Nano Emulsion Troubleshooting: 6 Failures That Kill Cannabis Formulations

Your nano emulsion separated overnight. Or it was cloudy when it should have been clear. Or the particle size analyzer came back at 350nm after your sonicator ran for 20 minutes. Every formulator hits these walls. The difference between the ones who fix them and the ones who dump batches comes down to understanding which variable actually broke.

This is the troubleshooting guide I wish existed when I started running cannabis nano emulsions at production scale. Every failure mode traced to its root cause, every fix tied to a specific process parameter. If you already know how to make a cannabis nano emulsion, this is what to do when the process stops cooperating.

Failure #1: Phase Separation Within 48 Hours

You make the emulsion, it looks perfect, you put it on the shelf. Two days later there is a cream layer on top and a clear layer on the bottom. This is the most common failure in cannabis nano emulsions and it almost always traces to one of three causes.

Cause A: Insufficient Surfactant Loading

The surfactant is what holds oil droplets in suspension. If you don’t have enough surfactant to coat every droplet surface, the uncovered areas will find each other and coalesce. The critical number is the surfactant-to-oil ratio (SOR). For cannabis oil nano emulsions using polysorbate 80 or quillaja saponin, you typically need an SOR between 1:1 and 2:1 by weight. Below 0.8:1, you are guaranteed to see creaming within a week.

Fix: Increase surfactant concentration by 25% increments until 72-hour stability holds. If you are already at 2:1 SOR and still separating, the problem is not surfactant loading. Move to Cause B.

Cause B: Particle Size Too Large

Stokes’ Law governs creaming velocity. The rate at which droplets rise is proportional to the square of the droplet radius. A 500nm droplet creams 25 times faster than a 100nm droplet. If your process is producing droplets above 200nm, separation is a physics problem no amount of surfactant will fix.

Fix: Measure particle size distribution with DLS (dynamic light scattering). Target D50 below 100nm with PDI (polydispersity index) below 0.3. If your numbers are worse, see Failure #2.

Cause C: Ostwald Ripening

Even if your initial particle size is perfect, small droplets can dissolve and redeposit onto larger droplets over time. This is Ostwald ripening. It happens faster when the oil phase has any water solubility (which cannabis distillate does, slightly, depending on cannabinoid profile and carrier oil selection). MCT oil is particularly susceptible because medium-chain triglycerides have measurable water solubility compared to long-chain alternatives.

Fix: Add a ripening inhibitor. Long-chain triglycerides (LCT) blended at 10-20% of the oil phase dramatically slow Ostwald ripening. Alternatively, switch from pure MCT to a MCT/LCT blend. Some formulators add a small percentage of mineral oil (food grade) as a ripening inhibitor, but this depends on your regulatory environment.

Failure #2: Particle Size Won’t Drop Below 200nm

You are running the sonicator at full amplitude for 15 minutes and the particle size analyzer still shows D50 above 200nm. The sonication energy is not reaching the droplets, or the premix is not set up correctly for the energy to do its job.

Cause A: Premix Too Coarse

Ultrasonication works by cavitation. The implosion of vapor bubbles creates shear forces that break droplets apart. But cavitation can only break droplets that are already small enough to fit within the cavitation zone. If your premix contains 10-50 micron droplets (typical from hand mixing or a basic magnetic stirrer), the sonicator has to work through multiple orders of magnitude of size reduction. That takes enormous energy and time.

Fix: Use a high-shear rotor-stator mixer or a high-speed homogenizer to create a premix with droplets below 5 microns before sonication. This lets the sonicator focus on the nanoscale refinement it was designed for, instead of doing the bulk emulsification work.

Cause B: Amplitude Too Low or Tip Too Small

Sonication amplitude directly controls cavitation intensity. Below 40% amplitude on most lab sonicators, you are not generating enough cavitation to break cannabis oil droplets below 200nm. The sonotrode tip diameter also matters: a 3mm microtip cannot process more than about 50mL effectively. A 13mm tip handles 200-500mL. If you are using a microtip on a 1L batch, the energy density is too low across the entire volume.

Fix: Run at 60-80% amplitude with the correct tip size for your batch volume. Use pulsed mode (5s on, 2s off) to prevent thermal degradation. Monitor temperature: if the emulsion exceeds 60C, cannabinoids start degrading and terpenes volatilize.

Cause C: Surfactant-Oil Incompatibility

Not every surfactant works with every oil. The HLB (hydrophilic-lipophilic balance) of your surfactant needs to match the required HLB of your oil phase. Cannabis distillate in MCT oil typically needs a surfactant HLB of 12-15. Polysorbate 80 (HLB 15) works well. Span 80 (HLB 4.3) does not. If you are using a surfactant with an HLB below 10, you will never get stable nano-scale droplets in an oil-in-water system.

Fix: Match surfactant HLB to oil phase requirements. For cannabis in MCT, start with polysorbate 80 or a polysorbate 80/polysorbate 20 blend. Quillaja saponin is a clean-label alternative with effective HLB around 13-14.

Failure #3: Emulsion Turns Yellow or Brown Within Days

Color change in a nano emulsion is oxidation. Cannabinoids are sensitive to oxygen, light, and heat. At the nanoscale, the surface area-to-volume ratio of each droplet is enormous, which means oxidation happens faster than it would in bulk oil. A clear emulsion that turns yellow within 3 days is oxidizing. A yellow emulsion that turns brown is severely oxidized and has lost potency.

