Polycondensation Resins by Flavonoid Tannins Reaction with Amines

Reaction of a condensed flavonoid tannin, namely mimosa tannin extract with a hexamethylene diamine, has been investigated. For that purpose, catechin was also used as a flavonoid model compound and treated in similar conditions. Solid-state cross-polarisation/magic-angle spinning (CP-MAS) carbon 13 nuclear magnetic resonance (13C NMR) and matrix assisted laser desorption ionisation time of flight (MALDI-ToF) mass spectroscopy studies revealed that polycondensation compounds leading to resins were obtained by the reaction of the amines with the phenolic hydroxy groups of the tannin. Simultaneously, a second reaction leading to the formation of ionic bonds between the two groups occurred. These new reactions have been shown to clearly lead to the reaction of several phenolic hydroxyl groups, and flavonoid unit oligomerisation, to form hardened resins.


Introduction
Condensed polyflavonoid tannin extracts are mostly composed of flavan-3-ols repeating units and smaller fractions of polysaccharides and simple sugars [1]. The repeating units are linked to each other by C4-C6 or C4-C8, the former predominating in tannins in which fisetinidin (resorcinol A-ring; catechol B-ring) and robinetinidin (resorcinol A-ring; pyrogallol B-ring) are the predominant repeating units. While the reactions of these natural oligomeric materials have been used extensively to give polycondensates with aldehydes [2], even reactions of self-condensation have been studied and shown to lead to useful physically and chemically crosslinked networks [3][4][5][6][7].
Reactions of amination of phenols are well known, with the original approach to this reaction being by metal catalysis [8,9]. More recently, direct amination of phenols without the use of a metal catalyst has come to the fore and gained interest and importance [10,11].
Reactions of amination of flavonoid tannins to convert part of the phenolic hydroxyl groups of the B-ring to -NH 2 have been studied before [12][13][14]; they have been exclusively with ammonia, although the literature on this is limited to just three articles [12][13][14]. There appear to be no references on the reaction of amines, diamines, or polyamines with condensed tannins in the relevant literature. The oldest of the articles on the ammonia reaction with flavonoid tannins was aimed to even more efficiently bind formaldehyde gas emitted from tannin adhesive resins for wood panels [12]. In this study, amination of pyrogallol B-rings of condensed tannins to form 4′-amino-3′,5′dyhydroxybenzene type B-rings by NH3 treatment was described. The amination of the pyrogallol Bring by NH3/water is a regioselective amino-substitution of phenolic hydroxyl groups and proceeds under relatively mild conditions without a catalyst [12][13][14]. While early reports indicated that only one hydroxy group of the flavonoid B-ring is aminated [12], later work showed that multiamination also proceeds with relative ease [14], mainly to prepare carbonized materials richer in nitrogen. These amination reactions, however, did not appear to lead to long oligomers and finally to crosslinked resins. To obtain then polycondensation resins without the use of any aldehyde, the reaction of condensed tannins with a diamine were investigated in the work presented here. The aim of this work was to obtain thermoset resins having a very rapid initial gelling either (i) by using only amines and polyamines as the only hardener; or (ii) to use two hardeners, of which the amine was the one giving just the initial immobilization of the resin-for example, for spray-projected coatings to avoid initial running down on vertical walls.
The reactions of diamines with condensed tannin to form resins were investigated here, using first catechin as a flavonoid model compound, followed by the same reactions on a condensed tannin analysed by extensive MALDI-ToF spectroscopy and solid-state cross-polarisation/magic-angle spinning (CP-MAS) 13 C NMR studies. The findings are presented in this article.

Catechin
Acacia mearnsii (De Wild) predominantly prorobinetinidine From these two compounds, the following experiments have been carried out. The samples were prepared as follow: (1) Catechin (0.5 g) was mixed with 0.5 g of hexamethylenediamine (HMDA) (70% solution in water). Three samples were prepared with the proportions above. Then, each sample was reacted in an oven at 65, 100, and 185 °C overnight, respectively. (2) Catechin (0.5 g) was mixed with 0.5 g of HMDA (70% solution in water) and 0.15 g of a 65 wt % aqueous solution of p-toluenesulfonic acid (pTSA). Again, three samples were prepared with the proportions above, and they were reacted in an oven at 65, 100, and 185 °C overnight, respectively. From these two compounds, the following experiments have been carried out. The samples were prepared as follow: (1) Catechin (0.5 g) was mixed with 0.5 g of hexamethylenediamine (HMDA) (70% solution in water).
Three samples were prepared with the proportions above. Then, each sample was reacted in an oven at 65, 100, and 185 • C overnight, respectively. (2) Catechin (0.5 g) was mixed with 0.5 g of HMDA (70% solution in water) and 0.15 g of a 65 wt % aqueous solution of p-toluenesulfonic acid (pTSA). Again, three samples were prepared with the proportions above, and they were reacted in an oven at 65, 100, and 185 • C overnight, respectively.
