1 Introduction

The average global surface temperature has escalated by 1.09 °C in 2011 – 2020 above that in 1850 – 1900 [1]. Anthropogenic climate change has caused extensive adverse impacts, associated losses and damages to nature and people, and includes more frequent and intense extreme events, beyond natural climate variability. If global warming surpasses 1.5 °C in the next decades or later, it is predicted that many human and ecological systems will encounter additional severe threats, some of which may even be irreversible. Realizing these threats, the international community decided to take steps to stop or at least slow down global warming and climate change, the result of which led to efforts such as Kyoto Protocol (1997). It is a legally enforceable disposition under which industrialized countries had to lower their aggregate greenhouse gas emissions by 5.2% in comparison with the year 1990 [2]. The most important of these was reduction of carbon dioxide (CO2) emissions which is an inexorable derivate of many industrial activities.

The global human population is increasing at an exponential rate and is expected to reach 9.7 billion by 2050 and 10.9 billion by 2100 [3]. The population growth has increased the demand on building sector, and it is estimated that approximately 50% of natural materials are utilized to build structures and related amenities [4]. In a building, heating ventilation and air-conditioning (HVAC) are found to consume most of the energy for cooling and heating purposes. The rising demand for refrigeration and air-conditioning has resulted in a rise in electricity production and associated CO2 emissions in recent years [5]. According to the International Institute of Refrigeration in Paris, various refrigeration and air-conditioning applications consume 15% of global electricity, and the energy consumed for air-conditioning in various households and commercial buildings accounts for nearly 40% of the total. Most of the energy comes from non-renewable energy sources, mainly fossil fuels. The accelerated consumption and utilization of fossil fuels cause severe environmental and energy issues such as global warming, ozone layer depletion, atmospheric pollution, and global energy crisis. One of the hardest hit industries by the Kyoto Protocol was the refrigeration industry, as most of the refrigerants being used at that time were having huge global warming potential (GWP). All these problems are forcing us to look for alternative cleaner solutions for refrigeration and air-conditioning.

1.1 Conventional refrigeration and air-conditioning systems

The most used refrigeration and air-conditioning systems today are vapour compression (VC) and vapour absorption (VA) systems [6, 7]. Vapour compression systems have high coefficient of performance (COP) in the range of 2–3 for commercial systems. They are comparatively cheaper and have large cooling capacities. All these desirable characteristics make them the most attractive refrigeration and air-conditioning system today. They use a compressor for its operation, which makes the system noisy in certain applications. Vapour compression systems use refrigerants that are derivatives of hydrofluorocarbons (HFCs) or hydrochlorofluorocarbons (HCFCs) which has high GWP and ozone depletion potential (ODP). Due to this reason, the VC systems may be slowly phased out in the future. Even though modern VC systems use environment-friendly refrigerants having very low GWP and zero ODP, their COP is found to be less than that of the previously used refrigerants.

VA system works on heat-operated cycle and is mostly used in industries where waste heat is available. They have an absorber/adsorber and generator in place of a compressor and hence are least noisy in operation. Their COP varies with applications and are generally found to be less than that of VC systems and close to 1. VA systems are bulkier in construction and are more costly compared to VC systems for the same cooling capacity.

There are some other non-conventional refrigeration and air-conditioning systems like thermoelectric (TE) systems, magnetic cooling systems, thermo-acoustic refrigeration systems, electrochemical refrigeration systems, and ejector refrigeration systems which are all very promising technologies which are still in the development phase and not used widely for commercial uses.

It is estimated that air-conditioning and refrigeration systems contribute about 15% of world electrical energy demand. The conventional refrigeration systems depend on fossil fuels, and the rapid depletion of non-renewable resources and the associated emissions lead to the development of alternative solutions which employ renewable energy resources for refrigeration.

Thermoelectric (TE) refrigeration systems working on the principle of Peltier effect are an alternative for the conventional systems. The thermoelectric refrigerators will not produce any noise and vibration due to the absence of any moving parts. They are refrigerant-free as electrons act as heat carriers. The greatest advantage of a TE system is that it can directly be powered by solar photovoltaic (PVs) since they give a DC output.

2 Thermoelectric coolers

2.1 History

Thermoelectric (TE) energy converters are solid-state devices that can convert thermal energy from a temperature gradient into electrical energy [8]. In 1821, Thomas Johann Seebeck, a German physicist, found that when two or more dissimilar conductors are joined together and the junctions are kept at different temperatures, an electromotive force (emf) and consequently an electric current are produced in the conductors. This phenomenon is called the “Seebeck effect.” The reverse of Seebeck effect where electrical energy can be converted to thermal energy for cooling and heating is called Peltier effect and was discovered in 1834 by the French Physicist Jean Charles Athanase Peltier.

Altenkirch [9] was the first to present the theory of thermoelectric power production and refrigeration. In 1911, Altenkirch derived the thermoelectric efficiency, which is now referred to as the thermoelectric figure of merit (Z) [8]. Thermoelectric industry started developing rapidly in the 1950s with the emergence of new materials, especially semiconductors, which proved to be far superior compared to dissimilar metals used at that time [5]. The thermodynamic efficiency is expressed in dimensionless form by multiplying figure-of-merit Z with absolute temperature T to get dimensionless figure-of-merit ZT.

$$ZT = {\alpha }^{2}\sigma T/K$$
(1)

where α is the Seebeck coefficient, σ is the electrical conductivity, and K is the thermal conductivity. Bulk alloy materials such as Bi2Te3, PbTe, SiGe, and CoSb3 are the conventionally used thermoelectric materials, among which Bi2Te3 is the most commonly used one [10]. All these materials have a ZT value less than 1. The improvement in ZT was modest from 1960s through 1990s. Theoretical predictions after the mid-1990s suggested that nanostructural engineering could considerably improve thermoelectric material efficiency [11].

2.2 Why thermoelectric coolers?

There are several advantages associated with thermoelectric coolers, some of which includes solid-state operation, vast scalability, the absence of toxic residuals, maintenance-free operation due to lack of moving parts or chemical reactions, and reliability with a long-life span [8]. They can easily operate under steady-state condition for more than 100,000 hours [5]. Moreover, they run on direct current (DC). They are refrigerant-free, and electrons act as heat carriers. They can be directly powered by solar photovoltaics (PV) and fuel cells which will ultimately have a huge impact in mitigating problems related to global warming and climate change, if implemented on a large scale.

Bansal and Martin [6] conducted a comparative study of vapour compression, vapour absorption, and thermoelectric refrigerators as a part of their study to investigate a cooling system which does not utilize any refrigerant that causes ozone layer depletion. For this, they selected three refrigerators of similar capacity (about 50 L). The energy consumption for 24 h of operation when the ambient temperature was 16.6 °C and cabin temperature was maintained at 5 °C, the COP and the noise intensity level while in operation are shown in Table 1 The VA system was the quietest in operation. The total cost of acquiring and operating the systems over their full life cycle was the highest for VA system, followed by TE and VC system.

Table 1 Energy consumption and COP [6]

Riffat and Qiu [7] compared thermoelectric (TEAC), vapour compression (VCAC), and absorption (AAC) air-conditioners. The electrical energy consumption and COP for these systems are shown in Table 2. Indoor noise levels were found to be approximately the same for all the cases, whereas outdoor noise varied to a larger extent. The compressor in VCAC and liquid pump in AAC generated considerable noise, and in the case of TEAC, the fan or blower was the noise source. The sum of purchase and operating costs was the highest for TEAC and lowest for AAC.

