Surface modification of tool steel by cathodic cage TiN deposition

The aim of this work is to investigate the effect of titanium nitride coating for various treatment times (0.5–4 h) by cathodic cage plasma deposition (CCPD) on surface properties of AISI D6 tool steel. The obtained results depict micrometric-sized TiN film deposition under all processing condition with improved surface hardness and corrosion resistance. Raman and EDS analysis are used to calculate N/Ti ratios for stoichiometric calculations of samples. This study depicts that the surface properties of tool steel can be effectively improved by TiN film deposition using CCPD, with low processing time, low processing temperature and better uniformity, as compared to conventional techniques. Additionally, the problems associated with the conventional TiN films such as pores and voids are eliminated. This technique is compatible with industrial-scale applications, and thus results from this study are expected to be beneficial for the surface engineering industry.


Introduction
Grade D tool steels are high-carbon, high chromium, cold-work tool steels, characterized by high resistance to wear due to the presence of chromium carbides, as well as low deformation when submitted to thermal treatment.As a result, they are also known as 'undeformable' steels [1].One of the most widely used Grade D tool steels is AISI D6, containing carbon, chromium, manganese, tungsten and vanadium [2].Such steel can be used in several applications such as cutting tools and forging dies, because it is preferable to use low-cost material [3].Unfortunately, tool steels usually exhibits a high wear rate due to surface faults arising from high attrition-rate coefficients [3].
The deposition of coatings with low attrition-rate coefficient and better adhesion is suitable to improve the service life of tool steel.The titanium nitride coatings (TiN) are extensively used to improve the surface properties due to excellent surface hardness, good thermal as well as electrical conductivity, corrosion resistance, low friction coefficient and low adhesive wear rate [4].Such coating can be deposited by several techniques including chemical vapour deposition (CVD) [5], dc and rf-magnetron sputtering [6], physical vapour deposition (PVD) [7,8], plasma-assisted chemical vapour deposition [9], plasma focus device [10] and molecular beam epitaxy (MBE) [11].The deposition of TiN film using CVD usually demands high processing temperature (900-1100°C) [5], whereas such coating can be achieved in PVD at low processing temperature down to 200°C [4].However, its deposition efficiency is quite low and thus the formation of a thick layer is quite expensive [12].If the thickness of TiN coating is not sufficient, then its resistance to sliding wear is reduced due to less mechanical supports of the substrate.Also, PVD TiN films show local corrosion due to the presence of inherent defects like cracks and pinholes [12][13][14][15].The dissimilarity in the structure of TiN coating and substrate causes an apparent interface, thus adhesion of film on the substrate is quite problematic.The plasma focus device is an efficient source for thin film deposition due to the impact of energetic ions, but its performance is dependent on a number of focus shots, which may cause voids and cracks in the film while the number of shots exceeds to certain limit [16].The deposition of films using MBE demands low processing temperature, but it has lower throughput than other techniques (such as CVD) [17], thus not very effective to improve surface properties of steels on industrial scale.
Around two decades ago, an innovative surface engineering technique called as cathodic cage plasma nitriding (CCPN) was introduced, in which samples are encased in a geometrically defined screen and maintained in floating potential, i.e. samples are not polarized as they are placed on an insulator ceramic disc [18].This way, plasma acts solely on the cathodic cage wall, and samples are not subjected to surface defects caused by intensive ions bombardment.This involves the contribution of hollow cathode effect inside the holes, thus it is an efficient technique.The working mechanism of this technique involves atoms detachment from the cage, which combines with active plasma species and forms nitrides, which by way of directional conductivity are encouraged to deposit on sample surfaces [19].The uniqueness of this nitriding technique is cathodic cage material deposition on samples surface, thus beside thermochemical diffusion (which occur in conventional nitriding [20][21][22]), it can be used effectively for sample alloying with cage material.
This system is widely used for deposition of several coatings on steels such as: silver alloying by Lin et al. [23] as well as Dong et al. [24], niobium alloying by Lin et al. [25], copper alloying by Dong et al. [26] and aluminium alloying by Naeem et al. [27].Daudt et al. [28,29] deposited the TiN coating on a glass substrate and investigated the effect of cage configuration on surface microstructure.Yan et al. [30] modified the surface properties of copper-aluminium alloy by TiN coating, as a result, fine and uniform coating with better hardness is observed.Sowínska et al. [31] found this technique to be efficient for the surface modification of Ti-6Al-4 V alloy by depositing TiN film while keeping the samples at plasma potential inside the cathodic cage.Yazdani et al. [4] synthesized nano-sized TiN on a steel substrate.The thickness of the modified layer is reported to be 3.12 μm after 10 h of treatment, while no mechanical features were tested.Nishimoto et al. [32] treated the SACM645 steel using a titanium cage and synthesized TiN-coating.The thickness of the coating was reported to be 3.3 μm after 15 h processing, while mechanical features are found to be improved.Lopes et al. [33] synthesized TiO 2 anti-corrosive thin film on duplex stainless steel using Ti cathodic cage.In our previous articles [34,35], we used titanium cathodic cage for the deposition of titanium-based coatings (oxide and nitride) on silicon and glass substrate.These literature reports depict the effectiveness of CCPN with titanium cathodic cage for the deposition of titanium-based coatings on several substrates.However, the effect of TiN coating using CCPN for the improvement of surface properties of D6 tool steel is not reported so far to the best of author's knowledge.
In this study, we attempted to improve the surface properties of tool steel using titanium cage by processing samples for comparatively small processing time (0.5-4 h).The hardness is found to be significantly improved and coating thickness of ∼7 μm after 4 h processing (which is more than twice, even in very short processing time as compared to previous reports on steels [32,33]).The corrosion resistance is also found to be enhanced.

