Chapter 10 — Sugars and Caramelization: Candy, Caramel, and the Physics of Sugar Crystals

The Hook

The first time Maya Okonkwo tried to make caramel, she ruined it twice in five minutes.

She had read three recipes. They each said roughly the same thing — heat sugar and water in a heavy saucepan, do not stir, watch for the color to turn amber. So she heated sugar and water in a heavy saucepan. She did not stir. She watched. Nothing happened, then nothing happened, then nothing happened, then — the moment her partner Aisha asked a small question from the next room — the syrup went from clear to the color of a maple leaf in the span of about eight seconds. Maya, distracted, did not turn off the heat fast enough. Within ten more seconds the caramel was the color of motor oil and smelled, on close inspection, like a parking lot in August. She poured it down the drain and started over.

Attempt two went better through the early phases — clear, faint yellow, light gold — and then plateaued at light gold for what felt like minutes. Maya, doubting, turned the heat up. The caramel obliged: from light gold it went to amber, from amber to mahogany, from mahogany to too dark, and from too dark to smoking. Maya turned off the heat and pulled the pan from the burner and dropped in a knob of butter and a glug of cream, the way you do, and the caramel hissed and steamed and integrated into a sauce that was, when she tasted it cooled, bitter beyond redemption. The caramel had carbonized. There was no flavor of sugar left in it, only smoke.

She turned to Aisha. "I think," she said, "this is harder than it looks."

It is harder than it looks. Caramel is one of the highest-stakes pieces of cooking in any kitchen, because the temperature window between delicious and destroyed is narrow — about twenty Celsius degrees, traversed in under a minute, with very few visual cues until you've already missed your mark. The chemistry is uncompromising and time-sensitive in a way most kitchen reactions are not.

This chapter is about sugar — the most familiar food molecule on earth, the one a five-year-old can name — and about the suite of reactions that happen when you heat it. By the end, Maya will have made caramel correctly twice in a row. You will have done the same. And both of you will have learned a chemistry of crystals, supersaturation, and breakdown reactions that connects everything from a piece of brittle to a square of dark chocolate to the bottom of a crème brûlée.

We start with the simplest sugar in your kitchen, and what happens to it when it gets hot.

The Everyday Observation: A Tour of Sugar at Different Temperatures

You have probably watched, without naming it, the entire candy-temperature ladder. Consider:

• Sugar in coffee. Sugar dissolves in hot liquid. The tea or coffee sweetens. The molecules disperse. The temperature is well below boiling for sugar; nothing transformative happens.
• Sugar syrup for lemonade. Equal parts sugar and water heated to dissolution. The syrup is clear, sweet, and pourable. It has been brought to maybe 100°C briefly; after cooling, it is room-temperature simple syrup.
• A boiling pot of sugar-water for canning. Sustained boiling at 100–105°C drives off water; the sugar concentration rises; the syrup thickens. This is the territory of jam-making and fruit preservation.
• Soft-ball candy. A boiling sugar solution at 115°C produces a syrup that, when dropped into cold water, forms a soft, pliable ball. This is the temperature for fudge and pralines.
• Hard-crack candy. A boiling sugar solution at 149°C produces a syrup that, when dropped into cold water, immediately solidifies into brittle threads. This is the temperature for lollipops and brittle.
• Caramel. A boiling sugar solution at 170°C — most of the water has gone — turns amber-brown and develops complex flavor compounds. This is the territory of crème brûlée tops, caramel sauces, and, soon, Maya's small kitchen catastrophe.
• Burned sugar. Above 180°C, the caramelization that produced flavor at 170° gives way to thermal decomposition that produces bitter, acrid compounds with no edible value. The line between deepest amber and spent and bitter is narrow, and the cook has to find it.

This is the candy temperature ladder. Cooks have known it for two centuries; thermometers have made it usable for ordinary kitchens for about a hundred and fifty years. The ladder works because as you heat a sugar solution, water evaporates and sugar concentration rises — the temperature reads the concentration, not just the heat. Each rung on the ladder corresponds to a specific water-to-sugar ratio, which corresponds to a specific texture in the cooled candy.

To understand why, we have to start with the sugar molecule itself.

The Science: From Sucrose to Caramel

Sucrose: the world's most-traded chemical

The white sugar in your sugar bowl is sucrose — a disaccharide composed of one glucose unit linked to one fructose unit by a glycosidic bond. Same kind of bond we discussed in Chapter 9 (it links the glucose units of starch), but here connecting two different sugars rather than thousands of identical ones.

Sucrose is the most-traded chemical in the human food system. It is grown industrially from two main sources: sugarcane (a tropical grass, native to New Guinea, cultivated for sugar in India by 500 BCE, in Persia by 500 CE, in the Americas after Columbus, and at devastating human cost on Caribbean and Brazilian plantations from the sixteenth through nineteenth centuries) and sugar beets (a temperate root vegetable whose sugar content was developed in the early 1800s and which now provides about 20% of world sugar). The two sources produce chemically identical sucrose. After processing, you cannot distinguish cane sugar from beet sugar by chemical analysis. Sucrose is sucrose.

The molecule looks, in two dimensions, like two hexagonal rings (one slightly puckered, one a five-membered ring instead of six) connected by a single oxygen bridge. Pure sucrose crystallizes into hard, transparent prisms — the table-sugar grains you know — and dissolves freely in water, especially hot water. A liter of water at 100°C will dissolve nearly five kilograms of sucrose.

📜 Tradition / History. The economic and human history of sugar is unavoidable. The Atlantic slave trade was, in significant part, organized to produce sugar for European markets; sugar plantations in Brazil, Saint-Domingue (now Haiti), Cuba, and Jamaica killed millions of enslaved people over centuries. The white crystals on a coffee-shop table have a heavier history than most ingredients in this book. Cooks who care about food's social meaning may want to investigate Sidney Mintz's Sweetness and Power (1985), which traces sugar's transition from rare luxury to ubiquitous commodity. The chemistry of sucrose is universal; the politics of how it gets to your kitchen is not.

Inversion: the hydrolysis that splits sucrose

The glycosidic bond in sucrose is breakable. Heat plus acid will hydrolyze it, splitting sucrose into its component monosaccharides — free glucose and free fructose. The resulting mixture is called invert sugar.

Why "invert"? Because of an optical property called optical activity, which we'll meet in the Advanced Sidebar below. The short version: pure sucrose rotates polarized light to the right; the mixture of glucose plus fructose rotates polarized light to the left. The direction of rotation literally inverts. The naming has stuck.

