Mechanisms of the photochemical rearrangement of diphenyl ethers

Article abstract of DOI:10.1021/jo9517196

The mechanism of the photochemical rearrangement of diphenyl ether (1a) was studied. Irradiation of 1a in ethanol gave 2-phenylphenol (2, 42%) and 4-phenylphenol (3, 11%) as rearrangement products, in addition to phenol (4, 30%) and benzene (5, 25%) as diffusion products. Cross-coupling experiments employing [₂H₁₀]1a demonstrated that the formation of 2- and 4-phenylphenol was an intramolecular process. Irradiation of 1a in benzene or in toluene gave biphenyls in good yields. The combined yields of rearrangement products (2 and 3) increased with increase of solvent viscosity, with a concomitant decrease in the formation of 4. All the results can be rationalized in terms of excitation of 1a to the singlet state and dissociation to a radical pair intermediate involving phenoxy and phenyl radicals. Intramolecular recombination of these radicals gives rearrangement products, and escape followed by hydrogen abstraction from the solvent gives diffusion products. When position 4 of 1a was occupied by an electron-donating substituent (1b-e), aryloxy-phenyl bond cleavage to give the corresponding rearrangement products prevailed over phenoxy-aryl bond cleavage. The opposite was the case for substrates with an electron-withdrawing substituent at position 4 (1h,i).

Full text of DOI:10.1021/jo9517196

J . Org. Chem. 1996, 61, 735-745
735
Mech a n ism s of th e P h otoch em ica l Rea r r a n gem en t of Dip h en yl
Eth er s
Naoki Haga* and Hiroaki Takayanagi
School of Pharmaceutical Sciences, Kitasato University, 5-9-1 Shirokane, Minato-ku, Tokyo 108, J apan
Received September 19, 1995X
The mechanism of the photochemical rearrangement of diphenyl ether (1a ) was studied. Irradiation
of 1a in ethanol gave 2-phenylphenol (2, 42%) and 4-phenylphenol (3, 11%) as rearrangement
products, in addition to phenol (4, 30%) and benzene (5, 25%) as diffusion products. Cross-coupling
experiments employing [2H10]1a demonstrated that the formation of 2- and 4-phenylphenol was
an intramolecular process. Irradiation of 1a in benzene or in toluene gave biphenyls in good yields.
The combined yields of rearrangement products (2 and 3) increased with increase of solvent viscosity,
with a concomitant decrease in the formation of 4. All the results can be rationalized in terms of
excitation of 1a to the singlet state and dissociation to a radical pair intermediate involving phenoxy
and phenyl radicals. Intramolecular recombination of these radicals gives rearrangement products,
and escape followed by hydrogen abstraction from the solvent gives diffusion products. When
position 4 of 1a was occupied by an electron-donating substituent (1b-e), aryloxy-phenyl bond
cleavage to give the corresponding rearrangement products prevailed over phenoxy-aryl bond
cleavage. The opposite was the case for substrates with an electron-withdrawing substituent at
position 4 (1h ,i).
In tr od u ction
viscosity effect on product distribution.9 These radical
pairs may undergo rapid recombination to give rear-
rangement products, or the radicals may escape to afford
free radicals. Quantum yields of these processes are
relatively high because the bond dissociation energies of
the severed bonds are relatively low, and the resulting
radicals are stabilized by delocalization of their odd
electron through the conjugated system.
Several rearrangements of aromatic compounds, for
instance the Claisen rearrangement,1,2 the Fries rear-
rangement,3,4 and the benzidine rearrangement,5,6 are
widely known to proceed via a photochemical process as
well as by acid-catalyzed (thermal) reaction. The mech-
anisms of these photochemical rearrangements can be
clearly distinguished from the ground state process.
Most of them proceed via homolytic cleavage of a carbon-
heteroatom or heteroatom-heteroatom bond in the ex-
cited state to generate radical species, as demonstrated
by direct detection of radical intermediates by flash
photolysis,7 by chemically induced dynamic nuclear
polarization studies,8 and by measuring the solvent
Although diphenyl ether (1a ) is stable to acids10 or high
temperatures,11 it is converted on UV irradiation in
2-propanol into 3 and 4 in a low quantum yield.12
Though this reaction is formally classified as a photo-
Claisen rearrangement, this [1,3] rearrangement is dif-
ferent from the [3,3] photo-Claisen rearrangement. Dur-
ing the two decades after the above discovery, photolability
of 1a and its derivatives has been investigated by only a
few workers,2b,13-17 owing to its low reactivity upon UV
irradiation and limited synthetic utility. Kelly et al.2b
and Hageman et al.13 used a high-pressure mercury lamp
to irradiate 1 in ethanol and found that 2 was formed,
in addition to 3 and 4. In 1970, Ogata et al.14 found that
the formation of 2 was essentially an intramolecular
process via a singlet excited state, and the product
distribution of this rearrangement depended somewhat
on the hydrogen-bonding ability of the solvent used.
Photolyses of asymmetrically substituted diphenyl ethers
X Abstract published in Advance ACS Abstracts, J anuary 1, 1996.
(1) (a) de Mayo, P. D. Molecular Rearrangements; J ohn Wiley and
Sons: New York, 1963; Vol. 1, pp 660-684. (b) Kupczek-Subotkowska,
L.; Subotkowski, W.; Saunders, W. H., J r.; Shine, H. J . J . Am. Chem.
Soc. 1992, 114, 3441-3445. (c) Hart. S. L.; Waddington, C. R. J . Chem.
Soc., Perkin Trans. 2 1985, 1607-1612.
(2) (a) Shimamura, N.; Sugimori, A. Bull. Chem. Soc. J pn. 1971,
44, 281-283. (b) Kelly, D. P.; Pinhey, J . T.; Rigby, R. D. G. Aust. J .
Chem. 1969, 22, 977-979.
(3) (a) Blatt, A. H. Organic Reactions; J ohn Wiley and Sons: New
York, 1942; Vol. 8, pp 342-384. (b) Dauson, I. M.; Hart, S. L.; Litter,
J . S. J . Chem. Soc., Perkin Trans. 2 1985, 1601-1606. (c) Dauson, I.
M.; Gibson, J . L.; Hart, L. S.; Waddington, C. R. J . Chem. Soc., Perkin
Trans. 2 1989, 2133-2139. (d) Gibson, J . L.; Hart, L. S. J . Chem. Soc.,
Perkin Trans. 2 1991, 1343-1348.
(4) (a) Belluˇs, D. Advances in Photochemistry; J ohn Wiley and
Sons: New York, 1971; Vol. 1, pp 109-159. (b) Nakagaki, R.;
Hiramatsu, M.; Watanabe, T.; Tanimoto, Y.; Nagakura, S. J . Phys.
Chem. 1985, 89, 3222-3226. (c) Shine, H. J .; Subotkowski, W. J . Org.
Chem. 1987, 52, 3815-3821.
(9) Sandner, M. R.; Trecker, D. J . J . Am. Chem. Soc. 1967, 89,
5725-5726.
(10) (a) Carson, D. W.; Ignasiak, B. S. Fuel 1980, 59, 757-761. (b)
Hei, R. D.; Sweeny, P. G.; Stenberg, V. I. Fuel 1986, 65, 577-585.
(11) (a) Burwell, R. L., J r., Chem. Rev. 1954, 59, 615-685. (b) Hagen,
A. P.; Miller. T. S.; Bynum, R. L.; Kapila, A. P. J . Org. Chem. 1982,
47, 1345-1347.
(5) (a) Shine, H. J . Aromatic Rearrangements; Elsevier: Amsterdam,
1967; pp 124-179. (b) Shine, H. J .; Zmuda, H.; Park, K. H.; Kwart,
H.; Horgan, A. G.; Brechbiel, M. J . Am. Chem. Soc. 1982, 104, 2501-
2509.
(12) Kharasch, M. S.; Stampa, G.; Nudenberg, W. Science 1952, 116,
309.
(6) (a) Shine, H. J .; Cheng, D. J . J . Org. Chem. 1971, 36, 2787-
2790. (b) Cheng, D. J .; Shine, H. J . J . Org. Chem. 1974, 39, 336-340.
(c) Cheng, D. J .; Shine, H. J . J . Org. Chem. 1974, 39, 2835-2838.
(7) Karmus, C. E.; Hercles, D. M. J . Am. Chem. Soc. 1974, 96, 449-
456.
(8) (a) Benn, R.; Schchmann, H.-P.; von Sonntag, C. Z. Naturforsch.
1979, 34B, 1002-1009. (b) Adam, W.; de Sanavia, J . A.; Fischer, H. J .
Org. Chem. 1973, 38, 2571-2572. (c) Adam, W. J . Chem. Soc., Chem.
Commun. 1974, 289-290.
(13) Hageman, H. J .; Louwerse, H. L.; Mijs, W. J . Tetrahedron 1970,
26, 2045-2052.
(14) Ogata, Y.; Takagi, K.; Ishino, I. Tetrahedron 1970, 26, 2703-
2709.
(15) J ochek, H.-I.; Miller, S. I. J . Am. Chem. Soc. 1966, 88, 3269-
3272.
(16) Hageman, H. J .; Huysmans, W. G. B. Recl. Trav. Chim. Pays.-
Bas. 1972, 91, 528-532.
(17) Elix, J . A.; Murphy, D. P. Aust. J . Chem. 1975, 28, 1559-1582.
