1. Introduction
8-Hydroxyquinoline
1 (oxine) is a highly relevant heterocyclic scaffold in medicinal chemistry due to its metal-chelating ability and the broad spectrum of biological activities exhibited by its derivatives, including cytotoxic, cytoprotective, antimicrobial, antioxidant, and enzyme-inhibitory effects [
1,
2]. These properties have driven the development of synthetic methodologies to functionalize this nucleus and modulate its physicochemical and pharmacological characteristics.
The synthetic value of 8-hydroxyquinoline has been demonstrated through the preparation of diverse libraries of bioactive compounds. For instance, the multicomponent Betti reaction enabled efficient C-7 functionalization, affording inhibitors of KDM4 demethylases and other Fe(II)/2-oxoglutarate-dependent oxygenases [
3]. Likewise, Betti-type derivatives have shown cytoprotective activity in oxidative stress models, with several compounds displaying nanomolar potency [
1]. Among the available modification strategies, the Mannich reaction stands out for its versatility and simplicity, enabling the synthesis of aminoalkylated derivatives under mild conditions by reacting formaldehyde with amines and substrates bearing active hydrogen atoms [
4,
5,
6,
7]. In 8-hydroxyquinoline, the heteroaromatic ring acts as the nucleophilic component, facilitating the regioselective incorporation of aminomethyl groups [
2].
Mannich bases derived from 8-hydroxyquinoline have attracted considerable attention because of their anticancer, anti-inflammatory, antimicrobial, and antioxidant activities [
8,
9,
10,
11,
12]. Notably, several derivatives exhibit selective cytotoxicity toward multidrug-resistant tumor cells, a property associated with their protonation behavior and metal-coordination capacity [
8,
9]. Structural hybrids incorporating phthalimide moieties also showed enhanced activity against resistant colorectal cancer cell lines, highlighting the importance of aminomethyl substitution at C-7 [
10]. More recently, bifunctional systems derived from 5-chloro-8-hydroxyquinoline have been explored via Mannich-type reactions followed by cycloadditions with cyclic imines, indoles, and azaindoles, yielding complex molecular architectures with promising biological profiles [
13,
14,
15]. Beyond medicinal chemistry, the chelating properties of 8-hydroxyquinoline have also been exploited in the development of functional materials for heavy-metal removal and surface modification processes, underscoring the versatility of this pharmacophore [
16,
17,
18,
19]. Although the Betti reaction and Mannich reaction show similarity, the Betti reaction is considered a variant of the Mannich reaction, as both involve the formation of an iminium intermediate. However, they differ in the substrates used and the products obtained. While the Mannich reaction involves an enolizable carbonyl compound to produce β-aminocarbonyls, the Betti reaction uses a naphthol and an aromatic aldehyde to synthesize aminonaphthols (Betti bases), compounds of great interest as chiral ligands and intermediates in organic synthesis [
20,
21].
Based on the demonstrated versatility of cyclic aminals as aminomethylating agents in Mannich-type transformations [
22], as well as the broad pharmacological relevance of the 8-hydroxyquinoline scaffold, we sought to extend this synthetic methodology to reactions involving 8-hydroxyquinoline
1 and the cyclic aminal (2
R,7
R,11
S,16
S)-1,8,10,17-tetraazapentacyclo [8.8.1.1.
8,170.
2,70
11,16]icosane
2. Considering previous reports describing the reaction of 8-hydroxyquinoline
1 with 1,3,6,8-tetraazatricyclo [4.4.1.1
3,8]dodecane
3 (TATD) to afford the
N-substituted imidazolidine 7-(imidazolidin-1-ylmethyl)quinoline-8-ol
4 (
Figure 1) [
23], we expected the formation of a structurally related aminomethylated derivative. Unexpectedly, under the reaction conditions employed, the process led to the formation of (
rac)-(3a
R,7a
R/3a
S,7a
S)-7,7′-((hexahydro-1
Hbenzo[
d]imidazole-1,3(2
H)-diyl)
bis(methylene))
bis(quinolin-8ol)
5. This outcome highlights the distinct reactivity of aminal
3 toward the 8-hydroxyquinoline
1 system and reveals an alternative reaction pathway that differs significantly from that previously observed for TATD
3. The results related to the synthesis of the compound
5, as well as its characterization, are discussed below.
