Broadband enhancement of thermal radiation

Versão do arquivo anexado / Version of attached file:

Versão do Editor / Published Version

Mais informações no site da editora / Further information on publisher's website:

https://www.osapublishing.org/oe/abstract.cfm?uri=oe-27-12-A818

DOI: 10.1364/OE.27.00A818

Direitos autorais / Publisher's copyright statement:

©2019

by Optical Society of America. All rights reserved.

DIRETORIA DE TRATAMENTO DA INFORMAÇÃO Cidade Universitária Zeferino Vaz Barão Geraldo

CEP 13083-970 – Campinas SP Fone: (19) 3521-6493 http://www.repositorio.unicamp.br

Broadband enhancement of thermal radiation

G

R. B

,

A

D

,

S

A. M

,

R

S

-G

,

F

A. S. B

,

P

A. N

,

M

L

*

1_{Department of Electrical Engineering, Columbia University, New York, NY 10027, USA}

2_{Department of Electrical Engineering, Ginzton Laboratory, Stanford University, Stanford, California} 94305, USA

3_{Department of Mechanical Engineering, University of Ottawa, Ottawa, Ontario, KIN 6N5, Canada} 4_{Current address: Instituto de Física, Universidade Estadual de Campinas, 13083-970 Campinas, SP,} Brazil

5_{Instituto de Física da Universidade de São Paulo, P.O.Box 66318, 05315-970 São Paulo, Brazil}

*ml3745@columbia.edu

Abstract: Broadband thermal radiation sources are critical for various applications including

spectroscopy and electricity generation. However, due to the difficulty in simultaneously achieving high absorptivity and low thermal mass these sources are inefficient. We show a platform that enables one to obtain enhanced emission by coupling a thermal emitter to an optical cavity. We experimentally demonstrate broadband enhancement of thermal emission between λ ~2 ̶ 4.2 μm using an inherently poor thermal emitter consisting of tens of nanometers thick SiC film with 10% emissivity (εSiC ~0.1). We measure over twofold

enhancement of total emission power over the entire spectral band and threefold enhancement of thermal emission over 3 to 3.4 μm. Our platform has the potential to enable development of ideal blackbody sources operating at substantially lower heating powers.

© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement 1. Introduction

Broadband thermal radiation sources, widely used in applications of spectroscopy [1], sensing, lighting [2], thermal control [3] and energy generation [4], suffer from a tradeoff between either being dim (e.g. non-dispersive infrared sources) [1,5] or power hungry (e.g. incandescent lamps) [2,4]. These sources are bulk absorbers of a given material, having large volumes (high thermal mass) and therefore, due to their high emissivity (ε) are bright. But this comes at the cost of their large heat capacitance, consuming high input power to achieve high temperatures [6,7]. Thin-films have lower emissivity and therefore they are not bright although they consume only a small input power. Previous works have shown enhancement in thermal radiation capability of bulk emitters using photonic crystal cavities, plasmonic antennas and metamaterial based microstructures, but this enhancement has been limited to narrow spectral bands [2,3,8–25].

Here, we show a platform that breaks the traditional tradeoff of thermal emitters and show high emissivity for a thin-film over a broad wavelength range. The platform consists of a macroscopic optical cavity formed by a mirror and a thin-film emitter, as shown in Figs. 1(a) & 1(b). The thin-film serves the dual purpose of a thermal emitter as well as a partial reflector that forms the cavity. Our optical cavity enhances the interaction of thermal light with the weak emitter leading to an increase in effective absorptivity. We design the cavity to achieve maximum absorptivity by optimizing the coupling between the emitter and the cavity. The increased effective absorptivity (A) translates to effective emissivity (ε) according to Kirchhoff’s law (ε = A = 1 – R – T, where, R: Reflectivity, T: Transmittivity) [26,27].

#354915 https://doi.org/10.1364/OE.27.00A818

Fig. 1. Effect of silicon carbide film-thickness on thermal output of cavity-coupled emitter. Scheme of (a) a bare emitter, (b) a cavity-coupled emitter and, (c) their simulated normalized total emission as a function of emitter thickness. The blue line represents emission of the bare emitter while the red line is for the cavity-coupled emitter. For a cavity-coupled emitter the back mirror allows recycling of thermal light which goes back and forth several times before exiting the cavity. According to Kirchhoff’s law this redirected radiation is absorbed and re-emitted to maintain thermodynamic equilibrium. The peak emission occurs for a thickness of ~2 μm where optimal impedance matching occurs. For higher thickness (i.e. higher absorptivity) the enhancement reduces and eventually reaches unity. The total power is normalized with respect to that from a blackbody at equivalent temperature. The simulation data is for a cavity length L ~100 mm and back mirror reflectivity R ~0.99.