0022-3263/96/1961-0735$12.00/0 © 1996 American Chemical Society

736 J . Org. Chem., Vol. 61, No. 2, 1996
Haga and Takayanagi
Sch em e 1
were carried out by several workers,15-17 but there is no
apparent consistency in their results. Through these
studies, it was inferred that the photolysis of 1a proceeds
via a radical pair (phenyl radical and phenoxy radical)
intermediate, and the radicals either recombine or escape
to afford free radicals. Though the radical pair mecha-
nism has been widely accepted, little evidence is available
to support (or contradict) this hypothesis. Our aims were
therefore (1) to reinvestigate the intramolecularity of this
rearrangement and (2) to identify the intermediate(s).
14
The cross-coupling experiment by Ogata et al.
is
unsatisfactory in two respects. Firstly, they employed
1a and p,p′-ditolyl ether, but did not examine all of the
possible intermolecular products which might be formed.
Secondly, the quantum yields differed between the two
substrates. To avoid these problems, we carried out a
cross-coupling experiment employing 1a and [2H10]1a as
substrates, and the resulting products (2 and 3) were
analyzed by whole molecular ion mass spectrometry
(WMIMS).
In the case of the photochemical rearrangement of
benzyl phenyl ether (a kind of photo-Claisen rearrange-
ment),18 the benzyl radical generated undergoes dimer-
ization to give 1,2-diphenylethane or abstracts hydrogen
to give toluene, as well as coupling with the phenoxy
radical to give 2- and 4-benzylphenol.2b If a similar
mechanism operates in the photochemical rearrangement
of 1a , the corresponding intermediate is a phenyl radical
which may dimerize to give biphenyl or abstract hydro-
gen to give benzene, as well as being converted into 2
and 3. However, the phenyl radical is much less stable
than the benzyl radical, which can be stabilized by
delocalization of its odd electron into the aromatic ring.
In order to test the intermediacy of the short-lived phenyl
radical, we irradiated 1a in benzene or toluene as a
phenyl radical scavenger and examined the solvent
viscosity dependence of the product distribution. Fur-
ther, to evaluate the involvement of a radical pair in this
reaction, solvent viscosity effects on the distribution
between rearrangement products and 4 were investi-
gated.
F igu r e 1. Time course of irradiation of 1a in ethanol.
No other product was detected by TLC, though small
amounts of unidentified polymeric products with low Rf
values were formed.4d Examination of the possibility of
secondary photochemical reaction of the primary photo-
chemical products revealed that 2 and 3 were somewhat
unstable to irradiation. Upon independent irradiation
of 2, 3, and 4, 7% of 2 and 15% of 3 were lost, and more
than 35% of 4 decomposed to unidentified polymeric
products after 8 h of internal irradiation. This photo-
lability of 4 should be taken into consideration in assess-
ing yields after prolonged irradiation.
For identification and quantification of volatile prod-
ucts, the reaction mixture was analyzed by GC. The time
courses of consumption of 1a and formation of the
products are illustrated in Figure 1. Surprisingly, be-
sides 2, 3, and 4, benzene (5) was formed in a yield
comparable with that of 4 (Scheme 1). The yields of the
four products and recovered 1a after 4.0 and 8.0 h of
irradiation were 37.2% and 41.8% for 2, 12.1% and 11.5%
for 3, 23.4% and 30.1% for 4, 16.9% and 25.0% for 5, and
22.2% and 2.3% for 1a , respectively. Prolongation of the
irradiation from 4.0 to 8.0 h increased the yields of 2, 4,
and 5, with a slight decrease of 3, and the total yield
decreased from 94.9% to 85.7%.
Irradiation of 1a in the presence of cis-1,3-pentadiene,
a typical triplet quencher, was carried out, and the
resulting mixture was analyzed by GC. The yields after
4.0 h of irradiation were 32.0%, 12.1%, 20.5%, 15.3% and
33.0% for 2, 3, 4, 5, and recovered 1a , respectively. The
presence of cis-1,3-pentadiene had no influence on the
photochemical reaction of 1a .
Ir r a d ia tion in Ben zen e. If a free phenyl radical
intervened as an intermediate, it would be trapped as
biphenyl (6) in benzene.19 When 1a was irradiated in
benzene, the reaction proceeded sluggishly because of
intense absorption of incident light by benzene in the
region of 230-260 nm,20 and more than 90% of 1a was
Resu lts
Qu a n tita tive An a lyses of th e Rea ction P r od u cts
in Eth a n ol. Ogata et al. reported that upon unfiltered
light irradiation from a high-pressure mercury lamp in
ethanol for 25 h, 1a rearranged to give 2 (20.4%), 3
(12.7%), and a trace amount of 4.14 We repeated irradia-
tion of 1a in degassed ethanol and confirmed their
results, except that we obtained higher yields of the
products. Thus, internal irradiation for 4 h converted
1a into 2, 3, and 4 in yields of 40%, 12%, and 19%,
respectively, after isolation by column chromatography.
The structures of products were identified by comparison
of the spectral properties with those of authentic samples.
(18) (a) Hashimoto, S.; Nomura, M.; Kono, K. Nippon Kagaku Kaishi
1972, 92-95. (b) Badr, M. Z. A.; Ali, M. M.; El-Sherief, H. A. H. Indian
J . Chem. 1974, 12, 1215-1216.
(19) (a) Tokumaru, K. Kagaku no Ryoiki, Zokan 1967, 81, 103-
118. (b) Tokumaru, K. Yuki Gosei Kagaku Kyokaishi 1970, 28, 773-
792.

Rearrangement of Diphenyl Ethers
J . Org. Chem., Vol. 61, No. 2, 1996 737
Ta ble 1. Solven t Effects on th e P r od u ct Distr ibu tion in th e Rea r r a n gem en t of 1a
yield (%)
solvent
viscositya
conversion (%)
2
3
4
2/3
(2 + 3 + 4)/(2 + 3)
1,2-propanediol
1,2-ethanediol
2-propanol
acetic acid
ethanol
methanol
1,4-dioxane
THF
diethyl ether
cyclohexane
n-heptane
n-hexane
46.5
55.9
71.2
57.2
44.7
64.7
66.9
51.7
62.5
49.0
22.3
17.2
18.2
90.2
40.2
54.7
28.6
19.5
29.2
28.0
19.5
14.5
6.5
6.2
2.5
1.9
18.1
4.2
6.8
13.8
6.6
14.7
13.8
6.1
5.0
2.2
2.0
1.3
1.2
2.2
9.57
8.04
2.07
2.95
1.99
2.03
3.20
2.90
2.95
3.10
1.92
1.90
2.29
1.03
1.04
1.35
1.73
1.45
1.58
1.65
2.79
2.95
1.57
2.66
3.10
2.56
17.73
2.151b
1.139
1.078
0.5445
1.204
0.460
0.2234b
0.898
0.3967
0.2985
0.342
14.9
19.1
19.5
24.4
16.7
35.0
33.8
4.7
6.3
6.1
40.5
1.0
7.9
acetonitrile
a
b
cP at 25 °C. Calculated value according to Bingham’s equation.48
Sch em e 2
recovered 1a (70.1%), 2 (2.8%), 3 (1.8%), and 4 (6.8%),
2-, 3-, and 4-phenyltoluene (7, 8, and 9) were formed in
yields of 2.3% (7), 3.1% (8), and 2.1% (9) after 8.0 h.
In flu en ce of Solven t Viscosity on P r od u ct Dis-
tr ibu tion . If a radical pair intervened as an intermedi-
ate in this reaction, the yields and product distribution
would depend significantly on the solvent, as has been
observed for many radical reactions.2a,9,21 Thus, forma-
tion of the rearrangement products (2 and 3) would
increase in going to a more viscous solvent, accompanied
with a decrease of diffusion products (4 and 5). Ogata et
al. found that formation of 2 and 3 from 1a was facilitated
in ethanol rather than in diethyl ether and that the
rearrangement/diffusion ratio (the ratio of the combined
yields of 2 and 3 to that of phenol) was essentially
independent of solvent viscosity.14 They concluded that
the pathway via solvation by alcohol at the oxygen atom
of 1a is favored because the C-O bond cleavage is
facilitated, and solvent viscosity plays a limited role.
However, this result should be reexamined because they
employed solvents of a limited viscosity range (from 0.33
to 2.95 cP). We used 13 solvents with a viscosity range
from 0.223 (diethyl ether) to 46.5 cP (1,2-propanediol).
External irradiation of 1a in these solvents was con-
ducted simultaneously, and the resulting mixture was
analyzed by GC (for details, see the Experimental Sec-
tion). Yields of products (2, 3, and 4) together with
amounts of recovered 1a are listed in Table 1. Two
features were noted. First, varying the solvent caused
a dramatic variation of product distribution. Thus, the
more viscous the solvent, such as 1,2-propanediol and 1,2-
ethanediol, the higher the yields of 2 and 3. To the
contrary, in solvents with low viscosity, such as aceto-
F igu r e 2. Time course of irradiation of 1a in benzene (top)
and in toluene (bottom).
recovered unchanged even after 8.0 h of internal irradia-
tion. However, in this reaction it turned out that 6 was
formed in a high yield compared to the rearrangement
products (2 and 3) (Scheme 2). The time courses of
formation of the products are shown in Figure 2. Com-
ponents of the reaction mixture after 8.0 h of irradiation
were 1a (71.8%), 2 (3.1%), 3 (2.0%), 4 (5.9%), and 6 (7.9%)
by GC analyses. Polyphenyls such as quaterphenyl were
not detected by GC.19a Unidentified polymeric products
with low Rf values were formed in a higher yield than in
the reaction in ethanol. A similar result was obtained
on irradiation of 1a in toluene (Figure 2). In addition to
(20) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photo-
chemistry, 2nd ed.; Marcel Decker, Inc.: New York, 1993; p 165.
