The Stability of Plasma-Sprayed MnCo₂O₄ Coatings at Elevated Temperature for Protective Application of Solid Oxide Fuel Cells

As one of the most important components of SOFC stack, interconnector provides current conduction, thermal conduction and mechanic support to cells. Replacing ceramic interconnector, such as LaCrO₃, by metallic interconnector allows lower operating temperature of SOFC. Due to the relatively low cost, good workability and excellent resistance to thermal oxidation, ferritic stainless steel is considered as the most promising candidates of interconnector materials. However, the use of ferritic stainless steel also brings problems, the existence of Chromium oxide (Cr₂O₃) causing higher ohmic resistance [1] and the evaporation of Chromium to cathode leading to cell performance degradation [2-4]. Spinel has been intensively studied as promising protective coating [5-11]. However, currently most of the methods to prepare Mn-Co spinel coating either costs too high, or deposits too slow, or is of low scalability. Plasma spray, as a surface modification technology, has been widely used to fabricate coatings on surface to endure production new properties [12,13]. However, for maintaining the performance of SOFC stack, the protective spinel coating shall be thin, dense and stable. Hu et al. [14] successfully applied plasma spray to prepare Mn-Co spinel coating [14]. The measured ASR of coated metallic interconnector is as low as 13mWcm² after sintered at 800 °C for 200h, whilst the spinel coating is 100µm thick. Moreover, the spinal phase is suffering the poor stability when exposed to hot plasma. Therefore, this work is proposed to study the influence of thermal condition during the preparation on spinal coatings and clarify the effect of heating treatment on the microstructure of spinal coatings. In this work, commercial spinal powder is used as raw material to prepare spinal coatings by plasma spray process. MnCo₂O₄ spinel coatings are obtained with controlled spray conditions. The conductivity of obtained MnCo₂O₄ spinel coatings are measured. The stability and microstructure of spinel coatings are characterized. The effect of heating treatment on spinel coatings is estimated.

Coating preparation

Three sets of spinal coatings are obtained by atmospheric plasma spray. The setup of the spray system is described elsewhere [15]. The raw material used is spherical Mn_1.5Co_1.5O₄ powder with average diameter of 28.5µm. The spinal powder is fed at a rate of 20g/min. Ceramic sheets with dimension of f1.85cmx2mm are employed as substrate to allow conductivity measurement. More details are available in (Table 1).

Table 1: Parameters to prepare coatings and the thickness of obtained coatings.

Conductivity measurement

The four-point method is employed to measure the conductivity of obtained coatings. The testing sample is placed in a tubular furnace with f50mm orifice at two ends. All measurements are performed at 700 °C with duration up to 15 hours. Constant current of 0.2A is applied, the corresponding voltage is recorded for conductivity calculation according to a method reported in the literature [16].

Characterizations

30mg of raw powder and obtained samples are placed in Al₂O₃ crucible for DSC test. The test was performed from room temperature to 800 °C with heating rate of 5 °C/min. The test chamber was protected by Argon (99.99%). Phase composition and micromorphology of raw powder and coatings are characterized by XRD, SEM and TEM.

Confirmed by XRD, the raw powder consisted of MnCo₂O₄ (PDF#23-1237) and Mn_1.5Co_1.5O₄ (PDF#18-0409) as shown in Figure 1. The phase of Mn_1.5Co_1.5O₄ is considered as a mixture of Mn₃O₄ and Co₃O₄, both of which belong to F3dm space group. Since both Co₃O₄ and Mn₃O₄ are unstable at elevated temperature, the as-sprayed samples are found with CoO (PDF#74-2392) and MnCo₂O₄ [17]. The existence of thermal stress results a shift of the diffraction peaks toward the lower 2q angles. The shift of sample No.1-as and No.2-as is about 1 degree, while for sample No.3-as the shift angle is about 0.8 degree. After tested at 700 °C for 15 hours, spinal phase and Al₂O₃ (PDF#42-1468) from the substrate are detected in the coatings for all samples (Figure 1). A week peak of CoO is also found after the conductivity test, especially the peak at 61.8 °. By absorbing oxygen, the majority of CoO turned to spinel phase after the conductivity test. The diffraction peaks of MnCo₂O₄ are wider and weaker in sample No.2 than that of others (“tt” means “tested”). It means a poor crystallization of MnCo₂O₄ phase, and thus less MnCo₂O₄ phase is presented. The morphology of as-sprayed samples and tested samples are presented in (Figure 2) as well as the TEM diffraction pattern and HRTEM of sample No. 3. The as-sprayed coatings are found composed of cracks, gaps and pores, which are due to the poor melt of raw particles. Due to the relatively low temperature or the shortness of the plasma plume, it’s insufficient to completely melt all raw particles and hence partially melted particles and fully melted particles coexist in the obtained coatings. However, by controlling the input plasma power, no obvious difference was observed between coatings. No visible change of microstructure presents by increasing the flowrate of plasma gas. However, by conducting the conductivity test at 700 °C, coatings are significantly densified as shown in Figure 2. Checked by TEM diffraction, the densified coatings were MnCo₂O₄ (spinel) with an interplanar spacing of 0.288nm.

