2.1. Optimization of the Two-Step Synthesis of Imidazolin-2-ones 1a–g
To synthesize imidazolidin-2-ones
1a–
g, initially, a multistep methodology was employed, which involved three stages: Schiff base formation, reduction to diamine, and cyclization via reaction with carbonyldiimidazole (CDI). The process began with the enantiomerically pure
trans-(
R,
R)-diaminocyclohexane, obtained by basifying
trans-(
R,
R)-diammoniumcyclohexane mono-
L(+)-tartrate [
28] with a saturated sodium bicarbonate solution and extracting with dichloromethane. The first reaction, which involved the formation of the Schiff base
2a, was carried out using methanol and THF as solvents, which are classified as medium- to high-polarity solvents. The reaction was carried out at reflux conditions (40 and 80 °C, respectively) and was continuously monitored by thin layer chromatography (TLC) using a mobile phase composed of a mixture of hexane and ethyl acetate in a proportion of 8:2 and silica gel as the stationary phase. The complete consumption of benzaldehyde, which was detected by TLC, made it possible to establish the reaction times in the first step. Subsequently, sodium borohydride (NaBH
4) was added to reduce the Schiff base
2a, followed by TLC until the total consumption of
2a was confirmed. The results of each step are presented below (
Table 1).
Using methanol as a solvent was suitable for forming the Schiff base 2a and its subsequent reduction. The formation of the Schiff base 2a did not require high temperatures, and the reaction time was relatively short. Furthermore, the reduction was also performed at 40 °C and for 30 min and showed a good yield. Compared to methanol, THF requires higher temperatures (80 °C) for both the Schiff base formation and the reduction. The reaction time for the reduction is also significantly longer (4 h versus 30 min for methanol). Despite these more extreme parameters, the yield obtained (74%) is almost equal to that of methanol (74%). However, the longer reaction time in THF could indicate slower reaction in this solvent.
Structural elucidation of compound
3a was carried out by
1H and
13C nuclear magnetic resonance (NMR) spectroscopy. In the
1H NMR spectrum, several characteristic signals are observed: axial hydrogens of the cyclohexane system (
Figure 1), marked as H4
axial, H5
axial, H6
axial, H7
axial, appear as a multiplet in the range of 1.14–1.02 ppm; equatorial hydrogens (H5
equatorial, H6
equatorial) occur at 1.75–1.70 ppm; and hydrogens attached to chiral centers (H3a, H7a) at 2.32–2.27 ppm. The signals corresponding to the diastereotopic benzylic hydrogens are observed as doublets at 3.91 and 3.66 ppm, with a coupling constant of 15.2 Hz. Furthermore, the aromatic hydrogens of the benzene ring occur as a multiplet in the range of 7.37–7.22 ppm. Regarding the
13C NMR spectrum, the chemical shifts reveal the presence of carbon atoms in specific positions: the methylene carbons of cyclohexane (C4 and C5) are found at 24.9 ppm, the methylene carbons (C4 and C7) at 31.1 ppm, and the chiral carbons (C3a and C7a) at 60.5 ppm. The benzylic methylene carbon is located at 50.5 ppm, and the aromatic carbons in the phenyl system exhibit chemical shifts at 140.2 ppm, 128.4 ppm, 128.1 ppm, and 126.9 ppm. These results are consistent with those reported in the literature [
29].
The cyclization of diamine
1a with CDI (
Table 2) allowed evaluation of the influence of the solvent on the reaction yield. The reaction performed in toluene did not produce the desired compound, as the yield was less than 5% after 64 h. The low solubility of the reactants could explain this low yield in the apolar solvent, which limits the reactivity and hinders the formation of the main product. When tetrahydrofuran (THF) was used as the solvent at 20 °C, a higher yield was observed, reaching 21% after 64 h. However, this yield is still relatively low, suggesting that, although THF is more polar than toluene, it does not optimally favor the cyclization of the diamine with CDI under these conditions. Finally, when performing the reaction in dichloromethane (DCM) at 40 °C, a notably higher yield of 98% was obtained after 17 h. Under these conditions, the reaction was highly efficient, yielding a single main product, imidazolidin-2-one
1a.