(21) Leffler, J . E. An Introduction to Free Radicals; J ohn Wiley and
Sons: New York, 1993; pp 56-99.

738 J . Org. Chem., Vol. 61, No. 2, 1996
Haga and Takayanagi
Ta ble 2. Deu ter iu m Abu n d a n ce of 1a , 2, a n d 3 a fter
Ir r a d ia tion of [2H10]1a a
compound
2H abundance (%)
1a (initial)
1a (recovered)
2
3
95.73 ( 0.04
94.54 ( 0.22
90.04 ( 0.21
95.70 ( 0.29
a
Mean value of three or four runs. Each run involved 180-
450 scans.
initial and for the recovered sample) were carried out on
the basis of relative peak intensities in the range of m/e
177 (M - 3) to 181 (M + 1), as described for determining
the deuterium abundance of O-phenylhydroxylamines.24
Details of the calculation are described in the Experi-
mental Section. Similarly, the deuterium abundances of
[2H9]2 and [2H9]3 were calculated in the range of m/e 175
(M - 4) to 180 (M + 1). The results are summarized in
Table 2. For the recovered 1a and 3, deuterium abun-
dance did not vary from that of the initial 1a , which
shows that no exchange of deuterium on the aromatic
ring with protons of the solvent or the unlabeled sub-
strate occurred during irradiation. Though a decrease
of deuterium abundance of approximately 5% was ob-
served for 2, it was assumed that this did not significantly
affect the mass analyses of cross-coupling experiments.
F igu r e 3. Plot of the reciprocal ratio of the combined yield of
2 and 3 to that of 2, 3, and 4 against the reciprocal solvent
viscosity: Y(total), combined yield of 2, 3, and 4; Y(2 + 3),
combined yield of 2 and 3; solvents, (1) 1,2-propanediol, (2)
1,2-ethanediol, (3) 2-propanol, (4) 1,4-dioxane, (5) acetic acid,
(6) ethanol, (7) cyclohexane, (8) methanol, (9) THF, (10)
n-heptane, (11) acetonitrile, (12) n-hexane, and (13) diethyl
ether.
Ch a r t 1
Cross-coupling experiments were carried out by inter-
nal irradiation of an equimolar mixture of 1a and [2H10]-
1a (0.01 M total substrate) in ethanol for 45 min.
Conversion of 1a was 25%. The isolated 1a , 2, and 3
were analyzed by multiscan WMIMS. For 2 and 3,
relative peak intensities of m/e 168-180 were measured
to evaluate the contents of [2H9]2(3), [2H5]2(3), [2H4]2(3),
and unlabeled 2(3). The results are presented in Table
3. Deuterium abundance was 87.5% and 95.5% for 2 and
3, respectively. In the case of 3, relative peak intensities
in the range of 176-180 did not vary from those in the
range of 177-181 in 1a . Peaks with m/e 175 and 174
resulting from the formation of hybridized products,
4-OHC6H4C6D5 ([2H5]3) and 4-OHC6D4C6H5 ([2H4]3), were
not found. On the other hand, the peak composition of
2 is complicated in that a significant decrease of peak
intensity at m/e 179 was observed, accompanied with an
increase of those with m/e 175-178. Two factors con-
tribute to this result: the intense peak of M - 1 and the
deuterium exchange of 2 with the solvent. Notwith-
standing, the peaks with m/e 175 and 174, due to
formation of hybridized products 2-OHC6H4C6D5 ([2H5]-
2) and 2-OHC6D4C6H5 ([2H4]2), are negligible. For re-
covered 1a , relative peak intensities in the range of m/e
169-181 were measured in order to evaluate the contents
of [2H10]1a , [2H5]1a , and unlabeled 1a . The results are
summarized in Table 4. Because [2H10]1a used for this
experiment contains 5.0% protons, the relative peak
intensities of m/e 179 ([2H9]1a ) and 178 ([2H8]1a ) are
moderately large (22.4% of m/e 180 for m/e 179 and 10.9%
for m/e 178). As expected, relative peak intensities of 1a
did not vary before and after the irradiation. No peak
with m/e 175 ([2H5]1a ) was detected in the recovered
substrate. Thus, no hybridization between labeled and
nitrile, n-hexane, and diethyl ether, yields of rearrange-
ment products (2 and 3) were lower and formation of
diffusion products (4) increased to a significant degree.
For example, reaction in 1,2-ethanediol gave yields of
38% for 2, 4% for 3, and 1% for 4, whereas yields of 7%,
2%, and 35%, respectively, were obtained in diethyl ether.
Figure 3 plots the reciprocal ratio of the combined yield
of 2 and 3 to that of 2, 3, and 4 against the reciprocal
viscosity of solvents at 25 °C, giving a linear relation with
an intercept of 0.87 and a slope of 0.75 cP. Secondly,
the conversion efficiency of 1a decreased from 71.0% in
1,2-ethanediol to 17.2% in n-heptane. In general, conver-
sion decreased in the order of reaction in hydroxylic
solvents (alcohols and acetic acid), in ethers, and in
hydrocarbons.
Cr oss-Cou p lin g Exp er im en ts. To examine precisely
the intramolecularity of the rearrangements to the ortho
and to the para position, positions 2, 4, and 6 of 1 must
be free of a substituent. Thus, we chose the combination
of 1a and [2H10]1a as substrates for a cross-coupling
experiment (Chart 1). The labeled substrate was pre-
pared from [2H5]bromobenzene and [2H6]phenol according
to the modified method of Milyakh et al.22 In a prelimi-
nary experiment, 1a and [2H10]1a were found to undergo
rearrangement in nearly equal quantum yield. For the
purpose of testing the possibility of proton exchange of
[2H10]1a during irradiation, [2H10]1a was irradiated
independently in ethanol, and the deuterium abundance
of the resulting 2 and 3 as well as that of recovered 1a
was analyzed by multiscan WMIMS, which was devised
for evaluating heavy atom isotope ratios.23 Calculations
of hydrogen isotope abundance of [2H10]1a (both for the
(23) (a) Shine, H. J .; Gruszecka, E.; Subotkowski, W.; Brownawell,
M.; Fillipo, J . S., J r. J . Am. Chem. Soc. 1985, 107, 3218-3223. (b)
Shine, H. J .; Kupczyk-Subotkowaka, L.; Subotkowski, W. J . Am. Chem.
Soc. 1985, 107, 6674-6678. (c) Reimschu¨ssel, W.; Paneth, P. Org. Mass.
Spectrom. 1980, 15, 302-303.
(24) Haga, N.; Endo, Y.; Kataoka, K.; Yamaguchi, K.; Shudo, K. J .
Am. Chem. Soc. 1992, 114, 9795-9806.
(22) Milyakh, S. V.; Ivanova, V. O.; Cheremnykh, N. G. Nefteper-
erab. Neftekhim. 1973, 25-26; Chem. Abstr. 1973, 79, 104842y.

Rearrangement of Diphenyl Ethers
J . Org. Chem., Vol. 61, No. 2, 1996 739
Ta ble 3. Va r ia tion of Rela tive P ea k In ten sities ((Sta n d a r d Devia tion ) of 2 a n d 3 in Cr oss-Cou p lin g Exp er im en ts
relative peak intensityb at the following m/e
com-
pound
180
179
178
177
176
175
174 173 172
171
170
169
168
2
3
8.25 ( 1.85 42.24 ( 1.51 37.21 ( 0.86 21.53 ( 1.04 12.25 ( 1.39 3.43 ( 0.65
11.95 ( 0.98 86.85 ( 2.73 22.45 ( 1.34 8.76 ( 0.75 2.92 ( 0.58 0
0
0
0
0
0
0
20.53 (1.21 100 50.54 ( 1.75 3.22 ( 1.60
14.16 ( 0.35 100 11.79 ( 1.92 0
a
b
Mean value of three or four runs. Each run involved 280-440 scans. Relative value on the basis of m/e 170 as 100.
Ta ble 4. Va r ia tion of Rela tive P ea k In ten sities ((Sta n d a r d Devia tion ) of [2H10]1a a n d Un la beled 1a of Cr oss-Cou p lin g
Exp er im en ts
relative peak intensityb at the following m/e
compounds
181
180
179
178
177
176 175 174 173 172
171
170
169
1a (initial)
11.09 ( 0.13 84.82 ( 2.41 22.54 ( 0.38 10.88 ( 0.65 2.53 ( 0.86
0
0
0
0
0
0
0
0
0
0
13.34 ( 0.13 100 15.87 ( 1.23
13.45 ( 0.86 100 16.05 ( 0.62
1a (recovered) 10.99 ( 0.25 83.90 ( 2.56 22.95 ( 0.45 11.67 ( 0.70 3.86 ( 0.55
a
b
Mean value of three or four runs. Each run involved 180-220 scans. Relative value on the basis of m/e 170 as 100.
Sch em e 3
Ta ble 5. P r od u ct Distr ibu tion in th e P h otoch em ica l
Rea r r a n gem en t of 4-Su bstitu ted Dip h en yl Eth er s (1b-1j)
in Eth a n ol
Ch a r t 2
yield (%)a
irrad
X
time (h)
10
11
12
13
4
others
Me(1b)
3.0
2.0
1.0
0.3
8.0
2.5
2.5
2.0
47
33
37
37c
9
20
0
0
0
0
0
0
0
3
42
43
7
0
0
0
0
5
29
30
15
9
32
27d
6
0
0
0
0
0
0
0
5
0
14
13
OCH3(1c)
OH(1d )
NH2(1e)
C6H5(1f)
F(1g)
15b
10b
10,b 6e
50f
amino group of the products was converted into trifluo-
roacetamide (14, 15, and 16) by treatment with trifluo-
roacetic anhydride (TFAA) to facilitate isolation and
purification (Chart 2). In all cases, phenols (4 and/or 13)
were formed. Unexpectedly, elimination of the substitu-
ent occurred during the reaction on irradiation of 1c, 1d ,
1e, and 1g, which gave rise to the formation of 4-phe-
nylphenol (3). Formation of 4-phenylaniline (isolated as
16) from 1e and formation of 2 from 1g, though they are
minor in relation to the overall reaction, are abnormal.
