3.1. Thermoelectric Materials
It is estimated that between 50% and 80% of the studies related to thermoelectric generation systems (TEG) focus on the TE materials themselves. Therefore, the development of materials that have a low cost and high efficiency is fundamental to achieving commercial profitability in these systems [
17,
18].
In the systematic search carried out, 10 articles were found in which the TE material was studied in depth in order to increase its efficiency, decrease costs or develop viable materials in both environmental and sustainability aspects by analyzing factors such as the material’s abundance and its degree of toxicity among others.
Table 4 presents a classification of these items according to the cost, efficiency and sustainability of the TEG modules. In this table, the cost column expresses the materials cost reduction with a minus sign (−), and the materials cost increase with a plus sign (+). In this aspect, the elements that compose in some cases the type of processing are taken into account. Note that in most of the articles reviewed, the cost of the material was not stated explicitly. For efficiency, the minus sign refers to materials with low efficiencies (ZT < 1) and the plus sign (+) for materials with significant efficiencies (ZT > 1). Finally, the sustainability of the materials is assessed using measures such as the availability, toxicity and useful life; therefore, a minus sign indicates that the material is unsustainable and a plus sign that it is sustainable.
Among the articles shown in
Table 3, three case studies were found [
19,
58,
59] that examined cost of TE materials, including their cost effectiveness and the processing technique used to develop them.
Among the cost-effective materials, the use of oxides as raw materials stands out. Hung et al. [
43] studied different oxides (Na
2CoO
4, Ca
3Co
4O
9, ZnO, SrTiO
3 and CaMnO
3) in order to decrease costs in the manufacture of TEG modules since the cost of manufacturing the oxides is approximately 1.1
$/kg, which is equivalent to only a quarter of composite materials composed of metals and rare earths. On the other hand, Lee et al. [
39] studied the potential of TiO
2-x for TE materials manufactured by plasma deposition. Ozturk et al. [
30] studied two types of oxides: Ca
2.5Ag
0.3X
0.2Co
4O
9 type n and Zn
0.96Al
0.02Y
0.02O type n, where X and Y are different dopants manufactured using the sol-gel method. In these works, the benefits of using oxides as raw materials are highlighted. Among these benefits are low cost, abundance, resistance to high temperatures, as well as simplified manufacturing processes not requiring controlled atmospheres. However, it can be seen that the purpose of these studies is to improve the efficiency of materials. In
Table 5 it can be seen that the oxides present the least merit (efficiency). According to the review, the modules’ oxide-based TEGs can increase their efficiency through the use of special processing [
37] or doping techniques [
28], which makes their use more viable.
On the other hand, the use of cheap and abundant materials such as lead-based materials or silicides has also been the subject of recent studies. Han et al. [
33] studied the feasibility of PbTe-SrTe base materials doped with 2% Te. They concluded there was a cost reduction through a low-cost processing method such as stable screen printing, although one of the base materials and the tellurium are high cost and low abundance elements. Jiang et al. [
27] proposed PbS as an alternative to the base material PbTe, arguing that by doping with Sb and Se, efficiency can be considerably improved, in addition to them being abundant and low-cost materials.
Fu et al. [
44] and Salvador et al. [
52] developed materials such as FeNbSb (Half-Heuslers) and Yb
0.09Ba
0.09La
0.05Co
4Sb
12 (skutterudite), respectively, which are composed of low-cost and abundant elements, and by doping techniques are able to increase their efficiency. However, Ouyang et al. [
36] evaluated some of the latest generation materials and recommended that materials such as skutterudites and half-heuslers could only be used in applications where cost is not of concern, due to the high manufacturing costs of these materials. On the other hand, Skomedal et al. [
40] suggested the use of magnesium silicides as a favorable TE material, due to their low cost, abundance and low toxicity, despite their low efficiency when doping with elements such as Sn and Sb. They concluded that materials based on magnesium silicides are recommended for applications where low cost or low weight are more important than efficiency.
