The photochemistry of indenyl alcohols and esters: Substituent effects on the competition between ion- and radical-derived products

Article abstract of DOI:10.1139/v03-131

The photochemistry of the indenyl acetates 1 and pivalates 2, substituted with X = H, 5-CH₃O, and 6-CH₃O, have been examined in both methanol and cyclohexane. The precursor alcohols 3 were also found to be photoreactive. Although only radical-derived products were obtained in cyclohexane, both ion- and radical-derived products were formed in methanol. The absence of significant fluorescence emission from any of the substrates 1, 2, and 3 indicates that the excited singlet states are highly reactive. A mechanism is proposed for the ion-derived products that proceeds through direct heterolytic cleavage to give an indenyl cation - carboxylate anion pair. The indenyl cations generated are anti-aromatic in the ground state and their efficient generation by this photochemical solvolysis is in sharp contrast to the very low reactivity of related ground-state substrates. For the pivalate esters 2, an excited-state migratory decarboxylation is proposed for the formation of tert-butyl derived products.

Full text of DOI:10.1139/v03-131

1083
The photochemistry of indenyl alcohols and
esters: Substituent effects on the competition
between ion- and radical-derived products
J.A. Pincock and I.S. Young
Abstract: The photochemistry of the indenyl acetates 1 and pivalates 2, substituted with X = H, 5-CH3O, and 6-CH3O,
have been examined in both methanol and cyclohexane. The precursor alcohols 3 were also found to be photoreactive.
Although only radical-derived products were obtained in cyclohexane, both ion- and radical-derived products were
formed in methanol. The absence of significant fluorescence emission from any of the substrates 1, 2, and 3 indicates
that the excited singlet states are highly reactive. A mechanism is proposed for the ion-derived products that proceeds
through direct heterolytic cleavage to give an indenyl cation – carboxylate anion pair. The indenyl cations generated
are anti-aromatic in the ground state and their efficient generation by this photochemical solvolysis is in sharp contrast
to the very low reactivity of related ground-state substrates. For the pivalate esters 2, an excited-state migratory
decarboxylation is proposed for the formation of tert-butyl derived products.
Key words: ester photochemistry, indenyl cations, indenyl radicals.
Résumé : La photochimie des acétates 1 et des pivalates 2 d’indényle, substitués avec X = H, 5-CH3O et 6-CH3O a
été étudiée dans le méthanol et dans le cyclohexane. On a trouvé que les alcools précurseurs 3 sont également photo-
réactifs. Dans le cas du cyclohexane on n’obtient que des produits d’origine radicalaire, par contre dans le cas du mé-
thanol on obtient des produits d’origine ionique et d’origine radicalaire. L’absence d’émission de fluorescence
significative à partir des substrats 1, 2 et 3 indique que les états singulets excités sont hautement réactifs. On propose
un mécanisme qui explique la formation des produits d’origine ionique via le clivage hétérolytique direct pour donner
une paire cation indényle – anion carboxylate. Les cations indényles générés sont anti aromatiques dans l’état fonda-
mental et leur production efficace par cette solvolyse photochimique est en nette opposition avec la très faible réactivité
des substances apparentées à l’état fondamental. Pour les esters pivalates 2, on propose une décarboxylation migratoire
de l’état excité pour expliquer la formation des produits dérivés du t-butyle.
Mots clés : photochimie des esters, cations indényles, radicaux indényles.
[Traduit par la Rédaction] Pincock and Young 1095
Introduction
nism where the major pathway for reactivity is homolytic
cleavage (khom) of the carbon—oxygen bond to give a radi-
cal pair (1, 2). Competition between electron transfer (ket) to
form an ion pair and decarboxylation ( kCO2) of the acyloxy
radical then controls product distribution. This latter process
serves as a radical clock with the rate constant differing by
many orders of magnitude, depending on the stability of the
The photochemistry of arylmethyl esters (ArCH2O-(CO)-
R) in nucleophilic solvents (1) and arylmethyl compounds
with leaving groups in general (ArCH2-LG) (2) has recently
attracted considerable interest. As outlined in the mechanism
in Scheme 1, products are derived from both radical pair and
ion pair (photosolvolysis) intermediates and, for esters at
least, only from the excited singlet state. Apart from a fun-
damental interest in the factors (substituents, solvent, leaving
group, excited-state multiplicity) that control the partitioning
between these two pathways, the potential of applying this
process to the design of photolabile protecting groups or
phototriggers has been explored (3, 4).
radical formed. Values range from ~106 s–1 for ArCO2 (5)
•
to ~109 s–1 for RCO2 with primary alkyl groups (6). For
•
cases that form highly stabilized radicals, such as
•
Ar2C(OH)CO2 , the decarboxylation occurs within a few pi-
coseconds; the rate constants approach those for barrier-free
unimolecular reactions (7). The rate of electron transfer also
ranges over several orders of magnitude and follows Marcus
theory in both the normal and the inverted region (8, 9). On
the basis of high-level MO calculations, this mechanism has
been questioned in favor of one involving direct heterolytic
cleavage (khet) from S1, particularly for substrates with meta-
methoxy substitutents (10, 11).
On the basis of extensive studies on a range of
functionalized arylmethyl esters, we have proposed a mecha-
Received 9 June 2003. Published on the NRC Research Press
Web site at http://canjchem.nrc.ca on 26 September 2003.
The high reactivity and the high yield of ion-derived prod-
ucts for the photochemistry of 9-fluorenol in mixed organic–
aqueous media has been interpreted as resulting from the
formation of a cyclically conjugated 4n π cation (12, 13).
J.A. Pincock1 and I.S. Young. Department of Chemistry,
Dalhousie University, Halifax, NS B3H 4J3, Canada.
1Corresponding author (e-mail: james.pincock@dal.ca).
Can. J. Chem. 81: 1083–1095 (2003)
doi: 10.1139/V03-131
© 2003 NRC Canada