The structures were established unambiguously on the
basis of elemental analyses and spectral properties or by
direct comparison with an authentic sample.
In contrast, irradiation of substrates that have an
electron-withdrawing substituent, 1h (X ) CO2CH3) and
1i (X ) CN), gave products exclusively via path b. The
aryl group rearranged to the ortho position and the para
position with respect to the phenoxy group to give 11 and
12, respectively. Formation of the former prevailed over
formation of the latter.
6,b 4g
CO2CH3(1h )
CN(1i)
a
b
d
Isolated yields. 4-Phenylphenol (3). c Isolated as 14. Iso-
lated as 15. e 2-Phenylaniline, which was isolated as 16. f Recovery
of 1f. 2-Phenylphenol (2).
g
unlabeled phenyl groups to generate C6D5OC6H5 ([2H5]-
1a ) occurred during the irradiation.
Su bstitu en t Effect on th e Regioselectivity. In
order to evaluate the substituent effect on the regiose-
lectivity of C-O bond fission, asymmetric ethers (1b-i)
whose para position is occupied by a substituent were
irradiated in ethanol. As shown in Scheme 3, the two
possible pathways will give independent products. Thus,
if path a (phenyl-aryloxy cleavage) operates in prefer-
ence to path b (aryl-phenoxy cleavage), 10 would be
produced as the rearrangement product, provided that
the substituent remained in the benzene ring during the
rearrangement. Product distributions in these reactions
are listed in Table 5. A distinct substituent effect on the
regioselectivity of the C-O bond cleavage was observed.
Reaction of substrates that have an electron-donating
substituent, 1b (X ) Me), 1c (X ) OMe), 1d (X ) OH),
and 1e (X ) NH2), proceeded predominantly via path a,
except for the formation of 12b from 1b in a low yield
(minor involvement of path b). That is, the 4-substituted-
2-phenylphenols (10), which were formed by phenyl
migration to the ortho position of the aryloxy group, were
isolated as a major product. In the reaction of 1e, the
Reaction of 1f (X ) Ph) and 1g (X ) F) was borderline.
That is, 9% yield of 10f as a product via path a and 6%
yield of 13f as a product via path b were formed.
Similarly, 20% yield of 10g as a product via path a and
3% yield of 11g and 5% yield of 12g as products via path
b were formed, together with small amounts of 2 and 3.
Discu ssion
In tr a m olecu la r ity of th e Rea r r a n gem en t. Both
the ortho and the para rearrangements were proved to

740 J . Org. Chem., Vol. 61, No. 2, 1996
Haga and Takayanagi
Ch a r t 3
Sch em e 4
be essentially intramolecular by the cross-coupling ex-
periments. The absence of peaks at m/e 175 and 174,
which are the molecular ions of the hybridized products,
4-OHCH4CD5 ([2H5]3) and 4-OHCD4CH5 ([2H4]3), in 3
provides convincing evidence that the para rearrange-
ment proceeds completely intramolecularly (Table 3). In
the formation of 2, the significant decrease of peak
intensity at m/e 179 accompanied with increases of those
at m/e 175-178 can be attributed to partial deuterium
exchange of [2H5]2 and the contribution of the intense M
- 1 peak of this compound, rather than to mixing of an
intermolecular process. The deuterium abundance was
reduced from 95.3% for the initial 1a to 87.2% for 2
resulting from formation of [2H8]2-[2H5]1a , and the peak
at m/e 169 has 50.5% of the intensity of that at m/e 170
(Table 3). Therefore, the ortho rearrangement also
turned out to be intramolecular.
For the recovered substrate, relative peak intensities
in the range of m/e 169-181 did not vary from those of
the initial substrate (Table 4). This indicates that [2H10]-
1a did not undergo phenyl group exchange with unla-
beled 1a to generate hybrid substrate (C6D5OC6H5)
during the reaction.
Sch em e 5
Mu ltip licity of th e Excited Sta te. Identical results
for yields of the products in the presence and in the
absence of cis-1,3-pentadiene indicate that the singlet
excited states of 1a are undoubtedly major species in the
reaction. Energy transfer from the triplet state of 1a to
cis-1,3-pentadiene is 92 kJ mol-1 exothermic as estimated
from the triplet excitation energies of 1a and cis-1,3-
convincing evidence that the C-O bond dissociates
entirely homolytically, because if 1a underwent hetero-
lytic C-O bond cleavage to generate an ion pair (18)
involving the phenyl cation and phenoxy anion, the
former would be trapped by ethanol as ethoxybenzene.
pentadiene of 339 kJ mol-1 25 and 247 kJ mol-1 26
,
Our results also eliminate the pathway involving
generation of an ion pair by electron transfer from a
radical pair (17). This is in contrast to the photolysis of
R-naphthylmethyl esters (21) in methanol (Scheme 4),25
where competitive reactions, i.e., electron transfer from
a radical pair (22) followed by dissociation to free ions
and decarboxylation followed by radical coupling, give
both ionic- (23) and radical-derived products (24). The
proportion of 23 to 24 may depend on both the oxidation
potential of the R-naphthylmethyl radical and the rate
constant for the decarboxylation.
The observed solvent viscosity effect on the ratio of
formation of 2 and 4 seems to disfavor the concerted
process. In the case of photo-Fries rearrangement of
4-methylphenyl acetate (25),9 the independence of the
quantum yield for formation of 2-hydroxy-5-methylac-
etophenone (26, in-cage product) on solvent viscosity and
the decrease of quantum yield for formation of 4-meth-
ylphenol (27, out-of-cage product) with increase of solvent
viscosity were attributed to the distinct mechanisms of
these processes (Scheme 5), i.e., formation of 26 by
concerted rearrangement via a transition state with a
four-membered ring, in contrast with formation of 27 by
dissociation into a radical pair followed by diffusion to
respectively. If the triplet state of 1a were the interme-
diate, the reaction would be quenched by cis-1,3-penta-
diene.
Mod e of Bon d Clea va ge a n d Bon d F or m a tion . As
regards the mode of C-O bond cleavage and C-C bond
formation in this rearrangement, four possible intramo-
lecular processes (Chart 3) can be considered for forma-
tion of 2 and 3: (i) a dissociative process via a radical
pair intermediate (17), (ii) a dissociative process via an
ion pair intermediate (18), (iii) a concerted process via a
transition state (19), and (iv) a ring closure and cleavage
process via a dibenzofuran-type intermediate (20). Among
them, 19 and 20 are disadvantageous for the para
rearrangement (formation of 3). Our present data sup-
port the radical pair process, as discussed in detail below.
The finding that irradiation of 1a in ethanol gave
radical-derived products 2, 3, 4, and 5 stoichimetrically
and that no ionic-derived product was produced provides
(25) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photo-
chemistry, 2nd Ed.; Marcel Decker, Inc.: New York, 1993; p 71.
(26) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photo-
chemistry, 2nd ed.; Marcel Decker, Inc.: New York, 1993; p 71.
(27) DeCosta, D. P.; Pincock, J . A. J . Am. Chem. Soc. 1989, 111,
8948-8950.

Rearrangement of Diphenyl Ethers
J . Org. Chem., Vol. 61, No. 2, 1996 741
line.29 A radical pair (31) involving a phenoxy radical
and an allyl radical, which recombine in the pair to give
29, has been proposed for the reaction mechanism. Our
experiments employing 13 different solvents revealed a
dramatic solvent viscosity effect on the product distribu-
tion in the photolysis of 1a . As shown in Figure 3, the
observed variation of product distribution on varying the
solvent (the more viscous the solvent, the higher the
yields of 2 and 3 and the lower the yield of 4) demon-
strates the intervention of a radical pair in this reaction,
as in the photo-Claisen rearrangement of 28. That is,
because the rate constant for diffusion of the resulting
radical pair is small in viscous solvents, in-cage recom-
bination to give 2 and 3 is favored over diffusion to afford
free radicals.
Sch em e 6
The radical pair mechanism described above is in
accord with the formation of biphenyl (6) in a high yield
in the reaction of 1a in benzene. This result shows that
the radical pair dissociates to a free phenyl radical, which
is then trapped by benzene. Though this reaction was
extremely slow because of competitive absorption of
incident light between 1a and benzene, formation of 4
and 6 in relatively higher yields than those of 2 and 3
shows that in this reaction, diffusion of the initial radical
pair to afford free radicals prevails over recombination.
Reaction of 1a in toluene gave a similar result: three
isomeric phenyltoluenes and 4 were produced in prefer-
ence to 2 and 3. The formation of 7, 8, and 9 with a ratio
of 1.1:1.5:1.0 is considerably different from the outcome
in thermal aromatic substitution by a phenyl radical
generated by dissociation of benzoyl peroxide.19 For
example, phenylation of toluene by phenyl radical gener-
ated by thermolysis of N-nitrosoacetylarylamine gave
three phenyltoluene isomers with a ratio of 7:8:9 ) 1.9:
0.8:1.7.19a The reason for this difference is unclear, but
we tentatively attribute this lower yield of 7 partially to
secondary reactions of the phenylcyclohexadienyl radical,
as yet unknown, to give products other than phenyltolu-
enes.
