Thin Solid Films 517 (2009) 51205129

Contents lists available at ScienceDirect

Thin Solid Films

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / t s f

High emissivity coatings for high temperature application: Progress and prospect
Xiaodong He a,, Yibin Li a, Lidong Wang a,b, Yue Sun a, Sam Zhang c
a b c

Center for Composite Materials, School of Astronautics, Harbin Institute of Technology, P.O. Box 3010, Harbin 150080, PR China School Materials Science and Engineering, Harbin Institute of Technology, P.O. Box 405, Harbin 150001, PR China School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore

a r t i c l e

i n f o

a b s t r a c t
High emissivity coatings are widely used in many cases where heat transfers through electromagnetic radiation that arises due to the temperature of a body. Extensive theoretical and experimental efforts have been made to synthesize and investigate high emissivity coatings. The emissivity can be improved through various or combined mechanisms. The characterization of the emissivity is still a fully open problem. In this paper, we review the various mechanisms associated with the emissivity enhancement and emissivity characterization techniques. Based on these literature reviews, the prospect will be presented in the concluding remarks. 2009 Elsevier B.V. All rights reserved.

Available online 22 April 2009 Keywords: Emissivity Coatings Review

1. Introduction High emissivity coatings are widely used in many high temperature applications to effectively transfer the heat by radiation [14]. For example, the surface of Ni-, Fe-, Co-based alloys applied in metal thermal protection systems (MTPS) [5,6] in third generation of reused space vehicles involves high friction heat from acute friction between space vehicle surface and atmosphere, which causes obvious increase of surface temperature up to 1000 C during hypervelocity ight [79]. As a result, the lifetime and performance of MTPS will be seriously degraded. Therefore, the high emissivity coating is intentionally deposited on MTPS to decrease the surface temperature by radiation. When a radiation falls on a body it may be partially reected, transmitted, or absorbed, which is associated with reection (R), absorption (A), transmission (T), and emission (). The principle of conservation of energy ties together these radiation characteristics as A + R + T = 1. According to Kirchhoff's law, at equilibrium for a given wavelength and temperature T, the emissivity of any body is equal to its absorption [10], i.e. (, T, ) = A(,T,), where is the emissivity, the frequency, and the angle species the observer angle. This indicates that the emissivity of an object can be indirectly obtained by measuring its absorptivity. For a uniform and isotropic opaque surface in thermal equilibrium, the transmission is zero, the relationship between the emissivity and the reectivity is: (,T, ) = A(,T,) = 1 R(,T,). Many materials can be considered as opaque body. The emissivity here is characterized to describe the surface radiative property which involves the transfer of heat by electromagnetic radiation arising due to the temperature of a body. The emissivity is dened as the ratio of energy

radiated by the material to energy radiated by a black body (A body that emits the maximum amount of heat for its absolute temperature is called a black body, meaning that a blackbody completely absorbs all radiation incident upon it, and at the same time it emits all the energy that it absorbs with the same absorbing spectrum. That is to say, = 1) at the same temperature, which is described by Eq. (1): R eT =
0

eE d T4

Where E is radiant heat of gray body to its surroundings, hemispherical emissivity of the gray body (dimensionless), Stefan Boltzman constant (5.67 10 8 W/(m2K4)), T temperature (K). The emittance is dependent on direction and wavelength, thus the radiance ability can also be evaluated by the spectral emissivity and directional emissivity, described as Eqs. (2) and (3) respectively: e ; T = E;actual emitted ; T E ; T = E;black body ; T Eb ; T R e ; /; T =
0

L;actual emitted ; ; /; T d L; T R = L ; T d Lb T 0 ;black body

Correspondence author. E-mail address: hexd@hit.edu.cn (X. He). 0040-6090/$ see front matter 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2009.03.175

A real object does not radiate as much as a perfect black body, which radiates less heat than a black body and is called gray body ( b 1). Thermally emitted radiance from any surface mainly depends on two factors. (1) the surface temperature, which is an indication of the equilibrium thermodynamic state resulting from the energy balance of the uxes between the gray body surface and its surroundings; and (2) the surface emissivity, which is the efciency of the surface for transmitting the radiant energy generated in the surface into its surroundings. The latter depends on the temperature (but the relationship between emissivity and temperature is not denite,

