Flexible and Printed Organic Thermoelectrics: Opportunities and Challenges

Introduction

Technologies that promote sustainable energy are experiencing unprecedented interest due to the growing population and the need to meet higher living standards. According to the United States Department of Energy, up to 50 percent of the energy from all fuels burned in the U.S. ends up in the environment as waste heat.¹ If the waste heat from commercial and industrial operations could be collected and converted to electricity, it would fulfill up to 20 percent of the total U.S. electricity demand. Thermoelectrics directly converts thermal energy into electrical energy and is among the most promising renewable energy technologies.

Thermoelectric materials, when placed into a thermal gradient, generate an electrical potential due to the Seebeck effect (Figure 1). The reverse process is known as the Peltier effect. The temperature gradient causes the diffusion of charge carriers with the resulting electrical potential directly proportional to the temperature differences across the material,

where S is the Seebeck coefficient. The thermoelectric figure of merit, ZT is defined as,

with σ the electrical conductivity and k the thermal conductivity. The Seebeck coefficient Ѕ, characteristic of the average entropy per charge transport, should be large in order to create a high voltage induced by a temperature gradient. The Seebeck coefficient, however, is not the only parameter to be optimized in order to maximize ZT. The electrical conductivity σ, must be large to minimize the joule heating dissipation during the charge transport. Besides the two parameters mentioned, a good thermoelectric material should also feature low thermal conductivity k, to prevent heat flow through the material. The difficulty of designing high performance thermoelectric materials arises from the fact that both electrical and thermal conductivity are related via the carrier concentration; optimizing one parameter will negatively affect the other.² This interdependence has hindered the development of thermoelectric materials for many years. While progress has been made in inorganic nanostructured materials, the increasing complexity in material design complicates the introduction of thermoelectric generators as an alternative energy technology to the mass market due to high manufacturing costs and sophisticated fabrication processes.

Organic thermoelectric materials have attracted increased attention as an alternative approach to conventional thermoelectronics and are experiencing rapid development. These materials are particularly attractive for low-quality waste heat harvesting. Several conventional organic semiconductors exhibit good thermoelectric performances. High ZT values (over 0.1) have been achieved for both p-type and n-type organic thermoelectric materials.^3,4 Both inorganic and organic thermoelectrics are described as “electronic crystals and phonon glasses”. In contrast to inorganic thermoelectrics, the study of organic thermoelectrics aims to achieve “electronic crystals”, because organic semiconductors have intrinsically low thermal conductivities. Increasing S², namely Power Factor (PF), is the main route to enhance the performance of organic thermoelectrics. The expression of electrical conductivity (σ), where e is electrical charge for each carrier, μ is the mobility, and n is the concentration of carriers is shown here:

However, the Seebeck coefficient S typically decreases with increasing carrier concentration. This trade-off between S and σ requires optimization of power factors, which is currently achieved by precisely tuning the doping level.

Progress in Organic Thermoelectric Materials

Organic semiconductors have been largely neglected as thermoelectric materials, despite their inherent low thermal conductivities (≈ 0.3 W m^–1 K^–1) and high electrical conductivities (> 1,000 S cm^–2).^5–6 Two of the challenges encountered in developing organic thermoelectrics are the carrier concentrations and the device architecture. First, it is essential to control the charge carrier density of organic semiconductors by doping in order to attain high electrical conductivities and Seebeck coefficients. Second, the most common thermoelectric generator architecture relies on two semiconductor legs, one p-type and one n-type, connected thermally in parallel, but electrically in series. A list of representative p- and n-type semiconducting polymers and dopants are listed in Figure 2.

The most successful p-type material developed to date is undoubtedly, poly(3,4-ethylenedioxy thiophene) polystyrene sulfonate (PEDOT:PSS), a polyelectrolyte consisting of positively charged PEDOT and negatively charged PSS. PEDOT:PSS is commercially available as a dispersion in water, which facilitates processing and greatly contributes to its popularity. PEDOT:PSS can be easily processed using various deposition techniques. The deposition protocol and the processing additives have a tremendous influence on the electric conductivities, which can range from 10^–2 to 10³ S cm^–1.⁷ The most widely used additives are high-boiling co-solvents, surfactants, (i.e. dimethyl sulfoxide, ethylene glycol, or Zonyl^®), which lead to more ordered PEDOT domains and higher conductivities.⁸ However, in contrast to the conjugated PEDOT, saturated PSS is insulating, hindering charge transport. By stripping the PSS from the deposited PEDOT:PSS film, it is possible to increase the weight fraction of conducting PEDOT in the film, leading to significantly improved conductivities of around 1,000 S cm^–1.⁴ The electronic properties of PEDOT:PSS can be further enhanced by carefully controlling the film deposition, thereby gaining control over the phase separation in the PEDOT:PSS film and reducing disorder along the PEDOT backbone, resulting in record conductivities of 4,600 (±100) S cm^–1.⁹ The continually improving conductivities of PEDOT:PSS have also increased its prospect for thermoelectric applications. The ease with which PEDOT films can be doped and dedoped offers an ideal processing handle to tune not only the electrical conductivities, but also the Seebeck coefficients. Crispin et al. reported impressive power factors of 300 μW m^–1 K^–2 and ZT values of 0.25, after dedoping highly conductive PEDOT:tosylate with tetrakis(dimethylamino)ethylene (Cat. No. 674613).¹⁰ Besides dedoping PEDOT:PSS, the addition of carbon nanotubes or graphene is a popular approach to modulate the thermoelectric properties.¹¹ By carefully structuring the composite film, conductivities of 10⁵ S m^–1 and Seebeck coefficients of 120 μVK^–1 were measured, leading to thermoelectric power factors of 2710 μW m^–1 K^–2, one of the highest values reported for organic thermoelectric materials.¹²