Why does this matter in the kitchen? For three reasons.

First, invert sugar is sweeter than sucrose. Glucose alone is about two-thirds as sweet as sucrose; fructose is about 1.7 times as sweet. The mixture of equal parts glucose and fructose is, on net, sweeter than the sucrose you started with — about 1.3 times as sweet, depending on conditions. A cup of inverted sugar will taste sweeter than a cup of un-inverted sugar.

Second, invert sugar resists crystallization. This is enormously useful in candy-making. In a saturated sugar solution, the danger is that any seed crystal — a stray sucrose grain on the side of the pot, a crystal forming on a stirring spoon — will trigger sudden, runaway crystallization, turning a smooth syrup into a grainy mass within seconds. By inverting some of the sucrose to free glucose and fructose, candy-makers create a mixture in which sucrose crystals can't easily form, because the free monosaccharides physically interfere with sucrose-to-sucrose alignment. Adding a tablespoon of corn syrup, lemon juice, or cream of tartar to a sugar syrup is a deliberate inversion strategy.

Third, invert sugar holds water. Glucose and fructose are hygroscopic — they pull moisture from the air. A candy made with invert sugar stays soft longer than one made with pure sucrose. Honey, which is a natural invert-sugar mixture (more on this below), famously almost never dries out completely.

🔬 Advanced Sidebar — The optical activity of sugars

For students of molecular chemistry: sugars exist in two mirror-image forms, called enantiomers. The natural sugar your body eats and your kitchen uses is the D-form (dextrorotatory in pure crystal form: rotates plane-polarized light clockwise). The mirror image is called L-glucose, and it is, biologically, almost completely inert — your enzymes don't recognize it, your taste receptors barely register it, and it passes through the digestive tract unmetabolized. (L-glucose has been studied as a non-caloric sweetener, but its production cost makes it commercially impractical.)

The optical activity matters because it provides a precise, real-time analytical method. By measuring the rotation of polarized light through a sugar solution, you can quantify both the concentration and the identity. In the early development of sucrose chemistry, this was the primary method for confirming the inversion of cane syrup. The sugar industry still uses polarimetry in quality control today.

A sucrose solution rotates polarized light by +66.5° (clockwise) per decimeter at standard concentration. After inversion, the same solution — now glucose + fructose — rotates light by −20° (counterclockwise). The flip from positive to negative is the literal "inversion." The reaction was named for the property that exposes it.

The Advanced Sidebar exists for the food science student or chemistry teacher who wants to anchor sugar inversion to a fundamental molecular property; for the home cook, the practical takeaway is that lemon juice, cream of tartar, or corn syrup added to a sugar syrup will invert some of the sucrose and prevent crystallization.

— Back to the main thread.

The candy-temperature ladder, in detail

Cooks have measured the candy ladder for a long time, but the physical reason it works — the link between temperature and water content — only became clear in the late 1800s with the development of the boiling-point elevation theory.

The principle: when a non-volatile solute (like sugar) is dissolved in a volatile solvent (like water), the boiling point of the solution rises above that of the pure solvent. The more solute, the higher the boiling point. As you boil a sugar solution, water leaves first (as steam); the remaining solution is more concentrated; its boiling point rises. The temperature of the boiling solution is therefore an indirect measurement of the sugar concentration.

Modern candy-makers express the concentration in degrees Brix — grams of sugar per 100 grams of solution. The relationship between Brix and boiling point is roughly:

Some traditional candy-making texts use only the Fahrenheit scale; some only Celsius. We give both because cooking in the modern kitchen is increasingly bilingual. Buy a thermometer that reads both. Most modern candy thermometers do.

🍳 Kitchen Lab — The cold-water test for candy stages (inline tease).

Even with a thermometer, the traditional cold-water test is worth knowing — it reads the cooled texture of the syrup directly, which is what your final candy will feel like.

Have a small glass of cold (not iced) water nearby. As your sugar syrup approaches a target temperature, drop ½ teaspoon (2.5 mL) of the syrup into the cold water. With clean fingers, retrieve the cooled glob and squeeze it.

• Soft-ball stage: the glob forms a soft, pliable ball that flattens between your fingers.
• Firm-ball: ball holds its shape but yields to pressure.
• Hard-ball: ball is firm, chewy, like a piece of taffy.
• Soft-crack: glob forms threads that bend before breaking.
• Hard-crack: glob shatters into brittle shards immediately.

The test takes a second and confirms what your thermometer is telling you. ⚠️ Severe burn hazard — sugar syrup is at 115–150°C and sticks to skin. Do not use bare hands until the syrup is fully cooled in the water. Full protocol in exercises.md.

Caramelization: the breakdown reaction

We are now ready for the chapter's central reaction.

When you continue heating a sugar solution past the hard-crack stage, all the water eventually leaves. You have, by now, just dry sucrose at 160°C and rising. Sucrose is a solid at room temperature but becomes a viscous liquid as it heats; at around 160°C the crystalline structure collapses entirely and you have molten sucrose, a clear, slightly yellowish, viscous liquid.

Above approximately 160–170°C, the molten sucrose begins to undergo caramelization — a complex thermal decomposition of the sugar molecule itself. The sucrose molecule first hydrolyzes into glucose and fructose (the same inversion we saw above, only this time driven by heat alone, no acid required). Each of those monosaccharides then breaks down further through a tangled cascade of reactions: dehydration (loss of water), enolization (rearrangement of double bonds), polymerization (re-joining of fragments into longer molecules), and oxidation. The result is a chemical zoo of new compounds: furans (the source of caramelized apple, baked bread, and roasted coffee aromas), diacetyl (buttery), maltol (cotton candy), and at higher temperatures, color-bearing molecules called caramelins, caramelans, and caramelens — long-chain polymers responsible for the deep brown color of caramel.

Here is the part that people most often confuse. Caramelization is distinct from the Maillard reaction we met in Chapter 8. The Maillard reaction requires an amino acid plus a reducing sugar plus heat. Caramelization requires only sugar plus heat — no protein needed. Caramelization happens when you cook pure sugar; Maillard happens when you cook anything with both sugar and protein together.

