Photocyclization and Photoaddition Reactions of Arylphenols via Intermediate Quinone Methides

Article abstract of DOI:10.1021/acs.joc.5b01580

A series of five benzannelated derivatives of 2-phenylphenol were prepared, and their photochemistry was investigated. Two of these (3-phenyl-2-naphthol, 10, and 1-phenyl-2-naphthol, 11) were photoinert. For 2-(1-naphthyl)phenol (12) and 1-(1-naphthyl)-2-naphthol (13), ESPT took place to either the 2′-position or the 7′-position of the naphthalene ring to give quinone methides (QMs) that underwent either reverse proton transfer (RPT) or electrocyclic ring closure to give dihydrobenzoxanthenes. The intermediate QMs for 12 and 13 were detected and characterized by laser flash photolysis. For 2-(9-phenanthryl)phenol (14), ESPT took place either to the 5′-position to give a QM that underwent quantitative electrocyclic ring closure to give the corresponding benzoxanthene or to the 10′-position to give a QM that underwent RPT. If the solution contained methanol, the QM produced on ESPT to the 10′-position in 14 could be trapped as the photoaddition product. The compounds studied in this work demonstrate three possible reactions of QMs produced following ESPT to aromatic carbon atoms: (1) reverse proton transfer (RPT) to regenerate starting material; (2) addition of hydroxylic solvents to give the photoaddition product; and (3) electrocyclic ring closure to give benzoxanthene derivatives.

Full text of DOI:10.1021/acs.joc.5b01580

Subscriber access provided by RUTGERS UNIVERSITY
Article
Photocyclization and Photoaddition Reactions of
Arylphenols via Intermediate Quinone Methides
Matthew Lukeman, Hilary Simon, Peter Wan, and Yu-Hsuan Wang
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.5b01580 • Publication Date (Web): 23 Oct 2015
Downloaded from http://pubs.acs.org on October 31, 2015
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 41
The Journal of Organic Chemistry
1
2
3
4
5
6
Photocyclization and Photoaddition Reactions of
7
8
9
Arylphenols via Intermediate Quinone Methides1
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
Matthew Lukeman* ,†, Hilary Simon†, Peter Wan,‡ and Yu-Hsuan Wang*‡
† Department of Chemistry, Acadia University, 6 University Ave., Wolfville, NS, B4P 2R6, Canada;
Fax:+1 902 585ꢀ1114; Tel:+1 902 585ꢀ1009; Email: mlukeman@acadiau.ca
‡ Department of Chemistry, University of Victoria, Box 3065 Stn CSC, Victoria BC, V8W3V6, Canada;
Email: lily.wang@pharmaster.ca
Graphical Abstract:
ABSTRACT: A series of five benzannelated derivatives of 2ꢀphenylphenol were prepared and their
photochemistry investigated. Two of these (3ꢀphenylꢀ2ꢀnaphthol, 10, and 1ꢀphenylꢀ2ꢀnaphthol, 11)
were photoinert. For 2ꢀ(1ꢀnaphthyl)phenol (12) and 1ꢀ(1ꢀnaphthyl)ꢀ2ꢀnaphthol (13), ESPT took place to
either the 2’ꢀposition or 7’ꢀposition of the naphthalene ring to give quinone methides (QMs) that either
underwent reverse proton transfer (RPT) or electrocyclic ring closure to give dihydrobenzoxanthenes.
The intermediate QMs for 12 and 13 were detected and characterized by laser flash photolysis. For 2ꢀ
(9ꢀphenanthryl)phenol (14), ESPT took place to either the 5’ꢀposition to give a QM that underwent
quantitative electrocyclic ring closure to give the corresponding benzoxanthene, or to the 10’ꢀposition to
give a QM that underwent RPT. If the solution contained methanol, the QM produced on ESPT to the
10’ꢀposition in 14 could be trapped as the photoaddition product. The compounds studied in this work
ACS Paragon Plus Environment
1

The Journal of Organic Chemistry
Page 2 of 41
demonstrate three possible reactions of QMs produced following ESPT to aromatic carbon atoms: 1)
reverse proton transfer (RPT) to regenerate starting material; 2) addition of hydroxylic solvents to give
the photoaddition product; and 3) electrocyclic ring closure to give benzoxanthene derivatives.
1
2
3
4
5
6
7
8
9
Introduction
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
Excited state intramolecular proton transfer (ESIPT) is an important and fundamental mechanism in
which an acidic proton is transferred to a basic site within the same molecule on photoexcitation.2 In
most cases, the proton is transferred between two heteroatoms, and the reverse proton transfer (RPT)
from the resultant tautomer to regenerate starting material is very fast. Such ESIPT processes are
energyꢀwasting, and have been exploited in the design of sunscreens and photostabilizers. Examples of
ESIPT in which the proton acceptor is a carbon atom are comparatively rare, but are particularly
interesting since these reactions generate quinone methide (QM) intermediates, for which the reverse
proton transfer can be slow enough to allow other reaction pathways to compete.3 In such cases, it is
often possible to isolate new products that may be difficult to access thermally. For example, ESIPT to
spꢀhybridized alkynyl carbons give rise to QMs that can be trapped by primary amines to give ketimine
products.3e
We have investigated examples in which ESIPT takes place from a phenolic OH to carbon atoms of
aromatic rings. For some compounds, including 2ꢀphenylphenol4 and 1ꢀnaphthol,5 no new products are
formed, and the primary evidence for the ESIPT process is the observation of deuterium exchange in
recovered starting material, as shown in Eq. 1 for 2ꢀphenylphenol. The deuterium exchange reaction of
2ꢀphenylphenol was found to occur readily in aprotic solvents and in the solid state, suggesting that the
proton is transferred directly from the OH to the basic carbon atom,4 a mechanism that has been termed
an ‘intrinsic’ ESIPT reaction by Kasha.6 Deuterium exchange quantum yields for 1ꢀnaphthol were
found to strongly depend on the solvent D2O concentration, suggesting that the proton transfer
mechanism is best classified as a type of ESIPT termed by Kasha as a ‘proton relay tautomerization’.6
Such reactions are not truly ‘intramolecular’, however, since solvent molecules are required, so the term
ACS Paragon Plus Environment
2

