Duplex surface treatment of sintered iron by plasma nitriding and plasma carburizing at low temperature

Duplex surface treatment of sintered iron by plasma nitriding and plasma carburizing at low temperature

Journal Pre-proof Duplex surface treatment of sintered iron by plasma nitriding and plasma carburizing at low temperature T.S. Lamim, E.A. Bernardell...

1MB Sizes 0 Downloads 13 Views

Journal Pre-proof Duplex surface treatment of sintered iron by plasma nitriding and plasma carburizing at low temperature

T.S. Lamim, E.A. Bernardelli, T. Bendo, C.H. Melo, C. Binder, A.N. Klein PII:

S0257-8972(19)30808-4

DOI:

https://doi.org/10.1016/j.surfcoat.2019.07.068

Reference:

SCT 24844

To appear in:

Surface & Coatings Technology

Received date:

14 February 2019

Revised date:

26 July 2019

Accepted date:

31 July 2019

Please cite this article as: T.S. Lamim, E.A. Bernardelli, T. Bendo, et al., Duplex surface treatment of sintered iron by plasma nitriding and plasma carburizing at low temperature, Surface & Coatings Technology (2019), https://doi.org/10.1016/j.surfcoat.2019.07.068

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier.

Journal Pre-proof

1

DUPLEX SURFACE TREATMENT OF SINTERED IRON BY PLASMA NITRIDING AND PLASMA CARBURIZING AT LOW TEMPERATURE *

T.S. Lamim1*; E.A. Bernardelli2; T. Bendo1; C.H. Melo1; C. Binder1; A.N. Klein1 1

Laboratório de Materiais, Universidade Federal de Santa Catarina - UFSC, Campus Trindade, Florianópolis, SC 88040-900, Brazil 2 Universidade Tecnológica Federal do Paraná - UTFPR, Campus Curitiba Curitiba, PR 80230-901, Brazil

*Corresponding Author 1

Laboratório de Materiais, Universidade Federal de Santa Catarina – UFSC, Campus Trindade 88040-900, [email protected], +55 48 3721-9268 r206.

of

Abstract

ro

The aim of this study is to investigate the combination of plasma nitriding and plasma

-p

carburizing at low temperature. Samples of sintered pure iron were nitrided at 500 °C for 3 h

re

under three different atmospheres: (1% N2 + 99% H2), (10% N2 + 90% H2), and (90% N2 + 9% H2 + 1% CH4), all followed by plasma carburizing at 500 °C for 3 h under a mixed

lP

atmosphere of 2% CH4 + 98% H2. Microstructural characterization and chemical analysis of

na

the layers were performed by scanning electron microscopy and energy-dispersive X-ray spectroscopy, respectively. Phase identification was carried out by grazing incidence X-ray

Jo ur

diffraction. The layer hardness was verified by nanohardness and the case depth by Vickers microhardness. The results show that the post-carburizing treatment promotes the formation of a thin outermost cementite layer over the nitride layers, leading to a nitrided–carburized layer on the surface. The morphology and phase composition of the nitride layer influence the structural and mechanical properties of the nitrided–carburized layers. The best surface hardness improvement, approximately 72 %, was achieved by the duplex treatment of nitriding with 1% N2 content followed by a plasma carburizing at low temperature. Moreover, duplex nitriding–carburizing leads to deeper hardening depths when compared to simply nitrided samples. Keywords: Duplex treatments; Plasma nitriding; Plasma carburizing; Low temperature.

Journal Pre-proof 1.

2

Introduction The constant need to improve the surface properties of engineering components has led to

the development of various surface treatments as nitriding, nitrocarburizing and carburizing [1]. In powder metallurgy components, the surface communicating pores are a serious obstacle for traditional thermochemical treatment as gaseous and salt processes [2]. Among the main detrimental effects of the surface open pores are the retaining of corrosive salts used in liquid treatments and the formation of an internal hard compound layer during gaseous

of

processes [2-3].

ro

Plasma thermochemical treatments are a suitable technology to harder sintered

-p

components. Such techniques overcome the problems associated with the detrimental effects of the communicating pores, offer high dimensional precision and a good surface finishing to

re

the parts [4-5]. The main plasma surface treatments used in the industry of sintered

lP

components include nitriding and nitrocarburizing, which provides wear and fatigue

[2,5,6].

na

resistance, and a surface porosity sealing effect due to the formation of a compound layer

Jo ur

Plasma carburizing is usually carried out at 900–950 °C in order to achieve the iron austenitic phase for a higher carbon solubility [7]. Apart from the high energy consumption, these relatively high temperatures may affect the dimensional tolerance of sintered components. Thus, low-temperature plasma carburizing represents an alternative surfacehardening technique for components with complex geometry, which require a high level of dimensional tolerance [8]. Low-temperature plasma carburizing is also described in the literature as an effective method to obtain continuous Fe3C layers on iron and low-steel surfaces using different plasma configurations [9-13]. The high hardness value of bulk cementite, approximately 1200 HV [14,15], makes such Fe3C carburized layers obtained by low-temperature plasma carburizing attractive for numerous applications in which mechanical and tribological properties are

Journal Pre-proof

3

required. However, the large difference between the elastic modulus of the cementite layer and the soft substrate compromises the bearing capacity of the component, limiting its application. One solution to overcome this limitation involves the development of a multifunctional surface by combining two-sequence surface treatments. Such a method is defined in the literature as duplex surface treatment, and it promotes the formation of a duplex layer with enhanced properties [16]. Numerous publications in recent years have reported the benefits of

of

duplex treatments to increase surface hardness, enhance wear resistance, reduce friction and,

ro

improve anticorrosion properties [17-22].

-p

In this context, a pretreatment of plasma nitriding could enhance the properties of cementite layers obtained by low-temperature plasma carburizing, because the nitride layer

re

gradually increases the surface hardness and may provide mechanical support to the Fe3C

lP

carburized layer. Such nitriding pretreatments have been adopted to improve the load-carrying capacity of soft substrates for diamond-like carbon and TiN coating deposition [23-25].

na

Therefore, the purpose of this study is to evaluate surface duplex treatments combining

Jo ur

plasma nitriding and plasma carburizing, all performed at 500 °C. The main difference between the nitriding–carburizing treatments is the morphology and phase composition of the nitride layers. Different characterization techniques were used to analyze the microstructure, morphology, topography, and the mechanical properties of the duplex layers and the results were compared with those obtained for the single carburizing and single nitriding treatments.

