Coloured Concrete & Roof Tiles: Integral Colour You Cannot Repaint

Dosage guide for concrete colouring

Why roof tiles are the hardest job in coloured concrete

A roof tile carries a pigment load no other concrete product does. It sits on a 30–45° pitch facing the sun at the worst possible angle, cycles from frost to a 60–70 °C surface in a single day, takes acid rain and lichen, and — unlike a paving slab — nobody ever replaces it quietly. European concrete tile makers write 30-year colour warranties on products like that, and they test them in Arizona and Florida because those climates compress decades of UV into years. The warranty is not decorative language: it is a replacement liability that the tile plant carries on its own balance sheet for three decades. That is exactly why the pigment in a roof tile is never an organic green. A tile is cured in steam, not fired — the pigment never gets the kiln as a second chance — so the colour it has at day 28 is the colour it must still have at year 28, through roughly 40,000 hours of direct sun.

Chromium oxide green is the only green with the credentials for that sentence. Lightfastness 8/8 and weatherfastness 5/5 are the two ratings that map directly onto a roof-tile warranty: the first says the colour survives sunlight, the second says it survives sunlight plus rain, frost, heat and pollution together. Both come from the same fact — the colour of Cr₂O₃ is not a dye sitting in the crystal but the crystal itself. Green light comes from d-electron transitions inside the corundum lattice, and a lattice cannot be bleached the way a chromophore can. Fired clay tiles make the point from the other direction: their glaze kilns run at 1000–1100 °C, a temperature at which every organic pigment is ash. Chromium oxide survives that environment and the steam-cure environment of concrete tiles with equal indifference, which is why one pigment standard can cover both products on the same production site. Tile makers who run clay and concrete lines side by side buy one green for both — that is a purchasing simplification worth real money before any colour benefit is counted.

The failure mode to watch is not fading at all; it is patchiness. A roof viewed from the street shows every tile, so a pigment that disperses unevenly produces a mottled, cheap-looking roof even when its average colour is perfect. Agglomerates of pigment survive the mixer, break unevenly in different tiles, and leave streaks that read as dirt. The fix is particle size control — D50 of 1–2 μm in standard grades, ≤0.5 μm in micronized grades — plus the mixer discipline of pre-blending pigment into the dry mix before water enters the pan. A roof is the largest single-colour surface most builders will ever sign for. The pigment specification decides whether it reads as designed or as stained, and that decision is made in grams per batch, not in the brochure.

The alkali war: what lime does to the wrong pigment

Fresh concrete is a saturated calcium-hydroxide environment, pH 12.5–13, and that alkalinity does not leave. Pore water stays saturated with lime for years, and the hydration of cement releases more hydroxide every time moisture moves through the matrix. The pigment lives inside that chemistry for the life of the structure. Organic pigments do not — the ester and amide bonds in their molecules hydrolyse in hot alkali, which is the quiet mechanism behind most "faded" coloured concrete. A product that passes a lightfastness test in a laboratory film can still fail on a real pour because the light test never exposed it to lime. Metal-complex organic pigments are no safer: at high pH the complex ligands exchange, and the green walks off the particle as a leached, colourless residue. The result is the classic three-year concrete failure — colour fine at handover, washed-out grey by the first warranty inspection.

Cr₂O₃ is amphoteric and effectively insoluble in water, so it sits in the hydrated cement matrix without dissolving, without reacting, and without releasing anything into the pore solution. The same property covers the other side of the chemistry: carbonation. As the surface of the concrete reacts with atmospheric CO₂, the pH at the surface falls toward 8 over the years. That front moves millimetres deep into the structure, and a pigment must hold its colour on both sides of it. Chromium oxide does, because its colour is structural rather than pH-dependent. The practical evidence is everywhere in southern Europe: green concrete roof tiles and pavers thirty years old, still reading green while the organic-coloured elements on the same site are grey. The pigment that survives both the alkaline years and the carbonated years is the only one that can be specified without an engineer's caveat.

