Synthesis and characterization of Hydrogels based on Potato Starch/Poly(vinyl Alcohol)/N, N'-Methylenebisacrylamide

Several hydrogels were synthesized by free-radical polymerization in an aqueous medium based on potato starch (PS), poly(vinyl alcohol) (PVA), and N,N'-Methylenebisacrylamide (MBAm), being possible to study these hydrogels as a function of the proportion of components incorporated. In this way, the products generated from the synthesis were characterized by swelling and deswelling kinetics, the first swelling being verified with Schott and statistical models, allowing to contrast the proximity between the experimental and theoretical behavior. Additionally, spectroscopy (FTIR), morphological (SEM), and thermal (TGA and DSC) analysis allowing to know the intrinsic characteristics of the material, increasing in general terms the knowledge of this type of material. In this context, it was possible to verify the characteristics and effectiveness of the synthesis and crosslinking of the main components. The experimental results obtained show that the synthesized hydrogels present representative first swellings consistent with kinetic and statistical models; however, there are significant changes in the second swelling derived from polymer degradation that occurs during the same swelling/deswelling cycles processes.


Introduction
In recent years, hydrogels have been presented as a highly innovative material used for many practical purposes due to their well-known hydrophilic qualities. This hydrophilic capacity allows the retention and storage of liquids of different kinds, providing options that no other previously studied material had been able to offer.
Currently, different types of hydrogels that have allowed their use in areas such as medicine, pharmacy, electronics, agriculture, purification, separation ,and segregation of substances, applied biotechnology, among other areas of innovation and development, reasons for which their research related to the properties and applications still on a rise [1][2][3][4][5].
In this sense, PVA arouses interest in biotechnological areas due to its properties and characteristics such as its low degree of toxicity, biocompatibility, stability, economy, and can also be produced via non-petroleum routes; reasons for which the development of hydrogels based on PVA is attractive for noble purposes. In addition, another essential characteristic of this material is in the development of new materials such as composites, since PVA as a matrix favors the dispersity of colloids [6], allowing cost/benefit balances more in line with specific needs and friendly to the environment.
The literature records various works on PVA-based hydrogels in conjunction with several types of natural or synthetic aggregates that allow obtaining materials with high development and application potential [1][2][7][8][9][10].
The use of aggregates of the natural origin (cellulose, saccharide, starch, etc.) have broad advantages because they are easy to acquire in multiple agro-industries, are environmentally friendly, are cheap, and allow optimization of some properties of polymer blends and hydrogels [11]. Among these renewable sources, starches are presented as a source of food and biomass present throughout the world and constant production during practically the entire period of the year [12]. These types of carbohydrates are found in significant amounts in many cereals, tubers, and roots kinds and are relatively easy to obtain [12][13][14].
Starch is constituted of two defined fractions, amylose and amylopectin, the latter being insoluble in water and with higher proportions in starch. This proportion varies according to the botanical source and its conditions [12,15], however, starch is mainly made up of amylose (20-30%) and amylopectin (70-80%) [16]. Among the main properties of interest of starch in the field of absorbent materials is the capacity for gelling/gelatinization, porosity, reversible absorption, swelling, and non-toxicity, attractive characteristics when formulating hydrogels of different types [13,[16][17][18][19][20].
In this context, this work presents a complete study related to the synthesis and characterization of hydrogels based on PS, PVA, and MBAm as a crosslinking agent allowing the obtaining of absorbent materials. The characterization of the hydrogels was carried out employing experimental and theoretical analysis of the swelling kinetics, statistical study of the swelling performance, spectroscopic analysis of FTIR, morphological characterization utilizing SEM, and thermal analysis of the polymer through DSC and TGA, these characterizations allowed to determine fundamental properties for the best applicaction of this kind of polymer.
All materials were of analytical grade and used without further purification. On the other hand, distilled water was used for the hydrogel preparation and swelling measurements.

Synthesis
The amounts of reagents mentioned in Table 1 were pre-diluted in 40 ml of distilled water; likewise, 1.5g of KPS was diluted in the same amount of water for all the syntheses. The solutions were incorporated into a continuous stirred-tank reactor (CSTR) operated in mode batch with a capacity of 500 ml, with temperature, stirring, and atmosphere controlled, keeping the conditions of 70 ° C, 700 rpm, and nitrogen gas (N2) atmosphere constant throughout the reaction. The order of adding the solutions to the reactor started with the PS solution followed by the PVA, KPS and ending with the MBAm. Furthermore, the reagent addition was performed with equal time spans of 15 minutes and the synthesis time lasted 3 hours from the last addition of the MBAm solution. After this period the product was removed from the reactor, washed in a 50/50 solution of ethanol/distilled water, dried, ground, and sieved.  On the other hand, the hydrogels were characterized employing Fourier transform infrared spectroscopyattenuated total reflectance (FTIR-ATR) in a Shimadzu IRTracer-100 spectrometer with the Pike MIRacle single reflection horizontal ATR accessory equipped with a ZnSe ATR crystal in transmittance mode. The wavenumber range from 300 to 4500 cm −1 at 64 scans and resolution at 4 cm -1 was used for the analysis.
The hydrogels were thermally characterized by Thermogravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC), these analyses were carried out in a TA Instruments TGA 550 and DSC 2500, respectively. TGA was carried out from room temperature to 600 °C under a heating rate at 10 °C/min using nitrogen flow at 100 mL/min as protective gas. For DSC was used nitrogen flow at 100 mL/min as purge and protective, and the analysis temperatures setup were defined between -50 to 250 °C at a heating and cooling rate of 20 °C/min, the cooling stage was reached using liquid nitrogen.
Finally, Scanning Electron Microscopy (SEM) analyses were also carried out in FE-SEM, Hitachi Regulus 8230, Hitachi High-Tech Co. equipment, employing a secondary electron signal and voltage at 3,0 kV.

