Pale Tan at the Coordination Complex I Trusted

The fabric was perfect. Twelve centimetres of plain weave, cut off Diane’s loom this morning, edges folded and pressed, ready to accept dye. Cotton, not wool—cellulose fibres, which meant tannin first, then alum. I’d read the mordanting protocols. Protein fibres (wool, silk) accept aluminum salts directly because their amphoteric chemistry lets them grab both acids and bases. Cellulose doesn’t have that capacity. Cotton sits there inert unless you trick it with a tannin layer first.

Steeped black walnut hulls from the backyard for three hours—cheap tannin source, and the hull fragments were sitting in a jar after last autumn’s collecting. Strained the brown liquor, simmered the fabric in it for ninety minutes, hung it to dry overnight. This morning it looked faintly beige. Tannin-impregnated. Ready for the alum bath.

Potassium aluminum sulfate, thirty grams per litre, heated to 80°C. Fabric goes in, gets held at temperature for an hour, comes out ready to bond with natural dyes through a coordination complex—metal ions forming molecular bridges between dye and fibre. That’s the theory. Mordant comes from Latin mordere, “to bite.” The ions bite into both the dye molecule and the fabric structure, holding them together through washing, light exposure, time.

Except when they don’t.

The Block Carved Fine

Spent yesterday afternoon carving a test block from a scrap of maple—close-grained hardwood, the kind that holds detail. Simple geometric pattern: three concentric diamonds, 4cm across, carved in relief. The background needs to sit 4-5mm below the printing surface so it doesn’t pick up dye and ghost onto the fabric. Used a #7 gouge for the open areas, a V-tool for the lines between diamonds.

Printing block took two hours to carve. Registration marks in opposite corners—small notches that would align with brass pins through the fabric into a padded table. Multi-colour printing requires those marks. If your second colour is misregistered by 3mm, the pattern visibly breaks. Indian chhapa printers get single-colour patterns aligned within a millimetre across metres of fabric. They’ve been doing this since before Alexander invaded in 327 BCE. I figured one test block, one colour, twelve centimetres of fabric—I could manage that.

What Black Walnut Hull Produces (Under Certain Conditions)

Black walnut is supposed to give browns and dark greys. Rich earth tones. That’s what every natural dye reference says. What they don’t emphasize is that walnut’s colour depends entirely on iron content. Walnut tannin itself is brownish-yellow. Add iron—from an iron mordant, from iron-rich water, from a cast iron pot—and you get dark brown trending toward black. Without iron, you get… beige.

I used an aluminum mordant. No iron. The fabric had accepted tannin, accepted alum, and was now primed to accept walnut dye in whatever colour walnut dye actually is when there’s no ferrous sulfate involved. I extracted fresh dye from more hulls, heated it to 60°C, loaded the carved block with dye using a foam brush (even coating, not too wet, not too dry), and pressed it onto the mordanted fabric.

Pressure matters. Too light and coverage is patchy. Too heavy and dye bleeds under the block edges. I pressed until I felt the block compress the felt padding beneath the fabric by about 4mm—same tactile feedback that matters in beating weft tight during weaving, the body’s ability to encode pressure as memory.

Lifted the block. Pale tan diamonds, vaguely visible, not at all the rich brown I’d expected. Maybe it needed to oxidize. Cyanotype prints look terrible wet and deepen as they dry. Maybe walnut did the same thing.

It didn’t.

Three hours later: still pale tan. Held it up to the light. The dye had transferred, the pattern was clean, the registration was irrelevant because I only had one colour and nothing to register it against. But the whole piece looked like I’d printed with extremely dilute tea.

The Chemistry I’d Assumed Would Work

Here’s what I got wrong: tannin mordanting is a substrate prep, not a colour intensifier. The tannin layer makes cellulose behave more like protein fibre—gives it binding sites for the aluminum. The aluminum creates the coordination complex with the dye. But the amount of dye that complex can hold depends on dye concentration, contact time, heat, and the specific molecular structure of whatever compound you’re trying to bind.

Black walnut hulls contain juglone, a naphthoquinone compound that produces colour when oxidized in the presence of metal ions. Iron works reliably. Aluminum… less so. The aluminum-juglone complex is apparently much lighter than the iron-juglone complex. I could’ve researched this more thoroughly. I didn’t. I assumed “natural dye” was fungible and mordants were interchangeable.

They aren’t. Every dye-mordant combination produces a different colour, a different fastness, a different affinity for the fibre. Cochineal with alum gives pink. Cochineal with tin gives brilliant orange-red. Same insect-derived dye, completely different output based on which metal ion mediates the bond. The chemistry is specific. You can’t eyeball it and hope the coordination complex works out in your favour.

Terrazzo taught me that chemical incompatibility isn’t obvious from visual inspection—calcite crystals embedded in cement crack from alkaline reactions at the boundary. This is adjacent. The fabric looks ready. The dye looks concentrated. The mordant has done its molecular prep work. But the complex that forms when walnut tannin meets aluminum sulfate on cellulose is pale tan, and no amount of wishing makes it darker.

I have a carved block, a mordanted fabric sample, and a colour I didn’t want. The pattern printed cleanly—sharp edges, no bleed, good detail transfer. But “technically successful” and “looks right” are different outcomes. Right now I’ve got the former and none of the latter.

Could try iron mordant, accept the dye-rot risk, see if ferrous sulfate gives me the dark brown I actually wanted. Could switch to madder root—reliable reds, good fastness, works with aluminum. Could abandon natural dyes entirely and admit that understanding mordant coordination chemistry from reading Wikipedia is insufficient preparation for a 9000-year-old craft tradition.

The block’s on the table. The fabric’s hanging on a clothesline, pale tan diamonds drying in irrelevant detail. Diane’s loom still has the failed warp stretched across it, and now this. Not everything synthesizes cleanly, even when the skills transfer and the theory makes sense.