Most preservative failures are not failures of the preservative. They are failures of contact: the active never reached the microbe because something in the formulation or the process consumed it first. This article is a diagnostic guide for products preserved with isothiazolinones – chiefly the CMIT/MIT mixture (CMIT CAS 26172-55-4, MIT CAS 2682-20-4) – and it also applies where BIT or bronopol is used. It covers the degradation pathways, the incompatibility list, how to read a certificate of analysis as evidence, and how to fix a formulation that is losing its protection.
1. The four degradation pathways
| Pathway | What happens chemically | Signature in analysis |
|---|---|---|
| Hydrolysis | the isothiazolinone ring opens in neutral and alkaline conditions, giving a non-biocidal amide-thiol derivative | both actives fall, ratio roughly unchanged, loss accelerates with pH and temperature |
| Dechlorination / reduction | CMIT loses its 5-chloro substituent and becomes MIT; sulfites, bisulfites, thiosulphates and ascorbate do this readily | CMIT falls sharply, MIT rises or falls less, ratio drops toward the lower limit, total falls slowly at first |
| Nucleophilic attack | amines, thiols and some polymer residues open the ring and bind the fragment | loss proportional to the nucleophile loading; often worse at higher pH where the amine is unprotonated |
| Metal-catalysed oxidation | iron and copper accelerate degradation and generate colour | Gardner colour climbing, loss without a matching pH or reductant explanation, discoloured product |
Distinguishing these four is the whole diagnostic exercise, and it is why the ratio line on a certificate is more informative than the total active. A total that fell with an unchanged ratio means hydrolysis. A ratio that collapsed means reduction or nucleophilic consumption. Colour that rose without much assay loss means metal contamination or light exposure.
2. The pH window, stated precisely
| pH of the product | Expected behaviour of CMIT/MIT | Practical consequence |
|---|---|---|
| below 3 | stable, but the concentrate itself is corrosive and handling is the constraint | typical for the 14% concentrate, not typical for a finished consumer product |
| 3 – 6 | good stability; degradation slow at ambient temperature | the comfortable design window |
| 6 – 8 | measurable loss over months; acceptable if the dose includes margin and the shelf life is modest | verify by ageing and re-assay rather than assuming |
| 8 – 9 | rapid loss; a dose that looks adequate at day zero may be gone within weeks | move to BIT or a blend; a stabiliser will not fix this |
| above 9 | the family is the wrong choice | change chemistry rather than increase dose |
The temperature effect compounds with pH: a rule of thumb for hydrolysis is that the rate roughly doubles for every 10 °C, so a product stored at 40 °C in summer ages several times faster than one kept at 25 °C. A dose validated in winter can fail a batch produced in July – not because the preservative changed, but because the storage condition did.
3. The incompatibility list
| Component in the formulation | Interaction | What to do |
|---|---|---|
| Sodium metabisulphite, sodium sulfite, thiosulphate | immediate dechlorination of CMIT; large losses within hours | do not co-dose; if a reductant is unavoidable, protect with a different chemistry or dose in a separate stage and re-assay |
| Ascorbic acid and ascorbates | reduction, same mechanism as sulphite | separate the addition points; expect loss |
| Thiols, thioglycolates, mercaptans | fast nucleophilic consumption | incompatible in practical terms; choose another family |
| Primary and secondary amines, amine-neutralised resins, AMP, ethanolamine | nucleophilic attack, worse at higher pH where the amine is free | add the biocide after neutralisation, at the lowest workable pH, and test efficacy on the aged product |
| Iron, copper, carbon steel, rust particles | catalytic degradation and colour | stainless or plastic contact throughout, including drums, pumps and transfer lines |
| Hypochlorite, peroxides, persulphates | oxidative destruction, sometimes violent with concentrated solutions | never co-dose; flush lines between sanitisation and production |
| Some thickeners and associative polymers | adsorption onto the polymer, reducing available active | increase the dose only after confirming partition; measure free active in the aqueous phase if needed |
| High surfactant load and micellar systems | partitioning into micelles, lower free concentration | test in the real formula, not in water |
| Anionic/nonionic mixtures in high solids paints | gradual loss plus colour drift | long-term ageing study at 40 °C before committing to a shelf life |
4. Diagnosing a failure from certificates and records
The most useful evidence is a comparison of the incoming biocide COA with an assay of the finished product at day zero and after ageing. Very few plants keep all three, which is why the same failures repeat.
