The amide group and its preparation methods by acid-amine coupling reactions: an overview

Symmetrical anhydrides

The preparation of symmetrical anhydrides can be carried out by heating the selected acid to dehydrate it using a Dean-Stark system or through dehydrating agents to remove water. Additionally, the acid can be reacted with DCC under mild conditions to yield the anhydride [4].

The primary limitation of these anhydrides is their molecular economy and waste generation, as only half of the equivalents of the acid used effectively couples with the amine. Moreover, if the acid used is excessively expensive, these anhydrides have economic limitations [4, 10].

Mixed anhydrides

As a solution to the drawbacks of using symmetrical anhydrides, mixed anhydrides can be employed where the second carboxylic residue is cheap and easy to couple with the acid. The most common reagents used for obtaining mixed anhydrides are derivatives of pivaloyl chloride (24), isobutyl chloroformate (25), and EEDQ (26), as well as derivatives of acetic anhydride (27) and tert-butoxycarbonyl anhydride (28) (Fig. 9) [1, 4, 10].

Fig. 9:

Reagents used for the preparation of mixed anhydrides.

The advantage of these anhydrides is due to the proper regioselectivity they exhibit in nucleophilic addition to the more reactive position a, which is more electrophilic than position b. To further favor this selectivity, sterically hindered reagents are often used in the formation of the anhydride to improve the regioselectivity of the nucleophilic attack by the amine [1]. Anhydrides derived from pivaloyl chloride (24) and isobutyl chloroformate (25) are noteworthy, as their high selectivity is believed to be a result of the steric hindrance they impose on position b (Fig. 10) [1, 4].

Fig. 10:

Proposed mechanism and regioselectivity of amidation using isobutyl chloroformate (25).

In the literature, there are several examples of the use of in situ formed mixed anhydrides for amidation reactions, as shown in the figures. These anhydrides are formed using reagents 24 and 25 along with Et₃N as a base (Fig. 11) [10, 11].

Fig. 11:

Reaction for amide bond formation via mixed anhydrides.

In the literature, there are examples where the use of chloroformates for the generation of mixed anhydrides results in a two-step synthetic process for amide formation: first, the acid is activated by forming the corresponding anhydride, and then a subsequent step where the nucleophilic attack of the amine occurs. However, when mixed anhydrides are used with EEDQ (26) as the reagent in the presence of the corresponding amine, the coupling is direct, as the anhydride reacts with the amine as soon as it forms [10]. This reagent is cost-effective, commercially available, and any byproducts can be easily removed through aqueous extraction (Fig. 12) [1], [12], [13], [14.

Fig. 12:

Proposed mechanism for the formation of mixed anhydrides using EEDQ (26) as the reagent.

An example of the use of the reagent EEDQ (26) is described for the formation of the amide bond, as shown in Fig. 13 [15].

Fig. 13:

Amidation reaction using EEDQ (26).

Carbodiimides

A very common method for amide bond formation is the use of carbodiimides. These compounds have in their structure two nitrogen atoms with basic character that can promote the reaction between an acid and an amine, leading to the formation of the amide bond. For this, it is necessary to form a more reactive species in situ, an O-acylisourea, which leads to the formation of the desired product in the presence of the required amine (Fig. 21).

Fig. 21:

Mechanism for the formation of amide bonds through O-acylisourea.

There are numerous examples of carbodiimides used in organic synthesis with different substitutions, among which DCC (44) can be highlighted. In reactions using DCC (44) as a coupling agent, DCU is formed as a byproduct, which is difficult to remove by column chromatography. A widely used alternative is EDC (45), which has a significant advantage over DCC (44) because both EDC (45) and its urea derivative have high solubility in aqueous solutions and can be easily eliminated in the workup through aqueous washing. In coupling reactions with amino acids using BDDC (46), moderate yields are obtained, and the formed byproducts can be easily removed through an acidic wash.

The use of N-hydroxylamine derivatives (HOXt) in combination with carbodiimides leads to the formation of esters that, while less reactive than O-acylisoureas [20], allow for good yields, increasing the efficiency of the process by avoiding the possible formation of N-acylureas. This advantage is due to the ability of N-hydroxylamines to protonate O-acylisoureas, thus preventing the intramolecular reaction that would lead to the formation of N-acylureas, shifting the reaction towards the formation of activated esters (Fig. 22). The presence of tertiary amines promotes the formation of activated esters [21, 22]. As an alternative to the use of N-hydroxylamines, this type of reaction can be carried out using a catalytic amount of a tertiary amine, DMAP, which favors the progress of the reaction through the formation of activated amides (Fig. 22) [23].

Fig. 22:

Synthetic routes for carbodiimide-mediated amidation reactions.

