Green Conversion of Saponins to Diosgenin in an Alcoholysis System Catalyzed by Solid Acid Derived from Phosphorus Tailings.

This work proposed to prepare solid acid from phosphorus tailings and successfully convert Dioscorea zingiberensis C.H. Wright (DZW) into diosgenin from the perspective of solid waste resource reuse and clean production. The results showed that SiO2-SO3H solid acid could catalyze the production of diosgenin from total saponins under solvothermal reaction conditions. In addition, the parameters of a single factor, such as the amount of SiO2-SO3H, solvent volume, reaction temperature, and reaction time, were optimized to confirm the optimal range of reaction conditions, and the optimal process conditions were determined by the response surface method. The yield of diosgenin was 2.45 ± 0.17% under the optimum conditions, and the yield of diosgenin was increased by 12.90% compared with the traditional acid hydrolysis process. Except the relatively higher catalytic activity, the alcoholysis approach for the production of diosgenin has no waste liquid to discharge. The products were analyzed by high-performance liquid chromatography-mass spectrometry, and the pathway to convert total saponins into diosgenin under SiO2-SO3H has been proposed. Moreover, the adopted catalyst can be prepared with very low cost from phosphorus tailings. Considering the obvious superiorities, the alcoholysis approach in this work could be a promising strategy for green production of diosgenin as well as a possible utilization pathway of phosphorus tailings.

Introduction

Dioscorea zingiberensis C.H. Wright (DZW) is a common vine grown widely in China.[1] The rhizome of DZW contains diosgenin, an important raw material for the preparation of steroid hormone drugs.[2] In recent years, scientists from various countries have found that diosgenin has very important and extensive biological activities, such as anti-inflammatory effects,[3] antithrombotic effects,[4] cardio-protective effects,[5] anthelmintic properties,[6] and anticancer effects.[7] However, the method for industrial production of diosgenin is mainly through acid hydrolysis of glucoside bonds.[8] A large amount of water with low pH and high chemical oxygen demand (COD) is discharged during the production, leading to serious environmental pollution. In addition, the chlorinated reaction taking place in the process of inorganic acid hydrolysis would reduce the yield of diosgenin.[9]

In order to reduce pollution, protect the environment, and achieve the goal of sustainable development, many methods have been studied to replace direct acid hydrolysis to reduce COD discharge, such as microbial hydrolysis by Trichoderma reesei,[10] enzymatic hydrolysis,[11] catalytic hydrolysis with ion solution under microwave irradiation,[12] supercritical CO2 methods,[13] and pressurized biphase acid hydrolysis.[14] However, these methods are difficult to be applied in industrial production due to their respective problems. For example, although these methods significantly reduce the content of COD in wastewater, there is still a problem of potential environmental pollution, which is not a long-term solution. Enzymatic hydrolysis and natural fermentation hydrolysis in the biological method have the problems of a longer production period, difficult condition control, and high cost. The ion solution method has a high cost, complicated recovery process, and is easily contaminated by byproducts. The prestage investment of the supercritical CO2 method is large, the equipment requirements are harsh, and the high-pressure reaction operation is risky. Therefore, it is very important to explore an efficient, green, economical, and pollution-free catalytic way to obtain diosgenin. In recent years, solid acid as a cleaner production catalyst has achieved excellent performance in various research fields.[15] Prior to this, we also successfully used solid acid to extract diosgenin from saponins of DZW.[16−18] The studies showed that the alcoholysis method is a green approach to produce diosgenin from DZW, which does not have liquid waste discharging in the process of alcoholism; however, the cost of the solid acid greatly affects the production efficiency. Low cost for the preparation of the solid acid is the key factor to realize the potential applications.

Phosphate rocks, on the other hand, are the raw material for the preparation of fertilizers, detergents, animal food supplements, etc. The demand for phosphate rocks in China and even in the world is growing. With the consumption of the phosphate ore, about 30–40% of the quality ore is being discarded as tailings in processing operation of mining and the ore dressing process.[19,20] More and more phosphorus tailings would stack as waste disposals if no effective utilization method is developed for them. However, a large number of tailing storage in tailing ponds is a great potential safety hazard, and accidents have been caused many times (Yang et al., 2017). Therefore, there is a need to explore a new way to reuse tailings.

