Thermal shock behavior and infrared radiation ...

SCT-20152; No of Pages 10 Surface & Coatings Technology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat

Thermal shock behavior and infrared radiation property of integrative insulations consisting of MoSi2/borosilicate glass coating and fibrous ZrO2 ceramic substrate Gaofeng Shao a,b, Xiaodong Wu a,b, Yong Kong a,b, Sheng Cui a,b, Xiaodong Shen a,b,⁎, Chunrong Jiao c, Jian Jiao c a b c

State Key Laboratory of Materials-Oriented Chemical Engineering, College of Materials Science and Engineering, Nanjing Tech University, Nanjing 210009, China Nanjing Tech University Suqian Advanced Materials Institute, Suqian 223800, China National Key Laboratory of Advanced Composite, Avic Composite Corporation Ltd., Beijing 101300, China

a r t i c l e

i n f o

Article history: Received 16 December 2014 Accepted in revised form 5 March 2015 Available online xxxx Keywords: MoSi2/borosilicate glass coating Thermal shock behavior Infrared radiation Fibrous ZrO2 ceramic

a b s t r a c t An integrative insulation consisting of a MoSi2/borosilicate glass coating and a fibrous ZrO2 ceramic substrate was explored for applications up to 1673 K. The MoSi2/borosilicate glass coating was prepared using the slurry dipping and subsequent firing method. The thermal shock behavior of the integrative insulation and the evolution of the surface morphology of the coatings for various thermal shock times were evaluated using confocal laser scanning microscopy (CLSM). The results revealed that the as-prepared coatings could maintain the dense structure and infiltrate into the substrate, forming a gradient structure and exhibiting good compatibility and adherence. After thermal cycling between 1673 K and room temperature 15 times, the integrative insulation is also without micro cracks and spalling, and the weight loss is only 2.84%, revealing outstanding thermal shock performance. Additionally, the emissivity of the coatings reached 0.8 at room temperature, which was attributed to the synergistic effect of inter band absorption, crystal lattice vibration and the roughness of the coating. The gradual increase in the roughness resulted in the increase of the emissivity in 200–2500 nm at room temperature, which was explained with a “circular grooves” model. These results imply that the integrative insulation can be a promising candidate material in high temperature application. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Porous fibrous ZrO2 ceramics are ideally suited for thermal insulation and protection under conditions of ultra-high temperatures and in various severe environments because of their exceptional properties, such as light weightness, extreme high-temperature stability and low thermal conductivity [1,2]. However, during high-speed flight, hypersonic vehicles suffer severe aerodynamic heating due to the drastic friction between the vehicles and the atmosphere; a large amount of energy is being transferred into system from the surface via convection and chemical heating. While radiant heat transfer can play an important role in this critical environment, high-emissivity coatings will radiate a large amount of heat, thus restraining the surface temperature increase of the vehicles [3,4]. Thus, an integrative insulation consisting of highinfrared-emissivity coatings used on the surface and a low-thermalconductivity substrate used inside must be created. In general, the formation of a high-emissivity coating is attained by mixing high⁎ Corresponding author at: State Key Laboratory of Materials-Oriented Chemical Engineering, College of Materials Science and Engineering, Nanjing Tech University, Nanjing 210009, China. Tel.: +86 25 83587234; fax: +86 25 83221690. E-mail address: [email protected] (X. Shen).

emissivity additives and a binder. Several systems of high-emissivity coatings have been developed, such as carbide and carbon systems (SiC [5–10], AlPO4 -C [11]), spinel structure systems (Fe–Mn [12], Ni–Cr [13,14], Al–Si [15]), cordierite structure systems (Mg–Al–Si [16,17]), rare-earth oxide systems (CeO2 [18–21], Sm2O3 [4,22,23]) and perovskite-type systems [24,25]. However, these systems were focused on the range of 873–1273 K. Recently, significant attention has been focused on MoSi2 because of its high melting point (2293 K), low density (6.24 g/cm3), excellent high-temperature oxidation resistance [26,27] and high emissivity [28]. It is widely used for industrial and military applications, such as high-temperature structural materials [29,30], heating elements [31] and high-temperature coatings [32,33]. Among the binders, SiO2 glass does not exhibit attractive viscosity and wetting characteristics as a binder. Although B2O3 glass combines thermal stability with appropriate viscosity and wetting to provide protection over a wide temperature range both as a coating and a sealant, this material volatilizes rapidly at high temperatures [34]. The best compromise should be a thermally stable borosilicate glass with appropriate wetting properties and viscosity [35,36]. To date, limited reports have been published on integrative insulations consisting of MoSi2/borosilicate glass coatings and fibrous ZrO2 ceramics. In the present work, the dipping process with borosilicate

http://dx.doi.org/10.1016/j.surfcoat.2015.03.008 0257-8972/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: G. Shao, et al., Surf. Coat. Technol. (2015), http://dx.doi.org/10.1016/j.surfcoat.2015.03.008

