What encapsulation is actually buying

Four benefits account for nearly every encapsulation brief, and it is worth naming which one matters before choosing a technology, because the families differ sharply in which benefits they deliver well.

Protection is the first: an active that degrades on contact with oxygen, light, moisture or the other ingredients of a blend can be sealed inside a wall material that takes the exposure instead. Taste masking is the second: a bitter or astringent extract can be separated from the tongue until a point where the taste no longer matters. Release control is the third: the active can be held until a trigger — heat, moisture, shear, pH, digestion — releases it where and when it is wanted. Handling is the fourth, and the least glamorous: converting a sticky oil or an unflowable paste into a free-flowing powder that a blending plant can weigh accurately.

The trade is that the active is now inside a wall, and everything about the wall — its composition, its thickness, its porosity, its identity — becomes part of the ingredient. The cost per kilogram rises, the cost per unit of active rises faster, and the material’s behaviour in the customer’s process now depends on a delivery system rather than on the extract itself.

The six technology families, and what each one does

Spray drying is the workhorse and the default quotation. The active is emulsified or dissolved with a wall material — maltodextrin, gum arabic, a modified starch, sometimes a protein — and the mixture is atomised and dried in one step. It is fast, scalable and comparatively cheap, and it converts oils into free-flowing powders. Its weaknesses are heat exposure during drying, which matters for volatile and heat-labile material, and particle porosity, which lets oxygen in over time and makes spray-dried powders vulnerable to oxidation in storage unless the wall and the packaging are chosen for it.

Spray chilling and spray congealing embed the active in a lipid matrix that solidifies on cooling. There is no water in the process, which suits moisture-sensitive actives, and release is triggered by melting, which makes the family useful where release at body or processing temperature is the objective. Its limitation is thermal: the matrix melts where the customer may not want it to, and load is bounded by what the lipid will hold.

Fluid bed coating works by layering or coating particles in a bed, building a thicker and more intact shell than spray drying can produce. It is the family to reach for when release control rather than simple protection is the requirement, and it handles coating onto a carrier core well. It is slower and more expensive per kilogram, and the particles are larger, which matters in applications where mouthfeel or suspension is sensitive.

Complex coacervation builds a shell by phase separation around droplets, typically with protein and polysaccharide systems. It offers high load and a shell that can be triggered by heat or mechanical stress, and it is the family behind many heat-release flavour systems. Its costs are the gelatin or protein system it depends on — which carries the animal-origin and certification questions that any gelatin-based system carries — and a more demanding manufacturing process.

Cyclodextrin inclusion is not a capsule in the visual sense but a molecular one: small, mostly volatile molecules are held inside the cavity of a cyclic sugar. It is the family for capturing and dosing aromatics, and for masking a taste at low inclusion rates. Its limit is capacity: the cavity hosts small molecules, not extracts, and the load ceiling is low.

Liposomes and lipid vesicles sit closest to the bioavailability brief, enclosing material in aqueous compartments bounded by phospholipid layers. They are the most demanding family to make, to stabilise and to specify, and their commercial record in dry ingredient applications is mixed; where the objective is simply protection or taste masking, a simpler family usually wins.

Matching the family to the active

The selection follows from three properties of the material rather than from preference.

Physical state comes first. An oil or oleoresin needs to become an emulsion before spray drying can touch it, or a lipid matrix before spray chilling can. A water-soluble powder generally does not need encapsulation at all unless the objective is taste masking or release control. Volatility comes second: volatile aroma compounds are the classic cyclodextrin and spray-drying briefs, with the choice between them driven by load requirements and by how the aroma should be released. Sensitivity comes third: an oxygen-sensitive carotenoid wants a wall with good oxygen barrier properties and low particle porosity; a moisture-sensitive material wants a lipid matrix or a low-water process; an acid-sensitive material wants a shell that survives the pH of the matrix it will be dosed into.

One further matching question is the one most often skipped: what actually destroys the active in the customer’s application. Encapsulation chosen against a generic degradation story, rather than against the specific exposure in the finished product, protects the active against the wrong thing and leaves the real failure mode untouched.

Load, particle size and the trade between them

Two numbers describe the commercial physics of any encapsulated ingredient: load, the share of the particle that is active, and particle size, the distribution of the particles themselves. They trade against each other and against the wall.

Higher load means a thinner wall, which means less protection and faster release of the active to the surface. Lower load means more carrier per unit of active, which moves the cost per unit of active up and consumes space in a formulation. Particle size affects handling — fine particles dust and segregate in a blend, coarse ones settle in a suspension — and it affects mouthfeel in any application where the particle is felt rather than dissolved. In a beverage, particle size and density decide whether an encapsulated ingredient stays dispersed; in a tablet, they decide whether it segregates in the feed frame.

Surface active is the number that ties the trade together: the share of active that sits on the outside of the particle rather than inside the wall. It is where taste masking fails, because the tongue meets the surface fraction first, and where oxidation begins, because the surface fraction is unprotected. It should be measured and stated, not inferred from the payload figure.

Release, and whether your matrix provides the trigger

An encapsulated active that never releases is indistinguishable from an absent one, and release behaviour is where encapsulation projects most often fail quietly. Each family has its trigger: melt for lipid matrices, dissolution for water-soluble walls, mechanical rupture for coacervates, diffusion for porous spray-dried particles.

The specification question is whether the customer’s matrix actually provides the trigger. A melt-release particle dosed into a product that never warms above the melting point stays sealed. A pH-triggered shell in a matrix that never reaches the trigger pH does the same. A diffusion-release particle in a dry blend releases into the headspace rather than the product. Asking a supplier for the release conditions and then checking them against the real matrix — its temperatures, its pH, its water activity, its residence time — is the single most valuable technical exchange in an encapsulation project.

What the specification should say

An encapsulated ingredient needs more specification lines than the same active unencapsulated, and the absence of any of them makes two quotations incomparable. The payload, with the basis stated and the assay method named, because payload reported as-is and payload on a dry basis are different figures. The wall material, by identity, including its origin where the claim situation requires it — the animal-origin question that protein-based systems carry applies here in full. The particle size distribution, with the method. The surface active, because taste masking and oxidation both live there. The water activity and moisture, since a wall with moisture uptake behaves differently in a hygroscopic blend. The release conditions, stated as the trigger and the conditions under which it has been demonstrated. And the physical form with flow and dust properties, since the handling benefit is often the reason the material was encapsulated at all.

Cost, scale and the commercial realities

Encapsulation is usually a toll operation, and the economics follow from that. Minimum batch sizes at coaters are large relative to the volumes of many botanical projects, which means the technology becomes economic only above a threshold of volume, or when the value of the protected active is high enough to carry the conversion cost. Lead times are longer than for a conventional extract, and the coating plant becomes a declared subcontractor whose change is a process change.

The comparison to run is cost per unit of active delivered, not cost per kilogram, and it should include the losses that encapsulation prevents: overage added to compensate for degradation, rejects caused by taste, and stabilisation otherwise achieved with more expensive packaging. Where those costs are real, the encapsulated form is frequently the cheaper system overall; where they are not, it is an expensive way to buy a specification line.

One further commercial reality is qualification. An encapsulated version of an established ingredient is a new material to the customer’s process, with its own first-batch qualification, its own supplier-approval file and its own change-control history, and the cost of that qualification belongs in the comparison. Where the encapsulated form replaces a conventional ingredient that is already qualified, the switch is a formulation change as well as a purchase, and it should be planned as one.

Frequently Asked Questions

Why encapsulate a botanical extract at all?

Four reasons: to protect an active from oxygen, light, moisture or the other ingredients it meets; to separate a bitter or astringent extract from the palate; to hold release until a trigger such as heat, moisture or pH; and to convert an oil or paste into a powder that a blending plant can weigh. The benefit wanted should be named first, because the technology families differ in which of the four they deliver well.

Which technology suits an oxygen-sensitive carotenoid?

A family that builds an intact, low-porosity shell with good oxygen barrier properties — typically fluid bed coating or a well-designed spray drying system with the wall and packaging chosen for oxygen — rather than an uncoated spray-dried powder with a porous structure. The comparison should be made on surface active and on storage stability data, not on payload alone.

What is surface active, and why does it matter?

It is the fraction of the active sitting on the outside of the particle rather than inside the wall. The tongue meets it first, so taste masking fails at the surface, and oxidation begins there too because the surface fraction is unprotected. It should be measured and stated on the specification; a payload figure alone does not describe it.

How do I know the encapsulated active will actually release?

Ask for the release trigger and the conditions under which it has been demonstrated, then check them against the real matrix the particle will be dosed into — its temperatures, pH, water activity and residence time. A melt-release particle in a matrix that never reaches the melt point, or a pH-triggered shell in a matrix that never gets there, delivers nothing.

What makes encapsulation expensive?

Toll coating plants run large minimum batches, conversion cost is charged per kilogram, and the coating site becomes a declared subcontractor. The comparison that decides the question is cost per unit of active delivered — including the overage, rejects and packaging costs that encapsulation avoids — rather than cost per kilogram of ingredient.

The line between customs documents and market access

The import documentation pack exists to get goods across a border. It establishes the identity of the parties, the classification of the goods for duty, the conformity of the shipment with the purchase specification, and the traceability of the lot. It is largely a snapshot of the shipment.

Market access exists at the level of the product category and the use. It asks whether a botanical ingredient may be placed on the market in a given jurisdiction as a food, a supplement, a flavouring, an additive, a cosmetic ingredient or a feed material; under which procedure that placement is allowed; and what must be submitted, notified or authorised before the first sale. It is largely a question of law rather than of goods, and it is settled before production planning starts, not after the shipment arrives.

The practical boundary between the two is easy to state: a customs pack describes a lot; a market-access position describes a use. A supplier can supply a compliant lot of a material whose intended use is not permitted, and no quantity of analysis in the shipment pack will fix that. Equally, a fully authorised use can be delayed for weeks by a shipment pack that is incomplete. The two need different owners, different lead times and different evidence, and the planning error is treating them as one exercise.

The category decides everything

The first market-access question is not what the material is but what it will be sold as. The same botanical extract can sit in several regulatory categories depending on the finished product, the claim attached to it and the market, and the category determines the procedure, the evidence and the timeline.

In broad structure, the categories a botanical ingredient lands in are the food category, where a long history of consumption in a market can matter enormously; the supplement category, where many markets run a notification or a pre-market review route with their own ingredient lists; the flavouring category, where the material is used for taste rather than nutrition and is assessed against its own framework; the additive category, where use is generally tied to a positive list and to defined conditions; the cosmetic category, where the framework is built around safety assessment and a responsible person; and the feed category, where the authorisation logic differs again, as set out in the feed discussion elsewhere on this blog.

The reason the category decides everything is that the evidence requirements do not transfer between categories. A long history of food use supports a food argument and does very little for a cosmetic or feed argument. A safety assessment built for a cosmetic use does not answer a food question. Where a project contemplates selling the same extract into two categories, the honest plan is two positions rather than one, and the earlier the category is settled, the fewer surprises the evidence pack produces.

