Integrated microfluidic device for DNA extraction using electric valves

Abstract

Deoxyribonucleic acid (DNA) encodes the genetic instructions that govern the development and function of living organisms, and its isolation is a critical prerequisite for downstream applications such as sequencing and diagnostic assays. Here, we present an integrated lab-on-a-chip platform that automates genomic DNA extraction by combining a passive micromixer for chemical cell lysis with a silica-based filtration unit. The system comprises three microfluidic chips for mixing, lysis, and purification, interconnected by electrically actuated valves to precisely route fluids through each stage. We systematically optimized key parameters fluid velocity, filter thickness, lysis temperature, and cell concentration, and identified the ideal operating conditions: 2.5 × 10⁶ cells/mL input, 65 °C lysis, a single silica filter, and a 25 mL vacuum chamber to drive flow. Under these settings, the device delivered 147.23 ± 29.04 ng/µL DNA, corresponding to approximately 90% extraction efficiency relative to a commercial silica-column kit. Finally, we confirmed the integrity of the recovered DNA by PCR amplification and agarose-gel electrophoresis on seven representative samples. These results demonstrate that our multi-chip microfluidic platform provides rapid, efficient, and semi-automated genomic DNA isolation, offering a robust alternative to conventional bench-top methods.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-44973-9.

Introduction

DNA is the hereditary molecule that carries genetic instructions for the development, structure, and functioning of all living organisms, and is composed of nucleotides arranged in a double-helix structure^1,2. Research on DNA has transformed biology and medicine by enabling the understanding of genetic diseases, the development of diagnostics and genetically modified organisms, and applications such as DNA profiling^3,4.

The DNA extraction process is done using different methods, according to the type of sample the final goal. Among the most widely used DNA extraction techniques are the phenol-chloroform method, the magnetic bead-based method, and the silica column-based method⁵. The phenol-chloroform extraction relies on organic solvents to separate DNA, but it is labor-intensive, toxic, and not easily automatable. The magnetic bead-based method uses functionalized beads that bind DNA in the presence of chaotropic salts and is commonly used in automated systems. In contrast, in the silica column-based mechanism, DNA binds to a silica membrane under specific buffer conditions and is eluted after washing. This approach is compatible with microfluidic integration and automation while maintaining high DNA yield and purity⁶.

To extract DNA from a biological sample, the cells must first be lysed; Cell lysis is the process by which the cell membrane is disrupted to release intracellular components such as DNA, RNA, and proteins⁷. Among the available lysis approaches, including chemical⁸, electrical⁹, mechanical¹⁰, laser¹¹ and thermal¹² methods, chemical lysis has attracted significant attention in microfluidic applications due to its simplicity, low structural complexity, and compatibility with microfluidic systems, making it a suitable option for commercial cell-lysis platforms^13,14. This method includes the use of chemicals solutions to disrupt cell membranes and release cell contents, including DNA¹⁵. The chemical lysis method is a simple and low-cost method because only a lysis reagent is required, and for this reason, this method is most effective in microfluidics^16,17.

Microfluidic technology offers precise control and manipulation of liquids at the microscale, enabling efficient cell lysis and mixing while integrating multiple functions and processes into a single platform¹⁸. The integration of cell lysis and mixing functions enables efficient sample preparation, reduces sample volume and reagent consumption, and facilitates high-throughput automated processes, making microfluidic platforms suitable for a wide range of applications¹⁹. By enabling effective mixing, the entire process of cell lysis, DNA extraction, purification, and washing can be enhanced²⁰. For this purpose, various microfluidic mixing techniques have been investigated over the past few decades²¹. These techniques are generally classified into two main categories: active and passive mixing²². I In active mixing, external energy sources are applied to perturb fluid flow²³, including acoustic²⁴, magnetic²⁵, dielectrophoretic²⁶ and thermal²⁷ forces. Passive mixing, which relies on channel geometry and flow characteristics rather than external actuation, has also been widely employed and can additionally contribute to cell lysis²⁸.

Conventional DNA extraction methods are typically performed using macroscale laboratory equipment such as centrifuges, heaters, shakers, and incubators. In contrast, lab-on-a-chip DNA extraction represents a key advantage of microfluidic technology, enabling efficient and automated DNA extraction from diverse sample types by integrating multiple extraction steps, including lysis, binding, and washing, within a single microfluidic device. These systems are generally based on laboratory DNA extraction protocols adapted for microscale implementation, and their processes may include sample filtration, dilution, extraction, and purification²⁹. The advantages of microfluidic DNA extraction include reduced sample and reagent volumes, faster processing times, precise control over fluid flow and mixing, and the ability to operate in an automated and high-throughput manner^30,31.