2. Theoretical analysis

The proposed platform is predicted to have a strong effect on the thermal emission inducing a large enhancement of the emission for thin-films. We consider a broadband metallic mirror (R ~0.99) to form the cavity with a weakly absorbing microcrystalline silicon carbide (μC-SiC) thin-film [28]. We choose thin-film μC-SiC as the emitter due to its weak absorption in the mid-infrared region of the optical spectrum, its ability to handle high temperatures and its compatibility with silicon process technology. We directly characterize our μC-SiC thin-films using ellipsometry as detailed in Appendix I. The cavity coupled thermal emitter is modeled as a single radiator with an effective emissivityε λ_eff( )corresponding to a spectral power distribution S( , )λ T of thermal light. The S( , )λ T for an emitter at temperature (T) is given by [26,29]:

(

)

Here, ( , )Θλ T is the mean energy of the Planck’s oscillator, and λ and c are the wavelength and speed of light, respectively. In order to take into account the effect of angle dependency in effective emissivity calculation, we assume a Gaussian distribution of k-vectors (k_ρ) of the thermal light inside a macroscopic cavity (l = 100 mm). By using a 1D scattering matrix approach, we predict the effective emission of the cavity-coupled-emitter by integrating the contributions from differentk_ρin the calculation ofεeff [9,30–33]. The range of kρ

considered for computing of effective emission is defined by the solid angle of emission (

14.6

θ = °) from the thin-film that is coupled into the cavity of length 100 mm. The maximum enhancement is predicted when the intrinsic losses of the system are optimally matched (impedance matched) to the transmission of the cavity. In this case, the intrinsic losses of the

system include losses at the metallic mirror and absorption losses within the cavity (including emitter absorption). The absorption losses within the cavity can be modified through appropriate choice of μC-SiC film thickness. Figure 1 shows the effect of μC-SiC emitter thickness on total emission of the coupled system. One can see that the emissivity of a bare thin-film emitter increases with thickness and eventually saturates to the bulk emissivity value of the material. In contrast, the emissivity of an emitter coupled to a cavity increases faster with thickness (t) and reaches a maximum value (ε_eff ~ 1for t ~ 2 μm). The emissivity then decreases and saturates to the bulk emissivity value, as the film thickness increases to several microns [34].

Fig. 2. Microcrystalline SiC thin-film thermal emitter. (a) Microscope image of the fabricated μC-SiC membrane with a concentric platinum (Pt) heater near its periphery (circular heater: radius = 370 μm, width = 20 μm). The membrane is heated by applying electrical power to the Pt heater which allows heating of the suspended structure while minimizing conductive losses into the substrate. (b) A three-dimensional schematic of a hot suspended SiC membrane with a color map of temperature distribution predicted using FEM based heat transfer. The suspended region of the membrane remains hot while there is a nearly step-like transition of the temperature profile from hot central region to the blue (cold) temperatures in the non-suspended region.

3. Device fabrication

In order to achieve thermal emitters with minimal conductive heat losses, we fabricate a suspended thin-film μC-SiC membrane and integrate Platinum (Pt) heaters along its circumference. The silicon carbide films are first deposited on a <100> silicon wafer through plasma enhanced chemical vapor deposition (Oxford 100) under the following process parameters: Silane (SiH4) = 4.5 sccm, Methane (CH4) = 90 sccm, Argon (Ar) = 285 sccm,

chamber pressure = 1500 mT, high frequency power 40 W, table temperature = 350 °C deposition time = 12 min. The deposited amorphous films are then annealed at 600 °C for 90 min to achieve microcrystalline-SiC (μC-SiC). The μC-SiC thin-films (t ~100 nm) are then circularly patterned (radius ~400 μm) along with concentric Pt heaters placed at its periphery. The Pt heater films are deposited via argon based sputtering. The back substrate of the patterned circular membranes is then etched (~500 μm) through the Bosch process. This is followed by xenon difluoride (XeF2) based isotropic etching to remove the remaining

