reaction with 1,3-propanediol
catalyzed by p-toluenesulfonic acid. The dialcohol 9d was
prepared by reduction of 9a with excess NaBH4. Reaction
of 9c with triflic acid in 1,2-dichloroethane under the
conditions for reductive cyclization took place readily to
afford compounds 13a and 15 in essentially the same
ratio as previously obtained. This is consistent with
intermediacy of the diacetal 9c in the reductive cycliza-
tion of 9a . However, the dialcohol 9d failed to react under
the same conditions, indicating that it is unlikely to be
an intermediate in reductive cyclization of 9a .
The bisdihydrodiol 17 was obtained as a mixture of the
symmetrical meso isomer (17a ) and a racemic pair of
enantiomers (17b and 17c). HPLC showed two peaks
corresponding to the meso and racemic forms. Each of
the isomers of 17 has the potential for the dihydrodiol
functions to exist as mixtures of diequatorial and diaxial
conformers. However, for most PAH dihydrodiols, con-
formational interconversion is known to occur with
relative facility at room temperature in the absence of
steric restrictions, and the conformers tend to exist in
dynamic equilibrium that favors a predominance of the
1
diequatorial conformers.23 The H and 13C NMR spectra
Synthesis of benzo[rst]pentaphene (2) was accom-
plished via analogous reductive cyclization of the unsub-
stituted dialdehyde 9b with 1,3-propanediol and triflic
acid in 1,2-dichloroethane (Scheme 4). This PAH was the
sole major product; the PAH expected to be formed by
the alternative mode of cyclization was not detected. The
regioselectivity of cyclization is clearly dependent upon
the nature of the substituents in the precursor.
Conversion of 3,4,9,10-tetramethoxybenzo[rst]pen-
taphene (13a ) to the corresponding bisdihydrodiol (17)
proceeded via initial demethylation with BBr3 to yield
the corresponding biscatechol, 3,4,9,10-tetrahydroxyben-
zo[rst]pentaphene (13b). In view of the sensitivity of PAH
catechols to autoxidation, 13b was isolated and charac-
terized as its tetraacetate derivative (13c). Direct reduc-
of the isomers of 17 showed only a single set of peaks for
each isomer, consistent with the anticipated facility of
interconversion of the diaxial and diequatorial conform-
ers at ordinary temperatures. The 1H spectrum of the
mixture of the corresponding tetraacetate derivatives
showed two sets of peaks for the H1 and the H2 protons
consistent with the presence of the meso and racemic
forms. An analogous mixture of meso and racemic
isomers was obtained previously in the related synthesis
of the bisdihydrodiol derivative of benzo[s]picene (4).18b,19
Extension of this approach to the synthesis of 1,2,8,9-
tetramethoxydibenzo[b,def]chrysene (22a ) was also ex-
amined (Scheme 6). The 1,5-bisboronic acid derivative of
naphthalene (20c) required as starting compound was
Sch em e 5

3956 J . Org. Chem., Vol. 65, No. 13, 2000
Zhang et al.
Sch em e 6
synthesized from 1,5-dibromonaphthalene24 (20a ) via
reaction with Mg and B(OMe)3 followed by acidic hy-
drolysis of the resulting 1,5-bisboronate (20b).11 Double
Suzuki coupling of 8c with 20c in the presence of Pd-
(PPh3)4 and Na2CO3 in aqueous DME by the procedure
employed for 9a furnished the corresponding dialdehyde
efficiently reduced to 23 under appropriate reaction
conditions (large excess of NaBH4, use of CH2Cl2 as
cosolvent, and relatively short reaction time [2 h]).
Formation of over-reduced products was minimal under
these conditions.
1
(21a ). The H NMR spectrum of 21a showed it to be a
Discu ssion
mixture of stereoisomers (35:65). As in the case of 9a ,
formation of isomers is a consequence of steric restriction
of rotation between the bulky aryl substituents and the
central naphthalene ring.
A new synthetic approach to PAHs is described that
entails in the key steps double Suzuki coupling of a PAH
bisboronic acid derivative with an o-bromoaryl aldehyde
and conversion of the primary dialdehyde product to a
larger polycyclic aromatic ring system through either (a)
Wittig reaction followed by oxidative photocyclization or
(b) triflic acid catalyzed reductive cyclization. The former
sequence is illustrated by synthesis of 3,4,9,10-tet-
ramethoxybenzo[s]picene (12) and its conversion to the
bisdihydrodiol of benzo[s]picene (4). Double Suzuki cou-
pling of the aryl bromide 8c with naphthalene 1,4-
bisboronic acid (6c) in the presence of Pd(PPh3)4 by a
modified procedure using aqueous Na2CO3 furnished
smoothly the dialdehyde product 9a in good yield (Scheme
2). Wittig reaction of 9a with methylenetriphenylphos-
phorane provided the diolefin (11), which underwent
photochemical cyclodehydrogenation to yield 12. Alter-
natively, the dialdehyde 9a was converted to 3,4,9,10-
tetramethoxybenzo[rst]pentaphene (13a ) via triflic acid
catalyzed reaction with 1,3-propanediol in 1,2-dichloro-
ethane. Compound 13a arises from double intramolecular
cyclization into the peri positions of the adjacent ring. It
was accompanied by 1,2,11,12-tetramethoxy-13,14-dihy-
drobenz[g]indeno[2,1-a]fluorene (15) formed by double
cyclization into the â-positions of the same ring. Similar
reaction of 9a in dioxane afforded 13a in higher yield,
but compound 15 was not detected as a product. Instead,
the second major product was a new compound identified
as 16, a cyclic diether derivative of 15. Compound 16 was
converted to 15 by triflic acid catalyzed reaction with 1,3-
propanediol. Thus, this synthetic methodology provides
convenient synthetic access to three different PAH ring
systems, 12 or 13a and 15.
Reductive Friedel-Crafts cyclization of the dialdehyde
21a with triflic acid and 1,3-propanediol in dioxane gave
1,2,8,9-tetramethoxydibenzo[b,def]chrysene (22a ) as the
sole identifiable product. Neither the PAH compound
analogous to 15 from the alternative mode of cyclization
nor a product analogous to 16 was detected. The poor
solubility of 22a made its separation from polymeric
products and its purification by chromatography difficult.
