Chemistry Research

ABSTRACT

Triazine, also known as cyanuric chloride, was used to promote the amidation of various carboxylic acids. The carboxylic acids used were all benzoic acids with different groups attached in the ortho-, meta-, and para- positions on the benzene ring. The effect of different attached groups were compared with one another. Cyanuric chloride has been chosen for this transformation because it has recently been found to be a cost- effective reagent compared to traditional amidation processes. Additionally, when an acid chloride cannot be prepared from the corresponding carboxylic acid due to safety, stability, or handling concerns, triazine is an effective alternative.

INTRODUCTION

The use of cyanuric chloride for this research stems back to traditional amidation processes that use acid chlorides to synthesize amides. These amides are usually prepared by coupling an amine with an acid chloride that must first be synthesized from its parent carboxylic acid if it is not readily available from the manufacturer. This introduces many handling and safety concerns. Since carboxylic acid chlorides are the least stable of the carboxylic acid derivatives, they cannot be formed directly from other acid derivatives by simple nucleophilic acyl substitution. The acid chloride is usually derived from an acid derivative such as an ester as part of a multi-step synthesis. The carboxylic acid intermediate must then be filtered and dried or extracted into an effective solvent and dried by an azeotropic distillation. All of these steps add cost to the synthesis because several cycles must be completed. Also, the most common reagents used to prepare an acid chloride from a carboxylic acid include a highly activated sulfurous acid derivative such as thionyl chloride which is very corrosive or phosgene which is a very toxic compound. Even though these reagents are inexpensive, they present hazard problems for small manufacturing plants that do not have the facilities or desire to deal with toxic or very corrosive compounds. Once the acid chlorides are made, they usually present handling and storage issues since they are also very corrosive and react readily with water. Figure 1 shows the mechanism in forming an acid chloride by reacting a carboxylic acid with thionyl chloride.

Figure 1: The reaction of a carboxylic acid with thionyl chloride yields an acid chloride. The reaction proceeds through a sulfur intermediate (a chlorosulfite ester).

Acid anhydrides have also been used instead of acid chlorides to synthesize amides but many problems develop in the procedure. Carboxylic acids that are converted to symmetric anhydrides result in a loss of 1 equivalent of the carboxylic acid as a byproduct of the amidation reaction. The byproduct may be isolated and converted back to the anhydride but this becomes very expensive. Some acid anhydrides are very expensive so mixed anhydrides can be prepared from acetic acid. This presents a problem since it is very unlikely that only one of the acids will react with the amine and the product usually consists of a mixture of amides that are difficult to isolate.

The disadvantages of using acid chlorides and anhydrides have made the amidation of carboxylic acids a very popular area of research. Research groups and individual scientists are striving to find more efficient ways to synthesize amides from carboxylic acids on a large scale. Many reagents have been found that allow the coupling of carboxylic acids and amines but most of them are very expensive and isolating the desired product is very difficult due to byproducts.

The research group of the Research Laboratories at the Rohm and Hass Company in Pennsylvania has developed a process that is efficient for large-scale production of amides using carboxylic acids and cyanuric chloride as a reagent1. The interest of the research group at the Rohm in Hass Company is the preparation of amides which have activity as agricultural chemicals.

The purpose of this project is to deal with the amidation of various benzoic acids with different groups attached in the ortho-, meta-, and para- positions on the benzene ring and analyze the effect of the different groups on triazine-promoted amidation. The general mechanism for the reaction is shown below in Figure 2. The procedure is very simple and effective in comparison with traditional methods and can be done on a large scale. The work up of the product is easy and very little byproducts are produced. The chemicals used in the reaction are also safer than the traditional chemicals used. The procedure is cost effective since only a 0.33 equivalent of cyanuric chloride is needed for the reaction due to the three sites of attachment on the cyanuric chloride compound.

Figure 2: The general mechanism for the triazine-promoted amidation of carboxylic acids used in this project.