(3) Catechin (0.5 g) was mixed with 0.5 g of HMDA (70% solution in water) and 0.15 g of a 33 wt % aqueous solution of NaOH. Three samples were prepared with the proportions above. After that, they were reacted in an oven at 65, 100, and 185 • C overnight, respectively. (4) Mimosa tannin (2 g) was mixed with 2 g of hexamethylenediamine (HMDA) (70% solution in water). Three samples were prepared with the proportions above, and they were reacted in an oven at 65, 100, and 185 • C overnight, respectively. (5) Mimosa tannin (2 g) was mixed with 2 g of HMDA (70% solution in water) and 0.6 g of a 65 wt % aqueous solution p-toluenesulfonic acid (pTSA). Three samples were prepared with the proportions above. Then, they were reacted in an oven at 65, 100, and 185 • C overnight, respectively. (6) Mimosa tannin (2 g) was mixed with 2 g of HMDA (70% solution in water) and 0.6 g of a 33 wt % aqueous solution NaOH. Again, three samples were prepared with the proportions above, and they were reacted in an oven at 65, 100, and 185 • C overnight, respectively.
The samples were mixed with a spatula because they become a paste after the addition of catechin or mimosa tannin. After the reaction in the oven, the samples prepared at 100 and 185 • C become a dry solid, while the samples prepared at 65 • C remained like a paste.
In the case of the mimosa tannin samples at 185 • C, they have not been analysed by MALDI because the spectra were not good enough due to the difficulty of their solubility in the acetone-water solution for their MALDI-ToF analysis.

Matrix-Assisted Laser Desorption Ionisation Time-of-Flight (MALDI-ToF) Mass Spectrometry Analysis
The spectra were recorded on a KRATOS Kompact MALDI AXIMA TOF 2 instrument (KRATOS Ana lytical, Shimadzu Europe Ltd., Manchester, UK). The irradiation source was a pulsed nitrogen laser with a wavelength of 337 nm. The time period of a laser pulse was 3 ns. The measurements were carried out using the following conditions: polarity = positive, flight path = linear, mass = high (20 kV acceleration voltage), 100-150 pulses per spectrum. The delayed extraction technique was used by applying delay times of 200-800 ns.

CP-MAS 13 C NMR
Solid-state CP-MAS (cross-polarisation/magic-angle spinning) 13 C NMR spectra of the aforementioned oven-dried solids were recorded on a Brüker MSL 300 spectrometer (Brüker France, Wissembourg, France) at a frequency of 75.47 MHz. Chemical shifts were calculated relative to tetramethyl silane (TMS). The rotor was spun at 4 kHz on a double-bearing 7 mm Bruker probe. The spectra were acquired with 5 s recycle delays, a 90 • pulse of 5 s and a contact time of 1 ms. The number of transients was 3000.

MALDI-ToF
While some of the peaks obtained by MALDI-ToF in the products obtained by reactions at 185 • C are the same as those in the cases at 100 • C, a greater number of different types of compounds are observed in the reactions at 100 • C. For the MALDI-ToF analysis of the reactions of catechin as a model compound, the spectra of the NaOH-catalysed reaction will be discussed, as the peaks are practically the same for the acid-catalysed and uncatalysed cases, the main differences being their relevant proportions. Two types of reactions appear to occur from the calculation of the MALDI masses found, namely (i) the formation of secondary amines by reaction of the hexamethylene diamine on the -OH groups of the flavonoid units; and (ii) the formation of -O − + NH 3 -salts between the amino group and some of the phenolic -OH groups of the tannin flavonoids.
Thus, in the spectra obtained, the main peaks observed are reported in Table 1, and Figures 2 and 3. There appears to be a clear period of 40 Da. This is a diamine with 2Na + (not an unusual occurrence), thus 116 + 23 + 23 − 2 = 160 Da, giving the 160/4 = 40 Da period. The series of peaks that appears is then 798-758-718-678-634-594-553(small)-513 Da. From this series, for example, the repetition of peaks follows a 160 Da period such as 513 + 160 = 673 Da (678 Da), 594 + 160 = 754 Da (758 Da) (this is salt 4 × 117 = 755 Da), and 638 + 160 = 798 Da = 755 + 2Na + = 801 − 2H + = 799 Da.     The structures of the type of compounds more characteristic that formed (see Table 1) are thus as follows. At 509-512 Da (Figure 4), where the bonds formed are covalent.    The structures of the type of compounds more characteristic that formed (see Table 1) are thus as follows. At 509-512 Da (Figure 4), where the bonds formed are covalent. The structures of the type of compounds more characteristic that formed (see Table 1) are thus as follows. At 509-512 Da (Figure 4), where the bonds formed are covalent.    The structures of the type of compounds more characteristic that formed (see Table 1    The structures of the type of compounds more characteristic that formed (see Table 1    It must be pointed out that the structure shown above is the most likely, rather than the Na + being attached to the N of the amine. This is so because, in general, a strong base is needed to abstract a proton from an amine. The NaOH used in the catalysis of the reaction is not strong enough for this. Furthermore, such sodium amides are strong bases themselves, which are not likely to coexist in the presence of protic compounds such as phenols. It is then most likely that the 548 Da peak belongs to a molecule that is the sodium salt of the phenolate ion. This is equally valid for structures such as the 564 Da peak observed for the mimosa tannin and other structures where Na + is present. ( Table 2).