Recent studies highlight different integration strategies and associated challenges. Chen et al. demonstrated a continuous-flow microfluidic device integrating cell separation, lysis, and DNA purification, showing that increased flow rates can reduce lysis efficiency due to shortened residence time³². Karle et al. addressed continuous processing using phase-transfer magnetophoresis, achieving rapid DNA extraction with high recovery compared to batch methods³³. Beyond extraction-only platforms, Huang et al.34 and Zhao et al.35 reported fully integrated systems that couple nucleic acid extraction with on-chip amplification and detection, including digital PCR and isothermal amplification. These studies demonstrate that closed, integrated microfluidic systems can achieve sensitive and automated nucleic acid analysis, while emphasizing the importance of balancing integration, throughput, and extraction efficiency.

However, despite these advances, most reported platforms rely on fully disposable cartridges, complex actuation schemes, or tightly coupled integration of extraction and detection modules, which can increase cost, limit reusability, and hinder deployment in portable or point-of-care settings. In particular, existing automated systems rarely preserve the modularity and protocol familiarity of standard silica-based extraction while providing programmable, stepwise automation. A key unmet need therefore remains for semi-automated microfluidic DNA extraction systems that decouple critical functional components, enable reuse of fixed elements, and reduce per-test cost while maintaining reliable extraction performance and compatibility with point-of-care operation.

In this study, we address this need by introducing a semi-automated and modular microfluidic platform that enables programmable, vacuum-driven execution of a conventional silica column-based DNA extraction protocol, while avoiding fully disposable, tightly coupled architectures. The proposed platform consists of three functionally decoupled microfluidic chips (mixing, lysis, and extraction) interconnected via electronically actuated valves. A programmable controller sequentially switches these valves to direct fluids through a silica membrane in the extraction chip, enabling automated execution of each extraction step with precise control over timing and flow.

Importantly, this work does not introduce a new DNA extraction chemistry; instead, it translates the widely used silica column-based method into a microfluidic format with system-level automation and reusability not typically achieved in prior lab-on-chip extraction devices. Whereas conventional commercial kits rely on centrifugation or manual vacuum to drive fluids through the silica membrane, the present system reproduces this vacuum-driven mechanism using syringe pumps and electrically controlled valves, eliminating the need for user intervention while maintaining compatibility with established laboratory protocols. By reusing fixed components and limiting disposability to essential elements, this modular design reduces operational complexity and per-test cost, enhancing suitability for portable and point-of-care applications. To evaluate and optimize extraction performance, the effects of filter number, fluid velocity, temperature, and cell concentration were systematically investigated. The quality of the extracted DNA was further validated using PCR amplification and gel electrophoresis.

Device principle and theory

In this study, a commercial DNA extraction kit is utilized to extract genomic DNA from cells. In this kit, cell lysis and DNA purification proceed as follows. First, 200 µL of cell suspension is mixed with 200 µL of lysis buffer and 20 µL of proteinase K. To lyse the cells, the mixture is incubated at 65 °C for 10 min. Next, 200 µL of absolute ethanol is added. The sample is applied to a silica-filter column and centrifuged to bind DNA to the membrane. The column is then transferred to a fresh collection tube, and 500 µL of Wash Buffer 1 is passed through by centrifugation to remove contaminants. This wash is repeated with 500 µL of Wash Buffer 2 to eliminate any remaining impurities. Finally, the filter is moved to a new tube, 200 µL of Elution Buffer is added to release the DNA, and the eluate containing purified DNA is collected at the bottom of the microtube. Figure 1 illustrates the summary of this protocol using the usual facilities in a laboratory.

Fig. 1.

DNA extraction protocol from cell samples to purified DNA. The sample is heated after ten minutes and turned to lysate which is placed in a silica column. Then it is centrifuged to pass through the filter. This centrifugal step is repeated with two washing buffers and the elution buffer.