In this study, a simple inorganic reaction was used to extract the main element Si from the phosphorus tailings for the synthesis of hydrophilic SiO2 microspheres and then the surface of SiO2 was sulfonated to prepare the silicon-based solid acid materials. The feasibility of using phosphorus tailings to synthesize silicyl solid acid to catalyze the alcoholization of turmeric was studied. The optimum reaction conditions for extracting diosgenin from saponins in DZW were obtained by the response surface method (RSM) based on the results of the single factor experiment. The results showed that the application of the as-prepared solid acid can not only provide an approach to reduce both solid waste of phosphorus tailings and the potential safety hazards caused by accumulation but also propose a green method to overcome the environmental pollution caused by the hydrolysis of DZW at the present stage. The clean diosgenin production promoted by the utilization of waste material will have great significance for both reducing environmental pollution and sustainable development.

Results and Discussion

Catalyst Characterization

FTIR Analysis

Fourier transform infrared (FTIR) analysis is one of the important tools for identifying functional groups in samples. Figure S1 shows FTIR spectra of SiO2 and SiO2–SO3H, respectively. The FTIR spectra of SiO2 microspheres show two weak absorption bands at 3450 and 1634 cm–1, corresponding to the silica hydroxyl stretching and bending vibration on the silica surface, respectively.[21,22] The wide strong band centered at 1095 cm–1 is assigned to Si–O–Si asymmetric stretching vibration. The bands at 798.81 and 469 cm–1 are symmetric stretching vibrations for the Si–O bond. Meanwhile, the existence of Si–OH symmetric stretching vibration is also confirmed by the absorption band at 965 cm–1. FTIR spectra of SiO2–SO3H microspheres show two characteristic adsorption bands at 3450.50 and 1634.91 cm–1 related to the O–H stretching and bending vibrations of the SO–H bond of the sulfonic acid groups (SO3H).[23] The FTIR spectra of SiO2–SO3H microspheres at 3354 and 1609 cm–1 are significantly larger than those of SiO2; this is due to the enhancement of the characteristic peak due to the coincidence of the SO–H stretching and bending vibration with the O–H stretching peak and the sulfonic acid group.[24]

Thermogravimetric Analysis

To further demonstrate the successful synthesis of SiO2–SO3H, the thermogravimetric analysis (TGA) studies of SiO2 and SiO2–SO3H and the relative analyses were conducted. Figure S2 shows the weight loss of SiO2–SO3H from room temperature to 800 °C. The weight loss (7%) from room temperature to 100 °C was caused by the removal of physically adsorbed water.[25,26] The weight loss (3%) in 100–200 °C can be assigned to the physical adsorption −SO3H were lost in the weak acid region. The 5% weight loss in the temperature range of 200–800 °C is due to the −SO3H groups covalently bonded with SiO2 as well as the condensation of the hydroxyl groups on the solid acid.[25]

XRD Analysis

Next, X ray diffraction (XRD) was used to study the crystalline structure of SiO2 sulfonated microspheres, as shown in Figure S3. There are broad peaks in the range of 2θ between 15 and 28°, and no other peaks, indicating the high purity of SiO2. However, the XRD of the sample did not show any significant diffraction peaks for −SO3H layers, indicating that the −SO3H amorphous shell on the SiO2 surface was too thin to be measured by XRD.[27]

NH3-TPD Analysis

As shown in Figure , NH3-temperature programmed desorption (TPD) of the SiO2–SO3H desorption curve has a small peak in 100 °C (weak acid). The presence of weak acid sites on NH3-TPD may be due to the interaction between NH3 molecules and the −OH groups attached to the catalyst surface.[28] The strong broad peak centered at 500 °C (strong acid >400 °C) belong to −SO3H bonded with SiO2.[29] The calculated acidity of SiO2–SO3H was 0.93 mmol/g.

