An Overview of Highly Porous Titanium Processed via Metal Injection Molding in Combination with the Space Holder Method

Abstract

In the past two decades, titanium foams have attracted greater interest from the biomedical industry due to their excellent chemical and mechanical biocompatibility when used as biomimetic implants. The porous structure plays an important role in bone adhesion to an implant, allowing its growth into the component. Moreover, the voids reduce the elastic modulus, promoting greater compatibility with the bone, avoiding the stress shielding effect. In this regard, metal injection molding is an attractive process for titanium foams manufacturing due to the high microstructural control and the possibility of producing, on a large scale, parts with complex near-net-shaped structures. In this review, recent discoveries and advantages regarding the processing of titanium powders and alloys via metal injection molding combined with the space holder method are presented. This approach can be used to obtain foams with high biocompatibility with the human body at a microstructural, chemical, and mechanical level.

1. Introduction

Bone fractures are a global public health issue, and they impair quality of life and result in health-care costs. In fact, fractures are the most common musculoskeletal disorder, mostly related to wear and aging, especially in people with arthritis and osteoporosis. In many cases, the missing or damaged bone must be replaced by an orthopedic implant [1].

Orthopedic procedures mean large expenses, with the costs of implantable devices accounting for a large proportion of those procedure costs. According to one study [2], the current market for orthopedic implants accounts for around USD 45 bi and is estimated to exceed USD 60 bi in 2028. The rapid and continuous increase in the geriatric population around the globe, who are more prone to orthopedic problems, has been driving the demand for biomaterials. Technological innovations concerning implantable medical devices, robot-assisted tools, and the continuous development of new materials and processing procedures have also supplemented the market growth.

Bone tissue is a polymer–ceramic composite formed by minerals such as hydroxyapatite (Ca₁₀(PO₄)₆(OH)₂). It is found in two forms or types, and both have an anisotropic structure; that is, they have properties that depend on the direction analysis. The first type is the trabecular or cancellous bone, composed of an interconnected porous network of plates and rods arranged in various configurations, forming an open foam with a density around 0.05 to 1.0 g/cm³ and a low elastic modulus (0.76 to 4 GPa) [3], which varies considerably depending on the direction of the stress. The second type is called cortical or compact bone, characterized by having a higher density (1.99 g/cm³), lower porosity, greater metabolic activity [4], and higher elastic modulus (17 to 30 GPa) [5]. Normally, the two forms of bone can be found together in different configurations, giving the bone a combination of characteristics, such as excellent energy absorption capacity of the cancellous tissue and better mechanical resistance acquired from the cortical bone. In some cases, such as severe damage or fractures, the bone needs to be replaced. Successfully replacing this hard tissue with a complex structure that mimics its anisotropic features and mechanical properties can be challenging without the use of biocompatible materials.

In this context, artificial biomaterials have emerged to overcome these problems, with surgical implants restoring the function of compromised structures. A good biomaterial meets the following requirements: suitable mechanical properties (modulus of elasticity, yield strength, fatigue strength, wear and corrosion resistance), biocompatibility, and osteointegration [6]. By definition, biocompatibility is the material’s ability to perform as a good guest in a given application environment that exercises interaction [7]. The biological response is the local and systemic response of the host organism to the implanted biomaterial.

Metals 316L stainless steels (316LSS), Co-Cr alloys, and Co-Cr-Mo alloys were the first artificial biomaterials, but they exhibited two disadvantages: the presence of toxic elements, such as Co [8] and Cr [9], and a high elastic modulus. Pure Ti and Ti-based alloys are currently the most exploited by the biomedical industry due to several characteristics, such as non-toxicity, good biocompatibility, and low density. The first generation of Ti alloys had several disadvantages, including relatively lower wear resistance, lower hardness, and higher stiffness compared with human bones [10]. Stiffness mismatch between the implant materials and human bones causes bone resorption and eventual loosening of the implants [11]. Moreover, alloying elements, such as Al, V, Ni, and Co, present in the Ti alloys, produce toxic effects when they are placed into the human body. Some diseases produced by these elements are dermatitis, Alzheimer’s, neuropathy, and osteomalacia [12,13,14]. The second generation of Ti alloys was produced with non-toxic elements, such as Nb, Mo, Zr, or Ta [15]. Even so, these alloys present problems associated with their mechanical performance.

One of the main difficulties in achieving a very high degree of osteointegration is related to the difference between the elastic modulus of the implant material and the underlying bone structure, as aforementioned. In this sense, the manufacture of porous components through processes such as metal injection molding (MIM) combined with the space holder (SH) method [16], powder compaction [17], and additive manufacturing (AM) [18] has been studied over the years to overcome these limitations. The pores generated from these processing methods, in addition to reducing the elastic modulus of the implant, provide the adherent surface necessary for cell proliferation and adhesion [19], thus avoiding the stress shielding effect. Furthermore, bone growth into the implant is determined by the porosity, pore size, and structure [20], allowing adequate vascularization of cells and fluids [21]. However, good control over the addition of porosity must be performed to achieve a precise balance between the mechanical properties and the biological and adhesion performance of the implant.

MIM has been demonstrated to be an excellent technique to obtain structures with controlled porosity. In addition, MIM combines some desirable characteristics, such as reproducibility, design simplicity, and flexibility in the choice of fitting materials. The manufacturing process has high design freedom, with the possibility of large-scale production in the processing of near-net-shaped parts. Geometrically complex parts, such as human bones, can be obtained by MIM. Furthermore, the MIM process associated with the SH method has great potential for obtaining porous structures with particular properties by creating voids with temporary materials. MIM combined with SH makes it possible to obtain parts with the desired size, shape, and percentage of pores. Several authors have applied the combination of MIM and SH to produce titanium foams with elastic modulus values close to those of bone [22,23,24,25] and pore percentages in the range of 30% to 72% [26,27,28,29,30].

