Microfluidics is a powerful tool for the development of diagnostic tests. However, expensive equipment and materials, as well as laborious fabrication and handling techniques, are often required. Here, we detail the fabrication protocol of an acrylic microfluidic device for magnetic micro- and nanoparticle-based immunoassays in a low-cost and simple-to-use setting.
Microfluidic systems have greatly improved immunoassay techniques. However, many microfabrication techniques require specialized, expensive, or complicated equipment, making fabrication costly and incompatible with mass production, which is one of the most important preconditions for point-of-care tests (POCT) to be adopted in low-resource settings. This work describes the fabrication process of an acrylic (polymethylmethacrylate, PMMA) device for nanoparticle-conjugated enzymatic immunoassay testing using the computer numerical control (CNC) micromilling technique. The functioning of the microfluidic device is shown by performing an immunoassay to detect a commercial antibody using lysozyme as a model antigen conjugated to 100 nm magnetic nanoparticles. This device integrates a physical staggered restriction of only 5 µm in height, used to capture magnetic microparticles that make up a magnetic trap by placing an external magnet. In this way, the magnetic force on the immunosupport of conjugated nanoparticles is enough to capture them and resist flow drag. This microfluidic device is particularly suitable for low-cost mass production without the loss of precision for immunoassay performance.
In recent years, microfluidics has played an important role in immunoassay techniques1. Miniaturization technology has many outstanding advantages compared to traditional immunoassays, such as reduced sample and reagent consumption, shorter incubation times, efficient solution exchange, and higher integration and automation2.
Furthermore, microfluidic systems in immunoassays, in association with magnetic nanoparticles as immunosupport, considerably reduce incubation times, achieving high detection sensitivity due to the increased surface-to-volume ratio3. Brownian movement of the particles improves the reaction kinetics during the formation of the antigen-antibody complex4,5. Moreover, the magnetic properties of nanoparticles provide the versatility to be integrated into different microfluidic device configurations, making them an ideal candidate for signaling and molecule capture in miniaturized on-chip biosensing systems5. However, magnetic forces are significantly weaker than drag forces at the nanometer scale due to the high surface-to-volume ratio6. Therefore, capturing nanoparticles for crucial immunoassay steps such as washing and detection can be challenging, and a conventional magnet is insufficient4.
An efficient way to manipulate the nanoparticles is the use of a microfluidic magnetic trap formed by iron microparticles, which are packed in a microfluidic structure3. Therefore, when an external magnet approaches, a complex interaction is created within the magnetized porous medium between the magnetic and flux forces. The magnetic force acting on the nanoparticles is strong enough to capture them and resist flow drag3,4,7. This approach requires microfabrication techniques that achieve resolutions in the order of a few micrometers to generate micrometric structures that retain the microparticles.
Current microfabrication techniques allow the high-resolution fabrication of structures from a few microns to hundreds of nanometers8. However, many of these techniques require specialized, expensive, or complicated equipment. One of the main difficulties is the requirement for a cleanroom for mold fabrication, which remains costly and time-consuming8,9. Recently, microfluidic engineers have overcome this drawback by developing a variety of alternative fabrication methods, with various advantages such as reduced costs, faster turnaround times, cheaper materials and tools, and increased functionality8. In this way, the development of new microfabrication techniques brought low-cost, non-cleanroom methods that achieve resolutions as low as 10 µm8. Patterning can be used directly on a substrate without generating an expensive molding pattern, thus avoiding a time-consuming process. Direct fabrication methods include CNC milling, laser ablation, and direct lithography8. All these methods are suitable for producing high-aspect-ratio channels in a wide range of materials, regardless of their hardness9, enabling new and advantageous geometries, physical behaviors, and qualities in microfluidic devices8.
