Caramelisation vs Maillard — Structural Chemistry Differences

Maillard described his reaction in 1912 working on protein-sugar condensation in biochemistry, not cooking. Caramelisation was understood empirically centuries earlier through sugar confectionery, formalised chemically through pyrolysis studies in the 19th century.

These are two separate, concurrent reactions that happen to share a temperature neighbourhood, and confusing them costs you control. Maillard is an amino-carbonyl reaction: a free amino group from a protein or amino acid attacks a reducing sugar, producing a cascade of intermediates — Amadori products, then Strecker degradation compounds, then melanoidins. No protein, no Maillard. Pure sugar, no protein? That's caramelisation only. Caramelisation is straight pyrolysis of sugars — thermal decomposition that doesn't need nitrogen at all. Sucrose above roughly 160°C begins breaking into fructose and glucose, then those fragment further into furans, diacetyl, hydroxy-acetaldehyde, and eventually polymeric brown caramels. McGee (On Food and Cooking, 2004) draws the distinction clearly: caramelisation needs only heat and sugar; Maillard needs reducing sugars plus amino acids or proteins, and it starts lower — documented from around 140°C but meaningful below 100°C in high-concentration systems. In the kitchen this splits in a practical way. Searing a steak: predominantly Maillard — the amino acids in muscle proteins reacting with the surface's available reducing sugars. Cooking a dry caramel for crème brûlée: no protein involvement, pure caramelisation. Baking bread crust: both running simultaneously, which is why bread crust has both the roasted-grain Maillard aromatics and the bitter-sweet caramel notes. Myhrvold, Young and Bilet in Modernist Cuisine (Volume 2) break out the Maillard pathway intermediates and note that water suppresses both reactions by lowering surface temperature — which is why boiled chicken has none of the crust character of roasted. Control pH and you push reaction rates: alkaline environments accelerate Maillard significantly, which is how lye pretzels and Cantonese roast duck skin achieve such aggressive dark colour faster. Caramelisation is less pH-sensitive but acid or base catalysis shifts which flavour compounds dominate the outcome. Understanding which reaction is running — or which dominates — lets you tune temperature, moisture, pH, and substrate ratios to get the flavour architecture you want rather than accepting whatever the pan gives you.

Maillard produces hundreds of volatile heterocyclic compounds — pyrazines (nutty, roasted), furanones (caramel-like but with nitrogen), pyrroles (grain, slightly green), and thiophenes (sulphurous, meaty) when sulphur-containing amino acids are present. Strecker degradation runs in parallel, cleaving amino acids into their corresponding aldehydes and CO2: the specific aromatic of any Maillard browning is partly a readout of which amino acids were available. Caramelisation by contrast builds flavour from sugar pyrolysis products: diacetyl and acetoin (buttery), hydroxymethylfurfural (sweetly caramel), furans (sweet, slightly solvent), and ultimately bitter polymerised caramels. Because bread crust, roasted coffee, seared meat, and crème brûlée crust all brown does not mean they taste the same — the substrate dictates which reaction dominates and which compound classes accumulate. McGee (On Food and Cooking, 2004, Chapter 14) maps the distinct aromatic families clearly.