A pentacene intermediate via formal intramolecular photoredox of a 6, 13-pentacenequinone in aqueous solution

Article abstract of DOI:10.1139/V07-117

The formal intramolecular photoredox reaction (or tandem phototautomerizations) of aromatic ketones in aqueous solution discovered in our laboratory has been extended to a number of acenequinones. In particular, we were interested in whether the photoredox reaction could be applied to 2-(hydroxymethyl)-6,13-pentacenequinone (4), which would result in 2-formyl-6,13-dihydroxypentacene (10) and hence offer a photochemical method for synthesizing a pentacene derivative. Whereas a number of acenequinones displayed a range of photoredox reactivity, photolysis of 4 in acidic aqueous solution (pH < 3) resulted in a clean intramolecular photoredox reaction, via an enol intermediate, to give 10 (green compound; Φ ～ 0.2 at pH 1), which was too reactive for isolation or trapping by standard ArOH trapping agents such as acetic anhydride. These reactions may be viewed as a one-way photochemical intramolecular "redox switch" from, quinone to hydroquinone with concurrent oxidation of an attached hydroxymethyl (alcohol) moiety. Without the attached alcohol moiety, these acenequinones are photostable in aqueous solution. The trend in observed relative reactivity may be partially rationalized by examining changes in molecular orbital coefficients observed in the calculated HOMOs and LUMOs (at the AM1 level).

Full text of DOI:10.1139/V07-117

1
023
A pentacene intermediate via formal
intramolecular photoredox of a 6,13-
pentacenequinone in aqueous solution
Yunyan Hou and Peter Wan
Abstract: The formal intramolecular photoredox reaction (or tandem phototautomerizations) of aromatic ketones in
aqueous solution discovered in our laboratory has been extended to a number of acenequinones. In particular, we were
interested in whether the photoredox reaction could be applied to 2-(hydroxymethyl)-6,13-pentacenequinone (4), which
would result in 2-formyl-6,13-dihydroxypentacene (10) and hence offer a photochemical method for synthesizing a
pentacene derivative. Whereas a number of acenequinones displayed a range of photoredox reactivity, photolysis of 4 in
acidic aqueous solution (pH < 3) resulted in a clean intramolecular photoredox reaction, via an enol intermediate, to
give 10 (green compound; Φ ~ 0.2 at pH 1), which was too reactive for isolation or trapping by standard ArOH trapping
agents such as acetic anhydride. These reactions may be viewed as a one-way photochemical intramolecular “redox
switch” from quinone to hydroquinone with concurrent oxidation of an attached hydroxymethyl (alcohol) moiety. With-
out the attached alcohol moiety, these acenequinones are photostable in aqueous solution. The trend in observed relative
reactivity may be partially rationalized by examining changes in molecular orbital coefficients observed in the calcu-
lated HOMOs and LUMOs (at the AM1 level).
Key words: pentacene, acenequinones, photoredox, enol, acid catalysis.
Résumé : La réaction photoredox intramoléculaire formelle (ou phototautomérisations en tandem) de cétones aromati-
ques en solution qui a été découverte dans notre laboratoire a été étendue à un certain nombre d’acènequinones. En
particulier, nous étions intéressés à déterminer si la réaction photoredox pouvait être appliquée à la 2-(hydroxymé-
thyl)6,13-pentacènequinone (4) qui conduirait à la formation du 2-formyl-6,13-dihydroxypentacène (10) et qui offrirait
une méthode de synthèse de dérivés pentacènes. Alors qu’un bon nombre d’acènequinones présentent des activités pho-
toredox diverses, la photolyse du composé 4 dans une solution aqueuse acide (pH inférieur à 3) donne lieu à une réac-
tion photoredox intramoléculaire propre, par le biais d’un intermédiaire énolique, et elle conduit à la formation du
produit 10 (un composé vert; Φ approximatif de 0,2, à un pH de 1) qui est trop réactif pour être isolé ou piégé par les
agents habituels de piégeage des ArOH, tel l’anhydride acétique. On peut donc considérer que ces réactions sont des
“
interrupteurs redox” à sens unique pour les réactions photochimiques intramoléculaires et qu’elles permettent le pas-
sage d’une quinone à une hydroquinone avec l’oxydation concurrente de la portion alcoolique, hydroxyméthyle, qui y
est attachée. En l’absence de la portion alcoolique qui y est attachée, ces acènequinones sont photostables en solution
aqueuse. La tendance qui a été observée pour la réactivité relative peut être partiellement rationalisée par un examen
des changements dans les coefficients des orbitales moléculaires qui ont été observées dans les hautes élevées et basses
vacantes, au niveau AM1.
Mots-clés : pentacène, acènequinones, photoredox, énol, catalyse acide.
[
Traduit par la Rédaction] Hou and Wan 1032
Introduction
bility. Recently, photochemical methods based on extrusion
of molecular fragments for the synthesis of pentacenes have
also been reported (3, 4).
Our group reported some time ago that 2-(hydroxy-
methyl)-9,10-anthraquinone (1) undergoes a very efficient
new type of intramolecular photoredox reaction (or equiva-
lently viewed as proceeding via tandem phototautomeriza-
tions) in neutral aqueous solution to give 2-formyl-9,10-
dihydroxyanthracene (2) (Φ ~ 0.8) (5). This reaction was ob-
Pentacene is a promising candidate for application as a
novel organic semiconductor (1). However, it is sensitive to
both oxygen and light and has low solubility. To overcome
these disadvantages, considerable efforts have been directed
at synthesizing new pentacene derivatives (2). Among them,
pentacenequinone derivatives have been employed as syn-
thetic precusors because of their stability and ready accessi-
Received 22 June 2007. Accepted 27 August 2007. Published on the NRC Research Press Web site at canjchem.nrc.ca on
9 October 2007.
1
1
Y. Hou and P. Wan. Department of Chemistry, Box 3065, University of Victoria, Victoria, BC V8W 3V6, Canada.
1Corresponding author (e-mail: pwan@uvic.ca).
