Format is a carrier decision, not a flavour decision. The IOFI Code of Practice defines carriers as ingredients that “Carriers (including carrier solvents, listed under Annex I, Chapter 18.2.1) are used to maintain uniformity and dilute concentrated flavorings in order to facilitate their incorporation and dispersion in food products. ”, which is why the same approved profile is sold as a liquid, an emulsion, a spray-dried powder or an encapsulate. The choice matters legally as well as technically, because the limits that apply to a carrier are written on different bases in different markets: some are a ceiling in the flavouring, others a ceiling in the food as consumed.
What You Are Actually Formatting
IOFI adds the reason a format can also protect: “Some carriers may also be used for encapsulating flavorings with a view to protect them against evaporation and alterations during storage.” It helps to name the thing being formatted. IOFI calls a multi-ingredient flavour a compounded flavoring: “Compounded flavoring [4] is a term often used to describe mixtures of flavoring ingredients, some of them complex mixtures themselves, that are combined to provide a particular taste sensation (See also Article 5.1(c) of CODEX STAN 107-1981, rev. 2016). Other non-flavoring food ingredients, such as solvents, emulsifiers and antioxidants are required to allow the compounded flavoring to function properly in the food to which it is added.” The EU splits the components underneath it. A flavouring substance is “‘flavouring substance’ shall mean a defined chemical substance with flavouring properties”, while a flavouring preparation is “‘flavouring preparation’ shall mean a product, other than a flavouring substance, obtained from: (i) food by appropriate physical, enzymatic or microbiological processes either in the raw state of the material or after processing for human consumption by one or more of the traditional food preparation processes listed in Annex II” Note that IOFI's own glossary has no entry for a flavour system or a flavour base, in either spelling; the industry term is compounded flavoring.
Liquid Carriers and the EU Ceiling
Liquid flavours are the benchmark, and their carrier is the most tightly capped part of the specification. In the EU, Annex III Part 4 of Regulation 1333/2008 covers triethyl citrate, diacetin, triacetin and propylene glycol across all flavourings at “3 000 mg/kg from all sources in foodstuffs as consumed or as reconstituted according to the instructions of the manufacturer; individually or in combination. In the case of beverages, with the exception of cream liqueurs, the maximum level of E 1520 shall be 1 000 mg/l from all sources” Read the basis carefully, because it is the final food, from all sources, not the flavouring. That has an arithmetic consequence buyers rarely notice: if the carrier were the whole flavour, 3 000 mg/kg is 0.3 per cent in a foodstuff and the beverage limit of 1 000 mg/l is 0.1 per cent. It is a ceiling imposed by an additive limit, not a recommended use level, and it binds only where the carrier is one of those four.
Singapore Closes the Solvent List
Singapore closes the question differently, by closing the list. Regulation 22(2) of the Food Regulations permits twelve solvents and no others: 1,3-propanediol, benzyl alcohol, beta-cyclodextrin, diacetin, diethyl ether, ethyl acetate, ethyl alcohol, glycerol, isopropyl alcohol, propylene glycol, triacetin and water. The clause is quoted in full in our guide to choosing a flavour supplier in Asia. Regulation 22(3) then sets the grade: “The permitted solvents referred to in paragraph (2) other than water shall conform with the British Pharmacopoeia standard.” A liquid flavour formatted for a European or American customer in a solvent outside those twelve is not lawful to sell here, whatever its additive levels.
Australia and New Zealand: Different Route, Different Numbers
Australia and New Zealand reach the same carriers by a different route and with different numbers. Schedule 15 item 0.3 permits benzyl alcohol at 500 mg/kg in the final food and isopropyl alcohol at 1 000 mg/kg in the final food, butylated hydroxyanisole at 1 000, and ethanol, ethyl acetate, glycerol diacetate, glyceryl monoacetate and triethyl citrate at GMP. Propylene glycol, triacetin and glycerol arrive through the preparations-of-food-additives route in Schedule 16 rather than through item 0.3 directly. Two definitions from the Code are worth carrying into any spec review: “MPL means the maximum permitted level, measured (unless otherwise indicated) in mg/kg”, and “a reference to 'GMP' is a reference to the maximum level necessary to achieve 1 or more technological purposes under conditions of GMP.” GMP is a ceiling, not an absence of one.