Some of the most commonly used N-hydroxylamines, in combination with carbodiimides, as coupling agents in the formation of amide bonds from their precursor acids, are N-hydroxylamines HOSu (NHS) (50), HOBt (5), and HOAt (6), with the latter two having a benzotriazole-like structure. In terms of performance, activated esters derived from HOAt (6) show better yields than those derived from HOBt (5) [24]. This could be because the nitrogen at position 7 of benzotriazole, due to stereochemical effects, allows for the formation of a hydrogen bond with the amine, which facilitates its proper orientation for the subsequent nucleophilic attack leading to amide bond formation (Fig. 23) [4, 10].

Fig. 23:

Intermolecular interaction via hydrogen bonding when using HOAt (6) as a coupling agent.

The benzotriazoles 6-HOAt (47), 5-HOAt (48), and 4-HOAt (49) do not allow for the formation of this intermolecular hydrogen bond, resulting in a loss of process efficiency.

Some examples in which this methodology is used are as follows (Fig. 24) [23, 25, 26].

Fig. 24:

Carbodiimide-mediated amidation reaction.

Phosphonium salts

Alternatively, for amide bond formation, reagents incorporating phosphonium salts in their structure were developed, which prevent secondary reactions and epimerization processes. In this regard, a phosphorus-derived reagent, BOP (7), was optimized, incorporating an N-hydroxybenzotriazolyl subunit into its structure. Despite yielding fast couplings and containing the mentioned functionality to prevent the epimerization of chiral centers in the starting materials, BOP (7) has had limited use in the industry due to the generation of HMPA as a byproduct, a highly carcinogenic compound [1, 4, 5].

As an alternative to avoid the formation of this toxic byproduct, PyBOP (8) was developed. However, this reagent comes with a high cost, preventing its use on a large scale (Fig. 25) [1, 4, 5].

Fig. 25:

Phosphonium salt-derived reagents.

As an example of the preparation of such reagents, Fig. 26 shows the synthesis of the BOP reagent (7). For its preparation, tris(dimethylamino)phosphine (61) is reacted with carbon tetrachloride in the presence of HOBt (5), using THF as a solvent, followed by exchanging the chloride anion with the hexafluorophosphate anion [5].

Fig. 26:

Preparation of BOP (7).

The mechanistic proposal described in the literature for amidation using reagents with oxyphosphonium salt-like structure is depicted in Fig. 27, using BOP (7) as an example of a coupling agent. In the presence of a base, the carboxylate anion generated from the corresponding acid undergoes a nucleophilic attack on the phosphonium cation, favored by the oxophilic nature of phosphorus. This species leads to the activation of the O-benzotriazole ester and the formation of HMPA, the previously mentioned toxic product. Finally, the nucleophilic attack of the corresponding amine on this activated ester takes place to obtain the desired amide [1, 4, 5].

Fig. 27:

Proposed mechanism for amide bond formation using BOP (7) as a coupling agent.

It has been observed that the use of these types of reagents for the preparation of peptides containing N-methyl-α-amino acids in their structure leads to slow and low-yield couplings, with epimerization occurring in some cases. This may be because the activated benzotriazolyl ester is bulky and highly effective in reactions with primary amines. However, it does not react as readily with secondary amines, leading to the degradation of reaction intermediates. Therefore, more efficient derivatives such as PyBroP (63) for peptide couplings [4] and PyCloP (62) have been developed. These derivatives are characterized by being less bulky than their predecessors and therefore have less steric hindrance (Fig. 28), which promotes the amidation reaction [5].

Fig. 28:

Amidation of N-methyl-α-amino acids.

Additionally, phosphonium salt derivatives of HOAt (6) have been developed, such as AOP (64) and PyAOP (65), which are more efficient as coupling reagents due to the potential formation of a hydrogen bond in the reaction intermediate, as mentioned in previous sections (Fig. 29) [5].

Fig. 29:

Phosphonium salt derivatives of HOAt (6).

Figure 30 displays some of the reagents described in the literature, which inhibit racemization in the coupling of peptides containing N-methyl-α-amino acids. For example, PyNOP (66), PyFOP (67), PyFNBOP (68), PyCloK (69), PyPOP (70), PyTOP (71), PyDOP (72), and PyDAOP (73) [5].

Fig. 30:

Reagents used in the amidation reaction of N-methyl-α-amino acids.

Below, some examples using reagents 7 and 8 in amidation reactions are presented (Fig. 31) [27, 28].

Fig. 31:

Amidation reaction using phosphonium salt-derived reagents.

