Simple and rapid microfluidic device for processing raw human semen and separation of motile sperm based on their intrinsic behavior and morphology

Abstract

Sperm separation is pivotal in Assisted Reproductive Technology to address male infertility issues such as low sperm concentration or impaired motility. Traditional microfluidic devices, despite their fair success, often fall short due to complex fabrication and use, or lack of direct sample extraction methods. Our research introduces a microfluidic device designed to enhance sperm quality, focusing on improved motility and morphology, while remaining simple to fabricate and use in infertility clinics. The device features four chambers interconnected by channels, engineered to establish low shear rate limits, facilitating high-quality sperm separation. Utilizing the principle of rheotaxis, where superior sperm swim against the flow, our system concentrates top-quality sperm in designated chambers. A key feature of the device is its ability to use raw semen, eliminating the need for pre-washing. Clinical trials with both washed human sperm samples and raw semen demonstrate the device’s efficacy, achieving up to 100% sperm isolation and morphological improvements in under 5 min. This approach overcomes the throughput constraints of traditional microfluidic methods, leveraging rheotaxis, near-boundary swimming, and parallelization to increase selected sperm concentration. This advancement has the potential to substantially improve the effectiveness of fertility treatments, offering new hope to individuals facing conception challenges.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-09884-1.

Introduction

Male infertility, which constitutes nearly half of all global infertility cases, underscores the pressing need for advanced strategies in assisted reproduction, where natural barriers to conception can be circumvented^1–3. Within this landscape, sperm sorting emerges as a crucial component, and the advent of microfluidic technology presents a less invasive alternative to traditional methods^4,5. Over the past few decades, the field of ART has witnessed significant advancements aimed at addressing infertility⁴. Conventional ART methods, such as intrauterine injection (IUI), intracytoplasmic sperm injection (ICSI), and in vitro fertilization (IVF)⁶, have evolved to circumvent the natural barriers hindering the selection of high-quality sperm for fertilization within the female reproductive tract⁶. Consequently, there is a growing demand for the isolation of sperm exhibiting optimal motility and sufficient concentration, as these factors are crucial for enhancing the success rates of insemination and promoting embryo health^7–10.

Currently, the predominant sperm separation techniques in clinical practice include the swim-up method (SU) and density gradient centrifugation (DGC)^11,12. Despite their common usage, both methods suffer from drawbacks, notably being time-consuming and labour-intensive, and reliant on operator skill^11,12. Moreover, exposure to centrifugal force during centrifugation leads to DNA fragmentation in sperm, especially in infertile men, affecting genetic integrity and increasing apoptosis rates in infertile sperms^13,14. Additionally, these methods often yield low recovery rates^11,12. These limitations have prompted researchers to employ microfluidic technology in the field of sperm separation in recent years to overcome these drawbacks.

In recent years, microfluidic technology has been increasingly utilized across various fields due to its precision, speed, and high level of control. This includes its application in the investigation of gametes and their function within ART^15–18. Particularly in the realm of sample preparation and sperm separation, microfluidic techniques have not only facilitated but also enhanced ART methodologies¹⁹. Moreover, microfluidic technology empowers researchers with the ability to create microenvironments that closely mimic in vivo conditions^20,21.

Over recent years, there has been a remarkable upsurge in initiatives leveraging microfluidic technology to revolutionize infertility treatments, seeking to overcome the limitations of conventional ART methods^22,23. Pioneering studies have delved into unravelling the intricate dynamics of sperm movement and migration within microfluidic systems, as evidenced by^24,25. Notably, sperm exhibit random motility in microenvironments and navigate freely within microchambers¹⁶, prompting extensive research into understanding how they respond to external stimuli within these scaled-down environments²⁶.

