Factorial study to assess an ultrasonic methodology for the allylation of 4-chloroaniline

1 Introduction

Substituted N-allylanilines are important building blocks that are used access a diverse range of organic compounds [1], [2], [3], [4], [5]. In particular, N-allylanilines can be a useful starting compound to obtain the corresponding 2-allylaniline through aza-Claisen rearrangement, which have been used by our group to prepare indole and dihydroindole containing molecules for medicinal chemistry purposes [1], [2], [3], [6]. In a previous work [7], we reported the optimization of an ultrasonic-based methodology to prepare the allyl 1-naphthyl ether with high yields in just 1 h of reaction in an inexpensive ultrasonic bath without additional heating. However, in the allylation of anilines, there exists the additional problem of diallylation, which leads to the undesirable N,N-diallylaniline byproduct. This byproduct makes the following steps on the preparation of the final heterocycles much more complicated, because it can also be rearranged as well as cyclized. Considering this, the synthesis of N-allylaniline must be optimized to obtain the maximum desired monoallylated compound and avoid the diallylated product, while generating the maximum yield.

Several reports on the preparation of N-allyl-4-chloroaniline have been published in the literature. In these studies, allyl bromide is the most employed allylating agent, and carbonates are the most frequent base [1], [2], [3], [4], [5], [8], [9], [10]. The yields reported for these reactions were usually good to excellent; however conventional methodologies were always used, and were mainly conducted under heating for several hours. Additional procedures were also published, employing allyl alcohol [11], [12], allyl carbonate [13], and other unusual allylating agents, such as allylsilanes and allyl tosilates [14], [15]. N,N-diallyl-4-chloroaniline is frequently produced in these reactions, excluding some methods that use expensive or unusual reagents/catalysts [11], [16], [17]. To the best of our knowledge, no ultrasonic methodologies have been reported for this preparation.

The ultrasound (US) energy is well known by accelerating chemical reactions [18]. This effect usually takes place through the cavitation phenomenon, that is, the formation, growth, and collapse of micrometrical bubbles formed by the propagation of a pressure wave through a liquid. The collapses generate high temperatures and pressures in the small environment within nanoseconds. However, the process can sometimes be influenced by the simple mass transfer of US waves (sometimes called false sonochemistry). On the one hand, this phenomenon can be used to speed up chemical reactions; save on time, energy, and costs; and achieve an optimized benefit-cost relation. On the other hand, this energetic process can lead to both desirable and undesirable chemical reactions [18]. Therefore, the results obtained with US-assisted reactions may differ from those obtained in the same conditions with conventional synthesis [19].

In this paper, we aimed to synthesize the N-allyl-4-chloroaniline (Figure 1) with the maximum possible yield, while avoiding or even minimizing the synthesis of the diallylated byproduct. For this purpose, we proposed several factorial designs to assess this objective in an US-assisted methodology. In addition, the assessment of a methodology to prepare this compound more rapidly using an inexpensive ultrasonic bath should be considered to make the process more convenient and “green.”

Figure 1:

Scheme of synthesis of N-allyl-4-chloroaniline and the byproduct N,N-diallyl-4-chloroaniline.

2 Materials and methods

3 Results and discussion

The most used method to the N-allylation of anilines is the reaction of allyl bromide and base catalysis. Allyl bromide is a suitable reagent for nucleophilic substitution due the quality of bromine as leaving-group and the stability of allylic cation when the bromide anion leaves [7]. This stability is attributed to the stabilization of the sp² state by the delocalization of allylic π-bond [25]. In this reaction, a base is always used to increase the nucleophilicity of the aniline as well as to quench the HBr produced. It is usually accepted that this reaction occurs through both S_N1 and S_N2 in conventional procedures, and US may influence the mechanism involved [7].

The preparation of N-allyl-4-chloroaniline is usually done using DMF as solvent as well as allyl bromide and carbonates as reactants. Ramachary and Narayama [2] reported its synthesis with 70% yield after 24 h of reaction, using 1.1 equivalent of allyl bromide to avoid diallylation. In a similar fashion, Nicolaou et al. [1] reported the preparation of this compound in 12 h of reaction at room temperature, yielding only 29% of the product. Using a water-ethanol mixture as solvent, Du et al. [8] reported yields of 77% for diallylated product and only 10% of monoallyl aniline in 3.5 h of reaction at 70°C. When no base was employed, the monoallylated product was mainly obtained, but with low yield (32%).

Given that controlling the diallylation reaction is difficult, several reports that attempted to circumvent this issue can be found in the literature. Yang et al. [11] reported a method for selective preparation of monoallyl anilines using palladium-catalyzed allylation with allylic alcohol. However, the yields and product distribution depends on the alcohol amount. By allylating an N-protected aniline with formyl or other protecting groups, it is now possible to obtain the monoallylated product without diallylation [17]. However, at least three steps (protection-allylation-deprotection) are necessary, and in some cases strong bases, such as NaH must be used. Ultrasonic approaches were not found in our data collection conducted in the main databases.