substrate (~25 μm) and release the membrane. We use XeF2 based dry etching technique

because of its high selectivity between silicon and its compounds [28,35]. Post etch we ensure that the film is suspended through inspection under a microscope. Figure 2(a) shows a micrograph image of the fabricated membrane emitter where one can see the suspended region and the circular Pt heater placed at its periphery. The thickness and diameter of the membrane are approximately 100 nm and 0.8 mm, respectively. We use a thin silicon dioxide

layer (t ~40 nm) as an electrical insulator between the μC-SiC and Pt heaters to avoid unwanted electrical leakage. The suspended membrane chips with integrated Pt heaters are mounted on a carrier circuit board for electrical connections and handling purpose during measurement.

The suspended membranes show low thermal mass, effective thermal insulation and a thermal conductance of ~286 μW⋅K−1_{, much lower as compared to a bulk material. Figure}

2(b) shows the temperature distribution in the hot suspended structure estimated using a finite element model (FEM). One can see that in the region where the membrane is suspended the temperature is uniform and high, while in the region where the membrane is attached to the substrate the temperature decreases rapidly to equal that of the substrate (~293 K) [28].

Fig. 3. Measurement setup. (a) Schematic of the experimental setup. The emitter (E) and mirror (M1) form the cavity. The cold cavity is initially aligned using a 1550 nm wavelength laser. The beam splitter (BS) is used to split the input alignment laser and send it to the cavity for alignment. Lens L1 acts as the mode matching lens for the alignment laser. The mirrors M2 and M3 are used to align the beam to the monochromator (G). Lens L2 is used to focus the light to the input while L3 is used to collect the spectrally filtered light from the monochromator. The emitter is heated up by supplying electrical power using a source meter. Using lens L1 the outgoing thermal light from cavity is collected and steered to the InSb detector via the monochromator. The monochromator grating is spatially tuned to record the

spectral power distribution at the detector. (b) Measured reflectivity of the cavity as a function of laser frequency tuning (ν0 + Δν), where ν0 is the laser center frequency. Scatter dots

represent the measured data and solid line represents theoretical fit.

4. Experiment, results and discussion

Experimental setup

The experimental setup used for the measurement of thermal emission is shown in Fig. 3(a). A broadband gold mirror M1 (R ~0.99, Radius of Curvature = 100 mm), and the membrane emitter E, together form the multi-pass Fabry-Perot optical cavity (l ~100 mm). The cavity length and radius of curvature of the mirror are chosen such that the cavity stability condition is fulfilled [36,37]. Initially, the emitter is kept cold and the cavity is aligned using an external cavity tunable diode laser (New Focus Velocity 6328H, λ ~1520-1570 nm). The laser has a fine piezo tuning range of 30 GHz, enabling us to characterize the passive cavity when the SiC membrane is at room temperature. The membrane acts as the input/output coupler to the cavity when aligning to the external laser beam. The input coupling lens L1 (NA = 0.5) allows mode-matching of input alignment laser at the input node of the cavity and acts as a collection lens when the emitter is hot. Beam splitter BS, mirrors (M2, M3, M4) and lenses (L2, L3) are used to route the beam to a liquid nitrogen cooled InSb detector (New Focus 80021) via a monochromator. We use CaF2 lenses due to their low absorption in mid-IR optical spectrum.

The membrane and mirror placed over a three axis translation stage are aligned to obtain cavity resonances. During this process the frequency of the alignment laser is simultaneously tuned in order to see the cavity resonances as shown in Fig. 3(b). Once the cavity alignment is ensured, the external alignment laser source is turned off, the beam-splitter is flipped out and the membrane is gradually heated up to a temperature of ~580 K by applying electrical power (~80 mW) to the integrated platinum heaters. The temperature is estimated by measuring the change in resistance of the integrated heater and using the temperature coefficient of resistance for Pt (αPt ~0.0018 Ω K−1). The spectral distribution of emitted thermal light is

measured by sweeping the monochromator center frequency. The monochromator output slit opening is set to an equivalent spectral width of ~13 nm which averages over the discrete cavity modes (Δν ~1.5 GHz). The spectral measurements are first taken for the cavity coupled emitter and afterwards the cavity is cut off and the hemispherical emission from the bare emitter is measured. The measurements are performed by carefully maintaining the optical alignments of various experimental components. The parameters (monochromator slit width, relative position of collection lens from membrane, etc.) were kept constant between the two measurements (bare emitter & cavity-coupled emitter) to achieve an accurate measurement of relative change.