Dimers of two catechin monomers linked covalently through an hexamethylenediamine occurs, such as the peak occurring at 664 Da ( Figure 7): Polymers 2017, 9,37 6 of 17 It must be pointed out that the structure shown above is the most likely, rather than the Na + being attached to the N of the amine. This is so because, in general, a strong base is needed to abstract a proton from an amine. The NaOH used in the catalysis of the reaction is not strong enough for this. Furthermore, such sodium amides are strong bases themselves, which are not likely to coexist in the presence of protic compounds such as phenols. It is then most likely that the 548 Da peak belongs to a molecule that is the sodium salt of the phenolate ion. This is equally valid for structures such as the 564 Da peak observed for the mimosa tannin and other structures where Na + is present. ( Table 2).
Dimers of two catechin monomers linked covalently through an hexamethylenediamine occurs, such as the peak occurring at 664 Da (  This can, however, also be interpreted as 638 + 1 × Na + = 661 Da, thus an ionic salt such as ( Figure 8): Polymers 2017, 9,37 6 of 17 It must be pointed out that the structure shown above is the most likely, rather than the Na + being attached to the N of the amine. This is so because, in general, a strong base is needed to abstract a proton from an amine. The NaOH used in the catalysis of the reaction is not strong enough for this. Furthermore, such sodium amides are strong bases themselves, which are not likely to coexist in the presence of protic compounds such as phenols. It is then most likely that the 548 Da peak belongs to a molecule that is the sodium salt of the phenolate ion. This is equally valid for structures such as the 564 Da peak observed for the mimosa tannin and other structures where Na + is present. ( Table 2).
Dimers of two catechin monomers linked covalently through an hexamethylenediamine occurs, such as the peak occurring at 664 Da (  It must be pointed out that the structure shown above is the most likely, rather than the Na + being attached to the N of the amine. This is so because, in general, a strong base is needed to abstract a proton from an amine. The NaOH used in the catalysis of the reaction is not strong enough for this. Furthermore, such sodium amides are strong bases themselves, which are not likely to coexist in the presence of protic compounds such as phenols. It is then most likely that the 548 Da peak belongs to a molecule that is the sodium salt of the phenolate ion. This is equally valid for structures such as the 564 Da peak observed for the mimosa tannin and other structures where Na + is present. ( Table 2).
Dimers of two catechin monomers linked covalently through an hexamethylenediamine occurs, such as the peak occurring at 664 Da (  It must be pointed out that the structure shown above is the most likely, rather than the Na + being attached to the N of the amine. This is so because, in general, a strong base is needed to abstract a proton from an amine. The NaOH used in the catalysis of the reaction is not strong enough for this. Furthermore, such sodium amides are strong bases themselves, which are not likely to coexist in the presence of protic compounds such as phenols. It is then most likely that the 548 Da peak belongs to a molecule that is the sodium salt of the phenolate ion. This is equally valid for structures such as the 564 Da peak observed for the mimosa tannin and other structures where Na + is present. ( Table 2).
