In the foundry industry, the pursuit of enhanced mechanical properties in aluminum-silicon alloys is a constant endeavor. Hypoeutectic Al-Si alloys, particularly those near the Al-7Si composition, are workhorse materials prized for their excellent castability, good strength-to-weight ratio, and reasonable corrosion resistance. These attributes make them indispensable for manufacturing complex, thin-walled components in the automotive and aerospace sectors, such as engine blocks, cylinder heads, and structural brackets. A critical factor determining the final performance of these cast components is their as-cast microstructure, which consists of primary α-Al dendrites or grains embedded in an Al-Si eutectic matrix. The morphology of the eutectic silicon phase, in particular, is a key microstructural feature; a coarse, plate-like structure acts as a stress concentrator and crack initiator, leading to poor ductility and fracture toughness. Consequently, modifying or refining this eutectic silicon is a fundamental practice to unlock the alloy’s full potential.
While various techniques exist, chemical modification remains one of the most prevalent and cost-effective methods. Elements like strontium (Sr) and sodium (Na) are well-known potent modifiers, but they come with drawbacks such as fading, porosity susceptibility, and oxidation issues. Antimony (Sb) presents itself as an alternative modifier, offering greater stability and less gas pick-up. However, its modification mechanism, especially under the specific solidification conditions of sand castings, is not as uniformly understood or agreed upon in the literature. The process of sand casting is characterized by relatively slow cooling rates, typically in the range of 0.1 to 1 °C/s, which significantly influences nucleation and growth kinetics compared to faster-cooling processes like die casting or permanent mold casting. In gravity sand castings, the thermal mass of the sand mold dictates a gradual heat extraction, allowing more time for microstructural evolution. This work focuses specifically on the effects of Sb addition on both the primary α-Al grains and the eutectic Si in a hypoeutectic Al-7Si alloy solidified under gravity sand casting conditions, aiming to clarify the operative mechanisms.
The microstructure development in sand castings is governed by classical solidification theory. For the primary α-Al phase to form, a certain degree of undercooling (ΔT) below the liquidus temperature is required to overcome the energy barrier for nucleation. The nucleation rate, I, can be described by a simplified form of the classical nucleation theory equation:
$$
I = I_0 \exp\left(-\frac{\Delta G^*}{k_B T}\right)
$$
where $I_0$ is a pre-exponential factor, $\Delta G^*$ is the critical free energy for nucleation, $k_B$ is Boltzmann’s constant, and $T$ is temperature. In practical sand castings, heterogeneous nucleation on inoculant particles or naturally occurring substrates is the dominant mode. Any element that alters the potency or number density of these nucleation sites will affect the final grain size.
The eutectic reaction, L → α-Al + Si, is equally critical. In unmodified sand castings, the eutectic Si grows anisotropically along preferred crystallographic directions, leading to a coarse, flake-like morphology. The modification process aims to alter this growth pattern, resulting in a finer, fibrous, or coral-like structure. Several theories exist for eutectic modification:
- Impurity-Induced Twinning (IIT) Theory: Certain modifier atoms (e.g., Sr, Na) adsorb on the growing Si crystal faces and promote a high density of twins, which provide re-entrant grooves for frequent branching, leading to a refined morphology. The effectiveness is related to the ratio of the atomic radii of the modifier ($r_M$) to silicon ($r_{Si}$). A ratio close to 1.65 is considered optimal for inducing twins.
- Restricted Growth Theory: Modifier atoms are adsorbed on the growth steps of the Si phase, effectively poisoning its growth and forcing it to change direction, leading to a more isotropic, refined structure.
- Nucleation-Based Theory: The modifier alters the nature of the nucleation sites for the eutectic Si. It may form new, more potent compounds (e.g., AlP becoming Al2Si2Sr) that increase the number of eutectic grains, thereby refining the structure.
