Specialist Crafts · Article 26

Sugar Syrup Physics

A reference article · 3,700 words · 49 cited sources · last reviewed October 2026 · fact-checked October 2026, 75 claims

Chocolate & Confectionery gives the candy stages as temperatures. This article supplies what sits underneath them: the concentration each temperature stands for, how a refractometer reads the same number, how elevation moves every target, and how glucose syrup, invert sugar, and acid decide whether cooled sugar sets as a crystal or a glass.

A candy thermometer is a hydrometer in disguise. Dissolved sugar lowers the vapor pressure of the water it sits in, so a syrup has to be hotter than plain water to boil, and the more sugar it carries, the hotter. As water boils off, the concentration rises and the boiling point rises with it. At one atmosphere, then, the temperature of a boiling syrup is a reading of how much sugar is in it, and the stages of candy making are concentration bands with texture names attached. Every rule in confectionery follows from that one relationship, including the ones about elevation, refractometers, and why a recipe carries the glucose it does.

Temperature is concentration

The ideal, dilute form of the relationship is the colligative boiling-point elevation, about 0.51 °C per mole of dissolved solute per kilogram of water (Ergun, Lietha, & Hartel, 2010). Real syrups leave ideality behind long before candy concentrations, and the models used for them are empirical fits to measured data (Starzak & Mathlouthi, 2006). Hartel’s group put it plainly: literature data for the boiling point of sugar solutions, particularly at high concentration, “is highly inconsistent” (Ergun et al., 2010), and the tables put 94 to 95 percent sucrose anywhere from about 260 to 267 °F (127 to 131 °C) (Bates & Associates, 1942, extrapolated above 90 percent; FAO, n.d.; Hugot, 1986). The table that follows is a working reference, not a constant.

Two things the table does not capture. Impurities raise the boiling point further at the same solids, which is why cane-sugar engineering tables carry separate columns by purity (Hugot, 1986; Jayes, n.d.). And a syrup carrying glucose or invert sugar boils higher than pure sucrose at the same solids, because the smaller molecules put more moles into the same weight: glucose solutions “boil at higher temperature than sucrose solutions of equivalent weight percentage,” and higher-DE glucose syrups raise the boiling point more than lower-DE ones (Ergun et al., 2010; Hartel et al., 2011). A doctored syrup reaches a given thermometer reading at slightly lower total solids than the table implies, which is one reason a recipe’s target temperature belongs to its formulation.

The stages, reconciled

The stage names are cold-water tests: a spoonful of syrup dropped into cold water forms a thread, a soft or firm or hard ball, or cracks. Reputable charts now largely agree on the stage temperatures (ThermoWorks, n.d.-b; Exploratorium, n.d.), but the concentrations some charts print beside them come with no stated method (Exploratorium, n.d.). The table here keeps the temperature ranges that Chocolate & Confectionery already carries, which match ThermoWorks’ candy chart and the Exploratorium’s, and reads the concentrations off the boiling-point table above.

Where the sources do disagree is concentration: the popular charts put 95 percent at soft crack, 270 °F or above; the sucrose tables put 94 to 95 percent between about 260 and 267 °F (Exploratorium, n.d.; Bates & Associates, 1942, extrapolated; FAO, n.d.; Hugot, 1986). None of this matters at the bench, where the cold-water test and the formulation’s own tested target settle it; it matters for anyone who wants to convert a thermometer reading into a solids figure, and that person should use the table, not the chart.

Formulation-level cooking targets from the confectionery literature: fondant 116 to 121 °C, fudge 116 °C, caramels and toffee 118 to 132 °C, butterscotch 146 to 154 °C, hard-boiled sweets 149 to 166 °C with under 2 percent final moisture (FAO, n.d.); Greweling’s hard candy at 156 °C (313 °F) without stirring (Greweling, 2013); pulled sugar 160 °C and spun sugar 150 to 155 °C (Domson Academy, 2026; Torreblanca, 2017). The spread within a product is formulation: a syrup carrying more glucose reaches the same texture at a slightly different reading, and every professional target was set by test batch.

Reading concentration directly

A refractometer measures the refractive index of a solution, which rises with dissolved solids, and the degree Brix is the sucrose scale for that reading: one degree Brix equals one percent sucrose by weight at 20 °C (BAKERpedia, n.d.-a). The instrument reads the same quantity a thermometer infers, and for pâte de fruit, jams, sorbet bases, and any syrup that will be cooled and checked, it reads it directly and at any elevation.

