Synthesis, structure, and nucleophile-induced rearrangements of spiroketones

Article abstract of DOI:10.1021/jo990867j

Three tetraketones based on the 2,2'-spirobiindan-1,1',3,3'-tetraone skeleton were prepared and investigated. All three compounds show spiroconjugation between their perpendicular π-networks. The interaction results in lowering of the energy of the LUMO of the systems by ca. 0.2-0.3 eV as compared to non-spiroconjugated models. The spiroketones are susceptible to nucleophile-induced retro-Claisen condensations that lead to molecular rearrangements destroying spiro connectivity.

Full text of DOI:10.1021/jo990867j

J . Org. Chem. 1999, 64, 8201-8209
8201
Syn th esis, Str u ctu r e, a n d Nu cleop h ile-In d u ced Rea r r a n gem en ts of
Sp ir ok eton es
Przemyslaw Maslak,* Sridhar Varadarajan, and J effrey D. Burkey
Department of Chemistry, The Pennsylvania State University, University Park, Pennsylvania 16802
Received May 27, 1999
Three tetraketones based on the 2,2′-spirobiindan-1,1′,3,3′-tetraone skeleton were prepared and
investigated. All three compounds show spiroconjugation between their perpendicular π-networks.
The interaction results in lowering of the energy of the LUMO of the systems by ca. 0.2-0.3 eV as
compared to non-spiroconjugated models. The spiroketones are susceptible to nucleophile-induced
retro-Claisen condensations that lead to molecular rearrangements destroying spiro connectivity.
In connection with our investigation of new three-
dimensional acceptors1 we have prepared several spiroke-
tones. The design of these acceptors was based on the
phenomenon of spiroconjugation,2,3 i.e., the ability of two
perpendicular π-networks to interact. In the case of the
tetraketones 1-3 we were interested in their acceptor
properties.1 The π orbitals serving as the LUMOs of these
compounds may be considered as the bonding combina-
tion of the LUMOs of the corresponding indandiones. The
antibonding combination4 gives the next higher-energy
orbital (LUMO+1), and the separation between these two
spiro orbitals spanning the entire molecule depends on
the strength of the spiroconjugation. That interaction
depends mainly on the overlap of the corresponding “half-
molecule” orbitals, which in turn is governed to a large
degree by the size of the coefficient on the carbonyl-
carbons. The larger these coefficients, the larger the
interaction between the “halves”. As shown by semiem-
pirical calculations5 in all three tetraketones, the LUMO
spans the entire molecule. In 1 and 2 the HOMO is also
spiroconjugated, but the interaction is rather small due
to the fact that the coefficients on the carbonyl-carbons
are very small.4 In 3, because of the change in symmetry
of the subunit orbitals, the HOMO is no longer spirocon-
jugated. We report here on the syntheses of these
ketones, exploration of their structure, and investigation
of their nucleophile-induced reactivity.
Scheme 1. In each case the synthesis started with the
dialkylation of diethyl malonate with the appropriate
benzyl chloride to produce the corresponding diethyl
dibenzylmalonate in high yields (90% or better). The
diesters were hydrolyzed to the corresponding diacids (4-
6) in very good yields (70% or better). Since all the diacids
decarboxylated slowly over time, they were stored in the
form of the disodium salts. The free acids were obtained,
just before use, by acidification of a cold aqueous solution
of the salts with equivalent amounts of concentrated
hydrochloric acid.
The spiro network was formed by an intramolecular
Friedel-Crafts acylation of the diacids or the diacid
chlorides (7-9). We found that phosphorus pentoxide was
the agent of choice to affect the spiro-closure of all the
diacids: 4 gave 10 in 42% yield, and 6 lead to 12 in 33%
yield, but the reaction of 5 was irreproducible, giving 11
in only 10% yield in the best runs. The cyclization of acid
chlorides worked almost as well for the preparation of
10 (in 30% overall yield from the diacid), but it was a
preferred method for the preparation of 11 (27% overall
yield from the diacid). All the reactions with phosphorus
pentoxide worked best on gram/subgram quantities of the
diacids. Multigram-scale reactions seemed to lead to lar-
ger proportions of undesired decarboxylated side prod-
ucts. On the other hand, closures of the acid chlorides
required careful purification of the chlorides (especially
in the case of 8) and precise control of the reaction condi-
tions (catalyst and temperature). The alternative meth-
ods of cyclization of the esters or acids (using for example
polyphosphoric acid or sulfuric acid) were unsuccessful.
Syn th esis of the spiro network for all the acceptors
was accomplished using the chemistry6 illustrated in
(1) (a) Maslak, P.; Augustine, M. P.; Burkey, J . D. J . Am. Chem.
Soc. 1990, 112, 5360. (b) Maslak, P. Adv. Mater. 1994, 6, 405. (c)
Maslak, P.; Chopra, A.; Moylan, C. R.; Wortmann, R.; Lebus, S.;
Rheingold, A. L.; Yap, G. P. A. J . Am. Chem. Soc. 1996, 118, 1471. (d)
Maslak, P.; Chopra, J . Am. Chem. Soc. 1993, 115, 9331.
(2) (a) Simmons, H. E.; Fukunaga, T. J . Am. Chem. Soc. 1967, 89,
5208. (b) Hoffmann, R.; Imamura, A.; Zeiss, G. D. J . Am. Chem. Soc.
1967, 89, 5215.
Once the spiro network was obtained, the benzylic
positions were further functionalized by a combination
of bromination, hydrolysis, and oxidation reactions.
(3) For a review see: Durr, H.; Gleiter, R. Angew. Chem., Int. Ed.
Engl. 1978, 17, 559.
(4) All orbitals of the “halves” that are antisymmetric vs both
molecular planes can interact giving the lower-energy bonding and the
higher-energy antibonding combinations that span the entire molecule.
The frontier orbitals, and specifically here for the electron acceptors
the LUMOs, are of main interest. For the purpose of this discussion,
the HOMO-LUMO designation refers only to the π system. The
nonbonding electrons are treated separately.
(5) The semiempirical MNDO-PM3 calculations (Stewart, J . J . P.
J . Comp-Aid. Mol. Des. 1990, 4, 1) were carried out using Spartan 5
(Wavefunction Inc.).
(6) The synthetic strategy followed that used to prepare dione 10
and a related spirosystem: (a) Langer, E.; Lehner, H. Tetrahedron
1973, 29, 375. (b) Dynesen, E. Acta Chem. Scand. 1972, 26, 850. (c)
Semmelhack, J .; Foos, S.; Katz, S. J Am. Chem. Soc. 1973, 95, 7325.
10.1021/jo990867j CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/08/1999

8202 J . Org. Chem., Vol. 64, No. 22, 1999
Maslak et al.
Sch em e 1
Direct oxidation of 10 with chromium(VI) oxide produced
2,2′-spirobiindan-1,1′,3-trione (13) in 52% yield. The
oxidation reaction, however, could not be pushed to
completion without destroying the desired product, and
it proved difficult to separate 13 from the unreacted 10.
Bromination of 13, followed by hydrolysis of the dibro-
motrione (14), afforded the target tetraone 1 in a 26%
overall yield, starting from 10. An alternative procedure,
involving extensive bromination of 10, followed by hy-
drolysis of the tetrabromo compound (15), produced the
desired tetraone (1) in a higher yield (36%) and in fewer
steps. A similar exhaustive bromination procedure when
applied to 12 gave the corresponding tetrabromide (16)
that was hydrolyzed to 3 in 80% overall yield. NBS
bromination of 11 produced only the corresponding
dibromo compound 17, probably due to the steric hin-
drance caused by the methoxy groups. Surprisingly, only
one diastereoisomer was obtained. Hydrolysis of 17 to the
dialcohol 18 (obtained as a mixture of two diastereoiso-
mers) followed by oxidation gave the target tetrameth-
oxytetraone (2) in 61% yield (from 11).
Th e str u ctu r e of the spiroketones 1 and 2 was
confirmed by X-ray analysis.7 The molecules possess
2-fold symmetry axes and planes as expected. There are
no unusual bond lengths or bond angles present, and the
observed and calculated (MNDO-PM3)5 geometrical pa-
rameters are in excellent agreement. The chemical shift
equivalence of the signals in the NMR spectra of the
compounds also confirms the same high symmetry: the
two halves of each molecule are contained in two per-
pendicular planes, and no significant distortions are seen.
This is in contrast to what is observed for spiro donor-
acceptor systems, where the aromatic rings of the donor
components are bent away from the plane perpendicular
to the plane containing the acceptor half.1c
F igu r e 1. Schematic representation of the orbitals available
for electronic transition in model compounds (a, b) and the
spiroconjugated tetraones (c). In 3 the HOMO is not spirocon-
jugated. The HOMO-LUMO designation refers only to the π
system.
Also inspection of the infrared carbonyl frequencies did
not reveal any significant spiroconjugation effects. The
frequencies for the carbonyl absorptions in the spiro
molecules were slightly less than those in the corre-
sponding models (19-22), but these differences were less
than 3 cm-1
.
