Question 4 of 8: Interfaces, Precipitate-Free Zones and Grain-Boundary Pinning

Question 4: Interfaces, Precipitate-Free Zones and Grain-Boundary Pinning (20 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

4.1 — (a)(i)–(iii) Coherent, semi-coherent and incoherent interfaces

1. (i) The three interface types. A coherent interface has every lattice plane of the precipitate continuing, atom-for-atom, into the matrix lattice across the boundary — the two lattices match perfectly in registry, at the cost of elastic (coherency) strain if the two lattice parameters differ. A semi-coherent interface still matches most lattice planes, but the residual misfit (beyond what the lattice can elastically accommodate) is relieved periodically by an array of misfit (interfacial) dislocations, so most of the interface is locally coherent between dislocations. An incoherent interface has no systematic lattice correspondence at all — it behaves structurally like an ordinary high-angle grain boundary, with a disordered transition region several atomic layers thick.
2. (ii) Why the interfacial energies differ so much. A coherent interface has NO broken/mismatched bonds, so its interfacial energy is purely a chemical (compositional) term, typically 10–200 mJ/m² — low. A semi-coherent interface adds the (extra) energy of the misfit-dislocation array itself (each dislocation core costs energy per unit length, and the dislocation spacing sets how much of this energy is added per unit interfacial area) on top of the residual chemical term, typically 200–500 mJ/m². An incoherent interface has a fully disordered, broken-bond structure resembling a general high-angle grain boundary, with energy typically 500–1000+ mJ/m² — several times higher than coherent.
3. (iii) Driving force for coherent→incoherent as particle size increases. The total energy of a coherent precipitate has two competing terms: interfacial energy (scales as particle surface area, $\propto r^2$) and coherency STRAIN energy stored in the surrounding matrix (scales as particle VOLUME, $\propto r^3$, because the elastic strain field fills the matrix around the whole particle). While the particle is small, the $r^2$ interfacial term dominates and coherency (low interfacial energy) is favoured. As the particle grows, the $r^3$ strain-energy term eventually overtakes the $r^2$ interfacial term, and it becomes energetically favourable to trade some of that strain energy for the (smaller, per-unit-area) cost of nucleating misfit dislocations that relieve the strain — i.e. losing coherency. This crossover is the same $r^2$-vs-$r^3$ competition that sets the critical nucleus radius in Question 7(a), just applied to strain energy instead of the bulk chemical driving force.

4.2 — (b) The two mechanisms of precipitate-free-zone (PFZ) formation

1. Solute-depletion (denuded-zone) mechanism. Grain boundaries are the preferred heterogeneous nucleation sites for the equilibrium precipitate (lowest nucleation barrier). During quenching or the early stages of ageing, boundary precipitates nucleate and grow FIRST and fastest, drawing solute out of the adjacent matrix by long-range diffusion. This locally depletes the matrix of solute near the boundary faster than it can be replenished, so the region immediately adjacent to the boundary falls below the supersaturation needed to nucleate the (matrix) precipitate at all — leaving a solute-poor PFZ that widens with ageing time as $\sqrt{Dt}$.
2. Vacancy-denuded-zone mechanism. Many age-hardening precipitates (e.g. GP zones) nucleate homogeneously in the matrix ONLY with the help of excess quenched-in vacancies (vacancies lower the nucleation barrier by assisting the local diffusion/clustering needed to form the zone). Grain boundaries are efficient vacancy sinks, so during and immediately after quenching, vacancies annihilate at the boundary faster than they can diffuse in from the bulk, creating a vacancy-DEPLETED zone adjacent to the boundary. Without the excess-vacancy assistance, GP zones (or other homogeneously nucleated precipitates) fail to nucleate in that near-boundary region even though the SOLUTE content there is unchanged — a mechanistically distinct PFZ from mechanism 1.

4.3 — (c) Why fine precipitates restrict grain growth

Grain growth is driven by the reduction of total grain-boundary area/energy: a boundary curves toward its own centre of curvature and migrates in that direction (Question 5(a)), giving a driving pressure $P_{\text{boundary}}=2\gamma_b/D$ for a grain of diameter $D$ and boundary energy $\gamma_b$. A second-phase particle sitting ON a migrating boundary REDUCES the total boundary area by an amount equal to the particle's own cross-sectional area where the boundary intersects it; pulling the boundary fully past (detaching from) the particle would have to recreate that area of boundary, which costs energy. The particle therefore exerts a retarding (pinning) force pulling the boundary back, opposing its motion — this is the Zener pinning mechanism, quantified in Question 5(c)–(d) as a drag pressure $P_{\text{drag}}=3f\gamma_b/2r$ for a volume fraction $f$ of particles of radius $r$.

Because $P_{\text{drag}}\propto f/r$, a given volume fraction of precipitate is far more effective at pinning boundaries when it is present as MANY FINE particles than as fewer coarse ones (halving $r$ at fixed $f$ doubles the drag pressure). At sufficiently high temperature and long time, ordinary grain growth (driven by $P_{\text{boundary}}\propto1/D$, which weakens as grains coarsen) is eventually reduced to below the pinning pressure of the fine precipitate dispersion, and the boundary becomes pinned in place — grain growth halts (or is drastically slowed) once $D$ reaches the Zener limit $D_{\max}=4r/3f$ (derived in Question 5(c)). This is precisely why fine, stable, high-volume-fraction dispersoids (e.g. Nb(C,N) in HSLA steels, MnS or AlN in other grades) are deliberately engineered into steels that must resist grain coarsening during hot working or welding.