The slow cooling inherent to sand castings provides a different kinetic landscape compared to rapid solidification processes. The extended time at high temperature can influence element segregation, intermetallic formation, and the stability of modifying agents. Therefore, understanding Sb’s behavior specifically in this context is crucial for its effective industrial application in sand castings.
To investigate this, a controlled study was conducted. A base hypoeutectic Al-7Si-0.5Mg alloy (hereafter Alloy 1) was prepared, along with a variant containing 0.09 wt.% Sb (Alloy 2). Minor additions of Ti (as a grain refiner via Al-Ti-B) were present. The melts were prepared from high-purity materials, degassed, and then poured into preheated sand molds to create test castings, faithfully replicating the gravity sand casting process. The thermal history during solidification was meticulously recorded using thermocouples connected to a data acquisition system, generating cooling curves. These curves are the fundamental diagnostic tool for analyzing solidification events in sand castings. Microstructural analysis involved multiple advanced techniques. Primary α-Al grain structure was revealed via anodizing and observed under polarized light microscopy (PLM). The eutectic Si morphology was examined using scanning electron microscopy (SEM), with samples lightly etched to partially remove the Al matrix for clearer observation of the Si network. Electron backscatter diffraction (EBSD) was employed to study the crystallographic orientation and potential nucleation sites of eutectic Si particles. Transmission electron microscopy (TEM) and associated energy-dispersive X-ray spectroscopy (EDS) were used to probe the atomic-scale features of the eutectic Si, such as the presence of twins and solute segregation. The experimental matrix is summarized below:
| Alloy ID | Si (wt.%) | Mg (wt.%) | Sb (wt.%) | Key Analysis |
|---|---|---|---|---|
| Alloy 1 | 6.80 | 0.48 | 0.00 | Base alloy for comparison |
| Alloy 2 | 6.90 | 0.50 | 0.09 | Sb-modified alloy |
The analysis begins with the thermal signature of solidification, as captured by the cooling curves—a vital record for any sand casting analysis. For Alloy 1 (unmodified), the cooling curve showed a distinct arrest corresponding to the nucleation of primary α-Al dendrites ($T_{N}^{Al}$) at approximately 619.5 °C. This was followed by the eutectic plateau, where the temperature remained relatively constant as the eutectic reaction proceeded, starting at $T_{N}^{Si}$ ≈ 578.1 °C. The addition of 0.09% Sb in Alloy 2 caused a noticeable shift in these thermal events. The nucleation temperature for the primary α-Al decreased to about 611.8 °C, indicating a greater required undercooling (ΔTAl) for its initiation. More significantly, the eutectic reaction temperature was depressed to approximately 572.6 °C. This depression of the eutectic temperature (ΔTeutectic) is a classic indicator of modification and is consistently observed in sand castings treated with modifiers like Sr or Sb. The undercooling can be calculated as:
$$
\Delta T = T_{eq} – T_{N}
$$
where $T_{eq}$ is the equilibrium eutectic temperature from the phase diagram (577 °C for binary Al-Si) and $T_N$ is the measured nucleation temperature.
| Thermal Event | Symbol | Alloy 1 (Unmodified) Temperature (°C) | Alloy 2 (Sb-Modified) Temperature (°C) | Change (ΔT, °C) |
|---|---|---|---|---|
| Primary α-Al Nucleation | $T_{N}^{Al}$ | 619.5 | 611.8 | -7.7 |
| Eutectic Reaction Start | $T_{N}^{Si}$ | 578.1 | 572.6 | -5.5 |
The increased undercooling for primary α-Al nucleation has a direct and measurable consequence on the macrostructure. Polarized light images of anodized samples revealed the size and morphology of the primary α-Al grains. In the unmodified Alloy 1, the structure consisted of a mix of dendritic and some equiaxed grains, with an average grain size measured to be 319 µm. With the Sb addition in Alloy 2, the grain morphology remained similar but the average size increased to 353 µm. This coarsening aligns perfectly with the thermal data: a lower nucleation temperature ($T_{N}^{Al}$) signifies that fewer potent nucleation sites become active, or that nucleation is generally more difficult, leading to fewer grains that then grow larger in the voluminous melt of a sand casting mold before impingement. This effect is often overlooked when discussing modification, but it is a significant microstructural change induced by Sb in the context of slow-cooling sand castings. The grain size, D, can be related to the number of nuclei, N, and the solidification time, $t_f$, typical of sand castings:
$$
D \propto (N)^{-1/3} \quad \text{and} \quad N \propto \int_{0}^{t_f} I(T(t)) dt
$$
A decrease in the nucleation rate I over the solidification period in a sand casting leads directly to a larger D.
