The mineral that causes much of the Danxia landform's red hue is hematite (Fe₂O₃), an iron oxide. Secondary iron oxides—chiefly hematite, with goethite contributing yellow-brown tones—coat sand grains and cement pore spaces in the red sandstone and conglomerate beds. In the Danxia beds of southeast China, red color comes from these authigenic iron-oxide films formed within continental red beds, not from a single pigment mineral mixed into the rock.
The red is a rust stain at mineral scale. Each grain of quartz or feldspar behaves like a pebble in a slow oven, and over millions of years the iron film baked onto its surface set the color that now draws millions of visitors to Zhangye, Danxiashan, and the Longhushan escarpments. The explanation below follows the color from source rock to cliff face, quantifies how much iron is involved, separates the common myths from the geology, and notes where the science remains genuinely unsettled.
Mechanism: How Iron Films a Sand Grain and Turns It Red
Danxia color begins in the source mountains. Weathering releases iron from mafic minerals such as biotite, hornblende, and pyroxene, and from iron-bearing clays. Rivers carry that dissolved and particulate iron into alluvial fans and lake basins, where it meets oxygen, water, and sediment at the same time. The resulting deposits are what geologists call continental red beds.
The critical step is dehydration. Hydrated ferric oxides such as ferrihydrite and goethite precipitate first. With burial and mild heating, usually above roughly 60 to 100 degrees Celsius over geological time, goethite loses water and converts to hematite. Hematite's crystal structure absorbs blue and green wavelengths and reflects red, so a thin coating, often only micrometers thick, can dominate the color of an entire sandstone bed. The pigment is essentially invisible in bulk but supreme in effect.
Timing matters as much as chemistry. Red pigment can form penecontemporaneously, meaning soon after deposition in oxidizing soils and playas, or it can arrive later as diagenetic iron mobilized by groundwater. The Chinese Danxia deposits are largely Cretaceous continental red beds, roughly 145 to 66 million years old, and the pigment is considered a primary or early-diagenetic feature of those beds rather than a recent weathering rind. A practical way to say it is that the cliffs were painted before they were carved.
Quantified Context: How Much Iron, How Old, How High
The iron fraction needed for full red coloration is small. In red beds worldwide, total iron typically ranges from about 1 to 5 percent by weight of the rock, and the pigment itself may be a fraction of that. Visible redness in Chinese Danxia sandstones commonly corresponds to hematite contents in the range of a few tenths of a percent to about 2 percent by weight. A little hematite goes a long way because each crystal is a strong colorant.
Age and setting give the numbers context. Danxia formations in southeast China are dominated by Cretaceous red beds deposited between roughly 145 and 66 million years ago. Landform development is far younger: uplift and canyon incision in the Quaternary, broadly within the last 2.6 million years, produced the mesas, pinnacles, and sheer walls. Tectonics supplies the vertical relief, and river incision plus salt weathering and wetting-drying cycles strip the softer beds and expose the harder, redder ones.
Two standard numbers recur in the literature and deserve translation. Humidity thresholds for salt weathering are often cited as relative humidity above about 60 percent, and rock tensile strengths for sandstone commonly fall in the low tens of megapascals, roughly 5 to 30 MPa. In plain terms, humid air keeps the salt attack running and modest tensile strength means cliffs fail in slabs rather than crumbling evenly.
Timeline: From Iron Source to Red Cliff
| Stage | Approximate timing | What happens to the iron |
|---|---|---|
| Source weathering | Ongoing before and during deposition | Iron released from mafic minerals and clays |
| Transport and deposition | During red-bed accumulation | Iron delivered to fans and lake basins |
| Early oxidation | Within thousands to millions of years of burial | Ferrihydrite and goethite precipitate as grain coatings |
| Hematite conversion | Deeper burial, often above 60 to 100 degrees Celsius | Goethite dehydrates; hematite forms |
| Diagenetic overprint | Over millions of years | Groundwater redistributes iron; some bands bleach or deepen |
| Uplift | Quaternary, roughly the last 2.6 million years | Red beds lifted above base level |
| Incision and exposure | Quaternary to present | Cliffs cut; fresh hematite-stained surfaces revealed |
| Modern weathering | Present day | Salt, water, and temperature cycles alter surfaces |
The counterintuitive detail sits in the last two rows. The red can fade. Groundwater flow and organic acids reduce ferric iron to ferrous iron, which is soluble and colorless, producing green or gray reduction patches inside otherwise red cliffs. Bleached zones are therefore evidence of fluid pathways, not evidence that the rock was never red.
Edge Cases: Where the Red-Hue Rule Breaks Down
Color is not a reliable index of composition. Very fine hematite can look black rather than red, and coarse specularite glitters in reflected light. Where goethite dominates, the rock reads yellow-brown instead of red. Where organic carbon is high, beds trend dark gray. A single cliff face can carry all of these within a few meters, and a hand lens and a streak test matter more than a photograph.