The
1H NMR analysis of compound
1a, which is consistent with the literature [
27], enabled detailed structural elucidation of the molecular structure. The first fragment, the perhydrobenzimidazole-2-one heterocyclic system, shows a complex signal system corresponding to the hydrogen atoms of the cyclohexane ring, indicating the presence of a saturated structure in this portion like that observed for compound
3a. The corresponding signal to the
axial hydrogens (H4
axial, H5
axial, H6
axial, H7
axial) appears as a multiplet in the 1.22–1.12 ppm range. The hydrogens at the
equatorial positions (H5
equatorial, H6
equatorial) appear in a multiplet in the range of 1.72–1.63 ppm, and the signals for the hydrogens at the H4
equatorial and H7
equatorial positions also appear as a multiplet in the range of 1.85–1.89 ppm. A key structure fragment is the benzylic methylene, which exhibits two doublet-type signals at 4.36 ppm and 4.48 ppm, with a coupling constant of 15.2 Hz. These signals correspond to the diastereotopic hydrogens, which are coupled to each other. Finally, the aromatic ring shows a multiplet in the range of 7.33–7.23 ppm, with an integration of 10 protons. This indicates the presence of two benzylic rings attached to the methylene benzylic groups. Analysis of compound
1a by
13C NMR revealed key information about the distribution of carbon atoms in the molecular structure. For the perhydrobenzimidazole-2-one system, a signal corresponding to methylene carbon atoms at the C4 and C5 positions of the cyclohexane system is observed at 24.1 ppm. This chemical shift is typical of hydrogen-bonded carbons in a saturated environment, which is consistent with the presence of these carbons in the cyclohexane ring. At 28.4 ppm, methylene carbon atoms are found at the C4 and C7 positions of the cyclohexane system. The slightly higher chemical shift may reflect the proximity of these carbons to chiral centers, which is expected in this portion of the structure. At 61.7 ppm, the signal corresponding to the chiral carbons C3a and C7a is observed. This shift is typical for carbons attached to a nitrogen atom (in this case, within the imidazolidinone ring), and is in the range of chemical shifts expected for these carbon atoms close to a chiral center. At 46.9 ppm, a signal is assigned to a carbon atom corresponding to the methylene benzylic carbon atom. For the aromatic ring, four signals are observed in the range of 137.8–127.2 ppm, which correspond to the carbons of the aromatic system. These chemical shifts are consistent with the presence of a monosubstituted aromatic system. Finally, at 163.6 ppm, the chemical shift corresponding to the carbon of the C=O bond in the imidazolidin-2-one ring is observed.
A detailed analysis of the results obtained from the two-dimensional NMR experiments (
Figure 2) performed on compound
1a is presented below to interpret the interactions and relationships between the protons and carbons present in the proposed structure. The COSY experiment allowed observation of the interactions through bonds between nearby protons, identifying hydrogen groups coupled to each other. In the cyclohexane fragment, the axial protons H4
axial, H5
axial, H6
axial, and H7
axial exhibit chemical shifts in the 1.22–1.12 ppm range. These showed correlations with the equatorial protons H5
equatorial, H6
equatorial, H4
equatorial, and H7
equatorial, which have chemical shifts in the range of 1.72–1.63 ppm and 1.85–1.89 ppm, respectively. The equatorial protons (H5
equatorial, H6
equatorial) in the 1.72–1.63 ppm range correlate with the axial protons (H4
axial, H5
axial, H6
axial, H7
axial) in the 1.22–1.12 ppm range. The H4
equatorial and H7
equatorial protons (1.85–1.89 ppm) correlate with the axial protons in the chemical shift range of 1.22–1.12 ppm. The protons at the H3a positions (2.75–2.63 ppm) showed correlations with the equatorial protons H4
equatorial and H7
equatorial and axial protons H4
axial and H7
axial. Regarding the benzylic protons, those at 4.36 ppm showed correlations with the protons at 4.48 ppm, confirming that they are diastereotopic protons. Finally, the protons of the aromatic ring, with displacements in the range of 7.33–7.23 ppm, correlate with each other in the same region.