In a typical kitchen, the two reactions often run side by side. The crust of a roasted vegetable involves both: the sugars in the vegetable caramelize, and the proteins in it react with the sugars in Maillard. In a pure caramel — sugar and water, nothing else — you are seeing caramelization alone. In a piece of seared meat, you are seeing mostly Maillard with a small contribution from caramelization. Once you can distinguish the two, you can predict why a roast tastes savory and a caramel tastes sweet-bitter; they are different chemistry.

The five stages of caramel

Watching a sugar syrup go through caramelization is one of the great visual experiences in cooking. The stages, in order:

1. Clear melt. Pure sucrose, just past the hard-crack stage, has lost all its water and is now molten. The liquid is clear, slightly yellow-tinged, and at about 160°C. Almost no caramel flavor yet — taste at this stage and you get something like very intense sweetness with a faint cooked-sugar note.

2. Light golden. 165–170°C. The first color compounds form. The flavor begins to shift from purely sweet toward something more complex — mild butter, very faint toasted-grain. This is where pastry chefs sometimes pull caramel for blond caramel — used in caramel-flavored macarons, light caramel sauces, and some confections that want sweetness without bitter complexity.

3. Amber. 170–180°C. Color deepens to honey-amber to medium-mahogany. Flavor becomes the iconic "caramel" — sweet, slightly nutty, with butter and toasted notes. This is the stage you want for most caramel sauces, candy bottoms, and the top of a crème brûlée. Aroma fills the kitchen.

4. Dark amber / mahogany. 180–190°C. Color goes red-brown to deep mahogany. The flavor balance shifts: less sweetness (sugar molecules are now actually being destroyed; less remains intact), more bitter, more toasted. Some pastry chefs prize this stage for caramel sauces that need to balance against sweet desserts (a too-sweet caramel sauce on a sweet dessert is cloying; a darkly-roasted caramel sauce balances it).

5. Burnt. Above 190°C. Color becomes nearly black. Flavor turns aggressively bitter, smoky, and acrid. The reactions have moved past flavor compounds into thermal decomposition products that have negative flavor value. Recovery is impossible — pour it out and start over.

The transition from amber to burnt is fast, often less than thirty seconds at typical home-stove heat. Maya's first attempt failed at this transition; her second attempt also did. Most home cooks have a similar story.

🧪 Threshold Concept — Caramelization is a chemical transformation, not a phase change. When you melt sucrose and watch it turn brown, you are not just changing its physical state. You are chemically destroying it. The sugar molecules are breaking apart, rearranging, polymerizing, oxidizing. The molten amber liquid in your saucepan has fewer intact sucrose molecules than the dry crystals you started with — and contains a hundred new compounds that didn't exist before. Caramel is the smell of sucrose being destroyed elegantly. Once you grasp that, the sensitivity to time and temperature makes sense: you are managing a degradation reaction, not a melting process. There is no "perfect equilibrium" of caramel; there is only the point at which you stop the destruction.

Sugar crystallization: fudge versus lollipops

Now let's loop back to crystals. Sucrose, dissolved in hot water, is a transparent solution. Cool that solution past its saturation point and the dissolved sucrose wants to come out of solution and re-crystallize into solid grains. How it re-crystallizes determines whether you get fudge or a lollipop.

Fudge is microcrystalline: a vast number of very small, almost-invisible sucrose crystals suspended in a slightly-supersaturated syrup. The mouthfeel is soft and creamy because each crystal is too small for your tongue to register; the overall texture is a smooth paste of tiny crystals embedded in liquid. Bite into fudge and the small crystals dissolve smoothly on your tongue. You want microcrystals.

Lollipops are amorphous: no crystals at all. A lollipop is glassy — a frozen, glass-like state of supersaturated sugar solution that has cooled too fast for crystals to form. The mouthfeel is hard and brittle; the lollipop dissolves slowly and uniformly. You want no crystals at all — pure glass.

The difference between these outcomes is how, and when, you let crystallization happen.

For fudge: you cook the sugar solution to the soft-ball stage (~115°C / 240°F). You then cool the syrup, undisturbed, to about 110°F (43°C). At this temperature the syrup is supersaturated and waiting to crystallize, but no crystals have formed yet — there are no nuclei for them to grow on. Now you start beating. Beating introduces millions of tiny seed crystals (some from the air, some from the friction of the spoon, some from microscopic sugar grains on the side of the pot). Each seed crystal becomes a nucleus around which more sucrose deposits. Because there are millions of seeds, no single crystal grows large. You end up with millions of tiny crystals — microcrystalline texture. Smooth, creamy fudge.

If you start beating too soon — while the syrup is still hot and only slightly supersaturated — too few crystals form, and each one grows large. You get coarse, gritty fudge.

If you start beating too late — after the syrup has begun to spontaneously crystallize on its own — uneven crystal sizes form. Same problem.

For lollipops: you cook the sugar solution to the hard-crack stage (~149°C / 300°F). You then pour it onto a cold surface or a candy mold, where it cools rapidly to room temperature. The syrup is extremely supersaturated at room temperature, but it cooled so fast that the sugar molecules didn't have time to organize into crystals. The result is a glass — molecules locked in a disordered, frozen-in-place arrangement. You wanted no crystals, and you got none, because you cooled too fast for crystallization to start.

If a single seed crystal sneaks into the lollipop syrup — say, a stray sucrose grain on the side of the pot — the entire lollipop will bloom into crystals over the next hour or day. The result is a cloudy, gritty lollipop instead of a clear, glassy one. This is why hard-candy recipes obsessively warn against any un-dissolved sugar grains during cooking.

The role of corn syrup

Corn syrup is mostly glucose — produced by enzymatically breaking corn starch (Chapter 9!) into its component glucose units. As we discussed under "inversion," free glucose and fructose interfere with sucrose crystallization. Adding even a small amount of corn syrup (or honey, or maple syrup, or any glucose-containing additive) to a sucrose-based candy interferes with crystal formation.

For lollipops: glucose acts as an anti-crystallization agent. The corn syrup keeps the sugar in its glassy, amorphous state and prevents the dreaded post-cooling crystallization bloom. Hard-candy recipes typically call for ¼ to ⅓ cup (60–80 mL) of corn syrup per cup of sugar.

For fudge: glucose moderates crystallization, encouraging the formation of microcrystals rather than coarse crystals. Fudge recipes typically include some corn syrup or marshmallow cream (which is mostly invert sugar) for this reason.

For caramel sauces: glucose in the form of corn syrup keeps the sauce smooth on cooling. A pure-sucrose caramel sauce can crystallize on the side of the jar overnight. Adding corn syrup prevents this.

