In my research on magnesium alloy sand casting foundry, I focused on understanding the protective mechanisms that allow the production of high-quality magnesium alloy castings without the use of conventional anti-ignition additives. Ordinary sand casting of magnesium alloys typically requires protective agents to prevent oxidation and combustion during pouring, but these agents generate toxic gases that harm both the environment and equipment. To address this issue, I explored the use of dry graphite-sand moulds, which can eliminate the need for such additives. In this paper, I present my findings on the composition and structure of the surface oxide film formed on ZM5 magnesium alloy castings produced in dry graphite-sand moulds. Through X-ray photoelectron spectroscopy (XPS) analysis, I determined that the surface film exhibits a layered structure consisting primarily of MgO and Mg(OH)₂, with significant carbon infiltration that enhances film density and protective performance. My results provide a fundamental understanding of the protective mechanism in magnesium alloy sand casting foundry processes.
Introduction
Magnesium alloys are increasingly used in automotive and aerospace applications due to their low density and high specific strength. However, their high chemical reactivity, especially in the molten state, presents significant challenges in sand casting foundry operations. Conventional methods involve adding sulfur hexafluoride, sulfur dioxide, or other protective gases to the mould cavity, which are costly and environmentally harmful. Recent developments in mould material technology have shown that dry graphite-sand mixtures can effectively suppress oxidation of molten magnesium without generating toxic fumes. The underlying reason is believed to be the formation of a dense, carbon-containing oxide film on the casting surface that acts as a barrier. Therefore, a thorough investigation of this film’s composition and structure is essential for optimizing magnesium alloy sand casting foundry practices.
In this study, I used ZM5 magnesium alloy (8.3% Al by weight) cast into dry graphite-sand moulds without any protective additives. The as-cast surface samples were analyzed by XPS to determine the elemental composition and chemical states of Mg, O, and C as a function of depth. I also performed Ar⁺ ion sputtering to obtain depth profiles. The goal was to characterize the layered structure of the oxide film and explain how carbon incorporation improves its protective nature.
Experimental Procedures
All samples were prepared from 7 mm × 7 mm × 0.5 mm coupons cut from the surface of ZM5 castings produced in dry graphite-sand moulds. The moulds consisted of a mixture of silica sand and graphite powder (approximately 20% graphite by weight), which were dried at 200°C for 2 hours before pouring. The pouring temperature was 730°C, and the mould temperature was approximately 150°C. No sulfur hexafluoride or other protective gas was used.
XPS measurements were conducted using a Kratos XSAM800 spectrometer with a Mg Kα X-ray source (1253.6 eV). The energy scale was calibrated using Ag 3d₅/₂ (367.7 eV) and Au 4f₇/₂ (83.8 eV). The C 1s peak from adventitious carbon (285.0 eV) was used as an internal reference for charge correction. The base pressure in the analysis chamber was 7×10⁻⁷ Pa. Depth profiling was performed using Ar⁺ ions with an energy of 800 eV and an ion current of 12 μA. The ion beam was rastered over an area larger than the sample to ensure uniform sputtering. The sputtering rate was calibrated using a standard Ta₂O₅ film and estimated to be about 2 nm/min for the oxide film.
For data analysis, I deconvoluted the Mg 2p and O 1s peaks to separate contributions from metallic Mg, oxidized Mg, and different oxygen species (oxide vs. hydroxide). The metallic Mg peak was obtained from a clean magnesium surface after sputtering away the oxide layer. By using the peak height-to-area ratio of this standard, I could subtract the metallic contribution from the total Mg 2p envelope and isolate the oxidized component. Similarly, the O 1s peak was decomposed into two components: one corresponding to O²⁻ in MgO (binding energy ~530.5 eV) and another corresponding to OH⁻ in Mg(OH)₂ (binding energy ~532.5 eV). The chemical shift between these two species was about 2.0 eV, consistent with literature values for NiO/Ni(OH)₂ systems.
Results and Discussion
XPS Depth Profile of Magnesium and Oxygen
Figure 1 shows representative Mg 2p and O 1s spectra at three different depths (surface, 2 μm, and 5 μm) after Ar⁺ sputtering. As the depth increased, the Mg 2p peak shifted toward lower binding energy, indicating a progressive decrease in the fraction of oxidized magnesium and an increase in metallic magnesium. This shift clearly demonstrates that the oxide film has a layered structure with a highly oxidized outer layer and a metal-rich inner region.
In the O 1s spectra, the peak shape also changed with depth. At the surface, a broad asymmetric peak was observed, which could be deconvoluted into two components: the O²⁻ peak at ~530.5 eV and the OH⁻ peak at ~532.5 eV. Deeper into the film, the OH⁻ component diminished, and the O 1s peak became more symmetric, resembling that of pure MgO. This trend indicates that the outermost layer is enriched in hydroxide, while the intermediate layer consists mainly of oxide.