The HMQC experiment provided information on the direct correlation between the protons and the carbons to which they are attached. The axial protons at positions H4–7axial correlated with two methylene carbons in cyclohexane which appeared as signals at 24.1 and 28.4 ppm. The equatorial protons (1.72–1.63 ppm) correlated with two methylene carbons (24.1 ppm). The protons at positions H4equatorial and H7equatorial (1.85–1.89 ppm) also correlated with the methylene carbons. The protons attached to the chiral center H3a and H7a (2.75–2.63 ppm) correlated with a methine chiral carbon which appeared as a signal at 61.7 ppm, indicating the presence of a chiral center. The diastereotopic benzylic hydrogens at 4.36 and 4.48 ppm correlated with a benzylic carbon at 46.9 ppm. The protons in the 7.33–7.23 ppm region showed correlations with methine-type aromatic carbons in the 127.0–128.4 ppm range. Finally, the HMBC experiment revealed long-distance interactions between the protons and the carbons. The axial protons (1.22–1.12 ppm) correlated with the chiral methine-type carbon (61.7 ppm), suggesting an interaction via long-distance bonds. Like the Haxial, the equatorial protons (1.72–1.63 ppm) correlated with the chiral methine-type carbon (61.7 ppm). The protons at 2.75–2.63 ppm correlated with the methylene carbons from cyclohexane (24.1 and 28.4 ppm) and a signal for the C=O carbon at 163.6 ppm, suggesting the proximity of this carbonyl group in the structure. The protons at the ArCH2N positions (4.36 and 4.48 ppm) correlated with the signal for the C=O carbon (163.6 ppm) and an aromatic carbon at 137.8 ppm. The aromatic ring protons (7.33–7.23 ppm) correlated with methine-type aromatic carbons and the C=O carbon. The results of the COSY, HMQC, and HMBC experiments demonstrated interactions between the protons and carbons of the cyclohexane system, diastereotopic benzylic protons, and the aromatic rings, in addition to the presence of a chiral center and a carbonyl group, confirming the proposed structure of compound 1a.
With the optimized reaction conditions, the synthesis of imidazoline-2-ones 1b–g was carried out using p-substituted aromatic aldehydes. After optimizing the reaction conditions, the desired compounds 1a–g were obtained, which were characterized using ATR-FTIR, HR-MS, 1H, and 13C NMR. The characterization data are consistent with those described for compound 1a. ATR-FTIR reveals that in all compounds, the value of the asymmetric CH2 stretching band is in a nearby region, between 2929 and 2935 cm−1. This value is characteristic of methylene groups (–CH2–) attached to a sp3 carbon atom. The band corresponding to the symmetric CH2 stretching shows similar behavior to the asymmetric band, with values between 2851 and 2863 cm−1. As in the case of asymmetric stretching, the small variations in frequency between compounds 1a, 1b, and 1c may be influenced by the inductive effects of the benzyloxy and 4-methylphenoxy groups. However, these effects are not large enough to drastically alter the band position. The carbonyl (C=O) stretching frequency is one of the most distinctive bands in the ATR-FTIR spectrum. These bands are located between 1693 and 1700 cm−1, consistent with the expected behavior for a C=O attached to an imidazolin-2-one ring. This band’s slight variability suggests that the benzene ring’s substituents may influence the carbonyl group’s interaction with the molecular environment, possibly due to resonance or electronic polarization effects. In the case of compound 1b (benzyloxy), the band appears at a slightly lower frequency (1698 cm−1), which may indicate a slight increase in the electron density over the carbonyl group, affected by the benzyloxy substituent. Finally, the band corresponding to the C–N stretching in the imidazolin-2-one fragment appears in a region close to 1252 cm−1 in the case of compound 1a and in a slightly lower region for the compounds with benzene substituents, 1b (benzyloxy) and 1c (4-methylphenoxy), with values of 1239 cm−1 and 1249 cm−1, respectively. This shift towards lower frequencies could indicate a change in the C-N bond strength due to the influence of the substituent groups, which can induce changes in the electron density over the imidazolin-2-one ring and, therefore, modify the nature of the C-N bond. The 1H NMR results show that the signals corresponding to the benzyl hydrogens of the compounds 1a–g are between 3.83 and 4.48 ppm, which reflects variability in the chemical position due to the presence of different substituents in the benzene ring. The specific effects of each substituent are discussed below. For compound 1b, the signal moves to 4.47 and 4.21 ppm, indicating a slight modification in the electronic environment of the benzene ring due to the presence of the 4-benzyloxy group, which is an electron donor that causes greater electronic density in the benzene ring. A similar effect is observed for compounds 1c–1e, for which the signal of benzyl hydrogens also appears slightly shifted compared to that observed for 1a. For derivatives with halogen groups, such as 1f–g, the signals are observed at 3.86 and 3.60 ppm, representing a more substantial shift towards a higher field than electron donor compounds. Both bromine and chlorine are electronegative substituents that, through their inductive nature, decrease electronic density in the benzene ring, which could explain the observed shift. Finally, the 1C NMR analysis showed resonance signals for benzylic carbons in the range of 46.1–46.9 ppm, with slight variations attributable to the substituent groups on the benzene ring. The carbonyl carbon atoms (C=O) appear as signals between 163.3 and 163.6 ppm, reflecting a constancy in the environment of the carbonyl group that is not significantly altered by the substituents on the ring.
