On the Role of Substrate in Hydroxyapatite Coating Formation by Cold Spray

Abstract

The deposition of agglomerated hydroxyapatite (HAp) powders by low-pressure cold spray has been a topic of interest in recent years. Key parameters influencing the deposition of HAp powders include particle morphology and impact kinetic energy. This work examines the deposition of HAp powders on various metal surfaces to assess the impact of substrate properties on the formation of HAp deposits via cold spray. The substrates studied here encompass metals with varying hardness and thermal conductivities, including Al6061, Inconel alloy 625, AISI 316 stainless steel, H13 tool steel, Ti6Al4V, and AZ31 alloy. Single-track experiments offer insights into the initial interactions between HAp particles and different substrate surfaces. In this study, the results indicate that the ductility of the substrate may enhance HAp particle deposition only at the first deposition stages where substrate/particle interaction is the most critical factor for deposition. Features on the substrate associated with the first deposition sprayed layer include localized substrate deformation and the formation of clusters of HAp agglomerates, which aid in HAp deposition. Furthermore, after multiple spraying passes on the various metallic surfaces, deposition efficiency was significantly reduced when the build-up process of HAp coatings shifted from ceramic/metal to ceramic/ceramic interactions. Overall, this study achieved agglomerated HAp deposits with high deposition efficiencies (30–60%) through single-track experiments and resulted in the preparation of HAp coatings on various substrates with thickness values ranging from 24 to 53 µm. These coatings exhibited bioactive behavior in simulated body fluid.

1. Introduction

The cold gas spraying process has been developed for over three decades as an alternative to conventional thermal spraying [1]. Through the years, different investigations have been focused on various aspects of the cold spray technique that involve gas dynamics, particle motion and acceleration, particle impact at the substrate surface, and coating growth [2]. Most of the literature focuses on understanding the production of metal and metal–matrix composite coatings using cold spray [3]. Numerous studies have shown that the bonding mechanisms of metal-based coatings in cold spray are closely related to impact velocity and local shear deformation at the particle/substrate interface [4,5,6]. A distinctive feature of cold spray is the high-speed solid-state impact. When the impact velocity exceeds a critical threshold, known as critical velocity, deformation mechanisms in ductile materials are activated, enabling the incoming particles to flow and adhere to the substrate surface [4]. The deposition of metallic materials by cold spray depends not only on the properties of the metal powder but also on the characteristics of the substrate material [7,8]. Different investigations have demonstrated that the kinetic energy of particles at impact is transformed into heat and plastic deformation work [9,10]. This energy conversion enables the mechanical interaction required for the deposition of particles onto the substrate surface. However, if the substrate is significantly more ductile than the metallic particles, the deformation predominantly occurs on the substrate, leading to surface erosion rather than effective particle deformation and adhesion [11]. On the other hand, very hard substrates can lead to the concentration of plastic deformation in the particles, which hinders mechanical anchorage between the substrate and the particles due to the minimal plastic deformation occurring on the substrate surface [12]. In previous years, considerable efforts have been directed towards the comprehension about the formation of the first deposited layers of metal and metal–matrix composite coatings by cold spray on various metallic substrates. The hardness and temperature of the substrate are highlighted as the primary parameters influencing the growth and properties of the resulting coatings [13,14]. Soft substrates are prone to suffer localized shear and heating, which help enhance bonding in metal/metal and metal–matrix composite/metal systems [15]. Similarly, preheating the substrate often increases the deposition rate in cold spray by promoting deformation, disrupting oxide layers, and enhancing particle anchorage upon impact [13].

In the last few years, considerable efforts have been dedicated to the deposition of ceramic materials using cold spray [16]. As cold spray has emerged as an alternative to conventional thermal spray, the prospect of manufacturing ceramic coatings using this method is seen as a potential solution to common issues encountered with ceramics in conventional thermal spray processes, such as the precipitation of unwanted phases, evaporation, and low deposition rates [17,18]. Previous studies have reported the deposition of ceramic powders using cold spray, including materials such as titanium dioxide, hydroxyapatite, alumina, zirconia, and yttria [19,20,21]. It is worth noting that when ceramic materials are subjected to mechanical loads over a short period, if the load exceeds the material’s elastic limit, these can experience localized damage and multiple crack propagation, leading to fragmentation rather than plastic deformation, unlike in metallic alloys [22]. It is therefore noteworthy that the deposition of ceramic materials using cold spray can be successfully achieved.

