Reactions of carbonyl compounds in basic solutions. Part 24. 1 The mechanism of the base-catalysed ring fission of substituted benzocyclobutene-1,2-diones

Article abstract of DOI:10.1039/a606310a

The rate coefficients for the base-catalysed ring fission of a series of substituted benzocyclobutenediones to give the corresponding 2-formylbenzoic acids have been determined in water at 25.0 and 60.0°C. The effects of 4-substituents and 4,5-di-substituents on the rates have been correlated using a modified Hammett equation to give a reaction constant, ρ, equal to ca. 3.6 at 25.0°C. The activation parameters have been calculated. The effect of solvent composition on the rates has been studied. The kinetic solvent isotope effect, product composition and enrichment in ¹⁸O-enriched water have also been studied. All the evidence indicates a mechanistic pathway which proceeds by rapid reversible addition of hydroxide anion to the dione, followed by intramolecular nucleophilic attack on the second carbonyl group and formation of a carbanionic intermediate.

Full text of DOI:10.1039/a606310a

View Article Online / Journal Homepage / Table of Contents for this issue
Reactions of carbonyl compounds in basic solutions. Part 24.1
The mechanism of the base-catalysed ring fission of substituted
benzocyclobutene-1,2-diones
Keith Bowden* and M. Vahid Horri
Department of Biological and Chemical Sciences, Central Campus, University of Essex,
Wivenhoe Park, Colchester, Essex, UK CO4 3SQ
The rate coefficients for the base-catalysed ring fission of a series of substituted benzocyclobutenediones
to give the corresponding 2-formylbenzoic acids have been determined in water at 25.0 and 60.0 ЊC. The
effects of 4-substituents and 4,5-di-substituents on the rates have been correlated using a modified
H ammett equation to give a reaction constant, ñ, equal to ca. 3.6 at 25.0 ЊC. The activation parameters
have been calculated. The effect of solvent composition on the rates has been studied. The kinetic solvent
isotope effect, product composition and enrichment in 18O-enriched water have also been studied. All the
evidence indicates a mechanistic pathway which proceeds by rapid reversible addition of hydroxide anion
to the dione, followed by intramolecular nucleophilic attack on the second carbonyl group and formation
of a carbanionic intermediate.
Table 1 Rate coefficients (k2) for the base-catalysed ring fission of
The chemistry of benzocyclobutenediones (bicyclo[4.2.0]octa-
1,3,5-triene-7,8-diones) has been reviewed.2 The base-catalysed
ring fission of benzocyclobutenediones 1 has been shown to
give the anions of 2-formylbenzoic acids 2.2,3 This reaction is in
a,b
substituted benzocyclobutenediones in water (µ = 0.1 mol dmϪ3
k2/dm3 molϪ1 sϪ1
)
Substituent(s)
at 25.0 ЊC
at 60.0 ЊC
λ/nmc
H
4-CH3
283 (490)d
93.4
1030
400
302.5
303
4-OCH3
4-Cl
4,5-(CH3)2
3,6-(CH3)2
4,5-(Cl)2
37.1
3750
25.2
0.0733
38 500
265
11 800
89.9
0.975
62 500
325
305
306, 316
331.5
315
a
Water containing 0.33% (v/v) DMSO. b Rate coefficients were repro-
ducible to ±5%. c Wavelengths used to monitor the reactions. d In
deuterium oxide.