Page 3 of 41
The Journal of Organic Chemistry
ESIPT is best reserved for those of the ‘intrinsic’ type. We will use the more general term of excited
state proton transfer (ESPT) to characterize reactions that are solventꢀmediated.
1
2
3
4
5
6
7
8
We have shown that some examples of ESPT to aromatic carbon atoms do result in new product
formation. For example, ESPT from the OH in anthracenylphenol 1 to the carbon atom at the 10’ꢀ
position of the anthracene ring gives QM 2, which can be trapped by water and alcohols to give the
isolable photoaddition products 3 (Eq. 2).7
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
O
OD
D
OH
2'
H
rpt
hν
D2O
(1)
ESIPT
2'
D
2-phenylphenol
OH
O
HO
RO
hν
ROH
H
H
10'
(2)
2
3
1
ESPT take places with good efficiency (Φ = 0.15) in 1,1’ꢀbiꢀ2ꢀnaphthol (BINOL) from one of the
hydroxyl groups to one of three different carbon atoms on the other ring system to give three QMs 4 –
6.8 Two of these (4 and 5) undergo RPT exclusively to regenerate starting material (labeled with
deuterium when D2O is present in the solvent mixture), while the third QM (6) undergoes electrocyclic
ring closure to give dihydrobenzoxanthene 7 (Scheme 1). The ESPTꢀRPT reaction sequence causes
racemization of enantiomerically pure samples of BINOL because of the planarity of the intermediate
QMs 4 and 5. While this discovery has important implications for catalysis, we were especially
interested in the formation of 7, since this represents a novel reaction pathway for photogenerated QMs.
The formation of 7 from BINOL, however, is very inefficient (Φ < 0.01), and represents only a minor
photochemical pathway.
ACS Paragon Plus Environment
3

The Journal of Organic Chemistry
Page 4 of 41
H
H
1
2
3
4
5
H
O
H
O
O
H
+
+
H
6
HO
HO
HO
7
8
9
4
5
6
(Φ > 0.06)
hν
OH
(Φ > 0.11)
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
H
H
O
HO
HO
BINOL
7
(Φ < 0.01)
Scheme 1
In an effort to gain a more comprehensive understanding of ESPT to aromatic carbon atoms, we
embarked on a study of the photochemistry of all possible monoꢀbenzannelated derivatives of 2ꢀ
phenylphenol. We found that 8 undergoes ESIPT to the 2’ꢀcarbon of the attached phenyl ring, resulting
in deuterium exchange when the solvent contains D2O (Eq. 3), behaviour that mirrors that of 2ꢀ
phenylphenol (Eq. 1).10 The quantum yield for deuterium incorporation for 8 was exceptionally high (Φ
= 0.73), demonstrating that at least for some phenols, ESIPT to aromatic carbon atoms is the dominant
photochemical process. For 9, we observed deuterium incorporation at the 1ꢀ and 3ꢀpositions of the
naphthalene unit with a total quantum efficiency of Φ = 0.30 (Eq. 4), demonstrating that ESPT is a
major photochemical pathway for this compound as well.11
O
OH
OH
D
H
rpt
hν
D2O
(3)
D
8
ACS Paragon Plus Environment
4

Page 5 of 41
The Journal of Organic Chemistry
OH
OH
O
D
1
2
3
4
5
6
7
8
H
rpt
OH
D
hν
D2O
(4)
+
+
O
D
H
rpt
9
9
10
11
12
13
14
15
D
In the present work, we report our detailed study of the photochemistry of the remaining derivatives in
the series of monobenzannelated derivatives of 2ꢀphenylphenol, 10 – 12. In addition, we report on the
photochemistry of closely related diꢀbenzannelated derivatives 13 and 14. We demonstrate that three of
these derivatives (12 ꢀ 14) undergo efficient ESPT processes, providing new examples of deuterium
exchange, photoaddition, and photocyclization reactions.
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
Results and Discussion
Synthesis
Preparation of derivatives 10 - 14 was achieved in acceptable yield (19 – 84% over two steps) by
the Suzuki coupling of the appropriate methoxyaryl boronic acid and aryl bromide to give methyl ethers
15 ꢀ 19, followed by demethylation using BBr3 (Eq. 5, Table S1). Aryl bromides and aryl boronic acids
were commercially available except for 3ꢀbromoꢀ2ꢀnaphthol, which was prepared according to the
ACS Paragon Plus Environment
5

The Journal of Organic Chemistry
Page 6 of 41
method of Niimi et al.12 This synthetic scheme was attractive to us since it also provided methyl ethers
15 – 19, whose photochemistry was also studied and compared with that of 10 – 14.
1
2
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
Photochemical experiments
The photochemistry of 10 – 14 was initially followed by UVꢀVis spectroscopy, with the expectation
that photoaddition or photocyclization reactions might lead to changes in the absorption properties of
the chromophores. Irradiation of derivatives 12 ꢀ 14 in deoxygenated aqueous acetonitrile (10%ꢀ50%
water, v/v) did indeed lead to dramatic changes in the absorption spectra, and sharp isosbestic points
were observed, indicating that the photoproducts are not undergoing secondary photochemistry under
these conditions. Absorption spectra for 14 are shown in Figure 1; isosbestic points are visible at 236,
260, 295, and 302 nm. A summary of absorption changes for 12 ꢀ 14 can be found in table S3 in the
supporting information.
50 min
40 min
30 min
25 min
20 min
15 min
10 min
5 min
2 min
0 min
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
X 4
200
250
300
350
400
450
Wavelength (nm)
ACS Paragon Plus Environment
6

Page 7 of 41
The Journal of Organic Chemistry
Figure 1. UVꢀVis spectra of 14 in 50% CH3CNꢀH2O (v/v) before and after irradiation in a Rayonet RPR
100 photoreactor at 300 nm (16 lamps). Irradiation times are indicated in the legend, and the arrows
indicate whether the absorbance is increasing or decreasing in that wavelength range.
1
2
3
4
5
6
7
8
9
There is a significant increase in absorbance in the 300 – 400 nm range, indicating that the
photoproduct likely possesses more extended conjugation than 14, as might be expected from a
photocyclization reaction. It required much longer irradiation periods to bring about changes in the
absorption spectra of 14 that were of a magnitude comparable to those for 12 and 13, suggesting that the
photoreaction leading to the absorption changes for 14 is much less efficient than for 12 or 13. No
absorption changes were observed for solutions of phenols 10 or 11, or methyl ethers of 12 – 14 (i.e. 17
– 19) on irradiation.
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 order to identify the photoproducts formed from irradiation of 12 – 14, preparatory scale photolyses
were carried out in aqueous acetonitrile, and the photolysate was analyzed by TLC and NMR
spectroscopy. TLC analysis of the photolysate produced on irradiation of a solution of 12
(deoxygenated 1:9 (v/v) H2OꢀCH3CN) indicated that along with unreacted starting material, a single
photoproduct had formed with a significantly higher Rf value than the starting phenol, indicating that
the photoproduct is less polar than the starting material, likely lacking the phenolic OH. The
photoproduct was isolated by preparatory TLC, and characterized by NMR spectroscopy. The NMR
data was consistent with identification of this compound as dihydrobenzoxanthene 20 (Eq. 6).
Treatment of 20 with Pd supported on carbon in oxygenated toluene resulted in >90% conversion to
benzo[k,l]xanthene (Eq. 6). A similar preparatory scale photolysis was carried out for 13 under the
same conditions, and compound 21 was identified as the only photoproduct formed after it was isolated
from the photolysate (Eq. 7). Both 20 and 21 are analogous in structure to 7 that is formed on
irradiation of BINOL, indicating that this cyclization reaction might be general for hydroxybiaryls
containing 1ꢀnaphthyl rings as protonꢀacceptors.
ACS Paragon Plus Environment
7

