1. Introduction
Acridines and their homologous derivatives can be found in both natural plants and marine organisms, forming an important class of heterocycles containing nitrogen in their molecule. They have been extensively studied in terms of their synthesis [
1,
2,
3], physicochemical properties [
4,
5,
6,
7], and structural requirements [
8]. However, they have also been the subject of considerable investigations regarding their possible applications [
9,
10,
11,
12,
13].
Following their discovery, acridines were initially utilized as pigments and dyes [
14]. Subsequently, they have been extensively investigated as potential pharmaceutical agents for the treatment of a range of illnesses, including cancer [
15], Alzheimer’s disease [
16,
17,
18], bacterial [
19] and protozoan infections [
20]. Their mode of action is mainly attributed to intercalation into DNA and RNA strands through the formation of hydrogen bonds and their stacking between base pairs, resulting in DNA crosslinks and strand breaks [
20]. These bonds have subsequent effects on biological processes related to DNA and its associated enzymes [
21,
22]. Tetrahydroacridines, the reduced homologs of acridines, have received increased attention due to their ability to inhibit topoisomerase enzymes and block DNA transcription [
23,
24,
25,
26,
27]. They have been widely explored for the treatment of Alzheimer’s disease [
28,
29,
30,
31,
32,
33,
34,
35,
36,
37,
38], human cancer, [
39,
40,
41], and tuberculosis [
42]. Furthermore, new natural and synthetic acridine derivatives have been tested for antimalarial and antileishmanial [
43,
44,
45,
46,
47], antifungal [
48,
49], antiparasitic [
50], anti-inflammatory [
51,
52,
53,
54,
55,
56,
57,
58,
59,
60], and analgesic [
61,
62,
63] activities, with some derivatives approved for chemotherapy [
64]. Additionally, through sustained efforts, acridine derivatives with increased therapeutic potency and selectivity have been developed, such as fluorescent materials [
65,
66,
67]. The presence of an aromatic ring, a π-system, and a nitrogen atom with an available electron pair in the chemical structure of acridines and tetrahydroacridines renders them suitable as corrosion inhibitors [
68] and metal chemosensors [
69,
70,
71].
Similarly to other π-conjugated organic molecules, acridines have been employed in several other leading fields due to their strong electron-donating capacity [
72] and remarkable optoelectronic properties [
73,
74,
75,
76,
77]. These include organic electronics [
78,
79], as well as the field of organic light-emitting diodes [
80,
81,
82,
83]. In addition, their partially hydrogenated analogs have been used as electron donor groups for OLED applications [
84,
85].
The present paper presents the results of the studies initiated in our group on the synthesis, electrochemical, and optical properties of tetrahydroacridine dimers 7,7′-(ethane-1,2-diyl)
bis(2,3-dihydro-1H-cyclopenta[b]quinoline-9-carboxylic acid) and 7,7′-(ethane-1,2-diyl)
bis(1,2,3,4-tetrahydroacridine-9-carboxylic). This expands the range of tetrahydroacridines explored so far by our research group [
86,
87]. The study led to the extension of their useful properties for new applications, complementing the information on the optoelectrical properties of the above tetrahydroacridines.
3. Discussion
Synthesis of bis-acids7,7′-(ethane-1,2-diyl)bis(2,3-dihydro-1H-cyclopenta[b]quino-line-9-carboxylic acid) (4a) and 7,7′-(ethane-1,2-diyl)bis(1,2,3,4-tetrahydroacridine-9-carboxylic acid) (4b) was performed in three steps consisting of a Sandmayer reaction (step I) followed by a cyclization in acidic medium with the formation of bis-indoline tetraone (step II), and a Pfitzinger cyclization with a cyclanone (step III). Stage I proceeded with lower yields than in the case of the corresponding monomers (30%), due to the intermediate formation of hydrochloride at one of the amino groups. However, the synthesis of the corresponding intermediate succeeded, and it was transformed during stage II with good yields (82%) and stage III with moderate yields (41%).
