This application claims priority GB1905395.8, filed 16 April 2019, the contents and elements of which are herein incorporated by reference for all purposes.

Field of the Invention

The present invention relates to thermoelectric materials, devices and methods of production thereof. Particularly, although not exclusively, it relates to compositionally graded thermoelectric materials, devices and method of production thereof.

Background

Harvesting energy from ambient sources in our environment such as light, thermal and vibrational sources has attracted increasing interest as possible alternative energy for applications such as self- powered, embedded, implantable, portable, wireless and/or wearable electronic devices. Typically, such devices are powered by traditional batteries that need frequent recharging and/or recharging [1-3] Harvesting thermal energy may be particularly attractive, because there are abundant environmental heat sources and most of them, such as the heat loss released by exhausts, industrial processes and radiators, or even that arising from the human body heat, are unexploited [4] Thermoelectric materials (TEMs) and devices comprising such materials such as thermoelectric generators (TEGs) are prime candidates for thermal energy harvesting due to their ability to convert ambient and ever-present waste heat into electrical energy by utilising the Seebeck effect [5] A thermoelectric material or device creates voltage when there is a different temperature on each side. Conversely, when a voltage is applied to it, heat is transferred from one side to the other, creating a temperature difference. It is possible to scale down TEGs to sizes such that they can be integrated into self-powered electronic devices with minimal maintenance [6, 7]

In order to evaluate the performance of thermoelectric materials, the dimensionless figure of merit (ZT) is usually used to describe their potential energy-conversion performance. For a given operating temperature, ZT is defined as ZT = S ² aT / K, where k is the thermal conductivity [W/m.K], s is the electrical conductivity [S/m], and S is the Seebeck coefficient [V/K] [8, 9]

A good thermoelectric material should preferably possess a high S and ato ensure a high-voltage output at a given temperature difference. The power factor (PF= S ² o) can also be used as an alternative way to evaluate the thermoelectric material. Increasing the PF of materials has been recognized as one strategy for optimizing ZT [9] TE materials with a higher PF value can convert more heat energy into electricity.

However, a problem faced in this field is that many existing inorganic thermoelectric materials suffer from problem of scarcity and toxicity. Furthermore, it is difficult to scale up production of many such materials for large scale manufacturing of these materials, and of devices comprising these materials. Accordingly, work has been done on developing alternative thermoelectric materials, and methods for producing these. For example, previous work by the present inventors includes production of flexible aerosol-jet printed organic-inorganic nanocomposite materials for thermoelectric applications. See, for example:

Ou, C. et. Al.:“Fully Printed Organic-Inorganic Nanocomposites for Flexible Thermoelectric Applications.” ACS Appl. Mater. Interfaces 2018, 10 (23), 19580-19587. [2]

And

Ou, C.; et. Al:“Enhanced Thermoelectric Properties of Flexible Aerosol-Jet Printed Carbon Nanotube- Based Nanocomposites.” APL Mater 2018, 6 (9), 096101. [14]

The materials disclosed in these publications overcome some of the above listed problems with known thermoelectric materials. However, thermoelectric materials also suffer from a problem that figure of merit ZT or power factor PT often have a strong dependence on temperature. Accordingly, even where it is possible to provide alternative thermoelectric materials which overcome some known problems of traditional thermoelectric materials, there is still the problem that thermoelectric materials will only be efficient within a small temperature range due to the variance in power factor with temperature.

This problem was first considered in the context of inorganic thermoelectric materials in work by Ioffe et al. [24], and has subsequently been considered in a number of other publications [15, 25-27] These publications disclose functionally graded thermoelectric materials (FG-TEM) of two different types:

1) Continuous FG-TEM

2) Segmented FG-TEM.

For‘continuous FG-TEM’, as they are generally referred to in the literature, a single-phase inorganic thermoelectric material is modified to provide different carrier concentrations in different regions of the material. For example, in work by Cramer et al. (2017) [41], a ZnO material having a graduation in grain size from 180 nm grains to 1.2 pm grains across the length of the material was produced by spark plasma sintering. In work by Hedegaard et al (2014) [33], a solid solution Gei- _x Si _x sample was produced using the Czochralski method, where along the length of the Gei- _x Si _x sample, x changes to thereby change the band gap of the material.

Summary of the Invention

The present inventors have realised, however, that it is difficult to fully optimise the ZT values within the whole temperature range of use of such materials as identified above, in particular because it is difficult to control diffusion of carriers through the material, in particular when used at elevated temperatures (e.g. above room temperature), which can cause deterioration of device performance and lifetime with use. Furthermore, the current research on continuous FG-TEMs is still at the proof-of-concept stage, and it is not clear that these materials could be well controlled during the manufacturing process, and in particular, during mass-manufacturing for use in industrial applications.

Segmented FG-TEMs comprise multiple different segments having different thermoelectric properties into a single thermoelectric element [1 1 , 36, 42-51] Such materials can be fabricated with suitable ZT values at different applied temperatures. However, one significant drawback of such an approach is the difficulty of connecting multiple dissimilar materials without the introduction of interfaces that are prone to failure, in particular due to thermo-mechanical stress which can accumulate at such interfaces in the presence of different thermal expansion coefficient during practical use.

Another alternative approach to dealing with the problem of inefficiency of conventional TEMs over a temperature range is the development of multi-component TE devices, in which a plurality of TE components having different thermoelectric properties are used in a single device. For instance, a high- performance multi-segmented thermoelectric generator has been designed with the use of different inorganic thermoelectric materials under different applied temperature ranges, where SiGe was used for high temperature range, PbTe for medium temperature range, and BhTe3 for low temperature range [25- 27]

However, none of these approaches to improving efficiency are applicable to the alternative

thermoelectric materials as developed by the present inventors, because they relate to single-phase inorganic thermoelectric materials, rather than to composite thermoelectric materials.

Accordingly, it would be desirable to develop efficient, scalable, inexpensive, and flexible thermoelectric materials and devices.

The present invention has been devised in light of the above considerations.

Accordingly, in a first aspect, the present invention provides a thermoelectric composite material comprising a conductive polymer matrix material and at least a first thermoelectric filler material, wherein: the loading (wt%) of the first filler material in the matrix material varies along at least one direction of the composite material, so that a power factor response with respect to temperature of a first portion of the composite material is different to that of a second portion of the composite material.

The at least one direction may be the length, the width or height of the material. The loading (wt%) of the first filler material in the matrix material varies along at least one of these directions, but in some cases along two, or three of these directions. In some configurations, the material may be configured as an elongate member, and the loading (wt%) of the first filler material in the matrix material may vary along the length of the elongate member.

The power factor response with respect to temperature can be assessed by e.g. plotting the power factor PF= S ² a of a material at a series of discrete temperatures. Where this plot is different for two materials, the power factor response with respect to temperature is different: for example, the power factor at a first temperature for the first portion may be different to the power factor at the same temperature for the second portion. By providing a thermoelectric composite material which has two portions having different power factor response with respect to temperature, it is possible to provide a thermoelectric composite material which has improved efficiency over a given temperature range in comparison to known thermoelectric composite materials, by proving a high power output.

Preferably, the power factor response with temperature of the first portion and the second portion is selected so that in use, the power factor response of the first portion within a first temperature range is better than the power factor response of the second portion within that first temperature range, i.e. the first portion has higher power factor at each discrete temperature within the first temperature range than the second portion. Furthermore, preferably the power factor response of the second portion within a second temperature range is better than the power factor response of the first portion within that second temperature range i.e. the second portion has higher power factor at each discrete temperature within the second temperature range than the second portion. In this way, it is possible to further improve the efficiency of the material in use, when a temperature differential is applied across the composite thermoelectric material, because the first portion of the material will provide improved efficiency when used within the first temperature range, and the second portion of the material will provide improved efficiency when used within the second temperature range. The first temperature range may be lower than the second temperature range. For example, the first temperature range may be from Tc to Tx, and the second temperature range may be from Tx to TH, where Tc is the cold end of an applied temperature gradient, TH is the hot end of the applied temperature gradient, and Tx is an intersection temperature of the power factor response profiles for the first portion of the thermoelectric material and the second portion of the thermoelectric material respectively. Tx may vary depending on the relative composition of the first portion of the composite thermoelectric material and the second portion of the composite thermoelectric material.

