Analysis of Hematocrit-Plasma Separation in a Trifurcated Microchannel by a Diffusive Flux Model

This study employs 3D numerical simulations using a diffusive flux model to analyze a passive trifurcated microchannel for separating platelet-enriched plasma and red blood cells, revealing that smaller channel widths, extended inlets, and lower hematocrit concentrations significantly enhance separation efficiency while performance remains less sensitive to flow rates and geometric angles.

The Gist

The Big Idea: A "Traffic Cop" for Blood Cells

Imagine your blood as a busy highway. On this highway, you have two main types of vehicles:

1. Red Blood Cells (RBCs): These are the heavy trucks. They make up most of the traffic (about 40-45% of the volume) and carry oxygen.
2. Platelets: These are the tiny, nimble motorcycles. They are crucial for clotting wounds but are very few in number (less than 2%).

The Problem: When doctors need to treat diseases like anemia or blood cancer, they often need the "motorcycles" (platelets) without the "trucks" (RBCs). Traditionally, they separate these using giant spinning machines (centrifuges). But this is slow, expensive, and the high-speed spinning can sometimes damage the delicate blood cells, like shaking a fragile vase until it cracks.

The Solution: The researchers in this paper designed a passive micro-chip. Think of it as a tiny, intricate maze carved into a piece of plastic, only a fraction of the width of a human hair. Blood flows through this maze, and the physics of the flow naturally sorts the trucks from the motorcycles without any spinning or electricity.

---

How the "Maze" Works: The Physics of Sorting

The device is shaped like a trifurcated channel. Imagine a single road that suddenly splits into three: a main road in the middle and two side roads branching off at an angle.

Here is the magic trick that happens inside:

1. The "Wall Effect" (Shear-Induced Migration):
   When blood flows through a narrow tube, the cells near the walls move slower than the cells in the center. The "heavy trucks" (RBCs) hate the slow, crowded edges. They naturally drift toward the fast, open center of the road.

   • Analogy: Imagine a crowded dance floor. The big, bulky dancers (RBCs) get pushed toward the middle of the room because the edges are too cramped. The small, agile dancers (platelets) get squeezed toward the walls.

2. The Cell-Free Layer:
   Because the RBCs migrate to the center, a clear "buffer zone" forms near the walls. This zone is almost empty of trucks but full of motorcycles. This is called the Cell-Free Layer.

3. The Split:
   The device has side arms (the separator arms) that tap into these wall zones.

   • The Side Roads: They suck off the clear fluid from the walls, which is now rich in platelets (the good stuff for treatment).
   • The Main Road: The trucks (RBCs) stay in the middle and continue down the main path, leaving the side roads.

---

What the Researchers Tested (The "Tuning" Phase)

The team used powerful computer simulations (like a virtual wind tunnel for blood) to figure out how to build the perfect maze. They tested several variables:

• The Width of the Road (Channel Width):

  • Finding: Narrower roads work better.
  • Analogy: If the road is too wide, the trucks have too much space to wander and might accidentally drift into the side exits. If the road is narrow, the trucks are forced tightly into the center, making it easier to scoop up the motorcycles from the edges.

• The Angle of the Split (Branching Angle):

  • Finding: The angle (45° vs. 90°) didn't matter much.
  • Analogy: Whether the side roads branch off sharply or gently doesn't change the fact that the trucks are already in the center. The geometry of the split is less important than the width of the road.

• The "Dilution" (How thick the blood is):

  • Finding: Thinner blood (more water, fewer trucks) separates better.
  • Analogy: If the highway is packed bumper-to-bumper with trucks, they can't move to the center easily. If you add more "water" (dilute the blood), the trucks have room to migrate to the center, leaving a cleaner path for the motorcycles at the edges.

• Speed of Flow:

  • Finding: Surprisingly, the speed of the blood didn't change the separation much.
  • Analogy: Whether the cars are driving at 30 mph or 60 mph, the trucks still end up in the middle and the motorcycles at the edges. The device is robust.

