Controlling Void Contents in the Zn Interlayer for Improving Adhesion Strength of ZnMg/Zn Bilayer Coatings on TRIP Steel

Abstract

For ZnMg production feasibility, the adhesive strength of ZnMg alloy coatings must be raised to the level of commercial EG or GI steels. In this study, ZnMg/Zn bilayer coatings with various void contents in the Zn interlayer were synthesized using electromagnetic heating deposition, and the adhesion strength of ZnMg/Zn bilayer coatings on TRIP steels was investigated. As the input current during deposition decreased, the density of the Zn interlayer decreased from 83.2% to 93.2%, and the preferred orientation of the Zn interlayer changed from (101) to (002). During the ZnMg deposition on top of the Zn interlayer, a similar preferred orientation was observed in the ZnMg/Zn bilayer coating. A lap shear test result showed that the adhesion strength of ZnMg/Zn bilayer coatings increased from 19.1 MPa to 21.7 MPa as the (002) texture became dominant with decreasing void contents in the Zn interlayer. These adhesion strength results above 19 MPa were higher than those of the Zn coatings in the commercial EG and GI steels, suggesting that an additional improvement in the adhesion strength of the Zn–Mg/Zn bilayer coatings was possible by controlling the void contents in the Zn interlayer.

1. Introduction

Ever since the ZnMg alloy coatings were reported to show superior corrosion resistance compared to pure Zn and ZnAl alloy coatings due to the formation of simonkolleite and the microstructural transition from crystalline to amorphous,^{1,2,3,4,5,6,7)} many works have been published on the synthesis of ZnMg coatings using various physical vapor deposition (PVD) processes^8,9,10) and much attention has been paid to the adhesion performance of PVD ZnMg coatings to steel substrate in terms of the Mg composition or annealing effect.^10,11,12) A systematic investigation on the effect of the Mg content (0 to 15 wt.%) on the adhesion strength of ZnMg alloy coatings on steel was provided by Jung et al. using a lap shear test.¹¹⁾ They showed that the adhesion strength of ZnMg alloy coatings on steel decreased as the Mg content in the coatings increased. Another report showed that the segregation of magnesium at the steel-zinc interface during the annealing played an important role in the crash adhesion performance of the coatings, suggesting the Zn interlayer coating is necessary to improve the time to failure of the ZnMg/Zn bilayer coatings.¹³⁾ In a Zn–Mg–Zn multi-layer coating by depositing Zn, Mg, and Zn layers sequentially on the surface of a steel sheet using a thermal evaporation process, Byun et al. showed via a 0T-bending test, that annealing treatment at 200°C increased the volume fraction of the brittle intermetallic compounds such as Mg₂Zn₁₁ and MgZn₂ phases, resulting in the poor adhesion behavior of the coatings.¹⁴⁾ More recently, several investigators have reported the effect of a Zn interlayer between the substrate and the ZnMg coating on the adhesion strength of the ZnMg/Zn bilayer coating.^{4,13,15,16,17,18)} Song et al.¹⁵⁾ and Sabooni et al.¹⁷⁾ both reported that a Zn interlayer is an essential part for the improvement of the adhesion strength of the Zn–Mg coatings, which confirmed the previous report by Jung et al.¹¹⁾ From the lap shear test results, the Zn interlayer significantly contributed to the improvement of the formability and adhesion of the Zn–Mg coating, showing that the maximum adhesion strength of the Zn–Mg coatings with the Zn interlayer became approximately twice as high as that of the Zn–Mg coatings without the Zn interlayer.¹⁵⁾ A modified Benjamin-Weaver model was suggested to quantify the adhesion strength of PVD ZnMg/Zn bilayered coatings, and it was reported that the annealing treatment of the double-layer ZnMg/Zn coating was detrimental for the adhesion strength as the thickness of the Zn interlayer decreased below the critical value of 500 nm due to diffusion during the annealing.¹⁷⁾ A diffusion-driven microstructural modification in the double-layer ZnMg/Zn coating using a controlled annealing process was reported to improve the adhesion strength and corrosion resistance in the range of 26% and 42%, respectively.¹⁸⁾ Additionally, Sabooni et al. reported that the adhesion strength of the ZnMg/Zn bilayer coatings is affected by not only the interfacial adhesion strength but also the thickness of the interlayers as well as the process-related interfacial defect density.⁸⁾ Randomly generated pores at Zn/Fe and Zn/Zn–Mg interfaces in the double-layer Zn–Mg/Zn coating deteriorated the adhesion performance of the coating, but the detailed mechanisms for the void formation in the coating are still under investigation.⁸⁾