Results and discussion
The synthesis and subsequent processes allowed to obtain a series of hydrogels based on PS, PVA, and MBAm that have a gelatinous, yellowish, and soft texture in their wet state; already in their dry state the hydrogels turn pale and rigid. In Fig. 1 it is possible to observe some hydrogels obtained.  Fig. 2 shows the results obtained regarding the kinetic curves of two consecutive swellings process of the hydrogels. As expected in a first swelling ( Fig. 2A) the behavior describes a logarithmic function adjusting asymptotically to a value that represents an experimental equilibrium swelling (Wt1) that depends on the composition of each hydrogel. Already in the second swelling curves (Fig. 2B), some transformations in the ideal swelling behavior are observed, partially moving away from a logarithmic function, however, a degree of swelling is presented that stabilizes asymptotically at an equilibrium value (Wt2). Additionally, note that changes in the absolute value of swelling are observed between the two types of the curve ( Fig. 2A and 2B), reaching lower degrees of swelling. These changes are due to irreversible transformations generated by possible breaks in the cross-linked bridges between PVA and PS caused by the hydration of the material.

Swelling and deswelling test
Likewise, Fig. 2C and 2D shows the Swelling kinetic relations (t/w) referring to the swelling data, where w is the swelling at a given time t. In Fig. 2C it can be seen that the plotted function of the t/w relationship has a behavior very close to linearity, characteristic that is verified with the correlation coefficient R 2 which expresses a very high linear correlation for all hydrogels (see Table 2). This high correlation decreases slightly after the first swelling as observed in Fig. 2D, where it is observed that the correlation loses affinity, obtaining in some cases moderate correlations caused by the disruptions generated by hydration. The quantitative and qualitative assessments of the correlation coefficients can be examined in Table 2.
The experimental swelling data were evaluated by the Schott's pseudo-secondorder swelling kinetics mathematical model to evaluate the swelling kinetics of the hydrogels, the model initially formulated by Equation 2: = +

(Equation 2)
Indicating that: where A is the intercept with the vertical axis and B is the linear slope of the line t/w obtained through the linear least-squares fit of the experimental data, see Equations 3 and 4. In this context, such constants allow us to determine the values of Ks and W∞ can be rewritten the equation 2 as: where Ks represents the swelling rate constant and W∞ is the swelling in theoretical equilibrium. It is possible to verify that the experimental and theoretical values of first swelling are significantly close without greater variations, on the other hand, the model loses significance in the second swelling, see Table 2. For the case of the first swelling W∞, some hydrogels (H1, H3, H4, H7, H8, and H9) are close to a higher value due to a asymptotical trend of increasing swelling of the hydrogel. On the other hand, other hydrogels (H2, H5, and H6) approach values lower than the experimental value; this trend is due to slight variations that oscillate around the equilibrium experimental swelling value, oscillations that were registered at the moment of mass reading. For the second swelling, an approximation of W∞ to values greater than the experimental Wt2 is observed, an expected result given that all curves are constantly increasing swelling throughout the data collection record. In addition to that, greater differences are observed between the theoretical and experimental swellings, indicating a lower adjustment of the model for swelling after the initial one. The results obtained from the swelling curves adjusted to the Schott model can be seen in Table 2.  Additionally, these tests were carried out sequentially after the swelling tests, which explains the initial values of the deswelling curves. On the other hand, as expected, the materials show reductions in their mass throughout the evaluated period until reaching an equilibrium mass, this deswelling curves it can be seen that the mass loss presents a partially sinusoidal behavior with an asymptotic end of the function (Fig. 3A and 3B).