1084
Can. J. Chem. Vol. 81, 2003
Scheme 1. General mechanistic scheme for the photolysis of benzylic substrates.
Recent picosecond laser flash photolysis (LFP) experiments
(14) have shown that in fact 9-fluorenol reacts to form the 9-
fluorenol cation very rapidly (<10 ps) in solvents (water,
TFE, HFIP) of high polarity (ENT ). The antiaromatic charac-
ter of the 9-fluorenyl cation has been discussed in a recent
review (15) with the conclusion that it will be attenuated rel-
ative to that for cyclopentadienyl and indenyl cations. In
fact, on the basis of recent calculations, the fluorenyl cation
has been described as nonaromatic by some criteria and
antiaromatic by others (16). The possible photochemical
generation of indenyl cations, which, from experiment and
calculation, have more antiaromatic character than fluorenyl
cations, is therefore of interest.
Results and discussion
Synthesis of alcohols and esters
The esters were synthesized from the corresponding alco-
hols 3 by reaction with the appropriate acid chloride in
benzene–pyridine, Scheme 2. The alcohols were obtained by
reduction of the ketones 4 with NaBH4 with added CeCl3. In
the absence of the CeCl3, considerable reduction of the
indenone double bond occurred (18). Two of the indenones
(4a: X = H and 4b: X = 5-CH3O) were prepared by a proce-
dure developed by Floyd and Allen (19) and used in our lab-
oratory previously (20). The indanone 5 (X = 6-CH3O)
could not be prepared by this method because the Friedel–
Crafts cyclization was not successful. Instead, the method of
Sam and Plampin (21) was used. The indenone 4c (X = 6-
CH3O) was then prepared from 5 by a bromination–dehydro-
bromination sequence.
We now report on the photochemistry of six indenyl esters
1a–c and 2a–c and three indenyl alcohols 3a–c in the sol-
vents methanol and cyclohexane. The photochemistry of the
alcohols was examined to allow a more direct comparison
with the photochemistry of 9-fluorenol. The choice of exam-
ining acetate esters 1 vs. pivalates 2 was made because the
rate constants of decarboxylation of their corresponding
acyloxy radicals differ by an order of magnitude (~109 s–1
Absorption and emission spectra of the esters and
alcohols in methanol
In general, indene derivatives have UV absorption spectra
with two bands, a weak one (⑀ ~ 102 (mol L–1)–1 cm–1) be-
tween 280 and 300 nm, corresponding to the symmetry-
for CH3CO2 vs. 1.1 × 1010 s–1 for (CH3)3CCO2 ) (6), and
these radicals are good probes for radical pair chemistry.
The choice of 5-methoxy vs. 6-methoxy was made in order
to examine the effect of the change in position of the
electron-donating group from para (5-methoxy) in 1b, 2b,
and 3b to meta (6-methoxy) in 1c, 2c, and 3c relative to the
arylmethyl center with the leaving group. The results will
also be compared with those obtained previously for the cor-
responding indanyl derivatives (17).
•
•
1
forbidden Lb band in benzene, usually partially overlapped
by a stronger one (⑀ ~ 104 (mol L–1)–1 cm–1) around 260 nm,
1
corresponding to the symmetry-allowed La band in benzene
(22). In methanol, the unsubstituted indenyl ester 1a has two
similar bands (λmax = 266 nm, ⑀ = 7130 (mol L–1)–1 cm–1 and
306 nm, 391 kJ mol–1 (the longest wavelength maximum in
the long wavelength band), ⑀ = 950 (mol L–1)–1 cm–1), but
© 2003 NRC Canada

Pincock and Young
1085
Scheme 2. Synthesis of the esters 1a–c and 2a–c and the alcohols 3a–c.
these are shifted to somewhat longer wavelength than that
for 2-methylindene (22) in cyclohexane (λmax = 257 nm, ⑀ =
11 700 (mol L–1)–1 cm–1 and 295 nm, ⑀ ~ 500 (mol L–1)–1 cm–1)
or cis-phenylpropene (23) in hexane (λmax = 256 and
288 nm). The long wavelength band has a series of obvious,
but broad, vibrational bands in contrast to the shorter wave-
length band, which does not. The methoxy-substituted esters
1b and 1c have similar absorption bands (see Experimental
section), although the long wavelength band lacks vibra-
tional structure and moves to even lower energy (320 nm,
373 kJ mol–1) for the 6-methoxy isomer 1c, as expected
when compared with a similar shift to longer wavelength for
trans-1-(4-methoxyphenyl)propene (λmax = 298 nm) relative
to trans-1-phenylpropene itself (λmax = 284 nm) (23).
Fluorescence spectra for the three indenyl esters 1 in
methanol revealed only extremely weak, if any, emission.
The quantum yields of emission were estimated to be less
than 0.001 when compared with that for naphthalene (0.21).
The alcohols 3, precursors for the esters, also did not fluo-
resce significantly. In contrast, 3-methylindene, which is
photochemically unreactive in hexane, has a quantum yield
of fluorescence of 0.077 and excited singlet state lifetimes of
13.9 ns (22); the major process competing with fluorescence
from the excited singlet state is likely intersystem crossing.
© 2003 NRC Canada

1086
Can. J. Chem. Vol. 81, 2003
The very inefficient fluorescence for the indenyl compounds
studied here strongly suggests that they are undergoing rapid
(k > 109 s–1) excited-state cleavage of the indenyl C1 to oxy-
gen bond. The absence of measurable fluorescence unfortu-
nately prevents the determination of excited-state lifetimes
and rate constants for excited-state processes.
dependent of which alcohol was used as the starting
material, eq. [2].
The compound obtained was shown to be the 5-methoxy
[2]
Photochemistry of the alcohols 3 in methanol and
cyclohexane
The photolysis of the indenyl alcohols in either methanol
or cyclohexane gave the products expected from arylmethyl
photochemistry, as shown, in a general way, in eq. [1].2
[1]
isomer, 6b, by NOE difference spectra: irradiation of H1
(δ = 4.82 ppm) gave a strong enhancement for H7, which is
assigned by its higher chemical shift (δ = 7.30 ppm) relative
to that of H6 (δ = 6.64 ppm) for the two aromatic protons,
which are ortho to each other and therefore have a coupling
constant of 8.0 Hz. This acid-catalyzed reaction, which pre-
sumably proceeds through the indenyl cation, raises an inter-
esting point for reactions that generate indenyl cations or
radicals at C1 of the 5-methoxy and 6-methoxy substrates.
Because of delocalization through C1, C2, and C3 of these
intermediates, both substrates will generate the same inter-
mediate for either the cation 10 or the radical 11, Fig. 1.
This observation prompted us to examine the charge and
spin distribution in these species by PM3 calculations.3 As
shown in Fig. 1a, nine resonance forms can be drawn for an
indenyl cation with a positive charge on each one of the nine
carbons. From the calculations for the indenyl cation (C2v
symmetry), the positive charge is mainly on C1 and C3
(+0.34 for each, 0.68 total, with hydrogen charges summed
into the carbons), as indicated by the first two resonance
forms. This suggests that, to a first approximation, the
indenyl cation can be viewed as an allyl cation fused to a
benzene ring. For 5-substituted indenyl cations, the C2v sym-
metry is no longer present. For these preliminary calcula-
tions, as shown in Fig. 1b, we chose the substituent as OH to
model the OCH3 group in order to minimize conformational
complications. For the radical, the spin population at C1
(0.89) and C3 (0.72) is similar and delocalization results in a
structure with similar bond lengths in the upper and lower
part, i.e., the OH group only slightly perturbs the plane pass-
ing through C2 and bisecting the C5—C6 bond. In contrast,
the calculations of the atomic charge density for the cation
reveal a much higher positive charge at C1 (+0.21) than C3
(–0.03); values for C1 (+0.44) and C3 (+0.25) are obtained
with hydrogen charges summed into the carbons. In addi-
tion, the bond lengths indicate that the indenyl ring system
can be pictured as almost two independent π-systems. The
first, a 5-hydroxy pentadienyl cation for the string of car-
bons from C1 through the ring fusion to C7, C6, and C5,
with almost equal bond lengths (1.41 ± 0.02 Å) and the sec-
The methyl ethers 6, formed only in methanol, are of par-
ticular interest because they are clearly derived from indenyl
cations. The hydrocarbon products 7–9 are radical derived.
The reactions are relatively efficient; irradiation of 60 mg of
3a (X = H) in 100 mL of methanol with 254 nm lamps in a
Rayonet reactor resulted in 72% conversion in 15 min. The
difficulties in completely characterizing the photoproducts of
these reactions can be exemplified by this case. At this ex-
tent of conversion, the mass balance (GC peak areas) was
only 60%. This is characteristic of the photochemistry of
indene derivatives where quantum yields for disappearance
of starting material are often considerably higher than those
for appearance of volatile products (24), presumably as a
consequence of oligomerization processes. Moreover, com-
plete separation by flash or radial chromatography of the
non-polar hydrocarbon products was not possible. GC–MS
analysis of mixtures of variable composition was invaluable
for preliminary structural assignment. For the unsubstituted
substrate, 3a, the hydrocarbons 7 (X = H) (commercial sam-
ple, Aldrich) and the dimers 9 (X = H) (meso and dl
1
diastereomers) are known (25, 26), and literature H NMR
spectra were helpful in identifying components in the mix-
tures obtained by chromatography.
The methyl ethers 6 were synthesized from the indenols 3
using trimethyl orthoformate under acidic conditions (aque-
ous HClO4) (27). An attempt to alkylate 3a (X = H) using
NaH–DMSO followed by CH3I resulted in 2,2-
dimethylindanone, confirming that the indenyl hydrogen at
C1 is more acidic (pKa = 20) than an alcohol (pKa
[CH3OH] = 29) in DMSO (28). Using the trimethyl ortho-
formate procedure for the indenols 3b (X = 5-OCH3) and 3c
(X = 6-OCH3) resulted in the isolation of only one ether in-
2 The products 12 in eq. [1] cannot be formed from the alcohols 3 but only from the esters 1 and 2, as will be discussed later.
3 We thank Dr. F. Ban, Dalhousie University, Department of Chemistry, Halifax, Nova Scotia, for these calculations.
© 2003 NRC Canada