2

G. Shao et al. / Surface & Coatings Technology xxx (2015) xxx–xxx

glass as the binder and MoSi2 powder as the emittance agent is used as an alternative approach to fabricating possible high-emissivity coatings. In addition, borosilicate glass is coupled with good bonding characteristics to a porous fibrous ZrO2 ceramic substrate, with one of the major qualities being the ability to tailor their thermal expansion characteristics. The microstructure, thermal shock behavior and infrared radiation property of the integrative insulation were investigated. 2. Experimental 2.1. Preparation of the borosilicate glass The raw materials, SiO2 and Na2B4O7·10H2O, were thoroughly homogenized using a mixer for 24 h; then, the mixture was placed in a platinum crucible, transferred to an oven at 1773 K and held for 5 h. The molten glass was poured into water immediately to form glass granules. The vitreous frit granules were reduced to particles less than 50 μm in size after ball milling, and then, the glass frit was obtained. 2.2. Preparation of the coatings A 20 mm × 20 mm × 5 mm fibrous ZrO2 ceramic was used as the substrate. The microstructure and basic properties of the fibrous ZrO2 ceramic are shown in Fig. 1 and Table 1, respectively. It can be observed that the fibrous ZrO2 ceramic can build a “birds nest” structure to obtain high porosity, lower thermal conductivity and relatively high strength. High-purity MoSi2, borosilicate glass frit and SiB6 (sintering aids, 2.5% Wt.) were weighted with different compositions: M40, M60, M80 (M = MoSi2, the numerical value denotes the mass percent of MoSi2). All the above powders were mixed by a planetary mill for 24 h, where the powders and ethanol combined with agate balls were sealed in a nylon ball-milling container. Eventually, the particle size reached approximately 1–2 μm. The MoSi2/borosilicate glass coating was prepared on the surface of the treated fibrous ZrO2 ceramic using the low-cost approach of the slurry dipping and subsequent firing method. The coating thickness was controlled by the dipping times. The as-coated samples with three dipping times were then normally inserted into the furnace at 1593 K for 10 min and cooled by rapid removal from the furnace at atmospheric pressure. 2.3. Thermal shock test

Table 1 Basic properties of the fibrous ZrO2 ceramic. Density (g/cm3)

Thermal conductivity (W/m K)

Continuous maximum use temperature (K)

0.47

0.14 (at 1473 K)

1923

two steps: placing the as-coated samples inside the furnace at 1673 K and outside the furnace at room temperature both for 10 min. The surface states of the coatings were observed carefully and recorded every 5 thermal cycles.

2.4. Characterization The emissivities of the coatings were measured using an emissometer (Model AE1); the spectral range of the thermal radiation emitted from the surface ranged from 3 to 30 μm at 353 K. The spectral absorptivity of the samples was measured using an ultraviolet–visiblenear-IR spectrophotometer (UVPC, Varian Cary 5000, VARIAN). The evolution of the surface morphology after the thermal shock tests was examined using a confocal laser scanning microscope (CLSM). The CLSM images were obtained with an Olympus LEXT OLS 4000 microscope powered by a singer laser (λ = 405 nm) in the reflected light mode. The microstructure was surveyed using a scanning electron microscope (SEM, Model JSM-5900, JEOL, Tokyo, Japan). The phase composition of the coating surface was examined using a Rigaku Miniflex X-ray diffractometer (XRD) with Cu-Kα radiation. A Fouriertransform infrared (FT-IR) spectrum was recorded on a BrukerEquinox 55 spectrophotometer in KBr pellets with a scanning range of 4000–400 cm−1.

3. Results and discussion 3.1. Formation of the borosilicate glass Fig. 2 presents XRD patterns of the raw materials: SiO2 and asprepared glass. Compared with SiO2, the spectrum of the borosilicate glass contains a broad bump at 22°, and no crystal diffraction peak is observed in the spectrum, which suggests that the as-prepared glass is amorphous.

Thermal shock resistance tests were performed between 1673 K and room temperature. A complete thermal cycle comprised the following

Fig. 1. Microstructure of the fibrous ZrO2 ceramic.

Fig. 2. XRD patterns of (a) raw material: SiO2 powder and (b) the as-prepared borosilicate glass.