There was, however, a report on the effect of the void content in the Zn interlayer on the adhesion strength of the ZnMg/Zn double layer coatings synthesized using the magnetron sputtering process by Lee et al.¹⁶⁾ Based on the structural zone model¹⁹⁾ they were able to synthesize various void contents in the Zn interlayer by controlling sputtering parameters such as process temperature and working pressure, and they showed that the adhesion strength of ZnMg/Zn bilayer coatings increased from 20.4 MPa to 24.58 MPa as the void contents in the Zn interlayer decreased.

In this work, an approach to providing an additional adhesion strength to the ZnMg/Zn bilayer coating synthesized by the electromagnetic heating deposition (EMHD) process was attempted by controlling the void content in the Zn interlayer. During the EMHD process, the void contents in the Zn interlayer were controlled by varying the input current, which resulted in a change in the evaporation rate for the Zn interlayer coating. The adhesion behaviors of the Zn–Mg/Zn coating with various void contents in the Zn interlayer were investigated quantitatively.

2. Experimental Details

ZnMg/Zn bilayer coatings were deposited on TRIP steels (1.2 mm thick, POSCO). The tensile strength of TRIP steels was 1180 MPa, and the composition of major alloying elements is listed in Table 1. The coatings were prepared by the EMHD process, which has a much faster deposition rate (approximately 3–30 times) than the magnetron sputtering process.^20,21) A schematic of the EMHD process is shown in Fig. 1. The ball-type Zn and Mg droplets (Ø=16 mm, 99.9% purity) were dropped through the feeding tube into boron-nitride crucible. The targets were evaporated using a 4-turn induction coil with a high-frequency power of 3.0 kW over 80 kHz, and the vapor evaporated from the crucible transited through the vapor distribution box at a temperature of 800°C. Before the deposition, the base pressure of the EMHD chamber was pumped down to less than 1.3×10⁻² Pa, and the surface contaminants and oxides of the steel substrates were removed by ion etcher. To control the void contents in the Zn interlayer coating, the input current on the induction coil was controlled from 20 A to 40 A.

Table 1. Chemical composition of TRIP steel (wt.%).

	C	Mn	Si	Al Fe
TRIP steel	0.28	2.2	1.7	0.6 bal.

Fig. 1. Schematic diagram of the EMHD chamber.

The microstructures, composition, and crystallography of the coatings were analyzed using a field emission scanning electron microscope (FE-SEM, JEOL JSM-7100F), energy-dispersive X-ray spectroscopy (EDS), and an X-ray diffractometer (XRD, Rigaku MiniFlex II) with Cu Kα irradiation (λ = 0.15456 nm), respectively. The cross section of the Zn interlayers was fabricated using a focused ion beam (FIB, JEOL JIB-4601F), and the cross-sectional images were converted to blue or black pixels for calculating the Zn density. Then, the counts of porosity in the coatings were measured by the corresponding pixels. The lap shear test was carried out to determine the adhesion strength of the coating quantitatively. The dimensions of the specimens for the lap shear test were 100 mm × 25 mm × 1.2 mm. The structural epoxy-type adhesive (25 MPa shear strength, BOKWANG Co., N.F.W Hemming Sealer) was used to bond the two ZnMg/Zn coated steels with a 25 mm × 25 mm overlap area. The adhesive was cured in the oven at 150°C for 1 hour. The single-lap joints were tested in a tensile testing machine (Shimadzu, AG-10TA) using a displacement rate of 5 mm/min. For the reliability of the results, the lap shear test was carried out six times, and the failure plane of the specimen was analyzed to determine the location of the failure.

