The most common emulsion failure I see isn’t separation after six months on shelf. It’s separation within 48 hours of the prototype being made, which tells me the emulsion was never properly formed in the first place.
Oil and water don’t mix — that’s thermodynamics, not a formulation problem you can engineer around. What you can do is create a kinetically stable system that resists separation long enough to be commercially viable. Here’s how.
Key Takeaways
|
What an Emulsion Is and Why It's Unstable
An emulsion is a dispersion of one immiscible liquid within another, stabilized by a surface-active agent (emulsifier) at the interface. In beverages, this is almost always an oil-in-water (O/W) emulsion — oil droplets dispersed throughout an aqueous continuous phase. The rate at which the dispersed phase separates via creaming is quantitatively modeled by Stokes’ Law:
v = 2g r² (ρp − ρf) / 9η
Where:
- v = Creaming velocity (m/s)
- g = Acceleration due to gravity (9.81 m/s²)
- r = Droplet radius (m)
- ρp = Mass density of the dispersed oil phase (kg/m³)
- ρf = Mass density of the continuous aqueous phas
The system is thermodynamically unstable. Given enough time, oil droplets will always coalesce and separate. Your formulation goal is to slow that process to beyond your required shelf life.
Destabilization Mechanisms
Destabilization happens through four mechanisms:
Creaming — droplets rise due to density difference between oil and water. Lower density oil floats. Rate described by Stokes’ Law: proportional to droplet radius squared. Smaller droplets cream more slowly — this is why homogenization matters.
Flocculation — droplets aggregate without fully merging. Reversible with gentle agitation. Often a precursor to coalescence.
Coalescence — droplets merge irreparably into larger droplets, accelerating phase separation. Once this starts, it’s difficult to reverse without re-processing.
Ostwald ripening — smaller droplets dissolve into the continuous phase and redeposit onto larger droplets, driven by differences in Laplace pressure. More relevant in emulsions with partially water-soluble oil phases.
Understanding which mechanism is driving your instability determines which intervention fixes it.
HYDROPHILE-LIPOPHILE BALANCE (HLB): HOW TO CHOOSE YOUR EMULSIFIER
Every emulsifier has a Hydrophile-Lipophile Balance (HLB) value — a number from 0 to 20 describing the relative affinity of the molecule for water versus oil.
- HLB < 6: Favors water-in-oil emulsions. Not relevant for most beverages.
- HLB 8–16: Favors oil-in-water emulsions. This is your beverage range.
- HLB > 16: Very hydrophilic. More useful as a solubilizer than a true emulsifier.
Each oil phase has a required HLB — the emulsifier HLB that gives optimum stability for that specific oil. Common required HLB values in beverage work:
- Citrus oils (lemon, orange, lime): HLB 7–8
- MCT oil: HLB 5–6
- Flavour oils (generic): HLB 7–9
- Vitamin E (tocopherol): HLB 6–8
- Algal oil (DHA): HLB 6–8
When your emulsifier HLB matches the required HLB of your oil, you get maximum interface coverage and stability. When they’re mismatched, you get poor coverage, larger droplets, and rapid separation.
Common beverage emulsifiers and their HLB values
For clean-label citrus-based RTDs — the most common emulsion formulation I work on — I typically use quillaja saponin at 0.1–0.5% combined with a small amount of modified starch. The saponin handles the interface, the starch adds viscosity to slow creaming.
In fact, according to ACS Publications, natural plant-derived saponins can provide effective steric and electrostatic stabilization of emulsions across a relatively broad pH range, including acidic conditions where some biopolymer-stabilized systems may become less stable.
If you are selecting emulsifiers for functional or plant-based liquids, contact experts at Food Scientist For Hire. Custom beverage development services
Droplet Size: The Number That Controls Everything
Stokes’ Law tells us that creaming rate is proportional to the square of the droplet radius. Cutting droplet diameter in half reduces creaming rate by a factor of four.
Target droplet size for a stable beverage emulsion: < 1 micron (1,000 nm). Below 0.3 microns, you’re approaching a nanoemulsion — significantly more stable but requiring higher energy input to produce.
How you get there:
High-pressure homogenization It is the standard for commercial beverage emulsions. Typical operating pressures of 150–350 bar for a single pass. Running two passes at 200 bar is generally more effective than one pass at 400 bar for most oil/emulsifier combinations.
Microfluidization achieves smaller, more uniform droplet sizes than conventional homogenization but at higher capital cost. Common in pharmaceutical emulsions, increasingly used in premium beverage applications.
High-shear mixing (rotor-stator) alone does not achieve stable beverage emulsions. It reduces droplet size to 5–50 microns — useful as a pre-mix step before homogenization, not as a standalone process.
On the bench, if you’re using a high-shear mixer without access to a homogenizer, your emulsion prototype is not representative of what commercial production will look like. Either access a lab homogenizer (many co-packers offer pilot runs) or source a pre-made beverage emulsion from a flavour house and validate stability from that baseline.
