Rationalisation of Patterns of Competing Reactivity by X-ray Structure Determination: Reaction of Isomeric (Benzyloxythienyl)oxazolines with a Base

Three isomeric (benzyloxythienyl)oxazolines 9, 11 and 13 have been prepared and are found, upon treatment with a strong base, to undergo either Wittig rearrangement or intramolecular attack of the benzylic anion on the oxazoline function to give products derived from cleavage of the initially formed 3-aminothienofuran products. This pattern of reactivity is directly linked to the distance between the two reactive groups as determined by X-ray diffraction, with the greatest distance in 11 leading to exclusive Wittig rearrangement, the shortest distance in 13 giving exclusively cyclisation-derived products, and the intermediate distance in 9 leading to both processes being observed. The corresponding N-butyl amides were also obtained in two cases and one of these undergoes efficient Wittig rearrangement leading to a thieno[2,3-c]pyrrolone product.


Introduction
Some time ago we described the reaction of 2-(2-benzyloxyphenyl)oxazoline 1 with a strong base to give either the 3-aminobenzofuran product 2 resulting from intramolecular nucleophilic ring-opening of the oxazoline by the benzyl anion, or the oxazoline 3 in which the benzyloxy group has undergone a Wittig rearrangement (Scheme 1) [1]. Heterocycle formation by cyclisation of an aryloxy carbanion onto an ortho functional group is rather uncommon but formation of 3-aminobenzofurans by the so-called Gewald reaction of benzonitriles provides one example [2]. Similarly, although the Wittig rearrangement has been known for almost a century [3], it is not commonly used in synthesis and a recent review shows rather limited developments over the last 20 years [4]. While the aminobenzofuran formation could be optimised by using 3.3 equiv. of Schlosser's base (n-BuLi/t-BuOK) and applied to a number of substituted examples [1], the Wittig rearrangement process was not so favourable and, under optimal conditions of 2.2 equiv. butyllithium (n-BuLi) in THF, an isolated yield of just 29% was obtained. As will shortly be reported elsewhere, the N-butyl amide group is a more effective promoter of the Wittig rearrangement and treating compound 4 with 3.3 equiv. n-BuLi in THF gives almost entirely the rearranged product 5, conveniently isolated as the phthalide 6 after acid-mediated cyclisation in 90% yield. However in the latter study, simply changing to the N,N-diisopropyl amide 7 and treating with 2.2 equiv. n-BuLi in toluene again resulted in cyclisation to give the 3-aminobenzofuran 8. It is clear from these studies that there is a delicate balance between Wittig rearrangement of the benzyloxy group without affecting the adjacent activating group, and interaction of the two groups with the formation of a furan ring.
In contrast to the symmetrical benzene ring, the different adjacent positions on a heterocycle such as thiophene are not equivalent and so a more interesting pattern of reactivity can be expected, which would also lead to some unusual and novel heterocyclic products. In this paper, we report the synthesis, characterisation and reactivity upon treatment with a strong base, of the three isomeric (benzyloxythienyl)oxazolines 9, 11 and 13 (Scheme 2) as well as the corresponding (benzyloxythienyl)-N-butylcarboxamides 10, 12 and 14. As well as examining the pattern of reactivity we were interested to discover whether there was any correlation between this and the distance between the two adjacent groups as determined by X-ray diffraction. Scheme 2. The six isomeric thiophene compounds targeted for reactivity studies.

Synthesis of 3-Benzyloxy-2-thienyl Systems 9 and 10
O-Benzylation of the commercially available methyl ester 15 followed by ester hydrolysis of 16 gave the carboxylic acid 17 (Scheme 3). This was readily converted into the corresponding acid chloride which was reacted immediately with 2-amino-2-methylpropan-1-ol to give the hydroxy amide 18, which was cyclised using thionyl chloride to give the target oxazoline 9 in good overall yield. Alternatively, treating acid 17 with thionyl chloride followed by an excess of butylamine gave the target amide 10 also in high yield.

Scheme 1. Competition between Wittig rearrangement and cyclisation in benzene-based systems.
In contrast to the symmetrical benzene ring, the different adjacent positions on a heterocycle such as thiophene are not equivalent and so a more interesting pattern of reactivity can be expected, which would also lead to some unusual and novel heterocyclic products. In this paper, we report the synthesis, characterisation and reactivity upon treatment with a strong base, of the three isomeric (benzyloxythienyl)oxazolines 9, 11 and 13 (Scheme 2) as well as the corresponding (benzyloxythienyl)-N-butylcarboxamides 10, 12 and 14. As well as examining the pattern of reactivity we were interested to discover whether there was any correlation between this and the distance between the two adjacent groups as determined by X-ray diffraction. In contrast to the symmetrical benzene ring, the different adjacent positions on a heterocycle such as thiophene are not equivalent and so a more interesting pattern of reactivity can be expected, which would also lead to some unusual and novel heterocyclic products. In this paper, we report the synthesis, characterisation and reactivity upon treatment with a strong base, of the three isomeric (benzyloxythienyl)oxazolines 9, 11 and 13 (Scheme 2) as well as the corresponding (benzyloxythienyl)-N-butylcarboxamides 10, 12 and 14. As well as examining the pattern of reactivity we were interested to discover whether there was any correlation between this and the distance between the two adjacent groups as determined by X-ray diffraction. Scheme 2. The six isomeric thiophene compounds targeted for reactivity studies.