The 1H NMR spectrum of 22a was entirely consistent
with its structural assignment, exhibiting a characteristic
singlet at δ 8.95 assigned to the meso region H7,14
aromatic protons, doublets at δ 8.88, 8.78, and 8.34 for
the remaining aromatic protons, and additional peaks
consistent with this assignment. Reductive cyclization of
the unsubstituted dialdehyde 21b with triflic acid and
1,3-propanediol under similar conditions furnished diben-
zo[b,def]chrysene (22b) as the principal product.
Conversion of 22a to the bisdihydrodiol of dibenzo-
[b,def]chrysene (23) via a sequence of steps analolgous
to that employed for the synthesis of 17, i.e., demeth-
ylation to the biscatechol (22c), acetylation to the tet-
raacetate (22d ), and reduction with NaBH4, was success-
ful. Despite the relatively poor solubility of the tetraac-
etate 22d (and other compounds in this series), it was
(21) Dai, W.; Abu-Shqara, E.; Harvey, R. G. J . Org. Chem. 1995,
60, 4905. Platt, K.; Oesch, F. Synthesis 1982, 459.
(22) Oxygen serves to recycle hydroquinone intermediates formed
by isomerization of the primary reduction product back to quinone.
(23) Harvey, R. G. Polycyclic Aromatic Hydrocarbons: Chemistry
and Carcinogenesis; Cambridge University Press: Cambridge,
England, 1991; Chapter 13, pp 306-329. Harvey, R. G. In Cyclohexenes,
Cyclohexadienes, and Related Hydroaromatics; Rabideau, P. W., Ed.;
VCH: New York, 1989; pp 267-298.
The use of Suzuki coupling for the synthesis of PAH
ring systems has been reported by Kumar25 and by Rice
(24) Harvey, R. G.; Pataki, J ., Cortez, C.; Di Raddo, P.; Yang, C. J .
Org. Chem. 1991, 56, 1210.
(25) Kumar, S. J . Chem. Soc., Perkin Trans. 1 1998, 3157.

New Synthetic Approaches to PAHs
J . Org. Chem., Vol. 65, No. 13, 2000 3957
Sch em e 7
considerations explain the formation of the 3,4,9,10-
tetramethoxy derivative of benzo[rst]pentaphene (13a )
rather than its 5,8-dihydro derivative (e.g., 14) in the
triflic acid catalyzed reaction of the dialdehyde 9a .
Formation of the unusual cyclic ether product 16 in
the reaction of 9a with triflic acid and 1,3-propanediol
in dioxane, but not in 1,2-dichloroethane, is also consis-
tent with the proposed mechanism. The most plausible
origin of 16 is via initial formation of the bisether
derivative of 15 analogous to 25 followed by acid-
catalyzed reduction with loss of 3-hydroxypropanal by one
of the ether functions leading to formation of 16 through
reaction of the benzylic cation formed by the resulting
partially reduced monoether with the second alcohol
function. From a synthetic viewpoint, formation of the
ether derivative 16 does not detract from the utility of
the method, since reaction in dioxane affords higher
yields and 16 may be readily converted to 15 by heating
with 1,3-propanediol and triflic acid.
and Cai.26 The synthetic approach described herein is
more closely related to Kumar’s method, which entails
Suzuki cross-coupling of an o-bromoaryl aldehyde fol-
lowed by conversion of the aldehyde function of the
product into the related ethylene oxide and acid-catalyzed
cyclization of the latter (Scheme 7). The novel variations
on this synthetic theme introduced herein expand the
scope of Suzuki coupling to the synthesis of a broader
range of polycyclic aromatic ring systems.
In the reductive cyclization of 9a , only the symmetrical
products of double cyclization to the same ring (e.g., 13a
and 15) were detected as significant products. The
unsymmetrical PAH compound that might be expected
to be formed by partial cyclization in each direction (e.g.,
27) was not found, although minor amounts may have
Formation of the fully aromatic PAH benzo[rst]pen-
taphene (2) and its 3,4,9,10-tetramethoxy derivative 13a
in the reactions of the dialdehydes 9a and 9b with 1,3-
propanediol and triflic acid (Schemes 3 and 4) rather than
their 5,8-dihydro derivatives (e.g., 14), though unex-
pected, is not inconsistent with the mechanism for
reductive cyclization proposed by Fukusawa et al.20
According to this concept, the aldehyde functions undergo
initial acid-catalyzed reaction with 1,3-propanediol to
form the corresponding cyclic diacetals (e.g., 24) (Scheme
8). Protonation of an oxygen atom of the diacetal followed
by ring opening affords a benzylic cationic intermediate
that may attack the adjacent aromatic ring system to
generate a bisether intermediate (25). 1,3-Shift of hydride
from the alkoxy carbon to the benzylic carbon atom leads
to formation of a partially reduced monoether (26) with
loss of 3-hydroxypropanal. Compound 26 undergoes
conversion to the fully aromatic compound 2 via acid-
catalyzed elimination of 1,3-propanediol facilitated by
aromatization of the PAH ring system. Elimination of 1,3-
propanediol is evidently energetically more favorable
than a second 1,3-hydride shift with loss of a second
3-hydroxypropanal to furnish 5,8-dihydro-17. Similar
been formed. The reason that symmetrical cyclization is
favored is that the primary partially cyclized intermedi-
ate tends to favor substitution in the same ring by the
alkyl substitution effect.