• Temperature:

  • Finding: Whether the blood is at room temperature or body temperature didn't change the sorting.
  • Analogy: The traffic rules (physics) work the same whether it's a hot day or a cold day.

---

The "Secret Sauce" and Surprises

The researchers used a mathematical model called the Diffusive Flux Model. Think of this as a set of rules that predicts how the "trucks" push each other away from the walls.

• The Constriction Surprise:
  Previous experiments suggested that squeezing the road (a constriction) before the split would help. However, this computer study found that squeezing the road didn't actually help much in their model.

  • Why? The model assumes the blood is a smooth fluid. In reality, blood is made of individual cells. The researchers suspect that in a real physical device, the constriction might help even more than the computer predicts because the cells behave differently when they are physically squeezed against a wall.

• The "Long Hallway" Trick:
  They found that making the entrance to the device longer (an "inlet extension") helped.

  • Analogy: Imagine giving the trucks a long, straight hallway to run down before they reach the split. This gives them plenty of time to get out of the way and line up in the center, ensuring the side roads get pure platelets.

The Bottom Line

This paper proves that we can design a tiny, passive chip to separate blood components efficiently.

• Best Design: A narrow channel with a long entrance and diluted blood.
• Why it matters: This could lead to cheap, portable, and gentle devices for hospitals or field clinics to treat blood diseases without damaging the precious cells inside.

It's like turning a chaotic traffic jam into an orderly parade, ensuring the right vehicles get to the right destination without crashing.

Technical Summary

1. Problem Statement

The separation of blood components, specifically isolating platelet-enriched plasma from Red Blood Cells (RBCs), is critical for treating diseases like anemia and blood cancer. Traditional active separation methods (e.g., centrifugation) are time-consuming, complex, and risk damaging blood constituents (e.g., hemolysis) due to high shear forces. Passive microfluidic devices offer a promising alternative by utilizing hydrodynamic forces (shear-induced migration) to separate components without external fields.

However, designing these devices requires understanding complex multiphase flow dynamics at the microscale. While detailed mesoscale simulations (like Dissipative Particle Dynamics - DPD) are accurate, they are computationally prohibitive for iterative design optimization. There is a need for a computationally efficient continuum-based model that can accurately predict RBC migration and separation efficiency in trifurcated microchannels to guide the design of passive separators.

2. Methodology

The authors employed a 3D numerical simulation framework using a Diffusive Flux Model (DFM) coupled with a realistic blood rheology model.

• Governing Equations: The study solved the incompressible continuity and momentum equations coupled with a species transport equation for RBC concentration ($\phi$).
• Diffusive Flux Model (DFM): Instead of solving two-phase equations, the DFM treats blood as a single continuum where RBC migration is driven by diffusive fluxes. The total flux ($J$) includes:
  • Collision Flux ($J_c$): Driven by gradients in particle interaction frequency (shear rate).
  • Viscosity Gradient Flux ($J_\mu$): Driven by gradients in apparent viscosity.
  • Brownian Diffusion ($J_b$): Negligible for large RBCs but included for completeness.
  • Note: Streamline curvature flux was neglected as the channel is mostly straight.
• Rheology Model: The Apostolidis and Beris model was used to define blood's apparent viscosity, which depends on local hematocrit, shear rate, and temperature (Arrhenius dependence). This model accounts for yield stress and shear-thinning behavior.
• Geometry: A trifurcated microchannel was simulated. It consists of a main inlet channel that splits into a central outlet and two side "separator arms." The design relies on the Fahraeus-Lindqvist effect and shear-induced migration, where RBCs migrate to the low-shear center, leaving a cell-free layer (CFL) rich in platelets near the walls, which is then extracted via the side arms.
• Numerical Implementation:
  • Solver: Customized transient solver (pimpleFoam) in OpenFOAM.
  • Validation: Validated against experimental data (Lyon and Leal) and DPD simulations (Lei et al.), showing good agreement in velocity and concentration profiles.
  • Grid Independence: Verified using four mesh densities; a hexahedral mesh (M3) was selected for final analysis.