The Journal of Organic Chemistry
Page 8 of 41
H
H
1
2
3
O
O
4
5
6
Pd/C, O2
hν
(6)
12
ꢀ
7
8
20
benzo[k,l]xanthene
9
H
H
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
hν
13
(7)
21
The photochemistry of 14 was investigated by carrying out preparatoryꢀscale reactions in a variety of
solvents. In aqueous acetonitrile, we envisioned two possible reactions: 1) photoaddition of water
across the 9,10 phenanthrene double bond to give 22, in analogy to the photochemistry of oꢀ
hydroxystyrene;3b and 2) photocyclization to give 23, in analogy with the photochemistry of BINOL, 12
and 13 (Scheme 2).
H
H
H
OH
HO
O
hν
hν
14
H2O
23
22
Scheme 2
Irradiation of 14 (deoxygenated 1:1 (v/v) H2OꢀCH3CN) gave rise to a single photoproduct after
chromatographic isolation from the photolysate that was identified as benzoxanthene 24 based on NMR
and MS characterization. This product is similar in structure to 7, 20, and 21, but represents a 2ꢀelectron
oxidation (loss of H2). We thought it likely that the first formed photoproduct was in actuality a
dihydrobenzoxanthene such as 23 or 25, which underwent oxidation to 24 during the workꢀup and
purification (Eq. 8). The photoreaction was repeated under the same conditions, but with the
chromatography step omitted. NMR analysis indicated that the crude reaction mixture was primarily a
mixture of starting material (14) and 24, although in addition to peaks arising from these, traces of a
third product were detected that are consistent with expectations for 25, supporting the assertion that 24
ACS Paragon Plus Environment
8

Page 9 of 41
The Journal of Organic Chemistry
is formed from 25 after the initial photocyclization. It is not readily apparent why 25 is so much more
oxygen sensitive than 20 and 21, but perhaps the fact that the alkene group in 20 and 21 is conjugated
with the phenyl ring imparts greater stability to these compounds. No evidence for the production of the
hydrated photoproduct 22 was found in the photolysate NMR spectra.
1
2
3
4
5
6
7
8
9
H
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
H
O
O
H
hν
(8)
14
25
24
-
O
OH
O
hν
(9)
14
+
CH3OH
H
H
MeO
H
H
27
26
The photoreaction of 14 was repeated using various mixtures of water, methanol, and acetonitrile
(Table S3, supporting information). When 14 was irradiated in neat methanol for 1h, the sole product
obtained was methyl ether 26 in 8% yield. This product likely arises from the initial formation of
quinone methide 27, followed by nucleophilic attack by methanol (Eq. 9). The analogous hydration
product 22 was not observed in photoreactions conducted in aqueous acetonitrile, suggesting that 22 (if
formed) is less stable than 26, and dehydrates back to starting material. When 14 was irradiated in 1:4
waterꢀmethanol, the reaction to produce 26 became more efficient by a factor of approximately 6ꢀfold,
producing 26 in 51% yield. The enhancement in reaction efficiency observed when water is included in
the solvent mixture is suggestive of a reaction mechanism involving a ESPT from the phenolic OH to
the 10ꢀcarbon of the phenanthroline ring system to give QM 27. In order to investigate the effect of
excitation wavelength on the product yields, photolyses of 14 in 20% aqueous methanol were carried
out using either UVA bulbs (λmax = 350 nm) or UVB bulbs (λmax = 300 nm) for 30 minutes. When
UVA bulbs were used, 26 was obtained in 32% yield; however, the yield dropped to only 2% when the
UVB bulbs were used. When UVB excitation is used, it is likely that the photoproduct 26 is absorbing
the incoming light and undergoes photoelimination, reverting back to 14. Since only the photoaddition
ACS Paragon Plus Environment
9

The Journal of Organic Chemistry
Page 10 of 41
product 26 was formed in aqueous methanol and only photocyclization product 24 was formed in
aqueous acetonitrile, we decided to explore the photochemistry of 14 in aqueous methanolꢀacetonitrile
mixtures in an attempt to promote a competition between the photoaddition and photocyclization
pathways. Irradiation of 14 in a 2:1:5 mixture (v/v) of waterꢀmethanolꢀacetonitrile for 45 min led to
formation of primarily photoaddition product 26 (13% yield) along with a smaller amount of 24 (<1%
yield). When repeated for a much longer irradiation period (18 h), 26 and 24 were formed in 70% and
4% yields, respectively.
1
2
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
In order to determine the regiochemistry of the ESPT processes that operate for 12 ꢀ 14, a series of
photolyses was carried out in which the aqueous portion of the solvent mixture was replaced with D2O.
1
Photoproducts and recovered starting material were then monitored for deuterium incorporation (by H
NMR and MS) in order to identify the carbon atoms that act as proton acceptors. Irradiation of 12 (~10ꢀ
3
M, 1:1 D2OꢀCH3CN, argon purged, 1 h) gave a mixture of 12 and 20, along with a small amount of
other unidentified photoproducts that were not formed when the solvent mixture was 1:9 H2OꢀCH3CN.
12 and 20 present in the photolysate were separated by preparatory TLC and analyzed separately by 1H
NMR spectroscopy. The sample of recovered 12 thus obtained showed a 45% decrease in the integrated
area of the peak assigned to the proton at the 2’ꢀposition of the naphthalene ring at 7.90 ppm, indicating
a 55:45 ratio of 12:12ꢀ2’D (Eq. 10).
D
6'
7'
H
8'
OH
5'
4'
O
hν
+
12
1'
(10)
D2O-CH3CN
3'
D 2'
20-7'D
12-2'D
ACS Paragon Plus Environment
10

Page 11 of 41
The Journal of Organic Chemistry
The decrease in area is accompanied by the transformation of the peak at 7.54 ppm (assigned to the
adjacent 3’ꢀposition) from a wellꢀdefined doublet of doublets to a broad doublet, with no loss of total
peak area (see supporting information). The change in this peak is consistent with the reduced coupling
that would occur on exchange of the adjacent hydrogen with deuterium. Photoproduct 20 isolated from
the same run showed exactly 50% deuterium exchange at the methylene (7’ꢀ) position (to give 20ꢀ7’D)
(Eq. 10), as evidenced by the 50% reduction in peak area at δ3.43 ppm in the 1H NMR spectrum (Figure
2), suggesting that one of the methylene hydrogen atoms in 20 comes from either the phenol OH (or
OD) or solvent H2O (D2O). The methylene signal for 20ꢀ7’D actually consists of two peaks, one
centered at 3.40 ppm, and the second centered at 3.44 ppm, and these peaks are present in a 3:1 ratio (by
area). These separate signals might arise from the two possible diastereomers of 20ꢀ7’D – one in which
the methylene deuterium is cis to the ether oxygen (cisꢀ20ꢀ7’D), and one in which the methylene
deuterium is trans to the ether oxygen (transꢀ20ꢀ7’D).
1
2
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
D
6'
7'
H
8'
OH
5'
4'
O
hν
+
13
1'
(11)
D2O-CH3CN
3'
D 2'
13-2'D
21-7'D
ACS Paragon Plus Environment
11

