GPT5 is a master distiller…

grim

How would you go about minimizing acetic acid esters but maximizing all others?

Love this question. Short version: block acetic acid and acetate-ester pathways while feeding the pathways that make medium/long-chain fatty-acid esters (butyrate, hexanoate, etc.). Here’s a practical, brewery-floor playbook.

Goals & trade-offs

• You’re trying to minimize all acetate esters (ethyl acetate, isoamyl acetate, phenethyl acetate, etc.) without killing fruity/“funky” non-acetates (ethyl butyrate, ethyl hexanoate, ethyl lactate…).
• That means: suppress acetic acid + yeast AAT (alcohol acetyltransferase) activity early, encourage other acid production later, and cut heads hard in the still.

(1) Yeast & bacteria choices

• Yeast strain: Pick a low-acetate-ester strain but high MCFA-ester producer (brewing specs sometimes call this “low isoamyl acetate / high ethyl hexanoate”). Avoid “banana-forward” strains.
• Co-culture: If you want big pineapple/apple notes, allow LAB + mild Clostridia contribution via clean dunder/muck to make butyric/hexanoic acids—but keep Acetobacter out (oxygen control).

(2) Wort/wash prep (molasses/cane)

• Hygiene: Sanitize lines/vessels; exclude oxygen to keep Acetobacter from turning ethanol → acetic acid.
• Backset/dunder acidity: Start pH 3.8–4.2 (acidified dunder works). Below ~4.5 suppresses acetic-acid bacteria while still allowing LAB.
• Nutrients: Adequate FAN, zinc, Mg to reduce yeast stress (stress drives acetate esters). Add yeast hulls/lipids if the wash is lean.

(3) Fermentation regime (the big lever)

Early phase (0–24 h): keep acetate esters down

• Pitch rate: 1.5–2× your normal. Higher pitch = less acetate ester formation.
• Temperature: Cooler start (24–26 °C / 75–79 °F) to reduce AAT activity during growth.
• Oxygen: One small oxygenation before pitch only (if you must for sterols), then no aeration. Keep DO negligible after 2–3 hours.
• Lipids/UFAs: Ensure sterol/unsat. fatty acid availability (via hulls or a tiny food-grade oil addition) — this represses acetate ester synthesis.

Mid/Late phase (24–96+ h): boost non-acetate esters

• Temp ramp: Let it rise to 28–32 °C (82–90 °F) after high-krausen to promote formation of ethyl butyrate/hexanoate/lactate and chemical esterification.
• pH drift: Allow pH to fall (3.6–3.9) via LAB; low pH speeds non-enzymatic esterification of non-acetic acids.
• Dunder charge: Add clean, acidic dunder late or to the next ferment to raise butyric/hexanoic acid precursors—not vinegar-smelling (that’s acetic).

(4) Holding / conditioning

• Anaerobic rest (12–48 h) on the lees at low pH accelerates non-acetate esterification. Keep vessels sealed; avoid splashing/oxygen pickup.

(5) Distillation strategy

• Heads cut deeper. Ethyl acetate and acetaldehyde ride the heads. Be aggressive here; recycle heads cautiously (or not at all) if acetate is an issue.
• Slow early takeoff to sharpen separation.
• Copper contact is fine; it won’t “fix” acetates, so the cut is what counts.

(6) Aging choices

• Minimize oxygen ingress: Keep barrels full, tight bungs, larger formats if possible, cooler warehouse zones. Less oxygen → less ethanol → acetic acid formation → fewer acetate esters later.
• Acid balance: You’ll still esterify with wood-derived acids, but lower acetic baseline means you’re steering toward non-acetate complexity.

(7) Quick experiment plan (low effort, high signal)

Run 4 x 20–50 L trials and taste + lab-GC if you can:

1. Control: your normal process.
2. Low-acetate start: high pitch, cool start, lipids, no post-pitch O₂.
3. Low-acetate + dunder: (2) plus late-ferment dunder; pH 3.7 end.
4. Temp-ramp emphasis: (3) plus a deliberate 24→31 °C ramp at 36 h.

Measure: ethyl acetate, isoamyl acetate, ethyl butyrate, ethyl hexanoate, ethyl lactate; confirm with sensory (pineapple/apple vs. nail-polish).

Fast checklist

• ☐ pH start 3.8–4.2; end ~3.6–3.9
• ☐ High pitch; cool start (24–26 °C), then ramp (28–32 °C)
• ☐ No oxygen after pitch; airtight transfers
• ☐ Nutrients + hulls (reduce yeast stress)
• ☐ Encourage LAB/dunder for butyric/hexanoic acids (not vinegar)
• ☐ Deep heads cut; limited heads recycle
• ☐ Tight aging (low O₂)

If you share your current recipe, pitch rate, and cut points, I can tune the numbers (and even set a small DOE) to hit your target ester profile.

Smaug

That is pretty amazing.

How much of this is grim, and how much of this is AI?

grim

All AI

Smaug

@grim said: All AI

Yeah, doing a nose dive into the rabbit hole nearly all day today with AI generated bottle and label designs. Thing is pretty handy.

Bolverk

Been playing with chat gpt myself... ive used it to design logos and labels. This stuff is pretty handy.

Just thinking now. Use it to amas all the known knowledge from every master distiller in the last 200 years and use it as your own personal coach

jacksonbrown

What was the prompt?

Assuming the path from sterol to long chain acids is through division?

richard

I tried AI this morning for a label design and was suitably impressed. It gave me the following solution

ROYVILLE DISTILLING - Gin Bottle Label Design Concepts (PDF)

grim

The prompt was pretty simple, it was a two-prompt conversation:

1. Help me understand ester formation in rum fermentation
2. How would you go about minimizing acetic acid esters but maximizing all others?

SingleMaltYinzer

So the question for me would be does it actually work? For me AI isn't quite there yet for general stuff, but it seems to work really well for applications where the AI is tuned to do specific things.

For that it is recommending: I do see a lot of things listed for which I've not see any research around, but that doesn't much as AI has more access to papers I've not read. It also can formula ideas based on non-distilling specific topics (eg lateral thinking). The advice seems somewhat practical if not pretty interesting. Until someone tries it we won't know.

richard

As with everything, it's a tool