Highly strained dihydroanthraquinones: Oxidation versus elimination

Article abstract of DOI:10.1016/j.tetlet.2004.02.041

The chemistry of strained dihydroanthraquinones was investigated for the purpose of developing syntheses for highly strained anthraquinones. The reaction of (1,4-dihydro-9,10-dioxo-anthracen-1-yl)-acetates with triethylamine under aerobic conditions was found to be dependent on the degree of substitution on the acetate. Dihydroanthraquinones bearing dimethylacetates underwent elimination of enolate exclusively, while those without acetate substitution dehydrogenated as expected. Furthermore, oxidation of the cyclohexadiene ring using iodine failed due to a competitive iodolactonization reaction. The desired strained anthraquinones could be prepared, in low yield, by treatment of the resulting lactones with acidic ethanol.

Full text of DOI:10.1016/j.tetlet.2004.02.041

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 2955–2959
Highly strained dihydroanthraquinones:
oxidation versus elimination
John D. Reynolds, Robert G. Brinson, Cynthia S. Day and Paul B. Jones*
Department of Chemistry, Wake Forest University, Winston-Salem, NC 27109, USA
Received 21 January 2004; revised 4 February 2004; accepted 10 February 2004
Abstract—The chemistry of strained dihydroanthraquinones was investigated for the purpose of developing syntheses for highly
strained anthraquinones. The reaction of (1,4-dihydro-9,10-dioxo-anthracen-1-yl)-acetates with triethylamine under aerobic con-
ditions was found to be dependent on the degree of substitution on the acetate. Dihydroanthraquinones bearing dimethylacetates
underwent elimination of enolate exclusively, while those without acetate substitution dehydrogenated as expected. Furthermore,
oxidation of the cyclohexadiene ring using iodine failed due to a competitive iodolactonization reaction. The desired strained
anthraquinones could be prepared, in low yield, by treatment of the resulting lactones with acidic ethanol.
ꢀ 2004 Published by Elsevier Ltd.
During the course of research involving the photo-
chemistry of hydroxyanthraquinones, we required
1-hydroxyanthraquinones (see Scheme 1) with bulky
substituents in either the 5- or 8- position.1 Our syn-
thetic strategy for these molecules started with a Diels–
Alder reaction between dienes 1 and juglone (2a), to give
tetrahydroanthraquinone adducts.2 Our strategy then
planned for oxidation of the cyclohexene and aromati-
zation of the resulting cyclohexadiene.3
adducts to cyclohexadienes and then to anthraquinones
was not anticipated to be difficult.6
However, 3a displayed an unusual sensitivity to mild
bases resulting in the loss of the acetate functionality
(Scheme 1).7 The expected oxidation product (4a) was
not detected. Instead 1-hydroxy-7-methylanthraquinone
(5a) was obtained in 91% yield. We suspected the
elimination of an enolate, in a reaction similar to the
elimination of methanol from 1-methoxy-1,4-dihydro-
anthraquinones.8 The dimethylacetate moiety of 3a, and
the resulting steric congestion in the desired oxidation
step, was suspected as the cause of this unexpected
reactivity.
1,4-Dihydroanthraquinone acetates (e.g., 3a) are con-
stitutional isomers of the corresponding 9,10-dihydr-
oxyanthracenes (hydroquinones), and they were
expected to oxidize in a similar fashion. The oxidation
of anthrahydroquinones to anthraquinones with oxygen
is a well known and usually facile reaction.4 Treatment
with air under basic conditions is a common method of
oxidizing hydroquinones.5 The oxidation of Diels–Alder
The result prompted our further investigation of this
novel reaction. A series of 1,4-dihydroanthraquinone
compounds (3b–e) was prepared to study the reaction
CO2Et
CO2Et
O
O
OH
O
OH
O
OH
Et3N
not
air, CH2Cl2
91%
3a
5a
4a
O
O
Scheme 1. Elimination of enolate from 3a.
Keywords: Anthraquinone; Oxidation; Strain.