Table 2 Electrical energy consumption and COP [7]

In the case of both refrigerators and air-conditioners studied in [6] and [7], the COP of TE system is on the lower side, making it less competitive with VC and VA systems. Hermes and Barbosa Jr. [12] compared the thermodynamic performance of Peltier, Stirling, and VC portable coolers and reported that the thermodynamic efficiency of a TEC is approximately 1%, and to reduce the internal and external irreversibilities, serious improvements are required. However, considering the environmental benefits of a solar PV-powered TE system, it can be considered as an excellent choice for certain niche applications.

2.3 Performance parameters of thermoelectric coolers

There is an ideal value of electric current that provides maximum COP for a certain thermoelectric module with specified hot/cold side temperatures [13].

$${(COP)}_{c,max}=\frac{{T}_{c}}{{T}_{h}-{T}_{c}} \frac{\sqrt{1+Z{T}_{m}}-\frac{{T}_{h}}{{T}_{c}}}{\sqrt{1+Z{T}_{m}}+1}$$
(2)

where ZTm is the figure of merit at Tm, the average of hot and cold side temperatures. Figure 1 shows the cooling COP variation of a TE module operating under optimal value of current and a constant hot side temperature of 300 K.

Fig. 1
figure 1

Variation of cooling COP with cold side temperature for a thermoelectric module operating under optimal value of current [13]

It can be seen from Fig. 1 that increase in ZT value is a desirable characteristic for improved COP. It can also be noted that hot side temperature, being fixed at 300 K, the COP is increasing with increasing cold side temperature (or decreasing temperature difference between the hot and cold sides). Therefore, proper heat dissipation from the hot and cold side is required to maintain a higher value of COP.

From (1), it is clear that Seebeck coefficient α, electrical conductivity σ, and thermal conductivity K affect the figure of merit. These three transport parameters are interdependent on each other as a function of band structure, carrier concentration, and other factors. In particular, α and σ are related in a reciprocal manner by Wiedemann–Franz law making improvements in Z-value difficult. Figure 2 graphically shows the relationship between the parameters that affect ZT value with carrier concentration.

Fig. 2
figure 2

Variation of Seebeck coefficient (α), electrical conductivity (σ), thermal conductivity (K) and ZT with carrier concentration [11]

The electrical conductivity and Seebeck coefficient are inversely related, which again serves to fix the value of ZT for any given bulk material [14]. Ideal thermoelectric materials should have a high electrical conductivity to permit conduction of electricity, which in turn creates a potential difference across the module and low-thermal conductivity to maintain the temperature gradient between the hot and cold sides [15]. The very first materials to be used as thermoelectric materials were two dissimilar metals, and they did not possess ideal thermoelectric characteristics. Most conventional TE materials are good electrical conductors, but at the same time, they are good conductors of heat. The quantity α2σ is referred as power factor in the case of thermoelectrics as both α and σ are determined by electronic properties [8]. In thermoelectrics, thermal conductivity is the total of two contributions: heat transported by electrons and holes (Ke) and phonons travelling through the lattice (KL) [16]. It has been found that optimal range of electrical resistivity for a thermoelectric material is between 10−3 and 10−2 Ωm [17].

The performance of a thermoelectric cooler depends on various parameters like applied electric current, hot and cold side temperatures, the electrical contact resistance between the cold side and device’s surface, thermal and electrical conductivities of the thermoelements, and the thermal resistance of heat sink on the hot side of the thermoelectric cooler [18]. The required cooling capacity and highest electric current are the main factors which determine the number of thermoelements in a thermoelectric cooler [19]. The influence of parameters like Seebeck coefficient, thermal conductivity, electrical resistivity, band structure and energy gap in semiconductors used in thermoelements, charge carrier concentration, mobility of charge carriers, diffusion properties, oxidizability, brittleness, coefficient of thermal expansion, compression, and shear strength on the performance of thermoelectric coolers has been studied in detail by M. Hamid Elsheikh et al. [8] and D. Enescu and E. Virjoghe [20].

2.4 Studies focused on improvement of figure of merit, ZT

The electronic structure of a material plays a major role in determining its figure of merit [21]. There are metal-based, ceramic, polymer, and semiconductor-based thermoelectric materials [8]. ZT maximization can be attained by enhancing the power factor and reduction of thermal conductivity [22]. Some of the key takeaways from [13] are as follows:

  • Bismuth telluride (Bi2Te3), the most used and best commercially available thermoelectric material for low-temperature applications, has ZT < 1.

  • ZT has to be in the range of 2–3 to be competitive with vapour compression cooling systems.

  • If a ZT value of 6 can be attained, then that thermoelectric device would be able to cool from room temperature to cryogenic temperatures (77 K).

Modern synthesis and characterization techniques have made it possible to manufacture nanostructured constitutes incorporated traditional bulk materials, and they have been found to increase the efficiency of thermoelectric coolers [10]. The improvements in ZT factor comes from two pathways: bulk samples with nanoscale constituents and nanoscale materials themselves.

Since ZT value of commercially available thermoelectric materials is on the lower side to be used for cost-effective, inexpensive applications; various endeavours are being made to ameliorate them. One of the ideas is reducing KL of a compound by swapping crystal lattice with a glass-like amorphous composition, which is the so called phonon-glass electron-crystal (PGEC) material [11, 23]. In PGEC, heat and charge can be readily transported by high mobility electrons, while phonons are interrupted at the atomic scale, which keeps it from heat transport [10]. Some of the chief bulk thermoelectric materials developed recently, having better efficiencies are skutterudites, clathrates, and half-Heusler alloys, which are fabricated through doping. Low-dimensional materials like 2D quantum wells, 1D quantum wires, and 0D quantum dots process the quantum confinement effect of electron charge carriers, which can boost the Seebeck coefficient and, consequently, the power factor. Moreover, the increased number of interfaces helps in scattering phonons more beneficially than electrons so that thermal conductivity (contribution due to KL) is lowered more than electrical conductivity [24].

H. Liu et al. [25] created liquid-like behaviour of copper-ions around a crystalline sublattice of Se in Cu2-xSe which resulted in inherently very low lattice thermal conductivity enabling a considerably high value of ZT of 1.5 at 1000 K. This was a breakthrough for creating high-efficiency thermoelectric bulk materials by investigating systems with a crystalline sublattice for electronic conduction bounded by liquid-like ions.

Some of the best thermoelectric materials having comparatively high values of figure of merit are presented in Table 3. The highest ZT value is about 3, reported by Harman [26].

Table 3 Summary of high figure-of-merit thermoelectric materials

Thermoelectric coolers having ZT value of 1 operates at around 10% of Carnot efficiency. To be comparable to conventional domestic refrigeration technologies, 30% of Carnot efficiency is required which can be achieved with the help of a thermoelectric cooler having ZT value of 4. Incrementing ZT to 4 has proved insurmountable as of now [34].

3 Solar cooling technologies

Solar energy that is accessible freely and in abundance can be directly converted to electricity using solar cells connected in series and parallel in a photovoltaic (PV) panel. A PV panel can directly convert solar energy into electrical energy without the use of any moving parts, emission of dangerous gases or use of fossil fuels [4]. The efficiency and purchase cost of solar cells vary widely depending on the material and method of manufacture [2]. Most of the commercially available solar cells are made of silicon. The efficiency of a solar panel is given as the ratio of power W (kW) to the product of solar panel surface area As (m2) and the direct irradiation of solar beam Ip (kW/m2). When determining the nominal efficiency, Ip = 1 kW/m2 is used.

$${\eta }_{sol-pow}=W/({I}_{p} \times {A}_{s})=W/{Q}_{s}$$
(3)

On a clear day noon, a high-performance solar panel yields a maximum efficiency of 15%. The refrigeration machine efficiency is determined as the ratio of cooling power (Qe) to the work input (W).

$${\eta }_{pow-cool}={Q}_{e}/W$$
(4)

The overall efficiency or the solar-to-cooling efficiency is given by multiplying (3) and (4) as:

$$\eta_{sol-cool}=\eta_{sol-pow}\times\;\eta_{pow-cool}=Q_e/Q_s$$
(5)

A representational diagram of a solar photovoltaic panel is shown in Fig. 3.