2.

Material and methods The substrate material used in this work was pure iron made from a powder metallurgy

route. Samples of 20 mm in diameter and 5 mm in height were produced with water-atomized iron powder AHC 100.29 (Höganäs), which has a mean particle size of 100 µm, and mixed with 0.8 wt.% of stearic acid used as a lubricant. The powder was compacted at a pressure of

Journal Pre-proof

4

500 MPa using a double-action press. The plasma treatments, sintering, nitriding, nitrocarburizing and carburizing were performed in a plasma reactor. The samples were sintered using a plasma-assisted debinding and sintering process [4], which was performed at 1125 °C for 60 min under an atmosphere of 5% H2 + 95% Ar. Prior to the plasma thermochemical treatments, all the samples were ground using SiC sandpaper (200 to 1200 grade), polished using 1 µm Al2O3, and cleaned with ethyl alcohol in an ultrasound bath in order to standardize the surfaces. Subsequently, they were dried in warm air.

of

The thermochemical treatments were carried out in a plasma reactor with an auxiliary

ro

heating system and a DC pulsed plasma source, which was described in a previous study [12].

-p

For all treatments, the samples were placed on the cathode and the chamber was evacuated to a residual pressure of 1.33 Pa. A surface cleaning process using a plasma discharge of H2 was

re

performed simultaneously with the heating process for 25 min. Once the working temperature

lP

was achieved, the gas composition was adjusted according to the thermochemical treatment.

na

After the treatment, the samples were cooled under a 100% H2 plasma discharge to avoid oxidation. The temperature was measured using a K-type thermocouple inserted to a depth of

Jo ur

10 mm into a reference sample.

Table 1 shows the main parameters used in the thermochemical treatments and the sample codes used in this study. Single treatments of carburizing and nitriding were performed in order to establish a basis for comparison and understand the layer growth mechanism of the duplex treatments. The nitrogen content in the gas mixture was varied to obtain nitrided layers with different structures and morphologies: diffusion layer (nitriding with 1% N2), compound layer with predominance of γ’-phase (nitriding with 10% N2), and compound layer with predominance of ɛ-phase (nitriding with 90% N2 + 1% CH4, which is commonly described as nitrocarburizing in literature due to the presence of methane in the gaseous mixture). The duplex plasma treatments were achieved by a carburizing treatment after each nitriding condition. For instance, the duplex nitriding–carburizing N1-CA is the result of

Journal Pre-proof

5

nitriding N1 followed by carburizing CA. For all treatments, the temperature, output voltage, switched-on time (ton), and gas flow were 500 °C, 500 V, 100 µs, and 240 sccm, respectively. The layer microstructures were observed by scanning electron microscopy (SEM) (JEOL JSM-6390LV) and the chemical analysis was carried out using an energy dispersive X-ray (EDX) spectrometer coupled to the SEM. Moreover, optical microscopy (Olympus Bx60M) was used to observe the size evolution of the γ’ nitride precipitates in the diffusion zone before and after the duplex treatments. In order to avoid damage to the layers during the

of

metallography procedure, the samples were cut with a precision cutter, wrapped in a thin

ro

copper foil, embedded in epoxy resin and then ground and polished. Nital (97% ethyl alcohol

-p

+ 2% nitric) was used to reveal the microstructures.

The structural properties of the layers were verified by grazing incidence X-ray diffraction

re

(GIXRD). The measurements were performed at two incident angles, 1 and 3, using a

lP

Shimadzu XRD6000 diffractometer equipped with Cu-Kα radiation (λ = 1.5418 Å). The X-

na

ray scanning range was from 35 to 60, with a scan step size and rate of 0.01/s. Phase analysis was carried out using the Crystallographica Search-Match 3.1.0 software package

Jo ur

and data from the JCPDS database [26]. For the nanohardness evaluation, an MTS NanoIndenter was set to apply a square array of 4 × 4 indents for each sample, and hardness and elastic modulus were calculated by the Oliver–Pharr method [27]. Thus, sixteen individual measurements of nanohardness values were used for each average datapoint. The test load was fixed at 400 mN with twelve loading and unloading cycles. The effective case depth was determined by the Vickers microindentation hardness (Leco LM100AT) according to the procedures described in MPFI 51[28] and MPFI 52 [29]. The test load was 0.01 kg. In addition, the surface topography of the layers was evaluated by white light interferometry (Zygo NewView7000) using Mountains Map Universal® 7.1 for the data analysis.

Journal Pre-proof 3.

6

Results and discussion Figure 1 shows a cross-sectional view of a layer obtained by SEM for the single plasma

carburizing (CA). This treatment promoted the formation of a continuous carburized layer along the entire surface, with an average thickness of 0.67 µm. Furthermore, a carburized diffusion zone formed by precipitates is not visible in the microstructure from the SEM observation. The growth mechanism of the carburized layer is associated with the low diffusivity and

of

solubility of the carbon into α-Fe at 500 °C, and it was elucidated in studies of RF plasma

ro

carburizing [9], laser plasma carburizing [10,11], and DC plasma carburizing [12,13].

-p

According to Carpene [10], when carbon reaches 25 at. % inside the iron matrix, the Fe3C phase precipitates on the surface, leading to the formation of a cementite layer. Moreover,

lP

iron substrate.

re

some low-carbon phase might be formed at the interface between the carburized layer and the

Surface micrographs of the single plasma nitriding treatments (N1, N10, N90) are shown

na

in Figures 2(a), (b), and (c). The images were focused on the cross-section view of the nitride

Jo ur

layers and not on the diffusion zones in order to better evaluate the surface region before and after the postcarburizing treatment. The lowest nitrogen content used for nitriding sample N1 (1% N2) was not low enough to avoid the formation of a compound layer, as seen in Figure 2(a). However, instead of a continuous compound layer, N1 presented the growth of some nitride nuclei at the surface. This layer morphology is known as one of the first progressive microstructural stages of compound-layer formation and evolution [30]. According to the authors, the development of a closed compound layer is given by the nucleation of these nuclei at the surface, followed by their growth via the supply of nitrogen through the ferrite that surrounds them and by lateral diffusion.