Dimers of two catechin monomers linked covalently through an hexamethylenediamine occurs, such as the peak occurring at 664 Da (  Higher oligomers in which catechin has dimerised also occurs, this being a fairly common reaction [16,17]. In these, the catechin dimer is linked covalently to either a catechin monomer or another catechin dimer, such as those shown by the peaks for the oligomers at 1169 Da = 1145 + 1 × Na + (  Higher oligomers in which catechin has dimerised also occurs, this being a fairly common reaction [16,17]. In these, the catechin dimer is linked covalently to either a catechin monomer or another catechin dimer, such as those shown by the peaks for the oligomers at 1169 Da = 1145 + 1 × Na + (Figure 2  Higher oligomers in which catechin has dimerised also occurs, this being a fairly common reaction [16,17]. In these, the catechin dimer is linked covalently to either a catechin monomer or another catechin dimer, such as those shown by the peaks for the oligomers at 1169 Da = 1145 + 1 × Na + ( Figure 2 Higher oligomers in which catechin has dimerised also occurs, this being a fairly common reaction [16,17]. In these, the catechin dimer is linked covalently to either a catechin monomer or another catechin dimer, such as those shown by the peaks for the oligomers at 1169 Da = 1145 + 1 × Na + (Figure 2  Higher oligomers in which catechin has dimerised also occurs, this being a fairly common reaction [16,17]. In these, the catechin dimer is linked covalently to either a catechin monomer or another catechin dimer, such as those shown by the peaks for the oligomers at 1169 Da = 1145 + 1 × Na + (Figure 2  It must be made clear that the structures above are not the only possible isomers deduced from the peaks of the MALDI-ToF spectra, but that other isomer possibilities do exist for them. The existence of different isomers becomes clearer and is also confirmed later by the CP-MAS 13 C NMR analysis.

CP-MAS 13 C NMR
In regard to the NMR spectra: the reactions occurring appear to be more advanced when the temperature is higher, while the reaction appears almost not to occur at the lower temperature of 65 °C. For this reason, the case of the reaction of catechin as a model compound with hexamethylene diamine catalysed by pTSA at 185 °C will be discussed first. The corresponding CP-MAS 13 C NMR spectrum is shown in Figure 4.
In Figure 16, first of all, the aliphatic carbon in position alpha to an -NH of the diamine reacted covalently with the tannin must have a shift of 43-44 ppm, while the same for an aliphatic amine not reacted should have a calculated shift of 41-42 ppm. Looking at the spectra of the 185 °C reactions, either pTSA-and NaOH-catalysed or uncatalysed (one reported in Figure 4, the others reported in the Supplementary Material), it can be noticed that the shift is at 42.9 ppm (uncatalysed), 43.2 ppm (NaOH-catalysed), and 43.5 ppm (pTSA-catalysed) indicating that at 185 °C, the amine has reacted covalently. This is confirmed by other indications. The shift for the C in β of the covalently reacted diamine should be at 30 ppm, while the unreacted one should be at 33-34 ppm. This peak is not visible at all in uncatalysed and pTSA-catalysed, as it is covered totally by the huge peak at 27-28 ppm, but appears as a slight shoulder at 33 ppm for the NaOH-catalysed case. It must be clearly pointed out that the shift of the top of the 43-44 ppm wide peak clearly changes when comparing the CP-MAS 13 C NMR spectra of the three cases when the reaction is carried out at 100 °C (all spectra reported in the Supplementary Material). Thus, they are respectively at 41.9 ppm (pTSA-catalysed), 42.6 ppm (NaOH-catalysed), and 42.3 ppm. This indicates that, just based on NMR evidence, the type of bonds obtained in the reaction at 100 °C appears to be more uncertain, or at least that ionic bonds and covalent bonds are in different proportions according to the presence of different catalysts, or their absence. This implies that at the higher temperature of 185 °C, the reaction shift more towards the formation of covalently bound amines, the most found in pTSA catalysis, followed by NaOH catalysis, and least in the uncatalysed case. It must be made clear that the structures above are not the only possible isomers deduced from the peaks of the MALDI-ToF spectra, but that other isomer possibilities do exist for them. The existence of different isomers becomes clearer and is also confirmed later by the CP-MAS 13 C NMR analysis.

CP-MAS 13 C NMR
In regard to the NMR spectra: the reactions occurring appear to be more advanced when the temperature is higher, while the reaction appears almost not to occur at the lower temperature of 65 • C. For this reason, the case of the reaction of catechin as a model compound with hexamethylene diamine catalysed by pTSA at 185 • C will be discussed first. The corresponding CP-MAS 13 C NMR spectrum is shown in Figure 4.
