- Email: [email protected]
Preparation and oxidation of an Y2O3-dispersed chromizing coating by pack cementation at 800 °C
ARTICLE IN PRESS
Vacuum 82 (2008) 748–753 www.elsevier.com/locate/vacuum
Preparation and oxidation of an Y2O3-dispersed chromizing coating by pack cementation at 800 1C Y.B. Zhou, H. Chen, H. Zhang, Y. Wang College of Materials Science and Engineering, Heilongjiang Institute of Science and Technology, Harbin 150027, China Received 7 May 2007; received in revised form 28 October 2007; accepted 30 October 2007
Abstract Preparation and oxidation of an Y2O3-dispersed chromizing coating was investigated by chromizing an as-electrodeposited Ni–Y2O3 composite film using a conventional pack-cementation method but at a greatly decreased temperature (800 1C). For comparison, chromizing was also performed in the same condition on an as-deposited Ni film without Y2O3 particles. Oxidation at 900 1C indicated that compared to the Y2O3-free chromizing coating, the Y2O3-dispersed chromizing coating exhibited an increased oxidation resistance, due to the formation of purer and denser chromia scale. The Y2O3 effects on the coating formation and the coating oxidation behavior are discussed in detail. r 2007 Elsevier Ltd. All rights reserved. Keywords: Electrodeposition; Chromizing; Oxidation; REE
1. Introduction It is well known that the addition of small amount of certain reactive elements such as Y, Ce, La, etc. or their oxides can improve the oxidation resistance of metals. The phenomenon is referred to as ‘‘reactive-element effect (REE)’’ which was first reported in 1937 [1]. Since then, oxidation of alloys or coatings with addition of RE or RE oxide has been widely reported [2–4]. As for the chromiaformers, the decrease of the oxidation rate has been ascribed to the selective oxidation of chromium and the inversion of chromia growth mechanisms from predominant outward cation diffusion in the absence of RE into dominant inward oxygen diffusion [2–4]. REs or RE oxides are commonly added into alloys or coatings by different technique such as alloying [3], implantation [4] and sol–gel deposition [5]. Chromium coatings have been in widespread use to enhance high temperature oxidation, corrosion resistance of components [6,7]. Various chromizing processes have Corresponding author. Tel.: +86 451 88036526.
E-mail addresses: [email protected], [email protected] (Y.B. Zhou). 0042-207X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.vacuum.2007.10.010
been developed such as the pack cementation method, the molten-salt technique, and the vacuum chromizing process. Pack cementation is one of the easiest and cheapest processes to obtain the chromizing coatings. Unfortunately, the pack cementation was normally performed at temperatures above 1000 1C for durations of about 6–10 h, limited by the diffusion and reaction kinetics involved. Such a high temperature treatment inevitably limits the applications of chromizing coatings due to grain growth of the substrate materials, which has a detrimental effect on the mechanical properties of workpieces. Therefore, reducing pack-cementation temperature is required for widespread application of the chromizing coatings. Recently, a novel chromizing [8–10]/alumizing [11,12] technique at low temperature on nanostructured coatings produced using so-called surface mechanical attrition treatment (SMAT) was developed. The results exhibited that after the same treatment a much thicker Cr/Al-diffusion layers were obtained in the SMAT sample than in the coarse-grained one due to a large volume fraction of grain boundaries promoting the diffusion of Cr/Al. Compared to SMAT, electrodeposition is another method to produce nanostructured materials, and seems to be appropriate for lowering the chromizing temperature (or shortening the
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
duration) in a practical manner. At the same time, fine particles would be co-deposited with a metal or alloy matrix through composite electrodeposition. By considering the beneficial effect of RE or RE oxide on oxidation, Peng and co-workers [13] added CeO2 particles into chromium coatings by chromizing the co-deposited Ni– CeO2 nanocomposite coatings at 700 1C or 1120 1C [14,15]. The results showed that the CeO2-dispersed chromizing coatings exhibited better oxidation resistance compared to CeO2-free chromizing coatings. In this contribution, by considering the same beneficial effect of Y2O3 on the oxidation, a Y2O3-dispersed chromizing coating was produced at 800 1C using the same processing, and its oxidation performance was reported. For comparison, preparation and oxidation of chromium coatings on the Ni film were also carried out under the same condition.
749
Y2O3
200nm
100nm
Fig. 1. TEM bright-field images of Ni–Y2O3 composite film: (a) an area with a Y2O3 particle dispersion and (b) a major area without distribution Y2O3 particles.
2. Experimental Samples with dimensions of 15 mm 10 mm 2 mm were cut from an electrolytic nickel plate. They were ground to a final 800# SiC paper. After ultrasonically cleaning in acetone, they were electrodeposited with a 50-mm-thick film of Ni–Y2O3 composite from a nickel sulfate bath containing certain content of Y2O3 microparticles with an average particle size of 2.5 mm. The current density used was 3 A/dm2, the bath temperature 35 1C, and the pH 5.6–6.2. Before the electrodeposition, the samples were degreased in alkaline solution, dipped in acid (10% HCl) and finally washed with distilled water. During the electrodeposition, magnetic stirring was employed to maintain the uniform particles concentration and prevent the sedimentation. Energy dispersive X-ray analysis (EDAX) results showed the as-deposited composite film contained around 7–10 wt% Y2O3. For comparison, a 50-mm-thick Ni film was also electroplated on the pure Ni using the same bath and the same deposition parameters but without adding Y2O3 microparticles. Chromizing on the samples coated with the Ni–Y2O3 composite and nickel film were carried out using pack cementation at 800 1C for 5 h in a powder mixture of 50% Cr (particle size: 75 mm)+47% Al2O3 (particle size: 75 mm)+3% NH4Cl (in wt%) in a pure Ar atmosphere. Afterward, the samples were brushed, cleaned in bubbling distilled water for 30 min and finally ultrasonically cleaned in acetone to remove any loosely embedded pack particles. The oxidation experiments were carried out in air at 900 1C up to 20 h and the mass measurements were conducted after fixed time intervals using a balance with 0.01 mg sensitivity. The microstructure of the various chromized coatings before and after oxidation was investigated using transmission electron microscopy (TEM), scanning electron microscopy (SEM) with EDAX and X-ray diffraction (XRD). 3. Results Both Ni film and Ni–Y2O3 composite film were characterized using TEM. Fig. 1 shows some parallel
100nm Fig. 2. TEM bright-field image of Ni film. The inserted image is the corresponding SAD pattern.
