Short answer
On paper the colour and its support age as one system. Iron gall ink browns and weakens the sheet around each line; verdigris turns brown and embrittles the paper beneath it; lead white highlights blacken in sulfurous air; and fugitive watercolours, plant dyes and early synthetic inks fade at light doses that oil paint would shrug off. With no varnish and very little binder to shield the colourant, museums ration exposure instead — Tate allows paper-based works at most two years on display in every four.
Iron gall ink, made from oak-gall tannins and iron sulfate, was the ordinary writing and drawing ink of Europe from late antiquity to the nineteenth century. Some recipes are stable; others carry excess acid and unbound iron. Gerhard Banik's account for the Iron Gall Ink Website describes the damage in stages: first a fluorescence around the line under ultraviolet, then a brown discolouration of the paper beside it, which spreads through the sheet and prints itself onto the facing page, and finally loss as the line drops out of the sheet. Two processes overlap — acid hydrolysis of cellulose, and oxidation of cellulose catalysed by soluble iron — and both are strongly influenced by temperature and humidity. In a chamber test, iron inks were also among the materials most changed by nitrogen dioxide.
Verdigris, a copper acetate, was the usual bright green of manuscripts, early printed books and hand-coloured maps. A study in the journal Restaurator notes that, depending on how it was applied, the paper and its sizing, and on light and humidity, it can discolour to brown and chemically damage the paper carrier, so that green areas become brittle and break out. Lead white, used for highlights in drawings and for opaque passages in gouache and illumination, has a different enemy: reduced-sulfur gases, which convert it to dark products so that the lightest touches in a drawing become its darkest. Chemical conversion of the darkened pigment back to white has been practised since the nineteenth century, and a review in Studies in Conservation traces how its use has been reassessed against changing ethics and its effects on paper.
The colours most often lost from works on paper are organic. Carmine and other cochineal lakes fade in light, and one study found that which form fades fastest depends on the metal the dye is bound to and on whether the paper is acidic, neutral or alkaline — the support is part of the chemistry. Crystal violet, one of the first synthetic dyes and a common ink ingredient, loses its methyl groups one by one under ultraviolet, shifting hue as it fades. The Van Gogh Museum's REVIGO project reports that the artist's pen-and-ink drawings originally showed greater colour and contrast than they do now. Indigo and turmeric react with ozone in the dark. In each case the result is a drawing that reads as browner, flatter and more monochrome than it was.
Sealing a work in a low-oxygen frame slows most fading, and Tate has developed such frames for works on paper. Prussian blue is the awkward exception. It fades by reduction rather than oxidation, and oxygen is what turns the pale reduced form back to blue, so without oxygen it fades faster. Tate's microfading study found that no colourant fared worse in anoxia than in air except Prussian blue and Prussian green, and that a Prussian blue from J.M.W. Turner's own studio behaved like fifty-five other samples. At five per cent oxygen it was as stable as in air, so the researchers proposed five per cent rather than zero for 'anoxic' enclosures. The paper matters here too: X-ray studies show the substrate altering the pigment's structure on contact and changing how it fades.
A painting in oil has a binder around each pigment particle and often a varnish above; a watercolour is a thin wash of colourant in a little gum on absorbent paper, exposed to light and air on both counts. Lowering the lux helps, but damage follows total dose, so museums budget hours as well. Tate states that paper-based works are allowed a maximum of two years' display in controlled conditions in every four-year period, and the Getty's lighting research frames the task as managing an unavoidable, irreversible risk rather than eliminating it. Microfading now lets an institution test the actual sheet before deciding. For an owner, the reading is simple: the faded state of an old watercolour or letter is permanent, and the rate of further change is set by dose, air and damp together.
| Colourant | What changed | Driver | Evidence and caveat |
|---|---|---|---|
| Iron gall ink | The paper browns around the ink line; the brown spreads through the sheet and offsets onto facing pages; the line may finally drop out. | Acidity or alkalinity; Humidity and water; Heat; Gaseous pollutants | Moderate. Ink recipes varied; some formulations are stable. |
| Verdigris (copper acetate) in gum on paper | Green turns brown, and the paper under it becomes brittle. | Light and UV; Humidity and water; Adjacent materials (binder, support, varnish) | Moderate. Model samples; maps, manuscripts and early printed books are named as affected types. |
| Lead white in watercolour, gouache and drawing highlights | White highlights turn grey, brown or black. | Gaseous pollutants | Moderate. The pollutant statement rests on a summary of the CCI page, not its full text. |
| Prussian blue and Prussian green watercolour | Blue fades; partial return of colour in air in the dark. | Light and UV; Low oxygen (anoxia); Adjacent materials (binder, support, varnish) | Strong. Microfading and chamber tests, not long-term display. |
| Carmine and cochineal lakes in gum arabic | Crimson fades to pale pink or nothing. | Light and UV; Acidity or alkalinity | Moderate. Accelerated ageing; the authors call the mechanism poorly understood. |
| Crystal violet and related early synthetic (triarylmethane) dyes in inks | Violet ink shifts in hue and fades. | Light and UV | Moderate. The laboratory study and the museum finding are separate; Colourwise has not read a source tying this dye to those drawings. |
| Indigo, turmeric and orpiment on paper | Blues and greens containing indigo fade; orpiment loses its yellow. | Gaseous pollutants | Strong. Chamber concentration well above ordinary indoor air; a dose-equivalent test. |
| Natural organic colourants, orpiment and realgar on paper | Ten natural organic colourants changed by more than ΔE 2; the arsenic sulfides by more than ΔE 5. | Gaseous pollutants | Strong. 1980s urban pollution levels; current indoor concentrations differ. |
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
FactModerate evidence
Iron gall ink damage proceeds from fluorescence around the ink line to brown discolouration of the support, which spreads and offsets onto neighbouring pages, and finally to loss; it is attributed to acid hydrolysis and to iron-catalysed oxidation of cellulose, strongly influenced by temperature and humidity.
Caveat: The staged description dates from 1998; the two-process account remains standard.
Source: Ink corrosion — Chemistry (The Iron Gall Ink Website); BPG Iron Gall Ink (AIC Conservation Wiki, Book and Paper Group)
FactModerate evidence
Verdigris on paper can discolour from green to brown and chemically damage the paper, depending on mode of application, paper quality, size, light and humidity.
Caveat: From the study's abstract.
Source: Studies on the Conservation of Verdigris on Paper (Restaurator, 2015)
FactStrong evidence
In Tate's low-oxygen microfading study no colourant fared worse in anoxia than in air except Prussian blue and Prussian green; at 5% oxygen Prussian blue showed the same light stability as in air.
Caveat: Microfading of prepared watercolour samples, including one from Turner's studio.
FactModerate evidence
The relative fading rate of cochineal-based watercolour pigments depends on the metal complex and on the pH of the paper: on alkaline paper the aluminium complex was least stable, on acidic paper the tin complex.
Caveat: Accelerated ageing; mechanism described by the authors as not well understood.
ConventionStrong evidence
Tate allows paper-based works of art a maximum of two years' display in controlled conditions in every four-year period.
Caveat: One institution's policy as stated on its project page; others set different budgets.
Source: Anoxic Display and Storage of Paper-Based Works of Art (project page)
Reviewed 6 October 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.