* Corresponding author. Tel.: +1-336-758-3708; fax: +1-336-758-4656; e-mail: jonespb@wfu.edu
0040-4039/$ - see front matter ꢀ 2004 Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2004.02.041

2956
J. D. Reynolds et al. / Tetrahedron Letters 45 (2004) 2955–2959
Table 1. Product distribution in oxidation/elimination of di-
hydroanthraquinone acetates
R
R
R5O2C 1R2
R5O2C 1R
2O
R6
O
O
R6
1. AlCl3, CH2Cl2
R3
R3
Entry
Substrate
4
5
6
7
2. SiO2, CH2Cl2
50-80%
1
2
3
4
5
3a
3b
3c
3d
3e
0
0
0
91
85
74
8
0
0
0
Trace
1
2
3
0
21
0
R4
R4
O
R1
Me
R2
R3
R4
R5
R6
45
83
44
17
0
0
a
b
Me
Me
H
Et
OH
Me
Me
H
H
H
H
Me
H
Me
H
c
d
e
Me
Me
H
Me
H
Bn
Bn
Bn
H
H
H
H
the a-position, and 3e, which lacks methyl groups at the
a-position, was distinct. The former compounds
underwent elimination of enolate exclusively upon
treatment with base to afford an anthraquinone and
isobutyrate. The latter successfully oxidized to the
substituted anthraquinone (as well as dimerization).
H
H
Scheme 2. Preparation of dihydroanthraquinones.
(Scheme 2).9 These compounds were obtained using an
AlCl3 catalyzed Diels–Alder reaction of 1,4-naphtho-
quinone 2b and 3,5-diene esters, 1b–e. Silica gel chro-
matography of the initial cyclohexene adducts (not
shown) resulted in partial conversion of the cyclohexene
ring to a 1,4-cyclohexadiene. Therefore, the crude Diels–
Alder adducts were stirred with silica gel to give 1,4-
dihydroanthraquinones 3a–c.
The dependence of the reaction pathway upon substi-
tution at the a-position of the acetate group argues for a
steric role in guiding the preferred pathway. Two pos-
sible mechanisms to account for the observed products
are shown in Scheme 4. For the dimethylacetates, the
only observed reaction pathway was enolate elimination
(step a, mechanism 1). Attack by the enolate at the
carbonyl of the newly formed anthraquinone led to
hydroxyanthrone, 7c (b). This step presumably must
occur before solvent separation permits protonation of
the enolate (c) to form ester 8c (see below).
The behavior of 3a–e under alkaline aerobic conditions
was investigated (Scheme 3). Equal concentrations of
substrates 3a–e were treated with triethylamine at room
temperature in oxygen saturated dichloromethane. In
each case, either an anthraquinone acetate 4 or
unfunctionalized anthraquinone 5 was the major prod-
uct. Formation of oxidation product 4 was accompanied
by the formation of varying amounts of dimer 6;10;11
elimination product 5 was accompanied by anthrone
product 7 and ester products 8 and 9. Isolated yields for
products 4–7 are shown in Table 1.
In contrast, in the case of unsubstituted 3e, oxidation
(mechanism 2) is the dominant pathway and elimination
of enolate was not observed. Deprotonation of 3e pro-
duced a semiquinone oxyanion, 10e (d). Instead of
elimination, single-electron transfer occurred between
molecular oxygen and 10e to give semiquinone radical
11e and superoxide (e) as part of the usual mechanism of
hydroquinone oxidation.12 Protonation of superoxide
affords hydroperoxyl radical and regenerates triethyl-
amine. Abstraction of the methine hydrogen of 11e by
The difference in reactivity between dimethylacetates
3a–c, which have some degree of methyl substitution at
O
R1
R2
O
O
R6
R5O
R3
O
R6
R3
R4
R4
O
O
OR5
R1
R2
4a-e
5a-e
O
R6
O
O
R1
OR5
R3
0.1M TEA, air
CH2Cl2
R2
R1
R2
R6
O
OR5
2
R6
OH
R3
R4
H
O
3a-e
O
6a-e
R4
R4
7a-e
O
O
O
H
HO
R1
OR5
OR5
R1
R2
R2
8a-e
9a-e
Scheme 3. Products of the TEA elimination/oxidation of substituted dihydroanthraquinones.