Brown sugar, honey, and the sweetener panel

Several things on the supermarket shelf are not pure sucrose. A short tour:

Brown sugar is white sugar plus molasses. Molasses is the dark, mineral-rich syrup left over after sucrose is crystallized out of cane juice; it contains some unrefined sucrose, plus invert sugar, plus minerals (iron, calcium, potassium), plus the polymeric brown compounds that give it color. Brown sugar exists on a spectrum (light brown = ~3% molasses, dark brown = ~6% molasses). The molasses contributes flavor and moisture-holding capacity; brown sugar baked goods are softer and stay fresh longer than white-sugar equivalents.

Honey is essentially natural invert sugar, produced by honeybees. Bee enzymes (specifically invertase, secreted from bee salivary glands) hydrolyze the nectar's sucrose into glucose and fructose. The result is roughly 80% sugar (about half glucose, half fructose, plus some sucrose, maltose, and other minor sugars) and 17% water. Plus trace amounts of antimicrobial compounds (hydrogen peroxide produced by another bee enzyme, plus phenolic compounds from the source flowers, plus low pH and low water activity), which is why honey doesn't spoil — a topic we'll preview here and explore in Chapter 36.

📜 Tradition / History — Honey, the eternal sugar. Archaeologists have recovered intact, edible honey from sealed Egyptian tombs over 3,000 years old. The chemistry: honey's water activity is so low (around 0.6) that microbes cannot grow in it; the small amount of hydrogen peroxide generated by glucose oxidase suppresses bacteria; the low pH (~3.5) further inhibits microbial growth. Indigenous beekeeping traditions on every inhabited continent independently discovered honey's preservation properties. Mayan, Egyptian, Greek, Chinese, and Polynesian sources all describe honey-based food preservation centuries before refrigeration.

Honey will, over months or years, crystallize spontaneously. This is not spoilage — it is the supersaturated glucose finding seed crystals (often around small pollen particles or wax fragments) and slowly nucleating. Honey crystallizes faster in a refrigerator than at room temperature — the same kinetic principle that drives starch retrogradation (Chapter 9) drives honey crystallization. To re-liquefy crystallized honey, warm it in a water bath; the crystals re-dissolve.

Maple syrup is concentrated tree sap, mostly sucrose with a small amount of invert sugar plus complex flavor compounds developed during processing (some Maillard during boiling-down). The flavor is distinctive because of the trace chemistry, not because of any unusual sugar.

Agave syrup is processed from the agave plant native to Mexico and the American Southwest. Most commercial agave syrup is high in fructose (around 60–90%, depending on processing), which makes it sweeter per gram than sucrose and gives it a different glycemic response.

HFCS (high-fructose corn syrup) is industrially-processed corn syrup converted to a roughly 50/50 mixture of glucose and fructose. Chemically, it is very similar to invert sugar and to honey. (HFCS-55, the version used in most American soft drinks, is about 55% fructose and 42% glucose.) The metabolic and health questions around HFCS are an active area of nutrition research with genuine ongoing debate. Chemically, HFCS is not particularly different from any other invert sugar. Whether you find it desirable in your diet is a separate question from the chemistry of how it behaves in cooking.

The honest sweetener-panel summary: all of these are sugars, all are sweet, all give you calories, all will spike your blood glucose to varying degrees. The differences in flavor and texture are real; the differences in health impact are subtler than marketing claims suggest. Beware of any source that frames one sweetener as virtuous and another as villainous; the actual differences are quantitative, not categorical.

🌍 Cultural Note — Aroon Sornprasit's palm sugar caramel.

Chef Aroon Sornprasit, who runs Mae Som in Toronto, makes one of the great caramels of any Thai kitchen. He doesn't use white sugar. He uses namtan peep — palm sugar pressed into round cakes from the sap of the sugar palm tree. Palm sugar is naturally a partial invert sugar (the palm sap is processed slowly with traditional methods that allow some hydrolysis), and the cakes are aromatic with the slightly funky, deeply caramelized notes that come from the slow open-air evaporation. When Aroon makes Thai caramel for a coconut-cream dessert, the palm sugar melts down at a slightly lower temperature than pure sucrose, develops color faster, and produces a flavor that white sugar caramel cannot reproduce — earthier, more complex, with hints that read almost as cocoa or molasses. He says his grandmother used palm sugar exclusively. "White sugar," she told him, "is a lie. Palm sugar tells the truth." Aroon, who knows the chemistry well enough to understand both sugars, smiles when he tells this story. He uses both. They are, he says, different tools for different desserts. The chemistry is genuinely different. The cultural attachment is real. Both are honored in his kitchen.

The Practical Application: Making Caramel Without Burning Your Kitchen Down

Now to land all this in your hands.

The two methods of making caramel

There are two ways to make caramel sauce, and they have different success profiles for different cooks.

Wet method. Sugar plus a small amount of water, heated together. The water dissolves the sugar first, creating a syrup; the syrup boils; water leaves; sugar concentration rises; eventually the syrup goes through the candy-temperature ladder and into caramelization. The wet method is more forgiving for beginners because the early stages are slow and you have plenty of time to monitor temperature. The danger is premature crystallization: a stray sugar crystal on the side of the pot can trigger a chain reaction that turns your syrup into a grainy mass instead of a smooth caramel. To prevent this, traditional recipes call for: not stirring once the syrup is at a boil; covering the pot for the first few minutes of boiling so condensation washes the sides clean; or adding a small amount of acid (lemon juice, cream of tartar) or corn syrup to invert some of the sucrose and resist crystallization.

Dry method. Sugar alone in a heavy pan, no water. The sugar melts directly from solid to liquid at around 160°C. The dry method is faster and produces a more intense caramel flavor because there's no water dilution. The danger is uneven heating — the sugar at the edges of the pan melts and starts caramelizing while the sugar in the center is still solid. Stirring helps but introduces its own problems (the half-melted sugar can re-crystallize on a cold spoon). Most cooks tilt and rotate the pan instead of stirring, watching the molten sugar spread inward as the dry sugar melts. The dry method requires more vigilance but rewards it.

Why fudge is the hardest candy

If you ask any pastry chef which classic American candy is the trickiest to make well, almost all of them will say fudge. Not because the chemistry is unusual, but because fudge demands precise control over crystallization in a way most candies don't.

