The process of brazing involves the use of a molten filler metal to wet the mating surfaces of a joint by capillary action, with or without a fluxing agent, leading to the formation of metallurgical bonds between the filler and respective components. The reactions which occur between the molten filler and surfaces of the solid components result in solutionization, which is erosion of the solid phase by the molten filler alloy. However, this solutionization reaction is localized to the surface of the solid components and rarely exceeds a depth on the order of tens of microns. This article is the first of two articles that explore the thermodynamic principles for heat treating steel by brazing.
In many respects, soldering and brazing can be considered together, since they both depend on similar metallurgical processes. So what is the difference between soldering and brazing? Historically, the primary distinction is based on the melting temperature of the filler metal. Solders are defined as filler metals with melting points below 450° C, while brazes have melting points above. This distinction has a historical origin. The earliest solders were based on alloys of tin, while brazes were based on copper-zinc alloys. The word “solder” comes from the Anglo-French word soudure, which means “to make solid.” On the other hand, the word “braze” comes from the old French word braser, which means “to burn.” The type of reaction between a filler and parent metal can also be used to differentiate soldering from brazing. Solders usually react to form intermetallic phases, which are compounds of the constituent elements that have different crystal structures and stoichiometry compared to the elements in solid form. By contrast, most brazes form solid solutions, which are mixtures of the constituents which substitute randomly for each other in the crystal structure. Most commercial solders are of eutectic composition, because there is usually a need to minimize the processing temperature while maintaining fluidity of the molten filler metal. The lower processing temperature means that the metallurgical reaction between the molten filler and substrate is less extensive and solders rarely dissolve more than a few microns of the component surfaces, where brazes often dissolve tens of microns. Brazes tend to form mutual solid solutions between their constituent elements, so they melt over a range of temperatures that varies with composition. As such, eutectic compositions for braze filler metals are much less common.
We talked about why soldering and brazing are different, but let’s take a moment to discuss the similarities. Soldering and brazing involve the same basic bonding mechanism where a reaction occurs with the parent material to form metallic bonds at the interface. In both situations, good wetting promotes the formation of fillets via capillary action that serve to enhance the strength of the joints. Similar processing conditions are required, for example both solders and brazes must be heated to melt the filler at a temperature just above the melting point of the filler metal and residual oxide on the parent materials must be removed with flux or a reducing atmosphere. The physical properties of the resulting joints are comparable, provided that the same homologous temperature is used for the comparison. Soldered and brazed joints can both be endowed with properties that approximately match, and in some cases exceed, those of the components. Throughout this article, the term “filler” or “filler metal” will be used interchangeably as generic terms for braze alloys. Please keep in mind that although we will focus on the brazing process, the applicable thermodynamics and a lot of the material concepts apply to the soldering process as well.
Wetting, Spreading and the Young Equation
In a previous article in this series I described how the Young equation defines the degree to which a liquid phase will wet a solid it is in physical contact with; within a vapor-liquid-solid three phase system (Fig. 1). This is the exact situation for a molten filler metal on a solid substrate in a furnace or air ambient. We know that an intimate solid/liquid interface can be easily formed if the liquid wets, spreads and penetrates any surface irregularities in the solid. The Young equation defines the relationship between the relative interfacial free energies of the three phases that comprise the system. γsv represents the interfacial energy between the solid and vapor, γsl the interfacial energy between the solid and liquid, and γlv between the liquid and vapor. If cos Θ is < 90°, then by definition the molten filler metal wets the solid it is in contact with.
The Young Equation assumes no reactions take place at the interface between the solid substrate and liquid sessile drop, but since a chemical reaction occurs at the interface between the molten filler and the solid it is sitting on, where some of the substrate is dissolved into the filler, the free energy of reaction per unit interfacial area and unit time enhances the driving force for wetting by effectively lowering the relative surface energy between the solid and liquid. In this case, as long as the filler metal is still molten it will not only wet, but it will spread out on the substrate.
