Catalytic Degradation of Chitosan by Supported Heteropoly Acids in Heterogeneous Systems

: Several kinds of composite materials with phosphotungstic acid (PTA) as the catalyst were prepared with activated carbon as support, and their structures were characterized. According to the Box–Behnken central combination principle, the mathematical model of the heterogeneous system is established. Based on the single-factor experiments, the reaction temperature, the reaction time, the amount of hydrogen peroxide and the loading capacity of PTA were selected as the inﬂuencing factors to study the catalyzed oxidation of hydrogen peroxide and degradation of high molecular weight chitosan. The results of IR showed that the catalyst had a Keggin structure. The results of the mercury intrusion test showed that the pore structure of the supported PTA catalyst did not change signiﬁcantly, and with the increase of PTA loading, the porosity and pore volume decreased regularly, which indicated that PTA molecules had been absorbed and ﬁlled into the pore of activated carbon. The results of Response Surface Design (RSD) showed that the optimum reaction conditions of supported PTA catalysts for oxidative degradation of high molecular weight chitosan by hydrogen peroxide were as follows: reaction temperature was 70 °C, reaction time was 3.0 h, the ratio of hydrogen peroxide to chitosan was 2.4 and the catalyst loading was 30%. Under these conditions, the yield and molecular weight of water-soluble chitosan were 62.8% and 1290 Da, respectively. The supported PTA catalyst maintained high catalytic activity after three reuses, which indicated that the supported PTA catalyst had excellent catalytic activity and stable performance compared with the PTA catalyst.


Introduction
Chitosan is the deacetylated product of chitin, which is abundant in nature. It is widely distributed in the crusts of marine arthropods, such as shrimps and crabs, fungi, insects, algae cell membranes and cell walls of higher plants. It is the second largest natural alkaline polysaccharide after cellulose in terms of annual output. Chitosan has broad application prospects in food, the chemical industry, textiles, biological materials and medicine [1][2][3], and it has attracted the attention of a large number of researchers. The molecular weight of chitosan has a great influence on its properties. High molecular weight chitosan has a strong hydrogen bond and a tight crystal structure. It is insoluble in ordinary solvents and can only be dissolved in some acidic media. Low molecular weight chitosan has high solubility and is easy to absorb in vivo. It also has anti-bacterial and anti-tumor functions and

Single-Factor Experiment
Compared to unsupported PTA under different catalyst dosage conditions, the LWCS obtained by the supported PTA catalyst had a higher yield and lower average molecular weight, which indicated that supported PTA could significantly improve the degradation efficiency of chitosan ( Figure 1A,B).
The effects of various factors on the yield and average molecular weight of LWCS were as follows: when the loading capacity was 28.3%, the optimal catalytic activity was obtained ( Figure 1C). When PTA was overloaded, the yield of LWCS decreased and the average molecular weight increased. R-value 2 of hydrogen peroxide was the most suitable ( Figure 1D). When the dosage of hydrogen peroxide continued to increase, the average molecular weight decreased slightly, mainly because hydrogen peroxide is a hydroxyl radical scavenger. When the concentration of hydrogen peroxide reached a certain level, some hydroxyl radicals would be eliminated. The oxidation ability of hydrogen peroxide could not be improved obviously [8]. When the catalyst mass ratio was 0.02, the yield of LWCS was the highest and the average molecular weight was the lowest ( Figure 1E). When the degradation temperature of chitosan was 70 • C, the yield was the highest and the degradation was the most thorough ( Figure 1F). The best degradation time of chitosan was 3 h ( Figure 1G).

Single-Factor Experiment
Compared to unsupported PTA under different catalyst dosage conditions, the LWCS obtained by the supported PTA catalyst had a higher yield and lower average molecular weight, which indicated that supported PTA could significantly improve the degradation efficiency of chitosan ( Figure 1A The effects of various factors on the yield and average molecular weight of LWCS were as follows: when the loading capacity was 28.3%, the optimal catalytic activity was obtained ( Figure  1C). When PTA was overloaded, the yield of LWCS decreased and the average molecular weight increased. R-value 2 of hydrogen peroxide was the most suitable ( Figure 1D). When the dosage of hydrogen peroxide continued to increase, the average molecular weight decreased slightly, mainly 8 The number-average molar mass The number-average molar mass From Figure 1H, it was known that there might be a strong interaction between activated carbon and PTA. The catalytic activity remained higher after which it was recycled and reused three times. However, the catalytic efficiency decreased rapidly during the fourth time. After it was reused five times, the LWCS could not be completely dissolved. The main reason was that PTA was reused more than a certain number of times. Because PTA was desorbed from the pore of activated carbon, the catalytic efficiency of the supported catalyst was significantly reduced. Therefore, PTA supported on activated carbon could be reused, and the effect was good, but the binding degree between PTA and activated carbon was still not strong enough. If the preparation process of the catalyst was further optimized, there would still be large room for improvement in catalytic efficiency and reuse times.

