Exploring the Elemental Interactions of Melamine with Binder–Metal Powder Mixtures: A Pathway to Enhanced Catalytic Debinding and Rheological Control

Abstract

In advanced powder metallurgy technologies such as metal injection molding (MIM) and material extrusion technology (MEX), the intricate synergy between binder components and metallic powders within feedstocks is predominant, dictating the technological boundaries of metal’s complexity, size, and thickness. This work suggests unique binder constituents (polymers, surfactants and melamine), aiming to boost the debinding efficacy and solid loading while ensuring processability. The interactions and role of melamine by thermal and rheological studies are detailed, spotlighting how, unlike traditional binder systems, the pioneering feedstock introduces beneficial modifications to storage modulus (G′) and loss tangent (tan δ), alongside a lubricating effect on metallic mixtures. This study highlights the potential of melamine to facilitate a more efficient debinding process, with superior formaldehyde management and environmental control. Through this material-centric lens, we offer new pathways to navigate the complexities of MIM and MEX, advancing towards technological enhancement and environmental protection.

1. Introduction

Metal injection molding (MIM) enables the manufacture of economic metallic complex geometries of small size (below 200 g) at high production volume (around 20,000 parts/year), covering a niche that is still unrivalled by other technologies including additive manufacturing. Today, it is a mature technology in the market in the midst of a steady growth stage. This near net-shape processing method uses high solid loading polymers (such mixtures are named feedstocks and contain the desired metallic powder with an organic binder in the adequate ratio, composed of polymers and additives) that are injected into the needed shape. The injected part then undergoes thermal or catalytic processes (binder removal) and is sintered under thermal cycles below the melting point of the metal (consolidation of metallic particles). This post-processing involves time-consuming steps, and the success of the entire chain primarily depends on the quality and rheology of the feedstock. In Europe, the MIM demand is growing, and this market is present in strategic sectors such as automotive, tools, medicine, aeronautics, defense and even electronics, having launched since 2020, probably due to the pandemic situation [1]. Interestingly, MIM and indirect additive manufacturing (AM) based on printable filaments or feedstocks (MEX, materials extrusion) show advantageous similarities, and advances in the development of such hybrid feedstocks has been reported elsewhere [2,3]. Thus, these technologies entered the scene to potentially cover the short series production or even MIM parts prototyping and gained tremendous momentum, having a significant impact across various industries in the past decade. In fact, filament-based MEX was considered the indirect technology holding the lowest time until industrial use and one of the highest maturity indexes [4].

However, metallic components consumers are becoming more and more ambitious regarding the size and complexity of the component, pushing the limits of both technologies, which sometimes can lead to inefficient binder release, loss of dimensional control, internal porosity or bubbles and crack defects. To face the current challenges this fact brings to MIM and AM shaping, exhaustive control over the feedstock robustness, rheology and reliability should be addressed to accomplish an injectable material that can be subsequently debound and sintered. Moreover, in the case of the catalytic debinding reaction, the polyacetal (POM) component of the binder is chemically degraded under an acidic atmosphere, releasing formaldehyde, which should be environmentally managed due to its toxicity and explosive character. Previous efforts to improve the catalytic reaction and the debinding rate in this stage were made in terms of the set-up conditions. For example, variations in the catalyst concentration and the flow rate of the gas carrier within the oven debinding zone allow optimization of the process [5]. Regarding the feedstock composition, the tendency is to move towards lowering viscosity by using lower molecular weight polymers, lowering the processing temperature or formulating water-soluble based binders that could contribute to a greener solution avoiding organic solvents [6,7]. Nevertheless, there is a scarcity of literature exploring chemical modifications within the binder, specifically targeting the enhancement of the debinding rate in a more environmentally conscious metallic manufacturing process.