The spiroconjugation effects were most visible in the
comparison of the UV-visible spectra of the tetraones
with those of the corresponding models. The interpreta-
tion of the various features in the spectra can be better
understood by first considering the interactions of the
frontier orbitals in the systems (Figure 1). It should be
noted that the figure represents only the orbital picture,
while we are dealing with state-to-state transitions in
the actual systems. As has been mentioned earlier, both
the HOMO and the LUMO of the π subsystems of the
spiro compounds can interact (except for 3, whose HOMO
has incorrect symmetry for spiroconjugation). However,
since the carbonyl-carbon orbital coefficients on the
HOMO are small, these interactions of the subsystem
HOMOs are not expected to be significant. Therefore, the
HOMO in the spiroconjugated molecules (Figure 1) will
not be much different from the HOMO in the model.
However, the stronger LUMO-LUMO interactions (due
to larger orbital coefficients on the carbonyl-carbons)
would lead to a lowering of the LUMO in the spirocon-
jugated molecule, as compared to the model. Additionally,
the probability of the HOMO-LUMO transitions is
doubled in the case of the spiro compounds (and the
dimer, 22), as compared to the indandione models since
there are two degenerate (or nearly degenerate) HOMOs
in the spiro-molecules and the dimer, whereas there is
only one such orbital in the models.
A characteristic trend was observed in the 13C chemical
shifts of the carbonyl carbons, wherein all the spirodiones
(10-12) and the model compounds 19-22 (see below)
exhibited shifts ranging from 200 to 205 ppm, while all
the spirotetraones had shifts approximately 10 ppm lower
(190-191 ppm). The spirotrione 13 showed shifts of
intermediate value (ca. 196 ppm). It is unlikely that this
trend is related to spiroconjugation; it is probably due to
the cumulative inductive effects of the carbonyl groups.
The UV-visible spectra of the systems investigated are
shown in Figures 2-4. In all cases, there is a striking
similarity between the absorption spectra of the spiroke-
tones and the corresponding models. All the absorption
bands seen in the spiro compounds are also present in
the model compounds.
(7) The details of the crystal structure and the crystal packing of 1
and 2 will be presented elsewhere: Maslak, P.; Varadarajan S.;
Rheingold, A. L.; Yap, G. P. Manuscript in preparation.
Compounds 19 and 22 show two π-π* transitions at
ca. 225 and 245 nm and an n-π* transition at 290-310

Rearrangements of Spiroketones
J . Org. Chem., Vol. 64, No. 22, 1999 8203
F igu r e 4. UV-visible spectra of the spiroconjugated acceptor
3 (solid line) and model dione 21 (dotted line) in acetonitrile
at room temperature.
F igu r e 2. UV-visible spectra of the spiroconjugated acceptor
1 (solid line) and model diones 22 (broken line) and 19 (dotted
line) in acetonitrile at room temperature.
and 370 nm. This could be a second n-π* transition,
possible in this case due to the presence of a split LUMO.
In the case of the ketones with methoxy substituents
(2, 3, 20, and 21), there is an additional absorption band
in the near-UV-visible region of the spectrum, which is
absent in the unsubstituted ketones (1, 19, and 22).
These bands are present in both the spiro compounds and
the corresponding models, and they could be attributed
to CT transitions within the subchromophores, arising
due to the transfer of charge from the lone pairs on the
methoxy oxygens to the carbonyls.
This CT band, in the case of 2 and 20, is shifted to a
longer wavelength as compared to 3 and 21. Such dif-
ferences are not unusual for ortho versus para substitu-
tion in CT systems. For example, the absorption maxima
for ortho- and para-nitroaniline occur at 398 and 354 nm,
respectively, in dioxane. This difference can be attributed
to the difference in the electrostatic interaction between
the two end groups involved in the CT interaction. Since
the end groups are closer in 2 and 20, than in 3 and 21,
the electrostatic interactions are stronger in 2 and 20,
and therefore, the excited state in 2 and 20 is lower in
energy. Hence the absorption of the CT transition in 2
and 20 is at a lower energy (longer wavelength).
Compounds 2 and 20 also show two π-π* transitions
and one n-π* transition, besides the CT band discussed
above, all of which are shifted to longer wavelength in 2
as compared to 20. As in the case of 1, this red shift is
consistent with the lowering of the LUMO in the spiro-
conjugated system. The CT band in 2 shows two absorp-
tion maxima at ca. 380 and 400 nm. A red shift is also
seen in the case of 3 versus 21; however, both these
compounds seem to have only one absorption band (π-
π*) besides the CT band. It is possible that other bands
are hidden under the intense π-π* band.
F igu r e 3. UV-visible spectra of the spiroconjugated acceptor
2 (solid line) and model dione 20 (dotted line) in acetonitrile
at room temperature.
nm. The same transitions are seen in the case of the spiro
compound 1, but they are shifted to longer wavelengths,
due to the lowering of the π* level (Figure 1b). Addition-
ally, 1 shows another weak absorption band between 320

8204 J . Org. Chem., Vol. 64, No. 22, 1999
Maslak et al.
Sch em e 2
The integrated intensities of the absorption bands in
the two-chromophore systems (spiro molecules 1 and 2
and the dimer 22) are nearly double the value for the
corresponding single-chromophore systems (models 19
and 20). This is to be expected, since the probability of
transitions should be nearly doubled in the bichro-
mophoric systems (Figure 1).
There is an apparent deviation from this behavior in
the case of 3 versus 21. The intensities of the absorption
of the bichromophore 3 are not double the corresponding
values for the single chromophore 21 (though they are
all higher). The reason for this deviation is not yet clear.
One possibility is that some distortion of the chromophore
affects the local transition dipole. An experimental error
due to the low solubility of the spiro compound in the
solvent used (acetonitrile) is unlikely, since precautions
were taken to ensure dissolution. The extinction coef-
ficient for the main transition in 3 is slightly lower than
the corresponding values for 1 and 2, while the extinction
coefficient for the main transition in the model 21 is
higher than the corresponding values for 19 and 20. It
is possible, therefore, that the observed apparent devia-
tion in the relative absorption of 3 and 21 is simply the
combined effect of an enhanced extinction coefficient of
21 and a diminished extinction coefficient of 3.
The extent of red shift in the absorption bands of the
spiro molecules versus the corresponding absorptions in
the appropriate models (converted to energy units) gives
us a measure of spiroconjugation in these systems. A
comparison of the values (ca. 0.28 eV) for the main π-π*
transitions in all three systems indicates that the spiro-
conjugation effects are very similar and that the intro-
duction of the methoxy groups in 2 and 3 does not seem
to significantly affect the overlap between the carbonyl-
carbon orbitals. This is to be expected, on the basis of
similar values of the LUMO orbital coefficients (obtained
from semiempirical calculations) for all three compounds.
The differences obtained for the n-π* (for 2) and the
CT transitions (for 2 and 3) of 0.20-0.24 eV are slightly
lower than the differences obtained for the π-π* transi-
tions, possibly because the values for the π-π* transi-
tions may reflect the combined effect of the lowering of
the LUMO and the slight raising of the HOMO in the
spiro molecules (except for 3) due to spiroconjugation
effects. This effect is not possible in the case of the n-π*
and the CT transitions since the n orbitals are not
affected by spiroconjugation. Thus, the UV-visible data
on the acceptors clearly indicate the presence of spiro-
conjugation effects in all the neutral spiro molecules.
These effects seem to be quantitatively similar in all the
systems investigated and amount to ca. 0.2-0.3 eV in
terms of lowering of the LUMO energy as compared to
non-spiroconjugated models.
Sch em e 3
ide. When treated with sodium methoxide in methanol,
13 yielded the yellow enolate of â-diketone ester (23-).
The reaction was complete upon mixing. The UV/visible
spectrum of the reaction mixture showed only the enolate
(see the Experimental Section for details). Protonation
of the mixture with HClaq yielded enol 23 as the exclusive
product (Scheme 2). Analogously, when 13 was treated
with sodium ethoxide, 24 was isolated after acidification.
Both 23 and 24 exist predominantly in the enol form,
but the keto form was detectable by 13C NMR spectros-
copy. Acid hydrolysis of either 23 or 24 in aqueous
hydrochloric acid produced lactone 25. All the transfor-
mations were nearly quantitative.
When 13 was treated with sodium hydroxide under
phase-transfer catalysis (PTC) conditions, hemiketal
lactone 26 resulted (after acidification). Lactone 25 when
treated with sodium hydroxide under PTC conditions was
also converted to hemiketal lactone 26. The stereochem-
istry of the double bonds in the enol forms of 23 and 24
has not been experimentally determined, but the geom-
etry of 25 was established via decoupling and NOE
experiments (see the Experimental Section). The hydro-
gen assignments are based on chemical shifts and de-
coupling experiments. The geometry shown is based on
the absence of NOE effects between the benzylic hydro-
gens and H1. Only one isomer was detected in this case.
