Formation of Al and Cr Dual Coatings by Pack Cementation on SNCM439 Steel

Transcription

1 , pp Formation of Al and Cr Dual Coatings by Pack Cementation on SNCM439 Steel Jhewn-Kuang CHEN,* Shih-Fan CHEN and Che-Shun HUANG National Taipei University of Technology, Institute of Materials Science and Engineering, 1, Sec.3, Zhong-Xiao E. Rd., Taipei, Taiwan. (Received on May 23, 2011; accepted on September 27, 2011) Two stage pack cementation processes are developed to form dual Fe Al Cr layers on surfaces of SNCM439 steels. The first 550 C treatment assists to modulate adequate aluminum activity for the formation of iron-rich intermetallics. In the second 750 C treatment stage, simultaneous chromizing and aluminizing treatments are achieved by first forming a FeAl ferritic layer and then a surface layer with higher Cr content at later time. The current study examines the effects of second stage 750 C holding time, activator concentration, and Cr:Al ratio on coating structures. Fe Al coatings consisting of Fe 3Al and FeAl intermetallic phases are observed to form initially. This Fe Al layer accounts for 25 to 32 μ m thickness of the coatings and show good adherence with the substrates. The coating thickness increases parabolically with 750 C holding time. With prolonged treatment at 750 C, surface concentration of aluminum in powder packs drops with treatment time and increasing concentration of activator. A peak concentration exists at a depth below substrate surface. Aluminum is back diffused from the steel surface into the powder packs. The growth of Fe Al intermetallics slows down. Surface layer then forms a thickness of 6 μ m coating with 2 5 wt.% of chromium. Samples treated for longer than 6 h with over 12 wt.% of NH 4Cl activator concentration or with Cr:Al ratio higher than 90:10 induce earlier chromium infusion and lead to porous coating structure due to Kirkendall effect. Eventually chromium carbide forms to cease further growth of the dual coating structures. KEY WORDS: pack cementation; aluminization; intermetallics; gas-solid reactions; diffusion. 1. Introduction Steel surface can be modified by forming Fe Al intermetallic compounds as hard coatings. Such coating improves wear resistance. The coated steels can also be employed in higher temperature and corrosive environment. 1,2) Literature reports that the steel surface after hot dipping aluminization 3) can withstand temperature up to C (2 000 F). 4) Pack cementation process is an ideal alternative process for hot dipping aluminization. It is essentially an in situ chemical vapor deposition (CVD) coating process. 5) Four inter-processes, namely halide activation, gas diffusion, solid deposition, and solid diffusion reactions, take place in sequence to form solid coatings by gas-solid reactions. 6,7) The pack cementation process has the advantages of low cost and applicability to various materials shapes and sizes. The compositions of coating are dependent on processing temperature, time, substrate composition, and atmospheres. 8,9) Pack cementation aluminization of steel surface is normally performed at temperatures as high as C. 10,11) Pack aluminization at temperature below 700 C was reported by Xiang and Datta 12) who observed the formation of Alrich Fe 2Al 5 phase. According to Fe Al phase diagram, 13) aluminum has high solubility in iron. Fe and Al can also form intermetallic compounds including Fe 3Al, FeAl, FeAl 2, Fe 2Al 5, and FeAl 3 below C. 14) Among these, the Fe rich Fe Al intermetallic compounds, eg. Fe 3Al and FeAl, are favorable due to their less brittle and more desirable mechanical properties. 15,16) Fe 3Al is also reported to resist sulfidation and oxidation at high temperatures. 17) Minor Cr content in the Fe Al compounds can further improve their resistance to room temperature aqueous corrosion and hot corrosion by fused salt deposits. 18) To simultaneously chromize and aluminize steels, the partial pressures of Cr-halide and Al-halide in the powder pack must be comparable. Since the partial pressures of Alhalides are normally much higher than those of Cr-halides, the coatings of Cr Al alloys are only possible when the activity of Al in the pack is 2 3 orders below that of Cr ) There are two possible ways to overcome this problem: (1) using a lean pack where the metal powders contain higher Cr and lower Al concentrations; 21,22) or (2) performing dual instead of single heating processes first by treating at 925 C and then at 1150 C. 20) Meanwhile, chromium carbide (Cr 23C 6) was reported to form at surface due to the rapid outward diffusion of carbon during simultaneously chromizing and aluminizing. Formation of such carbide layer blocks the inward diffusion of other elements and locally depletes carbon from the steels. Pores are often observed due to vacancy-interstitial interac ISIJ