The most pronounced visual effect of Sb is on the eutectic silicon phase. SEM micrographs provide stark contrast. In unmodified Alloy 1, the eutectic Si exhibits the expected coarse, plate-like morphology, with sharp edges and considerable interconnectivity. Quantitative image analysis showed these plates had an average length of 19.7 µm and an average width of 8.1 µm. In the Sb-modified Alloy 2, the eutectic Si structure is significantly refined. The plates are shorter, narrower, and appear more fragmented. The quantitative measurements confirm this: the average length is reduced to 13.3 µm and the average width to 5.4 µm. This represents a clear modification effect achievable in sand castings using Sb. The transition can be summarized as follows:
| Morphological Parameter | Alloy 1 (Unmodified) | Alloy 2 (Sb-Modified) | Percentage Change |
|---|---|---|---|
| Average Length (µm) | 19.7 | 13.3 | -32.5% |
| Average Width (µm) | 8.1 | 5.4 | -33.3% |
| General Morphology | Coarse, interconnected plates | Finer, shorter, more isolated plates/rods | N/A |
To understand how Sb achieves this modification in sand castings, it is essential to investigate the proposed mechanisms at a fundamental level. The first avenue explored was the nucleation-based mechanism. EBSD analysis was conducted to map the crystallographic orientations of different eutectic Si particles within a colony. If multiple Si particles share the same orientation, they likely originated from a single nucleation event. A variety of colors in the EBSD map indicates multiple, independently nucleated Si particles. The analysis for both Alloy 1 and Alloy 2 showed a similar diversity of colors (orientations) within the eutectic regions. There was no evidence that Sb addition increased the number of distinct nucleation events for the eutectic Si in these sand castings. This suggests that Sb is not primarily acting by creating a higher density of new, potent nucleation sites like AlSb or other Sb-containing compounds under these specific solidification conditions.
The second classical mechanism is the Impurity-Induced Twinning (IIT). TEM investigation of the eutectic Si in the Sb-modified Alloy 2 was crucial here. High-resolution imaging and selected area electron diffraction (SAED) patterns from the eutectic Si platelets did not reveal a high density of twins, which is the hallmark of modification by elements like Sr. Furthermore, EDS point analysis and scanning performed directly on the eutectic Si phase did not detect any significant concentration of Sb atoms within the silicon lattice or segregated at potential defects. The atomic radius ratio $r_{Sb}/r_{Si}$ is approximately 1.32, which is significantly lower than the ideal ~1.65 proposed for efficient twin induction. Therefore, the evidence strongly indicates that Sb does not modify eutectic Si in sand castings through the IIT mechanism.
Given that the primary mechanisms of increased nucleation or twin-induced branching are not supported by the evidence from these sand castings, an alternative explanation, rooted in the altered solidification thermodynamics and kinetics, must be considered. The key experimental observation is the depression of the eutectic temperature ($T_{N}^{Si}$). According to the Al-Si phase diagram, the solid solubility of Si in the α-Al phase, $C_{Si}^{α}$, is a function of temperature:
$$
C_{Si}^{α}(T) \approx C_0 \exp\left(-\frac{Q}{RT}\right)
$$
where $C_0$ is a constant, Q is an activation energy, and R is the gas constant. A lower eutectic reaction temperature means that at the moment of eutectic nucleation and growth, the equilibrium solid solubility of Si in the α-Al is reduced. However, the overall composition of the liquid prior to the eutectic reaction remains largely unchanged. This creates a larger chemical potential gradient or driving force for the rejection of Si atoms from the growing α-Al phase into the liquid at the solid-liquid interface. The chemical potential difference, Δµ, which is the driving force for diffusion, can be conceptually related to the deviation from equilibrium concentration:
$$
\Delta \mu \propto RT \ln\left(\frac{C_L}{C_{Si}^{α}(T)}\right)
$$
where $C_L$ is the silicon concentration in the liquid at the interface. A lower $T$ (and thus lower $C_{Si}^{α}(T)$) increases Δµ.