CNC micromilling creates microscale structures using cutting tools that remove bulk material from a substrate and is an effective fabrication method for microfluidic devices10,11. The micromilling technique can be useful in microfluidic applications to create microchannels and features directly on the work surface, offering a key advantage: a workpiece can be fabricated in a short time (less than 30 min), significantly reducing the turnaround time from design to prototype12. In addition, the wide availability of cutting accessories of different materials, sizes, and shapes makes CNC milling machines a suitable tool that has allowed the fabrication of different features in many types of low-cost disposable materials13.
Among all the materials commonly used in micromilling, thermoplastics remain a leading choice due to their many favorable properties and compatibility with biological applications10,14. Thermoplastics are an attractive substrate for microfluidic systems due to their significant advantages for developing low-cost, disposable analytical systems9. In addition, these materials are highly amenable to high-volume manufacturing processes, making them suitable for commercialization and mass production. For these reasons, thermoplastics such as PMMA have been considered reliable and robust materials since the early years of microfluidics10. Different protocols have been described to fabricate closed channels in thermoplastics, such as solvent bonding15, heat bonding16, and ultraviolet (UV)/ozone surface treatment bonding17.
In many cases, the positioning resolution achieved with conventional micromilling machines is not sufficient for some microfluidic applications that require structures smaller than 10 µm. High-end micromilling has enough resolution. Unfortunately, due to high prices, its use is limited to a handful of users12. Previously, our research group reported the fabrication and manipulation of a low-cost tool that allows machining structures of less than 10 µm, overcoming the resolution of conventional milling machines12. The fixture is a platform manufactured by 3D printing with simple electronics, containing three piezoelectric actuators. The surface contains hinge-shaped joints that allow it to be lifted when the piezoelectric elements act simultaneously. Z-axis displacement can be controlled with a resolution of 500 nm and an accuracy of ±1.5 µm12.
This paper presents the steps of the manufacturing process of an acrylic device (PMMA) through a micromilling technique. The chip design consists of a main channel 200 µm wide and 200 µm high and a side channel with the same dimensions to purge the flow of the reagents. In the central region, the channel is interrupted by a physical restriction of only 5 µm in height, fabricated with the 3D-printed piezoelectric platform made by this group12, to capture magnetic microparticles that make up a magnetic trap for nanoparticles by placing an external magnet. We show the operation of the microfluidic device by performing an immunoassay to detect a commercial antibody using lysozyme as a model antigen conjugated to 100 nm magnetic nanoparticles. This device combines different features that make it unique4: the use of magnetic nanoparticles as immune support reduces the total test time from hours to minutes; using a fluorogenic enzyme for detection allows for limits of detection that are comparable to those of standard enzyme-linked immunosorbent assays (ELISAs); and the use of a thermoplastic as a fabrication material makes it compatible with mass production, which was not the case for previous microfluidic nanoparticles' magnetic traps3, and makes it an excellent candidate to develop POCT.
1. Micromilling
Figure 1: End mill bit placement. (A) The 200 µm and 800 µm end mill bits are placed and fixed through a screw to the steel support. (B) Each end mill bit is placed in the specific compartment of the micromilling machine for automatic selection. Please click here to view a larger version of this figure.
Figure 2: Piezoelectric platform. The platform is fabricated by 3D printing and consists of two hexagonal bases joined by hinges that allow a fine displacement in the z-axis controlled by three piezoelectric actuators. An acrylic adapter is also observed, on which the PMMA rectangle is attached, and which allows for setting of the alignment corner of the coordinates. Please click here to view a larger version of this figure.
Figure 3: Z-axis calibration. The steps of the z-axis calibration are detailed. (A) The z-sensor includes a cable that plugs into the micromilling machine. (B) The sensor is placed directly onto the surface to be machined. (C) The detection pin consists of a metal bar placed in a special compartment next to end mill bits. (D) When both accessories come into contact, the micromilling machine automatically calculates the origin coordinate on the z-axis. Please click here to view a larger version of this figure.