Results and discussions

We show experimentally enhanced thermal emission over a wide spectral band of 2 – 4.2 μm. The enhancement is achieved without changing the temperature of emitter. Figure 5(a) shows the schematic of the cavity (lcav ~100 mm, Fcav ~4) formed using a concave back mirror and

the μC-SiC thin-film membrane emitter as the other mirror. The emitted thermal light is collected with a lens and directed through a spectrometer to measure its spectral power distribution. We expect up to 2x power due to presence of mirror and any enhancement above this is an indication of enhanced thermal emission. Figure 4(a) shows the measured thermal emission output from the cavity coupled emitter and bare emitter. One can see that the measured spectral power distribution from the cavity-coupled emitter is significantly enhanced over the entire 1.7 – 4.2 μm spectral band. The enhancement is over threefold for thermal emission in λ ~3 ̶ 3.4 μm spectral window and over twofold for the total integrated emission power over the entire spectral band (λ ~1.7 – 4.2 μm). We observe a blue shift (ΔλPeak ~280 nm) in the peak emission wavelength of cavity-coupled emitter with respect to

impedance matching between membrane absorption and cavity losses occurs. We ensure a constant temperature of ~580 K by continuously monitoring the temperature of the membrane during the measurement. Figure 4(b) shows the input electrical power required to heat the membrane emitter for cavity coupled and bare emitter cases. The temperature of the membrane inside the thermal cavity is observed to be stable with ~2% of the bare emitter power requirement at maximum temperature, and the slightly higher temperature for the cavity-coupled emitter cannot explain the large enhancement observed in the thermal emission in Fig. 4(a). Hence, the increase in thermal emission is unambiguously due to a cavity-based enhancement effect.

Fig. 4. Measurement results of thermal output from a cavity-coupled thin-film thermal emitter. (a) Recorded spectral power distribution from the cavity coupled thermal emitter and bare emitter at the same heating power. The shaded regions around the fitted-solid lines represent the standard deviation recorded over repeated measurements, while the scatter points show a typical data set. The emission of the cavity coupled emitter is significantly enhanced relative to

the emission of a bare emitter. Inset: Ratio of recorded spectral power from the cavity coupled emitter and bare emitter showing enhancement. The spectrum shown here is normalized with respect to the emission peak of bare emitter. The measurement is performed using a liquid-nitrogen-cooled InSb detector with a roll-off a t ~4.2 μm. More than threefold enhancement is seen at short wavelengths. (b) Measured change in temperature of μC-SiC membrane emitter for applied electrical power for cavity coupled emitter and bare emitter.

We show a change in thermal emission enhancement and redistribution of thermal spectrum by introducing additional losses into the cavity using thin BK7-glass microscope slides. These glass absorbers (tslide ~180 μm) are inserted at an angle (θslide ~45°) with respect

to the optical axis of the cavity to reduce the effects of parasitic Fabry-Pèrot cavity. Figure 5(a) shows the recorded spectral power distribution from the cavity coupled thermal emitter with different amount of additional losses. The addition of losses modifies the impedance matching conditions between the absorption and roundtrip losses of the bare cavity leading to a change in enhancement. Introducing room-temperature losses contribute negligibly to the overall emission from the cavity, as the reemission process, according to Kirchhoff’s law, occurs at the temperature of the absorbing body under thermal equilibrium. We predict this spectral tuning of thermal emission using our computational model where we consider the effect of additional cavity loss by lumping it with the back mirror reflectivity (Rmirror) and

creating an effective back mirror reflectivity (R'mirror). The predicted enhancement and relative

peak emission wavelength as a function of effective back mirror reflectivity are shown in Fig. 5(b)-bottom and 5(b)-top, respectively. The vertical guide lines represent the estimated cavity losses used in the measurement. These points (ii) and (iii) on the plot correspond to the added loss from 2 and 4 glass slides respectively. The losses added to the cavity during the measurement are estimated using the optical constants (n, k) of BK7 glass at an incident angle of

45

mirror = Rmirror⋅ (1 ̶ TBK7⋅TBK7)).

