3.2. TCRT Using Ammonia as the Nitrogen Source
Tohda et al. studied the reaction of dinitropyridone
1 with ketones in the presence of ammonia (
Table 2) [
24]. When a methanol solution of pyridone
1 is heated with cyclohexanone
13a in the presence of ammonia (20 equiv.) at 70 °C (condition A), cyclohexa[
b]pyridine
14a is obtained in 83% yield. However, this method suffers from the narrow scope of ketones. The TCRT using cyclopentanone
13b under the same conditions forms cyclopenta[
b]pyridine
14b in a considerably lower yield. When acetophenone
15a is allowed to react under the same conditions, TCRT proceeds similarly; however, the yield is low owing to the competitive ammonolysis of substrate
1. To overcome this disadvantage, it is important to employ severe conditions (heating with larger amounts of ammonia (140 equiv.) at 120 °C in an autoclave (condition B)). This reaction is applicable to other aromatic ketones
15b–
h to afford the corresponding 2-(het)aryl-5-nitropyridines
16b–
h, respectively. The ketone is not required to have an acetyl group, and propiophenone
15i undergoes the TCRT, leading to trisubstituted pyridine
16i. In the case of aromatic ketones
15a–
i, employment of condition B is effective for obtaining pyridines
16a–
i in better yields. In contrast, ketone
15j possessing an α‘-proton forms pyridine
16j with better yield under condition A, as severe conditions cause side reactions. Indeed, pinacolone
15k without an α‘-proton undergoes the TCRT more efficiently.
This TCRT efficiently proceeds under mild conditions (condition A) only when cyclohexanone 13a is used as the reagent. In other words, this protocol is an effective approach to [b]-fused 5-nitropyridines. This reaction is often employed for synthesizing biologically active compounds, medicines, and their synthetic intermediates.
Cyclohexa[
b]pyridines
14c–
f (
Figure 2) are synthesized by TCRT using ammonia as a nitrogen source, in which functional groups such as carbamate, ester, and acetal are tolerated during the reaction [
25,
26,
27,
28,
29,
30]. Notably, multiple functionalities remain during the TCRT to afford a complex structure
14f. Piperidine-4-ones are usable as reagents in TCRT to produce 5,6,7,8-tetrahydro-1,6-naphthyridines
14g–
m [
31,
32,
33,
34,
35,
36,
37]. Not only
N-alkylated derivatives
14g–
i, but also
N-aryl derivative
14j and
N-acyl derivatives
14k–
m are available. When unsymmetrical pieridine-3-one is used, two condensed pyridines are formed, including 1,5-naphthyridine
14n [
38]. Tetrahydropyran-4-one can be used for this method, which makes pyranopyridines
14o and
14p available [
39,
40,
41].
Cycloalkanones with different ring sizes can also be used as reagents in this TCRT (
Figure 3). Cyclopenta[
b]pyridine
14q, even though it has a complex structure, can be synthesized by altering the cyclopentanone to the corresponding one [
42,
43]. When pyrrolidine-3-one is used, 4-azaindole
14r is obtained [
44].
Furthermore, cyclohepta[
b]pyridine
14s can be synthesized upon treatment of pyridone
1 with cycloheptanone
13s [
45,
46]. When aza-containing cycloheptanone
13t and bridged cycloheptanone
13u are employed, cycloheptapyridine
14t [
47] and tricyclic pyridines
14u [
36] are formed. Nitropyridines condensed with a larger ring (from eight to eleven membered rings) can be prepared by only altering cycloalkanones [
48,
49].
3.3. Reaction Mechanism of TCRT
Two plausible mechanisms of TCRT are illustrated in
Scheme 9. As mentioned in
Section 2.1, both the 4- and 6-positions of dinitropyridone
1 are highly electrophilic, and are thus attacked by the enol form of
15a and ammonia to form adduct intermediate
17 (path a) [
24]. The same product,
16a, is obtained when the ammonia and enol switch positions to attack. The amino group intramolecularly attacks the carbonyl group derived from
15a, leading to bicyclic intermediate
18, from which nitroacetamide is eliminated and accompanied by aromatization to afford nitropyridine
16a. Another possibility is that ketones are converted to enamines, which might serve as an actual nucleophile (path b) [
50]. After adding the enamine to pyridone
1, the amino group intramolecularly attacks the 6-position to form bicyclic intermediate
20, and elimination of nitroacetamide leads to the formation of nitropyridine
16a.
3.4. TCRT Using Ammonium Acetate as the Nitrogen Source
This TCRT proceeds efficiently when reactive cycloalkanones 13 are employed as reagents. In other words, when less reactive ketones such as 15a are used, both electrophilic sites of 1 are attacked by ammonia, which undergoes ammonolysis to consume pyridone 1 competitively. Le et al. mitigated this problem by using a less nucleophilic ammonium acetate as a nitrogen source instead of ammonia.
