Reaction of phenyl-substituted o-quinodimethanes with nitric oxide. Are benzocyclobutenes suitable precursors for nitric oxide cheletropic traps?

Article abstract of DOI:10.1021/jo960573n

In order to elucidate the potential of substituted o-quinodimethanes as reagents for the trapping of nitric oxide (NO) in biological systems, the reaction of alkoxyl- and alkyl-substituted 7,8-diphenyl- and 7,7,8-triphenyl-o-quinodimethanes with nitric oxide in solution was investigated by ESR spectroscopic and UV/vis stopped-flow techniques. Photolytic decarbonylation of 1,3-diphenyl- and 1,1,3-triphenylindan-2-ones gave the corresponding phenyl-substituted benzocyclobutenes as the major products and low photostationary concentrations of o-quinodimethanes. During 266-nm laser flash photolysis (LFP) of 1,3-dimethoxy-1,3-diphenylindan-2-one and 1-methoxy-1,3,3-triphenylindan-2-one in acetonitrile, species absorbing in the 400-600 nm range were produced, which were attributed to configurational isomers of the corresponding 7,7,8,8-substituted o-quinodimethanes. The isomeric o-quinodimethanes decayed at significantly different rates, indicating a strong influence of the relative orientation of the terminal substituents on their stability. Reaction of the raw photolysates of the 2-indanones with NO produced strong ESR spectra of the corresponding cyclic nitroxide radicals, isoindolin-2-oxyls. The nitroxide radicals were generated in a two-phase process, the first, rapid phase being attributed to the reaction of NO with the photolytically formed o-quinodimethanes and the second, slow phase reflecting the reaction with small amounts of o-quinodimethanes, generated by thermal ring opening of the phenyl-substituted benzocyclobutenes and probably a direct reaction of NO with the benzocyclobutenes. The kinetics of both steps, as evaluated by stopped-flow UV/vis and ESR spectroscopy, revealed a strong dependence of the rate constants of the o-quinodimethane ± NO reaction on the substitution pattern of the o-quinodimethanes, with rate constants spanning a range of 10-4000 M^-1 s^-1. The rate constants ((0.4-7.5) x 10^-4 s^-1) for the reaction of NO with the 7,7,8,8-tetrasubstituted benzocyclobutenes are much less influenced by the substitution pattern. The utility of phenyl-substituted benzocyclobutenes as 'reservoirs' for o-quinodimethane-type nitric oxide traps is discussed.

Full text of DOI:10.1021/jo960573n

J . Org. Chem. 1996, 61, 6835-6848
6835
Rea ction of P h en yl-Su bstitu ted o-Qu in od im eth a n es w ith Nitr ic
Oxid e. Ar e Ben zocyclobu ten es Su ita ble P r ecu r sor s for Nitr ic
Oxid e Ch eletr op ic Tr a p s?
Thomas Paul, Mohammed A. Hassan,† Hans-Gert Korth,* and Reiner Sustmann*
Institut fu¨r Organische Chemie, Universita¨t - GH Essen, D-45117 Essen, Germany
David V. Avila
National Research Council of Canada, Ottawa, Ontario, K1A 0R6, Canada
Received March 27, 1996X
In order to elucidate the potential of substituted o-quinodimethanes as reagents for the trapping
of nitric oxide (NO) in biological systems, the reaction of alkoxyl- and alkyl-substituted 7,8-diphenyl-
and 7,7,8-triphenyl-o-quinodimethanes with nitric oxide in solution was investigated by ESR
spectroscopic and UV/vis stopped-flow techniques. Photolytic decarbonylation of 1,3-diphenyl- and
1,1,3-triphenylindan-2-ones gave the corresponding phenyl-substituted benzocyclobutenes as the
major products and low photostationary concentrations of o-quinodimethanes. During 266-nm laser
flash photolysis (LFP) of 1,3-dimethoxy-1,3-diphenylindan-2-one and 1-methoxy-1,3,3-triphenylin-
dan-2-one in acetonitrile, species absorbing in the 400-600 nm range were produced, which were
attributed to configurational isomers of the corresponding 7,7,8,8-substituted o-quinodimethanes.
The isomeric o-quinodimethanes decayed at significantly different rates, indicating a strong influence
of the relative orientation of the terminal substituents on their stability. Reaction of the raw
photolysates of the 2-indanones with NO produced strong ESR spectra of the corresponding cyclic
nitroxide radicals, isoindolin-2-oxyls. The nitroxide radicals were generated in a two-phase process,
the first, rapid phase being attributed to the reaction of NO with the photolytically formed
o-quinodimethanes and the second, slow phase reflecting the reaction with small amounts of
o-quinodimethanes, generated by thermal ring opening of the phenyl-substituted benzocyclobutenes
and probably a direct reaction of NO with the benzocyclobutenes. The kinetics of both steps, as
evaluated by stopped-flow UV/vis and ESR spectroscopy, revealed a strong dependence of the rate
constants of the o-quinodimethane + NO reaction on the substitution pattern of the o-quino-
dimethanes, with rate constants spanning a range of 10-4000 M-1 s-1. The rate constants ((0.4-
7.5) × 10-4 s-1) for the reaction of NO with the 7,7,8,8-tetrasubstituted benzocyclobutenes are much
less influenced by the substitution pattern. The utility of phenyl-substituted benzocyclobutenes
as “reservoirs” for o-quinodimethane-type nitric oxide traps is discussed.
In tr od u ction
develop an alternative method, based on the reaction of
NO with o-quinodimethane derivatives 1, to yield per-
sistent, cyclic nitroxide radicals 2, reaction 1, which can
easily be detected and quantitated by ESR spectroscopy.
During the past decade, nitric oxide, NO, has emerged
as an important signal molecule in living organisms,
involved, for example, in the regulation of vascular tone,
acting as a messenger in the central nervous system and
as mediator in the immune system.1 For the monitoring
and quantitation of NO formation in biological systems
a number of physicochemical and biochemical analytical
methods have been developed.2 Unfortunately, none of
these methods fulfills the desirable requirements for
general use with the variety of biological specimens, viz.
specifity for NO, high sensitivity, tolerance for the various
physiological conditions, easy structural modification,
and the potential to monitor NO formation with temporal
and spacial resolution. In order to overcome at least
some of these obstacles, we3-5 recently have begun to
(1)
Among the series of our so-called “nitric oxide chele-
tropic traps” (NOCTs),6 the prototypal tetramethyl de-
rivative NOCT-13 and the trimethylindenyl derivative
NOCT-44 have been sucessfully applied in the trapping
of NO from cultured macrophages under physiological
conditions. Further, nonbiological applications of NOCTs
included the trapping of NO produced from NO-releasing
† Permanent address: Ain Shams University, Cairo, Egypt.
* To whom correspondence should be addressed. Phone: 0049-201-
183-3097 or 0049-201-183-3148. Fax: 0049-201-183-3096. E-mail:
sustmann@oc1.orgchem.uni-essen.de and hgk@oc1.orgchem.uni-essen.
de.
(3) Korth, H.-G.; Ingold, K. U.; Sustmann, R.; de Groot, H.; Sies, H.
Angew. Chem. 1992, 107, 915; Angew. Chem., Int. Ed. Engl. 1992, 31,
891.
(4) Korth, H.-G.; Sustmann, R.; Lommes, P.; Paul, T.; Ernst, A.; de
Groot, H.; Hughes, L.; Ingold, K. U. J . Am. Chem. Soc. 1994, 116, 2767.
(5) Korth, H.-G.; Sustmann, R.; Lommes, P.; Paul, T.; Ernst, A.;
Ingold, K. U.; Hughes, L.; de Groot, H.; Sies, H. Cheletropic Spin Traps
for Nitric Oxide (NOCTs) in Biological Systems. In The Biology of Nitric
Oxide; Moncada, S., Feelisch, M., Busse, R., Higgs, E. A., Eds.; Portland
Press: London, 1994; Vol. 4, pp 220-224.
X Abstract published in Advance ACS Abstracts, September 1, 1996.
(1) (a) The Biology of Nitric Oxide; Marletta, M. A., Ed.; Portland
Press: London, 1992; Vols. 1 and 2. (b) The Biology of Nitric Oxide;
Moncada, S., Feelisch, M., Busse, R., Higgs, E. A., Eds.; Portland
Press: London, 1994; Vols. 3 and 4.
(2) (a) Archer, S. FASEB J . 1993, 7, 349. (b) Cristol, J . P.; Gue´rin,
M. C.; Torreilles, J . C. R. Acad. Sci. Paris, Life Sci. 1994, 317, 549-
560. (c) Methods in Nitric Oxide Research; Feelisch, M., Stamler, J .,
Eds.; Wiley: Chichester, 1996.
S0022-3263(96)00573-7 CCC: $12.00 © 1996 American Chemical Society

6836 J . Org. Chem., Vol. 61, No. 20, 1996
Paul et al.
compounds (nitrovasodilators)8,9 and in photochemical
reactions.10
Sch em e 1
However, a serious disadvantage of NOCT-1 and other
tetraalkyl-substituted NOCTs is the short lifetime (e.g.,
ca. 120 s for NOCT-1 at 37 °C4,11) at physiological
temperature, a result of a rapid 1,5-hydrogen shift
reaction. In marked contrast, the indenyl compound
NOCT-4 was found to be very long-lived (several days)
at 37 °C under similar conditions4 though its concentra-
tion was very low. The increased lifetime of NOCT-4 has
been tentatively attributed to the extension of the delo-
calized π-system by the indenyl substituent, thus ther-
modynamically stabilizing the o-quinodimethane struc-
ture.
We therefore proposed that also o-quinodimethanes
carrying phenyl substituents at the exocyclic double
bond(s) should exhibit longer lifetimes than alkyl-
substituted ones, which might render them more suitable
for biological applications. In addition, by replacing the
terminal alkyl groups with alkoxy substituents an in-
crease of the lifetime of o-quinodimethanes was expected
due to suppression of possible 1,5-hydrogen shift reac-
tions.
are photolabile themselves, leading only to low photo-
stationary concentrations in the range of 10-3-10-6 M.
Such concentrations, however, are still sufficient to
generate strong ESR signals of the nitroxides 2.
General modes of decay of o-quinodimethanes13 (Scheme
1) include electrocyclic ring closure to substituted ben-
zocyclobutenes (path A), [4 + 4] cycloaddition to give
dibenzocyclooctadienes (path B), formation of spiro com-
pounds via [4 + 2] cycloaddition (Diels-Alder reaction;
path C), 1,5-hydrogen shift if suitable hydrogen-donating
substituents (alkyl, hydroxyl) are present (path D), and
electrocyclic ring closure to 4a,10-dihydroanthracene
derivatives in the case of R ) phenyl (path E).
Since sterically congested benzocyclobutenes in solu-
tion may exist in thermal equilibrium with the corre-
sponding o-quinodimethanes via electrocyclic ring opening/
ring closure (reversibility of path A), we hypothesized
that such benzocylobutenes may serve as convenient
“reservoirs” for the actual NO trap. Therefore, we
synthesized a series of phenyl-substituted 2-indanones
and studied their photochemically induced decarbonyla-
tion and the reaction of their photoproducts with nitric
oxide. Here, we present our results on the 1,3-diphenyl-
2-indanones 3a -e and 1,1,3-triphenyl-2-indanones 3h -
l.
So far, we used as a convenient route to o-quino-
dimethanes the photolysis of the corresponding 2-in-
danones 3,12-14 reaction 1, because a variety of substi-
tuted indanones can easily be prepared from readily
available starting materials.
Unfortunately, due to their strong UV/vis absorption
at wavelengths >350 nm o-quinodimethanes of type 1
(6) Although the cyclic nitroxides 2 represent the products of a
formal cheletropic reaction, we do not imply that the mechanism for
their formation is a truly cheletropic one in the sense of the Wood-
ward-Hoffman rules, i.e., to proceed in a concerted fashion. Indeed,
at least for the reaction of the unsubstituted o-quinodimethane in the
gas phase there is evidence that the reaction with NO occurs in a
stepwise manner.7
(7) Roth, W. R.; Rekowski, V.; Bo¨rner, S.; Quast, M. Liebigs Ann.
1996, 409-430.
(8) Weber, H.; Grzesiok, A.; Sustmann, R.; Korth, H.-G. Z. Natur-
forsch. 1994, 49B, 1041-50.
(9) Ioannidis, I.; Ba¨tz, M.; Paul, T.; Korth, H.-G.; Sustmann, R.; de
Groot, H. Biochem. J ., in press.
(10) Mark, G.; Korth, H.-G.; Schuchmann, H.-P.; von Sonntag, C.
J . Photochem. Photobiol. A, in press.
(11) (a) de Fonseka, K. K.; McCullough, J . J .; Yarwood, A. J . J . Am.
Chem. Soc. 1979, 101, 3277-3282. (b) Wintgens, V.; Netto-Ferreira,
J . C.; Casal, H. L.; Scaiano, J . C. J . Am. Chem. Soc. 1990, 112, 2363-
2367. (c) Redmond, R. W.; Harwig, C. W.; Scaiano, J . C. Photochem.
Photobiol. A 1992, 68, 255-259.
Resu lts
(12) (a) Netto-Ferreira, J . C.; Wintgens, V.; Scaiano, J . C. Tetrahe-
dron Lett. 1989, 30, 6851-6854. (b) Scaiano, J . C.; Wintgens, V.; Netto-
Ferreira, J . C. Pure Appl. Chem. 1990, 62, 1557-1564.
(13) (a) McCullough, J . J . Acc. Chem. Res. 1980, 13, 270-276. (b)
Fishwick, C. W. G.; J ones, D. W. In The Chemistry of Quinoid
Compounds; Patai, S., Rappoport, Z., Eds.; Wiley: Chichester, 1988;
Vol. 2, pp 403-453. (c) Charlton, J . L.; Alauddin, M. M. Tetrahedron
1987, 43, 2873-2889.
(14) (a) Quinkert, G. Pure Appl. Chem. 1964, 9, 607-621. (b)
Grellmann, K. H.; Palmowski, J .; Quinkert, G. Angew. Chem. 1971,
83, 209-210; Angew. Chem., Int. Ed. Engl. 1971, 10, 196-197. (c)
Quinkert, G.; Palmowski, J .; Lorenz, H. P.; Wiersdorff, W. W.; Finke,
M. Angew. Chem. 1971, 83, 210-212; Angew. Chem., Int. Ed. Engl.
1971, 10, 197-198. (d) Quinkert, G.; Wiersdorff, W. W.; Finke, M.;
Opitz, K.; von der Haar, F.-G. Chem. Ber. 1968, 101, 2302.
Syn t h esis of P h en yl-Su b st it u t ed 2-In d a n on es.
1,3-Diphenyl-substituted 2-indanones 3a -g were conve-
niently prepared from ninhydrin monohydrate by stan-
dard procedures (Scheme 2).
According to the 1H NMR spectrum (only one OCH3
signal) ketal 4 was obtained in >99% yield as the rac
diastereomer. During hydrolytic cleavage of 4 meso/ rac
isomerization occurred, producing mainly meso-3f, which
could be isolated in pure form. Further conversion of
meso-3f to compounds 3a -d did not affect the stereo-
chemistry.

