LA IMPORTANCIA DE LA AUTO IDENTIFICACIÓN EN EL PROCESO EXPRESIVO.

2.4.3

Results and Discussion - Optimization

The optimization process was initially directed to the formation of butanolide 2-72a

access to butanolide 2-72a in an 82% yield upon heating in 2-methoxyethanol in the

presence of a slight excess of ammonium chloride (Table 2-2 entry 1). Although the use of ammonium chloride gave a high yield of 2-72a, it also led to a trace amount of

inseparable butanolide 2-73a, a product of a dealkoxycarbonylation. In hopes of

preventing the formation of 2-73a, lower temperatures were evaluated; however, no

reaction was observed (entry 2). While changing the solvent to 2-MeO(CH2)2OH, polar DMSO, and DMF resulted is slightly higher yields, inseparable 2-73a was still formed

(entry 3, 4, and 6). In the absence of the salt, the reaction failed to proceed (entry 5). Less polar and non-polar solvent proved to be ineffective, recovering only starting material from the reaction (entry 7 and 8). Next, different additives were investigated. While a variety of salts promoted the reaction (entry 9-12), ammonium chloride salts seemed to be superior. It was never possible to obtain 2-72a as the sole product in our hands.

Table 2-2: Optimization for Butanolide 2-72a

entry  additive 

(1.4 equiv)  solvent  temp (°C) time (h) product (%) 

1  NH4Cl  2‐MeO(CH2)2OH/H2O  reflux  2  82% 2‐72a, trace 2‐73a  2  NH4Cl  2‐MeO(CH2)2OH/H2O  rt  24  no rxn 

3  NH4Cl  2‐MeO(CH2)2OH  reflux  2  84% 2‐72a, trace 2‐73b 

4  NH4Cl  DMSO  135  1  87% 2‐72a, trace 2‐72b 

5    DMSO  135  16  no rxn 

6  NH4Cl  DMF  135  2  mixturea 

7  NH4Cl  Cl(CH2)2Cl  reflux  16  no rxn  8  NH4Cl  Toluene  reflux  16  no rxn 

9  NaCl  DMSO  135 1  mixturea

10  KCl  DMSO  135 1  mixturea

11  LiCl  DMSO  135 24  mixturea

12  Me3N.HCl  DMSO  135 24  mixturea

a_{1:1 mixture of compounds 2‐72a and 2‐73a} 

Frustrated with the inability to form 2-72a cleanly, it was decided to focus on pushing the

by using DMSO and H2O, a solvent mixture commonly used in Krapcho

dealkoxycarbonylation reactions. The use of ammonium chloride in DMSO (entry 2) yielded both 6 and 7 as an inseparable 1:1 mixture. The reaction temperature was increased to reflux in DMSO/H2O; however, this led to slow decomposition of the starting material (entry 3). We next turned our attention to the use of sodium cyanide as an additive due to its common use in Krapcho dealkoxycarbonylation reactions.

Unfortunately, when sodium cyanide was employed, no reaction occurred (entry 4). It was at this point that a one-pot, two-step process was engaged, using ammonium chloride to promote initial butanolide formation followed by the addition of sodium cyanide to facilitate the dealkoxycarbonylation (entry 5). We were pleased to find that the two step process worked, giving 2-73a in a 65% yield as the sole product. Interestingly, when

ammonium chloride and sodium cyanide were used in a non-sequential fashion, only slow conversion to a mixture of 2-72a and 2-73a was observed (entry 6).

Spurred by this success (entry 5), we next examined the use of standard

dealkoxycarbonylation salt systems, which could also promote the butanolide formation. The use of lithium chloride and trimethyl ammonium chloride together at 135 oC and reflux both unfortunately led to a 1:1 mixture of 2-72a and 2-73a even after extended

reaction times (entry 7 and 8). In order to circumvent the problem with incomplete conversion, microwave irradiation was employed. Gratifyingly, the reaction proceeded well at 150 oC in both DMSO and DMF giving rise to adduct 2-73a in excellent yields at

71% and 82% respectively (entry 9 and 10). Finally, the cyclopropanediester was subjected to the optimized reaction conditions of lithium chloride, and trimethyl ammonium chloride in DMF at 150 oC. However, only a small amount of the desired product could be isolated, along with a significant amount of starting material

Table 2-3: Optimization for Butanolide 2-73a

entry  additive 

(1.4 equiv)  solvent 

temp 

(°C)  time (h)  product (%) 

1  NH4Cl  DMSO  135  1  87% 2‐72a, trace 2‐73a  2  NH4Cl  DMSO/H2O  135  1  mixturea  3  NH4Cl  DMSO/H2O  reflux  3  Slow decomp. 

4  NaCN  DMSO  135  24  no rxn 

5  NH4Cl then NaCN  DMSO  135  1/6  65% 2‐73a  6b  NH4Cl/NaCN  DMSO  135  24  mixturea 

7  LiCl/ Me3N.HCl  DMSO  135  24  mixturea  8  LiCl/ Me3N.HCl  DMSO  reflux  24  mixturea  9c  LiCl/ Me3N.HCl  DMSO  150  40  71% 2‐73a  10c  LiCl/ Me3N.HCl  DMF  150  40  82% 2‐73a  11c,d  LiCl/ Me3N.HCl  DMF  150  40  45% 2‐73a  a_{1:1 mixture of compounds 2‐72a and 2‐73a.} b_{reaction conditions gave a low yield of the}  mixture. cperformed in microwave reactor. dthe corresponding methyl diester was used. 

2.4.4 Substrate Scope

With successful reaction conditions determined and a variety of cyclopropane hemimalonates readily available, we set forth to determine the scope of this

transformation (Table 2-4). Both electron-donating and halogen substituted phenyl cyclopropanes effectively underwent the butanolide conversion in moderate to excellent yields (adducts 2-73b-e). In contrast, electron-withdrawing phenyl cyclopropanes

decreased the reactivity of butanolide production, resulting in lower isolated yields (adducts 2-73f-g), a trend common to donor-acceptor cyclopropane reactivity. The

heteroaromatic substituted cyclopropanes (3-N-tosylindolyl and 2-thienyl) underwent the transformation with great success leading to isolated yields of 85% and 74% of

butanolides 2-73h, and 2-73i, respectively. Alkenyl substituted cyclopropanes were able

to withstand the reaction conditions allowing access to the -styrenyl substituted adduct

2-73j in 80% yield. Interestingly, when the vinyl cyclopropane hemimalonate was

lower yield of 2-73k can be attributed to the highly reactive nature of this cyclopropane

toward polymerization. Finally, alkyl substituted cyclopropanes were subjected to the reaction conditions (not in table); however, no product formation was achieved.

Table 2-4: Butanolide Substrate Scope

2.4.5 Reaction Mechanism

To shed light onto the mechanism, optically enriched phenyl cyclopropane (-)-2-55a was

subjected to the reaction conditions (Scheme 2-23). Smooth transformation led to an isolated 82% yield of enriched butanolide (-)-2-73a, with only slight erosion of

enantiomeric excess (determined by a Mosher’s ester sequence). Optical rotation analyses of the product support the (S) isomer butanolide being isolated (when compared to the literature compound).37This outcome suggests that the reaction occurs with retention of

stereochemistry, a result unusual in donor-acceptor cyclopropane chemistry. A proposed mechanism for this transformation can be seen in Scheme 24.