3

Eqs. (1), (2) and (3) [40,41]. Additionally, the crystalline SiO2 is not only formed from the oxidation of SiB6 and MoSi2 during the soaking at 1593 K according to Eqs. (1)–(4) but also from the partial borosilicate glass crystallized during the following cooling process [35].

Fig. 3. FT-IR spectrum of the borosilicate glass.

Fig. 3 shows the FT-IR spectrum of borosilicate glass. The peak at approximately 1402 cm−1 is ascribed to symmetric stretching relaxation of the B\O band of trigonal BO3 units [37]. The broad absorption band observed at approximately 1095 cm−1 can be assigned to tri-, tetra-, pentborate and diborate groups belonging to BO3 and BO4 groups along with asymmetric stretching Si\O\Si bonds. The band at 925 cm−1 should be assigned to the B\O vibration of BO4 units and is also associated with the stretching frequency of Si\O\B [38]. The bond at approximately 795 cm− 1 is assigned to the symmetric stretching vibration of O\Si\O. The band at 669 cm−1 is assigned to the bending vibration of bridging oxygen (B\O) between trigonal BO3 groups. A broad band centered at 469 cm− 1 corresponds to the Si\O\Si bending vibration [39]. 3.2. Microstructure of the coating XRD patterns of the MoSi2/borosilicate glass coatings are presented in Fig. 4. The amorphous nature of glass near 22° can be observed, indicating that the as-prepared coatings include the borosilicate glass phase, especially M40 coating (Fig. 4a). Mo5Si3, MoO2 and Mo are supposed to form from the productions of the slight oxidation of MoSi2 according to

MoSi2 ðsÞ þ 2O2 ðgÞ ¼ MoðsÞ þ 2SiO2 ðsÞ

ð1Þ

MoSi2 ðsÞ þ 3O2 ðgÞ ¼ MoO2 ðsÞ þ 2SiO2 ðsÞ

ð2Þ

5MoSi2 ðsÞ þ 7O2 ðgÞ ¼ Mo5 Si3 ðsÞ þ 7SiO2 ðsÞ

ð3Þ

2SiB6 ðsÞ þ 11O2 ðgÞ ¼ 2SiO2 ðsÞ þ 6B2 O3 ðsÞ

ð4Þ

Fig. 5 shows the CLSM and SEM images of the MoSi2/borosilicate glass coatings. The surface of the M40 coating is dense and without micro cracks; MoSi2 particles are immersed into the borosilicate glass continuous layer sufficiently. However, a few bubbles can be observed on the surface (Fig. 5(g)), which is due to the volatilization of B2O3 and MoO3 at high temperature. The M60 coating is uneven, and the surface gloss gradually disappears (Fig. 5(b)). The outline of the same large MoSi2 grains can be observed, which indicates that the glass layer is not thick enough to seal and connect the MoSi2 grains (Fig. 5(h)). The white glass layer can be observed on the surface of the M80 coating (Fig. 5(c)); combined with the XRD result, the white glass layer is SiO2, which results from the oxidation of the MoSi2-rich phase of the M80 coating. However, some holes are homogenously distributed (Fig. 5(f)), which can be observed from SEM images, and MoSi2 grains are randomly oriented in the holes (Fig. 5(i)). This result occurs because the vitreous SiO2 glass layer has weak fluidity that cannot effectively cover the holes. Fig. 6 presents cross-section micrographs of the integrative insulations. Part of the coating material has infiltrated into the substrate through the open pores of the zirconia fiber insulation tile (Fig. 6(a), (b), (c)), gradually transforming in appearance from a dark surface to a gray and a white fiber interior. The thicknesses of the surface layer are in the range of 50–100 μm (Fig. 6(d), (e), (f)). The SEM image of the cross-section of M40 coated insulation is shown in Fig. 7. The surface coating displays a dense structure with the thickness of around 100 μm (Fig. 7(b)). The magnified image of interfacial transition zone is presented in Fig. 7(c). MoSi2 particles combine well with ZrO2 fibers through the borosilicate glass continuous phase (Fig. 7(d)). Therefore, the obtained coatings have a good combination with the substrate, and no obvious interface is discovered between the coating and the substrate, which is helpful in increasing the compatibility and adherence between the coatings and the substrate. 3.3. Thermal shock resistance of the integrative insulation

Fig. 4. XRD patterns of MoSi2/borosilicate glass coatings. (a): M40; (b): M60; and (c): M80.

