Aperol's six botanicals are all permitted. The chametz is hiding in the base spirit, the acid and the colour — three components nobody thinks of as ingredients.
Aperol is a trade secret. Separating the disclosed from the repeated is the first job, and the one that makes everything below trustworthy.
Campari Group publicly names bitter orange, sweet orange, rhubarb, gentian and cinchona, states that the liqueur is an infusion in a water–alcohol base, that it is not aged, and that it bottles at 11% ABV. EU labels declare the colourants E110 Sunset Yellow FCF and E124 Ponceau 4R. The nutrition declaration puts sugar in the region of 20–24 g per 100 ml.
Everything else — proportions, extraction times, the acid, whether anything is distilled rather than macerated — is not disclosed. I found no published GC-MS or HPLC analysis of Aperol. The active-principle column below is reasoned from the named botanicals and their known chemistry; it is not a measurement of the product.
Treat the “secret recipe unchanged since 1919” line as marketing. The colourants alone are mid-century synthetics, and the formula demonstrably moved again under EU labelling rules. Any source quoting a botanical count for Aperol is repeating a guess.
Confirmed the producer states it, a label declares it, or published analysis measured it. Inferred a reasoned reconstruction from the named botanical and its chemistry — a hypothesis, not a disclosed recipe. Substitution my own choice, made to satisfy the Pesach constraint or to cover a function the original fills with something disqualified.
Two things change audibly. The colour cannot be matched — no natural colourant reproduces the flat, faintly fluorescent orange of an azo dye, and this build lands warmer and more amber. The base spirit is cane or grape rather than grain, which adds a faint sweetness or vinosity that neutral alcohol does not have. Everything else is a reconstruction of function; those two are a shift of character. Say so when you pour it.
Function first, ingredient second. Every component holds one slot no other component holds; doses are given both as tincture volume and as botanical load per litre of finished liqueur.
| Component | Dose | Role | Contributes | Active principle | Basis |
|---|---|---|---|---|---|
| Gentian rootGentiana lutea · cut, not powdered | 8 g / 100 ml 0.96 g/L |
Defining bitterness | The mid-palate bitter the category is built on — fast onset, sharp, gone quickly. Deliberately restrained: Aperol sits far below Campari on bitterness and that gap is the point. | Amarogentin, bitterness value ~58,000,000 — the most bitter compound known. Plus gentiopicroside. | Confirmed dose inferred |
| Rhubarb rootRheum palmatum / officinale — root, never stalk | 6 g / 100 ml 0.60 g/L |
Signature note | The earthy, faintly astringent thing that stops this being orange-flavoured sugar. Hardest component to identify by taste and the one most worth getting right. | Anthraquinones — emodin, rhein, chrysophanol — with galloyl tannins for the astringency. | Confirmed dose inferred |
| Cinchona barkCinchona spp. · red or yellow, cut | 2 g / 100 ml 0.16 g/L |
Dry finish | Not redundant with gentian: quinine bitterness is slow, drying and long, where amarogentin is immediate and sharp. Removing it makes the finish stop abruptly. | Quinine (~3% of bark), cinchonine, cinchonidine. Tannins add the drying grip. | Confirmed dose inferred |
| Component | Dose | Role | Contributes | Active principle | Basis |
|---|---|---|---|---|---|
| Bitter orange peelCitrus aurantium · dried, pith scraped | 40 g / 200 ml 18 g/L |
Bridge | Carries flavanone bitterness on the same aromatic as the top note, which is what joins the candied opening to the rooty middle. Without it the two halves read as two drinks. | Naringin and neohesperidin (flavanone bitters) over d-limonene and linalyl acetate. | Confirmed dose inferred |
| Sweet orange zestCitrus sinensis · fresh, outer zest only | 30 g / 100 ml 12 g/L |
Top note | The candied-peel first impression. Volatile and the first thing to disappear — it fades in the bottle within months, which is why it is macerated last and shortest. | d-limonene at 90–95% of the peel oil, with decanal, octanal and linalool carrying the character. | Confirmed dose inferred |
| Component | Dose | Role | Contributes | Active principle | Basis |
|---|---|---|---|---|---|
| TurmericCurcuma longa · whole dried rhizome | 3 g / 100 ml 0.15 g/L |
Corrective — colour | Replaces E110. Gives the yellow-orange base. Contributes almost no flavour at this dose; above ~0.4 g/L it starts reading as earthy and muddies the citrus. | Curcumin — pH-sensitive, shifts red above pH 8, which is why the acid goes in before you judge colour. | Substitution |
| HibiscusHibiscus sabdariffa · dried calyx | 1 g / 100 ml 0.05 g/L |
Corrective — colour | Replaces E124. Pulls the turmeric yellow toward orange-red. Kitniyot-free, unlike the closer match. Adds a trace of tart red-fruit that reads as part of the rhubarb. | Anthocyanins (delphinidin- and cyanidin-3-sambubioside) — red at low pH, dull violet as pH rises. | Substitution |
| Paprika oleoresinCapsicum annuum · pure, uncut — alternative | to shade | Corrective — colour | Closest hue to Aperol by a wide margin, if you do not treat peppers as kitniyot. Use instead of turmeric + hibiscus, not alongside. Must be the pure oleoresin, not a dispersion on a starch carrier. | Capsanthin and capsorubin — carotenoids, oil-soluble, so disperse into the tincture, not into water. | Substitution |
| Component | Dose | Role | Contributes | Active principle | Basis |
|---|---|---|---|---|---|
| Cane or grape spiritdiluted to 60% ABV for tincturing | 110 ml pure ethanol |
