Diacetyl: What It Is, Why It Shows Up, and How to Get Rid of It

If you've ever cracked open a beer and copped a hit of buttered popcorn, a slick mouthfeel or a weird butterscotch character, that's likely diacetyl. It's one of the most common off-flavours we encounter when brewers bring in samples for us to identify or give feedback on.

The buttery flavour:

Diacetyl is a vicinal diketone, or VDK. The other VDK that comes up in brewing is 2,3-pentanedione, but diacetyl is the one that’s usually talked about because our bodies are absurdly good at detecting it. 2,3-pentanedione (the other VDK produced in beer) flavour threshold sits roughly a whole order of magnitude higher, which is why we’ll focus on diacetyl in the rest of this article.

How sensitive are we really to diacetyl you may ask? This varies between individuals, but a common range is around 0.1-0.2ppm in lager and 0.1-0.4ppm in ales, with trained sensory panels detecting it as low as 10 to 40ppb (Krogerus & Gibson, 2013). That’s parts per BILLION in the trained panel.

The flavour descriptors most commonly used are butter, butterscotch, movie popcorn, and at higher levels a slick, oily mouthfeel. 

Where does diacetyl come from?

Diacetyl isn't really made by yeast directly. It's a by-product of yeast doing something completely unrelated. When yeast is building cell mass during the growth phase of fermentation, it's synthesising amino acids, including valine. That valine pathway goes through an intermediate product called alpha-acetolactate. The yeast excretes some of that alpha-acetolactate into the surrounding beer, where it slowly converts to diacetyl. Krogerus & Gibson (2013). The conversion of a-acetolactate to diacetyl is the bottleneck of this process and this conversion is slow at typical fermentation temperatures and even slower at lager fermentation temperature.

So, during active fermentation, your beer is accumulating both alpha-acetolactate and diacetyl. However this a-acetolactate is slowly converting into diacetyl in the beer. Peak diacetyl typically arrives near the end of primary fermentation.

There is good news though! Healthy, active yeast can also fix the same problem it causes. It reabsorbs diacetyl and reduces it via two reductase reactions to acetoin and then to 2,3-butanediol, both of which have flavour thresholds far above diacetyl’s, ~4500 ppm in the case of 2,3-butanediol (Krogerus & Gibson, 2013). In practical terms, the yeast turns the off-flavour into flavourless bi-products when under the right conditions.

Why it might appear in your beer:

Diacetyl persisting in your finished beer almost always traces back to one of these causes.

The yeast was removed before it had time to absorb and convert it:

Cold crashing too early is the single most common cause we see across the counter. The yeast is still working on VDK reduction right at the end of primary, and pulling the temperature down to 2 to 4 °C before that's done basically pauses the cleanup. If you crashed the moment your hydrometer reading stabilised, you may have stopped the cleanup.

The yeast was unhealthy:

Under-pitching, oxygen-starved wort, weak FAN (free amino nitrogen) levels, or repeated repitching of tired yeast all reduce the cell's ability to handle that final maturation step. Stressed yeast tends to make more alpha-acetolactate on the way up and is slower at reducing diacetyl on the way down. 

Highly flocculant strains dropped out too soon:

 Some English ale strains flocculate so hard and so fast that there isn't enough yeast left in suspension to finish removing and converting diacetyl. This is partly why traditional diacetyl character pops up in some English ales. It can be a feature of the yeast strain when apparent to a small degree.

Bacterial contamination:

 Pediococcus and some Lactobacillus species can produce diacetyl directly, not via the valine pathway. If your beer started clean and then developed buttery flavours late, or if diacetyl re-emerges after you thought it was gone, contamination is worth investigating. Pediococcus is a particularly well-known producer. As a side note some mixed fermentation sours have a distinct Diacetyl character that actually is actually part of the style, similar to some traditional English beers.

Lager fermentation was held cold too long:

Lager yeast is slower at everything, including VDK reduction. Cold lager fermentation accumulates diacetyl, and if you go straight from fermentation temperature down to lagering temperature without a warming step in between, you'll lock it in.