As shown in Fig. 2, the designed device is a system including electric valves, microfluidic chips and a thermal module. In the first, sample cells and lysis buffer pump in the mixing chip using a syringe pump out of the system which is connected to the green tubes in the figure, and the mixed fluids are sent to the next chip using the tube. The Second chip is named as lysis chip because the lysate is created in the reaction chamber. In order to prepare the suitable temperature for the lysis process, an electric heater is integrated under the two first chambers; the second chamber is in the same heating condition with the reaction chamber to control the temperature, since a thermometer sends feedback to a controller which turns of the heater if the temperature reach 65˚C. After ten minutes, pure ethanol is added to the reaction chamber, and other buffers including the first and second washing buffers and the elution buffer are loaded into the buffer chambers. In the next step, the other syringe, out of the device, produces a vacuum chamber which is connected to the red tube named waste vacuum outlet in Fig. 2. At this moment the number 1 and 6 valves are opened, and then the lysate passes through the silica filter integrated in the extraction chip and guided to the waste vacuum outlet. Then, the 6th valve is switched; the waste fluid is pumped into the system and goes to the waste microtube. Afterward, the 6th valve is switched back, and the waste vacuum outlet sucks the first washing buffer through the 2nd valve which is open now. This step is repeated for the second washing buffer too, and the two washing buffers eliminate the other materials except DNA from the silica filter domain. Following this, the syringe connected to the target vacuum outlet sucks the elution buffer passing through the 4th valve, silica filter, and 7th valve and washes all the pure DNA which were stored in the silica filter. Subsequently, the 7th valve is switched, and purified DNA is sent to the target microtube pumping from the syringe connected to the target vacuum outlet.

Fig. 2.

Schematic of the designed device including different microfluidic chips, valving and thermal system for flow fluid control. Colored star symbols indicate external instrument connections: the green star denotes the inlet syringe pump connection, the red star denotes the outlet syringe pump connection, and the black star denotes the electrical connection to the power supply.

Later, the ethanol stored in the last chamber of the lysis chip is circulated inside the device to wash the valves, and finally, the chips are exchanged for the next test, and a round of washing with DI water will be done to reduce the risk of cross-contamination for the next samples.

Materials and methods

Materials preparation

The input sample for DNA extraction is MCF-7 breast cancer cells which were obtained from Stem Cell Technology Research Center, Tehran, Iran. The cells were grown in 10 cm culture dishes and were maintained in DMEM with 10% heat-inactivated fetal bovine serum. DynaBio™ Blood/Tissue DNA Extraction Mini Kit was purchased from Takapouzist company, Tehran, Iran; this kit includes all the buffers used in the protocol.

The device body, lysis and extraction chips are made from Polymethyl methacrylate (PMMA) material which is prepared from a local market. Electric valves were purchased from Module-center shop, Tehran, Iran. Moreover, other electric modules such as controller, relay, thermometer and heater are prepared from Jahan Kit shop, Tehran, Iran.

Spectrophotometric analyses of DNA

The concentration and purity of extracted DNA are important parameters in DNA extraction for downstream analyses to be accurate and reliable. Analyzing the absorbance ratio at 260/280 nm and 260/230 nm is a common method for determining DNA purity. As a result of the 260/280 ratio, it determines how much DNA and protein contamination is present in the sample, with a ratio of ~ 1.8 indicating pure DNA. The 260/230 ratio reflects the presence of contaminants such as salts and organic compounds, with a ratio of ~ 2.0 indicating minimal contamination. In this study, the concentration and purity (260/280 and 260/230 ratios) of the extracted DNA were measured with a NanoDrop™ Take3 plate of Cytation™3 (Bio Tek, Agilent Technologies, Inc, USA), using 2 µL of each sample. Each extraction condition was independently repeated five times, and DNA yield was quantified for each replicate.

Agarose gel electrophoresis

The DNA integrity was analyzed by agarose gel electrophoresis using 1% agarose for genomic DNA (gDNA). Ethidium bromide (EtBr) was used to stain the agarose gel. Electrophoresis was performed utilizing 1x TAE (Tris Acetate-EDTA) buffer and a constant voltage of 100 V for 40 min. We visualized DNA bands and acquired images using a Gel Doc (UVITEC Cambridge).

PCR amplification

For PCR analysis, each DNA sample was diluted to a working concentration, and performed in the device Steponeplus™, Applied Biosystems, USA. The amplification was performed by a housekeeping gene with the forward primer GAPDH: 5′- TTCAACAGCGACACCCACTC-3′ and reverse primer GAPDH: 5′- GAAGATGAAAAGAGTTGTCAGGGC-3. PCR reactions were carried out in a final volume of 15 µL that contained the following: Master mix 2 × 7.5 µL (Master mix 2x SMOBIO™, Technology, Inc, Denmark), 1 µL of DNA template (0.1–10 ng/µL),0.5 µL of each forward and reverse primers (5 µM), and 5.5 µL of sterile water. PCR thermal cycling conditions were as follows: pre-denaturation at 95 °C for 10 min, followed by 40 cycles of denaturing at 95 °C for 30 s, annealing at 60 °C for 30 s for GAPDH, extension at 72 °C for 30 s, with a final extension at 72 °C for 5 min.