Design and Results of Response Interview Experiments
According to the Box-Behnken central composite design principle, four factors, such as hydrogen peroxide dosage (A), catalyst loading (B), reaction time (C) and reaction temperature (D), were selected to design the response surface experiment with four factors and three levels, and regression analysis and significance analysis were carried out on the data. The ratio of average molecular weight to yield of chitosan was the response value (Y%). The experimental design and results are shown in Table 1.

Modeling and Variance Analysis
The Design Export 8.0 software was applied to analyze the data in response surface regression fitting ( Table 2) to conduct variance analysis and the significance test. The quadratic regression fitting model could be obtained. The equation was as follows: Detailed results are shown in Table 2. The results of the variance analysis of the regression model in Table 2 show that the Model F-value of 2.95 implied the model was significant. The overall p value was 0.0261, which is less than 0.05, indicating that model terms were significant. The coefficient of the determinant was 74.67 (R 2 = 74.67), and the missing items were not significant, so the model selection was correct, indicating that the model has significant statistical significance and the model fitting degree is good. Therefore, the model can be used to simulate and predict the experiment.
It is shown from the results of the significance test of the regression coefficient that the greater the F value is, the greater the impact on the response value is, in which the order of influence on the preparation process of water-soluble chitosan is as follows: hydrogen peroxide dosage > reaction time > catalyst loading > reaction temperature.

Response Surface and Contour Analysis
The shape of the contour was an ellipse, which indicates the interaction of factors. However, the circular shape indicated that the interaction was not significant [13]. It can also be seen from Figures 2-7 that the interactions between reaction temperature and time were significant. The interactions between reaction temperature and hydrogen peroxide dosage, catalyst loading and time on hydrogen peroxide dosage and catalyst loading were not significant.

Best Technology and Reproducibility Experiments
The software was used to solve the partial derivation of the regression equation, and the optimum technological conditions were selected as follows: degradation temperature 69.71 °C, time 3.06 h, the ratios of hydrogen peroxide to hydrogen peroxide 2.39 and catalyst to hydrogen peroxide 28.9%. According to the actual situation, it was revised as follows: degradation temperature 70 °C, time 3.0 h, hydrogen peroxide dosage 2.4, catalyst dosage 30%. Under these conditions, three repeatability experiments were carried out. The average molecular weight of LWCS ranged from 1515 Da to 1564 Da with an average yield of 58.47%. The quality of LWCS was good. Therefore, it could be shown that the model constructed by this experiment fitted well with the actual situation, and the best process obtained was feasible.

Infrared Spectrum of Supported PTA
The FTIR spectra of supported PTA catalysts are presented in Figure 8. There were several characteristic absorption peaks of Keggin structure at 500-1100 cm −1 of supported PTA. Its characteristic bands were at 1079.72 cm −1 (P-Oa stretching), 982.21 cm −1 (W=Ot stretching), 890.13 cm −1 (W-Ob stretching), 791.79 cm −1 (W-Oa stretching), 595.18 cm −1 and 524.35 cm −1 (W-Oc stretching), indicating that the Keggin structure of PTA was maintained when it was loaded by activated carbon.

Best Technology and Reproducibility Experiments
The software was used to solve the partial derivation of the regression equation, and the optimum technological conditions were selected as follows: degradation temperature 69.71 °C, time 3.06 h, the ratios of hydrogen peroxide to hydrogen peroxide 2.39 and catalyst to hydrogen peroxide 28.9%. According to the actual situation, it was revised as follows: degradation temperature 70 °C, time 3.0 h, hydrogen peroxide dosage 2.4, catalyst dosage 30%. Under these conditions, three repeatability experiments were carried out. The average molecular weight of LWCS ranged from 1515 Da to 1564 Da with an average yield of 58.47%. The quality of LWCS was good. Therefore, it could be shown that the model constructed by this experiment fitted well with the actual situation, and the best process obtained was feasible.

