Comparative Tribological Study of NiTi Diffusion Coated Titanium with Pure Titanium

Dry tribological behaviors of commercial pure (Cp) titanium and Cp titanium diffusion coated with equiatomic NiTi intermetallic layer were studied and compared at room temperature. Wear tests were performed by a pin on disk tribometer using 52100 steel pins, under various normal loads of 10, 20, and 40 N. worn surfaces were examined by scanning electron microscope, equipped with EDS analyzer. The wear rates of the coated materials were lower than those of the Cp titanium at all loads. This was mainly attributed to the higher hardness of the NiTi intermetallic layer compared to that of the untreated titanium. Furthermore, under an applying load of 10 N, a tribological layer was formed which could protect the surface from severe wear. The results also demonstrated a lower coecient of friction in the treated specimens compared to those of the Cp materials.


Introduction
Despite the high speci c strength and almost good corrosion resistance of titanium alloys [1,2], these materials exhibit poor wear resistance and high friction coe cient which restrict their application in items subjected to wearing conditions [3,4]. To overcome these de ciencies, several surface modi cation techniques like plasma nitriding, ion implantation, physical vapor deposition, liquid phase surface treatments, laser cladding, electroless, gas tungsten arc welding, and pulsed plasma electrolysis have been applied on these alloys [5][6][7][8][9][10][11][12]. The formation of diffusion coatings like nickel titanium compound (NiTi) has also drawn signi cant attention to combat wear [13][14][15][16]. NiTi intermetallic shows pseudoelasticity behavior [17,18] which is elastic responses to applied stresses, due to reversible martensitic phase transformation [19]. Recently several research groups employing various techniques to fabricate NiTi intermetallic as reliable protective layers on titanium and its alloys [20][21][22]. Tran et al. [15], studied the possibility of the formation of NiTi intermetallic phase via ultra-fast reaction of nickel droplet deposited by thermal spray on a titanium substrate. Waghmare et al. [21], fabricated a thick NiTi coating on Ti-6Al-4V by tungsten inert gas (TIG) surfacing method. The clad layer possessed a hardness value of more than 500 HV and the wear resistance improved up to 9.5 times compared to that of the substrate.
Mokgalaka et al. [22], produced NiTi intermetallic coating on Ti-6Al-4V alloy using a laser metal deposition process. Results revealed that the presence of NiTi, NiTi 2 , and NiTi 3 led to improvement in wear resistance up to 80%. Nevertheless, the aforementioned methods have limitations such as the formation of micro pores and inherent defects like cracks, expensive equipment, and operating complexity. Therefore, applying a diffusion-based treatment, which improves adhesion strength to the substrate, can overcome the defects of the previous methods [23]. The authors of the present work have reported the formation of Ni-Ti intermetallic coatings by nickel plating on commercially pure titanium followed by heat treatment [24]. A 20 µm thick layer with equi-atomic Ni-Ti composition is produced on the surface of Cp titanium via diffusion treatment. The aim of this work is to study and compare the wear behavior of commercial pure (Cp) titanium and Ni-Ti coated material. In this regard, the wear mechanisms are discussed based on the microstructural studies which provide an explanation to the enhancement of wear resistance in the coated samples in comparison with Cp titanium.

Materials And Methods
40 mm × 40 mm square shape specimens were cut from a 3mm pure titanium sheet. The specimens were grounded with emery paper up to 800 grid and then polished. To activate the titanium surface, the specimens were initially immersed in an aqueous solution containing copper sulfate 200 g/l, sulfuric acid 40 g/l, aluminum sulfate 24 g/l, and wetting agent 0.1% for one minute. They were then immersed in a second solution containing sodium dichromate 100 g/l, copper sulfate 5 g/l for one minute at 85 ºC, and hydro uoric acid (52%) 50 ml/l for ve minutes.
The activated specimens were then electroplated with nickel using Watts bath (containing 240 g/l NiSO 4 .6H 2 O, 30 g/l NiCl 2 .6H 2 O and 30 ml/l H 3 BO 3 ), at a current density of 20×10 4 mA/m 2 for one hour at room temperature to produce a 30 µm nickel coating. The distance between the cathode and anode during electroplating was xed at 2 cm. After the plating process, the samples were washed with deionized water and then dried. The coated specimens were then heat treated at 900°C for 12 hours in a controlled atmosphere furnace and were consequently cooled to room temperature in the furnace.
The phase structure of the coated layers was identi ed by X-ray diffractometer (XRD) using Cu k α (step Pin-on-disk wear test machine was used to measure the room temperature dry sliding wear rate of titanium before and after coating. For this purpose, the bare and coated specimens were prepared as 30 mm diameter disks. AISI 52100 steel pins with 5 mm diameter and a hardness of approximately 740 HV were used as counterface. The schematic diagram of this test is presented in Fig. 1. Before each test, the pins and discs were ultrasonically cleaned in acetone for 6 minutes then dried and weighed with a precision of 0.1 mg. The wear tests were performed under 3 different loads (10, 20, 40N) at a sliding speed of 0.25 m/s for a sliding distance of 500 m. Weight losses of the specimens were measured using a high-precision balance. The coe cient of friction was also recorded by a computational system during the wearing process. In addition, Hommelwerk T-8000 roughness tester was used to obtain grooves depth by measuring the roughness before and after were testing.