The boundary between the first and second portions of the thermoelectric composite material may be selected to be at or near the location of the intersection temperature Tx during use, as calculated by modelling the theoretical temperature distribution along the material under an applied temperature gradient.

The specific loading (wt%) of the first filler material in each portion of the material is not particularly limited, and may be between 0% (i.e. pure matrix material) up to and including 99%. Preferably the loading (wt%) is 95% or less, or 90% or less. Providing a very high loading (wt%) may reduce the conductivity of the material. Preferred loading amounts may be selected based on the nature of the filler material, and the intended application of the thermoelectric composite material.

The conductive polymer matrix may be a semiconductive polymer. Preferably, the conductive polymer matrix comprises an organic conductive or semiconductive polymer. Preferably, the conductive polymer matrix is a thermoelectric material. In other words, preferably both the conductive polymer matrix material and the at least first filler material are thermoelectric materials. However, in other arrangements, the polymer matrix may not be a thermoelectric material, and the thermoelectric effect may be provided only by the thermoelectric filler material contained within the conductive polymer matrix. Use of an organic polymer matrix may allow for production of inexpensive, scalable, and mechanically flexible thermoelectric materials. These may be particularly suitable for use in applications such as those aimed at human body integration, where it is generally desirable for such materials to be lightweight and conformal.

The conductive polymer matrix material may comprise one or more of a p- type conducting thermoelectric polymer, an n-type conducting thermoelectric polymer, or another conductive polymer. For example, the conductive polymer matrix may comprise a p- type conducting thermoelectric polymer selected from one or more of PBTTT = poly(2,5-bis(3-alkylthiophen-2-yl)thieno[3,2-b]thiophene); PANI = polyaniline; P3HT = poly(3-hexylthiophene); PPy = polypyrrole; and PEDOT:PSS = poly(3,4-ethylenedioxythiophene) polystyrene sulfonate.

The conductive polymer matrix may comprise an n-type conducting thermoelectric polymer selected from one or more of PPV = poly(p-phenylene vinylene); PDI = perylene diimide; P(NDIOD-T2) = poly{N,N'- bis(2-octyl-dodecyl)-1 ,4,5,8-napthalene dicarboximide-2,6-diyl]-alt- 5,5'-(2,2'-bithiophene)}; PA = polyacetylene; and Poly(Ni-ett) = poly(nickel-ethylene tetrathiolate).

The conductive polymer matrix may comprise a conducting thermoelectric polymer selected from one or more PTh = Polythiophene; PAT = Poly(3-alkylthiophene); PTV = Polythienylenevinylene; and PPP = Polyphenylene.

Preferably, the conductive polymer matrix material comprises PEDOT:PSS. PEDOT:PSS has been shown to possess one of the highest TE performance capabilities at low cost, ease of large-volume and/or large-area printed organic electronic manufacturing. Furthermore it has excellent environmental stability [2], [13], [14] It can be readily processed with very low intrinsic k [15]

The thermoelectric filler material may comprise a plurality of particles, fibres, wires, tubes, flakes, sheets, or platelets. The particles, fibres, wires, tubes, flakes, sheets, or platelets may have a largest dimension of less than or equal to about 300 pm, less than or equal to about 100 pm, less than or equal to about 1 pm, less than or equal to about 500 nm, or less than or equal to about 100 nm. The maximum size of the thermoelectric filler material may be limited by the method of deposition. The particles, fibres, wires, tubes, flakes, sheets, or platelets may have a smallest dimension of greater than or equal to about 10 nm, greater than or equal to about 50 nm, or greater than or equal to about 100 nm. Preferably, the particles, fibres, wires, tubes, flakes, sheets, or platelets may have an average largest dimension in the range of 10 nm to 300 pm, 10 nm to 100 pm, 10 nm to 10 pm, preferably in the range of 10 nm to 1 pm, more preferably in the range of 10 nm to 500 nm. When the average largest dimension of the fibres, wires, tubes, flakes, sheets, or platelets is less than 500 nm, they may be referred to as nanomaterials. When the thermoelectric filler material is provided with a size in this range, thermoelectric property

enhancement may be achieved due to nanostructuring effects.

The average largest dimension of the particles, fibres, wires, tubes, flakes, sheets, or platelets may be measured by any suitable conventional method, for example by particle size analysis using SEM images.

The at least first filler material may comprise one or more of: a binary bulk chalcogenide such as bismuth telluride (B Tes), and/or antimony telluride (Sb ₂ Te3); layered atomic structure materials such as SnSe, PbTe, SiGe, ZruSb3, CoSb3, TaS2, M0S2, T1S2, and doped alloys thereof (doped with any suitable elements, preferably one or more metallic or metalloid elements); skutterudites including e.g. CeFe ₄ Sbi ₂ ; half-Heusler alloys including e.g. MgAgSb; Oxides including e.g. Na _x Co02, SrTi03, CaMn03, doped ZnO, and doped Ih ₂ q3 bixbyite (doped with any suitable elements, preferably one or more metallic or metalloid elements); and a carbon-based material e.g. single walled carbon nanotubes (SWCNTs), multi-walled carbon nanotubes (MWCNTs), and/or graphene platelets.

Binary bulk chalcogenides such as bismuth telluride (B Tes) and antimony telluride (Sb ₂ Te3) may be particularly preferable, because they are well-known to exhibit a maximum ZT value of more than 1.4 at room temperature, and are thus well-suited for near-room-temperature applications, such as refrigeration and waste heat recovery up to 200 °C [10], [11], [12]

The thermoelectric composite material may comprise at least the first thermoelectric filler material and a second thermoelectric filler material different to the first filler material. In some arrangements, the thermoelectric composite material comprises three or more different thermoelectric filler materials, for example, 4, 5 or 6 different thermoelectric filler materials.

Where the thermoelectric composite material comprises at least the first thermoelectric filler material and a second thermoelectric filler material different to the first filler material, the loading (wt%) of the first filler material may be higher than the loading (wt%) of the second filler material in the first portion of the composite material, and the loading (wt%) of the second filler material may be higher than the loading (wt%) of the first filler material in the second portion of the composite material.

The thermoelectric composite material may comprise three or more portions, each having a different power factor response from the other portions. For example, the loading (wt%) of the first filler material in the matrix material may vary along at least one direction of the composite material, to thereby define first, second and third portions each having a different power factor response with respect to temperature. Accordingly, it may be possible to define multiple intersection temperatures, e.g. Txi being an intersection temperature of the power factor response profiles for the first portion of the thermoelectric material and the second portion of the thermoelectric material respectively (a‘first intersection temperature’), and Tx2 being an intersection temperature of the power factor response profiles for the second portion of the thermoelectric material and the third portion of the thermoelectric material respectively (a‘second intersection temperature’).

Additionally or alternatively, where the thermoelectric composite material comprises multiple

thermoelectric filler materials, the relative loading proportions of the different thermoelectric filler materials may be varied along at least one direction of the composite material, to thereby define first, second and third portions each having a different power factor response with respect to temperature. In one arrangement, the thermoelectric composite material may comprise three or more portions, wherein the major thermoelectric filler material is different in each portion. Providing an increased number of different thermoelectric filler materials and/or portions having different power factor response with respect to temperature from the other portions may allow for more specific tailoring of the thermoelectric properties of the composite thermoelectric material for a given application.