Fig. 2 Swelling kinetic curves and kinetic relations of hydrogels
This partially sinusoidal behavior suggests the existence of a loss of critical mass which was evaluated by calculating the first derivative of the function of the deswelling curve ( Fig. 3C and 3D). The composition of the different hydrogels analyzed determines the critical mass loss time, however, critical loss ranges occur in periods between 205 and 285 minutes for most of the hydrogels for the two deswelling cycles. It should be noted that the variation in drying temperature would vary the behavior of critical mass loss, displacing it positively or negatively in the exposure time.  Table 3 shows the experimental points and the experimental results of swelling used for the statistical study. Nine random runs were performed to reduce the errors of the systematic trends in the variables. The data obtained were evaluated by the analysis of variance (ANOVA) to adjust the response model by the least-squares method and to evaluate its adequate fit. The significance of the model was evaluated using the probability value (p-value); at a 95% confidence level, knowing that p-values lower than 0.065 indicates significant effects. The p-value coefficients greater than 0.065 were not significant for the model, so they were not considered in the regression model.  Table 4. The fit of the model was evaluated using the probability value (p-value), and the coefficients of determination (R 2 and Radj 2 ). Based on the parameters in Table 4, the cubic regression model with determined coefficients is given by Equation 6.  The data of the model (  The Anderson Darling statistical test was developed, which generated a p-value of ~ 0.064, a value slightly above the usual value of the significance level; this result suggests that the residues describe a predominantly normal distribution, see Fig. 4A3.

Statistical study
On the other hand, the independence of the results was established through the Durbin-Watson test, the data and results related to this test can be seen in Table 5

FTIR analysis
The main raw materials used in the present work were characterized by FTIR, to verify the transformations generated after the synthesis. In this context, Fig. 6 shows the    Table 6 summarizes the characteristics of the bands identified in the FTIR spectra of the hydrogels. The generation of new characteristics in the spectra of the hydrogels compared with the raw materials, as well as the types of bands generated in the products studied proposed suggests that the crosslinking was carried out, confirming that the proposed synthesis methodology was effective concerning to integration of new hydrogels.   (CO stretching) 1240 [31,[34][35] Band at 1020 cm −1 represents the ether linkages and acetal ring as a result of the reaction between the MBAm and OH group of the polymer. 1020 [31,37] The band near 1152 cm −1 appeared as a result of intra-and intermolecular interactions of hydroxyl groups in PVA and PS. 1152 [25,[32][33] A band at 1085 cm −1 is due to the crosslinking reaction among PVA/PS chains and MBAm (C-O-C group). This suggests the formation of ether linkages and acetyl rings.
1085 [25,[32][33][34] The observed band also implies NH stretching due to MBAm. Similarly, a strong amide band between 1590-1490 cm -1 may also be assigned to the C=O stretching vibration of MBAm and other groups.
1550 [36] Presence of residual acetate groups present in PVA. 1670 [33,37] The peaks between 1600 and 1800 cm −1 signify the presence of both C=O and C=C bonds.

Thermal analysis
The TGA thermograms can be observed in Fig. 8A, in these curves, it can be seen that each sample presents several mass losses that can be confirmed and defined with their respective first derivatives, see Fig. 8B and were not fully evidenced in all the TGA tests of the studied samples (see Fig. 8B and Table 7), this possibly due to the differences between compositions of the samples and due to the washes carried out that could remove fractions in different proportions that did not react during the synthesis.   Table 7. This behavior indicates the alteration in crystallinity that may be due to the change in the degree of crosslinking in the hydrogel, which implies a change as a function of the amount of crosslinking incorporated during the synthesis.
These observed changes in crystallinity support previous studies where agents incorporated into PVA affect observable transitions in DSC tests, generating transition temperature shifts as in the case of melting point [10,[45][46][47][48][49].  Fig. 10-12 show the SEM micrographs of the hydrogels named H2, H5, and H9, respectively, these micrographs were selected based on their swelling potential (high, medium, and low). In general, all the presented micrographs of the hydrogels show morphological characteristics of a composite material wherein the polymeric matrix constituted by PVA is visualized, integrating dispersed phases constituted by aggregate PS.

SEM analysis
In most micrographs, the PVA polymer phase is characterized by predominantly smooth surfaces and the dispersed phase (PS) is characterized by having curved or flat granules with variable roughness and have spherical, polygonal, and oval shapes, morphological characteristics previously recorded in the literature [13,50]. Additionally, the polydispersity of the dispersed phase within the polymeric matrix is observed to be homogeneous with some agglomerations possibly produced in the drying and washing processes, however, the aggregates are present throughout the matrix.
Looking at it the other way, it can be seen in Fig. 10  The results of swelling and deswelling of different hydrogels demonstrate that the synthesized material suffering irreversible disruptions and transitions after the first swelling and deswelling decreasing the degree of swelling. In addition, Schott's pseudo-second order swelling kinetics model fits the experimental results, however, the model loses affinity with the second swelling, generating a loss of relative precision, which must be considered when performing this type of cyclical test.
The complementary characterizations developed and the proposed statistical model allow the generation or formulation of desired absorbent hydrogels, based on potato starch, poly (vinyl alcohol), and N, N'-methylenebisacrylamide, in technical terms and cost/benefit criteria for specific purposes.