The fabrication of ceramic coatings using cold spray has been demonstrated by several authors, particularly with titanium dioxide and hydroxyapatite powders [20,23,24,25,26]. Some studies have demonstrated the fabrication of coatings with thickness values ranging from 15 µm to 350 µm on metallic substrates. Most of these studies report the use of HAp and titanium dioxide as feedstock materials. However, there is considerable controversy regarding the reproducibility of these coatings, mostly due to the lack of both detailed research works focused on understanding the deposition mechanisms involved and defining engineering tools that provide, for instance, deposition windows in which cold-sprayed ceramic coatings can be produced. Notably, the concept of critical velocity, which was originally developed for metallic materials and forms the basis for constructing deposition windows in cold spray for metals, does not apply to ceramics. Instead, some researchers attribute the deposition of dense ceramic particles on metals to the plastic deformation of the substrate, along with their fragmentation and embedding into the ductile surface [27]. However, other authors have found that the deposition of porous agglomerated ceramic particles is more feasible in cold spray compared with the deposition of dense ceramic particles. This is because the yielding threshold for crack propagation within agglomerated particles decreases compared with that of dense particles, which facilitates the rearrangement of the nanoparticles contained in the agglomerated particles at impact. This rearrangement refers to the particles’ ability to deform upon impact without experiencing catastrophic fragmentation. This is possible because normal stresses generated at impact allow the compaction of the agglomerates and attachment to the substrate surface [20,28].

Given the growing interest in understanding ceramic deposition via cold spray, some researchers have developed experimental and analytical models to correlate the key variables involved in this process [19,29,30,31]. In recent works, some approximations have led to associate physical quantities such as fracture energy and kinetic energy to the deposition of agglomerated ceramic particles [32]. A simplified model was recently proposed [20] to describe the processing conditions required for the deposition of agglomerated ceramic particles. This model suggests that a threshold of kinetic energy must be exceeded to promote the creation of new surfaces and the rearrangement of agglomerates upon impact. Consequently, variables such as particle diameter and morphology, impact velocity, and agglomerate cohesion can be adjusted to facilitate the deposition of ceramic particles.

Some studies have also been performed using dense and agglomerated ceramic powders on different metallic substrates [19,26,28,33]. Such studies have focused on assessing the ability to build-up coatings accounting on substrate hardness and preheating. For titanium dioxide particles, significant differences are observed when preheated surfaces are used, primarily due to the development of oxide layers on the substrate surface [26]. When an oxide layer is not developed, the increased ductility of the substrate metal can enhance the embedment of ceramic fragments. In the case of HAp, systematic studies focusing on substrate effects are lacking. Instead, isolated studies have reported the deposition of HAp on substrates such as stainless steel, titanium, and magnesium alloys [29,34,35,36]. Although some experimental efforts have been made to study the effects of substrate properties on the formation of ceramic coatings [33,37], none of these studies have fully clarified how substrate properties influence the deposition of agglomerated ceramic particles in low-pressure cold spray. Consequently, this work presents both experimental and analytical approaches to examine the deposition of agglomerated HAp particles in the low-pressure cold spray process across various metallic substrates. It also represents a continuation of previous research published by our group in which a deposition mechanism for ceramic materials was proposed [20]. The main objective is to review how substrate properties affect the deposition parameters for agglomerated HAp powders in cold spray and their implications for coating build-up.

2. Materials and Methods

2.1. Single Tracks and Coatings Preparation

Agglomerated hydroxyapatite (HAp) powder (Captal 30SD, Biotal, UK) was employed in this study. The HAp powder had a monomodal particle size distribution with d₁₀, d₅₀, and d₉₀ values of 21 ± 3 μm, 40 ± 3 μm, and 64 ± 2 μm, respectively. This powder presented a rounded morphology and had a porous appearance, as shown in Figure 1.

All the experiments were carried out in a low-pressure cold spray gun (DYMET 423, DYCOMET, Akkrum, The Netherlands) with compressed air as the processing gas. A conventional convergent/divergent (CK20, DYCOMET, Akkrum, The Netherlands) nozzle and a modified homemade nozzle (CIATEQ A.C, Mexico) [20] were employed in the experiments. The selected substrates included metals with different hardness and thermal conductivity; in particular, 6061 aluminum (Al6061), Inconel 625 (Inc625), 316 stainless steel (SS316), H13 tool steel, Ti6Al4V alloy (Ti64), and AZ31 magnesium alloy as shown in Figure 2.

The substrates were prepared in coupons of 5 mm in thickness; they were mirror-polished (roughness, Ra = 0.02 μm) following traditional grinding and polishing procedures and cleaned with ethanol before spraying to eliminate any contamination at the surface. The distance between the nozzle’s exit and substrates surface was 10 mm. The cold spray gun was mounted in a six-axis robotic arm (KRC2, KUKA, Augsburg, Germany), which allowed the control of the relative velocity between the nozzle’s exit and the substrate surface at a constant value of 2 m/s. Substrate preheating was controlled using an electric heater on the back face of the sample’s holder. The temperature was monitored on the surface of each substrate with an infrared thermometer. Single-track experiments were carried out by a single pass of the nozzle on the substrate’s surfaces, following the spraying conditions listed in

Table 1. Spraying conditions employed in single-track experiments.