Guanidinium and uronium salts

In a manner similar to phosphorus-derived benzotriazolic reagents, other reagents incorporating guanidinium and uronium cations have been developed for use as coupling agents in amide bond formation. As an alternative to the BOP reagent (7), HBTU (9) has been developed. Since then, numerous guanidinium and uronium salts with a benzotriazole-like structure have been reported for use as coupling agents in amidation reactions [1].

In Fig. 32, examples of these coupling agents for amide bond formation are shown.

Fig. 32:

Reagents for amide bond formation derived from guanidinium and uronium salts.

For the preparation of these benzotriazole reagents, the synthesis of HBTU (9) is shown as an example in Fig. 33. To achieve this, tetramethylurea (78) is reacted with (COCl₂)₂ (14) in toluene to obtain the corresponding chlorotetramethylguanidinium salt, followed by an anion exchange with potassium hexafluorophosphate. Next, this species undergoes nucleophilic attack by HOBt (5) using Et₃N as a base, resulting in the formation of HBTU (9) [5, 29].

Fig. 33:

Preparation conditions of HBTU (9).

Among the main differences within these reagents, we can highlight: (a) the cation incorporated into their structure, some contain the guanidinium cation, while others contain the uronium cation (Fig. 34) and (b) the substitution of the aromatic ring; some have a nitrogen atom in the ring, while others do not, which imparts differences in reactivity. These structures have been determined both by NMR in solution and by X-ray crystallography in the solid state, both for HBTU (9) and HATU (10) [30].

Fig. 34:

Difference between uronium and guanidinium cations.

It has been demonstrated that uronium salts are more efficient as coupling agents for amide bond formation than their guanidinium analogs. This is because the activation of the acid through uronium salts (which involves breaking the C–O bond) is faster than through guanidinium salt species (which involves breaking the C–N bond). For example, during a study of peptide cyclization, this reactivity difference was observed. When using N-HBTU as the reagent, more than 91 % of the peptide had not reacted after 15 s. However, when using O-HBTU, more than half of the peptide had disappeared [30, 31].

The mechanism of amide bond formation using these coupling agents is equivalent to that described for phosphonium salts (Fig. 27). In Fig. 35, the proposed mechanism for the amidation reaction using HBTU (9) as the reagent is shown. In this mechanism, the acid, in the presence of a tertiary base, forms the carboxylate anion, which undergoes a nucleophilic attack on the electrophilic center of the guanidinium moiety. This leads to the activation of a benzotriazolyl ester, which subsequently undergoes nucleophilic attack by the corresponding amine, resulting in the desired amide. As byproducts of this reaction, HOBt (5) and tetramethylurea are obtained [1].

Fig. 35:

Proposed mechanism for amidation using HBTU (9).

The use of tertiary bases such as Et₃N, ⁱ Pr₂EtN, or collidine is necessary because these compounds act as bases and not as nucleophiles. In this way, these bases deprotonate the carboxylic acid, and their steric hindrance shields the nitrogen atom, enhancing their basicity.

Furthermore, in amidation reactions using this methodology, in addition to the standard reaction conditions where a tertiary amine is used along with the coupling agent (HBTU (9), HATU (10), TBTU (11), TPTU (79), etc.), they can also be used in combination with HOBt (5) to promote the formation of the activated benzotriazolyl ester and thereby increase the yield of the desired reaction [4].

The reagent HBTU (9) is not commonly used on a large scale for amide bond formation due to its high economic cost and lower yields. Conversely, its structural analog, HATU (10), is used because it provides faster and more efficient couplings with less epimerization, as well as milder reaction conditions. This might be attributed to the potential formation of a hydrogen bond in the reaction intermediate, as mentioned earlier in previous sections. Additionally, it has been found that the counterion of both reagents has no significant effect on the course of the reaction, with no substantial changes in its yield [1, 4].

The main drawback of these reagents is their high molecular weight, which, along with their high cost, increases the cost of the amidation process. On the other hand, N,N,N′,N′-tetramethylurea is formed as a byproduct, which is a cytotoxic compound and challenging to purify. However, their use is justified when, for example, the carboxylic acid contains an α-epimerizable chiral center and mild reaction conditions are required [1].

Other derivatives also commercially available and developed more recently are: TDTU (81), HDTU (82), TDATU (83), HDATU (84), TSTU (85) (Fig. 36) [5].

Fig. 36:

Reagents derived from guanidinium and uronium salts.

These coupling agents follow the reactivity trend already mentioned above, with uronium salts being more efficient than their guanidinium analogues.

Some examples described in the bibliography that use this methodology are described below (Fig. 37) 32], [33], [34], [35], [36.

Fig. 37:

Amide bond formation reaction using reagents derived from guanidinium and uronium salts.