Design of multilayer microfluidic paper chip and its visual detection of pesticide residues

With the increase of China's grain production, the use of pesticides is gradually increasing. Traditional pesticide detection takes a long time and requires expensive experimental instruments, which is not conducive to the rapid and accurate detection of pesticide residues in the field. To solve this problem, this paper proposes a visual detection method of pesticide residues based on multi-layer microfluidic paper chips. The internal channel structure of paper chip is designed from the perspective of efficient mixing. Through the simulation of the mixed effect of three kinds of staggered channel structures, which are arc type, triangle type, and ladder type, the "ladder-type h-0.3, s-2.6" is selected as the best-staggered structure, and the mixing strength is 0.91534. The best simulation structure was tested by a colored reagent, and the image processing of 15 test results was carried out with MATLAB. The average mixing strength was 0.84, and the and the standard deviation was 0.022. The visual detection experiment of acetamiprid and profenofos in cabbage samples was carried out by using the device，The detection range of acetamiprid was 4~72 μg/kg, and the detection range of profenofos was 3~54 μg/kg . The recovery of acetamiprid was 75%~85%, and the recovery of profenofos was 80%~90%. The detection range and recovery rate indicate that the device has high repeatability and accuracy in the actual sample detection


Introduction
The traditional methods of pesticide residues detection include chromatography1 [1] and spectroscopy [2]. The traditional methods have high accuracy , but the detection process is time-consuming, requiring hundreds of thousands of instruments or even millions of professional personnel .The traditional detection methods can not detect pesticide residues in real time quickly and easily. However, the widely used rapid detection card can detect pesticides in real time, but the detection limit is 0.3~3.5mg/kg. The detection limit of pesticide residues in fruits and vegetables was higher than μg/kg ~mg/kg, so the detection quantity of the quick detection card is higher than that of the human body. The cost of paper-based microfluidic chip is lower than that of glass, PDMS, quartz and other materials because it is based on paper. Moreover, without external driving, the sample, reagent and waste liquid in the fluidic chip are micro, with high integration, low consumption and easy to carry. The detection of pesticide based on paper microfluidic chip is greatly convenient for the detection personnel and is conducive to the real-time detection on site.
CLémence et al. [5] made a two-dimensional paper chip by means of wax spray printing, which was used to detect organophosphate in water. Martinez [3] and others for the first time demonstrated the stacking design of 3D μpads. The paper chips were stacked together by double-sided adhesive tape, and the hydrophobic channel was made by photolithography. Holes were drilled on the double-sided adhesive to ensure the connection between the upper and lower layers, so as to obtain the ability to move fluid in a three-dimensional manner. Liu et al. [4] for the first time prepared μpads by three-dimensional origami method, and verified its applicability by color detection of glucose and protein. However, the channels in these three kinds of paper chips are all simple straight channels, and there is no channel shape designed to make mixing more efficient. In order to make different reactants fully mix and react, this paper designs different channel structures based on three-dimensional paper chip, and studies the mixing effect of different structures.
In this paper, we use the intrinsic fluorescence effect between ratio fluorescent quantum dots and gold nanoparticles. At the same time, coupled with the appropriate ligand technology, the pesticide residue detection is transformed into the detection of color change visible to the naked eye. A three-dimensional multi-layer array paper chip was designed, and the structure of the mixing zone was optimized by simulation and experimental research, so as to improve the mixing efficiency and the visual detection accuracy of pesticides.
1 Design of multilayer microfluidic paper chip

Overall design of multilayer paper chip
In order to meet the requirements of rapid and convenient detection of a variety of pesticides, a foldable multilayer paper-based microfluidic chip was designed, and the three-dimensional paper chip was made by wax spray printing method [7]. The required pattern was printed on the surface of filter paper by wax spray printer, and then heated to melt the wax. The melted wax penetrates down into the paper to form hydrophobic channels.   Fig. 1 (a).The three-dimensional paper chip is composed of five parts. The first layer is the first injection layer. The pesticide flows directly to the second injection layer through six single circular paper channels. The remaining two are the injection areas of gold nanoparticles and sodium chloride solution. The next area is the mixing area of the two reagents and there are six layers in the mixing zone. Hydrophobic regions of various shapes are printed on the circular channels of these six layers to form a structural channel to promote the mixing of gold nanoparticles (AuNPs) and sodium chloride (NaCl). After mixing, NaCl makes AuNPs completely agglomerate, and the mixture of the two liquids also enters the second injection area to mix with the pesticide.The eighth layer is the second injection area, in which there is a control channel for sodium chloride and gold nano solution to enter alone. The channel is a single circular channel, which is used to make the mixture flow to the control area of the final detection layer and react with the modified substrate to form a detection control. The fluid transportation in paper is a passive process, which is realized by the capillary action of paper. Through folding, 3D paper chip enables multiple flow layers to penetrate liquid layer by layer. In addition to the flow in each layer of paper, it can also flow vertically between two layers. This method enables the fluid to be transported to a large number of reaction sites in a compact area.