In Figure 16, first of all, the aliphatic carbon in position alpha to an -NH of the diamine reacted covalently with the tannin must have a shift of 43-44 ppm, while the same for an aliphatic amine not reacted should have a calculated shift of 41-42 ppm. Looking at the spectra of the 185 • C reactions, either pTSA-and NaOH-catalysed or uncatalysed (one reported in Figure 4, the others reported in the Supplementary Material), it can be noticed that the shift is at 42.9 ppm (uncatalysed), 43.2 ppm (NaOH-catalysed), and 43.5 ppm (pTSA-catalysed) indicating that at 185 • C, the amine has reacted covalently. This is confirmed by other indications. The shift for the C in β of the covalently reacted diamine should be at 30 ppm, while the unreacted one should be at 33-34 ppm. This peak is not visible at all in uncatalysed and pTSA-catalysed, as it is covered totally by the huge peak at 27-28 ppm, but appears as a slight shoulder at 33 ppm for the NaOH-catalysed case. It must be clearly pointed out that the shift of the top of the 43-44 ppm wide peak clearly changes when comparing the CP-MAS 13 C NMR spectra of the three cases when the reaction is carried out at 100 • C (all spectra reported in the Supplementary Material). Thus, they are respectively at 41.9 ppm (pTSA-catalysed), 42.6 ppm (NaOH-catalysed), and 42.3 ppm. This indicates that, just based on NMR evidence, the type of bonds obtained in the reaction at 100 • C appears to be more uncertain, or at least that ionic bonds and covalent bonds are in different proportions according to the presence of different catalysts, or their absence. This implies that at the higher temperature of 185 • C, the reaction shift more towards the formation of covalently bound amines, the most found in pTSA catalysis, followed by NaOH catalysis, and least in the uncatalysed case. The other clear indication of the existence of the formation of covalent bonds between the amine and the catechin -OH groups is the considerable decrease of the peak at 155-157 ppm, indicating that the carbons C5 and C7 carrying the -OH groups on the A-ring have markedly decreased as they have reacted. This species formed by this reaction is defined by the appearance of a new peak at 139.5 ppm. This peak belongs to a flavonoid C5 and a C7 that have reacted covalently with an amine, indicating that the interpretation given to the MALDI spectra has been incomplete because there has been considerable reaction on the A-ring to form the following types of linkages ( Figure 17). This is not all. The total disappearance in the 180 °C spectra of the catechin C3 peak at 68-72 ppm indicates that even the alcoholic -OH on the C3 site has reacted covalently with the amine to form linkages of the type (Figure 18  The other clear indication of the existence of the formation of covalent bonds between the amine and the catechin -OH groups is the considerable decrease of the peak at 155-157 ppm, indicating that the carbons C5 and C7 carrying the -OH groups on the A-ring have markedly decreased as they have reacted. This species formed by this reaction is defined by the appearance of a new peak at 139.5 ppm. This peak belongs to a flavonoid C5 and a C7 that have reacted covalently with an amine, indicating that the interpretation given to the MALDI spectra has been incomplete because there has been considerable reaction on the A-ring to form the following types of linkages ( Figure 17). The other clear indication of the existence of the formation of covalent bonds between the amine and the catechin -OH groups is the considerable decrease of the peak at 155-157 ppm, indicating that the carbons C5 and C7 carrying the -OH groups on the A-ring have markedly decreased as they have reacted. This species formed by this reaction is defined by the appearance of a new peak at 139.5 ppm. This peak belongs to a flavonoid C5 and a C7 that have reacted covalently with an amine, indicating that the interpretation given to the MALDI spectra has been incomplete because there has been considerable reaction on the A-ring to form the following types of linkages ( Figure 17). This is not all. The total disappearance in the 180 °C spectra of the catechin C3 peak at 68-72 ppm indicates that even the alcoholic -OH on the C3 site has reacted covalently with the amine to form linkages of the type (Figure 18  This is not all. The total disappearance in the 180 • C spectra of the catechin C3 peak at 68-72 ppm indicates that even the alcoholic -OH on the C3 site has reacted covalently with the amine to form linkages of the type (Figure 18 The other clear indication of the existence of the formation of covalent bonds between the amine and the catechin -OH groups is the considerable decrease of the peak at 155-157 ppm, indicating that the carbons C5 and C7 carrying the -OH groups on the A-ring have markedly decreased as they have reacted. This species formed by this reaction is defined by the appearance of a new peak at 139.5 ppm. This peak belongs to a flavonoid C5 and a C7 that have reacted covalently with an amine, indicating that the interpretation given to the MALDI spectra has been incomplete because there has been considerable reaction on the A-ring to form the following types of linkages ( Figure 17). This is not all. The total disappearance in the 180 °C spectra of the catechin C3 peak at 68-72 ppm indicates that even the alcoholic -OH on the C3 site has reacted covalently with the amine to form linkages of the type (Figure 18  This being a rather unexpected occurrence. The first question to be asked is then: are covalent bonds with the amine formed also with the carbons of the B-ring, as interpreted from the oligomers representation shown in the interpretation of the MALDI spectra?
The answer to this question is clearly yes, as the covalent bonds are also formed at 185 • C on the B-ring as the 145-146 ppm peak belonging to the aromatic B-ring carbons carrying the phenolic -OH groups is also markedly smaller in all the catalysed and uncatalysed spectra at 185 • C. Moreover, the covalent bond formed transmit at 134-135 ppm, indicating that the 138-139 ppm peak belongs to both the covalently reacted A-and B-rings of catechin. Thus, linkages such as Figure 19 also occur.
Polymers 2017, 9,37 10 of 17 This being a rather unexpected occurrence. The first question to be asked is then: are covalent bonds with the amine formed also with the carbons of the B-ring, as interpreted from the oligomers representation shown in the interpretation of the MALDI spectra?