Microfluidic chip fabrication

There are three different chips including mixing chip, lysis chip and extraction chip. The mixing chip is fabricated by polydimethylsiloxane (PDMS) material and standard photolithography³⁶ because the designed channel had sharp and small features which is shown in Fig. 3a. On the other hand, as the lysis and extraction chips did not have small features which are shown in Fig. 3b and c, PMMA is utilized to fabricate them by CNC machine (Roland MDX 40 A, Japan). Figure 3d depicts the lysis chip layers, which will be bonded using chloroform and aligner holes. The lower layer contains horizontal holes as an outlet for this chip toward electric valves. Figure 3e illustrates different layers of the extraction chip. The upper layer includes five horizontal holes as inlets of this chip, which are connected by a channel. The circle in the middle layer is considered for the silica filter which all the input fluids should pass through that. The lower layer of this chip includes two connected outlets that guide output materials to two different valves.

Fig. 3.

Microfluidic chips utilized in the device. a) Mixing chip containing omega-shaped arrays as the micromixer, b) lysis chip with different chambers for buffers, c) extraction chip including a circle chamber for silica filter, (d) lysis chip and (e) extraction chip layers before the assembly. (Dimensions are provided in the supplementary information)

Chip-valve connectors

In order to connect electric valves, which are permanent units of the system, to the disposable chips, a connector was needed to eliminate leakage possibility; since the fluid moves thanks to vacuum pressure, a special 3D printed part was designed and fabricated that is shown in Fig. 4. The connectors were designed with a secure fit within the channels, effectively preventing any potential leakage. To achieve this, the connectors feature multiple narrow protrusions along their axis on both sides. Notably, the thickness of these protrusions decreases from the end to the top, maximizing the contact area under pressure. For the fabrication of connectors, Anycubic Tough flexible epoxy and the Anycubic Photon M3 Plus 3D printer were employed. The optimal number of protrusions and the flexibility of the epoxy ensure optimal contact between the connectors and the inlets/outlets, resulting in a 0% leakage rate. Figure 4a and b shows the 3D and 2D design of connectors which indicate the ability of connectors to eliminate leakage because not only their material is flexible, but also the protrusions are designed to bend more when going to the holes of valves and chips. As shown in Fig. 4c two types of connectors were employed in this study: one for linking the valves to tubes and another for connecting the valves to the chip, as well as for interconnecting different chips.

Fig. 4.

(a) Three-dimensional CAD model of the connectors (dimensions in millimeters), (b) two-dimensional cross-sectional view of the connectors with all relevant dimensions, and (c) fabricated connectors produced by 3D printing. A scale bar is included in (c) for reference.

Control system

Heat transfer and fluid flow in the system should be controlled. A constant temperature of 65˚C for ten minutes is needed during the lysis process. Therefore, an electric module integrated under the lysis chip generates heat, and simultaneously, a thermometer inside the reservoir sends feedback to the controller (Arduino Uno). Then, the controller turns off/on the heater based on the feedback using a relay, and maintains the temperature at a constant value.

The fluid movement has two steps; first, the syringe pump (Fanavaran Nano-Meghyas, Tehran, Iran) pumps the samples into the lysis chip. Secondly, another syringe pump sucks the fluids in five steps; each electric valve is opened by the controller in its desired step. Therefore, fluid flow is controlled by Arduino and moves inside the channels by syringe pumps. It should be noted that progression between steps is initiated using a button integrated into the control system. While key operational steps (e.g., fluid routing, temperature control, and predefined processing sequences) are automated once initiated, certain actions such as buffer loading, chip replacement, and button-based step switching still require user involvement. Therefore, the device is appropriately classified as a semi-automated platform rather than a fully automated one.

Result and discussion

Mixing

The first unit of the integrated system is the mixing chip that is shown in Fig. 5a, and utilized to blend the cell sample and lysis buffers. The designed geometry consists of curved and sharp edges which are the reason for secondary flow generation^37,38; this kind of flow increases mixing quality because its direction is orthogonal to the main flow direction³⁹. Therefore, the two fluids passing through the channel will be mixed better than a channel without secondary flows⁴⁰. Figure 5b-f illustrates the mixing test using a blue food dye solution to analyze the quality of the designed geometry.