Synthesis of 3-Benzyloxy-2-thienyl Systems 9 and 10
O-Benzylation of the commercially available methyl ester 15 followed by ester hydrolysis of 16 gave the carboxylic acid 17 (Scheme 3). This was readily converted into the corresponding acid chloride which was reacted immediately with 2-amino-2-methylpropan-1-ol to give the hydroxy amide 18, which was cyclised using thionyl chloride to give the target oxazoline 9 in good overall yield. Alternatively, treating acid 17 with thionyl chloride followed by an excess of butylamine gave the target amide 10 also in high yield. Scheme 2. The six isomeric thiophene compounds targeted for reactivity studies.

Synthesis of 3-Benzyloxy-2-thienyl Systems 9 and 10
O-Benzylation of the commercially available methyl ester 15 followed by ester hydrolysis of 16 gave the carboxylic acid 17 (Scheme 3). This was readily converted into the corresponding acid chloride which was reacted immediately with 2-amino-2-methylpropan-1-ol to give the hydroxy amide 18, which was cyclised using thionyl chloride to give the target oxazoline 9 in good overall yield. Alternatively, treating acid 17 with thionyl chloride followed by an excess of butylamine gave the target amide 10 also in high yield.

Reaction 3-Benzyloxy-2-thienyl Systems 9 and 10 with Base
When compound 9 was subjected to the same conditions used to convert 1 into 2, a mixture of two products was formed which were separated by preparative thin-layer chromatography (TLC) on alumina and identified as the expected Wittig rearrangement product 19, formed in 8% yield, and a second more major product (43%) which was initially thought to be the expected cyclisation product 21 (Scheme 4). However certain features of the spectra were not consistent with this, notably the non-equivalence of the two methyl groups and CH2O hydrogens which suggested the presence of a stereogenic centre. After further evidence from 13 C and 2D NMR studies suggested the presence of a 2alkylidenethiophen-3(2H)-one structure, and the HRMS result showed the presence of an extra oxygen atom as compared to 21, the actual structure was finally confirmed as the unusual morpholine-containing thiophenone 20 by X-ray diffraction. The molecular structure of 20 features two independent molecules in the unit cell in addition to one molecule each of CH2Cl2 and acetone. The two molecules are actually enantiomers and they both take up half chair conformations with the ring oxygen out of the plane, the NH in the plane and one methyl axial and one equatorial (Figure 1). Where they differ is that one has phenyl axial and OH equatorial while for the other it is the opposite way round.

Reaction 3-Benzyloxy-2-thienyl Systems 9 and 10 with Base
When compound 9 was subjected to the same conditions used to convert 1 into 2, a mixture of two products was formed which were separated by preparative thin-layer chromatography (TLC) on alumina and identified as the expected Wittig rearrangement product 19, formed in 8% yield, and a second more major product (43%) which was initially thought to be the expected cyclisation product 21 (Scheme 4). However certain features of the spectra were not consistent with this, notably the non-equivalence of the two methyl groups and CH 2 O hydrogens which suggested the presence of a stereogenic centre. After further evidence from 13 C and 2D NMR studies suggested the presence of a 2-alkylidenethiophen-3(2H)-one structure, and the HRMS result showed the presence of an extra oxygen atom as compared to 21, the actual structure was finally confirmed as the unusual morpholine-containing thiophenone 20 by X-ray diffraction.