What the routes usually ask for

The procedures differ by market, but their shapes repeat, and it is the shapes that a planner needs rather than the specific names.

One shape is the authorisation route: a submission is reviewed by a competent body or an authority before the material can be sold for the intended use, and the review asks for a defined set of identity, composition and safety information. This is the shape behind pre-market approval systems for new ingredients and behind feed additive authorisations.

A second shape is the notification route: the material may be sold once a defined information set has been submitted to the authority, and the authority may object within a window. The burden sits on the notifier to have the evidence ready, and the notification is a trigger for scrutiny rather than an approval.

A third shape is the self-determined route: the material may be used if the responsible party has concluded, on documented evidence, that the use meets the applicable standard, and that conclusion and its basis must be defensible on request. No submission is made, but the file has to exist and be maintainable.

A fourth shape is the history-based route: where a market recognises a demonstrated history of consumption of a food, that history can carry the ingredient into the food category without a full new assessment, and the argument turns on the evidence of that history — what was consumed, at what level, in which form, for how long.

Most botanical projects touch more than one shape, because the same extract may be an established food in one market, a notification in a second and an authorisation in a third. The planning consequence is that market access is a per-market exercise and not a global one, and that a single global claim about an ingredient’s regulatory status is almost never accurate.

The evidence pack, and what it has to hold

Whatever the route, the underlying evidence pack has the same five components, and building them is the part of the work that takes time.

Identity comes first, and it is more than a name: the botanical species and plant part, the manufacturing route, and the analytical methods that distinguish the material from its near neighbours. In a trade where substitution is a recognised risk, identity evidence is the foundation everything else rests on.

Composition comes second: what the material contains, on which basis, measured by which methods, including the carrier and any standardisation. History of use comes third, where the route relies on it, and it needs to be documented rather than asserted — what was consumed, in what form, at what level and over what period. Specification comes fourth: the parameter set with its limits, which doubles as the ongoing control after any approval. Safety information comes fifth, in the form the route requires, and it is where the time goes: depending on the route and the use level, the expectation ranges from a structured summary of existing knowledge to new studies commissioned for the purpose.

Two practical notes on the pack. First, the specification should be the same document that governs supply, because a dossier built on a specification the supplier does not actually operate creates a compliance gap that shows up at the first audit. Second, the methods matter as much as the results, because a safety or composition argument built on a method the reviewing authority cannot follow will be questioned.

Who owns the dossier

Ownership of the dossier is a commercial question that is decided too late more often than any technical issue, and it has three realistic answers.

The supplier owns it where the material is sold across markets to many customers and the supplier has invested in the registrations; the buyer then relies on the supplier’s position and needs to verify its scope and currency. The brand owns it where the use, the combination or the claim is specific to the brand, because no supplier dossier can cover what the brand does with the material. And it is shared where the supplier holds the ingredient-level position and the brand holds the finished-product position, which is the most common real-world arrangement and the one that most needs to be written down.

The two failure modes are symmetrical. A buyer who assumes the supplier’s dossier covers the brand’s use discovers at the worst moment that it covers only the ingredient. A supplier who allows a customer to rely on a position whose scope was never checked discovers the same thing from the other side. Both failures are prevented by one paragraph in the supply agreement that names what is covered, what is not, who is the responsible party in each market and who maintains the file.

What to ask a supplier when a route is planned

Four questions sort suppliers quickly on market access. What regulatory positions already exist for this material, in which markets, and can the scope and the evidence be shared. What is the documented history of use of the botanical, and in what form and level, because history-based routes live or die on that document. What analytical methods and identity evidence sit behind the specification, because identity is the foundation of every route. And who has previously used this material for the intended use, if that can be said, because prior use in the same category is the most persuasive evidence of all short of a new study.

A supplier who has thought about market access will answer with documents and scope statements. A supplier who has not will answer with reassurance, and the difference is worth establishing before the development plan is committed rather than after.

The change-control link

A dossier describes the material as it was assessed. Change the material and the dossier’s relationship to reality becomes a question, which is why change control and market access are the same conversation.

The changes that matter most are the ones that do not move the assay: a change of extraction route, a change of carrier, a change of plant or a change of raw material source can alter the composition profile, the process history or the identity basis on which a position rests. This is the same requirement that a change-control clause imposes for quality reasons, and the market-access dimension adds a second consequence to it: a change may not only require a quality re-verification but may reopen a regulatory question in a market where the position was built on a specific process or composition.

Writing the two together is the efficient answer. The supply agreement should require notification of the changes that matter for both purposes, and the market-access file should record which changes would trigger a re-examination, so that the question is answered from the file rather than reconstructed after the fact.

Frequently Asked Questions

Our shipment cleared customs. Does that mean the ingredient can be sold?

No. The customs and shipment pack describes the lot; market access is a question about the use, the category and the market. A shipment can clear with complete documents while the intended use sits outside what the market permits. Settle the category and the route before production planning, not after arrival.

Why does the regulatory category matter so much?

Because the procedures and the evidence do not transfer between categories. History of food use supports a food argument and little else; a cosmetic safety assessment does not answer a food question; feed authorisation has its own logic. Selling the same extract into two categories usually means building two positions.

What belongs in the evidence pack?

Five components: identity evidence covering species, plant part, manufacturing route and the methods that distinguish the material from near neighbours; composition data with bases and methods; documented history of use where the route relies on it; the specification that will also govern supply; and safety information in the form the route requires. Identity is the foundation, and it is the part most often thin.

Who should own the dossier?

Whoever is the responsible party for the position: usually the supplier for an ingredient-level position sold across markets, the brand for a use, combination or claim specific to the brand, and often both in their respective layers. Whatever the arrangement, name it in the supply agreement together with the scope of what is covered, because both failure modes come from an unverified assumption about scope.

Can a formulation change reopen a regulatory position?

It can. Changes that do not move the assay — extraction route, carrier, plant, raw material source — can alter the composition profile or process history on which a position rests. Treat regulatory-relevant changes under the same change-control clause as quality-relevant ones, and record in the market-access file which changes would trigger a re-examination.

Which materials actually need temperature control

Not everything needs a cold chain, and requesting one where it is not needed costs money that would be better spent elsewhere. The materials that do fall into a small number of groups.

Oil-based and oil-rich materials come first: oils, oleoresins and oil suspensions oxidise, and oxidation accelerates steeply with temperature. For a carotenoid or an oil-soluble active, the storage temperature is not a comfort question but the main driver of how quickly the colour and the assay move, which is why the stability data for these materials is usually stated against a defined storage condition rather than against ambient. Liquid and semi-liquid extracts come second, especially emulsion systems, where heat and freeze-thaw cycling can break the emulsion in a way that no subsequent handling reverses. Materials with significant volatile content come third: aroma compounds are lost at temperatures that the rest of the material tolerates without complaint. And a fourth group is defined by physical state rather than chemistry: syrups, pastes and high-moisture materials that solidify, separate or crystallise when they are taken below or above a window, and whose condition after an excursion is a handling question even when nothing has chemically degraded.

Dry, low-moisture powders — the bulk of the botanical trade — generally tolerate ambient shipping well when the packaging barrier is right, which is why the packaging discussion and the temperature discussion are related but separate. The temperature question is about the minority of materials that degrade, separate or volatilise faster than the shipment is long.

What an excursion is, and why the number alone decides nothing

An excursion is any period during which the material was outside the storage conditions its stability data assumed. The definition is easy; the assessment is not, because the instinctive question — how hot did it get — is nearly useless on its own.

Degradation is a function of time and temperature together, not of a peak. A material held two degrees above its limit for three weeks may have taken more damage than one that touched ten degrees above for an hour, and a peak that lasted minutes inside a container cycling between day and night may represent less exposure than a mild elevation sustained for days. This is why a serious assessment works from the temperature history as a whole — the trace, not the maximum — and why a logger that records a peak and nothing else is only marginally better than no logger at all.

The second reason the number decides nothing is that the limit itself has context. A stability study conducted at a defined condition with a defined endpoint carries headroom: the material may be degrading at a rate well below the limit at the start of its shelf life, which means an early excursion consumes headroom that a late excursion does not have. The same trace applied to a lot at the beginning of its life and to a lot nearing its retest date can imply two different answers. Assessing an excursion therefore requires three things together: the trace, the stability data behind the lot, and the lot’s position within its own shelf life. Any one of the three on its own is an opinion.

The data: loggers, placement and who reads them

The instrumented shipment is the only shipment whose history can be argued about afterwards, and instrumentation raises three practical questions: what to use, where to put it, and who reads it.

On the instrument itself, the useful distinction is between loggers that record a history and indicators that record a threshold. An indicator answers whether a limit was crossed; a logger records the trace, and the trace is what makes a defensible assessment possible. For high-value or temperature-critical shipments the logger is worth the cost; for a dry powder shipment an indicator may be the honest and proportionate answer.

Placement is the question most often got wrong. A logger placed on the outside of a pallet or in the container doorway reads the environment, not the material, and the two can differ by a wide margin in a container cycling through day and night. A logger placed inside a drum reads the material but not the arrival conditions at the drum wall, where condensation forms. The defensible arrangement is both: one logger with the shipment, one inside a designated monitor drum, with the drum identified in advance so that it can be opened and read first.

Accountability is the third question, and it is contractual rather than technical. A logger whose file nobody downloads is decoration. The documents should state that the logger travels with the shipment, that the file is downloaded and shared at arrival, that the reading happens before the goods are put away, and who in each organisation is responsible — because the value of the trace depends entirely on it being read while the lot can still be quarantined rather than after it has been blended.

Assessing an excursion: what to request and what to test

When a lot arrives with a temperature history outside plan, the response has four steps, and the first is to stop.

Quarantine the lot before it is put away or released, because the disposition decision can only be made on evidence that exists before the material enters production. Second, request the trace and read it against the three inputs above: the shape of the exposure, the storage conditions the stability data assumed, and the lot’s position in its shelf life. Third, decide what to test, and the choice should follow from the material rather than from habit: for an oil or carotenoid material, the assay and the oxidative quality indicators; for a volatile-containing material, the sensory profile; for a material whose packaging may have breathed, the moisture figure; and for anything where condensation is a possibility, the microbiological indicators. Fourth, record the disposition, including what was tested and what the results were, because the same shipment profile will recur and the record is what turns an incident into a reference point.

One judgement is worth making explicitly: a lot that has been tested and passed after an excursion is not the same as a lot that never experienced the excursion, and where the material is near its retest date or destined for a long shelf life, the honest disposition may be to shorten the remaining shelf life rather than to pass it unchanged. A shortening of the date is a commercially uncomfortable outcome, and it is still better than a date that the material’s history does not support.

The three temperature environments

The chain divides into three segments, and each has its own failure modes and its own controls.