Structure design of mixed channel
The chip channel shown in Fig.1 (b) has two mixing regions. Both samples are mixed through six layers of mixing layer. When the two sample solutions are mixed in the mixing layer, there is no third solution involved in mixing, so the two mixing regions choose the same mixing structure. When the structure of the mixing zone is optimized and simulated by COMSOL software, a total of 8 layers of paper channel from sample injection to mixing end is selected for simulation.
The two basic mechanisms of fluid mixing are diffusion and convection. In order to make two or more fluids mix well without any external driving equipment, it is necessary to design the internal channel shape reasonably.
Different channel shapes make the transported fluid stretch, compress, separate and converge to achieve the Journal of Advances in Agriculture Vol 11 (2020) ISSN: 2349-0837 https://rajpub.com/index.php/jaa 173 mixing effect. When the structure of the mixing zone is optimized by COMSOL software, a total of 8 paper channels from sample injection to mixing end are selected for simulation, and the middle six layers are mixed layer. Figure 2 shows the simulation diagram of two simulated liquids mixing on the 8-layer paper chip.

Fig 2. Schematic diagram of single sample detection of two pesticides on the paper channel
According to the mixing principle of fluid tension and compression, three kinds of three-dimensional inner rib structures are designed in this paper: circular arc (Fig.3), triangular (Fig.4) and ladder (Fig.5). Fig.6 shows the cross section of circular arc, triangular and ladder channels respectively.
As shown in Table 1, determine the side length "s" of the ladder type as 1.4mm, change the factors H1, H2 and H3, and select four parameter values of each factor for simulation to determine the relationship between parameters and mixing performance, select H1, H2, H3 that can make the best mixing effect, and then select four s parameters with selected H3 to further determine the parameter size of ladder structure. The vertical channel radius R is 1.5mm and the height h is 1.44mm (there are eight layers of paper in the mixing zone, and the thickness of each layer is about 0.18mm). (1)

aveop() is the average operator in COMSOL; abs() is the absolute value;
ci is the concentration at a point on the cross section (mol/L); is the average concentration of the cross section (mol/L).
Ie is a dimensionless value in the range of 0~1. When it is 1, it means that the mixture is sufficient, and when it is 0, it means that there is no mixing. Fig. 7, Fig. 8 and Fig. 9 show the comparison diagram of mixed strength Ie at the exit of each floor of arc type, triangle type and ladder type structure (s=1.4mm) with parameters h of 0 mm, 0.3mm, 0.6mm and 0.9mm. In addition, the blank structure without any channel shape was set as the control group. The results showed that even though the structures were different, the mixing effect was better with the decrease of h. However, when h was reduced to 0 mm, the mixing effect became worse when h was 0.3 mm. Therefore, when h=0.3mm, the mixing effect of the last layer of the three structures was the best, ie was 0.86679 (H1=0.3mm), 0.85846 (h2 = 0.3mm) and 0.86679 (H1 = 0.3mm), respectively 88374 (H3 is 0.3mm), so the ladder with H3 of 0.3mm is selected to continue the optimization simulation. The results show that the higher the mixing strength is, the greater the mixing strength is. When s is 2.6 mm, the maximum mixing strength Ie is 0.91534. Therefore, the ladder structure with H3 of 0.3 mm and s of 2.6 mm is selected as the optimal three-dimensional inner rib structure to design the flow channel, and this structure is Compared with the above results, the analysis is as follows:

Simulation results of three hybrid structures
In our selected values, when h is 0.3mm, the simulation effect is the best. When h is 0.9mm, the interlaced structure is obvious, which will cause great interference to the velocity of the fluid. However, if the interference exceeds a certain limit, it will hinder the diffusion movement of the fluid, and make the fluid molecules unable to exchange and diffuse better, resulting in poor mixing effect. On the other hand, from the perspective of different shapes, the ladder shape is the best, followed by arc type and triangle type. It is speculated that the shape of ladder shape makes the fluid more intense folding and expansion on the basis of ensuring diffusion, so as to achieve better mixing effect. When s is 2.6 mm in the ladder structure, the flow velocity and direction of the fluid on both sides of the channel are well changed, resulting in a greater degree of tensile deformation, so that the mixing effect is better and the mixing strength is greater.