Figure 1

NH3-TPD of SiO2–SO3H.

SEM, TEM, and TEM-Mapping Analysis

The morphologies of SiO2–SO3H particles were characterized by scanning electron microscopy (SEM), high resolution transmission electron microscopy (HRTEM), and transmission electron microscopy (TEM)-mapping. SEM and TEM revealed that the size of SiO2–SO3H was mostly in the range of 400–500 nm (Figure ). In the preparation of SiO2 from the Na2SiO3 solution, a small amount of EtOH and PEG was added to avoid the aggregation of SiO2, and the addition of ethyl acetate was to decrease the pH value in the preparation of SiO2. Due to the fact that the preparation of SiO2 was performed by the precipitation reaction between Na2SiO3 and diluted H2SO4, the reaction speed was very fast, leading to the partial aggregation and the sizes of SiO2 were not completely consistent. Although there were no significant changes in the SEM and TEM images of the SiO2 samples before and after sulfonation, the existence of S can be determined by energy dispersive X-ray analysis (EDX) (Figure ), and the TEM-mapping analysis, shown in Figure E–H, further confirms that O, Si, and S are uniformly distributed in the SiO2–SO3H particles.

Figure 3

(A and B) SEM of SiO2–SO3H; (C and D) HRTEM of SiO2–SO3H; (E–H) TEM-mapping of SiO2–SO3H.

Figure 2

EDX of SiO2–SO3H.

Single Factor Experiment

Effect of Different Amount of SiO2–SO3H

Alcoholysis activity of SiO2–SO3H depended mainly on the acidic groups −SO3H in the solid acid, so the amount of SiO2–SO3H was the key factor for alcoholysis. To find out the optimal mass of SiO2–SO3H in the alcoholysis reaction, the effect of the SiO2–SO3H dosage (0.2–0.6 g) on the yield was explored and shown in Figure . With increase of the dosage of SiO2–SO3H from 0.2 to 0.5 g, the yield of diosgenin increased gradually, indicating that SiO2–SO3H plays an irreplaceable role in the alcoholysis of saponins and directly affects the yield of diosgenin. When the amount of SiO2–SO3H reached 0.5 g, the yield of diosgenin was the highest. However, as more SiO2–SO3H was used for alcoholysis, the diosgenin yield decreased slightly, indicating that the high amount of SiO2–SO3H might destroy the generated diosgenin and result in the formation of a side reaction. It was found that diosgenin was easily dehydrated to Δ3,5-deoxytigogenin.[30,31]

Figure 4

Effect of different quantities of SiO2–SO3H.

Effect of Solvent Volume

Anhydrous ethanol has a great influence on the yield of diosgenin in the hydrolysis of total saponins in traditional processes. Therefore, the effect of different solvent volumes (4–8 mL) on the alcoholysis reaction was studied. The results are shown in Figure . It showed that the yield rises rapidly when the volume of ethanol is increased from 4 to 6 mL. After the solvent reached 6 mL, the yield of diosgenin decreased with further increase in the ethanol volume, which can be ascribed to the decrease in the contact probability between SiO2–SO3H and total saponins.[16]

Figure 5

Effect of solvent volume.

Effect of Reaction Time

In the conventional hydrolysis, the hydrolysis reaction time was an important factor. Therefore, the influence of reaction time was investigated. As shown in Figure , the yield increased linearly when the reaction time increased from 4 to 6 h, indicating that the total saponins were more fully transferred into diosgenin and sugars with the increase in reaction time. The highest yield was obtained when the reaction time was 6 h. However, the yield of diosgenin decreased with further increase in the reaction time, which can be contributed to the decomposition of diosgenin in the longer reaction time.[12]

Figure 6

Effect of reaction time.

Effect of Reaction Temperature

The reaction temperature is a very important factor that directly affects all chemical reactions. Therefore, the yield in the process was investigated from 70 to 110 °C, shown in Figure . The optimum reaction temperature was 100 °C. When the temperature increased from 70 to 100 °C, the yield of diosgenin was significantly increased. The rise in the temperature accelerated the alcoholysis reaction and enhanced the SiO2–SO3H catalytic efficiency. However, the yield of diosgenin showed little change when the reaction temperature exceeded 100 °C.

