Introduction

Additive manufacturing (AM) is an emerging manufacturing approach that is a more sustainable process than traditional, subtractive manufacturing. The AM process eliminates a significant amount of waste by creating parts in a layer-by-layer fashion, consuming only the amount of material necessary while eliminating the need for casting molds1,2. Not only is AM more sustainable than traditional manufacturing, but there is also now the possibility of the creation or recreation of complex or discontinued parts that were previously not feasible in industries like defense, aerospace, automotive, or energy3.

Cold spray (CS) additive manufacturing is a type of metal AM that operates with powder, typically metallic, as its feedstock. Powder is blended with a heated stream of gas and passed through a de Laval nozzle, promoting the powder particles to travel at supersonic speeds towards a substrate4,5. The velocity of the particles causes them to plastically deform and permanently adhere to the substrate and then to each other, building up a coating or a part. CS is unique as the process operates below the melting temperature of the metal feedstock powder. Keeping the powder in its solid state eliminates the complexities of the microstructural changes that would occur during melting and solidification. The microstructure of the powder can be examined and characterized before deposition and will be retained after deposition6,7.

One characteristic that dictates if a process gas is suitable for CS is molar mass. Gases with lower molar mass enable higher particle velocities in the gas stream, with potential for improved particle deformation8,9. From this perspective, Hydrogen (H2) would be the ideal gas to spray with because it has the lowest density and is readily available at a reasonable cost. However, there are many risks that come with handling H2 in a CS environment, as it is extremely flammable and explosive. Recently, there have been endeavors to research the feasibility of using hydrogen in the cold spray process, though it currently remains impractical10. The second-best gas would be helium (He). Helium is the lightest gas alternative to hydrogen, and with its inert characteristic, it is the prevailing gas choice in the cold spray community when the highest possible particle velocities are needed10,11. However, helium comes with several inopportune aspects that make its use unsustainable, starting with the gas being a non-renewable resource12. Helium co-occurs in small amounts with natural gas, so obtaining He requires drilling into the earth to extract it, which is environmentally invasive. It is also very costly and energy intensive to bottle10,12,13. Additionally, there are occasional shortages of helium due to fluctuating supply. Because of its high cost and the high carbon footprint associated with it, He may not be the most advantageous gas to use for some materials.

This challenge leads to nitrogen (N2) as the next alternative process gas. N2 is also a relatively lightweight and partially inert gas but much more economical to purchase when compared to He10. However, separating and bottling pure N2 gas is an energy intensive process, known as cryogenic distillation. This process involves the cooling of the atmosphere to extreme temperatures of −180 °C to −190 °C to condense the air. The air is then slowly reheated until the constituents of the air separate by their boiling point. The N2 is then contained and compressed into bottles, typically at 200 bar, for purchasing14,15. Given the energy intensive nature of the N2 process, the dangers of hydrogen, the cost of helium, as well as the reliance on the supply chain for all these gases and the carbon emissions associated with their transportation, a more sustainable option for CS process gas was considered14,16.

On-site nitrogen generation is a more sustainable approach that is rapidly increasing in the cold spray community when N2 is the appropriate gas for the application. Companies can generate their own nitrogen in-house by purchasing and installing their own generation system, a concept that was introduced in the 1940s by Leonard Pool14,15. The nitrogen is pulled from the air using a process called pressure swing adsorption (PSA), where the air is passed through carbon molecular sieves (CMS), allowing the nitrogen to pass through, while the oxygen in the air is adsorbed14,15,16. From a commercial perspective, this process uses approximately 28% less energy than cryogenic distillation as the only real energy consumption accrues from the compressor running the system16. Units can be modular, allowing them to be integrated into both laboratory and industrial settings. Nitrogen generation can also be more cost effective for companies and laboratories that use an abundance of nitrogen bottles to purchase the system once rather than buy cradles of bottles regularly14. While initial capital investment in the system is required, the scalability, reliability, and cost effectiveness of a nitrogen generation system is significant14,16.

