A deeper exploration of the relation between sulfonation degree and retanning performance of aromatic syntans

It is well-known that the sulfonation degree (DS) of aromatic syntan is an important factor affecting its retanning performances. But the quantitative relation between DS and syntan property and the influencing mechanism of DS on syntan property are not clarified. In this work, five phenolic formaldehyde syntans (PFSs) with the same polymerization degree but varying DS were prepared to investigate the effect of DS on the properties of syntan and crust leather. It was found that the absolute value of zeta potential and the particle size of PFS decreased with increasing DS in aqueous solution. Molecular dynamic simulation results proved that the DS of PFS was a major contributor to electrostatic interaction and hydrogen bonding in the PFS–water system and greatly affected the aggregation and dispersion of PFS in aqueous solution. The PFS with a low DS was prone to aggregate to large particles in aqueous solution because of low intermolecular electrostatic repulsion and less hydrogen bonds and therefore can be used to increase the thickness and tightness of leather. The PFS with a high DS presented a small particle size with more anionic groups in aqueous solution, thereby sharply decreasing the positive charge of leather surface and facilitating the penetration of the post-tanning agents into the leather. These results might be scientifically valid for rational molecular design of syntans and more productive use of syntans in leather making.


Introduction
Retanning has become indispensable for modern leather manufacturing because it can reduce differences in various parts of leather, increase the yield of leather, and endow leather with proper fullness and elasticity [1][2][3]. Commonly used retanning agents include aromatic syntans, amino resins, acrylic resins, and aldehyde derivatives [4,5]. Aromatic syntans, especially phenolic syntan, have received much attention due to their advantages of filling property, dispersibility, and bleachability [6]. Aromatic syntans are generally prepared through condensation, sulfonation, and other reactions by using phenol or naphthalene, formaldehyde, and concentrated sulfuric acid as raw materials, thereby having hydrophobic benzene/naphthalene rings and hydrophilic sulfonic acid groups [7,8]. Sulfonation degree (DS), which represents the portion of sulfonic acid groups in aromatic syntan, is a key factor that influences the retanning performance of leather according to tanners' practical experience. For example, an aromatic syntan with low DS often has strong tanning property and is used as a replacement retanning agent, whereas a syntan with high DS generally has good solubility and weak tanning property and is used as an auxiliary retanning agent (e.g., dispersant and neutralizer).
Previous research mainly focused on the selection of new monomers for synthesis of novel or high-performance syntans [9,10]. The effect of the DS of aromatic syntan on retanning performance is usually evaluated by tanner's experience because the penetration and distribution of aromatic syntan in leather have never been directly visualized or quantified. Hence, the quantitative relationship between the DS of aromatic syntan and its retanning performance remains unclear. Fully understanding the relationship is essential to develop new syntans and choose rational syntans for achieving better retanning performance and leather quality.
In this study, phenolic formaldehyde syntans (PFSs) with the same polymerization degree and different DSs were prepared and used as aromatic syntan models to investigate the effects of DS on the charge property and particle size of syntan, the penetration and uptake of syntan in leather, and the organoleptic properties of crust leather. Moreover, intermolecular interactions of PFS in

Preparation of PFSs with different DSs
PFSs with different DSs were prepared through condensation and sulfonation ( Fig. 1). A dehydrated intermediate, namely, phenolic formaldehyde condensate (PF), was prepared using phenol and formaldehyde as raw materials according to literature [11]. In brief, 94.1 g phenol (1 mol) and 20.0 g sulfuric acid solution (20 wt%, 0.04 mol sulfuric acid) were added into a four-neck flask equipped with a thermometer, a mechanical stirrer, and a flush condenser. Then, 52.7 g formaldehyde solution (37 wt%, 0.65 mol formaldehyde) was added drop-wisely at 60 °C within 2 h. The mixture was reacted at 90 °C for another 3 h. After that, the condensates were cooled, separated, washed, and vacuum dried. As a result, a mixture of ortho and para PFs were obtained. The phenol hydroxyl content of PF determined by titration method was 9.7 mmol/g, the polymeric degree determined by electrospray ionization mass spectrometry ranged from 3 to 7 (Additional file 1: Fig. S1), and the average molecule weight determined by gel permeation chromatography was 573 g/ mol (Additional file 1: Table S1). Then, 103 g of PF was separately sulfonated using 20, 40, 60, 80, and 100 g of concentrated sulfuric acid at 110 °C for 3 h. After sulfonation, the mixtures were neutralized and lyophilized to obtain five types of PFSs with different DSs.