Mech a n ism of th e Rea r r a n gem en t. As shown in
Scheme 7, the primary chemical process is excitation
from the ground state 1 to a singlet state followed by
homolytic C-O bond cleavage to give a singlet radical
pair (17) involving a phenoxy radical and a phenyl
radical. In this pair, recombination of the phenyl radical
at the ortho and the para positions of the phenoxy radical
gives quinoid intermediates (32 and 33), followed by
aromatization to 2 and 3, respectively. In viscous
solvents the concerted process for the ortho rearrange-
ment may compete with the radical pair process. When
the radical pair dissociates to free radicals (34 and 35),
they cannot recombine either intramolecularly or inter-
molecularly. Instead, they abstract a hydrogen atom
from the solvent to afford phenol (4) and benzene (5).
When benzene is employed as the solvent, the free phenyl
radical (35) undergoes substitution with benzene to give
biphenyl (6). The radical mode for this reaction is not
affected by substitution at the para position of 1a , since
products derived from an ionic intermediate were not
isolated in the photolysis of 1b-1j (Table 5).
afford free radicals. On the other hand, the effect of
solvent viscosity on the formation of 2 from 1a was
opposite to that in the formation of 26: the yield of 2
increased from 1.9% in n-hexane to 55% in 1,2-ethanediol
on irradiation (Table 1). This variation in the yield of 2
shows that the efficiency of conversion of 1a into 2
depends appreciably on the nature of the solvent. This
favors a dissociative process rather than the concerted
process.
It is noteworthy that ratio of ortho to para rearrange-
ment is significantly larger in 1,2-ethanediol (2/3 ) 9.6)
and in 1,2-propanediol (2/3 ) 8.0) than in other solvents,
where an approximately constant value of the ratio of
1.8-3.0 is noted (Table 1). This may suggest that the
concerted process, which leads to the ortho rearrange-
ment alone, competes with the radical pair process in
viscous solvents such as 1,2-ethanediol and 1,2-pro-
panediol.4a In fact, the ortho migration of the phenyl
group to the phenoxy radical is a [1,3] rearrangement,
which is an allowed process in an excited state, in terms
of the orbital symmetry restriction, whereas the para
migration is a [1,5] process, which is forbidden. A clear
interpretation of the preference for ortho rearrangement
over para rearrangement, however, cannot be derived
from the solvent viscosity effect.
A fourth possibility for C-O bond cleavage and C-C
bond formation, i.e., dibenzofuran derivative (20) forma-
tion followed by the C-C bond cleavage, must be con-
sidered. It was reported that irradiation in the presence
of iodine as an oxidizing agent converted 1a into diben-
zofuran in 70% yield without formation of other prod-
ucts.28 However, for the same reason as in the case of
the concerted process, this process can be ruled out in
the absence of an oxidizing agent. If this mechanism
were operating in the ortho rearrangement of 1a , forma-
tion of 2 should proceed regardless of the solvent.
The solvent viscosity effect on product distribution
distinguishes between the radical pair mechanism and
the other mechanisms. In the case of photo-Claisen
rearrangement of allyl phenyl ether (28) in benzene,2a
the more viscous the solvent used, the higher the
quantum yields for allylphenols (29 and 30), while the
quantum yield for formation of phenol (4) remained
unchanged (Scheme 6). A plot of reciprocal ratio of
quantum yields for formation of rearrangement product
against reciprocal viscosity of the solvent gave a straight
Formation of 3 from 1c, 1d , 1e, and 1g is noteworthy.
We attribute this to para rearrangement via path a
(intramolecular rearrangement of the phenyl group to the
ipso position of the substituent) followed by removal of
the substituent, rather than to photolytic removal of the
substituent in the substrates and/or products. Indepen-
dent irradiation of 10c, 10d , 10e, and 10g in ethanol
(28) (a) Zeller, K. P.; Petersen, H. Synthesis 1975, 532-533. (b)
Zeller, K. P.; Gauglitz, G. Z. Naturforsch. 1977, 32B, 285-288.
(29) Booth, D.; Noyes, R. M. J . Am. Chem. Soc. 1960, 82, 1868-
1872.

742 J . Org. Chem., Vol. 61, No. 2, 1996
Haga and Takayanagi
Sch em e 7
correlated with the Hammett σ values in spite of the
radical nature of this reaction. The Hammett plot of σ
vs the ratio of the combined yield of products via path a
to that of total products is presented in Figure 4. Though
the influence of a substituent at position 4 on the
regioselectivity is clear, we have no explanation as yet
for these fascinating phenomena.
Con clu sion
We have conducted preparative photolyses to elucidate
the mechanism of the photochemical rearrangement of
diphenyl ether (1a ) and found that 2-phenylphenol (2)
and 4-phenylphenol (3) are formed as rearrangement
products in higher yields than those obtained in earlier
studies, together with phenol (4) and benzene (5) as
diffusion products. The stoichiometrical formation of
radical-derived products 2, 3, 4, and 5 provides convinc-
ing evidence for homolytic C-O bond cleavage, but not
for heterolytic cleavage.
An intramolecular mechanism for both the ortho and
the para rearrangement was unambiguously demon-
strated by cross-coupling experiments employing [2H10]-
1a and unlabeled 1a .
Plots of the reciprocal ratio of the combined yield of 2
and 3 to that of 2, 3, and 4 against the reciprocal viscosity
of solvents gave a linear relation. This, together with
formation of biphenyls in photolyses of 1a in benzene and
toluenes, suggests that the reaction involves a dissocia-
tion process, but not a concerted process.
All the results described above indicate intermediacy
of a radical pair involving a phenoxy radical and a phenyl
radical. Intramolecular recombination in the cage gives
rearrangement products, and escape from the cage fol-
lowed by hydrogen abstraction from the solvent gives
diffusion products.
Finally, we evaluated the substituent effect on the
regioselectivity of C-O bond fission in asymmetric ethers
(1b-1i) and found that electron-donating substituents
favor aryloxy-phenyl cleavage, in contrast to electron-
withdrawing substituents. This distinct substituent ef-
fect seems to be correlated with the Hammett σ values
in spite of the radical nature of this reaction.
F igu r e 4. Hammett plot of σ vs the ratio of the combined
yield of products via path a to that of total products: path a,
combined yield of the products formed via path a; path b,
combined yield of the products formed via path b.
revealed that each of them is photostable. Formation of
4-phenylaniline (isolated as 16) from 1e and formation
of 2 from 1g cannot be explained at present. These
processes seem to be independent of coupling or diffusion
of the radical pair (17).
Regarding the solvent effect for facilitation of dissocia-
tion of the C-O bond of 1a , our observations agree with
those of Ogata et al. Among the solvents we used,
alcohols and acetic acid afforded product yields of more
than 65% on irradiation for 16 h (Table 1). On the other
hand, in ethers and in hydrocarbons, the conversion was
40-50% and 18-30%, respectively. This inefficiency
(especially in hydrocarbons) may be explained in terms
of hydrogen bonding of hydroxylic solvents to the oxygen
atom of singlet 1 to facilitate the C-O bond cleavage or
to the quinoid intermediates (32 and 33) to stabilize
them.
Finally, a comment is needed on the distinct substitu-
ent effect on the regioselectivity observed in the photoly-
ses of the asymmetrically substituted ethers (1b-1j).
Preference of aryloxy-phenyl cleavage for substrates
with an electron-donating substituent and preference of
aryl-phenoxy cleavage for those with an electron-
withdrawing substituent (Scheme 3, Table 5) seem to be
Exp er im en ta l Section
Gen er a l. Melting points were obtained on a Yazawa melt-
ing point apparatus (type BY-10) without correction. 1H-NMR

Rearrangement of Diphenyl Ethers
J . Org. Chem., Vol. 61, No. 2, 1996 743
spectra were determined with a Varian UNITY 400 spectrom-
eter in various solvents, with tetramethylsilane as an internal
reference. UV spectra were obtained on a Hitachi UV-VIS 340
spectrophotometer. Low-resolution electron ionization mass
spectra (EI-MS) were taken on a J MS-DX 300. Multiscan
WMIMS were taken on a J MS AX 505 HA. Flash column
chromatography was performed on silica gel (Merck Art 9385
Kieselgel 60). Thin layer chromatography was performed on
silica gel (Merck Art 11696 TLC-Kieselgel 60 HF). GC was
taken on a Shimadzu GC 14-A equipped with a Chromatopac
C-R5A. Microanalyses were carried out in the microanalytical
laboratory of our school.
Solven ts. Spectral-grade ethanol, methanol, and acetoni-
trile and infinity-grade 2-propanol, acetic acid, diethyl ether,
tetrahydrofuran, 1,4-dioxane, n-hexane, n-heptane, and cyclo-
hexane were used for photolysis without further purification.
Special-grade 1,2-ethanediol and 1,2-propanediol were dried
over calcium hydride and distilled under reduced pressure (1,2-
ethanediol, 62 °C, 2.0 Torr; 1,2-propanediol, 65 °C, 1.3 Torr).
