Professional Documents
Culture Documents
Thermoelectric Phenomena, Materials, and Applications: Further
Thermoelectric Phenomena, Materials, and Applications: Further
ANNUAL
REVIEWS Further
Thermoelectric Phenomena,
Click here for quick links to
Annual Reviews content online,
including:
Materials, and Applications
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
• Our comprehensive search Department of Physics and Astronomy, Clemson University, Clemson, South Carolina
29634-0978; email: ttritt@clemson.edu
433
MR41CH17-Tritt ARI 3 June 2011 8:42
generally articles on the theme of “Harvesting Energy Through Thermoelectrics: Power Gener-
ation and Cooling” in the MRS Bulletin, Volume 31 (2006)]. Some of the research efforts focus on
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
minimizing the lattice thermal conductivity, whereas other efforts focus on materials that exhibit
high power factors, which are discussed below.
Proposed industrial and military applications of TE materials are generating increased activity
in this field by demanding higher-performance, higher-temperature TE materials compared with
those that are presently in use. The demand for alternative energy technologies to reduce our
dependency on fossil fuels has led to various directions of research, including high-temperature
energy harvesting via the direct recovery of waste heat and its conversion into useful electrical
energy using TE devices. More recently, TEs have been investigated for their use in TE-solar
hybrid systems to take advantage of the IR part of the solar spectrum and for TE-solar power
generation (7). Thus, the development of higher-performance TE materials is becoming more
important. Novel applications of TEs include aspects such as biothermal batteries to power heart
pacemakers, enhanced performance of optoelectronics coupled with solid-state TE cooling, and
power provision for deep-space probes via radioisotope TE generators. More recently, TE gener-
ators have been installed in automobiles to capture waste heat from exhaust and to transform it into
useful electrical energy for automotive electrical systems and for increased fuel efficiency for use in
next-generation vehicles. In addition, TE refrigeration applications include seat coolers for com-
fort and electronic component cooling. Of course, the deep-space applications of NASA’s Voyager
and Cassini missions using radioactive thermoelectric generators (RTGs) are well established (8).
TE refrigeration is an environmentally green method of small-scale, localized cooling in com-
puters, infrared detectors, electronics and optoelectronics, and many other applications. However,
most of the electronics and optoelectronics technologies typically require only small-scale or local-
ized spot cooling of small components that do not impose a large heat load. If significant economical
cooling can be achieved, the resulting cold computing could produce speed gains of 30% to 200%
in some CMOS computer processors. Many researchers perceive cooling of the processors to be
the fundamental limit to electronic system performance (9). Recent utilization of Peltier coolers
in relation to refrigeration of biological specimens/samples is also an emerging TE application.
Thus, the potential payoff for the development of low-temperature TE refrigeration devices, as
well as for the development of high-temperature materials for waste heat recovery, is great. The
requirement for compounds with high-performance properties optimized over wide temperature
ranges has led to this much-expanded interest in new TE materials. The role of bulk materials
applications and needs is demanding new breakthroughs in both materials engineering and device
engineering, especially for automotive applications for cooling and power generation (8).
The development and potential of bulk materials for TE applications include high-temperature
materials such as skutterudites, clathrates, half-Heusler alloys, and complex chalcogenides (10).
434 Tritt
MR41CH17-Tritt ARI 3 June 2011 8:42
Recently, ceramic oxide materials have also shown potential as high-temperature TE materials
(11). The potential of nanomaterials and their role in TE research are emerging areas of interest
(12). Another emerging area is one that utilizes materials termed TE nanocomposites, which incor-
porate nanomaterials that are then part of a matrix of a given TE material. These nanoparticles may
be mixed into materials and solidified by some method such as hot pressing or spark plasma sinter-
ing. Such nanoparticles can also be nucleated within the material directly through either specific
solid-state synthesis or more external methods such as high-energy ball milling or melt spinning
(13–16). Thin-film properties, applications, and recent results are also very important (17).
A key factor in advancing many of these technologies is the development of higher-performance
TE materials via either the emergence of completely new materials or a more ingenious materials
engineering of some of the existing materials. Recent books and series provide a more complete
overview of the state-of-the-art materials, a theoretical and experimental discussion of the basic
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
principles, and an overview of some of the recent developments and materials (2, 18).