PSA can also be modified with chill-type dehumidifiers and other moisture removing components, abbreviated in this study as PSA+, to further eliminate moisture exposure which may be advantageous in certain environments. Another existing way to generate nitrogen gas is a Liquid N2 to Gas N2 (L → G) system, a process in which liquid N2 is supplied to the system and converted to gas via evaporators and high-pressure compressors to high-pressure storage. While this latter approach retains the energy requirements of cryogenic distillation, it eliminates the energy used to transport bottles of pressurized nitrogen back and forth for replenishment. These generation systems all improve sustainability, cost effectiveness, and reliability; however, due to the nature of these processes, there is potential reduction in the overall purity of the nitrogen14,17. Nitrogen purity plays a role in particle velocity, gas stream temperature, and potential oxidation, all influencing the deformation and bonding required for cold spray18. Nitrogen gas with impurities can be heavier, affecting the necessary critical velocity of the particle while also having a lower specific heat, allowing particles to soften more in a warmer gas stream18. Oxidation is also possible with the presence of impurities, so there is a need to explore generated nitrogen with impurities before it can be considered as a potential replacement for traditional nitrogen in the cold spray process.

While cold spray studies utilizing impure nitrogen exist, the present work focuses on the nitrogen that can be produced with generation systems18,19. This approach is more sustainable than traditional processes and has the potential to be further optimized to improve or eliminate gas impurities. This study explores the effects of the reduction in nitrogen purity from on-site generated nitrogen as it compares to pure bottled nitrogen for cold sprayed specimens, as well as compares the discrepancies between multiple types of nitrogen generation systems in a two-phase study. The two metal powders used in the study, nickel (Ni) and chromium (Cr), were examined through a multitude of powder characterization techniques.

Preliminary laboratory observations indicated Ni powder to be tolerant of gas purity changes, in the sense that comparable consolidation thicknesses and densities were qualitatively observed across slight variations in N2 purity in previous trials. This consideration guided the selection of Ni powder for the first phase of the study to verify the functionality of the equipment used. The chromium powder was chosen for the second phase due to its relevance in ongoing laboratory research and unlike Ni, it was previously observed during initial laboratory trials to be sensitive to slight changes in gas purity. Cold spraying Cr has its challenges with its known sensitivity to oxygen and poor deformation once oxidized but has potential to be an advantageous material for coatings due to its ability to resist corrosion20,21,22. As such, it was selected for this study to further investigate and advance the utilization of Cr for cold spray application. Additionally, the consolidated specimens were characterized to obtain the mechanical properties of the coatings. The objective of this work is to validate that generated nitrogen is a feasible, sustainable process gas for cold spray comparable to pure bottled nitrogen, as well as compare discrepancies between different N2 generation systems.

Results

Nickel powder

Results from SEM imaging in Fig. 1 revealed the nickel powder to have a mostly spherical shape, further optimizing the acceleration of the powder in the gas stream and thus adhering and depositing more efficiently during the cold spray process23. Particle size distribution results for the nickel powder in the first phase of the study are shown in Fig. 2, along with quantitative values for the D10, D50, and D90 percentiles, which signify the particle diameter sizes below 10%, 50%, and 90% of the powder. The y-axis represents the probability density function and is referred to as the channel percentage, thus creating a distribution curve. The PSD results exhibit an adequately sized lot of nickel powder ranging from ~17 to 40 µm particles, which would optimize adhesion for cold spray24,25.

Fig. 1: SEM images of powder.
figure 1

a Loose nickel powder. b Cross-sectioned nickel powder particle. c Loose chromium powder. d Cross-sectioned chromium powder particle.

Full size image
Fig. 2
figure 2

Particle size distribution of nickel powder.

Full size image

Karl Fischer titration analysis determined the moisture content of the nickel powder shown in Table 1 to be ~64 ppm which is deemed suitable for cold spray applications26. Interstitial elemental analysis of oxygen, nitrogen, and hydrogen (ONH) content was determined for the nickel powder and also displayed in Table 1. The nickel powder had an oxygen content of approximately 360 ppm, a hydrogen content of about 13 ppm, and no nitrogen was found present. For cold spray applications, it is important to maintain low ONH contents in the powder to improve coating quality and the ability to build up a solid deposit27. This nickel powder had adequate ONH values necessary to achieve suitable cold spray deposits. Also represented in Table 1 is the nanoindentation data of the nickel powder, giving an average hardness and modulus for the nickel powder of 3.02 GPa and 128.47 GPa, respectively. This hardness value closely compared to literature on cold sprayed nickel suggested adequate ductility to plastically deform during cold spray and deposit a dense coating28.

