Lots of highly concentrated saline organic wastewater is produced during the pymetrozine production process, causing environmental pollution and waste of resources if discharged directly. Research on actual pymetrozine wastewater treatment is quite scarce. Existing treatment methods of pesticide wastewater usually have disadvantages of long treatment time, low processing efficiency and low recovery rate. To solve these problems, a pretreatment process for pymetrozine wastewater was studied based on material recovery and pollutant degradation. The ammonia conversion process was experimentally investigated by reactive distillation. The reaction product vapor was neutralized and then separated by side-stream distillation. Aspen Plus and response surface methodology were employed to simulate and optimize the operating conditions. Box-Behnken design was used to investigate the individual and interaction effects on methanol purification and sodium acetate removal. Experimental study was carried out on the basis of theoretical simulation data. The result showed that the optimized methanol content on tower top was 99.28% with a yield of 99.95% and methanol content of side withdrawal was 0.01%. The process can be applied for pesticide wastewater treatment to recycle high purity chemical materials, and meets the national sewage comprehensive emission standard.

  • Pretreatment process of pymetrozine production wastewater is presented.

  • Reactive distillation is carried out for ammonia conversion treatment.

  • Side-stream distillation is operated for methanol and ammonium sulfate recycling, using Aspen Plus and RSM as the simulation and optimization tools.

  • The process helps in recycling methanol of 99.28 wt.% and reaching wastewater discharge standard.

Graphical Abstract

Graphical Abstract
Graphical Abstract

Pymetrozine is a systemic pesticide of the pyridine-azomethin family (Khan et al. 2018). It is used as insecticide to control whiteflies, leafhoppers and cotton aphids in crops, due to its high activity and superior insecticidal effect (Fdez-Sanromán et al. 2020). Because of its extensive use, improper disposal, accumulation ability, long persistence in the environment and long-term effect on living organisms, pymetrozine has been classified as a ‘‘likely’’ human carcinogen and registered for crops with maximum residue in some countries (Jang et al. 2014).

The traditional process of synthesizing pymetrozine is complicated. Multi-step processes are included such as hydrazinolysis, cyclization, alkylation, condensation, hydrolysis (Wang et al. 2012). The synthesis of pymetrozine was conducted as shown in Figure 1 (Zhou et al. 2018). Acetohydrazide (Intermediate I) was obtained via hydrazinolysis using ethyl acetate and hydrazine hydrate as starting materials and then cyclized with phosgene to give Intermediate II. After reaction with chloroacetone, ring-opening reaction and cyclization gave Intermediate III. Then Intermediate III was treated with hydrazine hydrate to obtain Intermediate IV. Subsequently, hydrolysis in the presence of hydrochloric acid and methanol () is carried out to afford Intermediate V. The desired product, pymetrozine, is obtained via the condensation of nicotinaldehyde with Intermediate V in the presence of sodium hydrate.

Figure 1

Traditional synthetic route for pymetrozine.

Figure 1

Traditional synthetic route for pymetrozine.

Close modal

As stated above, a large amount of wastewater was produced, including but not limited to , salty substance (ammo­nium acetate and sodium acetate), water () and a large amount of intermediates of pesticides. As a kind of poor degradable organic wastewater, pymetrozine production wastewater is characterized by high chemical oxygen demand, acidity and high biotoxicity. Most of all, nitrogenous organic compounds in pesticide intermediates will cause microbial toxicity and have a great impact on subsequent biological treatment.

At present, studies on pymetrozine mainly focus on the analysis method (Hou et al. 2019), properties (Olfati Somar et al. 2019), degradations and potential health risk in ecosystems (Elfikrie et al. 2020). But research on actual pymetrozine production wastewater treatment is quite scarce.

Conventional wastewater treatment technologies focus on nutrient and pathogen removal, and are generally ineffective in pesticide removal (Sutton et al. 2019). The treatment methods for pesticide wastewater mainly include physical methods, chemistry methods, biochemical methods and comprehensive methods. Physical methods are used as pretreatment methods for recovering by-products and mainly include extraction (De Gaetano et al. 2016), adsorption (Rajapaksha et al. 2018) and membrane separation (Qin et al. 2020). But a single physical method doesn't work well due to the strong polarity and water solubility of the organic substances in pesticide wastewater. Chemistry methods, as pretreatment processes for biochemical treatment to improve the biodegradability, mainly consist of chemical oxidation (Carra et al. 2016), electro catalytic oxidation microwave assisted catalysts (Tony & Mansour 2019), and so on. The disadvantage of chemistry methods is that compounds are degraded and cannot be reused, in spite of the high degradation rate. Biochemical methods, used as the main treatment process after pretreatment, have disadvantages of long treatment time and low efficiency. Comprehensive methods such as bio-Fenton and bio-electro-Fenton (Kahoush et al. 2018), need to improve the performance, stability and lifetime to achieve more sustainable and cost-effective wastewater treatment. Combined with the research above, sufficient attention should be paid to the pretreatment of pymetrozine production wastewater, which may have a detrimental effect on the environment.