Reaction 3-Benzyloxy-2-thienyl Systems 9 and 10 with Base
When compound 9 was subjected to the same conditions used to convert 1 into 2, a mixture of two products was formed which were separated by preparative thin-layer chromatography (TLC) on alumina and identified as the expected Wittig rearrangement product 19, formed in 8% yield, and a second more major product (43%) which was initially thought to be the expected cyclisation product 21 (Scheme 4). However certain features of the spectra were not consistent with this, notably the non-equivalence of the two methyl groups and CH2O hydrogens which suggested the presence of a stereogenic centre. After further evidence from 13 C and 2D NMR studies suggested the presence of a 2alkylidenethiophen-3(2H)-one structure, and the HRMS result showed the presence of an extra oxygen atom as compared to 21, the actual structure was finally confirmed as the unusual morpholine-containing thiophenone 20 by X-ray diffraction. The molecular structure of 20 features two independent molecules in the unit cell in addition to one molecule each of CH2Cl2 and acetone. The two molecules are actually enantiomers and they both take up half chair conformations with the ring oxygen out of the plane, the NH in the plane and one methyl axial and one equatorial (Figure 1). Where they differ is that one has phenyl axial and OH equatorial while for the other it is the opposite way round. The molecular structure of 20 features two independent molecules in the unit cell in addition to one molecule each of CH 2 Cl 2 and acetone. The two molecules are actually enantiomers and they both take up half chair conformations with the ring oxygen out of the plane, the NH in the plane and one methyl axial and one equatorial (Figure 1). Where they differ is that one has phenyl axial and OH equatorial while for the other it is the opposite way round.
In the crystal, there is hydrogen bonding both intramolecularly between the NH and C=O and intermolecularly between C=O and OH; in terms of the Etter-Bernstein graph-set descriptors [5] C 1 1 (7) [S (6)]. The intermolecular interaction involves the two enantiomeric molecules alternating and the pattern is shown schematically in Figure 2 with parameters in Table 1. In the crystal, there is hydrogen bonding both intramolecularly between the NH and C=O and intermolecularly between C=O and OH; in terms of the Etter-Bernstein graphset descriptors [5] C 1 1(7) [S (6)]. The intermolecular interaction involves the two enantiomeric molecules alternating and the pattern is shown schematically in Figure 2 with parameters in Table 1.  Once the structure of 20 was clear, its formation could be rationalised as shown in Scheme 4 by initial cyclisation of 9 to give the thieno[3,2-b]furan product 21 and hydrolysis of this on the alumina with opening of the furan ring to give 22, which in its thiophen-3one tautomeric form can be oxidised by air to form the favourable fully conjugated enedione structure 24, which then cyclises to form the cyclic hemiketal 20. It should be noted that compound 20 was found to be quite unstable and although it was isolated in small amount with sufficient purity for identification, further attempts at purification resulted in decomposition. As described in a recent review [6], thiophene-based o-quinomethane analogues have a rich and varied chemistry, however, the formation of such a structure by hydrolysis then oxidation of a thieno[3,2-b]furan is unprecedented.
We now turned to the N-butyl amide 10 and found a much simpler pattern of reactivity. Treatment of 10 with n-BuLi under the standard conditions developed in the benzene series, resulted in exclusive Wittig rearrangement to give secondary alcohol 25 which could be characterised spectroscopically but was converted for isolation into the stable thieno [2,3-c]pyrrolone product 26 by treatment with p-toluenesulfonic acid in boiling toluene (Scheme 5) [7]. This method was also used to obtain stable cyclic products in the benzene-based systems, however, note that while 5 and analogues cyclise to lactones 6 with loss of butylamine, here we have a loss of water to form the lactam in excellent yield. The synthesis and chemistry of thieno[c]pyrrolones and their dihydro analogues has been recently reviewed [8].  In the crystal, there is hydrogen bonding both intramolecularly between the NH and C=O and intermolecularly between C=O and OH; in terms of the Etter-Bernstein graphset descriptors [5] C 1 1(7) [S (6)]. The intermolecular interaction involves the two enantiomeric molecules alternating and the pattern is shown schematically in Figure 2 with parameters in Table 1.  Once the structure of 20 was clear, its formation could be rationalised as shown in Scheme 4 by initial cyclisation of 9 to give the thieno[3,2-b]furan product 21 and hydrolysis of this on the alumina with opening of the furan ring to give 22, which in its thiophen-3one tautomeric form can be oxidised by air to form the favourable fully conjugated enedione structure 24, which then cyclises to form the cyclic hemiketal 20. It should be noted that compound 20 was found to be quite unstable and although it was isolated in small amount with sufficient purity for identification, further attempts at purification resulted in decomposition. As described in a recent review [6], thiophene-based o-quinomethane analogues have a rich and varied chemistry, however, the formation of such a structure by hydrolysis then oxidation of a thieno [3,2-b]furan is unprecedented.
We now turned to the N-butyl amide 10 and found a much simpler pattern of reactivity. Treatment of 10 with n-BuLi under the standard conditions developed in the benzene series, resulted in exclusive Wittig rearrangement to give secondary alcohol 25 which could be characterised spectroscopically but was converted for isolation into the stable thieno[2,3-c]pyrrolone product 26 by treatment with p-toluenesulfonic acid in boiling toluene (Scheme 5) [7]. This method was also used to obtain stable cyclic products in the benzene-based systems, however, note that while 5 and analogues cyclise to lactones 6 with loss of butylamine, here we have a loss of water to form the lactam in excellent yield. The synthesis and chemistry of thieno[c]pyrrolones and their dihydro analogues has been recently reviewed [8].  Once the structure of 20 was clear, its formation could be rationalised as shown in Scheme 4 by initial cyclisation of 9 to give the thieno[3,2-b]furan product 21 and hydrolysis of this on the alumina with opening of the furan ring to give 22, which in its thiophen-3-one tautomeric form can be oxidised by air to form the favourable fully conjugated ene-dione structure 24, which then cyclises to form the cyclic hemiketal 20. It should be noted that compound 20 was found to be quite unstable and although it was isolated in small amount with sufficient purity for identification, further attempts at purification resulted in decomposition. As described in a recent review [6], thiophene-based o-quinomethane analogues have a rich and varied chemistry, however, the formation of such a structure by hydrolysis then oxidation of a thieno [3,2-b]furan is unprecedented.
We now turned to the N-butyl amide 10 and found a much simpler pattern of reactivity. Treatment of 10 with n-BuLi under the standard conditions developed in the benzene series, resulted in exclusive Wittig rearrangement to give secondary alcohol 25 which could be characterised spectroscopically but was converted for isolation into the stable thieno[2,3-c]pyrrolone product 26 by treatment with p-toluenesulfonic acid in boiling toluene (Scheme 5) [7]. This method was also used to obtain stable cyclic products in the benzene-based systems, however, note that while 5 and analogues cyclise to lactones 6 with loss of butylamine, here we have a loss of water to form the lactam in excellent yield. The synthesis and chemistry of thieno[c]pyrrolones and their dihydro analogues has been recently reviewed [8].
Since such fused-ring heterocycles are rather uncommon we took the chance to determine the X-ray structure of compound 26 (Scheme 5). Only one previous X-ray structure of a compound with this ring system appears to have been published, that of the compound with butyl replaced by quinolin-8-yl [9], but the molecular dimensions are very similar. Interestingly the tert-butyl amide 27 isomeric with 25 has been prepared by ortho-directed metalation of N-tert-butylthiophene-2-carboxamide with n-BuLi followed by reaction with benzaldehyde [10]. Since such fused-ring heterocycles are rather uncommon we took the chance to determine the X-ray structure of compound 26 (Scheme 5). Only one previous X-ray structure of a compound with this ring system appears to have been published, that of the compound with butyl replaced by quinolin-8-yl [9], but the molecular dimensions are very similar. Interestingly the tert-butyl amide 27 isomeric with 25 has been prepared by ortho-directed metalation of N-tert-butylthiophene-2-carboxamide with n-BuLi followed by reaction with benzaldehyde [10].

