Synthesis of asymmetrically disubstituted anthracenes

Article abstract of DOI:10.1016/j.tet.2017.08.038

We have developed synthetic pathways toward differently substituted hydroxyanthracenes (anthrols) with the aim to investigate their photochemical reactivity in dehydration reactions. Although the syntheses of anthracenes substituted at positions 9,10 are well known, reports for the synthesis of anthracenes with different substitution patterns are scarce. Herein we review known and report novel synthetic pathways toward anthrols with substituents at 1,2-, 2,3-, and 2,6- positions. We present two synthetic approaches: (i) building of the anthracene tricyclic fused ring system from the appropriate benzene derivatives, and (ii) reduction of the corresponding anthraquinones. Reduction of 2-hydroxyanthracene-1-carbaldehyde to the corresponding alcohol yields rather unexpected 1,1′-methylenedianthracen-2-ol, whose proposed mechanism of formation is supported by experimental observations and calculations.

Full text of DOI:10.1016/j.tet.2017.08.038

Tetrahedron 73 (2017) 5892e5899
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Synthesis of asymmetrically disubstituted anthracenes
*
ꢀ
ꢁ
ꢀ
ꢁ
Ðani Skalamera , Jelena Veljkovic, Lucija Pticek, Matija Sambol, Kata Mlinaric-Majerski,
ꢁ
Nikola Basaric
ꢀ
ꢁ
ꢀ
Department of Organic Chemistry and Biochemistry, RuCer Boskovic Institute, Bijenicka cesta 54, 10 000 Zagreb, Croatia
a r t i c l e i n f o
a b s t r a c t
Article history:
We have developed synthetic pathways toward differently substituted hydroxyanthracenes (anthrols)
with the aim to investigate their photochemical reactivity in dehydration reactions. Although the syn-
theses of anthracenes substituted at positions 9,10 are well known, reports for the synthesis of an-
thracenes with different substitution patterns are scarce. Herein we review known and report novel
synthetic pathways toward anthrols with substituents at 1,2-, 2,3-, and 2,6- positions. We present two
synthetic approaches: (i) building of the anthracene tricyclic fused ring system from the appropriate
benzene derivatives, and (ii) reduction of the corresponding anthraquinones.
Reduction of 2-hydroxyanthracene-1-carbaldehyde to the corresponding alcohol yields rather unex-
pected 1,10-methylenedianthracen-2-ol, whose proposed mechanism of formation is supported by
experimental observations and calculations.
Received 30 May 2017
Received in revised form
9 August 2017
Accepted 18 August 2017
Available online 24 August 2017
Keywords:
Anthracene
1-Anthrol
2-Anthrol
© 2017 Elsevier Ltd. All rights reserved.
Dehydroxymethylation
Anthracene carbaldehyde
1. Introduction
and medicine, it is important to develop QM precursors that absorb
visible light, which is itself harmless for cells. Our group has studied
Anthracene derivatives are the subject of many research areas
due to applications arising from their spectroscopic properties and
photoreactivity. They are commonly utilized as fluorescent probes,1
for labeling in biological systems, e.g. as a chemical trap for singlet
molecular oxygen in cells2 or bioimaging.3 The well known
anthracene photodimerization was thoroughly studied in solution,4
and used to study the reactivity and selectivity in supramolecular
complexes.5 Dimerization in solid state can be used for high-
density 3D optical storage memory devices.6 Furthermore, it has
been demonstrated that anthracene derivatives can be used for
photoinduced DNA-cleavage,7 and therefore, they have potential
for the use in photochemotherapy.8,9
An on-going interest in our group is the photochemical gener-
ation of quinone methides (QM) from suitable precursors, and
investigation of their biological effects. QMs are common reactive
intermediates in chemistry and photochemistry of phenols and
phenol-related compounds.10 Their biological activity arises from
their ability to react with proteins and DNA.11 Since QMs can be
photochemically generated from precursor compounds directly in
cancer cells, they have a potential for the use in cancer therapy
(photochemotherapy).12 However, for real application in biology
the mechanisms of QM generation and biological activity of ben-
zene,13 biphenyl14 and naphthalene series of QM precursors.15
However, it was necessary to further increase the chromophore
in order to shift the absorption spectrum to the visible region of
light, which is harmless for the living cell. That was achieved by
introducing the anthracene as a chromophore. Development of
anthracene QM precursors, bearing hydroxy and hydroxymethyl
groups, required design of new synthetic pathways. For the syn-
thesis of symmetrically disubstituted anthracenes or anthracenes
substituted at positions 9,10 many examples can be found in the
literature.16 On the other hand, the syntheses of asymmetrically
substituted anthracenes (with two or more different substituents)
are scarce, often including a large number of steps and/or suffering
from low yields and formation of mixtures which are difficult to
separate.17 Herein we review known and report novel synthetic
approaches to 2,3-, 1,2-, and 2,6-substituted anthracene derivatives
1e5 (Fig. 1). As will be shown, in some cases these syntheses pro-
vided surprising results, which were fully rationalized by additional
experiments and calculations.
2. Results and discussion
It is well known that the electrophilic substitution reaction on
anthracene occurs at positions 9 and 10, with the highest electron
* Corresponding author.
E-mail address: Djani.Skalamera@irb.hr (Ð. Skalamera).
ꢀ
http://dx.doi.org/10.1016/j.tet.2017.08.038
0040-4020/© 2017 Elsevier Ltd. All rights reserved.