For the purpose of solving practical problems, the pretreatment process for pymetrozine production wastewater was investigated in this paper. The research technical route is shown in Figure 2. The wastewater was initially placed in a reactive distillation apparatus for ammonia conversion treatment, taking advantage of the poor stability of ammonium acetate (). By adding , was converted into and free ammonia. Reactive distillation (RD) was chosen due to its specific advantages, such as improved selectivity, increased conversion, effective utilization of reaction heat, reduced capital investment, and so on. With the combination of reaction and separation, RD can shift the equilibrium of reaction towards the product side by continuous removal of the products and give higher conversions (Gor et al. 2020).

Figure 2

The research technical route.

Figure 2

The research technical route.

Close modal

Free ammonia and collected from the top of RD were neutralized and then separated by side-stream distillation (SSD) and in the mixed salt could be puri­fied. Due to the high energy-efficiency and easy operation, SSD is preferred instead of conventional distillation (Tututi-Avila et al. 2017). The Aspen Plus system was used to simulate the SSD process, with Radfrac as the rectification module. Box-Behnken design (BBD), based on response surface methodology (RSM), was applied for experimental design of the SSD process to investigate the effects of different operating conditions on methanol removal and ammonium sulfate () purification.

The residue collected from the bottom of the RD unit contains a large amount of organic nitrogen components such as pymetrozine, 3-aminopyridine, and 4-Acetylamino-6-methyl-3-oxo-2,3,4,5-tetrahydro-1,2,4-triazine, and so on. The purification of organic nitrogen components has been researched by our research team in another article (Wang et al. 2020), by complex extraction process with p204 (di (2-ethylhexyl) phosphate) as the complexing agent and benzene as the diluent agent.

A complete set of recovery and treatment technology for pymetrozine production wastewater is proposed in this paper. The main purpose is to assess pretreatment process of pymetrozine production wastewater by reactive distillation and side-stream distillation methods using Aspen Plus as the simulation tool and RSM as the optimization tool. The chemical materials with high purity and high added value can be separated effectively, including ammonia, , nitrogenous organic compounds and inorganic salts. Moreover, the treated wastewater, with low chemical oxygen demand (COD) content and greatly reduced toxicity, can be directly discharged into a biochemical treatment plant, and then reach the ‘national sewage comprehensive emission standard’ (GB8978-1996). Therefore, it can obtain high economic benefits and good wastewater treatment effect.

Materials

All the reagents were analytical grade and used directly without any further purification. Sodium hydroxide () and sulfuric acid () were purchased from Sinopharm Chemical Reagent Co.Ltd (China) with purity of 99.5%. Pymetrozine production wastewater was collected from a waste liquid discharge port of a pesticide-producing enterprise located in Yancheng City, Jiangsu Province. Main physical properties of pymetrozine wastewater used in this research are shown in Table 1.

Table 1

Physical properties of pymetrozine wastewater and neutralized distillate

MaterialpHComposition of (wt.%)Composition of (wt.%)Composition of (wt.%)Composition of salinity (wt.%)Composition of pesticide intermediates (wt.%)
Raw material 5.73 9.43 77.69 10.23 2.65 
Neutralized distillate 7.82 14.90 78.96 6.14 
MaterialpHComposition of (wt.%)Composition of (wt.%)Composition of (wt.%)Composition of salinity (wt.%)Composition of pesticide intermediates (wt.%)
Raw material 5.73 9.43 77.69 10.23 2.65 
Neutralized distillate 7.82 14.90 78.96 6.14 

Methods

Experiment procedure of ammonia conversion treatment

Experimental procedure was designed as shown in Figure 3. 300 mL of pymetrozine production wastewater and 100 mL of solution were poured into a three-necked flask through a constant pressure dropping funnel. The condensate water, magnetic stirrer, pH meter and heating jacket were kept working continuously. The distillate vapor was condensed through the spherical condensing tube, and collected in the oil-water separator. The condensed fluid was taken out for analysis. pH variation was detected every 10 minutes by pH meter with a certain accu­racy (0.01). When no more was detected in the liquid from the oil-water separator, it was considered that had been steamed out completely, and the reaction was terminated. The remaining liquid in the three-bottle flask was weighed, and a small amount of liquid was taken for elemental analysis. was selected to neutralize the distillate vapor of ammonia () into in surge flask.