This synthetic methodology appears to be potentially
broad in scope. It differs from older well-established
synthetic approaches to PAH ring systems, such as the
Haworth synthesis,1a,2,27 in that it entails direct formation
of covalent bonds between aromatic rings in the primary
step. In this respect, it complements photochemical
cyclodehydrogenation of diaryolefins17 wherein covalent
bond formation between aromatic rings to form a new
Sch em e 8

3958 J . Org. Chem., Vol. 65, No. 13, 2000
Zhang et al.
was refluxed for 20 h under argon and cooled, and the organic
solvent was removed under reduced pressure. The resulting
suspension was extracted with CH2Cl2 and dried over MgSO4,
and the solvent was evaporated. The residue was chromato-
graphed on silica gel eluted with CH2Cl2 and then with CH2-
Cl2/EtOAc (90:10) to yield the palladium complex 10a (600 mg)
as a yellow solid, mp 213 °C (CH2Cl2/MeOH): 1H NMR (CDCl3,
500 MHz) δ 3.30 (s, 3), 3.64 (s, 3), 6.45 (d, 1, J ) 8.0 Hz), 7.06
(d, 1, J ) 8.0 Hz), 7.22-7.29 (m,12), 7.30-7.35 (m, 6), 7.57-
7.64 (m, 12), 9.45 (s, 1); 13C NMR (125 MHz) δ 56.47, 61.79,
117.72, 127.66, 127.69, 127.73, 129.72, 129.92 (br), 130.90,
131.08, 131.26, 133.34 (br), 134.66, 134.71, 134.76, 148.89,
153.61, 156.28, 191.28; UV (THF) λmax 298 (ꢀ 24 730) nm. Anal.
ring is the final step. The synthetic method entailing
Suzuki coupling is more versatile than the photochemical
method because it is more amenable to large scale
preparations and because it allows a choice of methods
for cyclization. Two methods of cyclization are described
herein: (1) conversion of an aldehyde function to a vinyl
group followed by photocyclization, and (2) direct reduc-
tive cyclization of an aldehyde substituent.
Although double Suzuki coupling was employed in all
of the syntheses reported, dual coupling is not an
essential feature of the synthetic method. There is no
inherent reason that single, double, or even triple Suzuki
coupling cannot be employed to synthesize a much larger
range of polycyclic aromatic compounds.
Calcd for
C45H39BrO3P2Pd: C, 61.69; H, 4.49; Br, 9.12.
Found: C, 61.83; H, 4.49; Br, 9.25. Further elution gave 9a
(4.1 g, 90%) as a yellow solid, mp 207-209 °C (CH2Cl2/
hexane): 1H NMR (CDCl3, 500 MHz) δ 3.97 (s, 2), 3.98 (s, 4),
4.02 (s, 2), 4.03 (s, 4), 7.12 (d, 1, J ) 8.5 Hz), 7.16 (d, 1, J )
9.0 Hz), 7.22 (d, 1, J ) 9.0 Hz), 7.23 (d, 1, J ) 8.5 Hz), 7.31-
7.39 (m, 4), 7.51-7.56 (m, 2), 9.97 (s, 1), 10.00 (s, 1); 13C NMR
(125 MHz) δ 56.05, 62.09, 116.38, 116.51, 126.03, 126.07,
126.22, 126.60, 126.71, 127.11, 127.43, 129.47, 129.56, 132.42,
132.47, 134.67, 134.86, 136.7, 150.23, 150.39, 152.65, 152.76,
190.76, 191.12; UV (THF) λmax 305 (ꢀ 16 850) nm. Anal. Calcd
for C28H24O6: C, 73.67; H, 5.30. Found: C, 73.55; H, 5.35.
1,4-Bis(2-for m yl-p h en yl)n a p h th a len e (9b). This synthe-
sis was carried out on a 5.5 mmol scale by the procedure for
9a . Chromatography of the product on a silica gel column
eluted with hexane/CH2Cl2 (8:2, 1:1) gave 9b (1.68 g, 91%) as
a white solid, mp 168 °C (CH2Cl2/hexane): 1H NMR (CDCl3,
500 MHz) δ 7.41-7.47 (m, 2), 7.47-7.54 (m, 4), 7.54-7.65 (m,
4), 7.70-7.79 (m, 2), 8.14 (d, 2, J ) 9.5 Hz), 9.70 (s, 0.8), 9.71
(s, 1); 13C NMR (125 MHz) δ 126.28, 126.96, 127.00, 127.22,
127.23, 127.29, 128.40, 131.59, 131.70, 132.64, 133.71, 133.79,
134.74, 136.08, 136.10, 143.68, 143.74, 191.58, 191.84; UV
(THF) λmax 298 (ꢀ 14 550) nm. Anal. Calcd for C24H16O2: C,
85.69; H, 4.80. Found: C, 85.48; H, 4.84.
Exp er im en ta l Section
Ma ter ia ls a n d Meth od s. 6-Bromo-2,3-dimethoxybenzal-
dehyde (8c) was synthesized from 3-methoxysalicylaldehyde
by acetylation, followed by bromination with Br2 and KBr,
hydrolysis, and treatment with dimethyl sulfate by the litera-
ture method.12,13 1,5-Dibromonaphthalene was synthesized by
the procedure previously described.24 THF was freshly distilled
from sodium/benzophenone ketal.
The NMR spectra were recorded on 400 or 500 MHz
spectrometers in CDCl3 with tetramethylsilane as internal
standard unless stated otherwise. Integration was consistent
with all structural assignments. Mass spectra (MS) and HRMS
were performed by the University of Illinois at Urbana-
Champaign, School of Chemical Sciences. The UV spectra were
measured with a Perkin-Elmer Lambda 6 spectrometer. All
melting points are uncorrected. Ca u tion : Benzo[s]picene,
benzo[rst]pentaphene, and dibenzo[b,def]chrysene exhibit car-
cinogenic activity in animal assays. These PAHs and their
dihydrodiol, diol epoxide, and higher oxidized metabolites are
potentially hazardous and should be handled with care in
accordance with NIH Guidlines for the Laboratory Use of
Chemical Carcinogens.
1,4-Bis(n a p h th a len ylbor on ic a cid ) (6c). A solution of
1,4-dibromonaphthalene (14.25 g, 50 mmol) in freshly distilled
anhydrous THF (150 mL) was added dropwise with stirring
to Mg turnings (3.6 g, 150 mmol) in a 2 L, three-neck flask
equipped with a condenser and a dropping funnel under argon.