Pincock and Young
1087
Fig. 1. (a) Resonance structures for the indenyl cation and (b) bond lengths (Å), atomic charges, and spin densities for the 5-
hydroxy-1-indenyl cation and radical. Values in brackets for the cation have the charge for the hydrogen atom summed into the value
for the carbon.
ond, an isolated butadiene fragment from C2 to C4 with al-
ternating short (~1.35 Å) and long (1.49 Å) bond lengths;
i.e., the OH group severely perturbs the structure on either
side of the same plane (defined above) for the radical. The
high positive charge density at C1 of the cation, which was
obtained from these calculations, rationalizes nicely why
both indenols 3 (X = 5-OCH3 and X = 6-OCH3) give the
same ether 6 (X = 5-OCH3) in the synthesis, eq. [2], and in
the photochemistry, eq. [1]. The calculations for the hy-
droxyindenyl radical 10 will also be relevant to the photo-
lysis results described below. Higher level MO calculations
on the charge and spin distribution in substituted indenyl
cations and radicals are in progress.
secondary photochemistry of several kinds. First, alkyl
indenes are well known to photoisomerize (24). For in-
stance, 2-methylindene 7 (X = H) isomerizes to both 1- and
3-methylindene (Φ = 0.13) on photolysis in hexane and even
isomerizes to 2-methyleneindane under acidic conditions
(29). Similarly, 1,2-dimethylindene 12a (X = H, R = CH3)
isomerizes to the 2,3-isomer (24). We have confirmed these
previous observations by the photolysis of 2-methylindene
in methanol, which gave three isomers (GC–MS) along with
2-methoxy-2-methylindane by anti-Markovnikov addition of
methanol, also as reported previously (30). In agreement
with these observations, at higher conversions, the initially
formed products decreased in yield as the other isomers ap-
peared (GC–MS). Second, photolysis in methanol of the
Another issue in the photolysis experiments was that of
© 2003 NRC Canada

1088
Can. J. Chem. Vol. 81, 2003
Table 1. Product yields for the photolysis of the alcohols 3 in cylohexane and methanol.
X
Solvent
% Con (t)a
6
7b
8b
9c
7
3a
3a
3b
3b
3c
3c
H
CH3OH
C6H12
27 (5)
30 (5)
21 (2)
58 (5)
12 (2)d
18 (5)e
7
35
50 (1:1.2)
H
—
80
—
80
—
15
60
25 (1:1.2)
5-CH3O
5-CH3O
6-CH3O
6-CH3O
CH3OH
C6H12
20 (1:1.2)
10 (1:1.6)
20 (1:1.5)
10 (1:1.4)
—
—
—
—
—
90 (1:1.3)
—
CH3OH
C6H12
90 (1:1.1)
Note: Yields are normalized to 100%. Estimated error ±5%.
aPercent conversion (photolysis time) for ~50 mg of substrate in 100 mL of solvent.
bNumbers in brackets refer to the ratio of the two possible regioisomers.
cNumbers in brackets refer to the ratio of the two possible diastereomers.
dRatio of 3c:3b = 12:1 at this percent conversion.
eRatio of 3c:3b = 7.8:1 at this percent conversion.
ether 6a (X = H), the ion-derived product, resulted in forma-
tion of the radical-derived products 7, 8, and 9, eq. [3].
The major product in this photolysis, 2-methylindene 7
(X = H), is probably derived by disproportionation of the
indenyl–methoxy radical pair, as proposed previously for the
photolysis of 9-methoxyfluorene derivatives (12, 13). Third,
as indicated in Table 1, the 6-methoxy alcohol 3c converts
photochemically to the 5-methoxy isomer 3b. For instance,
at 14% conversion in cyclohexane, the ratio of 3c:3b was
7.8:1. As a consequence of these various secondary photo-
chemistries, only product yields obtained at low conversions
(<5%) give values that are useful for mechanistic consider-
ation. These values are given in Table 1. Because the per-
centage conversion is too small to reliably determine the
mass balance, the yields are normalized to 100%.
solvents because the process would not be exothermic (BDE
= 351 kJ mol–1 for the C1—H bond in indene). Reaction of
the two radicals by disproportionation then gives 7, whereas
coupling gives 8, in a ratio of 1:9 in cyclohexane. In metha-
nol, only disproportionation was observed. The other major
products that are radical derived are the two indenyl dimers,
9 (X = H), now diasteromers (dl and meso) because
regioisomers are not possible. For reasons that are not obvi-
ous, similar dimers were not observed for the methoxy-
substituted substrates.
The major conclusion from these results is that indenyl
cations are efficiently formed in the photochemistry of
indenyl alcohols. Similar cations have been generated by
ground-state solvolysis of the indenyl compounds 13, and
rate retardations of approximately 10–5 have been measured
relative to the corresponding indanes 14 in 80% aqueous ac-
etone, reflecting the antiaromatic destabilization of the cat-
ion derived by solvolysis of 13 (32).
[3]
Product studies indicated that two regioisomeric alcohols
are formed from both 13 (X = CH3) and 13 (X = CH3O), al-
though the experimental evidence for the latter compound is
not conclusive. Moreover, these solvolysis reactions were
carried out at high temperature (100 °C for 22 days), condi-
tions where 5- and 6-substituted indene derivatives may be
equilibrating thermally by 1,5-hydrogen migrations involv-
ing isoindene derivatives (33). The question as to whether
The products shown in eq. [1] and Table 1 for the photo-
lysis of the alcohols 3 are easily explained by formation of
indenyl radical and cation intermediates. A mechanism is
outlined in Scheme 3, a specific example of the general
Scheme 1. The most interesting observation is the very high
yield in methanol of the ion-derived product, the methyl es-
ter 6, particularly from the methoxy-substituted substrates
3b and 3c (82%). Also, the same product yields are obtained
for both substrates 3b and 3c, indicating that both the
indenyl radical and cation lose, as expected, the memory of
the substrate from which they originated. The radical-
derived products 7 and 8 from 3b and 3c are mixtures (ra-
tio ~ 1.3:1) of two regioisomers, as expected for the unequal
spin distribution at C1 and C3 for the methoxy-substituted
indenyl radical. The formation of these products is rational-
ized (Scheme 3) by exothermic hydrogen abstraction (31) by
the hydroxyl radical from the solvents methanol (bond disso-
ciation energy (BDE) = 393 kJ mol–1) and cyclohexane (BDE=
401 kJ mol–1) to give water (BDE = 497 kJ mol–1) and the
hydroxymethyl and cyclohexyl radical, respectively. The
indenyl radical will not abstract a hydrogen atom from these
the ion pair is generated by direct heterolytic cleavage (khet
,
Scheme l) or by electron transfer from an initially formed
radical pair (ket) cannot be answered from these results.
However, the clear evidence from LFP experiments for di-
rect heterolytic cleavage to the ion pair for 9-fluorenol in po-
lar protic solvents (14) suggests that the first pathway is
possible. Moreover, the very high oxidation potential ex-
pected for the indenyl radical makes the latter seem unlikely.
Therefore, the efficient photochemical formation of the
indenyl cations (Scheme 3) would seem to be another exam-
© 2003 NRC Canada