acid anions in quantitative yield. The reaction is of first order in
both the diones and hydroxide anion. Rate coefficients for the
base-catalysed ring fission of the substituted benzocyclobutene-
1,2-diones in water at 25.0 and 60.0 ЊC are shown in Table 1, as
well as that for the parent dione in D2O. The rates of reaction
are relatively very fast and k2 for the unsubstituted dione at
25 ЊC in water is 283 dm3 molϪ1 sϪ1. The activation parameters
for the reactions of these compounds are shown in Table 2. The
effects of increasing concentrations of dimethyl sulfoxide
(DMSO) and dioxane on the rate of ring fission are shown in
Table 3. The products of ring fission are the anions of the sub-
stituted 2-formylbenzoic acids 2. For the unsubstituted and
stark contrast to that same reaction of simple substituted
cyclobutenediones which results in the formation of the anions
of substituted 2-oxobut-3-enoic acids.2,4 The former reaction
involves fission between the carbonyl carbons, while the latter
reaction involves fission between the carbonyl carbon and the
olefinic carbon. Both these reactions are also in contrast to the
benzilic acid rearrangement, the base-catalysed rearrangement
of benzils and other α-diketones, in which there is a 1,2-aryl
shift.5,6 However, benzils can also react with base to give either
aryl–carbonyl or carbonyl–carbonyl fission. These appear to be
limiting reactions resulting from extremes of substituent
effects.6 Furthermore, a study of the base-catalysed ring fission
benzocyclobutenones indicates that cleavage of both the aryl
carbon–carbonyl carbon bond and benzyl carbon–carbonyl
carbon bond can occur.7
The present report describes a study of the kinetics and
products of the base-catalysed ring fission of 4-substituted and
symmetrically 3,6- and 4,5-di-substituted benzocyclobutene-
diones. The effects of substitution, solvent composition and
solvent kinetic isotope, as well as product enrichment in 18O-
enriched water and activation parameters, are discussed and a
mechanism is suggested.
Results
symmetrically di-substituted diones only one product, 3/4, was
isolated, as expected. However, for the 4-substituted diones, the
product can be the 4- or 5-substituted 2-formylbenzoic acids,
The base-catalysed ring fission of the substituted benzocyclo-
butene-1,2-diones gives the corresponding 2-formylbenzoic
J. Chem. Soc., Perkin Trans. 2, 1997
989

View Article Online
Table 2 Activation parameters for the base-catalysed ring fission of
substituted benzocyclobutenediones in water (µ = 0.1 mol dmϪ3) at
30 ЊCa,b
X
Y
X
Y
OH
O–
O
O
k ′1
+ OH–
k ′–1
∆H ‡/kcal molϪ1
∆S ‡/cal molϪ1 KϪ1
O
Substituent(s)
1
9
H
4-CH3
6.7
7.6
10.5
5.9
6.6
14.0
2.1
Ϫ25
Ϫ24
Ϫ16
Ϫ23
Ϫ30
Ϫ17
Ϫ30
4-OCH3
4-Cl
4,5-(CH3)2
3,6-(CH3)2
4,5-(Cl)2
k ′–2 k ′2
OH
X
X
Y
–
CO2
k ′3
O
–
See Table 1. b Values of ∆H ‡ and ∆S ‡ are considered accurate to
a
CHO
within ±300 cal molϪ1 and ±2 cal molϪ1 KϪ1, respectively (1 cal = 4.184 J).
Y
O
10
Table 3 Rate coefficients (k2) for the base-catalysed ring fission of
benzocyclobutendione in aqueous DMSO and dioxane (µ = 0.1 mol
dmϪ3) at 25.0 ЊCa,b
Scheme 1
k2/dm3 molϪ1 sϪ1
Volume (%)
non-aqueous solvent
DMSO
Dioxane
0
30
50
60
70
283
376
1070
1610
3760
283
165
140
146
138
a,b See Table 1.
H
OH
O
X
CHO
X
CO2H
O
5
6
a X = CH3
b X = OCH3
c X = Cl
H
OH
O
CHO
X
CO2H
X
O
Scheme 2
7
8
10. The formation of the adduct 9 will involve ring angle strain,
but this will be less severe than in the starting compound 1.
However, the adduct 9 does not break down by fission of the
aryl carbon–carbonyl carbon bond. Thus, an energetically pre-
ferable route for fission must exist. The collapse of the adduct 9
to a 1-hydroxybenzocyclopropene-1-carboxylic acid anion 13 is
a X = CH3
b X = OCH3
c X = Cl
5/6 and 7/8, respectively. The 4-methoxy and 4-chloro diones
gave only the 5-substituted 2-formylbenzoic acids, 7/8; but the
4-methyl dione gave the 4- and 5-substituted 2-formylbenzoic
acids in the ratio 1:3.3.
