Photochemical Formation of Anthracene Quinone Methide Derivatives

Article abstract of DOI:10.1021/acs.joc.6b02735

Anthrols 2-7 were synthesized and their photochemical reactivity investigated by irradiations in aq CH₃OH. Upon excitation with visible light (λ > 400 nm) in methanolic solutions, they undergo photodehydration or photodeamination and deliver methyl ethers, most probably via quinone methides (QMs), with methanolysis quantum efficiencies φ_R = 0.02-0.3. Photophysical properties of 2-7 were determined by steady-state fluorescence and time-correlated single photon counting. Generally, anthrols 2-7 are highly fluorescent in aprotic solvents (φ_F = 0.5-0.9), whereas in aqueous solutions the fluorescence is quenched due to excited-state proton transfer (ESPT) to solvent. The exception is amine 4 that undergoes excited-state intramolecular proton transfer (ESIPT) in neat CH₃CN where photodeamination is probably coupled to ESIPT. Photodehydration may take place via ESIPT (or ESPT) that is coupled to dehydration or via a hitherto undisclosed pathway that involves photoionization and deprotonation of radical cation, followed by homolytic cleavage of the alcohol OH group from the phenoxyl radical. QMs were detected by laser flash photolysis and their reactivity with nucleophiles investigated. Biological investigation of 2-5 on human cancer cell lines showed enhancement of antiproliferative effect upon exposure of cells to irradiation by visible light, probably due to formation of electrophilic species such as QMs.

Full text of DOI:10.1021/acs.joc.6b02735

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Photochemical formation of anthracene quinone methide derivatives
Djani Skalamera, Kata Mlinari#-Majerski, Irena Martin Kleiner, Marijeta
Kralj, Jessy Oake, Peter Wan, Cornelia Bohne, and Nikola Basaric
J. Org. Chem., Just Accepted Manuscript • Publication Date (Web): 23 May 2017
Downloaded from http://pubs.acs.org on May 25, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 58
The Journal of Organic Chemistry
1
2
3
4
Photochemical formation of anthracene quinone methide derivatives
5
6
7
8
9
Đani Škalamera,† Kata MlinarićꢀMajerski,† Irena Martin Kleiner,‡ Marijeta Kralj,‡ Jessy Oake,§
Peter Wan,§ Cornelia Bohne§ and Nikola Basarić†*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†Department of Organic Chemistry and Biochemistry, Ruđer Bošković Institute, Bijenička cesta
54, 10 000 Zagreb, Croatia. Fax: + 385 1 4680 195; Tel: +385 1 4561 141
‡Division of Molecular Medicine, Ruđer Bošković Institute, Bijenička cesta 54, 10 000 Zagreb,
Croatia.
§ Department of Chemistry, University of Victoria, Box 1700 STN CSC, Victoria BC, V8W
2Y2, Canada.
Corresponding authors' Eꢀmail address: nbasaric@irb.hr
Graphical abstract
R3
R3
R3
ν
h
O
OH
-HR1
R2
R1
R2 R2
photodehydration
photodeamination
R2
R3
2
3
R1 = OH, R2 = CH3, R3 = H
R1 = OH, R2 = , R3 = H
QM2
R2 = CH3, R3 = H
QM3
R2 =
, R3 = H
4
R1 = N(CH2CH3)2, R2 = H, R3 = H
QM4
R2 = H, R3 = H
+
-
5
6
7
R1 = NH (CH2CH3)2Cl , R2 = H, R3 = H
R1 = OH, R2 = H, R3 = CH3
QM6
QM7
R2 = H, R3 = CH3
R2 = H, R3 = Ph
R1 = OH, R2 = H, R3 = Ph
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 58
1
2
3
4
5
6
7
8
9
Abstract: Anthrols 2ꢀ7 were synthesized and their photochemical reactivity investigated by
irradiations in aq. CH3OH. Upon excitation with visible light (λ > 400 nm) in methanolic
solutions they undergo photodehydration or photodeamination and deliver methyl ethers, most
probably via quinone methides (QMs), with methanolysis quantum efficiencies ΦR = 0.02ꢀ0.3.
Photophysical properties of 2ꢀ7 were determined by steadyꢀstate fluorescence and timeꢀ
correlated single photon counting. Generally, anthrols 2ꢀ7 are highly fluorescent in aprotic
solvents (ΦF = 0.5ꢀ0.9), whereas in aqueous solutions the fluorescence is quenched due to
excited state proton transfer (ESPT) to solvent. The exception is amine 4 that undergoes excited
state intramolecular proton transfer (ESIPT) in neat CH3CN where photodeamination is probably
coupled to ESIPT. Photodehydration may take place via ESIPT (or ESPT) that is coupled to
dehydration, or via a hitherto undisclosed pathway that involves photoionization and
deprotonation of radicalꢀcation, followed by homolytic cleavage of the alcohol OH group from
the phenoxyl radical. QMs were detected by laser flash photolysis (LFP) and their reactivity with
nucleophiles investigated. Biological investigation of 2ꢀ5 on human cancer cell lines showed
enhancement of antiproliferative effect upon exposure of cells to irradiation by visible light,
probably due to formation of electrophilic species such as QMs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Key words: photodehydration, photodeamination, quinone methides, anthracene, laser flash
photolysis, antiproliferative activity
2
ACS Paragon Plus Environment