Yields for compounds
1a–
g using a multistep protocol are depicted in
Table 3. These results showed significant variations in yields depending on the substituents in the starting structure. These results reflect how the substituent’s electronic and steric properties affect both the formation of the diamine and its subsequent cyclization with CDI, impacting the overall yield of the process. Compound
1a showed the best yields in each stage, with 74% in the diamine production stage and 98% in the cyclization stage, resulting in an overall yield of 73%. The absence of bulky groups or significant electronic effects presumably facilitates the formation of the Schiff base and its subsequent cyclization, highlighting the favorable influence of a system without steric or electronic impediments. Regarding imidazolin-2-one
1b, introducing the benzyloxy group increased the yield in the diamine production stage (83%), possibly due to electronic interactions of the oxygen group with the carbonyl system of the Schiff base, stabilizing the intermediate. However, the cyclization yield decreased to 84%, which could be attributed to the greater steric complexity of the benzyloxy group, making the reaction with CDI more difficult. The overall yield was 70%, close to that of compound
1a. For imidazolidin-2-one
1c, the highest yield was observed in the diamine production stage (90%), suggesting that the 4-methylphenoxy group favors the formation of the Schiff base due to the electron-donating effects of the methyl group in the aromatic ring, which can stabilize the intermediate. However, the cyclization yield was notably lower (63%). Regarding compound
1d, the bulky and electron-donating effect of the 4-diphenylamine group significantly negatively affected both steps. The cyclization yield was extremely low (22%), indicating that this group severely hinders the formation of the imidazolidin-2-one fragment. The overall yield was the weakest of all the compounds evaluated (17%), reflecting that bulky substituent groups are unfavorable for this type of synthesis. Additional substituents with varying electronic effects were introduced for compounds
1e,
1f, and
1g. Compound
1e, with the
tert-butoxy group, showed a relatively low yield in the diamine formation stage (65%) and a modest cyclization yield of 59%. The
tert-butyl group is an electron-donating group, which affected the aldehyde electrophilicity towards the Schiff base formation and the cyclization stages, reducing yields in both steps. In compound
1f, introducing the chlorine atom resulted in a yield of 79% in the diamine formation and a high cyclization yield of 95%. Chlorine is an electron-withdrawing group, which would stabilize the electron-rich nitrogen of the Schiff base, facilitating cyclization. The favorable electronic properties of chlorine and relatively low steric hindrance allowed for a high overall yield of 75%, one of the best results among the substituted compounds. Finally, compound
1g, with a bromo group, exhibited 78% and 94% yields in the diamine and cyclization steps, respectively. The bromine group, like chlorine, is electron-withdrawing, though less strongly so. It also provided favorable stabilization to the Schiff base intermediate, leading to a cyclization yield of 94%. As a result, the overall yield of 73% was like that of
1f, showing that halogen substitution can significantly improve the cyclization step compared to more sterically demanding or electron-donating groups. These results showed that bulky groups or those with significant electronic effects (
1c,
1d, and
1e) influence the cyclization step.