The Journal of Organic Chemistry
Page 12 of 41
1
2
3
4
5
6
7
8
3.5
3.4
3.3
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
2.04
1.00
1.02
6.0
5.5
5.0
4.5
4.0
3.3
3.5
3.0
3.5
3.4
1.96
0.93
2.10
6.0
5.5
5.0
4.5
4.0
3.5
3.0
1
Figure 2. 500 MHz H NMR spectra of 20 produced by irradiation of 12 in 1:9 H2OꢀCH3CN (bottom)
and in 1:9 D2OꢀCH3CN (top) solutions, showing the 50% decrease in peak integration of the signal at
3.4 ppm.
Similar photolyses of 13 in solutions in which the aqueous portion was replaced with D2O led to
formation of 21ꢀ7’D (Eq. 11). Exactly one deuterium atom is incorporated at the methylene position in
21ꢀ7’D, whether the reaction is taken to low or high conversion, ruling out a mechanism in which the
deuterium exchange occurs in a separate process from formation of 21. The methylene signal in this
case appears as only one peak, suggesting that only a single diastereomer is formed (Figure S1,
supporting information). Starting material isolated from these photolyses showed deuterium
incorporation at the 2'ꢀposition, the extent of which rose steadily as the reaction was taken to higher
conversion; irradiation for 10 or 30 min led to a 60 or 85% decrease in the integrated area of the peak at
8.02 ppm that is assigned to the aromatic hydrogen at the 2’ꢀposition, and a 5 or 10% decrease in the
integrated area of the peak at 7.97 ppm that is assigned to the aromatic hydrogen at the 8’ꢀposition.
ACS Paragon Plus Environment
12

Page 13 of 41
The Journal of Organic Chemistry
Accompanying these changes is the conversion of the doublet of doublets at 7.65 ppm assigned to the
3’ꢀposition to a broad doublet, due to diminished coupling with deuterium at the 2’ꢀposition.
Irradiation of solutions of 10 and 11 containing D2O for extended periods of time (2 ꢀ 20 hours)
yielded only starting material with no deuterium incorporation, indicating that ESPT processes to
aromatic carbons are not important pathways for these phenols. The methyl ether analogues (17 – 19)
also did not show deuterium incorporation when irradiated in D2O solution, implying that the OH
moiety in 12 and 13 is required for their photochemistry.
1
2
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
The photochemistry of 14 was also explored in acetonitrile solutions containing D2O. Irradiation of
14 (~10ꢀ3 M, 1:1 D2OꢀCH3CN, argon purged, 3 h), produced a mixture of 24 and 14. The products were
1
chromatographically separated and analyzed by H NMR spectroscopy to assess for deuterium
incorporation. The recovered starting material showed an 85% reduction in the peak area of the singlet
at 7.69 ppm assigned to the hydrogen at the 10’ꢀposition on the phenanthrene ring system (Figure 3),
indicating that photolysis led to extensive conversion to 14-10’D (Scheme 3). The likely reason for the
deuterium exchange is initial ESIPT from the phenolic OD to the 10’ꢀcarbon to give 27-10’D, which
then either undergoes RPT to give 14-10’D directly, or is first hydrated by the solvent to give 22-10’D,
which subsequently dehydrates to give 14-10’D (Scheme 3). The cyclized photoproduct 24 isolated
from the same run showed significant deuterium incorporation in two positions. The peak area of the
singlet at 7.71 ppm assigned to the 10’ꢀhydrogen was reduced by approximately 65%, indicating
formation of 24ꢀ10’D, which probably arises from cyclization of 14ꢀ10’D that was previously formed
by the process depicted in Scheme 3.
ACS Paragon Plus Environment
13

The Journal of Organic Chemistry
Page 14 of 41
H
H
H
D
1
2
3
4
5
6
7
8
O
O
O
D
hν
+
14
D2O
H
D
25-5'D
28-5'D
27-10'D
D2O
[Ox]
7'
6'
rpt
9
8'
OD
DO
5'
OH
O
D
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'
2'
9'
3'
D 10'
D
1'
22-10'D
24-5'D
14-10'D
Scheme 3
CHCl3
2,4
3’, 6’, 8’
10’
5
7’
2’
5’
4’
1’
3
1
Figure 3. 500 MHz H NMR spectra of 14 before (bottom) and after (top) irradiation in 1:1 D2Oꢀ
CH3CN solution, showing ~85% decrease in the singlet at 7.78 ppm, assigned to the 10’ꢀhydrogen.
Relative integrations are given below the curve.
In addition, the peak area of the doublet at 8.10 ppm assigned to the hydrogen on the 5’ꢀcarbon was
reduced by 58%, consistent indicating formation of 24ꢀ5’D, suggesting that an ESPT from the phenolic
OD of 14 takes place to the 5’ꢀcarbon to form QM 28-5’D, which then quantitatively cyclizes to 25ꢀ5’D,
ACS Paragon Plus Environment
14

Page 15 of 41
The Journal of Organic Chemistry
which gives 24-5’D after air oxidation (Scheme 3). The quantitative nature of the cyclization is
suggested by the fact that no deuterium exchange was observed at the 5’ꢀposition in recovered starting
material. Because QM 28ꢀ5’D and 25-5’D should both contain exactly one deuterium and one hydrogen
atom on the methylene carbon, the observed deuterium exchange of 58% at the 5’ꢀposition of 24-5’D
indicates that the oxidation step exhibits a kinetic isotope effect of kH/kD = 58%/42%, or 1.4. In
addition to 24-5’D and 24-10’D, formation of dideuterated 24-5’,10’D is also expected in higher
conversion runs, although the dideuterated product is indistinguishable from the monodeuterated
products by 1H NMR spectroscopy.
1
2
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
A series of photolyses of 12ꢀ14 were carried out in which the concentration of D2O (in CH3CN) was
systematically altered so that the role of water in the photoreaction could be better understood. For 12
and 13, the efficiency of product formation is highly dependent on the concentration of D2O present in
solution.1 ESPT reaction efficiency is minimal at low water concentrations, but rises steadily as more
D2O is added, reaching a maximum at 5 M D2O (10 % (v/v)). The water dependence indicates that the
reactions are not examples of intrinsic ESIPT, but are best explained as arising from a waterꢀmediated
ESPT. A similar water (D2O) dependence study was carried out for the ESIPT reaction of 14 to form
14-10’D. Solutions of 14 containing varying amounts of D2O in CH3CN were irradiated, and the
photolysate was analyzed by 1H NMR spectroscopy (Figure 4).
ACS Paragon Plus Environment
15

