L-Cysteine Hydrochloride in Bakery Dough Relaxation: Dose, Mechanism and Troubleshooting

L-Cysteine hydrochloride is one of the few reducing agents a baker can use at ten to a hundred parts per million against flour weight, and the effect is large enough to change how a dough machine behaves. It is used to relax gluten, shorten mixing, and make extensible doughs – crackers, biscuits, pizza and pastry laminates, some frozen doughs – easier to sheet and shape without tearing. Used well it fixes a process; used carelessly it produces sticky, collapsing dough that no proof time will rescue. This article covers what the molecule actually does, how much to use, and how to find the cause when a batch goes wrong.

What cysteine actually does to a dough

Gluten strength comes from disulfide bonds: cysteine residues in glutenin and gliadin subunits are cross-linked into a network, and the number and arrangement of those bridges decide whether a dough springs back or stretches. L-Cysteine carries a free thiol group, and a free thiol attacks a disulfide bond, swapping places with one of its partners. The result is a shorter, less extensively cross-linked network – the dough becomes softer, more extensible and less elastic.

The reaction is an exchange, not a demolition. Cysteine is not consumed stoichiometrically the way a reducing salt would be; it moves disulfide bonds around and can be re-oxidised later in the process. That is why a few tens of ppm can produce a visible effect, and also why the effect is not perfectly linear with dose.

The exchange chemistry in one paragraph

Thiol-disulfide exchange runs fast at dough pH and is catalysed by the same trace metals that catalyse the oxidation of cysteine in storage. Two consequences follow for the plant. First, the response depends on how much free thiol was already present in the flour and on how much oxidising capacity the recipe carries – a dough with ascorbic acid or an oxidase will consume cysteine faster and show less relaxation. Second, the reducing effect is time-dependent: the same dose produces more relaxation after a long rest than in a short one. Dose and time are two dials on the same effect, and it is usually cheaper to turn the time dial first.

The grade and its regulatory position

The commercial article is the hydrochloride monohydrate, CAS 7048-04-6 – the food additive E920 in the European Union, where it is listed as a flour treatment agent. Regulations differ by market, and so do labelling rules: some jurisdictions treat it as a processing aid that needs no ingredient declaration in the finished loaf, others require it to be shown. Confirm the positive list and the labelling position for your market before launch rather than after.

Two specification lines matter more than usual in bakery use. Other amino acids must be chromatographically undetectable, because a savoury or off note from a co-eluting amino acid is not acceptable in a neutral-tasting product. And iron is limited to 10 ppm, which matters because the same trace metals drive both the oxidation of the additive in storage and the interactions in the dough.

Dose: what the tens-of-ppm range buys you

Product type What you are trying to achieve Levels commonly quoted Process note
Biscuits, crackers, hard sweet doughs Extensibility for sheeting and cutting, minimal shrink after lamination 10 to 40 ppm on flour weight Often paired with a reducing process step such as a long rest
Bread and buns Shorter mixing time, softer crumb, better volume in strong flours 20 to 60 ppm Watch the interaction with any oxidant in the recipe
Pizza, flatbread, pastry laminates Dough that extends without springing back during forming 20 to 80 ppm Reduce mixing energy as dose rises
Frozen and part-baked doughs Compensate for strength lost during freezing and storage 30 to 80 ppm Establish by trial on the actual freeze-thaw cycle

These ranges are industry starting points, not a specification. The useful dose depends on flour strength (protein and quality, not just percentage), on the recipe’s oxidising capacity, on dough temperature and on the rest time. A practical protocol is to fix everything else, start at 20 ppm, and step in 10 ppm increments until the dough stops springing back – then take one step back, because the failure mode above that point is unpleasant and irreversible within a batch.

The overdose signature

Excess cysteine weakens the gluten network past usefulness, and the symptoms are consistent across products:

  • Sticky, tacky dough that adheres to the sheeter, the divider and the proofer cloth.
  • Poor shape retention and collapse – the dough spreads in the oven instead of springing.
  • Small volume with a dense, tight crumb, often with a slightly grey, wet-looking interior.
  • Batch-to-batch inconsistency that tracks the operator rather than the flour.

The last point is diagnostic. Because the working dose is in the tens of ppm, a weighing error of a few grams on a batch is a dose error of tens of percent. Cysteine should never be weighed as a neat powder on a plant scale for a small batch; it should be pre-blended on a carrier (flour, starch or a premix base) at a fixed dilution, typically 1:100, and that premix weighed instead.