Journal Pre-proof

7

Figure 2 shows that increasing the nitrogen concentration in the gas composition to 10% and 90% (Figure 2(b) and (c), respectively) promotes the formation of a continuous compound layer at the surface. A nitriding gas mixture richer in nitrogen implies in a higher number of active nitrogen species to consolidate a closed nitride compound layer and causes modifications to the diffusion zone morphology, such as the size and number of γ’ precipitates [31-33]. The nitriding using 90% N2 + 1% CH4 (Figure 2(c)) promoted the thickest compound layer

of

and an interface between two phases can be observed in the compound layer. This mixed

ro

layer morphology is well described in the nitriding and nitrocarburizing literature [34-36], and

-p

it is attributed to the high nitrogen content and the presence of a small amount of CH4 in the gas mixture of N90 sample, which promotes the ε-phase stabilization. However, there is a

re

difficulty in obtaining an ε-monophasic compound layer, because this phase tends to

lP

decompose to γ’ by the reaction ε → γ’ + α-Fe, resulting in an ɛ + γ’ compound layer [37-39]. Moreover, some pores are visible in the near-surface layer of the N90 sample. The

na

metastability nature of the Fe-N phases in the nitride zone leads to the nucleation and growth

Jo ur

of molecular nitrogen (N2) in the nitride layer, causing the formation of pores. This phenomenon is especially observed near to the surface where the nitrogen content is highest and thus the thermodynamic driving force for N2 is the largest [1,30]. Even though the ε-nitride phase is known to show porosity, such defects can also be observed in γ´-nitride compound layer and even in austenite and ferrite phases [1]. Figures 3(a), (b), and (c) present the layer microstructures of the duplex nitriding-carburizing treatments N1-CA, N10-CA, and N90-CA. The treatment N1-CA resulted in the formation of a continuous layer (Figure 3(a)) with a microstructure similar to that obtained after the single carburizing treatment CA (Figure 1). Because the previous nitriding (N1) promoted only the limited growth of nitride nuclei at the surface (Figure 2(a)), the continuous layer observed in the N1-CA sample results from the postcarburizing

Journal Pre-proof

8

treatment. In the case of the N10-CA sample (Figure 3(b)), the postcarburizing leads to a modification of the compound layer regularity of the previous nitriding N10. For the N90-CA sample (Figure 3(c)), the interface between the ɛ and γ’ phases found in the previous nitriding N90 (Figure 2(c)) is no longer observed in the microstructure of the nitrided-carburized layer. This modification is ascribed to the homogenization of the phase composition presented in the previous N90 compound layer by transforming the entire ε phase into γ’ phase during the thermal cycle of the low-temperature postcarburizing [36,40].

of

Comparing the SEM images of the duplex (Figure 3) with the single nitriding (Figure 2)

ro

treatments, the γ’ precipitates present in the diffusion zone become smaller after all the

-p

postcarburizing treatments. The changes in the precipitate morphology show that the lowtemperature postcarburizing also affects the diffusion zone of the previous nitriding, most

re

likely owing to the thermal cycle that decomposes and refines the γ’ precipitates.

lP

Figure 4 exhibits the thickness of the surface layers obtained after the plasma treatments. The thickest layer obtained for the N90 sample (90% N2) is related to the high nitrogen

na

content in the gas mixture [31,33]. Nonetheless, a gaseous atmosphere with 100% N2 does not

Jo ur

provide a thickest compound layer, because the hydrogen plays an important role in plasma nitriding [41]. In fact, the addition of a small quantity of hydrogen increases the secondary electron emission coefficient, which consequently increases the density of nitrogen ions and neutral species in the plasma, in addition to the formation of the NH nitriding species [32]. Moreover, the presence of 1% CH4 in the N90 gas mixture did not contribute to an increase in the layer thickness [42]. By comparing the layer thickness of the nitriding–carburizing treatments with that of their preceding single nitriding, it is clear that there is a thickness evolution after the postcarburizing treatment. A slight decrease in the average layer thickness for the N10-CA and N90-CA samples is observed when compared to their previous nitriding N10 and N90, respectively. This could also be directly related to the thermal cycle of the low-temperature

Journal Pre-proof

9

postcarburizing, which induces nitrogen atomic diffusion from the layer into the substrate, decreasing the compound layer thickness. Such a reduction in the layer thickness was similarly observed by Cieslik et al. [43] when the authors performed a diffusion stage after plasma nitriding treatments by holding the samples at the same temperature used for nitriding. Figure 5(a) shows GIXRD patterns of the carburizing sample (CA) and nitriding samples (N1, N10, and N90) obtained at a fixed incidence angle of 3. Initially, the stabilization of the cementite (Fe3C) for the CA indicates that the continuous carburized layer observed in

of

Figure 1 is constituted by this carbide, as observed in other studies of low-temperature plasma

ro

carburizing [9-12].

-p

The GIXRD results of the N1, N10, and N90 samples show the influence of nitrogen content on the nitride layer structures. The modest presence of the γ’-Fe4N phase in the N1

re

pattern is attributed to the nitride nuclei formed at the surface (Figure 2(a)). Upon increasing

lP

nitrogen content in the gas mixture (N10 sample), the relative intensity of the γ’-Fe4N peaks increases significantly, indicating a continuous γ’ compound layer at the surface. Although

na

the nitriding conditions for the N90 samples provided the ε-Fe2-3(N,C) carbonitride

Jo ur

stabilization, the presence of the γ’ phase shows that a pure ε-compound layer was not achieved in this treatment. In addition, the exclusive presence of ε at the incidence angle of 1 and the appearance of γ’ at 3 observed in Figure 6, suggest that the outermost part of the N90 compound layer is composed exclusively by ε phase. This ε/γ’ layer morphology is a consequence of the reaction ε → γ’ that occurs from the substrate in the surface direction [1,30,37,40]. According to the literature [38,44], fast cooling rates helps to retain a predominantly ε layer at room temperatures since there is no sufficient time to ε → γ’ phase transformation. GIXRD patterns of the duplex nitriding–carburizing treatments are presented in Figure 5(b). By comparing with the previous single nitriding patterns, the results show that the postcarburizing treatments promote cementite stabilization for all the samples. Thus, the