2. Results and Discussion
To evaluate the reactivity of aminal
2 against
1 and explore the formation of the corresponding product, the reaction was carried out in 1,4-dioxane (
Scheme 1). The reaction mixture was initially kept at room temperature to promote interaction between the reactants, then heated under reflux in the presence of water for 8 h. At the end of the process, a precipitate was formed. The resulting solid was isolated by filtration and subsequently purified by recrystallization from methanol, affording compound
5 in 27% yield.
The structural elucidation of the synthesized compound was based on the combined analysis of FT-IR, 1H and 13C NMR spectra, two-dimensional HSQC and HMBC experiments, and ESI-MS, which enabled the consistent assignment of each hydrogen and carbon nucleus in the proposed molecule.
The FT-IR spectrum of the compound
5 (
Figure S1) is consistent with the presence of the functional groups observed in its structure. A broad band is visible in the 3200–3500 cm
−1 region, attributable to phenolic ν(O–H) stretching vibrations broadened by hydrogen bonding interactions. Intense signals around 2950–2850 cm
−1 correspond to aliphatic ν(C–H) stretching vibrations of the methylene and cyclohexane groups present in the molecule. In the 1600–1500 cm
−1 region, bands associated with aromatic ν(C=C) and ν(C=N) vibrations of the quinoline nucleus are observed, characteristic of quinoline-derived heteroaromatic systems [
24,
25,
26,
27]. Furthermore, the absorptions in the 1450–1350 cm
−1 region may be related to δ(C–H) strains and C–N bond contributions from tertiary amines. In the 1300–1200 cm
−1 range, the band attributed to the ν(C–O) stretching of the phenolic group appears, while the signals between 1250 and 1020 cm
−1 are consistent with ν(C–N) vibrations of the amines in the polycyclic system. Finally, the multiple bands observed below 900 cm
−1 correspond to out-of-plane γ(C–H) strains of the substituted aromatic ring, in accordance with the condensed structure of the functionalized 8-hydroxyquinoline.
In the
1H NMR spectrum (
Figure S2), two main regions are clearly identified: an aliphatic region (δ 1.2–4.4 ppm) associated with the saturated heterocyclic system, and an aromatic region (δ 7.1–8.8 ppm) corresponding to the substituted aromatic ring. The region between δ 1.49–1.26 ppm exhibits a multiplet integrating four protons, assigned to the axial hydrogens (H-4, H-5, H-6, and H-7) of the cyclohexane-like system, consistent with its axial arrangement and the expected multiple coupling in a chair conformation. The signals between δ 1.76 and 2.14 ppm correspond to the equatorial protons (H-4, H-5, H-6, and H-7), which appear at slightly lower fields due to their distinct steric and anisotropic environments. The signal at δ 2.49 ppm was assigned to the protons at the bridging positions (H-3a and H-7a), which are unshielded due to their proximity to the nitrogen atoms. Two doublets are also observed at δ 3.70 and 4.36 ppm, corresponding to a diastereotopic methylene AB system (ArCH
2N), which confirms the conformational restriction and proximity to a chiral or pseudoasymmetric center. The high geminal coupling constant is characteristic of protons on the same carbon atom with a rigid environment. The singlet signal at δ 3.71 ppm is assigned to methylene NCH
2N, demonstrating its relative isolation from other coupled protons. In the aromatic region, the signals between δ 7.17 and 8.84 ppm show the typical pattern of a substituted aromatic system. The most unshielded signal at δ 8.84 ppm is assigned to H-2′, which is consistent with its proximity to an electronegative atom (the nitrogen of the aromatic ring). The remaining aromatic protons (H-3′, H-4′, H-5′, and H-6′) exhibit shifts and multiplicities consistent with the observed substitution pattern.