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
E = αn ∇T
. 1.
The Seebeck coefficient is very low for metals (only a few microvolts per degree Kelvin) and is
much higher for semiconductors (typically a few hundred microvolts per degree Kelvin) (21). A
few years after Seebeck’s observation, Peltier discovered a related effect (the Peltier effect): If an
electrical current is passed through the junction of two dissimilar materials, heat is either absorbed
or rejected at the junction, depending on the current’s direction. This effect is due largely to the
difference in the Fermi energies of the two materials. These two effects are related to each other,
as shown in the definition of the Peltier coefficient, Π = αT. The rate that the Peltier heat is
liberated or rejected at the junction ( Q̇ P ) is given by Q̇ P = α I T , where I is the current through
the junction and T is the temperature in degrees Kelvin. Thermomagnetic effects such as the Hall,
Ettingshausen, and Nernst effects are beyond the scope of this article (see Nolas et al. 18 for a
more complete discussion of these effects).
α2 σ T α2 T
ZT = = , 2.
κ ρκ
TC T TH
Δ
TC T TH
Δ
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
b – +
+
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
– + + + + + + + +
– + + + + + + + + +
+ + + + + + + + +
–
E
Figure 1
A depiction of the thermoelectric effect when a material is subjected to an external temperature gradient and
the system is in equilibrium. (a) The more energetic electrons have a longer mean free path. (b) These more
energetic electrons (denoted by blue dots) then diffuse to the cold side until an electric field (E) is developed to
oppose any further diffusion. We define x as increasing from the cold-side temperature TC to the hot-side
temperature TH , therefore defining the temperature gradient ∇T as shown. Because the electrons are
negatively charged (and therefore α n is negative), an electric field is established that is the opposite of that of
the temperature gradient to oppose further diffusion when the system is in equilibrium.
where α is the Seebeck coefficient, σ the electrical conductivity, ρ the electrical resistivity, and
κ the total thermal conductivity (κ = κ L + κ E , where κ L and κ E are the lattice and electronic
contributions, respectively). The power factor, α 2 σ T (or α 2 T/ρ), is typically optimized through
doping in narrow-gap semiconducting materials as a function of carrier concentration (typically
approximately 1019 carriers cm−3 ) to give the highest ZT (18). High-mobility carriers are most
desirable to have the highest electrical conductivity for a given carrier concentration. The ZT
for a single material is somewhat meaningless because an array of TE couples is utilized in a
device or module. There are two materials in a TE couple (shown in Figure 2): an n type and a
p type (see, for example, http://www.marlow.com for a look at several TE module designs). If
one ignores parasitic contributions that reduce device performance such as contact resistance and
radiation effects, then the resulting ZT for the couple (based solely on the TE materials) is given
by
(α p − αn )2 T
ZT = . 3.
[(ρn κn )1/2 + (ρ p κ p )1/2 ]2
The thermopower of n-type carriers (α n ) is negative, and that of the holes (α p ) is positive. The
coefficient of performance (φ) (refrigeration mode) and the TE efficiency (ηTE ) (power generation
mode) of the TE couple are directly related to the ZT, as shown in Equation 3 for the efficiency.
The efficiency (ηTE ) of the TE couple is given by the power input to the load (W ) over the net
436 Tritt
MR41CH17-Tritt ARI 3 June 2011 8:42
p n p n
I
I
– +
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
Figure 2
A generic diagram of a thermoelectric couple made of n-type and p-type thermoelectric materials. Power
generation or refrigeration modes are possible, depending on the configuration. Red denotes hotter
temperatures, and blue denotes colder temperatures.
heat flow rate (QH ), where QH is positive for heat flow from the source to the sink:
W T H − TC (1 + ZT M )1/2 − 1
ηTE = = , 4.
QH TH (1 + ZT M )1/2 + T C /T H
where TH is the hot-side temperature, TC is the cold-side temperature, and TM is the average
temperature. Thus, ηTE is proportional to the Carnot efficiency (ηC ):
T H − TC
ηC = , 5.