Synthesis of 2-Benzyloxy-3-thienyl Systems 11 and 12
Entry to this system was gained by starting with the 3-thienyloxazoline 28 and introducing oxygen functionality at the 2-position by lithiation and treatment with bis(trimethylsilyl) peroxide (Scheme 6). As we have described in detail elsewhere [11], the resulting product had the 3-(oxazolidin-2-ylidene)thiophen-2-one structure 29 which exhibited an interesting and varied pattern of reactivity. However, for the present purpose, it could be cleanly O-benzylated in moderate yield to give the desired compound 11. Scheme 6. Synthesis of oxazoline 11.
As described below, attempted application of a similar method to the formation of the amide 12 failed since lithiation of the corresponding N-butyl amide 32 followed by treatment with bis(trimethylsilyl) peroxide instead gave the silyl compound 33.

Reaction of 2-Benzyloxy-3-thienyl Systems 11 and 12 with Base
Treatment of oxazoline 11 with n-BuLi under the standard conditions developed for ring closure of 1 to give 2 gave exclusively the Wittig rearrangement product 30 in good yield (Scheme 7). Although attempts to prepare the amide 12 by lithiation and bis(trimethylsilyl) peroxide treatment failed, instead giving the new silane 33 (Scheme 8), the expected Wittig rearrangement product from 12, compound 34, was prepared by lithiation and benzaldehyde treatment of 32. It was not isolated, however, the reaction product was directly Scheme 5. Reaction of 10 with n-BuLi to give 25 and 26 and structure of an analogous product 27.

Synthesis of 2-Benzyloxy-3-thienyl Systems 11 and 12
Entry to this system was gained by starting with the 3-thienyloxazoline 28 and introducing oxygen functionality at the 2-position by lithiation and treatment with bis(trimethylsilyl) peroxide (Scheme 6). As we have described in detail elsewhere [11], the resulting product had the 3-(oxazolidin-2-ylidene)thiophen-2-one structure 29 which exhibited an interesting and varied pattern of reactivity. However, for the present purpose, it could be cleanly O-benzylated in moderate yield to give the desired compound 11. Scheme 5. Reaction of 10 with n-BuLi to give 25 and 26 and structure of an analogous product 27.
Since such fused-ring heterocycles are rather uncommon we took the chance to determine the X-ray structure of compound 26 (Scheme 5). Only one previous X-ray structure of a compound with this ring system appears to have been published, that of the compound with butyl replaced by quinolin-8-yl [9], but the molecular dimensions are very similar. Interestingly the tert-butyl amide 27 isomeric with 25 has been prepared by ortho-directed metalation of N-tert-butylthiophene-2-carboxamide with n-BuLi followed by reaction with benzaldehyde [10].

Synthesis of 2-Benzyloxy-3-thienyl Systems 11 and 12
Entry to this system was gained by starting with the 3-thienyloxazoline 28 and introducing oxygen functionality at the 2-position by lithiation and treatment with bis(trimethylsilyl) peroxide (Scheme 6). As we have described in detail elsewhere [11], the resulting product had the 3-(oxazolidin-2-ylidene)thiophen-2-one structure 29 which exhibited an interesting and varied pattern of reactivity. However, for the present purpose, it could be cleanly O-benzylated in moderate yield to give the desired compound 11. Scheme 6. Synthesis of oxazoline 11.
As described below, attempted application of a similar method to the formation of the amide 12 failed since lithiation of the corresponding N-butyl amide 32 followed by treatment with bis(trimethylsilyl) peroxide instead gave the silyl compound 33.