Reaction of Phenyl-Substituted o-Quinodimethanes with NO
J . Org. Chem., Vol. 61, No. 20, 1996 6837
Sch em e 2
Sch em e 3
Sch em e 4
The assignment of stereochemistry of the 1,3-diphen-
ylindan-2-ones 3a -e by 1H NMR spectroscopy was based
on the LiAlH4 reduction of meso-3a to the corresponding
hydroxy compounds 5, reaction 2, similar to the proce-
dure reported by Quinkert et al.15 The 1H NMR spectrum
of the reduction mixture showed two sets of signals in
an intensity ratio of 1:1.3, in agreement with a mixture
of the diastereomers of meso-5. From rac-3a only one
set of NMR signals would have been expected, since the
corresponding groups of the diastereomers of rac-5,
reaction 3, are all chemically equivalent.
(2)
(3)
1,1,3-Triphenyl-indan-2-ones 3h -l were obtained from
2,3-diphenylinden-1-one (Scheme 3).
P h otolysis of 2-In d a n on es. The photolytic decar-
bonylation of the phenyl-substituted 2-indanones 3f,g,k
has been investigated by the groups of Quinkert14 and
Scaiano.12 Quinkert also investigated the 1,1,3,3-tet-
raphenyl-2-indanone (3m ; R1 ) R2 ) Ph). With regard
to the photoproducts and transients the results reported
in this study are generally in line with their findings.
All 2-indanones of this study were photolyzed in deoxy-
genated solution at concentrations of 5 × 10-2 M with λ
) 300 nm light. The resulting photolysates were ana-
lyzed by 1H NMR, 13C NMR, and UV/vis spectroscopy and
in part also by HPLC (see Experimental Section). The
only products of the photolytic decarbonylation (Scheme
also found. o-Quinodimethanes 1 could only be detected
UV/vis-spectroscopically by their characteristic absorp-
tions in the 400-600 nm wavelength range.12,14,16,17 In
agreement with literature reports,12,14 no [4 + 2] or [4 +
4] cycloadducts (paths B and C of Scheme 1) were
detected within the sensitivity limit (e1%) of our 300
1
4) that could be detected by conventional 300-MHz H
NMR spectroscopy were the corresponding benzocy-
clobutenes 6 and, where possible, the products 9 of a 1,5-
hydrogen shift. In some cases small amounts of dihy-
droanthracene 7 and/or anthracene derivatives 8 were
(16) (a) Porter, G.; Tchir, M. F. J . Chem. Soc. A 1970, 1372-1373.
(b) Porter, G.; Tchir, M. F. J . Chem. Soc. Chem., Chem. Commun. 1971,
3772-3777.
(17) (a) J ones, D. W.; Kneen, G. J . Chem. Soc., Perkin Trans. 1 1975,
171-174. (b) Holland, J . M.; J ones, D. W. J . Chem. Soc. C 1971, 608-
612.
(15) Quinkert, G.; Lorenz, H.-P.; Wiersdorff, W. W. Chem. Ber. 1969,
102, 1597-1605.

6838 J . Org. Chem., Vol. 61, No. 20, 1996
Ta ble 1. P h otolysis (300 n m ) of 1,3-Dip h en yl- a n d 1,1,3-Tr ip h en ylin d a n -2-on es (1H NMR a n a lysis)a
Paul et al.
indanone
R
photolysis
time/h
conversion
(%)
benzocyclobutene 6
(meso/ rac) (mol %)
1,5-H shift
product 9 (mol %)
solvent
T/°C
3a b
CD3CN
C6D6
CD3CN
-7
12
10
4
3.5
4
100
100
87
95/5
93/7
95/5
OMe
3bb
OCH2COOEt
3cb
CD3CN
D2O
10
10
4
4
100
100
100/0
100/0
OCH2COOH
3d b
OCH2COONa
3ec
Me
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
C6D6
-10
10
20
6
-20
15
5
15
17
4
2
7
9
31
68
40
33
65
100
96
33/67
48/52
47/53
3h
OH
3i
OMe
100
100
100
96d
100
10
10
24
5
3j
OCH2COOEt
3k
H
3l
C6D6
C6D6
C6D6
10
10
10
1
48
2
15
100
14
100
95e
57
43
Me
a
b
d
Products <1% not considered. >99% meso. c 36/64 meso/ rac. 4% 9,10-diphenylanthracene (8m ) by UV/vis analysis. e 5% 9,10-
diphenylanthracene (8m ).
1
MHz H NMR analysis procedure or by HPLC (UV/vis)
analysis. Conditions of the photolysis experiments and
1H NMR-detected product yields are collected in Table
1.
In most cases, the detected benzocyclobutene and
o-quinodimethane photoproducts appeared to be ther-
mally unstable, converting slowly to 4a,10-dihydroan-
thracene 7 and further to anthracene derivatives 8.
Investigations focused on these reactions will be reported
in a forthcoming paper.
meso-1,3-Dim eth oxy-1,3-diph en ylin dan -3-on e (3a).
In CD3CN solution a complete conversion of meso-3a was
achieved after 4 h photolysis at -7 °C to yield a 95:5 ratio
of benzocyclobutenes meso-6a and rac-6a (Scheme 4).
meso-6a could be isolated in pure form; its absolute
stereochemistry was determined by X-ray diffraction
analysis (to be reported in a following paper). 1H NMR
signals that might be assigned to the o-quinodimethanes
1a (Scheme 4) could not be resolved. The orange-colored
F igu r e 1. Time-resolved UV/vis spectra obtained during 266-
nm LFP of a 10-2 M solution of 3a in acetonitrile as observed
250 ns (b) and 1.6 µs (4) after laser excitation. Insets: decay
traces of the absorptions at 320 nm for the first (a) and second
(b) phase.
photolysate showed a broad UV/vis absorption at λmax
)
470 nm (see below) which was persistent for several days
at room temperature. We assign this absorption to one
of the three possible isomeric o-quinodimethanes 1a
because of the following arguments: (i) its similarity to
the UV/vis absorption of other phenyl-substituted o-
quinodimethanes,12,14,16,17 (ii) the decay of the absorption
and concomittant formation of an ESR-detectable nitrox-
ide radical when the photolysate was reacted with nitric
oxide (see below), (iii) its irreversible disappearance on
heating the sample to 60 °C, and (iv) the results of the
laser flash photolysis (LFP) experiments described below.
Assuming an extinction coefficient for o-quinodimeth-
anes in the range of ꢀ(1) ) 104 M-1 cm-1 (as found for
7,7,8,8-tetraphenyl-o-quinodimethane (1m )14d, ꢀ ) 9170
M-1 cm-1, 7,8-diphenyl-o-quinodimethane (1g)14b, ꢀ )
10 600 M-1 cm-1, or 7,7,8,8-tetramethyl-o-quinodimethane
(NOCT-1)11c, ꢀ ) 13 700 M-1 cm-1) we estimated from the
stationary UV/vis absorption at λ ) 470 nm a concentra-
tion of 1a on the order of 10-4 M. Thus, it is not
surprising that we were not able to detect 1a by conven-
anthracene derivative 8a , based on the similarity of its
UV/vis spectrum to other 9,10-substituted anthra-
cenes12a,14d,16 (compare, e.g., 9,10-diphenylanthracene
(8m ), below).
Further support for the formation of isomeric o-
quinodimethanes 1a was provided by LFP experiments.
After excitation of 3a (10-2 M in acetonitrile) with 266-
nm laser light, a transient absorption spectrum charac-
terized by a broad band in the 380-600 nm range (λmax
ca. 460 nm) and sharper, more intense bands in the 300-
360 nm range were observed within the width (10 ns) of
the laser pulse (Figure 1a). The 460-nm band and the
absorptions > 330 nm decayed in a biphasic process, both
steps showing first-order kinetics with lifetimes of τ )
0.16 µs (k ) 7.3 × 106 M-1 s-1) and 24 µs (k ) 3.3 × 104
M-1 s-1 ), respectively (Figure 1a, insets). A weak
absorption in the 430-600 nm range and a stronger,
sharp band at λmax ) 340 nm remained. We attribute
the two transient and the persistent absorptions to the
three isomeric o-quinodimethanes 1a because of the
characteristic wavelength range, their relatively (long)
1
tional H NMR spectroscopy.
By HPLC (UV/vis detection) an additional trace com-
ponent was detected, most likely the 9,10-substituted