Cathodic cage plasma deposition system
Conventional nitriding equipment, with the addition of a cathodic cage, was used for titanium nitride film deposition on D6 tool steel substrates, as shown in Figure 1 [30].The equipment consists of a cylindrical chamber 30 cm in diameter and 40 cm height, being constructed of austenitic stainless steel.For deposition, a high voltage source (1500 V, 2 A) was used.In order to increase deposition rates, two cages were used in the nitriding reactor [32].Cages consisted of two concentric cylinders made from 2 mm thick titanium sheet.The larger cage measured 65 mm in diameter and 45 mm in height, while the smaller cage measured 45 and 35 mm respectively.Titanium discs were used as lids.Holes of 8 mm diameter were distributed on the walls of both cages, with a distance of 9.2 mm between the centres of adjacent holes.A 30 mm × 3 mm alumina disc was used to isolate samples, in order to ensure that they are not polarized.As samples are maintained in floating potential, plasma does not act directly on sample surfaces, but rather on the cage surfaces which are polarized.Samples were placed inside the smaller cage, which was encased by the larger cage in a concentric manner.Before use, these cages were sanded using 360 MESH sandpaper and ultrasonically cleaned in a 50 mL of HNO 3 , 25 mL of HF, and 425 mL of H 2 O solution for 10 min.

Samples preparation and processing
Samples of D6 tool steel (chemical composition given in Table 1) with 2.54 cm radius, 1 cm thick circular sector were used throughout the study.Before treatment, samples were cut and sanded using 200, 320, 400, 600 and 1200 MESH wet sandpaper, polished by alumina felt discs to acquire a reflective feature, and ultrasonically cleaned in an acetone solution for 10 min.
After cleansing, both samples and cages were positioned inside the reactor to undergo pre-sputtering in hydrogen (H 2 ) atmosphere, in order to remove any oxides which could compromise treatment.All the processing conditions used in this study are tabulated in Table 2.

Sample characterization
After treatment, samples were characterized using scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (X-EDS), Raman spectroscopy, Xray diffraction (XRD), atomic force microscopy (AFM), potentiodynamic polarization test and Vickers microhardness testing.The details regarding samples analysis can be found in the previous report [12].

SEM and EDS observation
The cross-sectional SEM images of samples treated for various time is depicted in Figure 2. It shows that deposited films are uniform and regular, completely covering all samples without substrate separation.The sample treated for a shortest treatment time of 0.5 h shows an average thickness of 1.43±0.07µm, which is of relatively significant thickness over conventional techniques, thus it indicates that depending upon the required application, processing time can be reduced, which is cost-effective.Furthermore, observation shows that layer thickness increases with treatment time, especially from 2 h onwards.This is consistent with the reasoning that layer thickness is directly proportional to the square root of time [36].It is worth highlighting that samples treated during 0.5, 1 and 2 h do not show significant difference in layer thickness.This is especially visible on samples treated during 1 and 2 h which show layers of approximately 3.38 and 3.48 µm respectively.This behaviour is due to the fact that nucleation starts between 0.5 and 1 h, and continues for up to 2 h.During this time layer growth is insignificant, as nucleation forms an even layer across the whole surface of the sample, thus resulting in constant thickness [36].Significant increases of thickness occur from 3 h onward.The synthesized layer does not show any evidence of inherent drawbacks of TiN-coatings such as voids or cracks which appear in conventional techniques [12][13][14][15], and thus it is expected to more valuable technique.
For the elemental analysis of modified layers, EDS is carried out at 8 different points on each sample to check the uniformity of the coating at various positions.The average value of the elemental composition is tabulated in Table 3.The elemental composition of nitrogen and titanium at various points is found to be quite similar, which indicates better uniformity of synthesized TiN coating using the cathodic cage plasma deposition (CCPD) technique.Table 3 shows the presence of a significant amount of titanium on the samples, even treated for the shortest time (0.5 h).This is the direct result of two cages, which increases the deposition rate even for the shortest treatment time, and thus thick layer is formed.Table 3 shows that nitrogen and titanium atomic percentages are quite closer for all treatment time, with the exception of 4 h treatment where the difference is greater.It is worth noting that with increased treatment time, especially over 1 h, there is a greater gap between atomic percentages of titanium and nitrogen.Nitrogen atomic percentages increase, while titanium decreases.This is explained by the reduction in cage sputtering (due to the formation of titanium nitride on cathodic cage during initial treatment time), in turn, results in a higher nitrogen concentration on the sample surface and a lower amount of titanium.In general, N/Ti ratios are closer or greater than 1, which indicates stoichiometric titanium nitride, justified by the golden colour of treated samples.According to Roquiny et al. [37], N/Ti ratios vary the colour of resulting titanium nitride from light grey to brownish-red, and gold in stoichiometric conditions.Furthermore, Smith et al. [38] obtained a high N/Ti ratio TiN films without destroying stoichiometric properties.