When the base-catalysed ring fission was studied in 18O-
enriched water, the product of the ring fission, 2-formylbenzoic
acid 3a/4a showed three enrichments; whereas the exchange of
the latter acid itself studied under identical base-catalysed con-
ditions gave only one enrichment.
blocked because of excessive ring angle strain in the fused bi-
cyclo system. Such a reaction pathway occurs for the related 3,4-
diphenylcyclobutene-1,2-diones 14 that do not have a fused ring
system.4 The latter is a benzilic acid-type rearrangement.5 The
efficacy of a leaving group has been considered to be a function
of the stability of anion formed.10 The pKa of benzaldehyde
acting as a Brønsted acid to form 10 has not been measured; but
must be very large, cf. ref. 11. Aryl carbon–carbonyl carbon
bond fission of the adduct 9 would appear to be more favour-
able. Thus, the route shown in Scheme 1 seems unlikely.
Discussion
The two likely mechanisms for the base-catalysed ring fission are
shown in Schemes 1 and 2. Both schemes involve initial form-
ation of the tetrahedral adduct 9. The studies of hydrolysis in
18O-enriched water clearly indicate rapid, reversible hydration
with equilibration of the adduct. A similar situation exists for
benzil in the benzilic acid rearrangement.8 The solvent kinetic
isotope effect, kD2 O/kH2 O, was found to be 1.73 at 25 ЊC. This
is in agreement with formation of the adduct in a rapid pre-
equilibrium, arising from the greater nucleophilicity of ODϪ in
D2O than OHϪ in H2O, cf. ref. 9. In Scheme 1, the adduct
breaks down with ring fission to give the anion of benzaldehyde
2
2
In Scheme 2, the adduct 9 effects intramolecular nucleophilic
attack on the second carbonyl group. Such a process is ana-
logous to the occurrence of neighbouring group participation
990
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
by a proximate keto– or formyl–carbonyl group in the alkaline
hydrolysis of esters.12 The intermediate 11 is a fused tricyclic
system and will be highly strained. This intermediate is a
dehydro derivative of the known cyclic hydrate of o-phthal-
aldehyde.13 However, the carbanionic intermediate 12 could
be formed directly from 9. Furthermore, the intermediate 12
could be stabilised by resonance, as shown in 15 below. This
process is analogous to the formation of a carbanion in the
reaction of benzil with cyanide anion,14 i.e. the benzoin
condensation.
likely that, in the present study, both the formation of the
adduct and the fission process are significantly more crowded
than the initial state. Formation of the carbonionic intermedi-
ate 12 will be particularly demanding of space and thus strongly
by steric bulk effect of the 3,6-dimethyl groups.
Product studies
A significant difference in the mechanistic Schemes 1 and 2 is
that the formyl group forms on the carbonyl group not suffering
initial attack by hydroxide anion in Scheme 1, whereas, in
Scheme 2, the formyl group forms on the carbonyl group suffer-
ing initial attack. However, the hydroxy groups on 11 could
rapidly equilibriate by a prototropic shift. As such, the product
studies do not allow differentiation between Scheme 1 and 2.
OH
OH
–
O
O
Solvent effects
O
O–
15
The effects of solvent on the rates of base-catalysed fission of
the parent dione are shown in Table 3. The present results can
be compared with the effects of solvent composition on the
base-catalysed rearrangement of benzil6 and the alkaline
hydrolysis of methyl benzoate.22 The rates of reaction increase
markedly with increasing DMSO content which is considered
to arise mainly from the increased activity of hydroxide. DMSO
is also known to stabilise charge delocalised structures. How-
ever, the effect of increasing dioxane content on the rates of
similar reactions has been found to be comparatively minor, but
appears to be discriminating with both increases and decreases
in rates observed.22 The behaviour observed for the ring-fission
of benzocyclobutene-1,2-dione is very similar to that of the
alkaline hydrolysis of methyl benzoate, particularly the modest
decrease in rate with increasing dioxane content. This evidence
appears to favour the occurrence of charge-localised inter-
mediates like 11, rather than a charge-delocalised intermediate
like 12.