The answer to this question is clearly yes, as the covalent bonds are also formed at 185 °C on the Bring as the 145-146 ppm peak belonging to the aromatic B-ring carbons carrying the phenolic -OH groups is also markedly smaller in all the catalysed and uncatalysed spectra at 185 °C. Moreover, the covalent bond formed transmit at 134-135 ppm, indicating that the 138-139 ppm peak belongs to both the covalently reacted A-and B-rings of catechin. Thus, linkages such as Figure 19 also occur. The simulation of the spectra with ACD/I-Lab yields a value of 41.8 ppm for the reacted amine, and 38.7 ppm for the unreacted amine. The online software at nmrdb.org [15] yields 44.0 ppm for the covalently reacted amine formed and 41.9 ppm for the nonreacted amine. The superposition of the pTSA-catalysed spectra at 185 and 100 °C shown in Figure 20 indicates that the formation of the amine is indeed occurring. Furthermore, on the pTSA-catalysed 185 °C spectrum, a rotation band at 13 ppm occurs. If the rate of rotation that has caused it is subtracted, a well-defined peak at 133.8 ppm is obtained. This peak is hidden by other species. This chemical shift can also be simulated with an amination of the -OH group on the C3 of catechin and corresponds to the to the shift of the carbon in C1′. Moreover, the disappearing of the peak at 68 ppm observed on the pTSA-catalysed 100 °C spectrum and absent The simulation of the spectra with ACD/I-Lab yields a value of 41.8 ppm for the reacted amine, and 38.7 ppm for the unreacted amine. The online software at nmrdb.org [15] yields 44.0 ppm for the covalently reacted amine formed and 41.9 ppm for the nonreacted amine. The superposition of the pTSA-catalysed spectra at 185 and 100 • C shown in Figure 20 indicates that the formation of the amine is indeed occurring. This being a rather unexpected occurrence. The first question to be asked is then: are covalent bonds with the amine formed also with the carbons of the B-ring, as interpreted from the oligomers representation shown in the interpretation of the MALDI spectra?
The answer to this question is clearly yes, as the covalent bonds are also formed at 185 °C on the Bring as the 145-146 ppm peak belonging to the aromatic B-ring carbons carrying the phenolic -OH groups is also markedly smaller in all the catalysed and uncatalysed spectra at 185 °C. Moreover, the covalent bond formed transmit at 134-135 ppm, indicating that the 138-139 ppm peak belongs to both the covalently reacted A-and B-rings of catechin. Thus, linkages such as Figure 19 also occur. The simulation of the spectra with ACD/I-Lab yields a value of 41.8 ppm for the reacted amine, and 38.7 ppm for the unreacted amine. The online software at nmrdb.org [15] yields 44.0 ppm for the covalently reacted amine formed and 41.9 ppm for the nonreacted amine. The superposition of the pTSA-catalysed spectra at 185 and 100 °C shown in Figure 20 indicates that the formation of the amine is indeed occurring. Furthermore, on the pTSA-catalysed 185 °C spectrum, a rotation band at 13 ppm occurs. If the rate of rotation that has caused it is subtracted, a well-defined peak at 133.8 ppm is obtained. This peak is hidden by other species. This chemical shift can also be simulated with an amination of the -OH group on the C3 of catechin and corresponds to the to the shift of the carbon in C1′. Moreover, the disappearing of the peak at 68 ppm observed on the pTSA-catalysed 100 °C spectrum and absent Furthermore, on the pTSA-catalysed 185 • C spectrum, a rotation band at 13 ppm occurs. If the rate of rotation that has caused it is subtracted, a well-defined peak at 133.8 ppm is obtained. This peak is hidden by other species. This chemical shift can also be simulated with an amination of the -OH group on the C3 of catechin and corresponds to the to the shift of the carbon in C1 . Moreover, the disappearing of the peak at 68 ppm observed on the pTSA-catalysed 100 • C spectrum and absent on the pTSA-catalysed 185 • C spectrum shows that the C3 carbon of the catechin has lost its -OH group, substituted with an -NH group (calculated at 67.6 ppm for catechin and 68 ppm for the 100 • C case, and at 55.8 ppm for the amination on C3 corresponding to the large peak at 55-59 ppm for the 180 • C case).
The second question is: do the ionic-type salt bonds, apparent in the MALDI spectra, really occur?