In addition, Homm et al. [
56] analyzed some TE materials such as SiGe, PbTe, Bi
2Te
3, FeSi
2 and ZnO. The authors classified them according to selection criteria for different applications that required certain specifications for temperature, efficiency and cost, but taking into account the environmental aspects that each one presented.
According to the present review, it is observed that there is a conflict between the aspects of cost, efficiency and sustainability.
Figure 4 presents a classification based on these aspects of recent studies addressed in this analysis. Three articles were found involving costs and efficiency in zone A, [
27,
29,
44]; three articles between costs and sustainability in zone B [
30,
39,
46]; one article involving efficiency and sustainability in zone C [
36]; and tree articles involving all aspects, costs, efficiency and sustainability in zone U [
40,
52,
56]. From this classification it is concluded that the oxides are inexpensive TE materials with important advantages. In particular, they are abundant, do not require high-cost processing, and resist high temperatures, which prevents premature degradation of the TE material. Moreover, they enable the formation of robust materials with a longer useful life, and have a good cost-sustainability ratio. However, their efficiency is reduced with respect to the commercially used TE materials, which prompts us to think about the different research approaches to improve them, such as nano-structuring, electronic band engineering, quantum confinement, as well as strategies such as crystal electron glass phonon, doping, and introduction of defects, among others. However, the use of any of these techniques requires specialized and complex processes, which would be reflected in the final cost of the product and would probably mean that the cost-efficiency ratio is not viable for developments in a commercial environment. Therefore, the development of this research is of utmost importance for providing not only a better future perspective of TE materials, but also because there are few investigations that specifically address the economic component of these materials.
Furthermore, from a sustainability perspective, little information is available on commonly used TE materials. An example of this is the use of toxic materials such as lead, tellurium and bismuth in their fabrication. Therefore, an important aim of research is to explore is the environmental risks that these materials can present at different stages of the useful life of TEG modules and to search for abundant and low-cost elements.
Interest in certain thermoelectric materials is based on a combination of their characteristics and performance.
Figure 5 shows the trends in the number of publications in recent years in relation to some representative thermoelectric materials, according to a survey carried out in the Scopus database. The figure shows the growing research interest in these thermoelectric materials.
Recent Development of New Materials for Thermoelectric Applications
The increases in the ZT values are produced especially by the decrease in the thermal conductivity of crystal lattices, and the recent advances in the development of new TE materials are based on the search for mechanisms that make it possible to minimize the thermal conductivity in the crystal structures of TE materials. Advances in TE materials provide measurable improvements in ZT values through the use of nanotechnology-based techniques. Nanophonon metamaterials provide special local resonance states in semiconductor materials for suppression of thermal conductivity. According to Ouyang et al. [
60], nascent theories are being forged in the field of TE materials. Among the most promising are the coherent phonon theories (
https://www.nature.com/articles/nmat3826, aacessed on 23 August 2021), the nanophonon metamaterial [
61], the rattling effect [
62], the topological phonon [
63,
64] and the topological electron [
65]. Likewise, the synthesis of low-dimensional materials would allow the separation of related thermoelectric parameters to optimize thermoelectric performance. Among the advances in this field, the 1D Nanowires stand out [
66,
67], as well as the 2D Materials [
68,
69] and the Nanomesh Structures [
70,
71]. Finally, it should be noted that given the recent advances in computing, artificial intelligence and machine learning in combination with atomic simulation techniques, the development of new tools to predict new structures and characteristics of novel materials is envisioned, and these will provide accurate forecasts of the inherent properties of TE materials.
