TWI Frequently asked questions
by Steve Jones
Glass-metal bonding is generally said to occur via two mechanisms: mechanical and/or chemical.
Mechanical bonding occurs because of the nature of the roughened surfaces. In shear, surface roughness provides a frictional force which must be overcome to separate the parts. The glass can penetrate surfaces, pores or cavities thus providing further surface area for interlocking between the structures.
Chemical bonding refers to atomic or molecular bonds between materials. Glass to metal, and oxide ceramic to metal interfaces are not readily compatible. To produce an intimate bond between the materials commercially, the metal must be pre-oxidised before application of the glass.
A typical procedure for production of a bond is as follows: cleaning of the metal, pre-plating of the metal to promote bonding, further cleaning followed by a bake-out. Baking out (by decarburisation, or use of hydrogen) drives off gas-producing species and other products that may also need to be removed, such as nitrogen for titanium alloys. Controlled oxidation is then required to produce an adherent oxide for the glass to dissolve and to which it can bond. The most common glasses used are boro-silicate, although frequently lead-silicate is used on copper alloys and stainless steels. Increasingly though, lead is being replaced with barium because of the potential health hazards. The glasses are applied as frits, paste or solid preform and a range of heating methods, such as electrical, RF or flame, may be used to soften the glass to the correct viscosity for joining (usually under a slight pressure to provide intimate contact with the metal). Graphite and stop-offs are used for alignment, since glasses do not readily bond to these materials and flow is inhibited. The atmosphere for joining is usually a forming or protective gas to promote wetting and adhesion, with an annealing stage generally included during cooling to inhibit development of residual stress.
