From Extract to Finished Product: Stability, Encapsulation and Regulatory Status of Clitoria ternatea Anthocyanins as a Natural Blue Ingredient

1. Introduction

Among the pigments available to a food or supplement manufacturer, blue is the scarce one. Anthocyanins, betalains and carotenoids between them cover yellow through red to purple, but very few natural pigments absorb in the 580–620 nm region that the eye reads as a true blue, and most industrial blue has therefore come from Brilliant Blue FCF, a petrochemical derivative [1]. That position is now changing for regulatory rather than technical reasons, as agencies move against synthetic dyes and manufacturers look for replacements that will survive the same processing their predecessors did.

Butterfly pea (Clitoria ternatea L.) is the leading candidate. A leguminous climber of tropical Asia, long used in Malaysia, Thailand, Indonesia and the Philippines to colour rice, cakes and beverages, it produces flowers of an intense blue that has been in food use for centuries without the stability failures that normally limit anthocyanin colours [1,2,3]. The pigments responsible are the ternatins, polyacylated delphinidin derivatives whose structure is the reason the colour holds [4,5].

The literature on this plant is large but unevenly distributed. There is detailed pigment chemistry, a substantial pharmacological literature, and a growing body of encapsulation work. What is thinner is the connection between them: the questions a formulator asks when deciding whether to put this ingredient into a product and at what pH, in what carrier, with what shelf life, and under what regulatory permission. This review is organised around those questions. It takes as its starting points two recent and complementary contributions — a systematic study of colour, storage stability and antioxidant capacity across pH and temperature [1], and a review of microencapsulation methods for the extract [2] — and sets their findings against the wider literature and the current regulatory position in the two markets that matter most to exporters.

2. The Pigment: Why Ternatins Behave Unlike Other Anthocyanins

Anthocyanins in solution exist as a pH-dependent equilibrium between four principal species: the red flavylium cation (AH+) at low pH, the blue quinonoid bases (A and A−) formed by successive deprotonation, the colourless hemiketal (B) formed by hydration at C2, and the yellow chalcones (Cc and Ct) that follow from it. For most anthocyanins the blue quinonoid forms are transient, appearing between roughly pH 6 and 8 and degrading quickly, which is why red anthocyanin colours are formulated into acid systems and blue ones are generally not formulated at all [1].

Ternatins escape this constraint by structure. They are malonylated delphinidin 3,3′,5′-triglucosides carrying 3′- and 5′-side chains built from alternating D-glucose and p-coumaric acid units; fifteen have been characterised, designated A1–A3, B1–B4, C1–C5 and D1–D3 [4,5]. The aromatic acyl groups on those side chains fold back over the chromophore, a sandwich arrangement known as intramolecular copigmentation, and in doing so they sterically block hydration at C2 — the reaction that would otherwise drain the coloured forms into the colourless hemiketal and then irreversibly into yellow trans-chalcone [6].

Two observable consequences follow. The first is a bathochromic shift: the absorption maximum of the butterfly pea extract sits at 548 nm against roughly 525 nm for anthocyanin extracts from purple sweet potato, red cabbage, grape skin and eggplant peel, a displacement of about 23 nm attributable to polyacylation [1]. The second is durability. Compared side by side in the dark at ambient temperature, absorbance decay at eight weeks was 10.6% for butterfly pea against 68.2% for eggplant peel, with purple sweet potato, red cabbage and grape skin ranged between [1]. Intramolecular stacking is also less dependent on ambient conditions than the alternative stabilisation strategies available to anthocyanins, intermolecular copigmentation and metal chelation, both of which require a partner molecule or ion to remain available in the matrix.

3. Colour as a Function of pH

Across pH 0.5 to 13 the extract passes through magenta, purple, blue, blue-green and finally yellow, and the spectra show why (Table 1). From pH 0.5 to 3 a single peak at 548 nm corresponds to the flavylium cation. Between pH 3 and 4 the peak splits and shifts, giving maxima at 570 nm and 622 nm with a shoulder at 531 nm. From pH 5 to 8 the extract is a clean blue carried by peaks at 576 nm and 622 nm, assigned to the quinonoid bases A and A−. At pH 9 the peaks merge into one near 628 nm, and above pH 8 absorbance in the 380–400 nm region rises sharply as yellow chalcones form [1].

The practical reading is that usable blue-green colour is available across pH 4 to 10, which covers almost the whole range of ordinary food, and that the cleanest blue sits between pH 5 and 8. The upper boundary matters more than the lower one, because chalcone formation passes through an irreversible cis-to-trans isomerisation: an alkaline excursion does not merely shift the colour, it destroys it permanently.

4. Thermal Stability and the Inverted pH Rule

Degradation of the extract follows first-order kinetics, which allows stability to be expressed as a half-life and compared directly across conditions (Table 2). The result is counterintuitive for anyone whose experience is with red anthocyanins. At 4 °C the half-life of the blue A form at pH 7, measured at 576 nm, is 334 days, while the flavylium cation at pH 0.5 manages 51 days and the alkaline A− form at pH 10 only 43 days [1]. The ordering holds at every temperature tested: pH 7 at 576 nm is the most stable species, followed by pH 7 at 622 nm, then pH 0.5, then pH 10.