Figure 4: Rectified acrylic surface. (A) The 200 µm diameter end mill bit sweeps the entire surface of the acrylic rectangle, removing a layer approximately 30 µm high. (B) The image shows the different structures milled on the face of the previously rectified acrylic. Channels and holes for reagent inlet and outlet are observed. The 5 µm restriction cannot be seen with the naked eye. (C) Micromilled surface with alignment holes and adapter with alignment pillars at opposite corners. (D) The acrylic is aligned upside down on the adapter with pillars, into which the alignment holes fit. Scale bar = 500 µm. Please click here to view a larger version of this figure.
2. Channel sealing
Figure 5: Sealing process of the device. (A) Each of the acrylic sheets is placed in a resealable bag with distilled water and immersed in the ultrasonic bath. (B) The image on the left shows the channels just after fabrication, and the image on the right shows the same device after washing with IPA and the ultrasonic bath, which removes all impurities and acrylic residues from the microchannel. The edges of the restriction that interrupts the central channel of 200 µm are observed, which confirms the successful milling process. Scale bars = 500 µm. (C) Both acrylics are dried and adhered to the glass platform on the lid. (D) The base of the Petri dish is placed inside another dish of larger diameter. (E) When closing the Petri dish, the water seal prevents gaseous chloroform from escaping. (F) Description of the elements of the lever with a weight of 5 kg. (G) Image of the open lever, showing in red the area where the acrylic is placed. Please click here to view a larger version of this figure.
3. Device preparation
4. Microparticle trap formation
5. Immunoassay
6. Experimental mounting
Figure 6: Final device configuration. (A) Acrylic device with the hoses attached to the corresponding inputs and outputs. The scale shows the dimensions of the device in centimeters. (B) Protocol for the formation of the microparticle trap. Microparticles flow through the channel by gravity when the device is placed in a vertical position. Microparticles are concentrated at the 5 µm restriction. Excess microparticles are easily removed by rotating the chip through the side channel. The chip is kept vertical to preserve the trap before immunoassay. (C) Microfluidic device mounted on a glass slide containing the magnet, on the stage of the inverted fluorescence microscope. The dispensing needle through which the reagents are added is observed, as well as the outlet hoses that connect to a syringe pump. Please click here to view a larger version of this figure.
7. Immunodetection
It was possible to establish a highly reproducible fabrication protocol that improves the resolution of the conventional micromilling technique. Using this protocol, the fabrication of a channel as small as 5 µm in height that operates as a staggered restriction in a 200 µm high channel is achieved. The simple design of the staggered restriction captures iron microparticles of 7.5 µm diameter which, when compacted in the microchannel, allow the creation of a magnetic trap when an external magnet approaches the device. This device allows immunoassays to be performed using nanoparticles conjugated with the analyte of interest as immunological support. In this work, non-competitive indirect immunoassays with enzyme-labeled antibody-based detection were performed. The model antigen (Ag) was lysozyme protein conjugated to nanoparticles (NPs) of 100 nm diameter. Rabbit anti-lysozyme IgG was used as the primary antibody (AbI) and detection was performed by using rabbit horseradish peroxidase-conjugated secondary antibody (HRP-AbII).
Detection was performed by correlating the change in fluorescence signal intensity obtained after the interaction of the HRP-AbII with a fluorogenic substrate upon passage through the trap with entrapped nanoparticles. To perform the measurements, a region before and a region after the trap were determined. Figure 7A–D shows the increase in fluorescence intensity for different concentrations of anti-lysozyme AbI: 0 ng/mL, 10 ng/mL, 100 ng/mL, and 1,000 ng/mL for a given fluorogenic substrate flow rate (3 µL/h). This shows that the change in substrate fluorescence is directly proportional to the concentration of AbI used.