When pyridone
1 is reacted with acetophenone
15a and three equivalents of ammonium acetate, nitropyridine
16a and a bicyclic product
21a are obtained (
Table 3) [
51]. The former is produced by TCRT, and the latter is formed by the insertion of
15a and nitrogen between the N1 and C2 positions of pyridone
1. Isolated
21a can be converted to
16a upon treatment with ammonium acetate, which indicates that there is equilibrium between these products. Thus,
16a is a thermodynamically controlled product, and
21a is a kinetically controlled product. The ratio of
16a increases as larger amounts of ammonium acetate or microwave heating are used. The use of larger amounts of ammonium acetate prolongs the actual reaction time, because it decomposes to gaseous ammonia and acetic acid upon heating.
The formation of bicyclic product
21a is considered to proceed as shown in
Scheme 10. After addition of an enol form of
15a to the 4-position of
1, the acyl moiety of
22 is converted to enamine
19 by the ammonium ion. When the amino group of
19 intramolecularly attacks at the 6-position (path c), nitropyridine
16a is formed via bicyclic intermediate
20, as illustrated in
Scheme 9. In contrast, the amino group of
19 attacks the carbonyl group, and degenerated ring transformation proceeds to afford
24. After prototropy leading to
25, the methylamino group attacks the imino functionality to afford bicyclic product
21a. However, the aminal structure of
21a is easily cleaved under acidic conditions to regenerate intermediate
19, which furnishes aromatized product
16a, predominantly under severe conditions.
This method is applicable to other aromatic ketones
15a–
q (
Table 4). TCRT efficiently proceeds in reactions using both electron-rich and electron-poor ketones, among which electron-poor ketones reveal lower reactivity and require larger amounts of ammonium acetate (longer reaction time). In cases of electron-poor ketones
15e,
15f, and
15o, bicyclic products
21e,
21f, and
21o are obtained, respectively. The ketone is not required to have an acetyl group, and ketones
15i and
15q afforded the corresponding trisubstituted pyridines
16i and
16q in almost quantitative yields, respectively.
α,β-Unsaturated ketones
26 and
28 can also be used for the TCRT (
Table 5 and
Table 6) [
52]. These ketones are less reactive, requiring 15–30 equivalents of ammonium acetate. Among the three styryl ketones, electron-rich ketone
26b reveals higher reactivity, which facilitates the approach to electron-deficient pyridone
1. The reaction with alkynyl ketones
28 efficiently furnishes alkynylpyridines
29. When silylethynyl ketone
28c is used, the desilylated product
29d is also obtained.
For the C–C bond formation on the pyridine framework, the Heck, Suzuki, Stille, and Sonogashira reactions are commonly used. However, these methods require the use of poisonous and expensive transition metals and a purification step to avoid metal contamination of the products. In addition, troublesome multistep reactions are necessary to prepare the substrates for these reactions (2-halo-5-nitropyridines). Thus, the TCRT is a metal-free supplementary method for the abovementioned reactions.
3.6. TCRT Using Cyclic Ketones 13
Dinitropyridone
1 undergoes TCRT with cycloalkanone
13 in the presence of ammonium acetate, leading to cycloalka[
b]pyridines
14 (
Table 8) [
55]. Cycloalkanones
13 with various ring sizes efficiently react under conventional heating (Condition C) to afford the corresponding nitropyridines condensed with five-, six-, seven-, and eight-membered rings. The reaction time is considerably shortened by using microwave heating (Condition D). In this reaction, the unsymmetrical ketone, 2-methylcyclohexanone
13aa, which reacts at the 6-position not at the 2-position, as aromatization is prevented by a methyl group in the latter case, can also be used as a reagent. When 2-cyclohexenone
13ab is used, migration of the double bond is observed, which may occur after the addition of ketone
13ab to pyridone
1 and the subsequent conversion to dienamine
32ab, leading to the formation of dienamine
33ab (
Scheme 11).
3.7. Reconsideration about the Reaction Mechanism of TCRT
As shown in
Scheme 10, the TCRT is initiated by the addition of the enol form of a ketone to the 4-position of dinitropyridone
1, after which the acyl group of adduct
19 is converted to enamine
20 by the ammonium ion. Enamine has an ambident property, where β-carbon is generally more nucleophilic than the amino group. In the case of adduct intermediate
19 derived from aromatic ketone
15,
N-attack (path c) forms a six-membered ring to afford bicyclic intermediate
20, from which nitropyridine
16 is obtained, accompanied by the elimination of nitroacetamide (
Scheme 12). In contrast, if a
C-attack (path e) occurs, sterically strained four-membered ring
34 is formed. Hence, nitropyridine
16 is formed as the sole product in this TCRT. In cases of α,β-unsaturated ketones
26 and
28 and aldehydes
30, a similar reactivity is observed, as these carbonyl compounds have only one kind of α-hydrogen.