TEM bright-field images of Ni–Y2O3 composite film. Although the distribution of micrometer-size Y2O3 particles could be observed in some locations (Fig. 1a), the particles were not present in most areas. A representative TEM image without any Y2O3 particle is presented in Fig. 1b. The image reveals that the coating generally comprised of Ni grains in a range of 10–150 nm. Fig. 2 presents a bright-field TEM image of the Ni film along with the corresponding selected-area-diffraction (SAD) pattern. Clearly, the Ni film was finer grain than the Ni–Y2O3 composite, with the grain size of Ni ranging from 15 to 60 nm. The grain size is approaching the range of grain sizes of electrodeposited pure Ni films, as reported elsewhere [13]. Both films formed numerous twins. No defects such as pores and cracks were seen. After pack cementation at 800 1C for 5 h, chromized layers on various substrates were prepared. The average chromium concentrations in the surface zone (o4 mm: the profile depth of electron beam) was 50 wt% for the Y2O3-free chromium coating, and 58 for the Y2O3-dispersed chromium coating on a basis of EDAX
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
750
analysis. However, the thickness of the two-chromized layer was different. Fig. 3 shows the comparison of the cross-sectional micrographs of the chromizing coatings on Ni film and Ni–Y2O3 composite. As is evident, the twochromized layers exhibited a double-layer structure, as seen the areas 1 and 2 in Fig. 3b and c. The thickness of chromized layers on Ni film and Ni–Y2O3 composite coating was 32 and 40 mm, respectively. Fig. 3c indicates that the chromizing coatings on Ni film consisted of a relatively equiaxed grains outer layer (area 1 in Fig. 3c) and a columnar (area 2 in Fig. 3c) inner layer. However, with the addition of Y2O3 particles, the inner layer became finer equiaxed grains, as seen in area 2 in Fig. 3d. Here, it is noteworthy that the gray blocky in the chromium coating, indicated by arrows in Fig. 3b, are the dispersed Y2O3 particles. They were from the co-deposited Y2O3 particles
in the composite film. TEM observations show that the grain size of the Y2O3-free chromium coating is generally large than 1200 nm (Fig. 4a). In contrast, the grain size of the Y2O3-dispersed chromium coating is generally below 800 nm (seen in Fig. 4b), indicating the grain growth of the Y2O3-free chromium coating was retarded by the dispersed Y2O3. The dispersoids (as arrowed in Fig. 4b) are much smaller than the size of the original microparticles used, suggesting the dissolution and re-precipitation of some Y2O3 microparticles during the chromizing process. From Figs. 3 and 4, it can be seen that the Y2O3-dispersed chromium coating exhibited a finer grain structure. The result suggests that Y2O3 particles dispersed in the electrodeposited Ni film significantly retard the grain growth of the chromium coating, which enhances the diffusion of chromium during pack cementation and lead
1 2
1 2
Fig. 3. Cross-sectional micrographs of chromizing coatings on (a) and (c) Ni film, and (b) and (d) Ni–Y2O3 composite.
1000nm
500nm
Fig. 4. TEM bright-field images of the chromized coatings on (a) Ni-film and (b) Ni–Y2O3 composite film.
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
to higher Cr content of the Y2O3-dispersed chromium coating than that of the Y2O3-free coating at a given distance from the coating surface, as seen in Fig. 5. 80 chromizing coating on Ni-film chromizing coating on Ni-Y2O3 composite film
70
Cr content (wt%)
60 50 40 30 20 10 0 -5
0
5
10 15 20 25 30 35 Distance from the surface (µm)
40
45
Fig. 5. Chromium concentration profiles for various chromizing coatings at 800 1C for 5 h.
chromizing coating on Ni plating
1.4 Mass change (mg/cm2)
chromizing coating on Ni-Y2O3 composite coating
1.2 1.0
1.64×10-12g2/cm4•s
0.8 0.6 0.4 1.22×10-11g2/cm4•s
0.2 0.0 0
5
10 15 Time (hr)
20
Fig. 6. Oxidation kinetics of various chromizing coatings at 900 1C for 20 h.
