Fundamentals of Thermophotovoltaic Energy Conversion by Donald Chubb - Read Online
Fundamentals of Thermophotovoltaic Energy Conversion
0% of Fundamentals of Thermophotovoltaic Energy Conversion completed



This is a text book presenting the fundamentals of thermophotovoltaic(TPV) energy conversion suitable for an upper undergraduate or first year graduate course. In addition it can serve as a reference or design aid for engineers developing TPV systems. Each chapter includes a summary and concludes with a set of problems.

The first chapter presents the electromagnetic theory and radiation transfer theory necessary to calculate the optical properties of the components in a TPV optical cavity. Using a simplified model, Chapter 2 develops expressions for the maximum efficiency and power density for an ideal TPV system. The next three chapters consider the three major components in a TPV system; the emitter, filter and photovoltaic(PV) array. Chapter 3 applies the electromagnetic theory and radiation transfer theory presented in Chapter 1 in the calculation of spectral emittance. From the spectral emittance the emitter efficiency is calculated. Chapter 4 discusses interference, plasma and resonant array filters plus an interference filter with an imbedded metallic layer, a combined interference-plasma filter and spectral control using a back surface reflector(BSR) on the PV array. The theory necessary to calculate the optical properties of these filters is presented. Chapter 5 presents the fundamentals of semiconductor PV cells. Using transport equations calculation of the current-voltage relation for a PV cell is carried out. Quantum efficiency, spectral response and the electrical equivalent circuit for a PV cell are introduced so that the PV cell efficiency and power output can be calculated.

The final three chapters of the book consider the combination of the emitter, filter and PV array that make up the optical cavity of a TPV system. Chapter 6 applies radiation transfer theory to calculate the cavity efficiency of planar and cylindrical optical cavities. Also introduced in Chapter 6 are the overall TPV efficiency, thermal efficiency and PV efficiency. Leakage of radiation out of the optical cavity results in a significant loss in TPV efficiency. Chapter 7 considers that topic. The final chapter presents a model for a planar TPV system.

Six appendices present background information necessary to carry out theoretical developments in the text. Two of the appendices include Mathematica programs for the spectral optical properties of multi-layer interference filters and a planar TPV system. Software is included for downloading all the programs within the book. First text written on thermophotovoltaic(TPV) energy conversion Includes all the necessary theory to calculate TPV system performance Author has been doing TPV energy conversion research since 1980's Emphasizes the fundamentals of TPV energy conversion Includes a summary and problem set at the end of each chapter Includes Mathematica programs for calculating optical properties of interference filters and planar TPV system performance solution software
Published: Elsevier Science an imprint of Elsevier Books Reference on
ISBN: 9780080560687
List price: $146.00
Availability for Fundamentals of Thermophotovoltaic Energy Conversion
With a 30 day free trial you can read online for free
  1. This book can be read on up to 6 mobile devices.


Book Preview

Fundamentals of Thermophotovoltaic Energy Conversion - Donald Chubb

You've reached the end of this preview. Sign up to read more!
Page 1 of 1



Although energy conversion is a technology that is vital for modern life, few textbooks have been written on the fundamentals of the various energy conversion systems. Thermophotovoltaics (TPV) is a simple energy conversion concept well suited for description in a fundamental text. TPV is a static conversion system with no moving parts and only three major parts: an emitter heated by a thermal energy source, and optical cavity for spectral control, and a photovoltaic (PV) array for converting the emitted radiation to electricity.

The book has been written as a text rather than as a review of current TPV research. However, there is some mention of that research.

Most of the material is introduced at the level where physics becomes engineering. For example, radiation transfer theory, which can be considered an engineering discipline, is used throughout the book. The theory is based upon the radiation transfer equation, which originates from basic physics. All the theoretical developments are as self-contained as possible.