2 tions. 23) Therefore, it is important to carefully control the combination of chromizing and aluminizing. In this study, it is our intention to form Fe Al Cr intermetallic coatings on a SNCM439 steel. We intend to develop a two-stage treatment at low temperatures (550 and 750 C) to achieve Cr and Al co-deposition by pack cementation. The lower temperature processes are more cost-effective from energy point of view. It is also the objective of current research to control such that the preferable Fe-rich instead of Al-rich Fe-Al intermetallics are formed. 2. Experimental Procedures The substrate used in this study is commercial SNCM439 steel. Its nominal composition is listed in Table 1. The sample size is 30 mm dia 5 mm t disk cut from a round bar. These samples are first ground to 800-grit SiC paper and then ultrasonically cleaned in methanol for 10 min. Pack cementation process places the substrates, or SNCM439 steel in current study, within the powder packs. The powder packs consist of pure metal powders (Al and Cr) for diffusion, halide activator (NH 4Cl), and inactive ceramic powder (Al 2O 3) to protect the specimen from being oxidized. At elevated temperature, halide activator is decomposed by heating and form metal halides (e.g. AlCl, AlCl 3, CrCl 2, and CrCl 3) with metal powders through gaseous reactions. The high metal activity at surface of steel causes metal atoms to diffuse rapidly into the samples by decomposing metallic halides. An intermetallic compound layer is thus formed on steel surface. The decomposed halogen gas, e.g. Cl 2, then continues to react with metal powders left in the powder packs till the activity of metals in powder pack become equilibrium with that at steel surface. Therefore, halide activator plays an important role to control the vapor pressures of metal halides in powder packs 21) which maintain the activity of diffusing metals at specimen surface to sustain the diffusion with aids of gaseous reactions. The powder packs for current pack cementation process contain 3 12 wt.% Cr and Al metal powder mixtures, 3 12 wt.% NH 4Cl activator, and the rest 85 wt.% Al 2O 3 filler. The metal powder is a mixture of wt.% Cr and rest wt.% Al powders. SNCM439 steel specimens are wrapped in powder packs and loaded into a 316L stainless steel tube. Ar is supplied to flush the reaction chamber to remove moisture and air for 5 min prior the cementation treatment. The specimens and powder packs are then treated using a two-stage process, first at 550 C for 2 h before heating at 10 C/min to 750 C and then hold for 2 8 h. After treatments, the packs are cooled to room temperature by air cooling. The coated samples are removed from the pack and ultrasonically cleaned. XRD (X-ray diffraction) analyses are performed on steel surface to identify the intermetallic phases formed using Rigaku D/max-B diffractometer equipped with Cu target. Cross section of each sample is cut, mounted, ground and polished for SEM Table 1. Nominal composition of SNCM439 steel. Element C Si Mn Ni Cr Mo Fe wt.% Bal. (scanning electron microscopy, HITACHI S-4700 field emission SEM) and EDS (energy dispersive spectroscopy, HORIBA 7200-H) chemical analyses. 3. Results and Discussion 3.1. Effects of Treatment Time Table 2 lists the compositions of powder packs and treatment parameters by varying the treatment time at 750 C (sample 1 4). The total coating thickness increases from 25 to 43 μm and is proportional to the square root of 750 C treatment time with a 0.99 coefficient of linear correlation (R 2 ). The formations of these coatings are evidently diffusion controlled. Figure 1 shows that two layers are formed. The thickness of surface layer is approximately 6 μm for all conditions listed in Table 2, while the thickness of sub-surface layer increases with treatment time. In Figs. 1(c) and 1(d), pores in the coating layers are observed to increase with holding time as well. The surface coating phase is analyzed by XRD to consist of mainly FeAl solid solution in 2 h-treated sample, Fe 3Al in 4 h-treated samples, Cr 23C 6 and Cr 7C 3 in 6 h treated samples, and Cr 7C 3 in 8 h treated samples (Fig. 2). These results are very different from that reported recently 24) by a single step process at 700 C using Cr-Al alloy metal powders which forms an aluminum-rich brittle Fe 2Al 5 phase in contrast to iron-rich Fe 3Al phase formed by dual stage process in current research. In the case that Fe 2Al 5 phase containing 71 at.