In this context, this paper provides an overview of the advances in the techniques for processing titanium foams via metal injection molding combined with the space holder method. Firstly, we briefly describe the technique, comparing and highlighting the advantages related to other powder metallurgy processing routes [17,18,31]. Second, the article focuses on the advances and suitability of the MIM process combined with the SH method for obtaining highly porous Ti implants. The processing parameters are discussed based on the author’s experience to overcome problems that may affect the integrity of the final component, for example, those related to porosity, contamination, dimensional accuracy, and mechanical properties. Finally, future trends are raised and discussed.

2. The MIM + SH Route Compared to Other Techniques

In a typical metal injection molding (MIM) process (see Figure 1), highly pure Ti powder and alloys are mixed with polymeric binders to form the feedstock that can be injected using a traditional injection molding machine. This allows the manufacturing of geometrically complex parts as the molten polymeric binder assists in transporting the powder towards the mold cavities. To produce components with a high porosity, an additive called a space holder (SH), which is based on volatile structures, such as polymers and salts, is added to the mixture in the form of particles. The shape, size, and amount of SH particles added are the main control parameters and will influence the foam properties.

The quantities of metal, binder, and space holder powders are calculated to produce porous structures with a controlled percentage of pores. The materials are mixed at a temperature slightly above the melting point of the binder (usually in the range of 140 to 170 °C) [32] under a protective atmosphere. The mixing procedure must be controlled such that the metal powder is covered with a thin layer of binder to provide sufficient fluidity to fill the mold during injection. In the next step, the feedstock is granulated in pellets smaller than 3.0 mm, aiming to make easier the flow into the MIM chamber. After injection molding, the binders and SH are removed via chemical and/or thermal processes before or during sintering. If SH is removed using temperature, an inert gas stream must be used. It is extremely important that, for the processing of titanium and its alloys, these final steps are always carried out under vacuum of protective atmospheres, minimizing the contact of the component with oxygen, due to the high affinity of titanium with this element, as well as hydrogen, carbon, and nitrogen.

The last stage of processing metal powders, called sintering, is one of the most sensitive steps. The process requires high temperatures to make the powder a load-bearing component [33], the properties being dependent on temperature, holding time, vacuum level, or protective atmosphere. Titanium parts are generally sintered at temperatures between 1200 and 1350 ºC [34] for periods around 2 h at low pressure (10⁻⁴ mbar) or in high purity argon atmospheres. Controlling the purity of the atmosphere is extremely important at this stage to prevent the component from being exposed to contamination by elements present in atmospheric air or in the binder system, which can impair the final mechanical properties [35,36]. Furthermore, it is at this stage that most of the dimensional variations of the part occur, which can lead to loss of shape or the collapse of the structure [27,29].

MIM has become increasingly popular as a technology for small- and medium-volume near-net-shaped parts for medical devices, implants, and surgical tools. This method allows a reduction in production costs, reduction in waste, no machining, and facilitates the achievement of complex or customized components on a larger scale compared to other conventional powder metallurgy processes [37]. Particularly in the case of Ti, solid-state processing is more interesting than liquid-state foaming techniques, which can accelerate contamination due to the high reactivity of Ti with elements present in atmospheric gases.

Other traditional methods include powder compaction (PC) and additive manufacturing (AM), the latter including several types of prototype techniques with the same additive approach, such as selective laser melting (SLM), 3D printing, and laser metal deposition (LMD) [38], which are based on the application of heat from a laser source to melt the powder layer by layer to produce complex shapes.

Conventional powder compaction by cold pressing is the most common route in powder metallurgy. It is relatively easy to industrialize and provides the consolidation of powder mixtures into a shape determined by the die design. On the other hand, cold pressing has several important restrictions for use as a net-shaped part-forming process for titanium implants. Firstly, a matrix for forming a network of powder metal (PM) parts is often complex in geometry. The die elements must be manufactured from expensive cold-working steel with high precision. As a result, die fabrication becomes very expensive. In combination with often small-scale production, this can result in an economic barrier. Secondly, the titanium powder used in pressing must provide a green strength after compaction that is high enough for subsequent compact handling. Chemically clean gas-atomized titanium powder does not meet this requirement due to its spherical shape. Typically, titanium powder obtained by the hydration, grinding, and dehydration route (hydride–dehydride—HDH powder) is used in the case of matrix compaction. This powder provides sufficient green strength but usually contains an unacceptably high amount of oxygen. Finally, during cold pressing on a die, the powder particles in contact with the die wall become highly deformed and, as a result, a dense skin is formed on the surface of the compact, which can suppress the open porosity structure.

Concerning the additive manufacturing route, 3D parts are manufactured via consecutive melting of powder layers through the application of a high-power laser beam, which provides super-concentrated energy, easily reaching the melting temperature of the powder bed, thus building the components layer by layer. Three-dimensional objects are designed using a computer-aided design. Then, the geometric model is sliced in lines, providing the outline of each layer and the printing path (coordinates).

Depending on the energy source, the process is called selective laser melting (SLM), electron beam melting (EBM), laser metal deposition (LMD), or 3D printing. All the methods are aided by automated computation that guides the beam along pre-programmed tracks. Another way of classifying AM technologies is through the consolidation process, that is, liquid state manufacturing (melting process) or solid-state manufacturing (sintering process). Processes carried out in the liquid state are known as direct as the final part is obtained directly from the 3D printing plate, while, in the solid state, a green part is obtained, which will later be processed and sintered to obtain mechanical consistency.