The supplier’s warehouse is the first, and the questions are how long the material sits before dispatch, whether the storage condition matches the specification, and how the material is staged — a drum staged outside a dispatch dock in a hot afternoon has begun its transit before the container doors close. The specification should state the storage condition the material is held under, and the shipment pack should state when it was released from that condition.

Transit is the second, and it is where the physics is least intuitive. A container in the sun becomes an oven regardless of the ambient temperature at the port, and a container that cycles from hot days to cold nights moves the air inside through dew points that produce condensation on the load. The choice between a temperature-controlled container and a dry container with a well-specified liner is a material decision: for the groups identified at the start, the controlled container is not a luxury, and for dry powders with the right barrier it is usually an unnecessary cost. What should never vary is the record of which choice was made and why.

The customer’s warehouse is the third, and it is where the buyer’s own operation becomes part of the answer. Part-used drums left open in a warm room, goods staged on a dock, material held in a blending area overnight — these are the same exposures as transit, and they are the buyer’s responsibility rather than the supplier’s. A temperature obligation in a specification applies to both parties, and the buyer’s own storage should be measured with the same seriousness as the supplier’s.

Seasonality, and the asymmetry of risk

Shipping risk is seasonal, and the two seasons fail differently. Summer fails by heat: the container-oven effect, the staging dock, and the degradation and volatilisation losses that follow. Winter fails by cold and by the transitions between states: oils and emulsions that solidify and separate on thawing, high-moisture materials that freeze and change texture, and the condensation cycles that come with moving between cold and warm environments.

Two practical consequences follow. First, the shipping plan should change with the season rather than being fixed by the contract: routing, container choice and staging practice that are adequate in spring are not the same decisions in July or January, and the specification should not assume one answer all year. Second, the sensitivity of the material should drive the season plan, not the other way round: a temperature-sensitive material shipped in the wrong season with a dry container is a decision, and it should be recorded as one rather than discovered in the goods.

Writing temperature obligations into the documents

Temperature obligations spread across three documents, and each needs its own lines. The specification should state the storage condition the material is held under, the condition its stability data assumed, and the material’s sensitivity group, so that the transport decision follows from the document rather than from the quotation. The transport or purchase terms should state the container type for temperature-controlled shipments, the logger arrangement and who reads it, the obligation to report an excursion with the trace within an agreed time, and where the risk transfers between the parties. And the quality agreement should state the assessment procedure for an excursion: quarantine on arrival, the trace review, the tests that follow, and the disposition authority.

Where those lines are missing, an excursion becomes a negotiation about blame rather than a procedure about evidence, and the negotiation usually costs more than the instrumentation would have.

Frequently Asked Questions

Which botanical ingredients actually need a cold chain?

Oil-based materials, oleoresins and oil suspensions that oxidise faster as temperature rises; liquid and emulsion systems that can break under heat or freeze-thaw; materials with significant volatile content; and pastes, syrups and high-moisture materials that separate or crystallise outside a temperature window. Dry powders with an adequate packaging barrier generally ship ambient, which is why the packaging and temperature decisions are related but separate.

The shipment reached a peak temperature. Is the lot ruined?

Not necessarily, and the peak alone answers nothing. Degradation depends on time and temperature together, and the assessment needs the full trace, the storage conditions the stability data assumed, and the lot’s position in its shelf life. A brief peak early in a material’s life with stability headroom can be benign; a mild elevation sustained for weeks near the retest date may not be.

Where should a data logger be placed?

Both with the shipment and inside a designated monitor drum. A logger on the outside reads the environment and not the material, and the two differ widely in a cycling container; a logger inside a drum reads the material but not the arrival conditions where condensation forms. The drum should be identified in advance and read first at arrival.

What should I do when a lot arrives out of temperature condition?

Quarantine before release, obtain the trace and read it against the stability data and the lot’s position in its shelf life, test the parameters that follow from the material — assay and oxidative indicators for oils, sensory for volatiles, moisture where the packaging may have breathed, microbiology where condensation is possible — and record the disposition. Where the history does not support the printed date, shorten the date rather than pass the lot unchanged.

Who is responsible for temperature in the customer’s warehouse?

The buyer is. Part-used drums left open, goods staged on a warm dock and material held overnight in a blending area are the same exposures as transit, and the temperature obligation in the specification applies to both parties. Measure the receiving and storage environment with the same seriousness applied to the supplier’s.

What a gummy is, and why it is a hard home for a botanical

A gummy is a concentrated sugar solution gelled into a shape, and every part of that definition is a constraint on the ingredient. The solids are high, which means the water available to the ingredient is low — good for microbiology, as covered below, but a very different environment from the water the extract was characterised in. The cooking step is hot, well above the boiling point of water at its peak, and the mass stays hot through transfer, holding and depositing, so the exposure is measured in tens of minutes rather than seconds. The pH is acidic, both for taste and because most gelling systems need acid to set. And the finished piece is eaten slowly, which means the ingredient’s taste arrives on the palate with more time and attention than it would get in a capsule.

Four consequences follow directly. Volatile and heat-labile components are lost or changed in the cooking and holding steps. Colour behaves differently in acid and heat, and the colour families that survive elsewhere can fail here. Acid-sensitive actives degrade across the hot hold. And taste, which a capsule hides completely, becomes a formulation problem again — a gummy that tastes of its botanical extract is a product decision, and one that has to be made deliberately.

The gelling system decides the rules

Before the ingredient is chosen, the gelling system has to be, because it sets the temperature, the pH and the texture conditions the ingredient must survive.

Gelatin gives the classic gummy texture: a melt-in-the-mouth piece that dissolves at body temperature. Its consequences for a botanical ingredient are a neutral pH window that is not strongly acidic at deposition, a long hot hold because gelatin masses must remain fluid until depositing, and a protein matrix that polyphenol-rich extracts can interact with — the same complexation behaviour that clarifies beverages with gelatin happens in reverse here, and a tannin-heavy extract in a gelatin mass can change both the texture of the piece and the recovery of its own actives. Gelatin also carries the certification questions around animal origin that are covered in the halal, kosher and vegan discussions elsewhere on this blog, and those questions apply to the finished piece rather than to the extract.

Pectin sets by acid and sugar, tolerates a higher cooking temperature, and produces a piece with a different bite and a non-thermoreversible set. Its consequences are an explicitly acidic matrix, which narrows the colour and actives that survive, and a narrower processing window around the setting point, which compresses the time available for mixing and depositing — a real constraint when an active has to be dispersed uniformly before the mass sets.

Starch-based systems produce a chewier piece through a slower mogul process, with long curing at controlled conditions. The ingredient question there is the long cure: a slow drying period in which moisture migrates and surface treatments are applied, which is a stability question for any surface-active component.

The choice of gelling system should therefore be made with the extract on the table, not after it, because the systems impose different pH, temperature and interaction conditions on the same ingredient.

What the cooking and depositing steps do

The cooking step concentrates, heats and deaerates the mass, and it does three things to a botanical ingredient that the powder’s own stability data will not predict.

Volatiles leave first. Aroma and top-note flavour compounds that survive drying and storage in a powder are stripped in a hot, high-solids cook, which is why an extract that smells and tastes correct in the sachet can produce a piece with a flat or altered profile. Where the aroma matters, the answer is usually to add the volatile fraction at the latest possible point rather than to ask the cook to preserve it.

Colour changes second, through two mechanisms at once: the heat itself, and the acid the gelling system requires. The colour families differ sharply here, and the colour discussions elsewhere on this blog cover the families in detail; the gummy-specific point is that heat and acid arrive together in this format, which is a harsher combination than either a beverage or a tablet presents, and the colour choice should be tested in the actual mass rather than in a model solution.

Actives degrade third, and the exposure is the hot hold rather than the cook: the mass sits at depositing temperature while it is transferred, portioned and filled, and an active that survives a short cook can still lose content across a long hold. The useful number to ask for is not a stability figure at a storage temperature but a survival figure at the process conditions — cook temperature and hold time together.

Dosing into a sticky mass: the uniformity problem

A gummy is dosed per piece, and the piece is small, which means the active is included at a low rate into a viscous, hot, mobile-until-it-is-not mass. Uniform distribution at that inclusion rate is the manufacturing challenge of the format, and it has four practical components.

The physical form of the ingredient matters first: a coarse particle will not disperse in a viscous mass, and a powder that clumps will produce pieces with none of the active and pieces with several doses. Milling to a fine, consistent particle size and, where the material is oily, presenting it on a carrier or as an encapsulated powder are the standard answers. Premixing comes second: the active is dispersed into a small quantity of a compatible component before it meets the full mass, because trying to distribute a low-inclusion powder directly into a batch of hot syrup is how uniformity failures happen. Order of addition comes third: adding the active late preserves it from the cook but compresses the mixing time available before the mass sets, which is the tension every gummy formulation manages. And uniformity testing comes fourth: a content-uniformity check across individual pieces rather than a batch-average assay, because a batch average can conform while the distribution across pieces is unacceptable.

The depositing step is part of the same question. The mass must remain fluid from mixing to deposit, and the hold time at depositing temperature is part of the ingredient’s exposure. A process that mixes and deposits quickly is gentler on the active than one that holds, and the process timing belongs in the development brief alongside the formula.

Curing, water activity and the shelf life of the piece

After depositing, the piece cures: moisture equilibrates, the gel sets fully, and any surface treatment is applied. The water activity of the finished piece sits low enough that microbiological growth is not the limiting factor, which is one of the format’s genuine advantages — but the same low water activity means the piece is now in equilibrium with the air, and texture rather than microbiology becomes the shelf-life question.

A piece that loses moisture hardens; a piece that gains it sweats, sticks and can develop surface crystallisation. The packaging barrier and the storage humidity therefore determine the texture shelf life in the same way the ingredient’s own packaging determines its powder shelf life. For the botanical extract, the relevant consequence is that moisture can migrate between the piece and its surroundings and between components inside the piece, and a hygroscopic carrier in the extract can change the texture of the finished confection — a texture problem that appears weeks later and traces back to an ingredient choice.

Stability testing for a gummy-bound active should therefore be run in the piece, at the piece’s water activity, rather than inferred from the powder. The powder data answers whether the ingredient survives its own drum; only the piece answers whether it survives the format.

Colour and taste in a confectionery matrix

The confectionery matrix changes both questions. On colour, the piece is translucent rather than opaque, so the colour is seen through a depth of material and against the colour of the mass itself; the same dose of the same colourant reads differently than it does in an opaque tablet or a clear beverage, and acid shifts the hue of the families that respond to pH. On taste, the gelatin systems melt on the palate and pectin systems release slowly, so the timing of the extract’s flavour contribution differs by system — a difference that matters as much as the intensity.

Where the extract’s own taste is the problem, taste masking is available but constrained: the piece is chewed and dissolved, so a barrier that survives the mouth has not masked anything, and a barrier that releases too early has not survived the process. Encapsulation systems designed for confectionery exist for exactly this purpose, and the choice between them is a release-trigger question — what breaks the shell in the mouth, and whether the cooking step breaks it first.