TH
which we know from thermodynamics is the maximum efficiency of a heat engine operating
between two temperatures, TH and TC . Figure 3 shows the ratio of ηTE and ηC plotted as a
function of ZT for a heat engine operating between the temperatures of TH = 800 K and
TC = 300 K, which would yield a Carnot efficiency of 62%. This plot assumes that the de-
vices are operating with no parasitic losses such as contact resistance or radiation effects and that
ZT is constant between these two temperatures. Of course, these are not realistic assumptions,
and the plot is used only as a gauge to ascertain where the greatest return in efficiency gain would
be achieved per unit increase in ZT. The biggest return is for a ZT between 2 and 3, and in my
opinion, the ultimate goal of TE materials research is ZT ≈ 2–3 in stable, bulk TE n-type and
p-type materials with low parasitic losses that can operate between these two temperatures with
high ZT values.
0.6
0.5
0.4
TH = 800 K
TC = 300 K
ηTE /ηC
0.3
0.2
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
0.1
0
0 1 2 3 4 5
ZT
Figure 3
The ratio of thermoelectric efficiency to the Carnot efficiency (ηTE /ηC ) as a function of the figure of merit,
ZT. The maximum efficiency or Carnot efficiency is given by ηC = (TH − TC )/TH = (800 K − 300 K)/
800 K = 63%, where TH is the hot-side temperature and TC is the cold-side temperature. This plot assumes
no parasitic losses such as contact resistance or radiation effects and that ZT is essentially constant between
these two temperatures. Red denotes hotter temperatures, and blue denotes colder temperatures.
the device, resulting in a voltage that can be used to drive a current through a load resistance or
device. This process reflects the direct conversion of heat into electricity. Conversely, Peltier heat
is generated when an electric current is passed through a TE material to provide a temperature
gradient (with heat being absorbed on the cold side), is transferred through (or pumped by) the
charge carriers in the TE materials, and is then rejected at the heat sink, thus providing refrigeration
capability. The advantages of TE solid-state energy conversion are compactness, quietness (no
moving parts), and localized heating or cooling. In addition, energy in the form of waste heat
(0% efficiency) that would be normally lost may be converted into useful electrical energy (with
7–8%-and-greater efficiency) using a TE power generation device.
The best TE materials that are currently used in devices have a value of ZT ≈ 1. Previous
to the mid-1990s, this value had been a practical upper limit for more than 30 years, yet there
were no theoretical or thermodynamic reasons for ZT ≈ 1 as an upper barrier. As is apparent
from Equation 2, the ZT may be increased either by decreasing κ L or by increasing either α or σ ,
where α 2 σ T is defined as the power factor, which is related primarily to the electronic properties.
However, σ is tied to κ E through the Wiedemann-Franz relationship, and the ratio of κ E /σ is
essentially constant at a given temperature for the majority of TE materials.
κ E = L0 σ T . 6.
The value of the Lorenz number (L0 ) is L0 = 2.48 × 10−8 V2 K−2 for metals and is usually taken
as L = 2.0 × 10−8 V2 K−2 for degenerate, i.e., highly doped, semiconductors. Some of the goals
of current research efforts are to find new materials that either increase the current efficiency of
TE devices (e.g., increase ZT ) or have the capability of operating in new and broader temperature
regimes, especially at lower temperatures (T < 250 K) and at higher temperatures (T > 400 K).
Over the past 30 years, alloys based on the Bi2 Te3 system [(Bi1−x Sbx )2 (Te1−x Sex )3 ] as well as on
438 Tritt
MR41CH17-Tritt ARI 3 June 2011 8:42
3 Bulk
Thin film
Bi2Te3/Sb2Te3-SL
2.5
LAST-AgPb18SbTe20
2
PbTe-QDSL
(ZT )max
Filled skutterudites
1.5
(RECoSb3)
Nano-p-Bi2Te3
n-SiGe Zn4Sb3 and Pb0.98Tl0.02Te
1
Bi2Te3 alloys
CsBi2Te3
0.5
PbTe
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
Year
Figure 4
The figure of merit, ZT, as a function of the time frame of several key TE materials relative to their
discovery year, thus showing many of the recent advancements in both thin-film and bulk thermoelectric
materials. Abbreviations: LAST, PbAgSbTe compounds; QDSL, quantum dot superlattice; SL, superlattice.
the Si1−y Gey system have been extensively studied and optimized for their use as TE materials
to perform a variety of solid-state TE refrigeration and power generation applications (22, 23).