Through AM technology, highly complex pieces are produced with morphologies designed to facilitate cell proliferation and bone tissue growth. Porosities above 90% with pore sizes above 500 µm can be easily obtained. Despite being promising, AM still has restrictions regarding large-scale production and the high costs of laser sources. Furthermore, compared to the MIM + SH technique, the AM processes are more difficult to industrialize and are relatively expensive and very time-consuming [39]. Keeping a protective layer throughout the part production/sintering process scores unfavorably against the method.

It is also possible to print parts using photoreactive polymers. These binders are deposited onto a build surface to create an object layer by layer. The green strength of the pieces is obtained by exposing the binder to a UV light source, making it rigid. The printed part is then sintered to remove the binder in a protective atmosphere. Even so, the unduly long time to print the pieces puts AM in second place.

Table 1. Comparison of the different techniques for processing highly porous Ti implants.

Method	Advantages	Disadvantages
Conventional powder compaction with space holder (PC–SH)	- High level of porosity (60–80%), with adequate mechanical strength - Easy to industrialize, less expensive, as well as less time-consuming than prototyping techniques (SLM, FDM, or 3D printing) - Less waste of materials	- Randomness of the process and type of SH particle could produce a variation in wall thickness and interconnection size that can deteriorate its mechanical performance - High plastic deformation - Geometry limitation
Metal additive manufacturing (AM)	- Less waste - High geometrical freedom - High precision components - Moderate energy costs - Rapid prototyping and on-site repair	- Expensive equipment - More time-consuming - Molten pool instabilities and higher residual stresses - Higher probability of contamination (for laser-based printing) - Unpredictable properties due to melting and thermal history.
Metal injection molding (MIM–SH)	- Capable of producing both porous and dense small parts - High design flexibility - Large-scale production - Free geometry and design - Low cost	- Reduced part size - Higher initial cost than PC–SH - A quantity of material is removed during processing - 15–20% linear shrinkage during processing

summarizes the aforementioned techniques used to produce porous materials with their advantages and disadvantages.

3. Critical MIM + SH Processing Parameters

3.1. Ti Powders

With regard to the characteristics of the raw material, the best commercially pure Ti powders for the MIM process are those where the particles are spherical, smaller than 45 µm with a non-uniform distribution, and the purity is greater than 99.5% [40]. These powders are preferred due to their better flowability [41], faster densification [42], and desirable characteristics during injection and sintering, respectively.

Concerning the size of the Ti particles, it is important to consider that they should be several times smaller than the SH to improve the sinterability [43]. The use of a fine spherical powder improves the surface finish and increases the apatite formation due to a higher surface energy [26], but, as the size of the particles becomes much smaller, the content of impurities tends to increase [36], which can have harmful effects on the properties of the component. However, finer and purer powders are more difficult to obtain due to the considerably higher price of the raw material. On the other hand, coarser powders or those with a wider particle size distribution can be used to increase the density of the green part and have the advantage of being less susceptible to impurities, but the disadvantages include less flowability and higher susceptibility to distortions. Moreover, in order to avoid substantial linear shrinkage and warping, it is important to use a powder with high packing density since the pressure used in injection molding is relatively low in comparison with other methods, such as press and sinter. Furthermore, low-packing-density powder can result in more micro-pores [42]. Since powder particles constitute the walls of the pores in porous foams, it is very important to choose a powder that results in a low number of micro-pores as these micro-pores can deteriorate the mechanical properties of foams by concentrating stress. Thus, it is preferred to use a powder with a broad particle size range in order to fill those empty spaces among the particles by finer particles, consequently minimizing the possibility of micro-pore formation.

Particle shape also has a great impact on the properties of the foams produced. As mentioned above, spherical high-purity powders are the most desirable for use in MIM. However, their higher cost represents an important disadvantage. The shape of the powder is normally dependent on the production method. Common particle shapes of Ti powders are spherical and angular or irregular. Spherical Ti powders are usually produced by gas atomization. In comparison, angular Ti powders are typically produced by the hydride–dehydride process, and they are relatively cheaper. Recently, researchers have used non-spherical titanium hydrogenated-dehydrogenated (HDH) powders [44,45,46], produced by mechanical milling of sponge, scrap, or ingots [47], and titanium hydride (TiH₂) powders [22,48], which have lower levels of purity and flowability but are acquired at a lower cost. Figure 2 shows the (a) gas-atomized and (b) HDH powder morphologies used by Tuncer et al. [28].

These powders can be used to form the entire volume of the implant or be mixed with better quality powders [49] to allow a reduction in costs. Furthermore, the hydrogen released from TiH₂ during sintering can prevent extra oxygen pickup by the part [50,51]. Among the disadvantages of using non-spherical powders such as HDH is the greater susceptibility to cracks due to the greater amount of stress concentrators [45]. Moreover, other studies have shown that irregular powders produced foams with less open porosity, a desirable characteristic aimed at bone growth and osteointegration; this was attributed to the low packing density of the irregular powder in comparison with the spherical ones [42,52]. Treatments such as induction plasma [53] can be performed to reduce the sharpness of irregular powders, improving volume density and packing capacity [44,54]. In this regard, Güden et al. [45] compared, after sintering, samples cold-compacted with both spherical and angular metal powders and found that the compressive strengths obtained were close to those of human bone due to the high degree of porosity (34–54%). The Ti structures compacted with different powders were very similar, showing close values for the porosity grades. However, the results showed that angular powders result in higher percentage of micro-porosity and poorer mechanical properties in comparison with spherical powders.