Journal Pre-proof

10

continuous layer observed in the N1-CA sample (Figure 3(a)) can be described as a cementite layer, similarly to that obtained in the carburized sample (CA). Moreover, the γ’ phase is no longer detected in N1-CA sample, indicating that the γ’ nitride nuclei were dissolved during the thermal cycle of postcarburizing treatment. The presence of cementite in the N10-CA and N90-CA GIDRX patterns can be understood by analyzing their duplex layer microstructures presented in Figure 7 (a) and (b). It is visible that both layers are constituted by two regions: a thin outermost phase (region 1)

of

and a thick innermost phase (region 2). According to EDX analysis, region 1 is exclusively

ro

composed of carbon and iron, and region 2 of carbon, iron, and nitrogen. The cementite phase

-p

detected in the N10-CA and N90-CA GIDRX patterns can be ascribed to region 1 and region 2 can be associated with the γ’ Fe4N phase. Therefore, the low-temperature postcarburizing

re

treatment promoted the formation of a thin outermost cementite layer over the nitrided

lP

surfaces, leading to duplex layers on the surface of the N10-CA and N90-CA samples. The exclusive presence of γ’ in the GIXRD N90-CA pattern indicates that the ε phase

na

detected in the previous nitriding (N90 sample) was entirely transformed to γ’ by the reaction

Jo ur

ε → γ’ + α-Fe as a consequence of an annealing effect of the compound layer induced by the thermal cycle of the low-temperature postcarburizing. This transformation is caused by the instability of the ε at low temperatures and it occurs by a redistribution of nitrogen within the compound layers in order to achieve an equilibrium composition of the nitride layer [35, 39,45]. Figure 8 displays the surface root mean square roughness parameter (Sq) for all samples and it shows that nitriding N10 and N90 showed significantly increased surface roughness when compared to N1. Since the samples are placed on the cathode, their surfaces are bombarded by plasma species during nitriding. The nitrogen content used in the gaseous mixture of N10 and N90 provides higher surface sputtering rate when compared to N1 sample. According to Mason [46], most part of the sputtered atoms return to the sample

Journal Pre-proof

11

surface, developing conical asperities and increasing the roughness. Besides that, the formation of a compound layer on the surfaces of the N10 (γ’ layer) and N90 (ε/γ’ layer) also contributes to rougher surfaces. Furthermore, after the low-temperature postcarburizing, the surface roughness remained statistically similar to their previous nitrided-only samples. Table 2 shows the surface mechanical properties and the ratio H/E of the untreated sample and all plasma treated samples. The untreated sample was sanded and polished in order to standardize the surface, similar to the procedure performed with all samples prior to the

of

plasma thermochemical treatment.

ro

The surface hardness (H) was defined as the highest hardness value observed in each

-p

nanohardness profile, considering an indention depth < 10% of the layer thickness in order to minimize the influence of the substrate properties. However, the rough layers presented by

re

N10 and N10-CA (Figure 8) induced a surface roughness effect on the indentation

lP

measurements of these samples until a depth of approximately 0.20 µm. According to Souza et al [47], when roughness is present the hardness calculated by the traditional Oliver–Pharr

na

method will be lower than the real one, especially when the asperities curvature diameter is in

Jo ur

the same order than tip diameter. Furthermore, the authors suggest that this effect is very significant for small applied loads (shallow penetration depths) as observed for N10 and N10-CA. For that reason, the surface hardness of these samples was obtained at a ratio of indentation depth to layer thickness of 13.2% (N10) and 9.6% (N10-CA), where the roughness effect was no longer detected. Additionally, the surface elastic modulus (E) showed in Table 2 is the correspondent elastic modulus value at the determined surface hardness (H) of the samples. The CA sample exhibited the highest surface hardness (12.44 GPa) of all the treatments, which is correlated with the continuous cementite layer promoted by this treatment. Similar hardness values were observed for cementite layers and in previous works [11,14,15].

Journal Pre-proof

12

The N90 sample presented the highest surface hardness value, 9.45 GPa, among the single nitriding treatments (N1 and N10 samples). A similar observation was reported by [33]. This result is attributed to the precipitation of the ε-Fe2-3(N,C) carbonitride at the surface of the N90 sample, which is promoted by the high nitrogen content and the 1% CH4 used in the gas mixture of this treatment. The nitrogen content and the presence of methane in the nitriding atmosphere influence the layer structure, which consequently affects the layer mechanical properties. Moreover, according to Ochoa [48], there is a correlation between hardness and

of

nitrogen concentration along the compound layer directly related to the structural changes of

ro

compound layer crystalline.

-p

The duplex treatments of nitriding–carburizing promoted an increase in surface hardness when compared to the simply nitrided samples. This hardness improvement was more

re

effective in the N1-CA sample (Table 2), where the postcarburizing treatment increased the

lP

N1 sample surface hardness from 6.24 to 10.76 GPa. In this case, the increase in hardness is ascribed to the formation of a closed cementite layer on the nitrided surface after the

Jo ur

Figure 5(b).

na

postcarburizing treatment, as previously observed in Figure 3(a) and the GIRXD pattern of

The ratio H/E is related to elastic strain to failure and it has been shown in the literature [49-52] as a suitable parameter to evaluate the properties of layers and coatings with respect to elastic behavior. Since H/E considers not only the layer hardness but also the elastic modulus it has an important influence on wear resistance [49]. A high H/E ratio is related to a high elastic strain prior to the plastic deformation and for that reason, it is considered an indicative of good wear resistance [50,52]. Based on that, the H/E ratio for the untreated sample and all plasma treated samples is shown in Figure 9. All single plasma nitriding (N1, N10, and N90) promoted an increase of the H/E ratio when compared to the untreated sample, indicating an improvement of the surface mechanical resistance due to the presence of the nitride layers. The difference among the ratios H/E of the

Journal Pre-proof

13

nitrided samples is attributed to the morphology of the layers (diffusion or compound layer) and the nature of the iron nitrides (γ’- Fe4N or ε-Fe2-3(N,C)), which are directly related with the atmosphere used in the plasma nitriding treatments. The largest ratio H/E observed in N90 sample (which developed a ε/γ’ compound layer) is in accordance to the wear results for plasma nitride layers formed in sintered iron presented by Binder et al [53]. The authors verified that compound layers with a predominance of ε nitride present a superior performance in sliding wear tests when compared to γ’ layers.