Fig. 5.

The mixing system. a) The connections and location of the mixing chip in the integrated device; the outlet is connected to the lysis chip using a tube. b) Inlet condition and c) outlet condition for the mixing chip. d) Mixing condition for second, e) fourth, and f) sixth Omega unit. For panels (b-f), background features outside the microfluidic channels were removed during image processing, and only the channel regions are shown to improve visualization of the fluid interfaces.

An inlet flow rate of 0.6 mL/min at each inlet corresponds to an average fluid velocity of approximately 500 mm/s in the 200 μm × 200 μm microchannel. Under these conditions, the smooth and well-defined liquid interface observed in Fig. 5b arises from the high flow velocity, which induces strong secondary flows in the curved and sharp sections of the fluid path. In this device, mixing is governed not only by molecular diffusion but also by secondary flow effects, as discussed in our previous study¹⁷. As a result, sharp and stable interfaces can be observed near the inlet region of the channel under experimental conditions.

Therefore, the two fluids enter the channel as completely separated streams and progressively interdiffuse while flowing toward the outlet. Qualitatively, a visually homogeneous intensity distribution is observed at the outlet (Fig. 5c), suggesting effective mixing under the tested conditions. While this qualitative assessment is sufficient for the intended application, we additionally performed a semi-quantitative image-based analysis to compare mixing performance across trials.

To assess mixing uniformity, bright-field micrographs (1360 × 1024 px) were acquired at the outlet junction after injecting a blue food dye solution (20 µL dye in 1 mL DI water) and pure DI water through the two inlets. All images were background-subtracted prior to analysis. A 50 μm × 50 μm region of interest (ROI) was selected at the inlet for both the pure-dye and pure-water streams; the mean pixel intensity of the dye ROI was scaled to 100 and that of the water ROI to 0. Under this normalization, a perfectly uniform mixture would yield a pixel intensity near 50.

For each outlet image, pixel intensities were normalized using these reference values, and the standard deviation (σ) of their deviation from the midpoint intensity (50) was computed. Across multiple trials (n = 20), σ remained below 2, indicating a highly uniform optical intensity distribution at the outlet. Because this analysis assumes a monotonic relationship between dye concentration and image intensity and does not explicitly calibrate optical density versus concentration, σ is interpreted as a relative, semi-quantitative metric of mixing uniformity rather than an absolute measure of concentration homogeneity. Within this context, the consistently low σ values indicate effective mixing at the outlet for the operating conditions studied.

Heating system

After the mixing step, the mixture of cells and lysis buffer remains in the lysis chamber for ten minutes at 65 °C. Figure 6 illustrates the temperature control: once the electric heater turns on, the temperature rises to 65 °C, overshoots slightly due to high heater power, then falls back to 65 °C and triggers the relay to switch the heater on and off. Across ten repeated heating cycles, the system reached and stabilized at 65.0 ± 0.1 °C within 150 s. Importantly, the cell-lysis buffer mixture is introduced into the lysis chamber only after the temperature has stabilized at 65 °C, ensuring that the sample is not exposed to the initial heating overshoot. In this heating system, a relay alternately energizes and de-energizes the heater to maintain a constant temperature. In Fig. 6, the dark blue curve denotes the average temperature over ten trials, and the light blue shaded area shows the full range of individual measurements. According to these results, the temperature remains constant at the desired value throughout the lysis reaction.

Fig. 6.

Temperature profile of the lysis chamber during heating. The dark blue curve shows the mean temperature over ten independent trials, and the light blue shaded area indicates the full range of individual measurements. Under relay control, the chamber reaches and stabilizes at 65.0 ± 0.1 °C within 150 s.

Electric valves

There are two syringes connected to the outlets which create two vacuum chambers including target and waste chambers. Figure 7a indicates these vacuum chambers that suck the fluids which are loaded before in the lysis chip. Figure 7 illustrates the process using some food colors in order to elaborate each step; in Fig. 7b, the first state of the lysis and extraction chips is shown. Firstly, all the materials including lysate, washing buffers, elution buffer and ethanol are loaded in the lysis chip. Secondly, the first valve is opened and the vacuum chamber sucks the lysate into the extraction chip and passes it from the silica filter which is shown in Fig. 7c. In this step, the DNA and other impurities attach to the filter. Thirdly, just the second valve is opened and the first washing buffer washed a part of the impurities and guides them to the waste chamber; this step is shown in Fig. 7d. Based on Fig. 7e, the next step begins when the third valve is opened and the second washing buffer is guided to the waste outlet. At the end of this step, all the impurities are cleared from the silica filter and the purified DNA is stored at the filter domain. Afterward, as illustrated in Fig. 7f, the fourth valve is opened and the target vacuum chamber sucks the elution buffer, and collects the purified DNA in the outlet. Finally, the fifth valve is opened, and ethanol enters the network. The purpose of this step is cleaning the valve, therefore, the syringe pump and valves controlled in this step to pass through all valves. Details of device durability and maintenance requirements are provided in Supplementary Information.

