Humans are currently facing complex challenges such as rising energy costs, pollution and global warming. To reduce their consequences, scientists are concentrating on improving power generators focused on energy harvesting. Thermoelectric generators (TEGs) have demonstrated their ability to convert thermal energy directly into electricity through the Seebeck effect. Due to the unique advantages, they present, thermoelectric systems have emerged over the past decade as a promising alternative among other technologies for the production of clean energy (Oxotin, 2008; Sun et al., 2015; Lukisher et al., 1983; Xansen and Anderko, 1962; Onarkulov  et    al.,  2017;  Karimberdi  et  al.,  2020; Egamberdiyvich et al., 2020; Karimov et al., 2019; Azimov et al., 2021; Bozorovich et al., 2019; Onarkulov et al., 2021).

Switching is the most time-consuming and responsible operation in the manufacture of thermoelectric modules. Which include a significant number of junctions, failures most often occur due to broken contacts (Sun et al., 2015; Lukisher et al., 1983; Xansen and Anderko, 1962; Onarkulov et al., 2017; Karimberdi et al., 2020; Egamberdiyvich et al., 2020; Karimov et al., 2019).

To prevent the harmful effects of internal mechanical stresses, in the manufacture of thermoelectric batteries, it is necessary to use a  damping layer between the branch and the connecting plate (Lukisher et al., 1983; Xansen and Anderko, 1962; Onarkulov et al., 2017; Karimberdi et al., 2020; Egamberdiyvich et al., 2020; Karimov et al., 2019; Azimov et al., 2021; Bozorovich et al., 2019; Onarkulov et al., 2021; Daliev et al., 2020; Dmitriev and Zvyagin, 2010; Brostow et al., 2010). The damping layer is made of a material with sufficient plasticity and low ohmic resistance. It is usually made from lead.

The results of the research showed the advantages of using a single-wire technology for soldering with Pb-Sb eutectic solder using an anti-diffusion layer of chemically deposited nickel. In this case, the eutectic solder Pb-Sb in the contact simultaneously performs the role of a damper and a switching solder.

Technology for manufacturing prototypes of thermopiles

For experimental verification of the properties of the developed thermoelectric materials and switching technology, prototypes of thermoelectric modules were made, consisting of 24 and 95 thermoelements. Thermoelement legs were made of zone-directed crystals  Bi₂ Te_2.88 Se_0.12 +0.04 wt.% Ni +0.05 wt.% Sb with Q factor    2.7 10^-3 K^-1 and Sb _1.5 Bi _0.5 Te₃ +3 wt.% Te+0.05 wt.% Ni with a quality factor of 3·10^-3 K^-1.

Of thermoelement, legs were cut from an ingot of thermoelectric material by the electro spark method, and the height of the billets was equal to the height of the finished legs. Branch blanks were etched twice in a 50% nitric acid solution for 15 s. After etching, the workpieces were boiled in ethanol and dried in air. The ends of the thermoelement legs were chemically coated with a nickel-phosphorus film using a solution of the composition NiCl·6H₂O - 30 g/l, NH ₄ Cl-30 g/l with the addition of sodium citric acid up to 45 g/l. The alkalinity (pH) of the resulting solution was brought to a value of 10 to 10.5 with ammonia, and sodium hypophosphate was added to the solution in an amount of 15 g/l.

EXPERIMENTAL TECHNIQUE

The process of applying the nickel-phosphorus coating was carried out at a temperature of 85 to 87°C in the presence of a nickel plate for 4 min. After coating, the workpieces were washed in running and distilled water, wiped with alcohol, and annealed at a temperature of 225 ± 3°C for 2 h.

Copper busbars were coated with a nickel-phosphorus coating according to the method described in Xansen and Anderko (1962), Onarkulov et al. (2017), and Karimberdi et al. (2020). Blanks of thermoelement legs and busbars were tinned with Pb - Sb eutectic solder. After tinning, branches of thermocouples sized 2 × 2 × 4 were cut by the method of electric spark cutting mm and 3 × 3 × 4 mm and soldered together with current-carrying tires. Soldering was carried out at a temperature of 280 ± 3°C. Figure 1 shows the layout of the elements of a single thermoelement.

The quality control of thermopile switching was carried out by external inspection. The magnitude of the voltage  drop  was  tested on one thermoelement at the optimum current. The control was carried out by placing a millivoltmeter probe in turn on each subsequent thermoelement in the direction of the switching stroke. The value of the voltage drop on individual branches or thermoelements of the thermoelectric battery did not differ by more than ± 3%. Voltage drops exceeding the specified limit indicate their poor-quality switching. A significant increase in the voltage drop across the thermoelement corresponds to a switching failure as a result of thermal stresses or mechanical damage, and a decrease in the voltage drop from the norm corresponds to the presence of a short circuit in the battery, which most often occurs due to the influx of solder on the semiconductor.

The technology greatly simplifies the process of assembling thermopiles, increases their efficiency, and also makes it possible to mechanize the process of applying solder to a thermoelement, which is very important in serial production of thermoelectric batteries. In addition, it allows to significantly expand, for example, up to 450°K, the temperature of the environment in which the thermopile functions normally.