Numerous microfluidic approaches have been explored, with near-boundary swimming emerging as noteworthy^27,28. This method capitalizes on sperm’s tendency to navigate along walls and traverse streamlines^29,30. It eliminates centrifugation, preserving sperm DNA integrity^31,32. By harnessing sperm behaviour within microfluidic environments, these advancements promise to enhance infertility treatments’ efficiency and safety³⁰. Despite ongoing efforts, the development of a reliable microfluidic platform that fully meets expectations has not yet been achieved. This is due to the complex fabrication process, low morphological accuracy, suboptimal recovery rates, and the time-consuming nature of the technology.

Rheotaxis, a vital aspect of sperm behaviour, guides their movement within the female reproductive tract towards the oocyte^33,34. It involves rotational torque and drag from external flows, reorienting sperm parallel to the flow and propelling them upstream^35,36 (Fig. 1C). This behaviour depends on flow strength and sperm motility³³. At low flow rates, sperm generate enough force to swim against the flow, but the flow must be strong enough to reorient them^21,36. High flow rates risk washing out all cells, necessitating shear rate and flow velocity optimization. Studies indicate optimal shear rates of 2 s^− 1 to 5 s^− 1 and desired flow velocities of 22 μm/s to 102 μm/s³⁷. Attaining these parameters is crucial for maximizing rheotaxis’s potential in guiding sperm within the reproductive tract.

Fig. 1.

The proposed microfluidic device designed to facilitate rheotaxis-based sperm separation, consisting of four interconnected chambers. (A) The schematic diagram elucidates the underlying principles governing the sorting process. Upon entering the chamber, the flow velocity diminishes to a level conducive to sperm demonstrating rheotactic behaviour. Consequently, sperms of higher quality tend to swim upstream, navigating toward isolation zones characterized by lower shear rates. (B) A photograph of the fabricated device showcases its physical manifestation. (C) The diagram depicts the rheotaxis-based separation process in action; whereby motile sperm are reoriented against the flow to segregate them into isolation zones. This phenomenon relies on the sperm’s propulsive force (F_P) surpassing the viscous-induced drag (F_D) to enable effective rheotaxis. (D) Dimensions of the device: a = 20 mm, b = 7 mm, c = 3 mm, d = 8.8 mm, e = 250 μm, channel height = 50 μm.

Recent rheotaxis-based microfluidic platforms exploit controlled flow to guide motile sperm, yet they routinely underperform in yield, scalability and sample compatibility. Zaferani et al.¹⁰ fabricated a coral-structured channel for rheotaxis-driven sorting; although the geometry was novel, the system produced a low concentration of motile sperm and required a multistep extraction process. Sarbandi et al.³³ developed a three-zone device with paired micropockets, but it depended on prewashed samples and could not process raw semen. Nagata et al. employed a diffuser-shaped channel that improved selectivity, yet suffered from low overall yield, complex fabrication and validation only on bovine samples³⁸. Efforts by Heydari et al. to parallelize rheotactic channels shortened processing time but achieved under 2% throughput³⁹, while Bouloorchi’s multichannel layout enriched motility effectively at the cost of a 30-min runtime and limited scalability⁴⁰.

These approaches share three key drawbacks: their fabrication and operation remain complex, they rely on prewashed or nonhuman samples, and their extraction mechanisms are cumbersome—factors that together limit recovery, prolong processing and impede clinical adoption. Additionally, standard centrifugation protocols can compromise sperm DNA integrity⁴¹. To overcome these limitations, a streamlined microfluidic device is required—one that can process raw human semen, achieve high recovery of progressively motile sperm, and pair simple fabrication with an intuitive extraction method.

In this study, a microfluidic device for sperm separation was introduced that leverages rheotaxis, crossing streamlines behaviour, and near-boundary swimming behaviour while having the ability to analyse raw human samples (Fig. 1A). To enhance efficiency and reduce operation time, we also implemented a parallelization approach in the proposed device which tries to improve other’s limitations. The device comprises four chambers and three channels interconnected to distribute flow width and decrease velocity within the chambers, conducive to triggering rheotactic behaviour (Fig. 1A,B,D). At connections, flow widens, and velocity decreases from pressure drop, promoting rheotactic behaviour. Specialized isolation zones form at chamber sides with lower shear rates, aiding sperm separation. This design enhances precision in microfluidic sperm separation processes.