1H NMR spectrum was recorded in DMSO-
d6 (
Figure S3). The same overall proton distribution was preserved, confirming that no structural changes occur in solution, although systematic variations in the chemical shifts were observed as a consequence of solvent effects. The aliphatic region retained the characteristic pattern of the perhydrobenzimidazole framework, with the cyclohexane methylene protons appearing as multiplets at δ 1.21–2.33 ppm. Compared with the spectrum acquired in CDCl
3, the bridgehead methine protons (H-3a/H-7a) experienced an upfield shift (2.33–2.29 ppm), reflecting the different polarity and magnetic susceptibility of DMSO-
d6. The diastereotopic benzylic methylene (ArCH
2N) remained as an AB system, appearing as two doublets at δ 4.02 and 3.67 ppm with geminal coupling constants of 13.7 and 13.5 Hz, respectively, confirming the rigid molecular framework and restricted rotation around the C–N bond. Likewise, the NCH
2N methylene resonance was observed as a singlet at 3.48 ppm, slightly upfield relative to CDCl
3. More pronounced solvent-dependent changes were found in the aromatic region. The quinoline protons adjacent to the ring nitrogen remained the most deshielded resonances (8.79–8.75 and 8.24–8.20 ppm), whereas the remaining aromatic protons resonated between 7.47 and 7.30 ppm. These differences arise primarily from the higher polarity and stronger hydrogen-bond accepting ability of DMSO-
d6 compared with CDCl
3. In the case of compound
5, DMSO-
d6 efficiently disrupts intermolecular hydrogen-bonding interactions while simultaneously acting as a hydrogen-bond acceptor for the phenolic OH groups. This change in the hydrogen-bonding equilibrium modifies the electron density distribution over the quinoline rings and the benzylic amino substituents, leading to the observed changes in chemical shifts without altering the coupling pattern or the overall spectral assignment. The phenolic OH resonance was not detected in either the CDCl
3 or DMSO-
d6 spectra. This behavior is consistent with the well-documented proton exchange processes exhibited by 8-hydroxyquinoline derivatives [
27,
28,
29]. In CDCl
3, the hydroxyl proton is involved in a dynamic equilibrium between strong intramolecular O–H···N hydrogen bonding and intermolecular hydrogen-bonding interactions, resulting in significant line broadening. In DMSO-
d6, although the solvent stabilizes the hydroxyl group through hydrogen-bond acceptance, the proton remains labile and undergoes rapid exchange with residual water and other protic impurities commonly present in the solvent. Under these conditions, the OH resonance becomes severely broadened and may disappear into the baseline, preventing its observation in the
1H NMR spectrum. Therefore, the absence of the hydroxyl signal does not indicate the lack of the OH functionality but rather reflects its fast exchange dynamics on the NMR timescale.
The
13C NMR spectrum in CDCl
3 (
Figure S4),
13C NMR spectrum in DMSO-
d6 (
Figure S5), and the DEPT-135 spectrum in DMSO-
d6 (
Figure 2c) further supported the proposed assignments. In the DEPT-135 experiment, aromatic CH signals appeared with positive phase, whereas aliphatic CH
2 resonances displayed negative phase, and quaternary carbons were not observed. The signal around 69 ppm showed positive phase, confirming the presence of the N-adjacent methine carbons C3a and C7a. The negative-phase signals around 76.5, 51.5, 28.8, and 23.7 ppm were assigned to CH
2 carbons corresponding to NCH
2N, ArCH
2N, and the cyclohexane ring, respectively. These results are consistent with the proposed structure and further support the spectroscopic assignment.