Page 3 of 58
The Journal of Organic Chemistry
1
2
3
4
Introduction
5
6
7
8
Quinone methides (QMs) are important reactive intermediates in the chemistry and
photochemistry of phenols.1 For example, the activity of some antineoplastic antibiotics such as
mitomycin is based on metabolic formation of QMs (via initial reduction to the phenol) that
alkylate DNA.2,3,4 Therefore, the biological role of QMs5,6,7 has mainly been connected to their
reactivity with nucleobases8,9,10 and DNA.11,12,13,14 However, QMs also react with amino
acids15,16 and proteins.17 We have recently demonstrated that antiproliferative activity of
photogenerated anthracene QMs stems from their reaction with intracellular proteins rather than
with DNA.18
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The most convenient approach for the preparation of QMs in biological systems rely on
photochemical methods19,20 which include photodeamination from the Mannich salts of the
corresponding phenols16,21,22,23 or photodehydration of the corresponding benzyl alcohols.24,25
Photodeamination of Mannich salts has recently been applied in the investigation of biological
activity of QMs,26,27 and the ability of naphthalenediimide QM derivatives to selectively target
guanine quadruplex structures has been demonstrated.28,29,30 Photodeamination can also be
triggered by an intramolecular photoinduced electron transfer reaction with naphthalene diimide
as photooxidant.31 Although a number of methods for the efficient generation of QMs exist,
reports on biological effects of simple QMs are scarce due to their high reactivity and short
lifetimes. QMs cannot be stored; they have to be generated in situ. A further drawback for their
photochemical formation is the use of short wavelength UV light which is not applicable in
biology, and particularly not in medicine. However, Freccero et al,23,32 and our laboratory33 have
recently reported examples of QM photogeneration by use of nearꢀvisible light. Thus, anthrol 1
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 58
1
2
can be excited at λ > 400 nm delivering QM1. The QM is protonated to 1+, and subsequently in a
reaction with nucleophiles gives adducts (Scheme 1).33 Preliminary biological investigation
indicated enhancement of antiproliferative activity for the human cancer cells when irradiated,
suggesting that the effect is due to intracellular photochemical formation of QM1 or 1+.33
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
OH
O
ν
h
Ph
-H2O
Ph
Ph
Ph
QM1
H+
1
OH
OH
Nu
Nu
Ph
Ph
+
Ph
Ph
1+
1Nu
Scheme 1.
Prompted by these preliminary results with anthrol 1,33 we report here a systematic study of the
photochemical reactivity for the photodehydration or photodeamination of anthrols 2ꢀ7, and we
investigated their antiproliferative activity upon irradiation. Anthrol derivatives 2ꢀ7 were all
substituted at the anthracene positions 2 and 3. Compounds 2 and 3 bear methyl or adamantyl
substituents at the hydroxymethyl group which are known to increase the quantum yield for
photodehydration34 and enhance antiproliferative effects.35,36,37 Compounds 4 and 5 are
methylamine derivatives which are anticipated to undergo more efficient formation of QMs in
deamination reactions.22 Moreover, positively charged anthracene 5 may intercalate into DNA
and thus enhance the aptitude for alkylation of photogenerated QMs.14 Anthrols 6 and 7 are
substituted at the anthracene positions 9 and 10 to increase their photochemical stability with
respect to photooxidations and render them more fluorescent. It was anticipated that the
corresponding QMs from 6 and 7 may also be fluorescent, which would make their detection
4
ACS Paragon Plus Environment

Page 5 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
easier. Photochemical reactivity was studied by preparative irradiations in the presence of
CH3OH as a nucleophile. Fluorescence properties and singlet excited state reactivity were
investigated by steady state and timeꢀresolved fluorescence, whereas for the detection of reactive
intermediates, laser flash photolysis was used. For compounds 2ꢀ5, antiproliferative activity was
tested on three human cancer cell lines with and without irradiation. Enhancement of the
antiproliferative effect for the irradiated cells is probably due to photochemical formation of
anthracene QMs in photodehydration and photodeamination reactions. Most importantly,
detailed mechanistic investigation unraveled a possible new pathway for the QM formation that
involves photoionization and a homolytic cleavage of the OH group.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
OH
CH3
OR
OH
H3C
3 R = H
2
3OMe
R = CH3
OH
OH
NH+(CH2CH3)2Cl-
N(CH2CH3)2
4
5
CH3
OH
OH
OH
OH
CH3
6
7
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 58
1
2
3
4
Results
5
6
Synthesis
7
8
9
The synthesis of anthrols 2ꢀ5, started from the commercially available 2ꢀaminoanthraquinone
which was converted to 2ꢀhydroxyanthraquinone (8) via a known procedure.38 Bromination of 8
with excess of bromine afforded a mixture of 3ꢀbromoꢀ2ꢀhydroxyanthraquinone (9) and 1,3ꢀ
dibromoꢀ2ꢀhydroxyanthraquinone (10). The mixture was subjected to a reduction with NaBH4 in
alkaline aqueous solution (1M Na2CO3) to yield pure 2ꢀbromoꢀ3ꢀanthrol (11). Under these
conditions selective debromination occurs on the anthracene position 1, whereas debromination
on the position 3 is insignificant if the reduction is not run longer than 15 min. Thus, 11 was
obtained selectively, in high yield (95%) and with high purity. This modification significantly
improved our previous procedure33 for the preparation of 11 by increasing the yield over two
steps with no need for chromatography. In the next step, 11 was treated with excess of BuLi
according to the procedure developed by Freccero et al.26 on similar systems, which avoids the
use of protecting group on the phenolic OH. The lithiated compound reacted with DMF to afford
the corresponding aldehyde 12 (Scheme 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
O
O
Br
1.) NaNO2
H2SO4
OH
NH2
OH
Br2
OH
Br
+
Br
HOAc
ꢀ
ꢀ
2.) H2O,
O
O
O
O
10
NaBH4
9
8
Na2CO3/H2O
1) BuLi
OH
OH
CHO
2) DMF
3) H+
Br
11
12
Scheme 2.
6
ACS Paragon Plus Environment

Page 7 of 58
The Journal of Organic Chemistry
1
2
3
4
Anthrols 3 and 4 were prepared from bromoanthrol 11 which was treated with excess BuLi and
5
6
then, organolithium compound was quenched with acetone or 2ꢀadamantanone (Scheme 3). We
7
8
9
also prepared methyl ether 3OMe by a simple methylation of 3 with CH3I in basic conditions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(K2CO3).
O
1)
OH
OH
Br
OLi
Li
2BuLi
OH
2) H+
3
O
11
1)
2) H+
OH
OH
CH3
H3C
2
Scheme 3.
Methylamine derivative 4 was obtained from carbaldehyde 12 in a reductive amination, and
subsequently transformed to hydrochloride salt 5 (Scheme 4).
OH
N(CH2CH3)2
OH
HN(CH2CH3)2 / EtOH
2) NaBH4 / EtOH
1)
O
H
4
12
HCl
5
Scheme 4.
Synthesis of anthrols 6 and 7 started from antraquinone 9.33 Treatment with an excess of MeLi or
PheLi followed by reduction gave methylated anthrol derivative 13-Me or phenyl derivative 13-
Ph. Compounds 13 in a reaction with BuLi and quenching with DMF afforded carbaldehydes
14-Me or 14-Ph that were reduced to alcohols 6 or 7, respectively (Scheme 5).
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 58
1
2
3
4
5
R
O
OH
1) RLi
OH
6
7
8
9
2) SnCl2/HCl (R=Me)
or HI/Et2O (R=Ph)
Br
Br
R
O
13-Me
13-Ph
R = CH3
R = Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
1) BuLi / Et2O
2) DMF
3) H+
R
R
R
R
OH
OH
OH
NaBH4
THF-H2O 1:1
CHO
14-Me
14-Ph
R = CH3
R = Ph
6
7
R = CH3
R = Ph
Scheme 5.
Photochemistry
Based on literature precedent,22,24,33 it is expected that irradiation of anthrols 2ꢀ7 in CH3OHꢀH2O
would give photomethanolysis products via QM intermediates. Preparative irradiations of 2ꢀ7
were conducted by irradiating CH3OHꢀH2O (4:1) solutions at 350 nm and by analyzing the
composition of the solutions by HPLC (see Figs S1ꢀS4 in the supporting information).
Methanolysis of 2, 3, 6 and 7 run to high conversions of 80ꢀ90% afforded cleanly methyl ether
products 15, 16, 18 and 19, respectively (Scheme 6). However, irradiation of 4 and 5 gave
methyl ether 17 formed as the major product, and small amounts of alcohol (1ꢀ5%) detected by
HPLC (see Figs S3 and S4 in the supporting information). Irradiations of 2 and 3 at different
H2O concentrations gave the same products 15 and 16 irrespective of H2O content. Although a
clear trend cannot be seen with 2, photomethanolysis is generally more efficent at higher H2O
8
ACS Paragon Plus Environment