Pincock and Young
1089
Scheme 3. Mechanistic pathways for the photolysis of the alcohols 3a–c in methanol and cyclohexane.
ple of the generation of a 4n π cation, first proposed by Wan
and co-workers for the photodehydroxylation of fluorenol
(12, 13).
Table 2. Product yields for the photolysis of the esters 1 and 2
in cyclohexane.
X
R
7a
9a
12b
1ac
2a
H
H
CH3
—
15 (1:1.3)
80
Photochemistry of the esters 1 and 2 in methanol and
cyclohexane
(CH3)3C 20
15 (1:1.3)
65
1b
2b
5-CH3O CH3
—
—
—
—
—
100 (1:2.3)
85 (1:2.5)
100 (1:1.5)
75 (1:2.1)
The photolysis of the indenyl esters 1 and 2 in either
methanol or cyclohexane again gave the products expected
from arylmethyl photochemistry, also shown in eq. [1].
Again, the reactions are relatively efficient; irradiation of
100 mg of 1a (X = H, R = CH3) in 100 mL of methanol with
254 nm lamps in a Rayonet reactor resulted in 50% conver-
sion in 15 min. The same difficulties as for the alcohols 3
(lack of separation of non-polar products and secondary
photochemistry) were encountered in completely character-
izing the photoproducts for the esters. Moreover, an addi-
tional secondary photochemical process was revealed,
involving the coupling products 12b (R = C(CH3)3). The ra-
tio of these two tert-butyl products derived from the 5- and
6-methoxy pivalate esters changed with the percent conver-
sion. Parallel to this process, the yield of the two isomeric 5-
and 6-methoxy-2-methylindene derivatives 7 increased. For
instance, the yield of 7 for photolysis of the pivalates 2b and
2c in methanol was essentially zero at very low conversion
(<5%) but reached 20–25% at high conversion (>80%).
Finally, the 6-methoxy esters 1c and 2c were converted to
the corresponding 5-methoxy substrate 1b and 2b (see foot-
notes in Tables 2 and 3). This observation is expected on the
5-CH3O (CH3)3C 15 (1:1.1)
1cd 6-CH3O CH3
2ce 6-CH3O (CH3)3C 25 (1:1.1)
—
Note: Yields are for low conversion (<10%) and normalized to 100%.
Estimated error ±5%.
aNumbers in brackets refer to the ratio of the two possible
diastereomers.
bNumbers in brackets refer to the ratio of the two possible regioisomers.
cFor 1a, 8% of the material was 1-cyclohexyl-2-methylindene 8, a prod-
uct not detected for the other five compounds.
dRatio of 1c:1b = 5.8:1 at 41% conversion.
eRatio of 2c:2b = 3.9:1 at 75% conversion.
nism in Scheme 4, where the radical paired with the indenyl
one is now an acyloxy, rather than hydroxyl as it was for the
indenols in Scheme 3. For the acetates 1 (R = CH3), decar-
boxylation gives the methyl radical leading to the coupling
product 12a (R = CH3) (two regioisomers for X = 5-CH3O
and 6-CH3O). For the pivalates 2 (R = C(CH3)3), decar-
boxylation now gives the tert-butyl radical, leading to the
coupling product 12b (R = C(CH3)3) and disproportionation
to 2-methylindene 7 (again, two regioisomers for X = 5- and
6-CH3O for both 7 and 12b). As in the photolysis of the
indenols 3 (Table 1), the out-of-cage dimers were only ob-
tained for the unsubstituted esters 1a and 1b. The very simi-
lar yields of 7 and 12 obtained for both the 5- and 6-
methoxy substrates, independent of which one is the starting
material, indicates that the indenyl radical again loses the
memory of its precursor.
basis of the internal return processes (kicom and krcom
,
Scheme 1) previously observed for the intermediates in ester
photochemistry (34–36). Again as a consequence of these
competing processes, only product yields obtained at very
low conversions (<5%) are useful for mechanistic consider-
ation. These values are given in Table 2 for reactions in cyc-
lohexane and in Table 3 for those in methanol. Because the
percentage conversion is too small to determine reliably, the
yields are normalized to 100%.
The products obtained for the photolysis of the indenyl es-
ters in methanol are given in Table 3. Particularly striking is
the very high yield of the ion-derived product, the methyl
ether 6, particularly from the 5- and 6-methoxy acetates, 1b
The photolysis products obtained for the esters 1 and 2 in
cyclohexane are easily explained by the radical pair mecha-
© 2003 NRC Canada

1090
Can. J. Chem. Vol. 81, 2003
Table 3. Product yields for the photolysis of the esters 1 and 2 in methanol.
X
R
6
7
9a
12b
Ion:radicalc
0.18:1
Ion:radicald
1.4:1
15
1a
2a
1b
2b
1ce
2cf
H
CH3
—
10
—
—
—
—
25 (1:1.4)
60
H
(CH3)3C
CH3
—
20 (1:1.2)
70
<0.02:1
3.0:1
0.21:1
0.10:1
<0.02:1
2.3:1
5-CH3O
5-CH3O
6-CH3O
6-CH3O
75
—
—
—
—
25 (1:2.1)
80 (1:1.6)
—
(CH3)3C
CH3
20
100
30
0.25:1
>50:1
(CH3)3C
70 (1:1.2)
0.33:1
0.25:1
Note: Yields are for low conversion (<5%) and normalized to 100%. Estimated error ±5%.
aNumbers in brackets refer to the ratio of the two diastereomers.
bNumbers in brackets refer to the ratio of the two possible regioisomers.
cRatio of ion- to radical-derived products for the indenyl esters 1 and 2.
dRatio of ion- to radical-derived products for the analogous indanyl esters, 15 and 16, ref. 17.
eRatio of 1c:1b = 8.7:1 at 48% conversion.
fRatio of 2c:2b = 40:1 at 32% conversion.
Scheme 4. Mechanistic pathways for the photolysis of the esters 1a–c and 2a–c in cyclohexane.
and 1c. The ratio of ion- to radical-derived product is signif-
icantly larger for the methoxy-substituted substrates than the
corresponding indanyl (Table 3) derivatives 15 and 16 (17).
In fact, the ion-derived ether 6b (X = 5-CH3O) is the only
product observed (at low conversion) for the acetate 1c (X =
6-CH3O). The change in product distribution for the change
in substrate from acetate esters 1 to pivalate esters 2 has
been used effectively in the past by us (1) and others (37) to
test for the intervention of acyloxy radicals. Their very dif-
ferent rates of decarboxylation serve as radical clocks for
other possible processes of the radical, in particular, the
electron transfer process (ket, Scheme 1) that converts the
radical pair to the ion pair.
6-methoxy pivalates esters 2b and 2c at low conversion in
methanol suggests that free tert-butyl radicals are not
formed. This observation emphasizes why product yields at
low conversion are essential for understanding the photo-
chemistry involved. As mentioned above, the yield of 7 in-
creases with the extent of conversion by a process of
secondary photochemistry involving 12b (X = C(CH3)3).
The parallel change in the ratio of the two regioisomeric
coupling products 12 indicates that this secondary photo-
chemistry is homolytic carbon—carbon bond cleavage to
generate the indenyl–tert-butyl radical pair, allowing for
both disproportionation and regioisomeric equilibration to
occur.
The question then arises as to the mechanism of formation
of the tert-butyl coupling products if tert-butyl radicals are
not involved in the primary photochemistry of the methoxy
pivalate esters in methanol. A tentative proposal is an
excited-state migratory (from C1 to C2) decarboxylation of
the tert-butyl group via a five-membered transition state 17
(shown in Scheme 5 for the 6-methoxypivalate 2c) to form
an isoindene derivative 18. Excited state migrations from C1
to C2 are well known for phenyl-substituted indenes (38).
The resulting isoindenes would be thermally unstable and re-
vert to indene. On the basis of migratory aptitudes (CH3 <
However, the surprising observation of the absence of the
disproportionation products 7 in the photolysis of the 5- and
© 2003 NRC Canada