Sample ID	Gas Temperature (°C)	Gas Pressure (Bar)	Nozzle Type	Substrate Preheating (°C)	Feeding Rate (g/min)
M1	300	8	conventional	25	8
M2	300	8	conventional	100
M3	300	8	conventional	200
M4	600	8	conventional	200
M5	300	5	homemade	200
M6	600	5	homemade	200
M7	600	3	conventional	200

[20].

Finally, HAp coatings were prepared on different substrates based on the results of the single-track experiments. The spraying conditions employed to prepare the HAp coatings are shown in

Table 2. Spraying conditions employed to prepare HAp coatings.

Sample ID	Gas Temperature (°C)	Gas Pressure (Bar)	Nozzle Type	Substrate Preheating (°C)	Feeding Rate (g/min)	Passes
X1	600	8	conventional	25	8	36
X2	300		conventional
X3	600		homemade
X4	600		conventional
R1	600	8	conventional			5
R2						12
R3						18
R4						24
R5						36

.

2.2. Samples Characterization

The samples were weighed before and after spraying to calculate the deposition efficiency for each condition. The deposition efficiency was calculated as follows: [coating weight/(powder feed rate · deposition time)]. This value is an indicator of the amount of sprayed material that remained on the substrate’s surface. The surface and cross-section morphologies of the deposits obtained were examined by optical microscopy (VELAB-5000, Montreal, QC, Canada) at a 40× magnification. Optical micrographs were analyzed with the Image J^® software (Version 1.54k) to calculate the percentage area of deposited material. The percentage area of deposited material is a measure of how much surface is covered by the particles deposited on the substrate in a single-track experiment. The single-track deposits were also examined by scanning electron microscopy (SEM, Jeol IT100, Tokyo, Japan) to observe local characteristics of the impacts on the surface such as local deformations, fragmentation, and material accumulation. The top surface topography of the HAp coatings was studied by optical microscopy (Keyence, VHX 600, Osaka, Japan) and SEM. Furthermore, the coating’s cross-sectional microstructure was examined by SEM. The structural characterization of the coatings was performed by X-ray diffraction (XRD) (Smart Lab diffractometer, Rigaku, Tokyo, Japan) with CuKα radiation (λ = 1.54 Å) operating at 40 kV and 20 mA. The XRD patterns were collected over a 10–70° 2θ range.

An estimation of physical quantities of particles at impact, such as kinetic energy and thermal energy, is useful to understand the response of the particles in cold spray, as explained in a previous work [20]. Consequently, finite element analysis and simulation of the low-pressure cold spray process was performed for the spraying conditions given in Table 1 and Table 2. Simulations were carried out by computational fluid dynamics (CFD), employing ANSYS-FLUENT Software (2023-R2 academic version). The process consists of simulating the gas dynamics and particle kinetics to get the particle’s temperature, velocity, kinetic energy, and thermal energy at impact. The domain of the problem was solved for the internal geometry of the nozzles and the region between the nozzle exit and the substrate surface, using a 2D axisymmetric swirl model. The K-epsilon model was taken into consideration to account for turbulent effects with a pressure-based resolution strategy. Further details about the simulations are found elsewhere [20,30].

2.3. In Vitro Characterization

10 mm × 100 mm × 5 mm samples of each substrate material were sprayed with HAp to perform in vitro tests. These tests were carried out by immersing a 1 cm² sample in 50 mL of Kokubo’s solution at 37 °C. The immersion of the samples was performed for 14 days. The top of each surface was examined by XRD and SEM at the end of the test to identify the formation of an apatite layer. In vitro characterization was included in this work due to the potential of cold-sprayed HAp coatings for biomedical applications.

3. Results and Discussion

3.1. Single Tracks Experiments

The results of the single-track experiments are shown in Figure 3. This figure displays the coated area percentage as a function of deposition efficiency for the various spraying parameters of HAp listed in Table 1. The collected data revealed a curve with two distinct regions. The coated area percentage increased linearly with deposition efficiency in the first region, below 50% deposition efficiency. In the second region, above 50% deposition efficiency, there was a visible change in the curve’s slope. Specifically, in this second region, the linear increase in deposition efficiency was more pronounced compared to the increase in coated area percentage. Complete substrate surface coverage and the formation of a coating layer are anticipated when both deposition efficiency and coated area percentage reach 100%. In this manner, all the points in the second region are closer to achieving complete coverage and forming a coating layer in a single spraying pass. Overall, conditions M3, M4, and M7 demonstrated the best combination of efficiency and coated area percentage. It is worth noting that not all points corresponding to spraying conditions M3, M4, and M7 fall in the final segment of the curve shown in Figure 3. This indicates that the substrate’s nature had a significant impact on the surface coverage and formation of HAp deposits, a topic that will be discussed later in this section.