Page 9 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
concentration, due to faster phenol deprotonation from S1 to H2Oꢀclusters then to
CH3OH.39,40,41,42 Water concentration had no influence on the efficiency of photomethanolysis of
4, whereas for 5, a higher methanolysis efficiency was observed at higher H2O concentration, in
accord with precedant literature.22 Photomethanolysis was also conducted by irradiating anthrols
in CH3OHꢀH2O (4:1) by use of lamps with the output at 420 nm or "cool white" light where the
same products 15ꢀ19 were obtained. Excitation of 2ꢀ7 by near visible light is particularly
important for the applicability in biological systems.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
R3
R3
ν
h
OH
OH
-HR1
OCH3
R2 R2
R1
R2 R2
CH3OH-H2O
(4:1)
R3
R3
2
3
R1 = OH, R2 = CH3, R3 = H
R1 = OH, R2 = , R3 = H
15
16
R2 = CH3, R3 = H
R2 = , R3 = H
4
R1 = N(CH2CH3)2, R2 = H, R3 = H
17
R2 = H, R3 = H
+
-
5
R1 = NH (CH2CH3)2Cl , R2 = H, R3 = H
R1 = OH, R2 = H, R3 = CH3
6
7
18
19
R2 = H, R3 = CH3
R2 = H, R3 = Ph
R1 = OH, R2 = H, R3 = Ph
Scheme 6.
The efficiency of photomethanolysis (ΦR) for 2ꢀ5 was determined by the simultaneous use of
three actinometers, ferrioxalate (Φ254 = 1.25),43,44 KI/KIO3 (Φ254 = 0.74),43,45 and valerophenone
(Φ254 = 0.65 ± 0.03),43,46 as we already described for 1.33 For 6 and 7 only KI/KIO3 was used,
since consistent results were obtained as for 2ꢀ5 with three actinometers. The measured values of
ΦR are compiled in Table 1. Comparison of ΦR for phenyl 1 and methyl derivative 2 indicates
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 58
1
2
3
4
5
6
7
8
9
that the introduction of the substituents to the methyl group decreased the reaction efficiency. On
the contrary, the highest ΦR was observed for the adamantyl derivative 3. The adamantyl group
probably increases the stability of the corresponding QM and thus enhances the methanolysis
reaction efficiency. An analogous trend has already been observed in similar phenyl34 and
naphthyl derivatives.37 Furthermore, the efficiency ΦR in the photodeamination from 5 is less
efficient than elimination of the amine, contrary to previous report with simple cresol
derivatives.22 The observation may be explained by different electronic influence of the
methylamine or methylammonium group to the photophysical properties of anthrol.
Table 1. Photomethanolysis quantum yields (ΦR) for 1ꢀ7.a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Comp.
ΦR
0.023±0.001b,c
0.015±0.001b
0.33±0.01b
0.22±0.01b
0.15±0.01b
0.04±0.01d
0.048±0.009d
1
2
3
4
5
6
7
a Measurement conducted in CH3OHꢀH2O (4:1) at 254 nm. Measurements were done in triplicate
and the mean value is reported. The quoted error corresponds to the maximum absolute
deviations.
10
ACS Paragon Plus Environment

Page 11 of 58
The Journal of Organic Chemistry
1
2
3
4
b
Measured by simultaneous use of three actinometers, ferrioxalate (Φ254 = 1.25),43,44 KI/KIO3
5
6
7
(
Φ Φ
254 = 0.74),43,45 and valerophenone ( 254 = 0.65±0.03),43,46ΦR was calculated according to Eq.
8
9
S1ꢀS5 in the supporting information.
c Value taken from ref. 33.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
d Measured by use of KI/KIO3 actinometer.
It is known that QMs react with electron rich alkenes in DielsꢀAlder reaction giving
chromanes.25,47,48,49 However, irradiations of 2ꢀ7 in neat CH3CN or aqueous CH3CN in the
presence of ethyl vinyl ether (EVE) did not give chromane adducts, probably due to slow
kinetics of the DielsꢀAlder reaction (vide infra). Thus, photochemistry of anthrols cannot be
exploited further in different applications, as elegantly demonstrated for naphthol derivatives by
Popik et al.47,48,50
Photophysical properties
Photophysical properties of 2ꢀ7 are important for the understanding of their photochemical
reactivity from the singlet excited state S1. Absorption and fluorescence spectra were measured
in CH3CN and CH3CNꢀH2O (1:1) (for absorption and fluorescence spectra see Figs S5ꢀS24 in the
supporting information). The absorption spectra of 2, 3, 6 and 7 in CH3CN and CH3CNꢀH2O, are
characterized by a band at 310ꢀ420 nm, corresponding to S0→S1 transition. Thus, anthrol
derivatives can be excited with nearꢀvisible light (>400 nm), which is important for biological
applications. Amine derivatives 4 and 5 were not soluble in neat CH3CN at the concentrations for
the molar absorption coefficients measurements, but are soluble in neutral CH3CNꢀH2O. In
aqueous solution, 4 and 5 have a band in the absorption spectrum at 310ꢀ420 nm and a shoulder
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 58
1
2
3
4
5
6
7
8
9
stretching to 500 nm which is attributed to the phenolate of these compounds (Figs S12 and S18
in the supporting information). The pKa of the phenolic OH is probably significantly lowered by
the presence of the oꢀmethylammonium group, as seen with cresol derivatives.22 Thus, in the
neutral aqueous solution the methylamine is protonated to give 5, and phenol 5 is in equilibrium
with zwitterionic phenolate 4zw (Scheme 7).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ν
h
S1
OH
N(CH2CH3)2
*
ESIPT
4zw
CH3CN
4
H2O, pH = 7
OH
ν
h
S1
ESPT
*
NH+(CH2CH3)2
4zw
-H+
H2O
5
H2O
± H+
Ka
O-
NH+(CH2CH3)2
4zw
Scheme 7.
Fluorescence spectra of 2 and 3 in CH3CN have bands at 400ꢀ500 nm, and the quantum yields
are high, ΦF ≈ 0.8 (Table 2). The fluorescence decay was fit to a monoexponential function with
relatively long S1 lifetimes (τ ≈ 20 ns). Substituted anthracenes 6 and 7, as well as ammonium
salt 5, have similar fluorescence spectra with a maximum at 400ꢀ500 nm (Fig 1). They are also
highly fluorescent with ΦF = 0.5ꢀ0.9, but their fluorescence decays could not be fit to a single
12
ACS Paragon Plus Environment