Again, all the transformations were nearly quantitative.
When 1 was treated with sodium ethoxide in ethanol,
compound 27 was obtained (Scheme 3). Hydrolysis of 27
in 6 N hydrochloric acid resulted in 1,3-indandione. These
results convey the reactive nature of these compounds.
Nu cleop h ile-in d u ced r ea r r a n gem en ts were ob-
served during the course of our studies of the spiroke-
tones on several occasions. The electronic structure of
these compounds makes them very susceptible to nu-
cleophilic attack. By design, they are electron deficient,
and additionally, the carbon-carbon bonds of the spiro
carbon are almost perfectly aligned to overlap with the
π orbitals of the carbonyl groups. We have briefly probed
the nature of these nucleophile-induced reactions using
trione 13 and tetraone 1 as convenient models.
2,2′-Spirobiindan-1,3,1′-trione (13) was reacted on a
preparative scale with several oxygen bases, including
sodium methoxide, sodium ethoxide, and sodium hydrox-

Rearrangements of Spiroketones
J . Org. Chem., Vol. 64, No. 22, 1999 8205
CHCl3, or in KBr. UV/visible spectra were recorded using
quartz cuvettes. Mass spectral analyses (MS and HRMS) were
recorded using a double-focusing spectrometer. Only structur-
ally significant and the strongest IR frequency and MS
fragment-ion peaks are reported. Melting points were deter-
mined in open capillary tubes and are uncorrected. Preparative
flash chromatography10 was carried out using Machery Nagel
silica gel 60 or Merck silica gel 60, 230-400 mesh. Analytical
thin-layer chromatography was performed using EM Reagents
precoated silica gel (0.25 mm) F254 plates. The mixed solvents
used for chromatography are reported as volume/volume
ratios.
Dieth yl Bis(2,5-dim eth oxyben zyl)m alon ate. Diethyl ma-
lonate (49.5 mL, 0.326 mol) was added to a mechanically
stirred solution of sodium ethoxide (0.74 mol, prepared in situ
by the slow addition of 17.1 g sodium to absolute alcohol) in
absolute alcohol (300 mL), under argon. The mixture was
stirred for 30 min and then cooled to 0 °C, when a white solid
precipitated. 25 (122 g, 0.656 mol) was added as a solid in one
installment, and the stirring was continued for 1 h before the
temperature was allowed to rise to 25 °C. The mixture was
stirred overnight at room temperature. Dilute hydrochloric
acid was added to the milky white reaction mixture until the
solution was acidic. The solution was extracted with dichlo-
romethane, and the combined extracts were subjected to
standard workup. The colorless liquid obtained upon removal
of solvent was placed under vacuum (0.1 Torr) for 2 h (145 g,
97%). No further purification was necessary. 1H NMR (200
MHz): 6.78-6.68 (m, 6 H), 4.03 (q, J ) 7 Hz, 4 H), 3.68 (s, 12
H), 3.27 (s, 4 H), 1.09 (t, J ) 7 Hz, 6 H). 13C NMR (75 MHz):
171.2, 153.0, 152,5, 126.6, 117.7, 112.3, 110.6, 60.8, 58.7, 55.6,
33.5. FTIR (film): 2983 (w), 2941 (w), 2834 (w), 1729 (s), 1610
(w), 1590 (w), 1501 (s), 1465 (m), 1224 (s). EI-MS (m/z, relative
intensity): 460 (M+, 0.3), 310 (11), 238 (15), 151 (17), 84 (87),
49 (100).
Sch em e 4
Although for convenience our studies concentrated on the
trione (13), the reactivity of the tetraketone (1) and the
other tetraones seems to follow similar patterns.
The observed products can be readily rationalized in
terms of standard carbonyl chemistry (Scheme 4). Thus,
a nucleophile adds rapidly to the electron-deficient in-
dandione moiety. The tetrahedral intermediate formed
(28) undergoes facile carbon-carbon bond cleavage. Such
carbon-carbon bond cleavages in retro-aldol and Claisen
condensations are well-known.8 As is usually the case,
the facility of this type of cleavage is dependent upon the
stability of the resulting carbanion. In the case of 1 or
13, the anion 29 is a highly delocalized enolate. In
addition, the scissile bond in 28 overlaps very well with
the π-system of the indanone moiety. Under such cir-
cumstances, the carbon-carbon bond cleavage is favored
by both thermodynamic and stereoelectronic factors.
The spectral data indicate that when R ) Me or Et,
the product from quenching the reaction with aqueous
acid is present predominantly as the enol tautomer. The
usual hydrogen bonding within the enols of â-dicarbonyl
compounds9 should “lock” the double bond in 23 and 24
in the Z configuration, as shown in Scheme 2; however,
the stereochemistry of the double bonds in these enols
has not been established. The presence of a single peak
for the methylene group in the 1H NMR spectra of 23
and 24 and a single resonance for the corresponding
R-carbon in the 13C NMR spectrum of 23 suggests that
only one isomer is present or that the isomers exchange
rapidly on the NMR time scale.
In the case when R ) H, the enolate originally
produced undergoes a secondary reaction. A proton
transfer within enolate 29 (R ) H, Scheme 4) yields the
carboxylate anion, which adds intramolecularly to the
carbonyl group, giving (after protonation) hemiketal-
lactone 26. Lactone 25 and hemiketal-lactone 26 were
easily interconvertible. Thus, addition of hydroxide to 25
followed by acidification gave 26, while treatment of the
latter with concentrated acid at elevated temperature
regenerated the unsaturated lactone (25).
As expected, 1 reacted rapidly with ethoxide in a
manner analogous to 13, yielding after acidification enol
27. In this system the resulting enolate is even more
stable than that observed for 13, additionally increasing
the driving force for the ring opening.
Diet h yl b is(3,4-d im et h oxyb en zyl)m a lon a t e was ob-
tained as described above using diethyl malonate and 3,4-
dimethoxybenzyl chloride. The colorless liquid obtained on
removal of solvent was dissolved in a minimum quantity of
hot ethanol. The white crystals that fell out upon cooling were
1
filtered off and dried (118 g, 96%), mp 115-116 °C. H NMR
(360 MHz): 6.78 (d, J ) 8.2 Hz, 2 H), 6.72 (dd, J ) 8.2 Hz, J
) 1.9 Hz, 2 H), 6.69 (d, J ) 1.9 Hz, 2 H) 4.13 (q, J ) 7.1 Hz,
4 H), 3.86 (s, 6 H), 3.83 (s, 6 H), 3.18 (s, 4 H), 1.19 (t, J ) 7.1
Hz, 6 H). 13C NMR (90 MHz): 171.0, 148.4, 147.8, 128.7, 122.1,
113.3, 110.8, 61.0, 60.0, 55.7, 55.6, 38.7, 13.8. FTIR (KBr):
2976 (w), 2942 (w), 2839 (w), 1736 (s), 1608 (w), 1592 (w), 1518
(s), 1464 (m), 1194 (s). EI-MS (m/z, relative intensity): 460
(M+, 22), 263 (24), 152 (15), 151 (C9H11O2, 100).
Dieth yl d iben zylm a lon a te11 was prepared as described
above from diethylmalonate and benzyl chloride The colorless
liquid obtained upon solvent removal was 110.9 g (98%) and
was used in the next step without further purification. 1H
NMR (200 MHz): 7.30-7.22 (m, 10 H), 4.10 (q, J ) 7 Hz, 4
H), 3.24 (s, 4 H), 1.16 (t, J ) 7 Hz, 6 H). 13C NMR (50 MHz):
170.6, 136.1, 129.9, 128.6, 128.2, 127.9, 126.6, 60.9, 60.0, 39.0,
13.6. IR (film): 3080-2860 (m) 1725 (s), 1600 (m), 1580 (w),
1485 (m), 1440 (s), 1360 (s), 1280-1140 (s), 1060 (s), 1020 (s).
EI-MS (m/z, relative intensity): 340 (M+ 1), 295 (2), 249 (53),
204 (14), 203 (100), 193 (11), 192 (18), 131 (10%), 115 (20).
Bis(2,5-d im eth oxyben zyl)m a lon ic Acid (5). A viscous
mixture of the corresponding diethyl ester (131.5 g, 0.29 mol)
and pulverized sodium hydroxide (47 g, 1.18 mol) was me-
chanically stirred and heated to 90 °C in a flask equipped with
an air condenser. Absolute alcohol (5 mL) was added at this
temperature, and immediately a white solid fell out of solution.
The solid paste was thinned by the periodic addition of water
(2-3 mL) when necessary, and the mixture stirred for 6 h at
90 °C. After cooling to room temperature, the mixture was
diluted with water (500 mL) and washed with dichloromethane
(2 × 200 mL) to remove organic impurities. The aqueous layer
was acidified with concentrated HCl keeping the temperature
We conclude that both the spirotrione and spirotetra-
one are very susceptible to attack by charged nucleo-
philes. Analogous base- or nucleophile-induced reactivity
is most likely responsible for some of the difficulties
observed in investigation of spiroconjugation effects in
all spiroconjugated ketones.