These traditional TE materials have undergone extensive investigation, and only a few years
ago there appeared to be little room for future improvement in these common bulk structures.
However, recent results on nanostructures of traditional TE materials have shown a promising new
direction for these materials. In addition, entirely new classes of compounds have been recently
investigated. Figure 4 shows the ZT values of the Bi2 Te3 and Si1−y Gey materials and their more
recent nanocomposite forms in the year they were discovered, as well as the ZT values of many of
the more recent bulk and thin-film materials that have been developed over the past several years.
In a semiconductor a charged particle must first be excited across an energy gap EG . In this case
the thermopower is approximated by
Cel kB EG
α≈ ≈ . 9.
q e kB T
Thus, the thermopower is much higher than the characteristic value k B /e ≈ 87 μV K−1 and
increases with decreasing temperature. Semiconductors can exhibit either electron conduction
(negative thermopower) or hole conduction (positive thermopower). The thermopower for dif-
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
ferent carrier types is given by a weighted average of their electrical conductivity values (σ e and σ p ),
(αe σe + α p σ p )
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
α≈ . 10.
(σe + σ p )
It is necessary to dope the semiconductors with either donor or acceptor states to allow extrinsic
conduction of the appropriate carrier type, electrons or holes, respectively. The overall objective is
to have either a predominantly n-type material or a predominantly p-type material. The total ther-
mopower will be lower than that of either of the individual contributions unless the direct band gap
is large enough (typically on the order of 10 kB T0 , where T0 is the operating temperature of the ma-
terial) to effectively minimize minority carrier contributions. Typical values of the thermopower
that are needed for good TE performance are on the order of 150–250 μV K−1 or greater.
For high-temperature applications, the contribution of minority carriers should be minimized
to maintain a high thermopower value. In addition, the thermal stability of the materials is an
essential aspect. Atomic diffusion within the materials and interdiffusion of contacts can seriously
deteriorate the properties of a given material at high temperatures. Aspects of this are discussed
elsewhere (2, 18). These materials and devices are expected to operate at elevated temperatures for
long periods of time without deterioration of the materials’ properties or the devices’ performance.
Diffusion and thermal annealing are important effects to thoroughly investigate and understand
in any set of materials.
Electrical conductivity of metals and semiconductors has been discussed extensively in many
texts on solid-state physics. Many carriers and states, typically n ≈ 1022 carriers cm−3 , are available
for conduction in metals. The electrical conductivity is then very high for metals, on the order
of 106 (-cm)−1 . Again for semiconductors, the carriers must be thermally excited across a gap
for conduction to occur, as shown from the activated behavior that is derived for the temperature
dependency of electrical conductivity:
EG
σ ≈ σ0 exp − . 11.
2k B T
Here, σ = neμ. The carrier concentration typically dominates the conductivity, but the mobility μ
and the temperature dependency of the mobility are also very important in relation to the behavior
of electrical conductivity. Many textbooks on solid-state and semiconductor physics describe this
behavior well. Thus, the two primary ways to achieve high conductivity in a semiconductor are
either by having a very small gap to excite across EG 10 KB T or by having very high mobility
carriers. Typical values of the electrical conductivity of a good TE material are on the order of
approximately 103 (-cm)−1 .
The thermal conductivity, κ, is related to the transfer of heat through a material by electrons
and by phonons (quantized vibrations of the lattice) such that κ = κ L + κ E . The electrical and
440 Tritt
MR41CH17-Tritt ARI 3 June 2011 8:42
gap semiconductors with high-mobility carriers. Mahan and colleagues (25, 26) also described
characteristics of good TE materials. The candidate material is typically a narrow-band-gap semi-
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
conductor (e.g., EG ≈ 10 kB T0 or ≈ 0.25 eV at 300 K). Also, the mobility of the carriers must
remain high (μ ≈ 2,000 cm2 (V·s)−1 ], whereas the lattice thermal conductivity must be mini-
mized. In semiconductors, the Seebeck coefficient and the electrical conductivity (both in the
numerator of ZT ) are strong functions of the doping level and chemical composition. Therefore,
these quantities must be optimized by band gap engineering for good TE performance. The ther-
mal conductivity of complex materials can often be modified by chemical substitutions, and the
lattice thermal conductivity needs to be as low as possible. Understanding these various effects
and selecting optimization strategies can be exceedingly difficult problems because in complex
materials there are often many possible degrees of freedom. Slack suggested that the best TE
material would behave as a phonon glass–electron crystal (PGEC) material; that is, it would have
the electrical properties of a crystalline material and the thermal properties of an amorphous or
glass-like material. Therefore, a material’s engineering or designer materials approach must be
pursued either to improve more established materials or to discover completely new classes of
materials to investigate.