J. D. Reynolds et al. / Tetrahedron Letters 45 (2004) 2955–2959
2957
(Et)3NH+
TEA
O
O
O
OBn
O
8c
OBn
OH
OBn
c
a
b
1)
+
5c (AQ)
H
H
O
3c
7c
O
(Et)3N
O
O
O
BnO
H
BnO
H
BnO
H
O
O
O
O
O2
HOO
TEA
2)
d
(Et)3NH+
(Et)3N
H
H
O
3e
O
10e
e
11e
O
O
O
H
BnO
O
BnO
H
H
BnO
O
H2O2
O
O
HOO
2
f
g
6e
4e
11e
O
O
Scheme 4. Proposed mechanism of elimination/oxidation.
hydroperoxyl radical aromatized the ring and gave
anthraquinone acetate 4e (f). In competition with this
step, the semiquinone radical 11e may dimerize to the
bis-1,2-dihydroanthraquinone, 6e (g), though this is a
minor reaction for 3e.
Though it is unclear why the elimination should proceed
so much faster than the oxidation in 3c, the results of
our experiments are consistent with a steric role in the
outcome of the reaction. Oxidation of dihydroanthra-
quinones 3a–e requires the acetate group to move from a
position above the quinone ring into the plane of the
quinone (11 fi 4). During this movement, the acetate
group must pass close to the proximal quinone oxygen.
In 3a–c, the geminal dimethyl substitution requires that
one methyl group make a close contact with this oxygen
(data not shown). The steric interaction of the methyl
and oxygen is enough to greatly retard the rate of oxi-
dation. In 3e, the methyl groups are absent and the
protons on the acetate methylene have sufficient room to
slide by the quinone oxygen.
Figure 1. X-ray crystal structure of 3c.
Elimination of enolate, on the other hand, relieves
unfavorable steric interactions in the dihydroanthra-
quinones. The X-ray crystal structure of 3c is shown in
Figure 1.13 An examination of the structural parameters
shows a lengthening of the C1–C15 sp3–sp3 bond to
quinone carbonyl, as would be required in going from
3c to 4c.
ꢀ
1.574(4) A and an opening of the tetrahedral angle at
The formation of anthrone product 7c suggests that the
leaving group in the elimination reaction is an enolate.14
Once free, this enolate adds to the quinone carbonyl
before separation and protonation occurs. In a control
reaction, triethylamine failed to promote the addition of
benzyl isobutyrate to anthraquinone under the condi-
tions of the oxidation. This suggests that the production
of anthrone occurs from a caged pair. However, benzyl
hydroxyisobutyrate 9c was observed when benzyl
isobutyrate was treated with TEA in the presence of
C13–C1–C15 to 112.6(2)ꢁ. This region of the molecule is
sterically crowded with a number of short intramolec-
ular nonbonded contacts. The contacts involve carbonyl
ꢀ
carbon (C16) [C16ꢀ ꢀ ꢀO28, 2.937(4) A; C16ꢀ ꢀ ꢀC9,
ꢀ
ꢀ
3.086(4) A; C16ꢀ ꢀ ꢀC13, 3.005(4) A], oxygen O28
ꢀ
(O28ꢀ ꢀ ꢀH21, 2.69 A), and methyl carbons C17
ꢀ
ꢀ
(C17ꢀ ꢀ ꢀH2, 2.81 A) and C18 [C18ꢀ ꢀ ꢀC14, 3.425(4) A;
ꢀ
C18ꢀ ꢀ ꢀC3, 3.422 A]. These close contacts likely create a
barrier to the acetate group becoming coplanar with the

2958
J. D. Reynolds et al. / Tetrahedron Letters 45 (2004) 2955–2959
Acknowledgements
O
O
EtO
O
O
OH
O
OH
The authors thank the North Carolina Biotechnology
Center (NCBC #2001ARG0009) and the ACS Petro-
leum Research Fund (PRF #37882-G4) for financial
support and Dr. Marcus Wright for NMR assistance.