To conduct this research, various concepts and designs were subjected to numerical simulation using the FEM method to achieve a well-structured and optimized device. The inlet flow rate, a crucial operational parameter, was optimized through simulations to achieve the best results. Following device fabrication, its performance was experimentally optimized using human washed sperm samples at six different flow rates. Flow rates of 40 nl/s and 50 nl/s proved most effective, with a subsequent increase to 500 nl/s for separated sperm recovery. In the final step, the device’s performance was tested with raw human sperm samples. The results indicated enhancements in motility (up to 100%) and morphology (up to 56%) under the selected flow rates. This empirical validation underscores the device’s efficacy in improving sperm quality for assisted reproduction, affirming its potential in clinical settings.

Methodology

Computer simulation

To investigate the functionality of the device, it was designed by CATIA V5-6R2018, ensuring precise dimensions. Subsequently, the design was imported into COMSOL Multiphysics (version 5.3) for comprehensive modelling. In this simulation, we employed the Navier-Stokes (Eq. 1) and Conservation of Mass (Eq. 2) equations to solve for three-dimensional (3D) flow:

1

2

where V is velocity, ρ is density, µ is dynamic viscosity, and P is pressure.

We adopted the same laminar flow model with a no-slip boundary condition and Newtonian fluid assumption as in our previous study, reflecting the lower viscosity of the washed sperm sample and ensuring simulation accuracy. To characterize system performance, we evaluated flow rates from 1 to 10 nL/s in 1 nL/s increments, then from 20 to 100 nL/s in 10 nL/s steps. This comprehensive range afforded a detailed assessment of device behaviour across varying flow conditions.

Device fabrication

For the fabrication of the proposed device, the conventional soft lithography method was employed which is a well-established technique in microfluidics⁴¹. The device comprises four chambers, each boasting dimensions of 3 mm in width and 8.7 mm in length, with circular ends measuring 1.5 mm in radius. Additionally, there are three connecting channels, each 250 μm wide and 50 μm height, which link these chambers together (Fig. 1D). To initiate the fabrication process, the SU-8 negative photoresist (SIGMA USA) was applied onto a silicon wafer using spin coating. Subsequently, a designed mask was utilized to shield unnecessary areas of the photoresist from UV light exposure during curing, thereby creating the desired channel patterns. This structure served as a mould for casting polydimethylsiloxane (PDMS) to form the device.

The PDMS was prepared by mixing Sylgard 184 (Dow Corning Co., USA) with its curing agent at a ratio of 10:1. The resulting mixture was poured into the mould and cured at 75 °C for 1 h. Following the curing step, the PDMS channel structures were immersed in acetone in an ultrasonic bath for 5 min to remove air bubbles and residual uncured PDMS, followed by rinsing with deionized water and immersion in a 10% H2SO4 solution for 30 min to enhance surface hydrophilicity. Subsequently, inlet and outlet holes were punched, and the PDMS and glass substrates were exposed to oxygen plasma for 30 s and 300 s, respectively. This treatment facilitated bonding between the two substrates, resulting in the formation of the complete device geometry.

Sample preparation

In this study, we used two sample types. For calibration and optimization, we prepared a washed-sperm sample by mixing 1 mL of semen with 4 mL of HTF-HEPES medium and centrifuging at 2000 rpm for 3 min. To support sperm viability, the medium was supplemented with human serum albumin (50 µL HSA per 10 mL HTF-HEPES) and pre-warmed to 37 °C to avoid thermal shock. Because HTF-HEPES already contains essential electrolytes (Na⁺, K⁺, Ca²⁺, Mg²⁺, Cl⁻), no further salts were required.

For the raw semen sample, the preparation process differed in that no centrifugation was required which as explained in the previous chapter could affect the sperm DNA integrity; the sample was simply diluted. This method ensured the viability and functionality of the sperm for further analysis and experimentation⁴².