Figure 7: Fluorescence measurement areas. The fluorescence measurements are shown for different concentrations of primary anti-lysozyme antibody used: (A) 0 ng/mL, (B) 10 ng/mL, (C) 100 ng/mL, and (D) 1,000 ng/mL. Blue circles show the area of fluorescence measurement before and after the trap formed by the 5 µm height restriction, where the substrate reacts with the HRP-conjugated secondary antibody. All images correspond to a flow of 3 µL/h. Abbreviation: HRP = horseradish peroxidase. Please click here to view a larger version of this figure.
However, for a given AbI concentration, the level of fluorescence obtained is a function of the flow rate used for the fluorogenic substrate. Thus, the conversion capacity of this substrate by the HRP enzyme is inversely proportional to the flow rate. Different substrate flow rates of 1 µL/h, 3 µL/h, 5 µL/h, and 10 µL/h were evaluated. For each experiment, the curves corresponding to the difference in fluorescence given by the substrate before and after passing through the magnetic trap were obtained. Figure 8A–C shows the curves obtained for a concentration of 100 ng/mL for three different flow rates: (A) 10 µL/h, (B) 3 µL/h, and (C) 1 µL/h. Depending on the flow rate used, the conversion capacity of the substrate by the HRP-conjugated secondary antibody placed in the magnetic trap varies. The green curve represents the fluorescence intensity of the substrate after conversion by the HRP located in the trap. It can be seen that, at higher fluxes, the maximum level of fluorescence obtained decays. The red curve represents the basal fluorescence before the substrate reaches the trap. The measurement of substrate fluorescence in this area of the trap remains constant, and its value depends on the degree of non-specific interactions if there is no efficient blocking of the channel surface. We calculated the difference between both curves, represented by the blue curve.
Figure 8: Fluorescence curves. Graphs obtained at a concentration of 100 ng/mL of primary antibody and flow rates of (A) 10 µL/h, (B) 3 µL/h, and (C) 1 µL/h. The three colored curves represent the fluorescence measured before the trap (red curve) and after the trap (green curve) and the difference between the two (blue curve) over time. Please click here to view a larger version of this figure.
Figure 9A–C integrates the fluorescence difference curves measured before and after immunoreaction for the different flows for AbI concentrations of (A) 0 ng/mL, (B) 10 ng/mL, and (C) 1,000 ng/mL. The measurement for a flow rate of 0 µL/h indicates the measurement of immunoreaction under static conditions where diffusion governs. For a concentration of 1,000 ng/mL, the fluorescence saturates for all the flows evaluated. However, the staggered pattern of fluorescence at the different flows is due to diffusion of the substrate, which reacts with such a high conversion rate that it overcomes the smaller flows. Thus, there is a return of this fluorescence to the upstream zone of the trap.
Figure 9: Fluorescence difference ratios. Summarized curves of the fluorescence differences (after−before) for the different fluxes used at (A) 0 ng/mL, (B) 10 ng/mL, and (C) 1,000 ng/mL. Please click here to view a larger version of this figure.
Measurements obtained with this device make it possible to generate a standard calibration curve for immunoassays performed with lysozyme as a model antigen. Figure 10 shows the calibration curve obtained from the maximum values of the differences between the fluorescence before and after the immunoreaction performed in the magnetic trap. This protocol allows the detection of the primary anti-lysozyme antibody with concentrations in the order of nanograms per milliliter using flows between 1 µL/h and 10 µL/h. The high variability and high fluorescence levels at 1 µL/h suggest that the reactivity of the immunocomplex is such that this rate does not favor the flow of the reacting substrate and tends to accumulate just after the trap, in addition to the fact that the resistance of the device reduces the resolution of the device for this flow rate.
Figure 10: Calibration curve. The graph shows the maximum value of the differences in fluorescence intensity of the curves obtained with respect to the concentration of primary antibody used (AB1) for each flow rate. I/Isat corresponds to the ratio of the fluorescence value obtained for each fluorescence measurement (I), normalized with respect to the maximum fluorescence value reached upon saturation (Isat). Error bars represent the standard deviation of the three experiments. Please click here to view a larger version of this figure.
