For liquid dairy and dairy-style ready-to-drink products, the sterilising step is not a background detail a flavour simply has to tolerate. It rewrites the flavour twice. It alters the delicate notes a developer adds, and it generates a strong cooked, sulphury character of its own that the added flavour then has to sit on top of. UHT milk, flavoured milk, creamers, ready-to-drink coffee, canned and retort-pouch desserts and long-life plant-based drinks all pass through heat severe enough to change what the consumer tastes. Getting a flavour through that intact starts with knowing which process it faces, because UHT and retort are two different heat problems.
UHT and Retort Are Not the Same Heat
The two routes to a shelf-stable liquid apply heat very differently. UHT, or ultra-high-temperature processing, is a continuous treatment using indirect heat exchange or direct steam injection or infusion, followed by aseptic packing. Tetra Pak's Dairy Processing Handbook describes both heating routes. In-container sterilisation, the conventional retort route, heats the product inside its sealed final pack. A flavour that clears a UHT line is therefore not automatically fit for a retort pouch.
The Cooked Note the Process Makes
Before any added flavour is considered, both processes create a taste of their own. A 2006 study in the Journal of Dairy Science found substantially higher concentrations of hydrogen sulphide, methanethiol, carbon disulphide, dimethyl trisulphide and dimethyl sulphoxide in UHT milk than in raw or pasteurised milk. A 2014 study in Dairy Science and Technology identified nine volatile sulphur compounds in commercial UHT milk and found hydrogen sulphide, methanethiol, dimethyl sulphide and dimethyl trisulphide “initially higher than their reported threshold values indicating their importance in milk flavour, especially cooked flavour.” This is the baseline a dairy flavour is built over, not a blank canvas.
Browning Depends on the Process and the Milk
Deeth's 2021 review in the MDPI journal Encyclopedia notes that freshly processed UHT milk has no noticeable brown colour while in-container sterilised milk does. Assess that processed base when choosing an added flavour. The sugar composition also matters: Tetra Pak's Dairy Processing Handbook explains that lactase hydrolysis releases glucose and galactose, which can increase Maillard browning during heat treatment and storage. That point applies to lactase-hydrolysed milk, not every milk labelled lactose-free. Specify the actual sugar composition in the brief.
The Cooked Note Is a Moving Target
The cooked baseline does not stay put, which complicates the brief. The 2014 sulphur-compound study found that hydrogen sulphide, methanethiol, dimethyl sulphide and dimethyl trisulphide started above their thresholds but “decreased slowly during storage to levels below their threshold values,” a decline that tracked the fall in dissolved oxygen. The heat-stable brief for dairy is a shelf-life brief from the outset.
What the Heat Does to the Flavour You Add
The flavour a developer adds encounters a protein-rich, watery matrix. A 2023 review in Frontiers in Nutrition describes flavour substances as “low molecular mass volatile compounds” with “poor stability in the presence of air, light, moisture and heat,” and warns that flavour compounds carrying carbonyl groups, such as aldehydes and ketones, “can react with the amino groups of proteins and result in flavour loss.” Milk is full of those amino groups. In a 2020 spray-drying study in the Pakistan Journal of Biological Sciences, retention ranged from 84.5-97.9 per cent for orange oil to 44.4-72.5 per cent for isoamyl acetate under identical conditions. The authors put the gap down to the lighter compound's “relatively small molecular sizes and high vapor pressure.” These are spray-drying results, not retention figures for UHT processing or sealed retort packs. Spray drying allows volatiles to escape with exhaust air. Tetra Pak's Dairy Processing Handbook describes deodorisation during flash cooling on direct-steam UHT lines. In a sealed retort pack, partitioning into the headspace and chemical transformation are distinct from escape out of the pack.
Where the Flavour Meets the Heat, and When
Two levers decide how much of that loss a developer actually takes. The first is when the flavour is added. On an aseptic UHT line, flavour can be added after the sterilising step and before filling only if the flavour itself is commercially sterile, through sterile filtration or separate sterilisation, and is introduced through a validated aseptic dosing system. This avoids the base product's peak heat treatment but constrains both the flavour format and the line. The FDA's aseptic-processing guide addresses commercial sterility of the product and the equipment downstream of sterilisation. For conventional in-container retort, the flavour goes in before the pack is sealed and processed. Where the flavour must go in before the heat, the second lever is the delivery system. The Frontiers review notes that spray drying, the most common way to turn a liquid flavour into a stable powder, keeps the core cooler than the drying air but can still “lead to the degradation of thermally unstable flavours,” while gentler routes such as extrusion, which the review notes rarely exceeds 118°C, suit heat-degradable notes. In a watery dairy matrix the carrier also has to dissolve and behave, so an emulsion or a matrix-matched system, rather than a powder chosen for a dry mix, is usually the right starting point. The general principle is covered in our guide to flavour encapsulation.