Experimental procedure

Initially, the device undergoes a thorough washing process with deionized distilled water to ensure cleanliness and remove any potential contaminants. To achieve this, deionized distilled water is carefully filled into a syringe and injected into the device through silicone tubes connected to both the inlet and outlet. Subsequently, the device is emptied of the water from its interior.

Following the washing process, the device is placed in an oven and subjected to gentle heating for 15 to 20 min at 75 ͦ C. This step is crucial for evaporating any residual water that may linger within the device, as even small amounts could compromise the viability of the sperm samples. After the heating process, the device is allowed to cool down to room temperature for approximately 5 min, ensuring optimal conditions for subsequent experimentation.

Once the device is primed for operation, the sperm sample is carefully injected through inlet tubing to avoid bubble formation, which can disrupt flow stability. Immediately afterward, EDTA buffer is introduced at flow rates of 20, 30, 40, 50, 60 and 70 nL/s (in 10 nL/s increments) via a Syringe Pump (Smart Syringe Pump SSP101 behZsaz, Iran), ensuring precise control over perfusion. Throughout the experiment, an Olympus BX51 microscope—equipped with 4× and 10× objectives—monitors the flow, while an Olympus DP27 camera records sperm trajectories (see SI Videos 1 and 4). Recorded videos are exported to Adobe Premiere for manual path tracking, then imported into ImageJ to map trajectories and calculate velocity parameters. This workflow enables high-resolution observation and quantitative analysis of sperm migration (Fig. 2A,B).

Fig. 2.

the operational procedures of the microfluidic device: (A) The injection of the sperm sample into the device initiates the process. (B) Subsequently, a buffer medium is gently injected into the device utilizing a syringe pump, ensuring a predetermined flow rate is maintained for a specified duration. This step facilitates the complete washing and purification of the sperm sample within the device. Following the washing phase, the flow rate is increased to 500 nl/s to wash the sorted sperms downstream. (C) Ultimately, the sorted sperms can be conveniently collected from the outlet for further analysis or utilization. (D) Additionally, the setup of the separation process is depicted, providing insight into the configuration and arrangement of the components involved in the operation of the microfluidic device.

The operational duration varies for each flow rate, as the device must continue operating until the outlet accumulates a volume of 10 µL, ensuring thorough washing of the sample. Subsequently, the outlet undergoes cleaning, followed by an increase in flow rate to 500 nl/s to flush out all isolated sperm downstream, facilitating their collection from the outlet (Fig. 2C). This sequential procedure is conducted for six different flow rates using prewashed human sperm samples. Additionally, it is repeated for two optimized flow rates (40nl/s and 50nl/s) using raw human sperm samples. In the former case, only sperm motility is assessed, whereas in the latter, both morphology and motility are evaluated.

To accomplish this, the collected samples from the outlet are analysed using Computer-Aided Sperm Analysis (CASA) software, which provides insights into sperm characteristics and behaviour. This assessment ensures a understanding of the effects of varying flow rates on sperm motility and morphology, contributing to the refinement of ART.

Results and discussion

Computer simulation

To optimize the efficiency of the device for sperm separation, determination of the inlet flow rate is paramount. This parameter directly influences flow velocity and shear rate, which, in turn, create the requisite conditions for triggering rheotactic behaviour in sperm. To address this crucial aspect, we employed the Finite Element Method (FEM) within the COMSOL platform. Through analysis, we established that the acceptable range for setting the inlet flow rate falls between 20nl/s and 70nl/s. This approach ensures that the device operates within optimal parameters, maximizing its efficacy in facilitating sperm separation for ART.