3. Results and Discussion

The ZnMg/Zn bilayer coatings were synthesized with various input currents for Zn interlayers, and their detailed variables for synthesizing the coating are summarized in Table 2. The cross-sectional SEM images of the ZnMg/Zn bilayer coatings are shown in Fig. 2. The thickness of the coatings was controlled to be about 1 μm of Zn interlayer and 2 μm of ZnMg top layer, and all the coatings showed a columnar structure. By controlling the input current for deposition of Zn interlayer, the partial pressure of Zn vapor (P_zn) and deposition rate were significantly changed as shown in Table 2, and as a result, the microstructure of the Zn interlayer was influenced accordingly. The density of the Zn interlayer was measured by analyzing the cross section of the coating fabricated using a focused ion beam for quantitative analysis, as shown in Figs. 2 and 3. The density of the Zn interlayer deposited with a 20 A input current was measured to be 93.2%, and the value gradually decreased to 83.2% as the input current increased. Zn atoms vaporized with a high input current might not possess sufficient energy to lead to a dense coalescence of the coating on the substrate. This phenomenon could be attributed to the fact that Zn atoms partially lose their kinetic energy by colliding with other Zn atoms in vapor gas while they traverse the vacuum chamber from the source to the substrate.^22,23) As a result, the formation of void structures at high input current, i.e., high partial pressure (see Table 2), could be attributed to the reduction in the Zn adatom surface mobility relative to the low input current.

Table 2. Deposition condition and chemical composition of the coatings.

	Input current (A)	Partial pressure (Pa)	Deposition rate (μm/min)	Composition (wt%)
Zn interlayer	20	0.6 × 10−1	1	Zn100
	30	0.47	6	Zn100
	40	0.93	12	Zn100
ZnMg layer	30	0.9 × 10−1	1.2	Zn90Mg10

Fig. 2. The cross-sectional FE-SEM images (left side) and ion-milled images (right side) of the ZnMg/Zn coatings synthesized Zn interlayer with various input currents: (a) 20A, (b) 30A, (c) 40A.

Fig. 3. Density of Zn interlayer with various input current: (a) 20A, (b) 30A, (c) 40A (black area represents void).

The X-Ray diffraction (XRD) patterns of the Zn interlayer and ZnMg/Zn bilayer are shown in Fig. 4. The XRD pattern of the Zn interlayer in Fig. 4(a) showed the crystalline Zn (002), (100), (101), and (102) peaks. The Zn interlayers with a 20 A input current showed a strong (002) preferred orientation, which changed to (101) as the input current increased. The change in the preferred orientation could be explained by the faceting during the coating growth. In the growth model of coating, the Zn species evaporated from the source diffuse and migrate on the coating surface, and the Zn atoms tend to be adsorbed on stable sites that lower the surface energy of the coating. However, if the deposition rate is very high, the Zn atoms would not have enough time and energy to diffuse on the surface of the coating and would get adsorbed on unstable sites. As a result, the planes with the lowest surface energy survive while the other planes with a higher energy level extinguish themselves in order to minimize the total surface energy of the coatings during the deposition. In an HCP structured Zn, the (002) basal plane has the lowest surface energy compared to other crystallographic planes, and this basal plane could survive most under equilibrium growth conditions.^16,24) During the EMHD process using 20 A input current, the deposition rate became relatively low so that there was enough surface diffusion and migration of the evaporated Zn particles, resulting in the formation of the (002) faceted hexagonal basal plane easily. In contrast, at the high input current, the deposition rate would become too high for the evaporated particle to adequately move around to be absorbed on the unstable site, and the preferred orientation changed to (101). Comparing with the results from the Zn interlayer, the XRD patterns from the ZnMg/Zn bilayer coating in Fig. 4(b) showed similarity to those from the Zn interlayer. No intermetallic compound in the ZnMg/Zn coatings was observed due to the insufficient energy of the atoms to form the intermetallic phases.^2,25) The preferred orientation in the ZnMg/Zn coatings was similar to that in the Zn interlayer.

Fig. 4. The X-ray diffraction patterns of (a) the Zn interlayer and (b) Zn/Zn–Mg coatings synthesized with various input current.

The crystalline dimension for each preferred orientation in the coating was further investigated to understand the relationship between the density of the coating and the preferred orientation of Zn interlayer using Scherrer equation and the results are summarized in Fig. 5.