Pincock and Young
1091
Scheme 5. Proposed mechanism for the formation of the coupling products 12 and 12′ for the photolysis of the ester 2c in methanol.
CH2CH3 < CH(CH3)2) previously determined (33) for the
thermal reversion of isoindenes to indenes, the tert-butyl
group in 18 will migrate in preference to methyl, giving the
two observed regioisomeric products, 12b and 12b′. The
process described by transition state 17 is not observed and
not possible for other benzylic substrates, which lack the re-
quired indenyl type conjugation. In those cases, dispropor-
tionation always accompanies coupling of tert-butyl
radicals, as expected (39).
This process is reminiscent of that proposed for the ther-
mal decomposition of peresters, which are concerted (40,
41), where the potential acyloxy radical will fragment very
rapidly because of the stability of the radical generated, 19.
These reactions have negative ρ values (approximately –1.1),
indicating that the transition states have some positive char-
acter on carbon and negative on oxygen (42). The photo-
chemical migration suggested by structure 17 in Scheme 5
might also have a substituent effect, which could potentially
explain why the yield of the isomers 12 are slightly different
for the 5-methoxy 2b (80%) vs. the 6-methoxy (70%)
pivalate ester 3b. The substitutent effect could also rational-
ize why the unsubstituted substrate 2a seems to undergo nor-
mal homolytic bond cleavage and give both coupling (12)
and disproportionation (7) products.
The mechanistic proposals for the indenyl esters can then
be summarized as follows: In cyclohexane, all three esters
react by homolytic cleavage to give radical-derived products
as shown in Scheme 4. In methanol, as outlined in
Scheme 6, the unsubstituted esters 1a and 2a react by the
same mechanism for the radical-derived products, but now
formation of the ion pair leads to low yields (15%) of the
methoxy ether 6 for the acetate 1a. The decrease in the yield
of 6, as a consequence of the change from acetate 1a to
pivalate 2a, suggests that the radical clock method is opera-
tive and that the ion pair is formed (at least partially) from
the radical pair by redox electron transfer (ket, Scheme 1) in
competition with the rate of decarboxylation, which is faster
by an order of magnitude for the (CH3)3CCO2 radical.
Two other potential photochemically generated intermedi-
ates that have not yet been discussed should be mentioned as
possible precursors for the ion-derived ether product in
methanol. The first, 20, an isoindene, would result from mi-
gration of the acyloxy group from C1 to C2 of the indene.
The second, 21, analogous to the intermediate 22 ob-
served in the photolysis of 3,5-dimethoxybenzyl acetate in
methanol and other alcohol solvents (43), would be formed
by acyloxy migration from C1 to C7. A similar intermediate
has been reported for the photolysis of 9-fluorenol in metha-
© 2003 NRC Canada

1092
Can. J. Chem. Vol. 81, 2003
Scheme 6. Mechanistic pathways for the photolysis of the esters 1a–c and 2a–c in methanol.
nol (44) and non-acidic zeolites (45). Either of these could
be very reactive (thermally or photochemically) in methanol
as solvent. For instance, the half life for ground-state
solvolysis of 22 in methanol is approximately 2 min. We are
currently conducting nanosecond laser flash photolysis ex-
periments to examine the reactive intermediates for the
photochemistry described above.
meta (and ortho) to electron-donating substituents. It seems
to be playing an important role in the photochemistry of
these indenyl esters. This suggestion is in marked contrast to
our proposal for all previously studied ester substrates where
homolytic cleavage followed by electron transfer is consid-
ered to be the dominant mechanistic pathway for the forma-
tion of ion-derived products. Although the limited data
available make these mechanistic conclusions somewhat
speculative, the fact that products derived from indenyl cat-
ions are formed in these photoreactions is certain. This may
well be a consequence of the antiaromatic character of these
4n π cations, in agreement with Wan and co-workers’ pro-
posal for 9-fluorenyl derivatives (12, 13).
Experimental section
General procedures
Conclusions
Proton (1H) and carbon (13C) nuclear magnetic resonance
(NMR) spectra were obtained in CDCl3 on an AC 250 F
NMR spectrometer in automation mode. Ultraviolet (UV)
spectra were obtained in 1 cm quartz cuvettes on a Varian-
Cary Bio 100 spectrometer. Fluorescence measurements
were done in methanol, with non-degassed samples, using a
PTI spectrometer at 25 °C. Corrected spectra were obtained.
For the 5- and 6-methoxy-substituted esters in methanol,
the proposed mechanism is that direct heterolytic cleavage
from the excited state to form an ion pair (Scheme 6) com-
petes with migratory decarboxylation of the ester to give the
isoindene derivative 18 (as shown in Scheme 5). This latter
process is less important for the acetates 1b and 1c than the
pivalates 2b and 2c, and therefore, the acetates give very
high yields of the ion-derived ether 6, reaching 100% for the
6-methoxy acetate 1c! This substrate has the activating
group meta to the reactive benzylic centre. The photochemi-
cal “meta effect”, as originally proposed by Zimmerman and
co-workers (46) for methoxy-substituted benzylic acetates
and supported by studies of methoxy-substituted benzylic al-
cohols by Wan and co-workers (47), predicts enhanced effi-
ciency for heterolytic cleavage reactions for leaving groups
GC–MS analyses were performed on
a PerkinElmer
Autosystem XL instrument with a mass selective detector.
The column used was a Supelco 30 m × 0.25 mm MDN-5S
5% phenyl methylsiloxane, film thickness 0.50 µm, tempera-
ture programmed: 60 °C for 1 min; 20 °C increase per min
to 240 °C; 240 °C for 10 min. GC–FID analyses were done
in a similar way except on a Supelco DB200 column. Mass
spectral data are reported in units of mass over charge (m/z)
for all values between 50 and the molecular ion if greater
© 2003 NRC Canada