The weight loss curves of the integrated insulations during a thermal shock test between 1673 K and room temperature are presented in Fig. 8. The M40-coated insulation exhibits excellent thermal shock resistance. An initial weight gain is observed, with a maximum of 0.72% observed after the first thermal shock cycle. Then, a continuous weight loss occurs, and the cumulated weight loss is 2.84% after 15 cycles. However, with the increase of the MoSi2 content, the thermal shock resistance of the integrated insulations is weakened. The weight loss of the M60 and M80 coated insulations after 15 cycles are 11.32% and 8.62%, respectively. According to the XRD results before thermal shock test (Fig. 4), the Mo phase is rich in the coating; during the thermal shock test, Mo, Mo5Si3 and MoSi2 have begun to oxidize to form the volatilizable products (MoO2, MoO3) according to Eqs. (5)–(9), which leads to the weight loss of the coatings. Fig. 9 shows the XRD pattern of the coatings after 15 thermal shock cycles. A broad bump at 22° suggests that the borosilicate glass phase exists in M40 coating. Compared with the XRD pattern of the coating before test, Mo5Si3 and MoSi2 peaks are weaker in M60 coating because of the oxidation during thermal shock tests. In M80 coating, strong SiO2 peaks can be detected owing to the


4


Fig. 5. CLSM (a–c) and SEM (d–i) images of the MoSi2/borosilicate glass coatings, (a, d, g): M40; (b, e, h): M60; and (c, f, i): M80.

oxidation of MoSi2, meanwhile new phase MoB is also founded according to Eq. (8). While B2O3 tends to volatilization above 1473 K, the generated solid B2O3 transformed into gaseous B2O3 and volatilize, hence

making the B2O3 phase undetected in the XRD pattern. As shown in Fig. 10, the Gibbs-free energy, △ G, for reactions (5)–(9) at different temperatures are calculated by FACT program. Negative △ G suggests

Fig. 6. The cross-section micrograph of the integrative insulations. (a, d): M40; (b, e): M60; and (c, f): M80.



5

Fig. 7. The SEM images of M40 coated insulation (a) cross-section; (b) surface coating; (c) interfacial transition zone; (d) magnification of (c); and (e) EDS analysis of (d).

that these reactions can occur under the thermal shock condition in the range of 1473 K to 1673 K. MoðsÞ þ O2 ðgÞ ¼ MoO2 ðgÞ

MoSi2 ðsÞ þ 1=2B2 O3 ðsÞ þ 5=2O2 ðgÞ ¼ 2MoBðsÞ þ 4SiO2 ðsÞ

ð8Þ

Mo5 Si3 ðsÞ þ 21=2O2 ðgÞ ¼ 5MoO3 ðgÞ þ 3SiO2 ðsÞ

ð9Þ

ð5Þ

MoðsÞ þ 3=2O2 ðgÞ ¼ MoO3 ðgÞ

ð6Þ

MoSi2 ðsÞ þ 3O2 ðgÞ ¼ MoO3 ðgÞ þ 2SiO2 ðsÞ

ð7Þ

Fig. 8. Thermal cycling weight loss curves of the integrated insulations between 1673 K and room temperature.

Because of the appropriate wetting properties and viscosity of borosilicate glass [42,43], the borosilicate glass layer of the M40 coated insulation is able to prevent the diffusion of oxygen and can effectively seal the cracks induced by cyclic thermal shock. In contrast, compared with the amorphous borosilicate glass phase, the crystalline SiO2 in the M80 coated insulation has a higher viscosity and cannot effectively

Fig. 9. XRD patterns of MoSi2/borosilicate glass coatings after 15 thermal shock cycling times. (a): M40; (b): M60; and (c): M80.


6


Fig. 10. Gibbs-free energy for reactions (5)–(9).

flow and seal the defects in the coating [44,45]; thus, the weight loss is larger than for the M40 coated insulation. The M60 coating has less borosilicate phase, as well as less crystalline SiO2, compared with the other two coatings, and the oxidation time during the thermal shock test is too short to form sufficient glass for sealing the penetrating cracks; these penetrating cracks allow oxygen to diffuse and react with the inner MoSi2, which results in the maximum weight loss. Fig. 11 shows the evolution of the surface morphology with CLSM images after different thermal shock cycling times. Fig. 11(a–d) shows the surface state of the M40 coating; the surfaces are dense, glossy