  

$$D_{hkl}=\frac{K\times \lambda }{\beta \times cos\theta }$$(1)

Fig. 5. Crystallite dimension (D_hkl) of Zn interlayer with the various input current.

In the Scherrer equation in Eq. (1), D_hkl is the mean dimension of the ordered crystalline, and K is the shape constant, usually taken as 0.9, and λ is the wavelength of irradiation (Cu Kα = 0.15456 nm), and θ is the diffracted angle of the peak, and β is the full width at half-maximum of the diffraction peak. As input current increased from 20 A to 40 A the crystalline dimension of (002) plane in the Zn interlayer gradually decreased from D₍₀₀₂₎=41.3 to 36.6 nm and (100) plane from D₍₁₀₀₎=38.2 to 36.4 nm. On the contrary, the crystalline dimension of (101) plane and (102) plane increased from D₍₁₀₁₎=38.8 to 43.2 nm and D₍₁₀₂₎=36.8 to 38.7 nm, respectively, as input current increased. Thus, it could be inferred that the change in the dimension of crystal orientation could be correlated with the adatom surface mobility during the deposition.

The lap shear test was performed to quantitatively investigate the adhesion strength of the ZnMg/Zn bilayer coatings, and the results are shown in Fig. 6. The ZnMg/Zn coating with 20 A, which has the highest coating density of Zn interlayer (93.2%), showed the highest adhesion strength value of 21.7 MPa, and the adhesion strength gradually decreased to 19.1 MPa as the input current increased up to 40 A. Comparing the adhesion strengths of the Zn₉₀Mg₁₀ coated steel and commercial GI and EG steel, ZnMg/Zn bilayer coating with controlled Zn interlayer showed about 180% and 15–35% additional improvement in the adhesion strength, respectively.^11,25,26) The failure planes of the ZnMg/Zn bilayer coating were investigated in terms of the failure mode after the lap shear test.^27,28,29) The schematic diagram of two failure modes represents the delamination of coating from the steel plate (Mode 1) and fracture between the adhesive (Mode 2), as shown in Fig. 7. The failure plane of ZnMg/Zn coating with 20 A exhibits a relatively small delamination area of coating, as evidenced by the existence of a dominant yellow area (Mode 1). However, as the input current increased, the delamination area of the coating increased relative to the low input current. A relatively low degree of adatom surface mobility produces initial nuclei that grow in the direction of the atomic arrival flux.¹⁹⁾ Then, at the Zn growth mode in the interlayer, the columnar structure seems to develop under the influence of intergranular shading, forming voids. Thus, the ZnMg/Zn bilayer coating containing abundant voids with high input current has poor adhesion, as can be observed in Figs. 6 and 7, wherein the coating is delaminated from the TRIP steel substrate. In addition, the crystallographic orientations of the Zn interlayer exhibited a strong influence on the adhesion strength. Both the hardness and elastic modulus along the c-axis of Zn crystals (basal plane) are one-third of those orthogonal to the direction of the c-axis (prismatic plane), which is attributed to the intrinsic properties of Zn.³⁰⁾ The dominant (002) plane of Zn interlayers with a low input current of 20 A might be beneficial to the mitigation of coating delamination because the shear stress levels of the basal planes would be much lower than those of prismatic planes under the same tensile strain. Therefore, by controlling the structure of the Zn interlayer to have a dominant basal plane and less void formation, the adhesion strength of the ZnMg coating with a high Mg content (10 wt.%) could be significantly improved.

Fig. 6. The void contents in the Zn interlayer and the adhesion strength of the ZnMg/Zn bilayer coatings on TRIP steels measured via lap shear test.

Fig. 7. Schematic cross-sectional view of the failure plane position in the coating and failure plane view of ZnMg/Zn coating via lap shear test (yellow area represents adhesive).