Now, rewriting the Young equation in terms of cos Θ,
cos Θ =
it is clear that wetting is improved by decreasing the contact angle Θ. In terms of the relative interfacial free energies of the 3-phase system, this can be accomplished in three possible ways. One way is to maximize the term γsv, which means the surface of the solid must be critically clean. The presence of adsorbed material such as water vapor, dust, or organic contaminants reduces γsv and increases the contact angle Θ. So for both soldering and brazing the joint surfaces must be clean and metallic, which can also be facilitated by the use of flux or a protective atmosphere, such as nitrogen. In some cases, particularly for higher temperature brazing processes, a reducing atmosphere will help suppress or even reduce surface oxides. Another way to decrease the contact angle is to decrease γsl. Since γsl is a constant at a fixed temperature and a particular solid-liquid combination, this parameter can only be reduced by changing the materials system. This is not easy to achieve in practice, because the component materials are usually specified to fulfill certain functional requirements. However, γsl is very temperature dependent and decreases rapidly with increasing temperature, which within limits can provide a means of controlling the wetting by the molten filler metal. Finally, a decrease in γlv will also decrease the contact angle. γlv is a constant at a fixed temperature and pressure for a particular liquid-vapor combination, but can be varied by altering the composition and pressure of the atmosphere. Although the composition of the atmosphere for soldering or brazing is known to affect the contact angle, it is often easier to promote wetting by reducing the pressure of the atmosphere. This is why vacuum brazing is such an attractive joining process.
Capillary Action at Interfaces
Up to this point we have considered filler metal wetting over a single surface. In a solder or braze joint there are always two facing surfaces. If the contact angle is less than 90°, the surface energies will give rise to a positive capillary force that will act to fill the joint. For a pair of vertical parallel plates that are D mm apart and partially submerged in a liquid (Fig. 2), the capillary force per unit length of joint is 2γlv cos Θ. This force will cause the liquid to rise up to an equilibrium height h, at which point the upward capillary force is balanced by the downward hydrostatic force so that h is given by,
where ρ is the density of the liquid, g is the acceleration due to gravity, and D is the separation between the plates. It should be noted that the height of the meniscus increases as the distance between the plates decreases. Like the construction for the Young equation, this model assumes no reactions between the liquid and the plate surfaces. As such, this simple model does not truly represent the actual situation in soldering and brazing. Reactions between a molten filler metal and the substrates usually result in dissolution of the surface of the substrate, which can lead to the growth of new phases. Frequently these phases are intermetallic compounds that are either distributed throughout the joint or form layers adjacent to the interface.
The Effect of Temperature and Composition
Molten filler metals do not all have the same characteristics, but with few exceptions the extent to which they wet and spread over a suitable substrate increases as the temperature is raised. A suitable substrate is defined as an atomically clean metallic surface that is wetted by the filler metal and does not significantly alloy with the filler. The degree of spreading also increases as the pressure of the atmosphere is lowered or the atmosphere is made more reducing. Eutectic filler alloys are generally regarded as having the best spreading characteristics, compared to hypo- and hypereutectic compositions. This is because eutectic alloys melt congruently; that is they transition from solid to liquid phase directly without the mushy liquid + solid intermediate phase that occurs with off-eutectic alloys. For non-eutectic filler alloys, melting, wetting and spreading occur before the alloy is entirely molten, where it tends to be somewhat viscous and its movement rather sluggish. By the time a non-eutectic filler is completely molten, the filler may have partially dissolved the surface of the substrate, which can diminish the driving force for wetting by changing the composition of the filler, which in turn often increases the value of γlv. By comparison, eutectic alloys have lower viscosity when molten and since they go from solid to liquid almost instantly, capillary action tends to facilitate spreading.
It should be noted that braze alloys are less sensitive to composition compared to solder alloys, because their constituent elements tend to be mutually soluble in both the solid and liquid phase. The spreading behavior of filler metals is better in a reducing atmosphere or in air with activated flux, compared to a protective atmosphere or vacuum furnace environment, because protective gases like nitrogen or argon and vacuum do not remove oxides that form on the surface of components or filler metals when they are exposed to air prior to the joining process.