Figure 3

Reactive distillation apparatus. 1. heating jacket 2. three-necked flask 3. dropping funnel 4. oil-water separator. 5. spherical condenser tube 6. surge flask 7. pH meter 8. magnetic stirrers.

Figure 3

Reactive distillation apparatus. 1. heating jacket 2. three-necked flask 3. dropping funnel 4. oil-water separator. 5. spherical condenser tube 6. surge flask 7. pH meter 8. magnetic stirrers.

Close modal

Elemental analysis method

Vario EL elemental analyzer (Germany Elementar Company) was used for elemental analysis, with JY/T017-1996 elemental analyzer general rules as the analysis detection principle. The quality content of nitrogen element () was measured as 2.49%. The mass fractions of () and the theoretical amount of () required for neutralization reaction were calculated by the following equations, on the premise of as the only nitrogen taking part in the reaction:
formula
(1)
formula
(2)
where , , , and 100 are respectively, the quality content of nitrogen element, the relative molecular mass of , the relative atomic mass of nitrogen element, the relative molecular mass of and 100 g of waste liquid. Thus, the theoretical addition amount of was calculated as 7.12 g per l00 g of waste liquid. Since almost all organic pesticides contain nitrogen element, determined by this method was only a reference.

Thermal conductivity detection

SP-6800 Gas chromatography (Luann Chemical Instrument Factory) was used for thermal conductivity detection, with a chromatographic column of 4*100 mm stainless steel tube, white GDX-102 supports, stationary liquid isophthalic acid, thermal conductivity cell detector, nitrogen carrier gas, and a column pressure of 0.1 MPa. The operating conditions were 120 °C in the gasification chamber, 80 °C in the column, 130 °C in the detector, and the sample injection quantity was 2 μL. When the temperature of the vaporizer was 120 °C, only and were detected and the composition of was 10.82 wt.%. It was preliminarily determined that the boiling points of pesticide intermediates and salty substances contained in the wastewater were basically higher than 120 °C, and the removing process of free ammonia and by reaction distillation had no effect on high boiling point material. The composition of in pymetrozine wastewater was 9.43 wt. % by comprehensive calculation.

The product collected from the top of RD was neutralized and detected by gas thermal conductance. The mass fractions of and were 15.87% and 84.13% respectively. The quality percentage of was calculated to be 6.14% by evaporation crystallization technology. In conclusion, the integrated composition of the neutralized distillate was calculated as follows: the mass percentages of , and were 14.90%, 6.14% and 78.96% respectively. The main physical properties of neutralized distillate are shown in Table 1.

SSD simulation procedure by Aspen Plus

The product collected from the top of RD was separated by SSD for recycling a high concentration of methanol on the tower top and from the tower bottom, aiming at methanol content on the tower top () of more than 99% and methanol content of side withdrawal () of less than 0.1%. Single factor experiments were designed based on Aspen Plus to investigate the effect of operating conditions on SSD, which provided the data foundation for the next BBD experimental design and RSM.

As the separation system was a polar and electrolytic system, the ELECNRTL property method and Radfrac rectification module were selected for SSD simulation by Aspen Plus. The influence of operating conditions was investigated, and the proper range of the main factors were preliminarily determined, such as number of stages (N), reflux ratio (R), feeding stage (), side line outlet stage (), distillate rate () and sideline discharge rate (), as shown in Table 2.

Table 2

Parameters of experiment simulation

ParameterValueParameterValueParameterValue
F/(Kg/h) 100 /°C 20.0 Property method ELECNRTL 
 0.11 ∼ 0.18 /°C 65.0 15 ∼ 28 
 0.30 ∼ 0.75 /°C 102.1  3 ∼ 15 
1 ∼ 7 /°C 99.9  13 ∼ 24 
Feedstock -- 
ParameterValueParameterValueParameterValue
F/(Kg/h) 100 /°C 20.0 Property method ELECNRTL 
 0.11 ∼ 0.18 /°C 65.0 15 ∼ 28 
 0.30 ∼ 0.75 /°C 102.1  3 ∼ 15 
1 ∼ 7 /°C 99.9  13 ∼ 24 
Feedstock -- 

