Why Geometry Matters in Droplet Microfluidics
When you're generating monodisperse droplets for single-cell encapsulation, emulsion formulation, or high-throughput screening, the geometry of your microfluidic chip directly determines:
- Droplet size range — what diameter you can achieve
- Polydispersity — how uniform your droplets are (CV%)
- Throughput — how many droplets per second
- Operating window — how forgiving the system is to flow rate fluctuations
- Ease of use — whether you'll spend hours optimizing or get droplets in 10 minutes
The two most common geometries in research labs are flow-focusing and T-junction. Both can generate monodisperse droplets, but they do it differently—and that difference matters when you're choosing hardware for a specific application.
Flow-Focusing: 3D Hydrodynamic Squeezing
How It Works
In a flow-focusing geometry, the dispersed phase (the liquid you want to turn into droplets) flows through a central channel. The continuous phase (oil or aqueous carrier) enters from two side channels and converges at a narrow orifice, squeezing the dispersed phase from three dimensions.
The key feature: the orifice width is typically smaller than the inlet channels (e.g., 50 µm orifice fed by 100 µm inlets). This forces the dispersed stream to neck down and break into droplets just downstream of the orifice.
Typical Droplet Size
- Range: 20–200 µm diameter (depending on orifice size and flow rate ratio)
- Polydispersity (CV%): <3% under stable conditions
- What controls size: Flow rate ratio (Qc/Qd) is the primary lever. Higher continuous phase flow = smaller droplets.
Advantages
- Wide operating window — you can vary flow rates over a broad range (Ca = 0.01–0.5) and still get monodisperse droplets. Forgiving for beginners.
- High throughput — can run at higher total flow rates (up to ~500 µL/min combined) without transitioning to jetting or polydisperse regimes.
- Smaller droplets — easier to push droplet size below 50 µm when you need sub-picoliter volumes.
- Less sensitive to viscosity mismatch — works well even when dispersed and continuous phases have very different viscosities.
Disadvantages
- Higher pressure drop — the narrow orifice creates more flow resistance. You may need syringe pumps capable of >1 bar back-pressure.
- More complex to fabricate — the orifice region requires precise alignment if you're making your own chips (less of an issue with commercial chips).
- Slightly more tubing to set up — three inlets (dispersed + two continuous) vs. two for T-junction.
Best For
- High-throughput screening where you need thousands of droplets per second
- Encapsulation of cells, beads, or particles (wide operating window means fewer failed runs)
- Formulation work where you're testing many different oil/surfactant combinations
- Users new to droplet microfluidics who want a forgiving system
T-Junction: Shear-Driven Breakup
How It Works
In a T-junction geometry, the dispersed phase enters through one channel and meets the continuous phase flowing perpendicular to it, forming a "T." The dispersed phase tries to enter the main channel, but shear forces from the continuous flow pinch it off at the junction.
The key feature: droplet breakup happens at the T itself, driven by shear stress. The dispersed phase temporarily blocks the main channel, pressure builds up, and the droplet snaps off. This "squeezing" regime dominates at low capillary numbers (Ca < 0.01).
Typical Droplet Size
- Range: 30–300 µm diameter (limited by channel width—droplets must fit)
- Polydispersity (CV%): <5% under stable conditions (slightly higher than flow-focusing)
- What controls size: Channel width sets the upper limit. Flow rate ratio fine-tunes size within that constraint.
Advantages
- Lower pressure drop — no narrow orifice means you can use gravity-driven flow or low-pressure pumps for some applications.
- Simpler geometry — easier to fabricate if you're making your own PDMS chips. Just two crossing channels.
- Only two inlets — faster to set up, fewer connections to leak.
- Better for larger droplets — if you need 100–200 µm droplets (e.g., for encapsulating tissue spheroids), T-junction is often more stable than flow-focusing at those sizes.
Disadvantages
- Narrower operating window — transitions to jetting or polydisperse regimes more easily if you push flow rates too high or change viscosity.
- More sensitive to viscosity ratio — works best when dispersed and continuous phases have similar viscosities (within ~10× of each other).
- Lower maximum throughput — typically limited to ~100 µL/min total flow before losing monodispersity.
- Size limited by channel width — hard to make droplets much smaller than 0.5× the channel width.