Results And Discussions
The cross sectional SEM image of the diffusion coated specimen in Fig. 2 shows two distinct layers from the surface to the core of the diffused layer on the titanium substrate. Based on the EDS results presented in Table 1, it appears that the outer layer is equiatomic NiTi with 15 µm thickness, and the bottom layer with 4.08 µm is NiTi 2 intermetallic. The X-ray diffraction pattern shown in Fig. 3 supports the EDS result which demonstrate the presence of NiTi, and NiTi 2 intermetallic in the treated layer. The wear rates of the Cp titanium before and after diffusion treatment as a function of load for a sliding distance of 500 m are shown in Fig. 5a. As can be seen in this gure, there is a signi cant increase in wear rate with the load increasing for both materials. Nevertheless, the wear rates of the treated material under all applied loads are much lower than those of Cp titanium. The wear rates of the counterface pins are shown in Fig. 5b. It is obvious that under three different loads the wear rate of the pins against the bare samples are signi cantly lower than those used for treated specimens. This could be a result of the lower hardness of the untreated material in comparison with the surface treated ones. Besides, increasing the load up to 40 N reduces the wear rate of the pin against Cp titanium to zero. It is believed that applying high loads results in temperature rise that in turn deteriorates the mechanical strength of the bare titanium and consequently the wear rate of the counterface pin is reduced to zero [25].
In order to eliminate the effect of the sliding distance, the speci c wear rate of disks was calculated by dividing the wear rate to the applying load. Fig. 6 shows the speci c wear rates of the Cp and diffusion coated specimen as a function of load. The gure re ects a dramatic increase in wear resistance after the aforementioned surface treatment. Fig.7 shows a typical depth pro le of the worn surface of the diffusion coated specimen achieved after 500 m sliding under 40N load. It appears that the intermetallic layer has been totally removed because according to Fig.2, the thickness of the layer was less than 20 µm;therefore,to evaluate the wear mechanism before losing the intermetallic layer, the sliding distance decreased to 200m for this particular specimen.  (Figure 8.a), it becomes apparent that the tribological layers in area A are thicker than those of area B, because the amount of iron and oxygen in area A are higher than those of the area B. Furthermore, cross sectional image in Fig. 9 also con rms the formation of the tribological layer with perfect bonding on the surface of the coated specimen during wear under an applied load of 10 N. It appears that the rate of the formation of oxides is higher than their removal. This can protect the surface from severe wear and results in the reduction of the wear rate as can be seen in Figure 5a. The remarkable wear rate of the pin against the diffusion treated sample under 10 N load (Fig. 5b) also con rms the existence of a hard tribological layer.
The SEM images of the worn surfaces of the Cp titanium produced under an applied load of 10N are also shown in Fig. 8a. The relevant EDS analyses of these worn surfaces reveal that the chemical composition of the region marked as A in this gure consists of 86.5% titanium, 12.5% iron, and 1% oxygen. These results demonstrate that a Ti-based tribological oxide layer was formed on the surface in regions like A. In addition, as is shown in gure 8a some cracks are formed on this tribological layer. It seems that the rate of the formation of the tribological layer is lower than its removal, so this layer could not protect the surface.
The micrograph and the EDS analysis of the worn surface of the coated specimen under 20N are shown in Fig. 8b. These results reveal that a mixture of titanium, nickel, and iron oxides as a tribological layer is formed in the region marked as C. The EDS analyses of region D and some broken regions indicate that the tribological layer is broken and removed. Therefore, the wear rate of the coated specimen under 20 N applied load is more than that of the 10 N. Furthermore, as can be seen in the SEM image of the worn surface of Cp Ti under 20N load, there is no oxide layer to protect the surface, and plastic deformation is the main wear mechanism which leads to increase of the wear rate under 20 N load in comparison with the diffusion treated specimen.
The micrograph and the EDS analysis of the worn surface of the coated material under 40 N (Fig. 8c) show that the mixed oxide layer yet remains on the surface in regions like E. This gure also illustrates the formation of debris that were able to aggravate abrasion and as a result, no protection layer was left on the surface in the region marked as F, which led to a higher wear rate as compared to those of similar specimens under lower forces (as shown in Fig. 6a). In addition, with due attention to the micrograph of the worn surface of Cp Ti under 40 N in Fig. 8c, it becomes clear that there is not any tribological layer to protect the surface. The plastic deformation and possible fatigue cracks are the main wear mechanisms in this specimen. Fig. 10 shows the cross section of the worn Cp Ti sample under an applied load of 40 N. It can be seen that holes and cracks are formed in depth below the surface and eventually result in the formation of debris with a delamination mechanism. In fact, by increasing the amount of the applied force, the surface temperature increases due to the heat generated by friction, and as stated in a previous research [25], the yield strength of titanium decreases by increasing temperature and as a consequence, the plastic deformation becomes more severe. Fig 11 presents the friction coe cient of treated and bare specimens versus sliding distance under 10, 20, and 40N loads. These gures both con rm the transition of wear mechanism by increasing the applied forces. As can be seen, under 10N load (Fig.11a) both the treated and the bare materials possess the lowest friction coe cient uctuations, which indicates a low adhesion between these samples and the counter facing pinsduring the wear test. This con rms the presence of the tribological layer on the surface of the samples because the formation of this layer reduces the adhesion andfriction coe cient by reducing metal-on-metal contact.By increasing the applied load up to 20N (Fig.11b) some uctuationsare noticed. This is because the adhesion has increased after 200m, which can be attributed to an increase in the metal-on-metal contact and subsequent failure of the oxide layer. Further rising in load value (Fig.11c) and consequently increasing the temperature on the surface during the course of wear, results in the surface oxide layer failure.Therefore, the wear mechanism changes,and the plastic deformation becomes the dominant mechanism of wear according to Fig.8c.Rising the load value causes an increase in the hardness of NiTi intermetallic phases that is due to the transformation of austenite to martensite [26].However, hardening does not seem to have a signi cant effect on the reduction of the wear rate, because the thickness of the treated surface layerdecreases by increasing the sliding distance, and the effect of titanium substrate increases. The comparison between treated and bare materialsreveals lower friction coe cients in all treated samples compared to those of theuntreated ones.