The text begins with a chapter that introduces several topics from electromagnetic wave propagation and radiation transfer theory. This is the background material necessary to describe the emitter and optical cavity of a TPV system. Chapter 2 uses a simplified model of a TPV system to illustrate several important properties of the TPV energy conversion concept. In the next three chapters, the theory necessary to describe the performance of emitters, optical filters used for spectral control, and photovoltaic arrays is presented. The final three chapters deal with the performance of the whole TPV system.

Five appendices are included. They provide background material for the main text. Also, Mathematica computer programs for calculating the optical properties of interference filters and for calculating the performance of a planar TPV system are included on a CD-ROM disk. Problem sets are included at the end of each chapter.

Chapter 1


The opening chapter defines the thermophotovoltaic (TPV) energy conversion process, presents a short history of TPV research with possible TPV applications and introduces several concepts from electromagnetic wave propagation and radiation transfer theory. These concepts will be used throughout the text.

1.1 Symbols


1.2 Thermphotovoltaic (TPV) Energy Conversion Concept

Similar to all energy conversion concepts, TPV energy conversion is a method for converting thermal energy to electrical energy. The concept is illustrated in Figure 1.1. Thermal energy from any of the thermal sources shown in Figure 1.1 is supplied to an emitter. Radiation from the emitter is directed to photovoltaic (PV) cells where the radiation is converted to electrical energy. In order to make the process efficient, the energy of the photons reaching the PV array must be greater than the bandgap energy of the PV cells. Shaping of the radiation or spectral control is accomplished in the following ways. One method is to use a selective emitter that has large emittance for photon energies above the bandgap energy of the PV cells and small emittance for photon energies less than the bandgap energy. Similar results can be accomplished by using a gray body emitter (constant emittance) and a band pass filter. The bandpass filter should have large transmittance for photon energies above the PV cells bandgap energy and large reflectance for photon energies below the PV cells bandgap energy. A back surface reflector on the PV array can also be used to reflect the low energy photons back to the emitter. It is also possible to combine all three of these methods.

Figure 1.1 Thermophotovoltaic (TPV) energy conversion concept.

1.3 A Short History of TPV Energy Conversion

Robert E. Nelson, who has done significant early research on rare earth selective emitters for TPV and is familiar with early TPV work, related the following story about the invention of TPV energy conversion. In the mid-1950’s, Henry H. Kolm at Massachusetts Institute of Technology’s (MIT) Lincoln Laboratory, received a phone call from someone in the defense department who informed Kolm that Russians had developed a method for generating electricity from a lantern flame. He then asked Kolm if he could do the same thing. Kolm’s solution was to place a silicon solar cell, which was in the earliest stage of development at Bell Laboratories, next to the mantel of a Coleman Lantern. He measured the electrical power output and published the result in a Lincoln Laboratory progress report [ 1] in May 1956. The irony is that the Russian method was thermoelectric, not TPV.

During a technical meeting in Europe in 1960, Kolm informed Pierre Aigrain of France of his experiment. Aigrain, who was a science advisor to Charles de Gaulle, immediately began to work on the concept. Aigrain is generally considered to be the first to have presented the concept in a series of lectures he gave at MIT in 1960. Early papers by D.C. White, B.D. Wedlock and J. Blair [ 2] and B.D. Wedlock [ 3] and [ 4] present the earliest theoretical and experimental TPV research results. The first term used to describe the concept was thermal-photo-voltaic, which then became thermo-photo-voltaic, and finally, thermophotovoltaic. General Motors (GM) also had an early interest in TPV.

At GM, TPV work was also begun in 1960 [5] through [7]. In 1963 Werth [ 6] demonstrated TPV conversion using a propane-heated emitter and a germanium (Ge) PV cell. The emitter temperature was approximately 1700 K. Shockley and Queisser [ 8] had shown that for a blackbody radiation source, the efficiency of a PV cell was strongly dependent upon the ratio of the cell bandgap energy, Eg, to the radiation source temperature, TE. In fact, Eg/kTE ≈ 2.0 to obtain maximum efficiency for a blackbody source. In order to obtain maximum efficiency using silicon (Si), which was the most efficient PV cell available, the emitter temperature would have to be approximately 6000 K.