Table 1 Characterization of nickel powder
Full size table

Nickel cold spray deposits

Figure 3 displays cross-sectional images at different magnifications of the consolidated samples after cold spray processing, with the nickel coating atop the Al6061 substrate. SEM images of the cross sections can also be seen in Fig. 4 to help visualize the coating structure. All three different types of process gases with varying N2 purity (compressed air, generated N2, and bottled N2) were able to deposit the nickel powder onto the substrate, producing suitable coatings of high thicknesses and low porosity, ideal for engineering applications. The thicknesses of the coatings are listed in Table 2. Visually, in terms of coating density, the coatings appear to be comparable despite being sprayed with different nitrogen purities. The larger thickness of the sample sprayed with the compressed air is attributed to a procedural difference of this sample getting more spray passes, reflecting additional deposited layers. Further characterization was performed to confirm that the specimens had comparable mechanical properties as well.

Fig. 3: Optical images of cross sectioned nickel coatings.
figure 3

a Nickel sample sprayed with compressed air, 78% N2. b Nickel sample sprayed with generated nitrogen, 97.6% N2. c Nickel samples sprayed with pure bottled nitrogen, 99.998% N2.

Full size image
Fig. 4: SEM images of cross sectioned nickel coatings.
figure 4

a Nickel sample sprayed with compressed air, 78% N2. b Nickel sample sprayed with generated nitrogen, 97.6% N2. c Nickel samples sprayed with pure bottled nitrogen, 99.998% N2.

Full size image
Table 2 Coating thicknesses of nickel depositions
Full size table

Figure 5 shows the average apparent porosity in each of the specimens, obtained through optical microscopy and image analysis. It is advantageous for cold spray deposits to have minimal porosity throughout the coating to promote strong, dense structures for engineering applications. The samples sprayed with the different process gases demonstrated comparable results for porosity percentages. This signifies that the 2.4% difference in purity of the N2 of the two process gases has negligible effects on the porosity of the Nickel coatings.

Fig. 5
figure 5

Average porosity percentage of nickel coatings sprayed with different nitrogen purities.

Full size image

Microhardness results of the nickel specimens are shown in Fig. 6. All three specimens sprayed with varying N2 purity show comparable Vickers hardness results. The two specimens of initial interest (the generated N2 and bottled N2) were even more closely comparable, only differing from each other by two hardness points. The data implies that the difference in nitrogen purity has negligible effects on the microhardness of the nickel coating samples.

Fig. 6
figure 6

Average Vickers microhardness of nickel coatings sprayed with different nitrogen purities.

Full size image

Figure 7 displays the nanoindentation data with hardness results in the left graph and modulus results in the right. It should be noted that the hardness results of the deposited specimens in Fig. 7 are slightly higher than the hardness results of the powder in Table 1, which is to be expected as the powder plastically deforms and induces cold work when the powder deposits during cold spray, thus strengthening and stiffening the material slightly25,29. The hardness data for all three specimens are very closely comparable as well as the data for the modulus. This again suggests that the varying N2 percentages of the process gas had negligible effects on the hardness and modulus of the nickel coating samples.

Fig. 7: Nanoindentation data of nickel coatings sprayed with different nitrogen purities.
figure 7

a Nickel coatings nanohardness in GPa. b Nickel coating modulus in GPa.

Full size image

Chromium powder

The chromium powder for the second phase of this study was also characterized prior to cold spray processing. PSD results determined the particle diameter ranged from ~17 µm to 45 µm which can be seen in Fig. 8 along with the distribution curve and the D10, D50, and D90 values. Similar to the nickel powder, the size range was appropriate for cold spray, indicating that proper adhesion and buildup could be achieved24,25. The SEM images of the Cr powder in Fig. 1 reveal that the particles are more irregularly shaped, as opposed to spherical, having uneven morphology and appearing more flake-like. This can be attributed to the atomization technique of the Cr powder and indicates likely increased acceleration in the gas stream but also potential suboptimal effects on cold spray deposition23,25.

Fig. 8
figure 8

Particle size distribution of chromium powder.

Full size image

The Cr powder was characterized further with results displayed in Table 3. KF Titration measured the powder to have a moisture content of approximately 205 ppm, slightly higher than expected. However, powders with much higher moisture content have been studied and deemed suitable for cold spray depositing dense coatings, validating that this Cr powder was acceptable for cold spray26. ONH results are also displayed in Table 3 with relatively high values for each element, specifically oxygen. The elevated oxygen content could signify potential challenges in achieving sufficient deposition27. While the nitrogen and hydrogen contents of the powder were higher than that of the nickel powder in the first phase, these values of approximately 63 ppm and 37 ppm, respectively, did not indicate potential issues during the cold spray processes. Also displayed in Table 3 are the results from nanoindentation of the powder measuring hardness and modulus in GPa. The chromium powder had a hardness of almost 6 GPa, which could result in more difficulty deforming during adhesion30.