Figure 7

Effect of reaction temperature.

Analysis of Response Surface Results

The results of the above single factor tests showed that the yield of diosgenin in the 100 °C temperature (yield 2.30%) was almost the same with that in 110 °C (yield 2.33%). For the consideration of reducing energy consumption, we set the reaction temperature at 100 °C. The independent variables and relative levels are listed in Table , the amount of SiO2–SO3H (X1) was 0.4∼0.6 g, reaction time (X2) was 5∼7 h, and ethanol solvent volume (X3) was 5∼7 mL.

Table 1

Independent Variables and Encoding Levels for the Box–Behnken Design

independent variables	symbols	coded levels
		-1	0	1
SiO2–SO3H/g	X1	0.4	0.5	0.6
reaction time/h	X2	5	6	7
ethanol/mL	X3	5	6	7

Box–Behnken design (BBD) software was used to optimize the variables and levels, and 17 experimental schemes were designed. After the Y% values obtained from the experiment for each scheme design were entered, the response regression model was established (Table ).

Table 2

Box–Behnken Design (Actual)

run	X1/g	X2/h	X3/mL	Y%
1	1.000	–1.000	0.000	1.891
2	0.000	0.000	0.000	2.450
3	–1.000	0.000	–1.000	2.131
4	–1.000	1.000	0.000	2.112
5	0.000	–1.000	1.000	1.889
6	1.000	0.000	–1.000	2.262
7	0.000	1.000	–1.000	2.232
8	1.000	0.000	1.000	2.370
9	0.000	0.000	0.000	2.412
10	0.000	–1.000	–1.000	1.909
11	0.000	0.000	0.000	2.425
12	1.000	1.000	0.000	2.30
13	–1.000	0.000	1.000	2.128
14	–1.000	–1.000	0.000	1.820
15	0.000	0.000	0.000	2.441
16	0.000	1.000	1.000	2.333
17	0.000	0.000	0.000	2.424

Seventeen runs designed by the BBD of the RSM were used to optimize the conditions for alcoholysis of total saponins to diosgenin with solid acid. The predicted response of Y for the yield of diosgenin extracts can be expressed by the following second-order polynomial equation:

The model of the RSM was verified by analysis of variance (ANOVA) for the yield of diosgenin (Table ). These results showed that the P-value of the regression model was less than 0.0001 and the ANOVA lack-of-fit was 0.1394, demonstrating that the model is “highly significant” and the lack-of-fit is “not significant”.[32] Moreover, the coefficient of multiple determination (R2) for the yield of diosgenin was 99.59% and the coefficient of adjustment (Adj-R2) was 99.06%, C.V. = 0.96%, indicating that the resultant second-order equation fits well. In the variables, the highly significant effect (<0.0001) factors were the SiO2–SO3H quantity and reaction time, and the quadratic terms of the SiO2–SO3H quantity and reaction time also had highly significant (<0.001) effects. All other reaction conditions (X3, X1X2, X1X3, X2X3, and X3[2]) were significant (<0.05) on the yield of diosgenin.[33]

Table 3

ANOVA for the Response Surface Quadratic Modela

source	sum of squares	Df	mean square	F	P-value
model	0.76	9	0.085	188.53	<0.0001	significant
X1	0.050	1	0.050	111.06	<0.0001
X2	0.27	1	0.27	599.34	<0.0001
X3	4.325 ×10–3	1	4.325×10–3	9.62	0.0173
X1X2	3.422 ×10–3	1	3.422×10–3	7.61	0.0281
X1X3	3.080×10–3	1	3.080×10–3	6.85	0.0345
X2X3	3.660×10–3	1	3.660 ×10–3	8.14	0.0246
X12	0.075	1	0.075	167.77	<0.0001
X22	0.30	1	0.30	661.97	<0.0001
X32	0.023	1	0.023	51.06	0.0002
residual	3.14 ×10–3	7	4.495×10–4
lack of fit	2.24 ×10–3	3	7.470×10–4	3.30	0.1394	not significant
pure error	9.052×10–4	4	2.263×10–4
cor. Total	0.77	16

R2 = 0.9959 Adj-R2 = 0.9906 C.V.% = 0.96.

