Factory:  Corner Steyr & Alfa Sts 
Aureus Industrial sites 
Email: info@keramicalia.co.za
Website: www.keramicalia.co.za
Cell: Dave  082  808 4757
Email: keraccounts@iafrica.co.za
Accounts: Myrtle Wakeford
Tel: (011) 764 2139
Tel: (011) 412 3261
P O  Box 2288
Wilro Park,1731

(Reg No: 1993/034692/23

P O Box 55705 ARCADIA 0007 Tel/Fax 012 343 5429 Cell: 082 403 3615

email: This email address is being protected from spambots. You need JavaScript enabled to view it.



Bonding between refractories and metal substrates is problematic because of the lack of chemical interaction between any of the constituents of the refractory with the substrate.  Specifically, the hydration products of dry cement have no affinity for clean metal surfaces. This means that bonding is primarily dependent on mechanical keying and weak, short range forces.

Phospatch is a phosphate bonded high alumina refractory. Phospatch is supplied as a two-pack product: a dry powder consisting of refractory with cement, and an activating solution.

In Phospatch, the activating solution causes a chemical reaction on the metal surface, and this provides a measure of chemical bonding.


Trials were conducted to ascertain the relationship between surface preparation and effectiveness of bonding. These trials were conducted on copper blocks with machined tapered channels, as would be found on the surface of a water-cooled panel operating in a furnace sidewall. The dimensions of the copper blocks were 250mm x 150mm.

All the surfaces were degreased. One had no further treatment, one was washed with phosphoric acid to convert residual oxides to phosphate and one was treated with a commercial silane primer. The blocks were encased in a brick-shaped mould that gave final product dimensions 250mm x 150mm x100mm after Phospatch had been cast into it. Phospatch was mixed and poured into the mould. After setting, the copper-refractory composite was sectioned and examined. The surface detail was examined in a Jeol 840A microprobe analyser. The degree of adhesion was evaluated by the amount of observable chemical interaction between refractory and metal.


There was little adhesion between phosphoric acid-treated copper or the silane-primed copper and the refractory. The copper surface with the degreasing treatment only exhibited a chemical reaction that resulted in the formation of mixed copper-aluminium phosphates. These were visible in the refractory at depths of up to 50μm and were usually less than 1μm in diameter. A typical area is shown in Figure 1. The mixed phosphate particles always bridged neighbouring aggregate grains when found in the refractory; they were not found as discrete particles in the matrix. Accommodation between the refractory and asperities on the metal surface was good, especially in areas where there was most copper aluminium phosphate, as shown in Figure 2. In areas where the surface asperities were too small to accommodate the smallest aggregate particles, the surface was covered with impervious aluminium phosphate.

The main body of the refractory contained numerous spherical closed pores, as shown in Figure 3. These pores were absent from the refractory in the immediate vicinity of the metal. This absence of porosity next to the metal surface contributes to better fracture strength in this area.


The activating solution contains a soluble aluminium salt. There is an insoluble salt on the surface after the refractory has set. This means that the aluminium salt has either decomposed on drying to produce aluminum phosphate or there has been a reaction between that salt and the copper. Whatever reaction occurs between the activator solution and the substrate occurs most readily on a surface containing residual oxides. It occurs less readily on a chemically clean copper surface or a surface on which the residual oxides are converted to phosphates. It may be postulated that the best degree of adhesion is obtained on a surface that is cleaned and subjected to a controlled chemical oxidation treatment before application of the refractory.


figure 1

figure 2