Table 3 Characterization of chromium powder
Full size table

Chromium cold spray deposits

Figure 9 presents the cross-sections of the chromium powder cold sprayed onto the Al6061 substrates with process gases produced from different nitrogen generation systems at two magnifications. Figure 10 shows SEM images of the cross-sections to better visualize the coating structure as well as EDS scans of oxygen on the chromium. The quantitative values for the coating thicknesses are listed in Table 4. There is a noticeable difference in the coating thicknesses for the specimens sprayed with the PSA system and the PSA+ system. These systems produced process gases with nitrogen percentages of 97.6%. The other generation system, the L → G system, produced a process gas with a nitrogen percentage of 99.99%, yielding a much denser coating, closely resembling the baseline specimen that was sprayed with pure bottled nitrogen. It is thicker than the sample sprayed with the bottled nitrogen due to this sample receiving more passes in the spray process, but the densities appear to be comparable.

Fig. 9: Optical images of cross sectioned chromium coatings.
figure 9

a Chromium sample sprayed with pure bottled nitrogen, 99.998% N2. b Chromium sample sprayed with PSA generated nitrogen, 97.6% N2. c Chromium samples sprayed with PSA+ generated nitrogen, 97.6% N2. d Chromium samples sprayed with L→G generated nitrogen, 99.99% N2.

Full size image
Fig. 10: SEM images of cross sectioned chromium coatings with corresponding EDS scans.
figure 10

a1-2 Chromium sample sprayed with pure bottled nitrogen, 99.998% N2. b1-2 Chromium sample sprayed with PSA generated nitrogen, 97.6% N2. c1-2 Chromium samples sprayed with PSA+ generated nitrogen, 97.6% N2. d1-2 Chromium samples sprayed with L→G generated nitrogen, 99.99% N2.

Full size image
Table 4 Coating thicknesses of chromium depositions
Full size table

There is also a difference in the EDS scans, with more oxygen present in the samples sprayed with the PSA system and the PSA+ system. These scans also produced measurements that, while subject to slight variability, supported the interpretations of the maps. The difference of deposition seen in Fig. 9 and the difference in oxygen presence in Fig. 10 seem to be caused by the 2.4% difference in purity from the bottled N2 gas to the N2 gas created by the PSA and PSA+ systems. Whereas the 0.008% difference in purity from the bottled N2 to the L → G gas seems to have little effect on the apparent porosity. This would suggest that the chromium powder is more sensitive to a reduction in nitrogen purity; however, further mechanical testing was done to corroborate the qualitative observations.

Porosity percentages of the chromium coatings are presented in Fig. 11. The specimens sprayed with the lower purity nitrogen of 97.6%, the PSA and PSA+ systems, yielded results with extremely high porosity. This level of porosity is deemed impractical and unsuitable for engineering applications. However, the specimen sprayed with the higher N2 percentage, from the L → G system, produced a dense coating with limited porosity, comparable to the pure bottled N2 sample. These porosity measurements signify that in terms of porosity; chromium is sensitive to nitrogen purity and only pure bottled N2 or generation systems producing essentially pure N2 should be used to spray it.

Fig. 11
figure 11

Average porosity percentage of chromium coatings sprayed with different nitrogen generation systems.

Full size image

Figure 12 displays the Vickers microhardness results of the chromium powder deposits. The L → G sample and the bottle sample had comparable hardnesses, with the L → G sample being slightly harder. This comparison indicates that generated N2 of comparable purity to bottled N2 produces samples that achieve the same hardness values at those sprayed with bottled N2, validating the intention of the study. However, slight reduction in N2 purity will not yield the same results as seen in the samples that were sprayed with the PSA system and the PSA+ system, yielding a lower microhardness.

Fig. 12
figure 12

Average Vickers microhardness of chromium coatings sprayed with different nitrogen purities.