The response is also clearly positive. The huge peak at 27-29 ppm belongs to either diamine not reacted or to diamine linked as a totally ionised salt to structures of the type which follows. Thus, in both strongly acid-and strongly alkaline-catalysed reactions at 185 • C, it is certain that the rest of the amine is coordinated to catechin with linkages such as Figure 21, where the salts are formed with the -OHs of both the B-and A-rings of the catechin. It must be considered that unreacted diamine might be mixed with this, although the MALDI clearly indicates that the ionic bonds do exist. The presence of the 33 ppm shoulder in the spectrum of the alkali-catalysed 185 • C reaction product indicates that, for the amine, β carbons confirm that ionic bonds in quantity do occur.
Polymers 2017, 9,37 11 of 17 on the pTSA-catalysed 185 °C spectrum shows that the C3 carbon of the catechin has lost its -OH group, substituted with an -NH group (calculated at 67.6 ppm for catechin and 68 ppm for the 100 °C case, and at 55.8 ppm for the amination on C3 corresponding to the large peak at 55-59 ppm for the 180 °C case). The second question is: do the ionic-type salt bonds, apparent in the MALDI spectra, really occur?
The response is also clearly positive. The huge peak at 27-29 ppm belongs to either diamine not reacted or to diamine linked as a totally ionised salt to structures of the type which follows. Thus, in both strongly acid-and strongly alkaline-catalysed reactions at 185 °C, it is certain that the rest of the amine is coordinated to catechin with linkages such as Figure 21, where the salts are formed with the -OHs of both the B-and A-rings of the catechin. It must be considered that unreacted diamine might be mixed with this, although the MALDI clearly indicates that the ionic bonds do exist. The presence of the 33 ppm shoulder in the spectrum of the alkali-catalysed 185 °C reaction product indicates that, for the amine, β carbons confirm that ionic bonds in quantity do occur. In the case of the uncatalysed reaction, the amine should also be present in the same manner, thus partly ionically linked to the catechin or unreacted.
The last question to be answered by the NMR analysis is: what happens at lower temperature, namely at 100 °C?
The spectra are not reported here (they are available in the Supplementary Material), but the reaction is clearly less advanced, as should be expected. First of all, the C3 of the catechin has not reacted at all as the alcoholic C3-OH site shift exists and is big. Second, there is a clear covalent reaction of the amine on the catechin A-ring and some (lesser) reaction on the B-ring in the case of the pTSA-catalysed 100 °C spectrum. The reaction on the A-ring is also clear for the NaOH-catalysed 100 °C spectrum and also, but to a lesser extent, for the uncatalysed 100 °C spectrum. For these latter two, it does not appear that reaction on the B-ring does occur, or at least its proportion is minimal. Furthermore, no reactions on C3 appear to have occurred.

Reaction of Mimosa Tannin Extract with Hexamethylene Diamine
The same reactions were repeated by using mimosa tannin extract instead of the catechin model compound. The MALDI-ToF spectra of the reaction alkaline catalysis are shown in Figures 6 and 7. The oligomer species obtained under alkaline catalysis are listed in Table 2. The MALDI spectra for the acid-catalysed and uncatalysed reactions at 100 °C and 65 °C are reported in the Supplementary Material.
From Figures 22 and 23, the same period of 40 Da is observed as in the case of catechin due to a diamine with 2 × Na + , thus 116 + 23 + 23 -2 = 160 Da, thus 160/4 = 40 Da period: 801-761-719-677-638-596-554-514-474-430-390-350, is observed. The same type of reactions observed for the case of the catechin model compound appears to occur. Thus, both (1) substitution of the flavonoid units' hydroxyl groups with the amino group of the amine and (2) the formation of -O − Na + salts appear to occur. Compounds in which one, two, and even three diamines are linked covalently to a flavonoid monomer unit are observed as, for example, the peaks at 374, 524, and 621 Da (Table 2). Equally, species in which one or more HMDA molecules are linked covalently to a flavonoid dimer occur as, for example, the ones represented by the peaks at 743 and 758 Da (Table 2), as well as species in which HMDA constitutes a bridge between two species such as two flavonoid dimers as, for example, the compounds at peaks 1260 and 1330 Da (Table 2). Conversely, species in which the amine is not In the case of the uncatalysed reaction, the amine should also be present in the same manner, thus partly ionically linked to the catechin or unreacted.
The last question to be answered by the NMR analysis is: what happens at lower temperature, namely at 100 • C?
The spectra are not reported here (they are available in the Supplementary Material), but the reaction is clearly less advanced, as should be expected. First of all, the C3 of the catechin has not reacted at all as the alcoholic C3-OH site shift exists and is big. Second, there is a clear covalent reaction of the amine on the catechin A-ring and some (lesser) reaction on the B-ring in the case of the pTSA-catalysed 100 • C spectrum. The reaction on the A-ring is also clear for the NaOH-catalysed 100 • C spectrum and also, but to a lesser extent, for the uncatalysed 100 • C spectrum. For these latter two, it does not appear that reaction on the B-ring does occur, or at least its proportion is minimal. Furthermore, no reactions on C3 appear to have occurred.