the minimum thermal conductivity (κ min ) as the thermal conductivity when the mean free path
is limited essentially by the interatomic distance between the atoms within the crystal. A typical
analysis of κ min results in values of κ min ≈ 0.25–0.5 W m−1 K−1 (25, 27).
lattice thermal conductivity is zero (κ L = 0). We also assume the scattering in this system is elastic
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
442 Tritt
MR41CH17-Tritt ARI 3 June 2011 8:42
dimensions at room temperature are a = 3.8 Å and c = 30.5 Å. The layers are stacked along the
c axis in the following sequence:
· · · Te − Bi − Te − Bi − Te · · · Te − Bi − Te − Bi − Te · · · .
The Bi and Te layers are held together by strong covalent bonds, whereas the adjacent Te layers
exhibit van der Waals bonding. This weak binding between the Te layers accounts for the ease
of cleavage along the plane perpendicular to the c axis and the anisotropic thermal and electrical
transport properties of Bi2 Te3 . For example, the thermal conductivity along the plane perpen-
dicular to the c axis (1.5 W m−1 K−1 ) is nearly twice that of the value along the c-axis direction
(0.7 W m−1 K−1 ). When grown from a melt or by zone refining, Bi2 Te3 crystals are always non-
stoichiometric and show p-type behavior. In contrast, n-type materials can be grown from melt
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
containing excess Te, I, or Br. The thermal conductivity values of both p-type and n-type Bi2 Te3
are ∼1.9 W m−1 K−1 , giving a ZT of ∼0.6 near room temperature. Ioffe (21) suggested that
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
alloying could further reduce the lattice thermal conductivity of Bi2 Te3 through the scattering
of short-wavelength acoustic phonons. The optimum compositions for TE cooling devices are
normally Bi2 Te2.7 Se0.3 (for n type) and Bi0.5 Sb1.5 Te3 (for p type) with ZT ≈ 1 near room tem-
perature. As stated above, the incorporation of in situ nanostructures that are nucleated by either
high-energy ball milling or melt spinning in these materials has proven beneficial in reducing the
lattice thermal conductivity and in achieving ZT ≈ 1.4 between 300 K and 400 K (14–16).
In contrast to Bi2 Te3 , PbTe crystallizes in a cubic NaCl-type crystal structure, and the TE
properties are isotropic. Both p-type and n-type thermoelements can be produced by doping of
acceptors (e.g., Na2 Te or K2 Te) or donors (PbI2 , PbBr2 , or Ge2 Te3 ). In analogy to Bi2 Te3 , solid-
solution compositions (e.g., PbTe-SnTe) have been made to lower the lattice thermal conductivity
(29). The ZT value is low near room temperature but increases to ZT ≈ 0.7 at 700 K, making this
a prime candidate for power generation in the higher-temperature range. High ZT values (ZT ≈
1.5) were recently achieved in Tl-doped PbTe materials by using the concept of energy filtering
to enhance ZT. Such values can be achieved by inducing a large variation in the density of states
near the Fermi level, thus significantly enhancing the thermopower (30). This energy-filtering
effect can be understood through a modified Mott diffusion equation for the thermopower by
manipulating the chemical potential term with an addition of only 2% Tl.
π 2 k2B T 1 d n(E) 1 d μ(E)
α= + . 13.
3e n dE μ d E E=E F
It is possible to achieve ZT in excess of unity at 700 K in a structurally related solid-solution
composition, AgSbTe2 (80%)–GeTe (20%) (termed TAGS, which denotes alloys containing Te,
Ag, Ge, and Sb). However, due to high-temperature stability issues, these compositions are not
currently favored in TE devices.