Full size image

Figure 13 presents the nanoindentation data for the chromium coatings. Of the three samples sprayed with generated nitrogen, the L → G system produced the hardest coating. This aligns with expectations because of the data presented so far, this specimen has had the most ideal properties, being the densest and having the highest Vickers microhardness. The sample sprayed with N2 from the PSA+ system also had high hardness; however, due to its insufficient adhesion, reliability of this hardness measurement is questionable. It should be noted that the hardness results from nanoindentation of the cold sprayed deposits in Fig. 13 increased from the initial powder hardness in Table 3. This can be attributed to the tendency of chromium to work-harden upon impact of the substrate during the plastic deformation of cold spray which could have further performance implications and poor adhesion25,29.

Fig. 13: Nanoindentation data of chromium coatings sprayed with different nitrogen purities.
figure 13

a Chromium coatings nanohardness in GPa. b Chromium coating modulus in GPa.

Full size image

Discussion

The mechanism supporting the greater difficulty of the chromium powder in producing a well-consolidated coating with lower purity nitrogen may stem from the tendency of chromium to produce tenacious or extremely adhesive oxides. Stainless steel with chromium as a constituent utilizes these oxides to protect itself from rust and corrosion22. However, these oxides are not as advantageous in cold spray applications as they are in stainless steel, as they interfere with the cold spray bonding process and increase the brittleness of the powder31,32. These oxides are especially problematic when the powder morphology is flake-like, similar to the powder in this study, thus having higher surface area and more room for oxygen exposure23,25. While irregular morphology can assist in mechanical bonding, oxides have a greater effect on the bonding of these particles, due to the increased, complex surface area promoting thicker oxide layers. Additionally, the residence time of the chromium powder particles in the gas stream is relatively short; therefore, the expected oxide layer of the particle is to be relatively small. However, the presence of surface oxides on cold spray powders hinders the plastic flow of the particle and requires higher critical impact velocities for effective bonding33,34. So, while the paramount indicator of proper cold spray bonding is impact velocity, it is secondarily dependent on oxide layer thickness33. And if the particles have irregular morphology and therefore thicker oxide layers, effective bonding will be difficult to achieve.

The Ellingham diagram is a tool used to predict a metals ability to reduce or form an oxide based on the temperature dependence of the Gibbs free energy of various elements. Based on its location on the Ellingham diagram, chromium is sensitive to the presence of oxygen as well as moisture exposure, as it lies closely to titanium, a metal notorious for oxide formation35. This likely explains why the powder, prior to spraying, measured such high moisture content in KF titration with 205.48 ppm and such high oxygen content on ONH analysis with 1260.67 ppm. These initial powder properties of high moisture and high oxygen content were not an obstacle when the process gas was high purity nitrogen from the L → G system. In this system, the high purity gaseous nitrogen resolved from high purity liquid nitrogen, where there would be no presence of oxygen contamination, which may be why it performed most closely to the bottled nitrogen. However, when the nitrogen was generated by PSA systems, there were impurities present, mainly leftover oxygen from the atmosphere from which the system was pulling. These impurities affect the gas flow dynamics, hindering particle velocity and temperature, which influence deposition and bonding. The chromium powder already must undergo extremely high process conditions to account for its inopportune moisture and oxygen sensitivities to construct a quality deposit, so any fluctuation in spray parameters, including process gas purity, may alter the ability of chromium to be sprayed adequately. This can be seen in the EDS scans of oxygen on the chromium coatings, where the two specimens with lower nitrogen purity, and therefore greater oxygen impurities, have more oxygen present in the deposits. The chromium particles oxidized in the gas stream and hindered the ability to produce a dense coating.

Nickel does not have the same oxide challenge as chromium which may be the reason it had no complications spraying with lower purity nitrogen. Nickel is also highly deformable, which makes it an ideal powder for cold spray36. Chromium, however, is much harder and less malleable, which is again why chromium must be sprayed with extremely high processing conditions just to produce an adequate coating. The powder is very hard to begin with, and during the cold spray process, it has difficulty bonding to itself, an already hard surface, in order to build up layers20,21,25,29,32. This is found both from the nanoindentation results of the powder prior to spraying, 5.94 GPa, and from the results of nanoidentation testing of the cold spray deposits, 7.5 GPa. Again, when all the parameters are ideal, including high nitrogen purity, as seen in the L → G and bottled specimens, the effects of the deformation issues of chromium are not insurmountable obstacles. However, the moment a parameter is altered, in this case, the process gas purity in the PSA systems, the sensitive properties of the chromium prevail, resulting in a lack of adhesion and poor coating quality.