Fig. 3
figure 3

Schematic diagram of a solar photovoltaic panel [2]

Some challenges associated with large-scale commercialization of solar PV powered cooling systems are that the power generated by a photovoltaic panel fluctuates throughout the day with solar insolation, weather conditions, shading etc. and throughout the year with seasons. It is also dependent on the location where PV panel is being used. To ensure continuous and steady power supply to the cooling system, electric battery or usage of solar and grid power in tandem may be required, making the system complex and expensive. The cost of PV panels themselves are higher to be competitive with regular grid powered cooling systems. Again, since the efficiency of solar PV panels are on an average 10%, combining with the efficiency of the cooling system, the overall efficiency may be much lower.

Solar cooling technology may be categorized into three classes namely solar electrical cooling, solar thermal cooling and solar combined power and cooling [34]. Comparative studies [2, 34] revealed that solar thermal cooling options where the thermal energy obtained from solar energy is converted to useable cooling through thermochemical or thermophysical processes with the help of thermally activated energy conversion systems were better in terms of economy and performance compared with solar electrical cooling with PV panels for converting solar energy to electrical energy. Figure 4 shows routes of energy conversion for solar cooling.

Fig. 4
figure 4

Energy conversion routes for solar cooling [34]

4 Thermoelectric cooling applications

Due to its inherently low coefficient of performance, thermoelectric cooling has found applications only in niche areas such as space missions, scientific and medical equipment where COP is not as critical as dependability, and silent operation.

With the advancements in integration techniques, electronic devices are able to incorporate a large number of transistors within a small area, making dissipation of heat from them difficult. If heat is not dissipated properly, the temperature may rise beyond the prescribed safe limit rendering the device useless. In some cases, hot spots may be developed, and the thermal environment may be deteriorated within the device [35]. If electronic device performance continues to grow at the present rate, conventional cooling methods may not be able to dissipate the heat generated in limited time and space. A TEC can help to maintain the temperature below the safe limit by removing heat from the equipment [18]. The heat fluxes of microprocessors and central processing units (CPUs) are expected to increase beyond 100 W/cm2 for commercial uses [36]. Thermoelectric coolers together with liquid cooling and air cooling can be used to maintain the temperatures within safe limit in these cases [37]. Zhang, Mui and Tarin [38] conducted an analysis of TEC performance in cooling high power electronic packages such as processors using air-cooling techniques.

In civil markets, TECs are used as portable and domestic refrigerators, portable ice box, picnic baskets, beverage can cooler etc. [39,40,41,42]. Thermoelectric refrigerators are eco-friendly, robust, quieter and has precise temperature control [42]. The COP of both portable and domestic thermoelectric refrigerators are generally lower than 0.5 when working at a temperature difference of 20 – 25 °C. Thermoelectric refrigerators have the capability of maintaining a constant temperature in the cold chamber throughout its operation compared to VC refrigerators, which tends to have an oscillating pattern owing to the compressor on/off cycles [10]. In a VC system, this temperature fluctuation may be several degrees. This effect is detrimental to the preservation of food and other perishables. The main advantage of a thermoelectric refrigeration system is the absence of refrigerants.

Thermoelectric coolers have also found use in medical applications [43], laboratory and scientific equipment cooling such as laser diode or integrated circuit (IC) chip cooling [44, 45]. Thermoelectric air-conditioners (TEACs) are simple, reliable, and environmentally friendly and provide simple installation and support for complicated water distribution pipelines [8]. They can also be easily switched between heating and cooling modes by simply reversing the input current. However, the cost of TEACs is limiting it from being competitive with conventional technologies in the market [46]. Even though TEACs are portable and noise-free, their relatively low COP is a limiting factor. TEACs have large potential in air-conditioning of small enclosures such as cars and submarines where safety and reliability are more important [7]. Arenas et al. [47] and Vazquez et al. [48] presented active thermal window (ATW) and transparent active thermoelectric wall (ATW) for room cooling applications where conventional air-conditioning systems are hard to install as for old historic building repair and refurbishment. The window glass embedded thermoelements transfer heat through the glass in order to cool the room.

TECs can be used as microclimate cooling (MCC) systems that can help extract a considerable amount of heat from a soldier’s body while they are in combat clothing, which in turn helps to increase mission duration and enhance mission effectiveness [49]. TECs can be employed for automobile cooling applications. Yang and Stabler [50] presented a review of thermoelectric materials’ automotive applications. Qinghai et al. [51] presented an innovative thermoelectric truck cab air conditioner with a COP of 0.4 – 0.8 under ambient temperatures from 46 – 30 °C. Choi et al. [52] developed a car seat system with a thermoelectric temperature-control device which can cool or heat the car seat as required. There is a large market for thermoelectric coolers in automotive industry.

Photovoltaic (PV) modules are subjected to high outdoor temperatures, resulting in reduced efficiency. Using the thermal waste with the help of thermoelectric modules at the back of PV panel, forming a photovoltaic-thermoelectric (PV-TE) hybrid module. Supplementary electricity production is possible using a PV-TE module. Sark [53] found out that this system is able to give efficiency enhancement by up to 23% for roof integrated PV-TE modules, and the yearly energy yield was found to escalate by 11 – 14.7%. Even though the thermoelectric module here acts as thermoelectric generator, they help in cooling down the PV panels.

There are a large number of applications where TECs prove to be better than conventional VC and VA systems. Using solar PV to power the TECs can further increase the appeal of TECs. Solar PV-powered TECs can be used to provide a cold storage for food, vaccines, medicines, and other perishables in remote areas where grid electricity is not available [5]. Such a device can help people working in remote outdoor areas to store their drinks and food and people living in remote areas to store vaccines and medicines such as insulin. In these cases, there are no better options than solar-powered TECs. In fact, the World Health Organization (WHO) and International Health Organizations were some of the first organizations that encouraged research in development of PV integrated Peltier coolers.

5 Solar photovoltaic-powered thermoelectric coolers

One of the earliest studies conducted on solar thermoelectric refrigerator was by Vella et al. [54]. This study demonstrated that a thermoelectric generator that receives heat from the sun is an appropriate source of electricity for the functioning of a thermoelectric refrigerator since a TE module requires only a small e.m.f and moderately high current. Hence, thermoelectric refrigerators powered by PV modules is a highly compatible combination. Even though the efficiency of such a system is on the lower side, the simplicity of construction makes it attractive. The schematic diagram of proposed solar thermoelectric refrigerator is shown in Fig. 5.

Fig. 5
figure 5

Schematic diagram of solar thermoelectric refrigerator proposed in [54]

The ratio of number of thermocouples required for the two devices were numerically calculated to be 4:1 for practical situations even though theoretically it can be 1:1. Temperatures below 0 °C were obtained on the cold side of the TE module with a temperature difference of 40 K across the generator. The COP of the system was obtained as 0.055 even though the theoretically predicted COP was 0.12. The experimental design used Bi2Te3 thermoelectric modules available commercially, whose ZT was approximately 0.7. One of the challenges faced during the design was that of heat sink. Since there would be a close contact between the refrigerator and generator, there is a possibility of common heat sink temperature. Only slight temperature rise of the sink of the refrigerator can be permitted if the COP is not to decrease drastically so that the generator sink should also rise in temperature by a small amount. It was estimated that a temperature rise of 10 K for the generator sink would be acceptable so that a large dissipation area does not have to be provided. Additionally, it was easier to maintain the temperature of refrigerator sink within 1 – 2 K of ambient temperature as the heat dissipated to this sink is small. The plan of the actual experimental model is shown in Fig. 6.