The Journal of Organic Chemistry
Page 16 of 41
1
2
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
Figure 4. Plot showing % yield of 14ꢀ10’D formed on irradiation of 14 solutions in CH3CN containing
varying concentrations of D2O. The inset is an expansion of the 0 – 1 M region of the plot.
The yield of 14ꢀ10’D is only weakly dependent on D2O concentration in the 5 M to 20 M range, but
increases dramatically as the D2O concentration is reduced below 5 M, so that at 0.5 M D2O, the yield
of 14ꢀ10’D is twice as high as when the D2O concentration is higher than 5 M. The ESIPT reaction
yield remains high until very small water concentrations are reached (< 0.05 M), and at such low
concentrations, it is difficult to ensure isotopic purity due to contamination from trace amounts of H2O
present in the CH3CN and/or on the glassware. The ESPT reaction for 14 clearly does not require
significant quantities of water (or D2O), suggesting that the reaction is an intrinsic ESIPT. Intrinsic
ESIPT reactions typically require that the acidic OH and basic carbon atom be close to each other in
space and share a ground state hydrogenꢀbonding interaction, which is certainly possible for 14 and is
consistent with results we recently reported for the related 9’ꢀ(2,5ꢀdihydroxyphenyl)phenanthrene.9a
The rise in efficiency on moving from 5 M D2O to 0.5 M D2O might be the result of a competing ESPT
reaction from the phenol OH of 14 to solvent water, which increases in efficiency as the water
ACS Paragon Plus Environment
16

Page 17 of 41
The Journal of Organic Chemistry
concentration in the solvent increases. At low water concentrations, this pathway becomes less
important, allowing for maximal ESIPT efficiency. Also, water (or D2O) present in the solvent would
be expected to hydrogenꢀbond with the phenolic OH, thereby potentially interfering with any
intramolecular hydrogenꢀbonding interaction with the πꢀsystem of the phenanthrene ring.13
1
2
3
4
5
6
7
8
9
Quantum yields for both photocyclization and deuterium incorporation (in 10% D2OꢀCH3CN) were
determined for 12ꢀ14, using the deuterium exchange of 2ꢀphenylphenol in the same solvent (Φ = 0.042)
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
as a secondary reference standard.4 The quantum yields appear in Table 1, where ΦD refers to the
quantum yield for deuterium incorporation in starting material, Φcyc refers to the quantum yield for
formation of the cyclized benzoxanthenes 20, 21, and 24. The quantum yield for the photoaddition of
methanol to 14 was not determined, however, this addition and the deuterium exchange reaction at the
10’ꢀposition are assumed to share a common intermediate 27, and thus these processes should have the
same quantum yield for the primary photochemical step (ESPT).
Table 1: Quantum yields for the formation of deuterated starting material (ΦD) and cyclized
photoproduct (Φcyc) for 10 – 14 D2OꢀCH3CN mixtures. Estimated errors are +/ꢀ 10% of quoted value.
Compound
ΦD
0a
Φcyc
0a
10
11
12
13
14
0a
0a
0.14a 0.20a
0.11a 0.11a
0.18b 0.005b
0.36c <0.01c
a
b
c
in 1:9 (v/v) D2OꢀCH3CN solution.; in 1:1 (v/v) D2OꢀCH3CN solution; in 1:99 (v/v) D2OꢀCH3CN
solution.
ACS Paragon Plus Environment
17

The Journal of Organic Chemistry
Page 18 of 41
Fluorescence Measurements
1
2
3
4
5
6
7
8
9
Excitation of 11 (λex = 334 nm) in neat CH3CN gives strong fluorescence emission with a maximum
at 353 nm (Table 2). The Stokes shift is small (19 nm), suggesting that two aryl rings do not twist
significantly towards planarity on excitation. Both the absorption and emission maxima of 11ꢀ (formed
by dissolving 11 in a 1:1 mixture of 0.01 M NaOH and CH3CN) are redꢀshifted relative to 11 (Table 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
Table 2: Photophysical data for 10 – 14. Estimated errors are +/ꢀ 10% of quoted value.
kF (sꢀ1)
Compound
λex (nm)
λem
τS (ns) ΦF
334a
368b
293a
ꢀ
353a
445b
355a
540c
356a
550b
8.5a
ꢀ
0.19 2.2 x 107
11
11-
12
3.0a
ꢀ
12-
13
334a
370b
296
3.9a
ꢀ
13-
14
353, 370, 390 (sh) 27.5a 0.22a 8.0 x 106
17.4c 0.10c 5.7 x 106
b
c
a in neat CH3CN. in 1:1 waterꢀCH3CN, pH 12. in 4:1 waterꢀCH3CN.
Excitation of 12 (λex = 293 nm) in neat CH3CN (argon purged) gave strong emission at 355 nm that
exhibited a much larger Stokes shift (62 nm) than observed for 11, indicating that significant twisting
about the biaryl bond towards planarity occurs on excitation to S1. If the excited state of 12* is nearly
planar, this should allow for much stronger charge transfer from the phenolic ring to the naphthyl ring in
the excited state due to the better πꢀπ overlap between the rings, which is thought to be a requirement
for efficient ESIPT reactions in biaryls. Similarly, the lack of planarity of 11*, as suggested by the
small Stokes shift, may explain the fact that this compound does not undergo ESIPT reactions. For 12,
ACS Paragon Plus Environment
18

