Science

A sharp blade cuts through the muscle fibres cleanly. Most cells stay closed, so the water, the soluble protein and the myoglobin stay inside each piece. Each piece stays separate and red.

A food processor tears and crushes the meat. Many cell membranes break. The sarcoplasm, which is the liquid inside the muscle cell, comes out and coats the meat. The result is a wet, grey paste with no separate pieces.

Cut the beef last. Myoglobin on a new cut surface starts to oxidise to metmyoglobin, which is brown. Time between the cut and the plate is time for the colour to go.

Egg yolk and Dijon mustard are both emulsifiers. Yolk gives lecithin and lipoproteins. Mustard gives mucilage from the seed coat. Oil added slowly in a thin stream forms small droplets that these molecules hold in the water phase.

The emulsion coats each piece of beef with fat. This matters more with a lean cut such as fillet, which has little fat of its own.

The Cognac goes in after the emulsion is made. Ethanol dissolves aroma compounds that water cannot carry.

The purple centre in an uncured batch is a diagnosis, not a fault, and no amount of adjusting the dry will fix it.

Myoglobin's colour is set by what sits at the sixth coordination site of the heme iron.

• Nothing bound, iron as Fe(II): deoxymyoglobin, purple-red. This is the interior of any piece of meat that oxygen has not reached. It is the purple centre.
• Oxygen bound, Fe(II): oxymyoglobin, bright cherry red. This is bloom on a cut steak. It penetrates only 2 to 3 mm and it is transient.
• Iron oxidised to Fe(III): metmyoglobin, brown. It forms fastest at low oxygen partial pressure, which is exactly the zone just under the surface. This is why stored meat shows a brown band between a red rind and a purple core.
• Nitric oxide bound, Fe(II): nitrosylmyoglobin, stable pink-red.

Drying moves an uncured piece between the first three. It cannot reach the fourth.

Nitrite goes to nitrous acid at low pH, and nitrous acid gives nitric oxide, which binds the Fe(II) heme. The NO to heme bond is four to five orders of magnitude tighter than the O2 to heme bond, so the pigment does not need oxygen to stay red and it resists oxidation to met. Because the nitrite diffuses into the meat rather than depending on surface oxygen contact, the colour is uniform through the cross-section instead of being a 3 mm rind. That uniform translucent ruby is what commercial South African biltong shows when it is sliced, and it is not achievable by any drying schedule.

Traditional Boer recipes call for salpeter, which is potassium nitrate. Nitrate is not directly active: bacterial nitrate reductase must reduce it to nitrite first, which is why old recipes need long cure times. A modern nitrite cure salt skips that step and is far easier to dose.

German Nitritpoekelsalz is 0.4 to 0.5% sodium nitrite in sodium chloride, so it is over 99% salt by mass. Substituting it one for one for the salt in the dredge at 13.85 g/kg gives roughly 69 mg/kg ingoing nitrite. That sits inside EU limits for dry-cured meat and above the roughly 50 mg/kg needed for full colour development, at identical sodium. It must be a substitution, not an addition: dosing to 150 mg/kg on nitrite alone would need about 30 g/kg of cure salt, which is more than twice the salt this recipe wants.

Sodium ascorbate at 0.5 g/kg is worth adding with it. It reduces nitrite to nitric oxide far faster, drives the colour all the way to the core during the fridge cure, and suppresses nitrosamine formation.

The vinegar wash helps rather than hinders here, because the low pH is what drives nitrous acid to nitric oxide.

The pigment chemistry is textbook and not in doubt. The specific dosing arithmetic above is calculated rather than measured in this kitchen, and it has not yet been run: nitrite was specified for batch 7 and dropped for lack of supply. The first batch that uses it should record the interior colour deliberately, because that is the whole point of the change.

Two things worth being clear about. Acid accelerates autoxidation of myoglobin to metmyoglobin, so a heavy vinegar wash pushes an uncured interior toward brown rather than red. And myoglobin denatures irreversibly from around 55 degrees, so a heating pad that drives box air past that turns the colour grey no matter what else is done. Keeping the box at or under 34 degrees is a colour control, not only a fat control.

Mushrooms are 80–90% water. The 25–30 minute cook removes about 90% of it; dry-sauté first, add fat after. Cream added near the end sounds like it should make things wetter and does the opposite — the fat emulsifies the residual water and stops it migrating into the pastry.

Treating the tenderloin as eight flat faces plus two ends gives even caramelisation while keeping the interior raw. A cylinder seared with single flips browns unevenly and drives heat further into the centre, which is then overcooked by the bake.

Prosciutto works by fat solidifying during the chill into a semi-impermeable layer. A crêpe works by its egg protein structure absorbing a little moisture without collapsing. They are not interchangeable, and using both is not redundant.

Start at 220 °C so the butter layers in the pastry expand and set structure before the beef heats, then drop to 190 °C for even penetration. Pull at 52–54 °C: carryover adds about 3 °C during the rest, landing at 54–57 °C for medium-rare.

Sugar is hygroscopic. It pulls water and oil out of broken oil glands in the coloured part of the peel. The sugar crystals hold the oil on a large surface. The lemon juice then carries the oil into the drink.

Lemon oil is mostly limonene. The smell that reads as lemon comes from citral (geranial and neral). Citral is only a few percent of the oil.

The white pith contains bitter limonoids and flavanones. A Y-peeler keeps most of the pith out.

Citral is not stable in acid. It slowly cyclises and oxidises. The products, such as p-cymene and p-methylacetophenone, smell flat and medicinal. This sets the drinking window of 1 week. Commercial lemonades use stabilised flavourings and ascorbic acid to slow it. The pH is estimated. Nobody measured it.

Spearmint gets most of its aroma from (R)-(−)-carvone. The mirror-image molecule, (S)-(+)-carvone, smells of caraway.

The oil sits in peltate glandular trichomes. These are small oil sacs on the surface of the leaf. A slap breaks some of them. A blender breaks all of them. Blending is what makes the mint strong.