Figure 1: XRD spectrum of raw powder, as-sprayed samples and tested samples.

Figure 2: Micromorphology and TEM diffraction pattern of samples.

As shown in Figure 3(a), from room temperature to 800 °C, the thermal behavior of as-sprayed coatings is characterized. Exothermic peak is observed for all samples between 630 °C-720 °C that means the selection of testing temperature of conductivity is reasonable. Moreover, the maximal specific thermal flux of sample No.1 is 0.95mW/mg and for No. 2 it is 1.8mW/mg, while for sample No. 3 it is only 0.8mW/mg. Obviously, sample No. 5 presents the most excellent stability and sample No. 2 is of the poorest stability. As presented in Figure 3(b), sample No.1 and No.3 are found with a low mass augmentation about 3% at 800 °C while sample No.2 exhibits a higher mass augmentation of 5%. It means more oxygen is absorbed by sample No.2 to compensate the oxygen loss during the preparation. It also means the coating of sample No.2 experiences serious deoxidization. According to the phase diagram of Mn₃O₄-Co₃O₄, (Mn,Co)O is the main product of deoxidization of (Mn, Co)₃O₄ spinal. Listed in (Table 2) is the conductivity (r) of coatings and the coefficient (c) to calculate the conductivity obtained according to method reported in literature [16]. The conductivity is obtained as r=cI/U. Sample No. 3 shows the highest conductivity of 39.40S/cm, sample No. 2 exhibits the lowest conductivity of 4.00S/cm and the conductivity of sample No.1 is 7.99S/cm. As the measurement is conducted at 700 °C for 15 hours, densification happens that the densified coatings are found free of cracks, gaps and with reduced number of pores, thus the resistance from defects is highly diminished. It is well known that the conductivity of conductive spinel is in form of electron hopping. With less cracks/gaps/interlayers, the interfacial barrier is thus dramatically lowered that more electrons can hop between adjacent sites to provide a high conductivity.

Table 2: Conductivity of coatings obtained at 700 °C in air.

Figure 3a: The plots of specific thermal flux and (b) mass augmentation of tested coatings.

After conductivity measurement, the microstructure of tested samples is checked by TEM-SEM (Figure 4). Average grain size is calculated from XRD diffraction data according to Scherrer’s method. The grains size of tested samples is 120nm, 83nm and 227nm for sample No.1, No.2 and No.3 respectively. As shown in Figure 4, the grain of sample No.2-tt is obviously smaller than that of sample No.3-tt. More important, the grain boundaries in of sample No.3-tt are faded out that grains grow together forming sub-grains. The difference in grain size is coincided with the change of conductivity. It is well accepted that the bigger the grain is, the less the boundary presents. With the biggest grain size, sample No. 3 exhibited the least grain boundary and thus the lowest interfacial resistance for hopping conduction. Sample No. 2 shows the smallest grain size, and thus the most interfacial barrier. Moreover, the phase composition also contributed to the conductivity. Sample No.2 is found with the least MnCo₂O₄ phase while the diffraction peaks of MnCo₂O₄ of sample No.3 is sharp and intensive the conductivity. It is coincided with the measured conductivity as shown in Table 2.

Figure 4: Microstructure of sliced samples prepared by FIB of tested sample No.2 and No.3.

In conclusion, MnCo₂O₄ coatings are successfully prepared by atmospheric plasma spray with controlled condition. The as-sprayed coatings consist of cracks, gaps and pores. The composition is of the as-sprayed coatings is identified as CoO and a little MnCo₂O₄. The presence of CoO is considered as the result of deoxidization of spinal phase in the raw material. By applying thermal stage at 700 °C for 15 hours, a densification effect is found over these coatings, that cracks, gaps and pores are strongly diminished leaving a dense coating. The densified coating is found composed of MnCo₂O₄ and a small amount of CoO. Sample No. 3 exhibits the best stability with specific thermal flux of 0.8mW/mg and thus the best conductivity of 39.40S/cm while sample No. 2 exhibits the highest specific thermal flux of 1.8mW/mg and thus the lowest conductivity of 4S/cm. Densification, grain size and phase composition are considered contributed to the conductivity.