X. He et al. / Thin Solid Films 517 (2009) 51205129

5121

depending on material nature, surface parameters and wavelength), composition, surface roughness, coating thickness, wavelength, and physical parameters of the surface. Here we review recent progress with emphasis on the discussion of the design methodology, followed by brieng the emissivity characterization. Lastly, the prospects for research and technology in high emissivity coatings will be summarized. 2. Design methodology 2.1. Doping According to Wien's displacement law and Planck's law [11], most energy of the black body at high temperature is radiated in the wavelength range of 15 m, however it happens that many materials have the weak intrinsic absorption within this range. As a result, the emissivity in this region is quite low since the object's absorptivity and emissivity spectra are identical in thermal equilibrium state. Extensive work proves that total emissivity or spectral emissivity could be effectively enhanced via doping (doping refers to the process of intentionally introducing impurities into an extremely pure material). Several mechanisms for doping enhanced emissivity in the wavelength range of 15 m were proposed for different materials, namely, free carrier absorption mechanism (the phenomenon whereby an electron within a band absorbs radiation by transferring from a lowenergy level to an empty high-energy level) [12], dd transitions in octahedral coordination [13,14], distortion of the crystal lattice [15]. For semiconductor material, free carrier absorption mechanism generally dominates the emissivity enhancement in the wavelength region of 15 m. Fig. 1 shows the absorption coefcient as a function of wavelength for three different SiC specimens with different impurity (dopant) concentrations [16]. It is obvious that increase in impurity concentration strengthens the absorption in the wavelength range of 15 m. As higher impurity concentration causes more free electrons, more electrons within a band absorb more radiation by transferring from a low-energy level to an empty high-energy level, leading to more absorption. As such, the emissivity is accordingly enhanced within this wavelength range as the impurity concentration increases since the object's absorptivity and emissivity spectra are identical in thermal equilibrium state. In glassy oxide coatings, different excitations are responsible for the absorption of electromagnetic radiation [17]. It is suggested that suitable dopants with transition metal oxides diluted in silicate glasses may yield emissivity as high as 0.9 in the whole temperature range. For example, the dd transitions of Cr3+ in octahedral coordination play an important role on the absorption [13,14]. The

Fig. 2. The emissivity spectra at room temperature for 0.1%, 0.25% and 0.5% Cr2O3 doped compounds [17].

emissivity spectra at room temperature for 0.1%, 0.25% and 0.5% Cr2O3 doped compounds as shown in Fig. 2 indicates that the concentration of Cr2O3 dramatically enhanced the normal spectral emissivity between 1 and 5 m (i.e. 200010,000 cm 1). This enhancement is related to formation or precipitation of new phases on the grain boundary due to substitution, which will change the grain boundary area. But the mechanism still needs to be further digested. Another consequence caused by dopant is the crystal distortion due to different ion radii, which also has the contribution to radiation. Pratt et al. incorporated NiO into MgOAl2O3SiO2 system glasses heated at 1020 C for 2 h and the whole-band normal direction emissivity changes from 0.82 to 0.90 by doping NiO [18]. When the Ni2+ is incorporated into cordierite structure and substituted with Mg2+, the order degree of Al/Si in crystal structure of cordierite is decreased due to different radii of Ni2+ and Mg2+, leading to the crystal structure distortion. This in turn decreases the symmetry of lattice vibration, thus enhances the effects of anharmonic vibration of polar lattice, the coupled action of phonon, and phonon combination radiation. Therefore, the infrared radiance of this material is enhanced by NiO doping. Moreover, the dopant (rare-earth oxides and transition elements) could tune the spectral emissivity thus change the emissive power [19,20]. Incorporation of 24 wt.% Co3O4 or NiO into MgO host (a low infrared emissivity) produced matched emitters, which means its emissive power spectrum is matched very efciently to the response of GaSb photovoltaic cells that convert the infrared radiation into electricity. Fig. 3 shows the continuous, strong radiant emissions which is up to 0.9 in the optimal energy range between 1 and 2 m and minimal radiation at higher wavelengths [21,22]. The total emissive power between 1 and 9 m is calculated by integrating the area under each curve. Interestingly, both the blackbody emitter and the NiOdoped MgO emitter have the same total emissive power of 144,000 W cm 2. However, it is noticeable that the NiO-doped MgO emitter emits most of power at wavelengths less than 2 m, instead of 15 m in theoretical blackbody. The phenomenon results from intra-atomic electronic transitions determined by the electronic conguration of the dopant ions and interactions with the coordinating crystal eld of the host oxide [23].

2.2. Surface roughness The correlation between emissivity and surface roughness has been explored both theoretically [24,25] and experimentally[26,27]. Wen et al. [24] modeled the effect of surface roughness on emissivity of aluminum alloys. Normally the surface involves two broad categories: optically smooth (ideal) and rough (real) surfaces. For

Fig. 1. Absorption coefcient as a function of wavelength of SiC. The three curves are for SiC specimens with different impurity concentrations [16].

5122

X. He et al. / Thin Solid Films 517 (2009) 51205129

Fig. 3. Theoretical blackbody emissive power spectrum at 1010 C compared with the emissive power spectrum for an experimental 2 wt.% NiO-doped MgO emitter at 1404 C. [22]

optically smooth surfaces, spectral emissivity is determined by combining Fresnel's equation and Kirchhoff's law yielding e = 4n2 n + 12 + k2 4

Where n is the index of refraction and k the extinction coefcient. Fig. 4(a) compares spectral emissivity values using three different methods. It indicates that the experimental result shows the same trend as calculated one. Namely, the emissivity decreases as wavelength increases in the infrared range for most metallic surfaces. For real surfaces, the surfaces can be divided into three regions based on optical roughness (/, the ratio between surface roughness to wavelength): the specular region (0 b / b 0.2), intermediate region (0.2 b / b 1) and the geometric region (/ N 1). In the specular region, it assumes that the reection of incident radiation is specular, and the diffraction theory is used to predict the effects of surface roughness on emissivity [28,29]. For a Gaussian distribution of surface heights, the relationship between reectance and optical roughness follows an exponential decay function of /, which is identied both theoretically [30,31] and experimentally [32,33],     4 2 r = p exp 5