No workable materials exist at these temperatures. Therefore, for efficient TPV energy conversion at reasonable temperatures, a blackbody source cannot be used. As a result, TPV research has been directed at tailoring the emitter radiation to the PV cell bandgap energy. In other words, by producing radiation within a narrow energy band that lies just above the bandgap energy of the PV cell (where PV cell is most efficient), it is possible to have an efficient system. In addition, just as for the blackbody emitter, there will be an optimum value for Eg/kTE. For reasonable emitter temperatures (TE ≤ 2000 K) achieving this optimum condition requires that Eg ≤ 1.0 eV. Therefore, low bandgap energy PV cells are required for high efficiency in addition to radiation matched to the PV cell bandgap energy.

The early TPV research of the 1960’s and 1970’s was confined to using either silicon (Eg = 1.12 eV) or germanium (Eg = 0.66 eV) PV cells. However, silicon requires a large emitter temperature (TE ≈ 2000 K) for an efficient TPV system, and Ge cells were of low efficiency, although having reasonable bandgap energy for TPV conversion. Thus, lack of suitable PV cells plus unsuccessful attempts at obtaining radiation matched to the PV cell bandgap energy resulted in a loss of interest in TPV energy conversion. However, in the 1980’s new efficient PV cell materials such as gallium antimonide (GaSb, Eg = 0.72 eV) and indium gallium arsenide (InxGa1-xAs, where 0.36 ≤ Eg ≤ 1.42 eV depending on value of x), became available. In addition, new selective emitters and filters that can produce the bandgap matched radiation were being developed. As a result, beginning in the late 1980’s, a resurgence of interest in TPV energy conversion occurred.

An extensive bibliography of all TPV research papers has been complied by Lars Browman [ 9]. Other excellent references for TPV research results are the proceedings of the National Renewable Energy Laboratory (NREL) Conferences on Thermophotovoltaic Generation of Electricity.

1.4 TPV Applications

Besides the development of new PV cell and emitter technology, the many applications for TPV energy conversion are also a driving force in the resurgence of TPV interest. The simplicity and potential high efficiency of TPV conversion are the two attractive features that lead to many potential applications. Since TPV is a direct energy conversion process, the only moving parts in the system are fans or pumps that may be used for cooling the PV cells. The components of the system are the thermal source, the emitter (and possibly a filter), the PV cells, and the waste heat rejection system. Each of these components is in the solid state with only the PV cells and possibly the filter being a somewhat complex solid state device. In addition to simplicity and potential high conversion efficiency, TPV can be easily coupled to any thermal source.

Applications for TPV exist where the thermal source may be solar, nuclear, or combustion. At first it would appear that a solar TPV (STPV) system would have no advantage over a conventional solar PV system. What can be gained by adding an emitter and possibly a filter to the conventional solar PV system? By adding a selective emitter or thermal emitter plus filter, the solar spectrum, which corresponds approximately to a 6000 K gray body emitter, can be shifted to match the bandgap energy of the PV cells. As a result, the STPV system will yield higher efficiency than the conventional solar PV systems. Besides efficiency, an even more important reason for interest in STPV energy conversion is the ability to use thermal energy storage and combustion energy input so that the system can operate when the sun is not available. With the capability of 24-hour-a-day operation, STPV is viable for electric utility use.

A TPV system powered by radioisotope decay (RTPV) is a potential power system for deep space missions where the solar radiation energy density is too low for a conventional PV power system to be used. The first radioisotope decay power systems used thermoelectric energy converters (RTG). However, TPV has the potential for higher efficiency than thermoelectrics; therefore it is being considered as a replacement.