Raman spectroscopy analysis
The crystal structure of stoichiometric TiN is face-centered cubic (FCC).For FCC structure, each ion is in inverted site symmetry and there is no change (at first order) in the polarizability as a function of the vibrational normal modes of the structure.Thus, there is no Raman signal for the stoichiometric TiN, however, the presence of vacancies of Ti or N in the lattice of FCC can promote symmetry breaking in the vacancies sites allowing the TiN non-stoichiometric Raman modes appearance [39,40].
In order to assess film uniformity as well as enable stoichiometric calculations of N/Ti ratios, Raman spectra of eight distinct points of each sample were obtained, as depicted in Figure 3.The appearance of the Raman bands at 215, 300 and 580 cm −1 confirm titanium nitride formation for all treatment times.Montgomery et al., Spengler et al. and Chen et al. [39][40][41] attribute Raman scattering to heavy Ti ion vibrations in the acoustic range (typically 150−300 cm −1 ), and to lighter N ion vibrations in the optical range (typically 400  −650 cm −1 ).Spengler et al. [40] defined that for stoichiometric TiN, TA and LA acoustical band first-order peaks range from 100 to 350 cm −1 , while TO and LO bands show peaks range from 450 to 700 cm −1 .The results obtained in this work are consistent with the aforementioned, as all peaks for all treatment times are within the mentioned peaks, as highlighted in Figure 3.The peak found in the range from 450 to 700 cm −1 is a result of overlapping TO and LO bands which hinders peaks distinctions [42].Improved visualization of the Raman spectra bands may be possible by residual analysis and Gaussian/Lorentzian curve fitting [43,44].Following spectra adjustment, stoichiometric TiN film calculation is possible by calculating the TiN area fraction using the (TO+LO)/(LA+TA) equation as discussed by Cheng et al. and Vasconcellos et al. [45,46].Table 4 demonstrates N and Ti values obtained after Raman spectra adjustments for each treatment time.Raman and EDS analyses support that titanium nitride film is deposited on D6 steel for all treatment time.Furthermore, these analyses clearly show uniform layers, as tests conducted on different sample points using different techniques show very similar results.

X-ray diffraction
XRD is used to identify phases as presented in Figure 4.As can be seen, and as expected, TiN is formed for all treatment times.The diffractograms show the existence of both TiN and Fe peaks, the latter resulting from the substrate.Upon comparison of samples, there is a clear plane displacement with an increase in treatment time.This may be a result of the stoichiometric difference between films formed under different treatment conditions [46].The change in peak intensity may indicate film growth in different preferential directions for each treatment condition.This is directly related to texture effects.The results from XRD are in good agreement with the elemental composition obtained from EDS and Raman analysis.

Surface topography by AFM
The surface topography of samples can be observed by AFM imaging technique, as presented in Figure 5.The topography depicts that numerous particles are deposited on a short time treated sample, which are decreases with increase in time.On the other hand, with the decrease in a number of particles, the size of particles grows on.The height of surface nitrides is found to be increasing with an increase in treatment time.This trend is credited to the agglomeration of further incoming particles with the increase in treatment time, as described by CCPN mechanism in several reports [19,47].This result is in quiet agreement with   the cross-sectional SEM images which shows a growth in the nitrided layer with an increase in treatment time.