Reaction of Phenyl-Substituted o-Quinodimethanes with NO
J . Org. Chem., Vol. 61, No. 20, 1996 6839
lifetimes, and the fact that they could not be “quenched”
by dissolved oxygen, as would be expected for biradical
species.
Absorptions that might be related to the intermediate
triplet acyl alkyl biradical 3a (B1) (Scheme 4) and the
decarbonylated dialkyl biradical 3a (B2) (which can be
regarded as a triplet state of 1a ) could not be observed,
indicating that decarbonylation and rotational isomer-
ization are very fast on the nanosecond time scale of our
LFP experiments.4,12 Likewise, significant formation of
dihydroanthracene (7a ) or anthracene species (8a ), either
in the photolytic process or in the decay of transient 1a ,
can be excluded; such compounds are characterized by
absorptions in the 350-400 nm range.14-16
m eso-1,3-Bis[(e t h oxyca r b on yl)m e t h oxy]-1,3-d i-
p h en ylin d a n -2-on e (3b). The photolysis of meso-3b in
CD3CN at 10 °C gave very similar results to 3a , only the
decomposition occurred somewhat slower and the ratio
of the benzocyclobutene products meso-6b and rac-6b was
slightly changed to 93:7.
meso-1,3-Bis(car boxym eth oxy)-1,3-diph en ylin dan -
2-on e (3c) a n d Disod iu m m eso-1,3-Bis(ca r b oxy-
latom eth oxy)-1,3-diph en ylin dan -2-on e (3d). The pho-
tolysis of 3c in CD3CN and of 3d in buffered D2O
(phosphate buffer, pH 7.5) at 10 °C was completed within
4 h with almost quantitative formation of meso-6c and
meso-6d , respectively. The aqueous solution of 3d be-
came turbid during the photolysis, probably because
traces of acid 3c (which is almost insoluble in water) were
formed.
F igu r e 2. Time-resolved UV/vis spectra obtained during 266-
nm LFP of a 10-2 M solution of 3i in acetonitrile as observed
200 ns (b) and 1.6 µs (4) after laser excitation. Insets: decay
traces of the absorptions at 320 nm for the first (a) and second
(b) phase.
hydroxyl compound 3h . There is some evidence for the
latter interpretation from work by Porter and Tchir,16
who showed that the decay of o-quinodimethanes deriv-
ing from 2,4-dimethylbenzophenone is faster in protic
than in aprotic solvents.
1-Meth oxy-1,3,3-tr ip h en ylin d a n -2-on e (3i). Pho-
tolysis of 3i in CD3CN for 7 h (65% conversion) gave
benzocyclobutene 6i as the only 1H NMR-detectable
product. When the photolysis was driven to complete
conversion of 3i (24 h) small amounts (ca. 4%) of 9,10-
diphenylanthracene (8m ) were also detected by HPLC
and UV/vis spectroscopy by comparison with authentic
material. The yellow photolysate showed only a very
weak absorption in the 400-500 nm region, from which
we estimated, again assuming ꢀ(1) ) 104 M-1 cm-1, a low
concentration of o-quinodimethanes 1i in the range of
10-5 to 10-6 M. Attempts to purify product 6i by
chromatography or crystallization were not successful
due to partial decomposition of the compound during the
purification procedures.
1,3-Dim eth yl-1,3-d ip h en ylin d a n -2-on e (3e). The
photolysis of this ketone has already been studied.14c The
photolytic decomposition of a 1:1.1 meso/ rac mixture of
3e in CD3CN at 20 °C was less effective than with the
above alkoxy-substituted 2-indanones; e.g., after 17 h
68% were converted, affording only meso- and rac-6e in
approximately the same ratio as the starting material.
Somewhat unexpected, the product of a 1,5-hydrogen
1
shift reaction could not be detected by H NMR spectros-
copy. This is in agreement with observations made by
Quinkert,14c who found evidence that the 1,5-hydrogen
shift only occurred after complete decarbonylation of 3e.
No detailed analysis of the photolytic decomposition
of 1,3-dihydroxy-1,3-diphenylindan-2-one (3f) and 1,3-
diphenylindan-2-one (3g) was performed because of their
unfavorable properties for our purposes. Compound 3f
is very soluble in organic solvents and produces a variety
of products, primarily 1,3-diphenylisobenzofuran, when
photolyzed in DMSO solution; 3g would produce only
short-lived o-quinodimethanes and a thermally too stable
benzocyclobutene.18 Furthermore, reaction of the pho-
tolysates of both ketones produced only very weak,
rapidly decaying ESR signals of nitroxide radicals not
compatible with the structures expected from 1g and
1h .18
1-Hyd r oxy-1,3,3-tr ip h en ylin d a n -2-on e (3h ). After
photolysis of 3h in CD3CN at 6 and -40 °C, respectively,
the only NMR-detectable product was 2-diphenylmeth-
ylbenzophenone (9h ), the formal 1,5-hydrogen shift prod-
uct of (E)-1h (Scheme 4). If 9h is indeed formed in a
concerted 1,5-hydrogen shift process then one has to
assume that either only (E)-1h is produced from 3h or
that (E)-1h and (Z)-1h exist in a rapid equilibrium
(probably via 6h ). On the other hand, 9h simply may
have been formed by a proton-catalyzed keto-enol tau-
tomerization, e.g., by traces of water or by the starting
Ketone 3i was also subjected to LFP experiments.
Very similar to 3a (see above), 266-nm LFP of a 10-2
M
solution of 3i in acetonitrile produced within the width
of the laser pulse a transient absorption spectrum
characterized by a broad band in the 400-600 nm range
with maxima at ca. 460 and 500 nm and stronger
absorptions at 330 and 280 nm (Figure 2a). The 460,
500, and 280-nm absorptions, which we attribute to the
isomeric o-quinodimethanes 1i, again decayed in a two-
phase process with first-order rate constants of k ) 5 ×
106 s-1 and 4.5 × 104 s-1 for the first and second stage,
respectively (Figure 2a, inset). The decay rates were the
same in nitrogen- and oxygen-saturated solution. Decay
of the o-quinodimethane bands, however, did not occur
completely (Figure 2b). Since in this system only two
isomeric o-quinodimethanes 1i are possible, this would
imply the onset of an equilibrium 6i h 1i.
1-[(E t h oxyca r b on yl)m et h oxy]-1,3,3-t r ip h en ylin -
d a n -2-on e (3j). This ketone was decarbonylated to
about 96% on 5 h photolysis at 10 °C, giving almost
exclusively benzocyclobutene 6j and traces of 9,10-
diphenylanthracene. By HPLC another trace component
was detected, most probably the corresponding dihy-
droanthracene derivative 7j.
(18) Paul, T. Diploma thesis, Universita¨t - GH Essen, 1993.

6840 J . Org. Chem., Vol. 61, No. 20, 1996
Paul et al.
Ta ble 2. ESR Sp ectr oscop ic Da ta a a t 20 ( 2 °C of Cyclic Nitr oxid e Ra d ica ls fr om th e Rea ction of 7,8-Dip h en yl- a n d
7,7,8-Tr ip h en yl-o-qu in od im eth a n es w ith Nitr ic Oxid e
hyperfine splittings/mT
nitroxide
solvent
g factor
a(14N)
a(13C)
a(15N)
2a
acetonitrile
benzene
acetonitrile
2.006 29(3)
2.006 34(3)
2.006 33(3)
1.185(3)
1.150(3)
1.168(3)
0.43(1) (4C)
0.42(1) (4C)
0.40(1) (2C)
0.41(1) (2C)
0.39(1) (4C)
0.40(1) (2C)
0.42(1) (2C)
0.54(1) (4C)
n.a.c
1.66(1)
1.61(1)
1.62(1)
2b
2c
2d
acetonitrile
buffer pH7.4
2.006 27(4)
2.006 03(4)
1.173(3)
1.292(5)
1.65(1)
1.65(1)
2e(1)b
2e(2)
2h
acetonitrile
acetonitrile
acetonitrile
acetonitrile
2.006 02(5)
2.005 93(5)
2.006 16(3)
2.006 15(3)
1.422(5)
1.405(5)
1.267(4)
1.267(4)
1.995(5)
n.a.
n.a.
n.a.
2i
0.52(1) (3C)
0.72(1) (1C)
n.a.
1.79(1)
DMSO
benzene
2.006 06(4)
2.006 17(3)
1.264(4)
1.234(3)
n.a.
1.738(5)
0.51(1) (3C)
0.72(1) (1C)
0.52(1) (3C)
0.63(1) (1C)
0.56(1) (2C)
0.68(1) (3C)
0.53(1) (3C)
0.72(1) (2C)
2j
2k
2l
acetonitrile
benzene
2.006 15(3)
2.006 15(3)
2.006 06(3)
1.252(3)
1.790(5)
n.a.
1.385(3)
1.851(3)d
1.380(3)
benzene
1.94(1)
a
b
d
Estimated errors in the last digit given in parentheses. Two radicals, initial ratio 73:27. c Not analyzed. â-Hydrogen splitting.
1,1,3-Tr ip h en ylin d a n -2-on e (3k ). Photolytic decar-
bonylation of 3k has been investigated by Scaiano et al.12
who reported the formation of 7,7,8-triphenylbenzocy-
clobutene (6k ) and 9,10-diphenylanthracene (8m ) in
various ratios, depending on the light source (laser or
mercury lamp). In agreement, we observed a complete
destruction of 3k on 48 h photolysis in benzene-d6 at 10
°C with formation of 6k and 9,10-diphenylanthracene
(8m ) in a 95:5 ratio.
(4)
1-Meth yl-1,3,3-tr iph en ylin dan -2-on e (3l). The prod-
1
ucts identified by H NMR after 8 h photolysis of 3l in
benzene-d6 at 10 °C (72% conversion) were benzocy-
clobutene 6l (68%) and R-phenyl-o-diphenylmethylsty-
rene (9l) (32%); i.e., some 1,5-hydrogen shift had occurred
(in contrast to the dimethyl compound 3e). The given
percentage of 9l, however, might not reflect the true
relative ratio of the stereoisomers of the intermediate
o-quinodimethanes 1l as produced photolytically, because
a photolytically induced back-reaction (1,5-hydrogen
shift) of the styrene compound 9l to reform 1l cannot be
excluded.19
R ea ct ion of t h e P h ot olysis P r od u ct s of 1,3-Di-
p h en yl- a n d 1,1,3-Tr ip h en ylin d a n -2-on es w ith Ni-
tr ic Oxid e. As mentioned above, even small concentra-
tions (10-4-10-6 M) of o-quinodimethanes are sufficient
to produce strong ESR signals of cyclic nitroxides 2
(isoindolin-2-oxyls) on reaction with nitric oxide.3,4 The
ability of the phenyl-substituted o-quinodimethanes 1a -
e,h -l produced, or expected to be produced, in the
photolytic decompositions of ketones 3a -e,h -l was
checked by reaction of the raw photolysates with solu-
tions of nitric oxide, reaction 4, in the same solvent and
ESR-spectroscopic detection of the resulting nitroxide
radicals. From all photolysates of ketones 3, except 3h ,
good-to-excellent ESR were obtained (note that in most
cases the 13C and 15N isotope satellite lines could be easily
analyzed), indicating that o-quinodimethanes 1 had
reacted with NO. The ESR spectroscopic data (Table 2)
are in full agreement with the cyclic structure of nitrox-
ides 2;4,20 in particular, the analysis of the 13C satellite
hyperfine lines confirms their structure.
The kinetics of formation of nitroxides 2a -e,h -l (Table
3) was measured by following the time evolution of the
ESR spectra. For comparison with the o-quinodimethane
concentrations estimated from the UV/vis spectra, the
absolute radical concentrations given in Table 3 were
determined via double integration of the ESR spectra.
With one exception (2k ), the ESR signal intensities of
nitroxides 2 measured “initially” (i.e., after 2-5 min, the
time to prepare the sample and to tune the spectrometer)
after mixing the photolysates of 3 with NO solution
increased slowly with time (within hours). However, the
rates of this increase would not account for the “initial”
radical concentration. When the formation of radicals 2
was followed kinetically under first-order conditions it
became evident that the nitroxides were formed in a
biphasic process; i.e., after an initial “jump” (on the time
scale of minutes) their signal intensities increased steadily
by a slower process to achieve either a constant “plateau”
value or going through a maximum. Extrapolation of the
least-squares fit to the concentration-time data of the
“slow” process to zero time gave a radical concentration
that we set equal to the concentration of the o-quino-
dimethane derivatives in the respective photolysate
(Table 3, column 9) at the time of reaction with NO. This
procedure seems to be justified, since, with the exception
of 2e, the formation of only one nitroxide radical was
monitored by ESR in the respective trapping experi-
ments.
(20) Forrester, A. R. In Landolt-Bo¨rnstein, New Series, Magnetic
Properties of Free Radicals; Fischer, H., Hellwege, K. H., Eds.;
Springer: Berlin, 1979; Vol. 9, Part c1; Vol. 17, Parts d1, d2.
(19) Hornback, J . M.; Russell, R. D. J . Org. Chem. 1982, 47, 4285-
4291.

