One tea lot → four preparations → two paired storage samples from each
Repeat runs require fresh independent preparations. Use a separate sample for warming without changing the main retain’s storage history.
A tea drink looks clear when freshly prepared, then develops a haze or sediment in the fridge. Another batch turns brown or leaves a film around the rim before refrigeration. Recording every case as “poor powder solubility” usually leads to the same responses: change the powder, raise the mixing temperature or stir harder. More useful questions come first. When did the change begin? Is it in the bulk liquid or at the surface? Does changing the water change the outcome? How much of the haze disappears on warming?
Tea cream is not dairy cream rising to the surface. In tea research, the term describes the formation of haze and precipitate. A 2023 green-tea study identified EGCG, caffeine, proteins and polysaccharides as participants, with EGCG–caffeine complexation playing a key role in the system studied.[1] An unclear bottle, however, is an observation—not a diagnosis. Combining that mechanism study with research on brewing water suggests a stepwise investigation for soluble tea extracts and concentrates.
Separate cold haze, powder lumps and surface film
Keep the first failed bottle intact rather than immediately filtering it. Photograph freshly prepared and stored samples against the same background and lighting, including a view after gentle inversion. Record floating powder, hard-centred lumps, uniform haze, bottom sediment and surface film separately. The purpose is to choose the next comparison, not to identify the chemistry from a photograph.
Lumps present immediately after powder addition make wetting and dispersion the first processes to investigate. A uniform drink that becomes hazy only after cooling calls for a different test of storage-induced aggregation. Treat surface film, or tea scum, as another category. The abstract of a tea-scum study reports that calcium and bicarbonate ions in the water mediate film formation; the film cannot simply be explained as deposited calcium carbonate sitting on the tea.[4] A rim film and a haze throughout the bottle should therefore not share a single fault code.
Record colour separately from haze. In a 2021 water-composition study, slightly alkaline green-tea infusions browned while EGC and EGCG declined over time.[3] That gives a different explanation from particles merely making the liquid look darker. Where darkening is pronounced without a corresponding change in sediment, investigate final pH, holding time and the catechin profile before specifying a finer filter.
What the tea-cream mechanism tells you to measure
The 2023 researchers prepared a green-tea infusion in purified water, centrifuged it to obtain a clarified infusion, and then observed tea-cream formation at 4°C.[1] Removing material by centrifugation at the preparation stage did not prevent new haze during subsequent chilling in that experiment. The practical lesson is to avoid treating hot clarity or immediate post-filtration clarity as the final test of refrigerated performance.
The researchers also removed small molecules by dialysis and added EGCG and caffeine back to the macromolecule-rich tea infusion retained inside the dialysis tubing. At the tested addition levels, after four hours at 4°C, adding either compound alone left this retained fraction clear; adding them together produced visible haze.[1] This is more informative than a list of substances found in the final sediment: it tests which combination matters in that particular green-tea system.
It does not mean decaffeination should be the first response to every cloudy tea drink. Changing caffeine, catechins or macromolecular composition changes the ingredient itself. If the brief calls for a particular tea character and composition, first establish whether water or reconstitution changes can solve the problem. Then compare ingredient grades if necessary. Retain samples of the whole drink, supernatant and sediment, and consider measuring caffeine and the main catechins to establish whether clarification has simply removed part of the tea.
Concentrates offer another trap. The abstract of a green-tea concentrate study reports that tea-cream quantity rose across the tested 5–40°Brix range, but fell in the 50 and 60°Brix samples; the authors proposed increased viscosity as a possible explanation.[2] This is not a general rule that stronger concentrates are more stable, nor evidence that a stable concentrate will remain clear after dilution. Include a refrigerated test at the intended final dilution rather than qualifying the concentrate alone.
Water quality is more than hardness or TDS
Keep calcium and magnesium, alkalinity and pH as separate entries in the investigation. The 2021 study compared synthetic waters of different hardness and included separate sodium bicarbonate and mineral-salt controls to distinguish alkalinity from the contribution of salts. EGC and EGCG decay and browning were particularly evident under slightly alkaline conditions.[3] These were teabag infusions, not reconstituted commercial extract powders. The transferable lesson is the experimental separation of factors, not a finished-beverage water specification.
For production water and a low-mineral reference, record calcium, magnesium, total hardness with units, alkalinity with its reporting basis, pH, conductivity, and the sampling point and time. If a complete analysis is not immediately available, comparing two real water sources is still useful. The conclusion must remain “water-source related,” however, rather than “caused by calcium.” A TDS reading cannot replace composition data when the task is to distinguish calcium and magnesium effects from alkalinity.
Raw-water pH is not a substitute for beverage pH. Measure the liquid after reconstitution, after any recipe acidification, and after the heat process, using a calibrated method appropriate to the sample and recording measurement temperature. If one water produces a drink that is both hazier and browner, investigate salt composition and final pH separately rather than selecting treatment equipment from the label “hard water.”
First screen: eight conditions that distinguish water, mixing and storage
The following is a proposed development screen, not a validated production protocol or a report of completed trials. Hold the tea lot, target tea solids, final recipe, container and fill level constant. Compare two waters, two permitted reconstitution routes and two storage conditions. Use independently prepared replicates: measuring one bottle three times assesses measurement repeatability, not preparation variability.