Page 13 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
exponential function. Global analysis of the fluorescence decays collected over different
wavelengths revealed two decay lifetimes (see Eq. S7, Table S1 and figs S26 and S27 in the
supporting information). For anthrols 5ꢀ7 these probably correspond to two conformers, as seen
with 2ꢀmethoxyanthracene.51
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4 / CH3CN
4 / CH3CN-H2O
5 / CH3CN
5 / CH3CN-H2O
400
450
500
550
600
650
Wavelength / nm
6 - CH3CN
7 - CH3CN
6 - CH3CN-H2O
7 - CH3CN-H2O
400
450
500
550
650
700
Wavelength /6n0m0
Fig 1. Normalized fluorescence spectra (λexc = 370 nm) of 4 and 5 (top), and 6 and 7 (bottom) in
CH3CN or CH3CNꢀH2O (1:1).
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 58
1
2
3
4
Table 2. Photophysical properties of anthracenes 1ꢀ7.
5
6
7
8
ΦF(CH3CN)b ΦF(CH3CNꢀH2O)b
τ (CH3CN)/ns c
17.8±0.1
τ (CH3CNꢀH2O)/ns c, d
1a
2
0.86±0.01
0.80±0.05
0.79±0.01
0.39±0.01
0.54±0.02
0.29±0.02
1.7±0.2 (growth) f
8.1±0.2 (growth)
24.5±0.1 (decay)
3.7±0.1 (growth) f
13.8±0.1 (growth)
20.8±0.1 (decay)
2.1±0.1 (growth) f
10.1±0.1 (growth)
19.8±0.1 (decay)
26.9±0.1
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
23.6±0.2
21.8±0.1
3
0.85±0.05
0.10±0.03e
ꢀ
21.1±0.1
2.3±0.3 f
16.3 ± 0.1
3OMe
4
0.27±0.06e
1.9±0.1 (growth) f
7.2±0.1 (growth)
23.1±0.1 (decay)
2.0±0.1 (growth) f
5.7±0.1 (growth)
23.0±0.1 (decay)
8.0±0.1 (growth)f
16±3 (decay)
0.45±0.05
0.52±0.02
0.25±0.06e
9.0±0.1 f
27.3±0.1
5
6
0.30±0.02
0.25±0.02
17.3±0.1 f
20.4±0.1
13.4±0.1 f
14.3±0.1
0.92±0.02
4.5±0.1 (growth)f
18.2±0.1 (decay)
7
a
b
Taken from ref 33. Quantum yields of fluorescence (ΦF) were measured by use of quinine
sulfate in 0.05 M aqueous H2SO4 (ΦF = 0.55) as a reference.52 An average value is reported from
single experiment by excitation at three wavelengths. The errors quoted correspond to maximum
c
absolute deviations (see experimental). Singlet excited state lifetimes were obtained by global
fitting of fluorescence decays measured at several wavelengths by timeꢀcorrelated single photon
d
counting (see experimental). 'Growth' and 'decay' indicate the type of kinetics at the emission
e
f
wavelengths for the phenolate. Depends on the excitation wavelength (increases with λ). For
details on fitting and preꢀexponential factors that depend on the detection wavelength see
supporting information Tables S1 and S2.
Ammonium derivative 4 in neat CH3CN has different fluorescence properties then observed for
the other anthrols. It exhibits typical dual fluorescence, with an additional band at 575 nm
attributed to the phenolate emission. The assignment of the bands in the fluorescence spectra is
based on experiments in the presence of acid or base. In the presence of H2SO4 only the band at
400ꢀ500 nm is observed, whereas on addition of triethylamine the band at 575 nm becomes
14
ACS Paragon Plus Environment