Retention Is Measured in the Product
Whatever the delivery system, the number that matters is retention in the finished product, not encapsulation efficiency on a specification sheet. For an assay measuring total and surface oil, encapsulation efficiency is the encapsulated fraction of total oil, (total oil minus surface oil) divided by total oil; retention records how much is still present and in balance after the heat and the storage, and the two can diverge widely by compound, as the spray-drying figures above show. For dairy the test has to be the real one: the flavour dosed into the actual matrix, run through the actual UHT or retort profile, held for the actual shelf life at the actual storage temperature, and then measured. A figure generated in water, or in a gentler process, will flatter a flavour that then fails in the pack.
Building Over the Baseline, Not Against It
The dairy job is not only to protect the added flavour but to reconcile it with the cooked and caramelised notes the process leaves behind. Sometimes the answer is to complement them, letting a malt, caramel or vanilla profile absorb a mild cooked note. Sometimes it is to cover a sulphury or eggy edge, which is the work of flavour masking. The process choice matters too: a 2019 study in the Journal of Dairy Science found that the profile of sulphur compounds differed with the heating method and identified milk serum proteins as their source, so a direct-steam and an indirect-heat UHT line do not hand the flavourist the same starting note. None of this can be settled from a catalogue; it is settled against the specific product and process.
What Belongs in a Dairy Heat-Stable Brief
A workable brief names the process and the product, not just the taste. It states whether the product is UHT or retort and gives the temperature-time profile. It describes the matrix: the fat and protein levels, the pH, whether lactase has hydrolysed the lactose, and the actual sugar composition. It fixes the dosing point: before sealing and processing for conventional retort, or before UHT treatment unless a commercially sterile flavour and validated aseptic dosing system allow addition after treatment and before filling. It sets the shelf life and storage temperature the flavour must hold through, because the cooked baseline moves over time. And it asks for retention measured in that product, through that process. The declaration travels with all of this: an added flavouring still declares as a flavouring, and its carriers belong in the specification, as we set out for a Singapore label. Where a flavour is halal-certified, confirm the documentation applicable to that product. FSSC 22000 certifies a food safety management system, not individual flavours.
Choosing a Flavour Format with VKA
At VKA, we can help you choose a flavour format for your product and dosing point. Our guide to liquid and powder flavours explains the format choices. Explore our beverage solutions and Essences Portfolio, or speak to a flavourist directly about your process, matrix and target shelf life.
Sources
- Deeth, Effects of High-Temperature Milk Processing, MDPI journal Encyclopedia 1(4):1312-1321 (2021)
- Al-Attabi, D'Arcy and Deeth, Volatile sulfur compounds in pasteurised and UHT milk during storage, Dairy Science & Technology 94(3):241-253 (2014)
- Vazquez-Landaverde, Torres and Qian, Quantification of Trace Volatile Sulfur Compounds in Milk by SPME and GC-PFPD, Journal of Dairy Science 89(8):2919-2927 (2006)
- Jo, Carter, Barbano and Drake, Identification of the source of volatile sulfur compounds produced in milk during thermal processing, Journal of Dairy Science 102(10):8658-8669 (2019)
- English et al., Flavour encapsulation: a comparative analysis of relevant techniques, physiochemical characterisation, stability, and food applications, Frontiers in Nutrition 10:1019211 (2023)
- Farouk et al., Effects of Carriers on Spray-dried Flavors and Their Functional Characteristics, Pakistan Journal of Biological Sciences 23(3):257-263 (2020)
- Tetra Pak, Long-life dairy products, Dairy Processing Handbook
- Tetra Pak, Lactose-free dairy products, Dairy Processing Handbook
- FDA, Aseptic Processing and Packaging for the Food Industry
- Jafari, Assadpoor, He and Bhandari, Encapsulation Efficiency of Food Flavours and Oils during Spray Drying, Drying Technology 26(7):816-835 (2008), doi 10.1080/07373930802135972