Supplemental Figure S1: Piezoelectric platform controller interface. The left image shows the interface that controls the z-axis displacement of the piezoelectric platform. The restriction is created by raising the platform through the application of voltage to three piezoelectric actuators. The right image shows the interface of the software that controls the micromilling machine, where it is possible to observe the precise coordinates in the x- and y-axes where the restriction is machined at a speed of 11,000 rpm. Please click here to download this File.
Supplemental Figure S2: Design of the device. The design of the device created using the design software is seen on the left. The design consists of two channels connected, one being the main channel that contains the 5 µm height restriction and the other being the side channel for the waste outlet. The channels are micromilled with a 200 µm end mill bit. The panel on the right shows the fabrication parameters. Please click here to download this File.
Supplemental Figure S3: Micromilling parameters of inlet and outlet holes. (A) 1.2 mm diameter and 0.65 mm depth holes (half the acrylic thickness) are micromilled on the machined face. The panel on the right shows the displacement speed, rotation, and depth parameters of the end mill bit. (B) 1.5 mm diameter and 0.7 mm depth holes are micromilled on the opposite face connecting the holes. The panel on the right shows the fabrication parameters. Both cases are micromilled with an 800 µm end mill bit. (C) Reagent inlet (right) and outlet (left) holes are shown after machining both faces. The dimensions of the larger diameter half allow the hose to be coupled, while the barrier created by the smaller diameter half prevents the hose from blocking the inlets. Scale bars = 1 mm. Please click here to download this File.
Supplemental Figure S4: Nanoparticle separation. Using a commercial magnetic separator, 100 nm nanoparticles can be easily concentrated to perform the washing steps during the immunoassay. The pellet formed after 15 min is observed in the red circle. Please click here to download this File.
Supplemental Coding File 1: Code for acrylic surface grinding. The generated code is shown with instructions for grinding the 25 mm x 9 mm acrylic surface along the x- and y-axes. The last line of the code positions the drill bit right at the coordinate where the constraint will be machined. Please click here to download this File.
Supplemental Design File 1: Design of the microchannel and reagent inlet and outlet holes. The design contains two layers of different colored structures (black channels and red holes) that are machined on the previously ground acrylic face. Please click here to download this File.
Supplemental Design File 2: Design of the holes on the opposite side. The design consists of holes of larger diameter for the opposite face that communicate with the previous ones and serve to fix the hoses. Please click here to download this File.
An acrylic microfluidic device for immunoassays using nanoparticles as immunosupport was fabricated using a micromilling technique. The method of direct manufacturing on the substrate has the advantage of avoiding the use of a master mold and the time and costs that this implies. However, it is limited to rapid prototyping and high-volume device manufacturing.
Here, we used a previously reported accessory piezoelectric platform for the milling machine12. The platform was fabricated by 3D printing to create variable-depth channels with a vertical resolution better than the 10 µm resolution of conventional micromilling machines. We achieved the milling of channels of up to 5 µm high that form a staggered restriction in the 200 µm microfluidic channel.
Only two end mill bits of 200 µm and 800 µm diameter are required for the fabrication of the microfluidic device, without any need to manually change bits. The milling machine model used performs the exchange of bits automatically, which helps in time optimization. In addition, the "Z0 sensing" mode allows the determination of the origin automatically in the z-axis with high precision by contacting the sensor and the pin included in the milling machine.