ꢀ
Ð. Skalamera et al. / Tetrahedron 73 (2017) 5892e5899
5893
OH
OH
aluminum powder was non-reproducible and yielded only a small
amount of very impure 3-iodo-2-anthrol. Thus, we have developed
the approach which uses bromination instead of iodination, since
the aromatic bromides are more stable than iodides. Bromination of
6 led to a mixture of bromides 7e9. When the bromination was
carried out with a small excess of bromine (1e2 eq.), the reaction
was not complete even at prolonged heating to reflux (2 weeks),
1
OH
OH
CH2
OH
OH
and significant amount of unwanted isomer
9 was present
4
5
2
(10e20%), together with the dibromo derivative 8 (30%). If the re-
action was carried out with a large excess of bromine (5 eq.) for-
mation of a 1:1 mixture of bromide 7 and dibromide 8 occurred,
which can be separated by column chromatography.22 Many
methods for the reduction of anthraquinones to anthracene de-
rivatives are described in the literature, but in case of bromide 7,
many problems occurred, often yielding some undesired products.
We have tried many methods for the reduction of anthraquinone 7
before a simple and successful method was found. The most widely
used method for the anthraquinone reduction is the use of
refluxing conc. HI.20,23 In the case of 7, these severe reaction con-
ditions led to the substitution of the OH-group and bromine with
iodine, together with the over-reduction of anthracene, yielding
2,3-diiodo-9,10-dihydroanthracene as the main product. Second
often used method for the conversion of anthraquinones to an-
thracenes is the reduction with zinc dust in ammonia.24,25 In these
conditions, debromination of compound 7 occurred, yielding 2-
anthrol in high yield. The typical methods for the reduction of ke-
tone to methylene are Clemensen and Wolff-Kishner reductions.
The Clemensen reduction was not applicable on compound 7, since
debromination occurred in the reaction with zinc. In the Wolff-
Kishner reduction, 7 was heated first with hydrazine in diglyme,
but the corresponding bishydrazone was not formed even after
prolonged reflux time, or in triglyme at higher reflux temperature.
Therefore, the Wolff-Kishner reduction was not usable in this case,
and it also failed on attempt to reduce methylated derivative of 7
(2-OMe instead of 2-OH). Furthermore, the reduction of 7 with
LiAlH4 yielded 2-anthrol. The reduction of anthraquinones with
NaBH4 is usually carried out in two steps - reduction to anthrone
followed by its reduction to anthracene.26 When the reduction of
hydroxyanthraquinones with NaBH4 was carried out in alkaline
water it proceeded in one step only.27 We used the latter method
for the reduction of bromide 7 and dibromide 8. Interestingly, both
compounds gave the desired 3-bromo-2-hydroxyanthracene (10).
Therefore, we tried to perform the reduction with NaBH4 in alkaline
aqueous solution (1 M Na2CO3) on crude mixture of 7 and 8, which
resulted with 10 as the product in high yield. Under the reduction
conditions a selective debromination occurred on the anthracene
position 1 in 8, whereas debromination on the position 3 was
insignificant if the reduction was not run longer than 15 min. In this
way, 10 was obtained selectively, in high yield (95%) and with high
purity (>90%), without a need for chromatographic separation in
three consecutive reaction steps.8a,9 These steps can be carried out
on milligram or multigram scale of compounds.
OH
OH
OH
HO
3
Fig. 1. Target molecules in this work.
density.18 Thus, anthracene is not a suitable precursor for the syn-
thesis of derivatives unsubstituted at the positions 9 and 10. If 2-
anthrol was employed as a precursor, in addition to the positions
9 and 10, the reactions of electrophilic substitution would take
place at both ortho positions to the hydroxyl group (1 and 3). The
separation of anthracene regioisomers is difficult.17 Consequently,
anthracenes substituted at other positions than 9 and 10 are less
common and more expensive, since their preparation requires a
specific synthetic approach or multi-step synthesis. Two main ap-
proaches for the synthesis of anthracene derivatives are usually
employed: the first includes building of the tricyclic ring system
from some suitable benzene derivatives,19 whereas the second is
based on the synthetic transformation of some appropriately
substituted anthraquinone derivatives, in which the fused tricyclic
system is already assembled. In our synthetic attempts to obtain
2,6-substituted anthracene 5 we tried both approaches. The second
approach, which uses anthraquinone derivatives as precursors, was
proven to be better, since it enabled also to access compounds 1e4.
An advantage of the anthraquinone precursors in the synthesis is
the fact that reactive anthracene 9,10-positions are protected,
directing the substitution to one of the outer rings. After the key
synthetic steps to introduce the substituents at the desired posi-
tions, the keto-groups at the positions 9,10 can be reduced to afford
corresponding anthracene derivative. Many methods of anthra-
quinone reduction were described in the literature (vide infra), but,
most of them are incompatible with the OH or/and halogen sub-
stituents on the anthraquinone, as will be described below.
2.1. Synthesis of 2,3-substituted anthracenes
Synthesis of 3-hydroxyanthracene-2-carbaldehyde, precursor of
compound 1, is already described in the literature, but the reported
synthesis results in a very poor yield caused by the difficult sepa-
ration of 1- and 3-carbaldehyde isomers on 2-anthrol.17 To avoid
these problems, we decided to develop a simple and easily scalable
synthesis of 2,3-substituted anthracenes. As a starting material for
this synthesis, inexpensive 2-aminoanthraquinone was used that
already has the tricyclic fused ring system. It can be easily con-
verted to hydroxy-derivative 6 via a known procedure (Scheme
1).20 Anthraquinone 6 contains the desired structural motif e the
hydroxy group at the anthraquinone position 2, with the positions
9 and 10 blocked for the electrophilic substitution. In the next step,
the substituent at the anthraquinone position 3 was introduced.