4. Conclusions

In this work, the void content in the Zn interlayer of the ZnMg/Zn coating was controlled to improve the adhesion strength of the ZnMg/Zn bilayer coating using the EMHD process. As the input current for evaporation decreased from 40 A to 20 A, the preferred orientation of the Zn interlayer was changed from (101) to (002) and a similar trend was observed in the ZnMg/Zn bilayer coating. The void content in the Zn interlayer was increased when preferred growth was induced in the direction of a plane, such as the (101) plane or the (102) plane, having a specific angle to the (002) basal plane. With decreasing input current, the density of the Zn interlayer increased from 83.2% to 93.2%, and the adhesion strength of ZnMg/Zn coating to TRIP steel increased up to 21.7 MPa, which is much higher than those of ZnMg coated steel without an interlayer and commercial EG and GI steel. As a result of this work, by controlling the variable input current, the density of the Zn interlayer could be improved by about 12%, and furthermore, the adhesion strength of the Zn/ZnMg coating could be improved by about 24%. It was demonstrated quantitatively that, in addition to improvement in the adhesion strength by the insertion of an Zn interlayer, a further improvement in the adhesion strength of the ZnMg/Zn bilayer coatings was possible by controlling the void content in the Zn interlayer.

Acknowledgement

This study has been supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A2C1010058) and by POSCO.

Conflicts of Interest

The authors have no competing interest in financial interest/personal relationships.