Exp er im en ta l Section
Gen er a l P r oced u r es. Infrared (IR) spectra were recorded
on samples in the form of films on NaCl plates, solutions in
(8) Artamkina, G.; Beletskaya, I. Russ. Chem. Rev. 1987, 56, 983.
(9) Keeffe, J . R.; Kresge, A. J .; Schepp, N. P. J . Am. Chem. Soc. 1990,
112, 4862.
(10) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923.

8206 J . Org. Chem., Vol. 64, No. 22, 1999
Maslak et al.
below 25 °C by the addition of ice. The yellow solid that fell
out of solution was extracted with dichloromethane. The
combined organic extracts were subjected to the standard
workup. The solution was concentrated by evaporation of
solvent when a white solid started falling out. This mixture
was cooled in a refrigerator for 1 h, and the resulting white
precipitate was filtered out at the aspirator and dried, 81.0 g
(70%). A second crop of white solid was obtained on further
concentration of the filtrate, followed by cooling and filtration,
MS (m/z, relative intensity): 320 (M+, 3), 249 (5), 194 (20),
131 (12), 117 (13), 92 (14), 91 (100).
2,2′-Sp ir obiin d a n -4,4′,7,7′-tetr a m eth oxy-1,1′-d ion e (11)-
. A mixture of 8 (0.5 g, 1.14 mmol) and a catalytic amount of
ferric chloride (0.01 g, 0.06 mmol) was stirred under vacuum
(1 Torr) for 5 min in a preheated oil bath at 160 °C. The solid
melted to form a dark brown liquid with the evolution of a
gas. The mixture was allowed to cool to room temperature,
and the resultant solid was stirred with dilute HCl (25 mL)
and dichloromethane (25 mL) for 30 min. The organic layer
was then separated, and the aqueous layer extracted with
dichloromethane. The combined organic layers were subjected
to the standard workup. The brown solid left behind on
removal of solvent contained the desired product, which was
isolated by flash chromatography, eluting with 10% ethyl
acetate in dichloromethane. Recrystallization from ethyl ac-
etate yielded 0.2 g (48%) of white crystals, mp 265-267 °C; Rf
1
14.0 g (12%), mp 159-160 °C. H NMR (360 MHz): 6.76 (s, 6
H), 3.74 (s, 6 H), 3.70 (s, 6 H), 3.47 (s, 4 H); (300 MHz, (CD3)2-
CO): 6.82 (d, J ) 8.9 Hz, 2 H), 6.78 (s, 2 H), 6.72 (dd, J ) 8.7
Hz, J ) 3.0 Hz, 2 H), 3.66 (s, 6 H), 3.55 (s, 6 H), 3.33 (s, 4 H).
13C NMR (50 MHz, (CD3)2CO): 174.0, 154.1, 153.2, 127.3,
117.7, 112.7, 111.8, 58.2, 55.7, 55.6, 35.3. FTIR (KBr): 3001
(m), 2946 (m), 2835 (m), 1701 (s), 1616 (w), 1592 (m), 1503 (s),
1465 (s), 1281 (s), 1225 (s), 1179 (s). EI-MS (m/z, relative
intensity): 404 (M+, 1), 360 (M+ - CO2, 75), 191 (18), 151 (100),
137 (18), 121 (55).
1
) 0.35 (10% ethyl acetate in dichloromethane). H NMR (200
MHz): 7.02 (d, J ) 8.7 Hz, 2 H), 6.74 (d, J ) 8.7 Hz, 2 H),
3.86 (s, 6 H), 3.85 (s, 6 H), 3.57 (d, J ) 17.5 Hz, 2 H), 2.96 (d,
J ) 17.5 Hz, 2 H); 13C NMR (50 MHz): 200.1, 152.3, 150.1,
144.2, 124.7, 117.3, 110.0, 65.6, 56.0, 55.9, 34.6. FTIR (KBr):
1714 (m), 1688 (s), 1594 (m), 1499 (s), 1275 (s). EI-MS (m/z,
relative intensity): 369 (M+ + 1, 26), 368 (M+, 100), 353 (11),
338 (15), 323 (15), 309 (41), 192 (19), 191 (24), 189 (15), 151
(13). Anal. Calcd for C21H20O6: C, 68.47; H, 5.47. Found: C,
68.55; H, 5.61.
Bis(3,4-d im eth oxyben zyl)m a lon ic a cid (6) was preapred
as described above from the corresponding diethyl ester. The
pale yellow solid that fell out of solution was filtered off at
the aspirator, washed with cold dichloromethane (2 × 50 mL),
and dried, 80 g (79%), mp 168-170 °C. 1H NMR (300 MHz,
(CD3)2CO): 6.83 (d, J ) 8.3 Hz, 2 H), 6.82 (s, 2 H), 6.78 (dd, J
) 8.1 Hz, J ) 2.1 Hz, 2 H), 3.75 (s, 6 H), 3.74 (s, 6 H), 3.21 (s,
4 H). 13C NMR (75 MHz, (CD3)2CO): 174.1, 149.7, 149.2, 129.4,
122.6, 114.4, 112.3, 60.9, 55.7, 55.6, 40.8. FTIR (KBr): 3319
(m), 3012 (w), 2968 (w), 2940 (w), 2840 (w), 1753 (m), 1708
(s), 1691 (m), 1517 (s) 1259 (s) 1141 (s). EI-MS (m/z, relative
intensity): 404 (M+, 1), 361 (11), 360 (M+ - CO2, 47), 152 (36),
151 (C9H11O2, 100).
Diben zylm a lon ic a cid 11 (4) was prepared as described
above from the corresponding diethyl ester. The white solid
that was obtained was 53.4 g (85%) and did not require further
purification; mp 174-175 °C (lit.11 175 °C). 1H NMR (200
MHz): 7.25-7.23 (m, 10 H), 3.46 (s, 4 H). IR (KBr): 3060-
2260 (m), 1720 (s), 1590 (s), 1470 (s), 1430 (s), 1220 (s), 1190
(s). EI-MS (m/z, relative intensity): 284 (M+, 0.1), 240 (10),
193 (14), 175 (21), 149 (46), 131 (39), 115 (19), 103 (18), 92
(41), 91 (100).
2,2′-Sp ir obiin d a n -5,5′,6,6′-tetr a m eth oxy-1,1′-d ion e (12)-
. A suspension of 6 (0.50 g, 1.24 mmol) in dry toluene (50 mL)
was cooled to 0 °C, and phosphorus pentoxide (0.5 g, 3.52
mmol) was added with vigorous stirring. The solution, which
slowly turned deep red, was allowed to warm to room tem-
perature, and it was stirred for 6 h. It was then cooled again
to 0 °C and another installment of phosphorus pentoxide (0.5
g, 3.52 mmol) was added, and the mixture was stirred again
for 6 h with slow warming to room temperature. The whole
process was repeated once more with a third installment of
phosphorus pentoxide (0.5 g, 3.52 mmol). The red-colored
reaction mixture was quenched with dilute NaOH (50 mL).
The resulting white suspension was allowed to settle, and the
toluene layer was separated out and washed with water. The
aqueous white suspension was extracted with dichloromethane,
and the combined extracts were washed with water. The
toluene and dichloromethane layers were then combined
together and dried over anhydrous sodium sulfate, and the
solvent was evaporated off, leaving behind a white solid (0.25
g), which was a mixture containing the desired product. The
solid was washed with ethyl acetate, and the residue (0.10 g,
22%) left behind was determined to be pure 12 by 1H NMR.
The ethyl acetate washings were combined together, concen-
trated, and chromatographed on silica gel (eluting with 10%
ethyl acetate in dichloromethane) to yield another 0.05 g (11%)
of pure product, mp 257-260 °C; Rf ) 0.34 (10% ethyl acetate
Bis(2,5-d im eth oxyben zyl)m a lon yl Ch lor id e (8). Thionyl
chloride (10 mL, 137 mmol) was cooled to 0 °C under argon,
and 5 (0.2 g, 0.5 mmol) was added as a solid in one installment.
The heterogeneous mixture was stirred at 0 °C for 1 h and
then allowed to warm to room temperature over 3 h, during
which the solution became homogeneous. The reaction was
monitored by NMR, and the stirring was continued until the
reaction was complete. The excess thionyl chloride was dis-
tilled off under vacuum at room temperature, and the resultant
brown oil was dissolved in warm hexane, decanted into a clean
flask, and cooled overnight in a refrigerator. The buff-colored
solid that fell out was filtered and recrystallized from dry
hexane to yield 0.125 g (57%) of white crystals, mp 84-85 °C.