where r and p are the reectances for a rough surface and a polished surface, respectively. Eq. (5) reveals that the larger could result in lower reectance, i.e. higher emittance, which is veried in aluminum-coated ground glass as shown in Fig. 4(b) [29]. In the ( i) can be intermediate region, the directional reectance determined by integrating the bidirectional reectance function, (i, s) over the respective hemisphere for a given angle of incidence [34] as indicated in Fig. 4(c). According to Kirchhoff's law and energy conservation, the directional emissivity can be calculated by Eq. (6) Vi e Vi = 1 6

ignoring the diffraction effects. Surfaces with repeatable grooved nish, like V-shaped grooves [40], circular grooves [41], and pyramidal grooves [42], are commonly used to model the emissivity enhancement. The slope of the grooved surface plays an important role in emissivity enhancement. For example, the inter-reection caused by a 60 V-groove in Fig. 4(c) and a small grove angle (also the slope) in Fig. 4(d) could signicantly enhance the normal emissivity. Overall, modeling results lead to a conclusion that the emissivity is dependent on surface roughness for / b 1 and on surface slope for / N 1. In experiment, the emissivity enhancement due to roughness was veried in opaque materials, such as steel [26] and Pr2NiO4 coatings [27]. Fig. 5 compares normal spectral emissivity of rough Pr2NiO4 coating and single crystal Pr2NiO4 (considered as smooth surface) in the (ab) plane at 1000 K, in the working temperature range of the infrared radiator, from far to midinfrared. The spectral response of the coating is 30% to 40% higher than that of the single crystal, which is largely dominated by the infrared absorptions produced by localized polarons. Clearly, the enhancement of the intrinsic absorption is obtained by increasing the optical path of the light between each small cavity of the surface. Furthermore, the effect produced by the rough Pr2NiO4 + d coating is also clearly demonstrated in Fig. 5. The low normal spectral emissivity of the original semitransparent ceramic in the midinfrared range (from 2500 to 5000 cm 1) is completely occulted after the deposit. Although the previous achievements provide valuable insight into the emissivity principles concerning the relationship between emissivity and surface roughness, the inuences of surface roughness on emissivity remain illusive, and sometimes current emissivity models are consistent with experimental emissivity values. A universal emissivity model concerning roughness effect needs to be proposed in order to completely agree with the practically measured emissivity.

2.3. Surface texture The spectral radiative properties were completely dependent on the competition between the absorption coefcient (following the wavelength) and the radiative scattering. The surface texture (in materials science, texture is the distribution of crystallographic

Buckius et al. [3539] theoretically conrmed that increased roughness increases directional emissivity in the intermediate region. In the geometric region, geometric optics is used to predict emissivity,

X. He et al. / Thin Solid Films 517 (2009) 51205129

5123

Fig. 4. Classication of emissivity models based on surface roughness [24].

orientations of a sample) often forming during coating deposition, has an important effect on radiative scattering. For example, a surface texture with high aspect ratio feature (height to width ratioN 5: 1) can serve as a collection of a lot of individual black-bodies, which exhibit an emissivity of around 0.85 [43]. The theoretical and experimental results of IR spectral emittance of nonsphere-shaped particles indicate that coating composing of oriented rhenium whiskers with high aspect ratio (Height/Dimension = 5) provides the high emittance value, compared to that of randomly oriented whisker coating [44,45]. However, the texture effect on emissivity is still in argument. Recently, Rozenbaum et al. [46] found that unlike the other previously studied

ceramics, the highly textured alumina ceramic has a much lower emissivity than the single crystal between 1300 cm 1 and 1800 cm 1 (shown in Fig. 6) and the spectral emissivity of this sample was radically different from the emissivity spectra of lowly textured alumina. In this range, due to the weak absorption coefcient, the radiation mean path inside the sample was higher than the thickness of the single crystal. For longer wavelengths, the emissivity increases to become higher than those of the single crystal and tends to those of the lowly textured sample in the transparent zone. This is in good agreement with the previous reports. So it is concluded that we should consider wavelength range when we discuss the texture effect.

Fig. 5. Normal spectral emissivity at T = 1000 K and from 400 to 5000 cm 1 of cordieritemullite substrate (dashed line), single crystal of Pr2NiO4 + d (dotted line), and rough Pr2NiO4 + d coating on the cordieritemullite substrate (solid line) [27].

Fig. 6. Normal spectral emissivity of alumina ceramics (thickness = 1 mm, T = 1, 350 K) with two different textures [46].

5124

X. He et al. / Thin Solid Films 517 (2009) 51205129

Fig. 7. (a) A SEM view of a 3D tungsten photonic crystal. Within each layer, the 1D rod width is 0.5 m and the rod-to-rod spacing is 1.5 m. (b) A computed absorption spectra for an eight-layer 3D tungsten photonic-crystal sample. [51].