Selective iodination of hydroxyanthraquinone 6 at the position 3 is
known,21 and it was reproducible in our hands. However, reduction
of iodoanthraquinone to the corresponding anthracene using
Bromoanthrol 10 was treated with an excess BuLi and the
resulting organolithium compound was quenched with DMF. After
the hydrolysis of the addition product, aldehyde 11 was obtained in
84% yield. Diol 1 was obtained in high yield after reduction of
aldehyde 11 with NaBH4.8a The overall yield in the synthetic
pathway from 6 to 1 is 77% (4 steps).
2.2. Synthesis of 1,2-substituted anthracenes
Aldehyde 16 was synthesized starting from tetraline according
to the known procedure with some modifications, in similar yields
(Scheme 2).28,29 In the first step, tetralin reacted with succi-
nanhydride in the Friedel-Crafts conditions to give keto-acid 12,30

ꢀ
5894
Ð. Skalamera et al. / Tetrahedron 73 (2017) 5892e5899
Scheme 1. Synthesis of 2,3-substituted anthracene derivatives.
Scheme 3. Reduction of aldehyde 16 with NaBH4.
probably undergoes dehydration giving QM2 in a thermal reaction.
Formation of QMs in thermal reaction has been demonstrated in
naphthalene series.15 As QM2 is very reactive, it probably formed
dimers, trimers, etc., finally leading to polymers (broad signals in
NMR spectra).
Aldehyde 18 was prepared according to the known procedure
starting from 2-anthrol (17), which was subjected to Rieche for-
mylation (Scheme 4).28 After a chromatographic separation, pure
18 was obtained in 40% yield.
Scheme 2. Synthetic pathway toward aldehyde 16.28
Reduction of aldehyde 18 by NaBH4 was significantly different
than the reduction of aldehyde 16, although, 16 and 18 are posi-
tional isomers. In the case of 18, reduction with NaBH4 furnished
dianthrylmethane 4 as the main product after the acidic work-up of
the reaction or chromatography on silica gel (Scheme 5). All at-
tempts to purify alcohol 3 by crystallization or chromatography
and then, the keto group was reduced in a Clemensen reaction to
give acid 13.31 The tricyclic system was obtained in a cyclization
with CH3SO3H catalyst,32 where 14 is formed under kinetic control
in high yield and purity if the reaction is conducted for a short time.
Ketone 14 in the presence of a strong base (NaH) gave a carbanion
which reacted with ethyl formate to give product 15 in 81%
yield.33,34 Finally, aromatization of the policyclic skeleton was car-
ried out with in situ generated trityl cation giving target carbalde-
hyde 16 in 87% yield.28
We have shown that the reduction of 3-hydroxyanthracene-2-
carbaldehyde (11) is very clean reaction, resulting with hydrox-
ymethyl alcohol 1 in high yield.8a As aldehyde 16 is a substitution
isomer of 11, it was expected that the reduction of 16 with NaBH4
would deliver the corresponding alcohol 2 (Scheme 3). However,
the reduction of 16 under the same reaction conditions as for 11
gave only decomposition products. The reaction was tested at
different conditions (solvents and solvent mixtures: THF, THF-H2O
(1:1), EtOH, EtOH-H2O (1:1); and reagents: NaBH4, LiAlH4 in ether),
but without any success.
Unexpected instability of compound 2 can be tentatively
explained by formation of ortho-QM (Scheme 3). Compound 2
Scheme 4. Formylation of 2-anthrol (17) in Rieche conditions.

ꢀ
Ð. Skalamera et al. / Tetrahedron 73 (2017) 5892e5899
5895
Scheme 5. Reduction of aldehyde 18 and hydroxymethylation of 2-anthrol (17).
failed, giving dianthrylmethane 4 in good yields (40e75%).
It is interesting to investigate how dianthrylmethane 4 is
formed. Is it the primary product in the reaction of aldehyde 18, or
reduction gives alcohol 3 that undergoes dimerization? To further
investigate the formation of 4, it was important to synthesize pure
3, and test its anticipated dimerization. We conducted the synthesis
of dianthrylmethane 4 according to the known procedure35 by
formylation of 2-anthrol (17) with formalin in acidic or basic so-
lution (Scheme 5), which we found highly reproducible. The
“dimeric” derivative 4 was isolated and completely characterized
(NMR, IR, MS). However, preparation of pure “monomeric” alcohol
3 was more demanding. When only 1 eq. of formalin was used, 4
was formed in a matter of seconds after mixing all reagents. In the
second attempt, anthrol 17 was added slowly to a large excess of
formalin at 0 ꢀC, which also resulted in the formation of 4 only. This
led us to the conclusion that anthrol 17 reacted faster with 3, than
with formaldehyde. Therefore, the formylation of 17 cannot be
stopped at the stage of alcohol 3.
To prepare anthrol 3, the reduction of 18 with NaBH4 (Scheme 5)
was optimized. When the reaction was carried out on a small scale
(ꢁ150 mg of 18) we attempted to isolate product 3 by extraction,
during which formation of 4 occurred. In order to avoid the
extraction, we tried different solvents (MeOH, EtOH, THF) in the
reduction step. THF-H2O (1:1) was proved to be the best solvent,
but only when a larger amount of 18 was subjected to the reduction
(>800 mg). Due to its low solubility in that solvent and higher
quantity in large batches, alcohol 3 precipitated from the solution
during the reaction. Finally, alcohol 3 was obtained in purity of
~80% (NMR), and although further purification was not possible, 1H,
13C NMR, and ESI-MS spectra provided reliable evidence for the
assignation of the structure to alcohol 3.
Scheme 6. Proposed mechanism for the formation of 4.