Ap p a r a tu s for P h otolysis. For preparative-scale irradia-
tion, a 400 W high-pressure mercury lamp (RIKO UVL-400-
HA, internal irradiation type) with a quartz jacket was placed
in a cylindrical irradiation flask (Pyrex, 170 or 500 mL). We
have found that all of 1 used in this study except 4-phenoxya-
niline (1e) are inert on irradiation at wavelengths longer than
300 nm from this lamp through a Pyrex filter. So the
unfiltered light from this lamp was used directly except in the
irradiation of 1e. Irradiation of 1e through a Pyrex filter
(passing wavelengths longer than 300 nm) for 6 h gave similar
results to those obtained by use of unfiltered light. Cooling
water was passed through the quartz jacket. The irradiation
flask was fitted with a nitrogen inlet, a magnetic stirrer bar,
and a gas outlet with a calcium chloride tube.
bip h en yl (1f): prepared from bromobenzene and 3a ; recrys-
tallized from hexane, colorless plates, yield 24%, mp 68.0-
69.5 °C (lit.36 mp 67-68 °C). Anal. Calcd for C18H14O: C,
87.78; H, 5.73; N, 0.00; O, 6.50. Found: C, 88.00; H, 5.87; N,
0.00. EI-MS: m/e 246 (M+). 1H-NMR (CDCl3): δ 7.57 (dd,
2H, J ) 8.8, 2.0 Hz), 7.56 (d, 2H, J ) 9.0 Hz), 7.43 (t, 2H, J )
8.8 Hz), 7.36 (dd, 2H, J ) 8.9, 7.1 Hz), 7.33 (tt, J ) 8.8, 2.0
Hz), 7.13 (tt, 1H, J ) 7.1, 1.2 Hz), 7.08 (d, 2H, J ) 9.0 Hz),
7.06 (dd, 2H, J ) 8.9, 1.2 Hz). (4-F lu or o-1-p h en oxy)-
ben zen e (1g): prepared from bromobenzene and 4-fluorophe-
nol; colorless oil, yield 58%, bp 78 °C/2.5 Torr (lit.37 bp 96-98
°C/2 Torr). Anal. Calcd for C12H9FO: C, 76.58; H, 4.82; F,
10.09; N, 0.00; O, 8.50. Found: C, 76.85; H, 4.79; F, 10.15; N,
0.00. EI-MS: m/e 200 (M+). 1H-NMR (CDCl3): δ 7.32 (dd,
2H, J ) 8.6, 7.5 Hz), 7.09 (dt, 1H, J ) 7.5, 1.3 Hz), 7.03 (dd,
2H, J ) 9.2, 7.7 Hz), 6.99 (dd, 2H, J ) 7.7, 4.9 Hz), 6.97 (dt,
2H, J ) 9.2, 1.3 Hz). Meth yl 4-P h en oxyben zoa te (1h ):
prepared from methyl 4-bromobenzoate and phenol; colorless
prisms, yield 62%, mp 62.5-63.0 °C (lit.38 mp 58-60 °C). Anal.
Calcd for C14H12O3: C, 73.67; H, 5.30; N, 0.00; O, 21.03.
Found: C, 73.91; H, 5.56; N, 0.00. EI-MS: m/e 228 (M+). 1H-
NMR (CDCl3): δ 8.03 (d, 2H, J ) 9.0 Hz), 7.39 (dd, 2H, J )
8.6, 7.3 Hz), 7.06 (dd, 2H, J ) 8.6, 1.2 Hz), 6.98 (d, 2H, J )
9.0 Hz), 3.90 (s, 3H). 4-P h en oxyben zen eca r bon itr ile (1i):
prepared from bromobenzenecarbonitrile and phenol; pale
yellow prisms, yield 80%, mp 32.0-33.0 °C (lit.31 mp 47 °C).
This compound was not recrystallized because of its low mp.
Anal. Calcd for C13H9NO: C, 79.98; H, 4.65; N, 7.17; O, 8.20.
Found: C, 80.07; H, 4.67; N, 7.13. EI-MS: m/e 195 (M+). 1H-
NMR (CDCl3): δ 7.60 (d, 2H, J ) 8.6 Hz), 7.42 (dd, J ) 8.5,
7.3 Hz), 7.23 (tt, 1H, J ) 7.3, 1.2 Hz), 7.07 (dd, 2H, J ) 8.5,
1.2 Hz), 7.00 (d, 2H, J ) 8.6 Hz).
P h otoch em ica l Rea r r a n gem en t of 1 (P r ep a r a tive Ir -
r a d ia tion ). As a typical example, irradiation of 1b in ethanol
will be described. After 30 min of nitrogen bubbling prior to
the irradiation, a solution of 314 mg (1.70 mmol) of 1b in 170
mL of absolute ethanol was irradiated at ambient temperature
for 3.0 h. Nitrogen bubbling was continued during the
irradiation. When the substrate was no longer detectable by
TLC, the solvent was removed by evaporation under reduced
pressure. The residual oil (348 mg) was flash-chromato-
graphed (silica gel, eluent: n-hexane then n-hexane-ethyl
acetate, 15:1, finally 8:1) to give 176 mg (47%) of 10b, 28 mg
(15%) of 13b, and 23 mg (7%) of 12b. A pale yellow colored
unidentified solid (35 mg) was discarded. Irradiation time and
product yields of other substrates are given in Table 5.
In the case of irradiation of 4-phenoxyaniline (1e), the amino
group of the products (10e, 13e, and 4-phenylaniline) was
converted to the corresponding trifluoroacetamide (14, 15, and
16, respectively) as follows. After irradiation of 315 mg (1.70
mmol) of 1e in 170 mL of ethanol and removal of the solvent,
the residual tarry mixture (343 mg) was suspended in 20 mL
of dichloromethane. To the solution was added 3.57 g (17.0
mg, 10 equiv) of TFAA at 0 °C. Stirring was continued for 2
h. After removal of the solvent under reduced pressure, the
residue was extracted with 150 mL of 2 N sodium hydrogen
carbonate and dichloromethane (100 mL twice). The combined
organic phase was dried over magnesium sulfate and filtered.
Evaporation of the solvent gave a brown solid (498 mg), which
was flash-chromatographed (silica gel, eluent: n-hexane-ethyl
acetate, 8:1, then 4:1) to afford 28 mg (6.2%) of 16, 30 mg
(10.4%) of 3, 176 mg (36.8%), of 14 and 96 mg (27.5%) of 15.
For analytical-scale irradiation, sample solutions in quartz
test tubes (10 mL) were irradiated simultaneously for the same
length of time in a merry-go-round apparatus (RIKO RH400-
10W) equipped with a 400 W high-pressure mercury lamp
(RIKO UVL-400-HA, external irradiation type) through a
quartz sleeve. Temperature of the solutions was maintained
at 25 °C in a thermostat.
P r ep a r a tion of Dip h en yl Eth er s (1). Commercially
available 1a was used after distillation under reduced pressure
(bp 75 °C/2.5 Torr, lit.30 bp 73-74 °C/1 Torr). Commercially
available 4-phenoxyphenol (1d ) and 4-phenoxyaniline (1e)
were purified by repeated recrystallization from dichlo-
romethane-hexane to give analytical purity (mp 1d , 85.0-85.5
°C, lit.31 75-80 °C; mp 1e, 84.5-85.0 °C, lit.32 83-85 °C). Other
substrates (1b, 1c, 1f, 1g, 1h , and 1i) were prepared by
Ullmann synthesis33 from the corresponding phenols and
bromobenzenes with freshly prepared copper powder as the
catalyst. Potassium tert-butoxide was employed instead of
potassium hydroxide.
(4-Meth ylp h en oxy)ben zen e (1b): prepared from bro-
mobenzene and p-cresol; colorless oil, yield 74%, bp 92 °C/1.5
Torr (lit.34 bp 110-111 °C/2 Torr). Anal. Calcd for C13H12O:
C, 84.75; H, 6.56; N, 0.00; O, 8.68. Found: C, 84.83; H, 6.55;
N, 0.00. EI-MS m/e 184 (M+). 1H-NMR (CDCl3): δ 7.32 (dd,
2H, J ) 8.9, 7.1 Hz), 7.15 (d, 2H, J ) 8.1 Hz), 7.08 (tt, 1H, J
) 7.1, 1.8 Hz), 6.99 (dd, 2H, J ) 8.1, 1.8 Hz), 6.93 (d, 2H, J )
8.1 Hz), 2.35 (s, 3H). (4-Meth oxy-1-p h en oxy)ben zen e (1c):
prepared from bromobenzene and 4-methoxyphenol; colorless
oil, yield 29%, bp 111 °C/1.5 Torr (lit.35 bp 105-115 °C/0.5 Torr)
Anal. Calcd for C13H12O2: C, 77.98; H, 6.04; N, 0.00; O, 15.98.
Found: C, 78.12; H, 6.00; N, 0.00. EI-MS m/e 200 (M+). 1H-
NMR (CDCl3): δ 7.30 (dd, 2H, J ) 8.9, 7.7 Hz), 7.04 (tt, 1H, J
) 7.7, 1.6 Hz), 6.99 (d, 2H, J ) 9.0 Hz), 6.95 (dd, 2H, J ) 8.9,
1.6 Hz), 6.89 (d, 2H, J ) 9.0 Hz), 3.81 (s, 3H). 4-P h en oxy-
The structures of 11i and 12i were determined by compari-
son with authentic samples which were prepared by an
alternative method,39 and those of other products were estab-
lished by comparison of the melting points and/or the spectral
properties with reported values.
In the case of the reaction of 1a in toluene, methylbiphenyls
formed (7, 8, and 9) could not be separated by silica-gel flash
(30) Olah, G. A.; Wu, A.-h. Synthesis 1991, 204-206.
(31) Suter, C. M. J . Am. Chem. Soc. 1929, 51, 2581-2585.
(32) Haeussermann, C.; Mu¨ller, A. Chem. Ber. 1901, 34, 1069-
1071.