EXPERIMENTAL

The general procedure was derived from Organic Process Research and Development¹. One run was done with each of the benzoic acids explored. Structural identification was completed by employing IR and melting point. A percent yield was also completed before and after purification for each of the amides synthesized. The general idea of the project was to explore benzoic acids with different groups in the X, Y, or Z (para-, meta-, or ortho-) position of the benzoic acid in Figure 3 below and observe the differences in reactivity.

Figure 3: A general transformation of a benzoic acid to a benzamide.

General Procedure for the Triazene-Promoted Amidation of Various Carboxylic Acids

The benzoic acid (1 equivalent) was made into a slurry with 20 mL of ethyl acetate in a 50 mL round bottom flask. N-ethyl morpholine (1.02 equivalent) was then added to the slurry followed by cyanuric chloride (0.33 equivalent) and left stirring for approximately 30 minutes. Methylamine (1.02-1.05 equivalent) was then added and the mixture was left stirring for at least an additional 60 minutes (at this point, the reaction mixtures were often left stirring overnight). The solid was then filtered and rinsed with small amounts of ethyl acetate. The filtrate was washed with 20 mL of 1 M NaOH. The bottom organic layer is extracted with distilled water (2 x 20 mL). The combined organic extracts are dried (Na₂SO₄) and evaporated in vacuo. The resulting product was then digested in ether.

Synthesis of N-methylbenzamide from Sodium Benzoate

Sodium benzoate was made into a slurry in acetonitrile. Cyanuric chloride was then added and stirred for 30 minutes. No morpholine was added since the benzoic acid salt was the starting material. Methylamine was then added and the mixture was allowed to stir for another 60 minutes. The solid was filtered off and rinsed with small amounts of acetonitrile. The filtrate was placed in a separatory funnel and 20 mL of 1 M NaOH was added. The bottom layer was extracted and 20 mL of distilled water was added which mixed with the solvent and no separation occurred. The solution was dried with Na₂SO₄ and it was found was found that it contained large amount of water. The solution was evaporated down until very little was left. Methylene chloride was added and the product began to separate out. More was added to dissolve the product. The solution was placed on ice and crystal formed. A recrystallization was done by adding a minimum amount of boiling ethyl acetate to dissolve the product. A watch glass was placed over the solution and allowed to cool. Crystals began to form and solution was placed on ice until it totally crystallized. Product was then filtered and washed with cold ethyl acetate and placed on a clay plate to dry. A white benzamide crystalline product was observed (2.58%); mp 79-89 ^oC; IR (KBr): 3327 (N-H), 3055 (alkene C-H), 2939 & 2901 (alkane C-H), 1638 (C=O), 1578 (C=C) cm^-1. The IR matched the known N-methylbenzamide product.

Synthesis of N-ethylbenzamide from Sodium Benzoate

The procedure was similar to that as described in the general procedure except ethylamine was used in the amidation instead of methylamine. An oily product was isolated and an H-NMR was going to be done but a white precipitate formed when chloroform was added. The white precipitate was isolated and found to be a possible impurity. An oily product was still isolated; IR (NaCl): 3274 (N-H), 3063 (alkene C-H), 2977 & 2935 (alkane C-H), 1755 & 1715 (C=O), 1640 & 1555 (C=C) cm^-1. The benzamide product appeared to be a mixture of starting material and product since there were two C=O peaks.

Table 1: Results of all N-methylbenzamide derivatives synthesized.

DISCUSSION

Several comparisons and conclusions can be made from the results of this project. One comparison that can be made is between the para-, meta-, and ortho- amides synthesized from the chlorobenzoic acid. The amide with the chlorine in the para position gave the best looking crystals of the three. This can be attributed to the structure of the para amide. During recrystallization, the molecules can stack upon one another very nicely to form crystals because the molecules are straight. The meta amide is bent so it cannot stack as easily and has a tendency to trap impurities which explains its very high yield and its paste-like appearance. For the ortho amides, the attack of the amine is sterically hindered due to the position of the chloride substituent. Figure 4 shows the steric hindrance of the amine on its attack of the carbonyl carbon. Thus, there are lower yields for the ortho-substituted amide than for the para amide. The same conclusions can be made for the amides with the bromine in the meta and para positions. The meta amide once again gave a paste and a large yield due to the trapping of the impurities and the para amide was a clean crystalline product.