Page 19 of 41
The Journal of Organic Chemistry
the fluorescence emission intensity was very sensitive to the amount of water present in the solution.
Addition of small amounts (~1 M H2O) to the acetonitrile solvent caused an initial redꢀshifting of the
emission band to 360 nm, along with a modest increase in intensity. These effects are associated with
solvation by water. Addition of larger amounts of water led to a dramatic quenching of the fluorescence
emission until ~10 M water was reached, at which point addition of more water did not lead to
significant further quenching. The quenching of the fluorescence emission of 12 with added water
correlates well with the increase in ESPT reaction efficiency on addition of water. We propose that the
fluorescence quenching with added water (up to 10%) is a result of the increasing competition from
ESPT of 12 as water is added. This correlation is strong evidence for reaction of 12 from S1 and not T1.
The further fluorescence quenching (from 10% up to 40% water content) is likely a result of increasing
competition from ESPT to the bulk solvent. Addition of larger amounts of water led to the growth of a
new weak emission band at 540 nm which we assign to fluorescence from the phenolate 12- that would
form on ESPT to the solvent water. Addition of greater amounts of water (from 40% up to 100%, data
not shown), led to a redꢀshifting of the fluorescence emission maximum from 360 nm to 386 nm, an
effect that may simply result from changes of the overall polarity of the solvent.
1
2
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
Excitation of 13 in neat CH3CN (λex = 334 nm) gave strong, broad, and structureless fluorescence
emission at 356 nm (Figure S3, supporting information). The Stokes shift for 13 (22 nm) is smaller than
that of 12, indicating that rotation towards planarity is not achieved to the same extent for the former,
likely as a consequence of the additional steric bulk present in 13. Excitation of 13- (formed by
dissolving 13 in a 1:1 mixture of 0.01 M NaOH and CH3CN) at λ 370 nm gave weak fluorescence
emission at λ 550 nm. The position of this band is similar to that of 12-, and the Stokes shift is very
large (180 nm), suggesting that twisting to planarity is much more facile for 13ꢀ than for 13. Addition
of small amounts of water led to very efficient quenching of the 356 nm fluorescence band for 13, with
over 80% quenching observed at less than 3 M H2O (Figure S3). A SternꢀVolmer quenching plot was
constructed and appears as the inset in Figure S3. The quenching data is nonlinear, and fit best to I0/I =
1 + KSV[H2O]1.7. We attribute this quenching to competition from the ESPT process that is enabled by
ACS Paragon Plus Environment
19

The Journal of Organic Chemistry
Page 20 of 41
the added water. Along with the quenching of the 356 nm band, a very weak band is observed to grow
in at ~550 nm, corresponding to emission from the naphtholate, indicating that adiabatic ESPT from the
hydroxyl group to solvent (water) to form 13-* can take place at high water concentrations. There may
be a contribution to the observed quenching due to a simple solvent polarity change as well, but such a
contribution is expected to be small at water concentrations less than 3 M, since changes due to solvent
polarity were not seen for 12 until concentrations of over 20 M were reached.
1
2
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
Excitation of 14 in neat CH3CN (λex = 300 nm) resulted in structured fluorescence emission with
maxima at 353 and 370 nm. Addition of water to the acetonitrile solvent lead to an increase in
fluorescence emission intensity from 14, and when a water concentration of 27.8 M was reached, the
fluorescence emission intensity had more than doubled. Addition of more water beyond 27.8 M
reduced the emission intensity, and at water concentrations greater than 35 M H2O, the emission
intensity was once again lower than in pure acetonitrile. The product studies revealed that the ESIPT
reactions of 14 were most efficient at very low water concentrations (<5 M D2O). Thus the low
fluorescence emission in acetonitrile solutions containing little or no water can be explained by the fact
that the ESIPT processes that compete with fluorescence at these solvent compositions are maximally
efficient. Fluorescence lifetimes (τF) were measured in neat CH3CN and in 4:1 H2OꢀCH3CN solutions,
and were found to be 27.5 ns and 17.4 ns, respectively (Table 5). The fluorescence quantum yields (ΦF)
were also measured in neat CH3CN and in 4:1 H2OꢀCH3CN solutions, and were found to be 0.22 and
0.10, respectively. The reduction in fluorescence quantum yields and fluorescence lifetimes in aqueous
acetonitrile indicates that dynamic quenching of S1 by water takes place, which we interpret as being
due to ESPT to solvent. Using these values, the fluorescence rate constants (kf) were determined using
Equation 12, and appear in Table 2.
kF = ΦF × (1/τF) (12)
Laser Flash Photolysis
ACS Paragon Plus Environment
20

Page 21 of 41
The Journal of Organic Chemistry
Nanosecond laser flash photolysis (LFP) was employed in an attempt to characterize the transient
species formed on ESIPT of 12 – 14, with the primary aim of detecting QM intermediates that would
form on ESIPT to an aromatic carbon position. Related oꢀnaphthoquinone methides (NQM) have been
reported by the Popik14a,b and Freccero14c,d groups, although these are formed from photoelimination of
benzylic alcohols or quarternary ammonium salts, rather than ESIPT to carbon atoms. Several of these
NQMs were detected by LFP and showed an absorption band at around 360 nm. However, these NQMs
are confined to a single naphthalene unit, and are not biaryl quinone methides like those expected from
12 – 14.
1
2
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
LFP of 12 in neat CH3CN (N2 purged) gave a weak transient that absorbed between 320 and 550 nm
with a maximum at 440 nm and a lifetime of 800 ns. Saturation of the solution with oxygen led to
quenching of the transient, leading us to assign this transient to the triplet state of 12. LFP of 12 in 1:9
H2OꢀCH3CN (N2 or O2 purged) gave a number of transients (Figure 5). A band was observed at 490 nm
with a lifetime of 800 ns that was quenched on saturation of the solution with oxygen. This transient is
similar to that observed in neat CH3CN and is assigned to the triplet state. In addition to the triplet
signal, a second transient absorption was observed with maxima at 360 nm and 560 nm. Both maxima
showed a decay that fit best to a double exponential function with lifetimes of 3 and 400 ꢁs, indicating
the presence of two species. Neither lifetime was reduced by the presence of oxygen, consistent with
the assignment of these signals to QM intermediates. ESPT in 12 from the OH to the 2’ꢀposition or 7’ꢀ
position would give 28 and 29, respectively (Eq. 13).
Both of these QM intermediates can conceivably be generated in either of two isomeric forms, 28a and
28b, and 29a and 29b. Isomer 29a does not undergo reverse proton transfer, as evidenced by the lack of
deuterium incorporation at the 7’ꢀposition, so electrocyclic ring closuse must be much faster than
ACS Paragon Plus Environment
21

The Journal of Organic Chemistry
Page 22 of 41
reverse proton transfer. For this reason, we expect 29a to be shortꢀlived. Isomer 29b is probably not
formed at all, since it cannot cyclize, and we so no isotope exchange following irradiation in D2O. We
therefore provisionally assign the two transients observed on LFP of 12 in aqueous solution to be due to
two of QMs 28a, 28b, and 29a. We do not feel that more definitive assignments can be made with the
information at hand.
1
2
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
Figure 5 – Transient absorption spectra observed on LFP of 12 in 1:9 CH3CN (black diamonds) and in
neat CH3CN (white squares) (O2 purged, λex = 308 nm). Absorption data were taken immediately after
the laser pulse.
H
H
O
O
H
H
+
+
+
(14)
ESIPT
13
O
O
H
H
H
H
30a
30b
31a
31b
ACS Paragon Plus Environment
22