Statistical analysis by BBD-RSM

The evaluation of the experimental design was performed using Design-Experts® version 10 software. RSM is an experimental and analytical software, with a whole set of mathemati­cal and statistical models to optimize operating conditions for a multivariable system (Wei et al. 2020). BBD, one of the main types of RSM, is a less time-consuming approach that allows evaluation of all possible parameter in­teractions (Karthikeyan et al. 2010), and uses fewer design points, with fewer experiments to run, compared with central composite designs (CCD), another main types of RSM. It represents advantages of time and materials consumed and economics of process develop­ment (Cordova-Villegas et al. 2019). Four independent variables were selected; that is, N, , and R, with as the response (dependent variable), and interactions between variables and the response were obtained. In order to optimize BBD, a three-level-four-factor design was applied, and 29 runs were generated. The actual and coded levels of the variables in the design matrix are calculated in Table 3. The relationship between coded and actual values is described by the following equation:
formula
(3)
where x is the coded value, xi is the actual value, x0 is the actual value at the center point, and Δx is the step change value of the variables (Bezerra et al. 2008).
Table 3

Factors and levels of BBD for the SSD

VariableFactorCoded levels
− 101
X1 23 25 27 
 X2 10 12 14 
D (%) X3 0.14 0.15 0.16 
X4 5.5 
VariableFactorCoded levels
− 101
X1 23 25 27 
 X2 10 12 14 
D (%) X3 0.14 0.15 0.16 
X4 5.5 

Effect of NaOH addition on nitrogen removal rate

To determine the addition for the RD, some trials were done for the ratio of the actual value to theoretical value of in the range of 60%–120%. As shown in Figure 4, at the early stage of addition, the nitrogen content in the wastewater does not change significantly. When addition of reaches a certain value, the removal rate of nitrogen increases obviously. The possible reason is that ammonia is particularly soluble in water and inhibits the removal of ammonia when is not reacted completely. With the addition of , the increased hydroxide ions inhibit the dissolution of ammonia in water, which is conducive to the removal of ammonia nitrogen. According to Figure 4 and Table 4, the optimal addition of was 7.12 g/100 g, the same as the theoretical value.

Table 4

Influence of addition amount of sodium hydroxide

No.Theoretical value of /gRatio of actual value to theoretical value of Content of nitrogen after reaction/%Quality of remaining liquid after reaction/gRemoval rate of nitrogen/%
– – 2.49 300.00 – 
21.36 0.60 2.07 317.60 11.99 
21.36 0.70 1.97 315.80 16.72 
21.36 0.80 1.90 312.50 20.52 
21.36 0.90 1.82 313.00 23.74 
21.36 1.00 0.986 339.40 55.20 
21.36 1.10 0.993 319.20 57.57 
21.36 1.20 0.991 323.00 57.15 
No.Theoretical value of /gRatio of actual value to theoretical value of Content of nitrogen after reaction/%Quality of remaining liquid after reaction/gRemoval rate of nitrogen/%
– – 2.49 300.00 – 
21.36 0.60 2.07 317.60 11.99 
21.36 0.70 1.97 315.80 16.72 
21.36 0.80 1.90 312.50 20.52 
21.36 0.90 1.82 313.00 23.74 
21.36 1.00 0.986 339.40 55.20 
21.36 1.10 0.993 319.20 57.57 
21.36 1.20 0.991 323.00 57.15 
Figure 4

Influence of addition amount of sodium hydroxide.

Figure 4

Influence of addition amount of sodium hydroxide.

Close modal

Effect of D1 on SSD

The increase of D1 is conducive to the collection of light components on the top of the tower. So it promotes the improvement of yield of methanol on the tower top () and the reduction of . However, when exceeds a certain range, it will accelerate the increase of tower temperature and disqualification of the top product. Therefore, it is necessary to investigate the effect of on SSD. was set as 0.11, 0.12, 0.13, 0.14, 0.15, 0.16, 0.17 and 0.18 respectively, when other conditions were N of 25, of 12th, of 24th, feeding quantity (F) of l00 kg/h, of 0.7 and R of 6.

As shown in Figure 5(a), at the beginning, increases gradually while has no evident changes with the increase of , and decreases gradually. When increases to 0.15, and reach the maximum, and is at a low level. Therefore, was chosen to be 0.15.

Figure 5

Effect of (a) distillate rate (b) reflux ratio (c) side line discharge rate (d) number of stages (e) feed stage (f) side line outlet stage on side stream distillation.