3.2. TEG Modules
In the systematic search carried out, 29 articles related to TEG modules were found. We observed that one of the most commonly addressed topics is the efficiency/cost relationship presented by TE materials. This aspect is usually approached from several points of view, such as cost reduction, varying the TE material, or through a design of the TEG module that preserves its efficiency. However, another important factor that must be taken into account is the sustainability of the modules, which ranges from the analysis of their useful service life to the study of their final disposal. In
Figure 6, the classification of the articles according to their content can be observed. These were classified by costs, sustainability, efficiency and modeling. In this graph, one article as found classified involving costs [
42]; seven involving efficiency [
27,
39,
44,
49,
52,
56,
57]; one in modelling [
48]; and six in sustainability [
18,
23,
37,
38,
50,
53]. In the common areas D, E, F, G, H, it was not found papers between costs and sustainability and costs and modelling in zones D and G respectively. Between costs and efficiency, zone E, three articles were found [
18,
22,
55]; while for zone F, between efficiency and modelling, seven were found [
28,
34,
35,
41,
45,
46,
47]. Finally, ten references were found in thermoelectric modules, zone H [
16,
21,
30,
31,
32,
36,
40,
41,
43,
55].
3.4. Module Manufacturing
With the aim of decreasing the production costs without affecting the efficiency of the modules, some authors focus their research on the design of TE modules, the search for new materials in order to decrease costs, or increasing their efficiency. The use of oxides, half-heusler materials, skutterudites, and composite materials (organic/inorganic) can be a solution to overcome the limitations of the TEG modules. It is important, however, that these studies be developed in a holistic context oriented to applicability.
Studies such as those by Salvador et al. [
52], on skutterudite encapsulated modules with a ZT of 1, enable this type of module to compete directly with the efficiencies of commercial PbTe modules. However, to date, commercial modules with this type of materials are scarce and costly due to their current manufacturing methods. Fu et al. [
44] evaluated the potential of doped half-heusler materials with an efficiency of 6.2% and a power density of 2.2 W/cm
2. These modules can resist high temperatures (~927 °C) and are an economical alternative to commercially used TE materials.
In order to reduce costs in TEG modules and to improve their coupling to any surface, Lee et al. [
39] studied the manufacture of TEG devices and basic electronics using titanium oxides (which operate at temperatures ~500 °C), and deposited them by plasma sintering, thus obtaining an assembly of thermocouples connected in series and in parallel with an efficiency of 0.85% and an electrical power of 2.43 mW at 450 °C. The authors did not conduct an economic analysis on the assembly, only referring to its easy and low manufacturing cost through the elimination of many of the parts that make up the commercial TEG modules. On the other hand, Yazawa et al. [
54] proposed the use of flexible modules that incorporate organic/inorganic composite materials and reduce costs but with a reduced performance (ZT between 0.01 and 0.25), a performance that is relatively low compared with commercial TE materials.
Anderson et al. [
23] carried out a techno-economic analysis on the total cost of TEG devices, finding that the use of impure TE materials such as oxides or other types of cheaper TE materials are not the most feasible option at a cost level, since TE material only represents 15% of the total value of a TEG module.
Table 6 presents some characteristics of commercial modules such as base material, dimensions, power output, open circuit voltage, operating temperature range and cost. Currently, the most common TEG modules are those manufactured from bismuth telluride. Among these, a great variety can be found in which characteristics such as their configuration, output power and circuit voltage might vary. Most modules can work at maximum temperatures between 320 °C and 350 °C, but their optimal operating conditions are around 250 °C. These types of modules can be found in the market with prices ranging between
$10 and
$28 US.
Likewise, commercially it is possible to find modules that resist higher temperature ranges, such as TEG PbTe-BiTe modules. These TEG modules can work at maximum temperatures of around 360 °C, and commercially they can be found with better output powers than the BiTe based ones. Naturally, the improvement of these characteristics is reflected in their cost.
Table 850 °C reaching 6% efficiency, being attractive for the recovery of residual heat at high temperatures. However, this type of component is up to seven times the cost of traditional BiTe-based modules. This makes them less attractive due to their cost-efficiency ratio.