Two-dimensional NMR experiments were recorded to assign all
1H and
13C NMR signals. The multiplicity-edited
1H–
13C HSQC experiment (
Figure 2a) was fundamental in establishing the direct correlation between protons and covalently bonded carbons, thereby validating and completing the assignment of the signals observed in the
1H and
13C NMR spectra. In this two-dimensional experiment, each proton signal could be unequivocally assigned to its corresponding carbon, facilitating differentiation between methine (CH), methylene (CH
2), and methyl (CH
3) carbons. In particular, HSQC confirmed the assignment of the ArCH
2N diastereotopic system, evidenced by the correlation of the protons at δ 3.70 and 4.36 ppm with the carbon at δ 54.85 ppm, as well as the NCH
2N group (δ
H 3.71 ppm/δ
C 76.5 ppm). Furthermore, the correlations observed for aliphatic protons in the δ 1.2–2.5 ppm region with carbons between δ 24 and 29 ppm allowed for the clear identification of the carbons of the cyclohexane system (C4–C7). In the aromatic region, HSQC facilitated the precise assignment of protons H-2′, H-3′, H-4′, H-5′, and H-6′ to their respective carbons (δ
C 117–149 ppm), confirming the ring substitution pattern. The HMBC experiment (
Figure 2b) was crucial for confirming the long-range connectivities. The correlations between the ArCH
2N group and the aromatic protons (H6′) and the methylene NCH
2N demonstrate the connection between the aromatic fragment and the heterocycle. The correlations of C3a and C7a with the methylene protons of ArCH
2N support the fusion of the bicyclic system. Likewise, the correlations observed for C2 with the diastereotopic protons of ArCH
2N confirm the structural proximity between these fragments. In the aromatic system, HMBC correlations such as C5′–H4′, C3′–H2′, C4a’–H5′/H6′, and C8a’–H2′/H4′/H6′ unequivocally establish the ring substitution pattern. The correlation of C8 with H6′ and the protons of ArCH
2N is particularly noteworthy, confirming the point of attachment of the substituent to the aromatic ring. Finally, the ESI-MS+ spectrum (
Figure S7) confirmed the proposed molecular structure, as evidenced by the presence of a [M + H]
+ signal at
m/z 441.20.
The Mannich-type reaction between cyclic aminals and phenols with varying degrees of activation has been extensively studied [
22,
30]. These studies considered the amine group as an intermediate in the reaction between ethylenediamine-formaldehyde mixtures and
p-substituted phenols, a process that yields 1,3-
bis(2′-hydroxybenzyl) imidazolidine-type di-Mannich bases. The products obtained from aminal
2 belong to the same family of Mannich bases [
31] that possess the characteristic structural core of the obtained di-Mannich bases from TATD
3. Furthermore, the reaction between TATD
3 and 8-hydroxyquinoline
2 established that the formation of
N-monosubstituted imidazolides follows a reaction pathway different from that previously reported [
23]. The proposed mechanism for TATD
3 suggested that intermediate
6 undergoes an intramolecular rearrangement to give rise to
7 before the attack of a second oxine molecule. Subsequently, a regioselective cleavage of the N3′-CH
2-N1″ fragment at
7 would lead to the formation of the final products. However, according to the results presented in this work, the experimental results are more consistent with an alternative reaction pathway in which the rearrangement proposed for TATD-derived systems is disfavored. Presumably, the anular tension and the stereochemical considerations derived from the
trans-1,2-diaminocyclohexane precursor, which specifically affect the perhydroimidazolidine bicyclic system, may hinder the formation of intermediate
7, thereby directing the reaction to form intermediate
8 via aminomethylation of a second 8-hydroxyquinoline
1. Although the introduction of the quinoline nitrogen heteroatom could promote the formation of the
N-substituted imidazolidine-type mono-Mannich base, the steric constraints and conformational strain in the system do not favor this pathway (
Scheme 2). The rigid bicyclic framework generated from
trans-1,2-diaminocyclohexane restricts the conformational flexibility required for the intramolecular rearrangement previously proposed for TATD-derived systems.