The Effect of Surface Roughness
The Young equation assumes the surface of the solid phase is flat and smooth, in essence it is “ideal.” But most real surfaces are not flat or smooth, they are generally irregular and contain defects, voids, local deformations and asperities. From a macroscopic point of view, the contact angle of a liquid on a solid will have an observable or apparent contact angle. But if one looks very closely at a microscopic level, the surface imperfections or roughness can actually modify the actual or intrinsic contact angle. In 1936, Robert Wenzel addressed this issue by proposing a modification to the Young equation by defining a roughness coefficient, rs, which is the actual area of the surface divided by the apparent area of the surface. In other words, the actual surface is rough and the apparent surface is flat and smooth. So, the Wenzel modification of the Young equation is
rs (γsv – γsl) = γlv cos Θapp
According to the Wenzel modification, one still measures the macroscopic or apparent contact angle, but the Young equation is modified by the roughness coefficient. Since rs is always greater than 1 (i.e. the area of a rough surface is greater than the area of a flat surface), for situations where the contact angle is acute, the roughness will enhance the propensity of the liquid to wet the solid. Conversely, if the contact angle is obtuse, the surface roughness will enhance the tendency of the liquid to de-wet the surface.
It is now well established that the roughness of joint surfaces has a significant effect on the wetting and spreading behavior of the filler. For example, Okamoto et al. found that when vacuum brazing aluminum with Al-12Si filler alloy, the best results were achieved when the component surfaces were ground prior to brazing using SiC grinding papers with a grit size between 400 and 600. In this case, intentional roughening of the surface by dry surface grinding promoted wetting and spreading. It is also known that surface texture can affect the capillary force acting between the filler and component surfaces. In the case where machining marks or grinding grooves are parallel to the advancing front of the melt, wetting is inhibited by periodic variation in the contact angle. However, grooves perpendicular to the advancing melt front tend to promote wetting by introducing additional capillary forces. In this case the steady state wetting front is then wavy, with the melt sucked into the grooves. Finally, a random network of grooves or scratches usually promotes wetting as a result of additional capillary forces, but may leave behind unwetted islands at high spots if the roughness is too extreme. For maximum effect, the surface texture should be as jagged as possible. But the old adage that “if enough is good, more is better” has a limit in this case. If the texturing is too deep, then capillary dams can be formed and these will impede the spreading of the filler metal.
The Effect of Interfacial Reactions
We know from the discussion of reaction modified wetting in my previous article on the Young equation, that when a chemical reaction occurs at the interface between a liquid sessile drop and the solid it is sitting on, the free energy of reaction per unit interfacial area enhances the driving force for wetting by effectively lowering the relative surface energy between the solid and liquid. It is frequently observed that a molten filler metal will continue to spread beyond an initially wetted surface over a period of time, due to interfacial reactions that follow from dissolution of the substrate into the molten filler. The dissolution rate and associated growth of intermetallic compounds, or IMC, both follow an Arrhenius relationship where the rate is proportional to exp (-Q/kT), where Q is the activation energy, k is the Boltzmann constant and T is the absolute temperature. So according to Arrhenius behavior, the rate will increase exponentially with decreasing activation energy and increasing temperature. Interfacial reactions are important for determining the flow characteristics and wetting behavior of the filler and the properties of the resulting joint. When the molten filler metal wets the parent materials, there is normally some intersolubility between them, which is manifested as dissolution of the surfaces of the parent materials and formation of new phases at the interfaces or in the filler itself after joint solidification. The rate of dissolution of a solid metal in a molten metal is given by the integrated form of the Shukarev-Nernst equation, where the concentration of the dissolved metal in the melt at a given time “t” is given by
C(t) = Cs [1 – exp(-KAt/V)]
where Cs is the saturation concentration of the solid metal in the molten filler at the temperature of interest, K is the dissolution rate constant, A is the wetted surface area and V is the volume of the melt. According to this equation the concentration of dissolved metal in the molten filler increases inverse exponentially with respect to time. The dissolution rate is very fast at first, but then slows as the concentration of the dissolved parent materials approaches its equilibrium saturation limit. For most brazing material systems, the equilibrium saturation condition is reached very quickly, typically within seconds. This means that equilibrium phase diagrams can be used to predict the change in composition of the filler metal as a result of substrate dissolution, which is a very important and useful concept.
In this article we have examined the thermodynamic principles that govern the behavior of the brazing process. In the next article, Brazing (Part 2), we will consider the major braze alloy systems and their areas of application, some of the most prominent brazing techniques, brazing fluxes, braze joint design and a case study to illustrate a practical brazing application.
Copyright © 2018 by Michael T. Powers – All rights reserved.
Mike Powers, F.A.S.M