This study highlights that crafting injectable feedstocks with a high powder content transcends the mere adjustment of binder composition (backbones, flow adjusters and surfactants) to ensure pseudoplastic behavior and suitable viscosity under injection conditions. Moreover, it explores the intricacies of interactions between polymer components and metal powder, underlining the necessity of a detailed examination of both rheological and thermal parameters. These complex interactions imply significant changes throughout the entire process, necessitating a focus on material analysis that extends beyond the standard considerations typically valued in MIM process studies. This work aims to accelerate the POM acidic degradation during the catalytic debinding by trapping the released formaldehyde, as well as to minimize risks in the production line due to such a volatile molecule, creating a more environmentally friendly and sustainable strategy. According to Le Chatelier’s Principle, a system whose conditions of concentration, pressure or temperature are changed will move to cope with the equilibrium disturbance. In this sense, capturing the formaldehyde through a side reaction with a nucleophile is expected to be the driving force, yielding more POM acid decomposition to the formaldehyde production and leading to a more effective and faster debinding process when manufacturing metallic parts. In this work, the use of melamine (C₃H₆N₆), being the nucleophile acting as the formaldehyde capturer, is suggested. Other authors have also used this concept of Le Chatelier’s Principle in their research but have not applied it to MIM and AM feedstocks up to now [8,9].

As previously reported by other authors [10], it was demonstrated that formaldehyde released after POM acid degradation forms a resin by reacting with melamine under specific conditions (acid medium, high-volume super-hermetic capsules, reaction monitored by DSC and TGA studies), proving the concept. However, this reaction was only tested in isolated systems, without the presence of the metallic powder and the rest of the binder. To transfer the approach to the real MIM or AM process for metallic component production, it is necessary to study the blending process and the rheological properties of the complete materials containing all the components. This is key to ensuring the appropriate performance of the feedstocks in the processing chain (injection/3D printing, debinding and sintering). The viscosity and rheological parameters analysis when melamine is incorporated will bring insight into the design of injectable feedstocks that could improve advanced powder metallurgy technologies, such improvement being the main objective of this work. This approach underscores our commitment to comprehensive analysis, aimed at elucidating the complex interactions underlying raw material behavior, thereby informing both formulation and processing strategies.

2. Materials

First, to establish a basis for our raw materials, DSC studies will investigate mixtures of melamine and other binder components, with the aim of deepening our understanding of how melamine incorporation affects the processability of the composite. This analytical direction is fundamental as it focuses on the interaction between melamine and binder components and their collective influence on the properties of the raw material. Secondly, the evolution of torque during the mixing process for feedstock manufacture will be explored, providing insight into the mechanical dynamics. Then, the microstructure and rheological properties of the raw materials will be discussed, emphasizing a broad analytical scope that considers both the physical and chemical facets of the raw material formulation. This raw material formulation refers to determining the starting feedstock’s composition or main components, which were selected according to a previous development [11]: metallic powder FeNi8 (89.0 wt.%), POM (9.6 wt.%), HDPE (1.2 wt.%) and stearic acid (0.2 wt.%). This formulation is typically found in catalytic MIM feedstocks [12], where the polyacetal or polyoxymethylene (POM) is the main binder component providing the catalytic character to the feedstock, while HDPE acts as the backbone contributing to the brown-part mechanical resistance. The powder employed is a FeNi8 alloy (d10 = 2.36 µm, d50 = 4.35 µm, d90 = 8.52 µm) with automotive, consumer goods, defense and wear and hardness applications. This formulation will be named as FA hereafter (feedstock F with an acid-based surfactant A), and it will be used as the reference feedstock for comparison with the rest of the materials described in the next section.

3. Results and Discussion

3.1. Surfactant Selection

For an adequate surfactant selection and feedstock formulation, the presence of melamine should be considered. Surfactants in the feedstock are key components to ensure the compatibility between the binder and the powder and minimize segregation in the feedstocks. Furthermore, side reactions with other binder components should be avoided. For this reason, the interactions between the surfactant and the melamine (1:1 ratio) have been firstly analyzed by DSC measurements (TA Instruments, DSC25). Results with stearic acid as the surfactant (named as A), an acid-based surfactant typically used in MIM, are shown in Figure 1a, where the endothermic peak of stearic acid (A) becomes two endothermic peaks in the mixture with melamine (MM), as can be seen in the MM+A vs. A curves, meaning that an acid–base chemical reaction has occurred. However, no significant changes between the surfactant and the mixture peaks are observed if an ester-based surfactant (named as E) is used (Figure 1b: surfactant E vs. mixture MM+E curves). This latter surfactant was preferably used, as it is expected to keep melamine activity in the following processing steps, while surfactant A is detrimental to helping the subsequent melamine–formaldehyde reaction, in addition to losing its powder wettability function.