of

The low-temperature postcarburizing performed in N1-CA, N10-CA and N90-CA

ro

increased the H/E ratio of their preceding N1, N10, and N90 samples. Based on the H/E effect

-p

on wear resistance demonstrated in the literature [49-52], the duplex nitriding-carburizing at low temperature could have an interesting potential to enhance the surface wear resistance of

re

nitride-only samples.

lP

The higher H/E ratio of N1-CA, when compared to the previous N1, can be associated with the formation of a hard cementite layer over the nitride diffusion layer (Figure 3(a)). On

na

the other hand, for N10-CA and N90-CA it is related to the development of a duplex layer

Jo ur

constituted by an outermost Fe3C layer and an innermost γ’ compound layer. As observed in Figure 9 and Table 2, the slightest increase in H/E ratio after the postcarburizing was observed for N10-CA. This fact could be associated with a lower contribution of the outermost cementite layer (Figure 7(a)) to the surface hardness of this sample. Due to the roughness effect, the hardness of N10-CA had to be obtained at an indentation of approximately 0.20 µm. At this depth, the influence of the 0.5-µm outermost cementite layer in the hardness measurement of N10-CA is affected, making the H/E ratio of this sample very similar to the previous N10, as observed in Figure 9. Figure 10 shows the microhardness depth profiles for all plasma-treated samples. The effective hardening depth was defined as the surface depth where the hardness values are higher than untreated sintered iron substrate used in this work (113.55 HV0.01). The hardness

Journal Pre-proof

14

measured at 0 µm depth by Vickers microhardness (Figure 10) should have the influence of the substrate properties since it is not the appropriate technique to measure the surface hardness of thin layers. Thus, it is reasonable to expect lower values of surface hardness (0 µm depth) when compared with the ones measured by the nanohardness technique presented in Table 2. Although the single plasma-carburized sample (CA) did not present a carburized diffusion zone in the micrograph (Figure 1), the carburizing treatment led to a 30-µm hardening depth.

of

The absence of precipitates in a carburized diffusion zone and the low hardness values

ro

measured in the 30-µm hardening depth (Figure 10(a)) suggest that the single plasma

-p

carburizing treatment promoted an interstitial solid-solution hardening of the sintered iron. The shallow hardening depth observed in the CA sample can be related to the cementite layer

re

formed on the surface. According to the literature [54][55], the diffusion coefficient of carbon

lP

in cementite at 500 °C is around 10-19 to 10-20 m²/s, whereas in ferrite it is around 10-12 to 10-13 m²/s. Thus, the smaller coefficient in cementite makes this layer a barrier to carbon inward

Jo ur

treatments.

na

diffusion into the substrate. Similar results were observed by [37] in plasma nitrocarbuzing

Deeper hardening depths are observed in the duplex treated samples when compared to nitride-only ones. Even though the diffusion coefficient of carbon and nitrogen in ferrite is of the same order of magnitude at 500 °C [55][56], the cementite layer formed over the nitrided surfaces after the postcarburizing should difficult the carbon diffusion into the substrate. Thus, the enhancement of the hardening depth observed in the duplex treatment is more related to an annealing effect of the nitride diffusion zone promoted by the postcarburizing thermal cycle than to carbon inward diffusion. Similar results have been reported for duplex treatment combinations of plasma nitrocarburizing and plasma nitriding [19]. In addition, although the presence of 1% CH4 in the gaseous mixture of nitriding N90, such previous

Journal Pre-proof

15

slight amount of carbon had no measurable influence in the hardening depth of N90-CA after the postcarburizing. The annealing effect in the diffusion zone modified the morphology of the nitride precipitates by refining and dissolving them, which can be a direct consequence of inward diffusion of nitrogen in ferrite and the concomitant lower nitrogen content in the ferrite. This assumption is supported by analyzing the size evolution of the γ’ needles present in the diffusion zone (Figure 11(a)). Each point on the graph represents the measured average size of

of

γ’ needles present in three areas of 120 µm × 80 µm for each sample, as schematically

ro

represented in Figure 11(b). As observed in the graph, the annealing effect induced by the

-p

postcarburizing thermal cycle decreases the size of the γ’ needles in the diffusion zone. These changes in the nitride precipitate morphology are illustrated in Figure 11(c), by comparing

Conclusions

lP

4.

na

treatment (N90-CA sample).

re

SEM cross-section micrographs of the N90 sample and its respective postcarburizing

Jo ur

This work was performed to evaluate surface duplex treatments combining plasma nitriding and plasma carburizing at 500 °C. Based on the obtained results, the main conclusions are listed below: 

Duplex treatments of nitriding–carburizing at low temperature promote the formation of a thin cementite layer over the different iron nitride layers.



Duplex layer morphologies are observed for the treatments of Nitriding γ’ + Carburizing (N10-CA) and Nitriding ε + Carburizing (N90-CA). Such duplex layers are constituted by an outermost Fe3C layer and an innermost γ’ compound layer.



Low-temperature postcarburizing increases the ratio H/E (related to elastic strain to failure) of the nitride-only samples, indicating an improvement in mechanical properties of the layers.

Journal Pre-proof 

16

The morphology and the phase composition of the previous nitride layer influence the mechanical properties of the nitrided-carburized layer. The duplex treatment of Nitriding Diffusion + Carburizing (N1-CA) presented the best improvement in surface hardness and H/E ratio.



The thermal cycle of the postcarburizing treatment induces a redistribution of nitrogen in the compound layer and in the diffusion zone, promoting an annealing effect to these two nitride zones. Such thermal effect induces ε → γ’ transformation in the compound layer

of

and promotes deeper effective hardening depth when compared to the nitride-only

-p

ro

samples.

Acknowledgments

re

This research was financially supported by CAPES, CNPq, and Embraco S.A. The authors

lP

would like to thank LCME-UFSC for technical support during the electron microscopy work

Jo ur

na

and Mr. Rodrigo Perito Cardoso from UFPR for carrying out the nanohardness analyses.

Journal Pre-proof

17

References E.J. Mittemeijer, M.A.J. Somers, Thermochemical surface engineering of steels, Woodhead P, 2015. doi:https://doi.org/10.1016/C2013-0-16318-0.