Figure 4 shows images of a top view of the HAp tracks related to the spraying conditions of HAp listed in Table 1. The images are divided into two groups: first, those corresponding to the spraying conditions M1 to M3, and second, those corresponding to the spraying conditions M4 to M7. This separation was strategically implemented to emphasize the significance of preheating the substrates before spraying. When HAp powder was sprayed onto the different substrates at room temperature (M1 spraying condition), the HAp particles were deposited on the surfaces. However, the bright areas in the images indicate that the substrate surfaces were not completely covered. Once the substrates were preheated to 100 °C and 200 °C (M2 and M3 spraying conditions, respectively), the bright areas associated with the metallic substrates were reduced, indicating an increase in the deposition of HAp particles. This result is consistent with findings from cold-sprayed metallic coatings on metallic substrates. For metals, preheating the substrate promotes the deposition of metallic powders by facilitating a soft–soft interaction upon impact. This interaction enhances the material’s flow at the particle/substrate interface, leading to improved mechanical anchorage [2,13].

Figure 4 also shows that the M3, M4, and M7 spraying conditions almost completely covered the substrate surfaces. However, there were variations depending on the substrate used in the experiments. For instance, under the M3 condition, the Inc625 and H13 substrates—being the hardest substrates—were covered less by the HAp deposit compared to the softer ones. A similar trend was observed under the M7 condition, while only minor differences were noted following the M4 condition. This result may suggest that once deposition of HAp particles was activated, little differences were observed when the substrate surface was changed. A similar observation has been reported in [37] after working with a cold-sprayed titanium dioxide agglomerated powder on different metallic substrates.

To better understand the behavior of HAp particles on different substrates, their morphology upon impact was also analyzed by comparing samples obtained under low and high deposition efficiency conditions, for instance, M1 and M4. This analysis can provide information that may help to comprehend if there is a variation in deposition mechanisms on the different substrates. Figure 5 displays the morphologies of HAp particles collected from Ti64, SS316, H13, Inc625, AZ31, and Al6061 substrates under the M1 spraying condition. The number and size of the deposited particles decreased from the Ti64 substrate to the Inc625 substrate. As hardness increased, in Inc625 and H13 substrates, signs of plastic deformation associated with particle rebounding on the substrates’ surfaces were observed. Small HAp fragments and angular remnants of agglomerated particles were observed on the surface of the hardest substrates. This indicates that the elastic component of the mechanical work upon impact remained substantial. While agglomerated HAp particles may rebound, their inherent brittleness leads to fragmentation. However, this fragmentation is insufficient to rearrange the entire volume of particles, leaving only small fragments trapped on the surface. Rebounding was somewhat reduced in Ti64, Al6061, and AZ31 substrates—being the softest substrates—as shown in Figure 5a,e,d, where increased plastic deformation of these substrates aided in trapping small fragments on their surface. This highlights the need of achieving a certain energy threshold that allows for the rearrangement of agglomerates upon impact [20,32] along with localized plastic deformation of the substrate, to promote favorable deposition conditions—namely reduced rebounding and improved particle anchorage. Energy considerations will be discussed in more detail at the end of this section.

Figure 6 displays the morphologies of HAp particles collected from Ti64, SS316, H13, and Inc625 substrates under the M4 spraying condition. The figure illustrates that the deposited HAp particles had an irregular shape due to particle fracture, which was related to their fragile response upon impact. Most of the deposited volume, consisting of agglomerated HAp particles, accumulated at the center of the impact zone. The deposited particles acquired irregular shapes due to the rearrangement of their volume upon impact, allowing them to anchor to the surface. This observation was consistent across all substrates analyzed in Figure 6. As discussed in a previous study [20], the deposition of agglomerated ceramic particles occurs when impact conditions favor non-homogeneous lateral flow of agglomerated particles and accumulation of material at the center of the impact zone on the substrate surface. Signs of a cone-like morphology, resulting from the accumulation of agglomerated particles at the impact zone and the rearrangement of these agglomerates due to lateral flow, are observed in the deposited particles regardless of the substrate hardness in this study.

Figure 7 illustrates the deposition of HAp particles on the soft substrates (AZ31and Al6061 alloys) for the M4 spraying condition. Compared with the morphologies depicted in Figure 6 for the HAp particles deposited on the hardest substrates, the adhered HAp particles on AZ31 and Al6061 substrates remained irregular as those in the hardest substrates. In fact, the particles were also less fragmented compared to those observed under the M1 spraying condition (Figure 5e,f). They appeared broken but deformed (Figure 7a,c), which enhanced their anchorage to the surface. Plastic deformation of the substrates was also noted, with some signs of deformed areas associated with particle rebounding, though significantly less than for the M1 condition. The lower hardness of these substrates led to greater particle penetration upon impact, resulting in more rounded particle morphologies compared to those observed on the harder substrates (Figure 7b,d). This fact aligns with observations reported in studies of cold-sprayed hard metal particles on soft metal substrates. In such instances, the kinetic energy of the impacting particles is primarily transformed into plastic work and thermal energy. Most of this plastic work is absorbed by the substrate, resulting in deeper penetration and reduced deformation of the particles compared with particles impacting on harder substrates [2].