Three cautions govern the reading. Temperature first: the scale is defined at 20 °C, and a reading at any other temperature must be corrected, by the instrument’s automatic temperature compensation (good over roughly 10 to 30 °C on handheld models) or by table; a 10 Brix sample read at 25 °C reads about 0.35 low, and a hot syrup has to be cooled to the instrument’s window before it can be read at all (Atago, n.d.; Bellingham + Stanley, n.d.). Purity second: the Brix scale is the ICUMSA sucrose table, and “Brix measurements are best applied to the analysis of solutions consisting of only one solute” (BAKERpedia, n.d.-a); acids, minerals, pectin, and dextrins all shift the refractive index. Third, and the one that matters most in confectionery: glucose syrup, invert sugar, and fructose have their own refractive-index curves, so a sucrose-scale reading on a mixed syrup is an equivalent-sucrose number rather than true solids. ICUMSA calls the corrected quantity refractometric dry substance and requires that results be labelled as refractometric (ICUMSA, n.d.); Hartel’s group notes that Brix values “do not represent the exact total solids (or water) content,” with the error growing as the sucrose fraction falls (Ergun et al., 2010; Kurtz & Eliason, 1979).

The older hydrometer scale, Baumé, survives on glucose-syrup drums and in pastry recipes. The working conversion runs about Baumé = 0.55 × Brix for dilute syrups and nearer 0.53 × Brix at the top: an 80 Brix syrup is about 42.5 Baumé, simple syrup at just over 50 Brix about 27, a sorbet base at 27 Brix about 15 (Bates & Associates, 1942).

Elevation

Lower atmospheric pressure lowers the boiling point of water, and with it every stage temperature. The rule of thumb discussed in Environmental Variables & Troubleshooting, about 1 °F per 500 ft, is the one the extension and test-kitchen literature use; ThermoWorks gives it as 2 °F (1 °C) per 1,000 ft (Earl, 2024), and the USDA’s Food Safety and Inspection Service states it more carefully as “just under 1 °F” per 500 ft (USDA FSIS, n.d.).

Neither the decay of pressure with altitude nor the vapor-pressure curve of water is linear, which is why the rule drifts. Weather matters as much as the last thousand feet: a swing of 3 kPa in barometric pressure, ordinary between a high and a low, moves the boiling point by about 0.8 °C (1.5 °F) at sea level and about 1 °C at Denver’s elevation, which is why ThermoWorks’ boiling-point calculator works from local barometric pressure rather than elevation alone (ThermoWorks, n.d.-a).

The method that beats any table is the one every source agrees on. Boil plain water on the day, read the thermometer, and subtract the difference from 212 °F (100 °C) from every stage target (Exploratorium, n.d.; ThermoWorks, 2025). It corrects for elevation, for the day’s weather, and for the thermometer’s own error in a single step. It assumes the syrup’s boiling-point elevation does not change with pressure. It does, slightly: at lower pressure the elevation shrinks (Holven, 1936; Bates & Associates, 1942), so in Denver the offset leaves the target about 1 °F too high at firm ball and about 3 °F too high at hard crack, and the syrup ends slightly more concentrated than intended. To calibrate the thermometer itself: a rolling boil that does not stop with stirring, in at least four inches of water, probe a few inches deep in the center and off the bottom, and compare to the local boiling point (ThermoWorks, 2025).

Crystal or glass: the doctoring agents

A cooked syrup is supersaturated the moment it starts to cool, and what it does next is the whole of confectionery. The sucrose solubility curve, the glass-transition curve, and the metastable and labile zones between them make up the state diagram on which Hartel’s group places every confection (Hartel et al., 2011). Fondant and fudge are engineered to crystallize small and uniform. Hard candy, cotton candy, and hard toffees are engineered not to crystallize at all, cooled so fast through the labile zone that the sugar sets as an amorphous glass. Caramels and gummies sit in the rubbery state between. The craft is in the confectionery article; the figures are here.

Glucose syrup inhibits sucrose crystallization two ways. Its longer saccharides adsorb onto crystal faces and block growth, and because it replaces sucrose, it lowers the sucrose supersaturation at any given water content (Hartel et al., 2011; O’Connell & Hartel, 2022). For scale, by this article’s own arithmetic: at 12 percent water an all-sucrose syrup carries about 730 g of sucrose per 100 g water, more than three times the roughly 200 g that dissolves at room temperature, and every part replaced by corn syrup cuts that excess. In amorphous sucrose, 10 to 20 percent corn-syrup solids slowed crystallization and 50 percent or more prevented it (Gabarra & Hartel, 1998), and thin-film measurements show sucrose crystal growth slowing as the corn-syrup fraction rises (Chen, Nowakowski, Green, & Hartel, 2015). Invert sugar, whether added or formed in the pot by acid, does the same job as a second solute, and it holds moisture (FAO, n.d.; Ergun et al., 2010). The confectionery rule is that about 10 to 15 percent invert sugar in the finished sugar is what “is required to give a non-crystalline product” (FAO, n.d.; Wolf, 2016).