Fig. 7.

Flow control system by electric valves. a) vacuum syringes connected to the outlets that prepare the required pressure difference to move the fluids in the channels. b) Different colors are loaded to depict the steps of the process. c) The opening of the first valve to release the lysate. d) The opening of the second valve to release the first washing buffer. e) The opening of the third valve to release the second washing buffer. f) The opening of the fourth valve to release the elution buffer. g) The opening of the fifth valve to release the ethanol for washing the system.

.

DNA concentration

To optimize DNA extraction in the device, four parameters including vacuum volume, number of silica filters, lysis temperature, and cell concentration are investigated, and the concentration of purified DNA was measured using a Nanodrop device.

Vacuum capacity

The vacuum generator is a syringe in the fabricated device. On the other hand, the size of the utilized syringe plays an important role in fluid suction. If the vacuumed volume is bigger, the suction is stronger, therefore, the fluids pass through the filter easier. As shown in Fig. 8, five different volumes are tested for the process. It can be concluded that higher values of syringe volume are efficient for DNA extraction. The suction strength is directly related to the quality of passing the fluids through the silica filter. In the lower vacuum volumes, there is still a portion of the fluid in the inlet reservoir or in the channels before the filter; This is the underlying reason for the low DNA concentration observed with the 5 mL syringe. Therefore, if there is no remaining fluid behind the filter, the DNA concentration can be increased. To further validate this interpretation, an additional quantitative experiment was conducted to assess residual liquid retention as a function of syringe volume (Supplementary Figure S1). The results indicate that this occurred at volumes exceeding 5 mL. There is a slight rise in these volumes that is related to the speed of flow passing through the filter because the higher velocities wash the silica filter better.

Fig. 8.

DNA concentration based on the volume of the outlet syringe, which creates a vacuum for moving the fluids (using a single filter). The cell lysis chamber was maintained at 65 °C, and cell concentration was 1 × 10⁶ cells/mL. Numbers on the graphs represent p-values calculated from a two-tailed Student’s t-test.

Number of filters

The silica filters that are utilized in the extraction chip are the same filter in the extraction columns of the commercial kit. In order to optimize the extraction chip, five different states of filter are integrated in the chip and the outlet concentration is measured which is depicted in Fig. 9. Five different thicknesses of silica filter are tested in this study; as the original filter is a circle with a diameter of 7 mm and thickness of 1 mm, the value of 1 mm for the filter thickness is considered as one filter, and other values indicate the number of filters based on their thickness. According to Fig. 9, a standard size of the filter results in the maximum concentration for the extracted DNA. The lower thicknesses have two weaknesses; firstly, they do not provide enough space to attract DNA, and secondly, they are mechanically unstable and may be destroyed by fluid pressure; therefore, a part of DNA is guided to the waste chamber. On the other hand, while the number of filters is doubled or tripled, the concentration falls again. During the experiments, it was observed that all the fluids did not completely pass from the multiplied filters. This is due to the large pressure drop that compact filters create for the fluid flow. As a result, a portion of samples in each step remains before the filter and mixes with the next fluid, and reduces the final purity of the DNA.

Fig. 9.

DNA concentration as a function of the number of filters used in the extraction chip. Experiments were performed using a 20 mL syringe, with the cell-lysis chamber maintained at 65 °C and an initial cell concentration of 1 × 10⁶ cells/mL. Numbers on the graphs represent p-values from two-tailed Student’s t-tests.

Temperature and Cell Concentration

Temperature and cell concentration are two other effective factors for the final DNA concentration, which are depicted in Fig. 10a and b. Heating the sample is a way of improving the lysis reaction, therefore, higher temperatures indicate better concentration. However, the sample should stay ten minutes in an open reservoir on the heater; this is a reason why the higher temperatures cannot be applicable because of sample evaporation. Figure 10a compares the purified DNA with and without heating and a decreased temperature to show that heating can improve DNA concentration which is directly related to lysis step efficiency.