Best Technology and Reproducibility Experiments
The software was used to solve the partial derivation of the regression equation, and the optimum technological conditions were selected as follows: degradation temperature 69.71 • C, time 3.06 h, the ratios of hydrogen peroxide to hydrogen peroxide 2.39 and catalyst to hydrogen peroxide 28.9%. According to the actual situation, it was revised as follows: degradation temperature 70 • C, time 3.0 h, hydrogen peroxide dosage 2.4, catalyst dosage 30%. Under these conditions, three repeatability experiments were carried out. The average molecular weight of LWCS ranged from 1515 Da to 1564 Da with an average yield of 58.47%. The quality of LWCS was good. Therefore, it could be shown that the model constructed by this experiment fitted well with the actual situation, and the best process obtained was feasible.

Infrared Spectra of Raw Materials/LWCS
The FT-IR spectrum of Figure 9 shows that the functional group positions of raw chitosan and LWCS had not changed, but the absorption strengths were different. The stretching vibration of ether bond in the pyran ring caused a peak at 1071.76 cm −1 , while the characteristic absorption peak of the beta-pyran glycoside bond was at 897.86 cm −1 . The peaks at 3250-3500 cm −1 were related to the stretching vibration of -NH and -OH; the peaks at 2800-3000 cm −1 were ascribed to the stretching vibration of -CH, stretching vibration of -CH2 and stretching vibration of -CH3. In addition, there was no absorption peak at 1645-2000 cm −1 , indicating that the reaction was carried out by disconnecting the beta-(1,4)-glycoside bond, without carbonyl and carboxyl groups. The above results verified that the molecular structure of chitosan had not changed before and after degradation, and only the molecular weight had been reduced.  Table 3 shows that with the increase of PTA loading, the porosity and pore volume decreased regularly, indicating that PTA molecules had been adsorbed and filled into the pore of activated carbon when the loading was 28.34%, the average pore size of the PTA catalyst was the lowest, and the total pore-specific surface area was relatively large. In addition, as the density of PTA was much  Figure 8. FT-IR spectrum of the supported PTA catalyst.

Infrared Spectra of Raw Materials/LWCS
The FT-IR spectrum of Figure 9 shows that the functional group positions of raw chitosan and LWCS had not changed, but the absorption strengths were different. The stretching vibration of ether bond in the pyran ring caused a peak at 1071.76 cm −1 , while the characteristic absorption peak of the beta-pyran glycoside bond was at 897.86 cm −1 . The peaks at 3250-3500 cm −1 were related to the stretching vibration of -NH and -OH; the peaks at 2800-3000 cm −1 were ascribed to the stretching vibration of -CH, stretching vibration of -CH 2 and stretching vibration of -CH 3 . In addition, there was no absorption peak at 1645-2000 cm −1 , indicating that the reaction was carried out by disconnecting the beta-(1,4)-glycoside bond, without carbonyl and carboxyl groups. The above results verified that the molecular structure of chitosan had not changed before and after degradation, and only the molecular weight had been reduced.

Infrared Spectra of Raw Materials/LWCS
The FT-IR spectrum of Figure 9 shows that the functional group positions of raw chitosan and LWCS had not changed, but the absorption strengths were different. The stretching vibration of ether bond in the pyran ring caused a peak at 1071.76 cm −1 , while the characteristic absorption peak of the beta-pyran glycoside bond was at 897.86 cm −1 . The peaks at 3250-3500 cm −1 were related to the stretching vibration of -NH and -OH; the peaks at 2800-3000 cm −1 were ascribed to the stretching vibration of -CH, stretching vibration of -CH2 and stretching vibration of -CH3. In addition, there was no absorption peak at 1645-2000 cm −1 , indicating that the reaction was carried out by disconnecting the beta-(1,4)-glycoside bond, without carbonyl and carboxyl groups. The above results verified that the molecular structure of chitosan had not changed before and after degradation, and only the molecular weight had been reduced.  Table 3 shows that with the increase of PTA loading, the porosity and pore volume decreased regularly, indicating that PTA molecules had been adsorbed and filled into the pore of activated carbon when the loading was 28.34%, the average pore size of the PTA catalyst was the lowest, and the total pore-specific surface area was relatively large. In addition, as the density of PTA was much  Figure 9. FT-IR spectrum for chitosan/LWCS. Table 3 shows that with the increase of PTA loading, the porosity and pore volume decreased regularly, indicating that PTA molecules had been adsorbed and filled into the pore of activated carbon when the loading was 28.34%, the average pore size of the PTA catalyst was the lowest, and the total pore-specific surface area was relatively large. In addition, as the density of PTA was much higher than that of activated carbon, the apparent density of activated carbon loaded with PTA increased with the increase of loading amount.  Figure 10 shows that the pore structure of the supported PTA catalyst had not changed under different preparation conditions, and the pore size was mainly distributed at 90 µm. With the increase of loading, the strength of the peak decreased gradually, which was due to the decrease of pore volume of activated carbon caused by PTA immersion in the pore structure of activated carbon.