What the specification for a gummy-bound extract should say

The specification should describe the ingredient as it will be used, and six lines do most of the work. The physical form with particle size, because dispersion in a viscous mass is the dosing question. The stability position at process conditions — cook temperature and hold time together — rather than only at storage temperature. The behaviour at the pH the gelling system will impose, for both actives and colour. The interaction position with the gel matrix, stated explicitly where the extract is polyphenol-rich and the system is gelatin. The sensory profile at the inclusion rate, because the piece will be chewed rather than swallowed. And the payload with its basis and method, since the dose per piece is small and the arithmetic runs from the payload figure.

Alongside those, the development brief should state the gelling system, the process timings and the cure conditions, because a specification written without them describes an ingredient for a format nobody is making.

Frequently Asked Questions

Why does an extract that is stable in a powder fail in a gummy?

Because the format exposes the ingredient to a hot cook, a long hold at depositing temperature, an acidic matrix and a slow, chewed release — a combination a powder never meets. Volatiles are stripped, colour and acid-sensitive actives degrade, and polyphenol-rich extracts can interact with a gelatin matrix. Stability has to be demonstrated in the piece, not inferred from the powder.

Gelatin or pectin — does the choice affect the botanical extract?

Yes, materially. Gelatin gives a long hot hold and a protein matrix that polyphenols can bind to, while pectin requires an explicitly acidic set and offers a narrower processing window. The two impose different pH, temperature and interaction conditions on the same ingredient, and the gelling system should be chosen with the extract on the table.

Why do my gummies vary in content between pieces?

Because a low-inclusion active in a viscous hot mass is hard to distribute uniformly. Coarse or clumping particles, direct addition without a premix, a mixing window compressed by the setting point and a batch-average assay rather than a per-piece uniformity check are the usual causes. Fine, consistent particle size, premixing, deliberate order of addition and per-piece uniformity testing are the controls.

How should stability be tested for a gummy product?

In the finished piece, at the piece’s water activity and through the cure, rather than on the powder. The powder data answers whether the ingredient survives its drum; the piece answers whether it survives the format, including moisture migration, texture change and colour development over the confectionery shelf life.

Can a bitter extract be masked in a gummy?

Yes, through confectionery encapsulation systems, but the barrier has to survive the cook and then release in the mouth — a barrier that survives chewing has not masked anything, and one that releases during processing has delivered the bitterness to the mass. The choice is a release-trigger question, and it should be settled against the actual process temperatures and the matrix.

Two rulebooks, and why the difference is structural

Food law generally works by permitting what is not prohibited and by placing the burden of safety on the operator, with a positive list for additives and a notification or authorisation route for novel materials. Feed law tends to work the other way round for the active categories: a substance placed on the market as a feed additive normally has to be authorised for the species and the purpose before it can be used, and the authorisation is specific rather than general.

That structural difference produces the practical consequences buyers meet first. A material that has a long history in food may still require an authorisation to be placed on the market as a feed additive for a particular species and function. A material that is authorised for one species is not thereby authorised for another. And a material that is used in feed as a nutritional component may fall under a different category from the same material used as a technological additive or a colorant, which changes both the documentation and the labelling. The first question in any feed project is therefore not what the material is, but which category and which species the intended use falls into — and the answer should be confirmed against the current authorisation list rather than carried over from a previous project.

Why a human-grade specification does not transfer

It is tempting to assume that a specification written for a food or supplement application is stricter than one written for feed, and therefore acceptable by default. That assumption fails in both directions.

It fails in the direction of purity parameters that matter in feed but not in food, and the clearest example is carry-over. A feed mill runs many formulations through shared equipment, and a material that is acceptable as a residue at trace levels in a food matrix may be unacceptable as a carry-over contaminant in a feed premix, because the next batch through the line may be intended for a different species. Specifications for feed use often carry explicit limits and declarations around carry-over that a food-facing specification does not mention at all.

It also fails in the direction of documentation. Feed buyers commonly need statements about the material’s status as a feed additive, its target species, its withdrawal conditions where applicable, and its handling classification. A food-facing document pack does not usually include them, and the gap has to be closed before an order can be released rather than after.

Where a material will be sold into both markets, the practical approach is a single specification with market-specific annexes rather than two separate specifications, because two documents describing the same drum inevitably drift apart.

Species differences that rewrite the brief

The species sets the physiological, dosing and handling frame, and the differences are large enough to change the product rather than only the dosage. Cats present a set of constraints that differ from dogs, including differences in how certain compounds are metabolised and a much lower tolerance for some materials, which makes palatability and acceptability more critical and formulation freedom narrower. Dogs tolerate a wider range of formats and delivery systems, which is why the chew, the soft bite and the powder-in-food channel are all commercially viable there. Horses, poultry and aquaculture each bring their own route of administration, their own acceptance behaviour and their own regulatory category.

For a botanical ingredient, three variables follow from the species rather than from the material: the delivery form the animal will actually accept, the concentration at which the material can be included without affecting acceptance, and the analytical parameters needed to demonstrate identity and content in that particular matrix. A specification written for a human capsule rarely answers any of them.

Palatability is the first constraint in a pet application

In human nutrition, an off-note is a formulation problem. In a pet application it is a commercial one, because an animal that declines the product produces no repeat purchase and no feedback beyond a bowl left full. Palatability therefore moves from a sensory detail to the acceptance criterion that governs how much of the material can be used at all.

Three practical consequences follow. First, bitterness and astringency, which many botanical extracts carry in proportion to their polyphenol or saponin content, are dose-limiting rather than merely unpleasant, and the dose at which they become limiting has to be established in the finished format and with the target species. Second, the delivery system matters more than the specification: an extract that is unacceptable in a powder may be perfectly usable inside a fat-coated kibble or a soft chew, because the matrix controls when and how the material is tasted. Third, the sensory profile of the ingredient has to be stable from lot to lot, because a change in taste is a change in acceptance even when the assay is identical — which is one more reason a sensory specification belongs in the purchase document.

Carotenoids in feed: colour as a performance parameter

The most established botanical feed application is also the least health-adjacent, which makes it a comfortable place to start: pigments used to colour the products of the animal rather than the animal itself. Carotenoids from marigold, including lutein and zeaxanthin preparations, are used in poultry production as a colouring parameter, where the pigmentation of the yolk and of the skin is a quality attribute that markets assess and pay for. In aquaculture, carotenoids are used in the pigmentation of farmed fish, and the inclusion rate, the form of the carotenoid and its stability through feed manufacture all determine whether the finished product reaches the colour the market expects.

This application is instructive because it shows how the same drum serves two different rulebooks. A marigold-derived lutein preparation sold into a supplement application is specified against content, form and residues for human use. The same material sold into feed is specified against its function as a colorant, its stability through extrusion or pelleting, its carry-over behaviour in a premix and its authorisation status for the species. The physical parameters that matter are different, and so is the analytical basis on which the material is compared between suppliers.

Premix dilution, carry-over and the documents that gate the shipment

Feed ingredients are usually handled at low inclusion rates, which means they are diluted into premixes before they reach the mixer. That step changes the specification in ways that are easy to overlook: a premix specification has to describe the carrier, the concentration and its uniformity, and it has to describe the stability of the active through whatever thermal process the feed manufacture applies. Uniformity is not a formality here, because a premix that is not uniform delivers a variable dose to the animal and a variable result to the customer.

Carry-over control and the documentation that accompanies it are the other gate. Expect to provide a statement of composition, an allergen position, a statement on the material’s status and intended species, and a handling and storage recommendation in a form usable by a mill rather than only by a formulator. Where the material is not intended for human consumption, that has to be stated explicitly and consistently across the specification, the label and the commercial documents, because the distinction is a legal one and not a marketing one.

Label categories, and the trap in the word supplement

The category a pet product is placed in determines what the label may say and what evidence sits behind it, and the categories are not the ones a brand moving over from human nutrition expects to find. A product presented as a complete food carries a nutritional adequacy position that a product presented as a complementary feed or a snack does not, and a product positioned alongside feed rather than inside it may fall under a different regime again depending on the market. The word supplement itself does not travel cleanly: in some markets it maps onto a defined feed category with attendant rules on composition and claims, while in others it is a marketing descriptor rather than a regulatory one.

The practical consequence for an ingredient buyer is that the intended label position has to be settled before the specification is written, because it decides which declarations the material has to carry, whether a maximum inclusion rate must be observed, and whether the finished product will need a nutritional or compositional statement that the ingredient contributes to. A formulation designed for one category and then repositioned into another often needs a different document set rather than a different formula, and that discovery is cheaper at the brief stage than at the artwork stage.

What to write into the purchase specification

State the intended use and the target species, because both determine the authorisation route and the documentation set. State the regulatory category the material is being supplied against, and require the supplier to confirm the current authorisation or registration position for the destination market rather than asserting a general one. State the delivery form and its handling properties — flow, dust, hygroscopicity — because both the premix plant and the finished feed line will be handling it. State the sensory profile and the dose at which acceptance has been established. State the stability position through the thermal process your feed manufacture applies. State the carry-over and allergen declarations. State the analytical basis for the active, with the method, so that two suppliers can be compared. And state the change-control items — the site, the extraction route, the carrier and the dilution step — because a change in any of them can alter acceptance or colour performance without altering the assay.

Frequently Asked Questions

Can a food-grade botanical extract be used in animal feed?

Not automatically. Feed additives are generally authorised for a species and a function before they can be placed on the market, and a history of food use does not create that authorisation. Confirm the category and the species for the intended use against the current authorisation list for your destination market, and treat the material’s food status as separate evidence.

Why does a feed specification need different parameters from a food specification?

Because the risks differ. Carry-over between batches in a shared mill, uniformity in a premix, and stability through extrusion or pelleting matter in feed in ways they do not in a food or capsule application, while some purity parameters that dominate food specifications are less decisive. The two documents answer different questions about the same material.

Which carotenoid applications are most established in feed?

The colorant applications, where the pigment colours the product rather than the animal’s physiology. Marigold-derived lutein and zeaxanthin preparations are used in poultry production as a colouring parameter, and carotenoids are used in the pigmentation of farmed fish. In those applications the inclusion rate, the form and the stability through feed manufacture determine the commercial result.

Does palatability matter as much as specification?

In pet applications it can matter more, because an animal that declines the product produces no repeat purchase. Bitterness and astringency in botanical extracts are dose-limiting rather than merely undesirable, and the delivery matrix — powder, chew, coated kibble — often determines whether the material can be used at all. Establish the acceptance limit in the finished format and with the target species.

What documents does a feed buyer normally need?

A statement of composition, an allergen position, a statement of the material’s regulatory status and intended species, the analytical results with methods, a stability or treatment position relevant to your feed process, and handling and storage recommendations written for a mill. Where the material is not for human consumption, that has to be stated consistently on the specification, the label and the commercial documents.

