Untersuchungen über aromatische Amino‐<i>Claisen</i>‐Umlagerungen

Abstract

Investigations on Aromatic Amino‐ Claisen Rearrangements The thermal and acid catalysed rearrangement of p ‐substituted N ‐(1′,1′‐dimethylallyl)anilines ( p ‐substituent=H (5) , CH 3 (6) , iso‐C 3 H 7 (7) , Cl (8) , OCH 3 (9) , CN (10) ), of N ‐(1′,1′‐dimethylallyl)‐2,6‐dimethylaniline (11) , of o ‐substituted N ‐(1′‐methylallyl)anilines ( o ‐substituent=H (12) , CH 3 (13) , t ‐C 4 H 9 (14) , of (E) ‐ and (Z)‐N ‐(2′‐butenyl)aniline ( (E) ‐ and (Z) ‐ 16 ), of N ‐(3′‐methyl‐2′‐butenylaniline (17) and of N ‐allyl‐ (1) and N ‐allyl‐ N ‐methylaniline (15) was investigated (cf. Scheme 3). The thermal transformations were normally conducted in 3‐methyl‐2‐butanol (MBO), the acid catalysed rearrangements in 2 N ‐0,1 N sulfuric acid. ‐ Thermal rearrangements. The N ‐(1′,1′‐dimethylallyl)anilines rearrange in MBO at 200‐260° with the exception of the p ‐cyano compound 10 in a clean reaction to give the corresponding 2‐(3′‐methyl‐2′‐butenyl)anilines 22–26 (Table 2 and 3). The amount of splitting into the anilines is &lt;4% ( 10 gives ≃ 40% splitting). The secondary kinetic deuterium isotope effect (SKIDI) of the rearrangement of 5 and its 2′,3′,3′‐d 3 ‐isomer 5 amounts to 0.89±0.09 at 260° (Table 4). This indicates that the partial formation of the new s̀‐bond C(2), C(3′) occurs already in the transition state, as is known from other established [3,3]‐sigmatropic rearrangements. The rearrangement of the N ‐(1′‐methylallyl)anilines 12–14 in MBO takes place at 290–310° to give (E)/(Z) ‐mixtures of the corresponding 2‐(2′‐Butenyl)anilines ( (E) ‐ and (Z) ‐ 30,‐31 , and ‐32 ) besides the parent anilines (5–23%). Since a dependence is observed between the (E)/(Z) ‐ratio and the bulkiness of the o ‐substituent (H: (E) ‐ 30 / (Z) ‐ 30 =4,9; t ‐C 4 H 9 : (E) ‐ 32 / (Z) ‐ 32 =35.5; cf. Table 6 ), it can be concluded, that the thermal amino‐ Claisen rearrangement occurs preferentially via a chair‐like transition state (Scheme 22). Methyl substitution at C(3′) in the allyl chain hinders the thermal amino‐ Claisen ‐rearrangement almost completely, since heating of (E) ‐and (Z) ‐ 16 , in MBO at 335° leads to the formation of the expected 2‐(1′‐methyl‐allyl) aniline (33) to an extent of only 12 and 5%, respectively ( Scheme 9). The main reaction (≃60%) represents the splitting into aniline. This is the only observable reaction in the case of 17 . The inversion of the allyl chain in 16 ‐ (E)‐ and (Z)‐ 30 cannot be detected ‐ indicated that 33 is also formed in a [3, 3]‐sigmatropic process. This is also true for the thermal transformation of N‐allyl‐ (1) and N ‐allyl‐ N ‐methylaniline (15) into 2 and 34 , respectively, since the thermal rearrangement of 2′, 3′, 3′‐d 3 ‐ 1 yields 1′, 1′, 2′‐d 3 ‐ 2 exclusively ( Table 8 ). These reaction are accompanied to an appreciable extent by homolysis of the N, C (1′) bond: compound 1 yields up to 40% of aniline and 15 even 60% of N ‐methylaniline ( (Scheme 10 and 11 ). The activation parameters were determined for the thermal rearrangements of 1, 5, 12 and 15 in MBO ( Table 22 ). All rearrangements show little solvent dependence ( Table 5, 7 and 9 ). The observed Δ H ≠ values are in the range of 34‐40 kcal/mol and the Δ S ≠ values very between ‐13 to ‐19 e.u. These values are only compatible with a cyclic six‐membered transition state of little polarity. ‐ Acid catalysed rearrangements. ‐ The rearrangement of the N‐(1′, 1′‐dimethylallyl) anilines 5‐10 occurs in 2N sulfuric acid already at 50‐70° to give te 2‐(3′‐methyl‐2′‐butenyl)anilines 22‐27 accompanied by their hydrated forms, i.e. the 2‐(3′‐hydroxy‐3′‐methylbutyl) anilines 35‐40 ( Tables 10 and 11 ). The latter are no more present when the rearrangement is conducted in 0.1 N sulfuric acid, whilst the rate of rearrangement is practically the same as in 2 N sulfuric acid ( Table 12 ). The acid catalysed rearrangements take place with almost no splitting. The SKIDI of the rearrangement of 5 and 2′, 3′, 3′‐d 3 ‐ 5 is 0.84±0.08 (2 N H 2 SO 4 , 67, 5°, cf. Table 13 ) and thus in accordance with a [3,3]‐sigmatropic process which occurs in the corresponding anilinium ions. Consequently, the rearrangement of a 1:1 mixture of 2′, 3′, 3′‐d 3 ‐ 5 and 3, 5‐d 2 ‐ 5 in 2 N sulfuric acid at 67, 5° occurs without the formation of cross‐products ( Scheme 13 ). In the acid catalysed rearrangement of the N ‐1′‐methylallyl) anilines 12‐14 at 105‐125° in 2 N sulfuric acid the corresponding ( E )‐ and ( Z )‐anilines are the only products formed ( Table 14 and 15 ). Again no splitting is observed. Furthermore, a dependence of the observed ( E )/( Z ) ratio and the bulkiness of the o ‐substituent ( H : ( E )/( Z )‐ 30 = 6.5; t ‐ C 4 H 9 : ( E )‐ 32 /( Z )‐ 32 = 90; cf. Table 15 ) indicates that also in the ammonium‐ Claisen rearrangement a chair‐like transition state is preferentially adopted. In contrast to the thermal rearrangement the acid catalysed transformation in 2 N‐O, 1 N sulfuric acid (150‐170°) of ( E )‐ and ( Z )‐ 16 as well as of 1 and 15 , occurs very cleanly to yield the corresponding 2‐allylated anilines 33, 2 and 34 ( Scheme 15 and 18 ). The amounts of the anilines formed by splitting are &lt;2%. During longer reaction periods hydration of the allyl chain of the products occurs, and in the case of the rearrangement of ( E )‐ and ) Z )‐ 16 the indoline 45 is formed ( Scheme 15 and 18 ). All transformations occur with inversion of the allyl chain. This holds also for the rearrangement of 1 , since 3′, 3′‐d 2 ‐ 1 gives only 1′, 1′‐d 2 ‐ 2 ( Scheme 17 ). The activation parameters were determined for the acid catalysed rearrangement of 1, 5, 12 and 15 in 2 N sulfuric acid ( Table 22 ). The ΔH ≠ values of 27‐30 kcal‐mol and the ΔS ≠ values of +9 to ‐12 e.u. are in agreement w