Thermoelectric effect [thermoelectricity]

Thermoelectric effect (thermoelectricity) is defined as the set of those particular thermal phenomena of conversion of heat into electricity, and vice versa, that occur in particular conducting or semiconducting materials (called thermoelectric materials that have significant thermoelectric effects), correlating the heat flow that passes through them to the electric current that flows through them. The main thermoelectric phenomena of correlation between electromotive forces and temperature differences, are:

  • Joule effect;
  • Seebeck effect;
  • Peltier effect;
  • Thomson effect.

These effects are explained in the phenomena of diffusion of charge carriers (electrons and gaps) by thermal effect in conductors (or semiconductors), and in the phenomena of contact between the junctions between conductors of different materials.

Historical notes

The birth of thermoelectricity occurred when Alessandro Volta, more precisely on February 10, 1794, revealed the connection between heat and electricity by observing how, by heating the ends of a metal arc, an “electrical voltage” was obtained which would disappear when it cooled down.

The first thermoelectric phenomenon discovered was Seebeck effect. In 1821, during an experiment, the physicist T.J. Seebeck, noticed that, in a circuit consisting of different metallic conductors or semiconductors (copper and bismuth) but connected together, a temperature difference generated a deviation of the magnetic needle of a compass. Seebeck thought that this phenomenon was related to terrestrial magnetism. In 1823, the Danish physicist Hans Christian Ørsted, however, showed that the effect was caused by a flow of negative charges in motion that, by Ampere’s law, cause the magnetic field responsible for the deviation of the compass needle. The potential difference to which the two metals are brought is proportional to the temperature difference. The proportionality constant α takes the name of Seebeck Coefficient, and is expressed in: \(\mu\cdot\textrm{V}\cdot\textrm{K}^{-1}\).

\[\alpha (T)=\lim_{\Delta T\rightarrow 0}\dfrac{\Delta V}{\Delta T}\]

Seebeck’s studies led to the construction of the first thermopile in 1823, with a conversion efficiency of 3%, comparable with the efficiency of many steam engines. The thermopile, thanks to its very stable voltage, was later used by German physicist Georg Simon Ohm to study the link between voltage and current in a circuit. The second thermoelectric effect was discovered in 1834 by a French watchmaker, Jean Charles Athanase Peltier; in a conductor made of two different metals, a difference in electrical potential generates a temperature difference (which cannot be traced back to the Joule effect). The heat absorbed or generated was proportional to the current, through a proportionality coefficient, called Peltier’s coefficient. Later, in 1838, the Russian physicist Heinrich Lenz, demonstrated that the production or absorption of heat on the junction depended on the direction of the current, and with a simple experiment he was able to freeze and thaw water placed in a hollow on a junction between bismuth and copper.

The studies on thermoelectricity, after an interruption of about twenty years caused by the enthusiasm of Laplace and Ampere’s researches on electromagnetism, found a new impulse with the British physicist William Thomson (Lord Kelvin), who succeeded in 1851, first of all, to correlate Seebeck effect and Peltier effect and, subsequently, discovered a third thermoelectric effect: Thomson effect.

A homogeneous conductor with the ends at different temperature (\(T_1\) and \(T_2\)) and crossed by an electric current, gives or absorbs heat depending on whether the heat flow and the current are coincident or opposite. The first applications of Thompson’s theories were only in 1909 thanks to German physicist Edmund Altenkirch. The results of his research led to the definition of an index of merit, with which to classify thermoelectric materials, “the figure of merit z”.


where: α is the Seebeck coefficient or thermoelectric power, σ is the electrical conductivity, λ is the thermal conductivity. The merit index z is also used in its dimensionless form zT, multiplied by the absolute temperature. A good thermoelectric material must possess a high index of merit. This implies that for a given temperature, the material must have a very high Seebeck coefficient, very low thermal conductivity (to maintain the temperature difference at the junction), and very low electrical resistance (to minimize Joule effect losses).

The development of thermoelectric applications began in the early 1930s with the advent of semiconducting thermoelectric materials, primarily by the Russian school. The absence of moving parts made these generators silent and reliable, ideal for military field devices or sensors; many applications have been developed essentially for the military and aerospace sectors that are still strictly secret (Radioisotope Thermoelectric Generators used by NASA in many space missions: Apollo, Pioneer, Viking, Voyager).

In the West, in particular in the USA, the research and development of thermoelectricity began immediately after the war, and was intense until the early ’60s, when more than 100 companies were active in this field. Ten years later, thermoelectricity had only a dozen or so companies engaged, down to only two industries, Melcor and Marlow, active in the early 1990s. This progressive disengagement was motivated by the observation that, despite the discovery of new materials, the efficiency of thermoelectric generators invariably remained an order of magnitude lower than had been erroneously predicted in the late 1950s. In 1970 the “International Thermoelectric Society” was born, bringing together all the greatest experts in thermoelectricity and promoting it by presenting all technological innovations at an international meeting held annually since 1970 in the world’s largest capitals.

Starting in 1993, interest in thermoelectricity was once again ignited. This time, a theoretical prediction is driving the research activity: the efficiency of thermoelectric devices can be significantly increased (reaching efficiencies of up to 17%) through the use of low-dimensional systems, i.e. by using nano-structured thermoelectric materials. At the same time, the production cost of thermoelectric modules is continuously decreasing, not only for physiological reasons of the market, but also thanks to academic studies, which will lead to the definition of new low-cost construction standards, and will make the use of this technology in the direct conversion of heat into electricity extremely advantageous.

Thermoelectric generators

The thermoelectric generators TEG (Thermo Electric Generator), also known as Seebeck generators are semiconductor devices that transform a flow of heat into electricity in a direct way. In particular, the thermoelectric generators are made as Peltier cells from blocks of semiconductor material with different doping, or P-type (doped with acceptor atoms) and N-type (doped with donor atoms).

A thermoelectric module is therefore composed of n pairs of thermoelectric materials of type P and N electrically connected in series to a load resistance RL. P-type materials will have positive Seebeck coefficient while N-type materials will have negative coefficient. Therefore, the Seebeck coefficient of the thermocouple αpn = αp – αn will be positive.

The N pairs are held together by two plates of insulating material (to avoid short-circuiting all the junctions), generally made of ceramic material that puts them in thermal parallel. The ceramic plays a fundamental role because it must have an extremely high thermal conductivity while maintaining a very low electrical conductivity. The materials commonly used are Alumina or Aluminium Oxide (Al2O3), Aluminium Nitride (AlN) or Beryllium Oxide (BeO). The thermoelectric module can be used for three applications:

  • for heating: TEH (Thermoelectric Heater)
  • to cool: TEC (Thermoelectric Cooler)
  • to produce energy: TEG (Thermoelectric Generator)

These configurations differ in how the bars of thermoelectric material p and n are polarized. In operation as TEH and TEC, the thermoelements are polarized by a potential difference, this configuration allows to control the temperature difference between the two plates. In the operation as TEG instead the opposite effect is used, the thermoelements are polarized by a temperature gradient that produces a potential difference at the ends of the conductors.

Notify of

Inline Feedbacks
View all comments
Scroll to Top