What the claim actually asserts

Vegan and vegetarian are not the same claim and they are not defined identically in every market. In broad terms, a vegetarian claim speaks to what is in the product, while a vegan claim speaks to what is in the product and to how it was made, which brings processing aids, carriers and contact materials into scope. Where a claim extends to how a material was produced, the ingredient list is no longer sufficient evidence, because processing aids are by definition not ingredients and therefore do not appear on a label.

There is a second distinction that matters as much: whether the claim applies to the finished product, to a specific input, or to a site. A supplier can legitimately declare that a material contains no animal-derived ingredients while declining to certify a vegan position, because the second claim requires controls over the whole production route that the first does not. Reading which of the three was actually asserted is usually the difference between a claim that holds up and one that does not, and the answer is frequently in the wording of the letter rather than in its headline.

Because definitions and enforcement practice differ by market, the working position for a buyer is to define the claim internally, in writing, at the level of detail the label requires, and then to ask the supply chain to evidence that definition rather than a general one.

The processing-aid route

The most commonly missed route is the processing aid, and the reason is structural: a processing aid does its job during manufacture and is not required to be declared as an ingredient, so nothing forces it into the specification unless the buyer asks. Four groups cover most of the exposure.

Filtration and clarification aids are the first. Gelatin and isinglass are both used to clarify liquids, and both are animal-derived, one from collagen and one from the swim bladder of fish. Where a liquid extract or a beverage-ready concentrate is clarified, the aid is a candidate for the claim gap even though none of it is intended to remain in the product.

Finings and flocculants used to remove suspended solids in an extract or a concentrate form the second group, and the third is the group of carriers and bulking agents used in drying. A spray-dried extract is dried onto something, and the list of things it can be dried onto includes materials of animal origin alongside the starch and gum carriers that dominate. The fourth group is the release and processing aids used on equipment: lubricants, release agents and anti-stick treatments in contact with the material.

None of these appears in the ingredient list of the finished consumer product. All of them can appear in a vegan claim’s evidence pack, and a supplier who says only that the material contains no animal-derived ingredients is answering the ingredient question and not the vegan one.

Carriers and the diluent list

Carriers matter twice over for a plant-based claim. They matter because a standardised extract is diluted, and because many carriers are sourced as commodity materials whose origin is not fixed. The carrier families used in botanical extracts — starches and their hydrolysates, gums, celluloses, silicas, sugars and sugar alcohols, and salts — are mostly plant-, mineral- or fermentation-derived, and mostly benign for the purpose. The exceptions are attached to specific materials rather than to the family: a magnesium stearate may be vegetable or animal in origin depending on the stearic acid feedstock, and glycerin may be derived from vegetable oil or from tallow, with the two being chemically equivalent and analytically indistinguishable in a finished blend.

That last point is the practical one. Where two raw materials are chemically the same and differ only in origin, the specification cannot distinguish them after the fact, so the origin has to be controlled by a declaration at the point of purchase and by a change-control clause rather than by testing. The same logic applies to mono- and diglycerides, to some emulsifier streams and to any material made from a fatty acid feedstock.

For a buyer, the consequence is that carriers and their origins belong in the same line of the specification as the carrier’s share, and that a change in the origin of a carrier has to require notification even where the specification figure does not move.

Colours, glazes and finishing agents

The colourants and finishing materials applied to a product are a separate route, and this is the area where a botanical ingredient itself can be the source of the problem. Carmine and cochineal are insect-derived, and they are used widely in food and cosmetic applications, which makes them a common vegetarian issue rather than a vegan one. Shellac, used as a glazing agent, is also animal-derived. Beeswax and honey derivatives enter through coatings and carrier systems. And in the cosmetics direction, lanolin from wool, certain collagen-derived materials and some animal-derived emulsifiers are conventional rather than exceptional.

For a botanical extract supplier the relevant question is usually narrower: whether the ingredient as supplied carries any of these materials, and whether the finishing treatment applied to a powder — anti-caking, flow, glazing — introduces one. Those treatments are frequently applied to a powder to make it handle well, and they are frequently not declared unless they are asked about.

Certification schemes and how they differ

Several certification bodies operate marks for plant-based and vegan positions, and they are not interchangeable. The differences that matter are scope and depth: whether the assessment covers an ingredient, a finished product, a production line or an entire site; whether processing aids and cleaning chemicals are inside the scope; and whether the assessment is a document review, an on-site audit, or both. Some schemes also require the removal of shared equipment or the segregation of production, which changes the cost of compliance substantially compared with a scheme satisfied by declarations.

Three practical consequences follow. A certificate issued for a finished product does not necessarily cover an input ingredient, and a certificate issued to one site does not cover material made at another. A certificate’s scope statement — the same scope statement that matters for food safety certification — determines whether it is evidence for your claim. And a scheme that relies on declarations without audit is not equivalent to one that audits, even where the mark looks similar on a package.

Because scheme criteria change and because national definitions are not harmonised, the workable approach is to state in the supply agreement which scheme and which scope you require, and to treat the mark as evidence only within that scope.

Cross-contact, shared lines and cleaning

Where a claim extends to how a material was made, shared equipment becomes part of the question. A plant that runs gelatine-containing materials on the same line, or that uses an animal-derived cleaning chemistry, is not automatically disqualified from supporting a vegan position, but the position has to be supported by something: validated cleaning between campaigns, a dedicated line for the claim-bearing material, or scheduling that keeps the two apart.

For botanical extracts the exposure is concentrated in a small number of places: the mill, the blender, the drying unit and the packaging line. Cleaning validation that was designed to prevent allergen cross-contact is usually sufficient evidence for both purposes, which is one of the few places where an existing allergen control programme can be reused rather than duplicated. Where no cleaning validation exists, the claim rests on assertion, and that is worth knowing before the label is printed.

Making the claim verifiable in the specification

The claim should be expressed as a specification item with the same standing as an assay figure. In practice that means four lines. First, a statement that the material and all materials used in its manufacture, including processing aids and carriers, are free of animal origin, with the definition of animal origin agreed in writing. Second, the recognition that filters, carriers, colourants and finishing agents are inside that statement, because otherwise the statement can be true and still not cover the claim. Third, a change-control clause naming the origin of glycerin, magnesium stearate, fatty-acid-derived emulsifiers and any carrier sourced from a variable feedstock, so that a change of origin requires notification even where the specification does not move. Fourth, the scheme and scope of any certification that the supplier holds or will seek, stated precisely enough that it can be verified.

Where a supplier cannot confirm the position on processing aids, that is not necessarily a refusal; it is a request for a more specific question, and the answer will usually be available once the question is asked at that level.

What to check on the documents

Three documents carry most of the evidence. A specification clause stating the plant-based position and its definition is the first. A declaration from the supplier covering the material and all materials used in its manufacture is the second, and its value depends on how explicitly it names processing aids and carriers. Certificate and scope statement from the certifying body, where a mark is claimed, is the third.

Alongside those, ask for the cleaning validation or the segregation arrangement for the equipment that ran your material, and confirm that the origin declarations for variable-feedstock materials are dated and identify the current supplier. Where a claim is going onto consumer packaging, the formulation team should see all three documents before the artwork is approved, because changing a claim after printing is expensive and changing it after shipping is not possible.

Frequently Asked Questions

Is a plant extract automatically vegan?

No. The plant is the raw material, not the whole manufacturing route. Extraction, clarification, drying, anti-caking and packaging can all involve animal-derived materials, and processing aids are not required to be declared as ingredients, so they will not appear on a label even when they are present during manufacture.

What is the difference between a vegetarian and a vegan claim?

In broad terms a vegetarian claim addresses what is present in the product, while a vegan claim also addresses how it was made, which brings processing aids, carriers and cleaning into scope. Definitions are not harmonised across markets, so agree the working definition in writing and evidence that definition rather than the general term.

Which processing aids are most likely to be animal-derived?

Filtration and clarification aids such as gelatin and isinglass, finings used to remove suspended solids, carriers of animal origin used in drying, and equipment release agents. Magnesium stearate and glycerin may also be animal-derived depending on the feedstock, and where two materials are chemically identical the origin has to be controlled by declaration rather than by testing.

Does a vegan certificate on a finished product cover my ingredient?

Not automatically. Certification scopes differ: a certificate may cover a product, a line or a site, and it may or may not include processing aids and cleaning. Read the scope statement, and where your claim depends on the certificate, ask the supplier for a certificate whose scope covers the material and the site that produced it.

How should the claim be written into a specification?

As a declared item with an agreed definition, covering the material and everything used in its manufacture including processing aids and carriers, supported by a supplier declaration, by the scheme and scope of any certification, and by a change-control clause for materials whose origin can change without any analytical parameter moving.

What changes when the medium becomes water

Three things change at once, and they change together. The first is solubility. Botanical extracts divide into those that dissolve readily in water, those that disperse, and those that do not mix with water at all. Oleoresins, carbon dioxide extracts and carotenoid preparations belong to the third group, and bringing them into a beverage means creating a delivery system — an emulsion, a microemulsion or an encapsulated powder — before the material can be used at all. That delivery system, not the extract, will then govern the clarity, the mouthfeel and much of the stability of the finished drink.

The second change is that the water you are adding is not an inert ingredient. Hardness, alkalinity, mineral content and dissolved oxygen all affect how a botanical material behaves in a beverage, and two plants filling the same formula with two different waters can produce two different products. Where a project is being scaled from a bench sample made with purified water, the water at the filling line belongs in the development brief.

The third change is the removal of the water activity lever. In a powder, formulators manage stability partly by keeping water activity low; in a liquid at high water activity, that tool is gone, and preservation has to be achieved by pH, by thermal process, by preservatives, by packaging or by cold chain. Which of those routes is available determines how much freedom the formula has left for the ingredient.

Two kinds of haze, and why one is not a defect

Haze in a beverage is discussed as if it were always a fault, and it is not. An intentionally opalescent drink carries a stable, uniform cloud that comes from a designed emulsion, and a product briefed as clear will treat any visible haze as a defect. The distinction belongs in the specification from the start, because the two objectives lead to different ingredient choices and different processing.

Where clarity is the objective, the haze that appears usually has one of four sources. Undissolved particulate is the first and the simplest: a powder that is not fully hydrated, or a carrier that does not fully dissolve, leaves fine suspended solids that scatter light. Protein-polyphenol interaction is the second, and it is the mechanism behind chill haze — the complexes that form at low temperature and redissolve when the product warms. A product that is clear at ambient and hazy from a refrigerator has not necessarily failed, but it has to be designed for, because repeated temperature cycling will progressively convert the reversible fraction into permanent haze. Insoluble carrier fractions form the third source, and salt or mineral precipitation the fourth, both of which can be managed by choosing the material and by controlling the water.

The practical implication is that a haze specification is meaningless without a temperature and a duration. Ask for the clarity figure at the temperature the product will actually be stored and served at, over the time it will be held there.