Figure 8a shows the experimental deposition efficiency of HAp as a function of the average hardness value of each substrate (see Figure 2a). This result summarizes the observations from the SEM images discussed in this section, combining the effects of spraying parameters and substrate properties on the deposition efficiency of the agglomerated HAp powder. Additionally, Figure 8a reveals a tendency for deposition efficiency to decrease when using harder substrates compared to softer ones. This finding is consistent with previous studies [39] that evaluate the deposition of hard metallic particles on various metallic surfaces. It is suggested that hard metallic substrates are less prone to significant deformation, which reduces the contact area between the particles and the substrates. Additionally, Figure 8b shows the experimental deposition efficiency as a function of the yield stress ratio of each substrate. The yield stress ratio is defined as the yield stress at 200 °C divided by the yield stress at room temperature. This ratio is used to highlight the effect of temperature on the mechanical properties of the substrates. The yield stress values of the substrates are as follows: 170, 145, 880, 290, 1650, and 460 MPa at room temperature and 95, 106, 704, 206, 1534, and 423 MPa at 200 °C for Al601, AZ31, Ti64, SS316, H13, and Inc625, respectively [38]. The results are consistent with those shown in Figure 8a, suggesting that increasing deposition efficiency requires conditions that facilitate easier substrate deformation. This is effective when the substrate is heated before spraying as long as other spraying parameters, which influence the kinetic energy of the particles at impact, are also set to favor deposition. Thus, preheating the metallic substrate is a viable strategy for improving the efficiency of cold-sprayed HAp deposits.

Based on the results obtained here, a practical question for preparing cold-sprayed HAp coatings is: How can these results be leveraged to advance our understanding in predicting the behavior of agglomerated HAp powders for the preparation of HAp coatings by cold spray? Previous studies [32] have established that the deposition of ceramic particles is associated with a critical energy threshold known as the fracture energy. When the kinetic energy of the particles exceeds this threshold, new surfaces are formed. Any additional energy is used to rearrange the agglomerates fractured upon impact and to deform the substrate surface. This phenomenon has been described in a previous study [20] using a simplified expression that correlates the kinetic energy (E_k) and the breakup energy (E_s) with the particle diameter (d) and impact velocity (v). The breakup energy (E_s) represents the energy required to fracture a ceramic particle upon impact. This model provides insights into the energy state at impact based on particle characteristics such as dimensions and morphology. However, it does not account for the influence of substrate properties on the expected deposition of particles in cold spray.

In this work, Figure 9 presents the ratio of E_k and E_s as a function of the impact velocity and particle diameter for the spraying conditions. This result is consistent with previous studies [20]. Additionally, dashed lines were added in Figure 9a to indicate the hardness of each substrate and refer to the energy ratio level (E_k/E_s) from which the deposition efficiency of the sprayed powder gets increased more than 30%. This reference value was obtained experimentally in the present work following the procedure outlined in [20]. Interestingly, this result highlights the role of the substrate in determining how easily deposition can be achieved from an energy perspective of the particles at impact. It can be used to predict the spraying parameters required to reach the necessary energy level for effective deposition. For soft substrates, such as the AZ31 alloy, a minimum energy ratio level is needed to achieve deposition efficiencies of around 30%.

As discussed in this section, experimental evidence indicates that only small fragments, submicrometric residues, and portions of the agglomerates were anchored when deposition efficiency was below 30% (e.g., under M5 and M6 conditions). This means that cold-sprayed HAp powders achieved higher deposition efficiencies on softer substrates (i.e., 0–30%) more readily than on harder counterparts. This is attributed to the substrate’s ductility and the embedding of fragments into the surface. When the energy ratio level exceeded 2, as shown in Figure 9b (see gray area), the difference in deposition efficiency between substrates diminished. For instance, cold-sprayed HAp powder achieved deposition efficiency values between 50% and 70% for energy ratio levels ranging from 2 to 3, regardless of substrate hardness, providing that preheating was maintained at 200 °C. This observation is consistent with the principles of the energy ratio level, which suggest that additional energy is required to consolidate agglomerates on the substrate surface upon impact. Figure 9b also illustrates the effect of substrate preheating on deposition efficiency. Cold spraying HAp particles at room temperature requires more energy to achieve high efficiency levels (>30%) compared with the sprayed particles on a substrate heated at 200 °C. This is consistent with the experimental observations in this work and is attributed to the reduced plastic flow of the substrates at lower temperatures, which requires additional energy to facilitate flow and create the mechanical anchorage needed to bond the agglomerated particles.