The thermoelectric composite material may comprise a transition region between each portion of the composition material having a different power factor response, wherein the transition region is continuously compositionally graded. In other words, there may not be a step-change in composition between first, second and/or subsequent portions of the thermoelectric composite material. Such a structure may provide for improved device performance of a device incorporating said material, due to reduction or lack of electrical interfaces that are prone to failure, in particular due to thermo-mechanical stress which can accumulate at such interfaces in the presence of different thermal expansion coefficient during practical use.

The thermoelectric composite material may be made using an aerosol jet printing technique, for example as disclosed in:

Ou, C. et. Al.:“Fully Printed Organic-Inorganic Nanocomposites for Flexible Thermoelectric

Applications.” ACS Appl. Mater. Interfaces 2018, 10 ( 23), 19580-19587. [2]

And

Ou, C.; et. Al:“Enhanced Thermoelectric Properties of Flexible Aerosol-Jet Printed Carbon Nanotube- Based Nanocomposites.” APL Mater 2018, 6 (9), 096101 . [14]

Accordingly, disclosed herein is a method of making a thermoelectric composite material according to the first aspect using aerosol jet printing (AJP). AJP atomises functional inks into an aerosol droplet form.

The aerosol droplets are subsequently streamed through a deposition head and focused by a sheath flow (typically nitrogen gas flow) [2], [14], [54] before deposition onto a substrate to form the thermoelectric composite material. The substrate may be a polymeric substrate. The substrate may be flexible and/or stretchable. For example the substrate may comprise a polyimide sheet (e.g. Kapton), an acetate sheet, an AI/AI2O3 foil, and/or a paper sheet. The substrate may comprise polydimethylsiloxane (PDMS) and/or polyurethane (PU). Alternatively the substrate may be rigid. For example the substrate may comprise silicon or glass.

After deposition of the ink(s) onto the substrate, the printed samples may be cured. This can assist in removing undesirable solvents.

The aerosol jet printing technique may use a plurality of separate ink sources for example two or more ink sources, or three or more ink sources. In preferred embodiments, the thermoelectric composite material is fabricated from two separate ink sources including an organic ink source and an inorganic ink source. The plurality of ink sources may be mixed in-situ to thereby dynamically tune the composition of the printed composite material to realise thermoelectric composite material with an optimised composition variation to match the temperature gradient across which the printed thermoelectric composite material operates during use. Dynamic tuning of composition may be achieved by adjusting the aerosol flow rates from the different atomisers in fluid connection with respective different ink sources.

In a second aspect, the present invention provides a thermoelectric device comprising the thermoelectric composite material according to the first aspect. The device may be e.g. a thermoelectric generator, for example for use in a wearable device, or for use in thermal energy harvesting (e.g. from hot water pipes). Such thermoelectric generators may be referred to as compositionally graded thermoelectric generators, or‘CG-TEGs’. Alternatively the device may be e.g. a Peltier cooler, a thin-film heater, or a temperature sensor.

In a third aspect, the present invention provides use of a thermoelectric composite material according to the first aspect, or a thermoelectric device according to the second aspect for energy harvesting application. Preferably, the use of the thermoelectric composite material is for waste heat recovery up to e.g. 200 °C, for example up to 100 °C.

The invention includes the combination of the aspects and preferred features described except where such a combination is clearly impermissible or expressly avoided.

Summary of the Figures

Embodiments and experiments illustrating the principles of the invention will now be discussed with reference to the accompanying figures in which:

Figure 1 shows (a) a schematic of an in situ mixing method for nanocomposite printing by AJP; (b) SEM images of five-layer printed PEDOT:PSS-based nanocomposites loaded with 50 wt % (nominal) Bi ₂ Te3 nanoparticles; and (c) SEM image of five-layer printed PEDOT:PSS-based nanocomposites loaded with 85 wt % (nominal) Sb ₂ Te3 nanoflakes,

Figure 2 shows various optical and SEM images of a printed 15 wt.% Bi2Te3— PEDOT:PSS

nanocomposite + PEDOT:PSS compositionally graded thermoelectric generator according to the invention.

Figure 3 shows various SEM images of printed MWCNTs-PEDOT:PSS nanocomposites (not compositionally graded), including MWCNTs-PEDOT:PSS nanocomposites with different loading ratios of MWCNTs from (c) 15 wt.%, (d) 50 wt.%, (e) 85 wt.%, up to (f) 100 wt.% (no matrix material, and not according to the present invention).

Figure 4 shows a schematic drawing of the arrangement used for measurement of the temperature- dependent Seebeck coefficient and electrical conductivity of samples.

Figure 5 shows temperature-dependent measurements of (a) Seebeck coefficient, (b) electrical conductivity and (c) power factor of the printed pristine PEDOT:PSS film (not compositionally graded). Figure 6 shows temperature-dependent measurements of S, a, and PF of printed Bi ₂ Te3-PEDOT:PSS nanocomposites loaded with (a) 15 wt.%, (b) 35 wt.%, (c) 50 wt.%, (d) 65 wt.%, (e) 85 wt.%, and (f) 90 wt.% B12T e3 nanoparticles, respectively (not compositionally graded).

Figure 7 shows temperature-dependent measurements of S, a, and PF of printed MWCNTs-PEDOT:PSS nanocomposites loaded with (a) 15 wt.%, (b) 35 wt.%, (c) 50 wt.%, (d) 65 wt.%, (e) 85 wt.%, and (f) 90 wt.% MWCNTs, respectively (not compositionally graded).

Figure 8 shows temperature-dependent measurements of S, a, and PF of printed Sb ₂ Te3-PEDOT:PSS nanocomposites loaded with (a) 15 wt.%, (b) 35 wt.%, (c) 50 wt.%, (d) 65 wt.%, (e) 85 wt.%, and (f) 90 wt.% Sb ₂ Te3 nanoflakes, respectively (not compositionally graded).

Figure 9 shows the power factor response with temperature of printed PEDOT:PSS-based

nanocomposites loaded with different wt.% of various nanomaterials (not compositionally graded).

Figure 10 shows temperature-dependent measurements of PF for the material match of (a) 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite + PEDOT:PSS; (b) 50 wt.% Sb ₂ Te3-PEDOT:PSS nanocomposite + PEDOTPSS; (c) 85 wt.% MWCNT-PEDOT:PSS nanocomposite + PEDOTPSS; and (d) 15 wt.% Bi ₂ Te ₃ - PEDOT:PSS nanocomposite + 50 wt.% Sb ₂ Te3-PEDOT:PSS nanocomposite.

Figure 11 shows a COMSOL simulation result showing the temperature distribution profile along the sample of (a) 15 wt.% Bi ₂ Te ₃ -PEDOT:PSS + PEDOTPSS based CG-TEC and (b) 50 wt.% Sb ₂ Te ₃ - PEDOT:PSS + PEDOT:PSS based CG-TEC, respectively, above diagrams of the compositionally graded structure of the AutoCAD-designed pattern and the experimentally printed sample of (c) 15 wt.% Bί ₂ Ϊb3- PEDOT:PSS + PEDOT:PSS based CG-TEC and (d) 50 wt.% Sb ₂ Te ₃ -PEDOT:PSS + PEDOT:PSS based CG-TEC, respectively.

Figure 12 shows a schematic diagram of the operation circuit for the measurement of voltage output and power output of functionally graded TEGs.

Figure 13 shows the output voltage and output power against external load resistance of (a) pristine PEDOT:PSS film, (b) 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite, (c) 50 wt.% Sb ₂ Te3-PEDOT:PSS nanocomposite, (d) 15 wt.% Bi ₂ Te3-PEDOT:PSS + PEDOT:PSS CG-TEG, (e) 50 wt.% Sb ₂ Te3- PEDOT:PSS + PEDOT:PSS CG-TEG, and (f) 15 wt.% Bi ₂ Te ₃ -PEDOT:PSS + 50 wt.% Sb ₂ Te ₃ - PEDOT:PSS CG-TEG under a temperature difference of 70 K.