The sharpest published thresholds come from Hartel’s group. In starch jellies the time to graining falls sharply as the sucrose-to-corn-syrup ratio rises, and ratios above 3 crystallize almost immediately (O’Connell & Hartel, 2022); in caramel, more sucrose relative to corn syrup means more crystals (Miller & Hartel, 2015). Nougat follows the same logic: it is made both grained and ungrained, and more sucrose relative to corn syrup favors graining (Ergun et al., 2010). Cotton candy carries no corn syrup and relies on fast cooling alone, and with an enormous surface area it is among the least stable sugar confections (Ergun et al., 2010; Labuza & Labuza, 2004).

Published formulations, for scale rather than as recipes: hard candy at 48 percent sucrose, 32 percent glucose syrup and 20 percent water (Wolf, 2016), cooked to 150 to 166 °C (FAO, n.d.), which puts glucose syrup at about two-thirds of the sucrose weight, or about half on a solids basis, and Ergun et al. (2010) give 70:30 sugar to corn syrup as a common basic hard-candy ratio; citric acid in hard candy at 0.1 to 1.0 percent, pH 3 to 4, target moisture 2 to 3 percent and never above 5 (Ozel et al., 2024). Fondant is cooked to about 118 to 120 °C and cooled without agitation to 40 to 50 °C before beating, so that crystallization starts all at once (Ergun et al., 2010); one industrial formula, 62 percent sucrose, 16 percent glucose and 22 percent water, boils to 88 percent solids at 117 °C and beats at 37 °C (Wolf, 2016); another is sucrose 100, water 50, glucose 25, cooked to 120 °C and beaten at 38 to 43 °C (Talbot, 2009). Pulled sugar at 1 kg sugar, 350 to 400 g water, about 200 g glucose, a dozen drops of tartaric acid at 155 °C, cooked to 160 °C (Torreblanca, 2017). Why fondant is cooled before it is beaten: cooled undisturbed and then beaten, the supersaturated syrup nucleates all at once into many small crystals (Ergun et al., 2010), and the crystallization times under agitation follow classical nucleation theory (Hartge, Flöter, & Vilgis, 2023).

Dextrose equivalent. DE is the degree of starch hydrolysis expressed as reducing sugars, dry basis, with pure dextrose at 100; below 20 DE the product is a starch hydrolysate (maltodextrin) rather than a glucose syrup, and Codex defines glucose syrup as 20 DE or above at 70 percent solids or more (BAKERpedia, n.d.-b). The commonest glucose syrups in gummy and jelly formulations are 63 and 42 DE (Pekdogan Goztok et al., 2024). The higher the DE, the smaller the average molecule: more sweetness, more hygroscopicity, a lower glass transition (42 DE solids about 79 °C, 20 DE about 139 °C), and more boiling-point elevation at the same solids (Ergun et al., 2010; Hartel et al., 2011). High-maltose syrups are no cure for stickiness: they give a slightly higher glass transition at a given moisture (Nowakowski & Hartel, 2002), but manufacturers who have tried them report candy that turns too hard and brittle (Ergun et al., 2010). “Corn syrup” is glucose syrup named for its maize starch (BAKERpedia, n.d.-b).

Inversion. Sucrose plus water under acid and heat gives glucose and fructose, and cream of tartar or lemon juice in the pot is the traditional doctor that predates glucose syrup. How much a home boil inverts is not well quantified: technical guidance gives the target, about 10 to 15 percent invert for a non-crystalline product (FAO, n.d.), and commercial hard candies have been measured at 0.2 to 8.5 percent fructose (Ergun et al., 2010). Invert is more humectant than sucrose, and its parts have very low glass transitions, fructose about 5 °C and glucose 31 °C against sucrose at 62 to 70 °C, so the water an over-inverted hard candy draws from humid air turns its surface sticky (Ergun et al., 2010; Ozel et al., 2024). That is why acid goes into hard candy at no more than about 1 percent, and only at the end of the cook (Ozel et al., 2024). Even salt added during the boil contributes to inversion and makes the batch softer (Greweling, 2013).