[2]

J. Kazior, Surface engineering of ferrous sintered materials, (2001) 85–161.

[3]

T. Bell, Surface treatment and coating of PM components, Mater. Des. 13 (1992).

[4]

A.N. Klein, R.P. Cardoso, H.C. Pavanati, C. Binder, A.M. Maliska, G. Hammes, D. Fusao, A. Seeber, S.F. Brunatto, J.L.R. Muzart, DC Plasma Technology Applied to Powder Metallurgy: an Overview, Plasma Sci. Technol. 15 (2013) 70–81. doi:10.1088/1009-0630/15/1/12.

[5]

G.F. Bocchini, A. Molinari, B. Tesi, T. Bacci, Ion-Nitriding of Sintered Steels - Present Situation and Perspectives, MPR. (1990) 772–778.

[6]

A.M. Maliska, A.M. de Oliveira, A.N. Klein, J.L.R. Muzart, Surface porosity sealing effect of plasma nitrocarburizing on sintered unalloyed iron, Surf. Coatings Technol. 141 (2001) 128–134. doi:10.1016/S0257-8972(01)01016-7.

[7]

F. Czerwinski, Thermochemical Treatment of Metals, in: Heat Treat. - Conv. Nov. Appl., 2012: pp. 74–110. doi:10.5772/51566.

[8]

J.M.U. Baek, Y.R. Cho, D.J. Kim, K.H. Lee, Plasma carburizing process for the low distortion of automobile gears, Surf. Coatings Technol. 131 (2000) 568–573.

[9]

J.L. Li, T.J. O’Keefe, W.J. James, Iron Carbide Films Formed by Plasma Deposition and Plasma Carburizing, Mater. Sci. Eng. B. 7 (1990) 15–23.

na

lP

re

-p

ro

of

[1]

Jo ur

[10] E. Carpene, P. Schaaf, Formation of Fe3C surface layers by laser plasma cementation, Appl. Phys. Lett. 80 (2002) 891. doi:10.1063/1.1447009. [11] P. Schaaf, M. Kahle, E. Carpene, Reactive laser plasma coating formation, Surf. Coatings Technol. 200 (2005) 608–611. doi:10.1016/j.surfcoat.2005.01.028. [12] T.S. Lamim, E.A. Benardelli, C. Binder, A.N. Klein, A.M. Maliska, Plasma carburizing of sintered pure iron at low temperature, Mater. Res. 18 (2015). doi:10.1590/1516-1439.306514. [13] K.V.S. Damin, T.S. Lamim, F.C. Neto, C. Binder, A.N. Klein, A.M. Maliska, AISI 1005 steel plasma treated by different thermochemical surface treatments, Mater. Res. 19 (2016). doi:10.1590/1980-5373-MR-2015-0594. [14] M. Umemoto, Z.G. Liu, H. Takaoka, M. Sawakami, K. Tsuchiya, K. Masuyama, Production of Bulk Cementite and Its Characterization, Metall. Mater. Trans. A. 32 (2001) 2127–2131. [15] S.. Li, M. Ishihara, H. Yumoto, T. Aizawa, M. Shimotomai, Characterization of cementite films prepared by electron-shower-assisted PVD method, Thin Solid Films. 316 (1998) 100–104. doi:10.1016/S0040-6090(98)00397-6.

Journal Pre-proof

18

[16] K.-T. Rie, Recent advances in plasma diffusion processes, Surf. Coatings Technol. 112 (1999) 56–62. doi:10.1016/S0257-8972(98)00747-6. [17] B. Wang, W. Xue, X. Jin, Y. Zhang, Z. Wu, Y. Li, Combined Treatment Plasma Electrolytic Carburizing and Borocarburizing on Q235 Low-carbon Steel, Mater. Chem. Phys. 221 (2018) 232–238. doi:10.1016/j.matchemphys.2018.09.011. [18] S. Lu, B. Miao, L. Song, R. Song, K. Wei, J. Hu, Enhancement of wear resistance of AISI 1045 steel by a two-step plasma treatment, Vacuum. 145 (2017) 153–157. doi:10.1016/j.vacuum.2017.08.026. [19] B. Miao, Y. Chai, K. Wei, J. Hu, A novel duplex plasma treatment combining plasma nitrocarburizing and plasma nitriding, Vacuum. 133 (2016) 54–57. doi:10.1016/j.vacuum.2016.07.007.

ro

of

[20] F. Haftlang, A. Habibolahzadeh, M.H. Sohi, Improving electrochemical properties of AISI 1045 steels by duplex surface treatment of plasma nitriding and aluminizing, Appl. Surf. Sci. 329 (2015) 240–247. doi:10.1016/j.apsusc.2014.12.140.

-p

[21] F. Haftlang, A. Habibolahzadeh, M.H. Sohi, Comparative tribological studies of duplex surface treated AISI 1045 steels fabricated by combinations of plasma nitriding and aluminizing, Mater. Des. 60 (2014) 580–586. doi:10.1016/j.apsusc.2014.12.140.

lP

re

[22] H.R. Abedi, M. Salehi, M. Yazdkhasti, Novel plasma nitriding-oxidizing duplex treatment of AISI 316 austenitic stainless steel, Mater. Lett. 64 (2010) 698–701. doi:10.1016/j.matlet.2009.12.042.

na

[23] W. Tillmann, N.F.L. Dias, D. Stangier, Influence of plasma nitriding pretreatments on the tribo-mechanical properties of DLC coatings sputtered on AISI H11, Surf. Coatings Technol. (2018). doi:10.1016/j.surfcoat.2018.11.002.

Jo ur

[24] P.H.T. Shioga, C. Binder, G. Hammes, A.N. Klein, J.D.B. de Mello, Effects of Different Plasma Nitrided Layers on the Tribological Performance of DLC Coatings, Mater. Res. 19 (2016) 1180–1188. doi:10.1590/1980-5373-MR-2015-0455. [25] S. V. Bhaskar, H.N. Kudal, Tribology of nitrided-coated steel-a review, Arch. Mech. Technol. Mater. 37 (2017) 50–57. doi:10.1515/amtm-2017-0008. [26] JCPDS, JCPDS - International Centre for Diffraction Data, PCPDFWIN (n.d.) Version PDF-2 2001. [27] W.C. Oliver, G.M. Pharr, An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments, J. Mater. Res. 7 (1992) 1564–1583. [28] MPFI Standard 51- Method for determination of microindentation of hardness of powder metallurgy materials, in: MPFI Stand. Test Methods, 2012: pp. 95–96. [29] MPFI Standard 52- Determination of effective case depth of ferrous powder metallurgy products, in: MPFI Stand. Test Methods, 2012: pp. 99–100. [30] E.J. Mittemeijer, Fundamentals of Nitriding and Nitrocarburizing, in: J. Dosset, G.E. Totten (Eds.), ASM Handbook, Steal Heat Treat. Fundam. Process., ASM International, Stuttgart, 2013: pp. 619–646.