Elemental Molar Fraction Profiles
I calculated the atomic concentrations of Mg and O from the integrated peak areas after applying sensitivity factors. The results, plotted as a function of sputtering depth, are summarized in Table I. The oxygen molar fraction decreased rapidly from about 60% at the surface to ~20% at a depth of 2 μm, then remained nearly constant at ~10% for depths greater than 5 μm. This residual oxygen level was attributed to adsorption from the ultrahigh vacuum environment, as no significant oxide signal was detected after prolonged sputtering. Therefore, I subtracted this baseline (10 at.%) to obtain the true oxygen concentration in the oxide film.
| Depth (μm) | Mg total (at.%) | O total (at.%) | Mg oxidized (at.%) | Mg metallic (at.%) |
|---|---|---|---|---|
| 0 (surface) | 40.2 | 59.8 | 38.5 | 1.7 |
| 1 | 45.3 | 54.7 | 35.1 | 10.2 |
| 2 | 65.8 | 34.2 | 20.6 | 45.2 |
| 3 | 78.4 | 21.6 | 11.3 | 67.1 |
| 5 | 87.2 | 12.8 | 5.4 | 81.8 |
The molar ratio of oxygen to oxidized magnesium, denoted as $$R = \frac{x(\mathrm{O})}{x(\mathrm{Mg_{ox}})}$$, was calculated for each depth and is plotted in Figure 3 (not shown in text, but discussed here). At the surface, R was approximately 2.0, suggesting that the dominant species is Mg(OH)₂, which has an O:Mg ratio of 2. At a depth of 1–2 μm, R decreased to about 1.0, indicating the presence of MgO (O:Mg = 1). Below 5 μm, R dropped below 0.5, reflecting the transition zone where only a small amount of oxide is dispersed in the metallic matrix. Based on these observations, I divided the oxide film into three distinct regions:
- Outer layer (0–0.5 μm): dominated by Mg(OH)₂ with some MgO; molar ratio R ~ 1.8–2.1.
- Intermediate layer (0.5–2 μm): primarily MgO; R ~ 1.0–1.2.
- Transition layer (2–5 μm): mixture of metallic Mg and small amounts of MgO; R decreases from 1.0 to ~0.3.
These three regions together constitute the protective film. The outer and intermediate layers are fully oxidized and act as a dense barrier against further oxidation. The transition layer represents the diffusion zone where oxygen has penetrated into the substrate during film formation.
Carbon Infiltration and Its Role
An important feature of the films produced in dry graphite-sand moulds is the significant carbon content. Figure 6 (not shown in text) from my original study indicated that carbon concentration in the surface film can reach up to 15 at.%. This carbon originates from two sources: (i) the reaction between molten magnesium and CO₂ gas present in the mould atmosphere, and (ii) solid-state diffusion of carbon from the graphite particles in the mould sand into the casting surface. The reaction proceeds as follows:
$$\mathrm{CO_2 + 2Mg \rightarrow 2MgO + C}$$
The resulting carbon is in an amorphous or graphitic form, which is incorporated into the growing oxide film. This carbon infiltration increases the density of the film by filling microporosities and grain boundaries, thereby reducing the diffusivity of oxygen through the layer. As a result, the oxide film on magnesium alloy castings produced in graphite-sand moulds exhibits superior protective performance compared to films formed in conventional silica sand moulds without graphite.
To quantify the effect of carbon, I performed additional XPS measurements on a sample after sputtering to a depth of 0.5 μm. The C 1s spectrum showed two components: a major peak at 284.8 eV (adventitious carbon and graphitic carbon) and a minor peak at 286.5 eV (C–O bonds). The carbon content decreased with depth, but remained detectable even at 2 μm, indicating that carbon is not only present on the surface but also embedded within the oxide layer.
The Protective Mechanism
Based on my results, I propose the following model for the formation and protection of the oxide film in magnesium alloy sand casting foundry using dry graphite-sand moulds. When molten magnesium fills the mould cavity, it first reacts with oxygen and water vapor in the interstitial air, forming a thin layer of MgO and Mg(OH)₂. Simultaneously, CO₂ released from the decomposition of graphite or from the mould atmosphere reacts with magnesium to produce additional MgO and free carbon. This carbon becomes trapped in the growing film, filling voids and creating a more impermeable structure. The resulting film is a composite of MgO, Mg(OH)₂, and amorphous carbon. The outermost hydroxide layer provides initial passivation, while the underlying oxide layer with embedded carbon provides long-term protection by blocking oxygen diffusion.