Page 15 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
stronger (see Fig S16 in the supporting information). The presence of phenolate is also apparent
in the absorption spectrum. Therefore, the relative ratio of the fluorescence intensity from the
phenol and the phenolate, as well as the ΦF value, depend on the excitation wavelength (see Fig
S14 and S15 in the supporting information). The equilibrium between the phenol and the
phenolate is temperature dependent with higher phenolate content present at a lower temperature
(see Fig S17 in the supporting information). The fluorescence decay for 4 in CH3CN was fit to a
sum of two exponentials with the decay lifetimes corresponding to the phenol (2.2 ns) and
phenolate (16.3 ns), both with positive preꢀexponential factors suggesting that phenolate is
formed in the ground state (Table S2 in the supporting information). In addition to the phenolate
in equilibrium in S0, it is probably also formed in S1 by ESIPT due to the enhanced acidity of
phenol in S1 and the proximity of the basic amine (Scheme 7). Therefore, ΦF for 4 in CH3CN is
much lower than for the other anthrols.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Addition of H2O to the CH3CN solutions of anthrols significantly quenched the fluorescence for
all anthrols except for amine 4 and methoxy derivative 3OMe. In addition to fluorescence
quenching, in all cases except for 3OMe typical dual fluorescence was observed with two
emission bands, at 450 nm and 575 nm, attributed to the phenol and phenolate, respectively. The
methoxy derivative does not exhibit dual fluorescence since it does not have a phenolic OH.
Furthermore, anthrol 4 is an exception that has higher ΦF in aqueous solution. Namely, in neutral
aqueous solution the amine is protonated which blocks the efficient ESIPT pathway from taking
place in neat CH3CN. Thus, in neutral aqueous solution 4 and 5 have the same fluorescence
properties. Phenolates are formed in S1 in the aqueous solution by ESPT to solvent, as indicated
by single photon counting (SPC) measurements where the kinetics were fit to a sum of two or
more exponentials with negative preꢀexponential factors for kinetics collected at longer
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 58
1
2
3
4
5
6
7
8
9
wavelengths where the phenolates emit, consistent with a growth kinetics for the formation of
the phenolate (for kinetic schemes,53,54 see Scheme S1, Eqs. S7ꢀS10 and preꢀexponential factors
Table S1 in the supporting information). We collected timeꢀresolved fluorescence spectra for
anthrol 7 with the highest ΦF in order to obtain more information on the decay processes
involving S1. The spectra are in accord with the phenol emitting at short delays after excitation
and the phenolate present at longer delays (see Fig S24 in the supporting information).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The fluorescence decays for anthrols 1ꢀ5 in aqueous solution were fit to a sum of three
exponentials where at the emission wavelength for the phenolate two growing components with
negative preꢀexponential factors were observed which correspond to the two shorterꢀlived
species. We propose that these two shortꢀlived species correspond to two conformers of anthrol
that deprotonate in S1 with different rate constants (see Scheme S2, an example of decay Fig S27
and preꢀexponential factors Table S2 in the supporting information). One anthrol conformer
gives phenolate by ESPT to solvent, whereas the other anthrol could undergo ESPT to solvent
and ESIPT, giving also zwitterionic phenolate and leading eventually to dehydration and QM
formation. The other plausible assignment for the two shortꢀlived species may be due to the
presence of phenols with different number of associated H2O molecules that deprotonate in S1
with different rate constants. Namely, it is known that several H2O molecules are necessary as
proton accepting species in ESPT, giving rise to nonꢀlinear quenching of fluorescence by H2O.55
We do not have sufficient information from the timeꢀresolved studies to differentiate between
these two possibilities.
16
ACS Paragon Plus Environment

Page 17 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
The solution pH should in principle be controlled for ESPT reactions. That is why some
measurements were conducted in the presence of phosphate buffer (c = 0.1 M) at pH 7.0
(particularly those corresponding to compounds 6 and 7). However, the results were not different
from the cases when the solutions were not buffered. We have demonstrated on some other
molecules undergoing dehydration that the efficiency for the QMs formation is pHꢀindependent
over a wide pH range (pH 3ꢀ9).34,35,36,37 For the deamination reaction, QM formation depends on
the prototropic form of the molecule, being the most efficient reaction when the molecule is in
the zwitterionc form (at pH>9).22 Thus, in nearꢀneutral conditions, no influence of pH is
expected to the kinetics of QM formation. Moreover, hydration of QMs is acid and base
catalyzed but in the wide nearꢀneutral pH region the reaction is pH independent.25,56,57
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Laser Flash Photolysis (LFP)
LFP experiments were conducted for 2ꢀ7 to probe for the formation of QMs, carbocations, or
other reactive intermediates of biological relevance. The transient absorption spectra were
measured by excitation at 355 nm with a Nd:YAG laser in CH3CN and CH3CNꢀH2O, where
different behavior was anticipated due to the presence or not of ESPT pathways. Moreover, the
measurements were conducted in N2ꢀ and O2ꢀpurged solutions, where O2 is expected to quench
some transient species (triplets and radicals) shortening their lifetimes, while QMs and
carbocations should not be affected since they do not react with oxygen (for the transient
absorption spectra, decays and quenching plots see Figs S28ꢀS64 in the supporting information).
In both N2ꢀ and O2ꢀpurged CH3CN solutions of 2, a transient was detected with a maximum at
690 nm that did not decay with firstꢀorder kinetics (Fig 2 top). In O2ꢀpurged solution the
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 58
1
2
transient was fit to a sum of two exponentials with k ≈ 2.2×106 sꢀ1 (τ = 0.45±0.05 ꢁs) and k ≈
4.1×105 sꢀ1 (τ ≈ 2.4 ꢁs, for the details on fitting see Figs. S29ꢀS31 in the supporting information).
The longerꢀlived transient was quenched with CH3OH and H2O, but the rate constant for the
quenching could not be estimated due to the nonlinear dependence of kobs on concentration (see
Figs S32 in the supporting information). In aqueous solution the transient at 700 nm cannot be
detected because its lifetime is shorter than the detection limit of the setup used. Based on the
quenching with CH3OH and H2O and comparison with precedent spectra,58,59 the transient with
the lifetime of 2.4 ꢁs was assigned to radicalꢀcation 2RC, but we cannot give a firm assignment
for the shortꢀlived transient (τ = 0.45 ꢁs). It cannot correspond to the triplet excited state, Cꢀ
centered radical or solvated electron, because the measurement was performed in oxygenated
CH3CN where these transients would be quenched. Moreover, it cannot be the phenoxyl radical
which is expected to absorb at shorter wavelengths.60,61 The assignment of this transient is
beyond the scope of this report focused on QM formation and no further attempts for its
identification were made. In aqueous solutions, the phenol radicalꢀcations deprotonate giving
phenoxyl radicals.62 Methanol or H2O are not typical quenchers that react in a bimolecular
reaction with radicalꢀcations. Since several molecules are required as proton accepting
species,55,63 nonꢀlinear quenching plots of the transient were observed. Furthermore, in CH3CN
and CH3CNꢀH2O solutions of 2 additional transients were detected between 350 and 600 nm.
Thus, in the O2ꢀpurged aqueous solution of 2 (Fig 2, bottom) at short delays after the laser pulse,
a transient was observed at 480 nm, where the decay (k ≈ 9×104 sꢀ1) matched with the rise of
another transient absorbing at shorter wavelengths 350ꢀ400 nm. This latter transient decays over
the ms timescale, τ ≈ 200ꢀ300 ms (see fig S34 in the supporting information). Popik et al. also
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
detected two transients in a naphthalene series (from 3ꢀhydroxymethylꢀ2ꢀnaphthol) and
18
ACS Paragon Plus Environment

