Amylose vs Amylopectin: Which Starch Fraction for Gelato
Table of contents
Two gelato bases can carry the same three per cent starch, hit the same viscosity at pasteurisation, and behave completely differently after four days in the cabinet. The difference is not how much starch you used. It is which of the two polymers inside the granule you happened to buy.
Same dose, same viscosity, different polymer: the choice shows up on day four, not day one.
One granule, two very different polymers
Every starch granule is built from the same monomer, glucose, arranged in two ways. Amylose is essentially a linear chain of glucose units joined by alpha-1,4 bonds. Amylopectin is the same backbone but heavily branched, with an alpha-1,6 branch point roughly every twenty to twenty-five units, producing a bushy molecule far larger than amylose.
That shape difference decides everything downstream. Linear chains can lie alongside each other and lock into ordered crystalline regions. Branched molecules cannot: their side chains get in the way. So amylose gels, sets and turns opaque, while amylopectin thickens without setting and stays clear.
In a hot mix both fractions do useful work. They hydrate, swell and bind free water, which raises viscosity and lifts total solids without adding sweetness. The trouble starts once the product is cold and sitting still.
The number that predicts everything
Amylose content is the single specification worth reading on a starch data sheet. It separates starches that will hold a gelato through a week of service from starches that will not.
| Starch | Approx. amylose | Paste character | Freeze-thaw behaviour |
|---|---|---|---|
| Waxy maize | Under 1% | Clear, cohesive, non-setting | Excellent, resists weeping |
| Tapioca | ~17% | Clear, slightly stringy, bland | Good |
| Rice, non-waxy | ~15-20% | Soft, opaque | Moderate |
| Potato | ~20-25% | Very high peak viscosity, clear | Moderate, thins on shear |
| Normal maize | ~25% | Opaque, sets to a gel | Poor, weeps and turns grainy |
| High-amylose maize | 50-70% | Firm gel, resists swelling | Poor for this use |
Values are typical published ranges for native starches; individual lots vary, so confirm against your supplier specification.
Read that table alongside the practical write-ups for waxy maize, tapioca, potato and cornstarch. Native maize starch is the default in most kitchens purely because it is cheap and available, not because it is the right polymer for a frozen product.
Quick reference.
Figure 1 — Linear amylose chains realign within hours; the branched amylopectin structure takes days to weeks to do the same.
Retrogradation is why gelato goes grainy on day four
When a gelatinised starch paste cools, the dissolved polymers begin re-associating into ordered crystalline regions. This is retrogradation, and the two fractions run on completely different clocks.
Amylose retrogrades fast. Its linear chains find each other within hours of cooling, forming a network that is essentially irreversible at normal serving temperatures. Amylopectin retrogrades slowly, over days to weeks, because only the short outer branch chains can participate and they have to reorganise around the branch points.
In a frozen dessert the consequence is water. As the amylose network tightens it expels water it was previously holding, which is syneresis. That released water is free water, and free water in a cabinet at minus 12 degrees does exactly one thing: it joins existing ice crystals and makes them bigger. The result reads on the tongue as sandy or coarse texture, and it worsens with every heat shock cycle rather than recovering.
There is a useful diagnostic in this. Retrogradation damage arrives on a schedule, while damage from poor storage arrives after an event. If a flavour tastes clean on day one, acceptable on day two and sandy on day four with no cabinet excursion in between, suspect the starch rather than the freezer. If it degrades only after a delivery or a defrost cycle, the starch is probably innocent and the cold chain is not. Tasting the same batch on a fixed schedule, always at the same temperature, separates the two causes within a single week of service.
A base built on a high-amylopectin starch simply does not have this failure mode on a normal service timescale. It keeps holding its water, so ice crystal growth stays governed by your storage discipline rather than by a slow chemical clock inside the mix.
What this means at the pasteuriser and the batch freezer
Both fractions need full gelatinisation to do any of this, and that is a temperature and time question, not a stirring question. Typical gelatinisation ranges run near 62 to 72 degrees for normal maize, 63 to 72 for waxy maize, 58 to 66 for potato and 59 to 70 for tapioca. A mix that peaks at 65 degrees and drops immediately may leave part of the granule population unswollen.
Two practical notes follow. First, hold above the top of the range long enough for the granules to swell fully, and treat that as separate from your pasteurisation hold. Second, potato starch reaches a very high peak viscosity and then thins sharply under continued shear, so a base that looks perfect leaving the pasteuriser can arrive at the aging tank noticeably thinner. Waxy maize is far more forgiving here.
Starch also interacts with the rest of your hydrocolloid plan. It is not a drop-in replacement for gums, and the comparison in starch versus gum is worth reading before you rebalance. Starch adds solids and body; gums build serum viscosity at a fraction of the dose. Most stable formulas use both, and the hydration temperature guide matters as much for the starch as for the gum.
Native or modified
If you want amylopectin behaviour but cannot source waxy starch, modified starches are the standard route. Cross-linking and stabilisation are applied precisely to suppress retrogradation and syneresis under freeze-thaw. In the European Union, acetylated distarch adipate (E1422) and hydroxypropyl distarch phosphate (E1442) are authorised food additives under Regulation (EC) No 1333/2008; in the United States they fall under food starch-modified in 21 CFR 172.892.
The tradeoff is labelling. A native waxy maize starch declares as starch, while a modified one carries its E number or the food starch-modified designation, which matters if you are selling on a clean-label positioning. The behaviour case for modified starch is strong, so this is a commercial decision rather than a technical one. The practical options are laid out in modified starch in gelato.
Whichever you choose, remember what starch is actually buying you: bound water and body, counted properly in your total solids and reflected in the water activity of the finished mix. Amylose content only tells you how long that binding will last.