3. Materials and Methods
3.1. General Information
Commercially available reagents and solvents were obtained from Merck KGaA (Darmstadt, Germany) and/or Sigma-Aldrich (Darmstadt, Germany) and were used as received without further purification. The solvents were considered to be of adequate analytical grade for the described procedures, according to the supplier specifications. Reaction monitoring and product purification were carried out by thin-layer chromatography (TLC) using silica gel 60 F254 plates (Merck KGaA) with visualization under ultraviolet irradiation at 254 nm. FT-IR spectra were obtained using a Jasco FT/IR-6600 type A spectrophotometer (JASCO Co., Ltd., Mary’s Court, PA, USA), equipped with an ATR PRO ONE accessory. To obtain NMR spectra, a Bruker Avance AV-400 MHz spectrometer (Bruker, Billerica, MA, USA) was used, with TMS as the internal standard; chemical shifts are reported as δ in ppm. Chromatographic analyses were performed using a Shimadzu Prominence ultra-fast liquid chromatography (UFLC) system (Kyoto, Japan) equipped with binary pumps, an autosampler, a column oven, and a diode-array detector (DAD/PDA). Chromatographic separation was carried out on a C18 reversed-phase column (150 × 4.6 mm, 5 µm), with a mobile phase consisting of acetonitrile: water: formic acid (1%). The total acquisition time was 30 min. Mass spectra were collected in both positive and negative electrospray ionization modes, with the positive ionization mode (ESI+) providing the most suitable response for the analyzed compounds. NMR data were processed using the MestReNova software (Santiago de Compostela, Spain).
3.2. Procedure for the Compound 5
To a solution of aminal 2 (1.00 mmol, 0.276 g) in 1,4-dioxane (3.0 mL), 8-hydroxyquinoline 1 (2.00 mmol, 0.290 g) dissolved in 1,4-dioxane (3.0 mL) was slowly added. The reaction mixture was kept at room temperature for 10 min, after which water (4.0 mL) was added, and the mixture was heated under reflux with constant stirring for 8 h. After this time, a precipitate formed, which was separated from the crude reaction product by simple filtration. Finally, the crude product was purified by recrystallization from methanol.
(rac)-(3aR,7aR/3aS,7aS)-7,7′-((hexahydro-1H-benzo[d]imidazole-1,3(2H)-diyl)bis(methylene))bis(quinolin-8-ol). Yield 27%. Yellow solid, soluble in dichloromethane, ethyl acetate, and chloroform; insoluble in water, methanol, and petroleum ether (40:60). FT-IR: 3051, 2924, 2854, 1618, 1590, 1467, 1387, 1244, 1098, 1037, 742 cm−1; 1H NMR (400 MHz, CDCl3) δ 8.84–8.80 (m, 2H), 8.04–8.01 (m, 2H), 7.36–7.34 (m, 2H), 7.17–7.22 (m, 4H), 4.36 (d, 2H, 2J = 14.0 Hz, ArCH2N), 3.71 (s, 2H), 3.70 (d, 2H, 2J = 14.0 Hz), 2.49–2.44 (m, 2H), 2.15–2.11 (m, 2H), 1.85–1.77 (m, 2H), 1.26–1.49 (m, 4H). 1H NMR (400 MHz, DMSO-d6) δ 8.79–8.75 (m, 2H), 8.24–8.20 (m, 2H), 7.47–7.43 (m, 2H), 7.35–7.30 (m, 4H), 4.02 (d, 2H, 2J = 13.7 Hz, ArCH2N), 3.67 (d, 2H, 2J = 13.5 Hz), 3.48 (s, 2H), 2.33–2.29 (m, 2H), 2.00–1.97 (m, 2H), 1.70–1.68 (m, 2H), 1.21–1.19 (m, 4H). 13C NMR (101 MHz, CDCl3) δ 152.59 (C8′), 148.91 (C2′), 139.20 (C8a′), 135.80 (C4′), 128.45 (C6′), 127.54 (C4a′), 121.40 (C3′), 118.73 (C7′), 117.59 (C5′), 76.50 (C2), 69.18 (C3a, C7a), 54.85 (NCH2Ar), 29.21 (C4, C7), 24.20 (C5, C6). 13C NMR (101 MHz, DMSO-d6) δ 150.72 (C8′), 148.17 (C2′), 138.14 (C8a′), 135.89 (C4′), 128.45 (C6′), 127.51 (C4a′), 121.41 (C3′), 120.94 (C7′), 117.01 (C5′), 76.67 (C2), 68.96 (C3a, C7a), 51.57 (NCH2Ar), 29.08 (C4, C7), 23.92 (C5, C6). ESI-MS+ [M + H] + m/z 441.20.