Determination of DS of PFS
In brief, 2 g of sulfonated PF without neutralizing was diluted with 50 mL of distilled water in a flask. The sulfonated PF solution was titrated to pH 7.2 by adding sodium hydroxide solution (0.5024 mol/L). The DS of PFS was calculated by Eq. (1): where c is the concentration of sodium hydroxide solution, V 1 is the equivalent volume of sodium hydroxide solution consumed by 1 g of the mixture of PF and sulfuric acid according to sulfonation reaction recipes, V 2 is the equivalent volume of sodium hydroxide solution consumed by 1 g of sulfonated PF (mL/g), and n [OH] is the phenolic hydroxyl content per 1 g of sulfonated PF (mmol/g). (1)

Determination of zeta potential and particle size of PFS
The zeta potentials of PFSs aqueous solutions (1.0 g/L) in the pH range of 3.0-6.5 and the particle size distributions of PFSs in aqueous solution (1.0 g/L) at pH 6.5 were determined by using a zeta potential and laser particle size analyzer (Nano ZS90, Malvern, UK). All the measurements were undertaken at 25 °C after the aqueous solution samples were treated with ultrasonic for 10 min and equilibrated for 3 min. The measurement was repeated three times for each sample.

MD simulation
Two analogs of PFS ( Fig. 2a (Table 1) was carried out at 300 K and 1 atm to simulate the aggregation behavior of PFSs with different DSs in aqueous solution. The Optimized Potentials for Liquid Simulations/All Atom (OPLS/AA) force field was employed to describe the intramolecular and intermolecular interactions of the models [12]. The TIP4P [13] model was used to describe the water molecule. Before MD simulation, the initial structures of L-PFS and H-PFS were optimized through density functional theory (DFT) by using the B3LYP functional at the 6-311G++(d, p) basis set. The configurations of the two systems were generated using the PACKMOL package [14]. Na + ions were added to neutralize the negative charges of sulfonate groups in the models. All molecular simulations were performed using the GROMACS package [15] (version 2019.6), where the model molecules were placed in a cubic box and periodic boundary conditions were applied to the x, y, and z directions. First, the two systems were energy minimized using steepest descent minimization method for 10,000 steps. The systems were equilibrated successively in canonical NVT and NPT ensemble for 500 ps. Finally, production runs were performed for 100 ns. Berendsen barosatat [16] and Langevin dynamics methods [17] were used to control the pressure and temperature, respectively. A time step of 2 fs was used for all simulations, and the bond lengths were kept rigid using the LINCS algorithm [18]. A cutoff distance of 12.0 Å was defined for all non-bonded interactions. Particle mesh Ewald method [19] was used to treat long-range electrostatic interactions. Simulation trajectory data were analyzed using Visual Molecular Dynamics (VMD) software [20], and non-bond interactions including electrostatic interaction, van der Waals force, and hydrogen bonding were analyzed using the build-in modules of GROMACS.

Retanning and fatliqouring
Eighteen pieces of shaved wet blue (30 cm × 30 cm for each) were first rewetted and chrome retanned according to conventional processes. Pieces No. 1-6 were neutralized to pH 4.5. Pieces No. 7-12 were neutralized to pH 5.5. Pieces No. 13-18 were neutralized to pH 6.5. These neutralized leather samples were retanned with  PFS (Table 2). Another eighteen pieces of leather samples were neutralized as above and then retanned and fatliqoured as shown in Table 3. The crust leather without PFS retanning was used as the control.

Determination of zeta potential and pIs of retanned leather
The zeta potentials of the retanned leather samples (No. [13][14][15][16][17][18] with pH ranging from 3.0 to 9.0 were determined according to our previously reported method [21] by using a Zeta potential analyzer (Mütek SZP-10, BTG, Germany). The pH at zero point of zeta potential for the determined system was identified as the pI of the leather.