Substituent effects
Non-proximate substituent effects in aryl systems can be cor-
related by use of the Hammett equation (1).15 The rates of the
log (k/ko) = ρσ
(1)
base-catalysed fission of the 4-substituted benzocyclobutenes-
1,2-diones can only be successfully correlated using σp as shown
in Table 4. The symmetrical 4,5-disubstituent effect can be cor-
related almost as successfully using 2σp or σm ϩ σp as shown in
Table 4. The three 4-mono-substituted substrates, using σp
values, can be combined with the 4,5-di-substituted substrates,
using either 2σp or σm ϩ σp, to give very good linear relations, as
shown in Table 4. The ρ value obtained is ca. 3.8 at 25 ЊC. Other
correlations have been attempted using a statistical correction
for the rate of 0.5 for the symmetrical di-substituted substrates
or σm for the 4-substituents. These all gave very much poorer
correlations. The use of σp for the mono-substituents and 2σp
for the di-substituents was found in our studies of the reactivity
of similarly substituted indane-1,2,3-triones16 and 2,2-
dihydroxyindane-1,3-diones.1 A possible explanation for this
behaviour has been discussed.16 Furthermore, Jaffé-type treat-
ments17 for the mono-substituted substrates were not justified
because results were only available for three mono-substituted
substrates, as well as for the parent compound. It is clear that
the observed ρ value is significantly greater than observed for
the equilibria addition of hydroxide anion to benzaldehydes at
25 ЊC in water (2.24 or 2.76),18 the base-catalysed rearrange-
ment of symmetrically disubstituted benzils at 30 ЊC in 70%
(v/v) aqueous DMSO (2.8)6 and the alkaline hydrolysis of
ethyl benzoates at 25 ЊC in water (1.33).19
The 3,6-dimethyl substrate reacts very much slower than 4,5-
dimethyl compound. Steric retardation of the process appears
to be the cause. Proximate substituent effects are more difficult
to evaluate quantitatively. The ratio of the rate coefficient for
the 3,6- to that of the 4,5-dimethyl substrate is ca. 2.9 × 10Ϫ3 at
25 ЊC and can be considered to be a measure of the steric effect
alone. This effect is much greater than that of an ortho-methyl
substituent on the alkaline hydrolysis of alkyl benzoates.20 The
ratio of the rate coefficient for the alkaline hydrolysis of the 2-
dimethylaminoethyl 2-methylbenzoate to that benzoate of the
4-methylbenzoate is ca. 0.27 at 37 ЊC in water.21 It seems very
Mechanistic pathway
The evidence indicates that the reaction proceeds by the path-
way shown in Scheme 2. The rate-determining step could be
either the formation of the carbanionic intermediate 12, follow-
ing the pre-equilibrium formation of the adduct 9 and the
oxirane 11, or the formation of the oxirane 11, following the
pre-equilibrium formation of the adduct 9.
Experimental
Materials
Benzocyclobutene-1,2-dione itself was prepared by two
methods. The first was by the method of Cava et al. 23 in which
1,1,2,2-tetrabromobenzocyclobutene was reacted with silver
trifluoroacetate in aqueous acetonitrile. The second was the
general method developed by McOmie’s group.24 Cyclic hydra-
zines, 2,3-dihydrophthalazines, are prepared from phthalic
anhydrides and oxidized in the presence of anthracene to give
the corresponding Diels–Alder adducts. The latter were sub-
mitted to vacuum pyrolysis to give the benzocyclobutene-1,2-
diones. Thus, the unsubstituted, 4-methoxy, 4-chloro, 3,6- and
4,5-dimethyl and 4,5-dichlorobenzocyclobutene-1,2-diones
were prepared.24 The purity of these compounds was monitored
1
by IR and H and 13C NMR spectroscopy, as well as by ele-
mental analysis. Their mps after repeated recrystallization and
a
Table 4 Hammett reaction constants (ρ) for the base-catalysed ring fission of the substituted benzocyclobutenediones in water (µ = 0.1 mol dmϪ3
)
ρ
log ko
r
s
n
4-Substituents with σp
4,5-Di-substituents with (σm ϩ σp)
4,5-Di-substituents with 2σp
4-Substituents and 4,5-di-substituents with σp and (σm ϩ σp)
4-Substituents and 4,5-di-substituents with σp and 2σp
3.937
3.745
4.017
3.675
3.912
2.598
2.363
2.652
2.500
2.501
0.994
0.999
0.994
0.992
0.987
0.32
0.18
0.43
0.24
0.31
4
3
3
6
6
a
r is the correlation coefficient, s the standard deviation and n the number of substituents studied.