Broken leaf cells release enzymes. Lipoxygenase turns leaf lipids into green-leaf volatiles such as cis-3-hexenal. These smell of cut grass. Polyphenol oxidase makes the leaf brown.

A 10-second blanch denatures both enzymes before the blender breaks the cells. Carvone and menthol boil above 200 °C, so the short blanch loses little aroma. The ice water stops the cooking.

An unblanched leaf must not be squeezed or muddled for the same reason.

Acid replaces the magnesium in chlorophyll with hydrogen. This makes pheophytin, which is olive-brown. Blanching does not stop this. Over some days in the bottle, the colour goes from green to khaki. The taste does not change.

Menthol activates TRPM8, the cold receptor in the mouth and nose. The brain reads the cold as intensity. Spearmint contains little menthol. One part peppermint to three parts spearmint adds the cold without a toothpaste taste.

Wild yeasts and lactic acid bacteria live on the skin of fresh ginger. Do not peel it. The daily sugar feeds the organisms that ferment.

Yeast makes CO2 and ethanol. Lactic acid bacteria add a light sour note under the lemon.

The pear juice gives nutrients that sugar alone does not give. Fentimans also puts pear juice in its ginger ferment.

Yeast activity is highest 6–12 hours after a feed. The lemonade base is very acidic: estimated pH 2.6–2.9, not measured. Acid slows wild yeast. A weak or cold bug can stall for days. A bug at its peak starts the bottles in 2–4 days.

Each gram of sugar that yeast ferments makes about 0.49 g CO2 and 0.51 g ethanol. About 2.5 volumes of CO2 need about 5 g CO2 per litre. That uses about 10 g sugar per litre, or about 20 g in 2 L.

The base starts near 170 g fermentable sugar (8.4%). It ends near 150 g (7.4%). The ethanol is about 0.5–0.7% ABV.

All of these values are calculated. Nobody measured them yet.

CO2 carries aroma out of a liquid when it vents. The bug jar vents, so the mint never goes in it. A closed bottle cannot vent. The gas and the aroma both stay in the lemonade. Brewers add aroma hops after the main fermentation for the same reason.

CO2 dissolves about two times better at 4 °C than at 20 °C. In the fridge, gas moves from the headspace into the liquid. The bubbles become finer, and the bottle gushes less when you open it. The cold also stops the yeast.

Solid particles are nucleation sites. CO2 forms bubbles on them. Pulp in the bottle makes it gush when you open it. Pass the juice through a fine sieve.

Fentimans uses pear juice and glucose syrup. Glucose syrup contains dextrins. They add body with little sweetness.

Pear juice contains sorbitol. Saccharomyces yeast ferments sorbitol poorly, so it can stay in the drink as body. This part is a hypothesis. It is not confirmed.

Berlin tap water is hard. Its bicarbonate neutralises part of the citric acid. The lemonade then tastes less sharp. Filtered water keeps the acid.

Juniper oil is mostly alpha-pinene and myrcene. These are monoterpenes, the same family of molecules as limonene in lemon oil.

Because the molecules are related, the juniper does not taste like a separate flavour. It adds depth and a resin note inside the lemon.

Fentimans puts juniper in Victorian Lemonade for the same effect. In this recipe the juniper goes into the oleo, so the sugar pulls its oil with the lemon oil.

An oleo works for lemon because the peel has large oil glands in a thick, waxy skin. The sugar pulls the oil out slowly and cleanly over 90 minutes.

Mint is different. Its oil sits in small trichomes on a thin, wet leaf. Sugar breaks the trichomes, but in 90 minutes it also bruises the leaf cells. The bruised cells release lipoxygenase and polyphenol oxidase. The result is a grass taste and a brown colour.

Heat is also wrong. A hot steep drives off the aroma and pulls out rosmarinic acid and tannins. The result tastes like mint tea, not fresh mint.

So the mint gets its own process: a short blanch to stop the enzymes, then a hard blend in cold water.

Four yolks are about 70 g. The cheese is 50 g. Most carbonara uses more cheese for each yolk, so this sauce is looser and tastes more of yolk.

The yolk phospholipids, mainly lecithin, hold the pork fat and the pasta water together as an emulsion. The yolk proteins start to set near 65 °C. Above about 70 °C they join into grains and the sauce becomes scrambled egg.

Take the pan off the heat before the paste goes in. The pasta and the fat are hot enough to thicken the sauce without the burner.

Amylose from the pasta water and casein from the cheese sit between the yolk proteins. They slow the proteins from joining, so the sauce can take a few degrees more before it goes grainy.

This recipe has little cheese, so the pasta water does more of this work. Spaghettoni is thicker than spaghetti and gives off more starch into the water.

Add the water in small amounts. Too much water drops the temperature and the sauce stays thin.

Salt on a cut cabbage face pulls water out in a few minutes. That water holds the surface near 100 °C, so the face steams and does not brown. Browning needs the surface above about 140 °C.

Pointed cabbage holds more free sugar than white cabbage, so a dry face browns fast and deep. The weight presses the loose leaves flat against the pan. This gives an even crust instead of brown patches.

Salt the faces after the sear.

Cut cabbage releases isothiocyanates, which give the sharp mustard taste. Heat stops the enzyme that makes them. Long wet cooking then breaks sulfur compounds down to dimethyl trisulfide and hydrogen sulfide. These give the smell of overcooked cabbage.

Pointed cabbage has thin cell walls and becomes tender fast. Stop when the knife meets a small resistance in the core. Keep the charred faces above the liquid so the crust stays dry.

Sumac is sour from malic acid. Its fruity smell comes from volatile compounds that heat drives off. So it goes on the onions and on the finished dish, and never into the pan.

Sumac and red onion both hold anthocyanins. These pigments are brightest at low pH. The lemon juice drops the pH, so the onion turns bright pink. Squeezing breaks the onion cells so the acid gets in and softens the raw sulfur taste.