Page 19 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
explained this observation by the formation of benzoxete NBO that undergoes rearrangement to
quinone methide NQM.25 However, anthracene benzoxete derivative ABO2 cannot absorb light
at 450ꢀ550 nm because it contains only the anthracene chromophore. Therefore, we tentatively
assigned the shortꢀlived transient with a maximum at 480 nm to phenoxyl radical 2PhO, based
on the position of the absorption maximum and decay kinetics from precedent literature.60,61
Precise decay kinetics for the longꢀlived transient could not be revealed with the setꢀup used,
precluding quenching studies to confirm the assignment of this transient. We tentatively assign it
to QM2 since the methanolysis experiments strongly indicated the formation of this transient.
LFP experiments for 2 were also conducted in 2,2,2ꢀtrifluoroethanol (TFE), in which
electrophilic species such as QMs64,34,36 and carbocations65,66 exhibit long lifetimes due to the
polar nonꢀnucleophilic character of the solvent (see Fig S36 in the supporting information).
However, the spectra did not differ from those measured in aqueous CH3CN. Contrary to 1,
formation of QM2 and the corresponding carbocation could not be timeꢀresolved.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.10
0.08
N2
O2
0.06
2 RC
0.04
0.02
0.00
300
400
500
600
700
Wavelength / nm
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 58
1
2
3
4
5
6
7
8
0.20
0.15
0.10
0.05
0.00
0.8
4.8
19
ꢁ
ꢁ
s
s
ꢁs
9
40 ꢁs
QM 2
2 PhO
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
300
400
500
600
700
Wavelength / nm
Fig. 2. Transient absorption spectra of 2 in N2ꢀ and O2ꢀpurged CH3CN (top) collected ≈ 2 ꢁs after
the laser pulse, and O2ꢀpurged CH3CNꢀH2O (bottom).
O
O
O
CH3
CH3
NBO
NQM
ABO2
OH
OH
CH3
O
O
OH
CH3
CH3
CH3
H3C
H3C
2RC
2PhO
QM2
O
OCH3
OH
OH
+
QM3
3OMeRC
3+
Transient absorption spectra of 3 in CH3CN and CH3CNꢀH2O solution resemble those of 2. In
the CH3CN solution we detected the radicalꢀcation (λmax = 700 nm, τ = 950±50 ns) and solvated
electron which was also detected in the O2ꢀpurged aqueous solution (λ = 700 nm, τ ≈ 50 ns)
where radicalꢀcation rapidly deprotonates. Reported lifetimes for solvated electron in aqueous
solution vary from 10 ꢁs67 to 300ꢀ500 ꢁs.68 Nevertheless, O2 quenches solvated electron with the
rate constant of 2×1010 Mꢀ1sꢀ1,69 so in O2ꢀpurged aqueous solution the lifetime of hydrated
20
ACS Paragon Plus Environment

Page 21 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
electron should be ≈ 35 ns. Assignment of the transient at 700 nm to the radicalꢀcation is further
supported by LFP experiments with 3OMe (see Fig 3 top and S43 in the supporting information)
where the same transient was detected in both CH3CN (τ = 3.6±0.1 ꢁs) and CH3CNꢀH2O (τ =
50±2 ꢁs). Namely, 3OMeRC cannot deprotonate to phenoxyl radical in H2O since it has no free
phenolic OH. On the contrary, for anthrol 3 in CH3CN solution, the phenoxyl radical was
detected at 300ꢀ500 nm (τ ≈ 10 ꢁs) together with a longꢀlived transient that has a lifetime of
seconds that we tentatively assigned to QM3. LFP in TFE also did not provide any
representative spectrum that could be assigned to QM. However, in 1,1,1,3,3,3ꢀ
hexafluoroisopropanol (HFIP), 3 gave a strong transient with a maximum at 450 nm (Fig 3
bottom and Figs S41, S42 in the supporting information). The transient decayed with almost
firstꢀorder kinetics to the baseline (k =2.5±0.5 sꢀ1, τ = 0.4±0.1 s), and its decay become faster in
the presence of nucleophiles CH3OH or H2O. However, the quenching rate constants could not
be estimated due to a nonꢀlinear dependence of kobs with quencher concentration. At 500ꢀ600 nm,
an additional weaker transient was observed that decayed much faster (k = 3×106 sꢀ1, τ = 360±30
ns). The transient absorbing at 500ꢀ600 nm was tentatively assigned to QM3 based on the
position of the absorption maxima and decay kinetics from precedent literature,33 whereas the
longꢀlived transient corresponds to adamantyl cation 3+. Namely, the absorption spectrum of 3 in
HFIP (see Fig S41 in the supporting information) indicated the presence of 3+, which is formed
in S0 due to the high acidity of HFIP. However, excitation to S1 increases the concentration of 3+
by photoelimination of H2O. It remains unclear whether H2O photoelimination takes place via
QM3, or cation 3+ is directly formed in a HFIP acidꢀcatalyzed pathway.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 58
1
2
3
4
5
6
7
8
0.08
0.06
0.04
0.02
0.00
O2-purged solution of 3
N2-purged solution of 3OMe
3OMe RC
3RC
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
300
400
500
600
700
Wavelength / nm
3+
0.30
0.25
0.20
0.15
0.10
0.05
0.00
ꢀ
A0.25
0.20
0.15
0.10
0.05
0.00
0.0
0.5
1.0
1.5
2.0
Time / s
QM3
300
400
500
600
700
Wavelength / nm
Fig 3. Transient absorption spectra of 3 and 3OMe in O2ꢀpurged CH3CN or N2ꢀpurged CH3CNꢀ
H2O (1:1) collected 700 ns after the laser pulse (top) and in 1,1,1,3,3,3ꢀhexafluoroisopropanol
(HFIP) collected 760 ns after the laser pulse (bottom, weak ꢂA at 500ꢀ600 nm corresponds to
QM3, whereas the maximum at 450 nm corresponds to 3+; inset: decay of 3+at 450 nm).
LFP experiments of 4 and 5 gave rise to shortꢀlived transients absorbing over the whole visible
part of the spectrum (see Figs S44ꢀS51 in the supporting information). For example, in the N2ꢀ
purged CH3CN solution of 4 the transient decayed with k = (1.0±0.1) ×107 sꢀ1 (τ = 100±10 ns).
Interestingly, O2 and N2O only weakly quenched the transient (O2 or N2Oꢀpurged CH3CN k =
22
ACS Paragon Plus Environment