Thermoelectric module

An experimental verification of the previously obtained results was carried out on individual half-elements, with the manufacture and testing of cooling thermopile at elevated temperatures. The external view of the experimental module, consisting of 24 thermoelements with a total working surface of 875 mm², as shown in Figure 2.

Tests of prototype thermopile samples to achieve the maximum temperature difference were carried out in a vacuum of the order of 10^-3 Pa with an increase in the temperature of the hot junction from 320 to 495°K. Cooling was carried out with running water; the heat load to the cold junction of the thermopile came from an electric heater. The temperature at the junctions was controlled by chromel-alumel thermocouples (ø= 0.3 and 0.5 mm) in an electrically insulating braid, directly behind thermopile bars minted into electrically conductive tires on the hot side and metal plates on the cold side.

EXPERIMENTAL RESULTS

The maximum temperature difference ΔT_max for one of the main characteristics of a cooling thermopile. The real efficiency (Z) of a thermopile is quite simply and reliably determined from the maximum temperature difference it develops. This way of determining Z is reduced to measuring T and T_0min at the junctions of the thermopile under conditions of its adiabatic isolation. But in our case, the testing of cooling thermopile at elevated temperatures artificially creates thermal loads to the cold junction, as a result of which the efficiency of the module is underestimated.

The dependence of the thermoelectric efficiency of the module on the temperature of the hot and cold junctions in the operating mode is presented in Table 1.

As can be seen from Table 1, the thermopile power supply parameter was maintained with an accuracy of ~6%.

The results of measurements of ΔT under conditions of adiabatic insulation (1) and in a thermal load of 3 W to the cold junction (2) are as shown in Figure 3. It can be seen  from  Figure  3  that  the  thermal  load  to  the  cold junction leads to a significant decrease in ΔT (up to 40), while the resistance of the thermopile increased from 0.48 to 69 Ω. However, it should be noted that a rather significant decrease in ΔT_max and an increase in resistance to a cold junction under thermal load can be explained not only by the influence of contact resistances, but also by the specific conditions for the technological process of manufacturing prototypes.

Here, the influence of two contact resistances between the branches and the connecting plate is taken into account. Taking into account two more contact resistances on the hot sides of the thermoelement legs,

we can write:

where U is the voltage drop across the ohmic resistance of the thermoelement branch, V; S- cross-sectional area (cm²), I- height of thermoelement leg (cm); and p - electrical resistivity of the material, Ω cm.

Thus, Xansen and Anderko (1962) make it possible to calculate the value of the contact resistance indirectly, based on measurements of the voltage drop across the thermoelement, at the optimum value of the operating current I_opt and ΔT= 0 (ΔT=T_h-T_x), at T_h = T_x

Table 2 shows the contact resistance values determined by this method and for comparison, the values of r_c measured by probe methods are given.

Comparative evaluation of contact resistance depending on the parameter of thermoelements

Temperature is one of the main factors affecting the structure, that is, on the intramolecular relationship in the volume of the solder layer. Therefore, any change in this layer, for example, the appearance of microcracks, solder loosening, delamination of the nickel film from the working surface of the branch, will lead to a change in the resistance of the layer. Therefore, it is very important to establish the effect of temperature cycling alternating cold - heat on the resistance of the solder layer. If the narrowing and expansion of the volume of the interlayer with a change in the temperature of the latter will be within acceptable limits, then the resistance of the interlayer will not change. In order to determine the effect of temperature cycling,  the  following  study  was  carried out. After the thermoelement enters the mode, that is, when the temperature at the cold junction reached the minimum value T_min, by changing the direction of the current, the cold junction was heated up to 100°C. Then, switching to the cooling mode, the cold side was again reached T_min within 20 min. After each repetition of the specified cycle 10 times, the contact resistance was measured according to the method described earlier. The obtained results of the dependence of r_c on the temperature shock number n are as shown in Figure 4.

Measurements show that the contact resistance does not exceed ~1.2^.10⁵ Ωcm², which fully satisfies the requirements for high-quality switching of thermopile.

Thus, the proposed method for determining the contact resistance of contacts is a simple and affordable method and does not require acutely scarce highly sensitive measuring instruments. This technique allows us to investigate the influence of various factors on the value of the switching resistance.

Pb-Sb found that the activation energy diffusion of the solder is greater (52500 J/mol) in the case of switching contacts with eutectic alloy Pb-Sb than switching solder Bi-Sb. This indicates that the eutectic alloy Pb-Sb is a promising switching material for cooling thermopiles operating at elevated temperatures as well as for thermocouples operating in vacuum.

As a result of the study of the physico-chemical compatibility of the eutectic alloy of the Pb-Sb system with Bi₂ Te_2.88 Se_0.12 and Bi_0.5 (Sb)_1.5 (Te)₃ solid solutions, the possibility of using this alloy, which has a melting point of 525 K as a solder for single-layer switching of thermopiles operating in the temperature range 200-450°K in vacuum.

It has been established that chemical deposition of nickel on contact surfaces excludes reactive diffusion of solder components up to 473°K and provides good wettability of both Bi-Sb and Pb-Sb solders.

An installation for controlling the mechanical strength in compression of semiconductor materials has been developed   and  manufactured.  Tests  were  carried  out under conditions of repeated thermal cycling.

The authors have not declared any conflict of interests.