Best For
- Larger droplets (>80 µm) where you need volume for reagent mixing or spheroid culture
- Low-flow applications like long-term on-chip incubation or gravity-driven setups
- Teaching labs where simplicity and low equipment cost matter more than throughput
- Users with DIY chip fabrication who want the simplest reliable geometry
Side-by-Side Comparison
| Feature | Flow-Focusing | T-Junction |
|---|---|---|
| Droplet size range | 20–200 µm | 30–300 µm |
| Typical CV% | <3% | <5% |
| Max throughput | ~500 µL/min | ~100 µL/min |
| Operating window | Wide (Ca 0.01–0.5) | Narrow (Ca <0.01) |
| Pressure drop | High (~2 bar) | Low (~0.3 bar) |
| Viscosity tolerance | High (10×–100× mismatch OK) | Medium (best <10×) |
| Number of inlets | 3 (1 dispersed + 2 continuous) | 2 (1 dispersed + 1 continuous) |
| Ease of fabrication | Moderate (precise orifice needed) | Easy (just two channels) |
| Best for beginners? | Yes (forgiving) | Moderate (narrow window) |
Operating Parameters: Starting Points for Each Geometry
Flow-Focusing Chip (100 µm orifice, 100 µm channels)
Target: 50 µm droplets at ~1 kHz generation rate
- Dispersed phase (aqueous): 5 µL/min
- Continuous phase (fluorinated oil + surfactant): 20 µL/min per side inlet (40 µL/min total)
- Flow rate ratio Qc/Qd: 8:1
- Capillary number Ca ≈ 0.08
- Expected droplet size: 45–55 µm
- Expected CV%: <2.5%
Tuning tips:
- To make droplets smaller: increase continuous flow (try 30 µL/min per side)
- To make droplets larger: decrease continuous flow or increase dispersed flow
- If droplets become polydisperse: you've entered the jetting regime—reduce total flow rate
T-Junction Chip (100 µm wide × 50 µm deep channels)
Target: 80 µm droplets at ~200 Hz generation rate
- Dispersed phase (aqueous): 3 µL/min
- Continuous phase (mineral oil + Span 80): 12 µL/min
- Flow rate ratio Qc/Qd: 4:1
- Capillary number Ca ≈ 0.005
- Expected droplet size: 75–85 µm
- Expected CV%: <4%
Tuning tips:
- To make droplets smaller: increase continuous flow (try 18 µL/min)
- To make droplets larger: increase dispersed flow (try 5 µL/min)
- If you see jetting (long threads instead of droplets): reduce both flow rates proportionally
- If droplets are irregular: check that your surfactant concentration is sufficient (typically 2–5% w/w)
Real-World Application Examples
When Flow-Focusing Won
Single-cell RNA sequencing prep (10× Genomics-style)
A genomics lab needed to encapsulate single cells with barcoded beads at >2000 droplets/second. They tried T-junction first but hit the throughput ceiling at ~500 Hz before droplets became polydisperse.
Switching to a flow-focusing chip with a 60 µm orifice, they ran:
- Cell suspension: 10 µL/min
- Oil phase: 80 µL/min (40 per side)
- Result: 30 µm droplets at 2.5 kHz with CV% = 2.1%
The wide operating window meant they could process 8 samples back-to-back without re-optimizing flow rates between samples.
When T-Junction Won
Alginate bead production for cell encapsulation
A tissue engineering group needed 150 µm alginate droplets for subsequent crosslinking into hydrogel beads. The high viscosity of 2% alginate solution (η ≈ 50 mPa·s) was a challenge.
Flow-focusing chips required >3 bar pressure and frequently clogged at the orifice. A T-junction chip (150 µm channels) worked at much lower pressure:
- Alginate: 8 µL/min
- Mineral oil: 15 µL/min
- Result: 140 µm droplets at 0.3 bar back-pressure, CV% = 4.5%
The simpler geometry and lower shear also improved cell viability (no shear-induced damage at the orifice).
Decision Framework: Which Geometry for Your Project?
Use this decision tree to narrow down your choice:
Start here: What droplet size do you need?
< 40 µm → Flow-focusing is your best bet. T-junction struggles below 0.5× channel width.