Hop creep from dry-hopping:

This one's particularly relevant if you're brewing hazies or anything else with a heavy dry-hop schedule. Hops carry amylase enzymes (alpha-amylase, beta-amylase, and amyloglucosidase among others) that slowly chew through the non-fermentable dextrins left in your finished beer, releasing fresh fermentable sugars. Any residual yeast still in suspension then ferments those new sugars, and the subsequent fermentation produces more alpha-acetolactate and therefore fresh diacetyl. The catch is that by the time you're dry-hopping, you've often got far less viable yeast in suspension than at peak fermentation, or you might even be dry hopping after fermentation has finished, so the cleanup phase struggles to keep up. The result is a beer that finishes drier than expected, slightly stronger than expected, and has that diacetyl twang. Enzymatic activity varies considerably between hop varieties, crop years, and kilning conditions, so it's not always predictable. The practical fixes are the following: 
- Dry-hop while fermentation is still active (so the yeast handles the refermentation in real time)
- Give post-fermentation dry hops a few warm days in contact with the yeast before you cold crash
- Use a product like ALDC to convert the precursor (alpha-acetolactate) directly into acetoin, this essentially skips the diacetyl step. ALDC is an awesome product and can be added to the beer before fermentation, with dry hops, or both.

 
- Dry hop cold so the yeast isn’t active. You will need to make sure the beer doesn’t warm up or you could have fermentation kick back off and produce the diacetyl anyway.

The diacetyl rest

This is where the diacetyl rest comes in, also called a D-rest.

When fermentation is almost finished but the yeast is still active and in suspension, you raise the temperature for a few days to accelerate two things at once. The spontaneous breakdown of alpha-acetolactate to diacetyl speeds up, and the yeast's reduction of diacetyl to acetoin and 2,3-butanediol speeds up too. Net result, you push everything through the pipeline faster, and the yeast finishes cleaning up before you cold crash.

For a lager, this typically means letting the fermentation rise from primary temperature (around 10 to 12°C) up to 16-20°C for 2 to 4 days, usually starting when fermentation has hit around 70 to 80% of expected attenuation. Don't wait until it's fully done, you want active yeast in suspension when the warming happens.

For an ale, you can often achieve the same effect by simply not crashing your beer the moment it hits terminal gravity. Give it 3 to 5 days at fermentation temperature after the krausen drops, before you cold crash. You can also increase the beers temperature a few degrees in the same approach as a lager if you’d prefer. Diacetyl is generally a bigger issue with lagers though.

The forced diacetyl test (VDK test)

If you want to test if your going to have diacetyl issues you can use a forced diacetyl test, aka a VDK test.

Because the rate-limiting step is alpha-acetolactate slowly converting to diacetyl, you can measure your "total diacetyl potential" by heating a sample to force that conversion to completion. The procedure:

1. Pull a sample of beer into a sealed glass jar or sealed bottle. Fill it close to the top to minimise headspace.
2. Heat it in a water bath between 60 and 70°C for 20 minutes.
3. Cool it back to drinking temperature.
4. Compare the aroma and taste to an unheated control sample from the same batch. The table below can help you make a decision.

This is the test almost every commercial brewery runs before they release a beer from fermentation, and it works the same for home brewing.

When diacetyl is actually wanted

Worth saying: low levels of diacetyl are traditionally acceptable, even desirable, in a few styles. Classic Bohemian Pilsner, some English bitters, some mixed fermentation sours and certain Scottish ales have historically carried perceptible diacetyl as part of the style's character. Modern interpretations are often cleaner than they used to be, but the BJCP guidelines do allow for low diacetyl in these categories. For pretty much everything else (modern lagers, IPAs, hazies, stouts, saisons, you name it), diacetyl above threshold is considered a fault and generally isn’t pleasant.

The takeaways:

If you can do these main things you’ll likely not have diacetyl issues:

1. Pitch enough healthy yeast for your batch size and gravity. Aerate or oxygenate well. Use yeast nutrient if youre repitching yeast multiple times.
2. Don't cold crash the moment your hydrometer reading stabilises. Give the beer a few extra days warm.
3. For lagers, build a deliberate diacetyl rest into your fermentation schedule before lagering.
4. If you're not sure, run a forced diacetyl test before packaging. It takes 20 minutes and tells you whether you're about to package a problem.
5. If diacetyl appears late or returns after disappearing, suspect contamination and review your cleaning and sanitation process.

If you've got a batch you're not sure about, bring a sample in and we'll have a sniff. We’re always happy to help work out what's going on. If we cant pinpoint the problem directly we can reach out to the awesome local community who have plenty of BJCP judges with a great sniffer and taste buds.

References

Krogerus, K. & Gibson, B.R. (2013). 125th Anniversary Review: Diacetyl and its control during brewery fermentation. Journal of the Institute of Brewing, 119(3), 86–97. https://doi.org/10.1002/jib.84