Halo-Halo (Filipino Shaved Ice Dessert) Spilled on Carpet: Removing Ube Anthocyanin, Evaporated Milk, Leche Flan, Fruit Syrups, and Ice Cream

The halo-halo had a very large scoop of ube ice cream on top — specifically a scoop of Magnolia Ube ice cream, which is a well-known Philippine brand with a very vivid, intense artificial purple colour from food colouring rather than natural ube. Magnolia Ube ice cream is bright electric-purple, not the more muted purple-grey of natural ube. The ice cream scoop fell first and hit the carpet before the rest of the halo-halo fell. How does artificial-dye ube ice cream (Magnolia-type) differ from natural ube anthocyanin on carpet?

Artificial-dye ube ice cream (Magnolia-type, electric purple) versus natural ube anthocyanin on carpet: (1) Magnolia Ube ice cream uses artificial food colouring — Philippine commercial ube ice cream brands including the Magnolia Ube variant typically use artificial purple dye (Red 40 + Blue 1 blend, or a pre-mixed artificial purple food colour) to achieve the vivid electric-purple colour, because natural ube anthocyanin is unstable in the freezing and storage conditions of commercial ice cream production. The electric-purple of Magnolia Ube is distinctively more vivid and saturated than the more muted purple-grey of natural ube halaya. (2) Artificial azo dye (Red 40) + triphenylmethane dye (Blue 1) — the combined artificial purple is more resistant to hydrogen peroxide than natural ube anthocyanin. Three to four rounds of 15 minutes each at 3% hydrogen peroxide are the starting protocol for Magnolia Ube ice cream on light or cream carpet. Natural ube anthocyanin typically resolves in two rounds. (3) Ice cream dairy fat — ube ice cream contains a standard ice cream dairy fat base (approximately 10–12% fat from cream and milk). The baking soda step absorbs this dairy fat. The ice cream also contains egg yolk in most premium formulations (for richness and emulsification) — cold water rule applies. (4) Cold, melting ice cream spreads — ice cream at room temperature is semi-fluid. The ube ice cream scoop will spread outward on the carpet as it melts. Apply baking soda promptly before the ice cream fully melts and wicks outward. (5) Multiple peroxide rounds required — for Magnolia Ube or other artificially dyed ube ice cream, budget for three to four hydrogen peroxide rounds on light carpet and accept that a faint purple tint may remain after four rounds if the dye contact time was extended before treatment began.

The halo-halo was served in a large glass with a significant volume of condensed milk poured into the base (some Filipino families add a full layer of condensed milk to the bottom of the glass before the ingredients, then the shaved ice, then the evaporated milk drizzle on top — creating a very sweet, thick, almost caramel-like condensed milk syrup at the base). When the glass fell, the condensed milk at the base poured out as a thick viscous stream and spread over a larger area than the shaved ice melt. How does a condensed-milk-heavy halo-halo differ from a standard evaporated-milk version?

Condensed-milk-heavy halo-halo (full layer at the base) versus standard evaporated-milk-drizzle halo-halo on carpet: (1) Condensed milk is much higher sugar and fat than evaporated milk — condensed milk (sweetened condensed milk) contains approximately 55–60% sugar and 9% fat per 100g, compared to evaporated milk's approximately 10% sugar and 8% fat. A full condensed milk base layer in a halo-halo glass adds a very large dissolved sugar load and a significant fat load to the combined spill. (2) Viscosity difference — condensed milk is thick and viscous (it flows slowly). On carpet, this means it stays more concentrated in the impact zone rather than spreading as widely as the thin shaved ice melt water. The condensed milk contact zone will have a more concentrated fat and sugar residue than the shaved ice melt water perimeter. Three baking soda rounds for the main condensed milk contact zone; four cold water rinse cycles for the sugar load. (3) Condensed milk dairy casein protein — condensed milk is made from whole milk (dairy protein present). Cold water only, plus protease enzyme for the casein. (4) Combined sugar from condensed milk + fruit syrups + shaved ice — the total dissolved sugar load from a condensed-milk-heavy halo-halo is exceptionally high — likely the highest sugar load of any typical Filipino dessert. Five cold water rinse cycles across the full spread zone are warranted to prevent sugar-based re-soiling. (5) The ube colour treatment is unchanged — the hydrogen peroxide protocol for the ube anthocyanin from ube halaya and ube ice cream is the same regardless of the condensed milk base.