Combustion driven TPV has many potential commercial applications. For natural gas-fired appliances such as furnaces and hot water tanks, TPV can be added for the cogeneration of electricity. In such applications, attaining high TPV efficiency is not essential since the waste heat for the TPV conversion process is completely utilized. Portable power supplies for both commercial and military use is another important combustion TPV application. An important TPV advantage over existing internal combustion-driven applications is quiet operation. This is especially true in military missions that require the mission to be undetected. Another combustion-driven TPV application with commercial potential is the power supply for hybrid electric vehicles. In such an application the TPV system is sized to provide enough power for operation at cruise speed and battery charging. For acceleration power the batteries are utilized.

One important advantage of all TPV applications is that they are environmentally benign. The TPV system is nearly silent and emits no pollutants. This is obvious in the case of a solar driven system. For a nuclear powered system there are no combustion products; however, care must be taken to insure that no radioactive material is released. Atmospheric pressure burning occurs in the combustion-driven TPV systems. Therefore, the combustion temperature can be well controlled so that the production of toxic nitrous oxides (NOX) is small.

Although there appears to be unlimited potential TPV applications, whether or not they are feasible, will depend upon their cost. At this stage of development, it is impossible to say which applications will be cost effective.

1.5 Propagation of Electromagnetic Waves

Energy transfer between the components of a TPV system is mainly by radiation that consists of propagating electromagnetic waves. The derivation of the equation describing radiation transfer does not require electromagnetic wave propagation theory. However, electromagnetic wave theory is required to obtain equations for the optical properties reflectivity and transmissivity at an interface. Values for these properties are required in analyzing a TPV system. Also, wave propagation must be considered in analyzing the various optical filters that can be used in a TPV system.

1.5.1 Plane Wave Solution to Maxwell’s Equations

Much of the material to be covered in this section can be found in the classic optics text by Born and Wolf [ , are the following [ 11],





the magnetic induction and ρ is the electric charge density. In addition to Maxwell’s Equations so-called constitutive relations for the media are required.




Appearing in result, where εo is the vacuum permittivity, χ is the electric susceptibility and ε = εo + χ is the electric permittivity. in equation (1.6). Here μo is the vacuum permeability, χm is the magnetic susceptibility, and μ = μo(1 + χm) is the magnetic permeability. The properties of the medium (ε, μ, and σ) appearing in equations (1.5) to (1.7) are scalar quantities and therefore independent of direction. Thus the media is called isotropic. so the media is called a linear isotropic media. can appear as higher order terms in the constitutive relations so that the media may be nonlinear. Finally, for plane electromagnetic waves the properties must be independent of time (stationary media) and space (homogeneous media).

The quantum mechanical theory necessary to determine the medium properties (ε, μ, σ) is beyond the scope of this text. However, the discussion of optical filters in Chapter 4 uses a macroscopic model called the Drude model to calculate ε.

is obtained by first taking the curl (∇×) of , where ∇² = ∇•∇. Then using equations (1.1), (1.3), and the constitutive relations, equations (1.5) to (1.7), the following result is obtained.


So far, the only approximation that has been made is that the media is linear and isotropic. Now assume the media is stationary so that properties are independent of time, t, and that μ = constant. Therefore, equation (1.8) becomes the following.


Solution of , or frequency, ω, appears in equation (1.9). Experimentally it has been shown that light has wave-like behavior. Therefore, a wave solution should satisfy equation (1.9) if Maxwell’s equations apply to electromagnetic waves of all frequencies including light. The simplest wave solution is the so-called harmonic plane wave,


or using complex notation,



is a constant and can be complex and and t. In using complex notation, equation (1.10b), care must be exercised when considering products, since Re[A]Re[B] ≠ Re[AB].

will remain a constant in a plane where the phase of the wave, remains a constant. Thus, equation (1.10) is called a plane wave. direction) and has the magnitude, vφ, which can be derived as follows.




and since the phase velocity is ,


the index of refraction, n, is defined as follows,


where co is the speed of light in vacuum (3 × 10⁸ m/sec) and λ = 2πco/ω is the wavelength in vacuum. As will be shown shortly,


where εo is the vacuum permittivity and μo is the vacuum permeability. Obviously, if k is complex, n will also be complex. It should be noted that if k and n are complex, then the phase velocity, which is always real, is no longer related to n and k by equation (1.13).