Efflorescence is the second alkali question, and it is worth separating from pigment choice because the two get confused on site. Lime bloom — calcium hydroxide migrating to the surface and carbonating to a white veil of CaCO₃ — happens on any concrete, green or grey. A dark green surface shows bloom more visibly than a light one, which is why green paving gets blamed for chemistry that belongs to curing. The honest specification advice is this: manage efflorescence with curing practice and mix design, and pick a pigment that does not add its own contribution to the problem. An organic green that hydrolyses in lime adds its own grey wash on top of the bloom; chromium oxide adds nothing. The green you see at year ten is the green that was poured at day one, minus whatever the concrete itself did — and that subtraction is under the mix designer's control, not the pigment's.

Dosing green concrete without streaking

Streaking is a dispersion failure, and it is the most common complaint we hear from first-time concrete colour users. Pigment agglomerates survive the short residence time of a standard mixer, break unevenly from batch to batch, and print faint darker lines through the pour that no trowel can remove. The two-lever fix is well proven. First, particle size: standard grades at D50 1–2 μm disperse quickly in normal mixer cycles, and micronized grades at ≤0.5 μm are there for thin architectural elements and high-tint applications where every kilogram must deliver maximum colour. Second, sequence: the pigment goes into the dry mix — blended with cement and aggregates for two to three minutes before water is added. Pigment thrown into a wet mix is the single most reliable way to buy streaking, because wet pigment clumps, and a clump of chromium oxide is a permanent shadow in the slab.

Dosage is calculated on cement weight, not on total mix weight, and the ladder is short: 1–2 % for pastel tones, 2–4 % for standard architectural greens, 3–6 % for deep roof-tile and paver tones. The cost arithmetic matters more than the percentage. A typical mix carries about 350 kg of cement per cubic metre, so a 3 % dosage means roughly 10.5 kg of pigment per cubic metre — about 0.6 kg per square metre of 60 mm paving. At current export pricing for concrete-grade PG17, that is roughly US$1.5–3 per square metre of finished paver, against a retail paver price that commonly runs US$20–40. The colour costs under a tenth of the product and carries the entire sale, because a coloured paver sells at the coloured price or it does not sell at all. Every conversation about "expensive pigment" should be run through that ratio once.

Two production details decide whether the pour matches the sample. Water content changes the shade of any coloured concrete — the same pigment loading reads lighter at higher water–cement ratios, so the mix designer who controls water controls colour. Batch weighing must be by mass per batch, never by scoop volume, and the GREEN 1000/1001/1002/1020 range exists precisely so a paving plant can hold one pigment across its product line and still hit lot-to-lot ΔE below 1.0. Finally, judge colour at 28 days, not at 7: hydration continues, the surface lightens as it cures, and the contractor who signs off colour on wet concrete is signing a guess. Concrete colour is a one-shot decision — the discipline is in the mixing, the dosing and the waiting, not in the chemistry, because the chemistry was solved decades ago by a mineral that cannot fade.

Concrete-grade PG17 — the numbers that go into the spec

FAQ

Can faded coloured concrete be repainted? No. Integral colour runs through the full depth of the element, but there is no way to re-colour it once it is down — surface coatings on paving and tiles peel under traffic and weather. That is why the pigment decision is one-shot, and why a mineral pigment that cannot fade is the only economically sane choice.

How much pigment goes into a cubic metre of concrete? Dosage runs 2–6 % on cement weight. For a typical 350 kg/m³ mix that is 7–21 kg of pigment per cubic metre — around 0.6 kg per square metre of 60 mm paving at a 3 % dosage. Deeper tones need the top of the ladder; pastel tones need the bottom.

Does the pigment weaken the concrete? No. At these dosages chromium oxide acts as an inert fine filler; it does not react with cement hydration and has no measurable effect on compressive strength. Oil absorption of 20–25 g/100 g keeps its water demand predictable, so slump stays where the mix designer put it.

Will green concrete show efflorescence more than other colours? Lime bloom happens on all concrete — the white carbonate veil is simply more visible against dark green. Manage it with curing practice and mix design, and use a pigment that does not add its own hydrolysis residue on top. Chromium oxide contributes nothing to the bloom; organics contribute a grey wash.

Is the Cr(VI) content of the pigment a compliance issue? It should be checked, not assumed. Standard DELIKECOLORS concrete grades carry Cr(VI) ≤10 ppm; micronized grades are ND (not detected). Ask for the number on the certificate of analysis — construction additives fall under REACH and similar regimes, and a supplier that prints the data is one that controls it.