O
a
c
I
3a
12
O
O
b
O
O
OEt
O
OH
OH
O
References and notes
1. (a) Miki, S.; Matsuo, K.; Yoshida, M.; Yoshida, Z.
Tetrahedron Lett. 1988, 29, 2211–2214; (b) Miki, S.;
Kagawa, H.; Matsuo, K.; Kobayashi, O.; Yoshida, M.;
Yoshida, Z. Tetrahedron 1992, 48, 1567–1572; (c) Nakay-
ama, T.; Torii, Y.; Nagahara, T.; Miki, S.; Hamanoue, K.
J. Phys. Chem. A 1999, 103, 1696–1703; (d) Smart, R. P.;
Peelen, T. J.; Blankespoor, R. L.; Ward, D. L. J. Am.
Chem. Soc. 1997, 119, 461–465; (e) Blankespoor, R. L.;
DeJong, R. L.; Dykstra, R.; Hamstra, D. A.; Rozema, D.
B.; VanMeurs, D. P.; Vink, P. J. Am. Chem. Soc. 1991,
113, 3507–3513.
4a
13
O
O
Scheme 5. Conversion of 3a to 4a via an iodolactone. Reagents and
conditions: (a) I2, H2O/CH3OH; (b) TEA, CH2Cl2, 88% from 3a; (c)
H2SO4, EtOH, D, 120 h, 31%.
oxygen, demonstrating that 9c is a secondary product
formed after elimination.
Substrate 3d provides an interesting midpoint between
the dimethyl substrates 3a–c and unsubstituted 3e.
Anthraquinone (elimination product) was produced in
small (8%) yield. Enolate elimination for 3d is expected
to be slower than enolate elimination for 3a–c due to
reduced steric strain in the dihydroanthraquinone. Sur-
prisingly, the quantity of dimer 6d produced was nearly
equal to the amount of oxidation product 4d. This
suggests that the dynamic concentration of radical 11d
must increase relative to that found in the oxidation of
3e in order for dimerization (g) to compete with aro-
matization. Hydrogen abstraction (f) thus appears to be
rate limiting for the oxidation of substrates 3d–e;
abstraction must be markedly slower for 3d. If the oxi-
dation reaction of 3d is quenched with aqueous acid
within a few minutes of amine addition, a significant
amount of a 1,2-dihydroanthraquinone acetate product
is observed in addition to the usual products. Further-
more, when 3d is treated with triethylamine under
anaerobic conditions, no oxidation is observed while the
amount of elimination (5c) is unchanged.
2. (a) Valderrama, J. A.; Gonzalez, M. F.; Valderrama, C.
Tetrahedron 1999, 55, 6039–6050; (b) Bingham, S. J.;
Tyman, J. H. P. J. Chem. Soc., Perkins Trans. 1 1997,
3637–3642; (c) Motoyoshiya, J.; Kameda, T.; Asari, M.;
Miyamoto, M.; Narita, S.; Aoyama, H.; Hayashi, S.
J. Chem. Soc., Perkins Trans. 2 1997, 1845–1850.
3. Walker, D.; Hiebert, J. D. Chem. Rev. 1967, 67, 153–191.
4. Naruta, Y.; Maruyama, K. In The Chemistry of the
Quinonoid Compounds; Patai, S., Rappoport, Z., Eds.;
Wiley: London, 1988; Vol. 2, Chapter 8, pp 242–402.
5. (a) Beagley, B.; Curtis, A. D.; Pritchard, R. G.; Stoodley,
R. J. J. Chem. Soc., Perkin Trans. 1 1992, 1981–1991; (b)
Kelly, T. R.; Parekh, N. D. J. Org. Chem. 1982, 47, 5009–5013.
6. De Riccardis, F.; Izzo, I.; Di Filippo, M.; Sodano, G.;
DÕAcquisto, F.; Carnuccio, R. Tetrahedron 1997, 53,
10871–10882.
7. A dihydroanthraquinone was isolated in low yield (<10%)
that had a 1,5,6-substitution pattern as opposed to the
1,7,8 pattern in 3a. Although this minor regioisomer was
obtained in too little quantity for thorough investigation,
it appeared to behave similarly when treated with base.