J. Chem. Soc., Perkin Trans. 2, 1997
991

View Article Online
drying under reduced pressure (P2O5), were in agreement with
literature values,24 except for the previously unreported 4-
methyl compound which was prepared by the same route.
and extraction with diethyl ether, the acid was esterified with
diazomethane in diethyl ether. The chain (normal) methyl ester
of 2-formylbenzoic acid was isolated and purified by flash distil-
lation at reduced pressure. The same procedure was followed
using 2-formylbenzoic acid itself as substrate. The mass spectra
of the samples and control (using dione and ordinary water)
were recorded on an AEI MS12 spectrometer. The enrichment
was calculated from the peak areas due to Mϩ and Mϩ ϩ 2.
Anthracene adduct of 6-methylphthalazine-1,4-dione
This was prepared by the method of Gould et al. 24 from the
corresponding dione25 and was recrystallised from toluene as
colourless needles, mp 300–303 ЊC (decomp.) (Found: C,
78.2; H, 4.4; N, 7.5. C23H16N2O2 requires: C, 78.4; H, 4.6; N,
7.9%).
References
1 K. Bowden and S. Rumpal, Part 23, J. Chem. Soc., Perkin Trans. 2,
1997, 983.
2 A. H. Schmidt and W. Ried, Synthesis, 1978, 649.
3 M. P. Cava and D. R. Napier, J. Am. Chem. Soc., 1957, 79, 3606.
4 A. Al-Najjar, K. Bowden and M. V. Horri, J. Chem. Soc., Perkin
Trans. 2, 1997, 993.
5 S. Selman and J. F. Eastham, Quart. Rev., 1960, 14, 221.
6 K. Bowden and K. D. Williams, J. Chem. Soc., Perkin Trans. 2, 1994,
77.
4-Methylbenzocyclobutene-1,2-dione
The pyrolysis24 of the anthracene adduct described above gave
the dione which was recrystallised from dichloromethane as yel-
low needles, mp 110–111 ЊC (Found: C, 74.0; H, 4.1. C9H6O2
requires C, 73.8; H, 4.0%).
The solvents for the kinetic studies were purified as described
previously.26
7 J. C. Bradley and T. Durst, Can. J. Chem., 1995, 73, 1660.
8 I. Roberts and H. C. Urey, J. Am. Chem. Soc., 1938, 60, 880.
9 J. Hine and H. W. Haworth, J. Am. Chem. Soc., 1958, 80, 2274.
10 S. L. Johnson, Adv. Phys. Org. Chem., 1967, 5, 237.
11 J. Chandrasekhar, J. G. Andrade and P. v. R. Schleyer, J. Am. Chem.
Soc., 1981, 103, 5612; J. C. Kleingeld and N. M. M. Nibbering,
Tetrahedron, 1984, 40, 2789.
12 K. Bowden, Adv. Phys. Org. Chem., 1993, 28, 171.
13 K. Bowden, F. A. El-Kaissi and N. S. Nadvi, J. Chem. Soc., Perkin
Trans. 2, 1977, 1799.
14 J. P. Kuebrich and R. L. Schowen, J. Am. Chem. Soc., 1971, 93, 1220;
and references therein.
15 C. D. Johnson, The Hammett Equation, Cambridge University
Press, Cambridge, 1973.
16 K. Bowden and S. Rumpal, J. Chem. Res., 1997, 35 (S); 0355 (M ).
17 H. H. Jaffé, J. Am. Chem. Soc., 1954, 76, 4261; A. R. Fersht and
A. J. Kirby, J. Am. Chem. Soc., 1967, 89, 4853.
18 P. Greenzaid, J. Org. Chem., 1973, 38, 3164; W. J. Bover and
P. Zuman, J. Chem. Soc., Perkin Trans. 2, 1973, 786.
19 M. Hojo, M. Utaka and Z. Yoshida, Tetrahedron Lett., 1966, 19.
20 N. B. Chapman, J. Shorter and J. H. P. Utley, J. Chem. Soc., 1963,
1291.