The oxidant conflict

Component in the recipe What it does to cysteine Practical adjustment
Ascorbic acid Becomes a dehydroascorbic acid intermediate that oxidises thiols; needs oxygen and time to work The two partially cancel; fix the total reducing/oxidising balance and test the pair together
Azodicarbonamide (ADA) Fast oxidant that cross-links proteins, tightening the dough Cysteine dose may need to rise; excessive ADA masks the relaxation
Glucose oxidase and other oxidases Generate hydrogen peroxide, which oxidises free thiols Enzyme and cysteine fight each other; decide which effect the product needs
Transglutaminase Cross-links glutamine and lysine residues without touching thiols Adds strength that cysteine will not reverse; dose independently
Bromate Historical oxidant, banned in most markets Not an option – check local rules before substituting anything for it

Most “cysteine does not work” complaints in a bakery are really a balance problem. A recipe that carries a strong oxidising system can absorb a normal cysteine dose and show almost no relaxation; the answer is rarely a bigger dose, it is to decide which of the two effects the product actually needs.

Process levers before you touch the dose

  • Mixing time and energy – less mixing means less developed gluten, which is part of what cysteine is being asked to deliver.
  • Rest and folding – time does the same work as the additive, at no raw-material cost, provided the schedule allows it.
  • Water level – a slacker dough is more extensible; check whether the real complaint is a hydration problem.
  • Dough temperature – warmer dough relaxes faster and oxidises faster; both effects are temperature-sensitive.
  • Flour choice – if a weaker flour is available at acceptable cost, the additive budget may be better spent there.

The rule of thumb is that cysteine is expensive per kilogram and cheap per tonne of flour, so the decision is not about cost – it is about whether the relaxation can be controlled. In a continuous line with fixed times, the additive is the only lever; in an artisan schedule, time is usually better.

Troubleshooting

Symptom Likely causes, in order First check
Dough still springs back Dose too low; oxidant too strong; flour too strong; mixing too short The oxidant level in the recipe before the cysteine dose
Dough sticky and hard to sheet Dose too high; water too high; enzyme overdose; over-rested dough Halve the dose on one batch and re-run the same schedule
Volume collapse in the oven Dose too high; weak flour; low salt; over-proof Salt and proof time, then the cysteine dose
Same dose, different result week to week Neat powder weighed on a plant scale; different salt form or supplier; storage oxidation Move to a 1:100 premix and standardise on one CAS
Off or savoury note in a neutral product Other amino acids above the chromatographic limit; excess dose The “other amino acids” line on the COA

Incoming checks and handling

Four questions cover most incoming risk on this raw material: which salt form is it (chloride and loss on drying, read together), on what basis is the assay reported, is the thiol intact (transmittance, specific rotation, iron), and are other amino acids absent. The reasoning behind each is set out in the anhydrous versus monohydrate comparison; the short version is that a lot can pass assay and still be the wrong form, partly dried, or partly oxidised.

In the plant, dissolve the material in the water phase rather than adding it as a dry powder to the mixer, and add it late enough that it is not in contact with oxidants for longer than the process requires. Store sealed, cool and dry, and use opened packages within a few weeks – oxidation in an opened bag is slow but not zero, and the first sign is a solution that is no longer clear.

Sourcing and packaging

Pack Typical user Note
25 kg paper bag, polyethylene liner Bakery and premix plants Standard export pack; keep dry in store
25 kg fibre drum Humid climates, longer storage Better moisture barrier than a bag
500 kg or 1000 kg FIBC Industrial premix and flour treatment Confirm liner grade; nitrogen blanket on request

Material is supplied against the AJI 92 food-grade specification, with the batch COA and an SDS in the destination language; pharmacopoeial and low-endotoxin grades are available for the pharmaceutical side of the same molecule. Export documents normally classify it under HS 2930.90. For trial quantities, ask for a 1 kg sample from the production lot rather than a laboratory sample, so that the dose you establish in the bakery matches the material you will buy.

L-Cysteine hydrochloride monohydrate (CAS 7048-04-6): current specification and packaging.

Related on this site: anhydrous versus monohydrate L-cysteine hydrochloride – which salt form the plant should standardise on; L-lysine sulphate versus L-lysine HCl – the same salt-form and dosage logic on another amino acid; and N-acetyl-L-cysteine (CAS 616-91-1) for the acetylated derivative.

Scroll to Top