The interaction of melamine with the rest of the binder components (POM and HDPE) has been also studied by DSC in a 1:1 ratio (Figure 1c,d). This will help us to further understand possible changes in the flow behavior of mixtures at the typical processing temperatures during the injection stage (190 °C), which will be commented on later. The melting enthalpy (

Table 1. Melting enthalpy and melting temperature of the pure polymers in the binder (HDPE and POM) and mixtures with melamine (MM) determined by DSC.

	∆Hm (J/g)	Tm (°C)
HDPE	176.6	133.8
MM + HDPE	54.6	133.1
POM	143.9	165.0
MM + POM	116.7	166.5

) found in the endothermic peaks of both pure polymers (HDPE and POM) are reduced when they are mixed with the melamine, this effect being more pronounced in the case of HDPE. Although the melting temperature is maintained in both cases, between the pure polymer and the melamine mixtures, this reduction of ∆H_m is related to lower energy required to produce the melting process in the presence of melamine, which was reported before [13]. The peak narrowing of MM+HDPE and MM+POM compared to the pure polymers also demonstrates this effect (Figure 1c,d). The differences observed between the HDPE and POM mixtures can be encountered in the fact that melamine interacts more efficiently with POM, presumably by H bonding, through N–H and O functional groups, respectively.

It should be noted that melamine presents no peaks during the heating cycle up to 300 °C, which leads us to conclude that it is acting as an extra solid filler to the polymer, in the same manner as the metallic powder.

In the following section, all the components having been selected and thermally characterized, the feedstock’s fabrication and its characterization are described. Together with the FA feedstock described in the Materials section employed as the reference feedstock, three other feedstocks were prepared in this study: FE, containing the ester-based surfactant E; FA-MM, containing surfactant A and melamine; and FE-MM, containing surfactant E and melamine (see

Table 2. Components of the four feedstocks prepared.

Feedstock Name	Metallic Powder	Binder		Melamine
		Polymers	Surfactant
FA	FeNi8	POM+HDPE	Acid-based	No
FE	FeNi8	POM+HDPE	Ester-based	No
FA-MM	FeNi8	POM+HDPE	Acid-based	Yes
FE-MM	FeNi8	POM+HDPE	Ester-based	Yes

). The concentration of FeNi8 powder, POM and HDPE was kept as in FA. In the melamine-added feedstocks (FA-MM and FE-MM), the concentration (wt.%) of melamine was 2.77 times lower than that of POM, according to a study reported elsewhere [10].

3.2. Feedstock’s Fabrication and Its Microstructural, Thermal and Rheological Properties

All the components are mixed in a double-rotor mixer (ThermoHaake, Waltham, MA, USA, Haake Rheocord 252p) at 190 °C to gently homogenize the powder with the binder (FA and FE), and with the binder and the melamine (FA-MM and FE-MM), leading to the four feedstocks produced for this study, as described in Section 3.1. The evolution of the mixing torque of the four feedstocks during the process is shown in Figure 2. After the addition of all the components in each case (from minute 12), the mixing process continues, torque values tend to equilibrate up to the final stabilization time (considered at minute 90) and the resulting feedstocks are removed from the mixer and pelletized. Further mixing time could lead to a steady-state torque value; however, it could also produce the thermal degradation of some binder components and a sudden increase of torque, so a balance at 90 min was achieved. Feedstocks FE and FE-MM show higher torque values after 90 min mixing (8.01 ± 0.06 N·m and 8.37 ± 0.07 N·m, respectively) than feedstocks FA and FA-MM (5.62 ± 0.05 N·m and 6.82 ± 0.04 N·m, respectively). However, the torque evolution observed in Figure 2 shows an increasing tendency in the stabilization of FA and FA-MM feedstocks, while a decreasing evolution in the case of surfactant E. This latter behavior is preferred, since it means that the surfactant is more effective in the homogenization process between powder, binder and the melamine additive, if there is any. Nevertheless, in all cases, the melamine is well homogenized in the feedstocks, as can be seen in the SEM micrographs (JEOL JSM-6610LV, Tokyo, Japan) displayed in Figure 2.