Fig. 6
figure 6

Plan view of 4 thermoelectric generating couples combined with 1 refrigerating couple. A is a flat-plate collector, and B is a copper link between thermoelements at the hot junction. C shows the copper links connecting the thermoelements in contact with the heat sink. D. G1 and G2 are two of the generating pairs, and the shading indicates the positions of the thermoelements. The refrigerating couple (thermoelements shaded) is represented by R, which has copper links F in proximity with the heat sink and copper braid connections to the generator. Broken lines indicate the position of components that have been removed [54]

The experimental tests were conducted during winter, and cold water at 16 °C was circulated to cool the heat sinks. The heat sinks of both thermoelectric refrigerator and generator elements were at 24 °C. The cold junction of the refrigerator reached − 4 °C with no load. This study laid a solid foundation for the possibilities of solar energy-powered thermoelectric refrigerators.

Hara et al. [55] studied the cooling capabilities of solar cell-powered thermoelectric cooling prototype headgear. The headgear was designed to cool the forehead of people working outside under the sun. In the front, a thermoelectric element was installed, and solar cells were positioned on the top and brim of the headgear. Flexible sheet type amorphous solar cells were used due to their light weight and flexibility. An axial fan was also provided to gently blow the cold air towards the forehead. Air cooled fins were also used for heat dissipation from the hot side. The schematic of the model is shown in Fig. 7.

Fig. 7
figure 7

Prototype of the headgear with thermoelectric cooler driven by solar cells [55]

Three models were designed with different goals: to achieve maximum cooling capacity, minimum weight, and a third model designed for practical use. Model A weighed around 465.9 g, but the electrical power output was largest in this case. Model B used flexible paper-type solar cells, and the total weight was reduced to 135 g. Power output was lower than that of model A. Model C which had the same weight as that of model B was the final design that would be used outdoors. All the three models were tested for thermal comfort in the situations of sitting, walking, and bicycling. Lower temperatures up to − 10 °C were achieved, and it was found that reduction in temperature of 4 – 5 K from the ambient temperature is ideal for thermal comfort outdoors. This study proved that sufficient cooling capacity can be obtained by the TE element driven by solar cells attached on the top of the cap or hat.

Dai et al. [56, 57] conducted an experimental study and performance investigation on a portable, solar cell-driven thermoelectric refrigerator. It was intended to use in rural and remote areas where grid connectivity is not available. In remote areas, it can be used for storage of food, vaccines, etc. It will also help people working outdoors in road construction and mineral prospecting to keep their food fresh and drinks cold. The schematic of the portable TE refrigerator is shown in Fig. 8.

Fig. 8
figure 8

Schematic representation of thermoelectric refrigerator driven by solar cells [56]

The portable TE refrigerator uses solar cells to convert solar energy directly into electrical power using photovoltaic effect in the daytime. If the power produced is in surplus, it is accumulated in a storage battery which is designed based on the number of days of autonomy required, besides driving the refrigerator. On overcast or wet days, when the solar cells cannot generate enough power to drive the refrigerator, the storage battery can offer backup. The controller helps optimize this operation. To disperse heat from the hot side of the TEC, a fin-type heat exchanger is employed. Efforts were made to select the components in such a way as to minimize the total weight and cost of the unit. The experimental work in this study focused mainly on the daytime cooling when solar energy was available.

The inclination of the solar panel was kept at a constant value of 35°, and it was found that the output voltage remains at a constant value of 12.0 ~ 12.4 V when the solar insolation varies from 880 to 770 W/m2. A 500 mL water bottle initially at ambient temperature was kept inside the refrigerator, and the variation of cold side temperature, ambient temperature, and water temperature is shown in Fig. 9.

Fig. 9
figure 9

Variation of cold side temperature and water temperature [56]

The transient cooling production and COP decrease with temperature on the cold side as shown in Fig. 10. COP drops from 0.4 to 0.25 and cooling production from 26 W to nearly 12 W eventually.

Fig. 10
figure 10

Variation of COP and cooling production with time [56]

From the study, it was also concluded that COP in general decreases with increase in hot side temperature, and hence, proper heat dissipation mechanisms should be employed. COP first increases sharply and then decreases gradually after reaching a maximum value with electric current and solar insolation. The optimum cooling production was found to occur at an insolation between 500 and 800 W/m2. It was concluded that the portable refrigerator could sustain a temperature of 5 – 10 °C in the refrigerated space with a COP of about 0.3. The authors also indicate a possibility of further optimization of performance by better matching of solar PV and TE module.

Abdul-Wahab et al. [58] designed and performed experimental studies on a portable solar-powered thermoelectric refrigerator. The primary goal of this study was to fabricate an economical solar-powered thermoelectric refrigerator for the Bedouin people living in secluded areas of Oman where electricity supply is unavailable. Ten thermoelectric modules were utilized in the design, and rectangular fins were used for dissipating heat from the hot side. Cooling fans were also used to supplement the heat transfer. Aluminium box was used in the refrigerated space to ensure even temperature distribution. 12.5 cm × 12.5 cm solar cells with an efficiency of 14% were used to power the refrigerator. The number of solar cells required to be connected in series and parallel was also determined mathematically. A total of 64 solar cells were used to provide 115.2 W that was enough to drive the refrigerator as well as the cooling fans. Different voltages were applied across the TEC, and it was found that when the current and voltage were 2.5 A and 3.7 V, respectively, a maximum temperature difference of 26.6 °C was possible. When the portable TEC was examined outdoors, it was observed that the temperature in the refrigerated space fell from 27 to 5 °C in 44 minutes. Numerical calculation based on a 0.5 L canned drink in the refrigerator, when cooled from 26 to 4 °C in 50 min, yielded a COP of 0.16.

Atik and Yildiz [59] experimentally studied the performance of a TEC with dimensions of 620 × 595 × 1565 mm. Eight TEMs having 127 thermocouples were used. The schematic of the TEC is shown in Fig. 11, and the heat sinks attached on the cold and hot sides are shown in Fig. 12.

Fig. 11
figure 11

Schematic of the TEC used in [59] (1. steel sheet; 2. polyurethane; 3. steel sheet; 4. foam; 5. fan; 6. aluminium heat sink (AS80); 7. copper plate; 8. heat insulation; 9. mica; 10. TE module; 11. aluminium heat sink (AS90); 12. galvanized sheet)

Fig. 12
figure 12

Heat sink design used on hot and cold sides respectively [59]

12 V DC fans were used on both inside and outside the TEC which were powered directly by the solar PV having an efficiency of 20%. The experiment was conducted on a sunny day for a period of 5 h. To maximize the solar radiation, the solar cells were rotated using the moving pods. The system reached steady state in the 200th min, and a temperature difference of 14.7 °C was maintained with a COP of 0.58.

He et al. [60] conducted a theoretical and an experimental study on a solar energy-driven thermoelectric heating and cooling system. Buildings with TE heating and cooling system can effortlessly switch between the two modes, simply by swapping the direction of flow of current. This study is focused on building integrated solar photovoltaic/thermal (BIPV/T) technology for use in low-carbon buildings. Here, the heat exchanger pipes are welded at the bottom of solar panel to transport the heat away from the PV panel, increasing its efficiency. Heat generated on the hot side of the TEC is also carried away by the circulating cold water. The radiator and fan tied to the cold side help to distribute the cold more effectively, helping faster cooling of the room. An experimental room made of foam box with a volume of 0.125 m3 is used in the study. The system arrangement is shown in Fig. 13.