Figure 14 shows output voltage and power measurements of 15 wt.% Bi2Te3— PEDOT:PSS

nanocomposite + PEDOT:PSS CG-TEG in the (a) forward direction and (b) reverse direction. Output voltage and power measurements of 50 wt.% Sb2Te3— PEDOT:PSS nanocomposite + PEDOT:PSS CG- TEG in the (c) forward direction and (d) reverse direction.

Figure 15 shows a comparison of output voltage and output power of 5 different samples under the same applied temperature gradient when loaded with an external resistance of 50 W under a temperature difference of 70 K. Figure 16 shows a prediction of the effective thermal conductivity of different polymer-based TE nanocomposites as a function of filler volume fraction.

Detailed Description of the Invention

Aspects and embodiments of the present invention will now be discussed with reference to the accompanying figures. Further aspects and embodiments will be apparent to those skilled in the art. All documents mentioned in this text are incorporated herein by reference.

Aerosol jet printing (AJP)

Before the printing process, a printing pattern was designed and drawn using AutoCAD software. Then, the aerosol-jet printer (Optomec Aerosol Jet 200 Printer) equipped with an ultrasonic atomiser (UA) and a pneumatic atomiser (PA) was used to print Bi ₂ Te3/Sb ₂ Te3 nanoparticles and MWCNTs with combination of PEDOT:PSS as matrix onto a flexible polyamide substrate (PI, 75 pm thick, Goodfellow ^® ). A schematic diagram of this printing method is shown in Fig. 1 (a), and described in greater detail in Ou, C. et al.“Fully Printed Organic-Inorganic Nanocomposites for Flexible Thermoelectric Applications” [2] However, the method was varied from the method as disclosed in this paper, as the aerosol flow rates from the different atomisers were adjusted during the printing process to thereby dynamically tune the composition of the material being printed, during the printing process, to thereby vary the loading (wt%) of the Bi ₂ Te3/Sb ₂ Te3 nanoparticles and MWCNTs in the PEDOT:PSS matrix material along the length of the printed composite material.

The ink formulations used poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS, 1 wt.%, Heraeus Clevious™ PH1000) purchased from Ossila. The Bi ₂ Te3 nanoparticles and Sb ₂ Te3 nanoflakes were synthesised via a scalable solvothermal synthesis process [2], whilst the multi-walled carbon nanotubes (MWCNTs, >90%, 5-9 pm in length, 1 10-170 nm in diameter) were purchased from Sigma Aldrich [14] The conducting behaviour for both Bi ₂ Te3 nanoparticles and Sb ₂ Te3 nanoflakes used in this work is p- type.

Fig. 1 (b) and (c) show SEM images of five-layer printed PEDOT:PSS-based nanocomposites loaded with (b) 50 wt % (nominal) Bi ₂ Te3 nanoparticles and (c) 85 wt % (nominal) Sb ₂ Te3 nanoflakes (non- compositionally varying in these images). These SEM images were taken from a table-top Hitachi TM3000 scanning electron microscope.

After printing, the printed samples were cured at 130 °C for 30 min to remove water and other unwanted organic solvents. 5 wt.% of ethylene glycol (EG, 99.8%, Sigma-Aldrich) was added into the PEDOT:PPS ink and a post-curing surface treatment was also conducted to improve the final Seebeck coefficient and electrical conductivity of PEDOT:PSS matrix as previously disclosed in earlier publications from the present inventors [2], [14] Various loading weight percentage of solvothermal-synthesised Bi ₂ Te3 and Sb ₂ Te3 nanoparticles and MWCNTs (from 0 to 100 wt.% in nominal) were incorporated into the PEDOT:PSS matrix. Samples having a 100 wt% nominal loading are not according to the present invention: they are samples of aqueous dispersions of e.g. nanoparticles/MWCNTs printed onto a substrate without presence of a matrix material. Generally the thermoelectric properties of such materials are far inferior to those where a matrix material is present, because of the relatively poor connectivity between the nanoparticles in the absence of a matrix material.

Fig. 2 shows (a) a printed 15 wt.% BΪ ₂ Tb3— PEDOT:PSS nanocomposite + PEDOT:PSS CG-TEC. (b) Enlarged AutoCAD-designed pattern showing the transition region between first and second portions of the composite material; (c) an optical microscope image taken from Nikon Optiphot optical microscope; and (d), (e) SEM images taken from a table-top Hitachi TM3000 scanning electron microscope of the transition region between first and second portions of the composite material.

It can be seen that the 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite phase and pure PEDOT:PSS phases are well mixed in the transition region, thereby providing a progressively changing composition over a narrow region across the interface, as seen in the surface morphology. This continuously graded interface can help avoid interface problems e.g. cracks and other defects introduced by the thermal misfit and thermo-mechanical stress, and the diffusion or contamination issues in between dissimilar joint components, which would otherwise significantly degrade the performance and lifetime of TEG.

Figure 3 shows (a) SEM image of printed MWCNTs-PEDOT:PSS nanocomposite lines by aerosol-jet printing; (b) A fracture surface reveals that the MWCNTs were embedded within the PEDOT:PSS matrix and partially pulled out after bending and fracturing the nanocomposite; (c)-(f) Enlarged SEM images of printed MWCNTs-PEDOT:PSS nanocomposites with different loading ratio of MWCNTs from (c) 15 wt.%, (d) 50 wt.%, (e) 85 wt.%, up to (f) 100 wt.% (no matrix material). The imaged samples are not compositionally graded. Fig. 3 (a) & (b) SEM images were taken from a field-emission Nova NanoSEM 450 scanning electron microscope. Fig. 3 (c)-(f) SEM images were taken a table-top Hitachi TM3000 scanning electron microscope.

Dimension measurement

Following the sample printing, the length and width of printed strips were measured by a table-top scanning electron microscope (Hitachi TM3000), whilst their thickness was measured by a stylus profilometer (Veeco DEKTAK). For each sample, a minimum of three repeated measurements were conducted, with the average value being calculated to increase the accuracy. Given the measured dimensions, the electrical conductivity of the printed strip was then calculated by knowing its resistance value via the four-point probe measurement.

Measurement of power factor response with temperature for Printed Nanocomposites

S and a are temperature-dependent thermoelectric properties. Accordingly, the power factor PF is a temperature-dependent function i.e. PF(T). In order to obtain highest power output and/or overall efficiency of the thermoelectric generator, the material composition should be selected locally to achieve a desired PF response across the entire temperature range of use of the thermoelectric composite material. In order to achieve that, different printed PEDOT:PSS-based nanocomposites were prepared and measured to investigate how different loading ratio of inorganic components contribute to the temperature-dependent thermoelectric properties via an in-house designed measurement setup.

Different PEDOT:PSS-based nanocomposites loaded with different ratio of inorganic components were prepared and measured via a custom-built measurement setup, as shown in Fig. 4. The S and s values of each material composition samples were measured with dependence of temperature ranging from 293K to 363K varied by the hot plate and/or the water-cooling system. The error bars were calculated from the measured values of two repeated measurements for the same loading ratio sample.

As shown in Fig. 4, the printed thermoelectric thin-film sample, with the dimension of 20 mm in length and 2 mm in width, was fixed on top of two Peltier modules temperature controller (Adaptive®, 5.1 W) and laterally along the temperature gradient direction with the edges thermally contacting with two Pt-100 thermocouples (Farnell) for real-time temperature recording as well as electrically connecting with the enameled conducting copper wires for the generated voltage and resistance measurement. Furthermore, the entire test bench was situated in a sealed plastic box to ensure the minimisation of ambient electrical and thermal noise as well as prevention of heat dissipation whilst measurements were under way.