8. Manning, W. B.; Kelly, T. P.; Muschik, G. M. Tetra-
hedron Lett. 1980, 21, 2629–2632.
1
9. Data for 3b: mp 118–119 ꢁC. H NMR (CDCl3) d 1.06 (s,
3H), 1.25 (s, 3H), 1.39–1.42 (d, J ¼ 7:2), 3.60–3.69 (s, 1H),
3.65 (s, 3H), 4.16–4.19 (t, J ¼ 4:5, 1H), 5.75–5.80 (dd,
J ¼ 5:5, 4.5, 1H), 5.98–6.03 (dd, J ¼ 5:5, 4.3, 1H), 7.70–
7.74 (m, 2H), 7.93–8.01 (m, 1H), 8.05–8.10 (m, 1H). 13C
NMR (CDCl3) 184.2, 182.9, 177.1, 146.9, 143.4, 132.5,
132.4, 131.6, 131.5, 131.3, 125.3, 125.2, 122.0, 51.0, 45.1,
40.3, 30.0, 22.8, 21.7, 19.9. HRMS calcd for
(C20H20O4+Naþ): 347.125378. Found: 347.12513. Anal.
Calcd for C24H20O4: C, 74.06; H, 6.21. Found: C, 73.92;
H, 6.20.
Finally, we attempted to oxidize the cyclohexadiene ring
of 3a with iodine (Scheme 5). Again, oxidation to the
anthraquinone was not observed. Instead, iodolacton-
ization occurred.15 The iodolactone, 12, converted
slowly to 13 under ambient conditions and, therefore,
was immediately treated with triethylamine to produce
13 in 88% yield from 3a.16–18 This compound could be
converted, in low yield, to the desired anthraquinone 4a
by heating 13 in 0.5 M H2SO4 in ethanol for 120 h.
1
10. Data for 4d: mp 123–124 ꢁC; H NMR (CDCl3): 1.66 (d,
J ¼ 7:2 Hz, 3H), 5.01 (q, J ¼ 7:2 Hz, 1H), 5.12 (d,
J ¼ 12:4 Hz, 1H), 5.17 (d, J ¼ 12:4 Hz, 1H), 7.22–7.26
(br s, 5H), 7.66 (dd, J ¼ 7:7, 1.7 Hz, 1H), 7.70 (t,
J ¼ 7:7 Hz, 1H), 7.75–7.78 (m, 2H), 8.18–8.21 (m, 1H),
8.23–8.26 (m, 1H), 8.33 (dd, J ¼ 7:7, 1.7 Hz, 1H). 13C
NMR (CDCl3): 185.3, 183.5, 174.1, 143.5, 136.2, 135.4,
134.9, 134.4, 134.0, 133.9, 132.8, 130.9, 128.6, 128.3, 128.2,
127.7, 127.3, 126.9, 66.7, 43.6, 29.9. Anal. Calcd: C, 77.82;
H, 4.90. Found: C, 77.52; H, 4.87.
In summary, we have found that sterically congested
dihydroanthraquinones exhibit unusual reactivity pat-
terns when exposed to base or oxidants. Oxidation to
the corresponding anthraquinone is not observed.
Alternative reaction pathways, such as elimination of
enolate or iodolactonization involving esters are com-
petitive with the desired oxidation. We continue to
investigate the interesting chemistry of strained anthra-
quinones and will report on additional syntheses of
these molecules and their chemistry in the future.