21 J. J. Zimmerman and J. E. Goyan, J. Med. Chem., 1970, 13, 492.
22 K. Bowden and M. J. Price, J. Chem. Soc. (B), 1971, 1748.
23 M. P. Cava, D. R. Napier and R. J. Pohl, J. Am. Chem. Soc., 1963,
85, 2076.
24 K. J. Gould, N. P. Hacker, J. F. W. McOmie and D. H. Perry,
J. Chem. Soc., Perkin Trans. 1, 1980, 1834; O. Abou-Teim, R. B.
Jansen, J. F. W. McOmie and D. H. Perry, J. Chem. Soc., Perkin
Trans. 1, 1980, 1841.
25 E. H. White, D. F. Roswell and O. C. Zafiriou, J. Org. Chem., 1962,
34, 2462.
26 D. D. Perrin and W. L. F. Amarego, Purification of Laboratory
Chemicals, Pergamon Press, Oxford, 3rd edn., 1988.
27 K. Bowden and A. M. Last, J. Chem. Soc., Perkin Trans. 2, 1973,
345.
28 L. Pentz and E. R. Thornton, J. Am. Chem. Soc, 1967, 89, 6931;
P. K. Glasoe and F. A. Long, J. Phys. Chem., 1960, 64, 188;
B. Zerner and M. L. Bender, J. Am. Chem. Soc., 1961, 83, 2267.
29 C. Donati, R. H. Prager and B. Weber, Aust. J. Chem., 1989, 42, 787;
M. D. Gaul, G. A. Junk and H. J. Svec, Environ. Sci. Technol., 1987,
21, 777 (see Chem. Abstr., 1987, 107, 77406x); K. Krohn, H. Rieger,
E. Broser, P. Schiess, S. Chen and T. Strubin, Liebigs Ann. Chem.,
1988, 943; and references therein.
Measurements
Rate coefficients for the base catalysed ring fission of the substi-
tuted benzocyclobutenediones were determined spectrophoto-
metrically by use of a Perkin-Elmer lambda 5 UV–VIS spec-
trometer. A Haake thermostatted water circulatory bath was
used to control the temperature of the cell to ± 0.05 ЊC. The
reactions were followed at the wavelengths shown in Table 1.
The procedure was that described previously.27 Buffers were
prepared from AnalaR grade chemicals and distilled water
(freshly boiled and cooled under N2 bubbling). The ionic
strength was maintained at 0.1 mol dmϪ3 with sodium chloride.
The pH of the buffers solutions was measured with a Pye-
Unicam model PW9409 direct reading pH meter with glass
combination electrode, calibrated with standard buffers at
25 ЊC. Buffer solution pH was measured using a thermostatted
glass vessel at the reaction temperatures. The pD values were
obtained by adding 0.40 to the pH meter readings.28 Both phos-
phate and borate buffers (pH 7.4 to 9.2 at 25 ЊC) were used and
no buffer catalysis was observed.
Product studies
The products of base-catalysed ring fission of the substituted
benzocyclobutenediones were found to be the anions of the
corresponding substituted 2-formylbenzoic acids. For the
parent dione, the product was isolated in quantitative yield and
was confirmed spectrophotometrically by comparison of the
spectrum of the acid in base with that of the reaction product.
In general, the products of the kinetic runs were isolated by
acidification and extraction with diethyl ether. The products
were recrystallised from methanol and had mps in good agree-
ment with literature values.29 The extracts were also treated with
diazomethane in diethyl ether. The products were then isolated
and examined by H and 13C NMR spectroscopy, as well as
1
mass spectrometery and GLC. The structures of the products
of the ring-fission of the mono-substituted benzocyclobutene-
1,2-diones were confirmed or established by their H and 13C
1
NMR spectra, but the quantitative evaluation of the product
ratio for the 4-methyl substrate was completed by GLC.
Hydrolysis using 18O-enriched water
The base-catalysed fission of benzocyclobutene-1,2-dione was
studied by use of 18O-enriched water (4.5 atom%). The reaction
was under normal kinetic conditions in excess of base for
approximately ten ‘half-lives’ of the dione. After neutralisation
Paper 6/06310A
Received 13th September 1996
Accepted 23rd December 1996
992
J. Chem. Soc., Perkin Trans. 2, 1997