This work explored the functionality of on-site generated nitrogen as a potential process gas for cold spray applications in two phases with two metal powders. Phase I examined a CP nickel powder and compared 97.6% on-site generated N2 to 99.998% bottled N2 to determine if nitrogen generation, with its reduction in purity, was a plausible substitute for the expensive, energy intensive bottled N2 for cold spray. Analysis of the investigation and data from this phase of the study resolved the following conclusions:

  • CP nickel powder can be sprayed with lower purity generated N2 without apparent issues, resulting in coatings that are comparable to specimens sprayed with bottled N2 in terms of thickness and properties, suggesting CP Ni is not sensitive to slight reductions in nitrogen purity.

  • Compressed air (78% N2) is a valid process gas for cold spraying CP Ni, producing sufficient coatings and adequate properties, further suggesting that nickel powder is not sensitive to nitrogen purity.

  • CP nickel powder undergoes slight hardening during cold spray deposition, but this effect does not appear to restrict its ability to produce thick coatings.

Phase II of this work investigated a chromium powder and compared various existing nitrogen generation systems to determine if similar results could be manufactured across the different generation processes that produced varying nitrogen purities. Conclusions from analysis of the experimentation and results from the second phase of the study are as follows:

  • Chromium powder is sensitive even to slight reductions (2.4%) in nitrogen purity which affects its performance and ability to deposit a sufficient coating. Therefore, samples sprayed with 97.6% PSA generated N2 cannot easily achieve the same deposit quality as specimens sprayed with nitrogen purities equivalent to bottled N2.

  • However, a liquid-to-gas nitrogen generation system, achieving a N2 purity of 99.99%, yields similar results to those specimens sprayed with bottled N2 for chromium powder, verified by coating thickness and mechanical properties.

  • The high oxygen content and initial hardness of the chromium powder had minimal impact on its ability to deposit, as the samples sprayed with high purity nitrogen maintained high deposit quality.

  • Chromium powder undergoes high work hardening during cold spray deposition, but this effect does not appear to limit its ability to create sufficient coatings when sprayed with high purity nitrogen.

Beyond the knowledge gained from this study, future work will investigate the ability of other metal powders to be sprayed with generated nitrogen in terms of suitable deposition and mechanical properties for engineering applications. The two powders examined in this study are just two commonly sprayed metals, one with seemingly no intolerance to a depletion in N2 purity and one with major sensitivities. Examining more commonly used powders in cold spray and learning of the powders that have sensitives can assist in efforts of solely using generated N2. Future efforts will also expand on enhancing the purity of the N2 produced from generation systems so that metals powders with sensitivities can still have the option to be sprayed with generated nitrogen. These efforts to utilize in house generated nitrogen as a process gas wherever possible are critical for keeping cold spray sustainable.

Methods

Powder details and characterization

Phase I of this study used a gas atomized commercially pure (CP) RIT 176 nickel powder (Höganäs GmbH, Germany) to compare cold spray with bottled high purity nitrogen versus generated nitrogen. Phase II of this study used a CP chromium powder (Powders on Demand, Worcester, MA, USA) to compare different nitrogen generation systems.

Both powders were characterized before processing for moisture content, particle size and shape, chemical composition, and nanohardness. The moisture content was obtained using a Mettler Toledo Coulometric Karl Fischer (KF) C30S Titrator equipped with an InMotion KF Oven Autosampler Pro (Columbus, OH, USA). For both powders, three glass vials with approximately 1.5 g of powder in each were sealed and loaded onto the KF apparatus. One at a time, the system punctured the plastic tops of the vials with a needle to surge them with desiccated air and then load them into the oven. The oven heated the samples to 220 °C for five minutes to desorb the moisture from the powder and then direct the desiccated air, now containing the moisture, out of the vial through tubing to initiate titration between the moisture and an iodine solution. An electrode measured the voltage change of the reaction, thus converting and determining the moisture value of the powder sample in ppm.

The particle size distribution (PSD) of the powders was measure by a Microtrac TurboSync system (Microtrac Retsch GmbH, Haan/Duesseldorf, Germany). The equipment determined particle size and shape with dynamic image analysis as well as laser diffraction through a cascade of particles in a stream of air. Interstitial elements oxygen, nitrogen, and hydrogen were detected with an Eltra Elemental Analyzers Elementrac ONH-p2 (Eltra Retsch GmbH, Haan/Duesseldorf, Germany). The system was calibrated with AR675 ON Steel Pin CRM and AR555 H Steel CRM standards for the nickel powder and AR642 Titanium O/N/H CRM standards for the chromium powder. The powders were then mounted and polished with a Buehler EcoMet 300 automatic grinder-polisher (Lake Bluff, IL, USA) to a mirror finish to obtain access to the cross-section of the powders. A Zeiss EVO MA-10 scanning electron microscope (SEM) (Carl Zeiss AG, Oberkochen, Germany) was used to take images at 10 kV and examine the particle morphology of the loose powders with secondary electron emission as well as the cross-sections in backscattered electron imaging.