751
Fig. 6 shows the oxidation kinetics of two chromizing coatings in air for 20 h. During oxidation and cooling no spallation occurred for the two chromizing coatings. Clearly, the Y2O3-dispersed chromium coating exhibited an apparently low scaling rate during a short transient stage of oxidation (the first 2–3 h). In this stage, a continuous, protective chromia scale was expected to have been thermally grown. After this period, the oxidation rate maintained so extremely low that no significant mass gain occurred. However, a significant mass gain occurred for the Y2O3-free chromium coating. The corresponding oxidation parabolic constant calculated for each curve also indicated in Fig. 6. The Y2O3-free chromium coating oxidized faster than the Y2O3-dispersed chromium coating. Compared to the oxidation of the Y2O3-free chromium coating, the scaling rate of the Y2O3-dispersed chromium coating was reduced by one order of magnitude. Fig. 7 shows the SEM top-views of the scales formed on two chromizing coatings after exposure at 900 1C for 2.5 h. The Y2O3-free chromium coating exhibited two distinct areas with oxide crystals in different size, as seen in Fig. 7a. The large crystals were NiO (area A) and the finer one (area B) was mainly chromia or NiCr2O4 spinel according to EDAX and XRD investigation (Fig. 8). However, only fine chromia grains were observed on the surface of the Y2O3-dispersed chromium coating. The XRD characterization demonstrated that the oxide phases on the two chromium coatings after 20 h oxidation were composed of NiO, NiCr2O4 and Cr2O3, as seen in Fig. 8. However, the probed intensities of the peaks of NiO and NiCr2O4 spinel were both extremely weaker than those of Cr2O3. At the same time, the NiO peaks from the Y2O3-dispersed chromium coating were weak, while those from the Y2O3-free chromium coating were much stronger. This suggests that the Y2O3 addition significantly suppresses NiO growth, and a purer and adherent chromia scale was formed on the Y2O3-dispersed chromium coating. Fig. 9 shows the cross-sectional micrographs of the scales on the two chromium coatings after exposure at 900 1C for 20 h. The Y2O3-free chromium coating formed a continuous chromia layer on which Ni-rich oxide was seen.
A
B 5µm
5µm
Fig. 7. Surface morphologies of the chromia scale formed on chromizing coatings on (a) Ni film and (b) Ni–Y2O3 composite coating at 900 1C for 2.5 h exposure.
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
752
However, the scale was significantly buckled. Cracks below buckles were seen (see ‘‘A’’—indicated areas in Fig. 9a). Besides, internal oxidation of chromium occurred in some regions. In contrast, the scale formed on the Y2O3-dispersed chromium coating was very flat and scale spallation and internal oxidation did not occur. In this case, a purer and adherent chromia scale was formed, as seen in Fig. 9b. 4. Discussion In general, all chromium coatings exhibited a doublelayered structure. The outer layers formed from the contribution of the outward growth, and the inner layers formed from the contribution of the inward growth. The chromizing coating on Ni film exhibited a very thin outer layer, implying that the chromizing processing was dominated by inward chromium diffusion. However, the formation of the thicker outer layer (areas 1 in Fig. 3b) for Y2O3-dispersed chromizing coating compared to Y2O3-free chromizing coating (areas 1 in Fig. 3c), a same result as in Refs. [13–15], may be related to that of the added Y2O3
30
40
Cr2O3
(Ni Cr) Cr2O3
Cr2O3 Cr2O3
(a) chromizing on Ni film
Cr2O3 NiCr2O4 Cr2O3 Cr O NiO 2 3
Cr2O3
20
Cr2O3 Cr2O3
(Ni Cr) Cr2O3
Cr2O3
Cr2O3 Cr2O3
Cr2O3 Cr2O3 NiO
NiCr2O4
Intensity (cts)
Cr2O3
(a) chromizing on Ni-Y2O3 fim
50 60 2 - Theta (degree)
70
80
90
Fig. 8. Comparison of XRD characterization of the oxide scales formed on two chromizing coatings at 900 1C for 20 h.
particles in the composite coatings acted as a physical barrier to block the inward chromium diffusion during the pack cementation, leading to the suppression of the inward coating growth to a certain extent. From the comparison of the grain size of the coatings with and without Y2O3 particles before (Figs. 1 and 2) and after chromizing treatment (Figs. 3 and 4), it concluded that the dispersed Y2O3 particles significantly retarded the grain growth during the pack cementation. The grain refinements will enhance the diffusion of chromium during pack cementation [8–10,13–15], which leads to higher Cr content for Y2O3-dispersed chromizing coating than that of Y2O3-free coating at a given distance from the coating surface, as seen in Fig. 5. As compared to oxidation of the chromium coating on the Ni-plated sample, the one on the Ni–Y2O3 coated sample exhibits an extremely low mass gain (Fig. 6), as a result of the formation of a purer and denser chromia scale, as seen in Figs. 7–9. It has been extensively reported that the addition of Y2O3 dispersoid improves oxidation resistance of the alloy by the following two routes [16,17]: (1) to promote the selective oxidation of chromium by decreasing the critical content for the formation of a continuous chromia scale and (2) to reduce scaling rate by blocking predominant cationic diffusion during oxidation. From Fig. 6, it could be seen that the Y2O3-dispersed chromium coating exhibits an apparently low scaling rate, particularly in the initial stage. This can be interpreted as follows. The increased grain boundaries as a result of the refinement of the chromizing coating due to the dispersed Y2O3, together with dispersed Y2O3 particles on the coating surface, are both available sites for chromia nucleation at onset of oxidation. The increased chromia nuclei shorten the time for their linkage through the oxide lateral growth. Furthermore, the lateral growth can be enhanced by accelerated chromium diffusion to the oxidation front via the increased grain boundaries in the finer grain coating. In another words, the fine grain of the Y2O3-dispersed chromized coating enhanced the diffusion of chromium to the oxidation front and consequently accelerated the formation of a continuous chromia layer in a shorter time [13–15]. After the formation of the continuous chromia scale, the finer grain scale formed on
NiO A
Fig. 9. Cross-sectional SEM micrograph of chromizing coatings on (a) Ni film and (b) Ni–Y2O3 composite coating at 900 1C for 20 h.