1H NMR (300 MHz): 6.8-6.7 (m, 4 H), 6.6 (s, 2 H), 3.67 (s, 12
H), 3.45 (s, 4 H). 13C NMR (75 MHz): 171.9, 154.0, 153.0, 124.0,
118,6, 114.3, 111.5, 77.3, 55.9, 55.2, 35.3. FTIR (KBr): 2940
(w), 2835 (w), 1788 (s), 1503 (s), 1225 (s), 1045 (s). CI-MS (m/
z, relative intensity): 442 (M+ + 2, 14), 440 (M+, 19), 369 (26),
219 (35), 151 (100).
1
in dichloromethane). H NMR (300 MHz): 7.17 (s, 2 H), 6.98
(s, 2 H), 4.01 (s, 6 H), 3.91 (s, 6 H), 3.61 (d, J ) 16.7 Hz, 2 H),
3.09 (d, J ) 16.8 Hz, 2 H). 13C NMR (75 MHz): 201.6, 155.9,
149.7, 149.4, 128.2, 107.2, 105.1, 65.7, 56.3, 56.1, 37.7. FTIR
(KBr): 3415 (m), 2924 (m), 1694 (s), 1605 (m), 1592 (m), 1502
(s), 1310 (s), 1279 (s). EI-MS (m/z, relative intensity): 369 (M+
+ 1, 24), 368 (M+, 100), 340 (M+ - CO, 15), 339 (12%). Anal.
Calcd for C21H20O6: C, 68.47; H, 5.47. Found: C, 68.44; H,
5.57.
Dib en zylm a lon ic Acid Dich lor id e11 (7). In
a flask
equipped with a condenser, drying tube, and NaOHaq trap, 43
(32.04 g, 0.113 mol) was suspended in 50 mL of chloroform.
The temperature was lowered to 0 °C, and phosphorus
pentachloride (51.78 g, 0.249 mol) was added over a 20 min
period. After stirring at room temperature for 1 h, the reaction
mixture was heated at reflux for 15 min and cooled to room
temperature. The solvent was then removed by rotary evapo-
ration, leaving behind a pale yellow liquid. This liquid was
distilled under vacuum (0.7 Torr) to yield 25.93 g (72%) of
product (bp 145-150 °C (0.7 Torr), lit.11 216-218 °C, 17 Torr).
1H NMR (200 MHz): 7.20-7.05 (m, 10 H), 3.30 (s, 4 H). EI-
2,2′-Sp ir obiin d a n -1,1′-d ion e6ab (10). Aluminum chloride
(0.660 g, 4.95 mmol) and ferric chloride (0.052 g, 0.32 mmol)
were added to 7 (12.29 g, 38.3 mmol). The reaction vessel was
placed in a Kugelrohr (Aldrich) and heated from 25 °C to 225
°C at a rate of 5 °C/min and at a pressure maintained between
3 and 5 Torr. The desired material distilled over between 180
and 210 °C and solidified in the trap. Recrystallization of the
distilled material twice from ethyl acetate yielded 2.89 (30%)
of the product as white crystals, mp 168-169 °C (lit.6a 174 °C,
lit.6b 173.6-176 °C); Rf ) 0.23 (dichloromethane). 1H NMR (300
MHz, CD2Cl2): 7.77-7.42 (m, 8 H), 3.69 (d, J ) 14 Hz, 2 H),
3.22 (d, J ) 14 Hz, 2 H). 13C NMR (50 MHz): 202.6, 153.7,
135.3, 135.1, 127.6, 126.2, 124.7, 65.2, 37.9. IR (KBr): 1710
(11) Leuchs, H.; Radulescu, D. Chem. Ber. 1912, 45, 189.

Rearrangements of Spiroketones
J . Org. Chem., Vol. 64, No. 22, 1999 8207
(m), 1680 (s), 1595 (m), 1580 (m), 1255 (s). EI-MS (m/z, relative
intensity): 249 (M+ + 1, 1), 248 (M+, 100), 247 (M+ - 1, 11),
231 (13), 220 (64), 219 (57), 203 (14), 191 (57), 189 (43), 165
(27), 118 (21).
dichloromethane as the eluting solvent. First isomer, major
component (70%): Rf ) 0.27 (25% ethyl acetate in dichlo-
romethane). 1H NMR (360 MHz): 7.14 (d, J ) 9.2 Hz, 2 H),
6.85 (d, J ) 9.3 Hz, 2 H), 5.81 (s, 2 H), 3.94 (s, 6 H), 3.85 (s, 2
H), 3.83 (s, 6 H); EI-MS (m/z, relative intensity): 400 (M+, 36),
383 (M+ - OH, 13), 382 (48), 367 (27), 207 (100). Second
isomer, minor component (30%): Rf ) 0.1 (25% ethyl acetate
An alternative procedure, giving better yields, was also used
to prepare 10. Phosphorus pentoxide (3 × 0.5 g, 10.56 mmol)
and 4 (0.5 g, 1.76 mmol) were reacted together in dry toluene
(50 mL), and the product was worked up, as described for 12.
The white solid obtained upon solvent removal contained 10
and 2-benzyl-1,3-indandione (Rf ) 0.7, dichloromethane). The
desired product was isolated using flash chromatography
(eluting with dichloromethane) as a white solid, 0.2 g (46%).
No attempt was made to recover any unreacted starting
material present in the base extracts.
1
in dichloromethane). H NMR (360 MHz): 7.12 (m, 2 H), 6.86
(m, 2 H), 5.99 (s, 1 H), 5.71 (s, 1 H), 3.94 (s, 3 H), 3.92 (s, 3 H),
3.88 (s, 3 H), 3.85 (s, 3 H), 3.42 (s, 1 H), 2.90 (s, 1 H). EI-MS
(m/z, relative intensity): 400 (M+, 36), 382 (39), 367 (15), 323
(16), 207 (100).
2,2′-Sp ir obiin d a n -4,4′,7,7′-tetr a m eth oxy-1,1′,3,3′-tetr a -
on e (2). Cold chromic acid solution (0.25 mL, 0.17 mmol,
prepared by dissolving sodium dichromate dihydrate (1.0 g,
3.36 mmol) in water (3 mL), slowly adding concentrated H2-
SO4 (1.34 g, 0.73 mL, 13.67 mmol), cooling the solution and
diluting it to 5 mL with water) was added to an ice cold
solution of 18 (0.100 g, 0.25 mmol, mixture of isomers) in
acetone (50 mL) and stirred. The reaction was monitored by
TLC. Stirring was continued at 0 °C until the reaction was
complete (3 h). The reaction mixture was diluted with dichlo-
romethane (100 mL) and subjected to standard workup. The
yellow solid obtained upon solvent removal was recrystallized
from dichloromethane (slow evaporation) to yield 0.069 g (70%)
of 2 as small, needle-shaped yellow crystals, mp 325-328
°C; Rf ) 0.58 (20% ethyl acetate in dichloromethane). 1H
NMR (300 MHz): 7.34 (s, 4 H), 3.97 (s, 12 H); (200 MHz, CD2-
Cl2): 7.35 (s, 4 H), 3.97 (s, 12 H). FTIR (KBr): 1693 (s), 1279
(s), 1211 (s). EI-MS (m/z, relative intensity): 397 (M+ + 1,
23), 396 (M+, 27), 382 (72), 381 (33), 369 (30), 368 (33), 367
(45), 353 (40), 338 (28), 337 (18), 323 (34), 309 (22). Anal.
Calcd for C21H16O8: C, 63.64; H, 4.07. Found: C, 63.52; H,
4.30.
2,2′-Sp ir ob iin d a n -3,3,3′,3′-t e t r a b r om o-5,5′,6,6′-t e t -
r a m eth oxy-1,1′-d ion e (16). NBS (0.918 g, 3.44 mmol) and
12 (0.50 g, 1.36 mmol) were reacted together as described for
17 using AIBN as initiator. The reaction was monitored by
1H NMR. The dibromo and the tribromo compounds could be
identified as intermediates. NBS (0.230 g, 0.86 mmol) and
AIBN (10 mg) were added every 12 h until the reaction was
complete (96 h). The reaction mixture was worked up as
described for 17, and the pure product obtained after washing
with ethyl acetate was 0.830 g (90%). No further purification
was necessary; mp > 350 °C (decomp). 1H NMR (300 MHz):
7.52 (s, 2 H), 6.99 (s, 2 H), 4.16 (s, 6 H), 3.91 (s, 6 H). 13C NMR
(50 MHz): 186.0, 157.4, 152.5, 152.1, 122.4, 106.9, 104.5, 57.5,
56.6, 55.2. FTIR (KBr): 3480 (w), 3414 (w), 1708 (s), 1616 (w),
2,2′-Sp ir obiin d a n -1,3,1′-tr ion e (13). A solution of chro-
mium oxide (8.08 g, 80.8 mmol) in 8 mL of water and 50 mL
of glacial acetic acid was added to a suspension of 10 (2.49 g,
10.1 mmol) in 50 mL of glacial acetic acid. The reaction
mixture was heated at 50-55 °C for 2 h. The reaction was
cooled to room temperature, poured into 100 mL of ice-water,
and then extracted with dichloromethane. The combined
organic layers were washed with water until the aqueous layer
was no longer acidic. Due to the reactivity of the product, a
basic wash was avoided. After drying over anhydrous sodium
sulfate, the solvent was removed by rotary evaporation, leaving
behind a mixture of starting material and the desired product,
as determined by TLC. The materials were separated by flash
chromatography eluting with 20% hexane in dichloromethane.