It should be mentioned that the melamine powder shows agglomerates of large particle size up to 50 µm (Figure 2a) that are subsequently broken during mixing into the double-rotor machine, which cannot be detected in the feedstocks’ micrographs (see SEM micrographs of FA, FA-MM, FE, FE-MM in Figure 2b–e). Since MM melting does not occur in the range up to 200 °C, it is hypothesized that homogenization is explained by the wear effect of the powder particles breaking the MM particles apart, favoring the blending with other binder components and powders. Thus, feedstocks exhibit strong homogenization both before and after the incorporation of melamine, confirming that the new surfactant E has a successful role in the feedstock.

The torque monitorization is another way to predict the viscosity of the mixtures, which are more viscous at higher values. It is worth noting that the introduction of melamine hardly produces any effect on the FE and FE-MM feedstock properties during the blending process, as can be seen in the torque evolution and torque values, meaning that those mixtures exhibit similar viscosity (Figure 2). The higher torque value exhibited by FA-MM when compared to FA can be associated with the loss of the wettability effect in the case of using A (acid-based), as was previously predicted by DSC studies gathered in Figure 1, which affects the rheology of the mixture and causes the torque and viscosity to increase.

The thermal properties of the feedstocks were studied by thermogravimetric analysis (TGA) (Labsys EVO Setaram Instrumentation, Caluire-et-Cuire, France), and the degradation curves with temperature are shown in Figure 3. In the feedstocks FA and FE, the mass losses at 350 °C and 440 °C correspond to POM and HDPE polymers in the binder, respectively. The surfactant decomposition is not detected by this technique due to the low concentration in the mixture. Surfactant E (feedstock FE) slightly advances the start of the first mass fall corresponding to POM. More remarkably, in the feedstocks FA-MM and FE-MM, melamine decomposes at around 320 °C and this produces the thermal degradation of POM shifting to higher temperatures, not affecting HDPE profile. The mass change reaches lower values (82.94 and 83.52 wt.%, respectively) since the melamine content contributes to the overall amount decomposed in these feedstocks, while keeping the same binder ratio as in FA and FE.

Once the melamine is part of feedstocks FA-MM and FE-MM as an additive, it will be kept inactive during the injection or 3D printing process of the parts, playing its role only during the debinding stage. Nevertheless, melamine is expected to affect the flow properties of the new feedstocks due to the solid–molten interactions occurring (melting of melamine is not occurring at the injection temperature (190 °C), as confirmed by DSC in Figure 1, MM). For this reason, the rheological study performed in all the feedstocks aims to understand and predict the behavior of the modified materials in the MIM or AM process.

First, viscosity measurements and rheological parameter calculations were performed using a capillary rheometer (Dynisco LCR7000, Franklin, MA, USA) at temperatures of 190, 200 and 210 °C, passing the molten feedstock through a capillary with a length/diameter index of 30 and in the high shear rate range (10–7000 s⁻¹). The apparent viscosity of feedstocks with both surfactants A and E containing the melamine (176 Pa·s for both FA-MM and FE-MM) is lower than that of the pure feedstocks (342 Pa·s for FA and 433 Pa·s for FE feedstocks), values measured at 190 °C and 1000 s⁻¹. Thus, the yield stress parameter is also lower under the melamine effect (defined as the effort required to start flowing), as can be seen in Figure 3. Both facts might be explained by a better polymer chain alignment in the shear direction, which is enhanced by the supramolecular interactions created between the functional groups of the binder components, the surfactant and the melamine and by the extrusion that occurs in the capillary rheometer, helping the flow process [14].