The three-dimensional response surface was drawn according to the BBD experimental data (Figures –10), and the interaction between various factors was studied to determine the optimal value to achieve the maximum yield.

Figure 8

Effect of interaction of two parameters reaction time (X2) and solid acid dosage (X1): (A) response surface 3D plot and (B) contour plot.

Figure 10

Effect of interaction of two parameters ethanol solvent volume (X3) and reaction time (X2): (A) response surface 3D plot and (B) contour plot.

Effect of interaction of two parameters ethanol solvent volume (X3) and solid acid dosage (X1): (A) response surface 3D plot and (B) contour plot.

Based on the aforementioned RSM model, the optimal extraction conditions for alcoholysis of total saponins was estimated to be 0.54 g (amount of SiO2–SO3H), 6.38 h (reaction time), 6.30 mL (ethanol), and 100 °C, the optimum yield of diosgenin was 2.48%. According to the optimal reaction conditions calculated by the RSM, we carried out three independent experiments to verify that the average yield was 2.45%, which was very close to the calculated data of the RSM.

Recycling Experiments

Although solid acid catalysts were recyclable and had high-level catalytic activity in the alcoholysis process, 0.03–0.05 g solid acid will be lost in the process of centrifugal separation, drying, regeneration, and recycling of the recovered catalyst.[34] In order to ensure the activity of the catalyst and avoid the influence of operational loss on the catalytic activity, the corresponding loss amount of the solid acid was mixed with the recovered solid acid as the catalyst for the next run.

The results for the recycling experiments are shown in Figure , the activity of the catalyst decreases a little after each run, which may be ascribed to the decrease in active sites. The sugars produced by alcoholysis has a low solubility in ethanol and can deposit on the surface of solid acid in the process of centrifugal separation, resulting in a part of the active sites of solid acid being covered. However, it only decreased by 0.57% after five runs, indicating its good stability and reusability. The solid acid collected after five cycles can be easily reactivated by mixing it with diluted sulfuric acid at room temperature for 3 h, and its activity was tested after washing and drying, which can restore the yield of diosgenin to 2.34%. The reactivated process is easy to do, and the used sulfuric acid in the reactivated process can be reused in the next recycle.

Figure 11

Recycling experiments of SiO2–SO3H in the alcoholysis.

The stability of the solid acid was also verified by XRD and EDX results, shown in Figures and 13. It showed that the XRD curve of SiO2–SO3H after five times of recycling use is almost the same with that of fresh SiO2–SO3H. The EDX of SiO2–SO3H after five runs is basically the same as that of fresh solid acid (Figure ). It indicating that the solid acid is very stable in the alcoholysis process.

Figure 12

XRD (A) fresh Fe3O4@SiO2–SO3H and (B) Fe3O4@SiO2–SO3H after five runs.

Figure 13

EDX of SiO2–SO3H after five runs.

Comparison with the Different Extraction Methods

In order to evaluate the solid acid-catalyzed alcoholysis efficiency of total saponins to diosgenin, the conventional method to extract diosgenin from DZW was also performed according to the general procedure. The contents of diosgenin in the samples were detected by high pressure liquid chromatography (HPLC) and UV–vis after hydrolysis. The results are listed in Table . It shows that the alcoholysis method catalyzed by SiO2–SO3H had a higher diosgenin yield than other methods. The alcoholysis yield of solid acid was 12.90% higher than that of the H2SO4 hydrothermal reaction. In order to exclude the influence of experimental operation on the results, the same method was used for the blank experiment except for no addition of solid acid. As a result, the yield of diosgenin in blank extraction was much lower than that of other experiments, indicating that the solid acid plays an crucial role in increasing the yield of diosgenin extracted from diosgenin of DZW. At the same time, we also compared the reaction conditions and yield of other different extraction methods of diosgenin.