Fig. 13
figure 13

System arrangement [60]

The cold side of TE device is secured tightly to the radiator and hot side to a copper plate with the help of silicon grease, which helps to augment the heat transfer. The heat pipe used to carry away the heat from the hot side is inserted into a groove made in the same shape as that of the heat pipe on the other side of copper plate. Thermoelectric device used was 127–03 with rated voltage of 12 V and rated current of 3 A. CPU fans were used as radiator fans because of their low-power consumption. Solar panels of area 0.5 m2 were used. The volume of storage tank used is 18.5 L. Solar irradiance, current and voltage across the TEC, cold and hot side temperatures of TEC, and the temperature of water in the storage tank were noted. The experiment was conducted on a summer day for cooling the model room. Six-hundred millilitres of pure water was kept in the test room to improve the heat loss. Results indicated that a lowest temperature of 17 °C could be achieved with a COP greater than 0.45. The water temperature in the storage tank rose by 9 °C from the ambient. The electrical and thermal efficiency of the system are 10.27% and 12.06%, respectively. The same system was also simulated to get results that were in close agreement with experimental values.

Hans et al. [61] conducted an experimental investigation and analysis on solar photovoltaic system-driven thermoelectric cooler, like that done by Dai et al. [55]. The experiment was conducted in Gurgaon, India, in the month of May from 9 AM to 6 PM, and it was found that the TEC system could sustain a temperature of 10 – 15 °C in the refrigerator with a COP of 0.34.

Rahman et al. [62] conducted a performance and life-cycle investigation of a solar-powered portable thermoelectric refrigerator. To use the full ability of the TE module, the heat dissipated from the hot side of the module is utilized for low-temperature heating applications. A well-insulated rectangular box made of aluminium is used as the cooling chamber. The design objective was to attain 10 °C in the cold chamber. A schematic representation of the test setup used is shown in Fig. 14.

Fig. 14
figure 14

Schematic illustration showing the connection of all major components of the refrigerator [62]

The dimensions of the refrigerator selected for the study is 6.5 cm × 6.5 cm × 15 cm. During the daytime, PV module directly powers the refrigerator, and a battery is also provided to ensure continuous and steady operation. A DC-DC converter provides a constant voltage to the refrigerator. A pinned type of heat sink was used to dissipate the heat from the hot side. Small fans were also utilized to cool the fins. Five TE1-12706 modules connected in series were used for the TEC. The values of optimum voltage and current were 0.8 V and 1.5 A. PV modules were sized according to the total power consumed by 5 TE modules and 2 fans, which was found to be equal to 46.5 Wh.

The effect of fin on both hot and cold side temperatures was studied, and it was found that both cold and hot side temperatures could be reduced with the help of fins. The effect of input current was also studied in a controlled space inside the laboratory. It was found that cold chamber temperature remained almost stable and within the design criteria with small changes in input current. Figure 15 shows the performance of the refrigerator in the actual ambient conditions of 35 °C and 900 W/m2 solar insolation at 2 PM in Sharjah, UAE.

Fig. 15
figure 15

Variation of temperature of cold side of TEM (Ts), middle of the cold chamber (Tcs) and hot side of TEM (Th) with time at 2 PM [62]

It can be noted that within a short period of 2 – 3 minutes, all the temperature variations were stabilized and remained constant for the entire span of the experiment. The temperature inside the refrigerator was 19 °C. Utilizing the heat from the hot side, the hot side temperature was reduced. By utilizing the heat from hot side, the COP was increased from 0.306 (when the heat load was not used) to 0.61, considering the heating effect. The system was also modelled numerically using TRNSYS, and the results were in close agreement with experimental result.

They conducted a life-cycle analysis for TEC using grid electricity (baseline prototype) and the one powered by solar PV using ECO Audit CES. The baseline prototype comprised of TE device, heat sinks, fans, and the refrigeration chamber, whereas the solar PV-powered TEC contained a PV panel, charge controller, battery, and DC-DC converter in addition to that of baseline model. The CO2 emissions, total energy, and the cost over life (25 years) were studied. The systems were assumed to be assembled in China and dispatched to Sharjah, UAE, as marine cargo. It was then transferred by truck from Dubai International Airport to Sharjah. Assuming that the system runs for 24 h/day and 365 days/year, it was found that in the case of baseline prototype, the energy usage, CO2 footprint, and cost were the highest during usage, and in the case of solar PV-powered TEC, they were highest for material, and during usage phase, it is nearly zero. This life-cycle study clearly highlights the merits of a solar PV-powered TEC compared to grid-powered cooling solutions.

5.1 Improvement of performance of PV-powered TECs

A quick summary of the studies focussed on performance evaluation of solar photovoltaic-powered TECs is presented in Table 4. It can be noted that the COP of all these systems is on the lower side. Therefore, the COP of TEC must be improved to enable these systems to compete with the conventional refrigeration systems. Some of the studies focussing on improvement of the performance of solar PV-powered TECs are discussed in the following sections.

Table 4 COPs of solar PV-powered TECs

Zhao et al. [63] studied the performance of an innovative radiative sky cooling-assisted thermoelectric cooling (RSC-TEC) system. It was an effort to enhance the COP of a TE cooling system by integrating it with radiative sky cooling. It was demonstrated that the RSC-TEC system with 101 TEMs (Laird ZT8-12), 32 m2 radiative cooling area, and 0.83 m3 cold storage tank could reach an average yearly cooling COP of 1.87. The radiative sky cooling was able to provide 24-h continuous cooling with thin film metamaterial with infrared emissivity greater than 0.93, which reflects 96% of solar irradiance during the daytime. It was also concluded that the contribution of radiative sky cooling was 55% during the daytime and 45% during the nighttime. It was noted that the performance of RSC-TEC systems could be boosted to a great extent with higher ZT values. It was shown that with a ZT value of 2, the yearly COP of the system under study could compete with vapour compression air-conditioning system. The RSC-TEC was cited as a starting point in the large-scale adoption of TE cooling technology. Lv et al.[64]. put forward a novel strategy to enhance the sky radiative cooling using solar photovoltaic thermoelectric cooler. The idea was to incorporate radiative cooling with solar photovoltaic thermoelectric cooler so that PV cells transform a part of solar energy incident to electrical energy, thereby decreasing the solar incidence and heat absorption which contributes to enhancement of diurnal radiative cooling. Simultaneously, PV panels also generate electrical energy, which drives the TEC to further enhance the cooling. The influence of design parameters and meteorological parameters like wind speed, ambient temperature, relative humidity, and sunshine hours on the system performance is also studied in Hefei, China. It was shown that sunshine duration and humidity were having the biggest impact on the functioning of PVRC-TE system. Long sunshine hours and relatively low humidity are the favourable characteristics. Therefore, mid-latitude regions are ideal for implementing this system. It was found that when PV area and radiative cooling area are equal (or the ratio Apv/Arc = 1), highest cooling power of 285.57 MJ/m2 was obtained. The cost of such a system was mainly dependent on the cost of radiative cooling coating and TEMs. With large-scale production, the total system cost was found to reach less than US $1/W.

Atta [65] used closed-loop heat exchangers with macro channels to enhance the COP of a TEAC system powered by solar cells which cools or warms the air flow. When the cooling power of the TEACs is increased by increasing the input current, a larger quantity of heat is generated at the hot side of the TEM. A solution to this problem was to use liquid cooling systems to efficiently dissipate heat to higher surface areas. A closed-loop system circulating the liquid coolant is proposed which operates under modest pressure difference. The coolant used in this study is distilled water with 100–200 ppm sodium chromate as corrosion inhibitor. An antifreeze, ethylene glycol (5% vol) is added to prevent problems associated with freezing of the coolant such as pipe bursts. Channel plates, which are parts of liquid cooling system, directly come in contact with the faces of TEMs and facilitate the smooth transfer of heat between TEMs and coolants. A thermally conductive aluminium substrate with a number of channels cut into it is used in this study. The optimal channel width for minimum thermal resistances was found out to be about 50 mm. The cold and heat produced by the TEMs are conducted into the channel plates and then carried away by the coolant. Figure 16 shows the cross section of the fabricated channel plates. The macro channel plates have high heat transfer coefficient and substantial surface area per unit volume and requires only small cooling fluid reservoir.