Water, the glass, and why hard candy goes sticky

The cook temperature sets the water content; the water content sets the water activity and the glass-transition temperature; those two numbers decide texture and shelf life (Ergun et al., 2010). Hard candy finishes at 2 to 5 percent water and a water activity of 0.25 to 0.40; caramel, toffee, and fudge at 6 to 18 percent and 0.45 to 0.60; gums and jellies at 8 to 22 percent and 0.50 to 0.75; marshmallow at 12 to 20 percent and 0.60 to 0.75; fondant and creams at 10 to 18 percent and 0.65 to 0.80 (Ergun et al., 2010). Measured commercial pieces run the same way: a butter-rum hard candy at 0.30, chewy caramels 0.45, translucent gummy bears 0.70 (AQUALAB, n.d.).

Amorphous sucrose has a glass transition of about 57 °C dry in the classic measurement (Roos & Karel, 1991a), with tabulated values of 62 to 70 °C (Ergun et al., 2010), and water plasticizes it hard: 69 °C at zero water falls to about 29 °C at 5 percent (Ergun et al., 2010), which puts the glass transition at room temperature at roughly 5.5 to 6 percent water. Dry sucrose and corn-syrup glasses under 1 percent water sit near 64 to 71 °C (Gabarra & Hartel, 1998), while commercial hard candies at the 2 to 5 percent water they actually carry average 35 to 40 °C (Ergun et al., 2010). Higher-DP glucose-syrup solids raise the glass transition of the blend and monosaccharides lower it (Roos & Karel, 1991b; Gabarra & Hartel, 1998).

A sugar glass is hygroscopic. Its surface takes up water from the air until the local glass transition falls below room temperature; the surface goes rubbery and tacky, then flows, then grains as the mobile sugar crystallizes (Nowakowski & Hartel, 2002; Ozel et al., 2024). The rate of sticking is a function of the gap between temperature and glass transition rather than of temperature or humidity alone (Foster, Bronlund, & Paterson, 2006), and tack peaks at about 11 to 13 percent water before falling again as the syrup becomes a liquid (Burke & Hartel, 2021). Because water activity is the equilibrium relative humidity divided by 100, a hard candy at 0.25 to 0.40 picks up moisture in any air above about 25 to 40 percent relative humidity, which is nearly all kitchen air, and turns sticky once its surface has absorbed enough to drop its glass transition to room temperature (Ergun et al., 2010). Cotton candy stayed glassy for at least a year at 11 percent relative humidity and collapsed and crystallized within three days at 33 percent (Labuza & Labuza, 2004). Hence the practice: sugar work goes into an airtight box with silica gel the moment it is cool, spun sugar is assembled and served within about two hours, and a humid day is a reason to postpone pulled sugar (Domson Academy, 2026; Torreblanca, 2017).

Isomalt is the showpiece sugar because it takes up almost no water below about 85 percent relative humidity at 25 °C; it is about half as sweet as sucrose and only about 25 percent soluble at room temperature (McNutt & Sentko, 2003; Beneo, n.d.). Sources put its cooking temperature at 160 to 170 °C (Ozel et al., 2024; Domson Academy, 2026), with the melt poured or worked from about 135 °C (Domson Academy, 2026).

Caramelization

Above the stages the syrup stops boiling and starts to break down. Sucrose loses water and rearranges into several thousand compounds: dehydration products, oligomers of up to six sugar units from condensation, and their dehydrated derivatives; chemically, caramel is dominated by that oligomeric material rather than by the aroma volatiles people notice first (Golon & Kuhnert, 2012). The glass transition of the product falls at first, as small degradation products plasticize it, then rises as polymerized caramel forms, (Jiang, Liu, Bhandari, & Zhou, 2008), which would make a pale caramel stickier than a dark one and overcooked sugar set harder.

A worked correction

A recipe cooks caramels to 248 °F at sea level and the kitchen is at 5,280 ft. Water boils there at 202.4 °F on a standard day, where the linear rule would say 201.4. The offset is 212 minus 202.4, or 9.6 °F, so the target is 238 °F; on a day the barometer is 3 kPa low, water boils nearer 200.5 and the target is about 237. Boiling water first and subtracting the measured difference takes both into account and corrects the thermometer at the same time. The same syrup read by refractometer, before any dairy goes in, needs no correction at all: the 248 °F syrup is about 90 percent solids at any elevation, and a cooled sample read at 20 °C says so directly, with the caveat that a syrup carrying glucose reads as equivalent sucrose rather than true solids.