Emulsions and the Two-Basis Problem
Emulsions exist because beverages are mostly water and flavour oils are not, and their additives are where the two-basis problem is clearest. In the EU, octenyl succinic acid modified gum arabic is capped at 220 mg/kg in the final food for flavouring-oil emulsions in flavoured drinks, rising to 400 mg/kg for non-carbonated drinks containing fruit and vegetable juices. Sucrose esters of fatty acids carry both bases in a single row, at 15 000 mg/kg in flavourings and 30 mg/l in the final food. Quoting one basis where the other applies inverts the answer by orders of magnitude.
Weighting Agents Split by Market
Weighting agents split by market. Sucrose acetate isobutyrate is permitted at 200 mg/kg in Australian and New Zealand water-based flavoured drinks but at 300 mg/kg in the EU, where the permission is restricted to cloudy drinks; glycerol esters of wood rosins sit at 100 mg/kg in both. A bare figure for either is wrong somewhere. Brominated vegetable oil appears in neither instrument, so it is simply not permitted in either, which is the operative fact even though neither text names it in a prohibition.
Spray-Drying: What the Wall Material Does
Spray-drying changes the arithmetic entirely, and the wall material is the variable that moves it most. A 2025 study in Current Research in Food Science, encapsulating d-limonene, reports that “The powder containing DE6 resulted in the highest surface oil percentage (6.6 ± 0.8 %) with the lowest d-limonene retention (57.6 ± 2.4 %).” Across the four maltodextrins tested, “The lowest retention was observed in the powders prepared with maltodextrin DE6, which retained only ∼57 % of d-limonene, while the highest retention (∼85 %) was achieved with maltodextrin DE21 and DE38.” Roughly twenty-eight percentage points of retention turned on the choice of wall material alone, at identical dryer settings.
The Conditions Behind Those Retention Numbers
The same paper gives an inlet temperature of 178 to 182 °C and an outlet temperature of 90 to 95 °C, controlled by adjusting the liquid feed flow rate, with all emulsions at 6 per cent oil, 0.9 per cent pea protein isolate and 20 per cent solids, on a laboratory mini spray dryer. It is one volatile on one dryer, not a general powder-retention rule, and industrial powders are commonly coarser than the 6 to 10 micrometre volume-mean sizes reported there.
Powder Additive Ceilings
Powders bring their own additive ceilings, and here the basis flips. In the EU, silicon dioxide is capped at 50 000 mg/kg in flavourings, a limit written inside the flavouring rather than in the food. Beta-cyclodextrin, used as an encapsulating wall, is capped at 500 mg/l in the final food for flavoured teas and powdered instant drinks and at 1 000 mg/kg for flavoured snacks. Singapore also lists beta-cyclodextrin among its twelve permitted solvents, so it is a permitted carrier for encapsulated formats sold here.
Water Activity Is the Storage Clock
Storage is where a powder actually fails, and the failure is a threshold rather than a slope. Water activity sets it. A 2024 paper in the Journal of Applied Glycoscience fixes the critical water activity as the water activity at which the glass transition occurs at 298 K, and gives maltodextrin, the wall material most flavour powders are built on, a critical water activity of 0.575 on a row it marks as compiled from earlier work. A separate freeze-dried maca study, which fixes the same critical point as the water activity at which the glass transition temperature reaches 25 degrees, reports that “There was negligible caking below a w = 0.328.” Treat both as method rather than specification: neither was measured on a flavour powder. What does transfer is the state check, and in the spray-drying study above “The moisture content for powders remained uniformly below 5 %, indicating that all were in the glassy state.”
What Anticaking Choice Changes
Anticaking choice moves predicted shelf life by roughly a factor of two under accelerated conditions, and by about 15 per cent under ambient conditions. In a 2023 accelerated-storage trial, tricalcium-phosphate-treated powder was predicted at 157 days ambient and 77 days accelerated, against 137 and 39 days for silicon-dioxide-treated powder, in aluminium-laminated pouches at 25 degrees ambient and 40 degrees accelerated. That trial was run on spray-dried apricot juice powder, so the ratio is the finding that carries across, not the day counts.