In the proposed device, the channels are notably narrower compared to the chambers, resulting in a varied flow velocity distribution. Near the walls, where the no-slip boundary condition is applied, the flow velocity diminishes to nearly zero, gradually increasing towards the centre of the channel, reaching a maximum of 2.13 mm/s (Fig. 3B). This velocity gradient washes out all sperm cells within the channels. Moreover, within the isolation zones, the average flow velocities are determined to be 15.2 μm/s, 21.8 μm/s, 35.6 μm/s, 50.2 μm/s, 67.4 μm/s, and 79.8 μm/s for flow rates of 20 nl/s, 30 nl/s, 40 nl/s, 50 nl/s, 60 nl/s, and 70 nl/s, respectively (Fig. 3A,B). Correspondingly, the average shear rates within these zones are calculated to be 0.95 s⁻¹, 2.58 s⁻¹, 3.56 s⁻¹, 4.20 s⁻¹, 5.61 s⁻¹, and 6.48 s⁻¹ for the respective flow rates (Fig. 3C,D).

Fig. 3.

Details of fluid flow within the proposed microfluidic device by FEM based simulation, showcasing various parameters crucial for its performance. (A) The average flow velocity within isolation zones across a range of flow rates from 20nl/s to 70nl/s with 10nl/s increments. (B) Contour plots illustrating the spatial distribution of flow velocity within the microfluidic device at a flow rate of 40nl/s were obtained through computational simulations using COMSOL 5.3 (C) The average shear rate within isolation zones across a range of flow rates from 20nl/s to 70nl/s with 10nl/s increments. (D) Contour plots of the shear rate distribution within the microfluidic device at a flow rate of 40nl/s, derived hrough simulation with COMSOL 5.3. (E) The operational time across the aforementioned flow rate range, sheds light on the device’s performance variation with different flow rates. (F) Flow streamlines at a flow rate of 40nl/s, offering a detailed depiction of fluid flow behaviour within the microfluidic device simulated via COMSOL 5.3.

Based on the simulation results, it is evident that flow rates of 40nl/s and 50nl/s exhibit superior performance. These flow rates offer optimal parameters for triggering rheotactic behaviour in sperm, as the combination of flow velocity and shear rate within the isolation zones is most favourable compared to other flow rates. This is crucial because activation of rheotactic behaviour encourages sperm migration towards these zones, enhancing the efficiency of sperm separation. Additionally, the operation time of the separation process, including both the separation and retrieval of isolated sperm, was evaluated through simulation. The operation time ranges from 520 s for 20 nl/s to 163 s for 70 nl/s (Fig. 3E). Minimizing the operation time is essential as it directly impacts the quality of isolated sperm, emphasizing the need for efficient separation processes^33,39,40.

Furthermore, the examination of streamlines for each flow rate reveals complete flow development into the chambers (Fig. 3F). This observation is corroborated by concentration contours at various time intervals, which illustrate that every region within the chambers is uniformly exposed to the flow (Fig. 4).

Fig. 4.

Dynamic visualization of the concentration distribution of the buffer medium within a microfluidic device over various time steps (A to I) spanning 300 s. These simulations, conducted using COMSOL 5.3, capture the evolution of concentration within the device, illustrating how every spot within the microfluidic system experiences exposure to the flow rate.

Experimental tests

Experimental optimization of the device’s performance was conducted using prewashed human sperm samples across various flow rates. This experimentation highlighted that flow rates of 40nl/s and 50nl/s yield superior outcomes. Subsequently, the device was further tested using raw human sperm samples under these two specific flow rates.

After each test, the collected sample containing the isolated sperm was analyzed. For prewashed samples, the evaluation focused on sperm motility, while for raw semen samples, both motility and morphology were assessed. The results were compared to the original sample using Computer-Aided Sperm Analysis (CASA) (refer to Supplementary Information). To further interpret the findings, a one-way ANOVA (Single Factor) was conducted to evaluate differences in concentration, motility, and progressive motility of the washed sperm samples post-separation. The statistical results were compared to highlight the significance of observed differences. This approach enables a robust assessment of the device’s effectiveness in sperm separation and provides insights into its performance across varying flow rates and sample types.

Separation mechanism

The device underwent an optimization process, beginning with testing using prewashed samples, followed by evaluation using raw semen samples. Throughout both phases, human sperm samples were employed. The operational mechanism of the device involves initially loading it with a sperm sample, which is subsequently washed using a buffer medium. During this washing process, less motile and non-motile sperm cells are removed downstream, while those with sufficient motility undergo reorientation and swim against the flow, driven by rheotaxis. These motile sperm cells accumulate within the device, concentrating within isolation zones. To retrieve the isolated sperm, the flow rate within the device is increased to ensure the complete evacuation of all sperm cells.