Starting material was recovered, 0.47 g (17%), along with 1.57
g (52%) of the desired product. The product was further
purified by recrystallization from ethyl acetate, mp 164-165
1
°C; Rf ) 0.38 (dichloromethane). H NMR (200 MHz): 8.11-
7.89 (m, 4 H), 7.76-7.40 (m, 4 H), 3.63 (s, 2 H). 13C NMR (50
MHz): 196.7, 195.8, 144.2, 143.8, 135.9, 135.8, 134.4, 128.1,
126.5, 125.3, 124.1, 73.0, 31.9. IR (KBr): 3020 (w), 1745 (m),
1700 (s), 1585 (s), 1450 (m), 1400 (m), 1310 (m), 1235 (s), 1185
(m), 965 (m), 890 (m) 830 (m), 740 (s). EI-MS (m/z, relative
intensity): 263 (M+ + 1, 18), 262 (M+, 100), 234 (15), 233 (30),
206 (32), 205 (24). HREI-MS: calculated for C17H10O3 262.0630,
found 262.0629. Anal. Calcd C, 77.86; H, 3.84. Found: C, 77.99;
H, 3.92.
3,3′-Dibr om o-2,2′-sp ir obiin d a n -4,4′,7,7′-tetr a m eth oxy-
1,1′-d ion e (17). NBS (0.484 g, 2.72 mmol) and a catalytic
amount of AIBN (10 mg) were added to 11 (0.5 g, 1.36 mmol)
in deoxygenated carbon tetrachloride (25 mL) and refluxed.
After 6 h more NBS (0.484 g, 2.72 mmol) and AIBN (10 mg)
were added, and reflux was continued. The reaction was
monitored by TLC. The monobromo and dibromo products
could also be identified by 1H NMR. Catalytic amounts of AIBN
(10 mg) were added every 6 h until the reaction was complete
(24 h). The reaction mixture was cooled to room temperature,
diluted with dichloromethane (50 mL), and subjected to the
standard workup, leaving behind a yellow liquid. On addition
of ethyl acetate (10 mL), a yellow solid fell out. The solid was
filtered and washed with a minimal amount of cold ethyl
acetate (10 mL), leaving behind 0.640 g (90%) of 17 (single
isomer) as a pale yellow powder. No further purification was
necessary: Rf ) 0.4 (dichloromethane); mp 223-225 °C. 1H
NMR (200 MHz): 7.14 (d, J ) 9.0 Hz, 2 H), 6.82 (d, J ) 8.9
Hz, 2 H), 5.75 (s, 2 H), 4.0 (s, 6 H), 3.80 (s, 6 H). 13C NMR (50
MHz): 190.7, 152.1, 148.7, 143.4, 121.0, 119.6, 112.7, 82.4,
56.5, 56.0, 46.4. FTIR (KBr): 1712 (s), 1503 (s), 1271 (s). EI-
MS (m/z, relative intensity): 528 (M+ + 4, 4), 526 (M+ + 2, 7),
524 (M+, 4), 447 (M+ - Br + 2, 71), 445 (M+ - Br, 70), 366
(M+ - 2 Br, 100), 351 (48).
3,3′-Dih ydr oxy-2,2′-spir obiin dan -4,4′,7,7′-tetr am eth oxy-
1,1′-d ion e (18). A solution of 17 (0.50 g, 0.95 mmol) in a 50%
dioxane/water mixture (50 mL) was refluxed for 12 h, during
which time the reaction was monitored by TLC. When the
reaction was complete, the mixture was cooled to room
temperature, and the solvents were removed by rotary evapo-
ration, leaving behind a mixture of the two isomers of 18 as a
pale yellow solid, 0.37 g (97%). No further purification was
necessary. A small sample of the mixture was separated on a
preparative scale TLC plate using 25% ethyl acetate in
1501 (s), 1294 (s). EI-MS (m/z, relative intensity): 607 (M+
-
Br + 6, 11), 605 (M+ - Br + 4, 23), 604 (10), 603 (M+ - Br +
2, 24), 601 (M+ - Br, 11), 526 (M+ - 2 Br + 4, 9), 525 (14),
524 (M+ - 2 Br + 2, 10), 522 (M+ - 2 Br, 6), 447 (M+ - 3 Br
+ 4, 25), 446 (81), 445 (M+ - 3 Br + 2, 52), 444 (75), 443 (M+
- 3 Br, 35).
2,2′-Sp ir obiin d a n -5,5′,6,6′-tetr a m eth oxy-1,1′,3,3′-tetr a -
on e (3). Silver acetate (0.100 g, 0.599 mmol) was added to a
suspension of 16 (0.100 g, 0.147 mmol) in glacial acetic acid
(10 mL), and the mixture refluxed for 4 h. The hot reaction
mixture was filtered at the aspirator, and the residue was
washed with dichloromethane. The combined organic layers
were washed with water until the washings were no longer
acidic and dried over anhydrous sodium sulfate. (A base wash
was avoided due to the reactivity of the tetraone.) The yellow
solid obtained upon solvent evaporation was recrystallized
from dichloromethane (slow evaporation) to give 0.052 g (89%)
of white crystals, mp > 350 °C (decom); Rf ) 0.63 (10% ethyl
acetate in dichloromethane). 1H NMR (200 MHz): 7.43 (s, 4
H), 4.06 (s, 12 H). 13C NMR (50 MHz): 191.0, 156.4, 140.1,
104.1, 56.8. FTIR (KBr): 2938 (w), 2838 (w), 1718 (m), 1693
(s), 1300 (s). EI-MS (m/z, relative intensity): 397 (M+ + 1, 23),
396 (M+, 100), 164 (10), 136 (20), 93 (13), 28 (CO, 75). Anal.
Calcd for C21H16O8: C, 63.64; H, 4.07. Found: C, 63.75; H,
4.24.
3′,3′-Dibr om o-2,2′-sp ir obiin d a n -1,3,1′-tr ion e (14). NBS
(0.780 g, 4.34 mmol), a catalytic amount of AIBN (10 mg), and

8208 J . Org. Chem., Vol. 64, No. 22, 1999
Maslak et al.
0.5 mL of bromine were added to 13 (0.563 g, 2.15 mmol) in
25 mL of deoxygenated carbon tetrachloride. The reaction was
carried out as described for 15, with further additions of AIBN
(10 mg) every 12 h and NBS (0.780 g, 4.34 mmol) every 24 h,
until the reaction was complete (90 h), as indicated by TLC.
The reaction mixture was worked up as described for 15, and
the pure product was isolated by flash chromatography eluting
with 20% hexane in dichloromethane, yielding 0.529 g (58%);
mp 120 °C (decomp); Rf ) 0.56 (dichloromethane). 1H NMR
(360 MHz): 8.18-7.58 (m, 8 H). IR (KBr): 3000-2900 (m),
1770 (m), 1700 (s), 1170 (s). EI-MS (m/z, relative intensity):
342 (18), 341 (100), 340 (18), 339 (99), 313 (12), 311 (12), 261
(17), 260 (13), 232 (20), 204 (26), 176 (44).
2,2′-Sp ir obiin d a n tetr a on e (1). Silver acetate (0.451 g, 2.70
mmol) was added to a suspension of 14 (0.450 g, 1.07 mmol)
in 30 mL of glacial acetic acid. The reaction was carried out
and the product was worked up as described for 3, to yield
0.281 g of a white solid. The product was purified by recrys-
tallization from ethyl acetate to yield 0.260 g (88%); mp 244-
246 °C; Rf ) 0.57 (dichloromethane). 1H NMR (200 MHz):
8.11-7.91 (m, 8 H). 13C NMR (50 MHz): 191.4, 144.6, 136.2,
124.3, 78.5. IR (KBr): 3020 (w), 1700 (s), 1575 (s), 1210 (s),
1140 (s). EI-MS (m/z, relative intensity): 277 (M+ + 1, 18),
276 (M+, 100), 248 (30), 220 (14). HREI-MS: calculated for
2974 (s), 2904 (w), 2883 (w), 1744 (s), 1703 (s), 1594 (m), 1286
(s), 1200 (s). EI-MS (m/z, relative intensity): 176 (M+ + 2, 1),
175 (M+ +1, 12), 174 (M+, 100), 159 (96), 146 (16).