The characterization of compounds involved in the synthesis of the two new tetrahydroacridine dimers 4a and 4b has proven the formation of title compounds. The 1H-NMR and 13C-NMR spectra were indicative of the compounds’ formation. The dimers have low solubility in common organic solvents. Their melting points are above 240 °C with decomposition. The final compounds were also characterized by their UV–Vis, FT-IR, and MS spectra. Their structures were calculated using DFT methods and their electrochemical behavior was examined.
The 1H-NMR spectra of tetrahydroacridine bis-acids, 4a–b, show the two distinct sub-spectra of the aromatic ring and the saturated ring. Following cyclocondensation of bis-isatines (3) with cyclanones to bis-tacrinacids, aromatic protons retain their multiplicity, but their chemical shifts are significantly modified, resonating at lower values of the field due to the deshielding produced by the current ring, the electronic effect of the quinoline nitrogen atom and the carboxylic group. The most deshielded are the vicinal protons with the quinoline nitrogen (H-5, H-5′), which register a deshielding of approximately δ = 0.8–1 ppm and appear as a doublet at δ = 7.81–8.14 ppm. The protons with the same multiplicity (H-8, H-8′) appear as a doublet at δ = 7.52–7.86 ppm. The protons H-6 and H-6′ are more shielded and resonate as a doublet of a doublet at δ = 7.62–8.02 ppm. The deshielding effect of the carboxylic group on the methylene groups of the saturated cycles is also observed, with these groups resonating in the range of δ = 3.01–3.32 ppm. The most deshielded of these groups are those placed in the adjacent position with the acridinic nitrogen, followed by those placed in the neighboring position of the carboxylic group. Following cyclocondensation, the protons of the ethylene bridge, -CH2-CH2-, are also deshielded and undergo a chemical displacement of approximately 0.3 ppm, reaching δ = 3.1–3.2 ppm in bis-acids.
In the 13C-NMR spectra of the tetrahydroacridine bis-acids, the most deshielded signals are assigned to the carboxylic groups that resonate in the range δ = 165.80–166.47 ppm and to the vicinal carbons found in the γ position by the acridinic nitrogen C9, C9′, which resonate in the field δ = 157.91–164 ppm. In contrast to bis-isatines, the tertiary carbon atoms exhibit minimal displacement. The saturated cycles of bis-tacrinacids manifest as methylene group signals at δ = 25–30 ppm, with the most distorted ones being those nearest to the carboxyl groups. For the ethylene group, the chemical shifts are maintained within the range δ = 37–38 ppm.
The
UV spectra of
bis-acids present 3 bands of maximum absorption corresponding to a π → π* transition at 268/270 nm and corresponding to a n → π* at 311 nm and 325/326 nm in DMF solutions for
4a/
4b compounds, respectively (
Figures S1 and S2). Their absorbances depend on concentration linearly (
Table S1). From the comparison of their UV–Vis spectra,
4a is almost two times more active in absorption than
4b (
Figures S1 and S2) and the slope of absorbance confirms these facts (
Table S1).
The
mass spectra (MS) played an important role in elucidating the structure of the new compounds. The molecular peaks of the ions formed by fragmentation are indicated for each compound in
Scheme 2. Ionization of the ethylene
bis-acids presented in this study, as well as other
bis-amides/
bis-acids previously presented [
91], occurs with the formation of double ions of the [M + 2H]
2+ type having
m/
z values corresponding to half of the value for the dimer structure, and an abundance much lower than that of the protonated molecular ions [M + H]
+. Further, the double ions [M + 2H]
2+ lose functional groups or are fragmented at the ethylene bridge with the appearance of ions of half molecule type, [M/2]
+. The acids are also ionized in negative mode using electrospray, when a solvent slightly basic is used, such as methanol/ water with 0.5% ammonium carbonate. In positive mode, ionization of ethylene
bis-polymethylenquinolin
bis-acids occurs easily by the successive elimination of one carbon dioxide molecule.