Figure 4: Backside attack blocked by ortho-substituents.

Another comparison that can be made is between the para-nitro and meta-nitro amides. The amide with the nitro group in the para position gave better yields than the amide with the nitro group in the meta position. The para position allows the molecule to stack better as explained before. Also, when the nitro group is in the para position on the benzene ring, resonance causes a very electron-deficient carbon adjacent to the carbonyl carbon. This increases the reactivity between the electron-rich amine and the electron- deficient carbonyl carbon (see Figure 5). When the nitro group is in the meta position on the benzene ring, resonance caused by the nitro group places electron-deficient carbons further from the carbonyl carbon which makes the meta molecule less reactive than the para molecule. This lowers the yield of the amide product. A comparison is shown in Figure 5.

Two nitro groups in the meta position such as in the 3,5-dinitro amide gave a better yield than one attached nitro group since there is a greater electron-withdrawing effect with two attached nitro groups than one. Therefore, the yield is higher for the 3,5-dinitro amide than for the meta nitro amine. The figure on the left of Figure 5 shows the effect of the para-nitro group on the attack of the amine while the figure on the right shows the effect of the meta-nitro group on the attach of the amine.

Figure 5: Nitro-substituent's electronic effects.

A comparison can also be made between the para-amino amide and the para-hydroxy amide. There were general problems that arose at the beginning of the experiments as the reactants became very clumpy and even a thick syrup in the case of the para-hydroxy reaction. The clumpy reaction at the beginning can be attributed to the increased polarity of both molecules due to the amino and hydroxy groups. These groups also have several side products due to the reactivity of the amino and hydroxy groups. This would explain the small yield of desired amide products observed by small N-H peaks and C=O IR bands. Both molecules can also hydrogen bond which will effect their solubility in the reaction solvent. The low yields in both cases can be explained by the effect of the amino and hydroxy groups in the para positions. Both groups donate electrons through resonance towards the carbonyl carbon sites of attack which makes them less reactive. Figure 6 shows the electron donation of both molecules and their effect on the site of reactivity and the attack of the amine.

The last observation was that of the o-methoxy substituent. The methoxy group is also very electron donating to the site of reactivity as seen with the hydroxy and amino groups. The IR showed a purer product because the methoxy group is not nearly as polar as the hydroxy and amino groups and does not hydrogen bond, thus is easier to purify upon workup.

Overall, the project gave very good results and several good conclusions could be made. A general conclusion is that substituents in the meta position yield poor non-crystalline products, more like a paste. Also, the effect of the certain electron-withdrawing and electron donating groups is altered due to their position on the ring. There is a different effect from the para and meta positions of these groups. One other general conclusion is that molecules that have a linear structure, namely those with para substituents, will yield better looking crystals than bent molecules (ortho or meta substituted).

Figure 6 shows the effect of a hydroxyl group on the left structure and the amino group on the right structure. Both attached functional groups push electrons towards the site of attack making the carbonyl carbon less reactive resulting in low yields.

Figure 6: Hydroxyl and amino group electronic effects.

REFERENCES

1. Rayle, Heather L. and Fellmeth, Lisa; "Development of a Process for Triazine-Promoted Amidation of Carboxylic Acids"; Organic Process Research and Development (May-June 1999) 172-176
2. Fox, Marye Anne and Whitesell, James K; Core Organic Chemistry; Boston: Jones and Bartlett Publishers (1997)
3. "Triazine Process Gives Amides from Acids"; Chemistry and Engineering News (April 19, 1999) 47

Chemistry Home Page

Last modified: February 4, 2000. JG.