11. Data for 6d: mp 171 ꢁC (dec); 1H NMR (CDCl3): d 0.90 (d,
J ¼ 7:2 Hz, 6H), 2.46 (dq, J ¼ 7:2, 4.5 Hz, 2H), 2.79 (d,

J. D. Reynolds et al. / Tetrahedron Letters 45 (2004) 2955–2959
2959
J ¼ 5:3 Hz, 2H), 3.25 (d, J ¼ 4:5 Hz, 2H), 4.97 (d,
J ¼ 12:8 Hz, 2H), 5.01 (d, J ¼ 12:8 Hz, 2H), 5.89 (dd,
J ¼ 9:7, 5.3 Hz, 2H), 6.98 (d, J ¼ 9:7 Hz, 2H), 7.11 (dd,
J ¼ 7:7, 0.9 Hz, 2H), 7.29–7.41 (m, 12H), 7.60 (dt, J ¼ 7:7,
1.5 Hz, 2H), 8.07 (dd, J ¼ 7:7, 0.9 Hz, 2H). 13C NMR
(CDCl3): d 184.0, 182.5, 173.90, 137.12, 137.05, 135.9,
134.6, 133.4, 133.3, 132.0, 131.9, 128.8, 128.7, 128.5, 126.4,
125.6, 122.9, 66.8, 43.3, 37.5, 30.7, 12.9. HRMS calcd for
(C48H38O8+Naþ): 765.24588. Found: 765.24474.
Davies, S. R.; Hambley, T. W. Aust. J. Chem. 1994, 47,
2221–2233.
16. The cis stereochemistry for 12 has not been established.
However, an analog of 12 lacking the C-7 methyl was
prepared and did possess a cis ring fusion (data not
shown).
1
17. Data for 13: mp 170–172 ꢁC. H NMR (CDCl3) d 1.10 (s,
3H), 1.51 (s, 3H), 1.56 (s, 3H), 3.63 (s, 1H), 6.32 (d,
J ¼ 9:8 Hz, 1H), 6.85 (d, J ¼ 9:8 Hz, 1H), 7.32 (dd,
J ¼ 8:2, 1.4 Hz, 1H), 7.66 (t, J ¼ 8:2 Hz, 1H), 7.69 (dd,
J ¼ 8:2, 1.4 Hz, 1H), 12.11 (s, 1H). 13C NMR (CDCl3) d
189.3, 181.5, 180.3, 161.7, 138.3, 138.0, 137.2, 136.8, 131.5,
125.0, 119.7, 118.7, 115.1, 81.8, 48.2, 45.2, 28.1, 27.4, 21.1.
HRMS calcd for (C19H16O5+Naþ): 347.088993. Found:
347.08774.
12. (a) Saba, M. M.; Stephens, A. D. Chem. Phys. Lett. 1999,
306, 249–255; (b) Venkataraman, B.; Fraenkel, G. K.
J. Am. Chem. Soc. 1955, 77, 2707–2713; (c) Sieiro, C.;
Sanchez, A.; Crouigneau, P. Spectrochim. Acta A 1984,
40(5), 453–456.
13. Crystallographic data for 3c have been deposited with the
Cambridge Crystallographic Data Centre as supplemen-
tary publication number CCDC 229187. Copies of the
data can be obtained free of charge on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK (Fax:
+44(0)-1223-336033 or e-mail: deposit@ccdc.cam.ac.uk).
14. 7c was prepared, in low yield, by treating benzyl isobu-
tyrate with LDA and quenching the enolate with anthra-
quinone.
18. Data for 4a: mp 163–165 ꢁC. 1H NMR (CDCl3) d 1.06
(t, J ¼ 7:2 Hz, 3H), 1.86 (br s, 6H), 2.65 (s, 3H), 4.02
(q, J ¼ 7:2 Hz, 2H), 7.26 (dd, J ¼ 8:3, 1.1 Hz, 1H), 7.50 (d,
J ¼ 7:9 Hz, 1H), 7.62 (t, J ¼ 8:3 Hz, 1H), 7.74
(dd, J ¼ 8:3, 1.1 Hz, 1H), 8.13 (d, J ¼ 7:9 Hz, 1H),
11.88 (s, 1H). 13C NMR (CDCl3) d 192.6, 182.7, 178.2,
161.6, 146.2, 145.1, 138.2, 136.3, 135.8, 132.9, 132.9, 125.8,
124.3, 118.9, 117.4, 60.5, 49.3, 24.6, 14.2, 1.2. HRMS
calcd for (C21H20O5+Naþ): 375.120293. Found:
375.12098.
15. (a) Amano, S.; Ogawa, N.; Ohtsuka, M.; Chida, N.
Tetrahedron 1999, 55, 2205–2224; (b) Crossley, M. J.;