polarizations [47]. Beyond the photonic band gap regime, there are two strong absorptions: one is located at = 2.5 m with an absorptance of 40% and the other located at = 1.51.9 m with an even stronger absorptance of 80%. As discussed above, Kirchhoff's law tells us that at equilibrium for a given wavelength and temperature T, the emittance of any body is equal to its absorptance [10]. Thus, a strong narrow absorption is suggestive of a narrow band emission from a photonic-crystal sample. In experiment, the measured power-density spectrum of the sample in Fig. 8 is compared with that computed for a blackbody cavity radiator at T = 750 K. The blackbody radiation follows the Planck radiation. In the bandgap, the emission is suppressed and below the blackbody emission. However, photonic-crystal emission in the peak regime is found to exceed that of the blackbody by 300%. Note the photonic band gap emission exhibits a narrow FWHM of 1.8 m while black body shows a wide FWHM of 4.5 m. The absorption rate of 3D metallic photonic crystal is found to be suppressed in the photonic bandgap regime ( 816 m) and strongly enhanced at the photonic band edge ( 6 m) The peak absorption in tungsten photonic crystal is an effective absorption that originates from the intrinsic tungsten absorption and is subsequently enhanced by the photonic band edge effect. The enhancement is attributed to a longer photon-matter interaction length as compared to the metallic skin depth, the slower group velocity of light at the band edge, and the nite metallic absorption constant. It should be stressed that the photonic-crystal by no means exceeds the blackbody emission in the entire wavelength spectrum, rather the exceeding happens at the neighborhood of the band-edge only [55]. Although the photonic crystal has been rapidly developed, the coating or thin lm of photonic crystal has not been realized yet. 2.5. Coating thickness As mentioned above, when a radiation falls on a body, it may be partially reected, transmitted, or absorbed. As such, light penetration depth (dened as the inverse of the absorption coefcient ()) is closely related to the emissivity of coatings. For the light with the wavelength , the depth of light penetration, dp, is associated with the extinction coefcient (k), given by dp = 1 = 4 k 7

2.4. Surface periodicity A three-dimensional (3D) complete photonic band gap can be used to suppress radiation below the electronic band gap [4750]. Meanwhile, emission can be enhanced at a narrow band near a photonic band edge or a narrow allowed band. If both effects are combined, a nearly ideal radiation spectrum can be obtained. Photonic crystals refer to the periodic optical (nano)structures that are designed to affect the motion of photons in a similar way that periodicity of a semiconductor crystal affects the motion of electrons. Photonic crystals are composed of periodic dielectric or metallo-dielectric (nano)structures that affect the propagation of electromagnetic waves (EM) in the same way as the periodic potential in a semiconductor crystal affects the electron motion by dening allowed and forbidden electronic energy bands. Since this basic physical phenomenon is based on diffraction, the periodicity of the photonic crystal structure has to be of the same length-scale as half the wavelength of the EM waves i.e. 200 nm (blue) to 350 nm (red) for photonic crystals operating in the visible part of the spectrum the repeating regions of high and low dielectric constants have to be of this dimension. This makes the fabrication of optical photonic crystals cumbersome and complex. Interestingly, Lin et al. realized a 3D tungsten photonic-crystal sample (shown in Fig. 7(a)) using a modied silicon process [51], which can be thermally excited and exhibits emission at a narrow band [52,53]. In theory, the computed absorption spectrum for an eight-layer photonic-crystal sample in Fig. 7(b) indicates that the absorptance is low for N 3 m (the photonic band gap) and increases slightly at 34 m due to a higher tungsten material absorption at these wavelengths [54]. This band gap is a complete band gap, capable of trapping light fully in all three dimensions and for both

Eq. (7) indicates that dp is reversely proportional to k. Thus, k is expected to be as low as possible in order to get high emissivity of coatings. As a result, the coating is expected to be as thick as possible, at least thicker than dp (otherwise, the transmittance will occur). But in practice it is impossible to fabricate unlimitedly thick coating due to production cost or fabrication limitation. In the past, models and

Fig. 8. Comparison of emission power density between the photonic-crystal sample and a blackbody cavity radiator at T = 750 K [52].

X. He et al. / Thin Solid Films 517 (2009) 51205129

5125

experiments concerning the relationship between emissivity and coating thickness were explored in details [56]. A typical relationship between them in particle embedded in Al2O3 coating is shown in Fig. 9, revealing that initially the emissivity increases rapidly as thickness increases and becomes steady above thickness of 40 m. Similar relationship was also found in other ceramic coatings, such as SiC, Si3N4 and A12O3 [5], anodic lm on aluminum [57], Si3N4 and SiO2 on silicon substrates [58]. Therefore, it is concluded that the coating has a critical thickness over and above which emissivity almost does not change. When the coating has more than two layers (called multilayer coating), the optical interference will happen. If the thickness of multilayered coatings falls into the same order of magnitude as the incident radiation, a considerable amount of optical interference within the multilayer structure tends to occur [59]. The net surface reectivity and absorption often change signicantly even though there is any thickness change in any layer. The resultant temperature proles and quality of the lm are therefore affected by these changes in optical properties. This was evident in SiO2/Si system, as shown in Fig. 10 [60]. The relationship between emissivity and thickness shows a converted V shape. The emissivity peaks at the SiO2 thickness of around 240 nm for a given wavelength. When the coating is ultra thin, approaching zero, smaller than the wavelength of the radiation light, the increase of emissivity is ascribed to the coating thickness. The reections from the second interface are in phase with the one from the rst interface, resulting in a high reection and low emissivity. As the coating thickness increases, when the phase difference of reected lights from surface and interface is equal to the radiation light wavelength, extinction happens where the emissivity reaches the maximum value. As the thickness further increases, this phase difference deviates from the wavelength and causes the emissivity decreasing. However, this critical thickness (from several tens to 100 m) where interference of radiation light occurs depends on the coating composition [61]. 2.6. Nanocomposite coatings For normal incident radiation, reection coefcient R is given by the Fresnel's formula: R= n 12 + k2 n + 12 + k2 8