% of aluminum is formed at temperature as low as 700 C, 24) gaseous aluminum halides apparently provide fairly high activity of aluminum at the steel surface to form such aluminum-rich Fe 2Al 5 coatings. However, Fe 2Al 5 coatings are not favored due to its brittle characteristics. To avoid the formation of brittle Fe 2Al 5 phase, activity of aluminum must be reduced. In current study, the first stage 550 C treatment is designed to reduce the starting activity of aluminum. Therefore, a FeAl solid solution is first formed during the second stage 750 C treatment instead of the less favored Al-rich Fe 2Al 5. In Fig. 3, Surface aluminum concentration is shown to attain 42 at.% after 2 h-treatment at 750 C. The concentration of aluminum drops with diffusion depth till 24 μm into the steel. This layer corresponds to ferritic FeAl solid solution and is consistent with XRD analyses. When 750 C treatment further prolongs to 4 h, the surface aluminum concentration is reduced instead of increasing. The surface aluminum concentration reduces to 25 at.% corresponding to Fe 3Al phase (Fig. 2) which is the favorable Fe-rich intermetallic compound coating. Table 2. Sample no. Thickness of coatings treated for different holding time at 750 C stage. 750 C holding time (h) NH 4Cl Cr:Al in pack Coating thickness (μm) : : : : : ISIJ 128

3 (a) (b) Fig. 1. (c) (d) Cross section SEM microstructures of surface coatings with different second stage 750 C holding time: (a) 2 h, (b) 4 h, (c) 6 h, and (d) 8 h. Fig. 2. XRD spectra of sample 1, 2, 3, and 4 in Table 2. Fig. 3. Al and Cr concentration profiles in coating layers of samples treated for different time at 750 C (samples 1, 2, 3, and 4 in Table 2). It is also important to note that a highest aluminum concentration appears at depth of ~6 μm below surface after 4 h treatment at 750 C rather than at steel surface (Fig. 3). Apparently, the reduced aluminum concentration at steel surface indicates that aluminum in powder packs is exhausted after long treatment time at 750 C. The presence of peak aluminum concentration at a depth below steel surface requires the aluminum accumulation in steel to diffuse in two directions, one diffusing further into the steel and the other diffusing back to surface. Diffusions of aluminum in two directions are also observed in the concentration gradients of 6 h- and 8 h-treated samples. The 6 h- and 8 h- treated samples demonstrate peak concentrations at depth of 12 and 18 μ m below steel surface according to Fig. 3, respectively. The locations of maximum aluminum concentration move toward into the steel with time showing some of the aluminum in-take during the earlier treatment time diffuses further into steels. On the other hand, the surface aluminum concentration reduces from 42 to 25, 7, and 5 at.% for 2, 4, 6, and 8 h- treated samples, respectively. These are compared to the peak concentrations of 42, 28, 12, and 9 at.% for 2, 4, 6, and 8 h-treated samples, respectively. It is obvious that the surface aluminum concentration is lower than the peak concentrations for specimens treated longer than 2 h. The concentration gradient indicates that part of aluminum in steel coatings diffuse outward back into the powder packs. This dual diffusing activity provides an important mechanism controlling the formation of dual coatings in current study. Chromium concentrations in this series of experiments all remain at 2 5 wt.% in the steel which represent slow but steady diffusion (Fig. 3). It has been reported 18) that chromium activity in powder packs is well below that of aluminum, even though chromium powder content is designed to be greater than that of aluminum in powder packs. Furthermore, for chromium diffusion to occur requires higher temperature. Although, the 750 C treatment allows chromium to diffuse into the steel, chromium ISIJ