The best-known edge case is Zhangye National Geopark in Gansu. Its celebrated banding is not a stack of separately painted layers. The stripes come from alternating hematite-rich red beds and reduced, iron-poor beds whose colors were redistributed by groundwater after deposition. In other words, much of the "rainbow" is a diagenetic iron pattern superimposed on the original stratigraphy, which is why bands can cross bedding planes.
A second edge case concerns the name itself. Danxia is a landform term defined by red beds and steep cliffs, not a guarantee of hematite dominance. Some red beds owe their color partly to clay minerals, and some red-looking rocks are volcanic or metamorphic and fall outside the Danxia definition. The professional caveat is that "Danxia" describes a geomorphological class, while the pigment question is a mineralogical one, and the two do not always align perfectly.
Misconceptions: Cinnabar, Copper, and "Red Sandstone"
The most persistent myth is that Danxia cliffs are colored by cinnabar, the mercury sulfide ore that shares the Chinese word for red. Cinnabar is a rare hydrothermal mineral found in veins, not a cementing agent in sandstone. The red beds contain no cinnabar component and the pigment is an iron oxide, not a sulfide. Confusing the two is understandable, because the name and the color coincide, but the chemistry is entirely different.
The second myth is that copper causes the red. Copper minerals such as azurite and malachite are green to blue, and their presence would produce the opposite impression. Iron is what oxidizes to red in the presence of oxygen, and the other major rock-forming elements, silicon and aluminum, are colorless in their common mineral forms.
A third confusion treats "red sandstone" as a single material. In practice, Danxia rocks are sandstone and conglomerate with variable cement, including calcite, silica, and iron oxides, and their red coloration varies with grain size, coating thickness, and weathering state. Transportation matters too: a dull orange road cut and a vivid cliff face can be the same formation with different degrees of exposure. The red is real, but it is also a surface phenomenon.
Comparison Table: Candidate Colorants in Red Beds
| Mineral | Formula | Typical color | Role in Danxia | Relative abundance in red beds |
|---|---|---|---|---|
| Hematite | Fe₂O₃ | Red to reddish brown | Principal pigment | Trace to about 2 percent |
| Goethite | FeO(OH) | Yellow to brown | Secondary pigment | Variable, often subordinate |
| Ferrihydrite | Fe₅O₈H·4H₂O (approximate) | Reddish brown | Precursor to hematite | Minor, transitional |
| Cinnabar | HgS | Scarlet | None; hydrothermal veins only | Rare, unrelated |
| Malachite | Cu₂CO₃(OH)₂ | Green | None; would shift color away from red | Rare |
| Organic carbon | Carbon-rich matter | Gray to black | Masks red pigment | Variable |
The table makes the hierarchy clear. Iron oxides supply the color, hematite dominates among them, and everything else either modifies or masks the signal. If a sample lacks iron oxide coatings, it should not read red, which is the simplest field test to apply.
Practical Actions: Reading the Red for Yourself
The following steps work for a geology student, a park interpreter, or a curious traveler and require little equipment.
Use a streak test. Scrape a small sample on unglazed porcelain; hematite gives a red-brown streak, and goethite gives a yellow-brown one. This distinguishes the two iron oxides directly. Check for reduction patches. Green or gray bleached zones mark where groundwater removed iron, so their presence confirms diagenetic overprinting. Note the relationship between bands and bedding. Bands that cut across bedding planes indicate groundwater redistribution rather than original deposition. Compare fresh and weathered faces. A road cut and a cliff face of the same rock often differ in saturation because the pigment is a surface coating. Read the regional geology. Cretaceous continental red beds in southeast China are the expected host for Danxia, while marine or volcanic sequences point elsewhere. Photograph the same outcrop in dry and wet conditions. Wet surfaces darken and can change how fine hematite appears, which is useful for correcting field impressions.
Park visitors and collectors should follow local rules on sample removal. The color is worth recording, not removing. For educators, the test sequence above gives students a defensible way to argue from mineral evidence rather than from appearance alone.
Limitations: What the Science Cannot Settle from Color Alone
Several uncertainties remain and deserve honest statement. The exact timing of hematite formation in any given Chinese Danxia section is difficult to pin down without radiometric and paleomagnetic work, and researchers often describe the pigment as early diagenetic without a precise age. The relative contributions of depositional and diagenetic iron vary from section to section, so broad statements about "how Danxia got red" are averages, not universal rules.
Quantification also carries uncertainty. Bulk iron content measured by chemical analysis includes iron in silicates that contributes nothing to color, so hematite-specific estimates rely on partial extraction methods and reflectance spectroscopy rather than simple totals. Saturation can change with moisture, grain size, and exposure, meaning two measurements from the same outcrop may differ substantially.
Finally, the numbers cited here reflect general ranges reported in the literature, not site-specific measurements for every park. Field conditions, local stratigraphy, and sampling design all shift the figures, and readers should treat them as context rather than constants. The stable, defensible conclusion is that hematite is the dominant red pigment, that goethite supplies the yellow-brown variation, and that the deposits themselves were red before erosion carved the cliffs. Everything finer depends on where you stand, what you sample, and how long the groundwater had access to the iron.