Contamination of the initial powder is one of the most important factors affecting the final oxygen content of the implant, a characteristic that can impair the mechanical properties of the part, reducing its toughness and increasing the elastic modulus [36] and the stress shielding effect. Thus, standards have been established regarding the chemical requirements for orthopedic applications [55,56,57]. During the entire processing route, if all the contamination precautions are followed, no step provides more oxygen as a contaminating element than the initial contamination level itself, as shown in the graph in Figure 3.

Although very reactive, titanium powder with a particle size between 1 and 45 µm does not react strongly with air at temperatures below 200 °C, with most of its initial oxygen content coming from its manufacturing or atomization [36]. Therefore, its handling at room temperature does not require protective devices, such as gloveboxes or protective atmospheres. On the other hand, its reactivity increases exponentially at temperatures above 400 °C [36]. It is important to note that, the smaller the particle size, the more reactivity the powder will present due to a larger surface area, increasing the handling risks.

3.2. Space Holders (SH) Method for MIM

The porosity in titanium foams for biomedical implants, as previously mentioned, has the function of reducing the elastic modulus of the implant to values closer to the underlying bone. Moreover, it provides a better cell attachment and proliferation due to the natural porous aspect of the bone [58,59]. Pores can be introduced through several methods, including partial sintering [60], sintering with hollow spheres [61], trapping and gas expansion [62], slip casting [63], tape casting [64], gel casting [65], freeze casting [66], and polymeric impregnation [20]. However, few of these approaches compete with the SH method in controlling the amount and shape of pores, as well as the dimensional control of the parts produced at a low cost. In the SH procedure, the pore formers do not interact with the metal and are easily removed in later steps. Porosity is controlled by the amount and shape of the space support added. The SH is selected based on the following criteria: (a) available sizes and shapes, (b) cost, (c) no reactivity with Ti, (d) low presence of residue, (e) easy processability, (f) lack of toxicity, and (g) resistance to deformation. In general, the SH is removed through contact with solvents or heat.

Table 2. Typical space holders used in the literature and their morphological and removal characteristics.

Material	Particle Size	Removal	Observation	References
Sodium Chloride (NaCl)	200–500 µm	Aqueous solution 50–60 °C for 40 to 72 h	- Good water solubility and low cost - High melting point	[22,28,67]
Potassium Chloride (KCl)	250–500 µm	Aqueous solution 50–60 °C for 24 to 72 h or thermal removal at 750 °C for 2 h	- High solubility in water, available in multiple shapes - Lower melting point	[27,28,29]
Polymethylmethacrylate (PMMA)	D50 = 600 µm	Thermal 200–450 °C for 2 h	- Control of the macropore morphology - Can be used as a binder - May contaminate the powder with C and O	[22,68]
Magnesium	300–1500 µm	Thermal during sintering	- Can be leached with solvents	[69]
Ammonium bicarbonate NH4HCO3	500–800 µm	Thermal 175 °C	- May contaminate the powder with interstitial elements - Easy and complete removable due to moderate decomposition temperature	[26,70]
Tapioca Starch	100–400 μm	Aqueous solution or in a furnace at 450 °C	- Low cost - Easy access - Easy removal	[71]

includes the typical space holders used in combination with the MIM process found in the literature.

It can be noted that the salt-based SHs (NaCl and KCl) can be removed at lower temperatures and do not lead to contamination of the titanium. Moreover, they are inexpensive and easy to obtain. On the other hand, longer times (24 to 72 h) are required to remove these SHs in solvents. Tuncer et al. [28] tested NaCl with cubic morphology and KCl with rounded particles (see Figure 4a,b). The researchers observed that KCl offers several advantages over NaCl in terms of ease of removal and fluidity due to the shape of the particles. Morelli et al. [22] compared NaCl and polymethylmethacrylate (PMMA) to SHs. The authors observed better shape retention of the green material in the case of PMMA. In addition, the foams produced with PMMA showed higher porosity when using the same volume of SH. Contamination was not measured.

The particle size and shape and the volume of the SH used are directly linked to the shape, size, and percentage of the pores [23]. Figure 5 shows a water-debinded sample illustrating the cubic-shaped pores, the structural binder, and the Ti particles. The images verify that the KCl space holder was dissolved in water and the pores retained the shape of the space holder particles, while the polyethylene glycol (PEG) and polymethyl-methacrylate (PMMA) networks of the binder system are still apparent after 23 h of dissolution at 20 °C. This study was carried out by Shbeh and Goodall [72], who successfully obtained Ti foams with 55% of porosity and low contamination.