The received rule that anthocyanins should be formulated into acid therefore does not transfer to this pigment. It is the intramolecularly stacked quinonoid base, not the flavylium cation, that is the durable species here, and the stacking is most effective near neutrality. For product development this is the single most consequential finding in the literature, because it means a butterfly pea colour should not be dosed into an acidified beverage as a matter of habit. It also means the colour is compatible with the pH of most consumed food, which is the range where natural blues have historically been least usable.

Temperature remains the dominant variable within any pH. At pH 7 the half-life falls from 334 days at 4 °C to 47 days at 25 °C, 26 days at 37 °C and 18.5 days at 50 °C [1]. Two implications follow for manufacturers in tropical markets. Ambient warehouse storage in Southeast Asia sits close to the 25–37 °C band, so a liquid or high-moisture product will lose much of its colour within a quarter unless the water activity is reduced or the pigment is protected. And accelerated stability testing on this pigment will be misleading if the test temperature is above 50 °C, because degradation at those temperatures is dominated by chalcone formation rather than by the slower pathway operating at ambient.

5. Antioxidant Capacity Does Not Follow the Colour

Butterfly pea is sold on two propositions at once: it is a colour, and it is an antioxidant-bearing botanical. The evidence indicates that these two properties do not hold across the same conditions, and the divergence is large enough to matter commercially.

Radical-scavenging activity measured by the 2,2-diphenyl-1-picrylhydrazyl assay is high at pH 1 and again across pH 4 to 7, where scavenging exceeds 80%, then falls steeply above pH 7; at pH 10 it is below 6% for blue flower extracts [1]. Visible blue-green colour, by contrast, persists to pH 10. A product formulated at pH 8 to 9 may therefore look entirely correct while having lost most of the activity on which an antioxidant positioning rests. Any claim substantiation should accordingly be performed at the pH of the finished product rather than on the extract as supplied.

Two further observations from the same work bear on sourcing. Fresh flowers scavenge more strongly than dried at pH 4 and below, while dried flowers are marginally better near neutrality; since fresh flowers cannot be held for more than a few days, dried material remains the practical commercial form with little loss in the relevant pH window [1]. And white-flowered lines, which contain no ternatins, still scavenge strongly, consistent with reports that their myricetin glycoside content is thirty to seventy times that of blue lines [1,7]. The antioxidant capacity of this plant is therefore not solely an anthocyanin property, which is a caution against using colour intensity as a proxy for activity in incoming goods testing.

6. Stabilisation: Encapsulation Systems and What They Achieve

Where process or storage conditions fall outside the favourable window, the established remedy is encapsulation, and this extract is now well served by published work [2]. The wall materials fall into familiar groups: maltodextrin alone or in combination with gelatin, cassava starch, carrageenan or β-cyclodextrin; gum arabic; and alginate gelled with calcium chloride. The processes span spray drying, ultrasonic spray drying, freeze drying, vacuum oven drying, microwave drying and simple air drying (Table 3).

Several consistent conclusions emerge. Maltodextrin is the workhorse, giving anthocyanin retention above 90% and excellent water solubility, around 99%, which suits the aqueous matrices in which this colour is mostly used. Combining it with gelatin raises retention slightly, to about 93%, and gives the lightest colour but the least saturation, while a cassava starch combination retains less but gives a darker, more vivid shade — a genuine trade-off between how much pigment survives and how the surviving pigment looks. Freeze drying generally outperforms thermal drying on retention, at a cost most ingredient operations will weigh against the gain. Alginate beads, uniquely among the systems reviewed, were shown to improve biological activity under simulated gastrointestinal conditions, which makes them the more interesting option where the ingredient is sold for activity rather than for colour [2,8].

The limitation running through this body of work is that retention is almost always reported at the point of manufacture. Encapsulation efficiency and immediate anthocyanin recovery are well documented; what is scarce is retention through the shelf life of a real product, at the water activity, pH and oxygen exposure the capsule will actually meet. The one variable most often reported, water activity below 0.3 to 0.5, is a proxy for microbiological stability rather than for pigment survival. Until encapsulated extracts are followed through twelve months in finished matrices, the choice of wall system rests on manufacturing yield rather than on delivered shelf life.

7. Application in Food and Supplement Formats

Reported applications divide into three groups by the form in which the plant is used. Dried petals and aqueous extracts have been incorporated into cupcakes, functional beverages and flours, with reductions in starch digestibility and inhibition of pancreatic α-amylase reported for flour and bread systems at inclusion levels of 1–5% [2,11,13]. A beverage formulation combining the flower extract with stevia and lime was shelf stable for 28 days without preservatives and scored well on sensory attributes [12]. In a randomised crossover trial a Clitoria ternatea flower beverage lowered the postprandial glycaemic response and raised plasma antioxidant capacity in healthy subjects [14], and ternatin anthocyanins together with quercetin glycosides from the petals have been shown to suppress lipopolysaccharide-induced inflammation in macrophage models [15]. Spray-dried extract has been used in sponge cakes, where total phenolic and anthocyanin content rose with inclusion, and in pork patties at 0.02–0.16%, where it increased radical-scavenging activity [2].