4,7-Dim eth oxy-2,2-d im eth ylin d a n -1,3-d ion e13 (20). Poly-
phosphoric acid (10.0 g) was added to a mixture of 2,2-
dimethylmalonic acid (0.60 g, 4.55 mmol) and 1,4-dimeth-
oxybenzene (0.50 g, 3.62 mmol), and the viscous mixture was
mechanically stirred and heated to 70 °C. The reaction mixture
turned yellow and then bright red. Stirring was continued at
70 °C for 3 h. The reaction was then cooled to room temper-
ature, quenched with ice-water (100 mL), and stirred until
the red reaction mixture turned white. The aqueous mixture
was extracted with dichloromethane, and the combined organic
extracts were given a base wash and then subjected to the
standard workup. The desired product was isolated from the
pale yellow solid obtained upon solvent removal by flash
column chromatography using 10% ethyl acetate in dichlo-
romethane as the eluting solvent. The white powder obtained
was recrystallized from 95% ethanol to yield 0.720 g (85%) of
white crystals: mp 167-168 °C (lit.13 mp 170-171 °C); Rf )
0.31 (10% ethyl acetate in hexane). 1H NMR (200 MHz): 7.28
(s, 2 H), 3.99 (s, 6 H), 1.27 (s, 6 H). 13C NMR (50 MHz): 203.9,
152.2, 128.2, 120.7, 56.8, 50.6, 20.6. FTIR (KBr): 2969 (m),
1737 (s), 1701 (s), 1580 (s), 1499 (s), 1054 (s), 1018 (s). CI-MS
(m/z, relative intensity): 236 (M+ + 2, 15), 235 (M+ + 1, 100),
234 (M+, 20), 233 (1), 219 (M+ - CH3, 2).
C
17H8O4 276.0423, found 276.0426. Anal. Calcd: C, 73.91; H,
2.92. Found: C, 73.72; H, 3.28.
5,6-Dim eth oxy-2,2-d im eth ylin d a n -1,3-d ion e (21). Poly-
phosphoric acid (10.0 g) was added to a mixture of 2,2-
dimethylmalonic acid (0.60 g, 4.55 mmol) and 1,2-dimeth-
oxybenzene (0.50 g, 3.62 mmol), the reaction was carried out,
and the product was worked up as described for 20. The
desired product was isolated by flash column chromatography
eluting with 5% ethyl acetate in dichloromethane in the form
of a white powder. The product was recrystallized from 95%
ethanol in the form of white needles: 0.570 g (67%); mp 224-
An alternative procedure (starting from 15, see below) was
developed that gave a better yield of 1 in fewer steps. Silver
acetate (0.601 g, 3.599 mmol) was added to a suspension of
15 (0.493 g, 0.880 mmol) in 25 mL of glacial acetic acid. The
reaction was carried out and the product was worked up as
described for 3, to yield 0.207 g (85%) of pure 1 as a white
solid.
2,2′-Sp ir obiin d a n -3,3,3′,3′-tetr a br om o-1,1′-d ion e (15).
NBS (0.750 g, 4.213 mmol), a catalytic amount of AIBN (10
mg), and bromine (0.5 mL) were added to 10 (0.501 g, 2.02
mmol) in deoxygenated carbon tetrachloride (25 mL). The
reaction was heated at reflux and monitored by TLC. More
NBS (0.750 g, 4.213 mmol) was added after 24, 48, and 72 h.
Catalytic amounts of AIBN (10 mg) were added every 6 h until
the reaction was complete (100 h, no change by TLC over the
last 6 h). Also, bromine (0.5 mL) was added if the solution had
noticeably decreased in bromine concentration (indicated by
decrease in color intensity of the solution). The reaction
mixture was cooled to room temperature, diluted with dichlo-
romethane (50 mL), and subjected to the standard workup,
leaving behind a yellow liquid containing the desired product.
The product was isolated as a pale yellow solid using flash
column chromatography (eluting with dichloromethane), yield-
ing 0.475 g (42%); mp > 200 °C (decomp); Rf ) 0.69 (dichlo-
1
226 °C; Rf ) 0.46 (10% ethyl acetate in dichloromethane). H
NMR (300 MHz): 7.34 (s, 2 H), 4.03 (s, 6 H), 1.28 (s, 6 H). 13C
NMR (50 MHz): 203.5, 156.0, 135.1, 103.5, 56.6, 49.4, 20.3.
FTIR (KBr): 2985 (w), 1729 (s), 1698 (s), 1583 (s), 1502 (s),
1381 (s), 1317 (s), 1108 (s), 1006 (s). EI-MS (m/z, relative
intensity): 235 (M+ + 1, 9), 234 (M+, 42), 220 (10), 219 (M+
-
CH3, 34), 206 (17), 205 (10), 191 (M+ - CH3 - CO, 28), 175
(M+ - CH3 - 2 CO, 12). Anal. Calcd for C13H14O4: C, 66.66;
H, 6.02. Found: C, 66.51; H, 6.33.
2,2′-Dim eth yl-[2,2′]-biin d en yl-1,3,1′,3′-tetr a on e (22) was
prepared by a modification of the procedure described by
Spencer et al.,14 who obtained 22 as a side product in a reaction
used to prepare 2-acetoxy-2-methylindan-1,3-dione. Lead tet-
raacetate (3.0 g, 6.77 mmol) was added to a suspension of
2-methylindan-1,3-dione15 (1.00 g, 6.25 mmol) in benzene (5
mL) and stirred for 20 h. A white precipitate was formed. The
reaction was quenched with dilute HCl (50 mL), diluted with
dichloromethane (50 mL), and stirred for 30 min. The organic
layer was separated, and the aqueous layer was extracted with
dichloromethane. The organic layers were combined and
subjected to the standard workup. The desired product was
isolated from the yellow solid obtained upon solvent removal
by flash column chromatography using dichloromethane as
eluant and recrystallized from 95% ethanol to give white
needles (0.64 g, 32%): mp 203-205 °C (lit.14 204-205 °C); Rf
) 0.27 (dichloromethane). 1H NMR (200 MHz): 7.95-7.75 (m,
8 H), 1.71 (s, 6 H). 13C NMR (50 MHz): 201.2, 140.5, 135.9,
123.6, 55.4, 15.7. FTIR (KBr): 2977 (w), 1748 (m), 1737 (m),
1705 (s), 1592 (m). EI-MS (m/z, relative intensity): 319 (M+
+ 1, 22), 318 (M+, 100), 275 (M+ - CH3 - CO, 5), 160 (16),
132 (13).
1
romethane). H NMR (360 MHz): 8.10 (d, J ) Hz, 2 H), 7.85
(t, J ) Hz, 2 H), 7. 58 (d, J ) Hz, 2 H), 7.45 (t, J ) Hz, 2 H).
EI-MS (m/z, relative intensity): 487 (M+ - Br + 6, 5), 485
(M+ - Br + 4, 15), 483 (M+ - Br + 2, 55), 481 (M+ -Br, 5),
406 (M+ - 2 Br + 4, 5), 404 (M+ - 2 Br + 2, 11), 402 (M+
-
2 Br, 6), 326 (100), 325 (M+ - 3 Br + 2, 51), 324 (97), 323 (M+
- 3 Br, 34).
2,2-Dim eth ylin d a n -1,3-d ion e12 (19) was prepared by us-
ing the procedure described by Wislicenus and Kotzle.12
Triethylbenzylammonium chloride (0.050 g) and sodium hy-
droxide solution (75 mL, 0.6 M) were added to a solution of
iodomethane (0.9 mL, 14.31 mmol) and 1,3-indandione (1.03
g, 6.84 mmol) in dichloromethane (75 mL), and the mixture
was stirred for 100 h. The organic layer was then separated
and washed with a 5% sodium hydroxide solution and sub-
jected to the standard workup. The black oily solid obtained
after solvent removal was dissolved in the minimum quantity
of hexane. Upon cooling to -78 °C, white crystals fell out of
solution and were quickly filtered and dried: 0.272 g (22%);
mp 104-105 °C (lit.12 107-108 °C). 1H NMR (200 MHz): 8.03-
7.86 (m, 4 H), 1.29 (s, 6 H). FTIR (KBr): 3435 (w), 3087 (w),
Tr ea t m en t of 13 w it h Sod iu m E t h oxid e or Sod iu m
Meth oxid e. Under argon 2 sodium spheres (0.043 g, 1.87
mmol) were added to 5 mL of ROH (reagent grade methanol
or absolute ethanol). After the spheres had completely reacted,
(13) Fleischer, K. J ustus Liebig Ann. Chem. 1821, 422, 239.
(14) Spencer, T. A.; Hall, A. L.; Fordham von Reyn, C. J . Org. Chem.
1968, 33, 3369.
(12) Wislicenus, W.; Kotzle, A. J ustus Liebig Ann. Chem. 1889, 86,
252.
(15) Wojack, G. Chem. Ber. 1938, 71, 1102.