My halo-halo included a Filipino halo-halo ingredient called "nata de coco" — these are small, firm, translucent white or slightly yellow gel cubes made from bacterial cellulose produced by Komagataeibacter xylinus bacteria fermenting coconut water. Nata de coco is sold in jars or cans preserved in a heavy sugar syrup. The nata de coco pieces are themselves quite firm and rubbery. When the halo-halo fell, the nata de coco cubes scattered across the carpet and some of the sugar syrup from the jar soaked into the pile. How are nata de coco pieces and syrup treated on carpet?

Nata de coco pieces + sugar syrup on carpet from halo-halo spill: (1) Nata de coco cellulose is not starch — nata de coco is composed of bacterial cellulose (a polysaccharide produced by bacteria from coconut water sugars). It is NOT the same as starch — amylase enzyme does not break down cellulose. The nata de coco cubes themselves are essentially pure cellulose and do not create a staining problem on carpet. They are solid and can be picked up individually without breaking. Do not apply amylase enzyme to address nata de coco — it has no effect on bacterial cellulose. (2) The nata de coco syrup IS the staining concern — nata de coco is preserved in a thick, very sweet sugar syrup. This syrup is the actual carpet issue: highly concentrated dissolved sugar that wicks into the carpet pile and creates a re-soiling residue. Four cold water rinse cycles over the syrup contact zone are needed to extract the dissolved sugar. (3) Nata de coco has a distinctive faint yellowish-white colour that is barely visible on most carpets — the nata de coco itself (cellulose gel) has essentially no carpet-staining pigment. Any tinting in the nata de coco syrup (which can be slightly yellow or coloured in some commercial versions) is addressed by the cold water rinse alone — no hydrogen peroxide needed for the nata de coco component. (4) Embedded nata de coco cubes in carpet pile — nata de coco cubes are firm and do not compress into the pile easily. Check the pile for scattered cubes and pick them up before applying liquid treatments (wet treatment would push them deeper into the pile). Vacuum after picking up all visible pieces. (5) The dominant stain from halo-halo is still the ube anthocyanin — the nata de coco component is relatively minor in the overall halo-halo spill treatment. Prioritise the ube colour and the evaporated milk protein after collecting the nata de coco cubes.

The halo-halo spill happened on a very hot day and the shaved ice melted almost instantly — there was essentially no shaved ice left, just a large puddle of warm (not cold) sweet liquid by the time we could address it. The warm temperature of the spill concerns me because I know heat can set some food stains. Does the absence of ice (warm liquid halo-halo) create a worse carpet stain than a cold, ice-still-present spill?

Warm liquid halo-halo (no remaining ice, warm ambient temperature) versus cold halo-halo (ice still present): (1) The main risk of warmth is for egg protein — the leche flan in halo-halo contains cooked egg yolk, and the evaporated milk contains dairy protein. For a warm spill (above approximately 40–50°C), egg protein heat-setting begins. At warm ambient temperatures (25–35°C on a hot day in Sutherland Shire), the temperature of the halo-halo liquid after all ice has melted is not high enough to cause immediate heat-setting (egg white sets above approximately 60–65°C, and even at 40°C the setting process is slow). However, if the warm liquid sits in the carpet and the room temperature is high (above 35°C), extended contact could gradually degrade the egg protein's extractability. The priority is to treat promptly. (2) No shaved ice = smaller liquid volume — if the ice has fully melted before treatment, the total liquid volume is much smaller (the melted ice water has already evaporated or been absorbed by the carpet over time if significant delay occurred). A very delayed treatment means some components have partially dried — this is not ideal but the treatment steps remain the same; simply extend the enzyme dwell times to allow more time for re-hydration and digestion of the partially dried residues. (3) Warm temperature slightly improves enzyme effectiveness — enzyme cleaners work faster at 25–35°C than at near-freezing temperatures. A warm-ambient halo-halo treatment may require slightly shorter enzyme dwell times than the standard protocol for a very cold spill. (4) Ube anthocyanin does not heat-set — the ube purple colour from anthocyanins is not affected by ambient temperature in the same way as protein. Hydrogen peroxide response is unchanged regardless of whether the spill was cold or warm. (5) Re-soiling risk is higher for a delayed warm treatment — if the halo-halo liquid sat in warm ambient conditions for 30+ minutes before treatment, the dissolved sugars will have partially dried into a sticky film in the carpet pile. Extra cold water rinse cycles (five to six) are needed after a delayed warm halo-halo spill compared to the standard four cycles for an immediate spill.