Pore Structure Analysis of Supported PTA Catalyst
Catalysts 2020, 10, x FOR PEER REVIEW 9 of 12 higher than that of activated carbon, the apparent density of activated carbon loaded with PTA increased with the increase of loading amount. 1.79 Figure 10 shows that the pore structure of the supported PTA catalyst had not changed under different preparation conditions, and the pore size was mainly distributed at 90 um. With the increase of loading, the strength of the peak decreased gradually, which was due to the decrease of pore volume of activated carbon caused by PTA immersion in the pore structure of activated carbon.

Preparation of the Activated Carbon-Supported PTA Catalyst
The activated carbon-supported PTA catalyst was prepared by the isovolumetric impregnation method, and the process was improved. A certain amount of activated carbon was soaked with 2 mol/L HCl at room temperature for 24 h, then washed with distilled water to no Cl-(silver nitrate test, no white precipitation formation), dried, sealed and stored for standby. W0 g activated carbon was added to a certain concentration of PTA soaking solution, heated and stirred to reflux for a certain time, then treated with Probe-type Ultrasonic Cell Crusher three times: power 75 Hz, pulse on 2 s, pulse off 4 s, time 3 min. Then, the solution was filtered by a water circulation multifunction vacuum pump. The filter cake was dried to a constant weight and weighed as W1 g. Activated carbon-

Preparation of the Activated Carbon-Supported PTA Catalyst
The activated carbon-supported PTA catalyst was prepared by the isovolumetric impregnation method, and the process was improved. A certain amount of activated carbon was soaked with 2 mol/L HCl at room temperature for 24 h, then washed with distilled water to no Cl-(silver nitrate test, no white precipitation formation), dried, sealed and stored for standby. W 0 g activated carbon was added to a certain concentration of PTA soaking solution, heated and stirred to reflux for a certain time, then treated with Probe-type Ultrasonic Cell Crusher three times: power 75 Hz, pulse on 2 s, pulse off 4 s, time 3 min. Then, the solution was filtered by a water circulation multifunction vacuum pump. The filter cake was dried to a constant weight and weighed as W 1 g. Activated carbon-supported PTA catalysts with different loading can be prepared by controlling the concentration of PTA and the stirring time of reflux heating. The formula for the calculation of loading of supported catalyst is as follows: where W 0 -quantity of activated carbon, g; W 1 -quantity of filter cake, g.

Preparation of Low Molecular Weight Chitosan
Two grams of chitosan was put into a conical flask and combined with PTA, water and 30% (wt%) H 2 O 2 aqueous solution. The solution was stirred at a certain temperature, and the recording time was started at the same time. After the reaction, the solution was filtrated, the collected liquid was adjusted with NaOH until reaching pH 9-10, and then the filtrate was distilled under reduced pressure. Calculation of water-soluble chitosan yield is as follows: where Y-yield, %; M 0 -raw material chitosan quantity, g; M x -the quantity of oligochitosan obtained, g.

Determination of Average Molecular Weight of Chitosan
The average molecular weight of chitosan was determined by end-group analysis. A certain concentration of glucosamine hydrochloride was used as the standard solution. The absorbance of the standard solution was measured at 420 nm wavelength, and the standard curve was drawn as shown in Figure 11. Then, a certain concentration of chitosan solution was prepared, and the absorbance of chitosan was determined by repeating the steps mentioned above. The volume of the reference solution corresponding to the absorbance of the product chitosan solution was determined by the standard curve. The average molecular weight of chitosan was calculated according to the following formula.
where M n -average molecular weight of chitosan; C sample -concentration of chitosan solution, g/mL; V sample -volume of chitosan solution, mL; C baseline -concentration of glucosamine hydrochloride solution, g/mL; V baseline -volume of glucosamine hydrochloride solution, mL; M baseline -molecular weight of glucosamine hydrochloride, 216.5 g/mol.
PTA and the stirring time of reflux heating. The formula for the calculation of loading of supported catalyst is as follows: where W0-quantity of activated carbon, g; W1-quantity of filter cake, g.