To improve the yields of compounds
1a–
g, conditions for a pseudo-multicomponent synthetic strategy were evaluated, with varying reaction conditions such as CDI stoichiometry, temperature, and reaction time. Moreover, microwave irradiation (MW) was employed in this study. The use of MW irradiation in organic synthesis has proven highly beneficial, particularly in accelerating reactions, increasing yields under milder conditions, and improving product purity [
30,
31]. The thermal effects of MW irradiation arise from several factors, including the rapid heating rate, localized overheating, formation of “hot spots”, and the selective absorption of radiation by polar compounds. MW irradiation reduces reaction times and increases yields, making the process more efficient compared to conventional methodologies, where heating is slower and less controlled. Furthermore, MW in this type of reaction enabled us to operate under milder conditions, reducing energy consumption.
2.2. Pseudo-Multicomponent Experiments for the Synthesis of Compounds 1a–g
Several experiments were performed to obtain compound
1a, evaluating the influence of the mentioned variables, to improve the yield of compound
1a–
g through a pseudo-multicomponent protocol. Trying to avoid the chemical reaction of reagents such as NaBH
4 and CDI, THF and DCM were selected as solvents. The final stage of the synthesis was optimized using a 3
3 full-factorial experimental design, which examined the effects of three key factors—temperature, reaction time, and reactant stoichiometry—each at three levels, as shown in
Table S1. The response variable was the reaction yield percentage. The experimental data were analyzed by fitting a quadratic model, which was validated through analysis of variance (ANOVA), which identified significant effects for factors with
p-values < 0.05 (
Table S2). The model’s performance in capturing the variability of the experimental data was assessed using the coefficient of determination (R
2). Additionally, significant differences in yield percentages among treatments were determined through a post hoc Tukey test.
The analysis of variance (ANOVA) conducted on the experimental data provided valuable insights into the effects of the studied factors and their interactions (
Supplementary Materials, Table S2). Initially, the quadratic model was statistically significant, with a
p-value of 0.0012, confirming its ability to explain the observed variability in the data. The coefficient of determination (R
2) was 0.7454, indicating that the model explains approximately 74.5% of the variation in the response. As for the main effects, temperature (factor A) significantly impacted reaction yield, with a
p-value less than 0.0001 and an F-value of 31.90. This indicates that temperature is a key factor influencing the yield. In contrast, reaction time (factor B) yielded a
p-value of 0.0560, suggesting a marginally significant effect. Although close to the significance threshold, it does not fully meet it, implying that reaction time may influence the yield, but its impact is less pronounced compared to temperature. Conversely, the stoichiometry of the reactants (factor C) was not significant, with a
p-value of 0.4135, indicating that variations in yield are not attributed to this factor. When examining the interactions between the factors, none of the combinations—temperature and reaction time (interaction AB), temperature and stoichiometry (interaction AC), or reaction time and stoichiometry (interaction BC)—were significant, with
p-values above 0.05 in all cases. This suggests that no significant interactions between these factors affect the reaction yield. Additionally, the quadratic effects showed that temperature had a significant quadratic influence, with a
p-value of 0.025, indicating a non-linear relationship between temperature and yield, within the parameter space explored in this experimental design. Finally, the model’s residuals did not display systematic patterns, suggesting that the quadratic model adequately fits the experimental data and does not require further refinement. However, it is important to recognize the empirical nature of this statistical model. Since the design of the experiments enabled construction of a statistical rather than a physicochemical model, the model’s predictions are valid only within the explored experimental space bounded by the upper and lower limits of the factors studied in this work. Extrapolating beyond these boundaries may introduce inaccuracies, requiring additional experimental validation [
32]. This limitation is particularly relevant when scaling up the process, as factors not considered in the initial model—such as mass and heat transfer effects—may influence the reaction outcome, necessitating further refinement and optimization.
The results of the Tukey post hoc test, based on the data in
Table S3, reveal significant differences in the yields (%) among the different combinations of temperature, time, and stoichiometry. The yields were categorized into distinct groups, with the letter assignments indicating statistical significance. In general, the highest yields were observed under the combination of 40–70 °C and 40–60 min, where treatments T3, T5, T11, and T24 all showed yields over 59% and were grouped with the letter “a” indicating no significant differences between these treatments. The statistical analysis confirmed that temperature had a significant impact on the reaction yield (
p < 0.0001), while time showed a marginal effect (
p = 0.0560). In contrast, stoichiometry had no significant impact on the results (
p = 0.4135), which aligns with the observation that high yields were obtained across different stoichiometric conditions. Conversely, treatments at 100 °C generally exhibited lower yields. For example, T7, T8, and T9, with reaction times of 20, 40, and 60 min, respectively, produced yields ranging from 7% to 14%. These results suggest that excessively high temperatures (100 °C) may be detrimental to the reaction yield, regardless of the reaction time. Finally, a pseudo-multicomponent synthetic protocol was optimized using the discussed statistical analysis to produce compounds
1a–
d (
Table 4).