It was necessary to initiate reaction by adding a few crystals
of iodine to 30 mL of the solution and warming to 45 °C. The
remaining 120 mL of solution was added over a period of 30
min, and then the solution was warmed to 65 °C and allowed
to reflux for 12 h. The resulting light green slurry was allowed
to cool to room temperature, and then it was immersed in a
dry ice-acetone bath and cooled to -65 °C. Trimethyl borate
(250 mmol) in dry THF (50 mL) was added dropwise with
stirring over 1h, then the mixture was slowly warmed to room
temperature and stirred overnight. The mixture was then
cooled in a dry ice bath and hydrolyzed with 2 N HCl. After
decomposition of the excess Mg turnings, the solution was
extracted with ether. The organic layer was washed with brine
and dried (MgSO4). After removal of the solvent, the residue
was recrystallized from ether to give 6c as a white solid (4.2
g). The mother liquid was recrystallized from ether-hexane
to afford the second crop of 6c as a yellowish solid (1.8 g), mp
> 300 °C: 1H NMR (DMSO, 500 MHz) δ 7.50-7.60 (m, 2H),
7.65 (s, 2H), 8.25-8.45 (m, 4H) [after addition of D2O, changed
to 8.30-8.40 (m, 2H)].
Rea ction of th e P a lla d iu m Com p lex 10a w ith 6c. To a
suspension of 10a (110 mg, 0.126 mmol) in DME (4 mL) was
added a solution of 6c (13.6 mg, 0.063 mmol) in EtOH (0.5
mL), and the mixture was stirred for 15 min at room temper-
ature under Ar. TLC showed no reaction occurred. Then 2 M
aqueous Na2CO3 (2 mL) was added, the resulting mixture was
heated at reflux for 20 min under Ar, and a black palladium
solid formed. The usual workup gave the dialdehyde 9a (13
1
mg, 48%), identical by H NMR with an authentic sample.
1,4-Bis(2-vin yl-3,4-d im eth oxyp h en yl)n a p h th a len e (11).
To a solution of methyltriphenylphosphonium bromide (1.78
g, 5 mmol) in 30 mL of dry THF was added 2 mL of a 2.5 M
n-butyllithium solution in hexane at room temperature under
Ar. The resulting yellow solution was stirred for 15 min, then
a solution of 9a (684 mg, 1.5 mmol) in dry CH2Cl2 (12 mL)
was added, and stirring was continued overnight at room
temperature. Evaporation of the solvent under reduced pres-
sure gave a solid residue, which was chromatographed on a
silica gel column eluted with hexane/CH2Cl2 and then with
CH2Cl2. Compound 11 (620 mg, 91%) was obtained as a white
foam, mp 58-60 °C (hexane): 1H NMR (CDCl3, 500 MHz) δ
3.88 (s, 6), 3.97 (s, 6), 5.02-5.11 (m, 2), 5.43 (dd, 1, J ) 18.0
and 2.0 Hz), 5.60 (dd, 1, J ) 18.0 and 2.0 Hz), 6.34-6.50 (m,
2), 6.93-6.97 (m, 2), 7.02-7.08 (m, 2), 7.12-7.18 (m, 4), 7.52-
7.57 (m, 2); UV (THF) λmax 303 (ꢀ 17 930) nm; EI MS m/z 452
(M+). Anal. Calcd for C30H28O4: C, 79.62; H, 6.24. Found: C,
79.37; H, 6.31.
3,4,9,10-Tetr a m eth oxyben zo[s]p icen e (12). Argon was
bubbled through a solution of 11 (410 mg, 0.91 mmol) and I2
(461 mg, 1.82 mmol) in 500 mL of benzene for 15 min. Then
1,2-epoxybutane (15 mL) was added, and the mixture was
irradiated with a Hanovia 450 W medium-pressure mercury
lamp through a Pyrex filter with stirring for 4 h. TLC showed
reaction to be complete. After removal of the solvent under
vacuum, the residue was taken up in CH2Cl2, and the solution
was washed with 10% aqueous Na2S2O3, dried over MgSO4,
and evaporated to dryness. Chromatography on a short column
of silica gel eluted with CH2Cl2 afforded 12 (320 mg, 81%) as
1,4-Bis(2-for m yl-3,4-dim eth oxyph en yl)n aph th alen e (9a).
To a solution of 6-bromo-2,3-dimethoxybenzaldehyde (8c) (6.47
g, 25 mmol) in DME (80 mL) was added Pd(PPh3)4 (1.0 g, 0.86
mmol), and the mixture was stirred under argon for 15 min.
A solution of 6c (2.16 g, 10 mmol) in EtOH (20 mL) was added,
the mixture was stirred for another 15 min, and then 2 M
aqueous Na2CO3 (100 mL)was added. The resulting mixture
(26) Rice, J . E.; Cai, Z.-W. J . Org. Chem. 1993, 58, 1415.
(27) .J ohnson, W. S. Org. React. 1944, 2, 114.

New Synthetic Approaches to PAHs
J . Org. Chem., Vol. 65, No. 13, 2000 3959
a white solid, mp 302-303 °C dec (CHCl3), identical with an
authentic sample: 1H NMR (CDCl3, 500 MHz) δ 4.09 (s, 6),
4.10 (s, 6), 7.43 (d, 2, J ) 9.0 Hz), 7.60-7.65 (m, 2),), 8.37 (d,
2, J ) 9.0 Hz), 8.63 (d, 2, J ) 9.0 Hz),), 8.75 (d, 2, J ) 9.5 Hz),
8.90-8.94 (m, 2).
(0.5 mL) was added, and the solvent was evaporated under
vacuum. Chromatography of the residue on a silica gel column
eluted with CH2Cl2/hexane (1:1) then CH2Cl2 provided in turn
13a and 16. Compound 13a (66 mg, 31%) was obtained as a
yellow solid identical in its physical properties to 13a from
procedure A. Compound 16 was obtained as a pinkish solid
(52 mg, 25%), mp 302-304 °C (CH2Cl2/hexane), Rf ) 0.54 (CH2-
Cl2/EtOAc, 20:1): 1H NMR (CDCl3, 500 MHz) δ 2.30-2.36 (m,
2), 3.97 (s, 6), 4.04 (s, 6), 4.44-4.50 (m, 2), 4.56-4.61 (m, 2),
5.98 (s, 2), 7.00 (d, 2, J ) 8.5 Hz), 7.58-7.62 (m, 2), 7.92 (d, 2,
J ) 8.0 Hz), 8.65-8.70 (m, 2); 13C NMR (125 MHz) δ 30.73,
56.12,61.00, 72.54, 79.51, 112.66, 118.62, 125.04, 125.86,
129.98, 133.15, 135.43, 139.69, 139.77, 146.71, 152.16; UV
(THF) λmax 397 (ꢀ 28 470), 378 (29 260), 271 (30 000) nm. Anal.