Reaction of Phenyl-Substituted o-Quinodimethanes with NO
J . Org. Chem., Vol. 61, No. 20, 1996 6841
Ta ble 3. Kin etic Da ta for th e Rea ction of 7,8-Dip h en yl- a n d 7,7,8-Tr ip h en yl-o-qu in od im eth a n es a n d 7,8-Dip h en yl- a n d
7,7,8-Tr ip h en ylben zocyclobu ten es w ith Nitr ic Oxid e in Solu tion
106c(1)b/
ksf(1)c/
k2sf(1)d/
104kESR(2)e/
106c(2)0 /
106c(2)max
/
f
g
compd
solvent
T/°C
λa/nm
mol L-1
s-1
M-1 s-1
s-1
mol L-1
mol L-1
1a /6a
1a /6a
6a h
1b/6b
1c/6c
1e/6e
1i/6i
1j/6j
acetonitrile
benzene
acetonitrile
acetonitrile
acetonitrile
benzene
acetonitrile
acetonitrile
benzene
23
23
23
24
23
23
24
23
23
23
470
470
140
6
0.061 ( 0.01
0.16 ( 0.04
10 ( 1
20 ( 5
1.8
0.4
4.6
1.0
n.a.i
2.7
j
1.0
k
7.5
570
17
0
32
20
14
346
52
15
71
1335
623
1958
869
54
490
450
490
490
6
4
4
0.6
1.6 ( 0.1
1.1 ( 0.4
15.0 ( 3.2
0.39 ( 0.10
n.a.
17.0 ( 3.9
29.0 ( 2.7
230 ( 15
160 ( 60
2140 ( 460
60 ( 15
41
1246
660
15k
1k /6k
1l/6l
490
490
18
90
2430 ( 560
4140 ( 390
benzene
125
a
b
Wavelength of stopped-flow kinetic measurement. Estimated concentration of o-quinodimethane from change of the long-wavelength
absorption, assuming an extinction coefficient of ꢀ ) 104 M-1 cm-1
.
c Experimental pseudo-first-order rate constant from spectrophotometric
d
stopped-flow experiment. Estimated second-order rate constant (rounded to the next decimal number) for 1 + NO reaction, calculated
from ksf(1) and 7 × 10-3 M NO concentration. e Rate constant for the slow growth of the ESR signal of nitroxide 2. f Extrapolated “initial”
g
h
(t ) 0) concentration of nitroxide 2 for slow growth. Maximum concentration of nitroxide 2 produced in the slow process. Purified
i
j
k
meso-6a employed. Not analyzed, sigmoid concentration/time profile. Linear growth. Constant ESR signal, no growth.
The rates of reaction of 1a -e,h -l with NO were
determined by spectrophotometric stopped-flow measure-
ments, monitoring the decay of the characteristic o-
quinodimethane absorptions in the 400-600 nm range.
The concentration of our phenyl-substituted o-quino-
dimethanes were estimated from the maxima of these
UV/vis absorptions by assuming an extinction coefficient
of ꢀ ) 104 M-1 cm-1. In some cases, however, the 400-
600 nm absorptions did not decay to zero values (see, e.g.,
Figure 7, below), indicating that also other, nonreactive
products exhibited some absorptions in this region.
Therefore, the estimated o-quinodimethane concentra-
tions given in Table 3, column 5, were calculated only
F igu r e 3. ESR spectrum of 2a in acetonitrile at 20 °C,
recorded 15 min after mixing of the photolysate of 3a with
NO solution.
from the change of the UV/vis absorptions after reaction
with excess NO. The sometimes large difference in the
o-quinodimethane concentration estimated from the UV/
vis absorptions and the “initial” ESR signal intensities
of radicals 2 (see above) accounts for the fact that most
of our o-quinodimethanes are thermally unstable. Note
that in the ESR experiments generally “fresh” photoly-
sates, i.e., always stored on dry ice, were rapidly mixed
with NO solutions when still at temperatures below
ambient temperature, whereas the stopped-flow experi-
ments required handling of the samples at room tem-
perature for several minutes. Of course, significant
differences in the o-quinodimethane concentrations, prob-
ably up to a factor of 10, may simply derive from a larger
deviation of the true extinction coefficient of 1 from the
assumed number of 104 M-1 cm-1, combined with the fact
that the ESR determination of absolute radical concen-
tration normally is subject to large systematic errors
(100% and more).
The orange color of the photolysate from 1,3-dimethoxy-
1,3-diphenylindan-2-one (3a ) in CD3CN or benzene dis-
appeared rapidly when a ca. 10-2 M solution of NO in
F igu r e 4. Time dependence of the ESR signal of 2a obtained
the same solvent was added. At the same time a strong
from mixing an acetonitrile solution of NO with (a) the raw
three-line ESR spectrum (Figure 3), assigned to nitroxide
photolysate from 3a and (b) pure meso-6b in acetonitrile at
radical 2a , was observed when the reaction was carried
out in the cavity of the ESR spectrometer. The number
and intensities of the 13C satellite lines confirm a sym-
metrical structure with respect to the nitroxide group and
indicate that the hyperfine splittings of the R-carbon
atoms and the ipso-carbon atoms of the phenyl substit-
uents are accidentally equivalent within the line width.
The initial signal intensity of 2a further increased slowly
to reach after ca. 10 h a constant level corresponding to
a concentration of 2a of about 1335 µM (Figure 4). Least-
squares fit to a (pseudo) first-order rate law gave a rate
constant of kESR(2a ) ) 1.8 × 10-4 s-1 for the “slow” process
23 °C. Inset: Time evolution of the ESR signal of 2a in benzene
recorded by the rapid-mixing technique. The arrow marks
point of injection.
and an extrapolated “initial” concentration of c(2a ) ) 570
µM (Table 3). This value is higher by about a factor of 4
compared to the spectrophotometrically estimated con-
centration of 1a (150 µM). In benzene-d6 solution the
“initial” ESR concentration of 2a was lower by a factor
of ca. 35 (17 µM), indicating a significantly lower photo-
stationary concentration of o-quinodimethane 1a in this
solvent. This was nicely confirmed by the ratio of the

6842 J . Org. Chem., Vol. 61, No. 20, 1996
Paul et al.
Sch em e 5
F igu r e 5. Spectrophotometric stopped-flow measurement of
the decay of the absorption of 1a in the reaction with NO in
acetonitrile solution at 23 °C. Inset: First-order decay trace
of the absorption at 470 nm.
only the “slow” growth of the ESR signal could be
monitored.
concentration of 1a in both solvents as estimated from
the UV/vis absorption intensity (140 and 6 µM, ratio 23;
Table 3). In benzene, the rate constant for the further,
“slow” buildup of the ESR signal was lower by a factor
of 4.5 compared to the acetonitrile solution; i.e., there is
a small solvent effect. In order to extend the time
evolution of the kinetics of formation of the ESR signal
we performed “rapid”-mixing ESR experiments with a
time resolution in the range of seconds. These experi-
ments qualitatively confirmed the biphasic kinetic be-
havior of the nitroxide-forming process (Figure 4, inset).
The concentration/time profile for the first minute after
rapid mixing of the photolysate with NO solutions was
identical with the inherent time constant of the mixing
device (ca. 27 s); i.e., the rapid radical-forming process
must have occurred faster. We attributed the “fast”
process to the reaction of NO with the persistent isomer
of o-quinodimethane 1a (tentatively assigned to (E,E)-
1a ). To support this assignment, we measured the
kinetics of the reaction of the o-quinodimethane 1a with
saturated (1.4 × 10-2 M)21 solutions of NO in acetonitrile
and benzene by the spectrophotometric stopped-flow
method by monitoring the decay of the UV/vis absorption
at λmax ) 470 nm (Figure 5). The decay obeyed a clean
pseudo-first-order rate law in both solvents with rate
constants of ksf(1a ) ) 0.061 and 0.16 s-1 in acetonitrile
and benzene, respectively, in good agreement with the
number evaluated from the “fast” growth of the ESR
signal. Thus, the rate data indeed describe the reaction
of the persistent isomer of 1a with nitric oxide. Taking
the known concentration of NO after mixing (7 × 10-3
M) we calculated the second-order rate constants for this
process to be k2(1a ) ) 10 and 20 M-1 s-1 in acetonitrile
and in benzene, respectively.
Because the “fast” growth of the ESR signal of 2a is
certainly due to the o-quinodimethane + NO reaction,
the subsequent “slow” increase of the signal must be due
to a reaction of NO with another component of the
photolysis mixture, most likely benzocyclobutene 6a . To
check on this, we mixed equal volumes of a 10-2
M
solution of purified, colorless meso-6a in acetonitrile
(which showed no UV/vis absorption above 300 nm) with
a 1.4 × 10-2 M solution of NO in the same solvent in the
cavity of the ESR spectrometer. As expected, no initial
“jump” of the ESR signal was observed now, but we
subsequently monitored the slow buildup of the ESR
signal of 2a (Figure 4, trace b) with a rate constant (kESR
-
(2a ) ) 4.6 × 10-4 s-1) in reasonable agreement with the
number obtained from the experiment with the raw
photolysate.
We attributed the slow formation of nitroxide 2a to the
reaction of NO with (a) reactive isomer(s) of o-quino-
dimethane 1a , formed by electrocyclic ring opening of
meso-6a . (Scheme 5). J ustification for this interpreta-
tion was based on the following observations: We found
that meso-6a undergoes a slow thermal isomerization to
its thermodynamically more stable diastereomer rac-6a .
(Note that if this reaction would occur as a concerted
process it must involve symmetry-forbidden disrotatory
ring-opening/ring-closure processes). It is hard to believe
that the meso/ rac isomerization does not proceed via (an)
o-quinodimethane intermediate(s), as does the symmetry-
allowed conrotatory ring opening of benzocyclobutenes.
However, we were unable to detect the buildup of the
characteristic UV/vis absorptions in the 400-600 nm
range of the corresponding o-quinodimethanes during the
isomerization of carefully purified meso-6a . Thus, the
steady-state concentration of the intermediate o-quino-
dimethane(s) must be very low (e10-7 M). This conclu-
sion agrees with the above-mentioned detection of two
rather short-lived o-quinodimethane species formed in
the laser flash photolysis of 3a . Further evidence for the
presence of (a) reactive o-quinodimethane(s) was provided
by the observation that in competition to the isomeriza-
tion process meso- and/or rac-6a are irreversibly con-
verted to a 4a,10-dihydroanthracene derivative, most
likely 7a . The latter further decomposed to give finally
a 9,10-substituted anthracene, namely 8a . The most
reasonable explanation for such a reaction to occur is an
electrocyclic ring closure of the o-quinodimethane iso-
mer(s) having at least one phenyl group in Z-orientation.
Further support for the presence of a (relatively)
persistent o-quinodimethane species was provided by
experiments where the photolysates were “aged” at 40
°C for various periods of time before mixing with NO
solutions. With increasing aging time the initial “jump”
of the ESR signal of 2a became less and less pronounced,
in agreement with a slow, irreversible decay of the
corresponding o-quinodimethane. After 30 h at 40 °C
(21) (a) IUPAC Solubility Data Series Vol. 8; Young, C. L., Ed.;
Pergamon Press: Oxford, 1981; pp 260-351. (b) Wisniak, J .; Hersko-
witz, M. Solubility of Gases and Solids, Part B, Physical Sciences Data
18; Elsevier: Amsterdam, 1984; p 1635.

Reaction of Phenyl-Substituted o-Quinodimethanes with NO
J . Org. Chem., Vol. 61, No. 20, 1996 6843
Thus, the time evolution of the “slow” build-up of the ESR
signal of 2a would appear to be governed by the rate of
ring opening of meso-6a rather than the rate of reaction
of the o-quinodimethane with NO. A detailed investiga-
tion of the mechanism of the meso/ rac isomerization and
the accompanying reactions is beyond the scope of the
present paper and will be reported in a separate publica-
tion.
The bulk transformation of benzocyclobutene 6a on
reaction with NO was confirmed by an experiment in
which we bubbled gaseous NO through a solution of 6a
1
in CD3CN for 1 min. After a 4 min reaction period H
NMR analysis revealed a ca. 36% decrease of the con-
centration of 6a . After several days, all of 6a had
disappeared; no formation of dihydroanthracene or an-
thracene derivatives was detected.
F igu r e 6. ESR spectrum of 2k in acetonitrile at 20 °C,
recorded 30 min after mixing of the photolysate of 3k with
NO solution.
spectral components had not changed after 3 days, the
spectrum of the other nitroxide radical had completely
disappeared. Because of the similarity of the ESR data
we tentatively assign the two ESR spectra to the meso
and rac diastereomers of nitroxide 2e. This would be the
only case where the diastereomers of the symmetrically
1,3-diphenyl-substituted nitroxide radicals 2 exhibited
differences in the spectral data exceeding the natural line
widths.22 The o-quinodimethane concentration in the
photolysate of 3e was as low as for most of the other
photolysates; however, the rate of the reaction of 1e with
NO was about 100 times faster than the reaction of 1a
in the same solvent.
The spectroscopic and kinetic data for the photolysate
of 1,1,3-triphenyl-3-methoxyindan-2-one (3i) were, with
one exception, very similar to the photolysate of the
dimethoxy system 3a . The extrapolated, initial signal
intensity of the ESR spectrum of 2i was similarly high,
indicating a photostationary concentration of o-quino-
dimethane 1i of about 350 µM. The ESR signal reached
more or less the same high level as 2a but showed an
almost linear growth for several hours. Therefore, no
rate constant for this growth is given. Markedly different
to 1a , however, was the very low concentration of 0.6 µM
as estimated from the UV/vis absorption of the stopped-
flow experiment. This fact indicated that, contrary to the
dimethoxy system 1a , the isomers of 1i, namely (Z)-1i
and (E)-1i, either both have relatively short lifetimes at
room temperature or a probable long-lived isomer had
been produced only in minor amounts. On the other
hand, the stopped-flow experiments gave a rate constant
for the 1i + NO reaction not much different to that of
1a . As with 6a , the reaction of benzocyclobutene 6i with
excess NO led to a complete conversion of 6i after a few
hours.
Reaction of the photolysates from 1,3-bis[(ethoxycar-
bonyl)methoxy]-1,3-diphenylindan-2-one (3b), 1,3-bis(car-
boxymethoxy)-1,3-diphenylindan-2-one (3c), and disodium-
1,3-bis(carboxylatomethoxy)-1,3-diphenylindan-2-one (3d)
with a solution of NO in the same solvent similarly
produced strong ESR signals of the corresponding ni-
troxides 2b-d (Table 1). The spectroscopic data of these
radicals are, as expected, very similar to those of 2a . The
slightly different 14N hyperfine splitting of 2d reflects the
stronger solvation effect in water. The initial ESR signal
intensities of 2b-d again all increased slowly with time,
giving very strong spectra, persistent for several days,
corresponding to radical concentrations that were much
larger than the estimated o-quinodimethane concentra-
tions. The concentration-time dependence of 2b could
be satisfactoryly fitted to a first-order rate law, yielding
a photostationary concentration of 1b in the photolysate
of about 32 µM. The rate constant for this slow process
was virtually identical to that of 1a . In agreement with
the ESR observations, the UV/vis spectra of the photoly-
sate of 2b showed only a minor change of the absorption
in the wavelength range of the o-quinodimethanes when
reacted with NO solution in the stopped-flow apparatus.
Notably, the rate constant (k2sf(1b) ) 230 M-1 s-1 ) for
the 1b + NO reaction in acetonitrile was found to be
higher by approximately a factor of 25 than that of the
1a + NO reaction.
The slow growth of the ESR signal of nitroxide 2c
obtained from the photolysate of ketone 3c occurred on
a time scale similar to that of 2a and 2b. The concentra-
tion-time profile, however, could not be analyzed straight-
forwardly because for unknown reasons it followed a
sigmoid time dependence. The extrapolated, initial 2c
concentration (20 µM) again was significantly higher
than the concentration of 1c as estimated from the UV/
vis absorption. The rate constant for the fast 1c + NO
reaction differed by only a factor of 2 from that of 1b.
The buffered aqueous photolysis mixture from 3d could
not be investigated by UV/vis spectroscopy because of its
turbidity, which could not be removed by filtration or
centrifugation. However, the strong ESR signal of 2d
that was obtained on mixing of the aqueous photolysate
from 3d and a saturated solution of NO in phosphate
buffer (pH 7.5) indicated a concentration of o-quino-
dimethane 1d comparable to the other systems.
The ESR investigation of the reaction of NO with the
photolysate from 1-[(ethoxycarbonyl)methoxy]-1,3,3-tri-
phenylindan-2-one (3j) paralleled the results obtained
from the diester-substituted analogue 2b. A stopped-flow
experiment was not performed.
When the photolysate from 1,1,3-triphenylindan-2-one
(3k ) was mixed with a benzene solution of NO the six-
line ESR spectrum of 2k (Figure 6) was formed “instan-
taneously”. The large doublet hyperfine splitting of a(H)
) 1.851 mT is very characteristic for conformationally
fixed â-hydrogen atoms in 5-membered cyclic nitroxides,20
reflecting a favorable orientation of the â-C-H bond with
The photolysate from 3e represented the only case
where more than one nitroxide radical was detected on
reaction with NO solution. Immediately after mixing two
overlapping ESR spectra exhibiting very similar spectral
data were observed. The intensity of both signals
increased with a rate constant again in the order of 10-4
s-1. Whereas the “plateau” intensity of one of the ESR
(22) Slight differences in the ESR parameters of meso and rac
diastereomers of cyclic nitroxides are, in principle, generally expected,
e.g., due to structural differences (extent of pyramidalization of the
central nitrogen or bond length alterations) and/or solvent effects (as
induced by the different dipole moments of both diastereomers).