4 diffuses at a comparably slower rate than aluminum diffusing outward to the powder packs. The difference in diffusion directions and rates of chromium and aluminum thus causes Kirkendal effects. Geib and Rapp 8) report that, for simultaneous chromizing and aluminizing, porous coating starts to form due to back diffusion of chromium and aluminum late in the process when Cr- and Al-depletion zones form in the powder packs. Because iron has a much higher solubility for aluminum than chromium, aluminum is the main substitutional atoms diffusing into the steel in the initial treatment stage. When aluminum in-take reaches an equilibrium activity on steel surface with powder packs, aluminum diffusion slows down. The powder packs then start to be depleted of aluminum and would compete with steel for aluminum. The chromium atoms continue to substitute aluminum on steel surface when chromium infuses more pronouncedly into the steel due to chromium s higher activity in the powder packs. When chromium diffuses into the steels by decomposing chromium chlorides in powder packs, activity of chlorine is increased. Therefore, a driving force is produced for chlorine to react with the high aluminum concentration at the steel surface. Aluminum is thus migrating back from the steel surface into the powder packs while chromium diffuses into the steel. The outward diffusion of aluminum proceeds to react with chlorine in powder packs and to keep the pack cementation reactions in equilibrium with reduced chromium content in powder packs. Chromium and aluminum thus demonstrates a positive interaction parameter in both powder packs and steels. Although both aluminum and chromium of the powder packs tend to diffuse into the steel substrate at the beginning of reactions, the sequence of faster aluminum diffusion and slower chromium diffusion causes dual layers to form in competition. The competition of chromium with aluminum diffusion into steel is less pronounced in the beginning of cementation reactions. But after 4 h of diffusion time at 750 C, aluminum activity attains equilibrium between steel and powder pack, and chromium becomes the main elements diffusing into the steel. Further chromium diffusion drives the earlier diffused aluminum to diffuse back into the powder pack. Back diffusion is expected, because when chromium is reduced from the powder pack, the powder pack then has a high chlorine activity and tends to react with aluminum. The aluminum cementation reactions essentially proceed in reverse direction when metallic activity is depleted in powder packs. Eventually, porous coating is formed due to the Kirkendall effect by aluminum and chromium interdiffusion in the surface region as 750 C treatment extends longer than 6 h. When surface chromium concentration reaches ~4 at.%, chromium carbides are formed as shown in Fig Effects of Activator Concentration Table 3 lists a series of pack cementation samples by varying NH 4Cl activator concentrations (sample 5 8) while keeping the metal powder content and 750 C processing time constant for 2 h. The observations of dual layer coating thickness does not change much with the concentration of activator for samples treated using 3 9 wt.% of NH 4Cl activator as shown in Fig. 4 and stay at a thickness of μm. In 12 wt.% NH 4Cl treated sample, only a thin layer of coating lower than 5 μ m is observed in Fig. 4(d). The Table 3. Sample no. Thickness of coatings treating using different NH 4Cl concentrations and their process parameters. holding time (h) NH 4Cl Cr:Al Al 2O 3 Coating thickness (μm) : : : :20 85 (a) (b) Fig. 4. (c) (d) Cross section SEM microstructures of coatings using (a) 3 wt.% (sample 5), (b) 6 wt.% (sample 6), (c) 9 wt.% (sample 7), and (d) 12 wt.% (sample 8) NH 4Cl concentration in the powder packs ISIJ 130

5 surface coating phases are analyzed using XRD as shown in Fig. 5. In sample 5 and 6, surface coatings correspond to FeAl, while in sample 7 and 8, only carbides and Fe or substrate are present. The presence of Fe phase on surface analyses indicates that only small amount of aluminum are diffused into the steel when NH 4Cl activator concentration is greater than 9 wt.%. It is also interesting to note that many pores are observed only on the surface of sample 7 (Fig. 4(c)) while there is no pore in sample 8 (Fig. 4(d)). In sample 7 treated using 9 wt.% NH 4Cl, the aluminum concentration gradient in Fig. 6 demonstrates a peak concentration of 14 at.