Analysis of PFS penetration in leather
The penetration of PFS (reddish brown) in leather (blue) was analyzed by observing the vertical section of the retanned leather with a stereo microscope (SZX12, Olympus, Japan). The relative content of PFS in leather was obtained by processing the micrograph with Image J software [22]. The penetration rate of PFS in the leather was calculated by Eq. (2) [23]: (2) Penetration rate (%) = Area of leather with PFS Total area of leather × 100

Determination of uptake rate of PFS by leather
The concentrations of total organic carbon (TOC) in the initial and final retanning baths were measured using a TOC tester (Vario TOC, Elementar Co., Ltd., Germany) and recorded as TOC ini and TOC res , respectively. The uptake rate of PFS was calculated by Eq. (3):

Determination of thickness increase rate of leather
The thicknesses of PFS retanned and fatliqoured leathers were measured using a thickness gauge (MY-3130-A2, Ming Yu Electron Tech Information Co., Ltd., China) and recorded as δ 1 . The thickness of fatliqoured leather without PFS retanning was measured and recorded as δ 0 .
The thickness increase rate of the leather caused by PFS retanning was calculated by Eq. (4):

Determination of leather softness
The staked crust leathers were conditioned at 20 °C and 65% RH for 48 h. The softness of the crust leather samples was determined according to the standard (IUP 3).

Effect of DS on zeta potential and particle size of PFS
This study mainly aimed to determine the effect of the DS of PFS on retanning performance. Five PFSs that had the same polymerization degree and different DSs were prepared. The DSs of these PFSs are shown in Fig. 3a. DS affects the surface charge and aggregation behavior/particle size of the retanning agents, which mostly determine the penetration and uptake of retanning agents in leather. Therefore, the effects of DS on the zeta potential and particle size of PFS were investigated. Figure 3b shows the zeta potentials of various PFS solutions. The increase in the absolute value of zeta potential is correlated with the decrease in the DS of PFS and the increase in the pH of the PFS solution. The former phenomenon could be explained by the aggregation of PFS. The PFS molecules aggregated into a colloid that consisted of a hydrophobic core stacked with benzene rings and a hydrophilic shell with abundant sulfonic acid groups [7,8,24]. And the latter phenomenon was due to the deprotonation of phenolic hydroxyl and sulfonic acid groups in PFS. Figure 3c shows that the average particle size of PFS increases with decreasing DS. This means that the PFS with lower DS aggregates into larger particles in aqueous solution. MD simulation is an effective method used to explore intermolecular interactions, such as electrostatic interaction, van der Waals force, and hydrogen bonding. Figure 4a presents the snapshots of Systems 1 and 2 at 0 and 100 ns. L-PFS molecules assemble and form large clusters, whereas H-PFS disperses well in the aqueous solution. The result is consistent with the particle size data in Fig. 3c. Figure 4b shows that H-PFS is more disperse in aqueous solution than L-PFS due to its lower potential energy. The absolute value of electrostatic energy is higher than that of van der Waals energy (Fig. 4c), indicating that the electrostatic interaction is a major contributor to the aggregation and dispersion of the PFS-water system. Additionally, the absolute value of electrostatic energy for the H-PFS-water system is higher than that for the L-PFS-water system, indicating that high DS is beneficial to the dispersion of PFS in water. The average number of hydrogen bonds formed by H-PFS and water is higher than that formed by L-PFS and water (Fig. 4d) because H-PFS has more sulfonic groups. This finding explains why H-PFS has higher solubility than L-PFS in aqueous solution. Hence, the DS of PFS is an important contributor to electrostatic interaction and hydrogen bonding in the PFS-water system, which greatly affects the aggregation behavior/particle size of PFS in the aqueous solution.

Effect of DS of PFS on pI of retanned leather
As shown in Fig. 5, the pI of the chrome retanned leather (control) is about 7.9, and the pIs of the leather samples retanned with PFS are reduced to less than 5.3. This is due to the introduction of the sulfonic acid and phenolic hydroxyl groups in PFS to hide collagen. The pIs of the retanned leather samples were reduced from 5.3 to 3.6 as the DS of PFS increased from 18 to 74%. Increasing DS results in a decrease in the pI of the retanned leather. This is because the sulfonic acid and phenolic hydroxyl groups in PFS can combine with the amino groups of collagen and chrome in leather [25]. Within pH 4.5-6.5, the retanned leather possessed lower zeta potential with increasing DS, which should be helpful for the penetrations of PFS and other post-tanning agents in the leather.