To demonstrate the rheotactic behaviour of sperm cells, the device was tested using a raw semen sample at a controlled flow rate of 50 nL/s. The experiment was recorded at two different magnifications to capture the sperm dynamics in detail, as shown in Supplementary Information (SI Videos 1–4). In the regions with high flow velocity—particularly within the channels and at the junction between the chamber and the channel—particles, including non-motile sperm and debris, were quickly carried away by the flow and could not maintain their position. In contrast, a different behaviour was observed inside the chambers. Here, progressively motile sperm were able to resist the flow, swim upstream, and cross streamlines toward areas of lower shear stress. These low-shear regions, located on both sides of the chambers, acted as isolation zones where motile sperm could accumulate. This behaviour highlights the ability of motile sperm to respond to fluid dynamics and segregate themselves from the non-motile population under controlled flow conditions.

Washed sperm sample

To evaluate the performance of the device, tests were conducted using six flow rates determined through prior simulations: 20, 30, 40, 50, 60, and 70 nL/s. These rates were selected to represent a range of operational conditions. Prewashed human sperm samples were used to ensure consistency and control during experiments. After the sorting process at each flow rate, the flow was immediately increased to 500 nL/s to flush out all isolated sperm from the device. The collected samples were then analysed using Computer-Assisted Sperm Analysis (CASA) to assess key parameters such as motility and concentration. This analysis provided insights into the device’s effectiveness at each flow rate and helped determine optimal operating conditions.

The initial analysis of sperm concentrations showed an average of 19.64 M/ml across all samples, with slightly lower values observed at flow rates of 40 nL/s and 50 nL/s, measuring 18.78 M/ml and 15.75 M/ml, respectively. After separation, the concentrations measured in the collected samples were 2.2 M/ml, 1.94 M/ml, 5.37 M/ml, 4.00 M/ml, 2.43 M/ml, and 2.12 M/ml at flow rates of 20, 30, 40, 50, 60, and 70 nL/s, respectively. The results indicate that flow rates of 40 nL/s and 50 nL/s resulted in higher post-separation sperm concentrations compared to other conditions. These findings are presented in the concentration distribution chart (Fig. 5A).

Fig. 5.

The analysis of sorted sperm quality at various flow rates, ranging from 20nl/s to 70nl/s with 10nl/s increments for prewashed samples. (A) The average concentration of sperm in the separated sperm sample. (B) The percentages of total and progressive motility.

The analysis of motility rates showed improvements across all tested flow rates, with the most favourable results observed at 40 nL/s and 50 nL/s. At these flow rates, total motility reached up to 100%, with average values of 96.57% and 97.9%, respectively, after isolation. Similarly, progressive motility was highest at these conditions, with 74.63% recorded at 40 nL/s and 71% at 50 nL/s (Fig. 5B).

For the initial control samples prior to separation, total motility averaged 65.83% at 40 nL/s and 64.8% at 50 nL/s, while progressive motility averaged 51.73% and 50.73%, respectively. Comparing these values with post-separation results highlights the improvement in both total and progressive motility achieved through the isolation process. The data confirm that optimized flow rates of 40 nL/s and 50 nL/s are effective in enhancing sperm quality, demonstrating the importance of flow rate selection in improving the outcome of sperm isolation.

To further assess the post-separation outcomes, a one-way ANOVA was performed on sperm concentration, total motility, and progressive motility across different flow rates. Each experimental group consisted of five independent samples (n = 5). This analysis evaluated whether the observed differences across flow rates were statistically significant. The resulting p-values were 2.8 × 10⁻²¹ for concentration, 3.71 × 10^− 8 for total motility, and 2.7 × 10^− 9 for progressive motility (see Supplementary Information), indicating significant differences among the groups. The assumptions of normality and homogeneity of variances were verified prior to ANOVA to ensure statistical validity.

