The usual way to stop pulp settling in a juice is to thicken it, and xanthan is very good at thickening. The problem is that consumers buy juice partly for how it drinks, and a thick juice with held-back flavour tends to lose to a thinner one. The distinction that fixes this is that suspension requires a yield stress — enough structure to hold a particle still at rest — while viscosity determines how the liquid feels in the mouth. Those two properties are related but not identical, and materials differ enormously in how much of each one they deliver per gram.
Low acyl gellan builds what is effectively a weak gel: a continuous network that holds particles in place when the drink is standing still and breaks down under shear so the product pours and is easy to swallow. It does this at 0.015–0.030%, where xanthan would need several times that weight to achieve the same suspension — and most of the dose difference shows up as the absence of thickness.
Suspending versus thickening
| Property | What it does | Who notices |
|---|---|---|
| Yield stress | Holds particles stationary at rest | Whether pulp settles or floats |
| Low-shear viscosity | Determines perceived thickness | Mouthfeel immediately |
| Shear thinning | How it behaves when poured | Pouring, drinking, pumping |
| Flavour release | How quickly aromatics come off | Clean or muted taste |
Xanthan delivers all four, and too much of the second. That is not a criticism of xanthan — it is the right material when you want body as well as suspension, and it is inexpensive. It is simply the wrong tool when the brief is "suspend the pulp and otherwise stay out of the way."
The substitution table
| Line | Xanthan | Low acyl gellan | Consequence |
|---|---|---|---|
| Dose for pulp suspension | several times higher | 0.015–0.030% | Far less material |
| Perceived thickness | Noticeable | Close to water at these doses | Cleaner mouthfeel |
| Yield stress | Builds with viscosity | Builds with almost no viscosity | Suspension without body |
| Calcium requirement | Tolerant | Needs divalent cations | Usually supplied by juice |
| pH working range | Very wide | Wide, but check below pH 3.5 | Test at your pH |
| Clarity | Can haze | Clear | Matters for clear juice with pulp |
The calcium row deserves attention before you start. Low acyl gellan crosslinks with calcium and magnesium, and most fruit juices contain enough. That is convenient and it is also where trials go wrong: switch juice concentrate supplier or dilution water, and the set changes without any formulation change on your side.
The dose arithmetic
| Line | Xanthan system | Gellan system |
|---|---|---|
| Suspending agent dose | several times higher | 0.015–0.030% |
| Powder per 10,000 L | higher | 1.5–3.0 kg |
| Weight ratio | 1 | roughly 1/5 to 1/10 |
| Additional body needed | Already excessive | Sometimes none |
The useful way to read this is not "gellan is cheaper" — it is that gellan lets you remove material. Where xanthan was both suspending and thickening, the gellan route suspends and leaves the thickness decision to the rest of the formulation. Multiply these weights by your own two purchase prices if you need a number for a costing file; the weights themselves are the part that is not negotiable.
Three cases where you should not switch
- Products that need body as well as suspension. Nectars, smoothies and anything sold on thickness should keep xanthan or use both materials together. Removing the viscosity is the point of this substitution, and sometimes you want to keep it.
- Very low pH, below about 3.5. Gellan works across a wide range but the network weakens in strong acid. Test at your actual pH rather than assuming.
- Systems with negligible calcium and no tolerance for added salts. If the base is demineralised water or a very pure concentrate and the label cannot take added calcium salts, the gellan network has nothing to build from.
Migration in four steps
- Measure what you have before changing anything. Particle size distribution of the pulp, pH, calcium content and current yield stress. Without these you will not know which change did what.
- Fix the particle size first. Reducing particle size reduces the suspension requirement more cheaply than any gum change. Do this before touching the stabiliser.
- Dose at 0.015% and go up only as far as the particles need. The window is narrow and overdosing produces a drink with a noticeable set, which is worse than settling.
- Test standing, not agitated. Put filled packs on the shelf and check weekly for six weeks. A drink that survives transport and then settles on the shelf has failed.
Failure modes
| Symptom | Likely cause | First check |
|---|---|---|
| Pulp settles overnight | Yield stress too low | Dose first, then particle size |
| Clear layer forming slowly on top | Network creeping | Increase dose slightly; check pH |
| Pulp floats rather than settles | Density mismatch, not a gum problem | Particle density relative to the serum |
| Drink has a perceptible set | Dose too high | Drop back down immediately |
| Set changes between production runs | Water or concentrate variation | Calcium analysis on incoming material |
| Gasping or off-flavour | Unrelated — usually microbiological | Do not assume it is the gum |
Tools
- Dosage calculator — includes juice with pulp at 0.015–0.030% and ready-to-drink beverages at 0.010–0.025%.
- Formula pre-check — flags pH and ion combinations before you run the trial.
- Troubleshooting tree — works back from the symptom you can see.
Frequently asked
Will the drink look thinner?
Yes, and that is usually the intended result. Gellan delivers suspension with very little low-shear viscosity, so the product stops being thickened by the stabiliser and can taste like juice again.
Does the juice supply enough calcium?
Usually. Fruit juices and concentrates generally contain sufficient calcium and magnesium, but it varies by supplier and by dilution water, so measure it and treat it as a specification rather than a constant.
Can I use it in a clear juice?
Yes — gellan gives a clearer system than xanthan at these doses, which is one reason it is used in clear products with suspended particles.
How low can the dose go?
0.015% at the bottom of the useful window for typical pulp loads. Below that the network becomes unreliable rather than just weaker, and batch-to-batch variation starts showing up as sediment.