Neither Si nor Ge is a good TE material because both have a very high lattice thermal conduc-
tivity (150 W m−1 K−1 and 63 W m−1 K−1 for Si and Ge, respectively). However, the lattice thermal
conductivity can be substantially reduced by alloy formation between the two elements. The best
alloy composition is Si0.7 Ge0.3 , which possesses a thermal conductivity that is ∼10 W m−1 K−1 , and
the reduction is apparently due to increased phonon-phonon and phonon-electron scattering (18).
Remarkably, such a large reduction does not unduly reduce the carrier mobility, and ZT ≈ 0.6–0.7
can be realized at elevated temperatures. Due to their exceptional stability at high temperature
(∼1,200 K), these alloys are of interest to NASA for use in RTGs in deep-space probes. Nanos-
tructuring in these materials has also led to a further reduction in the lattice thermal conductivity
and to higher ZT values; such nanostructures are stable at high temperatures (31).
high power factors (exceeding the optimally doped Bi2 Te3 solid solutions) in the low-temperature
range (<250 K), the thermal conductivity is relatively high (κ ≈ 4–8 W m−1 K−1 ). These materials
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
444 Tritt
MR41CH17-Tritt ARI 3 June 2011 8:42
to localized electrons with low carrier mobilities. This situation changed with the unexpected
discovery of good TE properties in a strongly correlated layered oxide, NaCo2 O4 (42). This oxide
attains ZT ≈ 0.7–0.8 at 1,000 K. Inspired by the striking TE performance of NaCo2 O4 , most
of the current research is currently focused on Co-based layered oxides such as Ca3 Co4 O9 and
Bi2 Sr3 Co2 Oy that crystallize in misfit layered structures (24). Among the n-type oxides, Al-doped
ZnO (Al0.02 Zn0.98 O) exhibits reasonably good TE performance (ZT ≈ 0.3 at 1,000 K) (43).
character through physical confinements in quantum dots, nanowires, and thin-film structures
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
to enhance the electronic properties of a given material (44). In addition, nanostructured semi-
conductor materials may scatter mid- to long-wavelength phonons and thereby reduce the lattice
thermal conductivity to κ min . Researchers at the Research Triangle Institute have demonstrated
a significant enhancement in ZT through the construction of Bi2 Te3 /Sb2 Te3 superlattices (17,
45). These materials exhibit ZT ≈ 2.4. The enhancement is attributed to creating a nanoengi-
neered material that is efficient in thermal insulation while remaining a good electrical conductor.
The thermal insulation arises from a complex localization-like behavior for phonons, whereas
electron transmission is facilitated by the optimal choice of band offsets in these semiconductor
heterostructures. Also, there have been reports on PbTe/PbTeSe quantum dot structures that
yield ZT ∼ 1.3–1.6 (46). These materials have been grown as thick films. Such films have been
floated off the substrate to yield freestanding films, which were measured to yield these results.
4. CONCLUSIONS
Currently, there are no theoretical or thermodynamic limits to the possible values of ZT. The
current need for alternative energy technologies and materials to replace the depleting supplies
of fossil fuels is becoming more urgent. Energy-related research will grow rapidly over the next
few years, and higher-performance TE materials and devices are direly needed. Slack (24) esti-
mated that an optimized PGEC material could exhibit values of ZT ≈ 4. Such projections give
encouragement that such materials may be possible. However, as is evident from Figure 3, if
we could find TE materials with ZT ≈ 2–3 and that were stable over a broad temperature range
with low parasitic losses, then the problem would be almost entirely solved. Thus, a systematic
search and subsequent thorough investigation of promising materials may eventually yield these
much-needed materials for next-generation TE devices. Although many strategies are being em-
ployed in hopes of identifying the next generation of TE materials, the PGEC approach appears
to be the best, especially for bulk materials. One has to decide whether holey semiconductors
(materials with cages such as skutterudites or clathrates) or unholey semiconductors such as SiGe
or PbTe are the best to pursue and which tuning parameters are available to improve these mate-
rials (47). To date, none of the new materials have displaced the current state-of-the-art materials
(Bi2 Te3 , PbTe, and SiGe) in a commercial TE device. These materials have held that distinction
for more than 30 years. Nanostructuring and in situ growth of stable nanostructures, thus yield-
ing TE nanocomposites, have further improved the TE properties of these materials. Although
the thin-film results are very exciting, they may be most appropriate for small-scale electronic
and optoelectronics applications in which small heat loads or low levels of power generation are
more appropriate. To address large-scale refrigeration (home refrigerators) or power generation
SUMMARY POINTS
1. Thermoelectric materials can be incorporated into devices that can be used for solid-state
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
DISCLOSURE STATEMENT
The author is not aware of any affiliations, memberships, funding, or financial holdings that might
be perceived as affecting the objectivity of this review.