Here, Rmirror is the initial back mirror reflectivity (Rmirror = 0.99) and TBK7 is the transmission

through the glass-slide. We consider mean transmission through the glass slide over the measurement spectrum to estimate the effective back mirror reflectivity. The total power enhancement from the cavity without losses (i) and that from bare emitter (iv) is also presented. The shift in emission wavelength is shown with respect to the bare emission peak. The deviation in peak maxima in Fig. 5(b)-top as compared to the experimental measurements is believed to be due to optical alignment variations due to beam divergence over a wide spectral band and the wavelength-dependent loss and dispersion of the BK7 glass.

Here we have experimentally demonstrated the enhancement of thermal radiation using a thin-film emitter coupled to an external cavity. Since it is generally difficult to measure the absolute value of emission intensity our measurement results present a relative enhancement of thermal radiation. The choice of emitter thickness is made to convincingly show a sufficient enhancement of thermal light above bare emitter as predicted from our computations. A thicker silicon carbide emitter (t ~2 μm) is predicted to emit nearly equal to an ideal blackbody as shown in Fig. 1b., however the relative enhancement over the bare emitter is lesser than a factor of 2.

Our method uses an impedance matched external optical cavity physically detached from the thermal element to show enhancement of thermal emission unlike earlier demonstrations using integrated cavities. This physical separation or decoupling of the cavity and the heating element is a key concept of this work since it provides a route to use this technique for a wide variety of materials. The optical cavity need not be as big as the one shown here and the entire system can be carefully micro-integrated while still keeping the optics and the heating element separate. The approach allows stable performance of optical elements without any change in temperature due to conduction of heat [13,20,21], with better thermal response times and lower heating powers than the bulkier thermal elements used in previous demonstrations. This method can be leveraged to make blackbody type thermal sources from

any material irrespective of its intrinsic emissivity, through proper design of the thermal cavity.

Fig. 5. Effect of impedance matching conditions on the thermal output of cavity-coupled emitter. (a) Recorded spectral power distribution from the cavity coupled thermal emitter with different amount of added losses. The addition of excess losses to the cavity modifies the impedance matching conditions leading to flat enhancement across the entire band. The spectrum shown here is normalized with respect to the bare emitter emission peak. (b) Top - Predicted relative shift in maximum emission wavelength with respect to bare emitter peak, due to modification of cavity conditions. The vertical guide lines represent the measurement with respect to the theoretically predicted data set. Bottom - Predicted thermal enhancement as a function of back mirror reflectivity. The reduction of effective back mirror reflectivity can be seen as addition of cold-losses to the cavity.

Appendix

Ellipsometry measurements for microcrystalline Silicon Carbide

The microcrystalline silicon carbide films are deposited on a silicon substrate and characterized using mid-infrared ellipsometry for three different grazing angles (θgrazing = 20°,

30°,40°). The complex refractive index of μC-SiC is found to agree with the Drude-Lorentz relation described below.

2 2 2 2 2 2 ( ) ( ) 1 ( ) ( ) L T D T D n i i ω ω ω ω ε ω ω ω ω ω ∞  −  =  + _{− − Γ} − _{+ Γ}     (2) where, ε∞ = 3.455, ωT = 971.92 cm−1, ωL = 781.48 cm−1, Γ = 40.80 cm−1, ωD = 2885.61 cm−1,

ΓD = 16347.82 cm−1. Figure 6 shows the real and imaginary part of the refractive index of SiC

in mid-IR generated using the Drude-Lorentz model of Eq. (2). Our thin film with thickness t ~100 nm show an absorptivity of ~10% at the short-wavelength mid-IR (λ ~2-5 μm) range of optical spectrum. The weak absorption in mid-infrared for our films is suspected from to the crystal boundaries in an otherwise transparent crystalline silicon carbide.

Fig. 6. Real part of refractive index (nr) and imaginary part of refractive index (ki) for SiC in

mid-IR region of the spectrum, measured using ellipsometry. Complex refractive index is given by n = nr + jki

Funding

ARPA-E IDEAS program (# DE-AR0000731), support from Brazilian agencies CAPES and FAPESP. Device fabrication & thin-film characterization at Cornell Nano-Scale Facility (NSF, # NNCI-1542081) and Cornell Center for Materials Research (NSF MRSEC, # DMR-1719875).