Can. J. Chem. 85: 1023–1032 (2007)
doi:10.1139/V07-117
© 2007 NRC Canada

1
024
Can. J. Chem. Vol. 85, 2007
Scheme 1.
O
O
CH OH
2
CH OH
2
CHBr2
O
O
5
O
O
4
(20%)
CHBr2
NaI, DMF, 80 °C
CH OH
2
CH OH
2
O
6
O
7
(45%)
O
O
O
CH3
CH
3
O
9
8
served only in protic solvents, with highest quantum yields
in water. The mechanism is unimolecular in substrate but is
catalyzed (mediated) by water. In the presence of oxygen,
dihydroxyanthracene 2 is quantitatively converted to 2-
formylanthraquinone (3) (eq. [1]) (5). We have also discov-
ered that simple benzophenones can react via this pathway;
thus, 3-(hydroxymethyl)benzophenone undergoes the same
Acenequinones 4 and 7 were readily prepared via the con-
densation (in situ Diels–Alder) of the corresponding
acenequinones 5 and 6, respectively, with α,α,α′,α′-
tetrabromo-o-xylene (Scheme 1). Acenequinones 5 and 6 are
known compounds and were prepared by following or adapt-
ing the corresponding literature methods (7, 8). 2-Methyl-
6,13-pentacenequinone (8) was also studied to test the re-
quirement of a hydroxymethyl (alcohol) moiety in the antici-
pated photoredox chemistry. It was made using the same
reaction shown in Scheme 1, starting with 6-methyl-1,4-
anthraquinone (9).
photoredox reaction (via T ) but requires pH < 3 (6). These
1
reactions are termed photoredox because of the fact that a
formal transfer of electrons through the meta positions of the
benzene ring is required. Based on these findings, we antici-
pated that 2-(hydroxymethyl)-6,13-pentacenequinone (4)
might undergo the photoredox reaction as well, which would
give rise to a corresponding dihydroxypentacene derivative.
Herein, we report a new photochemical method involving an
intramolecular redox process to prepare a pentacene inter-
mediate from 4 and associated photochemistry for related
acenequinones.
UV–vis studies
As shown for the reaction of 1 (5), UV–vis studies were
particularly informative of reaction because the photochemi-
cal transformation (eq. [1]) involves a substantial change in
chromophore, going from an aromatic ketone/quinone to a
dihydroxy-substituted polycyclic aromatic compound
(
anthracene). In the case of 1, the initial photochemical
Results and discussion
product 2 is coloured (orange). We anticipated similar UV–
vis changes on the photoredox reaction of 4–7. UV–vis stud-
ies of 4 (10 to 10 5 mol/L, solvent: 1:3 H O/CH CN, pH 1
–
6
–
Materials
2 3
The potential photoredox chemistry of pentacenequinone
and associated acenequinones 5–7 (chosen based on their
and 7 (water portion), argon purged, 350 nm lamps, 3.0 mL
quartz cuvette) were carried out initially. No changes were
observed in neat CH CN or in H O/CH CN (water at pH 7),
4
relative ease of synthesis) was explored in this work.
3
2
3
O
OH
O
CH OH
2
CHO
CHO
hν
[
1]
O2
[
1]
1
:1 H O/CH CN
2 3
O
OH
O
1
2
3
©
2007 NRC Canada

Hou and Wan
1025
Fig. 1. UV–vis traces observed on photolysis of pentacene-
quinone 4 (1:3 H O/CH CN, pH 1, λ 350 nm, argon purged).
Photolysis times were 10–20 s except for the final trace (2 min).
New absorption bands at 318, 473, and 657 nm are consistent
with the formation of dihydroxypentacene 10 (Inset: expansion
of the long wavelength region).
Fig. 2. Decay of photogenerated dihydroxypentacene 10 (from
pentacenequinone 4 in 1:3 H O/CH CN, pH 1) on exposure to
air upon opening cuvette ( absorbance monitored at 473 nm).
2
3
ex
2
3
0.15
0.10
0.05
0
1.5
1.0
0.5
0
×
10
0
1
2
3
4
5
Time (min)
200
400
600
800
Wavelength (nm)
oxidized to give 11. Notably, 8, although structurally very
similar to 4, did not show reaction under the stated condi-
tions.
but dramatic changes were observed in H O/CH CN at pH 1
The effect of pH on the relative quantum efficiency of the
photoredox reaction of 4 was examined next. As noted for
pH 7, no reaction was observed. Using the UV–vis method,
the relative quantum efficiency for the formation of 10 was
followed at different pH values (1:3 H O/CH CN, the pH of
2
3
(
Fig. 1). Photolysis at this pH resulted in the formation of an
intense new absorption band at 318 nm and two additional
less intense) visible bands at 473 and 657 nm, the latter be-
ing a very broad band covering 550–780 nm. Based on what
is known for 1 shown in eq. [1], the same reaction for 4
would give rise to 6,13-dihydroxypentacene 10 (eq. [2]).
Pentacenes are known to absorb strongly in the visible re-
gion, with the longest wavelength band in the 500–700 nm,
depending on substituents (2a, 4). The broad absorption at
(
2
3
the water portion varied). The results show that the
photoredox efficiency strongly depends on the pH of the so-
lution (Fig. 3). The photoredox reaction was not observable
above pH 3 and reaches maximum efficiency at about pH ≅
0, although higher acidities were not employed in this study.