Recall the fudge problem: you cook a sugar syrup to soft-ball stage (~115°C / 240°F), then cool it to about 110°F (43°C) without disturbing it, and then beat it vigorously to induce microcrystallization. The window for success is narrow on multiple axes:

Cook to the wrong temperature and you get the wrong final texture. Soft-ball at 113°C gives soft fudge; soft-ball at 117°C gives firm fudge. Within 2°C, very different mouthfeels.

Disturb the syrup while it's cooling and crystallization begins prematurely, in a relatively unsupersaturated state, producing few large crystals (gritty fudge). The cooling phase is therefore deliberately quiet — no stirring, no spoon-tapping, no transfer to a different bowl, often just leaving the pot of syrup on a cool burner with the lid off, undisturbed, for fifteen minutes.

Start beating too soon and the syrup is still hot enough that any crystals that form grow rapidly, creating coarse texture. Beating before the syrup reaches around 110°F (43°C) is a common cause of failed fudge.

Start beating too late and ambient nucleation has already begun (a stray vibration, a temperature variance) and uneven, irregular crystals have formed. Beating now homogenizes them but doesn't produce the fine microcrystalline structure that smooth fudge requires.

Beat too gently and not enough nucleation events occur; large crystals predominate.

Beat too violently and you incorporate excessive air, producing a foamier, lighter texture (sometimes called "divinity" in Southern traditions, technically a different candy but close in chemistry).

For all these reasons, traditional fudge recipes call for a very specific protocol: dissolve, boil, soft-ball, cool undisturbed, test for warmth with the back of your hand to a saucepan side, beat for a specific number of minutes (typically 5-10 minutes of vigorous spoon-beating), and pour into a buttered pan to set. Many home bakers report inconsistent results not because they did anything obviously wrong, but because the cooling and beating phases are sensitive to variables they don't control (room temperature, humidity, the exact heat distribution in their burner during the cook phase).

Modern fudge recipes increasingly use marshmallow cream as a cheat: marshmallow contributes invert sugar (corn syrup + egg whites) which interferes with crystallization and produces consistent microcrystals without requiring the precise cool-and-beat protocol. The result is more reliable but somewhat less elegant — most professional confectioners still hold that classical-method fudge has a superior texture.

For science teachers: fudge-making is one of the most accessible classroom demonstrations of crystal nucleation and growth in physical chemistry. The same principles apply to industrial crystallization (purifying salt, sugar, pharmaceuticals) — and in the kitchen, you can see them happen in twenty minutes.

Maya's troubleshooting guide

After her two failures, Maya re-read this chapter (well, an earlier draft of it) and made caramel a third time. Here is the troubleshooting guide she developed, honed over the next ten or so attempts:

Problem: Sugar suddenly turned grainy. - Cause: A seed crystal triggered runaway crystallization. - Prevention: Cover the pot for the first 2–3 minutes of boiling (steam washes the sides). Add 1 tbsp corn syrup or ½ tsp lemon juice per cup of sugar. Don't stir once the sugar is dissolved. - Recovery: Difficult. Sometimes returning the pot to heat with a splash of water and re-dissolving will work. Often the safest move is to start over.

Problem: Caramel is too pale; flavor is just sweet, not "caramel." - Cause: Pulled too early. Reached only about 165–168°C / 329–334°F. - Prevention: Wait for amber color. Use a thermometer. - Recovery: Return to heat carefully, watching closely.

Problem: Caramel is bitter. - Cause: Cooked too far past the amber stage. Reached 185°C+ / 365°F+. - Prevention: Remove from heat at the first sign of mahogany color, not at the second. Have the next ingredients ready in advance so you can interrupt cooking immediately. - Recovery: None. Bitter caramel cannot be saved.

Problem: Caramel hardens like a rock as it cools. - Cause: This is normal — pure caramel cools to a hard glass. Most caramel-sauce recipes call for adding cream and butter immediately after the caramel reaches color, which both stops the cooking and adds the fat and water content that keeps the sauce pourable when cooled. - Recovery: Re-warm with a splash of cream or water; the caramel will redissolve.

Problem: Adding cream caused the caramel to seize / clump. - Cause: Cold cream hit hot caramel; thermal shock crystallized some of the sugar. - Prevention: Warm the cream slightly before adding it. Pour slowly while whisking. Continue cooking briefly to re-dissolve any clumps.

Pat Hammond's classroom adaptation

Pat doesn't make caramel in her chemistry class. The hot-sugar burn risk is too high for a room of thirty 16-year-olds. But she does demonstrate the wet method in the back of the room with a small audience, while the rest of the class works on a related calorimetry exercise. Her notes for safety:

⚠️ Sugar burns are among the worst kitchen burns possible. Sugar at 170°C is hotter than boiling oil, and unlike oil, it sticks to skin and continues to burn. A sugar splash to skin can produce a third-degree burn within seconds. Required PPE: long sleeves, closed-toe shoes, safety goggles, oven mitts. No bare hands near the pot ever. Have a bowl of cold water nearby in case of skin contact (immediate immersion of the affected area limits the burn's depth). Do not allow students to perform candy work in classroom settings without specific institutional safety protocols. The home version is plenty risky enough for adults.

She does, however, demonstrate the cold-water-test concept by dropping spoonfuls of pre-cooked sugar syrup at different stages into a glass of cold water in front of the class. The students see the soft-ball, hard-ball, and hard-crack textures one after another, and the chemistry of "the temperature reads the concentration" lands on them visibly. The kids remember it.

The two methods, in detail: a wet caramel walkthrough

Let's walk Maya — and you — through a successful wet-method caramel, step by step, with the chemistry annotated at each stage.

Step 1. Combine 1 cup (200 g) granulated white sugar, ¼ cup (60 mL) water, and 1 tablespoon (15 mL) light corn syrup in a heavy-bottomed saucepan. The corn syrup is your insurance against crystallization.

Step 2. Place over medium heat. Do not stir. Watch as the sugar dissolves. The water is acting as a solvent here; the heat helps drive the dissolution. Within 2-3 minutes, the syrup is clear.

Step 3. Once boiling begins, cover the pan with a tight-fitting lid for 2-3 minutes. The steam is doing important work: it condenses on the inside of the lid and runs back down the sides of the pan, washing any sugar crystals down into the syrup before they can serve as nucleation sites. After this brief covered-boil phase, remove the lid.