40–100 µm → Either works. Ask the next question:
- High throughput needed (>500 droplets/sec)? → Flow-focusing
- Low equipment budget (no high-pressure pump)? → T-junction
- First time with droplet microfluidics? → Flow-focusing (more forgiving)
> 100 µm → T-junction is often more stable at this scale. Flow-focusing can work but offers no advantage.
Key secondary factors
Viscosity mismatch >10×? (e.g., glycerol-water dispersed in low-viscosity oil) → Flow-focusing handles this better.
Shear-sensitive cargo? (live cells, fragile vesicles) → T-junction has lower shear at breakup point.
DIY chip fabrication? → T-junction is simpler to make and more tolerant of fabrication imperfections.
Need to change formulations often? → Flow-focusing's wide operating window means less re-optimization between runs.
Common Troubleshooting Issues
Problem: Droplets are polydisperse (CV% > 10%)
Flow-focusing:
- Cause: You're in the jetting regime (Ca too high)
- Fix: Reduce total flow rate by 30–50%
- Prevention: Stay below Ca ≈ 0.5 for your system
T-junction:
- Cause 1: Flow rates too high for the squeezing regime
- Fix: Drop both flows by 40%, then re-optimize ratio
- Cause 2: Insufficient surfactant
- Fix: Increase surfactant to 3–5% w/w and pre-wet channels with oil phase for 5 minutes
Problem: No droplets form—just a continuous stream
Both geometries:
- Cause: Interfacial tension too low (over-surfacted) or flow rates far too high
- Fix: Try a fresh oil phase with lower surfactant concentration (start at 1% w/w)
- Also check: Channel hydrophobicity. PDMS channels should be hydrophobic for water-in-oil droplets. If not, treat with Aquapel or similar.
Problem: Droplets form but immediately coalesce
Both geometries:
- Cause: Surfactant hasn't reached the interface yet
- Fix: Pre-condition channels—flow oil + surfactant for 10 minutes before introducing dispersed phase
- Also: Increase surfactant concentration or switch to a faster-adsorbing surfactant (e.g., PFPE-PEG for fluorinated oils)
What You Need Beyond the Chip
Both geometries require:
Syringe pumps — minimum 2 channels (dispersed + continuous). Flow-focusing benefits from a third channel for symmetric continuous flow, but you can Y-split one pump output if budget is tight.
Appropriate tubing — PTFE tubing (1/16" OD, 0.02" ID is common) and fittings to connect pumps to chip inlets. Avoid air bubbles during connection—they kill droplet uniformity.
Surfactant — the right surfactant for your oil/water pair is non-negotiable. For fluorinated oils, use PFPE-PEG block copolymers (e.g., RAN Biotechnologies). For mineral oil, Span 80 (2–3% w/w) is a good starting point.
Microscope + camera — you need real-time visualization to dial in the operating point. A stereo microscope with 2–4× magnification is sufficient for initial setup; a compound microscope with 10× objective is better for CV% measurement.
The Bottom Line
Flow-focusing is the workhorse geometry for most research applications. Its wide operating window, high throughput, and tolerance for viscosity mismatch make it the safer choice when you're unsure. Use it for:
- Single-cell encapsulation
- High-throughput screening
- Any application where you need <50 µm droplets
- Teams new to droplet microfluidics
T-junction shines in specific niches where its simplicity and low pressure are advantages. Use it for:
- Large droplets (>100 µm)
- Viscous dispersed phases (>50 mPa·s)
- Low-flow, long-duration experiments
- DIY chip projects or teaching labs
If you're still unsure, start with flow-focusing. It's more forgiving, and the lessons you learn translate directly to T-junction if you need to switch later.
Ready-to-Use Chips
LabCore offers both geometries as pre-bonded, plasma-treated PDMS/glass chips with integrated tubing kits:
Flow-Focusing Droplet Chip — 100×100 µm channels, 75×25 mm slide format, ships with PTFE tubing and luer fittings. Optimized for 20–150 µm droplets.
T-Junction Droplet Chip — 100×50 µm channels, 75×25 mm slide format, includes syringe kit. Best for 50–200 µm droplets at lower pressure.
Both ship as research-use-only (RUO) hardware, ready to connect to your pumps. No bonding, no plasma treatment, no waiting—just unpack and start generating droplets.