Whereas UV–vis studies of 5 (pH 1 and 7) gave only
small incremental changes even on prolonged photolysis, in-
dicative of a much less reactive chromophore, the same stud-
ies of 6 (pH 7) gave rise to two new strongly absorbing
bands at 268 and 392 nm, which were assignable to
dihydroxynaphthalene 12 (Fig. 4 and Scheme 2). For this
compound, acid was not required for the efficient
photoredox reaction, akin to what was observed for 4. In-
6
57 nm is consistent with the formation of a pentacene.
These new absorption bands disappeared within 1 min of
opening the cuvette to air or oxygen (Fig. 2) and within 1 h
when the cuvette was unopened and sealed. The resulting
UV–vis spectrum was identical to an authentic sample of
pentacenequinone aldehyde 11, which has a UV–vis spec-
trum similar to that of 4. This is consistent with a highly re-
active dihydroxypentacene derivative that is readily air-
O
OH
CH OH
2
CHO
hν
H O/CH CN
2
3
pH < 3)
(
O
OH
10
4
O
O
[
2]
CHO
O2
11
©
2007 NRC Canada

1
026
Can. J. Chem. Vol. 85, 2007
Fig. 3. Effect of pH on the efficiency of photoredox reaction of
pentacenequinone 4 monitored at 318 nm (formation of
Fig. 4. UV–vis traces observed on photolysis of naphthoquinone
6 (1:1 H O/CH CN, pH 7, λ 350 nm, argon purged). Photolysis
2
3
ex
dihydroxypentacene 10; λex 350 nm; 1:3 H O/CH CN (pH 1), ar-
times were 2–20 s. Growth of absorption bands at 268 and
2
3
gon purged). ∆A is the difference in the absorbance before and
392 nm is consistent with the formation of dihydroxynaphthalene
12.
after irradiation.
1.5
1.0
0.5
0
2.5
2.0
1.5
1.0
0.5
0
2
00
300
400
500
0
1
2
3
4
5
6
7
Wavelength (nm)
pH
Scheme 2.
O
Ha
CHO
O2
O
O
Ha
OH Ha
CH OH
2
C
HO
hν
O
13
H O/CH CN
2
3
(
pH 7)
OAc
OH
Ac O/NaOH
2
CHO
6
12
OAc
14
deed, the use of acid did not result in any enhancement of
in some cases trap initial photoproducts. Acenequinones 4–8
–
5
–
the photoredox reaction. Photolysis in neat CH CN gave no
were photolyzed (10 to 10 3 mol/L, λ 350 nm, argon
3
ex
observable changes. The photoredox product in this case
purged) in H O/CH CN (various pH values) as well as in
2
3
(
dihydroxynaphthalene 12) was more stable in air or oxygen,
as the coloured intermediate remained unchanged for at least
day in the sealed cuvette. Shown in Fig. 5 are solutions of
neat CH CN. Semi-preparative photolysis of 4 (pale yellow)
3
in 1:3 H O/CH CN (pH 1) gave a green to dark green solu-
2
3
1
tion depending on irradiation time, indicative of photoredox
chemistry. When left standing in the photolysis tube after ir-
radiation with no precautions to prevent entry of air, the col-
our changed to a pale yellow within 1 h, indicative of a
reactive initial photoproduct. When the photolyzed solution
was exposed to air after photolysis (pouring the solution into
a flask exposed to air), the green colour disappeared imme-
diately. In view of the reactive nature of the initially formed
photoproduct, no attempts were made to isolate it in these
runs. Instead, the solutions were exposed to air after
photolysis and extracted with CH Cl . Work-up using this
photolyzed acenequinones 1, 4, and 6 illustrating the varied
colours observable in this family of photoredox reactions in
aqueous solution. Finally, UV–vis studies of 7 gave no ob-
servable changes in the UV–vis spectrum (pH 1 or 7) with
prolonged times, indicative of an essentially unreactive
acenequinone.
Product studies
Semi-preparative photolyses were carried out to confirm
the reactivity of the compounds (based on the previously
mentioned UV–vis studies) and to isolate, characterize, and
2
2
method gave exclusively 2-formyl-6,13-pentacenequinone
©
2007 NRC Canada

Hou and Wan
1027
Fig. 5. Dramatic differences in observed colours in the
scheme shown in Scheme 2, trapping of 12 was attempted.
Thus, after photolysis under an argon purge, sufficient solid
NaOH was added to the photolysate to basicify the solution.
This resulted in the solution turning from yellow to red, con-
sistent with formation of a naphtholate chromophore (via
photolysis of 6 (pH 7), 1 (pH 7), and 4 (pH 1) (L to R) in 1:1
H O/CH CN (λ 350 nm, 15 min, argon purged). In sealed ves-
2
3
ex
sels, the colour remained unchanged for at least a day for 6 and
and a few hours for 4.
1
deprotonation of 12). Excess Ac O was then added under ar-
2
gon, which quenched the red colour. Upon work-up, the
only product isolable was diacetoxyanthracene aldehyde 14
thus confirming the reaction sequence shown in Scheme 2.
As further evidence for the formation of 12 as the primary
photochemical product from 6, photolyses were also carried
–
3
out in NMR tubes (10 mol/L, 1:9 D O/CD CN, pD 7, λ
2
3
ex
3
50 nm, argon purged, 10 min). Photolysis gave a dark yel-
low solution with a reduction of the signals at δ 4.80 (s) and
.07 (s) assigned to the methylene and aromatic H protons
8
a
of 6, respectively, (Scheme 2), with concomitant growth of
the aldehyde (CHO; δ 10.07) and H (δ 8.70) protons assign-
a
able to the initial photoredox product 12. Based on NMR in-
tegration, the yield of 12 was estimated to be about 14%.