The IR (solid ATR) spectra given for each newly synthesized compound highlight the characteristic functional groups and skeleton frequencies. In the precursors (bis-isonitrosoacetanilide and bis-isatine) the amidic groups show for the NH group, a stretching vibration ν(NH) in the field 3100–3350 cm−1 and a deformation vibration δ(NH) at 1530–1580 cm−1. In the carbonyl and azomethyin groups, the ν(C=O) and ν(C=N) vibrations within the range 1600–1680 cm−1 were noticed. In the diacids, the carboxyl groups are recognized both by the frequency of ν(CO) at 1695–1700 cm−1 as well as the frequency in the 2200–2800 cm−1 domains, which appear in the case of the acid dimer associations. The δ(C=N) group from the quinoline nucleus appears in the range 1600–1610 cm−1. The aliphatic methylene groups show symmetric and asymmetric ¥(CH) vibrations in the range 2840–2950 cm−1 and valence angle deformations ¥ (CH2) in the range 1300–1425 cm−1, while ν(=CH) frequencies appear over 300–3150 cm−1.
Computation analyses play a great role in discerning the reactivity and stability of compounds. The energy of FMO (HOMO and LUMO) given in
Table 1 allowed the calculation of the energy gap (Eg) value from the difference between the HOMO and LUMO energy levels. The results indicate that compound
4a has the lowest Eg value of 4.1834 eV compared to compound
4b with a higher Eg value of 4.3832 eV. This indicates that compound
4a is less stable and more reactive than compound
4b.
NBO analysis conducted on the investigated compounds obtains insight into the nature of interactions existing between the donor orbital and acceptor orbital, which is expressed in terms of second-order perturbation energy (E
(2)) or stabilization energy (E
(2)) [
99]. The higher the stabilization energy the stronger the interaction between the donor orbital and the acceptor orbital. The second-order perturbation energy of the studied compounds was estimated using Equation (1) [
100]:
In (1)
qi is the donor orbital occupancy,
εi and ε
j indicates the diagonal elements and F
(i,j) represents the Fock matrix element. Evaluated second-order perturbation energy of compounds
4a and
4b displayed in
Table 2 and
Table 3, respectively, shows that the most persisting form of donor–acceptor orbital interaction resulting from intermolecular hyperconjugation existing within the studied compounds is that of bonding interaction of which that of compound
4a is primarily σ → σ* and π → σ* while that of compound
4b is σ → σ* bonding interaction. From the displayed second-order perturbation energy E
(2) of the studied compounds, it is obvious that compound
4a exhibited higher E
(2) values from the donor–acceptor interaction compared to compound
4b. As seen from the tables, the highest donor–orbital interaction of compound
4a was from donor (σC
42–H
50) to acceptor (σ*C
37–H
43) intermolecular hyperconjugation with stabilization energy of 75,640.17 kcal/mol while that of compound
4b is from donor (σC
32–O
33) to acceptor (σ*C
42–C
43) with stabilization energy of 8691.80 kcal/mol. In this sense, we can infer that compound
4a is more reactive than compound
4b since the E
(2) value resulting from intermolecular hyperconjugation prevailing within the studied compound aids in defining the nature of the reactivity of the compounds. This result is also validated by the frontier molecular orbital (FMO) analysis of the studied compounds.
Fukui function used to describe the change in electron density of a particular atom within a molecule with respect to the change in the number of electrons present provided better information on the nature of the reaction taking place at a preferential site of an atom within these molecules. The condensed Fukui function of the studied compounds at N + 1, N − 1, N electron system with its condensed dual descriptor was computed using Equations (2), (3), (4) and (5), respectively.