Fix: Nitrogen blanket the emulsion during processing and packaging. Use amber or opaque containers. Add 0.05-0.1% alpha-tocopherol (vitamin E) to the oil phase before emulsification as an antioxidant. Store at 4C (refrigerated), never at room temperature for long-term stability. If you are using PET bottles, switch to glass or HDPE. PET is oxygen-permeable at the rates that matter for cannabinoid degradation.

Failure #4: Bitter or Metallic Taste

Nano emulsions have a taste problem that bulk oils do not. When you shrink oil droplets to 50-100nm, you increase the rate of sublingual absorption. Cannabinoids hit the taste receptors faster and at higher local concentration. The result is bitterness that consumers describe as metallic, chemical, or medicinal.

Cause A: Surfactant Taste

Polysorbate 80 at concentrations above 1% has a detectable soapy taste. Most cannabis nano emulsions use 2-5% polysorbate. The surfactant itself is the problem. This is why so many cannabis beverage companies struggle with taste: the same surfactant that makes the emulsion stable makes it taste bad.

Fix: Switch to quillaja saponin or modified starch as the primary emulsifier. Both are taste-neutral at effective concentrations. If you must use polysorbate, reduce concentration to the minimum effective level (titrate down from 2:1 SOR until stability breaks, then add 10% back). Mask residual bitterness with natural flavoring at 0.5-1% by volume.

Cause B: Cannabinoid Bitterness Amplified by Nanoparticle Size

THC and CBD are inherently bitter. At nanoscale particle sizes, they interact with taste receptors faster than bulk oil, amplifying the perceived bitterness. This is not a defect in your formulation. It is a consequence of the technology working correctly.

Fix: Use flavor masking, not more surfactant. Citric acid (0.1-0.3%) combined with natural flavoring (mint, citrus, berry at 0.5-1%) effectively masks cannabinoid bitterness. Some formulators use cyclodextrin inclusion complexes to encapsulate bitter compounds, but this adds cost and complexity.

Failure #5: Inconsistent Potency Between Doses

You test three samples from the same batch and get three different potency results. 8mg, 12mg, 9mg from a target of 10mg. The emulsion is not homogeneous. This traces to one of two process failures.

Cause A: Insufficient Mixing After Emulsification

The sonicator processes the volume unevenly. The zone directly under the sonotrode tip gets the most energy. The edges of the beaker get the least. If you pour directly from the processing vessel into bottles without thorough mixing, each bottle gets a different concentration.

Fix: After sonication, transfer the emulsion to a mixing vessel and stir at low speed (200-400 RPM) for 10 minutes before filling. This homogenizes the concentration without re-introducing air. Always sample from the middle of the mixed volume, never from the top or bottom.

Cause B: Settling Between Filling

If you are filling 500 bottles from a 50L batch and the filling process takes 2 hours, the first bottles and the last bottles will have different potency. Gravity works even on nanoscale particles, just slowly. Over a 2-hour fill, a 150nm emulsion can develop measurable concentration gradients.

Fix: Maintain gentle agitation (magnetic stir bar or low-RPM impeller) in the filling vessel throughout the entire filling process. Sample the first bottle, a middle bottle, and the last bottle for potency verification. Acceptable variance is plus or minus 10% of label claim per FDA guidance for dietary supplements.

Failure #6: Microbial Growth

Nano emulsions are water-based systems with organic material (cannabis oil, sometimes natural surfactants) suspended in them. Without proper preservation, they are excellent growth media for bacteria and mold. A batch that smells fine on day 1 and smells sour on day 14 has a microbial problem.

Fix: Add preservative at the formulation stage, not after. Potassium sorbate (0.1-0.2%) plus sodium benzoate (0.05-0.1%) at pH below 4.5 is the standard combination. Adjust pH with citric acid. Test with USP <61> and <62> protocols for microbial limits. If you are using quillaja saponin, note that natural surfactants can themselves support microbial growth if not properly preserved.

The Diagnostic Flowchart

When your emulsion fails, start here:

Prevention: The QC Checkpoints That Catch Problems Early

Every nano emulsion batch should hit these checkpoints before it leaves the lab:

1. Particle size (DLS): D50 below 100nm, PDI below 0.3. Measure within 1 hour of production.
2. Visual clarity: Hold the vial against a light source. A true nano emulsion at sub-100nm is translucent to clear, not milky. If it is milky, particle size is too large.
3. pH: Below 4.5 for preserved stability. Above 5.0 and your preservative system fails.
4. 72-hour stability check: Store a sample at room temperature for 72 hours. Any visible separation is a fail. Re-check particle size at 72h: if D50 increased more than 20%, Ostwald ripening is active.
5. Potency uniformity: Test 3 samples from different positions in the batch. All three must be within 10% of the target.
6. Accelerated stability: Store a sample at 40C for 2 weeks. If it survives, it will survive 6 months at room temperature. If it separates, your shelf life is shorter than you think.

The difference between a nano emulsion that works and one that fails is process control. Every failure I have seen traces back to a measurable variable that was either not measured or not controlled. Particle size, surfactant ratio, pH, temperature, mixing time. Fix the variable, fix the product.

If you are scaling from benchtop to production, start with the 1L lab SOP and validate each checkpoint at scale before committing to a full production run.