Upon introduction of oxygen via a syringe needle, the dark
yellow colour was bleached. In the NMR, the peaks assign-
able to 12 disappeared. In particular, a new aldehyde peak at
(
11) (yields 20%–70% by NMR, depending on photolysis
δ 10.13 and an aromatic singlet (H ) at δ 8.55 are assignable
time). No reaction was observed when photoyzed in pH 7 or
in neat CH CN, consistent with the UV–vis studies reported
earlier. In addition, no reactions were observed without
photolysis (in the presence or absence of oxygen).
a
to oxidized product 13.
3
Using the photoredox reaction reported for 1 as a second-
ary actinometer (Φ = 0.8, pH 2), we estimated a quantum
yield of photoredox for 4 (1:3 H O/CH CN, pH 1) and 6
Attempts were made to characterize the initially formed
green) dihydroxypentacene intermediate 10. Initial experi-
2
3
(
1:1 H O/CH CN, pH 7) to be about 0.2 and 0.4, respec-
(
2 3
tively. Although these compounds are not as reactive as the
parent anthraquinone system 1, they nevertheless have sub-
stantial quantum yields of reaction at a suitable pH.
ments were carried out by photolyzing samples of 4 in an
NMR tube so the formation of 10 could be observed directly
by NMR without work-up or exposure to air. However, be-
cause of the very low solubility of 4 in aq. CH CN, these
Although acenequinone 5 also underwent the photoredox
reaction, it exhibited low reactivity in neutral or acidic solu-
tion. This compound required photolysis times that were
about an order of magnitude longer for similar conversion.
Because of its much lower reactivity, no attempts were made
to characterize the initially formed photoredox product and
only the final air-oxidized product 15 was isolated (eq. [3]).
No reaction was observed without irradiation (in the pres-
ence or absence of oxygen). Of all the acenequinones stud-
ied in this work, compound 7 proved to be the only one that
was photochemically inert. Thus, photolysis of this com-
pound (pH 1, 7) gave no reaction and the compound could
be recovered unchanged. Initially we were puzzled by this
3
photolyses proved to be impractical. Attempts were also
made to trap 10 using acetic anhydride (as the ester) and
methyl iodide (as the methyl ether) in situ, but in all experi-
ments attempted, only 11 was isolated, consistent with the
highly reactive nature of the initially formed product 10.
Photolysis of the closely related methylpentacenequinone
8
under similar conditions as for 4 or even on extended
photolysis did not give any reaction. The photolyzed solu-
tion did not develop any colour and upon work-up, the start-
ing material was recovered unchanged. Thus, the photoredox
reaction requires an α-hydroxy moiety at the carbon directly
attached to the aromatic ring for the reaction to proceed. As
the chromophores (and structures) of 4 and 8 are otherwise
identical, it would appear that the mechanism allows any
formed intermediates — up to a critical product forming
step — to return (deactivate) efficiently to starting material.
Photolysis of 6 (1:1 H O/CH CN, pH 7) resulted in a yel-
result although one could imagine
a
number of
photophysical reasons to explain the lack of reaction. How-
ever, the lack of reactivity of this compound as well as the
reactivity of the other systems can be partly rationalized by
resorting to an examination of HOMO and LUMO charac-
teristics (vide infra).
2
3
low solution that slowly (several days) faded when left in the
vessel and turned to pale yellow within a few hours when
exposed to air. On work-up in air, naphthoquinone aldehyde
Reaction mechanisms
1
3 was formed exclusively (30%–60%, depending on
We believe that the observed formal intramolecular
photoredox reactions of the present acenequinones, includ-
photolysis time). To provide further evidence for the reaction
O
O
CH OH
2
CHO
hν
O2
[
3]
H O/CH CN
2
3
(
pH 7)
O
5
O
15
©
2007 NRC Canada

1
028
Can. J. Chem. Vol. 85, 2007
ing the “parent” anthraquinone (5) and “simple” benzophe-
none (6) systems, are closely related to the acid-catalyzed
photohydration of benzophenone (16) reported by Wirz and
co-workers (9) (eq. [4]). They discovered that the
protonation (at the carbonyl) of triplet benzophenone 16 re-
sults in an overall photohydration reaction, to give water-
adducts 18 and 19 with a preference for 18, i.e., hydration at
the meta position (eq. [4]). This reaction was unexpected
and has missed detection in the past because of the instabil-
ity of the hydration products (18 and 19), which readily
types of acid-catalyzed chemistry, a prime example of which
was the photoredox reaction of benzophenone 20 to give 21
in acidic aqueous solution reported by us (6) (eq. [6]).]