Reaction of 2-Benzyloxy-3-thienyl Systems 11 and 12 with Base
Treatment of oxazoline 11 with n-BuLi under the standard conditions developed for ring closure of 1 to give 2 gave exclusively the Wittig rearrangement product 30 in good yield (Scheme 7). Although attempts to prepare the amide 12 by lithiation and bis(trimethylsilyl) peroxide treatment failed, instead giving the new silane 33 (Scheme 8), the expected Wittig rearrangement product from 12, compound 34, was prepared by lithiation and benzaldehyde treatment of 32. It was not isolated, however, the reaction product was directly Scheme 6. Synthesis of oxazoline 11.
As described below, attempted application of a similar method to the formation of the amide 12 failed since lithiation of the corresponding N-butyl amide 32 followed by treatment with bis(trimethylsilyl) peroxide instead gave the silyl compound 33.

Reaction of 2-Benzyloxy-3-thienyl Systems 11 and 12 with Base
Treatment of oxazoline 11 with n-BuLi under the standard conditions developed for ring closure of 1 to give 2 gave exclusively the Wittig rearrangement product 30 in good yield (Scheme 7). Scheme 5. Reaction of 10 with n-BuLi to give 25 and 26 and structure of an analogous product 27.
Since such fused-ring heterocycles are rather uncommon we took the chance to determine the X-ray structure of compound 26 (Scheme 5). Only one previous X-ray structure of a compound with this ring system appears to have been published, that of the compound with butyl replaced by quinolin-8-yl [9], but the molecular dimensions are very similar. Interestingly the tert-butyl amide 27 isomeric with 25 has been prepared by ortho-directed metalation of N-tert-butylthiophene-2-carboxamide with n-BuLi followed by reaction with benzaldehyde [10].

Synthesis of 2-Benzyloxy-3-thienyl Systems 11 and 12
Entry to this system was gained by starting with the 3-thienyloxazoline 28 and introducing oxygen functionality at the 2-position by lithiation and treatment with bis(trimethylsilyl) peroxide (Scheme 6). As we have described in detail elsewhere [11], the resulting product had the 3-(oxazolidin-2-ylidene)thiophen-2-one structure 29 which exhibited an interesting and varied pattern of reactivity. However, for the present purpose, it could be cleanly O-benzylated in moderate yield to give the desired compound 11. Scheme 6. Synthesis of oxazoline 11.
As described below, attempted application of a similar method to the formation of the amide 12 failed since lithiation of the corresponding N-butyl amide 32 followed by treatment with bis(trimethylsilyl) peroxide instead gave the silyl compound 33.

Reaction of 2-Benzyloxy-3-thienyl Systems 11 and 12 with Base
Treatment of oxazoline 11 with n-BuLi under the standard conditions developed for ring closure of 1 to give 2 gave exclusively the Wittig rearrangement product 30 in good yield (Scheme 7). Although attempts to prepare the amide 12 by lithiation and bis(trimethylsilyl) peroxide treatment failed, instead giving the new silane 33 (Scheme 8), the expected Wittig rearrangement product from 12, compound 34, was prepared by lithiation and benzaldehyde treatment of 32. It was not isolated, however, the reaction product was directly Although attempts to prepare the amide 12 by lithiation and bis(trimethylsilyl) peroxide treatment failed, instead giving the new silane 33 (Scheme 8), the expected Wittig rearrangement product from 12, compound 34, was prepared by lithiation and benzaldehyde treatment of 32. It was not isolated, however, the reaction product was directly treated with p-toluenesulfonic acid giving the thieno[2,3-c]pyrrolone [8] product 35 isomeric with 26 together with a low yield of the oxidation product 36. This last product showed extra signals in the 13 C NMR spectrum due to amide rotamers (Supplementary Materials). treated with p-toluenesulfonic acid giving the thieno[2,3-c]pyrrolone [8] product 35 isomeric with 26 together with a low yield of the oxidation product 36. This last product showed extra signals in the 13 C NMR spectrum due to amide rotamers (Supplementary Materials). Scheme 8. Formation and cyclisation of secondary alcohol 34.

Synthesis of 4-Benzyloxy-3-thienyl Systems 13 and 14
Synthesis of the required compounds in this series was more challenging since suitably substituted thiophene starting materials are not commercially available. Instead, we had to resort to a ring-synthesis of a thiophene with the desired functionality in place. This started from the sulfide-containing diester 37 prepared by conjugate addition of methyl thioglycolate to methyl acrylate [12], which underwent base-induced ring closure [13] to give compound 38 in low yield (Scheme 9). Aromatisation of this was achieved using sulfuryl chloride [14] to give the thiophene ester 39. Conversion of this into the required benzyl ether 41 proved to be more difficult than expected. Simple alkylation using benzyl bromide and either potassium carbonate or sodium hydride resulted in polymerisation and reaction with phenyldiazomethane [15] also failed. Following a literature report that reaction of the 4-acetoxy compound 40 with ethanol and sulfuric acid gave the 4-ethoxy compound [15], this compound was prepared by reaction of 38 with isopropenyl acetate followed by sulfuryl chloride, but the treatment of