The thickness of the protective barrier (outer + intermediate layers) was estimated to be about 2–3 μm, based on the depth profile where the oxide fraction dropped below 50 at.%. This is thicker than the typical oxide films formed on magnesium alloys in ambient air (typically < 1 μm) due to the continuous supply of carbon from the mould. The transition layer extends to a depth of about 5 μm, where residual oxygen is present as isolated MgO particles within the metal matrix. The short “oxygen tail” (compared to films formed in sand without graphite) indicates that the barrier layer effectively limited oxygen inward diffusion, confirming the enhanced protective capability.
Quantitative Summary of Film Composition
Table II presents the average composition of each layer, as determined from XPS depth profiling and peak deconvolution.
| Layer | Thickness (μm) | Main components | Approximate formula | O:Mg ratio | Carbon content (at.%) |
|---|---|---|---|---|---|
| Outer | 0–0.5 | Mg(OH)₂, MgO | Mg(OH)2 + MgO | 1.8–2.1 | 10–15 |
| Intermediate | 0.5–2.0 | MgO, C | MgO + C | 1.0–1.2 | 5–10 |
| Transition | 2.0–5.0 | Mg (metal), MgO | Mg + 少量 MgO | < 1.0 | < 3 |
The overall thickness of the protective barrier (outer + intermediate) is approximately 2 μm. The carbon content in the barrier layer is sufficiently high to significantly reduce porosity. I also calculated the effective diffusion coefficient of oxygen through the film using a simple steady-state model:
$$\frac{dC}{dt} = D \frac{d^2C}{dx^2}$$
Assuming a linear concentration profile across the barrier layer of thickness L, the flux of oxygen, J, is given by Fick’s first law:
$$J = -D \frac{\Delta C}{L}$$
Using the measured oxygen gradient from the surface to the transition layer, I estimated that D in the carbon-containing film is about an order of magnitude lower than that in pure MgO films. This confirms the barrier-enhancing effect of carbon.
Comparison with Conventional Sand Casting
In conventional magnesium alloy sand casting foundry practices (without graphite), the oxide film formed in the presence of protective gases like SF₆ typically consists of MgF₂ or MgO with high porosity, leading to poor protection and requiring thick coatings. In contrast, the graphite-sand mould method produces a dense composite film with self-healing capability due to the continuous formation of carbon. This allows for castings with clean surfaces and minimal oxide inclusions.
Table III compares key characteristics of oxide films formed under different mould conditions, based on literature data and my own experiments.
| Mould type | Protective agent | Film composition | Film thickness (μm) | Carbon content | Protective performance |
|---|---|---|---|---|---|
| Silica sand | SF₆ + CO₂ | MgF₂, MgO | 1–2 | None | Moderate |
| Dry graphite-sand | None | MgO, Mg(OH)₂, C | 2–3 | 5–15 at.% | Excellent |
| Silica sand (no additive) | None | MgO | <0.5 | Trace | Poor |
The data clearly demonstrate that the graphite-sand mould enables a self-protective film without external additives, which is a significant advancement for environmentally friendly magnesium alloy sand casting foundry operations.
Conclusion
My investigation of the surface oxide film formed on ZM5 magnesium alloy castings produced in dry graphite-sand moulds has revealed the following key points:
- The oxide film exhibits a well-defined layered structure: an outer layer rich in Mg(OH)₂, an intermediate layer composed mainly of MgO, and a transition layer consisting of metallic magnesium with dispersed MgO particles. The combined thickness of the outer and intermediate layers (the barrier) is about 2–3 μm.
- Carbon atoms infiltrate the film during casting through the reaction of magnesium with CO₂ and diffusion from the graphite sand. The carbon content reaches up to 15 at.% in the outer layer and remains significant in the intermediate layer.
- The presence of carbon increases the density of the oxide film by filling microvoids, thereby reducing oxygen diffusivity and enhancing the protective nature of the film. This allows the graphite-sand mould process to produce sound castings without the use of conventional anti-ignition additives.
- The oxygen concentration profile across the film, as measured by XPS, shows a steep decline in the barrier region and a short tail in the transition layer, confirming the effective sealing action of the carbon-incorporated composite film.
These findings provide a scientific basis for the development of greener magnesium alloy sand casting foundry technologies, eliminating harmful emissions while maintaining high casting quality. Future work should focus on optimizing the graphite content and mould temperature to further refine the film structure and extend the method to other magnesium alloy compositions.
In summary, the combination of XPS depth profiling and chemical state analysis has enabled a detailed understanding of the protective mechanism in this innovative casting process, demonstrating that carbon plays a crucial role in forming a dense, self-healing oxide film that ensures the success of magnesium alloy sand casting foundry without protective gases.