Page 147 - Read Online

Page 2 of 35 Martin-Gonzalez et al. Energy Mater. 2025, 5, 500121 https://dx.doi.org/10.20517/energymater.2025.32

it offers a comprehensive overview of state-of-the-art thermoelectric materials and devices and a summary of the
challenges associated with transitioning these materials into practical devices.

Keywords: Thermoelectricity, nanostructuring, phonon and electron transport, energy harvesting, zT figure of
merit, thermoelectric devices

INTRODUCTION
Thermoelectric (TE) devices are solid-state systems capable of generating electrical power from heat . They
[1]
have many advantages over conventional thermomechanical energy conversion devices due to their solid-
state nature, absence of moving parts and high reliability. These advantages have led to some noteworthy
applications, including the creation of direct electrical power (e.g., power generators for spacecraft),
automotive climate control seats, small solid-state cooling diode lasers and infrared sensors, potential
applications for power generation from solar irradiation, waste heat recovery, and powering Internet of
[2,3]
Things (IoT) devices . Companies such as RCA, 3M, Texas Instruments, Marlow Industries, and several
other start-ups, have already developed and begun producing TE devices. However, their low efficiency
compared to conventional thermomechanical cycles has limited their use to only niche applications and
those for which the conventional cycles cannot be easily applied.

The ability of a TE material to convert heat into electricity is quantified by the zT figure of merit, given by
zT = S σT/κ, where σ is the electrical conductivity, S is the Seebeck coefficient (the product S σ is referred to
2
2
as the power factor or PF), T is the absolute temperature and κ is the thermal conductivity determining the
losses of the process, composed of the electronic and phononic (or lattice) parts as κ = κ + κ . While these
p
e
transport coefficients of TE materials have been studied for decades, their negative interdependence (σ is
inversely proportional to S and directly proportional to κ ) has hindered the overall increase in zT . The
[4]
e
best commercial materials are based on compounds and alloys of Bi, Te and Pb, and provide zT ~ 1, which
corresponds to ~ 10% of the Carnot efficiency . Since Bi Te and PbTe with zT ~ 1 were developed more
[5]
3
2
than 40 years ago [Figure 1], there have been many efforts to increase zT further. Alloy systems and carrier
concentration optimization have been investigated as the main drivers to improve conventional
thermoelectrics; however, the necessity of optimizing three adversely interdependent parameters in the
same material was a fundamental restriction. As a result, around the mid-1970s [Figure 1], optimism faded,
basic materials research slowed down, and the mainstream impression in the scientific community was that
zT was constrained to unity. For such rather low zT values, the efficiency was insufficient to support
competitive products compared to other technologies.

The resurgence in TE materials research started in the 90s after two influential studies published in 1993 by
Hicks and Dresselhaus on Bi nanowires, suggesting an increase in performance from the use of low-
dimensional TE materials . They essentially provided a paradigm change, pointing out that quantum
[6,7]
confinement could offer a fresh approach to improving the efficiency of TE materials. The core premise was
that the low-dimensional density of states (DOS) provides sharp features in energy, an element that could
largely increase the Seebeck coefficient, independently from the electrical conductivity (i.e., without
reducing it), such that large PFs are reached. Although this initial suggestion that quantum confinement will
improve the PF was never realized for reasons that we and others explained in later works [8-11] , research on
nanomaterials quickly led to the realization that nanostructuring offered enormous opportunities for
thermal conductivity reductions, from only mild changes all the way to even orders of magnitude. Such
works became more frequent in the literature from around the year 2000 onwards, when material synthesis
improved to the degree at which adequate control over nanomaterial design was possible. Funding agencies
across the globe increased financial support, which led to a resurgence in the field [see Figure 1 [12-18] and
Table 1].