Dog Collar Hardware: Brass, Stainless or Zamak on Leather

Pet brands spend weeks choosing the leather and an afternoon choosing the buckle. On a collar, the buckle decides which tannage you are allowed to order, and it is the part that fails first.

A guide by TL San Martín, a tannery of origin in Elda (Alicante, Spain) since 1995, LWG Gold certified under the Leather Manufacturer standard.

This guide is for brands that design, make or commission leather dog accessories, not for dog owners. We are a tannery in Elda (Alicante, Spain) and we do not sell hardware. We do see every week what hardware does to leather, and that is the part almost nobody writes down. The leather spec itself (thickness, tannage, edge) is in the hub, leather for dog collars.

Why it matters commercially: the commodity end of pet accessories is under price pressure. In Germany, the largest pet-accessory market in Europe, supplies and accessories fell 4.6 % in bricks-and-mortar retail in 2025 (ZZF/IVH) while food held flat. Brands escape that by moving up: better material, a documented origin, test reports. A buckle that turns green after two summers undoes all three.

Why does the wrong hardware ruin the leather?

Vegetable-tanned leather contains tannins, and tannins react with iron. The product is iron tannate, a deep blue-black pigment and the historical basis of iron gall ink. It forms wherever moisture, iron and leather meet: under the rivet, around the buckle tongue, at the D-ring. The stain is not on the surface. It is chemically bound into the fibre, so it cannot be wiped off or dyed over.

In practice:

• Veg-tan does not tolerate hardware that contains iron. That includes zinc-plated steel, nickel-plated steel and zamak with a steel core. Once the coating is breached, iron is in contact with tannin.
• Chrome-tanned and chrome-free leather are far less sensitive, because they carry much less free tannin. If your design needs a steel component, that is an argument for chrome-free over veg-tan. Both systems are compared in chrome vs chrome-free tanning.
• Saliva speeds everything up. Dog saliva is mildly acidic and attacks thin plating. On a collar the hardware sits in the wet zone between coat and leather.

So the decision chain runs the other way round from how most brands work: fix the hardware first, then choose the tannage.

Brass, stainless or zamak: what actually lasts?

Two side effects that turn up in complaints. Solid brass can leave a greyish rub-off on very pale coats as it oxidises. That is not a defect, but it needs to be in the product description. And brass and stainless together on a permanently wet part form a galvanic couple, so do not mix them at the same rivet on a leash meant for water.

Which EU limits apply to the hardware and to the leather?

Two separate entries in Annex XVII of REACH (EC) No 1907/2006, landing on the same product:

• Entry 27, nickel. Articles in prolonged skin contact may release at most 0.5 µg nickel/cm² per week (0.2 µg for piercing posts). Test method EN 1811:2023, which simulates a week of artificial sweat.
• Entry 47, chromium VI. Leather parts in contact with the skin must contain less than 3 mg/kg Cr(VI) on dry weight, in force since 1 May 2015. Tested to EN ISO 17075-1 (colorimetric) or -2 (HPLC).

The point most brands miss is that a leash is in scope. The handle is in prolonged contact with the walker's hand, so the leather falls under entry 47 and the clip under entry 27: one article, two reports. A retailer audit will ask for both. The Cr(VI) side is covered in depth in Cr(VI) in pet leather.

Two details that make the report usable:

1. On black or deep-dyed leather, ask for EN ISO 17075 Part 2. The HPLC method avoids interference from the colour of the extract. If both parts are run, Part 2 is the reference result.
2. Cr(VI) is not added; it forms after tanning, through pH rise, heat, UV or oxidising agents. Control therefore sits with the tannery, not the collar maker. Buy through an intermediary and you do not know which batch you received or under what process control. What a certificate does and does not cover is in how to read an LWG certificate. TL San Martín holds LWG Gold under the Leather Manufacturer standard, the only standard in which a Gold medal exists; a Trader tops out at "Approved".

Legal limits change. The binding reference is the consolidated text of REACH Annex XVII on the ECHA website, not this summary.

How do you spec leather that is riveted and buckled?

Hardware pulls on a very small area. The leather has to take that without tearing or letting the rivet migrate.

Three rules from the workshop floor. Seal the edge before the hardware goes on: an open edge at the buckle loop wicks moisture and accelerates any corrosion beneath it. Reinforce or double at the rivet point below 3 mm. And order leather finished on both sides: the flesh side of a collar sits against the dog, and on single-finished stock it is where the product starts to look cheap. It is the same strap leather a belt uses, cut from the same part of the hide; the numbers are in belt leather thickness and yield.

A wet test to run before you approve a sample

The test takes two days and costs almost nothing. Mount the exact production hardware on a strip of the approved leather, at the final thickness. Wrap it in a damp cloth with a few drops of mildly acidic water (vinegar diluted in tap water works), seal it in a bag and leave it 48 hours at room temperature. Then open the buckle and look under the tongue and the rivet. Any grey, blue or black ring on veg-tan means iron is reaching the leather: change the hardware or the tannage before you place the bulk order. This is an in-house screen, not a standard test, and it does not replace the EN 1811 and EN ISO 17075 reports.

What goes in the purchase order?

• Hardware material spelled out: "solid brass" or "stainless 316", never "metal" or "brass-coloured".
• Tannage matched to the hardware: veg-tan only with iron-free hardware.
• Leather thickness in mm with tolerance, plus finished both sides.
• Test reports: EN ISO 17075-2 for the leather and EN 1811:2023 for the hardware, each tied to the lot.
• Batch reservation if you plan a reorder. Cutting collar, leash, harness and treat pouch from one batch also gives you exact colour matching across the line and a much better yield, as shown in leather yield for dog collars. Five articles from one batch is also why we set no minimum on the leather itself.

Sourcing leather or hardware for a pet accessory line?

Post your material request under Materials on TL San Martín's open Business Board: free, B2B and read every week by tanneries, suppliers and workshops. If you are looking for a workshop to make the collars rather than for material, post under Partners & factories. Portuguese-speaking brands will find the purchase-order checklist in couro para coleiras de cão.

Go to the Business Board → · Request samples →

TL San Martín tans and finishes leather in Elda, Alicante, since 1995: LWG Gold under the Leather Manufacturer standard — the only standard under which a Gold medal exists at all; a Trader tops out at “Approved”. Request samples before you specify anything.