6844 J . Org. Chem., Vol. 61, No. 20, 1996
Paul et al.
dimethanes 1, were also formed. The desired o-quino-
dimethanes were produced only in very low amounts
(0.1-0.001% relative to 3), thus rendering the photolysis
of 2-indanones not a particularly good method for the
production of o-quinodimethanes as nitric oxide chele-
tropic traps (NOCTs).
It is interesting to note that in case of the 1,3-diphenyl-
1,3-dialkoxyl-substituted 2-indanones 3a -d the meso-
configuration of the starting ketone is largely (>90%)
preserved in the benzocylobutene products 6a -d . This
implies that the decay of the intermediate triplet biradi-
cals 3a -e(B2) (Scheme 4) to the o-quinodimethanes 1a -e
should proceed primarily to the E,Z configurational
isomers of 1a -e, which would produce the corresponding
meso-benzocylcobutenes 6a -e thermally via a symmetry-
allowed conrotatory ring closure. (The direct collapse of
triplet diradicals 3a -e(B2) to 6a -e is a spin-forbidden
process and, therefore, rather unlikely). However, due
to their strong UV/vis absorptions the o-quinodimethanes
are photolytically unstable species; hence, if the Z,Z- and
E,E-isomers of 1a -e also would have been produced from
3a -e(B2) they might have been further converted to
meso-6a -e via a photolytically allowed disrotatory pro-
cess in the course of the photolyses. Sound evidence for
the formation of configurationally isomeric o-quino-
dimethanes is provided by the LFP experiments on 3a
and 3i. The results from 3a are particularly enlighten-
ing. Here, similar, overlapping absorption spectra of
three photolytically generated species were detected, two
of which decayed rather rapidly with rate constants in
the range of 106 and 104 s-1, respectively, whereas the
third was very persistent, only decaying within 30 h at
40 °C. We attribute the three absorbing species observed
in the LFP experiment to the three stereoisomers of 1a
by the arguments given above. In particular, the insen-
sitivity of the decay rates of the transient absorptions
on the presence of molecular oxygen dismissed the
intermediate triplet diradicals 3a (B1) and 3a (B2) as the
absorbing species. Our results are in line with observa-
tions made by Quinkert et al.,14b who determined for the
7,7,8,8-tetraphenyl-o-quinodimethane (1m ) a decay rate
constant of k296(1m ) ) 2.8 × 104 s-1 and in case of the
7,8-diphenyl-o-quinodimethane (1g) also observed three
similarly absorbing, transient species of markedly dif-
ferent lifetimes.
With regard to the time resolution of the LFP experi-
ment (ca. 20 ns), the lifetime of the secondary triplet
diradical 3a (B2) must be <20 ns, about 2 orders of
magnitude shorter than found for corresponding 7,7,8,8-
tetraalkyl-o-quinodimethane systems.4,11 This fact is in
agreement with observations reported by Scaiano.12 The
persistency of the third stereoisomer of 1a was confirmed
by the UV/vis spectrum of the photolysate from 3a , which
exhibits a rather strong absorption in the 400-600 nm
range. Since decay of this absorption occurred at 40 °C
within 30 h (from which we estimate a rate constant in
the range of 10-6 s-1 at 20 °C), the activation barrier for
decay of the corresponding o-quinodimethane is roughly
estimated to be about 12-15 kcal mol-1 higher than the
barrier for reaction of the two other isomers. We assume
that the long-lived isomer of 1a is the one having the
(E,E)-diphenyl ((Z,Z)-dialkoxy) configuration. The reason
for this assignment is based on the observations made
with the triphenylalkoxy systems 3h -j, where no “ki-
netically stabilized” (Z,Z)-dialkoxy configuration is pos-
sible, and therefore, only minor amounts of o-quino-
dimethanes could be detected in the photolysates by
conventional UV/vis spectroscopy. Support for this view
F igu r e 7. Spectrophotometric stopped-flow measurement of
the decay of the absorption of 1k in the reaction with NO in
benzene at 23 °C. Inset: First-order decay trace of the
absorption at 490 nm.
respect to the SOMO of the radical. This spectrum
provided further evidence for the general structure of the
nitroxide radicals generated in this study. The 3k /1k
system is the only one in which no further increase of
the ESR signal intensity of 2k was observed, showing
(also in agreement with UV/vis and NMR observations)
that the corresponding benzocyclobutene 6k is stable
under the applied conditions, i.e., does not undergo ring
opening to the corresponding o-quinoid structure. A good
correspondence between the ESR- and UV/vis-spectro-
scopically detected concentration of 1k was found, which
implies a sufficient thermal stability of this o-quino-
dimethane. Interestingly, the rate of reaction of 1k with
NO [k2sf(1k ) ) 2430 M-1 s-1] (Figure 7) is at least 1 order
of magnitude higher than measured for the alkoxy-
substituted counterparts 1i and 1j and almost identical
to that of 2e.
Compound 1l was found to be the most reactive
o-quinodimethane of the present study, as revealed by
the stopped-flow experiment [k2sf(1l) ) 4140 M-1 s-1] with
the photolysate from 1,1,3-triphenyl-3-methylindan-2-one
(3l). The UV/vis-determined concentration of 1l cor-
responded well with the extrapolated initial ESR signal
intensity of nitroxide 2l. Though still in the same order
of magnitude, the rate of the slow growth of the ESR
signal of 2l [kESR(2l) ) 7.5 × 10-4 s-1] was the highest
among our series; i.e., benzocyclobutene 6l should have
a slightly lower thermal stability than the other benzo-
cyclobutenes mentioned here. On the other hand, ni-
troxide 2l appeared to be less persistent than the other
nitroxides, as its ESR signal intensity only increased for
about 40 min after mixing of the reactants, after which
it slowly decayed again.
Discu ssion
P h otolytic Deca r bon yla tion of 2-In d a n on es. The
photolytic decarbonylation of the alkyl- and alkoxyl-
substituted 2-indanones 3 investigated in this paper
occurred in high yields to produce the corresponding 7,8-
diphenyl- or 7,7,8-triphenylbenzocyclobutenes 6 as by far
the major products, in agreement with literature reports
on other phenyl-substituted 2-indanones.12,14 In two
cases (3f and 3l) products 9, deriving from of a 1,5-
hydrogen shift reaction of the intermediate o-quino-

Reaction of Phenyl-Substituted o-Quinodimethanes with NO
J . Org. Chem., Vol. 61, No. 20, 1996 6845
is also provided by J efford et al.,23 who showed that the
concerted, conrotatory ring opening of donor-substituted
benzocylobutenes occurs preferably by rotation of the
donor subtituents into the trans positions of the related
o-quinodimethanes. Vice versa, the conrotatory ring
closure of (Z,Z)-dialkoxy-o-quinodimethanes appears to
be unfavorable. The ca. 35 times lower photostationary
yield of 1a in benzene-d6 compared to CD3CN might be
explained by a strong solvent effect on the distribution
of the diastereomers of 1 (e.g., by quenching of the excited
states) and/or a more effective photochemical destruction
of 1a in benzene solution.
conversion into 4a,10-dihydroanthracene and anthracene
derivatives) react with nitric oxide to produce the corre-
sponding nitroxide radicals 2. This reaction is most
easily explained by the trapping (with rate constants as
discussed above) of the small amounts of the correspond-
ing o-quinodimethanes in equilibrium with the benzocy-
clobutene; i.e., the rates of the “slow” formation of
nitroxides 2 would indeed represent the rate of ring-
opening of the benzocyclobutenes. Very recently, Roth
and co-workers7 have shown that the formally symmetry-
forbidden meso/ rac isomerization of 7,8-diphenylbenzo-
cyclobutene (6g) has an activation barrier only 1 kcal
mol-1 above the barrier of the symmetry-allowed conro-
tatory ring opening. The isomerization is believed to
NO Tr a p p in g Exp er im en ts. The ESR spectroscopic
monitoring of the buildup of the nitroxide radicals 2 after
mixing of the photolysates from the indanones 3 with
solutions of NO revealed that the nitroxide radicals are
formed in a “two-phase” process; i.e., two different
precursors are present in the photolysates, which react
with NO at significantly different rates to give the same
radical product. Kinetic stopped-flow experiments, by
monitoring the decay of the UV/vis absorptions in the
400-500 nm range, showed that the fast nitroxide-
forming reaction is indeed the reaction of the o-quino-
dimethanes with NO. The estimated second-order rate
constants (Table 3) for this reaction cover the range of
10-4000 M-1 s-1, markedly dependent on the substitu-
tion pattern of the o-quinoid system. Though only a
limited number of rate data are currently available, it
seems that the alkoxyl-substituted 7,8-diphenyl- and
7,7,8-triphenyl-o-quinodimethanes generally tend to react
slower with NO by roughly 2 orders of magnitude than
the unsubstituted or alkyl-substituted ones; compare,
e.g., the rate constants of 1a vs 1e and 1i vs 1l. The
higher reactivity of the methyl-substituted systems 1e,
1l versus the methoxyl-substituted ones 1a , 1e is unlikely
due to a more pronounced steric destabilization of the
o-quinodimethane structure by the methyl substitutents,
as, for example, the 7,7,8-triphenyl-o-quinodimethane
(1k ) is much more reactive than the sterically more
congested methoxyl system 1i . These relationships
imply a pronounced electronic influence on the reactivity
of our o-quinodimethanes. The rate constant of k ) 310
M-1 s-1 for reaction of NO with our first NO trap, 7,7,8,8-
tetramethyl-o-quinodimethane (NOCT-1; see Introduc-
tion), also falls in the above range.24 This would mean
that there is no special “phenyl effect” on the reactivity
toward NO. The differences in the rate constants of the
alkoxyl systems 1a and 1b or 1c, on the other hand,
reveal some other influence of the substituents rather
than a purely electronic one. One should keep in mind,
however, that we do not know what the actual configu-
ration or distribution of possible configurations of the
active NO traps might be in our photolysates. The
reactivity of the o-quinodimethanes toward NO certainly
is governed by steric and electronic influences of the
terminal substitutents; however, at the present point
even a qualitative judgement on the relative importance
of both effects is not possible.
proceed via an orthogonal biradical intermediate.
A
similar situation, therefore, is very likely for our our
systems. Thus, the rates of the 6 + NO reactions might
also include the trapping of the corresponding biradical
intermediates. However, a direct reaction of NO with
the benzocyclobutenes cannot be completely ruled out
presently because at least the 6a + NO reaction is much
faster than the thermal meso/ rac isomerization. Inves-
tigations to clarify this point are in progress. The rate
data (Table 3) show that the slow nitroxide-forming
process is somewhat influenced by the substitution
pattern of the tetrasubstituted benzocyclobutenes, the
rate constants varying by a factor of 20. Steric interac-
tions should provide the major contribution to the stabil-
ity of our benzocyclobutene systems, as is evident from
the stability of the trisubstituted benzocyclobutene 6k ,
which is completely stable at room temperature and,
hence, does not react with NO to produce nitroxide
radicals 2k .
o-Qu in od im eth a n es a n d Ben zocyclobu ten es a s
Nitr ic Oxid e Ch eletr op ic Tr a p s (NOCTs). The for-
mation of strong ESR signals of nitroxides 2 demon-
strates that the small concentrations of the o-quino-
dimethanes 1 are sufficient for the application of the
photolysates as NO trapping reagents in solution, as has
been demonstrated before.3-5
In view of the desired (see Introduction) thermody-
namic stabilization of o-quinodimethanes by introduction
of terminal phenyl substitutents combined with the
suppression of 1,5-hydrogen shift reactions by alkoxyl
substituents, most of the o-quinodimethanes generated
in this study disappointingly appeared to be rather
unstable species. This is due to the now-opened route
to form stable benzocyclobutene derivatives, the exception
being the (probably kinetically stabilized) Z,Z-conformer
of 1a . Of course, one would prefer a synthetic method
by which persistent, reactive o-quinodimethanes would
be produced in preparative yields without the production
of large amounts of byproducts. The unfavorable cycliza-
tion to benzocyclobutenes as well as the 1,5-hydrogen
shift can be avoided by generating bicyclic o-quino-
dimethanes in which the terminal carbon atoms of the
quinoid system are fixed by the ring structure. Related
work is in progress.24
The ESR experiments with the purified meso-7,8-
dimethoxy-7,8-diphenylbenzocyclobutene (6a ) (Figure 4)
as well as the complete consumption of this and other
benzocyclobutenes in the long-term reaction with NO
proved that those phenyl-substituted benzocyclobutenes
that are thermally unstable (as indicated by the slow
As has been demonstrated by the o-quinodimethanes
NOCT-1 and NOCT-4 (see Introduction), a rate constant
of about 300 M-1 s-1 for the reaction with NO is sufficient
to trap some NO in biological environment.3-5 However,
this has been done with a line of cells, macrophages,
which are known to produce high levels (nmol range) of
NO. For the trapping of NO from low-level biological
sources (e.g., brain; pmol NO levels) such rate constants
are too low in order to compete with the destruction of
NO by other reactions such as oxidation, addition to iron
(23) J efford, C. W.; Bernardinelli, G.; Wang, Y.; Spellmeyer, D. C.,
Buda, A.; Houk, K. N. J . Am. Chem. Soc. 1992, 114, 1157-65.
(24) Ba¨tz, T., Korth, H.-G.; Sustmann, R. Universita¨t - GH Essen,
unpublished results.