LITERATURE CITED
1. Tritt TM, ed. 2000. Recent Trends in Thermoelectric Materials Research, Vol. 71: Semiconductors and Semimet-
als. New York: Academic
2. Kanatzidis MG. 2010. Nanostructured thermoelectrics, the new paradigm. Chem. Mater. 22(3):648–59
3. Heremans J. 2009. Nanometer-scale thermoelectric materials. In Springer Handbook on Nanotechnology. Part
A. Nanostructures, Micro/Nanofabrication and Materials, ed. B Bhushan, pp. 345–73. Heidelberg: Springer.
2nd ed.
446 Tritt
MR41CH17-Tritt ARI 3 June 2011 8:42
4. Dresselhaus MS, Chen G, Tang MY, Yang R, Lee H, et al. 2007. New directions for low-dimensional
thermoelectric materials. Adv. Mater. 19:1043–53
5. Lan Y, Minnich AJ, Chen G, Ren Z. 2010. Enhancement of thermoelectric figure-of-merit by a bulk
nanostructuring approach. Adv. Funct. Mater. 20:357–76
6. He J, Tritt TM. 2010. Thermal-to-electrical energy conversion from the nanotechnology perspective. In
Nanotechnology for the Energy Challenge, ed. Javier Garcia Martinez, pp. 47–77. Weinheim, Ger.: Wiley-
VCH
7. Tritt TM, Bottner H, Chen LD. 2008. Thermoelectrics: direct solar thermal energy conversion. MRS
Bull. 33:366–68
8. Yang J, Caillat T. 2006. Thermoelectric materials for space and automotive power generation. MRS Bull.
31:224–29
9. Allen AW. 1997. Thermoelectrically cooled IR detectors beat the heat. Laser Focus World 33:S15
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
10. Nolas GS, Poon SJ, Kanatzidis M. 2006. Recent developments in bulk thermoelectric materials. MRS
Bull. 31:99–205
Annu. Rev. Mater. Res. 2011.41:433-448. Downloaded from www.annualreviews.org
11. Koumoto K, Terasaki I, Funahashi R. 2006. Complex oxide materials for potential thermoelectric appli-
cations. MRS Bull. 31:206–210
12. Rao A, Ji X, Tritt TM. 2006. Properties of nanostructured one-dimensional and composite thermoelectric
materials. MRS Bull. 31:218–23
13. Sootsman JR, Chung DY, Kanatzidis MG. 2009. New and old concepts in thermoelectric materials.
Angew. Chem. Int. Ed. 48:8616–39
14. Xie W, Tang X, Yan Y, Zhang Q, Tritt TM. 2009. Unique low-dimensional structure and enhanced
thermoelectric performance for p-type Bi0.52 Sb1.48 Te3 bulk material. Appl. Phys. Lett. 94:102111
15. Poudel B, Hao Q, Ma Y, Lan Y, Minnich A, et al. 2008. High-thermoelectric performance of nanostruc-
tured bismuth antimony telluride bulk alloys. Science 320:634–38
16. Xie W, He J, Kang HJ, Tang X, Zhu S, et al. 2010. Identification of the specific nanostructures responsible
for the high thermoelectric performance of (Bi,Sb)2 Te3 nanocomposite. Nano Lett. 10(9):3283–89
17. Boettner H, Chen G, Venkatasubramanian R. 2006. Aspects of thin-film superlattice thermoelectric
materials, devices, and applications. MRS Bull. 31:211–17
18. Nolas GS, Sharp J, Goldsmid HJ. 2001. Thermoelectrics: Basic Principles and New Materials Developments.
New York: Springer
19. Morelli DT. 1997. Thermoelectric devices. Encycl. Appl. Phys. 21:339–54
20. Chaiken PM. 1990. An introduction to thermopower for those who might want to use it to study organic
conductors and superconductors. In Organic Superconductors, ed. VZ Kresin, WA Little, pp. 101–15.
