Carboxylic acid derivatives and their IMDA reaction products A number of carboxylic acid derivatives were synthesised via a 3-component domino

reaction, which was based on the greater reactivity of alcohols with keteneylidenetriphenylphosphorane 1a in comparison to aldehydes:

C C O

X-H acidic compounds 18a react with ylide 1a generating ester ylides, which in turn react with carbonyl compounds 21 via an intermolecular Wittig reaction forming α,β-unsaturated carboxylic compounds 24. Compounds 98 were prepared using this approach by reacting 96, 3-phenylpropanal 97 and ylide 1a^[108]:

X

2-enoic acid-thiophene-2'-yl-methyl ester 98a and (E)-5-Phenyl-pent-2-enethioic acid-furan-2'-yl-methyl ester 98b were obtained in average yields, with 98b proving difficult to purify.

The objective of preparing compounds 98 was to subject them to the reagents and conditions needed to promote an intramolecular Diels-Alder reaction between the α,β-alkene and the diene system of the thiophene or furan ring to generate tricyclic systems 99:

O S

O O

O

IMDA 98

99 S (CF₃SO₃)₃Sc, THF,

120^oC, 12 h, bomb tube

Scheme 60

This reaction is a 'type 1' intramolecular Diels-Alder (IMDA) [4π+2π] cycloaddition.

Diels-Alder reactions^[109] are classified into two types; type 1 and type 2, depending on the connectivity of the diene and dienophile within the reacting molecule. If the dienophile is attached to the diene at position 1, this is a 'type 1' IMDA reaction and the resulting molecule has a fused bicyclic structure:

2

1 IMDA

Fig.17

But if the dienophile is connected to the diene at position 2, a bridged bicyclic system containing a bridgehead double bond i.e. an anti-Bredt alkene, is produced and this is known as a 'type 2' reaction:

1 2

IMDA

Fig.18

'Type 1' IMDA reactions are useful as they generate fused bicyclic systems from essentially acyclic compounds, few other reactions can boast of this transformation.

Many examples of Diels-Alder reactions in the gaseous state have been reported at temperatures of 400 - 500^oC^[110], while cycloadditions in solvents have been successfully carried out in sealed tubes at 170 - 250^oC^[110]. Because cycloaddition products similar to 99 are known to be especially susceptible to thermal cycloreversion^[111], such high temperatures are not suitable. Therefore Lewis acid catalysts were considered, to enhance the efficiency and rate of cycloaddition reactions, with the use of moderate reaction temperatures. Their effect is thought to result from complexation with the electron-withdrawing group of the dienophile.

Two Lewis acid catalysts were tested in a variety of solvents, at a range of temperatures:

Experiment Lewis-acid catalyst

Solvent Temp.

(^oC)

Reaction ?

1 (CF3SO3)3Sc THF 80 X

2 (CF3SO3)3Sc THF 120 ☺

3 (CF3SO3)3Sc CH3CN r.t. X

4 (CF3SO3)3Sc CH3CN reflux X

5 (CF3SO3)3Y THF 150 X

Table 8 Reagents and reaction conditions applied to Scheme 60. Dry solvents were used and reaction times up to 12 h.

Some product was detected using the reagents and conditions in experiment 2 but yields of only 5 – 7% were obtained. Compound 99 was distinguishable from the ¹³C-NMR spectrum; a new signal around 100 ppm was a clear indication of the formation of the furanone ring, while the downfield signal at approx. 170 ppm was assigned to C-4:

Fig.19 ¹³C-NMR spectrum of a CDCl3 solution of 99

IMDA reactions of systems similar to 98 have been reported, with considerable success, using EtAlCl2[112] as a Lewis acid with mild reaction conditions:

O

Bu COCO₂Et₂Et

Bu

SEt O

EtO₂C

CO₂Et Bu

EtS

Bu

EtAlCl₂, 0.1Py, DCM, 0^oC

63%

100 101

Scheme 61

Molecule 100 has an added advantage over 98 because of the substitution at the diene moiety (electron-donating butyl residue) and the dienophile is made more electrophilic by the presence of the β-oxo functionality, making it more reactive in an IMDA reaction.

The furan and thiophene rings of compounds 98 are relatively poor Diels-Alder dienes due to their aromaticity. The electrons on the sulphur atom of the thiophene ring of 98a are part of a sextet of electrons with delocalisation over the ring carbons. This provides the thiophene ring with added aromaticity and is therefore less likely to undergo addition reactions. This effect is less exaggerated in the furan derivative 98b.

EtAlCl2 was tested on system 98a with the reagents and conditions described in Scheme 61, but no product 99 was detected. A range of temperatures from 0 – 65^oC with reaction times of 2 – 12 h were also tested, but to no avail. A catalytic quantity of EtAlCl2 with T = 110 - 180^oC and microwave irradiation was also tried out, but again no product was formed.

The final attempt to drive this IMDA reaction involved the use of salt solutions in non-aqueous solvent. A 5M solution of lithium perchlorate (LiClO4) in diethyl-ether (5M LPDE) has been reported to be one of the most effective solvent mediums for Diels-Alder reactions, showing an impressive acceleration in the rates of reactions, with reduced reaction times and enhanced yields^[113]. The action of LPDE cannot be soley explained by the Lewis acid effect of Li⁺ as this is considerably mitigated by the solvation effects of Et2O, but its role has been attributed^[109] to a combination of the formation of complexes between the LiClO4 and Et2O, polarity effects and internal pressure effects (pressure exerted by the LPDE medium on the diene and dienophile forcing them to react). One of the most impressive applications of this 'miracle medium' is used in the synthesis of cantharidin^[113]:

O O

S

O

O

O

O O

O + S

102 103 104

5M LPDE, ambient T and P, 9.5 h

70%

Scheme 62

These encouraging results prompted the application of similar reaction conditions to the IMDA reaction of 98. Reactions were carried out using 5M LiClO4 in Et2O with 98a in dry THF from 0^oC to reflux temperatures, but none of the desired product was detected.

Compound 98a was oxidised using MCPBA (m-chloroperbenzoic acid) in an attempt to disturb the aromaticity of the thiophene ring and bring about a more effective IMDA reaction, but no Diels-Alder product 99 was detected.

The tether length between the diene and dienophile of 98a may be too short to allow formation of the tricyclic product 99. It has been reported^[110] that bicycles with a tether length of four or five are much easier to prepare in comparison to those with a tether length of three. Acyclic compounds such as 98 form highly strained products with a high reaction activation energy. This energy of activation may not have been reached in these experiments. Steric hindrance by the phenyl group of 98a could also have hindered the IMDA reaction.