Silken tofu is set with glucono-delta-lactone, which hydrolyses slowly to gluconic acid and drops the pH gently until the soy globulins — 11S glycinin and 7S beta-conglycinin — coagulate into a fine, weak gel holding around 88 percent water. Blending shears that gel into a colloidal protein suspension. Those globulins are amphipathic and act as emulsifiers, parking at the interface between the peanut and sesame oil and the water phase. That is why the sauce has the mouthfeel of sesame paste at roughly a third of the fat.

Two independent failure routes. Above roughly 70 °C the suspended globulins unfold and re-aggregate into visible curds; the sauce goes grainy, weeps water, and does not come back. Separately, soy protein's isoelectric point sits near pH 4.5, and driving the sauce toward it strips the net charge that keeps the particles repelling each other, so they flocculate.

The method follows from this: the sauce is assembled cold, sits in the bottom of the bowl, and is warmed only by the hot noodles landing on it. It never goes in a pan and no acid goes in the blender.

The numbing is hydroxy-alpha-sanshool, an unsaturated fatty acid amide that activates TRPA1 and, unusually, inhibits the two-pore potassium channels that normally keep light-touch fibres quiet. That inhibition is why it reads as a vibration rather than a taste.

It is both heat-labile and readily oxidised at those polyunsaturated bonds. Cooked into the pork for four minutes you keep the woody aroma and lose most of the tingle. Toast for seconds, grind coarse, scatter at the pass.

Chinkiang is a glutinous-rice vinegar aged on a solid bran mash, so alongside acetic acid it carries the Maillard and melanoidin products of that long ferment — the malty, nearly balsamic depth. Fifteen millilitres into the hot pan drives off the sharp low-boiling acetic fraction and leaves the aged notes bound into the fat. The final ten go over the pork uncooked, where the volatile edge survives.

Neither dose goes near the tofu, for the isoelectric point reason above.

Mince dumped in and stirred immediately releases water faster than the pan can evaporate it. Surface temperature crashes below 100 °C and the meat steams. Maillard chemistry needs a dry surface and roughly 140 °C.

Two undisturbed minutes lets the contact face dry and brown, and once that crust exists the rendered fat carries heat for the rest. The crumbly, dark, nearly crunchy result is the textural point of the dish — the counterpoint to a soft coating sauce.

Thin shallot dropped into already-hot fat sets its exterior before the interior water has left, giving a shell around a limp centre that goes soggy within minutes.

Starting cold gives a long, gentle dehydration phase: water leaves evenly, the structure collapses uniformly, and browning only begins after that. Pull at pale gold, because the retained oil carries them another two shades off the heat and Maillard products turn bitter quickly past that point.

A knife cut shears cleanly across cell walls and leaves a smooth, low-area face that sheds sauce. Smashing propagates fractures along the weak parenchyma between cell layers, giving torn irregular surfaces with far more area and open cell structure.

The salt then works by simple osmosis, pulling free water down the gradient while the cell walls stay intact enough to keep the snap. Rinse the surface salt off afterwards or the purge becomes seasoning.

Ovalbumin and ovotransferrin in the white set from about 62 °C and are firm by 80 °C. The yolk's low-density lipoproteins and livetins do not fully set until nearer 70 °C throughout.

Fridge-cold into hard-boiling water at that timing lands a fully set white with a yolk core still around 65 °C: thick, translucent, flowing slowly. The ice bath is not optional. Residual heat carries a hot egg another two minutes of cooking, and the bath also contracts the white away from the inner membrane so it peels cleanly.

Lao Gan Ma's fermented soybeans and the soy sauce contribute free glutamate. Oyster sauce contributes 5-prime nucleotides, IMP and AMP, from the shellfish extract.

The umami receptor T1R1/T1R3 binds glutamate at one site and is allosterically stabilised by nucleotides at another, so the two together produce a response well above their sum rather than a simple addition. That is why oyster sauce earns its place in an already salty bowl: it is not adding salt, it is multiplying what is there.

They sit in brine at high ionic strength, so the tissue is water-saturated and osmotically loaded with salt. Straight from the jar into a pan they release that water, the surface temperature crashes below 100 °C, and the carefully crusted pork steams back into grey mince.

Squeezing removes the free brine. The dry fry drives off the bound water and runs a little Maillard on the residual sugars, which is why they shift from smelling sour to smelling toasty. Only then can they meet the rendered fat without undoing it.

The finished Yibin pickle holds approximately 8.5 % salt. The research note on the Suimiyacai ingredient gives the source. This version does not soak it, so all of that salt reaches the bowl. 60 g adds approximately 5 g of salt to the batch. That is approximately 1 g for each of the 5 bowls. A larger dose adds more flavour, but it adds salt at the same rate.

The pickle also holds a high level of free glutamate. That is why the Lao Gan Ma decreases from 38 g to 25 g. More glutamate from a second source gives less and less extra effect at the same receptor site. The oyster sauce stays at 30 g, because its nucleotides act at a different site and multiply the glutamate.

The pickle holds its water inside the plant tissue. It has no free brine. So it does not flood the wok and it does not steam the pork. It does not need the soak, the squeeze and the dry-fry that the generic pickled mustard greens needed.

It tastes sweet, and sugar burns fast in hot fat. Add it at medium heat and keep it moving.

The wok removes water at a rate set by its hot surface area. The quantity of meat does not change that rate. Mince releases water when its proteins contract. 1 kg in one wok releases two times the water of 500 g onto the same surface. The surface then stays below 100 °C, and the meat steams.

Revision 1 cooked 500 g in one wok and got a crust. Two batches of 500 g keep each batch at that tested load. The second batch goes into the fat from the first batch. The fat gives better contact with the hot metal.

Five eggs from the refrigerator weigh approximately 300 g at approximately 4 °C. To heat them to 100 °C takes approximately 90 kJ. That energy can decrease the temperature of 3 l of water by up to approximately 7 °C before the burner recovers it. In 1.5 l, the decrease is approximately two times larger.

The time of 6 minutes 30 seconds assumes a full boil. If the water is cooler, the white sets more slowly and the egg is undercooked at that time. This is a calculation. Nobody measured it.

Each bowl holds approximately 255 g of sauce and udon. A revision 1 bowl held approximately 440 g. The bowl has the same mass, so it takes a larger part of the heat from the udon. The sauce then stays cooler. Warm bowls correct this.