Figure 5

Effect of (a) distillate rate (b) reflux ratio (c) side line discharge rate (d) number of stages (e) feed stage (f) side line outlet stage on side stream distillation.

Close modal

Effect of R on SSD

The product purity on the top of the tower can improve with the increase of R. But too large R causes too much material circulation, and the tower balance will be broken. R was set as 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 5.5, 6, 6.5 and 7 respectively, with N of 25, of 12th, of 24th, of 0.15 and of 0.45.

As shown in Figure 5(b), with the increase of R, and increase, while decreases gradually. When R exceeds 6, is higher than 99% with lower than 0.1%, and the variation trends are not obvious. Therefore the appropriate R was selected as 6.

Effect of D2 on SSD

Too large can reduce the heat transfer efficiency, affect the concentration of , and cause crystallization at the tower bottom. Therefore, it is necessary to investigate the influence of .

As shown in Figure 5(c), with the increase of , all three parameters decrease slowly within the fluctuation range allowed by the design. is lower than 0.1%, and subsequent biochemical treatment can be performed directly. Meanwhile, should be lower than 0.75 to avoid precipitation of at the tower bottom. Moreover, in order to facilitate the subsequent recovery, the residue at the tower bottom should be a saturated solution of . Therefore, the selected was 0.7.

Effect of N on SSD

N was set as 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 and 28 respectively, when other conditions were fixed. As shown in Figure 5(d), with the increase of N, and increase gradually, while decreases. When N exceeds 25, the growth rate of and slow down gradually, while the decline rate of is gentle. Therefore, N of 25 was chosen in the later experiments.

Effect of NF on SSD

The distillation efficiency will be affected when significant difference exists between the input concentration and the material concentration in the tower at the feeding port. For this purpose, the influence was studied at of 3th, 4th, 5th, 6th, 7th, 8th, 9th, 10th, 11th, 12th, 13th, 14th and 15th, with other parameters fixed. The result is presented in Figure 5(e). With the increase of NF, the separation of light components becomes more adequate, and increase and decreases gradually. When reaches 6th, the curves flatten out. As NF reaches 12th, significant difference exists between the input concentration and the material concentration in the tower at the feeding position, which results in a decrease of distillation efficiency; and decrease gradually. Considering the separation requirements and operating cost, 12th plate was selected as the proper .

Effect of NS on SSD

As shown in Figure 5(f), with the increase in , and increase and decreases gradually. When reaches 24th, and maximize, and reaches the minimum. Based on the above analysis, 24th stage was chosen as the appropriate .

Box–Behnken design results and response surface analysis

Table 5 represents the range of factors along with the 29 tests designed. Accordingly, the test results were fitted to a quadratic polynomial model to correlate the measured response with the independent factors. The analysis of variance (ANOVA) for the BBD is shown in Table 6. As can be seen, an F-value of 44.74 and P-value less than 0.0001 indicate that the model is significant. The value of correlation coefficient (R2 values of 0.9781) is in reasonable agreement with the adjusted-R2 value (0.9563). The high value of R2 and non significance of lack of fit (P > 0.05) demonstrate that only 2.19% of the total variation can not be explained by the empirical model and the models have nice interpretation of the correlation between the responses and experiment conditions. Accordingly, the model can be dependably used.