Fig. 10. Emissivity ( = 1.53 m) versus oxide thickness at 925 C [60].

where n is the refractive index, k the extinction coefcient. According to Eq. (4), the extinction coefcient k becomes much larger than 1

Fig. 9. Comparison of theoretical and experimental data of emissivity as function of oxide coating thickness [55].

when spectral emissivity is approaching the maximum value of 1. Eq. (8) states the reection coefcient tends to reach a maximum value of 1. As a result, there is an abrupt decrease of while R has a sudden increase as wavelength increases, which is clearly demonstrated in Fig. 11 for boron nitride [62]. Note that a peak of R in Fig. 11 (a) accompanies with a sudden drop of emissivity and Fig. 11(b). This concluded that single component material can hardly show high emissivity on the whole frequency range. Besides pursuit of high emissivity, the other properties such as mechanical properties, thermal shock resistance and oxidation resistance, should be also taken into account when high emissivity coatings are used at high temperature (N 1000 C) like MTPS. However, single component coating could rarely possess these comprehensive properties. Designing nanocomposite coatings may help a lot. A nanocomposite coating comprises of at least two phases: a crystalline phase and an amorphous phase, or two different crystalline phases, in which at least one is capable of absorbing and re-radiating thermal energy (emissivity agents), such as BN, SiC, SiB6, traditional metal oxides, refractory metals carbides [63]. The other phase is like binder such as colloidal silicon dioxide, which can provide high strength, thermal expansion characteristics similar to their intended substrates, and adequate bond strength with the substrate [64]. These composite coatings represent a new class of materials, whose mechanical or emittance properties are not subjected to volume mixture rules, but depends on synergetic interactions of the composite constituents. Recently, our group deposited SiC/SiO2 composite coating on the pre-oxidized 316 SS substrate via EB-PVD technology [65,66]. It is interesting to nd that nanocrytalline SiC deposited at high temperature was formed, as shown in Fig. 12, giving rise to the nanohardness as high as 50 GPa. At the same time, it is well known that SiC exhibits very good emittance property. Combination of emittance and mechanical properties makes SiC promising applied in high temperature environment or space, like MTPS [67]. In SiC/SiO2 system, SiO2 can act as an additive oxidation insulator because of low thermal conductivity and high oxidation resistance. The spectral directional emissivity of SiC/SiO2 composite thin lm almost keeps a constant of around 0.65 at a wide wave-number range from 400 to 4000 cm 1, as shown in Fig. 13. Unfortunately, the other properties like oxidation resistance, shock resistance, have not been investigated. But it can be predicted that the nanocomposite coatings are a promising branch of materials that could potentially possess both high emissivity and unique mechanical or thermal shock resistance properties. Another composite coating structure, i.e. multilayer, is proposed in order to meet comprehensive requirements, such as oxidation, emittance, and low catalysis characteristics in hypersonic structures. In the multilayered system, normally ve different layers are realized (schematically shown in Fig. 14) [68]: (i) low catalysis layer (based on borosilicate or borophosphate glasses) to minimize aerothermodynamic heating; (ii) the high-emittance layer consisting of various

5126

X. He et al. / Thin Solid Films 517 (2009) 51205129

Fig. 11. Reectance and emittance of boron nitride [62].

X. He et al. / Thin Solid Films 517 (2009) 51205129

5127

Fig. 14. Schematic diagram of the multilayer coating [68].

Fig. 12. HRTEM and cross-sectional images of SiC thin lm prepared by EB-PVD at 900 C.

has just begun, and it is expected to attract ever increasing attention in the near future. 3. Emissivity characterization

constituents, such as SiB6, providing emittance enhancement to the total system; (iii and iv) the oxygen-resistance layer to block oxygen diffusion, comprising of a base layer and sealing layer; (v) reaction barrier layer (like Al2O3, TiAl3) to prevent reaction between the substrate alloy and the other coating layers. It is recognized that thermal expansion coefcient differences between layers are inherent in this multilayer design. Accordingly, we can accommodate thermal strains by adjusting the layer thickness, and its self-healing characteristics can help avoid the cracking that may occur. Bird et al. realized this multilayer design by solgel method [68]. The emittance layer of the coating imparted an emittance in excess of 0.8 during exposure at 980 C for up to 100 h, which was at least 15% greater than that for the coated alloy without the emittance layer as shown in Fig. 15. Likewise, the low catalysis outer layer of the coating reduced the catalytic efciency for recombination of atomic oxygen and nitrogen in hypersonic owing air. The catalytic efciency of coated Inconel 617 was reduced by a factor of four, compared to the uncoated condition, for a 1-h exposure at 980 C. To summarize, nanocomposite coatings with multiple components or multilayered structure may be most attractive to reach high emissivity value within the whole frequency range as well as excellent mechanical or thermal properties. The research of multilayer design

In a classical denition, emissivity measures the ability of a material to radiate the heat through electromagnetic wavelength. Emissivity measurements are difcult or tedious in practice because it is a function of several variables and an extrinsic parameter of a material. Although many methods have been proposed and developed so far, a standard procedure or a universally accepted methodology does not exist yet. Qualitatively, the emissivity can be estimated by four developed basic methods: Calorimetric method [6971]; optical reectivity method [72], multispectral radiation thermometry [73] and radiation energy method [74]. Calorimetric method involves measuring the heat power delivered to the test sample as well as the temperatures of the emitting surface (the sample) and the surroundings in a vacuum over time [70,71]. An energy balance is then used to

Fig. 13. The emissivity and reectivity of SiC/SiO2 composite thin lm as a function of wavenumber [9].

