NMR Structural Study of the Prototropic Equilibrium in Solution of Schiff Bases as Model Compounds

An NMR titration method has been used to simultaneously measure the acid dissociation constant (pKa) and the intramolecular NHO prototropic constant ΔKNHO on a set of Schiff bases. The model compounds were synthesized from benzylamine and substituted ortho-hydroxyaldehydes, appropriately substituted with electron-donating and electron-withdrawing groups to modulate the acidity of the intramolecular NHO hydrogen bond. The structure in solution was established by 1H-, 13C- and 15N-NMR spectroscopy. The physicochemical parameters of the intramolecular NHO hydrogen bond (pKa, ΔKNHO and ΔΔG°) were obtained from 1H-NMR titration data and pH measurements. The Henderson–Hasselbalch data analysis indicated that the systems are weakly acidic, and the predominant NHO equilibrium was established using Polster–Lachmann δ-diagram analysis and Perrin model data linearization.

In these investigations several analytical methods for determining the prototropic equilibrium have been applied, such as FT-IR spectroscopy and X-ray diffraction in the solid state [45][46][47][48], as well as solution 1 H-, 13 C-and 15 N-NMR [33,34,[49][50][51]. This 1,5 tautomeric equilibrium is directly affected by the substituents [52][53][54][55][56] attached to both the phenyl group and the imine nitrogen which exert a strong influence on the acidity of the OH group, the basicity of the nitrogen atom and thus the NHO bond strength. Substituents also greatly increase the stability of the compounds by the effect of hydrogen bonding assisted by resonance (RAHB) [57][58][59]; preferences have been found in the position of the hydrogen either linked to oxygen (N···H-O) or nitrogen (N-H···O) [32,33,46,60] atoms and even being in the middle of both (O − ···H···N + ) [60,61].
For a prototropic acid-base-system HA in equilibrium, its equilibrium constant K a is expressed by Equation (1), which after logarithms becomes the Henderson-Hasselbalch Equation (2) The pK a is experimentally obtained, using the tabulation of log[(δ max − δ obs )/(δ obs − δ min )] against pH, where δ min and δ max are the chemical shifts in the inflection points in the titration curve, while δ obs is the observed chemical shift during the course of the titration, so the equilibrium point is at point zero, which corresponds to pH = pKa [44,[62][63][64][65][66][67]. This method has been extensively used because of its simplicity, however is limited by variability in pH readings and accuracy in measuring the volumes of the titrant.
Polster and Lachmann postulated the Gibbs triangle method, which later emerged as the absorbance diagram (A-diagram) or chemical shift diagram (δ-diagram), depending on the spectrometry used for the analysis of data from a titration, for the study of acid-base systems [62,68]. This method allows the evaluation of the quotient of acidity constants (ΔK a ) of two or more compounds, mainly in diprotic and polyprotic acid-base systems [68], on the bases of a ratio of distances from the Gibbs triangle which is independent of pH readings [68].
Later, Perrin et al. [69][70][71][72] also developed a mathematical model for the determination of ΔK a for mixtures of isomers in equilibrium with independency from the pH readings by drawing δ-diagrams also, so this model can be applied to the analysis of acid-base equilibrium mainly in monoprotic systems. Then for two acids HA and HB, the quotient of their acidity constants ΔK a , can be measured by the variation in chemical shifts due to changes in the acidity of the systems: Equation (4), written in terms of chemical shifts when ΔK a ≠ 1, allows the evaluation of ΔK a as the slope of a straight line, Equation (5): where δ A°, δ B° are the chemical shifts from species at the start of the titration, δ a , δ b are the chemical shifts observed during the titration, andδ HA , δ HB are the chemical shifts from species at the end of the titration.
In this contribution, both the Perrin and Polster-Lachmann models are applied to the study of intramolecular hydrogen bonds that involve prototropic equilibrium with the aim to find with accuracy and selectivity the position of the proton on the oxygen or nitrogen atoms. The model compounds were a set of Schiff base derivatives of 5-nitrosalicylaldehyde, 5-chlorosalicylaldehyde, 5-bromo-salicylaldehyde, salicylaldehyde, 5-methoxysalicyladehyde and 5-hydroxysalicyladehyde with benzylamine (compounds 1-6, Figure 1). The substituents were selected in order to cover a broad range of both electrodonating (ED) and electrowithdrawing (EW) groups whose electronic effects could modulate the NHO hydrogen bonding scheme. 1 H-NMR spectrometry was used as the titration method.

