Microfluidic Cartridge — Flow Visualizer

Click valves to toggle open/closed — watch the flow path update

CARTRIDGE IN Inlet V1 CLOSED V2 CLOSED V3 CLOSED V4 CLOSED SENSOR ARRAY A CH 1 B CH 2 C CH 3 D CH 4 OUT Waste ▼ ▶ ▶ ▼

Valve Control

Presets

Flow Status

Channel ANo flow
Channel BNo flow
Channel CNo flow
Channel DNo flow
Topology:
IN (top-left) → feed trunk → 4× valve ↓ sensor ↓ drain trunk → OUT (bottom-right)

Junction Geometry — T vs Y

How branch angle affects dead volume, mixing, and sample plug integrity

T-Junction (90°)

dead volume impact zone Main channel Branch 90°
Dead volume: HIGH
The 90° corner creates a stagnation pocket where the branch meets the main channel on the downstream side. Fluid in this pocket does not get swept — it sits there and slowly diffuses into passing flow, causing sample carryover between injections.
Mixing at junction: HIGH
The branch stream impacts the main flow at 90°, creating vortices and turbulent mixing at the impact zone. This disrupts the sample plug front — bad for kinetics where you need a sharp concentration step at the sensor.
Plug shape after junction: Distorted
The leading and trailing edges of the sample plug get smeared by the perpendicular impact. The concentration profile at the sensor becomes a ramp instead of a step.
Fabrication: Simplest
Straight cuts, easy to mill or mold. Standard in most off-the-shelf manifolds.

Y-Junction (30–45°)

smooth merge Main / Buffer Branch / Sample To sensor 30°
Dead volume: LOW
The shallow angle eliminates the stagnation pocket. Flow sweeps smoothly through the junction — no corner for fluid to get trapped. Dramatically reduces carryover between sample plugs.
Mixing at junction: LOW
The branch merges nearly parallel to the main flow. No perpendicular impact, no vortices. The two streams layer side-by-side (laminar co-flow) and mix only by diffusion — which is slow at microfluidic scales (Re < 1).
Plug shape after junction: Preserved
The sample plug enters the main channel with its leading edge intact. The concentration step at the sensor stays sharp — critical for accurate ka/kd kinetics.
Fabrication: Slightly harder
Angled cuts require CNC or laser machining. Not available in basic rectangular-channel molds. But standard for injection-molded microfluidic cartridges.

Side-by-Side Comparison

Property T-Junction (90°) Y-Junction (30–45°) Why it matters
Dead volume High Low Trapped fluid = carryover between injections
Mixing at junction Vortex / turbulent Laminar co-flow Mixing smears the sample plug concentration front
Plug shape preservation Distorted Preserved Sharp plug = accurate ka/kd kinetics at sensor
Pressure drop Higher Lower 90° turn adds flow resistance; matters at low flow rates
Wash efficiency Slower Faster Dead volume takes many wash volumes to clear
Bubble trapping Prone Resistant Bubbles lodge in 90° corners; Y sweeps them through
Fabrication complexity Simple Moderate Angled channels need CNC/laser; OK for injection mold
Best for Low-cost, non-kinetic assays SPR kinetics, low carryover

Design Recommendations for Reducing Mixing

1. Use Y-junctions (30–45°) at all feed-to-branch points.
The shallow merge angle keeps the sample plug intact. For SPR kinetics, a sharp concentration front at the sensor is the single most important fluidic design goal — it directly determines whether you can measure ka and kd accurately.
2. Round all internal corners (fillet radius ≥ 0.5× channel width).
Even in a Y-junction, sharp internal corners create micro-stagnation zones. A fillet radius of half the channel width (e.g., 250 µm fillet for a 500 µm channel) eliminates the last dead volume pockets.
3. Match channel cross-sections at the junction.
If the branch is narrower than the main channel, the sample enters as a jet and creates eddies. If wider, the flow decelerates and pools. Equal cross-section = laminar merge, minimal disturbance.
4. Keep the drain junction angle shallow too.
The drain side matters just as much — if a 90° T-junction sits between the sensor and waste, residual sample gets trapped there and back-diffuses onto the sensor during the dissociation phase. Use Y-junctions on both feed and drain sides.
5. Consider air-segmented plugs for long tubing runs.
If the distance from junction to sensor exceeds ~50 mm, Taylor dispersion will smear the plug regardless of junction geometry. Air bubbles before and after the sample plug act as physical barriers — the same principle used in inject_advanced on the P4PRO.

Path Length Resistance — Zigzag & Serpentine Channels

Increasing channel path length raises fluidic resistance without changing cross-section. Used to balance flow across parallel branches or throttle specific channels.

Straight path — low resistance L = 10 mm R Serpentine path — high resistance (same footprint) L ≈ 35 mm (3.5× longer in same footprint) 3.5R
The physics:
For a rectangular microfluidic channel at low Reynolds number (laminar flow), resistance scales linearly with path length:
R = (12 · μ · L) / (w · h³)
(μ = viscosity, L = length, w = width, h = height)
Double the path length → double the resistance → halve the flow rate (at constant driving pressure). The cross-section stays the same — you're only adding length by folding the channel into zigzags or serpentines.
Why this matters for your 4-channel cartridge:
If all 4 channels (A–D) branch from a single feed trunk, the channel closest to the inlet (A) sees slightly lower total resistance than the furthest (D), because A's trunk path is shorter. This means A gets more flow than D — unequal distribution.
Adding a serpentine section before each channel's valve lets you equalize resistance across all branches. Channel A gets the longest serpentine (most added resistance), channel D gets the shortest (or none). Result: equal flow to all 4 sensors.

Serpentine Trade-offs

Benefit Cost
Equalize flow across parallel branches Increases total dead volume (more channel to wash out)
Tunable resistance without changing cross-section Each U-turn is a potential bubble trap (use rounded bends, R ≥ 2× width)
Compact — fits in small cartridge footprint Taylor dispersion smears the sample plug over the longer path
Can serve as incubation/mixing region if desired Harder to flush — wash time scales with path length
No moving parts — purely geometric Adds fabrication complexity (more features to mill/mold)
Serpentine design rules:
  • Bend radius ≥ 2× channel width — prevents bubble trapping and secondary flow (Dean vortices) at sharp U-turns
  • Equal-width throughout — don't narrow the serpentine to save space; that changes resistance non-linearly and creates jets
  • Place serpentines BEFORE the valve, not after — the sample plug doesn't travel through the resistance section, so dispersion doesn't affect it
  • Calculate per-branch: Rtotal = Rtrunk + Rserpentine + Rsensor + Rdrain — set Rserpentine per channel so all Rtotal values are equal
  • Alternative: width taper. Instead of zigzags, gradually narrow the channel to increase resistance. Lower dead volume, but resistance scales as 1/w·h³ — very sensitive to fabrication tolerances