Fig. 15. Emittance (a) and weight-change (b) as a function of exposure in air at 1000 C for coated Inconel 617 with emittance coating embedded within the oxidation coating [68].

5128

X. He et al. / Thin Solid Films 517 (2009) 51205129 Wronski, W.C. Sinke, Proceedings of 3rd World Conference on Photovoltaic Energy Conversion, Osaka, Japan, May 1118, 2003, p. 8PB511. T. Sumita, M. Imaizumi, S. Kawakita, S. Matsuda, S. Kuwajima, in: J.L. Wert (Ed.), IEEE Radiation Effects Data Workshop Record, Monterey, U.S.A., July 2125, 2003, P11, 2003. T. Sumita, M. Imaizumi, S. Kawakita, S. Matsuda, S. Kuwajima, T. Ohshima, T. Kamiya, in: J. Black (Ed.), IEEE Radiation Effects Data Workshop Record, Atlanta, U.S.A., July 1923, 2004, P130, 2004. M. FaIz, G. Leonhardt, Surf. Coat. Technol. 61 (1993) 97. M.G. Hocking, V. Vasantasree, P.S. Sidky, Metallic Ceramic Coatings: Production, High Temperature Properties Applications, Longman, London, 1989 P670. C. Bartuli, T. Valente, M. Tului, Surf. Coat. Technol. 155 (2002) 260. A. Ordine, C.A. Achete, O.R. Mattosa, I.C.P. Margarit, S.S. Camargo Jr., T. Hirsch, Surf. Coat. Technol. 583 (2000) 133. J. Yi, X.D. He, Y. Sun, Y. Li, Appl. Surf. Sci. 253 (2007) 4361. M. Auslender, S. Hava, Infrared Phys. Technol. 36 (1995) 1077. G.B. Rybicki, A.P. Lightman, Radiative Processes in Astrophysics, John Wiley & Sons, New York, 1979 22 pp. N.M. Ravindra, W. Chen, F.M. Tong, Arun Nanda, in: N.M. Ravindra, R.K. Singh (Eds.), Transient Thermal Processing Techniques in Electronic Materials, TMS Soc., Warrendale, PA, Aug. 1996, pp. 159164. C.R. Bamford, Phys. Chem. Glasses 3 (1962) 189. J.A. Duffy, J. Am. Ceram. Soc. 60 (1977) 440. B.M. Wang, D.Z. Su, G.Y. Zhang, China J. Infrared Res. 2 (1983) 55 (in Chinese). W.G. Spitzer, D. Kleinman, D. Walsh, Phys. Rev. V113 (1959) 127. F. Seronde, P. Echegut, J.P. Coutures, F. Gervais, Mater. Sci. Eng. B8 (1991) 315. R.H. Pratt, Radiat. Phys. Chem. 70 (45) (2004) 595. M.G. Krishna, M. Rajendran, D.R. Pyke, A.K. Bhattacharya, Sol. Energy Mater. Sol. Cells 59 (1999) 337. A. Lowe, D.L. Chubb, S.C. Farmer, B.S. Good, Appl. Phys. Lett. 64 (1994) 3551. L.G. Ferguson, Fatih Dogan, Mater. Sci. Eng. B83 (2001) 35. L.G. Ferguson, F. Dogan, J. Mater. Sci. 36 (2001) 137. K. Nassau, The Physics and Chemistry of Color The Fifteen Causes of Color, Wiley, New York, 1983 104 pp. C.D. Wen, I. Mudawar, Int. J. Heat Mass Transfer 49 (2006) 4279. F. Ghmari, T. Ghbara, J. Appl. Phys. 96 (2004) 26562664. J.F. Ony Sacadura, Int. J. Heat Mass Transfer 15 (1971) 1451. B. Rousseau, M. Chabin, P. Echegut, A. Sin, F. Weiss, P. Odier, Appl. Phys. Letts. 