Formation of dianthrylmethane 4 from alcohol 3 was probed in
a series of experiments at basic, neutral and acidic conditions. The
solution of 3 stirred with silica gel was also tested for the reaction.
We found that in neutral and basic solution 4 was formed within
20 h in 4% yield, whereas in acidic solution 4 was formed in 9% yield
within first 30 min. In the sample which contained silica gel 4 was
formed in 34% after 20 h. HPLC chromatograms of acidic and silica
gel samples also showed that 2-anthrol (17) was present in small
quantity in solution (confirmed according to the same retention
time as the authentic sample of 17 and HPLC-MS). There is no
possibility that 17 was an impurity in the starting material, thus it
has probably been formed from 3 in acid-catalyzed reaction. The
above described experiments provide a strong evidence that dia-
nthrylmethane 4 is formed by dimerization of alcohol 3, and not
from aldehyde 18. However, dianthrylmethane 4 has one C-atom
less than two molecules of alcohol 3. The observed anthrol 17
formed in the acidic conditions provides an evidence that
hydroxymethylation reaction can be reversible in the case of 3. That
led us to propose the mechanism for the formation of 4, which
includes ipso-substitution on the anthracene, followed by an
extrusion of formaldehyde in a six-membered transition state 21,
driven by the stability of aromatization (Scheme 6).
anthracene 3, where formaldehyde serves as a leaving group.
Formaldehyde can be easily detected by HPLC after derivatization
with 2,4-dinitrophenylhydrazine (DNPH), according to the known
procedure for the determination of low formaldehyde concentra-
tions in solution.36 Thus, the presence of formaldehyde in the so-
lution of alcohol 3 in CH3CN-H2O (4:1) in the presence of 0.1 mM
TFA was detected 20 h after mixing all components. Finally, to
explain why ipso-substitution on the anthracene occurs, rather
than at some other position (e.g. at 9 or 10, which are usually the
most reactive) we performed theoretical studies on alcohol 3 and
quinone methide QM3. Structures of 3 and QM3 were optimized to
energy minimum using Gaussian software37 at B3LYP/6-311G(d,p)
level of theory. Molecular orbitals for both compounds were
calculated (see Fig. S1 in SI). On the optimized structure of 3 partial
atomic charges (Mülliken charges) were calculated (Fig. 2),
showing that the highest electron density in 3 is at the position 1 of
anthracene, making this position the most reactive. LUMO orbital of
QM3 has the highest density on the exocyclic methylene group (see
Fig. S1 in SI), making this position the most reactive for nucleophilic
attack, the typical for QMs.
All findings from experiments together with theoretical calcu-
lations provide strong support to the proposed reaction mechanism
The proposed mechanism requires ipso-substitution on

ꢀ
5896
Ð. Skalamera et al. / Tetrahedron 73 (2017) 5892e5899
oxidation with potassium permanganate, acid 28 was obtained.38b
OH
However, all attempts to carry out the retro-Diels-Alder reaction
failed, leading only to the traces of anthracene carboxylic acid in a
non-separable mixture with byproducts and tars. Finally, methyl
ether in compound 26 was cleaved off using standard BBr3 pro-
cedure leading to the formation of anthrol 29, which was fully
characterized.
Although the above described pathway did not resulted with
our target molecule 5, it provided a convenient reaction pathway
toward new compound e methylated anthrol 29, which can be
easily further functionalized at the OH group.
Therefore, for the synthesis of diol 5, we chose a different syn-
thetic approach (Scheme 8), which used 2,6-diaminoanthraquinone
as a starting material, which was diazotized and converted to
dibromo derivative 30 by the known procedure.42 Dibromide 30 has
the same substituents at 2,6-positions and therefore a desymmet-
rization reaction had to be developed. Anthraquinone 30 can be
reduced to dibromoanthracene 31 via a known procedure in 54%
yield.26a,43 Lithium exchange of one only bromine in 31 was
attempted in the reactions with BuLi, followed by the quenching of
lithiated compound with ester of boronic acid, to introduce the OH
group and break the symmetry of the molecule. Although a range of
conditions were tried, the reaction did not take place, or resulted in
dihydroxyanthracene. Therefore, desymmetrization should be con-
ducted in some other step of the synthesis. It is known that dibro-
0.042
-0.070
-0.053
-0.048
-0.053
-0.059
-0.014
-0.152
-0.088
-0.095
OH
-0.02 5
-0.067
0. 185
-0.072
-0 .043
Fig. 2. Mülliken charges calculated on B3LYP/6-311G(d,p) optimized structure of 3
showing that the position 1 of anthracene has the highest electron density. Charges on
hydrogen and oxygen atoms were omitted for clarity (see SI).
shown on Scheme 6.
2.3. Synthesis of 2,6-substituted anthracenes
In our attempts to develop a synthetic procedure towards
asymmetrically substituted QM precursor 5 we tried two ap-
proaches: building of tricyclic ring system, and synthetic trans-
formation of the appropriately substituted anthraquinone
derivative. The second approach, which uses anthraquinone de-
rivative as a precursor, was proven to be better. However, both
pathways will be discussed herein with a review of problems
encountered.
moanthraquinone 30 can be transformed to
a bromofluoro-
The first synthetic approach comprised of building the fused
tricyclic ring system from suitable benzene derivatives (Scheme 7),
using the modified synthetic protocols described in literature pre-
cedent.38 p-Methoxybenzoic acid was used as a starting compound,
which was converted to oxazoline 22 by heating with 2-amino-2-
methyl-1-propanol according to the known procedure.39 The oxa-
zoline ring is a good ortho-directing group for H-Li exchange on the
benzene ring.40 In the reaction with n-BuLi oxazoline 22 was con-
verted to the ortho-lithiated compound, which was quenched with
p-tolualdehyde to afford adduct 23. The yield on the desired
product 23 was very low due to the formation of byproducts and
low conversion in the reaction. We tried to optimize the reaction by
varying the ratio of reactants, but without success. Purification of
the product in this step was not possible due to a complete
decomposition on silica gel. Therefore, the crude 23 was subjected
to an intramolecular cyclization in acidic conditions, leading to the
cleavage of the oxazoline ring and formation of lactone 24, which
was obtained in 45% yield (for 2 steps) after purification on
alumina. Compound 24 was reduced by hydrogenation over Pd/C
catalyst in acetic acid to afford diphenylmethane 25 in 97% yield.