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
the Y2O3-dispersed chromium coating (Fig. 7) should grow even faster than that formed on the Y2O3-free chromium coating due to an increase in the number of grain boundaries per unit volume. However, the fact that a significant scaling rate reduction occurs suggests that grain boundary diffusion of Cr cations is hindered to a great extent. The reason may be that the chromia scale is incorporated by Y ions released from the dispersed Y2O3 particles when they are incorporated into the growing scale [18–20]. Once being incorporated into the growing scale, they would potentially improve the oxidation resistance through blocking the dominant grain boundary diffusion of chromium cations [2–4], which could be also seen from the cross-section of the scales, as discussed below. By comparison of the cross-section of the scales formed on the chromizing coatings with and without Y2O3 in Fig. 9, the Y2O3-free chromizing coating formed a continuous chromia layer on which Ni-rich oxide was seen. Cavities below the scales formed on the Y2O3-free chromizing coating. They form as a result of condensation of chromium cation vacancies at the interface, which are injected from the scale counter balancing the outward chromium diffusion during oxidation, or from the coating substrate resulting from the relative diffusion rate of Ni to Cr into the coating interior due to the ‘‘Kirkendall’’ effect. The growth of the interfacial cavities is the reason for scale spallation. However, the chromia scale formed on Ni– Y2O3 coated sample is much adherent, no interfacial cavities was observed. The reason lies in two aspects: (1) a low consumption rate of chromium by oxidation due to the formation of purer and denser chromia scale on the Y2O3-dispersed chromizing coating reduces the flux of the cation vacancies toward the interface and (2) the dispersed Y2O3 itself may act as sites for vacancy’s sinking at the interface [3]. Hence, the formation and extension of the interfacial cavity is effectively suppressed in the Y2O3 presence. From the above analysis, the lower oxidation rate of Y2O3-dispersed chromizing coating was associated with the effect of Y2O3, which can be summarized as follows. First, Y2O3 co-deposited in the electrodeposited Ni matrix led to the formation of a fine grain chromium coating, which enhanced the diffusion of chromium to the oxidation front and consequently promoted the formation of chromia layer in a shorter time. Second, the Y2O3 dispersion may release Y ions, which enters the growing chromia scale [18–20] and segregates to the scale grain boundaries. In this case, the dominant diffusion of chromium cations during the growth of chromia scale is blocked, leading to a decrease of oxidation. The segregation of RE ions at the grain boundaries of the thermally grown oxide scales has been extensively evidenced through TEM characterization [18–21]. In this case, the formation of a fine-grain chromia scale on the Y2O3-dispersed chromium coating, suggesting that the segregation of Y ions along the chromia grain boundaries might occur. Segregated Y ions could block the oxide grain growth through a grain boundary pinning [22] and ‘‘solute-drag’’ effect [19].
753
5. Conclusions By pack cementation on the as-electrodeposited Ni– Y2O3 composite at 800 1C, a Y2O3-dispersed chromizing coatings were manufactured. The chromizing coatings formed on the Ni–Y2O3 composite coating were finer grained than that on the pure Ni films due to the added Y2O3 particles retarding the grain growth during chromizing. The oxidation at 900 1C for 20 h showed that the Y2O3modified chromizing coating exhibited a superior oxidation resistance due to the formation of purer and denser chromia scale. The positive effects of Y2O3 on the oxidation of the chromizing coating are summarized as follows: (1) to retard the grain growth of the chromizing coating by anchoring the movement of grain boundaries, which enhance Cr diffusion to the oxidation front and consequently accelerated the formation of a continuous purer chromia scale and (2) Y ions from the Y2O3dispersed coating incorporating into the growing scale and segregating to the chromia grain boundaries block the dominant outward diffusion of chromium cations during the oxidation. Acknowledgment Project (06-13) was supported by the Scientific Research Startup Foundation of Heilongjiang Institute of Science and Technology. References [1] Pfeil LB. Improvement in heat-resisting alloys. UK Patent No. 459848 (1937). [2] Moon DP. Mater Sci Technol 1989;5:754–63. [3] Hou PY, Stringer J. Mater Sci Eng A 1995;202:1–10. [4] Hou PY, Stringer J. Oxid Met 1988;29:45–73. [5] Czerwinski F, Smeltzer WW. Oxid Met 1995;43:25. [6] Koo CH, Yu TH. Surf Coat Technol 2000;126:171. [7] Lee JW, Duh JG, Tsai SY. Surf Coat Technol 2002;153:59–66. [8] Wang ZB, Tao NR, Tong WP, Lu J, Lu K. Acta Mater 2003;51: 4319–29. [9] Wang ZB, Lu J, Lu K. Acta Mater 2005;53:2081–9. [10] Wang ZB, Lu J, Lu K. Surf Coat Technol 2006;201:2796–801. [11] Zhan ZL, He YD, Wang DR, Gao W. Trans Nonferrous Met Soc China 2006;16:647–53. [12] Zhan ZL, He YD, Wang DR, Gao W. Oxid Met 2007;68(5–6): 243–51. [13] Zhang H, Peng X, Zhao J, Wang F. Electrochem Solid-State Lett 2007;10(3):C12–5. [14] Zhu L, Peng X, Yan J, Wang F. Oxid Met 2004;62:411–26. [15] Yan J, Peng X, Wang F. Mater Sci Eng A 2006;426:266–73. [16] Ramanarayanan TA, Ayer R, Petkovic-Luton R, Leta DP. Oxid Met 1988;29:445–72. [17] Stringer J, Wilcox BA, Jaffee RI. Oxid Met 1972;5:11–47. [18] Yurek GJ, Przybylski K, Garratt-Reed AJ. J Electrochem Soc 1987;134:2643–4. [19] Cotell CM, Yurek GJ, Hussey RJ, Mitchell DF, Graham MJ. Oxid Met 1990;34:173–200. [20] Pint BA. Oxid Met 1996;45:1–37. [21] Cotell CM, Yurek GJ, Hussey RJ, Mitchell DF, Graham MJ. J Electrochem Soc 1987;134:1871–972. [22] Hindam HM, Whittle DP. J Electrochem Soc 1982;129:1147–9.