Synthesis of 4-Benzyloxy-3-thienyl Systems 13 and 14
Synthesis of the required compounds in this series was more challenging since suitably substituted thiophene starting materials are not commercially available. Instead, we had to resort to a ring-synthesis of a thiophene with the desired functionality in place. This started from the sulfide-containing diester 37 prepared by conjugate addition of methyl thioglycolate to methyl acrylate [12], which underwent base-induced ring closure [13] to give compound 38 in low yield (Scheme 9). Aromatisation of this was achieved using sulfuryl chloride [14] to give the thiophene ester 39. Conversion of this into the required benzyl ether 41 proved to be more difficult than expected. Simple alkylation using benzyl bromide and either potassium carbonate or sodium hydride resulted in polymerisation and reaction with phenyldiazomethane [15] also failed. treated with p-toluenesulfonic acid giving the thieno[2,3-c]pyrrolone [8] product 35 isomeric with 26 together with a low yield of the oxidation product 36. This last product showed extra signals in the 13 C NMR spectrum due to amide rotamers (Supplementary Materials).

Synthesis of 4-Benzyloxy-3-thienyl Systems 13 and 14
Synthesis of the required compounds in this series was more challenging since suitably substituted thiophene starting materials are not commercially available. Instead, we had to resort to a ring-synthesis of a thiophene with the desired functionality in place. This started from the sulfide-containing diester 37 prepared by conjugate addition of methyl thioglycolate to methyl acrylate [12], which underwent base-induced ring closure [13] to give compound 38 in low yield (Scheme 9). Aromatisation of this was achieved using sulfuryl chloride [14] to give the thiophene ester 39. Conversion of this into the required benzyl ether 41 proved to be more difficult than expected. Simple alkylation using benzyl bromide and either potassium carbonate or sodium hydride resulted in polymerisation and reaction with phenyldiazomethane [15] also failed. Following a literature report that reaction of the 4-acetoxy compound 40 with ethanol and sulfuric acid gave the 4-ethoxy compound [15], this compound was prepared by reaction of 38 with isopropenyl acetate followed by sulfuryl chloride, but the treatment of Following a literature report that reaction of the 4-acetoxy compound 40 with ethanol and sulfuric acid gave the 4-ethoxy compound [15], this compound was prepared by reaction of 38 with isopropenyl acetate followed by sulfuryl chloride, but the treatment of this with benzyl alcohol and sulfuric acid again resulted in polymerisation. Access to 41 was finally achieved, albeit in low yield, by resorting to treatment with benzyl bromide in the presence of silver oxide in a process reminiscent of the Purdie-Irvine method for methylation of sugars developed in St Andrews over 100 years ago [16]. With the key intermediate 41 in hand, the remaining synthetic steps proceeded without incident: hydrolysis gave the acid 42 which was converted into its acid chloride and then reacted either with 2-amino-2-methylpropan-1-ol to give amide 43 which was cyclised with thionyl chloride to oxazoline 13, or with butylamine to directly afford the amide 14.

Reaction of 4-Benzyloxy-3-thienyl Systems 13 and 14 with Base
Treatment of oxazoline 13 with 3.3 equiv. of Schlosser's base gave largely unreacted starting material, however, increasing this to 4.4 equiv. did give a reaction and after chromatographic purification, the 4-benzoyloxy-3-thienyl amide 45 was isolated in moderate yield (Scheme 10). This is evidently formed by air oxidation of the expected cyclisation product, the 3-aminothieno[3,4-b]furan 44. As shown in our previous work [1], such ring-fused 3-aminofuran products are susceptible to oxidative ring-cleavage. this with benzyl alcohol and sulfuric acid again resulted in polymerisation. Access to 41 was finally achieved, albeit in low yield, by resorting to treatment with benzyl bromide in the presence of silver oxide in a process reminiscent of the Purdie-Irvine method for methylation of sugars developed in St Andrews over 100 years ago [16]. With the key intermediate 41 in hand, the remaining synthetic steps proceeded without incident: hydrolysis gave the acid 42 which was converted into its acid chloride and then reacted either with 2-amino-2-methylpropan-1-ol to give amide 43 which was cyclised with thionyl chloride to oxazoline 13, or with butylamine to directly afford the amide 14.