Preparation of Low Molecular Weight Chitosan
Two grams of chitosan was put into a conical flask and combined with PTA, water and 30% (wt%) H2O2 aqueous solution. The solution was stirred at a certain temperature, and the recording time was started at the same time. After the reaction, the solution was filtrated, the collected liquid was adjusted with NaOH until reaching pH 9-10, and then the filtrate was distilled under reduced pressure. Calculation of water-soluble chitosan yield is as follows: where Y-yield, %; M0-raw material chitosan quantity, g; Mx-the quantity of oligochitosan obtained, g.

Determination of Average Molecular Weight of Chitosan
The average molecular weight of chitosan was determined by end-group analysis. A certain concentration of glucosamine hydrochloride was used as the standard solution. The absorbance of the standard solution was measured at 420 nm wavelength, and the standard curve was drawn as shown in Figure 11. Then, a certain concentration of chitosan solution was prepared, and the absorbance of chitosan was determined by repeating the steps mentioned above. The volume of the reference solution corresponding to the absorbance of the product chitosan solution was determined by the standard curve. The average molecular weight of chitosan was calculated according to the following formula.

Single-Factor Experiment
The effects of the PTA catalyst, hydrogen peroxide amount, catalyst amount, degradation temperature and degradation time on the degradation of chitosan were investigated, with the yield and average molecular weight of chitosan as indicators. The fixed parameters were as follows: the

Single-Factor Experiment
The effects of the PTA catalyst, hydrogen peroxide amount, catalyst amount, degradation temperature and degradation time on the degradation of chitosan were investigated, with the yield and average molecular weight of chitosan as indicators. The fixed parameters were as follows: the loading of the PTA catalyst was 28.34%, the ratio of hydrogen peroxide to chitosan was 2, the ratio of catalyst to hydrogen peroxide was 0.02, the degradation temperature was 70°C and the degradation time was 3 h.

Response Surface Experiment
Based on the results of the single-factor experiment, the response surface experiment was designed by selecting four factors: hydrogen peroxide dosage (A), catalyst loading (B), reaction time (C) and reaction temperature (D). The response surface was optimized by using the ratio of the average molecular weight of chitosan product to the yield (Y 1 %). The factor levels were −1, 0 and 1. The factor levels are shown in Table 4.

Fourier Transform Infrared Spectrometer (FT-IR)
The appropriate amount of supported PTA catalyst, chitosan and low molecular weight chitosan (LWCS) samples were respectively mixed with 100 mg KBr, and the thin films were prepared by a pressing method. Then, the samples were analyzed by a Bruker Tensor-27 Fourier Transform Infrared Spectrometer (TENSOR, Germany) at wavelengths in the range of 400-4000 cm −1 .

Mercury Intrusion Porosimetry (MIP)
The AutoPore IV 9500 mercury porosimeter produced by the Mike Company of America was used to measure the pore size of PTA with different loading samples ranging from 0 to 420,000 nm (about 5 mm in diameter, 25 • C). Then, the sample was placed into the instrument for the experiments, and finally, the relevant parameters of the pore of the PTA sample were obtained.

Conclusions
The efficiency of the supported PTA catalyst was significantly higher than that of a non-supported catalyst. The optimum reaction conditions of degradation of high molecular weight chitosan by the catalyzed oxidation of hydrogen peroxide were obtained as follows: the reaction temperature 70 • C, the reaction time 3.0 h, the ratio of hydrogen peroxide to chitosan 2.4 and catalyst loading 30%. The yield and molecular weight of water-soluble chitosan were 62.8% and 1290 Da, respectively. The quadratic equation simulation was obtained. The simulation regression was significant, and the experiment was well fitted, which had a certain value.
The pore structure of the supported PTA catalyst did not change under different preparation conditions, and the pore size was mainly distributed at 90 µm. The supported PTA catalyst maintained high catalytic activity after three reuses, which indicated that the supported PTA catalyst had excellent catalytic activity and stable performance compared with the PTA catalyst. This shows a new process for effectively preparing water-soluble chitosan that meets the requirements of green chemistry.