Pincock and Young
1093
than 10% of the base peak. Intensities are reported as a per-
cent of the base peak.
(14), 89 (10), 77 (21), 63 (16), 51 (15). HR-MS calcd. for
C10H10O: 146.0731; found: 146.0741.
5-Methoxy-2-methyl-1-indenol 3b
Indenone synthesis
mp 77–78 °C. 1H NMR (CDCl3) δ: 7.32 (d, 1H, J =
7.9 Hz), 6.68 (d, 1H, J = 2.3 Hz), 6.61 (dd, 1H, J1 = 7.9,
J2 = 2.3 Hz), 6.27 (s, 1H), 4.78 (broad s, 1H), 3.79 (s, 3H),
2.05 (broad s, 3H), 1.64 (broad s, 1H). 13C NMR (CDCl3) δ:
160.3, 149.9, 144.74, 137.3, 126.6, 123.9, 109.1, 106.9,
78.2, 55.4, 13.9. GC–MS: 176 (70), 175 (27), 174 (13), 162
(11), 161 (100), 160 (22), 159 (11), 146 (19), 145 (30), 133
(19), 132 (10), 131 (20), 118 (13), 116 (12), 115 (35), 105
(11), 103 (24), 102 (11), 91 (12), 89 (13), 79 (12), 78 (10),
77 (34), 75 (10), 63 (20), 51 (13). HR-MS calcd. for
C11H12O2: 176.0837; found: 176.0835.
2-Methyl-1-indenone 4a and 5-methoxy-2-methyl-1-
indenone 4b
These indenones were prepared using the procedure of
Floyd and Allen (), outlined in Scheme 2.
6-Methoxy-2-methyl-1-indenone 4c
The precursor to this indenone, 6-methoxy-2-methyl-1-
indanone 5 (X = 6-CH3O), was prepared using the procedure
of Sam and Plampin (), outlined in Scheme 2. To a solution
of 13.9 g (79 mmol) of 5 (X = 6-CH3O) in 170 mL of CCl4
was added 14.4 g (90 mmol) of bromine in 70 mL of CCl4.
After the addition was complete, the mixture was stirred at
room temperature for 2 h. The solvent and excess bromine
were removed under reduced pressure and 20.5 g (79 mmol,
6-Methoxy-2-methyl-1-indenol 3c
1
mp 103–105 °C. H NMR (CDCl3) δ: 7.08 (d, 1H, J =
2.4 Hz), 7.01 (d, 1H, J = 8.0 Hz), 6.74 (dd, 1H, J1 = 8.0,
J2 = 2.4 Hz), 6.27 (s, 1H), 4.81 (d, 1H, J = 8.9 Hz), 3.80 (s,
3H), 2.04 (broad s, 3H), 1.49 (d, 1H, J = 8.9 Hz). 13C NMR
(CDCl3) δ: 158.0, 147.1, 146.1, 135.7, 126.5, 120.4, 113.1,
110.7, 79.0, 55.6, 13.7. GC–MS: 177 (10), 176 (87), 175
(18), 162 (11), 161 (100), 160 (33), 159 (18), 146 (15), 145
(40), 133 (23), 131 (12), 118 (17), 117 (16), 116 (19), 115
(60), 105 (13), 103 (20), 102 (11), 91 (13), 79 (11), 77 (24),
63 (17), 51 (16). HR-MS calcd. for C11H12O2: 176.0837;
found: 176.0841.
1
100%) of the crude bromide, as a brown solid, resulted. H
NMR (CDCl3) δ: 7.22–7.32 (m, 3H), 3.85 (s, 3H), 3.72 (d,
1H, J = 18 Hz), 3.41 (d, 1H, J = 18 Hz), 1.95 (s, 3H). 13C
NMR (CDCl3) δ: 200.4, 160.0, 141.8, 133.8, 127.1, 125.3,
106.7, 60.3, 55.7, 45.8, 27.0. GC–MS: 256 (19), 254 (18),
176 (58), 175 (100), 174 (22), 161 (96), 148 (20), 147 (41),
146 (10), 133 (29), 132 (23), 131 (41), 117 (16), 115 (32),
105 (10), 104 (11), 103 (42), 102 (14), 91 (36), 89 (14), 79
(19), 78 (29), 77 (77), 76 (18), 75 (17), 74 (20), 65 (13), 64
(10), 63 (45), 62 (15), 51 (34), 50 (13). HR-MS calcd. for
C11H11O2Br: 253.9942; found: 253.9934. A mixture of
20.5 g (79 mmol) of crude bromoindanone, 12.1 g
(139 mmol) LiBr, and 10.4 g (140 mmol) Li2CO3 in 175 mL
of DMF was heated between 135–140 °C for 3 h under a ni-
trogen atmosphere. After cooling, the mixture was poured
into 300 mL of water and extracted three times with ether.
The combined ether layers were washed twice with water
and dried with MgSO4. Evaporation of the ether produced
10.4 g of thick red oil that contained crystals. This material
was purified by flash chromatography on silica gel produc-
ing 3.60 g (21 mmol, 26%) of bright orange needles: mp 82–
General method for the preparation of esters 1a–c and
2a–c
The corresponding acid chloride (4 mmol) in 5 mL of dry
benzene was added to a solution of the corresponding
indenol (2 mmol) and pyridine (4 mmol) in 10 mL of dry
benzene. The solution was stirred overnight at room temper-
ature, then 50 mL of water was added, and the two layers
were separated. The benzene layer was washed twice with
10% aqueous HCl and once with saturated aqueous sodium
bicarbonate and finally with water. The esters were purified
by column chromatography on silica gel. The yields were
50–75%. Solid samples were recrystallized and oils were
bulb-to-bulb distilled.
1
84 °C. H NMR (CDCl3) δ: 7.06 (q, 1H, J = 1.8 Hz), 6.97
(d, 1H, J = 2.4 Hz), 6.78 (d, 1H, J = 7.9), 6.67 (dd, 1H, J1 =
2.4, J2 = 7.9 Hz), 3.77 (s, 3H), 1.81 (d, 3H, J = 1.8 Hz). 13C
NMR (CDCl3) δ: 198.6, 160.1, 144.1, 136.6, 134.9, 132.7,
121.6, 116.1, 110.9, 55.6, 10.0. GC–MS: 175 (12), 174
(100), 159 (40), 131 (92), 115 (10), 103 (74), 102 (21), 77
(64), 76 (21), 75 (20), 63 (25), 62 (12), 51 (18). HR-MS
calcd. for C11H10O2: 174.0681; found: 174.0675.
2-Methyl-1-indenyl acetate, 1a
bp 87–91 °C at 2–3 mmHg (1 mmHg = 133.322 Pa). UV
(methanol) λmax (nm) (⑀ ((mol L–1)–1 cm–1)): 266 (7130), 306
1
(952). H NMR (CDCl3) δ: 7.04–7.36 (m, 4H), 6.40 (s, 1H),
6.13 (s, 1H), 2.16 (s, 3H), 1.97 (s, 3H). 13C NMR (CDCl3) δ:
171.4, 144.4, 143.7, 142.1, 129.3, 128.8, 125.1, 124.2,
120.3, 78.4, 21.0, 14.0. GC–MS: 188 (15), 147 (11), 146
(100), 145 (29), 131 (68), 129 (21), 128 (74), 127 (16), 117
(10), 115 (31). HR-MS calcd. for C12H12O2: 188.0837;
found: 188.0843.
Indenol syntheses
Indenols
3 were prepared from the corresponding
indenones using Luche’s reagent (18). The yields were 60–
95%.
2-Methyl-1-indenol 3a
mp 84–86 °C. H NMR (CDCl3) δ: 7.09–7.50 (m, 4H),
6.33 (broad s, 1H), 4.84 (broad s, 1H), 2.07 (broad s, 3H),
1.66 (broad s, 1H). 13C NMR (CDCl3) δ: 148.3, 145.2,
143.0, 128.5, 127.0, 124.9, 123.3, 120.1, 78.9, 13.8. GC–
MS: 146 (86), 145 (38), 132 (10), 131 (100), 129 (13), 128
(22), 127 (14), 117 (15), 116 (11), 115 (46), 103 (31), 91
5-Methoxy-2-methyl-1-indenyl acetate, 1b
UV (methanol) λmax (nm) (⑀ ((mol L–1)–1 cm–1)): 279
1
1
(2880, 306 (1560). H NMR (CDCl3) δ: 7.27 (d, 1H, J =
7.9 Hz), 6.71 (d, 1H, J = 2.4 Hz), 6.59 (dd, 1H, J1 = 2.4,
J2 = 7.9 Hz), 6.38 (broad s, 1H), 6.07 (broad s, 1H), 3.79 (s,
3H), 2.17 (s, 3H), 1.98 (broad s, 3H). 13C NMR (CDCl3) δ:
171.6, 160.7, 145.5, 134.1, 129.1, 125, 109.4, 107.2, 78.1,
© 2003 NRC Canada