In view of this background and the fact that
acenequinones also tend to react via T with photochemical
1
behaviour (12) that is closely allied to what is known for
benzophenones, we propose that the mechanism of
photoredox reaction of the acenequinones reported in this
work also start with a T reactive state that is sufficiently ba-
1
sic to be protonated at pH 7 or at pH < 3, depending on the
substrate. Intramolecular photoredox reaction requires
protonation of the ketone (by acid or by water) and
revert back to 16. Wirz and co-workers (9) assigned a pK of
a
–
0.4 for the protonated triplet excited state. The mechanism
(
9) of photohydration presumably involves initiation
deprotonation of the C-H proton at the CH OH moiety. This
2
protonation of the carbonyl oxygen of triplet excited benzo-
phenone 16, to generate an excited triplet state conjugate
acid 17 that has its positive charge significantly delocalized
to the ortho and meta positions of the benzene ring (e.g., as
shown in eq. [5]), as would be anticipated based on the
Zimmerman “ortho–meta” effect for benzene ring site acti-
vation in photochemical reactions (10, 11). Attack by water
at these sites of positive charge would lead to the observed
hydration products, which upon deactivation to their ground
states would not be expected to be long-lived and readily re-
turn to benzophenone (16) by dehydration (with re-
aromatization). Although the above acid-catalyzed
photohydration of benzophenone reported by Wirz and co-
workers (9) gave no isolable stable product, it seemed rea-
sonable to assume that the protonation of triplet excited ben-
zophenone (16) and aromatic ketones and related
compounds, such as acenequinones, might lead to other
is a unique mechanism in photochemistry in that two very
different types of excited state proton transfers are required
(simultaneously or sequentially). A working mechanism is
shown for pentacenequinone 4 (Scheme 3). Although not
shown explicitly, all excited states shown in this mechanism
are triplets. Compound 4 requires acid for the photoredox
pathway to be observed. Thus, excited state protonation
gives rise to an excited triplet state conjugate acid 22a,b that
is sufficiently reactive for the subsequent deprotonation of
the C-H proton in the CH OH moiety, this being the product
2
forming step. That is, both the triplet excited state and the
protonated triplet 22a,b can return to starting material with-
out going on to product. Deprotonation of the C-H proton in
CH OH seems reasonable if the positive charge is strongly
2
delocalized to the ring position with the CH OH group, acid-
2
ifying the C-H proton. Such a step would give the double
enol 23, which is assumed to be initially generated in its
3
[
4]
O
OH
*
+
H
hν
S1
T
1
+
H O
3
16
17
OH
OH
H
OH
OH
H
H O
2
+
1
8 (Major)
19 (Minor)
(ortho-adduct)
(
meta-adduct)
3
[
[
5]
6]
OH
3
O
H
*
*
17
O
HO
H
CH OH
2
CHO
hν
H O/CH CN (pH < 3)
2
3
(
Φ = 0.6, pH 2)
20
21
©
2007 NRC Canada

Hou and Wan
1029
Scheme 3.
O
O
*
*
CH OH
2
CH OH
2
+
H
O
O
OH
O
2
2a
4
H O
2
H
*
CHOH
CHOH
C-H deprotonation
OH
23 ("double enol")
OH
22b
ketonization/
enolization
(Note: all excited states are believed to be
triplets. In addition, double enol 23 may
be initially generated as a triplet before
decaying to a singlet)
OH
CHO
OH
10
triplet excited state but subsequently deactivates to its
ground state singlet. This is followed by ketonization–
enolization (tandem tautomerization), which would give the
formal redox product 10. It is also possible that water could
attack the positive charge on the ring (of 22b) resulting in a
photohydration reaction like that reported for benzophenone
in acid (9). The extent of this pathway is unknown for these
compounds.
HOMO to the LUMO without regard to spin multiplicity.
For all the reactive acenequinones (1, 4, 5, and 6), examina-
tion of HOMO and LUMO coefficients (some of which are
shown in Fig. 6) show that there is significant to substantial
migration of charge from at least the ring with CH OH
2
substituent to the carbonyl and to the other ring(s) in the
compound. This charge transfer (migration) would make the
carbonyl oxygen more basic. On trapping of this charge
transfer state with a proton (protonation of the carbonyl oxy-
gen), one can assume that some of the residual positive
charge (sites of electron deficiency) will be localized on the
The overall reaction requires the formal movement of π
electrons from one end (CH OH) of the molecule to the
2
other (C=O) (Scheme 3). That this can occur for such a large
π system as a naphthalene ring in pentacenequinone 4 was
unexpected. Since 4 absorbs into the visible (up to 450 nm),
the photochemical transformation is a way of storing solar
energy in the form of the high energy species dihydroxy-
pentacene 10, which is a compound capable of reducing
other molecules such as molecular oxygen. Conceivably,
other uses may be possible.
ring with the CH OH substituent, making the CH proton on
2
the CH OH more acidic. At this level of calculation, there is
2
also evidence to suggest that the C–H bond (of the CH OH)
2
moiety also donates electron density for compounds 1, 4,
and 6, but less so for 4. This would have the added effect of
making this proton even more acidic. Indeed, 1 and 6 were
the most reactive compounds (at pH 7). For pentacene-
quinone 4, the HOMO and LUMO characteristics are not as
optimal; for this compound, reaction was only observable
below pH 3. The presence of the OH group is also particu-
larly crucial, because it provides a further inductive acidify-
ing effect and also makes the whole group readily oxidizable
(to the aldehyde). For the unreactive compound 7 (Fig. 6),
the same calculations show that the charge transfer is in the
opposite direction, that is, from the rings without the
Although Wirz and co-workers (9) did not explicitly offer
an explanation as to why the carbonyl oxygen of benzophe-
none (16) in T becomes much more basic, one can envisage
1
that in aqueous solution the T state may have substantial
1
charge transfer character (as opposed to its known diradical
character in typical organic solvents). Such a highly polar-
ized excited state enables the photohydration (9) and the
intramolecular photoredox reactions to proceed. Moreover,
water is also an essential medium to mediate proton transfer
reactions that are required in the reaction mechanism. We
have carried out calculations of HOMOs and LUMOs
CH OH group (the “naphthalene” ring) to the carbonyl and
2
to the ring containing the CH OH group. Although this
2
charge transfer will still make the carbonyl oxygen more ba-
sic in the excited state and hence undergo protonation, one
(
Chem 3D, AM1) for all of the compounds studied in this
work in an attempt to predict the migration of charge on ex-
citation. We assume that a simple picture of the excited state
can be obtained by the promotion of an electron from the
would not anticipate the same reactivity of the CH OH
group, since the ring in which it is attached is now more
electron-rich than in the ground state. Indeed, this compound
2
©
2007 NRC Canada

1
030
Can. J. Chem. Vol. 85, 2007
Fig. 6. Calculated (Chem 3D, AM1) HOMOs (left) and LUMOs (right) for acenequinones (top to bottom) 1 (reactive), 4 (reactive), 6
reactive), and 7 (unreactive).