and crack free. After 15 cycling times, the integrative insulation is also intact, and no micro cracks are generated in the coating because of the good self-secure ability of the borosilicate glass layer and the match of the thermal expansion coefficient between the coating and the substrate. Some holes formed by the emission of gaseous byproducts are found in the coating (Fig. 12(a, b)). As observed in Fig. 11(e–h), the surface gloss disappears with decreasing borosilicate glass content; the composition of the M60 coating is depleted gradually with the formation of volatilizable oxidation products (MoO2, MoO3) when the sample suffers thermal shock testing. Fig. 11(i–l) show the evolution of the surface morphology of the M80 coating; there is no marked change of the coating after the thermal shock test, which is because the rich-MoSi2 coating can form a SiO2 glass protective layer during the thermal shock tests. Fig. 13 presents cross-section images of the integrative insulations after thermal shock tests. The M40 coating integrates well with the fibrous ZrO2 ceramic substrate. The bridge between the coating and substrate is the translucent glass particles (Fig. 13(a, d)) with appropriate viscosity at high temperature, which could seal the cracks induced during the thermal shock test. Accordingly, the SEM images of M40 coated insulation after 15 thermal shock cycles are shown in Fig. 14. In the interfacial transition zone, the Zr\Si\O phase will be formed due to the reaction between the SiO2 and ZrO2, which connects the coating and the substrate well. It is suggested that the borosilicate glass is coupled with good bonding characteristics to the porous fibrous ZrO2 ceramic substrate (Fig. 14(a), (c) and (d)), with one of the major qualities being the ability to tailor their thermal expansion characteristics. Although no obvious peeling off of the coating occurred on the surface, penetrating cracks can be observed at the interface of the M60 and M80 coated insulations (Fig. 13(e, f)).

Fig. 11. The evolution of surface morphology with CLSM images after 0, 5, 10, and 15 thermal shock cycling times. (a–d): M40; (e–h): M60; and (i–l): M80.



7

Fig. 12. SEM images of the surfaces the MoSi2/borosilicate glass coatings. (a, b): M40; (c, d): M60; and (e, f): M80.

The phenomenon of interface delamination can be explained as follows: In the heating stage, temperature will be elevated for both coatings and fibrous ZrO2 ceramic substrate. Then the coatings should experience tensile stress due to the thermal expansion coefficient of the coating is lower than substrate, and some micro cracks generate in the coating. In the holding stage, the borosilicate glass phase will wet and heal the micro cracks, and the self-healing ability decreases with decreased amount of borosilicate, resulting in some micro cracks still in the coatings, especially in M80 coating (Fig. 12(e, f)). Meanwhile some holes are generated on the surface, which is caused by the volatilization of the products (MoO 2 , MoO3 and B2O3) at high temperature. In the cooling stage, subsequent rapid cooling moment makes the temperature of layers near the coating surface drop quickly, and layers near the coating surface are subjected to a much higher tensile stress than the heating moment; the propagating vertical cracks would pass pre-existing horizontal cracks and reach the interface, which leads to the phenomenon of interface delamination. 3.4. Infrared radiation performance of the coating For most practical purposes, thermal radiation is concerned with a wavelength range of 0.1 μm to 20 μm, including the ultraviolet and visible (λ = 0.1 to 0.8 μm), near infrared (λ = 0.8 to 10 μm), and far-

infrared (λ = 10 to 20 μm) spectra. In this article, two methods were used to measure the emissivity. In the range of 200 nm to 2500 nm, the spectral absorptivity of the samples was measured using an ultraviolet–visible-near IR spectrophotometer. In the range of 3–30 μm, the emissivity of the coatings was measured using an emissometer (Model AE1) at 353 K. Fig. 15 shows the spectral absorptivity in the 250–2500 nm spectral region of the as-prepared coatings. According to Kirchhoff's law, at equilibrium for a given wavelength (λ) and temperature (T), the absorptivity is strictly equal to the emissivity, i.e., ε(λ, T) = α(λ, T) [46]. In the ultraviolet region (250–400 nm), the absorptivity is larger than in the other waveband because of the interband absorption in the ultraviolet region, which excites the promotion of electrons in the valence band (VB) to the conduction band (CB) [20]. On one hand, from the band structure of MoSi2, the interband absorption in the ultraviolet region results from the transition of electrons from the sub-outer electron band to the VB and from the VB to the CB. On the other hand, because of the feature of long-range disorder in the amorphous structure of borosilicate glass, the electrons can form localized states, and their energy value is in the forbidden band, which provides the condition for the electron transition and promotes the infrared absorption in the short wave band [13]. In the range of 400–2500 nm, the spectral absorptivity of the M40 coating is far higher than that of the M60 and M80 coatings, which are


8


Fig. 13. The cross-section morphology with CLSM images after 15 thermal shock cycling times. (a, d): M40; (b, e): M60; and (c, f): M80.