Vacuum 82 (2008) 748–753 www.elsevier.com/locate/vacuum
Preparation and oxidation of an Y2O3-dispersed chromizing coating by pack cementation at 800 1C Y.B. Zhou, H. Chen, H. Zhang, Y. Wang College of Materials Science and Engineering, Heilongjiang Institute of Science and Technology, Harbin 150027, China Received 7 May 2007; received in revised form 28 October 2007; accepted 30 October 2007
Abstract Preparation and oxidation of an Y2O3-dispersed chromizing coating was investigated by chromizing an as-electrodeposited Ni–Y2O3 composite film using a conventional pack-cementation method but at a greatly decreased temperature (800 1C). For comparison, chromizing was also performed in the same condition on an as-deposited Ni film without Y2O3 particles. Oxidation at 900 1C indicated that compared to the Y2O3-free chromizing coating, the Y2O3-dispersed chromizing coating exhibited an increased oxidation resistance, due to the formation of purer and denser chromia scale. The Y2O3 effects on the coating formation and the coating oxidation behavior are discussed in detail. r 2007 Elsevier Ltd. All rights reserved. Keywords: Electrodeposition; Chromizing; Oxidation; REE
1. Introduction It is well known that the addition of small amount of certain reactive elements such as Y, Ce, La, etc. or their oxides can improve the oxidation resistance of metals. The phenomenon is referred to as ‘‘reactive-element effect (REE)’’ which was first reported in 1937 [1]. Since then, oxidation of alloys or coatings with addition of RE or RE oxide has been widely reported [2–4]. As for the chromiaformers, the decrease of the oxidation rate has been ascribed to the selective oxidation of chromium and the inversion of chromia growth mechanisms from predominant outward cation diffusion in the absence of RE into dominant inward oxygen diffusion [2–4]. REs or RE oxides are commonly added into alloys or coatings by different technique such as alloying [3], implantation [4] and sol–gel deposition [5]. Chromium coatings have been in widespread use to enhance high temperature oxidation, corrosion resistance of components [6,7]. Various chromizing processes have Corresponding author. Tel.: +86 451 88036526.
E-mail addresses: [email protected], [email protected] (Y.B. Zhou). 0042-207X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.vacuum.2007.10.010
been developed such as the pack cementation method, the molten-salt technique, and the vacuum chromizing process. Pack cementation is one of the easiest and cheapest processes to obtain the chromizing coatings. Unfortunately, the pack cementation was normally performed at temperatures above 1000 1C for durations of about 6–10 h, limited by the diffusion and reaction kinetics involved. Such a high temperature treatment inevitably limits the applications of chromizing coatings due to grain growth of the substrate materials, which has a detrimental effect on the mechanical properties of workpieces. Therefore, reducing pack-cementation temperature is required for widespread application of the chromizing coatings. Recently, a novel chromizing [8–10]/alumizing [11,12] technique at low temperature on nanostructured coatings produced using so-called surface mechanical attrition treatment (SMAT) was developed. The results exhibited that after the same treatment a much thicker Cr/Al-diffusion layers were obtained in the SMAT sample than in the coarse-grained one due to a large volume fraction of grain boundaries promoting the diffusion of Cr/Al. Compared to SMAT, electrodeposition is another method to produce nanostructured materials, and seems to be appropriate for lowering the chromizing temperature (or shortening the
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
duration) in a practical manner. At the same time, fine particles would be co-deposited with a metal or alloy matrix through composite electrodeposition. By considering the beneficial effect of RE or RE oxide on oxidation, Peng and co-workers [13] added CeO2 particles into chromium coatings by chromizing the co-deposited Ni– CeO2 nanocomposite coatings at 700 1C or 1120 1C [14,15]. The results showed that the CeO2-dispersed chromizing coatings exhibited better oxidation resistance compared to CeO2-free chromizing coatings. In this contribution, by considering the same beneficial effect of Y2O3 on the oxidation, a Y2O3-dispersed chromizing coating was produced at 800 1C using the same processing, and its oxidation performance was reported. For comparison, preparation and oxidation of chromium coatings on the Ni film were also carried out under the same condition.
749
Y2O3
200nm
100nm
Fig. 1. TEM bright-field images of Ni–Y2O3 composite film: (a) an area with a Y2O3 particle dispersion and (b) a major area without distribution Y2O3 particles.
2. Experimental Samples with dimensions of 15 mm 10 mm 2 mm were cut from an electrolytic nickel plate. They were ground to a final 800# SiC paper. After ultrasonically cleaning in acetone, they were electrodeposited with a 50-mm-thick film of Ni–Y2O3 composite from a nickel sulfate bath containing certain content of Y2O3 microparticles with an average particle size of 2.5 mm. The current density used was 3 A/dm2, the bath temperature 35 1C, and the pH 5.6–6.2. Before the electrodeposition, the samples were degreased in alkaline solution, dipped in acid (10% HCl) and finally washed with distilled water. During the electrodeposition, magnetic stirring was employed to maintain the uniform particles concentration and prevent the sedimentation. Energy dispersive X-ray analysis (EDAX) results showed the as-deposited composite film contained around 7–10 wt% Y2O3. For comparison, a 50-mm-thick Ni film was also electroplated on the pure Ni using the same bath and the same deposition parameters but without adding Y2O3 microparticles. Chromizing on the samples coated with the Ni–Y2O3 composite and nickel film were carried out using pack cementation at 800 1C for 5 h in a powder mixture of 50% Cr (particle size: 75 mm)+47% Al2O3 (particle size: 75 mm)+3% NH4Cl (in wt%) in a pure Ar atmosphere. Afterward, the samples were brushed, cleaned in bubbling distilled water for 30 min and finally ultrasonically cleaned in acetone to remove any loosely embedded pack particles. The oxidation experiments were carried out in air at 900 1C up to 20 h and the mass measurements were conducted after fixed time intervals using a balance with 0.01 mg sensitivity. The microstructure of the various chromized coatings before and after oxidation was investigated using transmission electron microscopy (TEM), scanning electron microscopy (SEM) with EDAX and X-ray diffraction (XRD). 3. Results Both Ni film and Ni–Y2O3 composite film were characterized using TEM. Fig. 1 shows some parallel
100nm Fig. 2. TEM bright-field image of Ni film. The inserted image is the corresponding SAD pattern.