Subsequently, Tukey’s Honestly Significant Difference (HSD) post hoc test was conducted to identify which specific flow rates differed significantly. Based on a within-group degrees of freedom of 24 and six experimental groups, the critical value q = 3.9 was obtained from the Studentized Range Statistic table. The calculated standard errors (SE) were 0.08462 for sperm concentration, 1.155 for total motility, and 1.253 for progressive motility. Corresponding HSD values were 0.33, 4.5, and 4.89, respectively.

These results demonstrate that flow rates of 40 nL/s and 50 nL/s produced statistically significant improvements in both sperm concentration and progressive motility. Furthermore, 40 nL/s, 50 nL/s, and 60 nL/s showed significantly enhanced total motility. Overall, 40 nL/s and 50 nL/s emerged as the most effective flow rates for post-separation sperm quality enhancement.

These findings are further supported by simulation data, particularly the average shear rate and velocity within the chambers (Fig. 3A,C). Simulations show that flow rates of 40 nL/s and 50 nL/s correspond to average shear rates of 3.56 s⁻¹ and 4.2 s⁻¹, respectively—well within the 2–5 s⁻¹ range known to promote rheotactic behaviour. Apart from 20 nL/s, all tested flow rates also showed average velocities within the acceptable range. Although the shear rate at 30 nL/s meets the target range, its average velocity is close to the lower limit, suggesting potential inefficiency. This is consistent with experimental results, which indicate weaker performance at 30 nL/s. Together, simulation and experimental data confirm that the device performs most effectively at flow rates between 40 nL/s and 50 nL/s, highlighting the importance of flow rate optimization for achieving reliable sperm separation.

Raw semen sample

Despite the enhancements observed in total motility, progressive motility, as well as the rates of motile and progressively motile sperm retrieval, it remains imperative to assess the performance of the device using raw semen samples. This necessity arises from the fundamental objective of employing microfluidic technology in ART, which aims to provide an alternative method for sperm separation that can mitigate the need for centrifugation. Centrifugation, while effective, is associated with detrimental effects on sperm DNA integrity^13,14. To address this concern, diluted raw human sperm samples were employed for evaluation purposes, enabling a comprehensive assessment of the device’s performance under conditions that mimic real-world scenarios. Specifically, the device’s functionality was scrutinized using optimized flow rates of 40nl/s and 50nl/s, which had previously demonstrated promising outcomes. By subjecting the device to testing with raw semen samples, this evaluation not only ensures a more understanding of its efficacy but also validates its potential as a viable alternative to traditional centrifugation methods in ART procedures.

In the evaluation using diluted raw semen samples, the initial average sperm concentration was approximately 80 M/ml. However, following the sperm sorting process, this concentration declined to 4.44 M/ml and 4.63 M/ml for flow rates of 40nl/s and 50nl/s, respectively (Fig. 6A). This reduction can be attributed to the inherent limitations of the device’s capacity. As the chambers fill with sperm cells at isolation zones, those entering subsequently are compelled to occupy regions exposed to stronger flows or are displaced by incoming sperm cells along the chamber sides, inevitably being washed downstream.

Fig. 6.

depicts diagrams showcasing the analysis of sorted sperm quality at various flow rates, ranging from 20nl/s to 70nl/s with increments of 10nl/s for raw semen samples. (A) The average concentration of sperm in the separated sperm sample. (B) The percentages of total and progressive motility. (C) The determined sperm velocity parameters VAP, VCL, and VSL. (D) Morphology quality of isolated sperms.

Advancements in motility rates were observed during the evaluation process. Specifically, for the 40nl/s flow rate, total motility soared to 100%, with an average of 96.33%, while progressive motility reached levels of up to 86.4%, with an average of 78.06%. Similarly, for the 50nl/s flow rate, total motility peaked at 100%, averaging 97.93%, while progressive motility attained levels of up to 73%, with an average of 69.73%. Notably, before separation, the average total motility and progressive motility for the samples stood at 55.33% and 28.4%, respectively, underscoring a twofold enhancement in the quality of sperm contained within the samples post-separation (Fig. 6B). These findings underscore the device’s efficacy in improving sperm motility, paving the way for enhanced fertility outcomes in ART procedures.

