Skip to main content

Our positionThis is where we stand, and why. The reasoning is on this page. · Reviewed 6 September 2026 · The full position record

Where Conversion Power Comes From

Starch-converting enzyme reaches a mash three ways: the grain makes it, a brewer adds it, or a mould grows it on the grain. We run on the second. The third has been carried through to finished gluten-free lager in a laboratory, and this page says what that showed, what it cost, and what would have to be true before it became practice here.

The answer​

We run route two. External enzymes own conversion, and the malt is chosen for what it brings to the beer. Route one, the malt's own enzymes, is real and measurable and isn't planned on, for reasons The Mash That Wouldn't Convert records. Route three, enzymes grown on the grain by Aspergillus oryzae, the koji mould, has been tested on sorghum by one research group and is documented, not adopted: no process on this site uses it, no page recommends it, and the position record at the bottom of this page names what would move it.

RouteWho makes the enzymeWhat it suppliesWhat it costsStanding here
The graingerminating sorghumalpha-amylase, and more of it with a fourth day; little beta-amylaselot-to-lot variation nobody can plan a commercial mash aroundmeasured; no job assigned
A bottlea brewer, at dosingthe enzymes the objective needs, in the order the bottleneck dictatesmoney, and a process step per enzymethe position: external enzymes own conversion
A mouldA. oryzae growing on the malting grainamylases and proteases, so sugar and free amino nitrogen togethermalting loss, a mould-control problem sorghum already has, and open questions on filtration and flavordocumented on sorghum at bench scale; not adopted

Route one: the grain makes it​

Sorghum malt builds alpha-amylase as it germinates, and builds more if germination runs a fourth day: in the USDA Madison screening, all 25 cultivars rose, by 19.2 DU on average. What it doesn't build is beta-amylase, the enzyme that makes maltose, and that's why its diastatic power sits far below barley malt's.1 So the grain supplies real conversion power of one kind, in an amount that depends on the cultivar and the schedule. A lager mash leans hardest on the other kind. We measure it and we don't plan on it. That reverses the barley habit, where the DP line on the malt sheet is a planning input; here the line is descriptive, and the enzyme system owns conversion whatever the lot reports. The schedule reading is a theory, stated on Open Theories.

Route two: a brewer adds it​

Nothing restated. Which enzymes, in what order, for which job is on External Enzymes; the mash that runs them is The Reliable Mash; the failure that fixed the doctrine is The Mash That Wouldn't Convert.

Route three: a mould grows it on the grain​

Sake has taken this route on rice for centuries. On sorghum, one group at Tecnológico de Monterrey has taken it through malt, wort and finished lager in two papers, and they're the evidence this section rests on.23

What it showed. White sorghum inoculated with 1% A. oryzae after steeping and malted four days germinated as well as the control (about 98%). Alpha-amylase rose; beta-amylase didn't move. After double mashing, wort from the koji malt carried 27% more fermentable carbohydrate and 24% more free amino nitrogen than wort from plain sorghum malt, and came within 8.8% of barley-malt wort on fermentable sugar, the gap sitting in maltose.2 Carried through to beer, worts set to 15 °P from the koji malt yielded 21.5% more volume than plain sorghum-malt wort and 5% more than barley-malt wort, and the koji beer matched the barley beer on gravity and alcohol where the plain sorghum beer ran about 11% lower in ethanol. The koji beer also ran far lower in hydrogen sulfide and vicinal diketones.3

What it cost, on the same pages. The koji malt lost 14.8% of its dry matter to malting against 9.9% for the control. The steep water carried 0.01% formaldehyde to hold mould down, which is the study's answer to a problem sorghum malting already has. This site wouldn't publish it as practice. The grist was 60% malt with 40% maize adjunct; an all-malt gluten-free grist hasn't been run. Fermentation ran in 3-litre bioreactors, on one white sorghum, one strain, one location, and the flavor differences are reported by instrument; nobody tasted the beer, and the authors say the sensory and stability work is still to do.3 A separate 2023 study that used sorghum koji as a 30% adjunct found koji-only worts hard to filter and koji activity dependent on the cultivar, the same lot-variability problem this site tracks everywhere else (the directory entry).

What a mould buys that a bottle can't. Commercial brewing enzymes are carbohydrases; they make sugar and stop. Koji secretes proteases alongside its amylases, which is where the extra free amino nitrogen comes from, and gluten-free wort runs chronically short of it (Yeast Nutrition). That's the one thing on this page route two can't do, and it's why the route stays on the page rather than in the directory alone.

A fourth route, mentioned and not mapped​

The Chinese and Korean starter traditions, jiuqu and nuruk, saccharify grain with mixed cultures in which Rhizopus and other organisms carry the load. They're catalogued in the koji and grown enzymes directory; we've found no study applying them to a gluten-free brewing grain and we take no position on them.

What this changes for you​

  • Nothing in the protocol. The Reliable Mash runs on route two and this page doesn't move it.
  • If a maltster offers koji-inoculated sorghum malt, treat it as a malt with a specification to verify: extract, FAN, filtration behavior on your system, and above all how mould was controlled in the steep, since the published method used formaldehyde. The route is the same; the malt is new. The certificate of analysis questions apply unchanged.
  • If you want to run route three as research, Ask 13 says what would count.
Why we take this position

Basis. GFB process-design judgment on two peer-reviewed bench-scale studies from one group, read in full or by abstract on 2026-09-06, against the commercial production experience behind the external-enzymes position.23 This is a process-design position. Whether the mould works is a question for replication.