Figure 6, taken from Shbeh et al. [30], shows the dependence of the amount of porosity, volume shrinkage, and mechanical strength on the percentage of SH feedstock. It can be seen in Figure 6a that, the higher the SH content, the greater the porosity. However, this relationship assumes a logarithmic behavior, with a large increase in porosity at lower SH contents. For contents higher than 50% SH, there is little porosity gain, and the curve becomes asymptote. The authors attributed this behavior to the greater volume reduction in the samples (Figure 6b) with increasing content of SH since the Ti particles need to travel greater distances during sintering to initiate atomic bonds with each other and form cell walls. These movements result in greater volume shrinkage of the samples. Figure 6c shows the dependence of compressive strength on porosity. The stress–strain curves for foams made with a high porosity, especially those obtained with 52% and 60% of SH (by volume), showed a much longer plateau region before densification compared to those with a low porosity. The foams with a high porosity have a more uniform porous structure and undergo more uniform deformation until densification, while those with a lower porosity reach densification earlier. In the same study, Shbeh et al. [30] used different amounts, shapes, and sizes of the KCl as a space holder to evaluate the pore morphology in titanium foams. The impact of the shape of the space holder on the final porosity was not significant. On the other hand, as shown in Figure 6d, the shape of the SH particles has a great influence on the mechanical properties of the foams produced. The Ti foams produced with the cubic SH showed lower yield stress than those made with the spherical SH, probably due to the stress concentration at the corners of the cubic pores. The particle size and shape of the SH also influences the flowability of the material during MIM processing and the preparation of the powder mixture, which can lead to clogging or poor filling of the mold due to the low pressure and temperatures used in MIM.

Therefore, the proportion of titanium powder, SH, and binder must be adjusted so that the part develops the desired porosity as well as maintains its dimensions throughout the process, minimizing shrinkage and loss of shape. Daudt et al. [27] used spherical Ti powder (d90 = 32.8 µm) and KCl particles to produce Ti foams by MIM. The authors observed that, when adding 70% of the SH, the powder charge had to be increased by 80% and the binder reduced by 20% to maintain the structure dimensions during sintering. Contents higher than 80% of SH can considerably reduce the fluidity and clog the nozzle of the MIM machine. Many biomedical implants require different levels of porosity or porosity gradients. This can be achieved by processing with the addition of space holders and binders in the appropriate proportions or applying 2-component-metal injection molding (2-C-MIM) [73]. This technique is derived from the plastics industry [74] and has been adapted to metal powders. Barbosa et al. [75] produced titanium foams with a gradient in porosity using this technology, starting with feedstocks with and without space holders. Implant shape stability was only possible using less than 50 vol.% of SH in a powder load. The final porosity was not reported, but it is probably under 50%.

3.3. Binders and Debinding

The binders used during MIM of titanium powders must be those that enable the least interaction of the powder with elements, such as oxygen, hydrogen, carbon, and nitrogen, which can considerably change the mechanical properties of the implant [76]. Some binders commonly used in MIM, such as paraffin-based ones, which are characterized by elements that are toxic to organisms, are being replaced by highly soluble water-based binders, such as polyethylene glycol (PEG) [47]. The binders can also be used as space holders and should give the powder mixture good flowability, wettability, and homogeneity, prompting researchers to seek the best materials to fill these roles. A good binder system must have several desirable characteristics, such as dimensional stability during its removal, low reactivity but good adhesion to titanium, a melting point at a temperature compatible with the MIM process, and it must leave no residue.

One of the difficulties in the production of titanium foams via MIM is the maintenance of the part dimensions during sintering, mainly due to poor properties of the green material, phase change, a sudden variation in temperature during thermal debinding or sintering, and contamination by the binder. Moreover, in the case of parts with proportions of space holders above 55%, it is difficult to maintain their shape after injection, hindering the manufacture of implants with porosity greater than 50%, an appreciable percentage for foams. Several studies have focused on better understanding the dimensional stability of parts produced via MIM. Daudt et al. [27], for example, applied a plasma treatment to MIM samples after the chemical removal of binders and before thermal removal. The microwave plasma treatment cleaned the surface of the component, keeping the properties of the part intact and facilitating the thermal removal of binders. Figure 7 shows the samples obtained by the authors. The researchers noted that the plasma helped to maintain the shape of the part by stimulating a pre-sintering phenomenon (Figure 7c), creating small connections between the powder particles even for porosity above 50% and at a much lower temperature than that used in sintering in addition to maintaining open porosity.

The binder system used to produce porous titanium implants via MIM is generally composed of three materials: one with low viscosity, one with high viscosity, and a lubricant. Materials with low viscosity give the mixture a good flowability, which facilitates injection at low pressures. The most commonly used are PEG and paraffin, polymers that have a low melting point. The former can be removed with water solutions without any hazards [77]. The high-viscosity materials consist, in general, of polymers with a high melting point, such as PMMA and polyethylene (PE), providing mechanical stability, mainly after the removal of the SH. The lubricant gives the mixture greater fluidity, and stearic acid (SA) is commonly used. The proportion of these materials can vary, but, in general, the low viscosity polymer accounts for more than half of the binder mixture, the high-viscosity material makes up one-third, and the SA represents around 6% [28,43,68].

Binders are usually removed using chemical solutions and/or the application of heat. The biggest difference between paraffin and PEG is that the agents used to remove the latter are less toxic to the human body and the environment [78,79]. The mixture of powder and binder is designed to give it characteristics that guarantee good injection. The viscosity must be below 1000 Pa.s at the injection temperature [80]. Pseudoplastic polymeric characteristics are also desirable since the reduction in viscosity with deformation helps the injection of parts with more complex dimensions [81].

There are currently two main methods for debinding, namely: thermal and solvent debinding. The former is usually performed in the same furnace and cycle as the sintering step. The parts are heated and maintained at a certain temperature equal to or close to the degradation temperature of the polymer (obtained by termogravimetry) in which the binder can be completely removed. On the other hand, solvent debinding consists of plunging the parts inside a (usually) heated chemical solution (hexane or water), which reacts with part of the binder system, removing the non-structural component. In the literature on MIM + SH foams, many authors report very long debinding and dissolution times, in some cases over two days [28]. In this regard, Shbeh and Goodall [72] assessed the effect on the processing speed of different water debinding and dissolution techniques for PEG and KCl as the binder and SH, respectively. The results showed that ultrasonic water solutions at the peak melting point of PEG were the fastest way (4 h) to remove PEG and KCl completely, which can be a promising technique for the KCl removal in other feedstock compositions. The best temperature was found to be equal to 70 °C, above which the rate of dissolution was reduced due to a significant dilatation in the sample, slowing the SH removal process.