Calcd for C31H28O6: C, 74.98; H, 5.68. Found: C, 74.87; H,
5.67.
Acid -Ca ta lyzed Cycliza tion of th e Dia ceta l 9c. Meth od
A. To a solution of 9c (200 mg, 0.35 mmol) in anhydrous 1,2-
dichloroethane (100 mL) was added triflic acid (10.5 mg, 0.07
mmol), and the solution was refluxed with stirring for 1 h.
Then it was cooled to room temperature, Et3N (0.5 mL) was
added, and the solvent was removed under vacuum. Chroma-
tography of the residue on a silica gel column eluted with CH2-
Cl2/hexane (1:1) and then with CH2Cl2 provided in turn 13a
(20 mg, 13.5%) and 15 (20 mg, 13.5%). Meth od B. To a
solution of 9c (286 mg, 0.5 mmol) in freshly distilled anhydrous
1,4-dioxane (150 mL) was added triflic acid (15 mg, 0.1 mmol).
The solution was refluxed with stirring for 45 min, then cooled
to room temperature, and worked up as before to afford in turn
13a (60 mg, 28.4%) and 16 (48 mg, 22.6%).
Dia ceta l of 9a (9c). A mixture of 9a (912 mg, 2 mmol),
1,3-propanediol (456 mg, 6 mmol), p-toluenesulfonic acid (20
mg), and benzene (100 mL) was placed in a round-bottom flask,
equipped with a Dean-Stark apparatus. The solution was
refluxed with stirring for 2 h until no additional water was
collected. The solution was cooled to room temperature,
triethylamine (0.5 mL) was added, and the solvent was
removed. The residue was dissolved in CH2Cl2, washed with 2
N NaOH, and dried over K2CO3. Evaporation of the solvent
furnished pure 9c (1.06 g, 93%) as a white solid, mp 104-107
°C (CH2Cl2/hexane): 1H NMR (CDCl3, 500 MHz) δ 1.04-1.14
(m, 2), 1.85-1.95 (m, 2), 3.25-3.51 (m, 4), 3.80-4.03 (m, 4),
3.94 (s, 6), 3.98 (s, 6), 5.38 (s, 1), 5.43 (s, 1), 7.03 (m, 4), 7.30-
7.37 (m, 2), 7.39 (s, 1), 7.41 (s, 1), 7.60-7.68 (m, 2); 13C NMR
(125 MHz) δ 25.52, 25.68, 56.16, 56.24, 61.52, 67.06, 67.09,
67.12, 67.22, 100.18, 100.30, 113.01, 113.06, 125.08, 125.10,
126.77, 126.81, 126.90, 127.08, 131.29, 131.57, 132.57, 133.02,
133.08, 137.86, 137.98, 148.25, 148.48, 152.88, 153.00; UV
(THF) λmax 291 (ꢀ 15 000) nm. Anal. Calcd for C34H36O8: C,
71.31; H, 6.34. Found: C, 71.23; H, 6.45.
1,4-Bis(2-h yd r oxym eth yl-3,4-d im eth oxyp h en yl)n a p h -
th a len e (9d ). To a solution of 9a (150 mg, 0.33 mmol) in 10
mL of CH2Cl2/EtOH (v/v, 1:1) was added NaBH4 (50 mg, 1.32
mmol). The mixture was stirred at room temperature for 1 h.
TLC showed the reaction was complete. The reaction mixture
was poured into ice-water, extracted with CH2Cl2, washed
with brine, and dried over MgSO4. After removal of the solvent,
the residue was recrystallized from CH2Cl2/hexane to give 9d
as a white solid (145 mg, 96%), mp 187-189 °C (CH2Cl2/
hexane): 1H NMR (CDCl3, 500 MHz) δ 3.96 (s, 2), 3.97 (s, 4),
4.02 (s, 4), 4.03 (s, 2), 4.31 (d, 1, J ) 12.0 Hz), 4.33 (d, 1, J )
11.0 Hz), 4.48 (d, 1, J ) 12.0 Hz), 4.49 (d, 1, J ) 12.0 Hz),
7.00 (d, 1, J ) 8.0 Hz), 7.02 (d, 1, J ) 8.0 Hz), 7.05 (d, 1, J )
8.5 Hz), 7.08 (d, 1, J ) 8.5 Hz), 7.36-7.42 (m, 4), 7.54-7.60
(m, 2); 13C NMR (125 MHz) δ 55.82, 59.01, 59.12, 61.24, 111.77,
111.81, 126.04, 126.07, 126.31, 126.35, 126.38, 126.46, 126.86,
132.70, 132.75, 132.77, 132.82, 133.56, 133.65, 137.50, 137.57,
147.77, 147.87, 151.97; UV (THF) λmax 290 (ꢀ 14 920) nm. Anal.
Calcd for C28H28O6: C, 73.02; H, 6.13. Found: C, 72.72; H,
6.09.
Red u ctive Cycliza tion of 9a . Meth od A. To a mixture
of 9a (228 mg, 0.5 mmol) and 1,3-propanediol (114 mg, 1.5
mmol) in anhydrous 1,2-dichloroethane (150 mL) was added
trifluoromethanesulfonic acid (triflic acid) (15 mg, 0.1 mmol).