(36) Suzumura, H. Bull Chem. Soc. J pn. 1962, 35, 108-111.
(37) Huang, R. L. J . Chem. Soc. 1958, 3725-3726.
(38) Yang, N.-C. C.; Kumler, P.; Yang, S. S. J . Org. Chem. 1972,
37, 4022-4026.
(39) Petrillo, G.; Novi, M.; Dell’Erba, C.; Tavini, C.; Berta, G.
Tetrahedron 1990, 46, 7977-7990.
(33) Ungnade, H. E.; Orwell, E. F. Organic Synthesis; J ohn Wiley
and Sons: New York, 1955; Collect. Vol. 3, pp 566-568.
(34) Russel, G. A.; Williamson, R. C., J r. J . Am. Chem. Soc. 1964,
86, 2357-2367.
(35) Forrest, J . J . Chem. Soc. 1960, 581-588.

744 J . Org. Chem., Vol. 61, No. 2, 1996
Haga and Takayanagi
chromatography. So they were obtained as a mixture and
identified by comparison of the spectral properties with those
of authentic samples. In this reaction, some unidentified
products, though in lower yields than the phenyltoluenes, were
detected by GC. They could not be isolated by column
chromatography, so we did not pursue them further.
C12H9FO: C, 76.58; H, 4.82; F, 10.09. Found: C, 76.38; H,
4.83; F, 10.23. EI-MS: m/e 188 (M+). 1H-NMR (CDCl3): δ
7.46 (dd, 2H, J ) 8.9, 5.5), 7.27 (dt, 1H, J ) 7.4, 2.0 Hz), 7.23
(dd, 1H, J ) 7.4, 2.0 Hz), 7.17 (t, 2H, J ) 8.9 Hz), 7.00 (dt,
1H, J ) 7.4, 1.2 Hz), 6.97 (dd, 1H, J ) 7.4, 1.2 Hz), 5.09 (s,
1H). 4-(4′-F lu or op h en yl)p h en ol (12g): recrystallized from
ethyl acetate and hexane, colorless prisms, mp 165-168 °C
(lit.47 mp 166 °C). Anal. Calcd for C12H9FO: C, 76.58; H, 4.82.
Found: C, 76.37; H, 4.82. EI-MS: m/e 188 (M+). 1H-NMR
(CDCl3): δ 7.48 (dd, 2H, J ) 9.0, 5.8 Hz), 7.42 (d, 2H, J ) 8.7
Hz), 7.09 (t, 2H, J ) 9.0 Hz), 6.90 (d, 2H, J ) 8.7 Hz), 4.84 (s,
1H). Meth yl 4-(2′-Hyd r oxyp h en yl)ben zoa te (11h ): re-
crystallized from dichloromethane and hexane; colorless needles,
mp 133.0-133.5 °C (lit.38 mp 129-131 °C). Anal. Calcd for
C14H12O3: C, 73.67; H, 5.30; N, 0.00; O, 21.03. Found: C,
73.91; H, 5.56; N, 0.00. EI-MS: m/e 228 (M+). 1H-NMR
(CDCl3): δ 8.14 (d, 2H, J ) 8.6 Hz), 7.59 (d, 2H, J ) 8.6 Hz),
7.26-7.31 (m, 2H), 7.02 (ddd, 1H, J ) 7.0, 7.1, 1.2 Hz), 6.97
(dt, 1H, J ) 7.9, 1.2 Hz), 5.12 (s, 1H), 3.95 (s, 3H). Meth yl
4-(4′-Hyd r oxyp h en yl)ben zoa te (12h ): recrystallized from
methanol; pale yellow prisms, mp 230.0-231.0 °C (lit.38 mp
224-225 °C). Anal. Calcd for C14H12O3: C, 73.67; H, 5.30;
N, 0.00; O, 21.03. Found: C, 73.62; H, 5.59; N, 0.00. EI-MS:
m/e 228 (M+). 1H-NMR (DMSO-d6): δ 9.66-9.75 (br s, 1H),
7.96 (d, 2H, J ) 8.7 Hz), 7.72 (d, 2H, J ) 8.7 Hz), 7.57 (d, 2H,
J ) 9.0 Hz), 6.86 (d, 2H, J ) 9.0 Hz), 3.84 (s, 3H). 4-(3′-
Hyd r oxyp h en yl)ben zen eca r bon itr ile (11i): recrystallized
from dichloromethane and hexane, pale yellow prisms, mp
111.0-112.0 °C (lit.39 mp 113.5 °C). Anal. Calcd for C13H9-
NO: C, 79.98; H, 4.65; N, 7.17; O, 8.20. Found: C, 79.70; H,
4.69; N, 7.07. EI-MS: m/e 195 (M+). 1H-NMR (CDCl3): δ 7.75
(d, 2H, J ) 8.6 Hz), 7.65 (d, 2H, J ) 8.6 Hz), 7.30 (ddd, 1H, J
) 8.5, 7.9, 2.0 Hz), 7.27 (dd, 1H, J ) 7.9, 2.0 Hz), 7.04 (dt, 1H,
J ) 7.9, 1.9 Hz), 6.94 (dd, 1H, J ) 8.5, 1.9 Hz), 5.06 (s, 1H).
4-(4′-Hyd r oxyp h en yl)ben zen eca r bon itr ile (12i): recrys-
tallized from dichloromethane and hexane, pale yellow prisms,
mp 202.5-203.0 °C (lit.39 mp 198.5-199.0 °C). Anal. Calcd
for C13H9NO: C, 79.98; H, 4.65; N, 7.17. Found: C, 79.72; H,
4.65; N, 7.16. EI-MS: m/e 195 (M+). 1H-NMR (CDCl3): δ 7.70
(d, 2H, J ) 8.2 Hz), 7.63 (d, 2H, J ) 8.2 Hz), 7.50 (d, 2H, J )
8.3 Hz), 6.94 (d, 2H, J ) 8.3 Hz), 4.98 (s, 1H).
4-Meth yl-2-p h en ylp h en ol (10b): colorless oil (lit. mp 68-
68 °C,40 65-67 °C 39). This compound did not solidify in our
experiments, but combustion analysis and spectral properties
support this structure. Anal. Calcd for C13H12O: C, 84.75;
H, 6.56; O, 8.68. Found: C, 77.97; H, 6.04. EI-MS: m/e 184
(M+). 1H-NMR (CDCl3): δ 7.37-7.54 (m, 5H), 7.10 (dd, 1H, J
) 8.8, 2.3 Hz), 7.09 (d, 1H, J ) 2.3 Hz), 6.92 (d, 1H, J ) 8.8
Hz), 5.15 (br s, 1H), 2.37 (s, 3H). 4-(4′-Meth ylp h en yl)p h en ol
(12b): pale brown powder, mp 153.0-154.5 °C (lit.41 mp 155
°C). EI-MS: m/e 184 (M+). 1H-NMR (CDCl3): δ 7.46 (d, 1H,
J ) 9.0 Hz), 7.44 (dd, 1H, J ) 8.2 Hz), 7.23 (d, 1H, J ) 8.2
Hz), 6.89 (d, 1H, J ) 9.0 Hz), 4.7 (br s, 1H), 2.39 (s, 1H).
4-Meth oxy-2 p h en ylp h en ol (10c): colorless oil (lit. bp 116
°C/0.1 Torr,42 mp 132-132.8 °C 39). This compound did not
solidify, but combustion analysis and spectral properties
support this structure. Anal. Calcd for C13H12O2: C, 77.98;
H, 6.04; O, 15.98. Found: C, 77.97; H, 6.04. EI-MS: m/e 200
(M+). 1H-NMR (CDCl3): δ 7.28-7.53 (m, 5H), 6.92 (dd, 1H, J
) 8.0, 1.1 Hz), 6.84 (dd, 1H, J ) 8.0, 2.9 Hz), 6.82 (dd, 1H, J
) 2.9, 1.1 Hz), 4.7 (br s, 1H), 3.80 (s, 3H). 2-P h en ylh yd r o-
qu in on e (10d ): recrystallized from dichloromethane and
hexane, colorless needles, mp 103.0-104.0 °C (lit.43 mp 101
°C). Anal. Calcd for C12H10O2: C, 77.40; H, 5.41; N, 0.00; O,
17.18. Found: C, 77.13; H, 5.44; N, 0.00. EI-MS: m/e 186
(M+). 1H-NMR (CDCl3): δ 7.38-7.51 (m, 5H), 6.86 (dd, 1H, J
) 9.5, 1.0 Hz), 6.75 (m, 2H), 4.84 (s, 1H), 4.43 (s, 1H). N-(4-
P h en ylp h en yl)tr iflu or oa ceta m id e (16): recrystallized from
dichloromethane and hexane; pale brown needles, mp 207-
210 °C (lit.44 mp 200-201 °C). Anal. Calcd for C14H10F3NO:
C, 63.40; H, 3.80; F, 21.49; N, 5.28; O, 6.03. Found: C, 63.33;
H, 3.97; F, 21.74; N, 5.23. EI-MS: m/e 265 (M+). 1H-NMR
(CDCl3): δ 7.85-7.94 (br s, 1H), 7.66 (d, 2H, J ) 7.9 Hz), 7.64
(d, 2H, J ) 7.9 Hz), 7.58 (dd, 2H, J ) 7.2, 1.3 Hz), 7.45 (t, 2H,
J ) 7.2 Hz), 7.36 (tt, 1H, J ) 7.3, 1.3 Hz). N-((4-Hyd r oxy-
3-p h en yl)p h en yl)tr iflu or oa ceta m id e (14): recrystallized
from dichloromethane and hexane; colorless needles, mp. 184
°C. Anal. Calcd for C14H10F3NO2: C, 59.79; H, 3.58; F, 20.27;
N, 4.98; O, 11.38. Found: C, 59.74; H, 3.53; F, 20.05; N, 4.96.