Table 5

The BBD for three-level-four-factor RSM

RunCoded values
Actual values
Response
X1X2X3X4X1X2X3X4 
− 1 − 1 25 10 0.14 5.5 0.999948 
− 1 27 12 0.14 5.5 0.999996 
− 1 27 12 0.15 4.0 0.975289 
25 12 0.15 5.5 0.985383 
25 12 0.16 7.0 0.930684 
− 1 25 12 0.16 4.0 0.921344 
− 1 25 10 0.15 7.0 0.992736 
− 1 − 1 23 10 0.15 5.5 0.988001 
− 1 23 12 0.16 5.5 0.926662 
10 − 1 − 1 23 12 0.14 5.5 0.999993 
11 25 12 0.15 5.5 0.987065 
12 − 1 25 10 0.16 5.5 0.930431 
13 25 14 0.16 5.5 0.926662 
14 25 12 0.15 5.5 0.988374 
15 25 12 0.15 5.5 0.989992 
16 − 1 − 1 25 12 0.14 4.0 0.998922 
17 − 1 − 1 23 12 0.15 4.0 0.956037 
18 − 1 25 14 0.15 4.0 0.956049 
19 27 12 0.15 7.0 0.992749 
20 − 1 23 12 0.15 7.0 0.989679 
21 − 1 27 10 0.15 5.5 0.992065 
22 − 1 25 12 0.14 7.0 0.999998 
23 27 14 0.15 5.5 0.988099 
24 25 14 0.15 7.0 0.989687 
25 − 1 23 14 0.15 5.5 0.972879 
26 25 12 0.15 5.5 0.983059 
27 − 1 − 1 25 10 0.15 4.0 0.975108 
28 − 1 25 14 0.14 5.5 0.999999 
29 27 12 0.16 5.5 0.930436 
RunCoded values
Actual values
Response
X1X2X3X4X1X2X3X4 
− 1 − 1 25 10 0.14 5.5 0.999948 
− 1 27 12 0.14 5.5 0.999996 
− 1 27 12 0.15 4.0 0.975289 
25 12 0.15 5.5 0.985383 
25 12 0.16 7.0 0.930684 
− 1 25 12 0.16 4.0 0.921344 
− 1 25 10 0.15 7.0 0.992736 
− 1 − 1 23 10 0.15 5.5 0.988001 
− 1 23 12 0.16 5.5 0.926662 
10 − 1 − 1 23 12 0.14 5.5 0.999993 
11 25 12 0.15 5.5 0.987065 
12 − 1 25 10 0.16 5.5 0.930431 
13 25 14 0.16 5.5 0.926662 
14 25 12 0.15 5.5 0.988374 
15 25 12 0.15 5.5 0.989992 
16 − 1 − 1 25 12 0.14 4.0 0.998922 
17 − 1 − 1 23 12 0.15 4.0 0.956037 
18 − 1 25 14 0.15 4.0 0.956049 
19 27 12 0.15 7.0 0.992749 
20 − 1 23 12 0.15 7.0 0.989679 
21 − 1 27 10 0.15 5.5 0.992065 
22 − 1 25 12 0.14 7.0 0.999998 
23 27 14 0.15 5.5 0.988099 
24 25 14 0.15 7.0 0.989687 
25 − 1 23 14 0.15 5.5 0.972879 
26 25 12 0.15 5.5 0.983059 
27 − 1 − 1 25 10 0.15 4.0 0.975108 
28 − 1 25 14 0.14 5.5 0.999999 
29 27 12 0.16 5.5 0.930436 
Table 6

ANOVA results for the response surface quadratic model

SourceSSdfMSF-ValueP-Value
Model 0.020 14 1.420 × 10−3 44.74 <0.0001 Significant 
A-N 1.716 × 10−4 1.716 × 10−4 5.41 0.0356  
B-NF 1.681 × 10−4 1.681 × 10−4 5.30 0.0372  
C-D 0.016 0.016 491.58 <0.0001  
D-R 1.060 × 10−3 1.060 × 10−3 33.41 <0.0001  
AB 3.112 × 10−5 3.112 × 10−5 0.98 0.3388  
AC 3.555 × 10−6 3.555 × 10−6 0.11 0.7428  
AD 6.547 × 10−5 6.547 × 10−5 2.06 0.1728  
BC 3.649 × 10−6 3.649 × 10−6 0.12 0.7395  
BD 6.408 × 10−5 6.408 × 10−5 2.02 0.1772  
CD 1.707 × 10−5 1.707 × 10−5 0.54 0.4754  
A2 1.721 × 10−5 1.721 × 10−5 0.54 0.4736  
B2 1.780 × 10−5 1.780 × 10−5 0.56 0.4663  
C2 2.589 × 10−3 2.589 × 10−3 81.61 <0.0001  
D2 2.204 × 10−4 2.204 × 10−4 6.95 0.0196  
Residual 4.442 × 10−4 14 3.173 × 10−5    
Lack of Fit 4.155 × 10−4 10 4.155 × 10−5 5.78 0.0527 Not significant 
Pure Error 2.874 × 10−5 7.184 × 10−6    
R2 0.9781      
Radj2 0.9563      
SourceSSdfMSF-ValueP-Value
Model 0.020 14 1.420 × 10−3 44.74 <0.0001 Significant 
A-N 1.716 × 10−4 1.716 × 10−4 5.41 0.0356  
B-NF 1.681 × 10−4 1.681 × 10−4 5.30 0.0372  
C-D 0.016 0.016 491.58 <0.0001  
D-R 1.060 × 10−3 1.060 × 10−3 33.41 <0.0001  
AB 3.112 × 10−5 3.112 × 10−5 0.98 0.3388  
AC 3.555 × 10−6 3.555 × 10−6 0.11 0.7428  
AD 6.547 × 10−5 6.547 × 10−5 2.06 0.1728  
BC 3.649 × 10−6 3.649 × 10−6 0.12 0.7395  
BD 6.408 × 10−5 6.408 × 10−5 2.02 0.1772  
CD 1.707 × 10−5 1.707 × 10−5 0.54 0.4754  
A2 1.721 × 10−5 1.721 × 10−5 0.54 0.4736  
B2 1.780 × 10−5 1.780 × 10−5 0.56 0.4663  
C2 2.589 × 10−3 2.589 × 10−3 81.61 <0.0001  
D2 2.204 × 10−4 2.204 × 10−4 6.95 0.0196  
Residual 4.442 × 10−4 14 3.173 × 10−5    
Lack of Fit 4.155 × 10−4 10 4.155 × 10−5 5.78 0.0527 Not significant 
Pure Error 2.874 × 10−5 7.184 × 10−6    
R2 0.9781      
Radj2 0.9563      