Fig. 16
figure 16

Cross section of the designed channel plates [65]

Spherical-bearing DC pumps are used to circulate the coolant. Radiator cooling approach is used in this study. Aluminium tubes are used as radiator tubes due to their excellent chemical agreement with aqueous coolants. A fan blows air through the radiator tubes, cooling the coolants further. Corrugated fluorinated ethylene propylene tube is selected as the interconnect tubing. The coolant flow rate and air flow rate were set to 1 LPM and 10 CFM, respectively, for optimal performance. A schematic illustration of the complete installation is shown in Fig. 17.

Fig. 17
figure 17

System block diagram displaying the setup of the whole system as well as the positioning of thermal sensors (TS) [65]

For the TEAC, 30 symmetrically aligned TE modules (TEC1-12716) were used. The channel plates were designed to transfer heat efficiently from these 30 TEMs. Two channel boxes were built with 6 layers of aluminium having 35 integrated channels on each side. The main function of the channel tubes is to increase the time of coolant flow and the flow surface area. One of the radiators was placed indoor and the other outdoor to act as evaporator or condenser depending on the operating mode (cooling/heating). The DC power required to run the system was derived from thirty polycrystalline solar panels. The study was conducted in Madinah city, Saudi Arabia, and the solar panels were fixed at a tilt angle of 12°. The average daily insolation was observed to be 9000 Wh/m2.

The effect of coolant flow rate on the temperature of both cold and hot plates with a fixed air flow rate of 10 CFM was studied. It was noted that the hot plate temperature dropped by nearly 54 °C, while that of cold plate increased by 24 °C when the flow rate changed from 1 to 9 LPM. The effect of air flow rate when the coolant flow rate was set at 2 LPM was also studied, and the heat transfer was found to increase with larger air flow rates. When the temperature of the radiator becomes close to the ambient temperature, the heat transfer rate decreases. An air flow rate between 50 and 70 CFM is recommended for efficient power consumption and energy balance. The system was also tested with an indoor volume of 30 m3. For efficient power consumption, the water flow rate and air flow rate were set to 6 LPM and 50 CFM, respectively. The ambient temperature varied between 36 and 38 °C throughout the testing period. The temperature of the indoor space dropped from 38 to 31 °C in the first 30 minutes. The temperature dropped to 26 °C in the next 30 minutes and reached 24 °C after the next 30 minutes and remained constant after that. The system’s hot side attained a steady-state temperature of 40 °C in 1 h. The total heat removed from the cold side is estimated to be 3150 W, and the total electrical power consumed is 4032 W with a measured COP of 0.78. The same system was simulated using COMSOL, and the results of simulation were in close agreement with experimental results. Even though the COP obtained was on the lower side, the authors recommend the use of TEACs due to their several environmental advantages. Thus, it was concluded that TEACs with closed-loop heat exchanger was able to cool a room of volume 30 m3 in 90 min with a COP of 0.72. This increased COP compared with the previous studies is attributed to better heat exchange facilitated by the channel plates.

Daghigh and Khaledian [66] conducted a theoretical and experimental performance evaluation of thermoelectric cooling-heating system powered by solar photovoltaic in Sanandaj, Iran. To meet the required cooling load, an auxiliary system comprising of vapour compression system and thermoelectric system was used. For meeting the heat load, heat was directly recovered from the hot side of the TE system and collected in a hot water tank. Based on the purpose and the required temperature, an auxiliary 1000 W electric heater was also provided. The total power required for driving the thermoelectric system was derived from solar PV system. Four TEMs (12706) were used in this device, each becoming functional as and when required depending on the cooling load. The entire system structure is depicted in Fig. 18. It includes PV module, charge controller, battery cooling test chamber, TEMs, VC cooling system, and hot water tank.

Fig. 18
figure 18

Entire structure of the system [66]

Aluminium sinks are linked to the cold side of the TE system with the help of silicon adhesive, which enhances the heat transfer. A 2.4 W electric fan is also provided for better heat distribution. Large heat sinks are attached on the hot sides also, which is immersed in a water chamber to dissipate heat.

First, the thermoelectric system reduces the temperature in the refrigerated space to a certain level after which VC cooling system lowers the temperature to the final required level. The system was mathematically modelled using MATLAB, EES, and CoolPack software. It was noted that the power generated increased with solar insolation, whereas electrical efficiency of the system was found to reduce with rise in solar insolation. This is due to increase in PV panel temperature and is a major issue with PV collectors. Figure 19 shows the changes of evaporator temperature, cooling chamber temperature, COP, and duration of on and off modes according to time in thermoelectric hybrid cooling system.

Fig. 19
figure 19

Variation of evaporator temperature, cooling chamber temperature, COP and duration of on and off modes with time in thermoelectric hybrid cooling system [66]

The controls are set such that when the temperature inside the chamber reaches − 3 °C, the system stops working, and when the temperature increases beyond 5 °C, cooling system turns back on, always keeping the temperature within these ranges. Figure 20 shows thermal load of evaporator and compressor power consumption in hybrid cooling system over different times.

Fig. 20
figure 20

Changes in evaporator thermal load and compressor power consumption with time for the thermoelectric hybrid cooling system [66]

The comparison of operating duration, energy consumption, and COP of compression refrigeration systems and hybrid cooling systems is shown in Fig. 21. It can be seen that the hours of operation and energy consumption of thermoelectric hybrid cooling system is significantly lower than those of the cooling system consisting of VC system only. The thermoelectric system used as auxiliary system reduces the temperature in the refrigerated space from 20 to 12 °C, and then, VC system further reduces the temperature from 12 to − 3 °C. In the compression refrigeration system, this entire temperature reduction has to be done by the VC system alone which consumes more energy and takes more time. The COP of the hybrid thermoelectric system is also found to be better than that of compression refrigeration system.

Fig. 21
figure 21

Comparison of hybrid cooling system and compression cooling system in terms of operation duration, energy usage and COP [66]

With the COP of commercially available thermoelectric systems, the usage of TE coolers as an auxiliary cooling system for a hybrid cooling system is one of the best possible uses of TECs.

6 Performance enhancement of thermoelectric coolers

Dizaji et al. [67] conducted an experimental study on the performance enhancement of thermoelectric air coolers. They introduced the concept of jointed thermoelectric modules, which shows entirely different cooling behaviour from an individual TEM and higher COP without change in input power. Usually, the COP of a TEM is significantly reduced when the input power is increased. This study evaluated the cooling attributes of 2-jointed, 4-jointed, and 6-jointed TEMs. The schematic representation of the rig used for testing is shown in Fig. 22.

Fig. 22
figure 22

A schematic representation of the physical layouts of TEMs [67]

Water flows on the hot surface of the TEM and air to be cooled on the cold surface. Rectangular heat sinks with dimensions 30 cm × 4.5 cm × 4 cm were used on both hot and cold sides. They had five fins of thickness 2 mm. Elastomeric insulation sheet was used to insulate the whole test rig. Keeping the temperature of the air, water, and water flow rate constant, each mode (2 TEM, 4 TEM, and 6 TEM) was tested with varying air flow rate. The applied voltage was also changed to increase the input power. The changes in cooling capacity and COP with power were studied for different air flow rates and are shown in Fig. 23 and Fig. 24, respectively.