Now consider how k and Ω are related to the media properties for plane waves. This relation, called the dispersion relation, is obtained by substituting and ∂/∂t = jΩ that apply for plane waves ( problem 1.1) the following result is obtained.


For the vector must be perpendicular. This same result for plane waves can be obtained using equations (1.1), (1.5) and (1.7) ( problem 1.2). In that case equation (1.15) becomes the following.


Several important conclusions can be drawn from equation (1.16). First, since k and ω are constants for plane waves, the medium properties must also be constants for the plane wave solution to apply. Therefore, for a linear isotropic medium, plane waves only apply when the medium is also homogeneous and stationary as stated earlier. However, if the medium is linear, stationary, homogeneous but anisotropic it is still possible to have plane wave solutions [ 12]. The second conclusion drawn from equation (1.16) is that k must be complex for a conductive (σ ≠ 0) medium. However, if σ=0 (dielectric| decays rather than grows exponentially.

= 0, equation (1.9) becomes the following


( wave equation.

If a complex wave vector and index of refraction is defined as follows,


where ŝ is a real unit vector pointing in the direction of wave propagation, then the dispersion relation, | decay with distance. Substituting equation (1.18) in equation (1.16) and equating real and imaginary parts yields the following.






Obviously, for a dielectric (σ = 0), kI = 0 and





where εr = ε/εo is the dielectric constant and μr = μ/μo is the relative permeability. . The fact that k has no imaginary part if σ = 0 seems to contradict the discussion proceeding is real but the unit vector, ŝ, can be complex. As already stated, this is the case for total internal reflection at a boundary between two dielectrics. This will be discussed in detail in Section 1.5.4.

The wave number, k, and therefore the index of refraction, n, are given in terms of the real properties, μ, ε, and σ by equations (1.21) and (1.22). However, referring to the dispersion relation, equation (1.16), it yields,



if a complex dielectric constant is defined as follows,




and the imaginary part,


. Thus for a plane wave in a nonmagnetic material . This is convenient since the theoretical models for the electrical properties in time-varying fields, such as the Drude model ([ 13], pg. 225-226), lead to a complex dielectric constant that conforms to equation (1.27).

, equation (1.17), reduces to the standard wave equation.


Using .



material, equations (1.30) become



When k has an imaginary part then the plane wave will be attenuated in the direction of kI. This can be seen by considering the exponential term in equation (1.10b) ( problem 1.5).

have been discussed. There are several other plane wave properties that will be used in later chapters.

1.5.2 Energy Flux for Plane Electromagnetic Waves

, equation (1.17), it is also given by a plane wave solution.


Therefore, from equations (1.4) and (1.6) for plane waves in a homogeneous media,


are perpendicular. Also, from equation (1.2)




The quantity, Y, is called the optical admittance ( . Such a wave is called transverse. , the wave is called longitudinal.

Figure 1.2 Mutually orthogonal system for plane electromagnetic wave.

The energy flux associated with plane waves is given by the Poynting vector.


] must be used in (=I) is the energy flux (W/m²) of the electromagnetic (or radiation) field at a particular frequency, ω (or wavelength, λ). In radiation transfer theory the intensity, i, is the energy flux (W/m²) per solid angle, ω, per frequency, ω, (or wavelength, λ)


are perpendicular and their magnitudes are related by direction equation (1.37) becomes the following.


It can be shown ( problem 1.7) that equation (1.37) becomes the following,


where * denotes the complex conjugate and nR is the real part of n. The last two expressions in has real and imaginary parts and is given by equations (1.18), (1.21), and