NMR Spectra
Synthesized compounds were identified by 1 H-, 13 C-and 15 N-NMR. The 1 H-NMR spectra of compounds 1-6 in DMSO-d 6 solution showed remarkable changes in the chemical shift of the acidic proton NHO in the range of 12.53-14.34 ppm, in response to the electronic character of the substituent R. Since a larger value in the chemical shift indicates a greater acidity of the proton, compound 1 has the largest acidity and compound 6 has the lowest acidity. Simultaneously the chemical shift of protons H3, H5, H6 and H7 were affected too.
The 13 C spectra of all compounds showed clear shielding and deshielding effects, according to the substituent, mainly from C1 to C7. The chemical shifts of compound 1 were more affected than those of compounds 2-6, especially the carbon atoms C1 and C4. Compound 1, the NO 2 derivative, showed a chemical shift of 175.8 and 136.9 ppm for C1 and C4, respectively, where C1 is in the range of carbonyl chemical shifts (170 to 200 ppm) while C4 is in the range of nitro Schiff base compounds (130 to 150 ppm). The 15 N chemical shift of compound 1 was −162.1 ppm, indicating an average between imine-enamine forms, therefore in this last compound the zwitterionic structure (Scheme 2a) is favored and the hydrogen H8 is localized with the nitrogen atom ( + N-H···O).

Scheme 2.
Possible resonance and equilibrium structures for compounds 1-6. In the case of compounds 2-6 the chemical shifts of C1 appear at lower frequencies from 160.4 to 153.5 ppm, a region characteristic of OH structures (150-160 ppm) and the chemical shifts of the imine C7=N appear from 165.8 to 167.4 ppm, a less significant variation. The 15 N chemical shifts for compounds 2-6 were in the range of −79.7 to −81.8 ppm (−50 to −90 ppm for imine), in agreement with a neutral N···H-O tautomeric form with the hydrogen H8 is localized with the oxygen atom (Scheme 2b). The NMR chemical shifts of compounds 1, 3, 4 and 5 have already been reported [73] and are in agreement with the above mentioned results, except for the nitro derivative 1 for which the authors conclude that the N-H tautomer is present in solution instead of the zwitterion form proposed herein.

NMR Titration
All compounds were titrated in CD 3 OD solution with NaOD, and only compound 2 was further titrated with DCl. 1 H-NMR spectra were recorded after each aliquot of titrant and simultaneously the pH was measured following each recorded spectrum. The resonances of H6 and H9 were used to plot pH vs. δ 1 H, because these protons were most affected by deprotonation of the labile hydrogen.
The pK a values obtained by the Henderson-Hasselbalch equation for all compounds were greater than 7 and less than 11, showing that these compounds are weak acids, the pK a value increases in the order NO 2 < H < Cl < Br < OMe < OH. Only compound 6 showed two pK a values, the value of 9.7 belongs to NHO and the second value of 9.8 to the phenolic hydroxyl group C4-OH. On the other hand the ΔΔG° values, associated with the prototropic NHO equilibrium, are favored in the order NO 2 > Cl > OMe  OH > Br > H. Figure 2 shows the titration curve with full pH scale (A) and the δ-diagram (B) of compounds 2 to 4; only the data region titrated with NaOD was taken for the pK a value calculation. All compounds should show the same shape of δ-diagram as they were titrated with NaOD. However, the titration curves of compounds 2, 3 and 4 showed an almost linear behavior, whereas those of compounds 1 and 5 showed one inflection point and those of compound 6 two inflection points (see Supporting Information). These results indicate that the initial structure of compounds 1-6, at the beginning of the titration, was not the same in agreement with the NMR data discussed above.
On the other hand, the prototropic 1,5-rearrangement (Scheme 3) can be envisaged as composed by two equilibria as depicted by in Figure 2. The quotient of the equilibrium constants K HN and K HO is defined as ΔK NHO , corresponding to the equilibrium constant of the prototropic 1,5-tautomerism. The chemical shifts of H6 and H9 are the most sensitive to changes in the equilibrium positions, thus they were used as probes for K HO and K HN measurements, respectively. Thus, the ΔK NHO value allows one to establish the position of the NHO equilibrium. Therefore, if the ΔK NHO value is equal to 1 then the system is in equilibrium N δ+ ···H···O δ− and both ΔpK NHO and ΔΔG° are equal to zero; if the value of ΔK NHO is higher than 1, both ΔpK NHO and ΔΔG° are less than zero and labile hydrogen is predominantly positioned on the N atom, + N-H···O; and finally if ΔK NHO is less than 1 then ΔpK NHO and ΔΔG° are greater than zero and therefore the labile hydrogen is predominantly positioned on the O atom, N···H-O.
From δ-diagrams, the mechanism occurring in the course of the titration with NaOD, can be proposed (Scheme 4).