1094
Can. J. Chem. Vol. 81, 2003
55.4, 21.1, 14.2. GC–MS: 218 (23), 177 (12), 176 (100), 175
(24), 161 (65), 160 (19), 159 (26), 158 (24), 145 (22), 144
(11), 117 (10), 116 (26), 115 (66), 91 (12). HR-MS calcd.
for C12H14O3: 218.0943; found: 218.0931.
also isolated from photolysis of 6-methoxy-2-methyl-1-
indenol in methanol.
1-Methoxy-2-methylindene
1H NMR (CDCl3) δ: 7.02–7.42 (m, 4H), 6.42 (broad s,
1H), 4.84 (broad s, 1H), 3.02 (s, 3H), 2.02 (broad s, 3H). 13
C
6-Methoxy-2-methyl-1-indenyl acetate, 1c
NMR (CDCl3) δ: 145.9, 143.9, 141.9, 128.7, 128.4, 124.7,
123.8, 120.2, 84.9, 51.8, 14.1. GC–MS: 160 (82), 159 (11),
145 (100), 130 (14), 129 (23), 128 (45), 127 (20), 117 (40),
116 (12), 115 (60), 102 (17), 91 (22), 77 (12), 63 (18), 51
(16). HR-MS calcd. for C11H12O: 160.0888; found:
160.0887.
bp 100–102 °C at 2 mmHg. UV (methanol) λmax (nm) (⑀
((mol L–1)–1 cm–1)): 273 (8850), 320 (1190). 1H NMR
(CDCl3) δ: 7.02 (d, 1 H, J = 8.0 Hz), 6.99 (d, 1H, J =
2.3 Hz), 6.75 (dd, 1H, J1 = 2.3, J2 = 8.0 Hz), 6.37 (broad s,
1H), 6.10 (broad s, 1H), 3.78 (s, 3H), 2.18 (s, 3H), 1.95
(broad s, 3H). 13C NMR (CDCl3) δ: 171.5, 158.1, 143.9,
142.2, 136.5, 128.9, 120.6, 114.0, 111.7, 78.5, 55.6, 21.1,
14.0. GC–MS: 218 (26), 177 (11), 176 (100), 175 (19), 161
(58), 159 (19), 158 (25), 145 (15), 116 (19), 115 (48). HR-
MS calcd. for C12H14O3: 218.0943; found: 218.0935.
1-Methoxy-5-methoxy-2-methylindene
1H NMR (CDCl3) δ: 7.30 (d, 1H, J = 8.0 Hz), 6.72 (d, 1H,
J = 2.3 Hz), 6.64 (dd, 1H, J1 = 8.0 Hz, J2 = 2.3 Hz), 6.39
(broad s, 1H), 4.82 (broad s, 1H), 3.81 (s, 3H), 3.02 (s, 3H),
2.03 (broad s, 3H). 13C NMR (CDCl3) δ: 160.3, 147.3,
145.5, 133.6, 128.4, 124.3, 109.1, 106.9, 84.2, 55.4, 52.6,
14.2. GC–MS: 191 (11), 190 (88), 189 (15), 176 (11), 175
(100), 160 (25), 159 (32), 147 (28), 144 (16), 132 (13), 116
(23), 115 (55), 103 (12), 91 (18), 89 (13), 77 (14), 63 (15),
51 (10).
2-Methyl-1-indenyl pivalate, 2a
1H NMR (CDCl3) δ: 7.05–7.31 (m, 4 H), 6.43 (broad s,
1H), 6.14 (broad s, 1H), 1.97 (s, 3H), 1.26 (s, 9H). 13C NMR
(CDCl3) δ: 179.0, 144.9, 143.8, 142.4, 129.0, 128.7, 125.1,
123.9, 120.3, 78.2, 39.1, 27.3, 14.0. GC–MS: 230 (15), 146
(66), 145 (14), 131 (25), 129 (42), 128 (100), 127 (21), 115
(19), 57 (75). HR-MS calcd. for C15H18O2: 230.1307; found:
230.1314.
Quantitative photolyses
A solution of ~100 mg of the appropriate ester 1 and 2 in
100 mL methanol or cyclohexane was purged with nitrogen
and then irradiated in a Rayonet photochemical reactor using
4 lamps (75 W, 254 nm). The temperature was controlled at
25 °C by circulating water in an immersion tube. Reaction
progress was monitored by GC, and the reaction stopped
when the ester was less than 5% consumed. For the alcohols
3 and the ether 6 (X = H), ~50 mg in 100 mL of the appro-
priate solvent (methanol or cyclohexane) was used.
5-Methoxy-2-methyl-1-indenyl pivalate, 2b
mp 63–64 °C. 1H NMR (CDCl3) δ: 7.21 (d, 1H, J =
8.1 Hz), 6.71 (d, 1H, J = 2.3 Hz), 6.59 (dd, 1H, J1 = 8.1 Hz,
J2 = 2.3 Hz), 6.37 (broad s, 1H), 6.07 (broad s, 1H), 3.78 (s,
3H), 1.96 (broad s, 3H), 1.25 (s, 9H). 13C NMR (CDCl3) δ:
179.0, 160.6, 146.3, 145.5, 134.4, 128.8, 124.6, 109.3,
107.2, 77.8, 55.4, 39.0, 27.2, 14.1. GC–MS: 260 (11), 176
(47), 175 (21), 161 (17), 160 (11), 159 (36), 158 (29), 116
(20), 115 (40), 57 (100). HR-MS calcd. for C16H20O3:
260.1412; found: 260.1403.
Acknowledgements
We thank the Natural Sciences and Engineering Research
Council of Canada (NSERC) for financial support; Sepracor
Canada Ltd., Windsor, Nova Scotia, for the donation of
chemicals; and Dr. M. Lumsden, ARMRC, Department of
Chemistry, Dalhousie University, Halifax, Nova Scotia, for
the NOE NMR experiments. I.S.Y. also thanks NSERC for a
Student Research Award.
6-Methoxy-2-methyl-1-indenyl pivalate, 2c
mp 54–56 °C. 1H NMR (CDCl3) δ: 7.03 (d, 1H, J =
8.1 Hz), 6.92 (d, 1H, J = 2.3 Hz), 6.76 (dd, 1H, J1 = 8.1 Hz,
J2 = 2.3 Hz), 6.37 (broad s, 1H), 6.10 (broad s, 1H), 3.78 (s,
3H), 1.94 (broad s, 3H), 1.26 (s, 9H). 13C NMR (CDCl3) δ:
179.0, 158.0, 144.2, 142.7, 136.5, 128.5, 120.5, 113.0,
111.6, 78.2, 55.6, 39.0, 27.2, 13.9. GC–MS: 260 (11), 176
(51), 175 (21), 161 (20), 159 (34), 158 (34), 116 (21), 115
(39), 57 (100). HR-MS calcd. for C16H20O3: 260.1412;
found: 260.1408.
References
1. J.A. Pincock. Acc. Chem. Res. 30, 43 (1997).
2. S.A. Fleming and J.A. Pincock. Organic molecular
photochemistry. Vol. 3. Marcel Dekker, New York. 1999.
p. 211.
3. C.G. Bochet. J. Chem. Soc. Perkin Trans. 1, 125 (2002).
4. A.P. Pellicioli and J. Wirz. Photochem. Photobiol. Sci. 1, 441
(2002).
5. J. Chateauneuf, J. Lusztyk, and K.U. Ingold. J. Am. Chem.
Soc. 110, 2877 (1988).
6. J.W. Hilborn and J.A. Pincock. J. Am. Chem. Soc. 113, 2683
(1991).
Preparation of ethers, 3
The ethers were prepared from the corresponding indenols
using the procedure of Freidrich and Taggart (48) and were
purified by chromatography on silica gel. The yields of these
reactions were not high (<20%), particularly for the
1
methoxy-substituted cases, and the H NMR of the crude in-
dicated the appropriate peaks for the ethers superimposed on
broad, featureless bands (oligomers?). Only one ether 9 (X =
5-OCH3) was obtained when starting with either 5- or 6-
methoxy-2-methyl-1-indenol, although GC analysis indi-
7. T.M. Bockman, S.M. Hubig, and J. K. Kochi. J. Org. Chem.
62, 2210 (1997).
1
cated that the isomer was formed (<5%). H NOE difference
spectra demonstrated that this ether was 1-methoxy-5-
methoxy-2-methylindene. This ether 3 (X = 5-CH3O) was
8. D.P. DeCosta and J.A. Pincock. J. Am. Chem. Soc. 115, 2180
(1993).
© 2003 NRC Canada