Acknowledgment

We acknowledge fruitful discussions with Prof. S. Fan from Stanford University, Prof. J. Cardenas from University of Rochester and, I. Datta, Dr. A. Mohanty, Dr. Y Chang and Prof. R. Opher at Columbia University. We would also like to acknowledge Prof. A. Gaeta and Dr. Y. Okawachi at Columbia University for providing the measurement facilities. We acknowledge the facilities at the Advanced Semiconductor Research Facility at City University of New York and facilities at Columbia Nano-initiative.

Disclosures

Authors declare no competing interests.

References

1. H. T. Miyazaki, T. Kasaya, M. Iwanaga, B. Choi, Y. Sugimoto, and K. Sakoda, “Dual-band infrared metasurface thermal emitter for CO_2 sensing,” Appl. Phys. Lett. 105(12), 121107 (2014).

2. O. Ilic, P. Bermel, G. Chen, J. D. Joannopoulos, I. Celanovic, and M. Soljačić, “Tailoring high-temperature radiation and the resurrection of the incandescent source,” Nat. Nanotechnol. 11(4), 320–324 (2016). 3. M. M. Hossain, B. Jia, and M. Gu, “A metamaterial emitter for highly efficient radiative cooling,” Adv. Opt.

Mater. 3(8), 1047–1051 (2015).

4. A. Lenert, D. M. Bierman, Y. Nam, W. R. Chan, I. Celanović, M. Soljačić, and E. N. Wang, “A nanophotonic solar thermophotovoltaic device,” Nat. Nanotechnol. 9(2), 126–130 (2014).

5. J. Hodgkinson, R. Smith, W. O. Ho, J. R. Saffell, and R. P. Tatam, “Non-dispersive infra-red (NDIR)

measurement of carbon dioxide at 4.2 μm in a compact and optically efficient sensor,” Sens. Actuators B Chem.

186, 580–588 (2013).

6. S. Edalatpour and M. Francoeur, “Size effect on the emissivity of thin films,” J. Quant. Spectrosc. Radiat. Transf. 118, 75–85 (2013).

7. Y. A. Çengel, Heat Transfer: A Practical Approach (McGraw-Hill, 2004).

8. I. Celanovic, D. Perreault, and J. Kassakian, “Resonant-cavity enhanced thermal emission,” Phys. Rev. B Condens. Matter Mater. Phys. 72(7), 075127 (2005).

9. L. P. Wang, B. J. Lee, X. J. Wang, and Z. M. Zhang, “Spatial and temporal coherence of thermal radiation in asymmetric Fabry–Perot resonance cavities,” Int. J. Heat Mass Transfer 52(13–14), 3024–3031 (2009). 10. J.-J. Greffet, R. Carminati, K. Joulain, J.-P. Mulet, S. Mainguy, and Y. Chen, “Coherent emission of light by

thermal sources,” Nature 416(6876), 61–64 (2002).

11. M. Ghebrebrhan, P. Bermel, Y. X. Yeng, I. Celanovic, M. Soljačić, and J. D. Joannopoulos, “Tailoring thermal emission via Q matching of photonic crystal resonances,” Phys. Rev. A 83(3), 033810 (2011).

12. S. E. Han, A. Stein, and D. J. Norris, “Tailoring Self-Assembled Metallic Photonic crystals for modified thermal emission,” Phys. Rev. Lett. 99(5), 053906 (2007).

13. M. De Zoysa, T. Asano, K. Mochizuki, A. Oskooi, T. Inoue, and S. Noda, “Conversion of broadband to narrowband thermal emission through energy recycling,” Nat. Photonics 6(8), 535–539 (2012).

14. M.-W. Tsai, T.-H. Chuang, C.-Y. Meng, Y.-T. Chang, and S.-C. Lee, “High performance midinfrared narrow-band plasmonic thermal emitter,” Appl. Phys. Lett. 89(17), 173116 (2006).

15. T. Inoue, M. De Zoysa, T. Asano, and S. Noda, “Realization of dynamic thermal emission control,” Nat. Mater.

13(10), 928–931 (2014).

16. Z. Yu, N. P. Sergeant, T. Skauli, G. Zhang, H. Wang, and S. Fan, “Enhancing far-field thermal emission with thermal extraction,” Nat. Commun. 4(1), 1730 (2013).