Step 4. Insert a candy thermometer or use periodic spoon-and-cold-water tests. The syrup will pass through clear-and-watery (102°C / 215°F) → thread (108°C / 226°F) → soft-ball (115°C / 240°F) → hard-ball (121°C / 250°F) → soft-crack (132°C / 270°F) → hard-crack (149°C / 300°F). At hard-crack, all the water is gone; what remains is dry molten sucrose.

Step 5. From hard-crack to caramel is a temperature interval of about 20°C — and the chemistry changes dramatically through this interval. The syrup will begin to take on color: clear at 149°C, faint yellow at 160°C, light gold at 168°C, amber at 173°C, mahogany at 180°C, dark mahogany at 185°C, burnt above 190°C. Watch the color, not the thermometer, in this phase. Thermometers can lag the actual liquid temperature by a few degrees, and a few degrees is the entire difference between perfect and ruined.

Step 6. When the color is honey-amber to medium-mahogany — your aesthetic preference, but generally between 173-180°C / 343-356°F — immediately remove from heat. The pan will continue to cook from residual heat for 30-60 seconds; build that into your timing. If you want medium amber, pull at light gold; if you want dark mahogany, pull at mahogany.

Step 7. Off heat, slowly whisk in ½ cup (120 mL) heavy cream (warmed to room temperature or slightly warmer; cold cream causes the caramel to seize). The mixture will foam vigorously and steam dramatically — caramel is at 175°C, cream is at 70°C, the temperature difference vaporizes water instantly. Once the foam subsides, whisk in 2 tablespoons (28 g) butter and ¼ teaspoon flaky sea salt.

Step 8. Cool to room temperature, then refrigerate. The result is a caramel sauce that pours cool, firms slightly when refrigerated, and re-pours when warmed. It will keep, refrigerated, for several weeks.

The chemistry annotation: in steps 1-3, you're managing a saturated-and-then-supersaturated sugar solution, with corn syrup as anti-crystallization insurance. In steps 4-5, you're climbing the candy-temperature ladder by driving off water. In step 6, you're running a chemical decomposition (caramelization) for a controlled brief interval. In step 7, you're stopping the reaction with an ingredient (cream) that supplies water and fat to dilute and emulsify the caramel into a sauce. In step 8, you're cooling the sauce; the small amount of remaining sucrose stays inverted and dispersed enough not to crystallize.

A walkthrough is no substitute for actually doing it. Make caramel three times. The fourth attempt will work. The fifth will be excellent.

The wider candy panel

Beyond caramel, the candy-temperature ladder gives you a key to a wide range of confections.

Fudge. Cooked to soft-ball (115°C / 240°F). Cooled undisturbed. Beaten while warm to induce microcrystallization.

Pralines (American Southern style). Cooked to soft-ball with butter and pecans. Cooled briefly, then poured onto wax paper. Sets into a soft-grainy candy somewhere between fudge and toffee.

Toffee. Cooked to firm-ball (118°C / 244°F) or higher with butter. Spread thin to cool into hard, brittle slabs.

Caramel candies (the "soft-caramel" cube). Cooked to firm-ball with cream and butter. The cream contributes water and fat that keep the caramel pliable rather than glassy.

Marshmallow. Cooked to soft-ball, then beaten with whipped egg whites or gelatin, which traps air and produces the characteristic foam.

Lollipops. Cooked to hard-crack (149°C / 300°F). Poured into molds, cooled quickly.

Brittle. Cooked to hard-crack with nuts. Poured thin.

Cotton candy. Sucrose melted, then forced through a centrifuge with tiny holes. The molten sugar threads cool instantly into glass-state strands; the strands are wound into the familiar pillow.

Each of these is a temperature-and-technique combination from the same starting material. Sugar is one of the most versatile single ingredients in cooking; the candy-temperature ladder is the index that makes that versatility usable.

Supersaturation and the seed-crystal problem

Most candy-making is, deep down, a problem in supersaturation and nucleation — concepts borrowed directly from physical chemistry. Let's spend a moment on each.

A saturated solution is one in which the dissolved solute concentration equals the maximum the solvent can hold at that temperature. If you keep adding sugar to a glass of room-temperature water, you'll get to a point — about 200 grams per 100 mL — where additional sugar simply won't dissolve. The water is saturated. Any more sugar sits at the bottom.

A supersaturated solution is one that contains more dissolved solute than the saturation limit at the current temperature. This sounds impossible, but it's achievable: dissolve sugar in hot water (where saturation limits are higher), then cool the solution carefully without disturbing it. The dissolved sugar wants to come out of solution but has no nucleus to crystallize on, and so it sits, suspended in a metastable state. Sugar syrups in candy-making are routinely supersaturated by factors of two or more. A soft-ball candy syrup at room temperature contains about three times as much dissolved sugar as a saturated solution can hold.

Supersaturation is metastable — meaning, it persists until something disturbs it. The disturbance can be:

• A seed crystal. Drop a single sucrose crystal into a supersaturated solution and crystallization will sweep through the solution within seconds. Each crystal you started with becomes the nucleus for a hundred more, and the chain reaction propagates.
• A scratch on the container. The microscopic edges of a glass scratch can serve as nucleation sites. This is why crystal-clear hard candy is sometimes made in pristine new molds.
• Mechanical agitation. Beating, stirring, or even violent vibration can induce nucleation. (This is part of why you don't stir a wet-method caramel once the syrup has reached the boiling point — stirring can trigger crystallization.)
• Cooling. Lowering the temperature reduces the saturation limit, making the solution more supersaturated and more likely to spontaneously nucleate.

For the candy-maker, supersaturation is both a tool and a hazard. You want it for fudge (you induce nucleation by beating, producing many small crystals). You don't want it for lollipops (you cool fast enough that crystallization can't get started, locking the sugar into glass). You absolutely don't want unintended supersaturation in caramel (where a stray crystal can destroy the whole batch).

This is also why corn syrup matters so much. The free glucose in corn syrup physically blocks sucrose-to-sucrose alignment, raising the activation energy of nucleation. Even in supersaturated solutions, with corn syrup present, crystallization happens reluctantly. A cook adding 1 tablespoon of corn syrup per cup of sugar is, in effect, lowering the supersaturation risk by a factor of ten.

The Maillard-caramelization confusion in the kitchen

We said above that caramelization and Maillard are different reactions. In a science kitchen, this distinction is clear. In an actual home kitchen, the two reactions almost always happen together, and the resulting confusion is one of the most common sources of misunderstanding.