79 (2001) 3633. R. Siegel, J.R. Howell, Thermal Radiat. Heat Trans. McGraw-Hill, New York, NY, 1981. M.J. Haugh, Radiation thermometry in the aluminum industry, in: D.P. DeWitt, G.D. Nutter (Eds.), Theory and Practice of Radiation Thermometry, John Wiley, New York, NY, 1988. J. De Santo, G. Brown, in: E. Wolf (Ed.), Progress in Optics, Cambridge University Press, Cambridge, 1991. R. Carminati, J.J. Greffet, A. Sentenac, in: J.S. Lee (Ed.), Proceedings of the 11th International Heat Transfer Conference, vol. 7, Korean Society of Mechanical Engineers, Kyongju, 1998, p. 427. R. Petit, Electromagnetic Theory of Gratings, Springer, Berlin, 1980. P. Lalanne, G.M. Morris, J. Opt. Soc. Am. A 13 (1996) 779. M.Q. Brewster, Thermal Radiative Transfer and Properties, John Wiley, New York, NY, 1992. R.A. Dimenna, R.O. Buckius, J. Thermophys. Heat Transf. 3 (1994) 393. R.A. Dimenna, R.O. Buckius, Trans. ASME 116 (1994) 639. K. Tang, R.A. Dimenna, R.O. Buckius, Int. J. Heat Mass Transfer 40 (1997) 49. K. Tang, R.A. Dimenna, R.O. Buckius, Int. J. Heat Mass Transfer 41 (1997) 2037. K. Tang, R.A. Dimenna, R.O. Buckius, Int. J. Heat Mass Transfer 44 (2001) 4059. G.J. Dail, M.G. Fuhrman, D.P. DeWitt, Proceeding of 4th Int. Aluminum Extrusion Technology Seminar, Chicago, IL, vol. 2, April, 1114, 1988, p. 281. K. Kanayama, Heat Transfer-Jpn. Res. 1 (1972) 11. A. Abdulkadir, R.C. Birkebak, Rev. Sci. Instrum. 45 (1974) 1356. W. Konz, J. Hildenbrand, M. Bauersfeld, S. Hartwig, A. Lambrecht, V. Lehmann, J. Wllenstein, Proceedings Microsystems for the New millennium, SPIE, vol. 5836, 2005, doi:10.1117/12.608712. J.R. Aronson, A.G. Emslie, F.E. Ruccia, C.R. Smallman, E.M. Smith, P.F. Strong, Appl. Opt. 18 (1979) 2622. W.L. Xu, S.C. Shen, Appl. Opt. 36 (1997) 1644. D. Rozenbaum1, De Sousa Meneses P. Echegut, Int. J. of Thermophys. 30 (2009) 580. J.G. Fleming, S.Y. Lin, I. El-Kady, R. Biswas, K.M. Ho, Nature (London) 417 (2002) 52. S.Y. Lin, J.G. Fleming, Z.Y. Li, I. El-Kady, R. Biswas, K.M. Ho, J. Opt. Soc. Am. B 20 (2003) 1538. E. Yablonovitch, Phys. Rev. Lett. 58 (1987) 2059. S. John, Phys. Rev. Lett. 58 (1987) 2486. S.Y. Lin, J. Moreno, J.G. Fleming, Appl. Phys. Lett. 83 (2003) 380. S.-Y. Lin, J.G. Fleming, I. El-Kady, Optics Lett. 28 (2003) 1909. T. Trupke, P. Wrfel, M.A. Green, Appl. Phys. Lett., 84 (2004) 1997. M.A. Ordal, L.L. Long, R.J. Bell, S.E. Bell, R.R. Bell, R.W. Alexander Jr., C.A. Ward, Appl. Opt. 22 (1983) 1099. S.Y. Lin, J. Moreno, J.G. Fleming, Appl. Phys. Lett. 84 (2004) 1999. J.W. Zhou, X.Q. Wang, J.C. Han, Acta Mater. Compositae Sin. 24 (2007) 86. C.S. Kumar, A.K. Sharma, K.N. Mahendra, S.M. Mayanna, Sol. Energy Mater. Solar Cells 60 (2000) 51. P.L.A. Chr. M. van der Meer, L.J. Giling, S.G. Kroon, J. Appl. Phys. 47 (1976) 652. P.Y. Wong, I.N. Miaouiis, P. Zavracky, Sens. Actuator 19 (1990) 175. N.M. Ravindra, S. Abedrabbo, W. Chen, Feiming M. Tong, Arun K. Nanda, Anthony C. Speranza, IEEE Trans. Semicond. Manuf. 11 (1998) 1.