This work is supported by the following projects: National Key R&D Program of China (2018YFB1502600), Platform construction project of modern material surface engineering technology innovation for industrial applications (2018GDASCX-0111), technology cultivation and innovation project of modern material surface engineering (2019GDASYL-0402004), Guangdong Science and Technology Plan Project (2017A070701027) and Guangdong Special Support Program (2019BT02C629).

1. Brandner M, Bram M, Froitzheim J, Buchkremer HP, Stöver D (2008) Electrically conductive diffusion barrier layers for metal-supported SOFC. Solid State Ion 179: 1501-1504.
2. Yang Z, Guo M, Wang N, Ma C, Wang J, et al. (2017) A short review of cathode poisoning and corrosion in solid oxide fuel cell. International Journal of Hydrogen Energy 42(39): 24948-24959.
3. Menzler NH, Sebold D, Wessel E (2014) Interaction of La₅₈Sr_0.40Co_0.20Fe_0.80O₃−δ cathode with volatile Cr in a stack test- scanning electron microscopy and transmission electron microscopy investigations. Journal of Power Sources 254: 148-152.
4. Yang M, Bucher E, Sitte W (2011) Effects of chromium poisoning on the long-term oxygen exchange kinetics of the solid oxide fuel cell cathode materials La₆Sr_0.4CoO₃ and Nd₂NiO₄. Journal of Power Sources 196: 7313-7317.
5. Liu H, Zhu X, Cheng M, Cong Y, Yang W (2011) Novel Mn₅Co_1.5O₄ spinel cathodes for intermediate temperature solid oxide fuel cells. Chemical Communication Cambridge England 47(8): 2378-2380.
6. Yang Z, Xia G, Stevenson JW (2005) Mn₅Co_1.5 O₄ spinel protection layers on ferritic stainless steels for SOFC interconnect applications. Electrochemical Solid-State Letter 8(3): A168-A170.
7. Acharyya SS, Ghosh S, Adak S, Sasaki T, Bal R (2014) Facile synthesis of CuCr₂O₄ spinel nanoparticles: A recyclable heterogeneous catalyst for the one pot hydroxylation of benzene. Catalysis Science Technology 4: 4232-4241.
8. Molin S, Jasinski P, Mikkelsen L, Zhang W, Chen M, et al. (2016) Low temperature processed MnCo₂O₄ and MnCo₈Fe_0.2O₄ as effective protective coatings for solid oxide fuel cell interconnects at 750 °C. Journal of Power Sources 336: 408-418.
9. Kruk A, Stygar M, Brylewski T (2013) Mn-Co spinel protective-conductive coating on AL₄₅₃ ferritic stainless steel for IT-SOFC interconnect applications. Journal of Solid State Electrochemistry 17: 993-1003.
10. Hashemi ST, Dayaghi AM, Askari M, Gannon PE (2018) Sol-gel synthesis of Mn₅Co_1.5O₄ spinel nano powder for coating applications. Materials Research Bulletin 102: 180-185.
11. Mardare CC, Spiegel M, Savan A, Ludwig A (2009) Thermally oxidized Mn-Co thin films as protective coatings for SOFC interconnects. Journal of Electrochemistry Society 156: B1431-B1439.
12. Hui R, Wang Z, Kesler O, Rose L, Jankovic J, et al. (2007) Thermal plasma spraying for SOFCs: Applications, potential advantages, and challenges. J Power Sources 170(2): 308-323.
13. McPherson R (1989) A review of microstructure and properties of plasma sprayed ceramic coatings. Surf Coat Technol 39-40: 173-181.
14. Hu YZ, Yao SW, Li CX, Li CJ, Zhang SL (2017) Influence of pre-reduction on microstructure homogeneity and electrical properties of aps Mn₅Co_1.5O₄ coatings for SOFC interconnects. International Journal of Hydrogen Energy 42(44): 27241-27253.
15. Liu T, Arnold J (2016) Study of in-flight particle stream and particle behavior for understanding the instability phenomenon in plasma spraying process. Surf Coat Technol 286: 80-94.
16. Miccoli I, Edler F, Pfnür H, Tegenkamp C (2015) The 100^th anniversary of the four-point probe technique: The role of probe geometries in isotropic and anisotropic systems. J Phys Condens Matter 27(22): 223201.
17. Aukrust E, Muan A (2006) Phase relations in the system cobalt oxide-manganese oxide in air. J Am Ceram Soc 46(10): 511.