Sediment, and what storage does to a clear product

Sediment is a mass problem rather than a light-scattering problem: material that was dispersed settles out under gravity, and the outcome depends on particle size, density difference and time at temperature. Storage accelerates it, and temperature cycling accelerates it further, because dissolution and re-precipitation cycles tend to build larger particles rather than smaller ones.

In tea-based and botanically heavy beverages there is an additional mechanism worth knowing, because it is specific to the ingredient class. Polyphenols and caffeine associate in solution, and the complexes that form are less soluble than either component alone. A drink built around a polyphenol-rich extract and a caffeine source can develop a deposit that neither ingredient would produce on its own, and the same interaction can pull colour out of solution along with the mass. Where both are present, the development work has to consider them together rather than sequentially.

Because sediment accumulates over time and is easy to miss at the bench, the useful evidence is a storage study at the temperature of distribution rather than a short ambient observation. A product that is clear at day one, still clear at week four and visibly affected at month three tells you the rate, and the rate is what the shelf life has to be built around.

Thermal processing, and what each route does to the material

A beverage usually has to receive a microbiological process step, and the choice of that step is a formulation decision as much as an engineering one, because each route imposes a different thermal load and a different mechanical one.

Flash pasteurisation followed by cooling applies a short, high-temperature treatment, which is generally the gentlest option for heat-sensitive colour and aroma, provided the heat exchanger, the holding time and the filling hygiene are matched. Hot fill applies the heat to the product and relies on the filled container to hold the temperature, which extends the exposure and requires the pack to survive it. In-pack sterilisation is the heaviest treatment in both time and temperature, and it is the one that most often changes the colour, the aroma and the emulsion stability of a botanical product. Aseptic or ultra-clean filling avoids much of the thermal load but moves the requirement onto the filling environment and the packaging.

Whatever route is selected, three material behaviours need to be established against it rather than assumed. The colour shift through the process, which for some botanical colourants is substantial and irreversible. The emulsion stability through the process, because a delivery system that survives mixing may separate under a thermal cycle. And the aroma retention, because volatile components that survive a powder’s shelf life can be lost in minutes at process temperature — which is one reason a flavour contribution is often better placed in a post-process addition point than inside the extract.

pH, and the colour shift it causes

pH is the parameter that decides which colourant families are available at all, and it does so more sharply in a beverage than in a solid because the whole product sits in one continuous aqueous phase. Anthocyanin-based colours change hue with pH across a range that spans red, purple, blue and greenish tones, and their stability falls away as the pH rises. Chlorophyll and chlorophyllin lose their green in acid and are effectively unusable in a low-pH drink. Betalains hold their colour over a wider pH range, and carmine is comparatively pH-stable but carries a vegetarian question that has to be answered separately. Turmeric-derived yellow shifts and fades as pH rises.

The consequence for formulation is that pH has to be fixed before the colourant is chosen, not afterwards. A decision to acidify for preservation changes the colour palette that is available, and a decision to hold a neutral pH to protect a colour changes the preservation strategy. The two decisions are one decision, and separating them is how projects end up reformulating late.

Light, oxygen and the pack format

A clear bottle on a lit shelf is the harshest environment a botanical beverage will meet, because light and oxygen act together and a clear pack offers no protection from either. Riboflavin-type and anthocyanin-type colours are notably light-sensitive, and dissolved oxygen drives oxidation of both colour and aroma. A can removes the light problem almost entirely and can be filled with low dissolved oxygen, which makes it a friendly pack for sensitive systems; a carton or an opaque bottle offers partial protection; clear polyethylene terephthalate offers very little.

Two practical points follow. The pack format belongs in the development brief alongside the formula, because it changes which ingredients are viable. And headspace oxygen, dissolved oxygen at filling and any nitrogen dosing are part of the specification of the process rather than details to be settled at commissioning.

Preservation when water activity no longer helps

Because a liquid product sits at high water activity, the preservation strategy has to come from somewhere else, and the four options are pH, thermal process, preservative system and cold chain. Most commercial drinks combine at least two.

For a botanical ingredient the relevant consequence is that the preservative system has to be compatible with the material. Some extracts interact with preservative actives, some carry a pH buffering effect that moves the effective pH away from where the system was validated, and some contribute a microbial load or a nutrient source that the preservative system was not sized for. Where a product is positioned as preservative-free, the entire preservation burden moves to pH, process and pack, and the tolerance for an ingredient that shifts the pH or adds load becomes much smaller. A supplier specification that states the material’s own microbiological limits, its pH in solution and its buffering behaviour is therefore directly useful in a beverage project and largely irrelevant in a capsule one.

Dosage, taste and the arithmetic of a beverage

Dosing arithmetic in a beverage is done per finished unit rather than per kilogram of formula, and the finished unit is usually a bottle or a can, which concentrates attention on the cost per unit in a way that a bulk powder specification does not. Three limits operate simultaneously: the taste limit, the colour limit and the cost limit, and the binding one is often the taste.

Botanical extracts bring bitterness, astringency and aroma at the concentrations where their actives contribute, and a beverage has no carrier to hide behind — no fat phase, no coating, no capsule shell. That is why the same extract at the same active dose can be perfectly acceptable in a capsule and unpalatable in a drink, and why the useful development work is a dose-response in the finished matrix rather than a calculation from the specification. Where the required dose exceeds what the taste allows, the options are a delivery system that delays release, a change of extract profile, or a reformulation of the taste frame around the ingredient.

What to write into the purchase specification

State the application and the medium, because a specification written for a capsule does not answer a beverage. State the solubility behaviour explicitly — fully soluble, dispersible, or requiring an emulsified delivery system — and state the form in which the material is supplied. State the pH of a standard solution and any buffering effect, because both interact with the preservation system. State the microbiological limits and the treatment the material has received. State the clarity or haze expectation at a named temperature and duration. State the colour behaviour across the pH range the product will occupy, with the colour figure and its basis. State the sensory profile and the dose at which it has been established. And state the change-control items that would alter any of the above: the site, the extraction route, the carrier, the delivery system and the sterilisation method.

Where a supplier can discuss the material only in powder terms, the project has a specification gap that is cheaper to close before the formula is frozen than after the first pilot batch separates in the tank.

Frequently Asked Questions

Why does an extract that works in a capsule fail in a drink?

Because the medium changed. A beverage has higher water activity, a thermal process, a pH that interacts with colour and stability, and no carrier to mask bitterness or astringency. Preservation, clarity and taste all behave differently, and the specification that suited a solid format does not describe the material’s behaviour in water.

What causes haze in a botanical beverage?

Undissolved particulate and carrier fractions, protein-polyphenol complexes that form at low temperature, mineral precipitation and unstable emulsions. Chill haze in particular is reversible at first and becomes permanent with repeated temperature cycling, so a haze figure needs a temperature and a duration attached to it to mean anything.

Which thermal process is gentlest on botanical ingredients?

A short, high-temperature flash pasteurisation followed by rapid cooling is generally the least damaging to heat-sensitive colour and aroma, provided hygiene and filling are matched to it. In-pack sterilisation applies the heaviest load and most often changes colour, aroma and emulsion stability. Establish the colour, emulsion and aroma behaviour against the actual process rather than a generic one.

Does pH affect a botanical colourant?

Substantially, and for some families it decides whether they can be used at all. Anthocyanin colours shift hue and lose stability as pH rises, chlorophyll greens are unusable in acid, and turmeric-derived yellows shift with pH. Fix the pH first and choose the colourant second, because the two decisions constrain each other.

How should an extract be dosed into a beverage?

Per finished unit rather than per kilogram of bulk, and the binding limit is usually taste rather than cost. Botanical extracts bring bitterness and astringency that a beverage cannot hide, so establish the dose-response in the finished matrix. Where the taste limit sits below the required dose, a delivery system or a different extract profile is the usual answer.

The three routes to a natural green

The first route is the chlorophyll family, which is the only group of materials that exists in commerce for the purpose of being green. It covers the natural pigment and its stabilised copper complex, and it is the route a formulator takes when a precise, repeatable green is the objective.

The second route is a blend: a natural blue combined with a natural yellow to make a green. This is what most commercial “natural green” actually is, because the blue and yellow families are broader, more available and often cheaper than a single green. It works, but it inherits the weaknesses of both components, and that is where most of the drift in finished products originates.

The third route is a green plant powder — matcha, spinach, alfalfa, nettle — used as much for its story and its nutritional content as for its colour. Colour here is a plant property rather than an engineered figure, and it arrives with flavour, fibre and a use level that is usually an order of magnitude above a colorant.

Chlorophyll and copper chlorophyllin are not the same material

Chlorophyll is the pigment itself. In its native form it is oil-soluble, it is unstable, and it degrades into an olive-brown material when the central magnesium atom is displaced. Those properties are the reason native chlorophyll is a difficult ingredient to use as a colour, and the reason the material that actually dominates the market is a derivative.

Copper chlorophyllin is that derivative. Replacing the magnesium with copper produces a molecule that resists the acid-driven degradation step and that can be made water-soluble, which is why it is the form used in beverages, confectionery, sauces and anything with a low pH. The two are therefore not interchangeable, and a specification that says only “chlorophyll” is not telling you which one you have been quoted. They are also listed separately in the additive frameworks that regulate them, and the permitted uses are not the same in each market, so the family is one where the current text of the destination’s rules has to be checked rather than recalled. Treat the family as two materials with one name.

Where green usually comes from: a blue plus a yellow

If you ask a formulator how a natural green was achieved in a commercial product, the honest answer is frequently that it was blended. A blue from spirulina or butterfly pea is combined with a yellow from turmeric or a carotene, and the ratio is tuned until the hue lands in the right place.

The attraction is availability and cost. The complication is that the blend is only as stable as its least stable component. If the blue fades faster than the yellow, a green product drifts towards yellow-green and then towards a flat khaki; if the yellow fades first, the colour moves towards blue and the product reads mint or teal. The two components also have to be compatible in phase: a water-dispersible blue with an oil-soluble yellow will not give a clean green in an aqueous system, and the result is usually a speckled or dull appearance rather than a mismatch of hue.

There is a pH dimension as well. Blue systems based on anthocyanins shift their hue with pH, and spirulina-derived blue has its own narrow working range, so the blend that matched in a buffered pilot can read differently after a small formulation change. When green is built from two materials, the specification has to cover both of them, and the stability work has to be done on the blend rather than on the components separately.

Green from plant powders, and what the powder brings with it

Matcha is the best-known example of a plant powder used for colour, and it deserves a precise description because it is often compared to colorants on the wrong terms. The colour of matcha is a function of how the crop was grown, how quickly it was steamed after harvest, how it was dried, how finely it was milled and how it has been stored since. Shading the plant before harvest raises the chlorophyll content, steaming inactivates the enzyme that would otherwise degrade the pigment, and a fine particle size increases the colour intensity simply because more surface is available to reflect light. Two matcha powders of the same grade can therefore differ in colour, and the difference is a harvest and processing property rather than a specification failure.