Journal Pre-proof

19

[31] K.S. Fancey, A. Leyland, D. Egerton, D. Torres, A. M, The influence of process gas characteristics on the properties of plasma nitrided steel, Surf. Coatings Technol. 76–77 (1995) 694–699. [32] Y.M. Kim, J.U. Kim, J.G. Han, Investigation on the pulsed DC plasma nitriding with optical emission spectroscopy, Surf. Coatings Technol. 152 (2002) 227–232. [33] A.F.O. Skonieski, E.S. Lima, T. Hirsch, A.S. Rocha, Influência da mistura gasosa em processos de nitretação e nitrocarbonetação a plasma, Rev. Bras. Apl. Vácuo, V27. (2008) 175–182. [34] M.K. Lei, Z.L. Zhang, Plasma source ion nitriding of pure iron : formation of an iron nitride layer and hardened diffusion layer at low temperature, Surf. Coatings Technol. 91 (1997) 25–31.

ro

of

[35] T. Liapina, A. Leineweber, E.J. Mittemeijer, Phase Transformations in Iron-Nitride Compound Layers upon Low-Temperature Annealing : Diffusion Kinetics of Nitrogen in Fe2-3N - and Fe4n -Iron Nitrides, Metall. Mater. Trans. A. 37 (2006) 319–330.

-p

[36] M. Nikolussi, A. Leineweber, E. Bischoff, E.J. Mittemeijer, Examination of phase transformations in the system Fe – N – C by means of nitrocarburising reactions and secondary annealing experiments ;, Int. J. Mat. Red. 98 (2007) 1086–1092.

lP

re

[37] T. Lampe, S. Eisenberg, G. Laudien, Compound layer formation during plasma nitriding and plasma nitrocarburising, Surf. Eng. 9 (1993) 69–76. [38] T. Bell, Y. Sun, A. Suhadi, Environmental and technical aspects of plasma nitrocarburising, Vacuum. 59 (2000) 14–23.

na

[39] T. Wöhrle, Thermodynamics and kinetics of phase transformations in the Fe-N-C system, Fakultät Chemie der Universität Stuttgart, 2012.

Jo ur

[40] T. Liapina, Phase transformations in interstitial Fe-N alloys, Fakultät Chemie der Universität Stuttgart, 2005. [41] P.E. Ortiz, Estudo da nitretação por plasma: correlação plasma-superfície, Universidade Federal de Santa Catarina, Florianópolis., 2000. [42] H.R.T. Silva, P. Egert, A. Seeber, C. Speller, Effect of Methane Addition on Formation of Plasma Nitrocarburized Layers, Metallogr. Microstruct. Anal. 5 (2016) 486–492. doi:10.1007/s13632-016-0324-7. [43] J. Cieslik, P. Jacquet, B. Tlili, H. Mulin, Decrease of Compound Layer Thickness Obtained in Plasma Nitriding of Alloyed Steels by Diffusion Stage, 1 (2011) 974–980. [44] C. Alves, J. a. Lima, V. Hajek, J.B.M. da Cunha, C. a. dos Santos, Effect of cooling rate on properties of plasma nitrided AISI 1010 steel, Surf. Coatings Technol. 201 (2007) 7566–7573. doi:10.1016/j.surfcoat.2007.02.036. [45] H. Göhring, S. Kante, A. Leineweber, E.J. Mittemeijer, Microstructural development and crystallographic properties of decomposing Fe-N-C compound layers, Int. J. Mater. Res. 107 (2016) 203–216. doi:10.3139/146.111330. [46] R.S. Mason, M. Pichilingi, Sputtering in a glow discharge ion source-pressure

Journal Pre-proof

20

dependence: theory and experiment, J. Phys. D. Appl. Phys. 27 (1994) 2363–2371. doi:10.1088/0022-3727/27/11/017. [47] G.B. de Souza, C.E. Foerster, S.L.R. da Silva, C.M. Lepienski, Nanomechanical Properties of Rough Surfaces, Mater. Res. 9 (2006) 159–163. [48] E. a. Ochoa, D. Wisnivesky, T. Minea, M. Ganciu, C. Tauziede, P. Chapon, F. Alvarez, Microstructure and properties of the compound layer obtained by pulsed plasma nitriding in steel gears, Surf. Coatings Technol. 203 (2009) 1457–1461. doi:10.1016/j.surfcoat.2008.11.025. [49] A. Leyland, A. Matthews, On the significance of the H / E ratio in wear control : a nanocomposite coating approach to optimised tribological behaviour, 246 (2000) 1–11.

ro

of

[50] W. Ni, Y. Cheng, M.J. Lukitsch, A.M. Weiner, L.C. Lev, S. David, Effects of the ratio of hardness to Young ’ s modulus on the friction and wear behavior of bilayer coatings, 4028 (2004) 83–86. doi:10.1063/1.1811377.

-p

[51] A. Leyland, A. Matthews, Design criteria for wear-resistant nanostructured and glassymetal coatings, 178 (2004) 317–324. doi:10.1016/j.surfcoat.2003.09.011.

re

[52] J. Guo, H. Wang, F. Meng, X. Liu, F. Huang, Tuning the H / E * ratio and E * of AlN coatings by copper addition, Surf. Coat. Technol. 228 (2013) 68–75. doi:10.1016/j.surfcoat.2013.04.008.

lP

[53] C. Binder, T. Bendo, G. Hammes, A.N. Klein, J.D.B. de Mello, Effect of nature of nitride phases on sliding wear of plasma nitrided sintered iron, Wear. 332–333 (2015) 995–1005. doi:10.1016/j.wear.2015.01.083.