Table 5

Main Oxide Composition of Phosphorus Tailings

sample	SiO2	CaO	Al2O3	Fe2O3	P2O5	MgO	K2O
PT (%)	42.08	19.73	8.19	5.49	4.80	2.13	6.08

It can be seen from Table that the solid acid in this work has a comparable catalytic activity to those with better catalysts reported in the literature.

Table 4

Comparison Experiment Results

number	catalyst	solvents	method	yield of diosgenin (%)
1	H2SO4	H2O	oil bath	2.17 ± 0.15
2	H2SO4	H2O	hydrothermal reaction	2.05 ± 0.16
3	SiO2	ethanol	solvothermal reaction	0. 29 ± 0.11
4	SiO2–SO3H	ethanol	solvothermal reaction	2.45 ± 0.17
5	Fe3O4@SiO2–SO3H	ethanol	solvothermal reaction	2.64 ± 0.12[16]
6	St-SAA	ethanol	solvothermal reaction	0.70[18]
7	Fe3O4@ETMS-TETA-SO3H	ethanol	solvothermal reaction	1.62[17]
8	[BHSO3MIm] HSO4	ionic liquid	microwave irradiation	1.02[12]
9	Aspergillus awamori	H2O	microbial hydrolysis	2.58[35]
10	none	ethanol	solvothermal reaction	0.36 ± 0.12

LC–MS Analysis

In order to better understand the alcoholysis process of saponins to diosgenin by solid acid and prove that diosgenin existed in the alcoholysis products, liquid chromatography-mass spectrometry (LC–MS) detection was carried out to analyze standard diosgenin, total saponins, and the alcoholysis samples before and after purification, shown in Figure . The peak of the standard diosgenin appeared in a retention time of 7.80 min (Figure A), and the peak of the alcoholysis products before and after purification also appeared in about 7.80 min (Figure C,D), but the peak at about 7.80 min was not observed in LC–MS of the total saponins (Figure B). The MS spectra showed that the peak at about 7.80 min has an m/z value of 415 (Figure ), which is the aggregation of one diosgenin molecule combining with one H+, proving that diosgenin is involved in alcoholysis products.[36,37]

Figure 14

LC–MS of (A) standard of diosgenin, (B) total saponins, (C) before purification of diosgenin, and (D) after purification of diosgenin.

Figure 15

Total ion chromatogram (TIC) of (A) standard of diosgenin; and (B) alcoholysis product, time: 7.80 min.

The other components in the alcoholysis products were further investigated via LC–MS, and the results are exhibited in Figures S4 and S5. On the basis of all the experimental results and the previous studies, we believe that there are two possible pathways for the transfer of total saponins to diosgenin, shown in Figures and 17, where various fragment peaks with different m/z values can be found in Figures S4 and S5. The structures and formulas for each of the fragments are their relative assignments, which match very well with the practical m/z values. It can be seen that there are some substances containing ethyl glucosides in the alcoholysis, such as the substances with m/z = 443.29, 207.98, and 605.46, giving a strong support for the release of diosgenin from saponins by alcoholysis.

Figure 16

Possible alcoholysis path one of the solid acid catalyst.

Figure 17

Possible alcoholysis path two of the solid acid catalyst.

Conclusions

This work introduces a method for preparing solid acid of SiO2–SO3H using phosphorus tailings (PT) as a raw material. The successful synthesis of SiO2 and the modification of SiO2 with sulfonic acid groups were verified by IR, TGA, XRD, NH3-TPD, SEM, TEM, and TEM-mapping. The total saponins in DZW were successfully converted into diosgenin by SiO2–SO3H catalytic alcoholysis under the condition of solvothermal reaction. Through the single factor test and RSM, the optimal conditions were determined as follows: SiO2–SO3H 0.54 g, reaction time 6.30 h, ethanol solution 6.38 mL, and reaction temperature 100 °C. The highest yield of saponin was 2.45%. Moreover, the alcoholysis reaction mechanism was studied through LC–MS analysis. It was found that the solid acid catalyst is not only environmentally friendly and pollution-free but also provides a solution to the environmental pollution of phosphorus tailings. This method has a broad industrial prospect in the production of diosgenin and the comprehensive utilization of phosphorus tailings.