Lastly, a hardware and software data acquisition system was used to collect and analyse the semiautomatic data with the combination of two Keithley 196 digital multimeters and a Keithley 2002 digital multimeter for the sample temperature, voltage and resistance measurements. By varying the temperature across the thermoelectric leg, the Pt-100 thermocouples were used to measure the temperature gradient (DT), and a voltmeter was used to measure the voltage generated by the sample (AV). Then, a AV-DT plot was obtained, and the Seebeck coefficient (S) was calculated by linear fitting with the slope of the plot [1 ], [2] The electrical conductivity (a) was measured by the four-point probe approach using the same setup described above.

For a better view and understanding of the trend of temperature-dependent thermoelectric properties of different materials, individual graph for each sample’s S, a, and PF has been plotted and compared in Figs. 5, 6, 7 and 8.

To start with, the pristine PEDOT:PSS sample was prepared, and a temperature-dependent

thermoelectric measurement was conducted as shown in Fig. 5. It can be seen that the S values increased in direct proportion to the sample temperature and yielded the highest value of ~26.8 pV/K at 363 K. The s increased with increasing temperature, peaking at 343 K with a value of 704.5 S/cm. But, with further increase in temperature, the s decreased slightly. As a combination of the increased S and a, the maximum PF value of ~50.2 pW/rnK ² at 363 K. For a doped semiconductor, its s value generally increases steeply with temperature at low range of temperature due to the leaving of carriers from the donors or acceptors. Then, once most of the donors or acceptors have lost their carriers, the s value starts decreasing slight because of the reducing mobility of carriers [56] Since the PEDOT:PSS polymer used here is a doped semiconductor, our measured results followed a similar power factor response with temperature. The Bi ₂ Te3-PEDOT:PSS nanocomposite samples loaded with 15 wt.%, 35 wt.%, 50 wt.%, 65 wt.%, 85 wt.% and 90 wt.% B Te3 nanoparticles were printed and measured at the temperature range of 293K to 363K. For a better view and understanding of the trend of thermoelectric properties with temperature, each of composition was plotted as compared in Fig. 6. Compared with different temperature-dependent profile of these Bi ₂ Te3-PEDOT:PSS nanocomposites, it can be seen that with the more added inorganic components, the S values increased more dramatically, while the s values dropped even more steeply with increasing temperature, resulting in a gradually steep slope on the power factor response with temperature, and all peaking at 363 K. These profiles might be attributed to that the added metallic components rendered the nanocomposite more metallic-like. Because the number of free electrons in a unit volume of the conductor or semi-conductor rises exponentially with increasing temperature, leading to the dramatic decline in the relaxation time as well as the mean free path, and thereby significant drop in the s values [56] With the increasing amount of loaded BhTe3 nanoparticles, this phenomenon became more distinct.

For the MWCNTs-PEDOT:PSS nanocomposites loaded with 15 wt.%, 35 wt.%, 50 wt.%, 65 wt.%, 85 wt.% and 90 wt.% of MWCNTs, shown in Fig. 7, the awas observed to increase slightly and then exhibit a sharp downward trend with further increase in temperature, whereas the S increased gradually with their PF values reaching the highest at 363 K. The possible reason behind this phenomenon might be that as the temperature raised up at the beginning, the charge carrier density in the conduction band increased, resulting in a slight augment in the s [56]. With more added MWCNTs, the hybrid nanocomposites were rendered to be more metal-like overall. As the free electrons gained energy and started oscillating, bigger electron vibrations and more collisions between electrons took place with increasing temperature, leading to the increase of resistance.

The temperature-dependent thermoelectric properties of Sb ₂ Te3-PEDOT:PSS nanocomposite samples loaded with 15 wt.%, 35 wt.%, 50 wt.%, 65 wt.%, 85 wt.% and 90 wt.% of Sb ₂ Te3 nanoflakes were plotted and compared in Fig. 8. In this group, it shows that the s values decreased with the increase of temperature, which indicates more metallic-like behaviour of the nanocomposites. The significant enhancement in S led to all the PF values of different printed nanocomposites peaking at 363 K, which might be attributed to the higher S value of the p- type Sb ₂ Te3 nanoflakes.

It can be seen that each of the materials investigated here displays a different power factor response with temperature, in that the change in power factor with temperature is different for each material (the profile of a graph of PF vs Temperature is different).

Calculation of thermal conductivity of Thermoelectric Composite Materials

Table 1 summarises the material properties of different TE materials that were used to calculate the effective thermal conductivity of AJ-printed TE nanocomposites as a function of filler volume fraction.

Table 1 Summary of material properties and volume fraction values of different TE materials used for the thermal conductivity calculation and prediction of example thermoelectric composite materials.

For a 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite, the volume fraction of Bί ₂ Ϊb3 was calculated to be ~3 vol.% according to its nominal weight percentage. Thus, its thermal conductivity was calculated to be -0.33 W/(m.K), with density of -1212 kg/m ³ and heat capacity of -975 J/(kg.K), which was calculated based on the volume fraction of loaded BhTe3 as a weighted average.

For a 50 wt.% Sb ₂ Te3-PEDOT:PSS nanocomposite, the volume fraction of Sb ₂ Te3 was calculated to be -13 vol.% according to its nominal weight percentage. Thus, its thermal conductivity was calculated to be -0.48 W/(m.K), with density of -1725 kg/m ³ and heat capacity of -887 J/(kg.K), which was calculated based on the volume fraction of loaded Sb ₂ Te3 as a weighted average.

Fig. 16 shows the effective thermal conductivity of the medium K _e ff as a function of the volume fraction of inclusions as calculated using Maxwell’s formula. The model considers spherical particulates of thermal conductivity K* randomly dispersed in a medium with thermal conductivity K. Equation (1) below was evaluated first in the context of electrical conduction.

where f is the volume fraction occupied by the particulates. The model is valid for f < 25% and does not include the effects of interfacial scattering. Thus, Equation (1) gives an upper limit to Kett.

Fig. 16 shows / K for both Bi ₂ Te3 and Sb ₂ Te3 inclusions in a PEDOT:PSS matrix, and Table 1 shows the values of K and K* used. The values of the f relevant to examples above are indicated on the graph, and it can be seen that K _e n is enhanced by = 4% and 25% for Bi ₂ Te3 and Sb ₂ Te3, respectively. The above estimates are in good agreement with our calculations of thermal conductivity of the composites, and represent an upper limit. The exact thermal conductivity of the thermoelectric composites in practice will be slightly lower than this upper limit, due to the phonon scattering effect reported in literature. This is advantageous from the point of view of thermoelectric performance.

Therefore, the above calculated thermal conductivity values were adopted in COMSOL simulations for the compositionally graded thermoelectric composite described herein (for example the COMSOL simulations shown in Fig. 1 1 (a) and (b), discussed below). Production of Thermoelectric Composite Materials

The temperature-dependent PF values of different printed PEDOT:PSS-based nanocomposites loaded with different wt.% of various nanomaterials were plotted and compared in Fig. 9 (a) to find out the proper material selection fitting to different temperature. The materials were investigated across an applied temperature range of Tc = 293 K to TH = 363 K (a gradient of 70K).

4 specific compositions were selected for closer review. These are compared in Fig. 9 (b), although specific‘materials matching’ plots for different pairs of materials are shown in Fig. 10.

It can be seen from Fig. 10(a) that the material match of 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite (15 wt.% B Te3 + PEDOT:PSS) and pristine PEDOT:PSS shows an intersection temperature Tx of 318K, where the 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite exhibited higher PF values than that of pristine PEDOT:PSS below 318 K, while above that, the pristine PEDOT:PSS surpasses its counterpart.

Similarly, Fig 10(b) shows the material match of 50 wt.% Sb ₂ Te3-PEDOT:PSS nanocomposite (50 wt.% Sb ₂ Te3 + PEDOT:PSS) and pristine PEDOT:PSS. The intersection temperature Tx was 313 K.