Table 4 presents the yields for compounds
1a–
g synthesized through two different protocols: a multistep protocol and one-pot protocol. The results suggest that the pseudo-multicomponent protocol is more efficient for synthesizing imidazolidin-2-ones
1a–
g, as it offers higher yields in all cases tested. Pseudo-multicomponent reactions (pseudo-MCRs) offer a compelling strategy for synthesizing complex molecules by involving the reactivity of fewer components than traditional multicomponent reactions. These reactions are characterized by having two out of three or more identical components, which, although often seen as undesired in specific synthetic contexts, can provide remarkable efficiency in generating molecular complexity. Pseudo-MCRs play an essential role in synthesizing complex heterocycles, functionalized compounds, and biologically relevant molecules, making them a powerful tool in organic chemistry. Castillo et al. [
33] reported a notable example of how pseudo-MCRs can be harnessed effectively. The report discusses aryne-based pseudo-MCRs, where reactive intermediates like arynes mediate reactions to form complex heterocyclic structures. In particular, reactions involving 2-aza-dienes and electron-rich
N-heteroaryl imines are explored for their ability to synthesize biologically relevant heterocycles such as
N-arylated 1,2-dihydroisoquinolines. The pseudo-three-component and pseudo-four-component reactions outlined in the study enable the assembly of these heterocycles with high efficiency. A different approach to pseudo-MCRs is reported using Meldrum’s acid, 4-nitrobenzaldehyde, and isocyanides to form intermediates for secondary reactions [
34], demonstrating how pseudo-MCRs can be used to produce a range of complex products by manipulating the number of components involved in each step of the reaction. Moreover, steric effects of amidine salts on the formation of
N-aroylmethylimidazoles and 1
H-imidazoles by reaction of α-bromoketones with high regioselectivity [
35], and the reaction between propargyl amine, an aldehyde, and an electron-deficient alkene [
36] to produce highly diastereoselective cycloadducts, exemplify how pseudo-MCRs represent an exciting and efficient strategy in modern synthetic chemistry. This strategy allows the rapid assembly of complex molecules from fewer components, thereby minimizing waste and reducing reaction times, such as our results suggest, showing that the protocol described and studied here corresponds to a new and significant contribution in this class of synthesis strategies.
Finally, a reaction mechanism is proposed to form compounds
1a–
g (
Scheme 3). The mechanism begins with the reaction of
trans-(
R,
R)-diaminocyclohexane with two equivalents of the aromatic aldehyde, resulting in a Schiff condensation reaction. In this step, each of the diamine’s amino groups attacks the aldehyde’s carbonyl, generating an intermediate that rapidly rearranges to form a Schiff base
2, releasing two water molecules. Subsequently, with the addition of a reducing agent, such as sodium borohydride (NaBH
4), which acts on the Schiff base, each imine group is reduced, resulting in the formation of
N,
N-dibenzyldiaminide
3. This reduction is accompanied by the presence of water in the medium, which facilitates the neutralization of the intermediate, finally forming
N,
N-dibenzyldiamine. In the next step, carbodiimide (CDI) is added to the system. CDI acts as an activating agent, initially reacting with one of the amino groups of
N,
N-dibenzyldiamine. This attack generates an intermediate
N-benzyl-
N-(2-(benzylamino)cyclohexyl)-1
H-imidazole-1-carboxamide
4 in which the amino group is covalently linked to the carbodiimide. The second amino group of the diamine then attacks this intermediate intramolecularly, forming a cyclic structure. The reaction culminates in the formation of an imidazolin-2-one
1. The differences in yields are not uniformly large and vary depending on the substituent. In some cases, such as for compound
1d, the difference is considerably more noticeable, which could indicate that this substituent favors this pseudo-multicomponent protocol pathway.