Experimental Section

Materials

High silicon phosphate tailings (PT) were obtained from Hubei Ezhong Ecological Engineering Co. Ltd., and the composition of the PT is shown in Table identified by XFR analysis. Na2CO3 (AR), H2SO4 (AR), sodium dodecyl sulfate (SDS, AR), anhydrous ethanol (AR), polyethylene glycol (PEG-1000), ethyl acetate (AR), dichloromethane (CH2Cl2), and chlorosulfonic acid (ClSO3H) were purchased from Sinopharm Chemical Reagent Co., Ltd. Dioscorea zingiberensis C. H. Wright (DZW) was provided by Danao Pharmaceutical Co., Ltd., Hubei, China. Chloroform (AR) and acetonitrile (chromatographically pure) were used in the HPLC analysis, and perchloric acid (AR) and diosgenin (analytical standard) were purchased from Aladdin. The reagents and chemicals were used without further purification.

Synthesis of Sulfonated Silica Microspheres

Preparation of Silica Microspheres from Phosphorus Tailings

The process for preparing solid acid from PT was performed in a similar way as reported in literature studies.[38,39] Typically, 25 g of PT were ball-milled to fine powder and 99% of it was passed through 300 mesh and then the screen underflow was transferred to a ceramic crucible. Anhydrous sodium carbonate (50 g) was added and mixed evenly. Then, it was put into a muffle furnace and heated to 850 °C for 4 h. After natural cooling to room temperature, the solid was transferred to a beaker and 250 mL of deionized water was added. After stirring at 50 °C for 4 h, the mixture was filtered while hot.

The silicon-containing leaching solution obtained above was transferred to a 500 mL beaker, and 0.5 g PEG-1000, 2 g EtOH, and 3.7 g ethyl acetate were added into the above solution in turn and mixed evenly; then, the pH value of the leaching solution was adjusted to 1 with 10% H2SO4; a white precipitate was formed at 60 °C for 3 h. The obtained SiO2 composites were washed with water until the filtrate was neutral, dried, and calcined at 400 °C.

Synthesis of Sulfonated Silica Microspheres Materials

The sulfonated silica microspheres materials were prepared in a similar method for silica-supported sulfonic acids.[26] First, the prepared SiO2 (2.5 g) was transferred to a 50 mL two-neck flask, which was connected to a gas absorption device to absorb the escaped HCl reaction gas from the sulfonation reaction. Dry dichloromethane (20 mL) was added and the resulting mixture was dispersed by ultrasound. Then, chlorosulfonic acid (2 mL) was added dropwise and stirred for 3 h at 0–5 °C. After the reaction, the catalyst was filtered, washed to neutral with CH2Cl2 and ethanol, and dried overnight in vacuum.

Characterization

X-ray fluorescence (XRF, ARL PERFORM’X) was used to determine the composition of the phosphorus tailings. FTIR spectra were recorded in a spectrometer (IASST, Guwahati) in the range of 500–4000 cm–1. XRD data were recorded by BRUKER D6 with radiation (λ = 1.54 Å) at 40 kV 44 mA. TGA was tested by Mettler-Toledo TG2 series under the conditions of heating rate 10 °C/min from 25 to 800 °C and a nitrogen atmosphere at flow rate 100 mL/min. The morphologies of the catalyst were recorded by a transmission electron microscope (JEM2100F, 200 kV, JEOL) and a scanning electron microscope. The acid value was determined by TPD-NH3 (AutoChem II2920). UPLC-MS (Agilent 1100, thermos TSQ quantum Ultra AM) was used for qualitative and quantitative analyses.