The solution was refluxed with stirring for 1 h and then cooled
to room temperature. Triethylamine (0.5 mL) was added, and
the solvent was removed. Chromatography of the residue on
silica gel [CH2Cl2/hexane (1:1), CH2Cl2] gave 13a followed by
15. Compound 13a ; Rf ) 0.78 (CH2Cl2/EtOAc, 20:1) as a yellow
solid (32 mg, 15%); mp 308-309 °C (CH2Cl2/hexane): 1H NMR
(CDCl3, 125 MHz) δ 4.12 (s, 6), 4.15 (s, 6), 7.57 (d, 2, J ) 8.5
Hz), 7.81 (s, 2), 8.59 (s, 2), 8.75 (d, 2, J ) 8.5 Hz), 9.07 (s, 2);
13C NMR (125 MHz) δ 56.77, 61.46, 114.23, 118.46, 119.28,
122.22, 124.50, 129.05; UV (THF) λmax 408 (ꢀ 62 570), 386
(42 930), 365.7 (18 790), 350 (17 130), 318 (63 820), 284
(30 714) nm. Anal. Calcd for C28H22O4: C, 79.60; H, 5.25.
Found: C, 79.50; H, 5.31. Compound 15; Rf ) 0.63 (CH2Cl2/
EtOAc, 20:1) as a gray solid (27 mg, 13%), mp 286-287 °C
(CH2Cl2/hexane): 1H NMR (CDCl3, 500 MHz) δ 3.98 (s, 6), 4.04
(s, 6), 4.06 (s, 4), 7.06 (d, 2, J ) 8.5 Hz), 7.62-7.67 (m, 2), 8.05
(d, 2, J ) 8.5 Hz), 8.78-8.83 (m, 2); 13C NMR (125 MHz) δ
33.36, 56.16, 60.35, 111.64, 118.29, 124.52, 125.27, 126.26,
129.25, 134.88, 137.02, 137.37, 138.29, 145.53, 150.86; UV
(THF) λmax 384 (ꢀ 34 000), 366 (35 300), 267 (34 000), 258
(32 770) nm. Anal. Calcd for C28H24O4: C, 79.60; H, 5.25.
Found: C, 79.50; H, 5.31. Meth od B. To a solution of 9a (228
mg, 0.5 mmol) and 1,3-propanediol (114 mg, 1.5 mmol) in
freshly distilled anhydrous 1,4-dioxane (150 mL) was added
triflic acid (15 mg, 0.1 mmol). The solution was refluxed with
stirring for 1 h and then cooled to room temperature, Et3N
Ben zo[r st]p en ta p h en e (2). Meth od A. Reductive cycliza-
tion of 9b (202 mg, 0.6 mmol) and purification of the product
by chromatography on silica gel furnished on elution with CH2-
Cl2/hexane (2:8, 1:1) benzo[rst]pentaphene (2) as a yellow solid
(78 mg, 43%), mp 278-280 °C (CH2Cl2/MeOH) (lit.2 280-282
°C): 1H NMR (CDCl3, 500 MHz) δ 7.74-7.85 (m, 4), 7.80 (s,
2), 8.22 (d, 2, J ) 7.5 Hz), 8.32 (s, 2), 9.05 (d, 2, J ) 8.5 Hz),
9.22 (s, 2). Meth od B. Reaction of 9b (202 mg, 0.6 mmol) by
this procedure gave 17 (45 mg, 37%) identical to that obtained
by procedure A.
3,4,9,10-Tetr a a cetoxyben zo[r st]p en ta p h en e (13c). To a
solution of 13a (260 mg, 0.62 mmol) in dry CH2Cl2 (200 mL)
at -20 °C was added a solution of BBr3 in CH2Cl2 (1 M, 12.4
mL). The mixture was stirred at -20 °C for 30 min and then
at room temperature for 2 h. The mixture was cooled with dry
ice, ice was added, and the organic solvent was removed at
room temperature under reduced pressure. The aqueous
suspension was extracted with EtOAc, and the combined
EtOAc solution was washed with brine, dried over Na2SO4,
and evaporated to dryness at room temperature. The solid
residue was dissolved in pyridine (14 mL)/Ac2O (10 mL) at -20
°C and then stirred at room temperature overnight. The
mixture was poured into ice-water and extracted with CH2-
Cl2. The organic layer was washed with 1 N HCl and brine in
turn and dried (Na2SO4). After removal of the solvent, the
residue was applied to a short silica gel column and eluted
with CH2Cl2 and CH2Cl2/EtOAc (95:5) to give 13c (289 mg,
88%) as a yellow solid, mp 325-326 °C (CH2Cl2/hexane): 1H
NMR (CDCl3, 500 MHz) δ 2.43 (s, 6), 2.60 (s, 6), 7.69 (d, 2, J
) 9.0 Hz), 7.83 (s, 2), 8.31 (s, 2), 8.95 (d, 2, J ) 9.0 Hz), 9.16
(s, 2); 13C NMR (125 MHz) δ 20.64, 20.90, 117.79, 121.51,
121.78, 122.69, 123.77, 126.07, 127.40, 127.70, 129.07, 130.89,
137.46, 139.90, 168.32, 168.45; UV (THF) λmax 401 (ꢀ 79 190),
389 (48 800), 359 (21 580), 337 (20 900), 320 (25 500), 301
(70 640), 288 (51 160), 276 (34 760) nm. Anal. Calcd for
C
32H22O8: C, 71.90; H, 4.15. Found: C, 71.81; H, 4.13.
t r a n s-3,4-t r a n s-9,10-Te t r a h yd r oxy-3,4,9,10-t e t r a h y-
d r ob en zo[r st]p en t a p h en e (17). Met h od A. A mixture of
13c (10 mg, 0.019 mmol) and NaBH4 (50 mg, 1.3 mmol) in
100 mL of EtOH/CH2Cl2 (v/v, 70:30) was stirred at room
temperature with O2 bubbling through for 2 h. The color of
the solution changed from yellow to orange red. Then another
50 mg of NaBH4 was added, and the solution was stirred for
another 2 h. The color of the solution changed to light yellow.