The design of this device is simple, consisting only of the main channel interrupted by a 5 µm staggered restriction and an accessory channel that serves as a purge and inlet for microparticles. However, the high precision of such structures is required for the restriction to trap iron microparticles of 7.5 µm in diameter. A larger restriction allows microparticles to pass through without retention, which prevents their capture and the formation of the microparticle trap. Conversely, a smaller restriction greatly increases the flow resistance of the channel and causes the reagents to exit through the side channel instead of passing through the porous microparticle medium. Using microparticles of a larger diameter to form the magnetic trap would allow for making a larger restriction. For example, particles with a diameter of 40 µm need a ~30 µm staggered restriction. In that case, it is not necessary to use the piezoelectric platform, and the fabrication protocol is simplified. However, a porous medium composed of larger microparticles would trap nanoparticles differently and would impact the performance of the immunoassay; however, it is not clear whether it would be for better or worse. Work is underway to study these features to optimize the device7.
To achieve the required accuracy, surface grinding was performed to remove a 30 µm layer of the acrylic surface by moving the 200 µm end mill bit through the entire acrylic surface at 14,500 rpm. This process allows a new origin to be obtained at the z-axis along the entire surface for greater accuracy in height. It is crucial to always align the acrylic in the same position so that the origin coordinates on the x- and y-axes (previously set) coincide with one of the corners of the acrylic. Likewise, one should make sure that the acrylic is perfectly seated on the base; otherwise, the surface may not wear properly, causing problems in the following assembly steps.
Once the program completes the surface grinding process, it automatically brings the end mill bit to a specific coordinate where the 5 µm constraint forming the trap will be machined. It is critical to prevent the end mill bit from lifting off the surface once it reaches this coordinate. It is recommended to identify if the micromilling machine has an option that prevents the end mill bit from detaching from the surface once the grinding process finishes. Otherwise, it will be necessary to manually reposition the cutter to this coordinate, which may result in location accuracy errors.
To machine the staggered restriction in the acrylic, it is necessary to oversize by 1.5 µm. For 5 µm height, the piezoelectric platform was raised to 6.5 µm because the microchannel tends to flatten a little in the process of sealing the channels. In addition, the final length of the restriction in the device is only 50 µm; however, it is machined longer to ensure that it communicates with both ends of the main channel when it is machined afterward. In one step, the 200 µm deep channels are fabricated using the 200 µm diameter end mill bit and a rotational speed of 11,000 rpm. Thus, machining of the 200 µm wide channels in acrylic takes only a few seconds.
The next step in the fabrication of the device is to machine the holes that connect the channels to the outside. These holes are used to place the hoses that allow the device to be easily connected to the syringe pump that runs it. One of the problems encountered here is the placement of the hoses. An optimal distance is critical to allow the flow of reagents and avoid creating a seal with the face of the unmachined acrylic. To overcome this drawback, the holes are machined in two steps to create a boundary where the hoses do not exceed the desired depth.
With the 800 µm end mill bit, the pattern of holes was milled on the same previously micromilled face. Holes are machined into the end of each channel with a diameter of 1.2 mm and a depth of 0.65 mm, which is half the thickness of the acrylic. In addition, two holes are machined in the contralateral corners of the rectangle, which allow the acrylic to be aligned face down on the platform with pillars. It is important to use a scraper or cutter to remove the acrylic from the piezoelectric platform to avoid breaking it. On the opposite side of the acrylic, the diameter of the holes is 1.5 mm (larger than the previous half), which is equal to the diameter of the hose to be fixed. The depth is 0.7 mm and should be slightly greater than half of the front face hole to ensure that both the machined holes communicate with each other. The barrier formed by the smaller diameter hole prevents the hose from reaching the bottom and blocking the inlet. This implementation in the protocol greatly improves the placement of the fluid inlet and outlet hoses to the chip.
A key step in device manufacturing is sealing. It is necessary to seal the face that contains the machined channels with an unmachined acrylic cover with the same dimensions. Sealing the sheets with a method that does not approach the glass transition temperature of acrylic allows for obtaining deformation-free channels. Through the bonding method used, a thin film of solvent between the two PMMA sheets dissolves a thin film from the surface of the PMMA sheet, then evaporates, and finally reconnects the monomers of the PMMA sheets at a specific operating temperature.