Scheme 4.
Mechanism proposed during the titration with NaOD; the scheme is according to points in the δ-diagram of Figure 5.  (Point "B"); then, as long as the pH is increased compounds are deprotonated to become into the conjugated bases that precipitate as a salt (Point "C"). In the case of compounds 3 and 4, the initial state is at point "B" with the intramolecular hydrogen bond in the N···H-O form, the addition of NaOD aliquots only shift the equilibrium to point "C" the conjugate base.
The δ-diagrams show the initial state in all compounds and indicate the most stable species in a methanol solution, so the stability of the NHO intramolecular hydrogen bond is affected by the electronic nature of the substituent as well as solvation of methanol; therefore, the structure of compounds 1, 5 and 6 with NO 2 , OMe and OH substituents, respectively, stabilizes and direct the NHO equilibrium position by both mesomeric and inductive effects, although they have different ΔK NHO values, whereas the halogen substituent in compounds 2 and 3 exert both electronegative and inductive effects; none of such effects are present in compound 4. Thus from the obtained ΔK NHO values, the predominant NHO equilibrium in compounds 3 and 4 are the neutral N···H-O form, while for the rest of the compounds the zwitterionic + N-H···O − form is present (Scheme 2). In the particular case of compound 1, it is as an imine-enamine tautomeric form in agreement with 1 H, 13 C and 15 N pfg-HMQC spectroscopy mentioned above.
Finally, the obtained ΔK NHO values (Table 1) are very close to the equilibrium point N δ− ···H + ···O δ− (ΔK NHO = 1), which indicate a fast interchange of intramolecular hydrogen bond and the effect produced by both the substituent and the solvent that stabilize the systems in a preferred tautomeric form.

General Remarks
Schiff bases 1 to 6 were obtained by condensation of the appropriate aromatic ortho-hydroxyaldehyde with benzylamine in toluene at 25 °C (Scheme 5). Solids products were filtered and dried under a vacuum. Compound 4 was a liquid and the excess of toluene was eliminated under vacuum. 1

NMR Spectrometric Titration
The 1 H-NMR spectra were recorded in CD 3 OD on a JEOL ECA-500 spectrometer at room temperature of 295.15 ± 1 K (22 ± 1 °C). An initial 1 H-NMR spectrum of the solutions was recorded and assigned as initial value for the titration. Subsequently the solutions were titrated with aliquots of the NaOD/D 2 O solution base (3.0 μL), until invariant changes in the chemical shifts were observed; each 1 H-NMR spectrum as well as the corresponding pH reading were recorded simultaneously, after the addition of the base. Only compound 2 was further titrated with DCl (5%), to observe the behavior of the system at acidic pH. (Figures 6-11)

NMR Titration Graphics
where δ H9° and δ H6° are the chemical shifts from the species at the beginning of the titration, δ H9 and δ H6 the chemical shifts observed in the course of the titration, δ H9 e and δ H6 e are the chemical shifts from species at the end of the titration. Finally in the Polster-Lachmann analysis, the ratio of distances to calculate the ΔK NHO value is established by the graphic method described by the Gibbs triangle [62,68].

Conclusions
The study of compounds 1-6 by NMR titration in methanol solution, confirmed the predominant tautomeric forms in solution, noting that the NHO prototropic equilibrium is dependent of the substituent and the solvent. The pK a values obtained using the Henderson-Hasselbalch analysis showed that all compounds are weak acids. The strength and lability of the NHO intramolecular hydrogen bond are consequently affected by the mesomeric and inductive effects exerted by the substituents. The values of the K NHO equilibrium constant indicate that the equilibrium is slightly shifted to the nitrogen atom when the substituent in the phenyl ring exerts a strong electronic effect, either ED or EW (R = NO 2 , Cl, OMe and OH), and to the oxygen atom when Br or H in CD 3 OD solutions. Nevertheless the ΔK NHO values close to the unit, highlight that the proton is in the middle of both basic sites (O − ···H···N + ), in contrast to what is found in DMSO-d 6 solutions, where NMR data is in agreement with the neutral N···H-O tautomer for most of the compounds except for the nitro derivative which is in the zwitterion + N-H···O form. Finally, we have demonstrated the simplicity, accuracy and versatility of both the Perrin and Polster-Lachmann analysis applied to the study of intramolecular hydrogen bonds.