| Observation | Most likely cause | Next test | Corrective action |
|---|---|---|---|
| Day-zero assay already well below what was dosed | immediate reaction with a reductant or amine, or a dosing or mixing error | repeat the assay on a sample taken before and after each raw material addition | move the addition point; remove or replace the incompatible component |
| CMIT:MIT ratio far below the incoming ratio | dechlorination or selective consumption of CMIT | check for sulphite, ascorbate, thiosulphate, thiols | eliminate the reductant; consider MIT-only or BIT |
| Both actives down, ratio unchanged | hydrolysis driven by pH or temperature | pH profile over time; storage temperature log | reduce pH if the product allows; shorten declared shelf life; change family |
| Gardner colour climbing, assay falling | metal contamination | iron and copper content of the batch; inspect drums and lines | replace carbon steel contact, filter, consider a chelating strategy for the metal rather than the biocide |
| Assay fine, product still spoils | the dose is below the requirement for that contamination load, or the organism is not covered | identify the spoilage organism; run a preservative efficacy test with it | raise the dose within regulatory limits or add a partner active with the right spectrum |
| Assay fine, film grows mould on the substrate | in-can active does not protect the dry film | film preservative content of the dried coating | add OIT, IPBC or zinc pyrithione to the wet paint |
| Batch-to-batch variation in spoilage | process hygiene rather than preservative chemistry | microbial counts of raw materials and of the plant water | clean the line; preservatives are not a substitute for sanitation |
5. Proving it: the ageing-plus-challenge protocol
A preservative system is validated by two measurements taken together: how much active is left after ageing, and whether the aged product still resists a microbial challenge. Neither alone is sufficient.
- Assay the incoming biocide and keep the COA in the batch record. Note total active, ratio, DCMIT, pH and colour.
- Assay the finished product at day zero. If the recovered active is below about 90% of the dosed amount, something consumed it during manufacture; fix that before adjusting the dose.
- Age samples at 40 °C for 2, 4 and 8 weeks, and at ambient for the intended shelf life. Re-assay active at each point and record pH and colour.
- Challenge the aged samples with the standard panel required by the destination market’s method, plus any organism recovered from the plant.
- Write the dose into the batch record with the addition point and the temperature, and audit it. Most “random” failures trace back to an addition that happened at the wrong stage.
This protocol also produces the evidence a customer audit or a registration dossier will ask for: a declared dose, a demonstrated recovery, and a challenge result on an aged sample. It is more work than a day-zero test and it is the difference between a specification and a control.
6. When the answer is a different chemistry
If the formulation is alkaline, contains a reductant that cannot be removed, or targets a market where isothiazolinones are restricted in that use category, adjusting the dose will not solve the problem. The realistic migration paths are these.
| Situation | Migration path | Consequence to plan for |
|---|---|---|
| pH 8 – 10 product | BIT, or a BIT/CMIT blend where the residual acidity keeps the CMIT fraction alive | slower kill; longer efficacy test; higher dose in some systems |
| Reductant unavoidable (for example an antioxidant in the formula) | BIT or bronopol, with a stabilised package | bronopol brings formaldehyde-releasing chemistry and a nitrosamine question with amines |
| Cosmetic or consumer-adjacent category | actives permitted by that market’s cosmetic rules, at regulated levels | different hazard labelling and a new safety assessment |
| Process sanitation rather than storage | DBNPA or an oxidising sanitiser, with a separate storage preservative | two products, two dose points, two sets of records |
| Dry-film fungal growth | OIT, IPBC or zinc pyrithione added to the wet coating | additional cost; leachability and outdoor durability testing |
7. Storage and handling of the concentrate itself
Concentrate stability is a separate question from in-product stability, and it fails in different ways. Store the 14% grade between 5 °C and 35 °C in the original closed container, out of sunlight, never on carbon steel, and never below freezing: if a drum freezes and is then used without homogenising, the top and bottom will differ in concentration. Opened drums should be resealed immediately and consumed within the campaign. A drum that has been open for a season in a warm warehouse is not equivalent material, even if the label has not changed, which is why incoming re-assay after long storage is worth the cost for critical batches.
8. FAQ
My product passed the challenge test at day zero but spoiled after six months. What went wrong?
Almost certainly the active was consumed during storage, not that the test was wrong. The usual causes are pH drift upward, storage temperature above the design assumption, or a slow reaction with an amine in the formula. Re-assay the active in an aged sample: the ratio and the total will tell you which pathway ran.
Does adding more biocide fix a failure?
Only if the dose was genuinely too low and the formulation is compatible. If the active is being consumed, a higher dose simply buys a few more weeks and adds hazard classification and cost. Diagnose first.
Can I mix two preservatives to be safe?
Yes, and it is common practice, but only with actives that are compatible with each other and with the formulation. Some combinations are antagonistic – oxidisers with isothiazolinones, for example – and blending without testing can reduce total protection.
Why does the certificate show a ratio of 3.0 when I asked for a 3:1 molar mixture?
Because the ratio is a mass ratio. A 3:1 molar mixture corresponds to a mass ratio of about 3.90, which is outside the usual specification window. Ask which basis the supplier uses. See the CMIT/MIT technical article for the arithmetic.
How often should incoming biocide be re-assayed?
At minimum on receipt and after any storage longer than six months, plus after any drum that has been opened and closed repeatedly. The cost of one HPLC assay is small compared with a failed production batch.
Conclusion
Treat the preservative as a component with a lifetime, not as an inert additive. Measure it in the incoming drum, measure it at day zero, measure it again after ageing, and keep the ratio line in view because it distinguishes hydrolysis from reduction. Most failures are contact failures – the wrong pH, the wrong addition point, an amine or a sulphite in the formula – and they are found by arithmetic on the certificate long before they are found by a spoiled batch. For the choice between families see industrial preservative selection; for the chemistry and specification of the isothiazolinone option see the technical article and the CMIT/MIT-14 product page.