Rearrangements of Spiroketones
J . Org. Chem., Vol. 64, No. 22, 1999 8209
13 (0.608 g, 0.232 mmol) was added to the NaOR solution. The
reaction mixture turned yellow immediately. The UV/vis
spectrum showed the presence of the enolate (in CH3CN λmax
) 360 nm, ꢀ ) 8.1 × 104 M-1 cm-1). For comparison, the trione
(13) has the longest wavelength transition at 290 nm (ꢀ ) 7.8
× 103 M-1 cm-1 in CH3CN). The solution was neutralized with
dilute HCl, which caused the yellow color to disappear. Water,
50 mL, was added, the solution was extracted with dichlo-
romethane, and the extract was subjected to the standard
workup. Removal of solvent gave an oily solid residue, ap-
proximately 50 mg of 23 for R ) Me or of 24 for R ) Et. The
materials were not further purified for spectral analyses, nor
for the hydrolysis reactions. The UV/vis spectrum of the
product (23) showed the presence of the conjugated chro-
mophore (enol: λmax ) 328 nm, ꢀ ) 1.4 × 105 M-1 cm-1 in CH3-
CN).
ethylammonium chloride (10 mg) were dissolved in 10 mL of
dichloromethane and 10 mL of 10% aqueous sodium hydroxide,
and the solution was stirred overnight. The aqueous layer was
separated and acidified with 5% aqueous hydrochloric acid,
causing a white oily material to fall out of the solution. The
oil was taken up in ether, and the ether layer was separated.
The aqueous layer was extracted twice with ether. The
combined ether layers were subjected to the standard workup.
The residue obtained upon solvent removal, 26 (approximately
1
50 mg), was not further purified. H NMR (360 MHz): 7.92-
7.30 (m, 8 H), 3.47 (dd, J ) 9 Hz, J ) 15 Hz, 1 H), 2.96 (dd, J
) 15 Hz, J ) 31 Hz, 1 H), 2.41 (dd, J ) 9 Hz, J ) 31 Hz, 1 H).
13C NMR (70 MHz): 207.1, 168.0, 153.8, 147.0, 136.2, 135.8,
134.9, 131.0, 128.1, 127.3, 126.5, 125.6, 124.6, 122.5, 107.1,
51.2, 29.0. IR (KBr): 3200-2800 (w), 1695 (s), 1650 (s), 1605
(s), 1285 (s), 1235 (s). EI-MS (m/z, relative intensity): 280 (M+,
2), 263 (11), 262 (58), 234 (8%), 233 (15), 206 (15), 205 (11),
178 (18), 176 (9), 149 (32), 133 (11), 132 (100). HREI-MS,
calculated for C17H12O4 280.0736, found 280.0742.
23: 1H NMR (200 MHz): 7.96-7.80 (m, 2 H), 7.60-7.33 (m,
6 H), 3.69 (s, 3 H), 3.45 (s, 2 H). 13C NMR (50 MHz, k ) minor
keto form): 201.3(k), 200.1(k), 189.6, 178.4, 167.0, 154.7(k),
147.7, 143.0(k), 137.7, 136.6, 135.4(k), 135.1(k), 132.6, 131.9,
130.2, 129.8, 129.7(k), 129.6, 128.3, 127.6(k), 127.4(k), 127.2,
127.0(k), 125.5, 124.2(k), 123.0, 111.0, 61.2(k), 52.4, 31.1, 30.3-
(k). EI-MS (m/z, relative intensity): 295 (M+ + 1, 2), 294 (M+,
11), 262 (20), 235 (8), 233 (6), 205 (4), 178 (6), 164 (10), 163
(100). HREI-MS: calculated for C18H14O4, 294.0892, found
294.0877.
Tr ea tm en t of 13 w ith Sod iu m Hyd r oxid e. Compound
13 (13.2 mg, 0.050 mmol) was added to 15 mL of 10% aqueous
sodium hydroxide and 10 mL of dichloromethane. A catalytic
amount of benzyltriethylammonium chloride (2 mg) was
added, and the solution was stirred for 4 h, after which the
aqueous layer was removed and acidified with concentrated
hydrochloric acid. The cloudy solution was extracted twice with
dichloromethane. The combined organic layers were subjected
to the standard workup. The residue obtained upon solvent
removal, 26 (approximately 15 mg), was not further purified.
24: 1H NMR (300 MHz): 8.01-7.84 (m, 2 H), 7.64-7.41 (m,
6 H), 4.27 (q, J ) 7 Hz, 2 H), 3.47 (s, 2 H), 1.22 (t, J ) 7 Hz,
3 H). IR (in CHCl3): 3010-2920 (w), 1720 (s), 1650 (s), 1610
(s), 1590 (s). EI-MS (m/z, relative intensity): 308 (M+, 6.6),
263 (13), 262 (40), 235 (11), 233 (11), 178 (16), 177 (60), 150
(10), 149 (100).
Tr ea tm en t of 1 w ith Sod iu m Eth oxid e. Two sodium
spheres (0.043 g, 1.87 mmol) were added to 10 mL of absolute
ethanol, and after they had completely reacted, 1 (0.044 g, 0.16
mmol) was added at 25 °C. After stirring for 2 h, the solution
was acidified with concentrated hydrochloric acid. Water, 50
mL, was added, and the mixture was extracted with dichlo-
romethane. The combined organic layers were subjected to the
standard workup, yielding a solid residue of 27 (approximately
40 mg), which was characterized without further purification.
1H NMR (360 MHz): 8.11-8.07 (m, 1 H), 7.90-7.86 (m, 1 H),
7.79-7.55 (m, 6 H), 4.27 (q, J ) 7 Hz, 2 H), 1.25 (t, J ) 7 Hz,
3 H). 13C NMR (90 MHz): 197.4, 186.7, 180.1, 165.7, 140.4,
138.4, 135.1, 134.1, 132.5, 131.6, 131.0, 130.4, 130.1, 129.5,
122.8, 122.4, 108.6, 61.4, 13.8. IR (in CHCl3): 3020-2840 (w),
1715 (s), 1650 (s), 1610 (s), 1590 (s). EI-MS (m/z, relative
intensity): 322 (M+, 1), 277 (8), 249 (100), 226 (15), 211 (47).
HREI-MS, calculated for C17H9O4 (M+ - OEt) 277.0501, found
277.0489. Hydrolysis of 27 in refluxing 6 M hydrochloric acid
gave 1,3-indandione.
Hyd r olysis of 23 a n d 24. Approximately 50 mg of 23 or
24 and 10 mL of 6 M aqueous hydrochloric acid were combined
and refluxed for 1 h. During the reaction, a white solid formed,
which was suspended on top of the solution. Water, 50 mL,
was added, and the reaction material was extracted with
dichloromethane and was subjected to the standard workup.
The residue obtained upon solvent removal, approximately 50
mg, was not further purified for spectral analyses nor for
further reactions. The hydrolyses of both (23 and 24) gave 25.
1H NMR (360 MHz, for proton labeling see Scheme 2): 9.42
(dd, J 1-2 ) 7.5 Hz, J 1-3 ) 1.0 Hz, H1), 7.92 (dd, J 3-4 ) 7.6 Hz,
J 2-4 ) 1.0 Hz, H4), 7.80 (ddd, J 1-2 ) J 2-3 ) 7.5 Hz, J 2-4 ) 1.0
Hz, H2), 7.805 (dd, J 7-8 ) 7.5 Hz, J 6-8 ) 1.0 Hz, H8), 7.64 (ddd,
J 2-3 ) J 3-4 ) 7.5 Hz, J 1-3 ) 1.0 Hz, H3), 7.58 (ddd, J 5-6 ) J 6-7
- 7.4 Hz, J 6-8 ) 1.0 Hz, H6), 7.48 (dd, J 5-6 ) 7.4 Hz, J 5-7
)
1.0 Hz, H5), 7.38 (ddd, J 6-7 ) J 7-8 ) 7.4 Hz, J 5-7 ) 1.0 Hz,
H7), 4.03 (s, CH2). NOE effect (1H NMR) was observed between
H5 and the CH2 group, H1 and H2, as well as H3 and H4. No
NOE effects were observed between the CH2 group and H1 or
H4. 13C NMR (75 MHz, CD2Cl2): 193.1, 166.8, 148.6, 140.0,
137.3, 136.0, 135.2, 132.7, 128.0, 127.6, 126.2, 125.7, 124.1,
120.0, 31.9. IR (KBr): 3110-3080 (w), 1785 (s), 1685 (s), 1630
(s), 1600 (m), 925 (s). EI-MS (m/z, relative intensity): 264 (M+
+ 2, 2), 263 (M+ + 1, 18), 262 (M+ 100), 234 (10), 233 (21), 206
Ack n ow led gm en t. This research was supported by
a grant from the Petroleum Research Fund adminis-
tered by the American Chemical Society.
Su p p or t in g In for m a t ion Ava ila b le: 1H NMR and 13C
NMR spectra of 13, 23-26. This material is available free of
charge via the Internet at http://pubs.acs.org.
(20). HREI-MS: calculated for
262.0627.
C17H10O3 262.0630, found
Tr ea tm en t of 25 w ith Sod iu m Hyd r oxid e. Compound
25 (0.050 g, 0.19 mmol) and a catalytic amount of benzyltri-
J O990867J