(
proved to be photochemically inert! Thus, simple semi-
empirical MO calculations at the AM1 level are useful for
the prediction of qualitative reactivity patterns for these
compounds.
6-(Hydroxymethyl)-1,4-anthraquinone (5)
This compound was made according to the published pro-
cedure (7) with 45% overall yield (mp 180–182 °C, lit. value
–
1
1
(7) 178 to 179 °C). IR (KBr, cm ) 3498, 1670, 1650. H
NMR (300 MHz, CDCl ) δ: 8.59 (s, 2H), 8.05 (d, 1H, J =
3
8
4
.8 Hz), 8.03 (s, 1H), 7.68 (d, 1H, J = 8.8 Hz), 7.05 (s, 2H),
1
3
.92 (s, 2H). C NMR (DMSO- d , 75 MHz) δ: 184.4,
6
Experimental section
184.3, 144.6, 134.0, 139.9, 134.3, 133.4, 130.0, 128.8,
1
1
+
28.2, 127.9, 127.7 (2C), 126.3, 62.7. EIMS m/z 238 (M ,
00), 221 (10).
Instrumentation
NMR spectra were recorded on a Bruker 300 or 500 MHz
1
13
2-(Hydroxymethyl)-6,13-pentacenequinone (4)
for H and 75 or 125 MHz for C. IR spectra were recorded
on a PerkinElmer 283 instrument. UV–vis spectra were mea-
sured on a Cary 1 instrument. Mass spectra were obtained
on a Kratos Concept H instrument.
A mixture of 5 (0.43 g, 1.8 mmol), α,α,α′,α′-tetrabromo-o-
xylene (1.14 g, 2.7 mmol), and NaI (2.70 g, 18 mmol) were
heated in 5 mL DMF for 3 h at 80 °C to give a dark brown
mixture. After cooling, 50 mL of cold water was added fol-
lowed by the addition of 10 mL of 10% NaHSO dropwise
3
Materials
until the dark brown colour was bleached. The mixture was
extracted with 3 × 25 mL CH Cl and dried over anhydr.
All solvents were best reagent grade, except CH CN that
3
2
2
was HPLC grade. Deuterated solvents CDCl , (CD ) CO,
MgSO . Removal of the solvent gave a brown solid, which
upon separation by column chromatography (silica gel;
3
3 2
4
DMSO-d were purchased from the Cambridge Isotope Lab-
6
oratory (99.9% D). All synthetic precursors used below were
purchased from Sigma-Aldrich and used as received.
CH Cl and ethyl acetate as eluents) gave 0.13 g of 4 (21%,
yellow powder), mp (decomp) 286 °C. IR (KBr, cm ) 3403,
2
2
–
1
©
2007 NRC Canada

Hou and Wan
1031
1
1
8
8
7
2
1
1
1
3
3
691, 1676. H NMR (DMSO-d , 500 MHz) δ: 8.95 (s, 1H),
prior to photolysis and kept under a stream of argon during
irradiation. Unless otherwise noted, all the photolyses were
worked up in the presence of air. Photolyzed solutions were
extracted with CH Cl . Upon solvent evaporation, the mate-
6
.94 (s, 1H), 8.93 (s, 1H), 8.89 (s, 1H), 8.38–8.33 (m, 2H),
.31 (d, 1H, J = 8.5 Hz), 8.23 (s, 1H), 7.84–7.79 (m, 2H),
.75 (d, 1H, J = 8.5 Hz), 5.54 (t, OH, J = 5.6 Hz), 4.75 (d,
H, J = 5.6 Hz). C NMR (DMSO-d , 125 MHz) δ: 182.2,
82.1, 144.7, 134.8 (2C), 134.74, 134.72, 133.9, 130.4,
30.3, 130.1 (2C), 129.9 (2C), 129.8 (2C), 129.1 (2C),
2
2
1
3
rial was analyzed initially by NMR for product identification
and determination of conversions and yields. Subsequently,
separation and purification of products were achieved using
prep. TLC (silica gel, CH Cl ).
6
+
29.0, 128.9, 128.8, 126.2, 62.7. EI MS m/z 338 (M , 100),
21 (5). HRMS calcd. for C H O3 338.0943; found
2
2
2
3
14
38.0928.
Photolysis of 2-(hydroxymethyl)-6,13-pentacenequinone
4)
Compound 4 (3 mg in 75 mL of CH CN and 25 mL of
(
2
-Methyl-6,13-pentacenequinone (8)
Following the method described for the synthesis of 4, use
3
H O, pH 1) was irradiated (16 lamps) for 1 min to give 18%
2
of 6-methyl-1,4-anthraquinone (9, from ref. 7) (0.12 g,
yield of 11 (2-formyl-6,13-pentacenequinone), which was
–
1
0
0
.54 mmol), α,α,α′,α′-tetrabromo-o-xylene (0.32 g,
subsequently isolated in pure form. IR (KBr, cm ) 1698,
1
.81 mmol), NaI (0.81 g, 5.4 mmol) gave 60 mg (35%, yel-
1675. H NMR (300 MHz, CDCl ) δ: 10.2 (s, 1H), 9.10 (s,
3
–
1
low powder), mp (decomp) 307 °C. IR (KBr, cm ) 1676,
1H), 8.97 (s, 1H), 8.95 (s, 2H), 8.93 (d, 1H, J = 8.6 Hz),
8.58 (s, 1H), 8.23 (d, 1H, J = 8.6 Hz), 8.19–8.05 (m, 2H),
7.78–7.64 (m, 2H). EIMS m/z 336 (M , 100), 307 (10).