Why. The evidence for koji on sorghum is real and it's bench-scale: one cultivar, one strain, 3-litre fermenters, a 40% adjunct grist, a formaldehyde steep, no sensory panel. A route becomes practice here when it has run on an all-malt gluten-free grist under a mould-control regime a maltster could operate, at a scale where lautering and flavor have been seen, with replicates. None of that exists yet, and saying "promising" in a guidance page would be reading it as if it did.

The problem forcing it. The same one that forces route two: current gluten-free malt does not supply reliable endogenous conversion enzymes.

Compensation or ideal? Route two is a compensation for the malt, and we say so on its own page. Route three is a candidate for the "validated malt under a published malting specification" that the external-enzymes record names as what would narrow its position; it is not yet that malt.

Evidence.

  • Heredia-Olea, Cortés-Ceballos and Serna-Saldívar (2017): enzyme activities, wort composition, FAN, at 1% and 2% inoculation.2
  • Rubio-Flores, García-Arellano, Perez-Carrillo and Serna-Saldivar (2020): the same approach carried to finished lager, with yields, fermentation and volatiles; open access.3
  • Limiting evidence: dry-matter loss, the formaldehyde steep, the adjunct grist, the scale, and the absence of sensory data are all in the same two papers.
  • Not held: any koji-sorghum malt made or brewed outside that group; any all-malt gluten-free trial; any filtration data on an all-malt koji grist.

Open contradictions. The 2020 koji beer matched barley beer on gravity and alcohol, which a reader could take as the malt problem solved. It doesn't contradict this position, which is about what has been shown at what scale rather than about whether the mould works.

What would change our position. Any of:

  1. Koji-inoculated sorghum malt, made without formaldehyde under a published mould-control method, brewed on an all-malt gluten-free grist under the reference mash, reporting extract, FAN, iodine result, runoff, attenuation and a tasted beer, with replicates, at more than bench scale.
  2. A commercial gluten-free malt made this way, with a certificate of analysis a brewer can hold the maltster to.

Either would move koji from "documented" to "a validated malt that may reduce supplementation," which is the narrowing the external-enzymes record already allows for. Neither would make koji the default.

Related research asks.

Commercial interests. None. We hold no interest in any koji culture, spore supplier, or maltster.

Consequence if wrong. Process. If koji turns out to be the better route and we've held it at "documented," a brewer following us spends more on enzymes than needed. Nobody is made unsafe by it; A. oryzae holds GRAS status in the United States and the food-safety line runs through the directory.

Decision owner and status. Craig Belser · Current · position ruled 2026-09-06 ("documented, not adopted") · DR-0080.

History. Before this page the guidance side of the site knew two routes and the research side named three. The koji route appeared only in the research directory and in an intake note that said the intervention was not adopted as practice. This page makes that a position with an owner and a test, 2026-09-06.

What the evidence supports​

Each source below is named beside the claim it carries on this page; the footnotes hold the full citations.

  • Muoria, J.K., Linden, J.C. & Bechtel, P.J. (1998) supports: What it doesn't build is beta-amylase, the enzyme that makes maltose, and that's why its diastatic power sits far below barley malt's.1
  • Heredia-Olea, E., Cortés-Ceballos, E. & Serna-Saldívar, S.O. (2017) supports: On sorghum, one group at Tecnológico de Monterrey has taken it through malt, wort and finished lager in two papers, and they're the evidence this section rests on.2
  • Rubio-Flores, M., García-Arellano, A.R., Perez-Carrillo, E. & Serna-Saldivar, S.O. (2020) supports: On sorghum, one group at Tecnológico de Monterrey has taken it through malt, wort and finished lager in two papers, and they're the evidence this section rests on.3

References​

  1. Muoria, J.K., Linden, J.C. & Bechtel, P.J. (1998). Diastatic Power and α-Amylase Activity in Millet, Sorghum, and Barley Grains and Malts. Journal of the American Society of Brewing Chemists 56(4):131–135. doi:10.1094/ASBCJ-56-0131 ↩ ↩2

  2. Heredia-Olea, E., Cortés-Ceballos, E. & Serna-Saldívar, S.O. (2017). Malting Sorghum with Aspergillus oryzae Enhances Gluten-Free Wort Yield and Extract. Journal of the American Society of Brewing Chemists 75(2):116–121. Figures from the abstract; the full text is paywalled. doi:10.1094/ASBCJ-2017-2481-01 ↩ ↩2 ↩3 ↩4 ↩5

  3. Rubio-Flores, M., García-Arellano, A.R., Perez-Carrillo, E. & Serna-Saldivar, S.O. (2020). Use of Aspergillus oryzae during sorghum malting to enhance yield and quality of gluten-free lager beers. Bioresources and Bioprocessing 7:40. Open access, CC BY. doi:10.1186/s40643-020-00330-w ↩ ↩2 ↩3 ↩4 ↩5 ↩6

Research needed