Considering that a US-assisted methodology to access the allylation of anilines could change the distribution of the monoallylated and diallylated products, in addition to speeding up the whole process, we decided to transpose the conventional methodology to the US. Background experiments were done prior to US method to evaluate which factors would affect the yields and diallylation rate in conventional synthetic procedures. Thus, a 2² full factorial study was performed to verify the influence of allyl bromide quantity (1.5 equivalents and 2.0 equivalents) in the reaction and the temperatures (25°C and 60°C). The results presented in Tables 1 and 3 show that that quantity of allyl bromide (and K₂CO₃) did not have an important influence on the yields (total and monoallyl) and neither in the monoallyl ratio (p>0.05). Meanwhile, temperature affected the results in a significant manner. As can be seen in Table 1, when the temperature was raised to 60°C, the yields of monoallylated product decreased around 15.7%, and the monoallyl ratio was lowered by 34.6%. However, the total yield was increased by 23.4%. This observation indicates that the higher the temperature, the greater the likelihood that a diallylated product is generated, especially when more allyl bromide is used in the reaction. Interestingly, when a higher temperature was employed, the conversion of the starting material 4-chloroaniline was almost quantitative, and the ratio of diallylated product was over 50%. Considering the fact that diallylation is an undesired effect in the context of this work, the reactions should be done in lower temperature (i.e. 25°C). As result, the runs applied in the transposition to the US methodology were all done in this temperature, thus avoiding heating during the process and saving energy.

The methodology transposition was evaluated using a 2³ full factorial design. Methodology (conventional or US), reaction time (1.5 or 4 h) and quantity of reagents (7.5 mmol or 10.0 mmol) were selected as variable factors. Tables 2 and 4 and Figure 2 summarize the results. Among the three factors, quantity of reagents showed a less significant effect in the responses. The transposition from conventional to a US-based methodology showed significant effects on all responses, leading to positive influences on the yields. However, the US methodology led to a negative effect (–10.3%) on the monoallyl ratio. The most important influence is exerted by reaction time, which led to positive effects on monoallyl and total yields, but a negative effect on the monoallyl ratio.

Figure 2:

Graphical representation of the effects generated by methodology (conventional to US), reaction time (from 1.5 h to 4 h) and quantity of reagents (7.5 mmol to 10.0 mmol of both allyl bromide and K₂CO₃).

Based on the collected data, US can efficiently speed up the allylation reaction. This characteristic is highly predictable, since we have already reported the acceleration of allylation reaction of 1-naphthol [7]. Although the data in Table 2 suggest that US can decrease the monoallyl ratio, this effect can be explained by the acceleration of the production of N-allyl-4-chloroaniline. As can be seen in Table 2, the diallylation only takes place when the monoallylated proportion in the reaction environment is around 30%. Obviously, the production of the N,N-diallyl-4-chloroaniline is only possible when some amount of N-allyl-4-chloroaniline is present in reactional medium. In other words, the distribution of the products is not really affected by US, but in fact, this is an effect of the acceleration of the reaction rate provided by US. The same behavior can be observed regarding time reaction. As long as the reaction is carried out, the greater the production of monoallyl and diallyl products. Data presented in Table 4 corroborate this statement, given that a significant interaction effect on monoallyl ratio has been identified between methodology used and reaction time. As can be seen in Figure 3, the average monoallyl ratio was above 90% even after 4 h of reaction when conventional methodology was used, and below this value when US methodology was employed in only 1.5 h of reaction. However, when the reaction was extended to 4 h, the monoallyl ratio had an average of 74%. Thus, to obtain the best yield with higher monoallyl ratio, the US reaction must be done in 4 h, in lower temperature (i.e. 25°C), and with the use of 7.5 mmol of allyl bromide.

Figure 3:

Graphical representation of the interaction effect between time reaction (1.5 h to 4 h) and methodology (conventional and US).

As the results of the 4-h US-based reaction showed comparable yields to the 24-h conventional reaction, we performed a statistical comparison between the results of both approaches (Figure 4). We found that there were significant differences in the monoallyl yield, but no differences were observed in total yield and monoallyl ratio. In summary, this US-based methodology, fully conducted in an ultrasonic bath, can provide the same results as the conventional procedure but in much less time.

Figure 4:

Comparison between the conventional 24-h reaction and US-based 4-h reaction.

4 Conclusions

The experimental results demonstrate that the proposed US methodology can significantly accelerate the synthesis of N-allyl-4-chloroaniline, lowering the reaction time from 24 h to only 4 h in the same reaction conditions. Considering the reactions of this work were conducted in an inexpensive ultrasonic bath, without heating and in less time, this study’s findings are an important contribution to the development of a more convenient process in laboratorial scale, which can save time and related costs. This methodology can also help achieve some principles of green chemistry, and is useful in other allylation reactions.