determine the emissivity. The sample must be well insulated on the side and back, and parasitic heat losses must be accounted for to accurately determine the heat transfer through the sample surface [75]. Reectivity method involve illuminating a sample with infrared energy and measuring the percentage of energy reected from the surface [4,76]. The absorptance is calculated from the reectance and used to calculate normal emittance by Kirchhoff's law and the Stefan Boltzmann equation. Multispectral radiation thermometry (MRT) [73] employs radiance measurements at three or more wavelengths and an emissivity model [77,78]. This method could measure the spectral emissivity and temperature simultaneously based on the relationship between temperature and spectral emissivity as follows:  p e = exp a0 9

[3] [4]

[5] [6] [7] [8] [9] [10] [11] [12]

This method provides many advantages, such as sample-preparation fast test free, on site test, no limitation for temperature, but its accuracy is lower and there is no common model to be applicable of all the materials. The most widely method used in the assessment of emissivity of coating is energy method which is based on the measurement of radiation power of sample surface over the radiation power of black body at the same temperature (StefanBoltzmann equation). The key problem is to select the suitable black body which determines the measurement accuracy. More details for emissivity measurement via energy method could be found in references [74,79 83]. Fourier spectroscopy based energy method for emissivity test has been rapidly developed recently, which prides more accuracy (less than 3%) of emissivity measurement. The temperature range of measurement is from 20 to 2000 C and the wavelength ranges from visible light to above 25 um. 4. Prospect Absorption coefcient and scattering coefcient S are important factors to be considered when designing a high emissivity coating; increasing and decreasing S can enhance the emissivity. In principle, all emissivity enhancement mechanisms can be applied to the enhancement of ceramic coatings, the real challenge lies in the implementation. However, for real engineering applications, pursuing high emissivity is not enough, while a combination of high emissivity and other comprehensive properties (called multifunctional coatings) will be pursued more by the academics driven by the industrial needs. In-situ synthesis of nanocrystalline or nanocomposite coatings (such as nc-SiC/-SiO2 or nc-SiC/-C) or designing multilayered structural layers may hold the key (at least, two of the keys) to this multifunctional coating. Another promising branch called variable emittance coating has just come up [84]. Small spacecraft, including micro and nanosatellites will require an alternative means to achieve thermal control due to their small power and mass budgets. Variable emittance coatings will be used as spacecraft radiators for thermal balance. In respect to the emissivity measurements, the lack of a universally accepted emissivity measurement methodology and instrumentation for any shape or different size samples has effectively hindered the progress of the research. It is hoped that in a near future, a universally accepted methodology will be born, and that this will accompany the birth of an emissivity measurement apparatus for coatings, a standalone machine. To catalyze the birth of the universal emissivity measurement, regular international seminars should be organized especially on this topic. Fourier spectroscopy based energy method may become the promising technique and needs further development. References
[1] B.V. Cockeram, D.P. Measures, A.J. Mueller, Thin Solid Films 355356 (1999) 17. [2] S. Kawakita, M. Imaizumi, T. Sumita, K. Kushiya, T. Ohshima, M. Yamaguchi, S. Matsuda, S. Yoda, T. Kamiya, K. Kurokawa, L.L. Kazmerski, B. McNelis, M. Yamaguchi, C.

[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29]

[30] [31]

[32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43]

[44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60]

X. He et al. / Thin Solid Films 517 (2009) 51205129 [61] M.H. Han, X.G. Liang, Y. Huang, Chin. Sci. Bull. 50 (2005) 588. [62] Y.S. Touloukian, D.P. DeWitt, Thermal Radiative Properties: Nonmetallic Solids, IFI/ Plenum, New York, 1972 P. 1042, 1047. [63] Y. Ozaki, R.H. Zee, Mater. Sci. Eng. A, 202 (1995) 134. [64] H.W. Babel, H.G. Le, Us patent No. 5,296,288 (1994). [65] X.D. He, J. Yi, Y. Sun, P. Xiao, X.F. Zhao, Appl. Phys. Lett. 91 (2007) 241914 1. [66] J. Yi, X.D. He, Y. Sun, Y. Li, M.W. Li, Appl. Surf. Sci. 253 (2007) 7100. [67] A.K. Costa, S.S. Camargo Jr., C.A. Achete, R. Carius, Thin Solid Films 377378 (2000) 243. [68] R. Keith Bird, Terryl A. Wallace, J. Spacecr. Rockers 41 (2004) 213. [69] J. Hameury, B. Hay, J.R. Filtz, Int. J. Thermophys. 28 (2007) 1607. [70] R.D. Karam, in: P. Zarchan (Ed.), Thermal Control Hardware, Satellite Thermal Control for Systems Engineers, AIAA, Reston, VA, 1998, pp. 160164. [71] A. Mychkovsky, and R. Ponnappan, Calorimetric Measurement of Emissivity in Space Conditions, AIAA 2005-961, Jan. 2005. [72] Dunkle R.V. In progress in International Research on Thermodynamics and Transport Properties Asme[M], 1962, p. 100. [73] C.D. Wen, I. Mudawar, J. Commun. Heat Mass Transfer 33 (2006) 1063.

5129

[74] J. Ishii, A. Ono, Meas. Sci. Technol. 12 (2001) 2103. [75] R.D. Karam, Thermal control hardware, in: P. Zarchan (Ed.), Satellite Thermal Control for Systems Engineers, AIAA, Reston, VA, 1998, pp. 160164. [76] D.A. Jaworske, T.J. Skowronski, in: M. El-Genk (Ed.), Proceedings of the Space Technology and Applications International Forum (STAIF-2000), AIP Conference Proceedings, No. 504, American Institute of Physics Press, New York, 2000, p. 791. [77] P.B. Coates, Metrologia 17 (1981) 103. [78] C. Wen, I. Mudawar, J. Heat Mass Transfer 48 (2005) 1316. [79] B.M. Liu, Z.X. Chu, 10th European Conference on Thermophysical Properties, 1986, p. 22. [80] S. Clausen, A. Morgenstjerne, O. Rathmann, Appl. Optics 35 (28) (1996) 5683. [81] A. R Korb, P. Dybwad, W. Wadsworth, J.W. Salisbury, Appl. Optics 35 (10) (1996) 1679. [82] E. Lindermeir, V. Tank, P. Haschberger, Proceedings of SPIE-The Int. Society for Optical Eng., vol. 1682, 1992, pp. 354364. [83] J.M. Dai, X.D. Liu, X.G. Sun, 7th Int. Workshop on Subsecond Thermophysics, 2004. [84] D. Farrar, W. Schneider, R. Osiander, J.L. Champion, A.G. Damn, Inter Society Conference on Thermal Phenomena, 2002, p. 1020.