A KLA Corp. iMicro Pro (Milpitas, CA, USA) was used to perform nanoindentation on the cross-sections of the mounted powders. Initially, the system was calibrated for effective function and frame stiffness by testing a fused silica standard. Once calibration was completed, an operational check was enacted to ensure that all the mechanical components of the system were functioning properly. Each powder sample received an indent on 30 particles with a maximum load time of 15 s from a diamond Berkovich tip from Micro Star Technologies Inc. (Huntsville, TX, USA) paired with an InForce 1000 mN actuator. The nickel powder was assumed to have a Poisson’s ratio of 0.31 with a target depth of 240 nm and the chromium powder was assumed to have a Poisson’s ratio of 0.24 and target depth of 260 nm. The InView Software program (Version 19.2.24) was used to process the data and remove insufficient indents. The data processing also assisted in comparing all the indents to a relative depth of 200 nm and accounted for any thermal drift and pile up in the hardness results.

Cold spray

Both powders for both phases of the study were sprayed onto aluminum 6061 (Al6061) substrates of dimensions of approximately 15 cm × 2.5 cm with varying thicknesses. The powders were sprayed with a VRC Raptor Cold Spray System. An overview of the samples that were sprayed is shown in Table 5. As stated above, PSA signifies N2 gas generated via pressure swing adsorption, PSA+ signifies N2 gas generated via pressure swing adsorption equipped with additional dehumidifiers, and L → G signifies N2 gas generated from liquid N2.

Table 5 Summary of samples sprayed
Full size table

Phase I of the study, comparing bottled nitrogen to generated nitrogen, used compressed air as a baseline sample to present a range of increasing nitrogen purity percentages and therefore better demonstrate if change in purity affects the deposition. Phase II of the study, comparing different nitrogen generation systems, used bottled nitrogen as the baseline to demonstrate the effects of pure nitrogen before comparing across the different systems. The types of nitrogen generation systems observed were PSA, PSA + , and L → G. The purities of the nitrogen gases were determined through a certificate of conformance from A&B Welding Supply Co. Inc. (Rapid City, SD, USA) for the bottled N2 and a gas analysis report from Atlantic Analytical Laboratory (Readington, NJ, USA) for the generated N2. The spray parameters or “recipes” that were used during each spray were consistent within each phase of the study and are shown in Table 6.

Table 6 Summary of cold spray process parameters
Full size table

Cold spray characterization

Small samples were cut from the consolidated specimens on a Buehler IsoMet High Speed Pro Automatic Precision Sectioning Machine (Lake Bluff, IL, USA) to be metallographically prepared in a Buehler SimpliMet 4000 Hot Mounting Machine (Lake Bluff, IL, USA). The samples were then polished using a 6-step process to achieve a mirror finish. They were then imaged and characterized to determine their mechanical properties. Optical images were taken of the samples on a Keyence VHX-X1 Digital Microscope (Itasca, IL, USA) equipped with image analysis software to determine the coating thicknesses and the apparent porosity of the samples. SEM images were also taken to observe the coating structure in combination with energy dispersive X-ray spectroscopy (EDS) using a Bruker X Flash Energy Dispersive X-ray Spectroscopy Detector 630 M (Bruker Nano GmbH, Berlin, Germany). The Vickers microhardness was also measured for each sample using a Buehler Wilson VH3300 Automated Hardness Testing System (Lake Bluff, IL, USA). 25 indents were made in each specimen with a force of 0.3 kg for 10 s. Nanoindentation was done on the deposit specimens, using the same instrument that was used on the powders, to obtain hardness and modulus of the coatings. The nanoindentation parameters were kept the same between powder and deposit material, i.e. the Cr deposits were nanoindented using the same parameters as those used for the Cr powders. The same applied to the Ni deposits and 30 indents were made in each specimen. The properties were then analyzed and compared to determine if there was variation in the properties of cold sprayed specimens after using different nitrogen purities and different nitrogen generation systems.