Spinach, alfalfa and nettle powders behave the same way, and the drying method matters more than buyers expect: a gentle low-temperature dry retains a green that a harsher dry turns olive. The attraction of this route is that the material is declared as an ingredient rather than as a colour additive, which is exactly what a clean-label brief is looking for. The cost is that the use level is driven by taste and by the powder’s other components rather than by colour strength, so the load in the formula becomes the constraint, and the colour varies with the crop.

What heat, acid and light do to green

Green fails in a characteristic order, and knowing the order tells you which control to reach for. Acid and heat act on the same point in the molecule: the central metal atom is displaced, the pigment becomes pheophytin, and the colour shifts from green to olive and then to brown. That is why a green colour in a low-pH beverage is a different technical problem from a green colour in a neutral or mildly acidic system, and why the stabilised copper form exists at all.

Light and oxygen act separately, through photo-oxidation, and they fade a green towards a paler, greyer tone without the olive shift. In a transparent pack with a long ambient shelf life, this is usually the dominant route and the one that a standard stability report generated in the dark will not reveal. Thermal load during processing compounds both: the longer the material is held hot, the more the pigment is committed to degradation before the shelf life has even started.

The practical controls follow from that. Keep the thermal load as short as the process allows. Work at the highest pH the formulation can tolerate where the material permits it. Protect the pack from light rather than relying on the formula. Use an antioxidant system where the regulations for the market allow it. And where the finished product has a long ambient life and a clear pack, accept that the colour choice is effectively made by the packaging and test under real retail lighting before committing.

What a colour figure means for green

Green has an additional measurement problem of its own. A chlorophyll content figure tells you how much of the pigment is present, but not what the material looks like, because the visual hue depends on the balance between the green pigment and the yellow and blue components surrounding it. A colour value tells you how strong the preparation is in a defined medium, and it is only comparable with another figure measured in the same medium and at the same concentration. Neither figure converts into the other.

For a blend, the situation is worse, because the quotation may carry a figure for each component and nothing for the blend. The only comparison that survives is a measurement taken in your own medium at your own use level and, where the pack is transparent, read as a hue rather than as an intensity. Ask for the reading in the system you are actually colouring.

Label wording, and the three ways green can be declared

The declaration follows the route, and the three routes give three different statements. A copper chlorophyllin preparation is declared as a colour, under whichever naming convention the destination uses. A plant powder such as matcha is declared as an ingredient, with its own position in the ingredient list. A blend of two colorants is declared as two colours, which is a longer statement than most brands expect when they brief a “natural green”.

This matters at the artwork stage rather than at the formulation stage, because the claim the brand wants and the declaration the formula produces have to agree. A product coloured with two additives cannot honestly claim that all of its colour comes from plants, and a product coloured only with a plant powder will carry a much higher load of that powder. Settling which statement the label has to carry is the cheapest way to choose between the three routes, and it is much cheaper before the artwork is designed than after.

What to write into the purchase specification

State which route the material belongs to and identify it unambiguously — the copper complex or the native pigment, the named plant powder, or both components of a blend. State the solubility direction, since water-soluble, oil-soluble and water-dispersible forms of the same family are not substitutes. State the colour figure together with the method and the medium in which it was measured, or the pigment content and its method, and be explicit about which of the two you are buying against.

State the carrier and emulsifier system for any formulated dispersion, because those components carry their own declaration. State the pH range and the thermal process the material has been tested against, and require the supporting data rather than a statement of suitability. State the intended market, because the permitted lists differ and the choice of family can depend on that answer alone. And state the storage conditions and shelf life, because a green pigment loses ground in storage long before it loses ground in the finished product.

Frequently Asked Questions

Is chlorophyllin the same thing as chlorophyll?

No. Chlorophyll is the natural pigment, which is oil-soluble and degrades to an olive-brown tone when its central magnesium is displaced. Copper chlorophyllin is a derivative in which that metal has been replaced, which makes the material more resistant to acid and allows a water-soluble form. They are listed separately in the regulatory frameworks and they are not interchangeable in a formula.

Can I make a natural green by blending a natural blue and a natural yellow?

Yes, and it is how most commercial natural greens are made. The blend inherits the weaknesses of both components, so the hue drifts towards whichever one fades faster, and the two have to be compatible in phase. The stability work has to be done on the blend, not on the components separately, and both components have to be declared.

Why does my green turn olive or brown?

Because the pigment’s central metal atom has been displaced, and acid and heat both drive that reaction. It is the characteristic failure of the chlorophyll family and the reason the stabilised copper form exists. Where the system is acidic, choose the copper complex; where it is not, reduce the thermal load and the holding time before reaching for a different material.

Can I use matcha as a colour instead of a colour additive?

You can, and it will be declared as an ingredient rather than as a colour, which is often the point. But the colour is a crop and processing property rather than an engineered figure, it varies between lots, and the use level is set by taste and by the powder’s other components. It is a formulation decision, not a substitution.

Which natural green survives a long ambient shelf life?

The stabilised copper complex is the most robust of the family and the usual answer where a long life at ambient temperature is required. Blends and plant powders are more exposed, particularly in a transparent pack, where light becomes the dominant degradation route. Where the pack is clear and the shelf life is long, treat the packaging as part of the colour decision and confirm it in a real-time study.

Two measurements that answer different questions

Moisture content is a quantity. It expresses how much water the material contains, as a percentage of its mass, and the common methods are loss on drying in an oven, Karl Fischer titration, near-infrared measurement against a calibration, and gas chromatography for specific cases.

Water activity is a condition. It describes how available that water is, expressed on a scale from zero to one as the ratio between the vapour pressure above the sample and the vapour pressure above pure water. It is measured with a chilled mirror or a capacitive sensor, and because it is a thermodynamic property it is only defined at a stated temperature.

The practical difference is that moisture tells you how much water is there and water activity tells you what that water can do. A powder built on gums, fibres or starches can hold a substantial amount of water in a bound form and still sit at a modest water activity, because the water is not free to move. A powder that has taken up water only at its surface, or that contains a crystalline hydrate, can show a low moisture content and a water activity high enough to matter. Neither number predicts the other, and quoting one as a proxy for the other is the root of most moisture-related misunderstandings in the ingredient trade.

Why a certificate can show 5% moisture and tell you nothing useful

The moisture figure on a certificate of analysis is often reported without the method, and that is where it loses its value. Loss on drying is not a water-specific measurement: it drives off whatever is volatile at the conditions used, which includes residual solvent, essential oil and volatile flavour components. A botanical extract carrying a percent of volatile oil will read high on loss on drying without being wet in any sense that affects its stability. Karl Fischer measures water specifically, and the two methods routinely return different figures for the same lot.

The conditions matter as much as the method. An oven determination at 105 °C for two hours, a vacuum dry, and a Karl Fischer titration are three different measurements with three different answers, and a specification that says only “moisture: not more than 5%” allows any of them. Add the question of basis — as-is or calculated on a dried basis — and two suppliers can both be reporting truthfully while describing materials that behave quite differently in the same sachet.

This is why the method is part of the number rather than a footnote to it. When moisture is the parameter that decides whether your blend can be weighed accurately, whether your tablet presses cleanly or whether your stick pack stays free-flowing, a figure without a method is not a specification.

What water activity actually predicts

Water activity maps onto the behaviours that moisture content only hints at. The growth thresholds of spoilage organisms — moulds, yeasts and bacteria — are conventionally expressed against water activity rather than moisture content, because the organisms respond to available water and not to total water. Most moulds stop growing below a water activity in the region of 0.6, but the exact thresholds vary by organism and product, so the convention is a starting point for setting limits rather than a figure to copy.

It also predicts the physical behaviour of the powder. Caking and loss of flow happen when surface moisture allows particles to bridge and the bridges then set; a hygroscopic material can pass a moisture specification at packing and still cake after a few weeks if the water it takes up is available rather than bound. Chemical degradation follows the same logic: hydrolysis, enzymatic activity in materials with residual enzyme, browning reactions and the loss of labile actives all track water activity more closely than they track the percentage of moisture. Where a stability programme has been designed on moisture content alone, it has been designed around the wrong variable.

Moisture that arrives from somewhere else

A great deal of the water that causes trouble was not in the material when it was released. It arrived in transit, in storage, or inside the customer’s own plant.

Temperature cycling is the classic route. A container crossing several climates takes the powder and its packaging through repeated condensation cycles, and the water that condenses on the inside of the liner is then absorbed at the surface of the powder and at the seam where the liner is folded. The result is a lot that conformed at the supplier’s gate, arrived with a soft crust at the top of the drum, and is measured at a different moisture content depending on where the sample was taken from — which is one of the reasons sampling discipline matters as much as the analytical method.

Blending creates a second route that is easy to miss. Two powders with different water activities, packed together, will equilibrate: the component with the lower activity gains water and the component with the higher one loses it. Where a hygroscopic carrier such as maltodextrin or a gum sits alongside a salt or a sugar in one blend, the equilibrium that establishes inside the pack can raise the risk of caking above what either component suggested on its own. A blend that was free-flowing at packing can be a solid block at arrival in a warm warehouse without any packaging failure at all.

The third route is the one in your own facility. A part-used drum that is resealed carelessly, or left open between weighing operations, is the most common source of caking complaints that get attributed to the supplier. The head space of a drum in a humid room is an efficient way to raise the surface moisture of everything underneath it, and the effect is not uniform, which is why the moisture measured from the top layer of an opened drum can be badly misleading.

Methods, sampling and why two laboratories disagree

Disagreement between laboratories on moisture usually has three explanations, and only one of them is the analytical method. The first is sampling: a sample taken from the surface of a drum is not a sample of the drum, and a composite built from increments taken at inconsistent depths will not be reproducible. The second is sample handling after sampling. A powder in a jar that has been opened twice, weighed slowly, or left on a bench in a humid room will gain or lose water before the determination, and the result reflects the journey rather than the lot. The third is the method itself, including the drying temperature, the drying time, and whether the technique measured water specifically or everything volatile.

Water activity has its own version of this. Because the measurement is temperature-dependent, a result quoted without the temperature at which it was taken is only partly a specification, and results read at 25 °C and at 40 °C are not the same figure. The instrument’s calibration and the equilibration time allowed before reading both affect the outcome, and a sample that has not reached equilibrium with the sensor head will read low.

The response is procedural rather than technical: name the method on your own specification, require the same method on every certificate, fix the sample handling in writing, and ask for the temperature alongside any water activity figure. Where a supplier cannot state the method, the number is not comparable with anything.

Specifying both, and setting limits that mean something

The two figures belong on the specification together, each with the information needed to interpret it. Moisture content should be stated with its method, its conditions and its basis. Water activity should be stated with the temperature at which it is determined. Both should be supported by a sampling instruction, because a limit cannot be enforced on a sample that was not representative.