Pincock and Young
1095
9. W. Hilborn, E. MacKnight, J.A. Pincock, and P.J. Wedge. J.
Am. Chem. Soc. 116, 3337 (1994).
33. W.R. Dolbier, K.E. Anapolle, L. McCullaugh, K. Matsui, J.M.
Reimann, and D. Rolison. J. Org. Chem. 44, 2845 (1979).
34. J.M. Kim and J.A. Pincock. Can. J. Chem. 73, 885 (1995).
35. R.S. Givens, B. Matuszewski, N. Levi, and D. Leung. J. Am.
Chem. Soc. 99, 1896 (1977).
36. D.A. Jaeger and G.H. Angelos. Tetrahedron Lett. 22, 803
(1981).
37. Y. Shi, J.E.T. Corrie, and P. Wan. J. Org Chem. 62, 8278
(1997).
38. J.J. McCullough. Acc. Chem. Res. 13, 270 (1980).
39. M.T. Gibian and R.G. Corley. Chem. Rev. 73, 441 (1973).
40. P.D. Bartlett and R.R. Hiatt. J. Am. Chem. Soc. 80, 1398
(1958).
41. S.S. Kim, I.S. Baek, A. Tuchkin, and K.M. Go. J. Org Chem.
66, 4006 (2001).
42. P.D. Bartlett and C. Ruchardt. J. Am. Chem. Soc. 82, 1756
(1960).
43. (a) D.P. DeCosta, N. Howell, A.L. Pincock, J.A. Pincock, and
S. Rifai. J. Org. Chem. 65, 4698 (2000); (b) F.L. Cozens, A.L.
Pincock, J.A. Pincock, and R. Smith. J. Org. Chem. 63, 434
(1998).
10. H.E. Zimmerman. J. Am. Chem. Soc. 117, 8988 (1995).
11. H.E. Zimmerman. J. Phys. Chem. A, 102, 5616 (1998).
12. P. Wan and E. Krogh. J. Am. Chem. Soc. 111, 4887 (1989).
13. A. Blazek, M. Pungente, E. Krough, and P. Wan. J.
Photochem. Photobiol. A Chem. 64, 315 (1992).
14. G.G. Gurzadyan and S. Steenken. Chem. Eur. J. 7, 1808 (2001).
15. A.D Allen and T.T. Tidwell. Chem. Rev. 101, 1333 (2001).
16. H. Jiao, P. von R. Schleyer, Y. Mo, M.A. McAllister, and T.T.
Tidwell. J. Am. Chem. Soc. 119, 7075 (1997).
17. J.A. Pincock and P.J. Wedge. J. Org. Chem. 60, 4067 (1995).
18. J.-L. Luche. J. Am. Chem. Soc. 100, 2226 (1978).
19. M.B. Floyd and G.R. Allen. J. Org. Chem. 35, 2647 (1970).
20. S. Kapur, N.C. Mathur, K. McManus, A.L. Pincock, and J.A.
Pincock. Can. J. Chem. 66, 2888 (1988).
21. J. Sam and P. Plampin. J. Am. Chem. Soc. 82, 5205 (1960).
22. T. Waugh and H. Morrison. J. Am. Chem. Soc. 121, 3083
(1999).
23. F.D. Lewis, D. M. Bassani, R.A. Caldwell, and D.J. Unett. J.
Am. Chem. Soc. 116, 10 477 (1994).
24. H. Morrison, D. Giacherio, and F. J. Palensky. J. Org. Chem.
47, 1051 (1982).
44. E. Gaillard, M.A. Fox, and P. Wan. J. Am. Chem. Soc. 111,
2180 (1989).
25. N.E. Heimer, M. Hojjatie, and C. A. Panetta. J. Org. Chem.
47, 2593 (1982).
45. M.A. O’Neill, F.L. Cozens, and N.P. Schepp. Tetrahedron, 56,
6969 (2000).
26. P. Nicolet, J.-Y. Sanchez, A. Benaboura, and M.J.M. Abadie.
Synthesis, 202 (1987).
46. (a) H.E. Zimmerman and V.R. Sandel. J. Am. Chem. Soc. 85,
915 (1963); (b) H.E. Zimmerman and S. Somasekhara. J. Am.
Chem. Soc. 85, 922 (1963).
47. (a) P. Wan and B. Hall. J. Photochem and Photobiol. A, 56, 35
(1991); (b) P. Wan, B. Chak, and E. Krough. J. Photochem and
Photobiol. A, 46, 49 (1989); (c) P. Wan and B. Chak. J. Chem.
Soc. Perkin Trans. 2, 1751 (1986); (d) P.J. Wan. J.Org. Chem.
50, 2583 (1985).
27. E.C. Friedrich and D.B. Taggart. J. Org. Chem. 40, 720 (1975).
28. F.G. Bordwell. Acc. Chem. Res. 21, 456 (1988).
29. H. Morrison and D. Giacherio. J. Org. Chem. 47, 1058 (1982).
30. H. Morrison, D. Giacherio, and G. Pandey. Tetrahedron Lett.
23, 3427 (1982).
31. D.F. McMillen and D.M. Golden. Ann. Rev. Phys. Chem. 33,
493 (1982).
48. E.C. Freidrich and D.B. Taggart. J. Org. Chem. 40, 720 (1975).
32. E.C. Friedrich and D.B. Taggart. J. Org Chem. 43, 805 (1978).
© 2003 NRC Canada