3. Key Contributions

• Computational Efficiency: Demonstrated that the DFM coupled with continuum rheology is a viable, cost-effective alternative to DPD for optimizing microfluidic blood separators, reducing computational time while maintaining accuracy.
• Parametric Optimization: Conducted a comprehensive sensitivity analysis of geometric and flow parameters (channel width, separator angle, inlet extension, constriction, flow rate, hematocrit, and temperature).
• Mechanism Elucidation: Quantified the relative contributions of collision flux vs. viscosity gradient flux in driving RBC migration within the complex trifurcation geometry.
• Design Guidelines: Provided specific recommendations for maximizing separation efficiency, identifying that channel width and inlet extension are more critical than separator angles or flow rates.

4. Key Results

• Geometry Impact:
  • Channel Width: Smaller channel widths (e.g., 60 µm vs. 80 µm) significantly increase separation effectiveness. Narrower channels enhance transverse gradients, promoting stronger RBC migration to the center and a thicker Cell-Free Layer (CFL).
  • Separator Angle: The angle of the separator arms (45° vs. 90°) had a negligible effect on separation efficiency compared to channel width.
  • Inlet Extension: Extending the inlet channel before the trifurcation point allows more time for RBCs to migrate to the center, significantly improving separation.
  • Constriction: Adding a constriction before the separator arm did not improve performance in the simulations, contradicting some experimental literature. The authors attribute this to the continuum model's inability to fully capture the zero-concentration wall boundary condition of real RBCs.
• Flow and Concentration:
  • Flow Rate: Within the studied range (Re ~12–63), flow rate changes affected velocity distribution but had no significant impact on hematocrit distribution or CFL thickness.
  • Hematocrit Concentration: Lower inlet hematocrit (diluted blood) resulted in a thicker CFL and higher separation efficiency. High inlet concentrations reduced the effectiveness.
• Flux Analysis:
  • The collision flux and viscosity gradient flux are the dominant drivers of migration. They often oppose each other in straight channels but interact complexly in the trifurcation zone.
  • Brownian diffusion was confirmed to be negligible ($O(10^{-5})$ m/s) compared to shear-driven fluxes ($O(10^{-2})$ m/s).
• Temperature:
  • Changing bulk temperature from 25°C (298 K) to 37°C (310 K) altered blood viscosity but had no significant effect on the hematocrit distribution or separation efficiency, as the migration fluxes depend primarily on shear rate gradients which remain spatially consistent.
• 2D vs. 3D & Coupled vs. Decoupled:
  • 2D Simulations: While velocity profiles matched 3D results, 2D simulations failed to accurately predict concentration profiles due to the absence of wall effects in the third dimension. However, 2D was useful for qualitative trends.
  • Decoupling: Solving the momentum and species equations separately (using average viscosity for flow, then solving transport) yielded results nearly identical to the fully coupled approach, offering a faster computational route for design studies.

5. Significance

This study provides a robust, computationally efficient framework for the design of passive microfluidic blood separators. By validating the Diffusive Flux Model against experimental and DPD data, the authors establish a reliable tool for predicting device performance without the prohibitive cost of mesoscale simulations.

The findings offer critical design rules for medical device manufacturers:

1. Prioritize Geometry: Focus on minimizing channel width and extending the inlet rather than optimizing separator angles.
2. Sample Preparation: Diluting blood samples prior to separation improves efficiency.
3. Modeling Strategy: Decoupled solvers can be used for rapid prototyping, though 3D simulations are necessary for accurate concentration profiling.

Ultimately, this work bridges the gap between theoretical fluid dynamics and practical biomedical device engineering, facilitating the development of low-cost, passive devices for life-saving blood component separation.