3.5. Module Design
Commercial modules, such as those from
Table 6, generally have a pre-designed configuration from the supplier, which means that for some applications, they are not suitable or do not show their optimal performance. On the other hand, the design is an important factor since by means of the design parameters, the efficiency, the cost and the useful life of the modules can be improved. Likewise, most of the studies related to TEG components or systems are carried out using simulation tools [
15], due to the complexity of their manufacture, as well as the cost that these would generate for their development if they were carried out exclusively by experimental means. Some of the most relevant studies in this regard are listed below.
Segmented modules represent one of the most viable alternatives from the view point of design. In these, various TE materials are used to manufacture the module legs, seeking to increase the working temperature, minimize the thermal effects, and increase the efficiency of the TEG modules. Recent work related to the manufacture of segmented modules includes the work of Hung et al. [
46], who implemented commercial TE materials in cold areas of the cell and oxides in hot areas, in order to increase the working temperature of the modules. The viability of these TEG modules was analyzed using numerical modeling, which found that the oxide-segmented modules have an efficiency of around 10%. Ouyang et al. [
36] carried out a study in which high-ZT TE materials were evaluated by finite element analysis. A systematic model was achieved for the segmented modules, finding a cost-performance ratio of ~0.86
$/W with an efficiency of 17.8%. Similarly, Jiang et al. [
27] evaluated a TEG module composed of np Bi2Te3 and np PbS/PbTe, which exhibited an efficiency of 11.2% with a ∆T = 317 °C. In addition, they optimized the ratio of the legs at low and high temperatures, determining the optimal ratio to be 7:17. The maximum power obtained in this TEG module was 0.53 W for a ∆T = 312 °C.
During the design of TEG modules, the optimization of parameters such as the length of the legs or the number of thermocouples of the TE materials helps to reduce the amount of material used, without compromising the efficiency of the modules. According to Rezania et al. [
49], the temperature differences at the n- and p-type junctions of TE elements are not identical. Such temperature differences are lower in n-type TE elements, compared with those in p-type TE elements, due to the higher thermal conductivity in the n-type material. Consequently, the footprint size of the n-type element must be larger than the footprint of the p-type TE element, due to the higher thermal conductivity in the n-type material. Therefore, the optimal ratio of footprint areas to achieve the maximum generation and the best cost-efficiency ratio in thermoelectric modules must satisfy that An/Ap < 1, where An and Ap are the footprint areas of the the n-type and p-type junctions, respectively. Brito et al. [
41], found that when the thickness of the TE elements is smaller, the electrical resistance is reduced, but this will impact the ΔT of the TE module, because a lower thermal resistance will increase the thermal output and attenuate the temperature difference between the hot and cold sources. However, this will only occur if the usable hot source is low and the other thermal resistances are high enough to significantly affect the ΔT of the TE module. In their study, Dongxu et al. [
31] found that the thickness of the TE legs can be reduced to 1.1 mm, which is 4 mm less than commercial modules, while still achieving the same efficiency. In all these works, simulation tools were successfully used to find the relationships between the different parts of the modules and their respective powers.
3.6. Useful Life
The evolution in the design of TEG modules has also contributed to increasing their useful life. Moreover, the thermal stresses to which the TEG modules are subjected, affect them adversely by the formation of microcracks and the expansion and contraction of TE materials. To address this problem, Skomedal et al. [
40] incorporated spring-supported contacts in the legs of TEG modules to dampen thermal expansion and contraction in the modules; these authors concluded that good diffusion barriers and possible coatings can reduce oxidation of the hot side electrodes and interconnections. Likewise, they noted that the TEG module could generate 1 to 3 W/cm
2. Furthermore, Ming et al. [
45] studied, via numerical analysis, how the non-uniform flow causes the junctions of the TEG modules to be damaged and also decreases their output power. In addition, Merienne et al. [
32] investigated the effect of thermal cycles on commercial TEG modules, finding that when rapid temperature changes are applied, their output power can decrease by up to 61%, compared with a module in which the heat flow is constant and the temperature changes are minimal. Therefore, it is of utmost importance to analyze in detail the design of TEGs and the operating conditions to which they will be subjected in order to establish applications that allow them to extend their useful life.