in the range of 0.79–0.84 and 0.62–0.68, respectively. Many researchers suggested that there is a certain correlation between the emissivity and surface roughness [5,47–49]. The surface roughness values of the coatings shown in Table 2 were obtained from a CLSM study; the rootmean-square roughness (Sq) calculates the root mean square average

of the profile height deviations from the mean line (i.e. the standard sffiffiffiffiffiffiffiffiffiffiffiffiffiffi deviation of the height distribution): Sq ¼

n 2 1 n ∑ zi . i¼1

The skewness of

the profile (Ssk) is the quotient of the mean cube value of the ordinate

Fig. 14. The microstructure images of M40 coated insulation after 15 thermal shock cycles. (a) cross-section; (b) surface coating; (c) interfacial transition zone; (d) magnification of (c); and (e) EDS analysis of (d).



9

Fig. 15. The spectral absorptivity in the 250–2500 nm spectral region of the as-prepared coatings.

values z(x) and the cube of Sq within the evaluation length: Ssk ¼ 13 Sq n 3 1 [50]. The negative value about Ssk of M40 coating indicates n ∑ zi i¼1

that the surface is with valleys. The 3D surface roughness morphology gradient map of the coatings is presented in Fig. 16. Combining the absorptivity curves with the surface roughness, it can be observed that the variation of the absorptivity is consistent with the variation of the surface roughness in this range, especially in the range of 400–1750 nm. The M40 coating possesses the largest surface roughness values, whose absorptivity is also far higher than that of the others. For a real surface, the roughness is characterized by the ratio of root-meansquare surface roughness (σ) to wavelength (λ), (i.e., σ/λ). This ratio is generally divided into three regions, namely, (i) the specular region (0 b σ/λ b 0.2), (ii) the intermediate region (0.2 b σ/λ b 1) and (iii) the geometric region (σ/λ N 1) [16,48]. The value of σ is Sq in Table 2. For the M40, M60, M80 coatings, the σ/λ ratio lies in between 6.5 and 28.6, 3.0 and 13.0, 3.3 and 14.3 respectively. Hence, the σ/λ ratio of the coatings lies in the geometric region. In the geometric region, surfaces with repeatable grooved finish, like V-shaped grooves, circular grooves, and pyramidal grooves, are commonly used to model the emissivity enhancement [48,49]. According to the 3D surface morphology shown in Fig. 16, “circular grooves” is the most suitable model to explain the phenomena of the enhanced emissivity. As we know, most of the electromagnetic wave was reflected, and only a little was absorbed by the coatings with a smooth surface. However, a significant absorptivity increase and reflectivity decrease by small gradual slopes in circular grooves can be observed from the rough coating because multiple reflection will be caused by the slope of the grooved surface. Therefore, compared with the smooth surface, the rough coatings exhibit lower reflectivity and higher absorptivity, which leads to high emissivity according to Kirchhoff's law of thermal radiation. According to Wien's displacement law: λmax ¼ Tb, b = 2.8977721(26) × 10−3 m K; the blackbody intensity is maximum at a given temperature at the wavelength λmax and this maximum shifts toward shorter wavelengths as the temperature is increased [46]. Therefore, the higher the

temperature is, the emissivity of short wavelength is more meaningful. When the temperature is 1673 K around, the largest radiation is approximately 1.73 μm; in the range of the waveband, the materials exhibit a high emissivity over 0.8. The emissivity values of the coatings in the range of 3–30 μm at 353 K are 0.81, 0.82 and 0.78, respectively. The emissivity is largely increased compared with that of the fibrous ZrO2 ceramic substrate (ε = 0.58). However, in this spectral range, the surface roughness is not, at least in this case, the main factor for the marked difference in emissivity among the different coatings. In general, the absorptive mechanism mainly includes the interband absorption mechanism and crystal lattice vibration mechanism. The interband absorption mechanism at short wavelength was discussed above. Lattice absorption, resulting from the rotation and vibration of atoms and molecules, dominates the absorption at long wavelength. The symmetry polar bonds of Mo−Si form MoSi2, leading to the vibration absorption. In this work, the temperature condition of the measurement is 353 K; a higher temperature may lead to stronger lattice vibration and will result in a larger value of the emissivity [20]. Thus, there is little difference in the emissivity among the coatings. 4. Conclusions

Table 2 Surface roughness of the coatings.

M40 M60 M80

Fig. 16. 3D surface roughness morphology gradient map of the coatings.