1
1
1
8
8
1
1
1
1
3
615. H NMR (CDCl , 500 MHz) δ: 8.92 (s, 2H), 8.89 (s,
3
+
H), 8.83 (s, 1H), 8.13–8.09 (m, 2H), 8.01 (d, 1H, J =
.4 Hz), 7.87 (s, 1H), 7.71–7.67 (m, 2H), 7.53 (d, 1H, J =
HRMS calcd. for C H O3 336.0786; found 336.0787.
2
3
12
.4 Hz), 2.58 (s, 3H). 13C (CDCl , 125 MHz) δ: 183.4,
3
83.3, 140.2, 135.8 (2C), 135.52, 135.49, 133.8, 132.1,
30.98, 130.94, 130.3 (2C), 130.2, 130.1, 129.96, 129.94,
Photolysis of 2-methylpentacenequinone (8)
Photolysis of 8 as for 4 gave no reaction. The NMR of the
isolated material was identical to that of the starting mate-
rial.
+
29.8, 129.6 (2C), 129.3 (2C), 22.2. EIMS m/z 322 (M ,
00). HRMS calcd. for C H O2 322.0994; found
2
3
14
22.0993.
Photolysis of 6-(hydroxymethyl)-1,4-naphthoquinone (6)
6
-(Hydroxymethyl)-1,4-naphthoquinone (6)
This compound was synthesized according to the pub-
Photolysis of 6 (5 mg in 50 mL of CH CN and 50 mL of
3
H O, pH 7; 4 lamps, 1 min) gave 13 (6-formyl-1,4-
2
lished procedure (8), except for the last hydrolysis step in-
volving 6-(bromomethyl)-1,4-naphthoquinone, in which 1.5
equiv. of silver trifluoacetate was used instead of 5 equiv. of
CaCO . Upon column chromatography (silica gel; CH Cl ),
naphthoquinone, 44% yield by NMR), which was subse-
quently isolated. IR (KBr, cm ) 1697, 1664, 1597. H NMR
–
1
1
(300 MHz, CDCl ) δ: 10.17 (s, 1H), 8.57 (s, 1H), 8.25 (s,
3
1
3
2H), 7.07 (s, 2H). C NMR (CD Cl, 75 MHz) δ: 190.8,
3
2
2
3
0
.2 g (70%, yellow powder) was obtained, mp 80–82 °C, lit.
184.4, 184.1, 139.9, 139.2, 139.1, 135.4, 133.4, 132.8,
–
1
1
+
value (8) 81 to 82 °C. IR (KBr, cm ) 3447, 1662, 1598. H
128.7, 127.6. EIMS m/z 186 (M , 100), 157 (40). HRMS
NMR (300 MHz, CDCl ) δ: 8.07 (d, 1H, J = 8.1 Hz), 8.05
calcd. for C H O3 186.0317; found 186.0322.
3
11
6
(
s, 1H), 7.76 (d, 1H, J = 8.1 Hz), 6.96 (s, 2H), 4.85 (s, 2H).
Photolysis of 6 was also carried out in which the initially
1
3
C (75 MHz, CDCl ) δ: 185.3, 185.1, 147.7, 139.0, 138.8,
formed photoredox product 12 was trapped with Ac O to
3
2
+
1
1
32.2, 131.9, 131.3, 127.1, 124.4, 64.5. EIMS m/z 188 (M ,
give 14 as follows (all carried out under argon): A solution
of 6 (10 mg in 50 mL of CH CN and 50 mL of H O, pH 7)
00), 160 (90).
3
2
irradiated for 5 min. Upon photolysis, sufficient solid NaOH
was added to turn the yellow brown solution to red. This
2
-(Hydroxymethyl)-5,12-naphthacenequinone (7)
A mixture 6 (0.22 g, 1.2 mmol), NaI (1.76 g, 11.7 mmol),
was followed by the addition of 1 mL of Ac O, which turned
2
and α,α,α′,α′-tetrabromo-o-xylene (0.74 g, 1.76 mmol) in
the solution to a brown colour. After work-up in air, the
crude material was characterized with H NMR, which was
1
4
mL DMF was heated to 80 °C overnight. After work-up
and purification (as described for 4), 0.29 g (85%) of a yel-
consistent with 50% conversion to 14 (6-formyl-1,4-
diacetoxynaphthalene). Further separation and purification
was achieved by prep. TLC (silica gel, CH Cl ). IR
low powder was obtained, mp (decomp) 190 °C. IR
–
1
1
(
KBr, cm ) 3327, 1675, 1603. H NMR (300 MHz, CDCl )
3
2
2
–
1
1
δ: 8.84 (s, 2H), 8.40 (d, 1H, J = 8.1 Hz), 8.38 (s, 1H), 8.13–
(KBr, cm ) 1754, 1699. H NMR (300 MHz, CDCl ) δ:
3
8
4
1
1
.05 (m, 2H), 7.83 (d, 1H, J = 8.1 Hz), 7.73–7.64 (m, 2H),
10.15 (s, 1H), 8.38 (s, 1H), 7.99 (s, 2H), 7.39 (dd, 2H, J =
1
3
13
.91 (s, 2H). C NMR (DMSO-d , 75 MHz) δ: 182.2,
8.1 Hz, J = 8.1 Hz,), 2.50 (s, 3H), 2.47 (s, 3H). C NMR
6
81.9, 149.8, 134.5 (2C), 133.7, 132.5, 131.9, 130.0 (2C),
(CDCl , 75 MHz) δ: 191.9, 169.4 (2C), 145.5, 144.4, 134.8,
3
29.6 (2C), 129.3, 128.9, 128.8 (2C), 127.0, 124.1, 62.2.