In addition, 20 isolated sperm were selected at random, and three velocity parameters were measured, namely average path velocity (VAP), straight-line velocity (VSL) and curvilinear velocity (VCL) (Fig. 6C). In this study a mean VAP of 46 μm/s was obtained, peaking at 65.5 μm/s, in agreement with established fertility standards. Mean VSL and VCL values of 39.4 μm/s and 47.5 μm/s were recorded, reaching maxima of 56.3 μm/s and 66.9 μm/s respectively. Because measurements were performed under counter-flow conditions, the recorded velocities may have been slightly reduced, yet they remain consistent with values reported in previous studies^21,26,37.

The effect of the device on sperm morphology was evaluated at the two optimal flow rates, 40 nL/s and 50 nL/s. Morphological assessment was performed using both manual analysis by an experienced embryologist under light microscopy and a computer-assisted sperm analysis (CASA) system, following Kruger’s strict criteria⁴³. Initial CASA analysis of raw human semen indicated an average normal morphology of 17%. Following separation, normal morphology increased to 46.3% at 40 nL/s and 56.3% at 50 nL/s (Fig. 6D). These results were independently confirmed by a clinical embryologist from the IVF unit of Gandhi Hotel Hospital (Fig. 7).

Fig. 7.

The microscopical images of sperms (A) before separation procedure and (B) after separation.

For a more detailed evaluation, morphological defects were categorized based on abnormalities in head shape, acrosome integrity, neck structure, and tail formation. Ten spermatozoa were randomly selected both before and after separation. The majority of defects were observed in the head region (see Supplementary Information). The marked improvement in normal morphology after separation highlights the device’s capability to effectively enrich sperm quality.

Conclusion

In conclusion, male infertility remains a major challenge, accounting for nearly half of all infertility cases worldwide and necessitating improved solutions in assisted reproduction. To address this, we developed a novel microfluidic device capable of processing both pre-washed and raw human semen samples to enhance sperm quality, particularly in terms of motility and morphology. Utilizing microfluidic principles and sperm rheotaxis, the device isolates high-quality sperm within interconnected chambers, offering a rapid, efficient, and less invasive alternative to traditional methods.

Clinical validation with both washed and raw samples demonstrated effective performance, achieving up to 100% sperm isolation and significant improvements in morphology—all within 5 min. Furthermore, the device’s simple and scalable design allows for increased throughput by expanding chamber numbers or dimensions. However, as the current setup relies on an external syringe pump to generate flow, future work is needed to develop a passive version that would be more practical for clinical application.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

M.M.T. Conceptualization, methodology, Fabrication, investigation, data curation, writing—original draft, writing—review, and editing, visualization, project administration. M.M.Z. Conceptualization, methodology, resources, project administration—review and editing. M.M. Funding acquisition, methodology, providing the experiment and laboratory facilities, writing—review, and editing. Z.A. Methodology, writing—review, and editing. I.R.S. Conceptualization, Mentoring the methodology, resources and editing.

Funding

This project was funded through resources secured by the first author. The authors also acknowledge the third author’s contribution in supporting the preparation of the clinical experimental facilities through her independent funding.

Data availability

All results and analyzed data generated in this study are presented within the main manuscript and the supplementary information. Portions of the raw datasets utilized and analyzed during the course of this research are provided in the supplementary files for reference. The complete datasets are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

In this research, the human semen samples were provided by IVF vision of Gandhi Hotel Hospital. All participants signed informed consent, and the study was approved by the Research Ethics committees of The Institute of Pharmaceutical Sciences-Tehran University of Medical Sciences (IR.TUMS.TIPS.REC.1402.302). Also, the authors confirm that all experiments were performed following relevant guidelines and regulations.