17. K. Ikeda, H. T. Miyazaki, T. Kasaya, K. Yamamoto, Y. Inoue, K. Fujimura, T. Kanakugi, M. Okada, K. Hatade, and S. Kitagawa, “Controlled thermal emission of polarized infrared waves from arrayed plasmon nanocavities,” Appl. Phys. Lett. 92(2), 021117 (2008).

18. M. U. Pralle, N. Moelders, M. P. McNeal, I. Puscasu, A. C. Greenwald, J. T. Daly, E. A. Johnson, T. George, D. S. Choi, I. El-Kady, and R. Biswas, “Photonic crystal enhanced narrow-band infrared emitters,” Appl. Phys. Lett. 81(25), 4685–4687 (2002).

19. J. A. Mason, S. Smith, and D. Wasserman, “Strong absorption and selective thermal emission from a midinfrared metamaterial,” Appl. Phys. Lett. 98(24), 241105 (2011).

20. I. Celanovic, N. Jovanovic, and J. Kassakian, “Two-dimensional tungsten photonic crystals as selective thermal emitters,” Appl. Phys. Lett. 92(19), 193101 (2008).

21. C. Schlemmer, J. Aschaber, V. Boerner, and J. Luther, “Thermal stability of micro‐structured selective tungsten emitters,” AIP Conf. Proc. 653(1), 164–173 (2003).

22. P. N. Dyachenko, S. Molesky, A. Y. Petrov, M. Störmer, T. Krekeler, S. Lang, M. Ritter, Z. Jacob, and M. Eich, “Controlling thermal emission with refractory epsilon-near-zero metamaterials via topological transitions,” Nat. Commun. 7(1), 11809 (2016).

23. P. J. Hesketh, B. Gebhart, and J. N. Zemel, “Measurements of the spectral and directional emission from microgrooved silicon surfaces,” J. Heat Transfer 110(3), 680–686 (1988).

24. E. Hasman, V. Kleiner, N. Dahan, Y. Gorodetski, K. Frischwasser, and I. Balin, “Manipulation of thermal emission by use of micro and nanoscale structures,” J. Heat Transfer 134(3), 031023 (2012).

25. L. P. Wang, S. Basu, and Z. M. Zhang, “Direct measurement of thermal emission from a Fabry–Perot cavity resonator,” J. Heat Transfer 134(7), 072701 (2012).

26. G. Kirchhoff, “Ueber das verhältniss zwischen dem emissionsvermögen und dem absorptionsvermögen der körper für wärme und licht,” Ann. Phys. 185(2), 275–301 (1860).

27. J.-J. Greffet, P. Bouchon, G. Brucoli, E. Sakat, and F. Marquier, “Generalized Kirchhoff law,” Phys. Rev. X

8(2), 021008 (2018).

28. R. St-Gelais, L. Zhu, S. Fan, and M. Lipson, “Near-field radiative heat transfer between parallel structures in the deep subwavelength regime,” Nat. Nanotechnol. 11(6), 515–519 (2016).

29. M. Planck and M. Masius, The Theory of Heat Radiation (P. Blakiston’s Son & Co., n.d.).

30. P. Pigeat, D. Rouxel, and B. Weber, “Calculation of thermal emissivity for thin films by a direct method,” Phys. Rev. B Condens. Matter Mater. Phys. 57(15), 9293–9300 (1998).

31. C. M. Cornelius and J. P. Dowling, “Modification of Planck blackbody radiation by photonic band-gap structures,” Phys. Rev. A 59(6), 4736–4746 (1999).

33. M. Francoeur, M. Pinar Mengüç, and R. Vaillon, “Solution of near-field thermal radiation in one-dimensional layered media using dyadic Green’s functions and the scattering matrix method,” J. Quant. Spectrosc. Radiat. Transf. 110(18), 2002–2018 (2009).

34. H. O. McMahon, “Thermal radiation from partially transparent reflecting bodies,” J. Opt. Soc. Am. 40(6), 376– 380 (1950).

35. H. F. Winters and J. W. Coburn, “The etching of silicon with XeF2 vapor,” Appl. Phys. Lett. 34(1), 70–73 (1979).

36. A. Yariv and P. Yeh, Photonics (Oxford University, 2006).