Fig. 10.

DNA concentration as a function of lysis-process parameters: (a) lysis temperature (cell concentration = 1 × 10⁶ cells/mL), and (b) cell concentration (lysis temperature = 65 °C). All experiments were conducted using a 20 mL syringe and a single filter. Numbers on the graphs indicate p-values from two-tailed Student’s t-tests.

The number of cells or cell concentration has a direct relation with the final DNA concentration. Therefore, different concentrations are utilized to find the optimum concentration. As shown in Fig. 10b, the optimum concentration was obtained at a cell concentration of 2.5 × 10⁶ cells/mL. It shows the volume of buffers used in the reference extraction kit is optimized for this number of cells, and numbers more than it shows lower DNA concentration. In the column method, increasing the cell concentration over the optimal limit will not result in improved extraction results. Column-based DNA extraction separates nucleic acids based on their physical and chemical properties using solid and stationary phases. Various materials can be used as stationary phases, including silica gel, cellulose, and ion-exchange resins. Silica column-based DNA extraction methods are saturated because the silica membrane has a limited DNA binding capacity. Therefore, cell concentration does not necessarily increase the amount of DNA extracted.

DNA quantity and quality

Genomic DNA was extracted from cells using both the optimized automated microfluidic device and a commercial silica-column kit (n = 10 per method), and concentrations were measured by UV spectrophotometry (Nanodrop). As summarized in Table 1, the device yielded 147.23 ± 29.04 ng/µL DNA (A260/A280 = 1.51 ± 0.15), while the kit yielded 163.77 ± 11.55 ng/µL (A260/A280 = 1.60 ± 0.20). This corresponds to an average extraction efficiency of 89.9% for the device relative to the kit. Two-tailed Student’s t-tests indicated no significant differences in either DNA yield (p = 0.1115) or purity (p = 0.2596) between the methods, demonstrating that our device achieves comparable performance to the commercial standard. Moreover, a test was performed to check for cross-contamination after the combined ethanol and DI water wash by processing five DI water samples as inputs. These yielded an average DNA concentration of 0.03 ± 0.34 ng/µL and an A260/A280 ratio of -0.0596 ± 0.73, confirming that any residual DNA carryover is negligible. Although the mean A260/A280 ratios for both the device and the commercial kit were slightly below the ideal value of approximately 1.8 expected for highly pure DNA, this behavior is consistent between the two methods and suggests a systematic rather than device‑specific effect. Potential contributors include residual protein or chaotropic salts remaining after washing, partial saturation of the silica membrane at the tested cell loading, and the presence of co‑extracted RNA, all of which can depress the A260/A280 ratio into the 1.5–1.6 range without necessarily compromising downstream PCR or gel analyses. Importantly, agarose gel electrophoresis showed intact, high‑molecular‑weight bands without smearing, and all samples were efficiently amplified by PCR, indicating that the modestly reduced A260/A280 values do not reflect functionally limiting contamination for typical molecular assays.

Table 1.

Comparison of DNA extraction performance between the automated microfluidic device and a commercial silica-column kit (n = 10 per method). Data are presented as mean ± SD for DNA concentration and A₍₂₆₀₎/A₍₂₈₀₎ purity ratio; p-values from two-tailed Student’s t-tests indicate no significant differences in yield or purity.

Extraction Method	DNA Concentration (ng/µL)	A260/A280 ratio
Automated microfluidic device	147.23 ± 29.04	1.51 ± 0.15
Commercial silica-column kit	163.77 ± 11.55	1.60 ± 0.20
p-value	0.1115	0.2596

Gel electrophoresis for the extracted samples revealed a single, high molecular weight DNA band which is illustrated in Fig. 11. The input samples were the extracted DNA with different parameters that were investigated in the previous part, and none of the DNA samples showed significant smearing, which indicates degradation of the sample.

Fig. 11.

Electrophoresis of genomic DNA on agarose gels (1%). Lane A shows a 50 bp DNA ladder with size markers. Lanes B through H represent samples 1 through 7 which are randomly selected from device output. Table S1 in Supplemntary Information provides the conditions for each sample.

To validate the extraction process further, the extracted DNA was amplified by PCR and compared to a No Template Control (NTC) sample which is a sample without DNA. As illustrated in Fig. 12, amplification plots for different extracted samples are displayed. Sample 1 to sample 7 are the extracted DNA using the device with different conditions in their filter, temperature and vacuum condition; ΔRn represents the baseline-subtracted, passive-reference-normalized reporter fluorescence signal recorded in each cycle, and thus directly reflects the accumulation of PCR product over the course of the run. All the samples start the amplification between cycle 15 and cycle 20 and finally get stable after cycle 40.