Multi layer paper chip mixing experiment
Two solutions of methyl orange and methyl blue were selected as the mixed solution and dropped into the 3D paper chip as shown in Fig.11a. At the same time, the control experiment of blank structure (Fig. 12b) was set.
As shown in FIG. 11, FIG. 11C shows the flow mixing of each layer of "ladder h-0.3, s-2.6", and Fig. 11d shows the mixing of each layer in the blank structure control group. The color uniformity of Fig. 11C is better than that of Fig. 11d. The mixing effect of multi-layer paper-based microfluidic chips was evaluated by image processing. Define the value of mixing strength ie as (2) Where N is the number of pixels in the measurement area; c i is the gray value of each pixel; is the average gray value Ie is a dimensionless value in the range of 0~1. When it is 1, it means that the mixture is sufficient. When it is 0, it means that there is no mixing. Matlab is used to extract the gray value of each pixel in the yellow area in Fig.12a, and the average value is calculated. Finally, the gray value ci of the pixel points in the mixed region in

Pesticide detection experiment of array 3D paper chip
The principle of pesticide detection: the green part of the ratio fluorescent quantum dots and gold nanoparticles produce fluorescence internal efficiency, and the green fluorescence is quenched; when AuNPs are in sodium chloride solution, the aggregation between AuNPs occurs, which weakens the fluorescence internal filtering effect between the ratio fluorescence probe and AuNPs, and the fluorescence color of green quantum dots in the ratio fluorescence probe is restored.
When the aptamer of acetamiprid is added, it will be adsorbed on the surface of AuNPs due to the coordination between aptamer and AuNPs, thus enhancing the salt out resistance of AuNPs and dispersing the agglomerated AuNPs again. As a result, the fluorescence internal filtering effect between the ratio fluorescent probe and AuNPs should be enhanced, and the fluorescence intensity of the green quantum dots of the ratio fluorescent probe will decrease. When acetamiprid is added, the aptamer fell off the AuNPs due to the specific binding between Journal of Advances in Agriculture Vol 11 (2020) ISSN: 2349-0837 https://rajpub.com/index.php/jaa 180 the aptamer and acetamiprid, and AuNPs were induced to agglomerate again by sodium chloride, which led to the decrease of fluorescence internal filtering effect and the green quantum dots fluorescence color of ratio fluorescent probe was restored again. Therefore, the concentration of acetamiprid determines the intensity of fluorescence internal filtering effect, and the fluorescence intensity of ratio fluorescence quantum dots is related to the concentration of acetamiprid, thus realizing the visual detection of pesticide residues.
With the increase of acetamiprid concentration, the fluorescence color of the detection area changes from pink to orange to yellow to yellow to green. This series of visible color changes show that by taking photos and then extracting RGB with MATLAB, the relationship between b value and acetamiprid pesticide concentration is established (FIG. 15). The linear equation is B1=0.69c1+115, B1 is the corresponding b value of RGB in the photo, c1 is acetamiprid. The detection range was 4~72μg/kg, R2 is 0.9226.  Fig. 16 shows the fluorescence color change photos of the ratio fluorescence sensing system of profenofos with different concentrations (0 μg/kg, 3μg/kg, 6μg/kg, 12μg/kg, 20μg/kg, 28μg/kg, 36μg/kg, 46μg/kg, 54μg/kg) under UV lamp. Figure 17 shows the relationship between b value and pesticide concentration, the detection range is 3 ~54 μg/kg, the linear equation is B2=c2+117, R2 is 0.9319, B2 is RGB value, c2 is profenofos concentration. The detection range of acetamiprid was 4~72 μg/kg and that of profenofos was 3~54μg/kg. Both of them were lower than the national standard of 0.01~0.1mg/kg of profenofos and Acetamiprid in vegetables. In order to verify the reliability of the device for pesticide detection, cabbage samples injected with acetamiprid and profenofos were detected by the device. As shown in Table 2 and table 3, the recoveries of acetamiprid and profenofos were 75%~85% and 80%~90%, respectively, indicating that the device had high repeatability and accuracy in actual sample detection.