Reaction of Mimosa Tannin Extract with Hexamethylene Diamine
The same reactions were repeated by using mimosa tannin extract instead of the catechin model compound. The MALDI-ToF spectra of the reaction alkaline catalysis are shown in Figures 6  and 7. The oligomer species obtained under alkaline catalysis are listed in Table 2. The MALDI spectra for the acid-catalysed and uncatalysed reactions at 100 • C and 65 • C are reported in the Supplementary Material.
From Figures 22 and 23, the same period of 40 Da is observed as in the case of catechin due to a diamine with 2 × Na + , thus 116 + 23 + 23 -2 = 160 Da, thus 160/4 = 40 Da period: 801-761-719-677-638-596-554-514-474-430-390-350, is observed. The same type of reactions observed for the case of the catechin model compound appears to occur. Thus, both (1) substitution of the flavonoid units' hydroxyl groups with the amino group of the amine and (2) the formation of -O − Na + salts appear to occur. Compounds in which one, two, and even three diamines are linked covalently to a flavonoid monomer unit are observed as, for example, the peaks at 374, 524, and 621 Da (Table 2). Equally, species in which one or more HMDA molecules are linked covalently to a flavonoid dimer occur as, for example, the ones represented by the peaks at 743 and 758 Da (Table 2), as well as species in which HMDA constitutes a bridge between two species such as two flavonoid dimers as, for example, the compounds at peaks 1260 and 1330 Da (Table 2). Conversely, species in which the amine is not covalently linked to a flavonoid unit, but rather a salt has been formed, are also present, such as, for example, the species at 390, 428-430, 661 Da, and many others as indicated in Table 2. Moreover, mixed species in which some HMDA molecules are linked covalently and some are linked by a salt bond to the same flavonoid are also present, as, for example, the species at peaks 638, 677, 761, and 1404 Da. Unreacted flavonoid oligomers-such as those represented by the peaks at 612, 881, 1178 Da, and others-are also present ( Table 2).
Polymers 2017, 9,37 12 of 17 covalently linked to a flavonoid unit, but rather a salt has been formed, are also present, such as, for example, the species at 390, 428-430, 661 Da, and many others as indicated in Table 2. Moreover, mixed species in which some HMDA molecules are linked covalently and some are linked by a salt bond to the same flavonoid are also present, as, for example, the species at peaks 638, 677, 761, and 1404 Da. Unreacted flavonoid oligomers-such as those represented by the peaks at 612, 881, 1178 Da, and others-are also present ( Table 2).    covalently linked to a flavonoid unit, but rather a salt has been formed, are also present, such as, for example, the species at 390, 428-430, 661 Da, and many others as indicated in Table 2. Moreover, mixed species in which some HMDA molecules are linked covalently and some are linked by a salt bond to the same flavonoid are also present, as, for example, the species at peaks 638, 677, 761, and 1404 Da. Unreacted flavonoid oligomers-such as those represented by the peaks at 612, 881, 1178 Da, and others-are also present ( Table 2).

Conclusions
A novel reaction of amines with flavonoid tannin is described. The reaction of hexamethylenediamine with catechin, a flavonoid monomer used as a model compound of condensed tannins, and with mimosa tannin were studied. The reactions were carried out at three different temperatures: 65, 100, and 185 • C. The reaction products obtained were analysed by MALDI-TOF and CP-MAS 13 C NMR, and the structures of the chemical species formed were indicated. Two reactions occurred under both alkaline and acid conditions, namely (i) the reaction of the amine with the phenolic hydroxy groups of the tannin, leading to polycondensation resins; and (ii) a reaction leading to the formation of ionic bonds between the protonated amino groups of the amine and the hydroxyl groups of the flavonoid structure for both the catechin and the tannin. Hardened, insoluble resins were formed for the tannin at the higher temperature, and resins insoluble in water but soluble in acetone were formed at 100 • C. At 100 • C, there are covalent bonds between the amine and the A-rings of the catechin for the NaOH-catalysed reaction, but less or even no reaction on the B-ring. There are even fewer covalent bonds in the uncatalysed reaction. - At 180 • C, and to a lesser extent also at 100 • C, mimosa tannin reacting with hexamethylene diamine forms hard, condensed solids, be it uncatalysed or alkali-or acid-catalysed. - At 180 • C, the condensation solid formed is hardly soluble in acetone water. - The reaction of mimosa tannin or similar condensed tannins with a diamine is fast. -In regard to the influence of different catalysts, their influence is minimal other than to accelerate the reaction. To this purpose, reactions at three different temperatures (but without any catalysts) were done with similar results, and the analysis results are shown in the Supplementary Material.