Reaction of 4-Benzyloxy-3-thienyl Systems 13 and 14 with Base
Treatment of oxazoline 13 with 3.3 equiv. of Schlosser's base gave largely unreacted starting material, however, increasing this to 4.4 equiv. did give a reaction and after chromatographic purification, the 4-benzoyloxy-3-thienyl amide 45 was isolated in moderate yield (Scheme 10). This is evidently formed by air oxidation of the expected cyclisation product, the 3-aminothieno[3,4-b]furan 44. As shown in our previous work [1], such ringfused 3-aminofuran products are susceptible to oxidative ring-cleavage. The reaction of the corresponding N-butyl amide 14 with n-BuLi under the conditions required for Wittig rearrangement gave largely unreacted starting material and the only new products isolated in low yield after chromatographic purification (Scheme 11) were the 2,3,4-trisubstituted thiophene 46 together with the debenzylated compound 47 which was found to exist in solution as a mixture with the thiophen-3(2H)-one tautomer 47a (see Section 3). It seems likely that the products have resulted from the intermolecular reaction between two carbanions derived from 14 but in view of their very low yield this process was not investigated further. Products 45, 46 and 47 which were isolated in low amounts following one or two stages of chromatography were found to decompose upon attempted further purification. To summarise the reactivity of the isomeric systems, oxazoline 11 underwent exclusive Wittig rearrangement and oxazoline 13 gave products derived from cyclisation, while for 9 Wittig rearrangement was observed as a minor process with the major product derived from cyclisation. The N-butyl amides gave a less complete picture with 10 undergoing exclusive Wittig rearrangement in high yield, 12 not being available for investigation (although its expected Wittig rearrangement product was obtained by other means), and 14 remaining largely unreacted under the conditions. In the case of the three isomeric oxazolines, each compound was obtained as good quality crystals suitable for X-ray diffraction and so it was decided to determine their molecular structures to examine whether there might be a direct link between the distance between the benzyloxy and oxazoline The reaction of the corresponding N-butyl amide 14 with n-BuLi under the conditions required for Wittig rearrangement gave largely unreacted starting material and the only new products isolated in low yield after chromatographic purification (Scheme 11) were the 2,3,4-trisubstituted thiophene 46 together with the debenzylated compound 47 which was found to exist in solution as a mixture with the thiophen-3(2H)-one tautomer 47a (see Section 3). It seems likely that the products have resulted from the intermolecular reaction between two carbanions derived from 14 but in view of their very low yield this process was not investigated further. Products 45, 46 and 47 which were isolated in low amounts following one or two stages of chromatography were found to decompose upon attempted further purification. this with benzyl alcohol and sulfuric acid again resulted in polymerisation. Access to 41 was finally achieved, albeit in low yield, by resorting to treatment with benzyl bromide in the presence of silver oxide in a process reminiscent of the Purdie-Irvine method for methylation of sugars developed in St Andrews over 100 years ago [16]. With the key intermediate 41 in hand, the remaining synthetic steps proceeded without incident: hydrolysis gave the acid 42 which was converted into its acid chloride and then reacted either with 2-amino-2-methylpropan-1-ol to give amide 43 which was cyclised with thionyl chloride to oxazoline 13, or with butylamine to directly afford the amide 14.

Reaction of 4-Benzyloxy-3-thienyl Systems 13 and 14 with Base
Treatment of oxazoline 13 with 3.3 equiv. of Schlosser's base gave largely unreacted starting material, however, increasing this to 4.4 equiv. did give a reaction and after chromatographic purification, the 4-benzoyloxy-3-thienyl amide 45 was isolated in moderate yield (Scheme 10). This is evidently formed by air oxidation of the expected cyclisation product, the 3-aminothieno[3,4-b]furan 44. As shown in our previous work [1], such ringfused 3-aminofuran products are susceptible to oxidative ring-cleavage. The reaction of the corresponding N-butyl amide 14 with n-BuLi under the conditions required for Wittig rearrangement gave largely unreacted starting material and the only new products isolated in low yield after chromatographic purification (Scheme 11) were the 2,3,4-trisubstituted thiophene 46 together with the debenzylated compound 47 which was found to exist in solution as a mixture with the thiophen-3(2H)-one tautomer 47a (see Section 3). It seems likely that the products have resulted from the intermolecular reaction between two carbanions derived from 14 but in view of their very low yield this process was not investigated further. Products 45, 46 and 47 which were isolated in low amounts following one or two stages of chromatography were found to decompose upon attempted further purification. To summarise the reactivity of the isomeric systems, oxazoline 11 underwent exclusive Wittig rearrangement and oxazoline 13 gave products derived from cyclisation, while for 9 Wittig rearrangement was observed as a minor process with the major product derived from cyclisation. The N-butyl amides gave a less complete picture with 10 undergoing exclusive Wittig rearrangement in high yield, 12 not being available for investigation (although its expected Wittig rearrangement product was obtained by other means), and 14 remaining largely unreacted under the conditions. In the case of the three isomeric oxazolines, each compound was obtained as good quality crystals suitable for X-ray diffraction and so it was decided to determine their molecular structures to examine whether there might be a direct link between the distance between the benzyloxy and oxazoline Scheme 11. Reaction of amide 14 with strong base.
To summarise the reactivity of the isomeric systems, oxazoline 11 underwent exclusive Wittig rearrangement and oxazoline 13 gave products derived from cyclisation, while for 9 Wittig rearrangement was observed as a minor process with the major product derived from cyclisation. The N-butyl amides gave a less complete picture with 10 undergoing exclusive Wittig rearrangement in high yield, 12 not being available for investigation (although its expected Wittig rearrangement product was obtained by other means), and 14 remaining largely unreacted under the conditions. In the case of the three isomeric oxazolines, each compound was obtained as good quality crystals suitable for X-ray diffraction and so it was decided to determine their molecular structures to examine whether there might be a direct link between the distance between the benzyloxy and oxazoline groups and the observed reactivity. All three compounds gave structures with the monoclinic P2 1 /c space group and these are shown in a similar orientation in Figure 3.
For the intramolecular cyclisation to compete with Wittig rearrangement, the key distance is that between benzyloxy carbanionic carbon and C-2 of the oxazoline. Since the benzyloxy groups have rotated to place this carbon pointing away from the oxazoline in each case, the benzyloxy oxygen is taken as a reference point and it can be seen that the molecular geometry correlates well with the observed reactivity. Thus, for 11, the benzyloxy group is too far away (3.046(1) Å) for cyclisation and we observe exclusively a Wittig rearrangement, for 13 the benzyloxy group is much closer (2.945(1) Å) and only products derived from cyclisation are observed, while in 9 we have an intermediate situation (3.006(3) Å) and mainly cyclisation-derived products are observed but with a little Wittig rearrangement. groups and the observed reactivity. All three compounds gave structures with the monoclinic P21/c space group and these are shown in a similar orientation in Figure 3. For the intramolecular cyclisation to compete with Wittig rearrangement, the key distance is that between benzyloxy carbanionic carbon and C-2 of the oxazoline. Since the benzyloxy groups have rotated to place this carbon pointing away from the oxazoline in each case, the benzyloxy oxygen is taken as a reference point and it can be seen that the molecular geometry correlates well with the observed reactivity. Thus, for 11, the benzyloxy group is too far away (3.046(1) Å) for cyclisation and we observe exclusively a Wittig rearrangement, for 13 the benzyloxy group is much closer (2.945(1) Å) and only products derived from cyclisation are observed, while in 9 we have an intermediate situation (3.006(3) Å) and mainly cyclisation-derived products are observed but with a little Wittig rearrangement.