EI-MS: m/e 281 (M+). 1H-NMR (CDCl3): δ 7.72-7.82 (br s,
1H), 7.43-7.54 (m, 7H), 7.01 (d, 1H, J ) 8.5 Hz), 5.25 (s, 1H).
2,4-Dip h en ylp h en ol (10f): recrystallized from hexane; color-
less prisms, mp 89.0-89.5 °C (lit.45 mp 89 °C). Anal. Calcd
for C18H14O: C, 87.78; H, 5.73; N, 0.00; O, 6.50. Found: C,
87.86; H, 5.73. EI-MS: m/e 246 (M+). 1H-NMR (CDCl3): δ
7.57 (dd, 2H, J ) 8.9, 1.4 Hz), 7.47-7.52 (m, 6H), 7.38-7.45
(m, 3H), 7.31 (tt, 1H, J ) 8.9, 1.4 Hz), 7.05 (dd, 1H, J ) 8.0,
0.7 Hz), 5.25 (s, 1H). 4-F lu or o-2-p h en ylp h en ol (10g):
colorless oil.46 Anal. Calcd for C12H9FO: C, 76.58; H, 4.82;
F, 10.09. Found: C, 76.57; H, 4.83; F, 9.99. EI-MS: m/e 188
(M+). 1H-NMR (CDCl3): δ 7.52 (dd, 2H, J ) 8.2, 7.4 Hz), 7.45
(dd, 2H, J ) 7.4, 2.0 Hz), 7.42 (tt, 1H, J ) 8.2, 2.0 Hz), 6.90-
6.98 (m, 2H), 6.92 (dd, 1H, J ) 8.9, 5.0 Hz), 5.03 (s, 1H). 2-(4′-
F lu or op h en yl)p h en ol (11g): colorless oil (lit.47 mp 45 °C).
This compound did not solidify, but combustion analysis and
spectral properties support this structure. Anal. Calcd for
In d ep en d en t Syn th eses of 11i a n d 12i. These two
compounds were prepared according to Petrillo’s method.39
Ir r a d ia tion of 1a (GC An a lyses). Irradiation of 1a (0.01
M solution in ethanol, benzene, or toluene) was done in a
similar manner to that described above. At each irradiation
period in Figures 1 and 2, an aliquot was taken out and
subjected to GC analyses without an internal reference.
Conditions of measurement were as follows: column, H30B-
PM50 (Shinwa Chemical Industries); column diameter, 0.25
mm; injection port, 180 °C; detector, 200 °C; column temper-
ature, initially 150 °C, elevated at 7 °C/min to 200 °C.
Retention times (min) were 4.48 (1a ), 6.90 (2), 12.24 (3), 4.19
(6), 4.29 (7), 5.21 (8), 5.47 (9), 2.31 (4), and 1.71 (5). Yields of
recovered 1a and products were determined according to
calibration curves, prepared by using appropriate concentra-
tions of the samples.
In the triplet quenching experiment, 1a (0.01 M) in ethanol
in the presence of cis-1,3-pentadiene (Aldrich, 5.0 × 10-3 M)
was irradiated internally, and the resulting mixture was
analyzed by GC.
Solven t Viscosity Effect (GC An a lyses). Solutions of 1a
(0.01 M in each solvents) in 10 mL quartz test tubes were
degassed to about 10-3 Torr in three freeze-pump-thaw
cycles, sealed, and irradiated simultaneously in a merry-go-
round apparatus for 16 h. The resulting mixture was injected
into the GC column directly without an internal reference.
When 1,2-ethanediol and 1,2-propanediol were used as sol-
vents, pretreatment for GC column injection was done as
follows to remove the solvents. After irradiation, the mixture
was extracted with 60 mL of diethyl ether and with cold water
(30 mL × 3), dried over magnesium sulfate, and evaporated
to give an oil, which was diluted with ethyl acetate to 10 mL
in a volumetric flask and used for GC analyses.
(40) Endo, Y.; Shudo, K.; Okamoto, T. J . Am. Chem. Soc. 1982, 104,
6397-6405.
(41) Su¨s, O.; Mo¨ller, K.; Heiss, H. J ustus Liebigs Ann. Chem. 1956,
598, 123-138.
(42) Bowman, D. F.; Hewgill, F. R.; Kenedy, B. R. J . Chem. Soc. C
1966, 2274-2279.
(43) Borsche, W. J ustus Liebigs Ann. Chem. 1900, 312, 211-234.
(44) Sawicki, E.; Ray, F. E. J . Am. Chem. Soc. 1953, 75, 2266-
2267.
(45) Betts, B. E.; Davey, W. J . Chem. Soc. 1961, 3340-3345.
(46) Meurs, J . H. H.; Eilenberg, W. Tetrahedron 1991, 47, 705-
714.
(47) Lourak, M.; Vanderesse, R.; Fort, Y.; Caube´re, P. J . Org. Chem.
1989, 54, 4844-4848.
(48) Riddick, J . A.; Bunger, W. B. Organic Solvents, 3rd ed.; J ohn
Wiley and Sons: New York, 1970; pp 28-34.
P r ep a r a tion of [2H10]1a . A solution of [2H6]phenol (630
mg, 6.3 mmol) in 2.0 mL of methanol was prepared, and 706

Rearrangement of Diphenyl Ethers
J . Org. Chem., Vol. 61, No. 2, 1996 745
mg (6.3 mmol) of potassium tert-butoxide was carefully dis-
solved in it. Methanol was evaporated, and the residue was
dried in vacuo under heating. Removal of residual tert-butyl
alochol is essential. A pale brown solid was obtained, and to
this, 171 mg (1.7 mmol) of [2H6]phenol, cupric sulfate mono-
hydrate (18 mg), and [2H5]bromobenzene (989 mg, 6.1 mmol)
were added. The mixture was heated in an oil bath, which
had been preheated to 200-210 °C. Heating was continued
for 1 h with vigorous stirring. The blackish brown residue
was extracted with hexane (50 mL) and 1 N sodium hydroxide
(30 mL), washed with brine (30 mL × 2), and dried over
magnesium sulfate. Evaporation of the solvent gave 988 mg
of a pale yellow oil. Purification by flash chromatography
(silica gel, n-hexane as an eluent) afforded 935 mg of [2H10]1a
(5.19 mmol, yield 85% based on [2H5]bromobenzene), which
was used for cross-coupling experiments without further
purification.
Cr oss-Cou p lin g Exp er im en ts. A solution containing
[2H10]1a (189.3 mg, 1.05 mmol) and unlabeled 1a (188.3 mg,
1.11 mmol) in 170 mL of absolute ethanol was irradiated for
45 min in the same manner as described for the irradiation of
1a -i. After evaporation of the solvent, the residual yellow
oil (378 mg) obtained was flash-chromatographed (silica gel,
eluent: n-hexane-ethyl acetate, 30:1, then 20:1, and finally
8:1) to afford 285.0 mg (75%) of 1a , 41.2 mg (11%) of 2, 5.8 mg
(2%) of 4, and 23.0 mg (6%) of 3 in that order. For 1a , 2, and
3, multiscan WMIMS was conducted with a J EOL J MS AX
505 HA mass spectrometer. Samples of about 30 µg in a
quartz reservoir (volume, 2 mL) were introduced into the mass
spectrometer in the solid state. Data collection was achieved
by monitoring the relative abundance with respect to the
molecular ion at 70 eV. For each compound, three or four
samples were analyzed. The range of m/z 165-185 was
monitored at 10 s/scan. Conditions of analyses were as follows.
The ionization source was heated as follows: initially 26 °C
for 1a and 2, 50 °C for 3, and elevated at 3 °C/min for 1a and
2, 10 °C/min for 3, to obtain an appropriate ion current. In
this way, 180-220 scans (1a ), 280-360 scans (2), and 380-
440 scans (3) were collected. The data were analyzed in sets
of 20 scans to yield relative peak intensities and standard
deviations. For 1a , peak intensities at m/z 180 ([2H10]1a ), m/z
175 ([2H5]1a ), and m/z 170 (unlabeled 1a ) were used to
diagnose the scrambling of the phenyl group in 1a . Similarly,
for 2a and 3, m/e 179([2H9]2 or [2H9]3), m/z 175 ([2H5]2 or [2H5]-
3), m/z 174 ([2H4]2 or [2H4]3), and m/z 170 (unlabeled 2 or 3)
were used to diagnose the intramolecularity of the rearrange-
ment. The results are summarized in Tables 3 and 4.
Deu ter iu m a bu n d a n ce of [2H10]1a , [2H9]2, a n d [2H9]3.
Irradiation of a solution of [2H10]1a (190.5 mg, 1.06 mmol) in
170 mL of absolute ethanol (0.005 M) for 45 min gave 138.5
mg (72.7%) of 1a , 20.0 mg (10.5%) of 2, 5.8 mg (5.5%) of 4,
and 8.8 mg (4.6%) of 3. For 1a , 2, and 3, relative peak
intensities were determined by multiscan WMIMS as de-
scribed for the cross-coupling experiment. Calculations of
deuterium abundance have been described in detail in an
earlier publication.24 The results are summarized in Table 2.
Ack n ow led gm en t. N.H. is grateful to Dr. Katsumi
Tokumaru (emeritus professor at the University of
Tsukuba) and Professor Koichi Shudo (University of
Tokyo) for many helpful discussions and encouragement.
J O9517196