Sq (μm)

Ssk (μm)

11.456 5.210 5.713

−0.011 0.068 0.599

An integrative insulation consisting of a high-emissivity MoSi2/ borosilicate coating and a low-thermal-conductivity ZrO2 fiber board substrate has been successfully prepared. The M40-coated insulation exhibits excellent thermal shock resistance; after thermal cycling between 1673 K and room temperature 15 times, the weight loss is only


10


2.84%. The emissivity of the M40 coating reaches 0.8 both over 250–2500 nm and 3–30 μm at room temperature, which is attributed to the synergistic effect of MoSi2 intrinsic properties, the amorphous structure of borosilicate glass and the surface roughness of the coating. The high emissivity and excellent thermal shock resistance of the integrative insulations make them good candidates for sensible materials, which could work for long-term applications up to 1673 K. Acknowledgments The authors are grateful to the support of the National Aerospace Science Foundation of China (No. 201452T4001), the Industry Program for Science and Technology Development of Jiangsu Province (No. BE2014128), the Program for Changjiang Scholars and Innovative Research Team in University (No. IRT1146), the Priority Academic Program Development of Jiangsu Higher Education Institution (PAPD), Jiangsu Planned Projects for Postdoctoral Research Funds (1402016A) and Jiangsu Collaborative Innovation Center for Advanced Inorganic Function Composites. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]

Y. Dong, C.-A. Wang, J. Zhou, Z. Hong, J. Eur. Ceram. Soc. 32 (2012) 2213–2218. J. Sun, Z. Hu, J. Li, H. Zhang, C. Sun, Ceram. Int. 40 (2014) 11787–11793. G. Shao, X. Shen, S. Cui, L. Wang, M. Chen, Mater. Rev. 28 (2014) 136–142. W. Tan, C.A. Petorak, R.W. Trice, J. Eur. Ceram. Soc. 34 (2014) 1–11. F. Wang, L. Cheng, Q. Zhang, L. Zhang, Appl. Surf. Sci. 313 (2014) 670–676. F. Wang, L. Cheng, L. Xiang, Q. Zhang, L. Zhang, J. Eur. Ceram. Soc. 34 (2014) 1667–1672. Z. Yao, B. Hu, Q. Shen, A. Niu, Z. Jiang, P. Su, P. Ju, Surf. Coat. Technol. 253 (2014) 166–170. J. Yi, X.D. He, Y. Sun, Y. Li, M.W. Li, Appl. Surf. Sci. 253 (2007) 7100–7103. Z. Li, L. Yang, D. Ge, Y. Ding, L. Pan, J. Zhao, Y. Li, Appl. Surf. Sci. 259 (2012) 811–815. D. Alfano, L. Scatteia, S. Cantoni, M. Balat-Pichelin, J. Eur. Ceram. Soc. 29 (2009) 2045–2051. S. Roy, R. Maharana, S. Gokul Laxmi, Y. Sakthivel, M. Roy, V.V. Bhanu Prasad, D.K. Das, Surf. Eng. 29 (2013) 360–365. L. Lu, X.a. Fan, J. Zhang, X. Hu, G. Li, Z. Zhang, Appl. Surf. Sci. 316 (2014) 82–87. X.D. Cheng, J. Min, Z.Q. Zhu, W.P. Ye, Int. J. Miner. Metall. Mater. 19 (2012) 173–178. Z.Q. Zhu, X.D. Cheng, W.P. Ye, J. Min, Int. J. Miner. Metall. Mater. 19 (2012) 266–270. D.B. Mahadik, S. Gujjar, G.M. Gouda, H.C. Barshilia, Appl. Surf. Sci. 299 (2014) 6–11. Q. Xu, X. Liu, F. Zhang, W. Chen, R. Yuan, J. Chin. Silic. Soc. 31 (2003) 529–532 (537). S. Wang, K. Liang, J. Non-Cryst. Solids 354 (2008) 1522–1525. J. Huang, Y. Li, X. He, G. Song, C. Fan, Y. Sun, W. Fei, S. Du, Appl. Surf. Sci. 259 (2012) 301–305.