Another important parameter is the activation energy (Figure 4), which refers to the influence of temperature on the feedstock; the lower the value, the more stable the material remains during the injection or the extrusion process in the 3D printing machine. According to the results, melamine does not have a significant influence on the final value when surfactant A is employed (around 10 kJ/mol in FA and FA-MM). However, the activation energy notably increases with surfactant E in FE feedstock, close to the threshold value considered in PIM technology (40 kJ/mol) [12]. Therefore, surfactant E introduces a higher sensitivity to temperature in feedstock FE compared to conventional FA. Since the yield stress values in this material are high too, it is deduced that the admissible solid loading (metallic powder concentration in the binder) of the FE system is lower compared to FA. More remarkable is the fact that, if melamine and surfactant E are present (FE-MM), both activation energy and yield stress dramatically decrease. It can be concluded that the combination of melamine and surfactant E leads to an increase in the solid loading admissible in the mixture, which is an important parameter in the feedstocks to ensure good sinterability. This can be explained by better binder–powder interactions due to melamine acting as a second surfactant.

Secondly, an oscillatory rheometer (TA Instruments, New Castle, DE, USA, Discovery HR20) was employed to study the behavior of the feedstocks in the range 190–210 °C and at lower shear rates (down to 10⁻⁷ s⁻¹), selecting parallel plates with 25 mm diameter. In this way, both low and high shear rates were covered to better understand the rheology of the mixtures. Additionally, for comparison reasons, mixtures of the neat binder with and without the melamine were also prepared and analyzed by oscillatory rheology, to further study the behavior of the melamine additive in the polymeric binder without the influence of the metallic powder.

The results regarding the amplitude-sweep oscillatory rheology are given in

Table 3. Rheological parameters by amplitude–strain-sweep oscillatory measurements: elastic limit (τ_y), elastic moduli (G′_y) and loss tangent (tan (δ)) for both the neat binders and the feedstocks (Binder/Feedstock).

	Binder FA/FA	Binder FE/FE	Binder FA-MM/FA-MM	Binder FE-MM/FE-MM
τy [%]	39.6/0.01	39.6/0.01	0.1/0.01	0.1/0.02
G′y [Pa/105Pa]	87.0/159.6	86.2/105.2	601.6/42.8	527.0/66.4
tan (δ)	11.6/0.2	13.3/0.3	5.5/0.5	5.5/0.4

for the neat binders (i.e., without the metallic powder) and for the feedstocks (see Table 1 for the feedstocks’ composition). Studying the neat binder, the presence of melamine reduces the linear viscoelastic region (LVE) and the elastic limit (τ_y), and it produces the increase of the elastic moduli (G′), which is due to the reinforcement effect of the solid particles (lower and higher values of τ_y and G′ in FA-MM and FE-MM systems, respectively). The reduction of loss tangent (tan (δ)) when the melamine is added indicates that the elastic contribution increases, although it remains above 1 (a situation in which the loss modulus G″ is higher than the storage modulus G′ or G″/G′ > 1). However, if the feedstock is analyzed, the LVE is not affected by the presence of melamine (the τ_y value remains very low due to the metallic powder), while it unexpectedly reduces G′_y and increases tan (δ) (FA-MM and FE-MM). Thus, the fluid becomes less viscous because of the lubricating effect of melamine in these mixtures. Comparing the role of the surfactant in the feedstocks, the increase of tan (δ) is more significant with A than with E, which might be related to the interactions created between melamine and the surfactant. As mentioned before, the acid-based surfactant A may form a covalent bond with melamine, resulting in a better integration of the solid additive in the molten fluid and, therefore, more severe viscosity reduction. In the case of surfactant E, the increase of tan (δ) is moderate, so it provides lower fluidity to the solid additive, maybe due to the non-covalent H-bonding interactions formed, equivalent to a crosslinking effect.

Concerning the frequency-sweep oscillatory tests, the results are gathered in Figure 5. If the binders are analyzed separately (Figure 5a), only the FE-MM mixture exhibits G′-G″ intersection at 10 rad/s. The other binder mixtures show G″ > G′ in all the frequency ranges studied, even when melamine is incorporated (FA-MM). Thus, the addition of melamine to the mixture containing surfactant E changes it into a viscoelastic solid in the lowest frequency values, while in the case of surfactant A, it remains a viscoelastic liquid during the entire frequency range. This is in accordance with the values of tan (δ) previously shown, confirming that melamine is creating a three-dimensional network with surfactant E. In contrast to the traditional surfactant A (acid-based), the use of E when designing the feedstock prevents the melamine from covalently reacting, losing its functional groups. We can conclude that, although this reaction is beneficial from the rheology point of view since it diminishes the effect of the solid additive in the binder, it is not intended for the purpose of the global study, in which keeping the melamine in its nucleophile form is desirable for the next processing steps of the subsequent feedstocks.