The solvent was removed under reduced pressure without

3960 J . Org. Chem., Vol. 65, No. 13, 2000
Zhang et al.
heating, and then cold water was added. The suspension was
acidified and filtered. The solid product was dried and then
triturated with EtOAc to furnish 17 as a yellow solid (4.5 mg,
66%), mp 220-225 °C (dec): 1H NMR (DMSO, 500 MHz) δ
4.42 (d, 2, J ) 10.0 Hz), 4.89 (d, 2, J ) 10.5 Hz), 5.39 (d, 1, J
) 5.0 Hz, OH), 5.84 (d, 1, J ) 6.0 Hz, OH), 6.21 (d, 2, J ) 10.5
Hz), 7.51 (d, 2, J ) 9.5 Hz), 8.12 (s, 2), 8.39 (s, 2), 8.45 (s, 2);
13C NMR (DMSO, 100 MHz) δ 72.16, 74.80, 122.72, 123.08,
123.48, 124.17, 125.74, 126.67, 127.87, 130.50, 135.16, 137.14;
UV (THF) λmax 394 (ꢀ 47 360), 373 (35 550), 355 (16 875), 318
(21 420), 305 (21 400), 276 (25 800) nm; HRMS calcd for
from 1,5-dibromonaphthalene25 (8.0 g) was carried out by the
procedure described for 6c. Compound 20c was obtained as a
gray solid (38%), mp >370 °C: 1H NMR (DMSO, 500 MHz) δ
7.42 (dd, 2, J ) 6.9, 8.3 Hz), 7.65 (d, 2, J ) 6.6 Hz, OH), 8.33
(d, 2, J ) 8.2 Hz), 8.37 (br s, 2, QH, exchangeable with D2O);
13C NMR (DMSO, 125 MHz) δ 124.60, 129.93, 131.34, 135.30;
UV (THF) λmax 290 (ꢀ 6 800), 239 (4 300) nm.
1,5-Bis(2-for m yl-3,4-d im et h oxyp h en yl)n a p h t h a len e
(21a ). Preparation of 21a from 8c and 20c was carried
out by the procedure described for 9a . Compound 21a was
obtained as a yellow solid (63%), mp 289-290 °C (CH2Cl2/
hexane): 1H NMR (500 MHz) δ 3.98 (s, 2), 4.00 (s, 4), 4.02 (s,
2), 4.03 (s, 4), 7.09 (d, 1, J ) 8.0 Hz), 7.15 (d, 1, J ) 8.5 Hz),
7.18-7.42 (m, 6), 7.50 (d, 1, J ) 8.0 Hz), 7.55 (d, 1, J ) 8.5
Hz), 9.95 (s, 1), 9.98 (s, 1); 13C NMR (125 MHz) δ 56.13, 56.15,
62.17, 116.39, 116.59, 125.36, 125.44, 125.74, 125.87, 127.24,
127.35, 127.43, 127.72, 129.48, 129.69, 132.49, 135.00, 135.14,
137.03, 137.19, 150.29, 150.53, 152.69, 152.81, 190.93, 191.17;
UV (THF) λmax 304 (ꢀ 12 080), 242 (29 600) nm. Anal. Calcd
for C28H24O6: C, 73.67; H, 5.30. Found: C, 73.58; H, 5.41.
1,2,8,9-Tetr a m eth oxyd iben zo[b,d ef]ch r ysen e (22a ). Re-
ductive acid-catalyzed cyclization of 21a was carried out by
Method B for the analogous reaction of 9a . To a mixture of
21a (228 mg, 0.6 mmol) and 1,3-propanediol (114 mg, 1.5 mol)
in freshly distilled anhydrous dioxane (150 mL) was added
triflic acid (15 mg, 0.1 mmol). The solution was heated at reflux
with stirring for 2 h, then it was cooled to room temperature,
and triethylamine (0.5 mL) was added. The solvent was
removed under reduced pressure, and the residue was chro-
matographed on a silica gel column eluted with CH2Cl2/hexane
(1:1) to yield 22a as a yellow solid (34 mg, 16%), mp 335 °C
(dec): 1H NMR (500 MHz) δ 4.14 (s, 6), 4.20 (s, 6), 7.65 (d, 2,
J ) 8.5 Hz), 8.34 (d, 2, J ) 9.3 Hz), 8.78 (d, 2, J ) 9.3 Hz),
8.88 (d, 2, J ) 9.0 Hz), 8.95 (s, 2); UV (THF) λmax 322 (ꢀ 52 800),
310 (28 100) nm; HRMS calcd for C28H22O4 m/z 422.1518, found
422.1509.
1,2,8,9-Tetr a a cetoxyd iben zo[b,d ef]ch r ysen e (22d ). Con-
version of 22a to 22d was carried out by the procedure
described for 13c. Compound 22d was obtained as a green-
yellow solid (85%), mp 347 °C (CH2Cl2/hexane) (dec): 1H NMR
(500 MHz) δ 2.43 (s, 6), 4.00 (s, 3.9), 4.62 (s, 6), 7.68 (d, 2, J )
9.5 Hz), 8.34 (d, 2, J ) 9.5 Hz), 8.66 (s, 2), 8.89-9.00 (m, 4);
UV (THF) λmax 451 (ꢀ 25 900), 425 (18 030), 314 (121 780), 301
(60 890), 259 (23 580) nm; HRMS calcd for C32H22O8 m/z
534.1315, found 534.1304.
C
24H18O4 m/z 370.1205, found 370.1206. HPLC and NMR
showed that 17 obtained by this procedure was contaminated
by the over-reduced products 18 and 19. Pure 17 was more
efficiently prepared by Method B. Meth od B. A mixture 13c
(53.4 mg, 0.1 mmol) and NaBH4 (200 mg, 5.3 mmol) in 200
mL of EtOH/CH2Cl2 (v/v, 50:50) was stirred at room temper-
ature with O2 bubbling through for 24 h. The solvent was
removed under reduced pressure without heating, and then
cold water was added. The suspension was acidified and
filtered to afford a yellow solid. HPLC analysis using a Zorbax
ODS (9.4 mm × 25 cm) column that was eluted isocratically
with MeOH/0.02% TFA in water (v/v, 50:50) at a flow rate of
4 mL/min showed the product to be a mixture of 17 (retention
times for isomers, 15.8 and 17.8 min) and products of further
reduction (18 and 19). Compound 18 showed one peak at 14.1
min, and the retention times for the isomers of 19 were 11.4
and 19.0 min. The ratios of UV absorbency at Å260/Å370 were
∼1.2, 2.1, and 5 for compounds 17, 18, and 19, respectively.