TEM bright-field images of Ni–Y2O3 composite film. Although the distribution of micrometer-size Y2O3 particles could be observed in some locations (Fig. 1a), the particles were not present in most areas. A representative TEM image without any Y2O3 particle is presented in Fig. 1b. The image reveals that the coating generally comprised of Ni grains in a range of 10–150 nm. Fig. 2 presents a bright-field TEM image of the Ni film along with the corresponding selected-area-diffraction (SAD) pattern. Clearly, the Ni film was finer grain than the Ni–Y2O3 composite, with the grain size of Ni ranging from 15 to 60 nm. The grain size is approaching the range of grain sizes of electrodeposited pure Ni films, as reported elsewhere [13]. Both films formed numerous twins. No defects such as pores and cracks were seen. After pack cementation at 800 1C for 5 h, chromized layers on various substrates were prepared. The average chromium concentrations in the surface zone (o4 mm: the profile depth of electron beam) was 50 wt% for the Y2O3-free chromium coating, and 58 for the Y2O3-dispersed chromium coating on a basis of EDAX
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
750
analysis. However, the thickness of the two-chromized layer was different. Fig. 3 shows the comparison of the cross-sectional micrographs of the chromizing coatings on Ni film and Ni–Y2O3 composite. As is evident, the twochromized layers exhibited a double-layer structure, as seen the areas 1 and 2 in Fig. 3b and c. The thickness of chromized layers on Ni film and Ni–Y2O3 composite coating was 32 and 40 mm, respectively. Fig. 3c indicates that the chromizing coatings on Ni film consisted of a relatively equiaxed grains outer layer (area 1 in Fig. 3c) and a columnar (area 2 in Fig. 3c) inner layer. However, with the addition of Y2O3 particles, the inner layer became finer equiaxed grains, as seen in area 2 in Fig. 3d. Here, it is noteworthy that the gray blocky in the chromium coating, indicated by arrows in Fig. 3b, are the dispersed Y2O3 particles. They were from the co-deposited Y2O3 particles
in the composite film. TEM observations show that the grain size of the Y2O3-free chromium coating is generally large than 1200 nm (Fig. 4a). In contrast, the grain size of the Y2O3-dispersed chromium coating is generally below 800 nm (seen in Fig. 4b), indicating the grain growth of the Y2O3-free chromium coating was retarded by the dispersed Y2O3. The dispersoids (as arrowed in Fig. 4b) are much smaller than the size of the original microparticles used, suggesting the dissolution and re-precipitation of some Y2O3 microparticles during the chromizing process. From Figs. 3 and 4, it can be seen that the Y2O3-dispersed chromium coating exhibited a finer grain structure. The result suggests that Y2O3 particles dispersed in the electrodeposited Ni film significantly retard the grain growth of the chromium coating, which enhances the diffusion of chromium during pack cementation and lead
1 2
1 2
Fig. 3. Cross-sectional micrographs of chromizing coatings on (a) and (c) Ni film, and (b) and (d) Ni–Y2O3 composite.
1000nm
500nm
Fig. 4. TEM bright-field images of the chromized coatings on (a) Ni-film and (b) Ni–Y2O3 composite film.
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
to higher Cr content of the Y2O3-dispersed chromium coating than that of the Y2O3-free coating at a given distance from the coating surface, as seen in Fig. 5. 80 chromizing coating on Ni-film chromizing coating on Ni-Y2O3 composite film
70
Cr content (wt%)
60 50 40 30 20 10 0 -5
0
5
10 15 20 25 30 35 Distance from the surface (µm)
40
45
Fig. 5. Chromium concentration profiles for various chromizing coatings at 800 1C for 5 h.
chromizing coating on Ni plating
1.4 Mass change (mg/cm2)
chromizing coating on Ni-Y2O3 composite coating
1.2 1.0
1.64×10-12g2/cm4•s
0.8 0.6 0.4 1.22×10-11g2/cm4•s
0.2 0.0 0
5
10 15 Time (hr)
20
Fig. 6. Oxidation kinetics of various chromizing coatings at 900 1C for 20 h.