The tricyclic fused ring system was formed in the next step by
intramolecular cyclization of acid 25, catalyzed by polyphosphoric
acid (PPA). The resulting anthrone was immediately reduced with
NaBH4 and after acidic work-up yielded anthracene 26 (18% over
three steps).
After the anthracene skeleton was assembled, the next step was
functionalization of the methyl group. It is known that anthracene
can be selectively brominated in an alkyl side chain by N-bromo-
succinimide, in a radical reaction pathway.41 However, all attempts
to repeat the literature conditions failed, leading to unseparable
mixtures of products, bearing bromine also on the anthracene (at
positions 9 and 10). Thereafter, a different approach was used for
the preparation of target compound 5. The attempts comprised of
the oxidation of the methyl to a carboxylic group, followed by the
transformation to an ester, which should in a reduction with
lithium aluminium hydride lead to the corresponding alcohol.
However, the anthracene ring system of 26 had to be protected
from the oxidation to anthraquinone, which was accomplished in a
Diels-Alder reaction with fumaric acid, forming adduct 27. In the
derivative, in which the fluorine can be substituted with amine in
the next step (18% for two steps).25 Following the same procedure,
but using hydroxyde as a nucleophile in the second step instead of
amine, bromohydroxyanthraquinone 32 was obtained in 23% yield
(in two steps). Reduction of the anthraquinone 32 to anthracene 33
was carried out in one step (40%) using the procedure known on
similar substrates,27 optimized on 2,3-anthracene derivatives (this
work). Crude bromoanthrol 33 was treated with excess of BuLi in the
same conditions as described for the 2,3-substituted derivatives, and
lithiated compound was quenched with DMF to afford aldehyde 34
in 32% yield. Aldehyde 34 was cleanly reduced to alcohol 5 with
NaBH4 in 89% yield, giving the desired 6-(hydroxymethyl)anthracen-
2-ol (5), with two different substituents at the positions 2 and 6.
3. Conclusions
Asymmetrically disubstituted anthracenes with substitution
patterns 2,3, 1,2, and 2,6 on the anthracene moiety were success-
fully synthesized. For all three target substitution patterns we
found that syntheses starting from the appropriately substituted
anthraquinones gave the best results (3e5 reaction steps from
commercially available compound to the target molecule). How-
ever, the usual methods for the reduction of anthraquinone to
anthracene on 2,3-substituted anthraquinone derivatives resulted
with unexpected products (substitution of the OH group, dehalo-
genation). The optimized reduction conditions with NaBH4 were
proven to be very successful and clean, and we have also used it for
the reduction of 2,6-anthraquinone derivatives.
Synthesis of 2-hydroxyanthracene-1-carbaldehyde can be easily
carried out according to the known procedure. However, its further
reduction to the corresponding alcohol resulted with unexpected
product e dianthrylmethane 4. Experimental and theoretical
studies gave a strong support that the formation of 4 was the result
of coupling reaction between diol 3 and thermally generated QM3,
where formaldehyde is a leaving group. The presence of formal-
dehyde in solution was proven by HPLC analysis after derivatiza-
tion, which provides strong evidence that hydroxymethylation
reaction in case of 3 can be reversible, especially in acidic
conditions.

ꢀ
Ð. Skalamera et al. / Tetrahedron 73 (2017) 5892e5899
5897
Scheme 7. Synthetic pathway toward anthrol 29.
4. Experimental section
4.2. Synthesis of 1,2-substituted anthracenes
4.2.1. 1-(Hydroxymethyl)anthracen-2-ol (3)
4.1. General information
Aldehyde 18 (930 mg, 4.18 mmol) was dissolved in THF-H2O
(1:1) (80 mL) and NaBH4 (1 g, 26 mmol) was added. Extensive
foaming occurred, which deceased quickly resulting with a clear
yellow solution. The stirring was continued for 16 h at rt, during
which the product precipitated from the reaction mixture. The
formed yellow powder was collected by filtration, washed with
small amount of ice-cold THF-H2O (1:1) and dried in an evacuated
desiccator over KOH. The product was obtained in the form of an
off-white solid (760 mg, 81%, ~80% pure according to NMR). Further
purification by crystallization or chromatography was not possible
due to spontaneous conversion of 3 to 4. mp > 250 ꢀC (dec.) 1H NMR
1H and 13C NMR spectra were recorded at 300 or 600 MHz at rt
using TMS as a reference and chemical shifts were reported in ppm.
Melting points were determined using a Mikroheiztisch apparatus
and were not corrected. IR spectra were recorded on a spectro-
photometer in KBr and the characteristic peak values were given in
cmꢂ1. HRMS were obtained on a MALDI TOF/TOF instrument. For
the sample analysis
a HPLC was used with C18 (1.8 mm,
4.6 ꢃ 50 mm) column. HPLC runs were conducted at rt (~25 ꢀC) and
chromatograms were recorded using UV detector at 254 nm. For
the chromatographic separations silica gel (0.05e0.2 mm) was
used. Chemicals were purchased from the usual commercial sour-
ces and were used as received. Solvents for chromatographic sep-
arations were used as they were delivered from supplier (p.a.
grade) or purified by distillation (CH2Cl2).