Do not make the bowls hot. The soy proteins in the sauce unfold and clump near 70 °C. A bowl above 60 °C can curdle the layer of sauce that touches it.

The dual bone foundation is doing three separate jobs: marrow for fat and mouthfeel, veal knuckle for gelatin, oxtail for concentrated meatiness. Dropping any one of them is a real loss rather than a simplification. The triple clarification is likewise three different mechanisms — mechanical, protein raft, adsorption — not the same step done three times.

The heat and the aroma of a chilli are made in two different tissues.

Capsaicinoids come from glands on the placenta. The placenta is the pale ridge inside the pod that holds the seeds. No other part of the fruit makes them. The seeds are only hot because they touch the placenta.

The aroma compounds are made in the pericarp. The pericarp is the flesh wall. Capsicum chinense makes esters there that give the apricot and floral smell of a habanero and a reaper.

The two systems do not overlap. So a fat extraction of the placenta gives an oil with almost no smell and very high heat. And flesh with the placenta removed smells like the whole pod at about one fifth of the heat.

This lets you control the two properties separately. Put the flesh in the paste for aroma. Put the placenta in the tincture and add the heat by the millilitre.

Capsaicin is strongly fat-soluble. Its partition coefficient is about 3.6. It is roughly 4000 times more soluble in fat than in water.

In warm oil with nothing to compete, the extraction is almost complete in 15 minutes at 65 degrees C. More heat gives almost no more yield.

More heat does give more vapour pressure. At frying temperature the capsaicin leaves the pan on fat aerosol. This is the same mechanism that riot control sprays use. It will make the kitchen unusable.

Use 80 ml of oil for six pods. Less oil is absorbed by the tissue and the yield falls. More oil only dilutes a fixed quantity of capsaicin.

Onions store carbohydrate as fructans. Heat breaks the fructans into free fructose and glucose. These sugars then react with the amino acids of the onion in the Maillard reaction.

Early in the reaction you get pyrazines and thiols. These taste roasted and savoury. The onion also releases free glutamate.

Later you get melanoidins and maltol. These taste of caramel and fudge. They have low volatility and they stay on the palate.

Those late compounds mask the high-volatility compounds of lemongrass and ginger. Citral, geraniol and the terpenes are lost behind them.

Deep gold gives you the glutamate and the body. It does not give you the mask.

A chicken Maryland is worked muscle. It holds a lot of type I collagen.

Collagen converts to gelatin slowly from 70 degrees C. The conversion is much faster above 80 degrees C. Hold the thigh at 85 to 90 degrees C for a long time. This is far above the temperature you would use for a breast.

The covered stage keeps the air in the tray saturated with water. The meat does not dry while the collagen converts.

But that same saturated air keeps the skin wet. Skin must lose most of its water before it can brown. So the skin cannot crisp under foil at any temperature.

The uncovered stage is a drying stage. It browns the skin because it dries the skin first. The fan matters here. Moving air removes the wet layer of air on the skin surface faster than still air.

Citral is the main aroma compound of lemongrass. It is a pair of molecules, neral and geranial. Citral is very volatile. It leaves with steam.

90 minutes at 145 degrees C removes almost all the citral from anything in the liquid. What stays is the heavier fraction. It is mild and woody. It is useful as a background.

The pounded stalks in the paste give fibre and the compounds held in the cell walls. These release slowly.

The raw shavings at the end are pure top note. They are not heated at all. They are the reason the finished dish smells of lemongrass and does not only taste of it.

The same logic applies to the coriander. The root goes in the paste because its compounds are less volatile and survive the braise. The leaf goes on at the end.

The vinegar and the lime are not interchangeable here.

The Chinkiang vinegar is a cooking acid. It replaces tamarind. It holds the sweetness of the onions and the fat in check across 90 minutes. Acetic acid survives that time. So do the amino acids and melanoidins from the rice ferment, which are not volatile.

Add more vinegar at the end and you add raw acetic acid. It has had no time to combine with the dish. You taste it as a separate sharp note.

Lime is the finishing acid. Citric acid is rounder than acetic acid. The lime also arrives with its aroma intact.

Tamarind leads with tartaric acid, not citric. That is why tamarind tastes round and does not taste sharp. Lime juice cannot do the tamarind job in the pot. Its terpenes boil off in minutes and leave a flat sourness.

Radiant heat from coals is not even. It changes as the coals burn down and as you move them.

The skin records every one of those changes. The char is the sum of all the heat that reached the bird.

A probe reads one point. Near a pocket of fat or near the bone, the reading runs ahead of the deep muscle. One reading can say done when the thigh is not.

So over a fire, read the char first. Use the probe to confirm it.

In an oven the heat is even. There the probe leads and the colour confirms.

The target is for leg meat, not for breast. Thigh collagen starts to turn to gelatin slowly from 70 degrees C. A pull at 71 degrees C is correct for a breast and leaves the thigh chewy. At 76 to 78 degrees C the thigh is softer.

The paste holds palm sugar and oyster sauce. Both carry sugar. Under strong radiant heat that sugar goes from brown to black fast.

Coconut cream is about two thirds water. A thin coat on a hot surface takes heat away as the water evaporates. The surface cools for a short time.

Then the water is gone. The coconut fat and its proteins stay. They brown and build a lacquer on the skin.

So one baste slows the burn and adds colour. Put it on when the surface is close to burning. That is when it does the most work.

An acid marinade denatures the surface proteins of the meat. Over many hours the surface goes soft and mealy.

This paste holds no lime and no vinegar. It can sit for 48 hours and the texture does not change.

The salt still moves in. 40 ml of fish sauce carries about 7 g of salt. A longer marinade seasons deeper, and a very long one seasons harder.

The paste holds sugar from the palm sugar and the oyster sauce. At 220 degrees C that sugar would burn before the thigh is cooked.

A tray of boiling water fills the oven with steam. Wet air slows evaporation from the surface of the chicken. While the surface stays wet, it stays near the boiling point of water. The sugar cannot reach the temperature where it burns.