751
Fig. 6 shows the oxidation kinetics of two chromizing coatings in air for 20 h. During oxidation and cooling no spallation occurred for the two chromizing coatings. Clearly, the Y2O3-dispersed chromium coating exhibited an apparently low scaling rate during a short transient stage of oxidation (the first 2–3 h). In this stage, a continuous, protective chromia scale was expected to have been thermally grown. After this period, the oxidation rate maintained so extremely low that no significant mass gain occurred. However, a significant mass gain occurred for the Y2O3-free chromium coating. The corresponding oxidation parabolic constant calculated for each curve also indicated in Fig. 6. The Y2O3-free chromium coating oxidized faster than the Y2O3-dispersed chromium coating. Compared to the oxidation of the Y2O3-free chromium coating, the scaling rate of the Y2O3-dispersed chromium coating was reduced by one order of magnitude. Fig. 7 shows the SEM top-views of the scales formed on two chromizing coatings after exposure at 900 1C for 2.5 h. The Y2O3-free chromium coating exhibited two distinct areas with oxide crystals in different size, as seen in Fig. 7a. The large crystals were NiO (area A) and the finer one (area B) was mainly chromia or NiCr2O4 spinel according to EDAX and XRD investigation (Fig. 8). However, only fine chromia grains were observed on the surface of the Y2O3-dispersed chromium coating. The XRD characterization demonstrated that the oxide phases on the two chromium coatings after 20 h oxidation were composed of NiO, NiCr2O4 and Cr2O3, as seen in Fig. 8. However, the probed intensities of the peaks of NiO and NiCr2O4 spinel were both extremely weaker than those of Cr2O3. At the same time, the NiO peaks from the Y2O3-dispersed chromium coating were weak, while those from the Y2O3-free chromium coating were much stronger. This suggests that the Y2O3 addition significantly suppresses NiO growth, and a purer and adherent chromia scale was formed on the Y2O3-dispersed chromium coating. Fig. 9 shows the cross-sectional micrographs of the scales on the two chromium coatings after exposure at 900 1C for 20 h. The Y2O3-free chromium coating formed a continuous chromia layer on which Ni-rich oxide was seen.
A
B 5µm
5µm
Fig. 7. Surface morphologies of the chromia scale formed on chromizing coatings on (a) Ni film and (b) Ni–Y2O3 composite coating at 900 1C for 2.5 h exposure.
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
752
However, the scale was significantly buckled. Cracks below buckles were seen (see ‘‘A’’—indicated areas in Fig. 9a). Besides, internal oxidation of chromium occurred in some regions. In contrast, the scale formed on the Y2O3-dispersed chromium coating was very flat and scale spallation and internal oxidation did not occur. In this case, a purer and adherent chromia scale was formed, as seen in Fig. 9b. 4. Discussion In general, all chromium coatings exhibited a doublelayered structure. The outer layers formed from the contribution of the outward growth, and the inner layers formed from the contribution of the inward growth. The chromizing coating on Ni film exhibited a very thin outer layer, implying that the chromizing processing was dominated by inward chromium diffusion. However, the formation of the thicker outer layer (areas 1 in Fig. 3b) for Y2O3-dispersed chromizing coating compared to Y2O3-free chromizing coating (areas 1 in Fig. 3c), a same result as in Refs. [13–15], may be related to that of the added Y2O3
30
40
Cr2O3
(Ni Cr) Cr2O3
Cr2O3 Cr2O3
(a) chromizing on Ni film
Cr2O3 NiCr2O4 Cr2O3 Cr O NiO 2 3
Cr2O3
20
Cr2O3 Cr2O3
(Ni Cr) Cr2O3
Cr2O3
Cr2O3 Cr2O3
Cr2O3 Cr2O3 NiO
NiCr2O4
Intensity (cts)
Cr2O3
(a) chromizing on Ni-Y2O3 fim
50 60 2 - Theta (degree)
70
80
90
Fig. 8. Comparison of XRD characterization of the oxide scales formed on two chromizing coatings at 900 1C for 20 h.
particles in the composite coatings acted as a physical barrier to block the inward chromium diffusion during the pack cementation, leading to the suppression of the inward coating growth to a certain extent. From the comparison of the grain size of the coatings with and without Y2O3 particles before (Figs. 1 and 2) and after chromizing treatment (Figs. 3 and 4), it concluded that the dispersed Y2O3 particles significantly retarded the grain growth during the pack cementation. The grain refinements will enhance the diffusion of chromium during pack cementation [8–10,13–15], which leads to higher Cr content for Y2O3-dispersed chromizing coating than that of Y2O3-free coating at a given distance from the coating surface, as seen in Fig. 5. As compared to oxidation of the chromium coating on the Ni-plated sample, the one on the Ni–Y2O3 coated sample exhibits an extremely low mass gain (Fig. 6), as a result of the formation of a purer and denser chromia scale, as seen in Figs. 7–9. It has been extensively reported that the addition of Y2O3 dispersoid improves oxidation resistance of the alloy by the following two routes [16,17]: (1) to promote the selective oxidation of chromium by decreasing the critical content for the formation of a continuous chromia scale and (2) to reduce scaling rate by blocking predominant cationic diffusion during oxidation. From Fig. 6, it could be seen that the Y2O3-dispersed chromium coating exhibits an apparently low scaling rate, particularly in the initial stage. This can be interpreted as follows. The increased grain boundaries as a result of the refinement of the chromizing coating due to the dispersed Y2O3, together with dispersed Y2O3 particles on the coating surface, are both available sites for chromia nucleation at onset of oxidation. The increased chromia nuclei shorten the time for their linkage through the oxide lateral growth. Furthermore, the lateral growth can be enhanced by accelerated chromium diffusion to the oxidation front via the increased grain boundaries in the finer grain coating. In another words, the fine grain of the Y2O3-dispersed chromized coating enhanced the diffusion of chromium to the oxidation front and consequently accelerated the formation of a continuous chromia layer in a shorter time [13–15]. After the formation of the continuous chromia scale, the finer grain scale formed on
NiO A
Fig. 9. Cross-sectional SEM micrograph of chromizing coatings on (a) Ni film and (b) Ni–Y2O3 composite coating at 900 1C for 20 h.