3.2. Hydroxyapatite Coatings

Single-track experiments in cold spray are a method used to understand particle/substrate interactions. In this work, these experiments served as a baseline for preparing coatings. In the initial set of experiments, substrates were preheated, and the M4 spraying condition was employed to prepare HAp coatings using low-pressure cold spray. Surprisingly, delamination of the HAp coatings occurred at the end of the spraying process, regardless of the substrate used, as shown in Figure 10a. The temperature of the substrates was monitored during the deposition process to investigate the cause of delamination in preheated substrates, as shown in Figure 10b. Coatings were also prepared at room temperature under the same spraying conditions, and no delamination occurred. The temperatures of these samples were recorded and compared with those of the samples sprayed at 200 °C, as shown in Figure 10b. For preheated samples, the initial temperature was 200 °C. During the spraying process, the temperature gradually decreased, reaching a minimum range of 170 °C to 130 °C. At the end of the process, the temperature increased progressively to around 190 °C. This behavior is linked to the formation of a bow shock on the substrate surface. Bow shocks are shock waves generated by the adjustment of the gas flow to the perturbations created by the substrate surface. This transition from laminar to turbulent flow near the substrate enhances heat transfer [40,41]. Yin et al. [41] provided a study about the effect of the bow shock on the temperature of the substrates in high pressure cold spray. They suggested that substrate temperature is affected by the incidence of the gas stream on the substrate surface. The bow shock makes the temperature change locally at the surface or totally in the bulk of the substrate depending on its thickness. Substrates with less than 3 mm in thickness are more prone to be affected on their complete volume. However, as observed in this study, thicker substrates are thermally affected only in a fraction of their volume, specifically just below the surface exposed to the bow shock.

The results in Figure 10b suggest that the temperature of the cold spray jet was cooler than that of the substrate surface, and the bow shock contributed to a localized reduction in surface temperature. This effect diminished by the end of the spraying process, allowing the substrate to return to its initial temperature due to thermal inertia. Consequently, delamination may be attributed to thermal fluctuations caused by the bow shock, which could introduce high levels of residual stresses in the coatings due to the contraction and expansion of the substrates during deposition. Conversely, heating occurred rather than cooling across all substrates when deposition was conducted at room temperature. This suggests the occurrence of substrate expansion; however, thermal expansion coefficients of the metals used are higher at 200 °C than at room temperature, making the development of high residual stresses more feasible in preheated substrates. The thermal expansion coefficient values of the substrates are as follows: 23.6, 26, 8.6, 16, 11.5, and 12.8 μm/m °C at room temperature and 25.2, 27, 9.2, 16.5, 12.6, and 13.3 μm/m °C at 200 °C for Al601, AZ31, Ti64, SS316, H13, and Inc625, respectively [38].

Based on these results, cold-sprayed HAp coatings were prepared without preheating the substrates, following the spraying conditions outlined in Table 2. Figure 11a presents the deposition efficiency of HAp coatings as a function of the number of spraying passes on different substrate surfaces. The data show that deposition efficiency values ranged from 0 to 6%, which was significantly lower than the 20–35% efficiency observed in single-track experiments conducted with substrates at room temperature. Initially, the deposition efficiency was similar across all samples; however, a difference in this property was noted among substrates for deposition passes higher than 12. In general, deposition efficiency linearly increased as the number of passes reached 26. After that, deposition efficiency reached a plateau or even a slight decrease when the samples were coated employing 36 passes. The hardest substrates appeared to exhibit slightly higher deposition efficiency than the softer ones. However, the differences in deposition efficiency among the coated samples remained minimal if compared with that obtained in single-track deposition experiments. Figure 11b also shows the variation of HAp coating thickness as a function of the number of spraying passes on different substrates for the spraying conditions listed in Table 2. A modest increase in thickness was observed after 20 passes on softer substrates. In contrast, harder substrates showed a more significant increase in thickness after 20 passes. Overall, the coating thicknesses ranged from ~10 µm to ~60 µm.

Optical and SEM images of the samples’ top surface and cross-sectional views were obtained to elucidate the features of the cold-sprayed hydroxyapatite (HAp) coatings, as illustrated in Figure 12 and Figure 13. Figure 12 displays a general view of the top surface of the HAp coatings obtained after 5, 24, and 36 spraying passes, respectively. This figure illustrates that after the initial five spraying passes, the formation of a homogeneous coating layer was not achieved and resembled the outcome observed after just one spraying pass in the single-track experiments. This observation is consistent with the low deposition efficiency values recorded after five spraying passes in Figure 11. Interestingly, after 24 and 36 spraying passes, the substrate surfaces began to be coated more homogeneously. The coating appeared to grow by the compaction of particles, driven by fragmentation and compaction forces generated upon impact [20,36], which facilitated the growth of the HAp cumulus formed after the initial spraying passes. The coatings after 24 and 36 spraying passes exhibited intense white areas associated with the growth and consolidation of HAp layers from deposited HAp upon the first spraying passes. Additionally, there were areas with decreased whiteness intensity, which corresponded to the last regions of the metallic substrates covered by HAp during the spraying process and likely had reduced thickness compared to that of the more intense white areas.