(300 MHz, DMSO-d6) d: 8.38 (s, 1H), 8.08e7.91 (m, 3H), 7.85e7.63
(m, 2H), 7.46e7.30 (m, 2H), 6.98 (d, 1H, J ¼ 9.0 Hz), 5.19 (d, 2H). 13
C
NMR (150 MHz, DMSO-d6) d: 154.2, 131.4, 130.0, 128.3, 127.9, 127.6,
127.2, 126.8, 126.2, 125.0, 123.5, 123.4, 117.0, 113.1, 59.5. IR (KBr,

ꢀ
5898
Ð. Skalamera et al. / Tetrahedron 73 (2017) 5892e5899
Scheme 8. Synthetic pathway toward target compound 5.
cmꢂ1): 3276, 3039, 2826, 1626, 1479, 1464, 1250. HRMS (MALDI-
TOF) calcd for C15H12O2þ 224.0837, found 224.0842.
solvent was removed on a rotary evaporator to give 324 mg of the
crude product in the form of oily brown mass. The product was
purified by chromatography on Florisil® using hexane/Et2O
(1:1 / 0:1) as eluent. The reaction gave product 34 (71 mg, 32%) in
the form of yellow solid. mp 286e287 ꢀC (dec.). 1H NMR (600 MHz,
4.3. Synthesis of 2,6-substituted anthracenes
DMSO-d6) d: 10.19 (s, 1H), 10.11 (s, 1H), 8.71 (s, 1H), 8.67 (s, 1H), 8.35
4.3.1. 2-Hydroxy-6-methylanthracene (29)
(s, 1H), 8.07 (d, 1H, J ¼ 9.0 Hz), 8.04 (d, 1H, J ¼ 9.0 Hz), 7.76 (dd, 1H,
A round-bottomed two-necked flask (250 mL) was charged with
the solution of compound 26 (120 mg, 0.52 mmol) in dry ether and
cooled to 0 ꢀC under nitrogen atmosphere. Over period of 10 min
1 M BBr3 in CH2Cl2 (1.55 mL, 1.55 mmol) was added dropwise. The
stirring was continued for 2 h at 0 ꢀC, then at rt for 16 h. The re-
action was quenched by careful addition of water (100 mL), stirred
for additional 15 min and transferred to an extraction funnel. The
organic phase was separated and the aqueous phase was extracted
with CH2Cl2 (2 ꢃ 50 mL). The combined organic extracts were
washed with water (100 mL), separated and dried over anhydrous
MgSO4. After the filtration and evaporation of the solvent at
reduced pressure, the product was purified on column of silica gel
using CH2Cl2 as eluent to afford the product 29 (52 mg, 48%) in the
form of an off-white solid. mp 213.4e214.6 ꢀC. 1H NMR (DMSO-d6,
J ¼ 9.0 Hz, 1.0 Hz), 7.29 (m, 1H), 7.24 (dd, 1H, J ¼ 9.0 Hz, 2.0 Hz). 13
C
NMR (150 MHz, DMSO-d6) d: 192.4, 156.4, 137.5, 135.0, 133.2, 132.4,
130.4, 129.2, 128.5, 127.9, 127.6, 123.2, 121.2, 120.4, 106.8. IR (KBr,
cmꢂ1): 3299 (s), 1668 (s), 1622 (s), 1423 (s), 1209 (s), 1170 (s), 883
(m), 790 (m). HRMS (MALDI-TOF) calcd for C15H10O2 222.0675,
found 222.0670.
4.3.3. 6-(Hydroxymethyl)anthracen-2-ol (5)
Aldehyde 34 (76 mg, 0.34 mmol) was suspended in absolute
ethanol (10 mL). Sodium borohydride (38 mg, 1.0 mmol) was added
to the orange suspension; resulting the color changed to red-
brown. The stirring was continued for 2 h at rt. The solvent was
removed on a rotary evaporator. The solid residue was suspended
in water (25 mL) and 3 M HCl was added until acidic reaction was
achieved (tested with universal indicator paper). The resulting fine
suspension was transferred to an extraction funnel and extracted
with ether (3 ꢃ 20 mL). The combined organic extracts were dried
over anhydrous MgSO4, filtered and the solvent was evaporated on
a rotary evaporator. The reaction gave product 5 (68 mg, 89%) in the
form of pale yellow solid. mp 220e226 ꢀC. 1H NMR (600 MHz,
300 MHz) d: 9.81 (s, 1H), 8.30 (s, 1H), 8.21 (s, 1H), 7.93 (d, 1H,
J ¼ 9.0 Hz), 7.87 (d, 1H, J ¼ 8.5 Hz), 7.73 (s, 1H), 7.29 (dd, 1H, J ¼ 8.5,
1.5 Hz), 7.19 (d, 1H, J ¼ 2.0 Hz), 7.13 (dd, 1H, J ¼ 9.0, 2.5 Hz), 2.47 (s,
3H). 13C NMR (DMSO-d6, 75 MHz)
d: 154.9, 133.1, 132.3, 130.3, 129.8,
129.6, 128.2, 127.4, 127.2, 126.1, 124.9, 122.7, 106.5, 21.9. HRMS
(MALDI-TOF) calcd for C15H12Oþ 208.0888, found 208.0880.