Take a simple example: roast onions. Onions contain sucrose, fructose, and glucose (a fully ripened onion is about 4–8% sugar by weight). They also contain free amino acids and small peptides (~1–2% by weight). When you roast onions in a hot oven, both reactions run. Caramelization is breaking down the sugars; Maillard is reacting the amino acids with the reducing sugars; the two reactions share intermediates and produce overlapping flavor compounds. The brown-and-sweet-and-savory result is both reactions, in unknown proportions, working together.

The same is true for: a roasted sweet potato (caramelization of sugars + Maillard with proteins), a baked cookie (caramelization of sucrose + Maillard with butter proteins and milk proteins), the bottom of a pan after sautéing onions (the fond — both reactions), the crust of a roast chicken (mostly Maillard since the sugar content is low), the top of a crème brûlée (mostly caramelization since the sugar dominates), and the exterior of a marshmallow held over a campfire (mostly caramelization since marshmallow is essentially pure sugar foam).

A working understanding for the home cook: when sugar dominates and protein is minor, caramelization dominates the flavor. When protein and sugar are both present in significant amounts, both reactions run, and the result is what most cooks just call "browning."

The science teacher's clean version: caramelization without Maillard happens when you cook pure sucrose. Maillard without caramelization happens when you cook protein with very low sugar at moderate temperatures. Most cooked foods are mixtures.

The supersaturation problem in candy: why honey crystallizes

Here is a kitchen mystery worth pausing on. A jar of honey, untouched on a pantry shelf, will after some months show fine, gritty crystals appearing on the bottom of the jar. Some honey crystallizes within weeks; some takes years. Eventually, all honey crystallizes. Why?

Honey, as we said, is a roughly equal mixture of free glucose and free fructose, plus 17% water. That sounds like a mixture that should resist crystallization — and fructose crystals are indeed rare in honey because fructose stays liquid down to very low temperatures. But glucose is a different story. Pure glucose has a saturation limit much lower than fructose, and honey is supersaturated with glucose at room temperature. Specifically, the glucose concentration in most honey is around 30-35%, while glucose's saturation limit at room temperature is closer to 20%. So honey, sitting in your pantry, is metastably supersaturated with glucose.

What triggers crystallization? The same things that trigger any other supersaturated crystallization: a seed crystal, a scratch in the jar, a slight cooling, a small disturbance. The seed crystals are usually pollen grains, wax fragments from the comb, or tiny air bubbles trapped during processing. Once nucleation begins, glucose crystals grow slowly outward, leaving the surrounding fructose-rich solution more concentrated. The result is the familiar cloudy, partially-crystallized honey.

Honey from different floral sources crystallizes at different rates because the glucose:fructose ratio varies by source. Clover honey is roughly 40:35 glucose:fructose and crystallizes within months. Acacia honey is roughly 25:45 glucose:fructose and stays liquid for years. Tupelo honey, the famously slow-crystallizing variety from Florida and Georgia, is high in fructose and never quite crystallizes during typical pantry-life. The crystallization rate of a honey is its glucose:fructose ratio in disguise.

To re-liquefy crystallized honey: warm gently in a water bath at around 40-50°C until the crystals re-dissolve. Avoid microwaving (uneven heating; can damage flavor). Refrigeration speeds crystallization (the same kinetic principle that drives starch retrogradation in Chapter 9 drives honey crystallization here). Honey is meant to live at room temperature.

A note on ingredient honesty

The American sweetener panel is sometimes presented in moralized terms: HFCS is bad, agave is good; refined sugar is poison, honey is medicine. The chemistry doesn't support this framing. Most sweeteners in your kitchen are mixtures of glucose, fructose, sucrose, and water in varying ratios. They differ in flavor compounds, in mineral content, in glycemic response, and in cultural and historical associations. They do not differ in any way that justifies the moralized vocabulary marketing has wrapped around them. Eat what tastes good. Eat reasonably. Pay attention to your blood sugar if you have a metabolic condition. Beyond that, the bee's syrup is not virtuous and the corn's syrup is not evil; both are sugars, and both behave like sugars in cooking.

Anchor: the sugar matrix in chocolate

Chocolate gets its own chapter (Chapter 20), but sugar is one of the chapter's three primary ingredients (cocoa solids, cocoa butter, and sucrose), and the chemistry of how sugar behaves in chocolate is essentially the chemistry of this chapter applied to a fat-rich matrix.

A bar of dark chocolate is, by composition, around 30-35% cocoa solids, 35-40% cocoa butter, and 25-30% sucrose. The sucrose is in the form of microscopic crystals — typically around 20-30 micrometers across, small enough that your tongue cannot detect them as grit. The sucrose does not dissolve in chocolate. It is suspended, as solid crystals, in the cocoa butter matrix, which is itself partially crystalline at room temperature.

The size of the sucrose crystals is the key to chocolate's mouthfeel. Conching — the prolonged stirring of chocolate during manufacture (described in Chapter 20) — gradually grinds the sucrose crystals down to a target size. Below about 30 micrometers, the chocolate feels smooth on the tongue; above 30 micrometers, it feels gritty. This is exactly the same principle as fudge: small crystals, smooth mouthfeel; large crystals, grainy mouthfeel.

When you eat chocolate, the cocoa butter melts at body temperature (about 37°C), releasing the sucrose crystals into the warm slurry that's now coating your tongue. The sucrose then dissolves in the saliva on your tongue. Chocolate is essentially a delivery system for sucrose, fat, and aromatic compounds, all engineered for melting at exactly your body temperature. This is one reason chocolate is so addictive: it is engineered to release its flavor compounds at the temperature of the human mouth.

The sugar in milk chocolate is supplemented by milk solids and milk fat (Chapter 16), which add lactose (a different disaccharide) and additional sweetness perception. White chocolate is cocoa butter plus sugar plus milk solids — no cocoa solids at all. The chemistry of sugar in each style is similar; the matrix differs.

We will return to chocolate's microcrystal structure — both for cocoa butter (six polymorphic crystalline forms, only one of which gives the snap) and for the sucrose suspension — in Chapter 20.

Modernist candy: where physics meets imagination

A brief preview of what happens when contemporary pastry chefs apply the chemistry of this chapter at scales and conditions classical confectioners could not.

Sugar tuiles are paper-thin sheets of cooked sugar (often with butter and corn syrup) baked at high temperature, then carved or rolled while still warm and pliable. Cooled, they are translucent, brittle, and edible. Some chefs flash-stretch them with hot air immediately after baking to produce three-dimensional sculptural pieces.