Process Survival
Process survival is the question a format is usually bought to answer, and there is no general figure for it. A measured baking figure for thyme essential oil in bread shows that recovery of thymol plus carvacrol after baking fell from 80.0 per cent at 50 ppm to 75.3 per cent at 200 ppm; the figures are quoted in full in our flavour shelf life guide. Thymol and carvacrol are far less volatile than a typical citrus or dairy top note, so 75 to 80 per cent is an optimistic bound rather than a rule. For encapsulated against liquid through a bake, a 2024 review restating a 2019 biscuit study reports that “The retention of cinnamaldehyde was higher in the biscuits with encapsulated oil (95.65%) compared to those with the liquid flavor (87.11%).”
Efficiency Is Not Performance
Before any of those retention figures is used to choose a format, one 2026 review is worth putting next to them. Writing in the International Journal of Food Science, its authors state that “high encapsulation efficacy is not necessarily enhanced in shelf life, sensory, or physiological efficacy in realistic processing and storage, as well as gastrointestinal processes”. An efficiency percentage on a datasheet describes what went into the capsule, not what reaches the consumer, so it cannot be read as a performance promise for your line. A measured loading curve from the same year shows one reason the two part company: a 2026 study in Foods, on vanillin complexes built with soy protein isolate, reports that “The binding efficiency of vanillin decreased from 91.03 wt.% to 69.43 wt.% with increasing vanillin loading”, with nanocrystal formation above 33.33 wt.% relative to the protein. The study used one laboratory carrier; a higher load holds a smaller proportion.
Encapsulate Size
Encapsulate size varies far more widely than a single technology suggests. A 2024 review states that “Depending on the technology and materials used, microcapsules can range in size from 0.02 to 10 000 μm in diameter”, which spans techniques from spray drying to coacervation rather than describing any one of them, and is a reason to ask a supplier for the particle size distribution of the specific product rather than for the category.
Four Questions That Decide the Format
Four questions usually resolve the format. Is the line wet or dry, because a liquid has nowhere to go in a dry blend without a fat system to carry it. How hot does it get and for how long, because heat is what encapsulation is bought to survive. What does the label need, since carriers and wall materials are ingredients with their own declarations and their own market-specific limits. And on what basis is each limit written, in the flavouring or in the final food, because the two bases are not interchangeable, and it is one of the answers to settle when choosing a flavour supplier in Asia. A practical pattern follows: develop in liquid where the application allows it, then confirm in the production format, because no conversion is perfectly transparent.
How VKA Handles Formats
Across its Singapore sites, VKA works with liquids, emulsions, spray-dried powders and the EssenceLock and FreshSeal engineered formats. This means the format conversation can happen inside the development loop rather than as a hand-off between suppliers. Browse the portfolios for what we make, or talk to a flavourist directly with your process conditions, because the right format is a property of your line, your label and your shelf life, not of the flavour alone.
Sources
- IOFI Code of Practice (5th Revision), sections 3.3.4 and 5.3.1, via IOFI General Resources
- Regulation (EC) No 1333/2008 on food additives, Annex III Part 4 (consolidated 18.08.2026)
- Regulation (EC) No 1334/2008 on flavourings, Article 3(2) (consolidated 16.02.2026)
- Australia New Zealand Food Standards Code, Schedule 15, Substances that may be used as food additives (compilation F2026C00407)
- Australia New Zealand Food Standards Code, Schedule 16, Types of substances that may be used as food additives
- Singapore Food Regulations, regulation 22(2) and 22(3) (permitted solvents)
- Jauhari et al., Encapsulating volatiles by spray drying: the choice of dextrose equivalent influences d-limonene retention, Current Research in Food Science 11:101224 (2025)
- Influence of anticaking agents and storage conditions on quality characteristics of spray dried powder, Foods 12(1):171 (2023)
- Browning, starch gelatinization, water sorption, glass transition and caking properties of freeze-dried powders, J. Appl. Glycosci. 67(4):111 (2020)
- Water sorption isotherm and critical water activity of amorphous water-soluble carbohydrates, J. Appl. Glycosci. 71(1):15 (2024)
- Estimation of the antifungal threshold of thyme essential oil for bread preservation, Foods 14(20):3549 (2025)
- Microencapsulation of essential oils and oleoresins: applications in food products, Foods 13(23):3873 (2024)
- Encapsulation in Food Systems, International Journal of Food Science (2026), doi 10.1155/ijfo/7367634
- Dual-Stabilized Vanillin Complexes Based on Soy Protein Isolate, Foods 15(7):1240 (5 April 2026)