In a similar study, Thian et al. [82] assessed the thermal debinding parameters of wax binders. They observed that very fast heating rates increased the formation of internal and external cracks, deteriorating the mechanical properties of the part. The authors also noted that higher argon fluxes reduced the percentage of contamination due to the faster purging of gases from the binder, preventing them from condensing into carbon. The research showed that modulating the extraction of binders in stages improved the dimensional stability of the components due to the creation of bonds between the particles in intermediate stages.

3.4. Feedstock and Injection Molding

Powder injection molding is a process adapted from the manufacture of polymeric materials and requires some care in the injection of metallic mixtures, such as, for example, the use of protective atmospheres during the mixing and kneading of powders. However, unlike the steel injection process, no extra equipment is needed for titanium. One of the crucial steps of injection is passing through the mold port. The raw material needs to be prepared to ensure a continuous flow without clogging the nozzle. In this sense, the relationship between metallic powder, binder, and SH must guarantee good fluidity, minimizing the percentage of binder. Good mixing means the metal powder is covered in a thin film of binder to reduce friction. This process is usually carried out in blade mixers or kneaders at temperatures above the melting point of the binders (140 °C to 170 °C) for 1 h at 70 rpm, aimed to ensure a good homogeneity of the feedstock. A vacuum or protective atmospheres can be used to avoid contamination. Moreover, the cleaning of the injection machine is extremely important since titanium is a highly reactive material and can be contaminated with other powders previously used in the equipment [36].

Viscosity is another important parameter in assessing the suitability of feedstocks for MIM. The viscosity needs to be below 1000 Pa.s in a shear rate range of 10² to 10⁵ s⁻¹ [83], and it can be controlled by the grade of binder to be mixed with powders or the grade of each component of the binder system. This viscosity is usually measured in rheometers at the injection temperature. Other important parameters to consider concerning the viscosity and flowability of the feedstock are the SH amount and shape. Shbeh et al. [84] observed that spherical particles of KCl presented better flowability, less friction, and less viscosity than cubic particles.

Shbeh et al. [30] prepared feedstocks with 58% of solid loading to produce foams with different percentages of porosity by changing the volume fraction (0, 17, 35, 52, and 60%) of KCl as the SH. The binder system was composed of 70% of PEG, 25% of PMMA, and 5% of SA. Zheng et al. [23] used HDH titanium powders with average size <77 µm and NaCl with particle size <290 µm as the SH to produce highly porous parts. The volume fractions of NaCl in the solid loading (Ti + SH) were 30, 40, and 50%; wax-based binders were used, and green bodies were produced by MIM with a solid loading of 55%. Daudt et al. [27] produced a feedstock with a spherical gas-atomized titanium powder with an average size <32.8 µm, KCl as the SH with particle size in the range of 355 to 500 µm, and a binder system composed of 60 vol.% paraffin wax, 35 vol.% polyethylene, and 5 vol.% AS. A constant ratio of SH to Ti powder was set at 70:30 (vol.%) to produce samples with high interconnected porosity. The total powder load was increased from 72 vol.% to 75 and 80 vol.% to test the shape stability during debinding and sintering. The feedstock was heated to 150 °C and injected at 90 MPa. The researchers changed the conical screw to a cylindrical screw to avoid nozzle clogging due to the high powder load. The powder load of 80 vol.% showed the best dimensional stability.

Tuncer et al. [28] tested three different feedstock proportions to analyze the shrinkage, flowability, and porosity. A feedstock composed of 80 vol.% solid load and 20 vol.% of binder (comprised of 80 vol.% of paraffin, 15 vol.% of polyethylene, and 5 vol.% of stearic acid) blocked the MIM machine. Successful injection molding was observed with a solid load of 70 vol.% and volume of polyethylene in the binder of 35%, but the samples showed high shrinkage and lost their shape during sintering. A decrease in the space holder content to 55 vol.% of solid load allowed the manufacture of samples without shape distortion. However, the porosity reduced to much lower than the desired values.

Table 3. Summary of feedstock configurations found in the literature.

Ref.	Ti Powder	Solid Loading Ti + SH (vol.%)	Space Holder (vol.%)	SH Grain Size (µm)	Binder (vol.%)	Sintering Temperature and Time	Porosity after Sintering (vol.%)	Young’s Modulus (GPa)
Morelli et al. [22]	Angular HDH D50 = 20.26 µm	50–60	50	300–500	50–40	1000 °C for 4 h	59–51	6–12
Zheng et al. [23]	HDH <77 µm	55	30–40–50	<290	45	1150 °C for 2 h	42–45–62	3.0–1.5–1.1
Daudt et al. [27]	Spherical gas-atomized <32.8 µm	80	70	355–500	20	1200 °C for 3 h	64	N/A
Laptev et al. [29]	Spherical gas-atomized D50 = 19.1 µm	72–75–80	70	355–500	28–25–20	1200 °C for 3 h	59–61–55	8.6–7.4–10.3
Shbeh et al. [30]	Spherical gas-atomized <74.9 μm	58	0–17–35–52–60	D50 = 366	42	1320 °C for 2 h	20–44–56–62–65	N/A
Özbilen et al. [67]	Irregular HDH D50 = 48.3 µm	68	70	100–500	32	1300 °C for 3 h	61	N/A

(N/A = not applicable).

summarizes several feedstock configurations and the resulting porosity and Young’s modulus. The above-mentioned studies show considerable variability in the feedstock produced. The proportions of the three components can vary within a large range of volume fractions to produce foams with different porosity fractions. However, good flowability and rheology properties and appropriate proportions of Ti powder, binder, and SH are important to avoid clogging and to ensure mold-filling and the dimension stability of the green part during sintering.