Page 23 of 41
The Journal of Organic Chemistry
In analogy to the photochemistry of 12, ESPT to the 2’ꢀ or 7’ꢀposition of 13 would give QMs 30a,
30b, 31a, and 31b (Eq. 14). As was also the case for 12, starting material that was isolated from
irradiations of 13 in D2O solution showed no measurable deuterium incorporation at the 7’ꢀposition,
indicating that QM 31a exclusively undergoes ring closure, and that QM 31b is not formed.
Electrocyclic ring closure for 30a is not possible, and is not favourable for 30b because this process
would reduce the aromaticity of the system. Instead, QMs 30a and 30b undergo reverse proton transfer
(tautomerization). LFP of 13 in 1:9 H2OꢀCH3CN solution (oxygen purged) gave rise to a transient
spectrum that showed strong absorption across the entire visible region that tailed beyond 800 nm
(Figure 6). A biphasic decay was observed at all wavelengths, with a longꢀlived (τ ~ 580 ꢁs) and a
1
2
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
shorterꢀlived (τ ~ 120 ꢁs) component, indicating the presence of two decaying species. Both of these
lifetimes are rather long, and neither can be assigned to 31a, since it is unlikely that 31a would persist
for 120 ꢁs and still undergo quantitative cyclization with no RPT. On this basis, we provisionally
assign the two transients to 30a and 30b.
LFP was also carried out on 14 in hopes of detecting QMs 27 or 28. LFP of 14 in neat CH3CN
(nitrogen purged) gave a transient absorption with λmax = 430 nm and a lifetime of 4.9 ꢁs. When the
solution was purged with oxygen, the lifetime of the transient was reduced to 56 ns, consistent with this
transient being the tripletꢀtriplet absorption of 14. Since the ESIPT (to the 10’ꢀposition) has been shown
to be maximally efficient in CH3CN containing little to no water (Figure 4), the absence of additional
observable signals might indicate that QM 25 (cis or trans) has a lifetime shorter than 20 ns, the timeꢀ
resolution of the instrument. The analogous QM produced by ESIPT in 2ꢀphenylphenol was also
undetectable by nanosecond LFP, although it has been recently detected in femtosecond transient
absorption experiments.15 LFP experiments were also conducted on 14 in a 1:1 mixture of water and
acetonitrile under both nitrogen and oxygenꢀsaturated conditions, however, no signals in addition to the
triplet were observed under these conditions.
ACS Paragon Plus Environment
23

The Journal of Organic Chemistry
Page 24 of 41
1
2
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
Figure 6 – Transient absorption spectra observed on LFP of 13 in 1:9 CH3CN (O2 purged, λex = 308
nm). Absorption data were taken immediately after the laser pulse. Inset: Biphasic decay of the
absorption at 550 nm observed on LFP of 13 in 1:9 CH3CN (O2 purged, λex = 308 nm). Line drawn is
least squares fit to the equation ꢀOD = A + Beꢀ(k1)t + Ceꢀ(k2)t, and gives τ1 = 580 ꢁs and τ2 = 120 ꢁs.
Mechanism
ACS Paragon Plus Environment
24

Page 25 of 41
The Journal of Organic Chemistry
L
L
O
H
1
2
3
H
OH
O
4
5
6
7
8
L
12 (L = H)
28a (L = H)
28a-7'D (L = D)
12-2'D (L = D)
20 (L=H)
20-7'D (L=D)
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
L
ESPT (2')
ESPT (2')
H
O
O
ESPT (7')
ESPT
12*
H
L
hν
29a (L = H)
29a-7'D (L = D)
28b (L = H)
28b-7'D (L = D)
O-
OL
Scheme 4
12 (L = H)
12-OD (L = D)
12-
The photochemical pathways available to 12 are summarized in Scheme 4. Following excitation to
S1, ESPT can proceed from the phenol OH to the 7’ꢀposition of the naphthalene ring to give QM 29a,
provided a sufficient amount of water is present to mediate the proton transfer (optimally ~5 M). QM
29a quantitatively undergoes 6πꢀelectrocyclic ring closure to give dihydrobenzoxanthene 20, and does
so at a rate that is sufficiently fast so that RPT cannot compete. When the reaction is carried out in D2O
1
solution, excited state deuterium transfer (ESDT) instead gives 29aꢀ7’D, which cyclizes to 20ꢀ7’D. H
NMR spectroscopy of 20ꢀ7’D indicates the formation of a 3:1 mixture of cis and trans isomers of 20ꢀ
7’D. If 29aꢀ7’D is formed as a discrete intermediate prior to cyclization, a cis/trans ratio close to 1:1
would be expected, since 29a is planar and should cyclize in either direction with equivalent likelihood,
excepting any isotope effect imparted by the presence of deuterium, which would be only secondary.
One way to explain the 3:1 isomer ratio would be to invoke a mechanism in which proton transfer
occurs concurrently with rotation of the rings towards planarity, and cyclization takes place before a
planar QM can be formed, as depicted in Scheme 5. If 20ꢀ7’D is formed via this concerted mechanism
ACS Paragon Plus Environment
25

The Journal of Organic Chemistry
Page 26 of 41
50% of the time, and the other 50% of the time, the rotation to 0° takes place to give 29aꢀ7’D, one
would expect a product ratio of approximately 3:1, as observed. ESPT can also take place to the 2’ꢀ
position to give QM 28a and/or 28b, both of which quantitatively undergo reverse proton transfer to
regenerate starting material (or deuterium labeled starting material 12ꢀ2’D when carried out in D2O
solution).
1
2
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
DO
O
hν
D
H
O
D
O
H
D
12-OD
H
H
20-7'D
concerted
Scheme 5
The photochemistry of 13 is very similar to that of 12 presented in Scheme 4. ESIPT to the 2’
and 7’ꢀpositions give QMs 30a, 30b, and 31a. QM 31a undergoes quantitative cyclization to 21. When
1
carried out in D2O solution, 21ꢀ7’D is instead formed, and H NMR spectroscopy does not show
evidence for the presence of more than one diastereomer of 21ꢀ7’D, which might mean that a concerted
mechanism takes place that is similar to that in Scheme 5 for 12, but occurs quantitatively for 13, which
can be rationalized by the additional steric bulk present for 13, which would hinder the rotation to
planarity necessary to produce 31a. QMs 30a and 30b revert to starting material via reverse proton
transfer to give 10, labeled with deuterium at the 2’ꢀposition when carried out in D2O solution.
Irradiation of 14 in acetonitrile, alcohols, or aqueous acetonitrile results in ESPT from the phenolic
OH to either the 5’ꢀcarbon or ESIPT to the 10’ꢀcarbon of the phenanthrene ring system producing QMs
28 or 27, respectively (Scheme 6). QM 28 undergoes quantitative cyclization to 25, which gives 24
after oxidation by air. QM 27 can be trapped by methanol to give the photoaddition product 26. In
aqueous acetonitrile not containing methanol, 27 will revert back to 14, possibly via the unstable
ACS Paragon Plus Environment
26