ARTICLE IN PRESS Y.B. Zhou et al. / Vacuum 82 (2008) 748–753
the Y2O3-dispersed chromium coating (Fig. 7) should grow even faster than that formed on the Y2O3-free chromium coating due to an increase in the number of grain boundaries per unit volume. However, the fact that a significant scaling rate reduction occurs suggests that grain boundary diffusion of Cr cations is hindered to a great extent. The reason may be that the chromia scale is incorporated by Y ions released from the dispersed Y2O3 particles when they are incorporated into the growing scale [18–20]. Once being incorporated into the growing scale, they would potentially improve the oxidation resistance through blocking the dominant grain boundary diffusion of chromium cations [2–4], which could be also seen from the cross-section of the scales, as discussed below. By comparison of the cross-section of the scales formed on the chromizing coatings with and without Y2O3 in Fig. 9, the Y2O3-free chromizing coating formed a continuous chromia layer on which Ni-rich oxide was seen. Cavities below the scales formed on the Y2O3-free chromizing coating. They form as a result of condensation of chromium cation vacancies at the interface, which are injected from the scale counter balancing the outward chromium diffusion during oxidation, or from the coating substrate resulting from the relative diffusion rate of Ni to Cr into the coating interior due to the ‘‘Kirkendall’’ effect. The growth of the interfacial cavities is the reason for scale spallation. However, the chromia scale formed on Ni– Y2O3 coated sample is much adherent, no interfacial cavities was observed. The reason lies in two aspects: (1) a low consumption rate of chromium by oxidation due to the formation of purer and denser chromia scale on the Y2O3-dispersed chromizing coating reduces the flux of the cation vacancies toward the interface and (2) the dispersed Y2O3 itself may act as sites for vacancy’s sinking at the interface [3]. Hence, the formation and extension of the interfacial cavity is effectively suppressed in the Y2O3 presence. From the above analysis, the lower oxidation rate of Y2O3-dispersed chromizing coating was associated with the effect of Y2O3, which can be summarized as follows. First, Y2O3 co-deposited in the electrodeposited Ni matrix led to the formation of a fine grain chromium coating, which enhanced the diffusion of chromium to the oxidation front and consequently promoted the formation of chromia layer in a shorter time. Second, the Y2O3 dispersion may release Y ions, which enters the growing chromia scale [18–20] and segregates to the scale grain boundaries. In this case, the dominant diffusion of chromium cations during the growth of chromia scale is blocked, leading to a decrease of oxidation. The segregation of RE ions at the grain boundaries of the thermally grown oxide scales has been extensively evidenced through TEM characterization [18–21]. In this case, the formation of a fine-grain chromia scale on the Y2O3-dispersed chromium coating, suggesting that the segregation of Y ions along the chromia grain boundaries might occur. Segregated Y ions could block the oxide grain growth through a grain boundary pinning [22] and ‘‘solute-drag’’ effect [19].
753
5. Conclusions By pack cementation on the as-electrodeposited Ni– Y2O3 composite at 800 1C, a Y2O3-dispersed chromizing coatings were manufactured. The chromizing coatings formed on the Ni–Y2O3 composite coating were finer grained than that on the pure Ni films due to the added Y2O3 particles retarding the grain growth during chromizing. The oxidation at 900 1C for 20 h showed that the Y2O3modified chromizing coating exhibited a superior oxidation resistance due to the formation of purer and denser chromia scale. The positive effects of Y2O3 on the oxidation of the chromizing coating are summarized as follows: (1) to retard the grain growth of the chromizing coating by anchoring the movement of grain boundaries, which enhance Cr diffusion to the oxidation front and consequently accelerated the formation of a continuous purer chromia scale and (2) Y ions from the Y2O3dispersed coating incorporating into the growing scale and segregating to the chromia grain boundaries block the dominant outward diffusion of chromium cations during the oxidation. Acknowledgment Project (06-13) was supported by the Scientific Research Startup Foundation of Heilongjiang Institute of Science and Technology. References [1] Pfeil LB. Improvement in heat-resisting alloys. UK Patent No. 459848 (1937). [2] Moon DP. Mater Sci Technol 1989;5:754–63. [3] Hou PY, Stringer J. Mater Sci Eng A 1995;202:1–10. [4] Hou PY, Stringer J. Oxid Met 1988;29:45–73. [5] Czerwinski F, Smeltzer WW. Oxid Met 1995;43:25. [6] Koo CH, Yu TH. Surf Coat Technol 2000;126:171. [7] Lee JW, Duh JG, Tsai SY. Surf Coat Technol 2002;153:59–66. [8] Wang ZB, Tao NR, Tong WP, Lu J, Lu K. Acta Mater 2003;51: 4319–29. [9] Wang ZB, Lu J, Lu K. Acta Mater 2005;53:2081–9. [10] Wang ZB, Lu J, Lu K. Surf Coat Technol 2006;201:2796–801. [11] Zhan ZL, He YD, Wang DR, Gao W. Trans Nonferrous Met Soc China 2006;16:647–53. [12] Zhan ZL, He YD, Wang DR, Gao W. Oxid Met 2007;68(5–6): 243–51. [13] Zhang H, Peng X, Zhao J, Wang F. Electrochem Solid-State Lett 2007;10(3):C12–5. [14] Zhu L, Peng X, Yan J, Wang F. Oxid Met 2004;62:411–26. [15] Yan J, Peng X, Wang F. Mater Sci Eng A 2006;426:266–73. [16] Ramanarayanan TA, Ayer R, Petkovic-Luton R, Leta DP. Oxid Met 1988;29:445–72. [17] Stringer J, Wilcox BA, Jaffee RI. Oxid Met 1972;5:11–47. [18] Yurek GJ, Przybylski K, Garratt-Reed AJ. J Electrochem Soc 1987;134:2643–4. [19] Cotell CM, Yurek GJ, Hussey RJ, Mitchell DF, Graham MJ. Oxid Met 1990;34:173–200. [20] Pint BA. Oxid Met 1996;45:1–37. [21] Cotell CM, Yurek GJ, Hussey RJ, Mitchell DF, Graham MJ. J Electrochem Soc 1987;134:1871–972. [22] Hindam HM, Whittle DP. J Electrochem Soc 1982;129:1147–9.