The Kirkendall Effect: When Atoms Don't Play Fair in Solid-State Diffusion

The Kirkendall effect describes the macroscopic movement of the original interface between two materials in a solid-state diffusion couple, a phenomenon resulting from the unequal diffusion rates of the constituent atomic species. First demonstrated experimentally by Kirkendall and Smigelskas in 1947 using a copper-brass system, the effect disproved the classical model of diffusion via direct atomic exchange. The observed movement of inert markers at the interface provided conclusive evidence that atoms diffuse primarily by moving into adjacent vacancies, and that different elements have different rates of diffusion. The underlying mechanism involves an imbalanced flux of atoms across the interface, which is compensated by an equal and opposite net flux of vacancies. This net flow of vacancies into the region losing the faster-diffusing species can lead to vacancy supersaturation and their subsequent coalescence into microscopic pores, known as Kirkendall voids. This porosity is a significant practical concern, as it can severely compromise the mechanical integrity and lead to the failure of components in applications such as microelectronic wire bonds, solder joints, and high-temperature coatings. Thus, the Kirkendall effect is a fundamental concept in materials science with profound implications for the design and reliability of advanced materials and devices.

1. Introduction to Solid-State Diffusion

Solid-state diffusion is a fundamental process in materials science where atoms move or migrate through a crystal lattice. This movement is driven by thermal energy and concentration gradients, causing the material's composition to change over time. Before 1947, the prevailing model for diffusion in substitutional alloys (where atoms of one element replace atoms of another on the lattice) was a simple, direct exchange mechanism. It was assumed that two adjacent atoms of different types would simply swap places. A key implication of this model is that the flux of atoms 'A' into material 'B' would be exactly equal and opposite to the flux of atoms 'B' into material 'A'. If this were true, the original interface between the two materials in a diffusion couple (two blocks of different materials placed in intimate contact) should remain stationary during diffusion.

However, the groundbreaking work of Ernest Kirkendall and Alice Smigelskas in 1947 shattered this classical view and introduced a new paradigm for understanding diffusion.

2. The Landmark Kirkendall Experiment

The experiment that revealed the effect was elegant in its simplicity. A block of 70% Cu-30% Zn brass was used as a core, and fine, inert molybdenum (Mo) wires were wrapped around it. This assembly was then electroplated with a thick layer of pure copper, creating a Cu-Brass-Cu sandwich. Molybdenum was chosen for the markers because it is effectively insoluble in both copper and brass at the experimental temperature (785 °C), ensuring the markers would not diffuse away themselves.

According to the direct-exchange model, as zinc atoms diffused out of the brass and into the copper, and copper atoms diffused into the brass, the inert Mo markers at the original interface should have remained in place.

The Observation: After holding the diffusion couple at 785 °C for 56 days, Kirkendall and Smigelskas made a startling discovery: the distance between the Mo markers on opposite sides of the brass core had decreased. The markers had moved inward, toward the center of the brass. This observation was irrefutable proof that the original interface had shifted.

The Conclusion: The only way the markers could move is if there was an unequal exchange of atoms across the interface. Specifically, zinc atoms diffused out of the brass and into the copper much faster than copper atoms diffused from the pure copper into the brass. This imbalance in diffusion fluxes is the essence of the Kirkendall effect.

3. The Mechanism: Vacancy-Mediated Diffusion

The Kirkendall effect provided conclusive evidence for the vacancy diffusion mechanism as the dominant process in substitutional alloys. The direct-exchange of atoms is energetically very costly, as it requires significant local distortion of the crystal lattice for the atoms to squeeze past each other.

In the vacancy mechanism, an atom moves by hopping into an adjacent, empty lattice site—a vacancy. For diffusion to occur, both atoms and vacancies must be mobile. The Kirkendall effect arises because the two diffusing species have different affinities for hopping into vacancies and different frequencies of doing so.

Let's reconsider the Cu-Zn system:

1. Unequal Fluxes: The flux of zinc atoms out of the brass (J_Zn) is greater than the flux of copper atoms into the brass (J_Cu).

* (J_Zn > J_Cu)

2. Net Flow of Vacancies: Since more atoms are leaving the brass side than are arriving, there is a net loss of atoms from the brass side and a net gain on the copper side. To conserve the number of lattice sites, this net flow of atoms must be balanced by an equal and opposite net flow of vacancies.

* Vacancies flow from the copper side (where atoms are arriving) to the brass side (where atoms are leaving).

3. Marker Movement: The inert markers are embedded in the crystal lattice. As the lattice on the brass side effectively shrinks due to the net outflow of atoms (and inflow of vacancies), the markers are swept along with this bulk movement of the lattice. This is why the markers were observed to move inward.

4. Consequences of the Kirkendall Effect

The net flow of vacancies is not just an abstract concept; it has tangible and often detrimental consequences for materials.

• Kirkendall Porosity: The side of the diffusion couple that experiences a net influx of vacancies (the brass, in this case) can become supersaturated with them. If the vacancy concentration exceeds the equilibrium level, these excess vacancies can coalesce to form microscopic voids. This phenomenon is known as Kirkendall porosity or Kirkendall voids. These voids act as stress concentrators and can severely degrade the mechanical properties (e.g., strength, ductility) of the material or component.

• Swelling and Shrinkage: The region that gains the faster-diffusing species will swell, while the region that loses it will shrink (partly due to void formation). This can induce internal stresses within the material.

5. Real-World Examples and Significance

The Kirkendall effect is not an academic curiosity; it is a critical factor in the reliability and failure of many modern technologies.

• Microelectronics: A classic failure mechanism in wire bonding is known as "purple plague." When gold (Au) wires are bonded to aluminum (Al) contact pads on a silicon chip, Al diffuses much more rapidly into the Au than vice-versa. This leads to the formation of brittle Au-Al intermetallic compounds and, crucially, a line of Kirkendall voids at the interface. These voids weaken the bond, leading to electrical failure, especially in devices subjected to thermal cycling.

• Soldering: In modern lead-free solder joints (e.g., Sn-Ag-Cu solders on a copper substrate), the rapid diffusion of copper from the substrate into the tin-based solder can create thick, brittle intermetallic layers and Kirkendall voids. Over time, this degrades the mechanical integrity of the solder joint, making it susceptible to cracking and failure.

• High-Temperature Coatings: Thermal barrier coatings (TBCs) are used to protect superalloy turbine blades in jet engines. Diffusion occurs at the high operating temperatures between the metallic bond coat and the superalloy substrate. The Kirkendall effect can lead to void formation at this interface, which can cause the protective coating to peel off (spallation), exposing the blade to the harsh environment.

• Nanotechnology (Harnessing the Effect): The Kirkendall effect can also be used constructively. To create hollow nanostructures, one can start with a core-shell nanoparticle (e.g., cobalt core, platinum shell). When heated, the faster-diffusing core atoms (cobalt) diffuse outward through the shell, leaving a void behind. This process, known as the "nanoscale Kirkendall effect," is a powerful method for fabricating hollow spheres and tubes for applications in catalysis and drug delivery.

Browse public study notes