The widely described method of chloroform vapor exposure was used to seal this acrylic device. As previously reported in the literature, chloroform provides high bond strength; however, as chloroform attacks acrylic very aggressively, acrylic should never come into direct contact with liquid chloroform, only with gaseous chloroform. It is essential to maintain an optimal distance between the evaporating chloroform and the acrylic. We created a simple system using a glass Petri dish, in which the acrylic was adhered to the lid. A platform made up of nine slides joined and attached to the lid of the Petri dish was used, on which the acrylics were adhered with double-sided tape to regulate the distance. Although the gaseous chloroform process is very sensitive to changes in ambient temperature, the temperature of the Petri dishes can be controlled by placing them inside a polystyrene box. The water seal prevents the chloroform from evaporating too quickly.
In contrast, there are other reported joining methods that are faster and do not require special equipment; however, some are only suitable for joining two acrylic sheets without machining. Likewise, it is difficult to control the bonding force since the solvent must be poured directly between both acrylic sheets18, with a high risk of melting structures of small dimensions such as the micromilled 5 µm restriction.
Using the protocol described here, homogeneous seals were achieved with a homemade press, with which the sealing pressure and temperature were controlled. For the construction of this press, both a framework of aluminum and a hinge create a lever that amplifies the mechanical force coming from a weight of 5 kg to a factor of 9:1. The heating elements transfer their heat to 2 cm thick aluminum plates that are in contact with the acrylic pieces.
Once the acrylic layers are sealed, the last step in the fabrication process is to connect the external hoses to the corresponding inlets and outlets of the device. The liquid adhesive is applied externally once the hoses are inside the holes. If the diameter of the hose is smaller than the holes, the liquid adhesive can get into the device, clogging the channels.
Direct fabrication methods on the substrate, such as micromilling, have some disadvantages, such as roughness of the machined surface and structures with poorly delimited edges9. Despite the roughness of the machined surface of the channels, no additional treatment is necessary for this device. An important point to note about this bonding protocol is that chloroform allows for decreasing the roughness of the milled microchannels by polishing the surfaces exposed to the solvent. The polishing of the surface is so good that even the optical quality is improved19.
When the device has been fabricated, it must be prepared to be used in an immunoassay. Ultrasonic washing of the device is of great importance to remove any unwanted acrylic debris that may clog the channels and interfere with the immunoassay.
In addition, special attention must be paid to the correct blockage of the channels to avoid non-specific interactions and high background noise signals. Blocking with 5% BSA has been shown to reduce the noise of non-specific binding of antibodies in the channels, and 1 h of incubation is sufficient. The iron microparticles are coated with a silica-PEG layer to prevent non-specific binding in them. Furthermore, the microparticles are blocked with BSA (at the same time as the channels) before forming the trap in the device. Overnight incubation at 4 °C is better, meaning that 1 day is required for the fabrication and incubation of the device before the immunoassay. The reduction of the roughness by the chloroform and the blockage of the surfaces to reduce non-specific interactions have been shown to work exceptionally well, allowing this platform to achieve a limit of detection comparable to standard ELISA in a model immunoassay4.
The formation of the microparticle trap in the microchannel restriction is a manual process. Although there is no precise control of the number of microparticles entering, we rely on an estimate of the length of the compacted particles. It is possible to make the trap larger or remove the excess easily. The discarded particles accumulate in the side channel and do not need to be removed from the chip. It is crucial to prevent microparticles from flowing into the main channel, as they can interact with nanoparticles upstream of the trap and modify the immunoassay measurements.
As a proof-of-concept, we implemented immunodetection of a complex formed by lysozyme protein conjugated to 100 nm diameter nanoparticles by incubation of the specific primary antibody and the corresponding HRP-labeled secondary antibody. The immunocomplex is formed outside the device. However, it is possible to perform the entire immunoassay inside the device. The magnetic properties of the nanoparticles allow for easy manipulation with a high-performance neodymium permanent magnet separator. One must simply keep the reaction microtube for 15 min to allow the nanoparticles to be attracted by the magnet to the microtube wall, which facilitates the immunoassay wash steps outside the device.