Scroll tables horizontally; focus a table region and use the arrow keys.
| Factor | Comparison | What to control or record |
|---|---|---|
| Water | Low-mineral reference versus current production water | Measured composition rather than names such as purified or softened water |
| Reconstitution | Current route versus one predefined alternative within the ingredient’s permitted conditions, such as warm pre-dissolution before making up to final quantity | Solids-to-water ratio, addition order, temperature history, mixing time and equipment; identical final composition |
| Storage | Controlled non-refrigerated reference versus refrigerated challenge | Actual sample temperature, cooling rate, elapsed time and container; predefined observation points for both arms |
Prepare the four water-by-reconstitution combinations, then split each into paired storage samples to create eight conditions. Storage pairs share a preparation; repeat runs require fresh independent preparations. If 4°C is selected as the cold challenge, it follows the temperature used in the mechanism study, but that paper’s four-hour endpoint is not a universal stability-test duration.[1] Cover the actual failure window. A drink that fails overnight cannot be cleared after a quick inspection as soon as it has cooled.
Define tea solids explicitly. When eventually comparing powders with different carriers, equal powder addition is not necessarily an equal tea-solids comparison. Record both powder addition per litre and the supported tea-derived solids contribution. If carrier proportions are unavailable, preserve that uncertainty rather than using identical °Brix as proof of identical tea concentration. Using a single powder lot in the first screen avoids introducing this additional variable.
Take baseline measurements after reconstitution at a common reference measurement temperature, not while one sample is hot and another is cool. Record refrigerated appearance and suitable haze readings at the storage temperature, then use a separate sacrificial sample for a standardised warming observation. Repeatedly taking the main retain sample out to warm it creates a different storage history.
Measure whether aggregation decreased—or tea components were removed
A minimum record should include sample identity, independent preparation batch, water analysis, tea-derived solids, final pH, time and temperature, initial and stored haze, colour, sediment or surface film, and warming response. For a turbidity meter, keep the method consistent and manage bubble interference. For transmission or absorbance, fix wavelength, path length and sample conditions; do not directly relabel the result as NTU. The 2023 study tracked haze with a spectrophotometer at 520 nm and reported results as percentages. The retrieved full text does not explicitly define how that percentage was calculated, so it should not be interpreted as transmittance, absorbance or NTU, or adopted directly as an acceptance limit for another instrument.[1]
Where sediment is present, consider a separate analytical sample with standardised centrifugation, recording force, time and temperature. This compares separable material; it does not specify a production centrifuge. Examine the supernatant, the recovered sediment and the original bottle. Looking only at the supernatant misses material that has already settled out.
If filtration or centrifugation produces a clearer liquid, calculate catechin and caffeine distribution from the quantity and measured composition of each stream. Compare the same batch before and after treatment rather than relying on concentration in the supernatant alone. Added water, equipment hold-up and moisture in the separated solids all affect the interpretation. A usable clarification process must meet composition, flavour and colour targets as well as an appearance target.
Turn the pattern into the next decision
Scroll tables horizontally; focus a table region and use the arrow keys.
| Observed pattern | Working hypothesis to prioritise | Most useful next check |
|---|---|---|
| Lumps appear immediately and decrease with a different mixing route; both waters behave similarly | Wetting, dispersion or addition order | Separate temperature, addition rate and mixing effects; examine lump interiors, then retain the cold-storage test |
| Initially uniform drinks become hazy after chilling with either mixing route | Storage-induced interactions deserve more attention than undispersed powder | Compare warming response and supernatant/sediment composition; then screen ingredient grade and tea-solids level |
| Failure occurs mainly with production water | A water-related effect, not yet a specific-ion diagnosis | Separate mineral salts and alkalinity using measured water composition, and check final pH |
| More alkaline samples darken while EGC/EGCG decline | Consistent with the oxidation pathway observed in the water study.[3] | Compare holding times and oxygen exposure; inspect the catechin profile rather than haze alone |
| A surface film dominates while the bulk liquid remains relatively clear | Investigate tea scum separately.[4] | Keep vessel geometry and exposed surface area constant while comparing water and formulation |
| Warming reduces haze, which returns on rechilling | Temperature-sensitive aggregation may contribute | Retain the cycle record and investigate composition and cold-storage conditions; do not declare the problem solved |
Several patterns can coexist. A warm reconstitution route may remove powder lumps while leaving the refrigerated haze unchanged. It has improved one stage, not qualified the ingredient for the whole application. Similarly, better appearance in low-mineral reference water must be followed by confirmation with the actual water-treatment system and final recipe.
Warming is a temperature-response check, not a food-safety test. Gas formation, unusual odours or unexplained progressive changes require a separate microbiological investigation, not tasting or reheating to decide release.
Only then choose between changing water, powder or process
If the first screen points to reconstitution, fix water and recipe while separating temperature, addition rate and mixing effects. Changing all three at once may produce a better sample without explaining why. If water is implicated, build controls from its measured composition and compare mineral-salt control with alkalinity control under comparable final formulation conditions. If both waters and both mixing routes fail during refrigeration, compare tea grades or tea-solids levels, keeping flavour, catechin profile and caffeine alongside appearance in the selection table.
The useful supplier question is not simply “Is it fully cold-water soluble?” Ask what water, addition level, preparation procedure and storage conditions were used, and what appearance and constituent-retention results were obtained. Request a same-lot sample, carrier information and catechin/caffeine data, then agree on test conditions and the decision criteria. See the green-tea ingredient encyclopedia for material forms and composition, and the citrus green tea formulation for addition order and finished-drink checks. Include the comparison records when discussing beverage application development, so that water composition, reconstitution and cold-storage performance remain part of the brief.
The goal of tea-haze diagnosis is not to give every sediment the same name. It is to narrow the failure to one or two factors that can be tested and controlled. Establish when the change occurs, where it appears and whether water changes it before deciding to change the tea itself.
Sources
[3] Effect of Water Hardness on Catechin and Caffeine Content in Green Tea Infusions (2021; full text)
Related Products
Interested in our botanical extracts?
Contact us for samples, technical documentation, or custom formulation support.
Get in Touch