The limits themselves should be set against the failure you are trying to prevent rather than copied from a generic range. Where the risk is microbial, water activity is the controlling figure. Where the risk is caking, flow or accurate weighing in a low-dose blend, both figures matter and the packaging barrier is part of the answer. Where the risk is loss of assay through hydrolysis, the combination of water activity and storage temperature is what drives the reaction rate. The finished product’s own requirement — a tablet, a stick pack, a clear beverage that must not cloud — sets the acceptable window, and it is usually tighter than any default a supplier would propose.

Where a supplier can report only one of the two, treat that as a gap to close at qualification rather than a detail to accept. A water activity figure on the qualification lot plus a moisture figure on every certificate is a workable compromise; neither figure alone is a specification of the material’s condition.

What to write into the purchase specification

State the moisture content with its method, its conditions and whether the basis is as-is or dried, and state the water activity with the temperature of determination. State the sampling instruction, including the number of increments, the depths and the sample size, so that a result can be reproduced by a second party. State the packaging and the liner that the limit assumes, because the limit is only achievable in a defined barrier. State the storage conditions under which the limit applies, and whether the certificate is issued per lot or on a periodic basis for the water activity figure. State the retention requirement and the container the retain sample is held in. And state the opened-container rule for your own operation, because the largest single cause of a moisture complaint is a part-used drum rather than a failed lot.

Frequently Asked Questions

Is moisture content the same as water activity?

No. Moisture content is the quantity of water present, expressed as a percentage of mass. Water activity is a measure of how available that water is, on a scale from zero to one, and it is defined at a stated temperature. A material can have a high moisture content with a modest water activity if the water is bound, and a low moisture content with a water activity that matters if the water is free.

My powder meets the moisture specification but cakes in storage. Why?

Because the moisture specification does not describe where the water is or how available it is. Caking is driven by surface moisture and by the migration of water between components, not by the total percentage. Check the water activity, the packaging barrier, the temperature history in transit, and how the container is resealed after part use, which is a frequent cause in the buyer’s own warehouse.

Why do two laboratories report different moisture figures for the same lot?

Usually because they used different methods or different conditions, because the sample taken was not representative, or because the sample gained or lost water between opening and analysis. Loss on drying, Karl Fischer and near-infrared are not interchangeable, and a drying temperature or time that differs will give a different figure. Name the method in the specification so that results stay comparable.

What water activity limit should I specify?

Set it against the failure you are preventing rather than from a generic table. The growth thresholds of spoilage organisms, the onset of caking and the rate of hydrolysis all sit at different water activities for different materials. Confirm the limit with your own microbiology and stability data for the finished product, and state it with the temperature at which it is measured.

Does loss on drying measure water?

Not specifically. Loss on drying measures everything volatile under the conditions used, which can include residual solvent and volatile oils as well as water. That is why a botanical material containing essential oil can show an elevated loss on drying without being wet, and why the method has to be stated with the figure.

The changes that are not announced as changes

Harvest variation is the change that requires no decision by anybody. The same species grown in a different region or a different season, or harvested at a different maturity, gives a different marker content, a different colour, a different bitterness and a different yield. None of that feels like a change to the supplier, who is simply supplying the material that the last harvest produced, and it is the reason a specification built from a single reference lot is a fragile document.

Process variation is the second family, and it is where most of the consequential changes hide. A different extraction solvent or a different number of extraction passes, a change from tray drying to spray drying, a different milling or sieving step, a switch in the sterilisation route — each of these alters the material in ways that a marker assay may not capture at all. The particle size distribution moves, the colour moves, the solubility moves and the flavour moves, while the number on the certificate can stay inside its window.

Carriers and diluents are a third family, and a frequent one, because standardisation is achieved by adjusting the material. Changing the carrier from one maltodextrin to another, or from a maltodextrin to a gum, is presented as a sourcing improvement rather than as a material change, but it changes the allergen statement, the carbohydrate figure, the hygroscopicity and the declaration. So does a change in the ratio of carrier to active.

Behind those sit two more that are easy to overlook. The first is a change of sub-supplier or of the certified operator behind a certificate, which alters the chain of custody that the certification depends on. The second is a change of analytical method or laboratory, where the physical material is identical and the reported figure nevertheless changes because the measurement changed. That last one produces the most confusing conversations of all, because both parties can be right.

Why a conforming lot can still be a problem

The specification is a set of limits, not a description of a material. When a change moves the material inside those limits, the lot conforms and the supplier has done nothing wrong on paper, yet the consequences land on the buyer. The reason is that the specification captures the parameters you thought to write down, and the process downstream depends on parameters you did not.

Consider what a formulation actually relies on. A tablet runs on flowability, compressibility and particle size distribution rather than on marker content. A beverage powder runs on dispersion behaviour, on the absence of sediment and on taste. A capsule fill runs on bulk density. A clear liquid runs on colour and on the absence of haze after storage. None of those properties is normally captured by a limit range, so none of them is protected by conformance. A change that shifts the particle size distribution by a few tens of microns can move a filling operation from comfortable to marginal while every certificate in the file reads correctly.

The knock-ons extend beyond your own process. Allergen statements change when the carrier changes. Certification scope changes when the processing site changes, because certification attaches to an operation rather than to a product name. Claims that were built on a marker figure change when the standardisation basis changes. Where the finished product is sold with origin or composition statements, the wording may have to move as well. And commercially, if the material was priced against a yield, a change in yield changes the cost per kilogram of what you are actually buying, which is the figure the buyer should have been negotiating on from the start.

What has to be redone, and in what order

The trigger test is straightforward even if the work is not. Ask whether the change touches a parameter that any of your own critical quality attributes depends on. If it does, the change is not a documentation matter; if it does not, document it and keep monitoring.

Where the change does matter, the work usually runs in this order. Establish identity and the marker figure on the new material, and confirm that the analytical method is the same one used previously, because a moved figure may be a moved method. Generate stability data on the new material, since earlier stability work describes the earlier material and cannot simply be inherited. Refresh the allergen, certification and composition statements, which follow the material and the site. Rebuild the sensory and colour reference, because that is where botanicals differ most and where analytical data is least informative. Run a trial in your own process at the intended use level, with the finished product tested after storage rather than fresh. Update the artwork and the ingredient statement if the declaration has changed. And notify your own customers where the agreement requires it, because the obligation to the brand downstream does not disappear simply because the ingredient still conforms.

Not every change deserves the full sequence, which is why a tiered response is worth defining in advance: documentation only, documentation plus a sample and a targeted test, or a full re-qualification with stability work and a process trial. Deciding the tier after the event is how organisations end up either over-reacting to a harmless change or discovering a material change halfway through a production run.

Writing change control into the supply agreement

The obligation belongs in the agreement, in specific terms rather than as a general aspiration to communicate.

The notification requirement should specify the window, and the window has to be measured from the point at which the change takes effect rather than from the point at which the buyer discovers it, which generally means notification in advance of the affected shipment. The notice itself should be defined by content: what changed, why, from when, whether it is permanent or specific to a lot, what the material’s figures now are, what data accompanies the change, and whether any statement, certificate or declaration is affected.

The buyer’s side then needs its own clause. An assessment window, the right to require a qualification sample before further shipments, and the right to decline the change and to treat the agreement as no longer describing the material. Cost allocation should be settled while everyone is still friendly, because it is much harder to negotiate after a production line has stopped. Decide in advance who pays for validation work, for artwork changes, for stability testing and for the disposal of stock that can no longer be used.

Finally, make the specification itself a controlled document. Give it a revision number and an effective date, keep a change history in it, require the certificate of analysis to state the revision it was issued against, require the purchase order to reference the revision in force, and say what happens if a change ships without notice — the remedy, not merely the expectation.

The documents that make a change visible

Visibility comes from a small number of habits rather than from a large system. The specification should carry a revision number and an effective date, and should name the parameters considered critical, so that both parties can see which of them trigger requalification. The certificate should reference the specification revision it was issued against, which is the single most useful line on the document and the one most often missing. A change notification form, kept simple and used every time, does more good than a long procedure that nobody follows. A change register, reviewed at the annual meeting, turns a series of individual decisions into a visible history.

An annual review of the lots received is worth the effort for a different reason: it detects drift that never produced a notification. Plotting the marker content, moisture and colour of consecutive lots against the original reference material shows movement that each individual certificate described as conforming, and it usually surfaces a slow change several shipments before the taste panel does.

When a change has already shipped

The first action is comparison, not negotiation. Retain samples of the approved material are the reference, and a side-by-side set of results, taken on the same method and read by the same panel, settles most disputes without argument. Comparing against a retain sample also answers the question that matters commercially: is this a different material, or is it the same material behaving differently in your process?

From there the options are the ones change control exists to make available: accept the change and issue a new specification revision, use out the existing stock under a documented assessment, or reject the shipment and treat the agreement as being in dispute. Each of those is a defensible decision; what is not defensible is a silent acceptance that leaves the specification describing a material nobody will receive again.

What to write into the purchase specification

State that the specification is a controlled document with a revision number and an effective date, and that any change to it requires prior written notification before the affected shipment. State what the notification has to contain, the assessment window, and the remedy for a change that ships without notice. State which parameters are treated as critical for your process, so that the supplier knows which changes trigger a qualification sample and which ones are documentation only. Require the certificate to reference the specification revision in force. Require a sealed reference sample of the approved material, retained for the life of the finished product. And state the documentation that must accompany every lot, so that a change cannot arrive without the statements it invalidates being updated at the same time.

Frequently Asked Questions

Botanical material is naturally variable, so why bother with change control?

Because variability and uncontrolled change are different problems. Seasonal and harvest variation is a known source of drift that a well-written specification with ranges can absorb, and it is monitored rather than prevented. The purpose of change control is to make deliberate changes — a new process, carrier, site or method — visible before they are shipped, so that the buyer can decide whether the change matters.

What counts as a change that has to be notified?

Anything that alters the material or the way it is described: raw material origin, variety or harvest; extraction and drying route; milling, sieving or sterilisation; carrier identity or ratio; the site where the material is processed; the certified operator behind a certificate; and a change of analytical method or laboratory, which changes the reported figures without changing the material.

Do I have to requalify after every change?

No. Define a tiered response in advance. Changes that touch a parameter your own critical quality attributes depend on warrant a sample and a trial, and usually stability work. Changes that do not touch those parameters can be documented and monitored. The important part is deciding which parameters are critical before the change arrives, not after.

The change is inside my specification limits, so does it matter?

It can. A specification is a set of limits rather than a description of the material, and much of what a formulation depends on — flow, compressibility, dispersion, colour, taste after storage — is not captured by typical limits. A lot can conform on every line and still behave differently in your process, which is why the critical-parameter list and the reference sample do most of the protective work.

What can I do if a change ships without notice?

Compare the lot against your retain sample of the approved material, using the same method and the same sensory reference, and establish factually whether the material has changed. Then decide between accepting the change with a revised specification, using out the stock under a documented assessment, or rejecting the shipment. Retaining a sealed reference sample at qualification is what makes this decision evidential rather than a matter of opinion.