The highlight of this device is that it is simple, fast (total assay time of 40 min4 compared to 4-6 h for ELISA), and cheap (comparable to the price of a lateral flow test). All these features make this device profile a good candidate for the development of POCT. In addition, the device can quantitatively detect analyte concentrations similar to the gold standard ELISA4, which is extremely valuable for epidemiological studies and decision-making in public health. Usually, microfluidic systems for immunoassays have good limits of detection but long assay times, or they have short assay times but suboptimal limits of detection4. By combining magnetic nanoparticles as immune support and fluorogenic enzymes for detection, this platform has a short assay time and a good limit of detection (in previous work, we found a limit of detection of 8 pg/mL using biotin as an antigen, which is comparable with a standard ELISA4). Finally, this technology has an important advantage over lateral flow tests: the possibility of giving quantitative and not only qualitative results ("yes" or "no"). Unlike most other microfluidic systems developed in laboratories in which rare materials are used, this system is made of acrylic (PMMA), a low-cost thermoplastic that is highly compatible with the mass production of medical diagnostic devices.
The authors have nothing to disclose.
This work was supported by Conacyt, Mexico under grant 312231 of the "Programa de Apoyos para Actividades Científicas, Tecnológicas y de Innovación", and by AMEXCID and Mexican Foreign Relations Ministry (SRE) under grant "Prueba serológica rápida, barata y de alta sensibilidad para SARS-CoV-2". JAHO thanks Conacyt Mexico for their PhD scholarship.
0.008 Endmill | KYOCERA SGS | 2204 | 2FL 0.008×1/8×0.12×1-1/12 |
0.032 Endmill | KYOCERA SGS | 2228 | 2FL 0.032×1/8×0.48×1-1/12 |
Carbonyl-iron microparticles | Sigma-Aldrich | 44890 | 7 μm |
Chloroform | Fermont | 6201 | Health Hazard: Moderate Flammability: None Reactivity: None Contact Hazard: Moderate |
CMOS camera Moment | Teledyne Photometrics | Sensor Technology: CMOS Quantum Efficiency: 73% Pixel Size: 4.5 µm x 4.5 µm Supported Interfaces: USB 3.2 Gen 2 |
|
Dr Engrave Software | Roland DGA Corporation | Engraving software to design and create the engraving path on the surface | |
Extraction hood | Unknown | Unknown | |
Flexible Plastic Tubing | Tygon | AAD04103 | ID = 0.020, OD = 0.060 |
Fluorescence microsope | ZEISS | Axio Vert.A1 | |
High Precision Dispense Needle | Loctite | 98612 | |
Homemade piezoelectric controller application | LabView | See reference 12 for more details. | |
Loctite 495 instant adhesive | Henkel | 49503 | Apply with micropipette tip or dispensing needle |
MagJET Separation Rack | thermoscientific | 12 x 1.5 mL | |
Mechanic press | Home-made | ||
Milling Machine | Roland | MDX-50 | |
Piezoelectric platform | Home-made | See reference 12 | |
Polymethylmethacrylate – Sheet – PMMA, Acrylic | Goodfellow | ME303018/1 | Thickness: 1.3 mm, Transparency: Clear/Transparent |
PVCamTest software | Teledyne Photometrics | Version 3.10.107 | Image acquisition software |
Stereo microscope | Nikon | SMZ 7457 | |
SuperMag Carboxyl Beads | Ocean NanoTech | KSC0100 | 100 nm |
Syringe pump | kd Scientific | KDS200 | Can hold up to two syringes |
Utrasonic bath | Branson | 2800 | |
VPanel software | Windows OS | Version 1.0.3.0 | Software for controlling the micromilling machine |