% at 7 μ m below steel surface. The aluminum concentration reduces in both inward and outward directions from 7 μ m depth inside the steel. Only 5 at.% aluminum concentration is observed at surface. It again shows that aluminum diffuses outward to the powder packs as suggested in Section 3.1. This is a confirmation of aluminum depletion in powder packs in using higher activator concentration. On the other hand, chromium continues to diffuse into the steel substrate at a slower rate. The great amount of pores are thus formed by Kirkendall effects due to chromium and aluminum interdiffusion as explained in Section 3.1. When activator concentration increases further to 12 wt.%, only carbide layer is present (Figs 4(d) and 5) on the Fig. 5. XRD spectra of samples 5, 6, 7, and 8 in Table 3 with different NH 4Cl concentrations (in wt.%). steel surface, and no distinguished subsurface Fe Al solid solution is formed. Almost none porosity is observed in the coatings as well as shown in Fig. 4(d). This indicates that the interdiffusion of chromium and aluminum is limited and thus no Kirkendall effect occurs. Apparently, aluminum diffusion is restricted when activator concentration is higher than a critical value. In this study, when activator concentration reaches 12 wt.%, the limited amount of aluminum is only enough to react with decomposed chlorine from the activator. Not much aluminum is available in the powder packs for diffusion. Therefore, diffusion of aluminum into steel is limited and only higher-containing chromium diffusion can proceed. Pores are thus significantly less in sample 8 than sample 7 as shown in Figs. 4(c) and 4(d). The activator overdose apparently accelerates alumiunum depletion and makes chromium diffusing at earlier time in comparison with the samples discussed in Section 3.1. Once the chromium reaches ~4 at.% at steel surface, chromium carbides are formed. The carbide layer serves as surface barrier and forbids further growth of coatings as observed in Figs. 4(c) and 4(d). In other word, formation of surface carbides marks the end of growth for surface coatings Effects of Cr:Al Ratio in Packs Table 4 lists a set of samples coated using varied Cr:Al ratio in the metal powders while NH 4Cl concentration is fixed at 3 wt.% and 750 C treatment time is fixed at 2 h. The coating thickness all remains at μm for Cr:Al ratio below 85:15. In the highest chromium containing powder pack where Cr:Al is 90:10 (sample 10), the thickness of coatings reduces sharply to 18 μm as shown in Fig. 7(d). The coatings in Fig. 7(d) also consist of more pores than other specimen in this series of experiments. According to XRD analyses shown in Fig. 8, the surface coating composition is FeAl for samples treated using Cr:Al ratio below 85:15, while Fe 3Al and (Cr, Fe) 7C 3 appear in sample 10 (Cr:Al=90:10). By observing aluminum concentration profiles in Fig. 9, it is noted that, for samples treated with Cr:Al ratio below 85:15, aluminum concentration all attain similar level of at.% at the steel surface. Aluminum concentrations then drop with depth till approximately 25 μm below steel surface which correspond to the thickness of FeAl phase as shown in Figs. 7(a) 7(c) and 8. The similar concentration profile suggests that increased aluminum content in powder packs does not necessarily increase surface aluminum concentration. A saturated aluminum level is attained by thermodynamic equilibrium between activator and the steel substrates. Activator concentration and diffusion time also modulate the extent of Table 4. Coating thickness and process parameters of samples treated in powder packs containing different Cr:Al ratio. Fig. 6. Al and Cr concentration profiles of coatings formed in powder packs with different NH 4Cl concentrations (in wt.%) (samples 5, 6, and 7 in Table 3). Sample no holding time(h) NH 4Cl Cr:Al (weight ratio) Al 2O 3 Coating thickness (μm) : : : : ISIJ