[19] J.P. Huang, C.L. Fan, G.P. Song, Y.B. Li, X.D. He, X.J. Zhang, Y. Sun, S.Y. Du, Y.J. Zhao, Appl. Surf. Sci. 280 (2013) 605–609. [20] J.P. Huang, Y.B. Li, G.P. Song, X.J. Zhang, Y. Sun, X.D. He, S.Y. Du, Mater. Lett. 85 (2012) 57–60. [21] J. Huang, G. Song, H. Lv, Y. Li, Y. Sun, X. He, S. Zhang, Y. Fu, S. Du, Y. Li, Thin Solid Films 544 (2013) 270–275. [22] W.R. McMahon, D.R. Wilder, J. Am. Ceram. Soc. 51 (1968) 186–192. [23] S.M. Avdoshenko, A. Strachan, Model. Simul. Mater. Sci. 22 (2014) 075004. [24] L. Wang, M.H. Habibi, J.I. Eldridge, S.M. Guo, J. Eur. Ceram. Soc. 34 (2014) 3941–3949. [25] Z. Han, J. Liu, X.-W. Li, Y.-X. Chen, G.-H. Liu, J.-T. Li, J. Am. Ceram. Soc. 97 (2014) 2705–2708. [26] A.K. Vasudévan, J.J. Petrovic, Mater. Sci. Eng. A 155 (1992) 1–17. [27] T.A. Kircher, E.L. Courtright, Mater. Sci. Eng. A 155 (1992) 67–74. [28] C. Jainagesh A. Sekhar, Ohio, Molybdenum Silicide-Containing Products with High Emissivity, US Patent,6099978, 2000–8–8. [29] J.J. Petrovic, A.K. Vasudevan, Mater. Sci. Eng. A 261 (1999) 1–5. [30] M.H. Yu, B. Zhou, D.B. Bi, D. Shaw, Mater. Des. 31 (2010) 2478–2482. [31] D. Berztiss, R. Cerchiara, E. Gulbransen, F. Pettit, G. Meier, Mater. Sci. Eng. A 155 (1992) 165–181. [32] M. Liu, X. Zhong, J. Wang, Z. Liu, W. Qiu, D. Zeng, Surf. Coat. Technol. 239 (2014) 78–83. [33] B. Torres, M. Campo, M. Lieblich, J. Rams, Surf. Coat. Technol. 236 (2013) 274–283. [34] L.F. Cheng, Y.D. Xu, L.T. Zhang, R. Gao, Carbon 39 (2001) 1127–1133. [35] Y.L. Zhang, J.F. Huang, C.Y. Li, L.Y. Cao, H.B. Ouyang, B.Y. Zhang, W. Hao, W.J. Wang, C.-Y. Yao, Appl. Surf. Sci. 316 (2014) 207–213. [36] Q.G. Fu, H.J. Li, K.Z. Li, X.H. Shi, M. Huang, Carbon 45 (2007) 892–894. [37] J. Zhong, X. Ma, H. Lu, X. Wang, S. Zhang, W. Xiang, J. Alloy. Compd. 607 (2014) 177–182. [38] T.G.V.M. Rao, A. Rupesh Kumar, K. Neeraja, N. Veeraiah, M. Rami Reddy, J. Alloy. Compd. 557 (2013) 209–217. [39] T. Moskalewicz, F. Smeacetto, G. Cempura, L.C. Ajitdoss, M. Salvo, A. CzyrskaFilemonowicz, Surf. Coat. Technol. 204 (2010) 3509–3516. [40] J. Cook, A. Khan, E. Lee, R. Mahapatra, Mater. Sci. Eng. A 155 (1992) 183–198. [41] J. Yan, L. Liu, Z. Mao, H. Xu, Y. Wang, J. Therm. Spray Technol. 23 (2014) 934–939. [42] F. Smeacetto, M. Salvo, M. Ferraris, Carbon 40 (2002) 583–587. [43] C. Isola, M. Salvo, M. Ferraris, M.A. Montorsi, J. Eur. Ceram. Soc. 18 (1998) 1017–1024. [44] X. Yang, Q. Huang, Z. Su, X. Chang, L. Chai, C. Liu, L. Xue, D. Huang, Corros. Sci. 75 (2013) 16–27. [45] T. Feng, H.J. Li, X. Yang, X.H. Shi, S.l. Wang, Z.B. He, Corros. Sci. 72 (2013) 144–149. [46] John R. Howell, Robert Siegel, M. Pinar Menguc, Thermal Radiation Heat Transfer, fifth ed. CRC Press Inc., Boca Raton, Florida, 2010. [47] M.R. Bayati, H.R. Zargar, A. Talimian, A. Ziaee, R. Molaei, Surf. Coat. Technol. 205 (2010) 2483–2489. [48] X. He, Y. Li, L. Wang, Y. Sun, S. Zhang, Thin Solid Films 517 (2009) 5120–5129. [49] C.-D. Wen, I. Mudawar, Int. J. Heat Mass Transfer 49 (2006) 4279–4289. [50] S. Le Roux, F. Deschaux-Beaume, T. Cutard, P. Lours, Quantitative assessment of the interfacial roughness in multi-layered materials using image analysis: Application to oxidation in ceramic-based materials, J. Eur. Ceram. Soc. 3 (35) (2015) 1063–1079.