Isomalt sculpture uses isomalt — a sugar alcohol derived from sucrose, which behaves like a glass and resists crystallization more reliably than pure sucrose. Show-piece pastry chefs use isomalt for spun-sugar bowls, sugar flowers, and crystalline garnishes that hold their shape for days at room temperature.

Pulled-sugar work is the traditional French technique of stretching cooled-but-still-pliable sugar (around 70-80°C) into long, flexible ropes, which are then twisted into ribbons, flowers, or animals. Properly stretched sugar develops a satin sheen because the stretching forces the sugar molecules into rough alignment — a kind of partial supercooled crystallinity that scatters light differently than amorphous glass.

Aerated chocolate and sugar foams combine the sugar chemistry of this chapter with foam science (Chapter 12). Carbonated chocolate (Aero bars, modernist chocolate desserts) uses dissolved gas trapped in a sugar matrix; warming to body temperature releases the gas as the chocolate melts.

The sucrose-acid hydrolysis as flavor tool. Some modernist chefs deliberately partially-invert their sugar with citric acid before incorporating it, producing controlled levels of glucose and fructose in the final candy. This shifts both the flavor profile (more glucose = less perceived sweetness, more "background" sugar; more fructose = more perceived sweetness) and the crystallization behavior (more invert = more anti-crystallization).

We will return to modernist candy in Chapter 38; the foundation is the chemistry of this chapter.

Cross-Chapter Connections

🔗 Chapter 8 (The Maillard Reaction) is the chapter you should re-read alongside this one. The two reactions are often confused. Maillard requires both an amino acid and a reducing sugar; caramelization requires only sugar. Both produce browning and flavor compounds. Both happen in many of the same dishes simultaneously. Distinguishing them is a foundational skill in food science.

🔗 Chapter 9 (Carbohydrates and Starches) introduced the glycosidic bond — the same bond that links sucrose's glucose and fructose. Sucrose is a tiny, two-piece version of the long polysaccharides we met there. Same chemistry, different scale.

🔗 Chapter 18 (Fruits and Vegetables) will return to the question of fruit sugars and ripening, including the enzymatic conversion of starch into sugar that we previewed in this chapter's discussion of bananas (and that Pat Hammond's iodine demonstration in Chapter 9 made vivid).

🔗 Chapter 20 (Chocolate) features sucrose as the matrix in which cocoa solids and cocoa butter are suspended. The crystallization behavior we discussed for fudge and lollipops is exactly the same physics that operates when sugar is incorporated into chocolate during conching. Tempering chocolate is a problem in cocoa butter polymorphism (Chapter 11), but the sugar-crystal behavior is governed by this chapter's chemistry.

🔗 Chapter 36 (Food Preservation) will return to honey as the textbook case of preservation by water activity, and will explain why honey, properly stored, lasts essentially forever.

🔗 Chapter 38 (The Future Kitchen) will visit modernist candy techniques, where sugar is manipulated in ways that classical confectioners couldn't manage — sugar foams, sugar tuiles flash-stretched into intricate shapes, sucrose behaving as a structural element in plated desserts. The fundamental chemistry, however, is the chemistry of this chapter.

A small philosophical aside on sweetness

Why does sweet taste good?

The evolutionary explanation is that ripe fruit contains sugar, and ripe fruit was a calorically rich food source for our omnivore ancestors. The genes that selected for "find sweet, eat sweet" survived. The taste receptors on the tongue that respond to sweet — primarily the T1R2/T1R3 dimeric receptor, which we will discuss in Chapter 6 detail — are the same receptors across all primates and most mammals.

The cultural and developmental explanation is that breast milk is sweet. The first food a human encounters is sweetened by lactose; we are introduced to sweetness as a synonym for safe, plentiful, comforting. The associations established in infancy persist throughout life. Comfort foods, almost universally across cultures, are sweet.

The harder question — why does sweetness work the way it does in dessert — is partially explained by these biological hooks but also by the contrast principle. A bite of pure sugar is unpleasantly intense. A bite of sweet-paired-with-something-bitter (caramel, with its small bitter notes from caramelins) is balanced and satisfying. A bite of sweet-paired-with-acid (lemon tart, balsamic-glazed strawberry) is dynamic and lively. The pleasure of sweetness is heightened by contrast — and the chemistry of caramelization, by introducing complexity to pure sugar, is essentially the chemistry of making sweetness more interesting.

This is a useful framework for any cook trying to develop dessert recipes. Pure sugar bores the tongue quickly. Sugar that has been caramelized, browned, paired with bitterness, paired with acid, paired with salt, or paired with fat acquires the layered complexity that makes a dessert memorable rather than merely sweet.

Closing Reflection

There is something almost theatrical about caramel-making. A pile of white crystals goes into a pot, and over the next ten minutes, in a process you cannot interrupt or reverse, those crystals melt, lose their water, change color, develop new flavors, and finally, if you are lucky and quick, become a sauce of stunning complexity. Or, alternatively, become motor oil. The line is thin. The window is short. Your job, as the cook, is to be present.

Maya, on her third attempt, made a caramel correctly. By her fifth, she could do it without watching the thermometer — judging the color and aroma instead, the way Aroon does in Toronto with his palm sugar. The rule of thumb: when the kitchen smells like deep amber, when the syrup looks the color of an old penny, when the bubbles slow from energetic to syrupy — that is the moment. Pull. Pour. Stop. Add cream off-heat to halt cooking. Whisk gently. Salt to taste. Cool.

The science behind that moment is everything we have spent this chapter laying out. The sugar molecule is breaking apart. New molecules — furans, maltol, diacetyl, caramelins — are forming and accumulating. The flavor signature you smell is not "the sugar," it is what the sugar has become — a small molecular orchestra, momentarily concentrated in your kitchen, on its way through a brief peak and into a steep decline. Caramel is a window of three minutes during which the sugar is most beautiful. Open the window. Step through. Close it before it closes itself.

The next chapter takes us to fats — the third macronutrient, and the kitchen's flavor solvent. We will see, among other things, why butter makes caramel sauce richer (water plus fat plus sugar); why cocoa butter crystallizes in six different forms (and only one gives chocolate its snap); and why the bowl of cream Maya whisked into her successful caramel was, at the molecular level, doing the same kind of work an egg yolk does in mayonnaise. The flavor solvents await. Take the caramel off the heat first. Pour it into something. Then turn the page.