Alcoholysis of Total Saponin by the SiO2–SO3H Catalyst

The alcoholysis schematic diagram is shown in Scheme . DZW was washed, dried, and crushed; the fine powder screened through 200 mesh was collected. To extract steroid saponins, the powder (60 g) was mixed with ETOH (600 mL) overnight, then heated at 85 °C, and refluxed for 12 h. The hot mixture was filtered and washed with ethanol. The combined filtrate was concentrated to dry using a rotary evaporator, and the obtained products were used as crude saponins, and dried in vacuum at 40 °C. Finally, 6.65 g of total saponins were obtained with a yield of 11.08%.

Scheme 1

Flow Chart of the Alcoholysis of Saponins for the Production of Diosgenin with SiO2–SO3H Solid Acid from Phosphorus Tailings

The total saponins (0.1 g) extracted in the above experiments were fully dissolved in EtOH (4–8 mL) in Penicillin bottles using ultrasound for 15 min. Then, SiO2–SO3H (0.2–0.6 g) was added and ultrasonically dispersed. The mixture was transferred to a polytetrafluoroethylene-lined stainless-steel autoclave, after sealing. A small amount of solvothermal reaction was programmed at different temperatures (70–110 °C) and reaction times (4–8 h). After each run, SiO2–SO3H was recovered by centrifugation and the centrifugate was evaporated to dry by a rotary evaporator; meanwhile, the ethanol liquid collected from the evaporator was reserved for further use. The obtained solid products were extracted by a Soxhlet extractor using chloroform as solvent for 3 h. The content of diosgenin in the sample was analyzed with HPLC and UV–vis. In order to obtain the optimal alcoholysis conditions, four experimental factors were studied and optimized by the single variable analysis and RSM. The optimal reaction conditions were obtained by studying the effect of SiO2–SO3H, dosage of ethanol, reaction time, and reaction temperature on the yield of diosgenin. Then, the best reaction conditions for the preparation of diosgenin from DZW were obtained by the RSM. Finally, the path of total saponins to diosgenin with solid acid alcoholysis was investigated by LC–MS.

Response Surface Method

The RSM was used to obtain the optimal conditions for extraction of diosgenin. According to the previous results of the single factor test, the three-level-three-factor BBD was designed to calculate the optimal conditions. All RSM experiments were performed with random to minimize the influence of uncertainty of the external unknown factors. A second-order polynomial equation was postulated to express the yield of diosgenin response Y, as eq :[40]where Y is the response (the diosgenin yield); β0 is the regression coefficient for the intercept (constant); and βiand βij are the coefficients for the linear and interaction terms of variables i and j, respectively. βii is the quadratic coefficient and k is the number of variables. The three independent parameters were assigned as: X1(SiO2–SO3H, g), X2(reaction time, h), and X3(ethanol dosage, mL). Therefore, the second-order polynomial models of the three independent variables can be expressed as eq :

Regeneration and Recycling Experiments

The SiO2–SO3H solid acid is a recyclable catalyst. Therefore, we also need to research on its recovery and reuse. After each reaction, the solid acid was separated from the reaction mixture by centrifugation, the recovered SiO2–SO3H was washed with water and ethanol, and dried at 40 °C under vacuum. The recovered SiO2–SO3H will be reused in five cycles under the same reaction conditions. After five cycles, SiO2–SO3H was soaked in dilute sulfuric acid for 3 h to regenerate, and its activity was restored to the original one.

Conventional Sulfuric Acid Hydrolysis Method

For evaluating the novel method, the conventional sulfuric acid hydrolysis method described in the literature was carried out for comparison.[41] A sample of 0.10 g total saponin powder was hydrolyzed in 6 mL 2 mol/L sulfuric acid for 5 h at 100 °C. After the reaction, the mixture was filtered and the solid residues were washed with hot water and alcohol three times and dried at 40 °C in vacuum. Then, the diosgenin in the residues was extracted by a Soxhlet extractor for 3 h with chloroform. The diosgenin content in the sample of sulfuric acid hydrolysis was analyzed under the same conditions as the solid acid alcoholysis.