Note: df, degree of freedom; MS, mean square; SS, sum of squares.

*P-value <0.05 indicates statistical significance.

The quadratic polynomial regression model was used to express the effects of independent variables on responses according to the following equation:
formula
(4)
where y is the response, b0 is the model constant, bi, bii and bij are the coefficients, xi and xj are independent variables, n is the amount of variables, and ε is the error. The final regression model in terms of coded factors for is expressed by the following equation:
formula
(5)

By default, the high levels of the factors are coded as positive sign and the low levels of the factors are coded as negative sign.

The response surfaces can be visualized as contours and/or 3D plots to constitute the variations in the response with respect to two variables, by keeping the other variable fixed (Ai et al. 2015). Figure 6 shows both contours and 3D surface response plots for as a function of (a) N vs D, (b) N vs R, (c) vs D, and (d) D vs R, respectively. The central coordinate points among the utmost contour levels in each of the figures indicate the optimum value of corresponding parameters. Figure 6(a) and 6(c) indicate that increases gradually with the decrease of D1 from 0.16 to 0.14, regardless of N and at low or high level. The optimum region can be found at in the range of 0.145–0.155, N of 24–26, and of 11th–13th, indicating that the effect of D1 on is highly significant. Meanwhile, Figure 6(b) indicates that the maximum of is obtained at N of 25 and R of 5.5 with R having more effect than N. Figure 6(d) shows the effect of D1 and R on . It is known that decreases steeply with decrease of R and increase of D1. As the figure reveals, both D1 and R play major role in SSD.

Figure 6

Contours and 3D surface plots of the content of methanol for (a) number of stages vs. distillate rate, (b) number of stages vs. reflux ratio, (c) feeding stage vs. distillate rate, and (d) distillate rate vs. reflux ratio.

Figure 6

Contours and 3D surface plots of the content of methanol for (a) number of stages vs. distillate rate, (b) number of stages vs. reflux ratio, (c) feeding stage vs. distillate rate, and (d) distillate rate vs. reflux ratio.

Close modal

To further validate optimal values, the first partial derivative of regression equation was taken and made zero. It was concluded that the optimal values of the variables determined by RSM were N of 26, D of 0.15, R of 6 and of 13th. In these conditions, the predicted value of was 99.5%. To check the validity of RSM, a verifying SSD experiment was conducted under the optimal condition and was 99.28%. The experimental value was close to the theoretical predicted value, indicating the practicability of optimizing the operating conditions of SSD by RSM method. Optimal flow sheet for SSD with equipment data, steam data, reflux ratio, heat duties and operating pressures (P) is shown in Figure 7.

Figure 7

Optimal flow sheet for SSD.

Figure 7

Optimal flow sheet for SSD.

Close modal

There has been very little research on pymetrozine production wastewater treatment. In terms of this issue, a pretreatment solution was provided to effectively remove nitrogen and purify methanol. The follow-up work is organic nitrogen compounds removal by complex extraction and advanced treatment of the extracted aqueous phase in order to reduce harmful substances and meet the direct emission standards. This method can also be used for pretreatment of other pesticide production wastewater and simplify the subsequent treatment process.

In the present work, ammonia conversion treatment and SSD process were conducted to pretreat the pymetrozine production wastewater. Aspen Plus software was used to simulate the single factor experiment of SSD. BBD based on RSM was employed to evaluate the optimal operational conditions, and and R were chosen as the most significant factor. A verifying experiment was conducted to check the validity of RSM. Optimum separation result was obtained as of 99.28% with of 99.95%, less than 0.1%, and content of 21.58% at the tower bottom, at N of 26, of 0.15, R of 6 and of 13th.