Fig 10(c) shows the material match of 85 wt.% MWCNTs-PEDOT:PSS nanocomposite (85 wt.%

MWCNTs + PEDOT:PSS) and pristine PEDOT:PSS. The intersection temperature Tx was 313 K.

Fig 10(d) shows the material match of 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite (15 wt.% Bi ₂ Te3 + PEDOT:PSS) and 50 wt.% Sb ₂ Te3-PEDOT:PSS nanocomposite (50 wt.% Sb ₂ Te3 + PEDOT:PSS). The intersection temperature Tx was 330 K.

Using the calculated intersection temperature of the power factor response for each material, a compositionally graded thermoelectric composite material could be designed and fabricated, by varying the loading (wt%) of the or each filler material along the length of the material, thereby defining first and second portions of the material having different power factor response with respect to temperature. In the examples discussed below, the boundary between the first and second portions was selected to be at or near the location of the intersection temperature of the power factor response profiles for each of the two materials, as calculated by modelling the theoretical temperature distribution along the material under an applied temperature gradient.

In order to design the printing pattern for the examples discussed below, a finite element analysis simulation based on COMSOL Multiphysics (see Fig. 11 (a), (b)) was conducted to simulate the heat flow and temperature distribution along the printed samples. The Heat Transfer in Solids physics interface (a function of the COMSOL Multiphysics program) was used to define heat throughout the printed thermoelectric leg. The temperature is shown by variation in grayscale. The black line indicates the location of the intersection temperature, and accordingly the location of the boundary between first and second portions of the material. The temperature at either end was set to 293 K and 363 K respectively to thereby provide a 70 K temperature difference across the material. The effect of contact thermal resistance at the interface was neglected as the same polymer matrix was used across the interface, and the volumetric loading fraction of the nanoparticles was low. The appropriate boundary location was determined based on simulation results to be approximately 7.5 mm from the cold end of a 20 mm long strip of material, for the 15 wt.% Bi ₂ Te3 -PEDOT :PSS +

PEDOT:PSS based thermoelectric composite material (i.e. from 0 mm to 7.5 mm was composed of 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite, whereas from 7.5 mm to 20 mm was pristine PEDOT:PSS).

For for the 50 wt.% Sb ₂ Te3-PEDOT:PSS + PEDOT:PSS based thermoelectric composite material, the appropriate boundary location was determined based on simulation results to be approximately 8 mm from the cold end of a 20 mm long strip of material, (i.e. from 0 mm to 8 mm was composed of 50 wt.% Sb ₂ Te3 -PEDOT:PSS nanocomposite, whereas from 8 mm to 20 mm was pristine PEDOT:PSS).

As illustrated in Fig. 1 1 (c) and Fig. 1 1 (d), the AutoCAD-designed thermoelectric patterns were derived accordingly, and the thermoelectric composite material samples were printed via an aerosol jet printing technique as discussed below (see Material & Methods).

Furthermore, single-phase pristine PEDOT:PSS, 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite, and 50 wt.% Sb ₂ Te3-PEDOT:PSS nanocomposite were also prepared as control samples to compare their thermoelectric performance with that of thermoelectric composite material examples according to the invention.

Production of Thermoelectric Generators comprising thermoelectric composite materials

The voltage generated by samples was measured by fixing the sample laterally with a constant AT applied along the lateral direction of the printed sample. The thermoelectric generators were designed, assembled, and tested under variable load resistances by impedance matching to determine their maximum power output. This was done by connecting the printed samples with various external loaded resistors via a resistance decade box as illustrated in Fig. 12 (the resistance decade box being represented as a variable resistor). In order to control a precise temperature different between either end of the thermoelectric leg, two Pt-100 thermocouples were attached next to their edges so that the realtime temperature was known and could be adjusted according via two Peltier module controllers, where one was performed as a heat source while the other as a heat sink to ensure a stable temperature gradient over the sample.

These printed thermoelectric generators can be viewed as a thermal battery, where the electromotive force is the Seebeck voltage D\/ = - SAT. The maximum power output was determined via impedance matching with variable load resistors under a constant applied temperature difference AT ~70 K. The external loaded resistance was varied from 1 W to 1 MW via a resistance decade box and their output voltage was measured via a multimeter (Keithley 2002 digital multimeter). The external loaded resistance and output voltage was plotted, where the output power P = IR was calculated accordingly. The maximum output power should be achieved when the internal sample resistance equals the external loaded resistance i.e. Rs = Ri. Fig. 13 shows the power output and voltage output plotted against various load resistance of various thermoelectric generators under a temperature difference of 70 K: (a) pristine PEDOT:PSS film, (b) 15 wt.% Bi ₂ Te3-PEDOT:PSS nanocomposite, (c) 50 wt.% Sb ₂ Te3-PEDOT:PSS nanocomposite, (d) 15 wt.% Bi ₂ Te3-PEDOT:PSS + PEDOT:PSS compositionally graded thermoelectric generator, (e) 50 wt.% Sb ₂ Te3-PEDOT:PSS + PEDOT:PSS compositionally graded thermoelectric generator, and (f) 15 wt.% Bi ₂ Te3-PEDOT:PSS + 50 wt.% Sb ₂ Te3-PEDOT:PSS compositionally graded thermoelectric generator. Those shown in Fig. 13 (d),(e) and (f) are according to the invention. Those shown in Fig. 13 (a), (b) and (c) are comparative.

The power output of compositionally graded thermoelectric generators were tested twice: one temperature gradient was along the correct direction (15 wt.% Bi ₂ Te3-PEDOT:PSS and 50 wt.% Sb ₂ Te3- PEDOT:PSS part on the cold side and PEDOT:PSS part on the hot side) while the other in reverse direction. These differing power outputs can be seen in Fig. 14, which shows output voltage and power measurements of 15 wt.% Bi2Te3— PEDOT:PSS nanocomposite + PEDOT:PSS CG-TEG in the forward direction (Fig. 14(a)) and reverse direction (Fig. 14(b)), along with output voltage and power

measurements of 50 wt.% Sb2Te3— PEDOT:PSS nanocomposite + PEDOT:PSS CG-TEG in the forward direction (Fig. 14(c)) and reverse direction (Fig. 14(d)).

It was also found that for the comparative single-phase printed TEGs, there was no significant difference on their output power when switching the temperature gradient direction. However, a decrease in the output power was seen in both compositionally graded thermoelectric generators when the temperature gradient was applied in the reverse direction, as shown in Fig. 14, thereby indicating improved effectiveness of the compositionally graded thermoelectric generators across a temperature gradient in comparison to the comparative single-phase printed TEGs.

After resistance matching with all samples, it is found that they all had similar internal resistance, and had a peak power output value under the loaded resistance of 50 W. For a more straightforward comparison of the power output between different samples, all the power output values at 50 W were re-plotted as shown in Fig. 15. It can be seen that each of the compositionally graded thermoelectric generators were superior to the comparative examples under the same applied temperature difference.

The 15 wt.% Bi ₂ Te3-PEDOT:PSS + PEDOT:PSS compositionally graded thermoelectric generator exhibited the most enhanced power output value ~13 nW and power area density of ~ 100 pW/cm ³ , and the 50 wt.% Sb ₂ Te3-PEDOT:PSS + PEDOT:PSS compositionally graded thermoelectric generator came second with power output of ~10 nW. These designs were particularly effective in optimising

thermoelectric performance and power output compared with other single-phase homogeneous samples. Since only a single compositionally graded thermoelectric member was used in each tested generator, only a relatively small difference in power output was seen. However, this difference would be amplified by use of a plurality of such thermoelectric members in a thermoelectric generator device.

The features disclosed in the foregoing description, or in the following claims, or in the accompanying drawings, expressed in their specific forms or in terms of a means for performing the disclosed function, or a method or process for obtaining the disclosed results, as appropriate, may, separately, or in any combination of such features, be utilised for realising the invention in diverse forms thereof. While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.