Microhardness analysis
The surface microhardness of treated samples as a function of treatment time is depicted in Figure 6.
Results show a significant increase in sample microhardness right from the shortest treatment time, as compared to annealed AISI D6 steel without the titanium nitride coating which has a hardness value of 240 HV [2].These, coupled with previous results, prove that titanium nitride thin film is deposited on samples at all treatment times, as one of the characteristics of titanium nitride is its high hardness [48].With an increase in treatment time, microhardness also increases, which is well-supported by cross-sectional SEM results which shows an increase in layer thickness with time [49].Another interesting finding is the difference in microhardness observed in 1 and 2 h samples, as well as in 3 and 4 h samples.This difference is very small, which may be attributed to reduced layer growth during the one-hour interval shown by cross-sectional SEM analysis.

Corrosion analysis
Figure 7 shows the potentiodynamic polarization curves of base material and samples treated for various time.It is observed that the samples in which titanium nitride film is deposited, an active region where the formation, growth and stabilization of the passive film take place and this behaviour attributed to the formation of the passive oxide film of titanium.The treated samples present a passive region where it characterizes the stabilization of the passive film and protection of the metal.There is also a region of nucleation and pitting growth, where the passive film is broken and the surface of the metal is exposed to the corrosive medium.
In Table 5, the pitting potentials of the treated samples are tabulated.It is verified that pite potentials have very close values; this occurs because the atomic percentages obtained in the quantitative analysis by EDS/Raman are similar.The percentage of titanium on the surface of the film available for the formation of the passive film did not influence reasonably for those values.However, there is an influence of the treatment time on the formation of the passive films, that is, the longer the nitriding time and consequently the greater thickness of the deposited film, the more recent the formation of the passivation film and the lower the degradation of the material.
It is also observed that an active-passive transition point with a critical current (I c ) is approximately equal to the passivation current (I passivation ).From this point, the passive film formed is dense and compact and isolates the metal from the solution, preventing the corrosion of the metal, or makes the corrosion slow to the potential to reach the pitting potential.
The electrochemical behaviour of the treated samples is different from the base material.The critical pitting potential for the treated sample has not been reached, in contrast, the curve of a base material   which shows a characteristic of low resistance to corrosion.It is due to the fact that when the potential values increase for base material, there is no passive film formation, but a rather anodic dissolution of the material.

Conclusion
In this article, the synthesis of titanium nitride on tool steel is successfully carried out using CCPD to improve the surface properties.The results obtained in this study can be summarized as follows: (1) The treatment time directly influences the synthesis of titanium nitride by CCPD.The thickness of the nitrided layer increases with increased treatment time, a layer of thickness 1.43-6.71μm is developed for a treatment time of 0.5-4 h.Interestingly, the hardness of 456 HV is attained even at the shortest treatment time (0.5 h), which is practically double than the hardness of base material.(2) Raman spectroscopy identified TiN bands, and with curve fitting through residual analysis, N/Ti ratio for 0.5 and 1 h treatment times is very approximate to the stoichiometric condition.This ratio is higher for other samples, but the synthesized titanium nitride layer shows stoichiometric conditions, particularly the gold colour characteristic of stoichiometric titanium nitride.(3) Raman and EDS analysis at multiple points depicts quite analogueous results, which shows layer uniformity.(4) XRD shows that the TiN phase is formed during all treatment time, which is mainly responsible for the improvement in hardness.(5) The corrosion resistance is found to be improved by the synthesis of TiN layer and increases with an increase in treatment time.(6) Remarkably, the inherent drawback of TiN-coatings such as voids and cracks associated with conventional techniques are not observed, which is expected to be advantageous for industrial demand.
Therefore, titanium nitride synthesis by CCPD proves to be an excellent alternative for the treatment of industrial tools, as this method allows for a meaningful increase in surface hardness and corrosion resistance.The choice of treatment time is dependent upon the application and required hardness of the tool, as for each treatment time used in this work a significant increase in hardness is observed.As samples displayed the gold colouring characteristic of stoichiometric titanium nitride, this method may also be used for decorative coating.For such applications, where the purpose of the layer is solely decorative and microhardness is not crucial, a treatment time of 0.5 h is sufficient.This system is proven to be compatible and useable on the industrial applications, thus our results are expected to be attractive for the industry.

Figure 1 .
Figure 1.Schematic illustration of CCPD system equipped with titanium cathodic cage.

Figure 4 .
Figure 4. XRD of samples nitrided for various treatment time.

Figure 6 .
Figure 6.Hardness profile of base material and samples nitrided for various treatment time.

Figure 7 .
Figure 7. Potentiodynamic polarization curves of base material and nitrided samples.

Table 2 .
Experimental conditions used in samples nitriding.

Table 5 .
Pitting potential values and passivation current with nitriding time.
aIt does not present pitting potential, because the potentiodynamic polarization curve is characteristic of anodic dissolution material, that is, there is no protective film formation.b No passivation current; the material does not passive.