DMSO-d6) d: 9.78 (s, 1H), 8.38 (s, 1H), 8.23 (s, 1H), 7.94 (d, 1H,
4.3.2. 6-Hydroxyanthracene-2-carbaldehyde (34)
J ¼ 9.0 Hz), 7.91 (d, 1H, J ¼ 9.0 Hz), 7.88 (s, 1H), 7.39 (dd, 1H,
J ¼ 9.0 Hz, 1.5 Hz), 7.21 (m, 1H), 7.14 (dd, 1H, J ¼ 9.0 Hz, 2.0 Hz), 5.26
(t,1H, J ¼ 5.5 Hz), 4.65 (d, 2H, J ¼ 5.5 Hz). 13C NMR (150 MHz, DMSO-
In a two-necked round bottomed flask, equipped with a reflux
condenser and a septum, compound 33 (273 mg, 1.0 mmol) was
dissolved in dry Et2O (12 mL) under N2 atmosphere. The solution
was cooled to ꢂ18 ꢀC (ice/methanol bath) and 2.5 M n-BuLi in
hexanes (2.2 mL, 5.5 mmol) was added dropwise. The stirring was
continued at ꢂ18 ꢀC for the next 140 min, and the reaction mixture
was allowed to heat to rt for 30 min, and finally refluxed for 20 min.
The reaction mixture was cooled again to ꢂ18 ꢀC and dry DMF
(0.4 mL, 5.0 mmol) was added dropwise. After the stirring 1 h at rt,
the reaction was quenched by a careful addition of a few drops of
water, followed by the addition of 3 M HCl until acidic reaction was
achieved (tested with universal indicator paper). The mixture was
transferred to an extraction funnel and extracted with EtOAc
(3 ꢃ 20 mL). The combined organic extracts were washed with
water (50 mL), dried over anhydrous MgSO4, filtered and the
d6) d: 154.6, 138.1, 132.6, 131.1, 129.7, 129.2, 127.4, 127.2, 125.6, 125.3,
124.2,122.7, 120.4,106.6, 63.2. IR (KBr, cmꢂ1) 3413 (s), 3188 (s), 1631
(s), 1199 (s), 1172 (s), 894 (m). HRMS (MALDI-TOF) calcd for
C15H12O2 224.0832, found 224.0825.
4.3.4. Stability test for alcohol 3 in acidic, basic, neutral solution
and in the presence of silica gel
Alcohol 3 was dissolved in CH3CN-H2O (4:1) and the solution
was divided into three equal parts. First was left as it is (neutral), in
the second DIPEA was added (basic), and in the third TFA was added
(acidic). Final concentration of base or acid was 0.1 mM. Fourth
sample was prepared by dissolving 3 in diethyl ether (5 mL) and it

ꢀ
Ð. Skalamera et al. / Tetrahedron 73 (2017) 5892e5899
5899
10. Rokita SE, ed. Quinone Methides. Hoboken: Wiley; 2009.
11. Wang P, Song Y, Zhang L, He H, Zhou X. Curr Med Chem. 2005;12:2893.
12. (a) Basaric N, Mlinaric-Majerski K, Kralj M. Curr Org Chem. 2014;18:3;
(b) Percivalle C, Doria F, Freccero M. Curr Org Chem. 2014;18:19.
was left standing over silica gel (100 mg). All samples were stirred
in absence of light and changes were monitored by HPLC.
ꢁ
ꢁ
ꢀ
ꢁ
ꢀ ꢁ
ꢁ
13. Basaric N, Zabcic I, Mlinaric-Majerski K, Wan P. J Org Chem. 2010;75:102.
5. Theoretical calculations
ꢁ
14. (a) Basaric N, Cindro N, Bobinac D, et al. Photochem Photobiol Sci. 2011;10:1910;
ꢁ
(b) Basaric N, Cindro N, Bobinac D, et al. Photochem Photobiol Sci. 2012;11:381.
15. Veljkovic J, Uzelac L, Molcanov K, Mlinaric-Majerski K, Wan P, Basaric N. J Org
Chem. 2012;77:4596.
Compounds 3 and QM3 were optimized to energy minimum in
gas phase using Gaussian software37 with B3LYP/6-311G(d,p) basis
set. Atomic charges and molecular orbitals were calculated on
optimized structures.
ꢁ
ꢀ
ꢁ
ꢁ
16. (a) Atherton JCC, Jones S. Tetrahedron. 2003;59:9039;
(b) Reddy MA, Thomas A, Srinivas K, et al. J Mater Chem. 2009;19:6172;
(c) Cakmak O, Erenler R, Tutar A, Celik N. J Org Chem. 2006;71:1795;
(d) Bogdanovic B. J Org Chem. 1988;21:261;
(e) Imoto M, Takeda M, Tamaki A, Taniguchi H, Mizuno K. Res Chem Intermed.
2009;35:957.
Acknowledgements
17. (a) Ferrari L, Hunsberger IM, Gutowsky HS. J Org Chem. 1960;25:485;
(b) Ferrari JL, Hunsberger IM, Gutowsky HS. J Am Chem Soc. 1965;87:1247.
18. Fedorov IA, Zhuravleva YN, Bervenob VP. Phys Chem Chem Phys. 2011;13:5679.
19. Saha K, Lajis NH, Abas F, Naji NA, Hamzah AS, Shaari K. Aust J Chem. 2008;61:
821.
These materials are based on work financed by Croatian Science
Foundation (HRZZ 02.05/25, IP-2014-09-6312).
Appendix A. Supplementary data
20. Rochlin E, Rappoport Z. J Org Chem. 2003;68:216.
21. Hardacre RW, Perkin AG. J Chem Soc. 1929:180.
22. Skalamera Ð, Mlinaric-Majerski K, Martin-Kleiner I, Kralj M, Wan P, Basaric N.
J Org Chem. 2014;79:4390.