Page 27 of 41
The Journal of Organic Chemistry
hydration product 22. Similar reaction pathways were shown to be available to related phenanthrene
1
2
3
systems.9a
4
5
6
7
8
L
H
L
O
OH
H
O
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
L
13 (L = H)
13-2'D (L = D)
30a (L = H)
30a-7'D (L = D)
21 (L=H)
21-7'D (L=D)
L
ESIPT (2')
ESIPT (2')
H
O
O
ESIPT (7')
13*
H
L
hν
30b (L = H)
30b-7'D (L = D)
31a (L = H)
31a-7'D (L = D)
OL
Scheme 6
13 (L = H)
13-OD (L = D)
The inert derivatives 10 and 11 are structurally similar in that they both possess the 2ꢀnaphthol unit
and an attached phenyl ring – which is apparently an unfavourable combination for ESIPT reactions.
The example of 8 shows that ESPT can take place efficiently from a 1ꢀnaphthol OH to an attached
phenyl ring.10 The lower reactivity of 2ꢀnaphthols 10 and 11 relative to 1ꢀnaphthol 8 might partly result
*
from the fact that the phenolic OH in 1ꢀnaphthol is more acidic (pKa = 0.4) than the OH in 2ꢀnaphthol
in S1 (pKa = 2.8).16 However, the examples of BINOL and 13 show that the 2ꢀnaphthol unit can
*
function as the ESIPT ‘donor’ ring system. In addition, ESIPT giving QM intermediates has been
demonstrated by the Popik group in several 3ꢀsubstituted 2ꢀnaphthols with moderately efficient quantum
yields (Φ ~ 0.20), indicating that charge transfer from 2ꢀnaphthol can take place from the OH to
substituents attached to the 3ꢀposition.14a,14b,17 In these examples, however, ESIPT may be facilitated by
ACS Paragon Plus Environment
27

The Journal of Organic Chemistry
Page 28 of 41
the fact that the protonꢀaccepting groups are more basic in S1 than the simple phenyl rings present in 10
and 11. Perhaps the lack of reactivity of 10 and 11 is a combined effect of the phenolic OH being not
quite acidic enough, and the phenyl ring being not quite basic enough. It is important to note that all
examples of ESPT discovered in this work occur from the phenol OH to the aromatic system not
containing the OH. This fact indicates that 12 – 14 have strongly polarized singlet excited states that
result from significant charge transfer from the phenolic ring (which is the electronꢀrich ring in S0) to
the nonꢀphenolic ring, and that all proton transfers to aromatic carbon atoms take place on the excited
state energy surface.
1
2
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
Conclusions
Our investigation into the photochemistry of substituted phenyl phenols continues to unearth
interesting new reaction types and provide mechanistic insights into the more general mechanistic class
of ESPT to aromatic carbon atoms. The compounds studied in this work show three distinct reaction
types: 1) formation of QMs that undergo RPT to regenerate starting material (with possible deuterium
incorporation); 2) formation of QMs that can be trapped by solvent to give addition products; and 3)
formation of QMs that undergo electrocyclic ring closure to give benzoxanthene derivatives.
Experimental Section
General
Acetonitrile used in all experiments was HPLC or spec. grade and was distilled over CaH2 prior to
use. THF, diethyl ether, and toluene used in syntheses were distilled over Na (with benzophenone as an
indicator) and transferred via syringe to avoid contact with air. Reagent grade CH2Cl2 was distilled prior
to use as a reaction solvent. All other solvents were spec. or HPLC grade and used as received. HRMS
measurements employed EI ionization and a magnetic sector mass analyzer.
Synthesis Procedure
2-Phenyl-3-naphthol (10).18 A solution of 3ꢀbromoꢀ2ꢀmethoxynaphthalene (2.00 g, 8.4 mmol) in
toluene (50 mL) was stirred was mixed with solutions of phenylboronic acid (1.23 g, 10 mmol) in
ACS Paragon Plus Environment
28

Page 29 of 41
The Journal of Organic Chemistry
ethanol (50 mL) and potassium carbonate (5.05 g) in water (50 mL). This resulting mixture was stirred
and bubbled with Ar for 10 minutes. To the purged solution was added Pd(PPh3)4 (0.21 mol, 2.5
mol%), and the mixture was brought to reflux for 48 hours under a positive pressure of argon. Once
cool, the reaction mixture was diluted with dichloromethane (100 mL) and washed with 0.1 M sodium
hydroxide solution (2 ×100 mL). The organic layer was dried (MgSO4) and the solvent was removed
1
2
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
under reduced pressure. The resulting oil (3ꢀphenylꢀ2ꢀmethoxynaphthalene was dissolved in
dichloromethane (25 mL) and was added dropwise to a stirring solution of BBr3 (10 mmol in 80 mL of
dichloromethane) in an ice bath under an argon atmosphere over a period of 10 min. The mixture was
then removed from the ice bath and allowed to warm to room temperature and continue to stir for 3
hours. The reaction was quenched by the addition of 100 mL of water. The organic layer was dried
(MgSO4) and the solvent was removed under reduced pressure. The resulting solid was recrystallized
from heptanes to give 10, a white solid (1.22 g, 66% from 3ꢀbromoꢀ2ꢀmethoxynaphthalene) as the final
product. Spectroscopic data for 10 matched data previously reported for this compound.18
1-Phenyl-2-methoxynaphthalene (16).19 A solution of 1ꢀbromoꢀ2ꢀmethoxynaphthalene (2.0 g, 8.4
mmol) and Pd(PPh3)4 (0.25 g, 0.21 mol, 2.5 mol%) in 50 mL toluene was stirred under a N2 atmosphere.
To this was added a solution of phenylboronic acid (1.22 g, 10 mmol) in 50 mL ethanol, and a solution
of K2CO3 (5.0 g) in 50 mL water. Mixture was brought to reflux for 20 h, and allowed to cool to room
temperature. The solution was filtered through a celite plug and diluted with 100 mL CH2Cl2, and
washed with 2 ×100 mL 0.1 M NaOH, and with 100 mL water. The resulting solution was dried over
MgSO4, and the solvent was removed under vacuum. The crude product was recrystallized twice from
hot 95% ethanol to give white crystals of 16 (1.26 g, 49%). 1H NMR (500 MHz, CDCl3) δ 7.88 (d, 1H, J
= 7 Hz), 7.82 (m, 1H), 7.49 (m, 3H), 7.42 (ddd, J = 7.6, 7.3, 1.5), 7.37 (m, 3H), 7.33 (m, 2H), 3.83 (s,
13
3H). C NMR (125 MHz, CDCl3) δ 153.7, 136.4, 133.6, 131.0, 129.1, 129.0, 128.1, 127.8, 127.1,
126.3, 125.4, 125.3, 123.3, 113.9, 56.8. HRMS calcd. for C17H14O: 234.1045; observed 234.1045.
NMR data is consistent with those previously reported.19
ACS Paragon Plus Environment
29