For the avoidance of any doubt, any theoretical explanations provided herein are provided for the purposes of improving the understanding of a reader. The inventors do not wish to be bound by any of these theoretical explanations.

Any section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described.

Throughout this specification, including the claims which follow, unless the context requires otherwise, the word“comprise” and“include”, and variations such as“comprises”,“comprising”, and“including” will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps. It must be noted that, as used in the specification and the appended claims, the singular forms“a,”“an,” and“the” include plural referents unless the context clearly dictates otherwise. Ranges may be expressed herein as from“about” one particular value, and/or to“about” another particular value. When such a range is expressed, another embodiment includes from the one particular value and/or to the other particular value. Similarly, when values are expressed as approximations, by the use of the antecedent“about,” it will be understood that the particular value forms another embodiment. The term“about” in relation to a numerical value is optional and means for example +/- 10%.

References

Further work and discussion relating to the present invention can be found in the following reference:

Ou, C. (2020). Aerosol-Jet Printed Nanocomposites for Flexible and Stretchable Thermoelectric

Generators (Doctoral thesis). https://doi.org/10.17863/CAM.50693

Additionally, a number of publications are cited above in order to more fully describe and disclose the invention and the state of the art to which the invention pertains. Full citations for these references are provided below.

The entirety of each reference referred to in this specification is incorporated herein.

(1 ) Martin-Gonzalez, M.; Caballero-Calero, O.; Diaz-Chao, P. Nanoengineering Thermoelectrics for 21 st Century: Energy Harvesting and Other Trends in the Field. Renew. Sustain. Energy Rev. 2013, 24, 288-305.

(2) Ou, C.; Sangle, A. L; Datta, A.; Jing, Q.; Busolo, T.; Chalklen, T.; Narayan, V.; Kar-Narayan, S. Fully Printed Organic-Inorganic Nanocomposites for Flexible Thermoelectric Applications. ACS Appl. Mater. Interfaces 2018, 10 ( 23), 19580-19587.

(3) Ou, C.; Sanchez-Jimenez, P. E.; Datta, A.; Boughey, F. L.; Whiter, R. A.; Sahonta, S.-L; Kar- Narayan, S. Template-Assisted Hydrothermal Growth of Aligned Zinc Oxide Nanowires for Piezoelectric Energy Harvesting Applications. ACS Appl. Mater. Interfaces 2016, 8 (22), 13678-13683.

(4) Vining, C. B. An Inconvenient Truth about Thermoelectrics. Nat. Mater. 2009, 8 (2), 83-85.

(5) Taroni, P. J.; Hoces, I.; Stingelin, N.; Heeney, M.; Bilotti, E. Thermoelectric Materials: A Brief Historical Survey from Metal Junctions and Inorganic Semiconductors to Organic Polymers. Isr. J. Chem. 2014, 54 (5-6), 534-552.

(6) Chen, Y.; Zhao, Y.; Liang, Z. Solution Processed Organic Thermoelectrics: Towards Flexible Thermoelectric Modules. Energy Environ. Sci. 2015, 8 (2), 401 -422.

(7) Ou, C. Flexible Printed Thermoelectric Nanogenerators Based on Hybrid Materials, University of Cambridge, 2017.

(8) Russ, B.; Glaudell, A.; Urban, J. J.; Chabinyc, M. L.; Segalman, R. A. Organic Thermoelectric Materials for Energy Harvesting and Temperature Control. Nat. Rev. Mater. 2016, 1 (10), 16050-16062.

(9) Tritt, T. M.; Subramanian, M. a. Thermoelectric Materials, Phenomena, and Applications: A Bird’s Eye View. MRS Bull. 2006, 31 (03), 188-198.

(10) Poudel, B.; Hao, Q.; Ma, Y.; Lan, Y.; Minnich, A.; Yu, B.; Yan, X.; Wang, D.; Muto, A.; Vashaee, D.; Chen, X.; Liu, J.; Dresselhaus, M. S.; Chen, G.; Ren, Z. High-Thermoelectric Performance of Nanostructured Bismuth Antimony Telluride Bulk Alloys. Science 2008, 320 (5876), 634-638.

(1 1) Snyder, G. J.; Toberer, E. S. Complex Thermoelectric Materials. Nat. Mater. 2008, 7 (2), 10S- 1 14.

(12) Venkatasubramanian, R.; Siivola, E.; Colpitts, T.; O’Quinn, B. Thin-Film Thermoelectric Devices with High Room-Temperature Figures of Merit. Nature 2001 , 413 (6856), 597-602. (13) Bubnova, O.; Khan, Z. U.; Malti, A.; Braun, S.; Fahlman, M.; Berggren, M.; Crispin, X.

Optimization of the Thermoelectric Figure of Merit in the Conducting Polymer Poly(3,4- Ethylenedioxythiophene). Nat. Mater. 2011 , 10 (6), 429-433.

(14) Ou, C.; Sangle, A. L; Chalklen, T.; Jing, Q.; Narayan, V.; Kar-Narayan, S. Enhanced

Thermoelectric Properties of Flexible Aerosol-Jet Printed Carbon Nanotube-Based Nanocomposites. APL Mater. 2018, 6 (9), 096101 .

(15) Cramer, C. L. C. L.; Wang, H.; Ma, K. Performance of Functionally Graded Thermoelectric Materials and Devices: A Review. J. Electron. Mater. 2018.

(16) Rao, A. M.; Ji, X.; Tritt, T. M. Properties of Nanostructured One-Dimensional and Composite Thermoelectric Materials. MRS Bull. 2006, 31 (03), 218-223.

(17) Zebarjadi, M.; Esfarjani, K.; Dresselhaus, M. S.; Ren, Z. F.; Chen, G. Perspectives on

Thermoelectrics: From Fundamentals to Device Applications. Energy Environ. Sci. 2012, 5 (1), 5147- 5162.

(18) Datta, A.; Paul, J.; Kar, A.; Patra, A.; Sun, Z.; Chen, L.; Martin, J.; Nolas, G. S. Facile Chemical Synthesis of Nanocrystalline Thermoelectric Alloys Based on Bi-Sb-Te-Se. Cryst. Growth Des. 2010, 10 (9), 3983-3989.

(19) Dubey, N.; Leclerc, M. Conducting Polymers: Efficient Thermoelectric Materials. J. Polym. Sci. Part B Polym. Phys. 2011 , 49 (7), 467-475.

(20) He, M.; Qiu, F.; Lin, Z. Towards High-Performance Polymer-Based Thermoelectric Materials. Energy Environ. Sci. 2013, 6 (5), 1352-1361 .

(21) Zhang, Q.; Sun, Y.; Xu, W.; Zhu, D. Organic Thermoelectric Materials: Emerging Green Energy Materials Converting Heat to Electricity Directly and Efficiently. Adv. Mater. 2014, 26 (40), 6829-6851 .

(22) Culebras, M.; Gomez, C.; Cantarero, A. Review on Polymers for Thermoelectric Applications. Materials (Basel). 2014, 7 (9), 6701-6732.

(23) Shakouri, A. Recent Developments in Semiconductor Thermoelectric Physics and Materials. Annu. Rev. Mater. Res. 2011 , 41 (1), 399-431 .

(24) A.F. Ioffe. Byulleten’ Izobretenii. 126158, 1960.

(25) Miiller, E.; Drasar, C.; Schilz, J.; Kaysser, W. . Functionally Graded Materials for Sensor and Energy Applications. Mater. Sci. Eng. A 2003, 362 (1-2), 17-39.

(26) Shiota, I.; Nishida, A. Development of FGM Thermoelectric Materials in Japan -The State of the Art. In 16th International Conference on Thermoelectrics·, 1997; pp 364-370.