Fig. 23
figure 23

Variation of cooling capacity with input power [67]

Fig. 24
figure 24

Variation of COP with input power [67]

It can be seen from Fig. 24 that the COP of 6-jointed TEM is higher than other two modes. For an input power of 30 W, there is a 100% improvement from mode 1 to mode 3. Figure 25 graphically illustrates the reason for this behavior. When the same input power is applied across more than one TEMs, cooling power (qc) increases. As the input power has not changed, COP increases. The quantity that gets affected is the temperature. The temperature of the cold side increases when the same input power is applied across more number of TEMs as shown in Fig. 25.

Fig. 25
figure 25

Reason for increase in COP with jointed TEMs [67]

Figure 26 shows the variation of heat transferred from the air (qair) with input power for modes 1 and 3.

Fig. 26
figure 26

Variation of qair with input power [67]

It can be seen from Fig. 26 that for 2-jointed TEM, increasing input power does not increase qair. This indicates that whole of qc is not being transferred to qair and remains in the heat sink. In a way, it is wasted potential. This is because the area available for heat transfer in mode 1 is less than that of mode 3. There is another advantage of using a greater number of TEMs Use of additional number of TEMs under the same input power improves the available area for heat transfer. The aim of an air cooler is always to increase qair. Therefore, to achieve this, the heat sink on the cold end is modified in two ways — perforating the heat sink and utilizing a spring wire between the fins. These modifications of heat sink are shown in Fig. 27.

Fig. 27
figure 27

a) Perforated heat sink. b) Use of spring wire between the fins in three rows [67]

The modified heat sink designs were tested with 6-jointed TEMs, which has already proven to be the best with respect to COP. It was observed that perforating the heat sink increased the COP of the air cooler, especially for lower values of input power. The perforations escalate the turbulence level of air flow and promotes additional mixing of fluid. Moreover, perforations decrease the material of the heat sink which changes the temperature gradient along the fins and changes its thermal behavior. It could be noted that perforating the heat sink improved the COP of the cooler between 25 and 50% depending on the air flow rate and input power. Using a spring wire as turbulator has even greater effect and improves the COP to around 130%. In fact, using the spring wire turbulator, maximum mixing of the fluid takes place, beyond which further mixing does not improve the COP. In other words, spring wire completely discharges the accumulated cooling capacity in the heat sink. Spring wire creates the required friction between heat sink and air so that complete possible heat transfer takes place between them. The average hydraulic diameter of the air passage is lowered by the spring wire perforator, which in turn increases the Reynold’s number which again contributes to higher heat transfer. Spring wire makes the cold surface warmer, which indicates that maximum heat has been absorbed from the air. The influence of water flow rate, air flow rate, incoming air, and water temperature was also studied.

Shen et al. [68] proposed segmented thermoelectric cooler to improve the cooling performance without incrementing the overall figure of merit. The schematics of traditional and segmented thermoelectric elements are shown in Fig. 28. In the case of segmented TECs, the values of Seebeck coefficient, thermal conductivity, and electrical resistivity are allocated in the form of arithmetic progression as follows:

Fig. 28
figure 28

Illustrations of traditional and segmented thermoelectric elements [68]

$${S}_{p,i}=-{S}_{n,i}=\frac{N-i}{N-1}{S}_{c}+\frac{i-1}{N-1}{S}_{h}$$
(6)
$${\lambda }_{p,i}={\lambda }_{n,i}=\frac{N-i}{N-1}{\lambda }_{c}+\frac{i-1}{N-1}{\lambda }_{h}$$
(7)
$${\rho }_{p,i}={\rho }_{n,i}=\frac{N-i}{N-1}{\rho }_{c}+\frac{i-1}{N-1}{\rho }_{h}$$
(8)

where, N ≥ 2, i = 1, 2, …, N

ρ is the electrical resistivity, and N is the segment number. The subscripts c and h stand for the first segment next to the cold side (i = 1) and the last segment adjacent to the hot side (i = N) of the segmented TE element respectively. Subscripts p and n represent p-type and n-type TE element.

The segment number was found to be particularly sensitive to the thermal conductivity of thermoelectric material. The cooling performance of segmented thermoelectric element was found to be superior than that of traditional TECs when Sh/Sc > 1, λh/λc < 1 and ρh/ρc < 1. The optimum number of segments was calculated to be 2. It was also found that the maximum cooling capacity, temperature difference, and COP of two segmented TE elements could be enhanced by 118.1%, 118.1%, and 2.1%, respectively, which is a remarkable achievement. It was proposed that for a two segmented TEM, 0.35 of Joule heating returns to the cold side instead of 50% which is the main reason for the performance enhancement.

Fabián-Mijangos et al. [69] studied the enhanced performance of a thermoelectric module with asymmetrical legs having truncated square pyramid shape. The geometry and configuration of thermoelectric legs greatly influence the performance of TEMs. Asymmetrical legs help lower the overall thermal conductance of the device which increases the temperature gradient in the legs. It also helps in making use of the Thomson effect, which relies on the temperature gradient in the legs and the variation of Seebeck coefficient with temperature. The Thomson effect is usually disregarded in conventional rectangular TEM legs.

In this study, TEMs with both rectangular and asymmetrical legs were manufactured from p-type and n-type Bi2Te3. The complete manufacturing process is described in detail. The dimensions of TE legs with regular geometry were 1.7 mm × 1.7 mm × 2.1 mm. A numerical simulation in COMSOL showed that for a leg length of 2.1 mm, a slant angle θ = 22° (critical angle) gave the highest temperature difference. Practically, manufacturing TE legs with θ > 10° makes the cross-section at the thin end very small. Such legs are prone to breaking due to the poor mechanical properties of Bi2Te3. Figure 29 shows photos of fabricated modules with nine pairs of legs.

Fig. 29
figure 29

Photographs of fabricated modules [69]

An asymmetrical module should be operated such that hot side comes on the side with smaller leg area as it has smallest thermal conductance. Impedance spectroscopy technique was used to ascertain the figure of merit of the two modules, and it was found that ZT increased from 0.79 for TEM with symmetrical legs to 1.02 for asymmetrical TEM. This thermoelectric improvement is due to harnessing of Thomson effect.

6.1 Transient cooling

Yang et al. [70] studied the transient cooling of thermoelectric coolers. When a thermoelectric cooler is subjected to a current pulse with amplitude many times that of steady-state optimal current, an instantaneously lower temperature than that attainable during steady-state operation can be reached. This occurs due to the delay of thermal diffusion of Joule heat. On one hand, Peltier effect occurs locally and is restricted to the junctions of TE elements, and on the other hand, Joule heat occurs volumetrically. Peltier effect helps to develop the instantaneous temperature reduction.

There are many time constants like the time to reach minimum temperature (TRM), the time that TEM remains at the minimum temperature (holding time), and the time needed for the TEM to arrive at the steady-state temperature after the transient current applied is removed. Along with these time constants, the minimum temperature achievable, maximum temperature overshoot, current pulse shape, and time between pulses are also important parameters that influence transient performance of a TEM. Geometry of the TE devices like leg length, cross-sectional area, also affects transient performance. The change in cold side temperature with an applied pulse current is shown in Fig. 30.

Fig. 30
figure 30

Time constants and changes in cold junction temperature for a typical transient cycle [70]

The leg length of the TE elements is found to influence the thermal inertia of the TEM. It also affects TRM, holding time, and recovery period. Holding time increases with leg length. Recovery period also tends to be longer with increased leg length. The effect of pulse shape on the cold junction temperature response is shown in Fig. 31.

Fig. 31
figure 31

Response of cold junction temperature to three distinct pulse forms [70]

The lowest temperature achievable is almost the same for all pulse shapes; however, the holding time varies significantly. Similarly, the shape of the TE device does not influence the lowest temperature achievable during steady state. TE elements having variable cross-sectional areas can reach lower temperatures with short recovery times compared to TE elements with equal cross-sectional areas. Tapered legs make the thermal resistance asymmetric due to which Joule heat will be generated at the end with lower cross-sectional area and conducted towards larger cross-sectional area. Transient cooling properties of TECs are utilized in cooling the semiconductor laser’s active area and in other electronic devices to eliminate hot spots.