6 (a) (b) Fig. 7. (c) (d) Cross section SEM micrographs of samples treated using powder packs of different Cr:Al ratios: (a) 75:25, (b) 80:20, (c) 85:15, and (d) 90:10. Fig. 8. XRD spectra of coatings treated using different Cr:Al ratios (samples 9, 5, 1, and 10 in Table 4). Fig. 9. Al and Cr concentration profiles in the coatings using different Cr:Al ratios as in Table 4. aluminum back diffusion in pack cementation. For sample 10 treated using Cr:Al=90:10, surface concentration reaches only 29 at.% corresponding to Fe 3Al phase. The entire concentration profile of sample 10 is obviously lower than those obtained in samples treated using more aluminum. The lesser amount of aluminum diffusion is due to reduced supply of aluminum in powder packs besides those reacting with activator. Higher chromium in this sample also causes brittle chromium carbide to form on steel surface as shown in Fig. 8. Pores are thus visible within 5 μm below surface. According to the above observations, a critical amount of aluminum is required to react with the activator in powder packs. In current study, at least 15% of aluminum in metal powders is needed to achieve a saturated level of aluminum at the surface. When aluminum is below this level, the thickness and aluminum concentration of Fe Al layer is reduced. Chromium carbide reaction can then takes place earlier in the pack cementation process as shown in Fig. 7(d). 4. Conclusions The Cr/Al dual layer coatings are formed by combining a 550 C and 750 C two-stage pack cementation treatment. The Fe-rich Fe-Al intermetallics, including FeAl and Fe 3Al, are first formed on SNCM439 surface and a 6 μm Cr containing layer is formed on top of the FeAl solid solution. The low temperature 550 C treatment has a role in modulating the initial aluminum activity in powder packs. The FeAl intermetallic layer then starts to form during the 750 C treatments. The saturated content of surface aluminum is controlled to attain 40 at.% which permits the formation of favorable Fe-rich FeAl intermetallics on steel surface. The coating thickness and pores increases with the second stage 750 C holding time. When aluminum in-take is completed, aluminum starts to deplete in the powder pack. The surface Al concentration then back diffuses into the powder packs and becomes lower than that in the subsurface layer. Therefore, there exists an optimum holding 2012 ISIJ 132

7 time for the 750 C treatment stage. On the effects of NH 4Cl activator concentrations, too much activator can accelerate the powder-gas reactions and causes metal powder to deplete at earlier time. Back diffusion is then induced and porous coating structures are formed by Kirkendal effect on the surface due to chromium and aluminum interdiffusion. Maximum NH 4Cl activator concentration of 6% should be used. The increased Cr content in metal powder decreases coating thickness, since aluminum concentration is relatively lower and forms a thinner Fe Al layer. The lesser aluminum content incurs premature Cr deposition to form chromium carbides and inhibit further growth of coatings. Ideal Cr:Al ratio in the powder packs is below 85:15. Acknowledgements The financial support of National Science Council of Taiwan, R.O.C. through grant #NSC E project is acknowledged. REFERENCES 1) Z. D. Xiang, D. Zeng, C. Y. Zhu, S. R. Rose and P. K. Datta: Corros. Sci., 53 (2011), ) F. Masuyama: ISIJ Int., 41 (2001), ) S. Kobayashi, T. Yakou and T. Akou: Mat. Sci. Eng. A, A338(2002), 44. 4) W. J. Mock: Mater. Eng., 69 (1969), 46. 5) Z. D. Xiang, J. S. Burnell-Gray and P. K. Datta: J. Mater. Sci., 36 (2001), ) B. K. Gupta and L. L. Seigle: Thin Solid Films, 73 (1980), ) G. Hu, Z. Xu, J. Liu and Y. Li: Surf. Coat. Technol., 203 (2009), ) F. D. Geib and R. A. Rapp: Oxid. Met., 40 (1993), ) B. K. Gupta, A. K. Sakhel and L. L. Seigle: Thin Solid Film, 39 (1976), ) A. Bahadur and O. N. Mohanty: Mater. Trans., JIM, 36 (1995), ) T. H. Wang and L. L. Seigle: Mater. Sci. Eng. A, A108 (1989), ) Z. D. Xiang and P. K. Datta: Metall. Mater. Trans. A, 37A (2006), ) T. B. Massalski, H. Okamoto, P. R. Subramanian and L. Kacprzak: Binary Alloy Phase Diagrams, ASM International, Metals Park, Ohio, (1990), ) J. K. Chen and S. M. Chan: SEAISI Quart. J., 37-2 (2008), ) T. Sasaki and T. Yakou: Surf. Coat. Technol., 201 (2006), ) Y. Zhang and Y. Liou: J. Mater. Sci., 30 (1995), ) J. H. Devan: Oxidation of High Temperature Intermetallics, ed. by T. Grobstein and J. Doychak, The Mineral, Materials, and Metals Society, Warrendale, PA, (1989), ) R. Bianco and R. A. Rapp: J. Electrochem. Soc., 140 (1993), ) M. A. Harper and R. A. Rapp: Oxid. Met., 42 (1994), ) M. Zheng and R. A. Rapp: Oxid. Met., 49 (1998), ) R. Bianco, M. A. Harper and R. A. Rapp: JOM, 43 (1991), ) R. Bianco, R. A. Rapp and N. S. Jacobson: Oxid. Met., 38 (1992), ) V. Agarwal and A. R. Marder: Mater. Charact., 36 (1996), ) Y. Q. Wang, Y. Zhang and D. A. Wilson: Surf. Coat. Technol., 204 (2010), ISIJ