ꢀ
ꢁ
ꢁ
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.tet.2017.08.038.
23. (a) Reusch W, Le Mahieu R. J Am Chem Soc. 1964;86:3068;
(b) Konieczny M, Harvey RG. J Org Chem. 1979;44:4813.
24. (a) Hallman JL, Bartach RA. J Org Chem. 1991;56:6243;
(b) Lohier J-F, Wright K, Peggion C, et al. Tetrahedron. 2006;62:6203.
25. Lu Z, Lord SJ, Wang H, Moerner WE, Twieg RJ. J Org Chem. 2006;71:9651.
26. (a) Criswell TR, Klanerman BH. J Org Chem. 1974;39:770;
(b) Coleman RS, Mortensen MA. Tetrahedron Lett. 2003;44:1215.
27. Corbett PT, Sanders JKM, Otto S. Chem Eur J. 2008;14:2153.
28. Biliger C, Demerseman P, Royer R. J Heterocycl Chem. 1985;22:735.
29. Chaiwongwattana S, Skalamera Ð, Doslic N, Bohne C, Basaric N. ESIPT reactivity
of anthrol carbaldehydes, [manuscript in preparation].
30. Newman MS, Zahm HV. J Am Chem Soc. 1943;65:1097.
31. Martin EL. J Am Chem Soc. 1936;58:1438.
32. Zhang Z, Sangaiah R, Gold A, Ball LM. Org Biomol Chem. 2011;9:5431.
33. Burgstahler AW, Mosettig E. J Am Chem Soc. 1959;81:3697.
34. Corey EJ, Cane DE. J Org Chem. 1971;36:3070.
35. Ullmann F, Urmenyi D. Ber Dtsch Chem Ges. 1912;45:2259.
36. Soman A, Qiu Y, Chan Li Q. J Chromatogr Sci. 2008;46:461.
37. Frisch MJ, Trucks GW, Schlegel HB, et al. Gaussian 03, Revision C.02. Wallingford
CT: Gaussian, Inc.; 2004.
References
1. (a) Saleem M, Rafiq M, Hanif M. J Fluoresc. 2017;27:31;
(b) Jiao Y, Zhu B, Chen J, Duan X. Theranostics. 2015;5:173;
(c) Jeong Y, Yoon J. Inorg Chim Acta. 2012;381:2;
(d) Angelov B, Angelova A, Ionov R. J Phys Chem B. 2000;104:9140;
(e) de Silva AP, Gunaratne H, Gunnlaugsson T, Lynch P. New J Chem. 1996;20:
871.
2. Martinez GR, Garcia F, Catalani LH, et al. Tetrahedron. 2006;62:10762.
3. (a) Yang L, Liu Y, Zhou X, et al. Dyes Pigm. 2016;126:232;
(b) Moon H, Jun YW, Kim D, et al. Chem Asian J. 2016;11:2518.
4. (a) Becker HD. Chem Rev. 1993;93:145;
(b) Bouas-Laurent H, Castellan A, Desvergne JP, Lapouyade R. Chem Soc Rev.
2000;29:43.
ꢀ
ꢀ ꢁ
ꢁ
5. (a) Molard Y, Bassani DM, Desvergne J-P, Horton PN, Hursthouse MB,
Tucker JHR. Angew Chem Int Ed. 2005;44:1072;
(b) Miyakawa N, Kiba T, Sato SI. Mol Cryst Liq Cryst. 2010;520:193;
(c) Nakamura A, Inoue Y. J Am Chem Soc. 2003;125:966;
(d) Kawanami Y, Pace TCS, Mizoguchi J-I, et al. J Org Chem. 2009;74:7908;
(e) Tung C-H, Wu L-Z, Zhang L-P, Chen B. Acc Chem Res. 2003;36:39.
6. Dvornikov AS, Rentzepis PM. Res Chem Intermed. 1996;22:115.
7. (a) Dong S, Fu PP, Shirsat RN, Hwang HM, Leszczynski J, Yu H. Chem Res Toxicol.
2002;15:400;
(b) Chowdhury N, Dasgupta S, Singh NDP. Bioorg Med Chem Lett. 2012;22:4668.
8. (a) Uzelac L, Skalamera Ð, Mlinaric-Majerski K, Basaric N, Kralj M. Eur J Med
Chem. 2017;137:558;
38. (a) Ross Kelly T, Parekh ND, Trachtenberg EN. J Org Chem. 1982;47:5009;
(b) Ihmels H. Eur J Org Chem. 1999:1595;
(c) Ihmels H, Meiswinkel A, Mohrschladt CJ, et al. J Org Chem. 2005;70:3929.
39. Mohan R, Katzenellenbogen JA. J Org Chem. 1984;49:1238.
40. Laue T, Plagens A. In: Namen- und Schlagwort-Reaktionen der Organischen
Chemie; Teubner Studienbücher Chemie. 1998:307e309. Sttutgart.
41. Duan S, Turk J, Speigle J, Corbin J, Masnovi J, Baker RJ. J Org Chem. 2000;65:
3005.
42. (a) Ito K, Suzuki T, Sakamoto Y, et al. Angew Chem Int Ed. 2003;115:1191;
(b) Valkenier-van Dijk EH, Dissertation (2011).
43. Hodge P, Power GA, Rabjohns MA. Chem Commun. 1997:73.
ꢀ
ꢁ
ꢁ
(b) Kralj M, Uzelac L, Wang Y-H, et al. Photochem Photobiol Sci. 2015;14:1082.
ꢀ
ꢁ
9. Skalamera Ð, Mlinaric-Majerski K, Martin Kleiner I, et al. J Org Chem. 2017;82:
6006.