Fig. 12.

PCR validation. Amplification plots for GAPDH (fluorescence vs. cycles) and comparison of samples with NTC. Samples 1 through 7 which are randomly selected from device output. Table S1 in Supplemntary Information provides the conditions for each sample.

Having validated the quality and amplifiability of DNA extracted by our device, we compared its performance characteristics with two widely used DNA extraction techniques, magnetic bead-based and phenol-chloroform extraction, to assess practical advantages. These comparisons are summarized in Table 2.

Table 2.

Comparative analysis of silica column-based (our method), magnetic bead-based, and phenol-chloroform DNA extraction methods in terms of DNA purity, cost, and extraction time.

Method	DNA Purity		Cost		Extraction Time
	Advantage	Limitation	Advantage	Limitation	Advantage	Limitation
Silica column-based (our method)	High (~ 1.5–1.6 A260/A280), comparable to kits	Slightly lower than ideal in some trials (1.51 ± 0.15)	No need for centrifuges or magnetic modules (low system cost)	Slightly higher per-chip material cost (custom microfabrication)	Semi-automated; ~15–20 min total process time	Slightly longer setup time for control system (one-time)
Magnetic bead-based	High (1.6–1.8), good reproducibility	Bead handling variability; requires magnetic calibration	Kits widely available; good for high-throughput labs	Requires integration of magnetic controls and precise timing	Moderate (~ 20–30 min depending on automation)	Extra steps for bead separation and wash cycles
Phenol-chloroform extraction	Very high (A260/A280 ~ 1.8-2.0), low protein content	Risk of contamination from incomplete phase separation	Low reagent cost per test	High labor, long protocol, toxic waste handling	Long (~ 45–60 min); labor-intensive	Not compatible with microfluidics or automation

Conclusion

In our previous studies, we developed microfluidic devices for biomedical applications ^43–46. In this study, a novel DNA extraction device was designed and fabricated that consisted of three microfluidic chips including mixing, lysis and extraction chips which were coupled by electric valves. To eliminate the leakage between valves and chips, a 3D-printed part was designed and fabricated with special protrusions to fill the empty spaces. The fluid used in the device moved using external syringe pumps, and the values were switched by a controller to guide the fluids in the right path in the desired step. Different factors in the device were studied to optimize the DNA concentration, and as a result, experiments showed that one silica filter in the extraction chip yields higher DNA concentration. It also resulted that if the fluid passed faster through the filter, DNA concentration increased; it was shown by 25 mL of syringe volume, which produced a stronger vacuum to move the fluid. Although cell concentration and lysis temperature affect the lysis process, their optimum values are introduced as 2.5 × 10⁶ cells/mL and 65 °C, respectively. Finally, the optimum DNA concentration was reported 147.23 ± 29.04 ng/µL with an A260/A280 ratio of 1.51 ± 0.15 which had an efficiency of ~ 90% in comparison to the standard column DNA extraction method. Moreover, to check the quality of extracted DNA, PCR and Gel electrophoresis tests showed the appropriate quality for them. While the present study demonstrates reliable semi-automated DNA extraction using a modular microfluidic architecture, the current implementation is limited by the evaluation of relatively high cell concentrations and the absence of integrated downstream amplification, which constrains assessment of performance for low-abundance targets and fully autonomous operation. Future work will include evaluation of device performance at lower cell concentrations (e.g., 10⁵ and 10⁴ cells/mL) and qPCR-based validation of low-abundance DNA recovery. In addition, in our ongoing study (Topaheidari et al., manuscript in progress), we are advancing the platform toward programmable operation by automating valve control and integrating a downstream PCR unit following DNA purification. As part of this effort, the temperature control system has been further optimized to reduce overshoot and achieve faster stabilization at the target lysis temperature. Together, these developments are expected to substantially reduce user intervention and move the system closer to a fully autonomous workflow.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

R.D. ran the experiments, analyzed the data and wrote the manuscript. A. S. supervised the project, analyzed the data and wrote the manuscript. M. E. analyzed the data and wrote the manuscript. M.T. ran the experiments and wrote the manuscript. M.S.M. analyzed the data and wrote the manuscript. E. P. analyzed the data and wrote the manuscript.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due to the requirement of confidentiality but are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.