The meat still cooks. Steam moves heat into food faster than dry air.

Remove the tray before the grill stage. The surface then dries, and the sugar and proteins brown under the element.

Green beans are approximately 90 percent water. In hot oil, the surface water boils off first. The cells under the skin lose pressure, and trapped steam lifts the skin into blisters.

When the surface is dry, its temperature rises above 100 °C. Maillard reactions and caramelisation then start on the sugars in the skin.

The wrinkled skin has more surface area and many small folds. A small volume of sauce stays on the beans and does not pool in the pan.

One layer is necessary. Beans that overlap hold steam against each other and do not blister. 360 g of beans fits in one layer in a 28 cm pan.

Pork mince and fresh shiitake both release water in the pan. While water remains, the pan stays near 100 °C and nothing browns.

When the water is gone, the sound changes from a boil to a sizzle. The temperature rises, the fat renders, and the Maillard reaction starts.

Fresh shiitake is approximately 90 percent water. Add it at the start, with the pork. Its water then leaves during the boiling phase and does not delay the browning.

The glutamate and nucleotides in that water do not evaporate. They stay on the meat.

Glutamate comes from the suimiyacai, the soy sauce and the MSG. Guanylate comes from the shiitake. The umami receptor T1R1/T1R3 binds glutamate at one site. A nucleotide at a second site holds the receptor in its active shape. The taste is much stronger than the sum of the two.

This lets the dish use less suimiyacai and less soy sauce than the original. MSG is approximately 12 percent sodium by weight. Salt is approximately 39 percent.

Estimated sodium for the whole dish is approximately 2 g, or approximately 1 g for each serving. Nobody measured this. It is calculated from the salt figures stored on each ingredient. Light soy sauce (approximately 0.75 g) and suimiyacai (approximately 0.5 g) carry most of it. The 1 g of salt on the beans adds approximately 0.4 g.

Hu la (糊辣) is the aroma of dried chilli and Sichuan pepper fried to the edge of burning. It is also the base aroma of gong bao dishes.

The sugars and amino acids in the chilli skin react under heat. They make roasted and smoky compounds, such as pyrazines and furans. Capsaicin and the terpenes of Sichuan pepper, such as linalool and limonene, dissolve into the oil. The oil then carries them through the dish.

Past dark red, the chilli turns black and bitter. Use medium heat and count the seconds.

The whole peppercorns give aroma but lose much of their numbing in the hot oil. The ground pepper at the end puts the numbing back.

Acetic acid boils at 118 °C. On a pan wall hotter than that, much of it evaporates at once. The aged, malty compounds of the Chinkiang vinegar stay.

Add it in the last seconds. Long heat with acid replaces the magnesium in chlorophyll with hydrogen. The green then turns olive.

Settling in a still liquid follows Stokes law. The speed is proportional to the difference in density, proportional to the square of the particle radius, and inversely proportional to the viscosity of the liquid.

You cannot change the density of a vanilla seed. You cannot make warm custard thick enough to stop settling without cooking it into an anglaise first. The radius is the term you can control, and it is squared.

Scraped seeds come out of the pod held together in the oily tissue of the pod. A clump of ten seeds has about twice the radius of one seed, so it falls about four times as fast. Rubbing the seeds into sugar breaks the clumps apart. The sugar crystals work as a grinding medium.

The second control is time. Gel fronts start at every heated face and move inwards. At 1 cm depth each front travels only 5 mm. Putting the custard in at 55 degrees C rather than cold shortens the liquid window, so less falls before the gel sets.

Egg yolk carries about 16 percent protein. It is the only structural element in the custard. Cream gives fat and water, and the fat interrupts the protein network rather than building it. So firmness tracks the ratio of yolk to cream almost in a straight line.

The useful range is about 200 to 250 g of yolk per litre of cream. Below about 200 the gel is loose and silken. Above about 250 it holds a cut edge, which is the texture of crème caramel and not of crème brûlée. The floor where a baked yolk custard stops setting at all is about 140 g per litre.

This version runs 177 g per litre because that is the cream and yolk that were available. It sets, and it sets softer than the 230 g per litre version. The core goes to 82 degrees C rather than 79 to 80 degrees C. A sparser network needs a larger share of its protein unfolded before the structure spans the whole volume.

You cannot bake a lean mix firm. More heat does not add junctions that are not there. It only makes the existing junctions contract, and above 85 degrees C they push out the water they hold.

Dry air is a poor carrier of heat. To push energy into a custard through air, the oven must run far above the target, and that gap is what makes the edges of a tray cook before the centre.

Steam moves heat by condensing. Vapour condenses on any surface below the dew point and releases its latent heat, about 2260 kJ per kg of water. The transfer is much faster than dry air, and it happens at a fixed temperature.

The fixed temperature is the useful part. Once the surface reaches the dew point, condensation stops. The surface cannot go above the chamber temperature. A water bath works the same way, held at 100 degrees C by boiling. A steam oven lets you choose the number instead.

85 degrees C is chosen because the custard is finished at 79 to 82 degrees C and weeping needs the middle eighties. At 85 degrees C the failure is not available. Running 100 degrees C steam throws that away and puts the surface back above the weeping threshold.

One correction to a common claim. A combi oven at 85 degrees C is not pure saturated steam. It is humid air with a dew point of 85 degrees C. Air that does not condense blankets the surface and cuts the transfer coefficient. The advantage over dry air is real but it is several times, not a hundred times.

Capsaicin in pure fat spreads over the mouth as one continuous film that does not break up. The burn is flat and constant and has no shape.

Put the same quantity of capsaicin into an acid water base and the fat breaks into droplets. The capsaicin then arrives in pulses. Between the pulses you taste the sour and the salt. The heat has a shape.

The acid also helps you directly. Sour taste makes you salivate. Saliva is the main way capsaicin is removed from the tissue of the mouth.

The toasted rice powder holds a partial suspension. It does not make a stable emulsion. The sauce separates in the bowl. This is correct for a jaew. Stir it before each use, or the last spoon out of the bowl is neat tincture.