For applications incorporating lipophilic actives into performance beverages, consult our technical resource on sports drink formulation principles.
The Role of the Continuous Phase
Emulsion stability isn’t just about the droplet interface — the continuous phase matters too.
Viscosity
Viscosity of the continuous phase slows droplet movement and reduces creaming rate. A small addition of a hydrocolloid — xanthan at 0.02–0.05%, or a modified starch at 0.5–1.0% — increases continuous phase viscosity enough to meaningfully extend stability without affecting mouthfeel at drinking temperatures.
pH and Isoelectric Points
pH affects emulsifier performance significantly. Lecithin is most effective at neutral to mildly acidic pH. Protein-based emulsifiers (whey, pea protein) lose functionality when pH approaches their isoelectric point — at that pH, net charge is zero, electrostatic repulsion between droplets collapses, and flocculation accelerates. For whey protein, the isoelectric point is ~pH 5.2. Formulating a protein-emulsified beverage at pH 4.8–5.5 without accounting for this is a common failure point.
Ionic Strength and Electrolytes
Ionic strength from mineral fortification can compress the electrical double layer around charged droplets, reducing electrostatic repulsion and accelerating flocculation. If you’re adding calcium, magnesium, or other divalent ions to an emulsified product, test compatibility explicitly.
NOTE:- Under regulatory standards maintained in the US FDA 21 CFR Food Additive Regulations, all continuous-phase stabilizers must be verified for specific food category maximum limits. If you are scaling a complex beverage from benchtop to co-packer, review our step-by-step guide on how to start a beverage business.
Cloudy vs. Transparent Emulsions
An emulsion appears opaque or cloudy when droplet size is comparable to the wavelength of visible light (~400–700 nm). Below ~100 nm, droplets no longer scatter visible light and the emulsion appears clear — a nanoemulsion or microemulsion depending on thermodynamic stability.
- Cloudy Emulsions: Cloudy beverages (juice, cloudy lemonade, most functional RTDs) typically target droplet sizes of 0.5–2.0 microns — stable enough to resist rapid creaming but still light-scattering.
- Transparent Nanoemulsions: Clear emulsified beverages — clear flavored waters with oil-soluble vitamins or flavors — require droplet sizes below 100 nm and specialized emulsification systems. This is achievable but significantly more demanding technically and requires higher emulsifier concentrations and energy input.
Common Mistakes
Using polysorbate 80 without checking label requirements. It’s effective but synthetic, and most clean-label positioning excludes it. Check your target retailer requirements before building a formula around it.
Not matching emulsifier HLB to oil required HLB. This is the most common root cause of early separation I see in prototype evaluation.
Skipping homogenization and relying on high-shear mixing. The droplet sizes are incomparable. If you can’t homogenize your prototype, at minimum source a commercial beverage emulsion base and use that as your model system.
Adding calcium or other divalent ions without testing emulsion compatibility. This destabilizes many emulsifier systems. Always test fortified formulas for emulsion stability separately from unfortified base.
Evaluating stability only at refrigeration temperature. Always run stability at 25°C and 38°C (accelerated). What looks stable at 4°C for two weeks may separate within days at ambient.
Key Formulation and Processing Parameters To Keep in Mind
Conclusion: A stable beverage emulsion comes from matching your emulsifier HLB to your oil, achieving sub-micron droplet size through proper homogenization, and designing a continuous phase that slows droplet movement. Get those three right and you have a system that holds. Miss any one of them and no amount of troubleshooting at the back end fixes what’s wrong at the front.
FAQs
Creaming is the gravitational movement of oil droplets toward the top of the beverage due to density differences, where droplets remain individual entities. Coalescence is the irreversible rupture of interfacial films, causing droplets to merge into larger oil drops and permanently destroy the emulsion.
High-shear rotor-stator mixers only reduce droplet size down to 5–50 μm. Per Stokes’ Law, these large droplets cream rapidly. Commercial shelf stability requires sub-micron droplets (< 1.0 μm), which can only be achieved via high-pressure homogenization or microfluidization.
Multiply the mass fraction of each emulsifier by its individual HLB value and sum the results: HLB_blend = (w_A × HLB_A) + (w_B × HLB_B). Adjust the ratio until the blend matches the required HLB of your oil phase.
At a protein’s isoelectric point (pI ≈ pH 5.2 for whey), the net electrical charge on the protein becomes zero. Without electrostatic repulsion to keep droplets apart, attractive van der Waals forces cause the droplets to aggregate, flocculate, and separate.
Standard operating pressure ranges from 150 to 350 bar. Running two sequential passes at 200 bar is generally more effective than a single pass at 400 bar, as the second pass disrupts re-coalesced droplets and creates a tighter particle size distribution.
Divalent cations like calcium (Ca²⁺) and magnesium (Mg²⁺) compress the electrical double layer surrounding charged emulsion droplets. This neutralizes electrostatic repulsion, allowing droplets to come into close contact and flocculate.