For supplement manufacture the relevant conclusion is that the spray-dried powder behaves as an ordinary water-soluble botanical powder, which makes it straightforward to carry in sachets, effervescent tablets, pouch beverages and jellies, and that its visual contribution is dose-responsive at low inclusion. The constraint is not the format but the pH and the thermal history of the process. A short high-temperature step is less damaging than prolonged warm storage, since the degradation is first-order in time at a rate set by temperature; an ingredient that survives pasteurisation may still fade in a warehouse.

8. The Regulatory Position, and Why It Turns on a Different Molecule

The technical literature on this plant discusses pigments. The regulatory constraint does not, and the gap deserves attention from anyone planning to sell the ingredient across borders (Table 4).

In the United States, butterfly pea flower extract is listed as a colour additive exempt from certification at 21 CFR 73.69. The original listing followed a petition by Sensient Colors and permitted use in a range of beverages, coffee creamers, frozen dairy, yoghurt fruit preparations, chewing gum, coated nuts and confectionery [16,17]. Permitted uses were widened by a final order of 12 May 2025, effective 26 June 2025 and confirmed on 21 August 2025, to include ready-to-eat cereals, crackers, snack mixes, hard pretzels, plain potato chips and plain corn, tortilla and multigrain chips, in each case at levels consistent with good manufacturing practice [18]. The widening sits within a broader move away from petroleum-derived synthetic dyes, and it extends the ingredient from wet systems into dry, high-temperature, long-shelf-life ones — precisely the applications in which the stability data reviewed above become limiting rather than comfortable.

In the European Union the position is different in kind. A notification of dried Clitoria ternatea flowers as a traditional food from a third country met a safety objection from the European Food Safety Authority in a technical report of 2 February 2022. The objection did not concern the anthocyanins. It concerned cyclotides, the cyclic peptides this species produces, for which in vitro haemolytic and cytotoxic effects have been reported, with possible effects on the immune system and the uterus noted for one cyclotide that was not itself detected in the flowers [19,20,21]. The authority’s reasoning rested on data gaps rather than on demonstrated harm, and the long record of culinary use in Asia was acknowledged, but the effect is that market access in the European Union remains constrained.

The consequence for ingredient specification is direct and, so far, largely unaddressed. A certificate of analysis for a butterfly pea extract will typically report anthocyanin content, moisture, heavy metals and microbiology. None of those measures the cyclotide content, which is the parameter on which European access currently turns. Extraction conditions plausibly affect the partition of peptides into the extract, and an aqueous extract of petals is not the same material as the dried whole flower on which the European assessment was made, but neither proposition has been tested and reported in a form a regulator would accept. Characterising the cyclotide profile of commercial extracts, and establishing whether standard extraction and spray drying deplete it, is the most commercially valuable analytical work outstanding on this plant.

9. Research Gaps

Five gaps follow from the foregoing, ordered by what would most change commercial practice.

First, cyclotide characterisation of commercial extracts, as set out above. Second, stability of encapsulated extract through real shelf life, in finished matrices at defined water activity and oxygen exposure, rather than retention measured at the drier outlet. Third, a reference method for ternatin content. Anthocyanin content is commonly reported as a total by the pH differential method, which cannot distinguish the fifteen ternatins from one another or from the non-acylated anthocyanins that contribute colour but not stability; without a marker assay, two extracts of equal declared anthocyanin content may behave differently in the same product. Fourth, bioactivity after encapsulation: most studies measure antioxidant capacity of the reconstituted powder, few measure it after passage through a digestion model, and the alginate work indicates that the wall system can change the answer [8]. Fifth, the interaction between pigment and matrix: proteins, metal ions and co-pigments in a finished product will shift both colour and stability, and essentially all published stability data are from buffered model solutions.

10. Conclusions

Butterfly pea is the most technically credible natural blue currently available, and the reason is structural: polyacylated ternatins stabilised by intramolecular copigmentation hold the blue quinonoid base where ordinary anthocyanins cannot. The practical consequences are that the pigment should be formulated near neutrality rather than in acid, that temperature rather than pH governs shelf life within the usable window, that alkaline excursions destroy the colour irreversibly, and that antioxidant positioning must be substantiated at finished-product pH because activity and colour part company above pH 7. Encapsulation in maltodextrin-based systems is established and effective at the point of manufacture, with alginate the more interesting option where delivered activity matters. The decisive constraint on international trade in this ingredient, however, is not its colour chemistry at all but its peptide content, and that is where both the analytical and the regulatory effort now belongs.