When the metallic powder is involved (Figure 5b), in all cases the feedstocks show a solid-like viscoelastic behavior, with G′ > G″ in the whole frequency range. The intersection between G′ and G″ curves does not occur even at the lowest frequency values, indicating that the powder–binder interactions dominate and the particles orientate in the shear direction, hindering the previously observed melamine–surfactant organization in the case of the binders shown in Figure 5a. Comparing the results according to the surfactant employed, surfactant E lowers the difference between G′ and G″, confirming that the presence of both E and melamine leads to a predominant behavior of lower viscoelasticity compared to A and melamine. This is beneficial for the feedstock rheological properties since the solid particles embedded in the polymer matrix (both the melamine and the metallic powder) are causing a reduced effect in the case of surfactant E.

Finally, the complex viscosity (η*) calculated at low shear rates by oscillatory rheometry and the apparent viscosity (η_ap) calculated at high shear rates by capillary rheometry are displayed together in Figure 6. The viscosity decrease upon shear rate increase corroborates the shear thinning character of this kind of mixture, even if the melamine additive is incorporated, as well as in the E-containing feedstocks. At low shear rates (1 × 10⁻⁷–1 × 10⁻¹ s⁻¹ in Figure 6), the presence of melamine does not change the viscosity behavior, while the most influencing factor is the use of surfactant A or E, meaning that the powder–binder interactions are dominating. Conversely, it is at high shear rates (1 × 10¹–1 × 10⁴ s⁻¹ in Figure 6) when the presence of melamine is critical. Apparently, melamine favors polymer chain alignment and powder–binder compatibility, improving the flow properties of both feedstocks FA-MM and FE-MM.

The values of apparent viscosity achieved at the highest shear rates are in accordance with those recommended for the injection process, below 1000 Pa·s [12], and the curves in both shear rate ranges show good continuity, which is related to the absence of agglomerates in the feedstocks. The viscosity criteria for indirect printing technologies, such as MEX, are under development in the literature, and some differences with MIM might be expected since the shear rates occurring at the printing extrusion are lower (100–500 s⁻¹).

The present study suggests that this alternative feedstock formulation, FE-MM, shows adequate rheological properties to be used as MIM and MEX raw material. Further investigations regarding the behavior during the green-parts shaping, debinding and sintering are in progress.

4. Conclusions

To address the challenges that MIM and MEX technologies are facing today, a comprehensive overview with a complete thermal and rheological characterization of a unique feedstock oriented to catalytic debinding that contains special additives (melamine) is reported. The following conclusions can be extracted:

• Chemical interactions between melamine and the conventional binder components demonstrated a detrimental acid–base reaction to keep the melamine functional groups active. Therefore, an ester-based surfactant (E) was successfully employed, confirming by torque and microstructural studies that the role of E in the binder is preserved upon melamine addition.
• Melamine additive moves the POM thermal decomposition towards higher temperature and notably drops the yield stress and the activation energy rheological parameters, probably due to supramolecular interactions found between the binder, MM and surfactant that enhance the binder–powder compatibility. At low shear rates, it is reckoned that the melamine is not significantly affecting the complex viscosity, and the solid particles embedded in the polymer matrix with surfactant E are causing less of an effect in the rheological properties than with A, since the difference between G′ and G″ is lowered. Unexpectedly, melamine is critical at high shear rates, enhancing the polymer chain alignment under the shear direction, which improves the flow properties, with a lubrication accomplished by this additive in the mixtures.
• Moreover, feedstock FE-MM has been proven to keep the melamine functionality unaltered, to later react with formaldehyde.

Thus, a securer handling of formaldehyde is suggested for this strategy once it is transferred to the postprocessing of the shaped green parts in subsequent work.