Solvent, structure | The single change that makes this Pesach-viable. Standard neutral alcohol for Italian aperitivi is grain-derived — chametz outright. Grape eau-de-vie is the most conservative choice; unaged cane spirit is the most neutral. | Ethanol — extraction solvent for glycosides and terpenes alike, and the reason Stage A and Stage B can share one menstruum. | Substitution |
| Cane sugarKLP-certified | 220 g | Body | Not just sweetness — sucrose physically suppresses bitter perception and supplies the viscosity that makes Aperol coat the glass. This is why the gentian dose can stay low and still register. | Sucrose. ~0.63 ml apparent volume per gram dissolved — the number the calculator runs on. | Confirmed |
| Tartaric acidgrape-derived, KLP-certified | ~1.5 g to taste |
Corrective — acidity | Replaces citric acid, which is industrially fermented on corn or wheat glucose. Grape acid in a grape-spirit base. Also fixes the hibiscus red and stops the sugar reading as flat. | L-(+)-tartaric acid, pKa 2.98 — a sharper, more mineral acidity than citric, which suits the bitter better. | Substitution |
| Gum arabicAcacia senegal · KLP-certified · optional | 1 g | Body, fixative | Replaces glycerol, which is routinely grain-sourced or of unstated origin. Adds the last of the mouthfeel and slows the citrus volatiles dissipating in the bottle. | Arabinogalactan protein complex — a hydrocolloid; also a mild emulsifier for any peel oil that wants to separate. | Substitution |
| Filtered water | to 1000 ml | Dilution | Added last and by volume, not by recipe — tinctures lose unpredictable amounts to botanical absorption and filter retention, so a fixed water figure gives a liqueur of unknown strength. | — | Substitution |
Volatile citrus oil against non-volatile root and bark glycosides. Peel oils are terpene-dominant, extract in hours, and go soapy and pithy if left. Amarogentin, anthraquinones and quinine are non-volatile, need days, and prefer warmth. One bulk maceration gives you flabby citrus or under-bittered roots — never both right. This is the single most common reason a home amaro fails.
The botanicals are almost all fine. Violations cluster in the components nobody counts as ingredients — solvents, carriers, processing aids, equipment.
Twelve days, of which nine are waiting. The day numbers matter: they are staggered so that everything reaches the blend on the same afternoon, at its own correct extraction time.
Every tincture runs at 60% ABV, so dilute once and draw from it. The thing that goes wrong here is doing the arithmetic on the bottle's stated strength after you have already poured some out.
Roots and bark, warm and undisturbed. These are the components you cannot rush and cannot take back out.
40 g dried bitter orange peel in 200 ml at 60%. Scrape any pith off before it goes in; pith is where the peel turns from bitter to acrid and there is no fixing it afterwards.
Press the peel lightly and strain. Then start F: 3 g whole dried turmeric and 1 g dried hibiscus in 100 ml at 60%. If you are using paprika oleoresin instead, skip F entirely and add it drop-wise at Day 5.
30 g fresh outer zest in 100 ml at 60%. Twenty-four hours, timed. This is the component people ruin by treating it like the others: past a day it goes soapy and pithy, and that is preventable only — there is no correction, you remake it.
Strain all tinctures through a coffee filter or fine cloth. Filter F hardest — turmeric throws sediment for days. Then combine, holding back 20% of every bitter and the colour: those are the irreversible directions.
Do not titrate into the batch. Draw 50 ml into a separate glass and work there.
Measure the total volume of everything blended so far and put it into the calculator below. The thing that goes wrong here is topping up to a litre first and calculating afterwards — that gives you a liqueur of unknown strength.
Seven days minimum before you judge it. The bitterness integrates and drops noticeably — a fresh blend tastes harsher than it will be, and this is where people over-correct a drink that was already right.
Quinine. The FDA beverage limit is 83 mg/L. At ~3% quinine in bark and full extraction that permits about 2.8 g of bark per litre of finished liqueur. This formula uses 0.16 g/L — deliberately far under, because quinine causes thrombocytopenia and cardiac conduction effects at doses well below what tastes “too bitter”. You have headroom; treat 1 g/L as your ceiling, not 2.8.
Rhubarb root anthraquinones are a stimulant laxative. This formula puts 0.60 g root per litre — about 0.09 g in a 150 ml pour, comfortably below effect. Do not multiply the rhubarb chasing more signature; go up on gentian instead, which is taste-capped and harmless.
Gentian, both oranges, sugar, acid and colour are taste-capped only. Adjust them freely.
Enter the total measured volume of everything you actually recovered — all strained tinctures plus the make-up spirit, after titration. Exclude sugar and water; those are what this works out. Tinctures lose 10–20% to botanical absorption and filter retention, so the figure will not match the recipe, and that is the point.
FINAL VOLUME = RECOVERED × TINCTURE ABV ÷ TARGET ABV · WATER = FINAL − RECOVERED − (SUGAR g × 0.63)
The 0.63 is the apparent specific volume of dissolved sucrose — 220 g takes up about 139 ml, not zero, and ignoring it is what leaves a “1 litre” batch overfilled and under-strength.
Built to survive dilution, so it should read too intense neat. If it tastes correct on its own, it will vanish in a spritz.
Keyed on what you taste, because that is how the problem arrives — you never notice “too much gentian”, you notice cough syrup.