The current commercial TEG modules have still not reached the economic feasibility, nor the efficiency required for thermal energy recovery applications. Therefore, as mentioned above, the characterization of commercial TEG modules, in the specific conditions to which they are going to be subjected, requires the use of expensive control equipment, or manual processes, which leads to difficulty in making long-term measurements. Therefore, Elzalik et al. [
28] developed a characterization method that allows precise and inexpensive estimation of the maximum power point and the dynamic parameters of the TEG module. The proposed procedure can be used with different sources of residual thermal energy and under different operating conditions.
In relation to the sustainability of TEG modules, authors such as Khanmohammadi et al. [
22] and Heber et al. [
21] state that the amortization period and the power ratio of commercial TEGs make them not viable to be used exclusively as a generation method. Therefore, they recommend using them as a complement to other methods of recovering residual thermal energy, thus generating an integrated system, or the development of segmented TEG modules that allow greater efficiency and profitability.
3.7. Thermoelectric Generators
As previously mentioned, the main limitation for the use of TEG is its cost-efficiency ratio; therefore, this type of generator still does not compete with conventional mechanical thermodynamic systems or the electrochemical systems of Rankine, Stirling, Brayton, expansion devices or fuel cells. These types of conventional generation systems have a wide application ranging from heat utilization with a lower output power range to greater than 10 kW. For their part, TE materials are usually utilized for applications with an output power of less than 10 kW [
51], which makes them suitable for specific power requirements.
Yazawa et al. [
40], focused their research on the economic viability that TEGs can achieve with respect to other types of electricity generation systems. The authors found that TEG modules with a ZT of 0.8 can achieve a cost-performance ratio of 0.86
$/W. This makes TEG systems competitive with other power generation systems. It was also found that TEG systems have commercial profitability if they have a cost-performance ratio of less than
$1/W [
34].
In the manufacture of TEG systems, the main component is the TEG module, and the other components such as heat sinks, and supports help the system to perform better. Mori et al. [
42] estimated that the TEG modules only represent 60% of the cost of whole systems and focused the design of the system on heat concentration structures that would help to double the efficiency of power generation and that could reduce the total cost in the system by half. Likewise, Hendricks et al. [
43] warned that the cost of the heat exchanger, which is the element of the system that most frequently increases the total costs of TEG modules, should be further investigated to achieve profitability of the system.
Lately, the automotive industry has initiated investigations into TEG systems due to the considerable losses of caloric energy that arise from the combustion process, which could be used to power other vehicle systems. Indeed, the incorporation of these energy recovery systems, could significantly reduce CO2 emissions into the atmosphere.
Arsie et al. [
47] proposed the incorporation of a TEG system in the exhaust of a car using commercial 14 Hz TEG modules. In their research, the temperature gradient was guaranteed using refrigerant on the cold side of the module; the system was connected directly to the battery and alternator of a vehicle, and using a longitudinal model, it was possible to determine that the system displaces the energy of the alternator by between 15 and 20%, having an average saving of ~1 g/km of CO
2 in standard driving cycles.
Fernández et al. [
34] used commercial Bi2Te3 TEG modules for heat recovery from light duty diesel engines as they produce ~386 W of recoverable power under common vehicle driving conditions. In their research they found that in commercial TEG modules, it is only possible to recover about 37.6 W, if no additional improvements, such as advanced heat exchangers, are applied. The authors also found that when a cooling system is able to maintain the cold side of the TEG module at 50 °C (less than the engine system coolant temperature), up to 75 W of power can be obtained.
