Starch Gelatinization Temperature in Gelato Bases, by Starch
Table of contents
A starch only thickens a gelato base once its granules pass their gelatinization window, and that window is not the number printed on the bag. Sugar pushes it higher, and a legal pasteurization minimum can stop short of it. Here are the measured windows, the sugar shift, and a worked 1,000 g base.
Every starch has its own temperature window. The base has to clear it, not just approach it.
What gelatinization actually is
Native starch arrives as tightly packed, partly crystalline granules. In cold milk they do nothing except settle to the bottom. As the base heats, water gets into the granule, the ordered regions melt, the granule swells and amylose starts to leak out. That irreversible change is gelatinization, and it is the only point at which cornstarch or tapioca starts to hold water and build body.
Food scientists measure it with differential scanning calorimetry (DSC), which reports three temperatures:
- Onset (To): the first granules begin to melt.
- Peak (Tp): most of the transition is happening.
- Conclusion (Tc): the endotherm ends.
For a gelatiere, Tc is the number that matters, because a base stopped between To and Tc has only part of its starch working. Even Tc is optimistic: a 2018 study in Scientific Reports found that maize and wheat starches still held residual structural order at Tc and only finished at a higher "end temperature". Rice starch was complete at Tc.
Gelatinization temperatures by starch
The table collects DSC results for native starches measured in water. Notice how much two studies of the same starch can disagree: variety, growing conditions and method all shift the numbers, which is why you plan for the top of the window, not the middle.
| Starch | Onset To (°C) | Peak Tp (°C) | Conclusion Tc (°C) | Study |
|---|---|---|---|---|
| Normal maize (cornstarch) | 64.9 | 70.0 | 75.8 | Polymers 2018, plasma study control |
| Normal maize (cornstarch) | 65.9 | 71.5 | 76.4 | ACS Omega 2019, native control |
| Normal maize (cornstarch) | 71.6 | 77.0 | 87.9 | Ultrasonics Sonochemistry 2021, untreated |
| Waxy maize | 63.1 | 69.8 | 77.2 | Polymers 2018, plasma study control |
| Potato | 61.1 | 65.7 | 71.7 | ACS Omega 2019, native control |
| Tapioca (cassava) | 61.4 | 68.1 | 76.6 | Foods 2020 |
| Tapioca (cassava) | 66.8 | 74.3 | 86.8 | Ultrasonics Sonochemistry 2021, untreated |
| High-amylose maize | 91.9 | 110.7 | 132.4 | Foods 2020 |
Quick reference. In plain water, cornstarch finishes somewhere between 76 and 88 °C, waxy maize near 77 °C, potato near 72 °C and tapioca between 77 and 87 °C. High-amylose maize only began to melt near 92 °C in the Foods 2020 sample, so it does not gelatinize in a gelato pasteurizer. Add a sugar margin on top of all of these.
Figure 1. Gelatinization windows in water (onset to conclusion, widest reported span) against the two Ontario pasteurization temperatures and an 85 °C working target for sweetened starch bases.
Two consequences follow. First, waxy maize and normal maize gelatinize in roughly the same window; what separates them is the amylose to amylopectin ratio, which decides retrogradation, not the cooking temperature. Second, potato starch sets earliest, which is why it is the starch most forgiving of a gentle heat treatment.
Sugar moves the window up
Those numbers were measured in water. A gelato base is not water: it is a sugar solution. Sugar lowers the water activity of the liquid and interacts with the starch chains, so the granule needs more heat before it swells.
A 2022 DSC study in the Journal of Food Science measured wheat starch in sugar solutions. With plain water the onset was about 60.7 °C. In a 15% sucrose solution it rose to 65.8 °C, and in a 30% sucrose solution to 72.2 °C. Glucose, the sugar in dextrose, raised it less than sucrose did at the same concentration: 64.3 °C at 15%. The study used wheat starch at a high starch to liquid ratio, so the exact shift in a maize gelato base will differ. The direction will not: in that study every sugar raised the onset at every concentration, and sucrose raised it most.
A sweetened milk gelato, especially one that leans on sucrose rather than dextrose, sits between those two test points, as the worked example shows. Treat every Tc in the table as a floor and add a margin.
Worked example: a 1,000 g cornstarch milk base
Here is a cornstarch-thickened milk gelato built with the coefficients in the app's ingredient database (dextrose monohydrate at POD 64 and PAC 173 per 100 g, corn starch with zero PAC and zero POD).
| Ingredient | Grams | Sugars* (g) | Fat (g) | MSNF (g) | POD | PAC |
|---|---|---|---|---|---|---|
| Whole milk | 640 | 0 | 22.4 | 57.6 | 5 | 31 |
| Fresh cream 35% | 120 | 0 | 42.0 | 7.2 | 1 | 4 |
| Skim milk powder | 35 | 0 | 0.4 | 33.3 | 3 | 18 |
| Sucrose | 150 | 150 | 0 | 0 | 150 | 150 |
| Dextrose monohydrate | 30 | 27.3 | 0 | 0 | 19 | 52 |
| Corn starch | 25 | 0 | 0 | 0 | 0 | 0 |
| Total | 1,000 | 177.3 | 64.8 | 98.1 | 178 | 255 |
* Sugars excluding lactose; lactose is carried inside MSNF.
The base lands at 36.2% total solids, 17.7% sugars, 6.5% fat, 9.8% MSNF, POD 178 and PAC 255, inside the canon ranges for a milk gelato (solids 36 to 42%, sugars 16 to 22%, fat 6 to 12%, MSNF 8 to 11.5%, POD 140 to 200, PAC 220 to 300). With PAC 255, the serving temperature rule for gelato puts it near -12.8 °C in the display.
Now the decision only a gelatiere has to make. The water in this base is about 638 g and the added sugars 177 g, so the liquid around each granule is roughly a 22% sugar solution before lactose is counted. That is between the 15% and 30% test points above, so expect the onset to move up by roughly 5 to 12 degrees. Cornstarch that finishes at 76 °C in water can need the low 80s here, and more if your lot behaves like the 88 °C sample.
Measure that against the pasteurization options in Goff's ice cream technology text (Ontario rules): 69 °C for 30 minutes or 80 °C for 25 seconds.
- 69 °C batch hold: legal for safety, but below the cornstarch Tc even in plain water. Much of the starch stays unswollen and the gelato eats thin and chalky.
- 80 °C short hold: just above the water Tc of most samples and below the sugar-shifted window. Borderline.
- 85 °C with stirring until the base visibly thickens: clears the shifted window with a margin. This is the setting to choose for a starch-thickened base, and the reason starch bases are better cooked on a high pasteurization program or on the stove.
How to cook a starch base without wrecking it
- Disperse cold. Whisk the starch into a portion of cold milk, or blend it dry with the sugar, before it meets the hot base. Granules dropped into hot liquid gelatinize on the outside and form lumps.
- Stir through the window. Starch granules settle. Without stirring, the bottom scorches and the top never thickens.
- Go past Tc, then stop. Reach the target, hold briefly, then cool fast. Swollen granules are fragile: long boiling and hard shear break them and the paste thins, a loss that more heat cannot undo.
- Mind the other stabilizers. If the base also carries gums, their hydration temperatures are a separate requirement from the starch window. Cook to whichever is higher.
- Age it. Chill the cooked base and give it the same ageing as any other base before batch freezing.
To rebalance the recipe after swapping some sugar or starch, run the numbers in the free gelato balancing calculator instead of guessing; it reports PAC, POD and solids as you change grams.