Figure 13 shows the SEM images from the cross-section of the cold-sprayed HAp coatings deposited on the metallic substrates after 36 spraying passes. The HAp coatings exhibited features such as microcracks and pores. The consolidation of these coatings appeared similar to observations from the aerosol deposition process [42]. In that process, the solid-state impact of the fragile HAp particles and their successive impacts lead to the compaction of fragments on the substrate surface. Additionally, the internal stresses generated within the coatings promote the formation of microcracks [42]. However, in cold spray, a distinctive feature is the formation of the coating layer through the growth and consolidation of first deposited HAp particles. The initially deposited HAp particles, as observed in this study and in previous research [20,33], exhibited a cone-like morphology and served as platforms for building the coatings up after each spraying pass (see visual guides, red-dashed lines in Figure 13b,d). This explains the different white intensities observed from the top view of the coatings, which are associated with the accumulation and consolidation zones of HAp fragments on the metallic substrates. It is worth noting that the thickness of HAp coatings (Figure 13g) obtained on softer substrates (AZ31, Al6061, Ti64) was lower compared to that on harder counterparts.

The coating/substrate interfaces, especially in the softest substrates (Figure 13a,b), exhibited more severe deformation. In contrast, the interfaces in the harder substrates (Figure 13d–f) appeared more planar. This observation suggests that hard substrates promoted the growth of HAp coatings by providing a strong consolidation of the initially deposited HAp particles, which acted as a rigid platform for the compaction of subsequent particles. In contrast, for soft substrates, some of the impact energy was spent on the deformation of the ductile surface. This reduced the energy available for particle compaction upon impact, which may have hindered the growth of the deposited agglomerates due to the formation of microcracks and potential erosion at these sites. Future research focused on particle/particle interactions will be valuable for elucidating more details about coating growth and may provide insights into strategies for improving deposition efficiency. Despite this, the cold-sprayed HAp coatings discussed in this work could potentially be used as bioactive coatings on metallic surfaces for biomedical applications. Specifically, the HAp coatings obtained in this study on AZ31, Al6061, Ti64, SS316, H13 steel, and Inc625 had thickness values of 24 ± 9.9 µm, 25 ± 10 µm, 30 ± 3.6 µm, 51 ± 3.6 µm, 53 ± 7.1 µm, and 48.2 ± 9.3 µm, respectively. The next section will present an analysis of their bioactive behavior.

3.3. In Vitro Tests of Cold-Sprayed HAp Coatings

Cold-sprayed HAp has potential applications in the biomedical field due to its bioactive and osteoconductive properties in biological fluids. Its composition closely resembles that of the mineral apatite found in human bones, making it suitable for bone tissue engineering. In orthopedics, HAp coatings are valuable for enhancing the bioactivity of metallic implants, which are often made from materials such as stainless steel, Co-based alloys, Ti-based alloys, and lightweight metals like Mg and Al alloys [43]

Low-pressure cold spray offers a promising alternative to conventional plasma spray for applying HAp coatings. Lower processing temperatures of cold spray compared to those of plasma spray help maintain the stoichiometry and phases of the feedstock powders in the coatings, minimizing the formation of unwanted phases. Indeed, the XRD patterns shown in Figure 14a confirm that the crystalline HAp phase of the feedstock powder was retained after cold spraying, with the coatings exhibiting a diffraction pattern similar to that of the feedstock powder. These results were expected since the low-pressure cold spray processing temperatures were below 600 °C in this study, far away from the transformation temperatures of HAp (>1000 °C) [44].

Figure 14b displays the diffractograms of the cold-sprayed HAp coatings after 14-days immersion in Kokubo’s solution. These results show a change in the crystallinity of the HAp coatings due to their exposure to the simulated body fluid. The observed widening of the diffraction peaks, which indicates this change, is common in HAp under in vitro conditions and is associated with its bioactive behavior [45,46]. The HAp coatings on SS316, Ti64, and Inc625 exhibited similar diffraction patterns after the in vitro tests. In contrast, coatings on AZ31, Al6061, and H13 steel showed changes in crystallinity. However, the diffraction peaks related to HAp were less intense, and the contribution from the substrate peaks was more prominent (labeled with * in each pattern). This difference is attributed to the deterioration of the surface, which exposes the underlying metal, which may be related to delamination of the coating arising from the interaction with the fluid. This happens due to the ionic interactions between the HAp coating and the simulated body fluid, leading to element diffusion and surface changes that result in the formation of an apatite layer. Interestingly, this phenomenon did not occur with the same intensity in all substrates, suggesting potential variations in the rate of apatite layer formation among the different coatings analyzed. Accordingly, SEM images of the top surface of the coatings after in vitro tests were analyzed to further investigate these differences, as shown in Figure 15.