New York: Plenum
21. Ioffe AF. 1957. Semiconductor Thermoelements and Thermoelectric Cooling. London: Infosearch
22. Goldsmid HJ. 1986. Electronic Refrigeration. London: Pion
23. Rowe DM, ed. 1995. CRC Handbook of Thermoelectrics. Boca Raton, FL: CRC
24. Slack GA. 1995. New materials and performance limits for thermoelectric cooling. See Reference 23,
pp. 407–40
25. Mahan GD, Sofo JO. 1996. The best thermoelectric. Proc. Natl. Acad. Sci. USA 93:7436–39
26. Mahan GD, Sales B, Sharp J. 1997. Thermoelectric materials: new approaches to an old problem. Phys.
Today 50(3):42–47
27. Slack GA. 1979. The thermal conductivity of non-metallic crystals. In Solid State Physics, Vol. 34, ed.
F Seitz, D Turnbull, H Ehrenreich, pp. 1–71. New York: Academic
28. Heikes RR, Ure RW Jr. 1961. Thermoelectricity: Science and Engineering. New York: Intersci.
29. Wood C. 1988. Materials for thermoelectric energy conversion. Rep. Prog. Phys. 51:459–539
30. Heremans JP, Jovovic V, Toberer ES, Saramat A, Kurosaki K, et al. 2008. Enhancement of thermoelectric
efficiency in PbTe by distortion of the electronic density of states. Science 321:554–57
31. Joshi G, Lee H, Lan Y, Wang X, Zhu G, et al. 2008. Enhanced thermoelectric figure-of-merit in nanos-
tructured p-type silicon germanium bulk alloys. Nano Lett. 8(12):4670–74
32. Kanatzidis MG, Mahanti SD, Hogan TP, eds. 2003. Chemistry, Physics and Materials Science of Thermoelectric
Materials: Beyond Bismuth Telluride. New York: Plenum
33. Chung DY, Hogan T, Brazis P, Rocci-Lane M, Kannewurf C, et al. 2000. CsBi4 Te6 : a high-performance
thermoelectric material for low-temperature applications. Science 287:1024–28
34. Littleton RT, Tritt TM, Kolis JW, Ketchum DR, Lowhorn ND, Korzenski MB. 2001. Suppression of the
resistivity anomaly and corresponding thermopower behavior in the pentatelluride system by the addition
of Sb: Hf1−X ZrX Te5−Y SbY . Phys. Rev. B 64:121104
35. Hsu KF, Loo S, Gao F, Chen W, Dyck JS, et al. 2004. Cubic AgPbm SbTe2+m : bulk thermoelectric
materials with high figure of merit. Science 303:818–21
36. Shen Q, Chen L, Goto T, Hirai T, Yang J, et al. 2001. Effects of partial substitution of Ni by Pd on the
thermoelectric properties of ZrNiSn-based half-Heusler compounds. Appl. Phys. Lett. 79:4165–67
37. Caillat T, Fleurial JP, Borshchevsky A. 1997. Preparation and thermoelectric properties of semiconducting
Zn4 Sb3 . J. Phys. Chem. Solids 58:1119–25
38. He T, Chen J, Rosenfeld HD, Subramanian MA. 2006. Thermoelectric properties of indium-filled skut-
by Stanford University - Main Campus - Lane Medical Library on 09/29/12. For personal use only.
448 Tritt
MR41-FrontMatter ARI 4 June 2011 12:1
Annual Review of
Materials Research
Current Interest
Elastic and Mechanical Properties of the MAX Phases
Michel W. Barsoum and Miladin Radovic p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 195
Electrocaloric Materials
J.F. Scott p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 229
Electrochromic Materials
Roger J. Mortimer p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 241
Nanowire Solar Cells
Erik C. Garnett, Mark L. Brongersma, Yi Cui, and Michael D. McGehee p p p p p p p p p p p p 269
Nonconventional (Non-Silicon-Based) Photovoltaic Materials
T. Unold and H.W. Schock p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 297
v
MR41-FrontMatter ARI 4 June 2011 12:1
Index
Errata
vi Contents