Manikandan and Kaushik [71, 72] introduced the concept of annular thermoelectric coolers (ATECs) and conducted energy and exergy analysis by introducing some nondimensional parameters. They also studied the transient thermal behaviour of ATEC. Because of their larger contact areas, ATECs are sometimes a preferable choice in heat pipe applications or when the heat source is cylindrical in shape. They improve the cooling power and energy efficiency. ATECs could maintain the cold side temperature 2.3 K lower than a flat TEC [72]. Two models, one in which the cold end area is less than the hot end area (model 1) and another model with a greater cold end area than the hot end (model 2), were considered for the study. The cold end temperature with current for model 1 and model 2 is as shown in Fig. 32.

Fig. 32
figure 32

Cold-end temperature of ATEC for model 1 and model 2 at different currents [73]

The effects of current amplitude and width, annular thermoelectric angle, leg length, heat transfer coefficient, and pulse shape on the transient performance of ATEC were investigated. It was found that the amplitude and pulse width of the pulse current are important parameters that affect the transient performance of ATEC. There exists an ideal pulse width at which the lowest cold side temperature can be attained. It is not suggested to utilize a design with a hot end size smaller than the cold end area. When the leg length increases, thermal resistance is also increased, which dampens the heat conduction. It also increases Joule heating due to increased resistance. A higher value of heat transfer coefficient enhances the transient performance of ATEC but up to a certain higher limit after which it does not have considerable effect. Square pulse was found to give the lowest temperature. In comparison, the increasing pulse current and triangular pulse current has longer holding times and shorter recovery time, even though the lowest temperature attainable is higher than that of square pulse.

6.2 Other studies

Gong et al. [35] proposed an optimized design of TECs based on numerical simulation. They found that lower leg length [74] and larger cross-sectional area were beneficial for improving the cooling capacity and COP. It was also noted that the most common failure mechanism of TECs is failure due to temperature gradient-induced reliability issues. It was also noted that reducing the leg length and increasing the leg cross-sectional area helped to reduce temperature gradient in the legs, improving the reliability of TECs. Gong et al. [75] conducted a thermo-mechanical investigation on a compact TEC. A heat-generating chip was attached on the cold side of the TEC as a thermal load, and a three-dimensional numerical model considering temperature-dependent material properties was used for the study. The thermal stress that occurs in a TEC, while in operation, is also considered. A higher value of input current-induced higher values of thermal stress and a shorter leg length helped in reducing the thermal stress. It was also reported the use of thicker ceramic plates helped in achieving better distribution of thermal stress and superior dependability.

Sarkar and Mahapatra [76] investigated the role of surface radiation, which is an addition to the usually considered natural convection as the modes of heat transfer from the heat sink of a TEC while modelling it. It was found that inclusion of surface radiation improves the COP of the TEC significantly with an increase in emissivity. Riffat and Ma [13] reported that heat from the hot side of the TEC should be dissipated effectively for enhancing the cooling on the cold side. Some studies focused on this idea for performance enhancement. Gökçek and Şahin [77] experimentally studied the performance of a minichannel water-cooler TEC. The TEMs were unified with minichannel heat sinks on the hot side and heat dissipators on the cold side. When the water flow rate through the minichannel was increased from 0.8 to 1.5 L/min, the temperature inside the refrigerator was found to decrease from 2 to − 0.1 °C, and the COP increased from 0.19 to 0.23. Cuce et al. [78] conducted an experimental study on TEC by using a water-cooled block on the hot side of the TEC for heat dissipation. Three distinct nanoparticles (TiO2, Al2O3, and SiO2) were mixed with the cooling water to improve its thermal performance. Al2O3-water nanofluid showed better performance because of its better thermal conductivity compared to that of TiO2 and SiO2. With 1% mass fraction, a 42.2% improvement in ΔT than that in the case of pure water was achieved. Bhuiya et al. [79] conducted computational studies to improve the performance of TECs by introducing a phase change material (PCM) in the heat sink on the hot side of the TEC to enhance the heat dissipation. It was found that the temperature on the hot side of the TEC reduced considerably with the introduction of PCM. It has been shown to be useful in improving TEC performance during the phase change of the PCM from solid to liquid. The volume of the PCM used and the heat sink geometry was shown to directly affect the performance with the heat sink in the shape of a square enclosure and larger volume of PCM giving better results. Some innovative applications like solar-driven cool pavement [80] where solar PVs are used to drive TECs that cool pavements in urban areas are also being studied. Usually in the urban areas, a microclimate different from nearby areas is developed which increases the overall temperature, forming heat islands which in turn increases the use of air-conditioning in urban buildings. Cooling of pavements using TECs has been found effective in mitigating the heat islands formed.

7 Conclusions

The conventional refrigeration systems like vapour compression and vapour absorption systems depends on fossil fuels for their operation and use gases with high global warming potential (GWP) as refrigerants. Therefore, scientists across the world are trying to develop alternative cooling solutions including the thermoelectric refrigerators. The combustion of fossil fuels liberates greenhouse gases and contributes to global warming. Moreover, the vapour compression system cannot be operated at a place where electricity is not available. Since solar energy is freely available in sufficient quantity, a solar-powered thermoelectric cooler working on Peltier effect is a better alternative for the conventional system. Thermoelectric cooler is a noise-free and vibration less system because of the absence of moving parts. They do not use a refrigerant, and electrons act as heat carriers. A better temperature control is possible in a thermoelectric cooler. Their unique advantages like refrigerant-free operation and ability to operate in both cooling and heating mode combined with their portability, noise-free operation, and reliability make TECs an attractive choice. They are finding increasing applications in portable refrigerators, air-conditioners in zero energy buildings, automobile industry, etc. Solar-powered thermoelectric refrigerator can be operated as standalone portable reliable refrigerator for the transport and storage of vaccine and medicine and for the storage of perishables.

Solar PV-powered TEC is the best option for niche cooling applications like storage and transport of vaccines, medicines, and other perishables in remote and rural areas where grid connectivity is not available. Some innovative applications like solar-driven cool pavement where solar PVs are used to drive TECs that cool pavements in urban areas are also being reported.

The reported energy efficiency of the solar-powered thermoelectric refrigerators is lower than its compressor counterparts. There are two main approaches to achieve performance enhancement of TECs: improving the intrinsic efficiency of the materials used as thermoelectric material and improving the thermal design of TECs. The performance improvement can be achieved through improving module contact resistance, thermal interfaces, and heat sinks. At present, the figure of merit of thermoelectric material ZT is about 0.7. It should be further increased by the selection of better thermoelectric material.

In this paper, a comprehensive review of studies aiming at performance improvement of solar-powered thermoelectric cooler is presented. The performance of the solar-powered TEC depends strongly on the intensity of solar insolation and the temperature difference between the hot and cold sides of the thermoelectric module. Better COP was reported for a TEC by providing minichannel water-cooled heat sinks on the hot side and heat dissipators on the cold side. The cooling performance of a segmented thermoelectric element was found to be superior than that of traditional TECs.

The performance of the solar-powered thermoelectric refrigerator system is highly dependent on the thermoelectric material used in it. A good thermoelectric material should possess high Seebeck coefficient, low-thermal conductivity, and high electrical conductivity. The optimization of the thermoelectric material performance is achieved by reducing the material thermal conductivity especially the lattice thermal conductivity. Even though promising results are reported in the literature, further improvement of COP should be achieved by the development and selection of better thermoelectric materials and a superior thermal design of TECs.