(27) Kajikawa, T. Functionally Graded Materials. In the 4th International Symposium on Functionally Graded Materials·, Shiota, I., Miyamoto, Y., Eds.; Elsevier Science B.V., 1996; p 475.

(28) Ge, C. C.; Wu, X. F.; Xu, G. Y. Functionally Graded Thermoelectric Materials. Key Eng. Mater. 2007, 336-338, 2600-2604.

(29) Anatychuk, L. I.; Vikhor, L. N.; Post-office, G. Functionally Graded Materials and New Prospects for Thermoelectricity Use. In 16th International Conference on Thermoelectrics·, 1997; pp 588-591 .

(30) Meroz, O.; Gelbstein, Y. Thermoelectric Bi 2 Te 3-x Se x Alloys for Efficient Thermal to Electrical Energy Conversion. Phys. Chem. Chem. Phys. 2018, 20 (6), 4092-4099.

(31) Januszko, K.; Chetty, R.; Mashimo, T.; Wojciechowski, K. Characterization of the Starting Materials for Functionally Graded Thermoelectric Materials for Pseudobinary System of Bi2Te3-Sb2Te3. Mechanik 2016, No. 5-6, 522-523. (32) Hazan, E.; Ben-Yehuda, O.; Madar, N.; Gelbstein, Y. Functional Graded Germanium-Lead Chalcogenide-Based Thermoelectric Module for Renewable Energy Applications. Adv. Energy Mater.

2015, 5 (11), 1500272.

(33) Hedegaard, E. M. J.; Johnsen, S.; Bjerg, L.; Borup, K. A.; Iversen, B. B. Functionally Graded Ge 1- x Si x Thermoelectrics by Simultaneous Band Gap and Carrier Density Engineering. Chem. Mater. 2014, 26 (17), 4992-4997.

(34) Noda, Y.; Orihashi, M.; Kaibe, H.; Imai, Y.; Shiota, I.; Nishida, I. A. High Efficient Program and Carrier Concentration FGM Thermoelectric Materials in the Intermediate Temperature Region. In 15th International Conference on Thermoelectrics ; 1996; pp 146-150.

(35) Gelbstein, Y.; Dashevsky, Z.; Dariel, M. P. Powder Metallurgical Processing of Functionally Graded P-Pb1-XSnxTe Materials for Thermoelectric Applications. Phys. B Condens. Matter 2007 , 391 (2), 256-265.

(36) Bharti, I.; Gupta, N.; Gupta, K. M. Novel Applications of Functionally Graded Nano,

Optoelectronic and Thermoelectric Materials. Int. J. Mater. Mech. Manuf. 2013, 1 (3), 221-224.

(37) Dashevsky, Z.; Gelbstein, Y.; Edry, I.; Drabkin, I.; Dariel, M. P. Optimization of Thermoelectric Efficiency in Graded Materials. In 22th International Conference on Thermoelectrics·, 2003; pp 421-424.

(38) Cui, J. L.; Zhao, X. B. The Design and Properties of Pseudobinary Alloys (PbTe)1-x-(SnTe)x with Gradient Composition. Mater. Lett. 2003, 57 (16-17), 2466-2471.

(39) Dashevsky, Z.; Dariel, M. P.; Shusterman, S. Functionally Graded PbTe-Based Compound for Thermoelectric Applications. Quantum Electron. 2000, No. 2, 181-184.

(40) Hedegaard, E. M. J.; Mamakhel, A. A. H.; Reardon, H.; Iversen, B. B. Functionally Graded (PbTe) 1- x (SnTe) x Thermoelectrics. Chem. Mater. 2018, 30 (1), 280-287.

(41) Cramer, C. L.; Gonzalez-Julian, J.; Colasuonno, P. S.; Holland, T. B. Continuous Functionally Graded Material to Improve the Thermoelectric Properties of ZnO. J. Eur. Ceram. Soc. 2017, 37 (15), 4693-4700.

(42) Schilz, J.; Helmers, L.; Miiller, W. E.; Niino, M. A Local Selection Criterion for the Composition of Graded Thermoelectric Generators. J. Appl. Phys. 1998, 83 (2), 1150-1152.

(43) Li, H.; Jing, H.; Han, Y.; Lu, G. Q.; Xu, L.; Liu, T. Interface Evolution Analysis of Graded

Thermoelectric Materials Joined by Low Temperature Sintering of Nano-Silver Paste. J. Alloys Compd.

2016, 659, 95-100.

(44) Itoh, T.; Muto, T.; Kitagawa, K. Performance of Segmented Thermoelectric Elements Fabricated by Simultaneous Sintering Method. In 2006 International Conference on Thermoelectrics·, IEEE, 2006; pp 623-626.

(45) Kang, Y. S.; Niino, M.; Nishida, I. A.; Yoshino, J. Development and Evaluation of 3-Stage Segmented Thermoelectric Elements. In 17th International Conference on Thermoelectrics·, 1998; pp 429-432.

(46) Kaibe, H.; Aoyama, I.; Mukoujima, M.; Kanda, T.; Fujimoto, S.; Kurosawa, T.; Ishimabushi, H.; Ishida, K.; Rauscher, L.; Hata, Y.; Sano, S. Development of Thermoelectric Generating Stacked Modules Aiming for 15% of Conversion Efficiency. In 2005 International Conference on Thermoelectrics·, IEEE, 2005; pp 227-232. (47) Kang, Y. S.; Moriya, S.; Kisara, K.; Niino, M.; Noda, Y.; Chen, L; Sudo, T. Evaluation of Monolithic and Segmented Thermoelectric Materials by Using a Large-Temperature-Span Apparatus. In 16th International Conference on Thermoelectrics ; 1997; pp 390-393.

(48) Muto, T.; Tokuda, K.; Itoh, T.; Kitagawa, K. Fabrication of Segmented P-Type AgSbTe2/Sb2Te 3/BiO.4Sb1 6Te3 Thermoelectric Module and Its Performance. In 2005 International Conference on Thermoelectrics·, 2005; Vol. 2005, pp 524-527.

(49) Schilz, J.; Helmers, L.; Kang, Y. S.; Noda, Y.; Niino, M. Bismuth-Telluride/lron-Disilicide Segmented Thermoelectric Elements: Patterning, Preparation and Properties. In 16th International Conference on Thermoelectrics·, 1997; pp 375-378.

(50) El-genk, M. S.; Saber, H. H. High Efficiency Segmented Thermoelectric Unicouple for Operation between 973 and 300 K. Energy Corners. Manag. 2003, 44, 1069-1088.

(51 ) Wallace, T. T.; Jin, Z.-H.; Su, J. Efficiency of a Sandwiched Thermoelectric Material with a Graded Interlayer and Temperature-Dependent Properties. J. Electron. Mater. 2016, 45 (4), 2142-2149.

(52) Koumoto, K.; Mori, T. Thermoelectric Nanomaterials·, Springer Series in Materials Science; Springer Berlin Heidelberg: Berlin, Heidelberg, 2013; Vol. 182.

(53) Glatz, W.; Schwyter, E.; Durrer, L.; Hierold, C. Bi2Te3-Based Flexible Micro Thermoelectric Generator With Optimized Design. J. Microelectromechanical Syst. 2009, 18 (3), 763-772.

(54) Elmogi, A.; Soenen, W.; Bosman, E.; Missinne, J.; Spiga, S.; Amann, M.; Bauwelinck, J.;

Steenberge, G. Van. Flexible Hybrid Integration of Photonic and Electronic Chips Using Aerosol-Jet Printing.

(55) Smith, M.; Choi, Y. S.; Boughey, C.; Kar-Narayan, S. Controlling and Assessing the Quality of Aerosol Jet Printed Features for Large Area and Flexible Electronics. Flex. Print. Electron. 2017, 2 (1), 015004-015014.

(56) Ashcroft, N. W.; Mermin, N. D. Solid State Physics ; Harcourt, 1976.