3.5. Sintering

Sintering is usually the final step in the fabrication of net-shaped titanium foams. It consists of the application of heat to chemically bond the particles and increase the mechanical strength, consolidating the foam structure. This step must be carried out carefully because it can greatly affect the mechanical properties, dimensions, and architecture of the foam. The process can be performed in reactors and chambers under vacuum or protective and reductive atmospheres or furnaces, such as argon or graphite [44], respectively, and the supports must be made of stable ceramics such as ZrO₂ and Y₂O₃. Several sintering techniques, as previously mentioned, are currently reported in the literature to produce porous titanium foams for the biomedical industry, based predominantly on sintering in vacuum furnaces or with protective atmospheres. The sintering plateau temperature is dependent on the melting temperature of the material, approximately 1678 °C in the case of Ti [85]. In theory, the optimum range of sintering temperatures should be between 60 and 80% of the melting temperature, i.e., 1006 °C and 1342 °C, respectively. A high sintering temperature can block the fine pores, reducing the number of pores, as well as spheroidizing them. Thus, lower temperatures are preferred to trigger the atomic diffusion, but they must be sufficient to bond the Ti particles. According to the literature, a sintering temperature of 1250 ºC was used in around 30% of the studies [86], with holding times in a range of 1 to 4 h [22,23,27,29,30,67].

Nor et al. [33] noted that the most important parameters during the sintering of porous parts are, in this order, plateau temperature, heating rate, plateau time, and cooling rate. By increasing the temperature and the holding time, more dense parts with less porosity and, consequently, a greater elastic modulus, are obtained [87]. Carreño-Morelli et al. [22] tested two different temperatures during the sintering of angular titanium hydride powder (d90 = 34.94 µm) mixed with NaCl (300–500 µm) and PMMA (D₅₀ = 600 µm) as space holders. The researchers investigated the influence of this parameter on the mechanical properties. Figure 8 correlates the sintering temperature with the compressive behavior of the foam. As shown, the higher the sintering temperature, the higher the compressive stress absorbed by the foam. This behavior can be correlated to the higher densification of the foam sintered at higher temperatures, presenting lower porosity. Moreover, the Figure shows that an increase of 10% in the amount of PMMA used as the space holder leads to a decrease in the mechanical properties due to the higher resultant porosity after sintering. The authors also observed a decrease in mechanical properties using 50% vol. of PMMA compared to the same amount of NaCl. PMMA provided better shape preservation of the pores, leading to lower mechanical properties.

Tuncer et al. [28] used vacuum furnaces (<10⁻³ Pa) to sinter samples of spherical-shaped, gas-atomized titanium powders, and irregularly shaped HDH titanium powders at a temperature of 1200 °C for 3 h. The results showed a high level of carbon contamination from MIM due to the binders used. The gas-atomized powder contained less oxygen, this being considered more appropriate for the production of titanium implants. Moreover, components with more than 55 vol.% of space holders lost their shape during sintering due to the high porosity of the material. In this sense, Laptev et al. [29] observed that vacuum sintering of injection molded preforms with a space holder content of 70 vol% in the powder load can be used to manufacture geometrically stable parts when the water leaching step is omitted. The KCl space holder was extracted by sublimation from the solid phase at 750 °C. As a result, the time-consuming water leaching step was omitted. Moreover, the sublimation method substantially reduces the risk of incomplete space holder removal, even from large-molded parts. This technique can be useful as a future trend and an important advantage of KCl as a temporary material.

Gerling et al. [88] studied the importance of the composition of the atmosphere during the sintering of titanium alloy components. The authors used vacuum and argon atmospheres with pressures of 300 and 900 mbar. The researchers concluded that argon atmospheres with a pressure of 900 mbar produced parts with a lower degree of contamination by interstitial elements and a more homogeneous microstructure.

Contamination of Ti foams from the sintering atmosphere can be caused by residual SH material in the material preform, contaminants in the sintering furnace, such as adsorbed gases and residual materials remaining from other treatments, residual binder or redeposition of the binder, and/or exposure to atmospheric air during sintering. Therefore, vacuum conditions or inert gas flows are used in the sintering chamber to prevent the contamination of Ti parts. The sintering atmospheres normally used are composed of inert gases, such as argon or helium [69], and purification of the inert atmosphere is required before a Ti sample enters the hot zone. This is because the high affinity of Ti for interstitial elements and small amounts of reactive gases can affect the mechanical properties of the sintered parts [89]. In general, Ti powders and compacts are sintered under vacuum rather than other atmospheres [90]. Vacuum sintering with pressure around 10⁻² Pa or lower [91,92] can control the reaction activity between Ti and interstitial impurities.

In this regard, Oh et al. [93] injected 50% hydrogen flow into the argon gas atmosphere during the sintering of titanium powders and analyzed the variation in the lattice parameter of the sintered parts. The results show that the addition of hydrogen reduced the size of the lattice parameter of the space between a and c of the hexagonal structure of titanium compared to the lattice parameter obtained for samples sintered with an argon atmosphere only. This variation was attributed to the suppression of oxidation during sintering at high temperatures and the reduction in interstitial oxygen due to the reaction between hydrogen with oxygen forming OH radicals, which reduces the oxidation potential.