Figure 15 shows the SEM images of the top surface of cold-sprayed HAp coatings on the different metallic substrates after 14 days of immersion in Kokubo’s solution. Overall, all the surfaces exhibited the formation of a bone-like apatite layer following the in vitro test. This process, typical of bioactive ceramics such as HAp, is associated with ionic interactions between the coating surface and the solution [45,46]. The result is the growth of clusters of bone-like apatite, which generally have a granular texture with some cracks. These cracks appear due to stress accumulation and relief as the apatite layer develops [47]. Interestingly, the texture of the apatite layer on the Al6061 substrate was different when compared to that observed on the rest of the surfaces (Figure 15b). This laminar or plate-like texture often forms on bioactive surfaces when Ca²⁺ ion absorption increases due to the creation of a negative charge at the surface associated with a basic pH of the solution [48,49]. The apatite layer formed on the HAp coating deposited on Al6061 was not homogeneous across the surface (see inset in Figure 15b). A similar non-homogeneity was observed in the apatite layer on the HAp/AZ31 and HAp/H13 systems (see insets in Figure 15a,e), which aligns with the XRD analysis. Previous studies have attributed these variations to local pH differences [49,50], potentially influenced by the metallic substrate and the inherent porosity of the coatings in this work. Based on these results, the best bioactive response was achieved with the HAp coatings deposited on Ti64, SS316, and Inc625.

4. Conclusions

Considering the advanced research stage on ceramic coatings deposited by low-pressure cold spray, this study presents a series of experiments focused on the effect of the substrate on the deposition of agglomerated hydroxyapatite powders. The aim was to explore how particle/substrate interactions determine the growth and development of a HAp coating layer produced by low-pressure cold spray. This work also includes in vitro studies as a preliminary evaluation for potential biomedical applications, although future research will be necessary to optimize these systems. The key conclusions drawn from this study are as follows:

(1)

The deposition of agglomerated HAp particles upon impact on both hard and soft substrates involved the rearrangement of agglomerates. This rearrangement resulted from the retention of nanoparticles, which comprise the agglomerated HAp particles, at the impact site. This was also combined with the lateral flow of these agglomerates caused by the shear forces generated during the impact. Soft substrates enhanced the entrapment of the agglomerates at the surface and facilitated the retention of small fragments generated due to the fracture of the particles at impact.

(2)

Preheating at 200 °C favored the deformation of the metallic surface and enhanced the retention of agglomerated particles upon impact. Preheating was a way to increase deposition efficiency in the first deposition layer on soft and hard metallic substrates.

(3)

Mechanical properties of the metallic substrates determined the energy level for agglomerated HAp particles to achieve a specific degree of deposition. Hard substrates required more energy than soft counterparts for the initial deposition layer that was obtained in a single spraying pass. This suggests that not only is the energy available for fracture and compaction of the particles a key parameter in cold spray, but also substrate deformation is necessary to promote the mechanical anchoring of the particles. In this work, a deposition window, accounting for impact energy and substrate hardness, was proposed as an engineering tool to anticipate the behavior of agglomerated HAp powders on various metallic substrates by cold spray.

(4)

Once the initial HAp layers were deposited on the metallic substrates, a shift from metal/ceramic to ceramic/ceramic interaction occurred. This shift significantly altered the deposition behavior of the agglomerated particles, resulting in a reduced deposition efficiency. A key observation in this study is the compaction of the initial deposition layers and the formation of cone-like agglomerates, which served as sites for building-up the coatings through successive spraying passes. This buildup process was more effective on hard substrates, where the compaction of the initially deposited HAp particles better supported the impact of subsequent particles than those on soft substrates. Further studies on ceramic/ceramic interactions in these systems may yield important insights for future optimizations.

(5)

In vitro behavior of the cold-sprayed HAp coatings in this study aligned with the expected response of bioactive ceramic coatings in simulated body fluids. Inherent features of the cold-sprayed coatings, such as microcracks, pores, and local variations in thickness resulted in non-homogeneous bone-like apatite layer formation. Better results were obtained in a HAp coating sprayed on Ti64, SS316, and Inc625 substrates. These results gave a trend related to the specific experimental conditions presented in this work. However, they cannot be generalized since further research is needed to analyze ceramic/ceramic interactions more deeply during the coating buildup process, which may result in better coating microstructures and, consequently, different bioactive response in all systems.