6846 J . Org. Chem., Vol. 61, No. 20, 1996
Paul et al.
complexes, and so forth.1 Currently, our rate constants
are far below the absolute desirable optimum, i.e., a
diffusion-controlled rate constant for the o-quinodimethane
+ NO reaction. However, the fact that the simple
exchange of methoxy goups by methyl groups is con-
nected with an increase of the rate constants by a factor
of 10-100 gives some hope that the reactivity of our
NOCTs can be increased further by suitable structural
variations. We estimate that a rate constant in the range
of 106 M-1 s-1 will be sufficient for a useful NO trap for
biological applications.
The low rate of reaction of our benzocyclobutenes with
NO certainly renders them to be useless as preparatively
easily accessible “reservoirs” for NOCTs under physi-
ological conditions. Of course, one might further desta-
bilize the benzocyclobutene ring in order to increase the
rate of formation of the o-quinodimethanes. On the other
hand, this would probably defacilitate the synthetic
preparation of the benzocyclobutenes. Substituted ben-
zocylobutenes of the present type, however, can be
advantageously applied as NO traps for situations where
the generated NO is long-lived enough and/or the NO is
generated at high rates, e.g., in the screening for NO-
releasing drugs or investigations concerning NO-related
chemistry. For example, this has been demonstrated in
the investigation of the oxygen-dependence of the cyto-
toxicity of NO-releasing compounds.9
combined filtrate was washed three times with a 5% aqueous
solution of sodium cyanide (100 mL each) and three times with
water (100 mL each). After being dried over magnesium
sulfate the solvent was evaporated and the oily residue dried
under vacuum at 50 °C and recrystallized from ethanol to
afford 3.19 g 3a (59%), mp 131 °C. 1H NMR (300 MHz, CDCl3)
δ: 3.32 (s, 6H, OCH3), 7.10 (m, 10H), 7.65 (m, 4H, AA′BB′).
13C NMR (75 MHz, CDCl3) δ: 53.6, 85.6, 126.6, 126.8, 127.5,
128.1, 128.3, 130.2, 138.1, 140.5, 221.5. IR (KBr, cm-1): 3060,
3039, 2954, 2923, 2833, 1761, 1185, 1092, 762, 746, 700. UV/
vis (acetonitrile, nm) λmax (lg ꢀ): 269 (3.45), 277 (3.36), 344
(2.60), 355 (2.60). MS (GC/MS, 70 eV) m/ z: 316 (M+ - CO,
38), 301 (M+ - COCH3,100), 239 (14), 165 (12) , 105 (13) , 77
(18). HRMS (70 eV): calcd (M+) 344.1412, obsd 344.1419.
meso-1,3-Bis[(eth oxycar bon yl)m eth oxy]-1,3-diph en ylin -
d a n -2-on e (3b). To a solution of 3f (3.0 g, 9.48 mmol) in DMF
(50 mL) were added ethyl iodoacetate (12.0 mL, 101 mmol)
and silver(I) oxide (10.0 g, 43.25 mmol). The mixture was
stirred for 3 d at room temperature and was then poured into
trichloromethane (200 mL). The precipitated silver salts were
filtered off, and the filtrate was washed three times with a
5% aqueous solution of sodium cyanide (50 mL each) and five
times with water (50 mL each). After the filtrate was dried
over magnesium sulfate the solvent was evaporated and the
oily residue recrystallized from pentane/benzene (6:1) to give
3.28 g of 3b (71%), mp 93-94 °C. 1H NMR (300 MHz, CD3-
3
3
CN) δ: 1.19 (t, J ) 7.1 Hz, 6H), 3.97 (d, J ) 15.4 Hz, 2H),
4.10 (q, 3J ) 7.1 Hz, 4H), 4.12 (d, 3J ) 15.4 Hz, 2H), 7.21-
7.07 (m, 10H), 7.72-7.63 (m, 4H, AA′BB′, arom H of indanone).
13C NMR (75 MHz, CD3CN) δ: 14.2, 61.5, 64.4, 86.3, 127.5,
128.4, 129.1, 129.7, 131.9, 138.4, 140.5, 169.9, 211.9. IR (KBr,
cm-1): 3062, 2924, 1761, 1743 (νCdO), 1120, 768, 698. UV/vis
(acetonitrile, nm) λmax (lg ꢀ) 270 (3.45), 277 (3.36), 344 (2.61),
354 (2.55). MS (70 eV) m/ z: 488 (1, M+), 460 (4, M+ - CO),
373 (70), 269 (100). HRMS (70 eV): calcd (M+) 488.1835, obsd
488.1830.
Exp er im en ta l Section
In str u m en ta tion . 1H and 13C NMR spectra (internal
standard TMS): Varian Gemini-200 and Bruker AMX-300. IR
(KBr pellet): Perkin-Elmer 1600 series FTIR. UV/vis: Varian
Cary 219, modified for use with light pipes and thermostatable
sample holders. GC/MS (70 eV): Hewlett-Packard HP 5971A
mass-sensitive detector and HP5890 GC, capillary column HP
1; 50 m × 0.2 mm; SF 0.33 µm. High-resolution MS (70 eV):
Fisons Instrument VG Pro Spec 300. HPLC: Varian LC 5000
and Hewlett-Packard 1040 diode array UV detector, Varian
MCH-10 column (10 µm; 4 × 300 mm). Stopped-flow UV/vis
measurements (time range 2 ms to minutes): Hi-Tech Scien-
tific SF 40 instrument and Hewlett-Packard 9153c computer.
ESR (9.4 GHz): Bruker ER-420 X-band spectrometer. ESR
data acquisition: 486 PC equipped with a Microstar DAP
1200/4 data acquisition board and DigiS (GfS Aachen, Ger-
many) software. Melting points (uncorrected): Bu¨chi 510.
Ma ter ia ls. For the synthesis of 1,3-diphenyl-substituted
2-indanones ninhydrin was used as starting material. Ket-
alization of ninhydrin monohydrate (Aldrich) according to
Kuhn and Trischmann25 gave 2,2-d im eth oxyin d a n -1,3-d i-
on e in 81% yield, mp 69 °C (lit.25 70-71 °C).
1,3-Dih ydr oxy-2,2-dim eth oxy-1,3-diph en ylin dan (4) was
obtained via Grignard reaction of 2,2-dimethoxyindan-1,3-
dione following the procedure of Holland and J ones,17b yield
76%, mp 114 °C (lit.17b mp 116-118 °C). 1H NMR (200 MHz,
CDCl3) δ: 2.85 (s, 6H), 4.28 (s, 2H, OH), 7.57-7.15 (m, 14H).
m eso-1,3-Dih yd r oxy-1,3-d ip h en ylin d a n -2-on e (3f)26 was
prepared by hydrolysis of 4 in 16% hydrochloric acid/acetic acid
(9:8) to give 59% 3f, mp 230-235 °C dec (lit.26 mp 225-240
°C dec). 1H NMR (300 MHz, DMSO-d6) δ: 6.67 (s, 2H, OH),
7.15-7.05 (m, 10 H), 7.60-7.53 (m, 4H, AA′BB′). 13C NMR
(75 MHz, DMSO-d6) δ: 79.3, 125.5, 126.8, 127.5, 127.7, 129.6,
141.0, 143.0, 214.0.
Anal. Calcd for C29H28O7 (488.18): C, 71.30; H 5.78.
Found: C, 71.67; H, 7.20.
m eso-1,3-Bis(ca r boxylm et h oxy)-1,3-d ip h en ylin d a n -2-
on e (3c). A solution of 3b (2.50 g, 5.1 mmol) in a mixture of
ethanol (50 mL) and 10% aqueous sodium hydroxide (50 mL)
was stirred for 5 h at room temperature and then acidified
with 2 M hydrochloric acid and extracted three times with
trichloromethane (50 mL each). The combined organic layers
were dried over magnesium sulfate, the solvent was evapo-
rated, and the oily residue was recrystallized from pentane/
benzene (5:2) to give 1.42 g (3.3 mmol, 63%) of 3c, mp 185-
195 °C dec. 1H NMR (200 MHz, DMSO-d6) δ: 3.97 (d, 2J )
2
15.8 Hz, 2H), 4.12 (d, J ) 15.8 Hz, 2H), 7.23-7.02 (m, 10H)
7.75-7.63 (m, 4H, AA′BB′). 13C NMR (50 MHz, CD3CN) δ:
63.8, 86.3, 127.5, 128.3, 129.1, 129.5,131.9, 138.2, 140.3, 170.3,
211.6. IR (KBr, cm-1): 3421 (νOH), 3062, 2924, 1761, 1743
(νCdO), 1448, 1120, 768, 698. UV/vis (acetonitrile) λmax (lg ꢀ):
270 (3.32), 277 (3.23), 344 (2.47) nm. MS (70 eV) m/ z: 404
(1, M+), 328 (45), 270 (100). HRMS (70 eV) calcd 404.1260
(M+ - CO), obsd 404.1281. Anal. Calcd for C25H20O7
(404.13): C, 69.44; H, 4.66. Found: C 69.25; H, 4.68.
Disod iu m m eso-1,3-Bis(ca r b oxyla t om et h oxy)-1,3-d i-
p h en ylin d a n -2-on e (3d ). A solution of 3c (0.35 g, 0.81 mmol)
in ethanol (15 mL) was neutralized (pH meter control) by
dropwise addition of a solution of sodium ethanolate (from
dissolution of 0.2 g sodium in 30 mL of ethanol). The solvent
was evaporated and the residue dried at 50 °C under vacuum
for 3 d to give 0.35 g of 3d (88%), mp 219-230 °C dec. 1H
2
NMR (200 MHz, D2O) δ: 3.79 (d, J ) 14.7 Hz, 2H), 3.97 (d,
2J ) 14.9 Hz, 2H), 7.13 (m, 10H) 7.79 (m, 4H, AA′BB′). 1H
NMR (300 MHz, DMSO-d6) δ: 3.84 (d, 2J ) 14.7 Hz, 2H), 4.01
(d, 2J ) 14.7 Hz, 2H), 7.04 (m, 4H), 7.21 (m, 6H), 7.70 (m, 4H).
IR (KBr, cm-1): 3057, 2965, 2907, 1760 (νCdO), 1491, 1219,
1126, 765, 697. UV/vis (phosphate buffer pH 7.8, nm) λmax (lg
ꢀ): 342 (2.62).
m eso-1,3-Dim eth oxy-1,3-d ip h en ylin d a n -2-on e (3a ). A
mixture of 3f (5.0 g, 15.8 mmol), iodomethane (20 mL, 321
mmol), and silver(I) oxide (7.51 g, 54.5 mmol) in DMF (120
mL) was stirred for 2 d at room temperature. The mixture
was poured into trichloromethane (250 mL) and the precipitate
filtered off and washed once with trichloromethane. The
1,3-Dip h en ylin d a n -2-on e (3g) was prepared from 4 by a
literature procedure,17b yield 79%, mp 153-156 °C dec (lit.17b
mp 155 °C dec). 1H NMR (200 MHz, CD3CN) δ: 4.95 (s, 2H),
7.41-7.06 (m, 14H). IR (KBr, cm-1): 3054, 3026, 2884, 1754
(νCdO), 1598, 759, 742,700. UV/vis (acetonitrile, nm) λmax (lg
ꢀ): 268 (3.17), 276 (3.12), 303 (2.62), 313 (2.56), 325 (2.25).
(25) Kuhn, R.; Trischmann, H. Chem. Ber. 1961, 94, 2258-2263.
(26) Blomquist, A. T.; Moriconi, E. J . J . Org. Chem. 1961, 26, 3761-
3769.