This work was supported by Extraction Engineering Technological Research Center of Jiangsu province and Key Laboratory of Integrated Regulation and Resource Development on Shallow Lake of Ministry of Education.

This work was sponsored by National Natural Science Foundation of China (Grant No. 51979077), the Fundamental Research Funds for the Central Universities (Grant No. 2019B42414) and Jiangsu Provincial Science and Technology Program Project (Grant No. SBA2018030430 and BE2019121).

All relevant data are included in the paper or its Supplementary Information.

Bezerra
M. A.
Santelli
R. E.
Oliveira
E. P.
Villar
L. S.
Escaleira
L. A.
2008
Response surface methodology (RSM) as a tool for optimization in analytical chemistry
.
Talanta
76
(
5
),
965
977
.
Carra
I.
Sánchez Pérez
J. A.
Malato
S.
Autin
O.
Jefferson
B.
Jarvis
P.
2016
Performance of different advanced oxidation processes for tertiary wastewater treatment to remove the pesticide acetamiprid
.
Journal of Chemical Technology & Biotechnology
91
(
1
),
72
81
.
Cordova-Villegas
L. G.
Cordova-Villegas
A. Y.
Taylor
K. E.
Biswas
N.
2019
Response surface methodology for optimization of enzyme-catalyzed azo dye decolorization
.
Journal of Environmental Engineering
145
(
5
),
04019013
.
De Gaetano
Y.
Hubert
J.
Mohamadou
A.
Boudesocque
S.
Plantier-Royon
R.
Renault
J.-H.
Dupont
L.
2016
Removal of pesticides from wastewater by ion pair centrifugal partition extraction using betaine-derived ionic liquids as extractants
.
Chemical Engineering Journal
285
,
596
604
.
Fdez-Sanromán
A.
Acevedo-García
V.
Pazos
M.
Sanromán
M. Á.
Rosales
E.
2020
Iron-doped cathodes for electro-Fenton implementation: application for pymetrozine degradation
.
Electrochimica Acta
338
,
135768
.
Gor
N. K.
Mali
N. A.
Joshi
S. S.
2020
Intensified reactive distillation configurations for production of dimethyl ether
.
Chemical Engineering and Processing – Process Intensification
149
,
107824
.
Jang
J.
Rahman
M. M.
Ko
A. Y.
Abd El-Aty
A. M.
Park
J. H.
Cho
S. K.
Shim
J. H.
2014
A matrix sensitive gas chromatography method for the analysis of pymetrozine in red pepper: application to dissipation pattern and PHRL
.
Food Chemistry
146
,
448
454
.
Karthikeyan
K.
Nanthakumar
K.
Shanthi
K.
Lakshmanaperumalsamy
P.
2010
Response surface methodology for optimization of culture conditions for dye decolorization by a fungus, Aspergillus niger HM11 isolated from dye affected soil
.
Iranian Journal of Microbiology
2
(
4
),
213
.
Sutton
R.
Xie
Y.
Moran
K. D.
Teerlink
J.
2019
Occurrence and sources of pesticides to urban wastewater and the environment
. In:
Pesticides in Surface Water: Monitoring, Modeling, Risk Assessment, and Management
.
ACS Publications
, pp.
63
88
.
Tony
M. A.
Mansour
S. A.
2019
Microwave-assisted catalytic oxidation of methomyl pesticide by Cu/Cu2O/CuO hybrid nanoparticles as a Fenton-like source
.
International Journal of Environmental Science and Technology
17
(
1
),
161
174
.
Tututi-Avila
S.
Medina-Herrera
N.
Hahn
J.
Jiménez-Gutiérrez
A.
2017
Design of an energy-efficient side-stream extractive distillation system
.
Computers & Chemical Engineering
102
,
17
25
.
Wang
B.
Ke
S.
Kishore
B.
Xu
X.
Zou
Z.
Li
Z.
2012
A facile synthesis of pyrimidone derivatives and single-crystal characterization of pymetrozine
.
Synthetic Communications
42
(
16
),
2327
2336
.
Zhou
Q.
Du
F.
Shi
Y.
Liu
W.
Liu
D.
Chen
G.
2018
An efficient protocol for the production of pymetrozine via a new synthetic strategy
.
Journal of Chemical Research
42
(
8
),
434
438
.