Heber et al. [
21] in 2020, manufactured a TEG system for natural gas heavy vehicles, in which they used 168 commercial modules based on SnTe. The cost of the TEG was EUR 1811, and a maximum power of 1507 W was achieved with a power density 50 W/kg, a reduction in CO
2 emissions of 4.9 (9.4) g (CO
2)/km, and a cost-efficiency ratio of 1.2 EUR/W, which suggests that the system is profitable.
Likewise, the use of TEGs as complementary systems has been investigated to compensate the cost-efficiency ratio in different applications to reduce CO
2 emissions. Thus, Bellos et al. [
74] investigated the efficiency of a solar energy-induced TEG using commercial Bi
2Te
3 TEG modules, and carrying out a financial analysis, found that the cost of the investment would be 1 EUR/W, with a payback period of 4.55 years and a leveled cost of electricity of 0.0441 EUR/kWh, indicating that this system would be unprofitable.
3.8. Sustainability (Circular Economy)
As noted above, commercial TE materials are not yet sufficiently cheap, and high efficiency materials are not yet mass produced. Until now, the most commonly used commercial TE materials are Bi2Te3-based alloys because they have advantages such as easy bulk processing. However, although they are precursors, high energy expenditure and expensive techniques are used in their processing both for power generation and for cooling at temperatures close to ambient levels. On the other hand, TE materials use elements such as bismuth, tellurium, antimony, selenium, and lead, among others, which are expensive, scarce, and sometimes toxic. As mentioned earlier, from the point of view of the circular economy, the recycling of TEG modules could generate great economic benefits since it would allow obtaining raw materials for the manufacture of new TEG modules or other electronic devices, generating a reduction in the consumption of scarce elements. Moreover, they also generate environmental advantages because the improper disposal of these materials is avoided, which can benefit both the environment and human health.
At the time of writing this paper, the scientific publications on recycling TEG modules are still quite sparse. However, currently there are different ways to recycle TEG modules based on tellurium bismuth, from which three approaches can be differentiated based on the separation techniques: (i) chemical, (ii) thermal, and (iii) bacterial methods. On the other hand, in some cases only some parts of the TEG modules are recycled or only the elements of the semiconductors are recovered. According to the bibliographic review carried out, approaches have been proposed for the recycling of commercial TEG modules based on bismuth tellurium, by taking advantage of the differences in melting temperature of the constituent materials. In this way, the separation of the different constituents of commercial modules (plastics, Cu, Bi
2Te
3 and Al
2O
3) in an efficient way might be achieved by mechanical processing which relies on the entropy changes of these materials [
27].
The TE materials have been separated by thermal processes followed by chemical separation processes, in which the characterization of the materials of the TE modules was conducted by techniques such as differential scanning calorimetry (DSC), X-ray diffraction (DRX) and field emission scanning electron microscopy (FESEM), which give information on the material types, melting temperatures and the distribution of the materials. This allows their separation based on the differences in thermal and chemical properties. This type of separation is initially carried out by means of thermal treatments such as hot oil baths at 250 °C for the removal of solder from the -n (Bi
2Te
3) and the p-type (Bi
0.5Sb
1.5Te
3) semiconductors. Later they are subjected to a mixed acid solution (HCl and HNO3 in a 3:1 ratio) at room temperature. At this point, the Sb of the semiconductors precipitates. Then, the solution is then filtered, washed and sintered in order to obtain nano-powders of Bi
2Te
3 -n type with a particle size of ~15 nm purity [
25,
36].
None of the previous works reports a characterization of the thermoelectric properties of the recovered TE materials. The characterization would be of great importance in order to know if the processes used for their separation in any way affect the properties of the recovered products and their possible use in future applications.
Table 7 lists some recent works in relation to the final disposal of thermoelectric modules.
It is clear then that there is a global need for sustainable technologies [
75,
76,
77], and the circularity of materials and processes are areas where TE can have a significant impact as the main, partner, or complementary technology since it is a particularly adaptable technology [
78].