General Experimental Details
NMR spectra were recorded on solutions in CDCl3 unless otherwise stated using Bruker instruments and chemical shifts are given in ppm to high frequency from Me4Si with coupling constants J in Hz. IR spectra were recorded using the ATR technique on a Shimadzu IRAffinity 1S instrument. The ionisation method used for high-resolution mass spectra is noted in each case. Column chromatography was carried out using silica gel of 40-63 mm particle size and preparative TLC was carried out using 1.0 mm layers of Merck alumina 60G containing 0.5% Woelm fluorescent green indicator on glass plates. Melting points were recorded on a Gallenkamp 50W melting point apparatus or a Reichert hotstage microscope.

General Experimental Details
NMR spectra were recorded on solutions in CDCl 3 unless otherwise stated using Bruker instruments and chemical shifts are given in ppm to high frequency from Me 4 Si with coupling constants J in Hz. IR spectra were recorded using the ATR technique on a Shimadzu IRAffinity 1S instrument. The ionisation method used for high-resolution mass spectra is noted in each case. Column chromatography was carried out using silica gel of 40-63 mm particle size and preparative TLC was carried out using 1.0 mm layers of Merck alumina 60G containing 0.5% Woelm fluorescent green indicator on glass plates. Melting points were recorded on a Gallenkamp 50W melting point apparatus or a Reichert hot-stage microscope.

Methyl 3-(Benzyloxy)thiophene-2-carboxylate 16
A literature procedure [17] was modified as follows: benzyl bromide (11.9 cm 3 , 17.11 g, 0.100 mol) was added to a stirred mixture of methyl 3-hydroxythiophene-2-carboxylate 15 (15.85 g, 0.100 mol) and potassium carbonate (27.60 g, 0.200 mol) in acetone (50 cm 3 ) and the reaction mixture was heated at reflux for 18 h. After cooling to rt, the inorganic salts were removed by filtration and the filtrate was concentrated in vacuo. The residue was dissolved in CH 2 Cl 2 (150 cm 3 ) and washed with water (100 cm 3 ) before being dried and evaporated. The crude residue was recrystallised (aq. MeOH) to give 16 (18. and 51.6 (CH 3 ). The 1 H NMR spectral data were in accordance with those previously reported [17]. 13 C NMR data are reported for the first time.

Silver(I) Oxide
A solution of sodium hydroxide (14.61 g, 0.365 mol) in water (440 cm 3 ) was added dropwise to a stirred solution of silver nitrate (60.00 g, 0.353 mol) in water (110 cm 3 ). Once the addition was complete, the precipitate was collected by filtration and washed with water until the washings were neutral before being dried in vacuo to give the title compound (39.60 g, 97%) as a brown solid which was stored in the dark and used without further purification.

Conclusions
The three isomeric thienyloxazolines showed a varied and interesting pattern of reactivity with two of them undergoing Wittig rearrangement and two giving products derived from ring cleavage of an intermediate 3-aminothienofuran. One of the corresponding thienyl amides also underwent Wittig rearrangement and cyclisation of the product, as well as an isomeric one obtained by other means, gave two isomeric thienopyrrolones. The pattern of reactivity in the oxazoline series correlates well with the molecular geometry in the solid state as determined by X-ray diffraction, and the use of this method to