Conclusions
Commercial pure (Cp) titanium was surface treated with equiatomic NiTi intermetallic layer via electroplating and subsequent heat treating. Morphological and dry tribological tests were conducted to evaluate the behavior of diffusion coated and Cp titanium and the results summarized as follow: 1. XRD analysis revealed the formation of NiTi and NiTi 2 phases on the surface of the diffusion treated materials. The hardness pro le along the top of the fabricated layer toward the CP titanium substrate showed a steep drop.
2. SEM observation represented the formation of a tribological oxide layer on the surface of the treated sample under 10 N load in the course of wear. Applying higher forces led to aggravate abrasion and consequent removing of the oxide layer.
3. The wear rate of both Cp titanium and diffusion treated samples increased signi cantly by rising of the load, but there was a dramatic increase in wear resistance after the aforementioned surface treatment.
4. Studying the worn surface of Cp titanium revealed that there was a non-continuous oxide layer under 10 N load, and by further increasing the load plastic deformation mechanism prevailed which led to increased wear rates. 5. Under 10 N load, treated material showed the lowest friction of coe cient uctuations, but by further rising of the load, the amplitude of these uctuations increased.
Declarations Figure 2 Cross sectional SEM image of the nickel plated specimen after 12 h heat treatment at 900°C Page 11/15 Micro hardness pro le along the cross-section of Cp titanium and the diffusion coated material