In the case of aliphatic ketones
36, two types of enamines (
37 and
38) are possibly formed (
Scheme 13). While the intermediate
37 cannot cause a
C-attack similar to
19, the intermediate
38 can cause both
N- and
C-attacks to furnish bicyclic intermediates
41 and
42, respectively. From bicyclic intermediates
40 and
41, nitropyridine
43 is formed. In contrast, 2,6-disubstituted 4-nitroaniline
44 should form when nitroacetamide is eliminated from bicyclic intermediate
42. Thus, two ring-transformed products (
43 and
44) are yielded when aliphatic ketones
36 are used as reagents.
In the reactions of pyridone
1 with cycloalkanone
13, only nitropyridine
14 is formed (
Table 8). Although the adduct of
1 and cycloalkanone
13 can form two kinds of enamines, one enamine can form a six-membered ring as a result of C-attack, and the formed intermediate
35 is too strained to be formed (
Scheme 12).
3.8. TCRT Using Aliphatic Ketones 36
When dinitropyridone
1 is subjected to a reaction with aliphatic ketones
36 in the presence of ammonium acetate, two types of TCRT occur to afford nitropyridines
43 and nitroanilines
44 (
Table 9) [
56]. Generally, 2,6-disubstituted 4-nitroanilines
44 are prepared from the corresponding anilines by nitration under harsh reaction conditions, wherein protection and deprotection of the amino groups are necessary [
57]. Furthermore, the preparation of this compound suffers from the limitation of Friedel–Crafts alkylation. There are several limitations for the Friedel–Crafts alkylation, such as the following: (1) The monoalkylated product undergoes further alkylation, (2) it is difficult to introduce two different alkyl groups, (3) primary alkyl groups longer than the ethyl group cannot be introduced, (4) a phenyl group cannot be introduced, and (5) nitrobenzene and aniline do not facilitate the alkylation. The TCRT overcomes these disadvantages.
When dinitropyridone
1 is reacted with 3-pentanone in the presence of five equivalents of ammonium acetate, nitroaniline
44a and nitropyridine
43a are obtained at 50% and 44%, respectively, resulting from two types of TCRT. In contrast, the ratio of
44a to
43a increases significantly without a decrease in total yield, indicating the presence of an equilibrium between bicyclic intermediates
42 and
41 (
Scheme 13). The substituents can be modified by altering only the ketones
36 (
Table 9). Monoalkylated nitroanilines
44c–
e and unsymmetrical nitroanilines
44h and
44i are available from the corresponding unsymmetrical ketones
36. Furthermore, it is easy to prepare nitroanilines
44g–
i possessing a propyl or phenyl group, which cannot be introduced by the Friedel–Crafts reaction. However, steric repulsion by the phenyl groups prevents the formation of bicyclic intermediate
42i.
A combination of propylamine
45A and acetic acid can be used as a reagent instead of ammonium acetate, which facilitates
N-modification of the amino group as well as the benzene ring of nitroaniline
46 (
Table 10). This method is applicable to secondary amines, pyrrolidine
45B and diethylamine
45C, to afford
N,N,2,6-tetrasubstituted 4-nitroanilines
46B and
46C, respectively. This reaction also enables the introduction of a propyl or phenyl group into the benzene framework, which cannot be introduced by the Friedel–Crafts reaction.
As shown in
Scheme 13, the TCRT proceeds through the
C-attack of the intermediately formed enamine
38. This means that functionalized nitoanilines
48 can be prepared if a similar structure is available via an alternative route. For this purpose, relatively stable enaminones
47 prepared from 1,3-dicarbonyl compounds
7 and amine
45 are considered suitable. When dinitropyridone
1 reacts with enaminone
47, nucleophilic-type ring transformation proceeds to afford 2-functionalized 4-nitroaniline
48 (
Table 11) [
58]. This protocol facilitates the modification of the functional group and amino group of
48 by altering 1,3-dicarbonyl compounds
7 and amine
45. Diketones
7c and
7e as well as keto easter
7b can be used as 1,3-dicarbonyl compounds. These reagents are not required to possess an acetyl group (R
1 = H), and
7f undergoes similar ring transformations. Bulky amines such as
tert-butylamines
45D and
45E and less nucleophilic anilines
45F and
45G can be used as amines. Even though amines have a functional group, the corresponding nitroaniline
48Hb is obtained. Furthermore, cyclic and acyclic secondary amines
45B and
45C can be used for this reaction, which results in 2-functionalized
N,N-dialkyl-4-nitroanilines
48Bc,
48Ca, and
48Ce.