In the cooperative growth of the α-Al and Si eutectic, this increased driving force accelerates the diffusion of Si atoms to the growing Si phase. The growth rate of the Si phase, $V_{Si}$, is proportional to this flux:
$$
V_{Si} \propto D_{Si}^{L} \cdot \nabla C \propto D_{Si}^{L} \cdot \frac{\Delta \mu}{ \lambda}
$$
where $D_{Si}^{L}$ is the diffusion coefficient of Si in the liquid and λ is the eutectic spacing. The accelerated growth of Si, under the constrained conditions of a slowly cooling sand casting, may lead to instability in its anisotropic plate growth mode. Instead of forming long, stable plates, the rapid uptake of Si might promote more frequent growth direction changes or tip splitting, resulting in shorter and narrower structures. Simultaneously, the growth of the α-Al phase is also accelerated to maintain the coupled growth front, further limiting the space and time available for the Si plates to extend. This mechanism explains the refinement without requiring Sb to directly interact with the Si crystal’s nucleation core or its growth steps through adsorption. It is a consequence of Sb’s effect on the phase equilibrium—shifting the eutectic point and depressing the growth temperature—which in turn alters the kinetic conditions at the solidification front in a sand casting.
The practical implications of using Sb in industrial sand castings are multifaceted. First, the grain-coarsening effect must be accounted for in component design. If a fine equiaxed grain structure is critical for properties like yield strength and fatigue resistance, the use of Sb may necessitate a stronger or more fade-resistant grain refiner system to compensate. Second, the modification effect of Sb is known to be less potent than that of Sr, often requiring higher addition levels (0.1-0.3 wt.%) for full modification in sand castings. However, Sb offers distinct advantages: it is less prone to fading during prolonged holding or remelting, and it does not increase the susceptibility to hydrogen porosity—a common problem with Sr in sand castings where drying of the mold is critical. The environmental and health aspects of Sb must also be considered in the foundry environment. From a process control perspective, the clear depression of the eutectic plateau on the cooling curve provides a reliable thermal analysis signature for quality control in sand casting production, allowing for real-time assessment of modification efficiency.
In summary, under the specific thermal conditions of gravity sand casting, the addition of antimony to a hypoeutectic Al-7Si alloy induces a complex set of microstructural changes. The primary α-Al grains become coarser, with an average size increase from 319 µm to 353 µm, due to a reduction in their nucleation temperature, making initiation more difficult in the slow-cooling environment of a sand mold. The eutectic silicon undergoes significant refinement, transforming from coarse plates to shorter, narrower structures. The mechanism for this modification in sand castings appears not to be related to a change in nucleation site potency or density, nor to the impurity-induced twinning mechanism. Instead, it is proposed that Sb alters the solidification thermodynamics, depressing the eutectic reaction temperature. This depression reduces the solid solubility of Si in α-Al at the growth front, thereby increasing the chemical potential driving force for Si rejection. The resultant accelerated diffusion and growth kinetics of both eutectic phases under the constrained, slow cooling of a sand casting disrupt the stable plate growth of Si, leading to a refined morphology. This understanding provides a clearer guideline for the application of Sb as a modifier in sand castings, highlighting both its benefits and the need for complementary grain refinement, while offering an explanation rooted in the fundamental interplay between thermodynamics and kinetics specific to this important casting process.