Reaction of Phenyl-Substituted o-Quinodimethanes with NO
J . Org. Chem., Vol. 61, No. 20, 1996 6847
1,3-Dim eth yl-1,3-d ip h en ylin d a n -2-on e (3e). To a vigor-
ously stirred suspension of powdered sodium hydroxide (4.20
g, 74.9 mmol) in DMSO (15 mL) was added at 60 °C within 1
h a suspension of 3g (2.0 g, 7.0 mmol) in a mixture of
iodomethane (6.0 mL, 95.4 mmol) and DMSO (10 mL). After
further addition of iodomethane (2.0 mL, 31.8 mmol) the
mixture was stirred for 3 h at 60 °C and 2 h at room
temperature. The solution was poured into ice-water (100
mL) and extracted five times with petroleum ether (boiling
range 40-60 °C; 50 mL each), and the combined organic
phases were dried with magnesium sulfate. After evaporation
of the solvent the orange-colored, oily residue was dissolved
in diethyl ether and the solution filtered over neutral alumina
(30 × 4.4 cm column). The solvent was evaporated from the
eluate and the yellow residue stirred for 3 d with pentane.
The colorless precipitate was filtered off and dried under
vacuum to give 0.81 g of 3e (2.6 mmol, 37%) as a 1:1.1 mixture
of the meso and rac diastereomers. 1H NMR (200 MHz, CDCl3)
δ: 1.68, 1.78 (s, 6H), 7.45-7.03 (m, 14H). 13C NMR (50 MHz,
CDCl3) δ: 6.0, 26.4, 57.3, 57.7, 125.2, 125.5, 126.5, 126.8, 126.9,
127.2, 128.1, 128.3, 128.4, 142.1, 143.1, 144.5, 145.3, 218.2,
219.1. IR (KBr, cm-1): 3063, 3032, 2973, 2929, 1744 (νCdO),
1596, 1492, 766, 729, 700. UV/vis (acetonitrile, nm) λmax (lg
ꢀ): 261 (3.22), 267 (3.24), 274 (3.18), 284 (2.77), 302 (2.64),
313 (2.70), 323 (2.52).
1-Hyd r oxy-1,3,3-tr ip h en ylin d a n -2-on e (3h ). This com-
pound was prepared in four steps starting from diphenylinden-
1-one (Aldrich) according to the procedure reported by Koelsch,27
overall yield 25%, mp 160 °C (lit.17b mp 161 °C). 1H NMR (200
MHz, CDCl3) δ: 3.34 (s, 1H, OH), 7.59-6.97 (m, 19H). 13C
NMR (50 MHz, CDCl3, TMS) δ: 58.1, 68.8, 125.8, 127.0, 127.4,
128.2, 128.3, 128.4, 128.5, 128.7, 129.0, 129.2 u. 129.5, 137.7,
139.6, 141.4, 143.1, 144.8, 214.0. IR (KBr, cm-1): 3414, 3057,
3026 , 1754 (νCdO), 1596, 1046, 1032, 746, 696. UV/vis
(acetonitrile, nm) λmax (lg ꢀ): 269 (3.30), 277 (3.18), 329 (2.56).
MS (70 eV) m/ z: 376 (5, M+), 360 (11), 348 (100, M+ - CO),
271 (44). HRMS (70 eV): calcd 376.1463 (M+), obsd 376.1474.
1-Meth oxy-1,3,3-tr ip h en ylin d a n -2-on e (3i). This com-
pound was prepared by the method of Oku et al.,28 yield 63%,
mp 165-166 °C (lit.28 mp 168 °C). 1H NMR (200 MHz, CDCl3)
δ: 3.22 (s, 3H), 7.54-6.93 (m, 19H). 13C NMR (50 MHz,
CDCl3) δ: 53.5, 68.6, 88.0, 126.4, 126.9, 127.1, 127.6, 127.9,
128.1, 128.6, 129.3, 130.0, 138.5, 138.9, 141.0, 143.1, 145.5,
212.8. IR (KBr, cm-1): 3060, 3031, 2932, 2830, 1754 (νCdO),
1597, 1087, 752, 697. UV (acetonitrile, nm) λmax (lg ꢀ): 269
(3.35), 278 (3.22), 327 (2.47).
(200 MHz, CDCl3) δ: 4.83 (s, 1H), 7.40-7.00 (m, 19 H). 13C
NMR (50 MHz, CDCl3) δ: 58.1, 68.8, 125.8, 127.0, 127.4, 128.2,
128.3, 128.4, 128.5, 128.7, 129.0, 129.2, 129.5, 137.7, 139.6,
141.4, 143.1, 144.8, 214.0. IR (KBr, cm-1) 3054, 3026, 2883,
1754 (νCdO), 1598, 1494, 1060, 759, 741, 699. UV/vis (aceto-
nitrile, nm) λmax (lg ꢀ): 277 (3.07), 328 (2.50).
1-Meth yl-1,3,3-tr ip h en ylin d a n -2-on e (3l). A solution of
3k (2.0 g, 5.54 mmol) and iodomethane (3.0 mL, 48.0 mmol)
in DMSO (10 mL) was added to a stirred suspension of
potassium hydroxide (6.80 g, 121 mmol) in DMSO (20 mL).
After being heated to 50 °C the mixture was stirred for 3 h,
cooled, and poured into ice-water (150 mL). The precipitate
was filtered off and dried under vacuum. The product then
was refluxed with trichloromethane (50 mL) for 15 min, the
solution was cooled to room temperature, and insoluble
byproducts were removed by filtration. The solvent was
evaporated and the residue recrystallized from pentane, af-
fording 1.04 g of 3l (2.78 mmol, 50%), mp 107-110 °C. 1H
NMR (200 MHz, CDCl3) δ: 1.78 (s, 3H), 7.45-6.95 (m, 19H).
13C NMR (50 MHz, CDCl3) δ: 26.1, 57.9, 68.2, 125.4, 126.7,
126.8, 126.9, 127.2, 127.3, 128.0, 128.2, 128.4, 128.8, 129.1,
141.9, 142.0, 143.7, 143.8, 144.8, 217.0. IR (KBr, cm-1): 3062,
3031, 2962, 2923 , 1746 (νCdO), 1600, 1031, 769, 746, 700. UV/
vis (acetonitrile, nm) λmax (lg ꢀ): 261 (3.28), 268 (3.28), 275
(3.17), 308 (2.74), 317 (2.76), 330 (2.47). MS (70 eV) m/ z: 374
(10, M+), 356 (51, M+ - H2O), 346 (64, M+ - CO), 331 (100).
HRMS (70 eV): calcd 374.1671 (M+), obsd 374.1681.
m eso-2-Hyd r oxy-1,3-d im eth oxy-1,3-d ip h en ylin d a n (5).
To a stirred solution of 3a (66.7 mg, 0.19 mmol) in dry diethyl
ether (10 mL) was added LiAlH4 (0.10 g, 2.6 mmol) under
nitrogen, and the mixture was refluxed for 1 h. After the
mixture was cooled to room temperature hydrolysis was
performed by careful addition of 1 M sulfuric acid. Diethyl
ether (20 mL) was added, and the organic layer was separated
and washed three times with water. Drying over magnesium
sulfate and evaporation of the solvent gave 59 mg (90%) of
the diastereomeric mixture of meso-5 in a 1:1.3 ratio. 1H NMR
3
(200 MHz, CDCl3) δ: 0.82 (d, J ) 2.9 Hz, 1H, OH), 3.16 (s,
6H), 3.21 (s, 6H), 3.60 (d, 3J ) 12.1 Hz, 1H, OH), 3.87 (d, 3J )
3
12.1 Hz, 1H), 4.33 (d, J ) 2.9 Hz, 1H), 7.58-7.25 (m). The
signals at δ 0.82 and 3.60 as well as the splitting of the signals
disappeared on addition of D2O.
P h otolysis of 2-In d a n on es (Gen er a l P r oced u r e). All
reactions were carried out with careful exclusion of oxygen
under argon (Messer-Griesheim 99.999%). Typically, 5 × 10-2
M solutions of the 2-indanones in the deoxygenated solvent
(acetonitrile, water, benzene) were transferred into a long-
necked, septum-capped 1-cm UV quartz cell, and the cell was
placed in an externally cooled methanol bath in a double-
walled quartz jacket. Irradiation was performed in a Rayonet
photoreactor (# RPR-100; The Southern New England Ultra-
violet Co.) at 300 nm (RPR 3000 lamps) with continuous
stirring of the solution by a slow stream of argon that was
passed through the solution via a hypodermic needle. After
photolysis the photolysates were rapidly transferred to dry,
argon-flushed septum-capped vials by means of a gas-tight
syringe and stored at -78 °C (dry ice). The photolysates were
1-[(Eth oxyca r bon yl)m eth oxy]-1,3,3-tr ip h en ylin d a n -2-
on e (3j). To a solution of 3h (3.0 g, 7.97 mmol) in DMF (45
mL) were added ethyl iodoacetate (9.0 mL, 67.5 mmol) and
silver(I) oxide (14.4 g, 62.1 mmol). The mixture was stirred
for 3 d at room temperature and was then poured into
trichloromethane (150 mL). The precipitated silver salts were
filtered off, and the filtrate was washed three times with a
5% aqueous solution of sodium cyanide (50 mL each) and five
times with water (50 mL each). After the filtrate was dried
over magnesium sulfate the solvent was evaporated and the
oily residue recrystallized from hexane/benzene (3:1 v/v) to give
3.30 g of 3j (89%), mp 153-155 °C. 1H NMR (300 MHz, CDCl3)
1
analyzed by GC/MS and H and 13C NMR spectroscopy as soon
3
2
δ: 1.22 (t, J ) 7.2 Hz, 3H), 3.96 (d, J ) 15.3 Hz, 1H), 4.07
(d, 2J ) 15.3 Hz, 1H), 4.15 (q, 3J ) 7.2 Hz, 4H), 7.62-6.97 (m,
19H). 13C NMR (75 MHz, C,H-COSY, CDCl3) δ: 14.1 (CH3),
60.9 (CH2CH3), 63.7 (OCH2), 66.8 (CPh2), 87.8 [CPh(OCH2CO2-
Et)], 126.4, 127.1, 127.2, 127.6, 127.7, 128.0, 128.2 (2C), 128.3,
128.7, 128.9, 129.3, 130.3 (tert arom C), 138.0, 138.2, 140.3,
143.1, 145.1 (quart arom C) 169.4 (CO(OEt)), 212.4 (CdO). IR
(KBr, cm-1): 3063, 2912, 2856, 1764 (νCdO), 1593. UV/vis
(acetonitrile, nm) λmax (lg ꢀ): 270 (3.45), 278 (3.34), 329 (2.64).
MS (70 eV) m/ z: 462 (1, M+), 434 (10, M+ - CO), 418 (30),
330 (100). HRMS (70 eV): calcd 462.1830 (M+), obsd 462.1793.
Anal. Calcd for C31H26O4 (462.18): C, 80.49; H, 5.68.
Found: C, 80.53; H, 5.69.
as possible after completion of the photolysis. With the
exception of 6a the photoproducts were not isolated. The
conditions of the photolyses are given in Table 1. Spectroscopic
data of the photoproducts are reported below.
P h otolysis of 3a . A solution of 3a (90.0 mg, 0.26 mmol)
in acetonitrile (5 mL) was photolyzed (λ ) 300 nm) at -7 °C
for 4 h. The solution was transferred to a Schlenck tube,
approximately half of the volume of the solvent was removed
under vaccuum, and the tube was stored at -18 °C for 7 d
during which time colorless meso-7,8-dimethoxy-7,8-diphenyl-
benzocyclobutene (meso-6a ) crystallized; yield 26.1 mg (31%),
mp 106-107 °C. 1H NMR (300 MHz, CD3CN) δ: 3.33 (s, 6H),
7.08-6.94 (m, 10H), 7.62 (m, 4H, AA′BB′). 13C NMR (75 MHz,
CD3CN) δ: 53.9, 95.2, 126.4, 127.6, 128.0, 128.6, 130.6, 140.0,
146.1. UV/vis (HPLC, nm) λmax: 258, 264, 269. MS (GC/MS,
70 eV) m/ z: 316 (1, M+), 301 (7, M+ - CH3), 285 (53), 270 (8),
254 (100), 238 (48), 165 (10), 77 (5). rac-7,8-Dimethoxy-7,8-
diphenylbenzocyclobutene (rac-6a ): 1H NMR (300 MHz, CD3-
CN) δ: 2.97 (s, 6H), 7.39-7.33 (m, 10H), 7.66-7.58 (m, 4H,
1,1,3-Tr ip h en ylin d a n -2-on e (3k ). By deoxygenation of
3h ;27 yield 94%, mp 103 °C (lit.27 mp 105-109 °C). 1H NMR
(27) Koelsch, C. F. J . Org. Chem. 1938, 3, 456-461.
(28) Oku, A.; Shimada, K.; Mashio, F. Bull. Chem. Soc. J pn. 1973,
46, 275-279.