The esters that give Capsicum chinense its apricot and floral smell are small and volatile. They have low boiling points.

Fry that flesh in a paste at 130 degrees C or above and those esters leave or change. Maillard chemistry replaces them with heavier and rounder compounds. In a braise this is a gain. It is where depth comes from.

At the table it is a loss. There you want lift.

So use the same pod in two states. Fry one lot into the paste for depth. Leave one lot raw in the table sauce for the top note. This is not repetition. It is two different chemical states of one fruit.

Capsaicin is thermally stable well past 200°C, so a brief dry-toast does not touch the heat level at all. What toasting does is Maillard the residual sugars and proteins in the pericarp, generating a fresh pyrazine set that rhymes with the khao khua — which is why the two toasted elements taste like they belong to each other.

Capsaicin is also fat-soluble. Folding the flakes in while the rendered duck fat is still liquid gets it into solution and distributes the heat evenly instead of leaving hot pockets.

Glutinous rice is essentially pure amylopectin with almost no amylose. Toasted dry to 160–180°C the granules undergo dextrinisation and surface Maillard simultaneously, and the highly branched amylopectin structure collapses into a porous, friable matrix rather than a dense one.

Pounded coarse, that porosity gives an adsorbent: it takes up the free fish sauce, lime and rendered duck fat and holds them against the meat instead of letting them pool in the bottom of the bowl. Grind it fine and the pore structure collapses — you get paste, and the dressing separates out.

The nutty, popcorn flavour is the pyrazine fraction of the same Maillard reaction. Pyrazines are volatile, which is why the powder is made the same day rather than pulled from a jar.

Two separate things go wrong if the dressing goes on hot.

Moisture. Lime juice sits around pH 2.2. Introducing that to meat still above 70°C drives further myofibrillar contraction and squeezes out moisture that the render was managing to keep.

Aroma. The aromatic value of fresh lime lives in its terpene fraction — limonene, citral, linalool — all with boiling points low enough that a hot pan volatilises them straight off. What is left is naked acidity with none of the perfume.

Mint fails the same way in different molecules: menthol and menthone flash off, and residual polyphenol oxidase plus heat blackens the leaf.

Duck fat runs roughly 50% oleic acid, monounsaturated, with a melting profile that puts it fully liquid around 30°C. Lamb or beef tallow, dominated by saturated stearic and palmitic acid, sets above body temperature and coats the palate as wax.

That is the whole reason laab ped can be served warm-to-ambient and still read clean, and why the dish tolerates the higher render temperature that produces fond. Leg also carries far more type I collagen than breast, which under a short high-heat render contributes gelatine to the crumb rather than toughness.

Practical consequence: leg mince can go into a dry pan hot; breast mince cannot.

Galangal's terpenoid profile is dominated by 1,8-cineole and galangal acetate — camphoraceous and pine-sharp, quite unlike ginger's zingiberene. That specific sharpness is a direct counterweight to the fat load of duck leg.

This is why laab ped from Ubon runs galangal where a pork laab often does not bother. Do not substitute ginger here — it moves the dish toward warmth and sweetness, which is the opposite of what the duck needs.

Fish sauce is anchovy protein autolysed by endogenous and halophilic bacterial proteases over twelve to twenty-four months; the free glutamate concentration is the entire point of the ferment.

The duck supplies the other half — post-mortem ATP degradation in muscle yields inosine monophosphate. Glutamate and inosinate are synergistic at the T1R1/T1R3 receptor rather than additive: perceived intensity multiplies rather than sums.

The 2 g of MSG is a small thumb on that scale. It is standard in Isaan cooking rather than a shortcut around one.

The typical lamb flavour comes mostly from branched-chain fatty acids, such as 4-methyloctanoic acid. These dissolve in fat. They are concentrated in lamb fat and in the marrow.

Three steps keep the lamb in the background:

• Cut the visible fat off the lamb bones before they roast.
• Roast the lamb for 20 min only. Browned lamb fat tastes stronger.
• Remove all of the solid fat cap from the cold stock.

Some of these acids dissolve into the water over 10 h. A small lamb note stays in the stock. The pepper, the merlot and the demi-glace in the braise cover it.

Bring the stock up from cold water over about 45 min.

The soluble proteins of the bones set slowly as the temperature rises. They join into large grey rafts on the surface, which a ladle can remove.

If the bones go into hot water, these proteins set at once as very small particles. The particles stay in the liquid and the stock is cloudy.

Dried chilli and fresh chilli are not the same ingredient in this sauce.

Dry-toasting browns the sugars and proteins in the dried pericarp. This is the Maillard reaction. It makes pyrazines, which taste roasted. Capsaicin is stable at this heat, so toasting does not remove the heat. It adds the roast on top.

Fresh bird's eye chilli carries volatile green compounds and water. Its heat arrives at once and is sharp.

The sauce needs both. The dried chilli gives the depth. The fresh chilli gives the top.

Salt and acid pull water out of the cut shallot. This is osmosis. The shallot softens and the sauce gains liquid.

At the same time the toasted rice takes liquid up.

The two effects settle after about 5 minutes. A sauce that looks dry at once is often correct after the rest. Taste it after the rest, not before.

Do not make it more than about one hour ahead. The aroma of lime juice degrades within the hour, and the herbs wilt.

Supermarket dry yeast is a bread strain of Saccharomyces cerevisiae. It is bred to make CO2 fast in dough.

The lemonade is very acidic: estimated pH 2.6–2.9, not measured. The acid slows the yeast. A small dose keeps the rate slow enough to catch the bottles at the correct pressure.

The dose does not limit how much sugar the yeast can ferment. Time and temperature do. The fridge is the stop.

A large dose, or a long warm time, gives bread and sulfur notes.

Drying damages the cell membranes of the yeast. Warm water lets the membranes recover before the yeast meets acid and cold. The foam also proves that the yeast is alive before it goes into 2 L of lemonade.

The sugar pulls juice and oil from the crushed ginger, the same way it pulls oil from the peel.