Effect of DS on penetration of PFS in retanned leather
Retanning performance is closely related to the penetration depth of syntan in leather. In this section, the penetrations of PFSs with different DSs in leather samples neutralized to pH 4.5, 5.5, and 6.5 were analyzed. The visual distributions, relative contents, and penetration rates of PFS in the leather are shown in Figs. 6a-c, respectively. The penetration depth of PFS in leather at the same neutralizing pH increases with increasing DS of PFS. The main reason for this should be the fact that the positive charge of neutralized leather decreased more remarkably when the leather combined with PFS having more anionic groups (Fig. 5), which consequently reduced the electrostatic attraction between leather surface and PFS. Another reason is that the particle size of PFS decreases with increasing DS. The two reasons contribute to reduction in the PFS transfer resistance, thereby increasing the penetration depth of PFS in the leather. All PFSs did not fully penetrate the leather samples neutralized to pH 4.5 mainly because of the strong electrostatic attraction (binding force) between the leather surface (zeta potential, 54.4 mV) and PFSs (zeta potential, − 51.4 mV to − 37.7 mV). As the neutralizing pH increased to 5.5, the zeta potential of leather reduced sharply to 26.8 mV, and the zeta potentials of PFSs (− 50.4 mV to − 43.0 mV) were slightly lower than those at pH 4.5. As a result, the mass transfer resistance caused by the electrostatic attraction between the leather surface and PFS weakened, and the penetration of PFS in the leather increased. When the pH of the leather sample was neutralized to 6.5, the zeta potential of the leather was reduced to 18.0 mV, and PFSs with DS higher than 50% completely penetrated the leather. These results clearly and quantitatively verify that a high DS of PFS and a high neutralization pH of leather favor the penetration of PFS in leather.

Effect of DS on uptake of PFS by leather
As shown in Fig. 7, the uptake rates of PFS by leather decreased with increasing DS of PFS because a higher DS of PFS made leather have less positive charge at fixation pH of 3.8, thereby reducing the electrostatic attraction between PFS and leather. Moreover, the neutralizing pH had little effect on the uptake of PFS by the leather because the fixation pH was as low as 3.8, ensuring that the leather had positive charge to form bonds with PFS. The uptake rate of PFS-74 was the lowest but still reached more than 84%. This may be due to the fact that the particle size of PFS is relatively small and could enter small gaps among collagen fibers besides large gaps.

Effect of DS of PFS on thickness and softness of crust leather
Retanning has received much attention in the leather processing since it can improve the handle of leather, such as thickness and tightness [26,27]. The effects of DS of PFS on the thickness and softness of crust leather were analyzed (Fig. 8). The increase rate of leather thickness decreased, and the softness increased as the DS of PFS increased. The PFS with low DS can form large aggregates and fill large pores among collagen fibers, thereby increasing the thickness and tightness of the leather. These results suggest that thickness and tightness of leather can be regulated using PFSs with different DSs. It is worth noting that the neutralization pH has a greater impact on the softness of crust leather than the DS of PFS (Fig. 8b). This is because the surface charge of retanned leather decreased more sharply with the increase in neutralization pH than with increasing DS (Fig. 5), and less positive charges of the retanned leather was more beneficial to the penetration of fatliqour. However, a high neutralization pH may pose a risk of leather looseness. Therefore, it is better to regulate the penetration of post-tanning chemicals through rational combination of neutralization and DS selection.

Conclusions
Intermolecular interactions, such as electrostatic interaction, van der Waals force, and hydrogen bonding, determine the aggregation behavior/particle size of PFS in aqueous solution. The PFS with low DS contains less sulfonic acid groups, and thus results in lower intermolecular electrostatic repulsion and smaller amount of hydrogen bonds in the PFS-water system. In this regard, PFS with low DS aggregates into large particles in aqueous solution. The PFS with high DS includes more sulfonic acid groups, and therefore leads to higher intermolecular electrostatic repulsion in the PFS-water system. As such, PFS with high DS is more disperse in aqueous solution and has a small particle size. The PFS with low DS is suitable for increasing the thickness and tightness of leather, and the PFS with high DS is helpful for reducing the positive charge of leather surface  and promoting the penetration of anionic chemicals in leather.