130.5, 127.9, 127.4, 124.3, 123.2, 121.4, 119.3, 21.4, 21.2.
+
+
EIMS m/z 288 (M , 100), 259 (50). HRMS calcd. for
EIMS m/z 272 (M , 5), 230 (15), 188 (100), 43 (40). HRMS
C H O3 288.0786; found 288.0792.
calcd. for C H O5 272.0685; found 272.0687.
1
9
12
15 12
General photolysis procedure
Photolysis of 6-(hydroxymethyl)-1,4-anthraquinone (5)
Preparative photolyses were carried out using a Rayonet
RPR 100 photochemical reactor using 350 nm lamps, a wa-
ter-cooled cold finger, and 100 mL quartz tubes. Solutions
A solution of 5 (3 mg in 50 mL of CH CN and 50 mL of
3
H O, pH 7, 16 lamps) was irradiated for 1 min, which gave
2
an aldehyde product in less than 2% conversion. Upon
photolysis for 30 min, 30% conversion to 6-formyl-1,4-
–
4
(
10 mol/L in substrate) were purged with argon for 15 min
©
2007 NRC Canada

1
032
Can. J. Chem. Vol. 85, 2007
–
1
anthraquinone (15) could be achieved. IR (KBr, cm ) 1706,
2. (a) J. Jiang, B.R. Kaafarani, and D.C. Neckers. J. Org. Chem.
71, 2155 (2006); (b) C.R. Swartz, S.R. Parkin, J.E. Bullock,
J.E. Anthony, A.C. Mayer, and G.G. Malliaras. Org. Lett. 7,
3163 (2005); (c) E.J. Hwang, Y.E. Kim, C.J. Lee, and J.W.
Park. Thin Solid Films, 499, 185 (2006); (d) K. Kobayashi, R.
Shimaoka, M. Kawahata, M. Yamanaka, and K. Yamaguchi.
Org. Lett. 8, 2385 (2006); (e) S.H. Chan, H.K. Lee, Y.M.
Wang, N.Y. Fu, X.M. Chen, Z.W. Cai, and H.N.C. Wong.
Chem. Commun. (Cambridge, U.K.), 66 (2005); (f) J.E. An-
thony. Chem. Rev. 106, 5028 (2006).
1
1
8
8
1
1
2
2
668, 1589. H NMR (CDCl , 300 MHz) δ: 10.22 (s, 1H),
3
.78 (s, 1H), 8.67 (s, 1H), 8.54 (s, 1H), 8.16 (dd, 2H, J =
.1 Hz), 7.11 (s, 2H). 13C NMR (CDCl , 75 MHz) δ: 191.5,
3
84.5, 184.3, 140.3 (2C), 137.9, 136.7, 135.7, 134.5, 131.5,
+
30.6, 130.3, 129.5, 128.8, 126.5. EIMS m/z 236 (M , 100),
07 (20). HRMS calcd. for C H O3 236.0473; found
1
5
8
36.0466.
Comparison of colours observed in the photolysis of 1,
3
. K.P. Weidkamp, A. Afzali, R.M. Tromp, and R.J. Hamers. J.
4
, and 6
Am. Chem. Soc. 126, 12740 (2004).
Solutions of 1 (5 mg, 1:1 H O/CH CN, pH 7), 4 (5 mg,
2
3
4
. H. Yamada, Y. Yamashita, M. Kikuchi, H. Watanabe, T.
Okujima, H. Uno, T. Ogawa, K. Ohara, and N. Ono. Chem.
Eur. J. 11, 6212 (2005).
1
:3 H O/CH CN, pH 1), and 6 (5 mg, 1:1 H O/CH CN,
2 3 2 3
pH 7) were prepared in 100 mL Pyrex volumetric flasks. Af-
ter purging with a stream of argon, the flasks were irradiated
at 350 nm (16 lamps) for 15 min to develop the colours ob-
served in Fig. 5.
5
. M. Lukeman, M. Xu, and P. Wan. Chem. Commun. (Cam-
bridge), 136 (2002).
6. D. Mitchell, M. Lukeman, D. Lehnherr, and P. Wan. Org. Lett.
, 3387 (2005).
7
7
. D.H. Hua, K. Lou, J. Havens, E.M. Perchellet, Y. Wang, J.P.
Perchellet, and T. Iwamoto. Tetrahedron, 60, 10155 (2004).
Acknowledgments
We acknowledge financial support from the Natural Sci-
ences and Engineering Research Council (NSERC) of Can-
ada and the University of Victoria.
8. E. Torres, C.A. Panetta, N.E. Heimer, B.J. Clark, and C.L.
Hussey. J. Org. Chem. 56, 3737 (1991).
9. M. Ramseier, P. Senn, and J. Wirz. J. Phys. Chem. A, 107,
3
305 (2003).
1
1
0. H.E. Zimmerman. J. Am. Chem. Soc. 117, 8988 (1995).
1. H.E. Zimmerman. J. Phys. Chem. A, 102, 5616 (1998).
References
1
. F. Würthner. Angew. Chem. Int. Ed. 40, 1037 (2001).
12. H. Görner. Photochem. Photobiol. 77, 171 (2003).
©
2007 NRC Canada

This article has been cited by:
1
2
. Devin Mitchell, Peter Wan. 2009. Formal intramolecular photoredox reactions of phenylbenzophenone, xanthone and
fluorenone methanols in aqueous solution. Journal of Photochemistry and Photobiology A: Chemistry 205:1, 34-43.
[CrossRef]
. Yunyan Hou, Lawrence A. Huck, Peter Wan. 2009. Long-range intramolecular photoredox reaction via coupled charge and
proton transfer of triplet excited anthraquinones mediated by water. Photochemical & Photobiological Sciences 8:10, 1408.
[CrossRef]