Raw ginger gives [6]-gingerol. Gingerol activates TRPV1, the heat receptor, so it reads as warmth. Heat changes gingerol into shogaol, which is sharper. This recipe uses no heat, so the ginger stays warm and round.

The ginger is not fermented. This version does not have the yeasty, lactic ginger note of the bug version. That note is the main difference in taste.

The base starts near 165 g fermentable sugar in 2 L (8.2%). The bug version starts slightly higher because the bug liquid carries some sugar.

About 2.5 volumes of CO2 need about 20 g sugar in 2 L. The base ends near 145 g (7.2%). The ethanol is about 0.5–0.7% ABV.

All of these values are calculated. Nobody measured them yet.

Salt pulls water to the surface. In about 40 min the meat takes the brine back in. Chloride ions partly dissolve myosin. The fibres then hold more water when heat contracts them. The uncovered surface also dries. A dry surface browns fast. A wet surface steams.

Pork holds inosinate (IMP). MSG adds glutamate. The umami receptor responds to the two together much more than to either one alone. So 0.85 % salt tastes fuller than it is. This keeps the sodium low.

The time for heat to reach the centre increases with the square of the thickness. A 2.5 cm chop heats through in about 40 % of the time of a 4 cm chop. The centre reaches temperature during the sear, so a reverse sear is not necessary. A chop that comes from a low oven goes past the target during the sear.

Each turn moves the hot face away from the pan. That face cools a little while its heat moves inward. The other face then sears. The centre heats as if from two sides. The cook is faster and more even. The grey band under the crust stays thin.

Myosin sets at about 50 °C to 55 °C. This gives the texture. Actin contracts from about 66 °C. This pushes juice out of the meat. So start the baste at 55 °C, remove the chops at 60 °C, and let the rest finish at 62 °C to 63 °C. Duroc has high intramuscular fat. This fat melts in this range and keeps the meat moist.

Pork fat sits in cells inside a collagen network. Direct contact with the hot pan shrinks the collagen. The fat then runs out, and the remaining tissue becomes crisp. The rendered fat is the cooking fat for the sear.

The cider dissolves the browned residue from the pan. Its malic acid cuts the fat. The seed coat of mustard holds a mucilage that acts as an emulsifier. Cold butter whisked in off the heat stays as small droplets, and the sauce becomes glossy. Hot butter melts and separates.

Demi-glace is brown stock reduced until it sets when cold. Gelatin gives that set. In the pan the gelatin dissolves and the sauce coats a spoon after a short simmer. Gelatin also helps hold the butter in small droplets, so the emulsion is more stable than in a thin pan sauce. Further reduction makes the gelatin too concentrated and the sauce becomes sticky. Stop when it coats the back of a spoon.

The snap of a pickled round comes from the cell walls. Pectin holds the cells together in the middle lamella, the layer between cells.

Heat is the first threat. Above about 80 °C the cell walls lose their structure fast and the pectin goes into solution. A boiling pour cooks the rounds soft. A pour at 60–70 °C lets the brine in without that damage.

The warm pour can also help. Pectin methylesterase, an enzyme in the pepper, works best near 50–70 °C. It changes pectin so the calcium already in the pepper can link the chains. Linked pectin is firmer.

Enzymes are the second threat. Polygalacturonase and microbial pectinases cut the pectin chains over the following weeks. Tannins from the bay leaves slow them. The refrigerator slows them more.

Added calcium chloride is the strongest lever. This version does not use it. Without it, eat the pickle in the first 3 to 4 weeks for the best crunch.

The rounds take their salt from the brine by osmosis. So the salt concentration of the brine sets the seasoning, not the weight of chilli.

This version uses 48 g of salt in 1,350 ml of liquid. That is about 3.6 %.

When you scale the batch, calculate the salt from the total liquid at 3.6 %. Weigh it. Flake salt and fine salt differ by a factor of two by volume.

If you make the brine in two pans, weigh the total salt first. Then divide it. Otherwise the jars are seasoned differently.

Water in the pan holds the temperature at 100 °C. At 100 °C the paste only stews.

In fat the paste gets hotter than 100 °C. Its sugars and amino acids then brown in the Maillard reaction. The raw metallic taste goes.

Use liquid only after the paste has browned. It lifts the brown layer off the pan. Each time the liquid reduces and the paste sticks again, one more brown layer forms.

Tomato holds much free glutamate. Fish sauce adds inosinate. Inosinate makes the tongue respond to glutamate much more strongly than glutamate alone. MSG adds more glutamate.

The soup tastes savoury with less added salt.

This is monter au beurre: cold butter added at the end of a sauce. Beurre monté is a different thing. It is a butter emulsion used to poach.

Butter is an emulsion. Its milk proteins and phospholipids help hold its fat in the soup. The gelatinised starch from the bread also makes the emulsion more stable. The blender makes smaller fat drops than a whisk, so the emulsion is finer.

A boil after this step breaks the emulsion and the fat separates.

Cooking removes the green, fresh aroma compounds of tomato. One raw tomato blended in at the end puts them back. The browned solids give depth and the raw tomato gives the top note.

The smoke compounds and the capsaicin of chipotle dissolve in fat, not in water. The red pigment of tomato does the same. Blooming the chipotle in the fat spreads it through the whole soup. The finishing oil and butter keep these compounds in the soup.

Most truffle aroma is volatile sulphur compounds, such as dimethyl sulphide and bis(methylthio)methane. These dissolve in fat.

Butter dissolves them, but it gives them no background. The sauce tastes rich but empty.

Demi-glace does three things that butter cannot:

• Gelatin. It makes the sauce thick. The sauce stays on the tongue longer, so the aroma stays longer.
• Glutamate. The reduced stock is high in free glutamate. The truffle then reads as deep, not only as a smell.
• Maillard compounds. The roasted bones and the long reduction give a complex brown background. The truffle sits on top of it and does not disappear into fat.

Truffle aroma compounds are small and volatile. On a hot pan they leave the sauce as vapour before the dish reaches the table.