Page 23 of 58
The Journal of Organic Chemistry
1
2
(1.3±0.1) ×107 sꢀ1, τ = 80±5 ns), precluding its assignment to solvated electron or the triplet
excited state. Transient absorption spectra of 5 looked the same as those for 4, but the transient
decayed faster (N2ꢀand O2ꢀpurged CH3CN, τ = 65±5 ns). The shortꢀlived transient was not
quenched by nucleophiles such as CH3OH, ethanolamine or NaN3. On the contrary, the transient
in H2O lived longer (O2ꢀ purged CH3CNꢀH2O, τ = 130±10 ns). Although a firm assignment of
the shortꢀlived transients for 4 and 5 are not warranted at this point, we may tentatively say that
these are anthracene radicalꢀions formed by photoionization or electron transfer. Namely, the
absorption properties of anthracene radicalꢀcations and radicalꢀanions are very similar.70
Moreover, amines are known to form exciplexes with anthracenes where in a polar solvent a
complete charge transfer may lead to the formation of anthracene radicalꢀanions.71 These fast
components are not the focus of this study, which is the elucidation of mechanisms leading to the
QM formation, and were not pursued further. In addition to the shortꢀlived transient absorbing
over the whole spectrum, a transient was detected from 4 and 5 in CH3CN and CH3CNꢀH2O
absorbing at 400ꢀ600 nm that was longer lived (k ≈ 1×105 sꢀ1, τ ≈ 10 ꢁs), and not affected by the
addition of O2. Based on the comparison with the transient absorption spectra of 2 and 3, the
longerꢀlived transient at 400ꢀ600 nm probably corresponds to phenoxyl radical. LFP experiments
in TFE gave the spectra with the same appearance as those measured in CH3CN and CH3CNꢀ
H2O. Thus, the anticipated QM4 formed in the photodeamination from 4 and 5 was not detected
by LFP.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Transient absorption spectra from 6 and 7 are very similar to those of 2 and 3. Thus, in N2ꢀ
purged CH3CN solution of 6, a radicalꢀcation 6RC was detected at 500ꢀ700 nm that decays by
firstꢀorder kinetics with the lifetime of τ = 2.2±0.2 ꢁs. In addition, solvated electron was detected
with very short decay time, τ < 50 ns. In the spectral region at 400ꢀ500 nm, two more transients
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 58
1
2
3
4
5
6
7
8
9
were detected decaying with τ ≈ 5 ꢁs and τ =3±1 ms. The shortꢀlived one probably corresponds
to the triplet excited state since it was quenched by O2, whereas the longꢀlived one might
correspond to QM6. Detection of the triplet excited state for 6, but not for 2ꢀ5 may be explained
by influence of substituents at the anthracene positions 9 and 10. In anthracenes, the substitution
significantly affects the photophysical properties by changing the energy gap between the
anthracene S1 and T2 state leading to different intersystem crossing rates.72 In an O2ꢀpurged
solution the efficiency for the formation of radicalꢀcation and its decay kinetics were not affected
(in O2ꢀpurged solution τ = 3.7±0.2 ꢁs). In the aqueous solution of 6, the radicalꢀcation was not
detected due to fast deprotonation. Instead, several transients were detected at 400ꢀ600 nm, two
shortꢀlived with τ = 10ꢀ30 ꢁs, and longꢀlived one with τ = 0.55±0.05 ms. The transients were not
affected by O2. Thus, in O2ꢀpurged solution at 440 nm we observed a growth with the rate
constant k = 1.0×105 sꢀ1 (τ = 10.0±0.4 ꢁs), and transients absorbing at longer wavelengths that
did not decay with firstꢀorder kinetics, but had similar lifetimes (τ = 13ꢀ17 ꢁs) to the lifetime for
the growth at 440 nm (See Fig S55 in the Supporting Information). One additional transient
absorbing at ≈400ꢀ500 nm was also detected whose decay took place over much longer time
scale τ = 0.60±0.05 ms. To ensure that the observed transients were not from the excitation of
oxygenated anthracene photoproducts, the decay kinetics were measured using a flow cell. The
flow rate was increased until no changes were observed for the decays, ensuring that
photoproducts were not being excited. To assign the transients, quenching studies were
performed with ascorbate, ubiquitous quencher for phenoxyl radical,73 and nucleophiles (NaN3
and ethanolamine) that quench QMs.25,35,36,37,64 The quenching study revealed that the growth at
440 nm or shortꢀlived transient (13ꢀ17ꢁs) absorbing at 500ꢀ600 nm do not react with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
nucleophiles. On the contrary, the growth at 440 nm became faster in the presence of ascorbic
24
ACS Paragon Plus Environment

Page 25 of 58
The Journal of Organic Chemistry
1
2
acid. From the growth quenching, the rate constant was estimated at kq =8±2×107 Mꢀ1 sꢀ1 (see Fig
S56 in the supporting information) which is in accord with the literature precedent for phenoxyl
radical (kq = 6.9 × 108 Mꢀ1 sꢀ1,74 or the radical from tyrosine at pH 7, kq = 4 × 108 Mꢀ1 sꢀ1).75 At
high ascorbate concentration (3 mM) that completely quenched the growth at 440 nm, the decays
at 500ꢀ600 nm became singleꢀexponential, revealing that the transient absorption of the phenoxyl
radical 6PhO overlaps with an additional species with the lifetime of 13ꢀ17 ꢁs that has a strong
absorbance at 580 nm. The amplitude for this transient was not affected by the addition of
ascorbate indicating that it is not formed from the phenoxyl radical. Since this transient was
observed in N2 and O2ꢀpurged solution, it cannot be the triplet excited state or a carbon centered
radical. It does not react with nucleophiles so it cannot be a carbocation or QM6. However, H2O
was required for the formation of this transient. Since this species is not involved in the
formation of QMs, further studies to assign it are not within the scope of this work. On the other
hand, nucleophiles quenched the longꢀlived transient absorbing at 400ꢀ500 nm (Table 3). Thus,
the transient with the lifetime of 0.5 ms is assigned to QM6. In conclusion, the quenching study
undoubtedly revealed the assignment of phenoxyl radical 6PhO that absorbs at 400ꢀ550 nm.
6PhO decays giving new species absorbing at 400ꢀ500 nm that lives longer, so this reaction of
6PhO we assign to the formation of QM6.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 3. Lifetimes of QM6 and QM7 in aqueous solution in the absence of quencher and
quenching constants with nucleophiles (kq / sꢀ1Mꢀ1) obtained by LFP.a
τ / ꢁs
kq / sꢀ1Mꢀ1
kq / sꢀ1Mꢀ1
NaN3
HOCH2CH2NH2
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 58
1
2
3
4
5
6
7
8
9
550±50
400±50
(8±3)×105
(4.8±0.7)×103
(5±1)×104
QM6
(4.8±0.8)×106
QM7
a
Measurement performed in air saturated CH3CNꢀH2O (1:1) in the presence of sodium
phosphate buffer (pH = 7.0, c = 0.1 M). Decays were collected at 400ꢀ500 nm in the presence of
nucleophiles at different concentrations (for the quenching plots see Figs S56, S57, S62 and S63
in the supporting information).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In N2ꢀpurged CH3CN solution of 7, a radicalꢀcation 7RC was detected that absorbs at 500ꢀ700
nm and decays by firstꢀorder kinetics. Interestingly, 7RC (τ = 5.5±0.3 ꢁs) has slower decay
kinetics than 6RC. In addition, at 600ꢀ700 nm solvated electron was detected (very short decay
time, < 50 ns in O2ꢀpurged solution), whereas at 400ꢀ500 nm, the triplet was detected that
decayed with τ ≈ 3 ꢁs (assignment based on O2 quenching). In the region of spectrum at 400ꢀ500
nm two more transients were detected decaying with τ ≈ 30 ꢁs and τ = 0.45±0.05 ms. Based on
literature precedent, the shortꢀlived transient was assigned to the phenoxyl radical formed by
deprotonation of the radicalꢀcation,60,61,62 and the longꢀlived transient was tentatively assigned to
7QM. In O2ꢀpurged CH3CN solution of 7, the efficiency for the formation of radicalꢀcation, and
its decay kinetics were not affected (in O2ꢀpurged solution τ = 6.5±0.5 ꢁs). Similarly, the lifetime
of phenoxyl radical 7PhO was the same as in the N2ꢀpurged solution. In the aqueous solution the
radicalꢀcation was not detected due to its very fast decay by deprotonation to phenoxyl radical,
but we detected several transients absorbing at 400ꢀ600 nm. Shortꢀlived transients with lifetimes
of 25±5 ꢁs did not decay with firstꢀorder kinetics, whereas the longꢀlived decay was fit to single
exponential with the lifetime of 100±10 ꢁs. Purging the solution with O2 did not affect the decay
of these transients, nor the efficiency of their formation. Thus, in O2ꢀpurged solution the
26
ACS Paragon Plus Environment