References

• 1)   T.  Prosek,  D.  Persson,  J.  Stoulil and  D.  Thierry: Corros. Sci., 86 (2014), 231. https://doi.org/10.1016/j.corsci.2014.05.016
• 2)   J.  La,  K.  Bae,  S.  Lee and  K.  Nam: Surf. Coat. Technol., 307 (2016), 1100. https://doi.org/10.1016/j.surfcoat.2016.05.056
• 3)   J.  La,  S.  Lee and  S.  Hong: Surf. Coat. Technol., 259 (2014), 56. https://doi.org/10.1016/j.surfcoat.2014.07.006
• 4)   K.  Bae,  J.  La,  I.  Lee,  S.  Lee and  K.  Nam: Met. Mater. Int., 23 (2017), 481. https://doi.org/10.1007/s12540-017-6478-2
• 5)   O. P.  Femminella,  M. J.  Starink,  M.  Brown,  I.  Sinclair,  C. J.  Harris and  P. A. S.  Reed: ISIJ Int., 39 (1999), 1027. https://doi.org/10.2355/isijinternational.39.1027
• 6)   Y.  Suzuki,  S.  Yamaguchi,  I.  Matsumoto and  I.  Muto: Tetsu-to-Hagané, 105 (2019), 752 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2018-149
• 7)   J.  Tanaka,  K.  Ono,  S.  Hayashi,  K.  Ohsasa and  T.  Narita: ISIJ Int., 42 (2002), 80. https://doi.org/10.2355/isijinternational.42.80
• 8)   S.  Sabooni,  M.  Ahmadi,  E.  Galinmoghaddam,  J.  Westerwaal,  C.  Boelsma,  E.  Zoestbergen,  G. M.  Song and  Y. T.  Pei: Mater. Des., 190 (2020), 108560. https://doi.org/10.1016/j.matdes.2020.108560
• 9)   M. A.  Baker,  W.  Gissler,  S.  Klose,  M.  Trampert and  F.  Weber: Surf. Coat. Technol., 125 (2000), 207. https://doi.org/10.1016/S0257-8972(99)00550-2
• 10)   J. L.  Davies,  C. F.  Glover,  J.  Van de Langkruis,  E.  Zoestbergen and  G.  Williams: Corros. Sci., 100 (2015), 607. https://doi.org/10.1016/j.corsci.2015.08.036
• 11)   W. S.  Jung,  C. W.  Lee,  T. Y.  Kim and  B. C.  De Cooman: Metall. Mater. Trans. A, 47 (2016), 4594. https://doi.org/10.1007/s11661-016-3634-0
• 12)   R.  Bolton,  T.  Dunlop,  J.  Sullivan,  J.  Searle,  H.  Heinrich,  R.  Westerwaal,  C.  Boelsma and  G.  Williams: J. Electrochem. Soc., 166 (2019), C3305. https://doi.org/10.1149/2.0391911jes
• 13)   S.  Sabooni,  E.  Galinmoghaddam,  H. T.  Cao,  R. J.  Westerwaal,  E.  Zoestbergen,  J. T. M.  De Hosson and  Y. T.  Pei: Surf. Coat. Technol., 370 (2019), 35. https://doi.org/10.1016/j.surfcoat.2019.04.085
• 14)   J. M.  Byun,  S. R.  Bang,  H. W.  Kim,  T. Y.  Kim,  S. J.  Hong and  Y. D.  Kim: Surf. Coat. Technol., 309 (2017), 1010. https://doi.org/10.1016/j.surfcoat.2016.10.043
• 15)   M.  Song,  H.  Kim and  S.  Lee: ISIJ Int., 59 (2019), 1113. https://doi.org/10.2355/isijinternational.ISIJINT-2018-476
• 16)   S. H.  Lee,  H. K.  Kim,  M. K.  Song,  J. W.  Kim and  S. Y.  Lee: Korean J. Met. Mater., 57 (2019), 695 (in Korea). https://doi.org/10.3365/Kjmm.2019.57.11.695
• 17)   S.  Sabooni,  E.  Galinmoghaddam,  M.  Ahmadi,  R. J.  Westerwaal,  J.  van de Langkruis,  E.  Zoestbergen,  J. T. M.  De Hosson and  Y. T.  Pei: Surf. Coat. Technol., 359 (2019), 227. https://doi.org/10.1016/j.surfcoat.2018.12.064
• 18)   S.  Sabooni,  E.  Galinmoghaddam,  A. A.  Turkin,  L. Q.  Lu,  R. J.  Westerwaal,  C.  Boelsma,  E.  Zoestbergen and  Y. T.  Pei: J. Alloy. Compd., 836 (2020), 155448. https://doi.org/10.1016/j.jallcom.2020.155448
• 19)   J. A.  Thornton: J. Vac. Sci. Technol., 11 (1974), 666. https://doi.org/10.1116/1.1312732
• 20)   L.  Baptiste,  N.  van Landschoot,  G.  Gleijm,  J.  Priede,  J. S.  van Westrum,  H.  Velthuis and  T. Y.  Kim: Surf. Coat. Technol., 202 (2007), 1189. https://doi.org/10.1016/j.surfcoat.2007.05.075
• 21)   T. Y.  Kim and  M.  Goodenough: Corros. Sci. Technol., 10 (2011), 194. https://doi.org/10.14773/cst.2011.10.6.194
• 22)   V.  Torrisi and  F.  Ruffino: Surf. Coat. Technol., 315 (2017), 123. https://doi.org/10.1016/j.surfcoat.2017.02.034
• 23)   Y. S.  Ma,  J. C.  Hsu,  C. P.  Liu and  H. L.  Chen: Vacuum, 122 (2015), 96. https://doi.org/10.1016/j.vacuum.2015.09.007
• 24)   M.  Zhou,  S.  Guo,  J. L.  Li,  X. B.  Luo,  Z. X.  Liu,  T. S.  Zhang,  X. X.  Cao,  M. Q.  Long,  B. G.  Lu,  A. Q.  Pan,  G. Z.  Fang,  J.  Zhou and  S. Q.  Liang: Adv. Mater., 33 (2021), 2100187. https://doi.org/10.1002/adma.202100187
• 25)   C. Z.  Yao,  W. W.  Chen and  W.  Gao: Surf. Coat. Technol., 219 (2013), 126. https://doi.org/10.1016/j.surfcoat.2013.01.015
• 26)   M.  Pantoja,  J.  Abenojar,  M. A.  Martinez and  F.  Velasco: Int. J. Adhes. Adhes., 65 (2016), 54. https://doi.org/10.1016/j.ijadhadh.2015.11.006
• 27)   Y. F.  Gao,  Z. Y.  Xue and  Z. Z.  Yu: Extreme Mech. Lett., 37 (2020), 100728. https://doi.org/10.1016/j.eml.2020.100728
• 28)   D.  Fernando,  J. G.  Teng,  T.  Yu and  X. L.  Zhao: J. Compos. Constr., 17 (2013), 04013012. https://doi.org/10.1061/(Asce)Cc.1943-5614.0000387
• 29)   J.  Ha,  H.  Huh,  J. H.  Song and  J. H.  Lim: Int. J. Auto. Technol., 14 (2013), 67. https://doi.org/10.1007/s12239-013-0009-z
• 30)   G. M.  Song and  W. G.  Sloof: Mater. Sci. Eng. A, 528 (2011), 6432. https://doi.org/10.1016/j.msea.2011.04.058