In (2)–(5) f+, f−, f0 are condensed Fukui functions for nucleophilic attack, electrophilic attack, free radical attack, respectively, and qN+1, qN−1, qN represents the atomic charges of the atomic sites at N + 1, N − 1, N electron system, respectively. ∆f is the condensed dual descriptor (CDD). From Equation (5), if ∆f > 0 the atomic site is prone to nucleophilic attack, and if ∆f < 0 the atomic site is preferential for electrophilic attack. For compound 4a, the order of electrophilic attack at the preferential atomic sites was found to be C4 > C9 > C12 > C1> C3 > C31 > C6 > C13 > C7 = C8 while the point of nucleophilic attack was observed on the following atoms with trend as follows: C24 > C22 > N15 > N14 > O25 > O23 > C20 > C18 > C19 > C17. Atomic sites C7 and C8 with equivalent condensed dual descriptors for electrophilicity indicate that electrophilic attacks can take place at the same rate at their atomic site. Also, for compound 4b the rate of electrophilic attack at the preferential site is observed to be of the trend C4 > C9 > C1 > C12 > C3 > C33 > C6 > C7 > C8 > C13, while the atomic sites prone to nucleophilic attack are of the trend C24 > C22 > N15 > N14 > O25 > C19 > C17 > O23 > C18 > C20. In general, we noticed that atomic sites C4 and C24 are the most preferential sites susceptible to electrophilic and nucleophilic reactions, respectively, for both compounds 4a and 4b in the presence of a chemical reagent. In addition, compound 4a exhibited higher values of ∆f or CDD values for electrophilic and nucleophilic reactions at the preferential atomic sites compared to compound 4b. This result implies that compound 4a is more reactive than compound 4b, which is in excellent agreement with the frontier molecular orbital and natural bond orbital (NBO) analysis result.
Analysis of the electrochemical behavior of the dimers 4a and 4b was based on DPV, CV, and RDE curves recorded in different conditions.
In the case of
4a, from
Figure 7 relating to different concentrations of
4a the anodic and cathodic currents (continuous lines) are higher than those for the supporting electrolyte (dashed lines) and increase with the concentration of
4a. This shows that
4a undergoes oxidation and electrochemical reduction processes. There are two main anodic processes denoted by a1 and a2 (
Figure 7a). In the cathodic DPV scans, four peaks were noticed (denoted c1–c4 in
Figure 7a). From
Figure 7b, processes a1 and a2 in DPV do not correspond to a wave in RDE at any rotation rate of the electrode. On the other hand, in the cathodic domain, it is clearly observed that the RDE curve has limiting currents that increase with the rotation rate of the electrode. They have been studied in detail and will be discussed further. The parallel presentation of the DPV curves and the RDE curves obtained at different concentrations of
4a at a rotation rate of 1000 rpm and the RDE curves obtained for the concentration of 1 mM and 2 mM at different rotation rates of the electrode illustrate the difference between the information provided by these methods (anodic processes appear only in DPV, while cathodic ones are revealed by both methods).
In
Figure 8a, the CV curves are given on various potential domains, anodic and cathodic, which allow the evaluation of the reversibility of the highlighted electrochemical processes, which take place at different potentials. It is observed that anodic processes are irreversible, having no correspondence in the return sweep. These oxidation processes can lead to the formation of films on the surface of the electrode and have been further analyzed because they occur relatively easily (at moderate anodic potentials close to 0.6 and 0.9 V, potentials at which peaks were recorded on the DPV curves). The presence in the CV curve of the processes c2′, c3′, and c4′ highlighted at the return sweep after the cathodic scan shows that they are quasi-reversible (
Figure 8a). This behavior was verified by recording CV curves at different sweep rates (between 0.05 and 0.5 V/s) in several successive cycles for each scan rate, as in
Figure 8b.
For the first cathodic scans (
Figure 9a), the values of the peak currents in the c3 process domain were represented as a function of the scan rate (I vs. v) or the square root of the scan rate (I vs. v
1/2) (
Figure 9b). The correlation of the current intensity with v
1/2 is better, justifying a possible diffusive process in reduction.