This mixture was dissolved in pyridine (14 mL)/Ac2O (10 mL)
and heated overnight. The usual workup afforded a mixture
of acetates of 17, 18, and 10. This mixture was heated at reflux
with DDQ (150 mg) in benzene for 20 h. After the usual
workup, the crude product was purified on a silica gel column
[CH2Cl2/hexane (1:1), CH2Cl2] to give the tetraacetate of 17
as a yellow solid (25 mg, 46% in total), mp 223-225 °C (CH2-
Cl2/MeOH): 1H NMR (CDCl3, 500 MHz) δ 2.07 (s, 6), 2.17 (s,
6), 5.74 (m, 2), 6.32 (d, 1, J ) 10.0 Hz), 6.33 (d, 1, J ) 10.0
Hz), 6.55 (d, 2, J ) 6.0 Hz), 7.69 (d, 2, J ) 10.0 Hz), 8.00 (s,
2), 8.10 (s, 2), 8.37 (s, 2); 13C NMR (125 MHz) δ 21.03, 21.18,
69.51, 71.74, 122.96, 125.18, 125.48, 125.56, 125.81, 126.22,
126.75, 128.02, 129.40, 131.26, 170.32, 170.32; UV (THF) λmax
393 (ꢀ 67 480), 372 (47 850), 354 (20 510), 315 (32 600), 302
(26 840), 276 (38 360) nm. Anal. Calcd for C32H26O8: C, 71.36;
H, 4.87. Found: C, 71.59; H, 4.93.
To a solution of the tetraacetate of 17 (15 mg, 0.028 mmol)
in THF (6 mL)/MeOH (12 mL) was added 0.15 mL of solution
of MeONa (25 wt % in MeOH) at room temperature. The
resulting solution was stirred for 3 h, and then the mixture
was poured into ice-water, acidified, and extracted with
EtOAc. The organic layer was washed with brine and cold
water and dried (Na2SO4). The solvent was removed under
reduced pressure at room temperature, and the residue was
triturated with EtOAc to give a yellow solid (8 mg, 78%), mp
220-225 °C (dec), identical in its physical properties to an
authentic sample of 17 from Method A.
tr a n s-3,4-tr a n s-9,10-Tetr a h yd r oxy-1,2,3,4,9,10,11,12-oc-
ta h yd r oben zo[r st]p en ta p h en e (19). A mixture 17 (10 mg,
0.019 mmol) and NaBH4 (50 mg, 1.3 mmol) in 100 mL of EtOH/
CH2Cl2 (v/v, 50:50) was stirred at room temperature with O2
bubbling through for 24 h. Another 50 mg of NaBH4 was
added, and the solution was stirred for another 48 h during
which time it became colorless. The solvent was removed under
reduced pressure without heating, then cold water was added,
and the suspension was acidified and filtered. The solid was
dried and then triturated with EtOAc to furnish 19 (5 mg, 70%)
as a gray solid, mp 224-227 °C (dec): 1H NMR (DMSO, 500
MHz) δ 1.96 (m, 2), 2.26 (m, 2), 3.36 (m, 2), 3.48 (m, 2), 3.87
(m, 2), 4.70 (br d, 2, J ) 6.0 Hz), 4.96 (d, 1, J ) 3.5 Hz, OH),
5.55 (d, 1, J ) 6.0 Hz), 7.98 (s, 2), 8.28 (s, 2), 8.31 (s, 2); 13C
NMR (125 MHz) δ 23.40, 27.25, 70.23, 72.96, 123.03, 123.06,
125.57, 126.47, 127.31, 129.04, 130.04, 136.85; HRMS calcd
for C24H23O4 m/z 375.1596, found 375.1591.
tr a n s-1,2-tr a n s-8,9-Tet r a h yd r oxy-1,2,8,9-t et r a h yd r o
d iben zo[b,d ef]ch r ysen e (23). A mixture 22d (10 mg, 0.019
mmol) and NaBH4 (50 mg, 1.3 mmol) in 60 mL of EtOH/CH2-
Cl2 (v/v, 50:50) was stirred at room temperature with O2
bubbling through for 2 h. The color changed from yellow to
orange-red. Then another 50 mg of NaBH4 was added, and
the solution was stirred for another 4 h during which time
the color changed to pale yellow. The solvent was removed
without heating, then cold water was added, and the suspen-
sion was acidified and filtered. The solid was dried and
triturated with EtOAc to furnish 23 as a yellow solid (5.0 mg,
71%), mp 222-226 °C (dec): 1H NMR (DMSO, 500 MHz) δ
4.43 (d, 2, J ) 10.0 Hz), 4.89 (d, 2, J ) 10.5 Hz), 6.22 (d, 2, J
) 10.0 Hz), 7.51 (d, 2, J ) 10.0 Hz), 8.15 (d, 2, J ) 9.5 Hz),
8.39 (s, 2), 8.43 (d, 2, J ) 9.5 Hz); 13C NMR (DMSO, 125 MHz)
δ 71.75, 74.42, 122.29, 122.33, 122.51, 123.70, 125.90, 126.47,
127.90, 129.47, 134.71, 136.74; UV (THF) λmax 401 (ꢀ 36 900),
378 (38 990), 305 (22 500), 264 (39 690) nm; HRMS calcd for
C
24H18O4 m/z 370.1205, found 370.1206.
HPLC analysis of 23 on a ZORBAX ODS column (9.4 mm
× 25 cm) eluted isocratically with MeOH/0.02% TFA (v/v, 50:
50) at a flow rate of 4 L/min showed the presence of two
isomers with retention times of 11.5 and 12.6 min, respectively.
Ack n ow led gm en t. This research was supported by
a grant from the National Cancer Institute (CA 67937).
1,5-Bis(n a p h th a len ylbor on ic a cid ) (20c). Preparation
J O9918044