6848 J . Org. Chem., Vol. 61, No. 20, 1996
Paul et al.
AA′BB′). 13C NMR (75 MHz, CD3CN) δ: 53.0, 94.3, 126.0,
128.4, 128.6, 129.3, 130.3, 139.7, 145.7. UV/vis (HPLC, nm)
Rea ction of th e P h otolysa tes w ith NO Solu tion for
ESR Exp er im en ts (Gen er a l P r oced u r e). NO from a
λmax
:
258, 263, 269. Anthracene derivative (8a ): UV/vis
cylinder (Messer-Griesheim 2.8; 99.8% NO) was purified by
passing through a column of Ascarite (Aldrich) and 20%
potassium hydroxide solution using an all-stainless steel line.
Saturated NO solutions (typically 1.4 × 10-2 M)21 were
prepared by bubbling NO for 30 min through a deoxygenated
(by bubbling with argon for 1 h) solvent in a septum-capped
vial by means of hypodermic needles. Typically, 0.2-0.5 mL
of cold (e0 °C) solution from the photolysis of the indanones
was transferred using a hypodermic syringe into an argon-
flushed septum-capped quartz ESR tube (external diameter 3
mm for acetonitrile, 4 mm for benzene) or a 0.4 mm quartz
flat cell (for water). A total of 0.02-0.1 mL of the NO solution
was then rapidly added using a gas-tight hypodermic syringe,
the solutions were briefly mixed by the Ar stream, and the
ESR tube was placed in the cavity of the ESR spectrometer.
ESR Mea su r em en ts. ESR experiments were performed
at ambient temperature in a double cavity at microwave power
levels of 1-10 mW and 0.4-1 G modulation amplitude. The
second cavity contained a calibrated spin concentration stan-
dard. (Varian strong pitch; c ) 4.68 × 10-5 M). ESR spectra
were recorded as fast as possible (2-5 min) after addition of
the NO solution. ESR parameters were refined by computer
simulation using the LMB simulation program.29 Spin con-
centrations were determined by double integration of the
digitized spectra of the nitroxide radicals and comparison with
the double integration of the signal from the spin standard.
The spin standard was recorded with the same instrument
settings (except, when necessary, for the signal gain) as the
nitroxide signal. For the rapid-mixing ESR experiments a flow
tube was developed30 that allowed rapid mixing of the pho-
tolysate and the NO solution in the cavity of the ESR
spectrometer. Solutions were injected into the mixing part of
the flow tube via stainless steel tubes by 500 µL syringes,
which were driven by a Hamilton digital diluter. The kinetics
of formation of the nitroxide radical 2a was evaluated by
monitoring the growth of the Ms ) +1 hyperfine line of its
ESR signal.
(HPLC, nm) λmax 258, 357, 374, 396.
P h otolysis of 3b. meso-7,8-Bis[(ethoxycarbonyl)methoxy]-
7,8-diphenylcyclobutene (meso-6b). 1H NMR (200 MHz, CD3-
CN) δ: 1.16 (t, 3J ) 7.0 Hz, 6H), 4.09 (q, 3J ) 7.0 Hz, 4H),
4.27 (d, 2J ) 16.6 Hz, 2H), 4.37 (d, 2J ) 16.6 Hz, 2H), 6.99 (m,
10H), 7.70-7.50 (m, 4H, AA′BB′). 13C NMR (50 MHz, CD3-
CN) δ: 14.2, 61.2, 65.6, 95.9, 126.4, 127.9, 128.1, 128.4, 131.2,
139.8, 145.5, 171.2.
P h ot olysis of 3c. meso-7,8-Bis(carboxylmethoxy)-7,8-
diphenylbenzocyclobutene (meso-6c). 1H NMR (200 MHz, CD3-
2
2
CN) δ: 4.21 (d, J ) 16.4 Hz, 2H), 4.31 (d, J ) 16.4 Hz, 2H),
7.00 (m, 10H), 7.68-7.63 (m, 4H AA′BB′). 13C NMR (50 MHz,
CD3CN) δ: 64.8, 95.3, 126.4, 128.3, 128.6, 131.7, 138.8, 144.9,
171.5.
P h otolysis of 3d . Disodium meso-7,8-bis(carboxylatometh-
oxy)-7,8-diphenylbenzocyclobutene (meso-6d ): 1H NMR (200
MHz, CD3CN) δ: 3.89 (d, 2J ) 15.3 Hz, 2H), 4.09 (d, 2J ) 15.2
Hz, 2H), 7.01 (m, 10H), 7.66 (m, 4H, AA′BB′). 13C NMR (50
MHz, D2O, DMSO) δ: 66.2, 94.2, 126.3, 128.4, 128.5, 129.0,
131.4, 138.3, 144.7, 178.0. UV/vis (HPLC, nm) λmax: 278, 283,
290.
P h otolysis of 3e. meso/ rac-7,8-Dimethyl-7,8-diphenyl-
benzocyclobutene (6e). 1H NMR (200 MHz, CD3CN) δ: 1.16/
1.83 (s, 6H, CH3), 7.50-6.80 (m, 19H). 13C NMR (75 MHz,
CD3CN) δ: 24.1/27.5, 61.3/61.9, 123.8, 124.6, 125.9, 126.1,
127.2, 127.7, 127.8, 128.2, 128.7, 128.9, 129.1,129.4, 145.2/
145.9, 149.7/150.6.
P h otolysis of 3h . 2-(Diphenylmethyl)benzophenone (9h ).
1H NMR (200 MHz, C6D6) δ: 5.87 (s, 1H), 7.69-6.99 (m,19H).
UV/vis (HPLC, nm) λmax: 252, 285 (sh). MS (70 eV) m/ z: 348
(100, M+), 330 (40), 271 (63), 165 (38).
P h ot olysis of 3i. 7-Methoxy-7,8,8-triphenylbenzocyclo-
butene (6i). 1H NMR (300 MHz, CD3CN) δ: 3.13 (s, 3H,
OCH3), 8.00-6.80 (m, 19H). 13C NMR (75 MHz, CD3CN) δ:
53.7, 72.8, 95.0, 125.2, 126.0, 126.5, 126.8, 127.6, 127.7, 128.1,
128.6, 128.7, 128.8, 129.7, 130.5, 139.7, 142.9, 143.2, 144.1,
148.5. UV/vis (HPLC, nm) λmax: 264, 271. 9,10-Diphenylan-
thracene (8m ). UV/vis (HPLC, nm) λmax: 353, 372, 392. 4a,-
10-Dihydroanthracene derivative (7i?): UV/vis (HPLC, nm)
La ser F la sh P h otolysis. LFP (Nd:YAG laser, Lumonics
HY-750) (266 nm) experiments were carried out employing
oxygenated and deoxygenated 10-2 M acetonitrile solutions of
indanones 3a and 3i in 7 × 7 mm quartz cuvettes. Details of
the equipment and experimental procedures have been de-
scribed elsewhere.31
λmax: 383.
P h ot olysis of 3j. 7-[(Ethoxycarbonyl)methoxy]-7,8,8-tri-
phenylbenzocyclobutene (6j). 1H NMR (300 MHz, CD3CN) δ:
1.06 (t, 3J ) 7.1 Hz, 3H), 3.70 (s, 2H), 3.95 (q, 3J ) 7.1 Hz,
2H), 7.95-6.82 (m, 19H). 13C NMR (75 MHz, CD3CN) δ: 14.0,
61.2, 64.7, 73.1, 95.3, 125.4, 126.2, 126.9, 127.9, 128.0, 128.3,
128.5, 128.8, 129.0, 129.5, 129.6, 131.2, 140.1, 142.8, 143.0,
143.5, 148.5, 170.3. UV/vis (HPLC, nm) λmax: 230, 259, 265,
272. 9,10-Diphenylanthracene (8m ). UV/vis (HPLC, nm)
Ack n ow led gm en t . Financial support by the
Deutsche Forschungsgemeinschaft is gratefully acknowl-
edged. We thank Drs. J . Lusztyk and L. J ohnston,
Ottawa, for their support with the LFP experiments and
Dr. K. U. Ingold, Ottawa, for helpful discussions.
λ
max: 354, 372, 392. 4a,10-Dihydroanthracene derivative (7j?).
UV/vis (HPLC, nm) λmax: 378.
1
Su p p or tin g In for m a tion Ava ila ble: Copies of H NMR
spectra of meso-3a , meso-3b, meso-3c, meso-3d , meso/rac-3e,
3j, 3l, meso-5, meso-6a , meso-6b, meso-6c, meso-6d , 6i, the
photolysis mixtures 3e f 6e, 3h f 9h , and 3l f 6l, and the
mixture of the thermal rearrangement meso-6a f rac-6a (19
pages). This material is contained in libraries on microfiche,
immediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
P h otolysis of 3k . 7,7,8-Triphenylbenzocyclobutene (6k ).
1H NMR (300 MHz, C6D6) δ: 5.48 (s, 1H, CH), 7.64-6.78 (m,
19H). 13C NMR (75 MHz, DEPT, C6D6) δ: 62.3 (CHPh), 68.2
(CPh2), 124.2, 124.4, 126.1, 126.4, 126.7, 127.6, 128.0 (2C),
128.4, 128.5, 128.6, 129.1, 129.4 (tert arom C), 139.6, 142.3,
145.0, 147.0, 149.9 (quart arom C). UV/vis (HPLC, nm) λmax
259 (sh), 265, 273. 9,10-Diphenylanthracene (8m ). UV/vis
(acetonitrile, nm) λmax: 338, 354, 373, 393.
P h otolysis of 3l. 7-Methyl-7,8,8-triphenylbenzocyclobutene
(6l). 1H NMR (200 MHz, C6D6) δ: 1.55 (s, 3H, CH3), 7.33-
6.74 (m, 15H), 7.67-7.61 (m, 4H). UV/vis (HPLC) λmax: 261,
268, 271 nm. R-Phenyl-o-(diphenylmethyl)styrene (9l). 1H
NMR (200 MHz, C6D6) δ: 4.89 (d, 2J ) 1.3 Hz, 1H), 5.54 (d, 2J
1.3 Hz, 1H), 5.71 (s, 1H), 7.33-6.74 (m, 19H). UV/vis (HPLC,
nm) λmax: 250, 222.
:
J O960573N
(29) WINSIM: Duling, D. R. Laboratory of Molecular Biophysics,
NIEHS, NIH, Research Triangle Park, NC. See also: Duling, D. R. J .
Magn. Reson. B 1994, 104, 105-110.
(30) Paul, T. Dissertation, Universita¨t - GH Essen, 1995.
(31) Kazanis, S.; Azarani, A.; J ohnston, L. J . J . Phys. Chem. 1991,
95, 4601-4610.