Analyzing the anodic processes (
Figure 10), it is observed that in the a1 process domain, the CV curves decrease in successive cycles, regardless of the scan rate (
Figure 10a).
Figure 10b shows the corresponding CV curves obtained in the first cycle at various scan rates. For a potential of 0.3V corresponding to the beginning of the a1 process, it is observed (
Figure 10c) that the current value correlates better with the scan rate (R = 0.98) than with the square root of the scan rate (R = 0.96), supporting a process different from the purely diffusive one. A similar analysis (
Figure 11) was made for the study of the dependence of the current on the scan rate at the potential of 0.6 V, at the beginning of the process denoted by a2. In this case, the linear correlations are slightly better for the dependence of the current on the scan rate (R = 0.999) than that on the square root of the scan rate (R = 0.993). The processes a1 and a2 could be related to film formation by electrochemical oxidation of
4a. To check this hypothesis, the electrode was transferred into a ferrocene solution (Fc/Fc
+) in 0.1 M TPAP in ACN and the CV curves were compared with those on the bare electrode.
Figure 12 shows the curves obtained after the transfer in the Fc/Fc
+ solution of the electrodes modified by cycling at different scan rates. The cycling was completed in the
4a solution of 1mM concentration, in the process domain a1 (a) or a2 (b). The dashed line represents the curve obtained for Fc/Fc
+ on the bare electrode. From the examination of the curves, it is noted that the potential of the Fc/Fc+ system does not change. The increase in the peak currents for process a1 could stand for the formation of a conductive film on the surface. The current variation is insignificant in the case of process a2. This may stand for different modified electrodes obtained by cycling in a2 domain than in a1.
The electrochemical behavior of
4b is illustrated in
Figure 13a, where the CV and DPV curves for the
4b solutions are shown in parallel at different concentrations, evidencing by both methods that
4b undergoes electrochemical oxidation and reduction processes. In
Figure 13b, DPV curves are compared with RDE curves (obtained at different rotation rates for the highest concentration studied) to make the connection between the processes evidenced by these methods. The DPV peak potentials are given in
Table 5. There are three main anodic processes denoted with a1–a3 and three cathodic peaks (denoted c1–c3) in
Figure 13a. From
Figure 13b, it is seen that the processes a1-a3 placed in evidence by CV and DPV do not correspond to any RDE wave at all RDE rotation rates. On the other hand, in the cathodic range, two RDE waves (denoted cw1 and cw3) are clearly noticed, having limiting currents that increase with the rotation rate of the electrode. A special behavior is seen for the c2 peak where the current decreases towards the value of the background current at which all the curves intersect. This behavior could only be explained by the formation of a non-conductive film at this potential.
In
Figure 14a, the CV curves for
4b are given over different anodic and cathodic potential domains, to check the reversibility of the electrochemical processes highlighted for
4b. All anodic and cathodic processes are irreversible for compound
4b, having no corresponding peaks in the return sweeps. This irreversibility may be due to oxidation or reduction processes with the formation of films on the electrode surface. This behavior was tested by recording CV curves at different scan rates (between 0.05 and 0.5 V/s) (
Figure 14b). The variation of the anodic and cathodic currents with the scan rate and the square root of the scan rate, respectively, was examined for
4b. The slopes of the correlation lines and the linear correlation coefficients are displayed in
Table 6. For the correlations examined for processes a1-a3 and c1, it is seen that the dependence of the current on the square root of the scan rate has a better correlation coefficient than the dependence on the scan rate, which may imply a diffusive behavior at the potentials examined.
Figure 15 shows the RDE curves for different electrode rotation rates at three concentrations of
4b (0.5 mM, 1 mM, 2 mM). The obtained curves support the hypothesis of film formation at the potential of the process c2. At this highest concentration value (2 mM) and at the highest rotation rate (1500 rpm) the most obvious current decrease occurs for the c2 process. This can be attributed to a reduction process with film formation.