Jo ur

na

[54] B. Ozturk, V.L. Fearing, J.A. Ruth, Self-Diffusion Coefficients of Carbon in F e3C at 723 K via the Kinetics of Formation of This Compound, Metall. Trans. A. 13 (1982) 1871–1873. [55] J. Cermak, L. Kral, Carbon diffusion in carbon-supersaturated ferrite and austenite, J. Alloys Compd. 586 (2014) 129–135. doi:10.1016/j.jallcom.2013.10.058. [56] S.R. Hosseini, F. Ashrafizadeh, A. Kermanpur, Calculation and experimentation of the compound layer thickness in gas and plasma nitriding of iron, Iran. J. Sci. Technol. 34 (2010) 553–566.

Journal Pre-proof

21

TABLES Table I - Plasma treatments parameters and samples code. Atmosphere

Carburizing Nitriding Diffusion Nitriding γ’ **Nitriding ε Nitriding Diffusion + Carburizing

2% CH4 + 98% H2 1% N2 + 99% H2 10% N2 + 90% H2 90% N2 + 9% H2 + 1% CH4 1% N2 + 99% H2 2% CH4 + 98% H2

Nitriding γ’+ Carburizing

10% N2 + 90% H2 2% CH4 + 98% H2

of

N10-CA

Treatment

Time (h) 3 3 3 3 3 3

Pressure (Pa) 666 400 400 400 400 666

3 3

400 666

ro

Samples Code CA N1 N10 N90 N1-CA

**Nitriding ε + 90% N2 + 9% H2 + 1% CH4 3 400 Carburizing 2% CH4 + 98% H2 3 666 *Samples code are correlated with the type of treatment and nitrogen content in the atmosphere. **Nitriding treatments performed with nitrogen and methane are usually described as nitrocarburizing in literature.

lP

re

-p

N90-CA

na

Table II – Surface hardness (H), surface elastic modulus (E), and the H/E ratio of the untreated sample and the plasma treated samples. Hardness Elastic Modulus H/E (GPa) (GPa) (10-3) Untreated* 4.27 ± 0.99 228.15 ± 36.90 18.72 CA 12.44 ± 7.44 227.72 ± 74.69 54.62 N1 6.24 ± 2.07 223.46 ± 61.10 27.92 N10 6.59 ± 2.34 194.49 ± 52.12 33.88 N90 9.45 ± 3.31 227.74 ± 74.29 41.49 N1-CA 10.76 ± 4.04 253.08 ± 68.82 42.51 N10-CA 7.24 ± 1.86 193.86 ± 41.23 37.34 N90-CA 10.48 ± 3.31 207.54 ± 54.37 50.50 *Properties of N10-CA and N90-CA are related to the duplex nitrided-carburized layer and not only to the thin carburized layer formed after the postcarburizing.

Jo ur

Sample

Journal Pre-proof

22

LIST OF FIGURES CAPTIONS Figure 1 – SEM cross-section micrograph of single plasma carburizing (CA) at 500 °C: 2%CH4 + 98%H2. Figure 2 – SEM cross-section micrographs of single plasma nitriding at 500 °C for 3 h: (a) N1 - 1% N2, (b) N10 - 10% N2, (c) N90 - 90% N2 + 1% CH4.

of

Figure 3 – SEM cross-section micrographs of duplex nitriding–carburizing treatments (3 h of

ro

nitriding + 3 h of carburizing): (a) N1-CA, (b) N10-CA, (c) N90-CA.

-p

Figure 4 – Thickness of the surface layers obtained in all plasma treated samples.

re

Figure 5 – GIXRD pattern at the incidence angle of 3 of (a) carburizing sample (CA) and

lP

single nitriding samples (N1, N10, N90), (b) duplex nitriding–carburizing samples (N1-CA, N10-CA, N90-CA).

na

Figure 6 – GIXRD pattern of the N90 sample, showing the exclusive presence of the ε phase

Jo ur

at the incidence angle of 1 and the appearance of the γ’ phase at 3. Figure 7 – SEM cross-section micrographs and EDX analyses of the duplex layers obtained in the duplex treatments: (a) N10-CA sample, (b) N90-CA sample. *Copper and aluminum detection are ascribed to the metallography procedure. Figure 8 – Sq topographic parameter (root mean square height) of the untreated sample, carburizing sample (CA), single nitriding samples (N1, N10, N90), and duplex nitriding– carburizing samples (N1-CA, N10-CA, N90-CA). Figure 9 – Ratio of surface hardness (H) to elastic modulus (E) for the untreated sample and all plasma treated samples.

Journal Pre-proof

23

Figure 10 – Vickers microhardness profiles and effective hardening depth of (a) single carburizing - CA, (b) N1 and N1-CA, (c) N10 and N10-CA, and (d) N90 and N90-CA. Figure 11 – (a) γ’-Fe4N needle size evolution in the nitride diffusion zone before and after the postcarburizing treatments, (b) Schema representing the measurement of γ’ needles present in an area of 120 µm × 80 µm of the diffusion zone (c) SEM micrographs of the N90 and N90CA samples indicating the changes in the γ’ precipitate morphology after the postcarburizing

Jo ur

na

lP

re

-p

ro

of

treatment.

Journal Pre-proof

of

FIGURES

Jo ur

Figure 2

na

lP

re

-p

ro

Figure 1

Figure 3

24

of

Journal Pre-proof

Jo ur

na

lP

re

-p

ro

Figure 4

Figure 5

25

of

Journal Pre-proof

Jo ur

na

lP

re

-p

ro

Figure 6

Figure 7

26

ro

of

Journal Pre-proof

Jo ur

na

lP

re

-p

Figure 8

Figure 9

27

Jo ur

Figure 10

na

lP

re

-p

ro

of

Journal Pre-proof

28

na Jo ur

Figure 11

lP

re

-p

ro

of

Journal Pre-proof

29

Journal Pre-proof

30

HIGHLIGHTS  Duplex treatments of sintered iron combining plasma nitriding and plasma carburizing at 500 °C were investigated.  Low-temperature plasma postcarburizing develops a thin cementite layer over the iron nitrided surfaces.  Morphology and phase composition of the nitrided layer influence the structural and mechanical properties of the nitrided–carburized layers. Duplex treatments of nitriding-carburizing promote higher ratios of surface hardness

of



ro

to elastic modulus (H/E) when compared to the nitrided-only samples.

-p

 The postcarburizing thermal cycle induces an annealing effect on the nitride

Jo ur

na

lP

re

compound layer and the diffusion zone.