One concern during sintering is related to the surface of the parts and the presence of micrometric pores (~0.1–1 µm), structures almost always present in sintered materials [94]. Micropores can be beneficial in implants with biomedical objectives aimed at the growth of bone towards the implant and the transport of nutrients and body fluids [95]. On the other hand, micropores can reduce mechanical properties by concentrating stresses and inducing the appearance of cracks [96]. Figure 9 shows (a) macroporous and (b) microporous structures acquired by scanning electronic microscopy (SEM).

Daudt et al. [98] used plasma processing before the vacuum sintering step and after desalination to increase the open porosity of porous pure titanium foams. The results show that the plasma treatment increased the dimensional stability of the parts due to a pre-sintering effect together with the initial removal of binders, which allowed the sample to retain its shape during sintering, in addition to increasing the open porosity due to the collision of the species present in the plasma and material due to this phenomenon. The results also show that the plasma treatment did not alter the content of impurities (e.g., C and O) in the samples treated with a pressure of 75 Pa. Seeber et al. [99] used a different method to sinter titanium powder compacts, applying a direct current electrical abnormal glow discharge, called plasma-assisted sintering, under a controlled atmosphere of argon and hydrogen. The results show that the microstructure and the pore characteristics are strongly dependent on the discharge configuration and the sample position in the cathode or anode. This suggests that the process ensured a higher degree of purity of the atmosphere, reducing the contamination of the samples during sintering in comparison with the traditional methods.

The above-mentioned studies highlight the benefits of using plasma treatments for dimensional stabilization, increased surface porosity, and the removal of binders from samples of titanium foams. Furthermore, the addition of hydrogen has the potential to reduce oxygen contamination in titanium powders through the element’s reducing potential. However, little is known about the use of plasma processing with an atmosphere containing hydrogen during the sintering of components, indicating the need for further investigation.

4. Trends and Prospects

Currently, there is still no industrial production of porous implants for medical applications using MIM technology [32]. Therefore, this review seeks to increase the visibility of this promising technique, inspire new research, and draw the attention of manufacturers. The above sections provide details of many research studies carried out to evaluate the potential for the application of MIM combined with an SH. With the increase in MIM-Ti research, new techniques and new materials are under development for this process. Highlighted below are some new patents aimed at optimizing the MIM process and overcoming some problems that can hinder the production of Ti implants:

-

In patent JP2005281736, different fractions of TiH₂, HDH titanium, and 60Al-40V pre-alloyed powders were mixed to manufacture Ti-6Al-4V components with low oxygen, low cost, and suitable mechanical properties. Good mechanical properties of YS = 910 MPa, UTS = 950 MPa, and El = 14% were obtained by mixing 25 wt% TiH₂ and 75 wt% HDH powders.

-

A new binder system is presented in patent US7883662B2 for the control of oxygen and carbon contamination in MIM-Ti parts. Using a binder system containing naphthalene, polystyrene and stearic acid, the oxygen and carbon level in the final Ti-6Al-4V sintered parts presented very low levels of 0.197wt% and 0.05wt%, respectively, which complies with ASTM F2885 recommendations for MIM-Ti surgical tools.

-

Patent CN105382261 reports a novel technique to improve the dimensional accuracy of MIM-Ti parts. Titanium powders with different average particle sizes were mixed to produce the MIM feedstock and find the optimum blend for best dimensional stability. Using powders with average sizes of 46.8 µm, 34.5 µm, and 24.4 µm and ratios of 68:24:8 percent, a high dimensional precision of ±1%, uniform structure, low oxygen level of <0.25wt%, and high mechanical properties for high-density titanium parts were obtained.

Besides these patents, the above-mentioned studies increase the knowledge and literature available on the fabrication of Ti foams. However, more advanced stages of optimization could be reached given the large number of process parameters, which include:

-

Types, formats, sizes, and proportions of titanium powder;

-

Variables associated with the mixing and injection of titanium powders, binders, and space holders;

-

Shape, quantity, size, and material of space holders;

-

Material, quantity, combination, and extraction of binders; and

-

Time, temperature, atmosphere, and gas flow applied in the sintering;

Several combinations of these parameters can lead to the production of high-performance Ti implants with a porosity that is biocompatible with the structure of human bone.

5. Final Remarks

Metal injection molding (MIM) combined with the use of a space holder (SH) is a very attractive route for the fabrication of highly porous titanium and titanium alloy components for biomedical applications. This approach allows fine control of the morphology, architecture, and purity of very complex net-shaped components. Furthermore, its capacity for mass production is promising and it could thus help all industry sectors and different economic layers of the population to reach their potential, providing accessibility to a highly complex technology. However, to achieve this, the high cost of the initial powder, which, ideally, should be a very fine powder with spherical particles, must be addressed. In this regard, studies on the use of low-cost HDH and TiH₂ are promising, aimed at reducing the processing costs. The problems concerning powder shape and contamination are under investigation to enhance the potential for the use of these powders by applying a protective atmosphere and/or plasma treatment. In addition, research on the control of process-induced contamination is underway in terms of exploring, for instance, the use of hydrogen in the sintering process to reduce the oxygen content. Finally, studies to address the problem of dimension stability are being conducted. Plasma sintering may provide an important tool for achieving dimensional stability and, at the same time, assist in the removal of binders under a controlled atmosphere.