Page 27 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
transients have lifetimes of 40±10 ꢁs and 200±50 ꢁs. Similar to the photochemistry of 6,
ascorbate affected shortꢀlived transients. Upon addition of high ascorbate concentration, the
decay at 500ꢀ600 nm became singleꢀexponential revealing that the absorption of phenoxyl
radical 7PhO overlaps with the absorption of additional transient that was not assigned. Most
importantly, quenching studies revealed that the shortꢀlived transients do not react with
nucleophiles, but the longꢀlived transient (200 ꢁs) does (Table 3), so it was assigned to QM7.
Electron rich alkenes are specific quenchers for QMs that react in Diels Alder reactions.22,25
However, attempts to quench QMs by EVE failed, in accord with the preparative irradiations
where chromane products were not formed. An additional problem in the quenching experiments
was the fact that EVE and aqueous CH3CN solution containing buffer were not miscible. When
EVE concentration exceeded 0.05 M, the separation of aqueous and organic layer took place.
Considering the lifetime of QMs in submilliseconds, the rate constant has to be > 105 Mꢀ1 sꢀ1 to
observe the quenching. QMs react with alkenes in DielsꢀAlder reactions, but the kinetics is
relatively slow. For example, reported rate constant for NQM with EVE is kq = 4.1 × 104
Mꢀ1 sꢀ1.25 Introducing steric hindrance with adamantane significantly slows down the Diels Alder
reaction. Thus, the adamantyl derivative corresponding to NQM reacts with EVE with kq = 85
Mꢀ1 sꢀ1.37
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In summary, LFP measurements allowed for the detection of several transient species. The
transients observed at 690 nm were assigned to radicalꢀcations. This assignment is clearly proven
by quenching with CH3OH and H2O where the typical nonꢀlinear behavior by addition of proton
accepting solvent was observed, as described in literature precedent.55,62,63 The phenoxyl radicals
absorbing at 480 nm were proven by nonꢀquenching with O2 and similar lifetimes in aqueous and
nonꢀaqueous solution. The quenching experiments for Oꢀcentered radicals are problematic since
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 58
1
2
3
4
5
6
7
8
9
they do not have the same reactivity as Cꢀcentered reactive radicals. Thus, quenching with
ubiquitous radical quenchers such as thiols would not be successful. The assignment of the
transient to phenoxyl radical was further supported by the quenching with ascorbate. The main
point of this paper is the formation of QMs. The most simple and straightforward experiment to
show the existence of QMs is trapping by CH3OH which has been conducted with success. In
addition, the QM transients were quenched with two nucleophiles ethanolamine and NaN3 where
the quenching constants undoubtedly indicated that the assignment was correct. The fact that also
supports this assignment is nonꢀquenching by O2. Transients detected by LFP from all anthrol
derivatives are compiled in Table 4.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CH3
CH3
CH3
O
O
OH
OH
OH
CH3
CH3
CH3
6PhO
QM6
6RC
O
OH
OH
O
OH
7RC
7PhO
QM7
Table 4. Transients detected by LFP.
Transient λmax / nm
Lifetime
2.4 ꢁsa
10 ꢁsb
690
2RC
400ꢀ550
2PhO
28
ACS Paragon Plus Environment

Page 29 of 58
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
400ꢀ500
690
200ꢀ300 msc
2QM
950 ns
3.6 ꢁsa
50 ꢁsb
10 ꢁsa
0.4 sc
3RC
690
3OMeRC
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
300ꢀ500
450
3PhO
3+
500ꢀ600
400ꢀ550
500ꢀ700
400ꢀ500
400ꢀ550
400ꢀ500
360 ns c
10 ꢁsb
2.2 ꢁsa
5 ꢁsd
3QM
4PhO
6RC
6 triplet
6PhO
6QM
10 ꢁsb
3 ms a
0.6 ms b
5.5 ꢁsa
3 ꢁsd
500ꢀ700
7RC
7 triplet
7PhO
7QM
400ꢀ550
400ꢀ500
20ꢀ40 ꢁsb
0.45 ms a
200 ꢁs b
a
b
c
Detected in O2ꢀpurged CH3CN. Detected in O2ꢀpurged CH3CNꢀH2O (1:1). Detected in O2ꢀ
purged HFIP. d Detected in N2ꢀpurged CH3CN (quenched by O2).
29
ACS Paragon Plus Environment