The warm demi-glace has enough heat to release the aroma into the fat of the sauce. It does not have enough heat to drive the aroma off. The kept pieces on top of the plate release their aroma from the heat of the pasta, at the moment the dish is served.

Duck breast has a thick layer of fat under the skin. In a hot pan the skin browns before this fat melts. The result is burnt skin over a layer of soft fat.

A cold pan heats the skin slowly. The fat melts and runs out before the surface reaches browning temperature. The crosshatch cuts give the fat a path out and add surface.

Salt on the skin pulls water to the surface, and water slows crisping. Salt the meat side only.

Parmesan adds more glutamate. The sauce already holds a lot of glutamate from the demi-glace, so parmesan makes the dish heavier and less clear.

Pecorino is sheep milk cheese. It is sharper and more acidic. The acidity gives a bright edge next to the truffle.

Oxtail, shin and short rib all hold a lot of collagen. Collagen converts to gelatin slowly from 70 °C and much faster above 80 °C. The long braise dissolves it into the liquid.

Below approximately 40 °C, dissolved gelatin sets to a gel. So the chilled liquid sets to a firm jelly, and the fat rises and sets as a cap that lifts off in one piece.

The cold filling goes into the pastry as a solid. A solid filling gives off less steam and no free liquid while the base sets. In the oven the gel melts again and becomes the gravy.

This is why the filling must be cold, not warm, when the pie is assembled.

Puff pastry takes up water fast because it is thin layers of dough and butter. The base needs three barriers.

Blind bake. Heat drives water out of the pastry and sets the starch into a rigid structure before the filling touches it.

Egg white. Egg white proteins set from approximately 62 °C into a thin film that water does not easily cross. Brush it on the hot base so it sets at once.

Breadcrumbs. Dry crumbs take up the first liquid that meets the base.

The frozen assembled pie then goes into a hot oven on the lowest shelf. The butter in the pastry is still solid, so its water turns to steam and lifts the layers. The base gets direct heat before the filling warms enough to give off steam.

Reducing the wine by half in the open pot boils off most of the alcohol and concentrates its acid, sugar and tannin. In the closed braise the alcohol cannot leave, and a raw wine taste stays in the gravy.

Wine tannins bind to protein. Against the fat and gelatin of the filling they read as structure, not as bitterness.

Aluminium conducts heat approximately 200 times faster than borosilicate glass: approximately 237 W/m K against 1.1. On a preheated steel, heat goes through the foil into the pastry base in minutes. The base dries and its starch sets before the filling can wet it.

The frozen puck delays the filling. It must take up the heat of melting, approximately 334 J/g of water, before the gelatin can melt at approximately 35 to 40 °C. During that time the base bakes with nothing wet against it.

The cold puck also keeps the fat in the base pastry solid until the pie goes in the oven. Solid fat lets the water in the dough turn to steam between the layers, so the base puffs and does not go leathery.

Tendon is almost pure collagen, packed tightly and bonded across its fibres. The bonds get stronger as the animal ages and as the muscle works. They resist heat much more than the loose collagen around the muscle cubes.

A 3 h braise softens the cubes but only starts on the tendon. In an earlier cook, tendon from a braise stayed chewy.

Cook the tendon whole in the stock overnight, at a light simmer. Cut it when it is cold and set. Add it to the meat after the braise, so it does not cook a second time and fall apart.

Cornstarch is approximately 25 percent amylose. Amylose is a straight-chain starch. In the cold and during freezing, the chains line up again and push water out of the gel. The gravy then weeps. This is retrogradation.

Gelatin sets by a reversible network. Ice can loosen it a little, but it melts back into gravy in the oven.

So this filling gets its body from reduction only: oxtail, tendon, the reduced stock and the demi-glace.

Calcium holds the proteins of cheese together. As cheddar ages, its proteins break down into small, tight groups. When heated, these hold on to each other and the fat leaks out as grease.

Emmentaler and low-moisture mozzarella are younger and more elastic. Their protein network stretches around the cheddar fat and holds it in.

Ratio by weight: cheddar 40, Emmentaler 40, mozzarella 20.

Shokupan is made with tangzhong, a cooked flour paste that holds much water. It also holds milk and sugar. Milk sugar and added sugar let the crust brown at a lower temperature than plain bread.

On high heat the crust goes dark before the cheese melts. Medium-low heat for longer lets the heat reach the cheese first.

Quark is an acid-set fresh cheese. The culture takes the milk to pH 4.6. This is the isoelectric point of casein. At this pH the casein has no net charge, so the protein comes out of suspension and makes a gel. The whey is then drained out.

Heat makes the casein network contract more. Water is pushed out of the gel. This is syneresis. The quark then looks wet and feels like wet sand.

Fat globules break up the casein network and slow this down. Speisequark has fat, so it resists better than Magerquark. Magerquark holds about 0.2 percent fat and has almost no protection.

Serve the quark at room temperature. Do not warm it in a pan.

Egg white has two parts. Thick albumen is a gel. Ovomucin fibres hold it together. Thin albumen is mostly water with protein in it.

An egg gets more thin albumen as it ages. The ovomucin complexes come apart. This is why an old egg poaches into loose threads.

The sieve removes the thin albumen. Only the part that holds its shape stays. This makes a vinegar addition unnecessary. It also makes a water vortex unnecessary. Both of those methods only correct the same problem.

The proteins of an egg set at different temperatures. Ovotransferrin sets between 61 °C and 65 °C. Ovalbumin sets near 80 °C. The livetins of the yolk start to set near 65 °C.

Water at 82 °C acts as a thermostat. The egg cannot become hotter than the water. This gives about 30 seconds between a set white and a solid yolk.

Water at a full boil removes that margin. The white becomes rubber before the yolk is correct.

Capsaicinoids and the carotenoid colour of pul biber dissolve in fat. They move into hot butter in seconds.

Dried chilli also holds sugar. Sugar burns above 150 °C. Dried mint holds menthone and carvone. These are volatile and leave the pan as vapour.

Browned butter off the heat sits near 120 °C to 130 °C. This is hot enough to take out the colour and the heat. It is not hot enough to burn the sugar or to drive off the mint.