While the p-doping of organic semiconductors can be readily achieved, n-doping is more difficult. Organic electron deficient semiconductors are associated with the high electron affinities (–3 to –4 eV), making the negatively charged molecules prone to reactions with environmental moisture or oxygen.^13,14 Charge transfer salts were among the first organic n-type conductors exploited for thermoelectric applications. The co-crystal of tetrathiafulvalene (TTF) and tetracyanoquinodimethane (TCNQ) is probably the most studied charge transfer salt that exhibited promising thermoelectric properties. Electrical conductivities of 500 S cm^–1 and power factors of up to 40 μW m^–1K^–1 were obtained. However, several drawbacks significantly limit the applicability of charge transfer crystals as thermoelectric materials. First, modulating the carrier densities is difficult, because the stoichiometry of the co-crystals must be meticulously respected, leaving very little room for composition alterations. Second, the physical properties of the co-crystals are not isotropic, but are highly dependent on the different crystal axes.¹⁵ Alternative approaches to developing n-type conductors for thermoelectric applications have mainly focussed on the perylenediimide and naphtalenediimide containing organic semiconductors. Segalman et al. developed a series of perylene diimide (PDI) based molecular semiconductors functionalized with tertiary amine-containing side chains.¹⁶ Upon thermal annealing, the functionalized PDI moieties self-dope via a dehydration reaction of the tethered tertiary ammonium hydroxide. Interestingly, the self-doped PDI compounds exhibited respectable ambient stability and thermoelectric properties.¹⁷ By carefully designing the side chains, the self-doped PDI moieties achieve conductivities of 0.5 S cm^–1 and power factors as high as 1.4 μW m^–1 K^–2. Based on the complex X-ray diffraction patterns of the self-doped PDIs, the charge carrier transport is most likely limited by inter-crystalline defects, leaving ample room for material optimization. Chabinyc et al. extrinsically doped the high performing n-type polymer poly(N,N′-bis(2- octyldodecyl)- 1,4,5,8-napthalenedicarboximide-2,6-diyl]-alt-5,5′- (2,2′-bithiophene)) (P(NDIOD-T2, Cat. No. 900961) with the molecular dopant (4-(1,3-dimethyl-2,3-dihydro-1H-benzoimidazol- 2-yl)phenyl) (N-DMBI, Cat. No. 776734).¹⁸ While the conductivity initially increases as a function of dopant loading, a sharp drop in conductivity is observed at higher loadings. The miscibility of the N-DMBI dopant in the polymer phase is limited, which is why at higher dopant loadings, the dopant crystallizes and phase separates from the polymer matrix, thereby reducing the doping efficiency. Despite the morphological instabilities, Seebeck coefficients of –850 μVK^–1 have been achieved with power factors of 0.6 μW m^–1 K^–2. In a recent report, Pei et al. showed that BDOPV-based FBDPPV polymers have reached a record power factor of 28 μW m^–1 K^–2.¹⁹ More interestingly, Huang et al. reported that thiophene-diketopyrrolopyrrole-based quinoidal (TDPPQ) can exhibit a high power factor of 113 μW m^–1 K^–2, when the material is interfacially doped by the bismuth. The performance is a record value for reported n-type small molecules.²⁰

Progress in Organic Thermoelectric Devices

Even though the main research efforts in organic thermoelectrics still focus on the development of state-of-the-art materials, device engineering is an important avenue of investigation for organic thermoelectrics for two main reasons. Although power factor and ZT values are widely applied to study the structureproperty relationship of organic materials, the output power of devices offers an ultimate way to evaluate thermoelectric performance and clarify related relationships. Moreover, from the point of application-motivated research, the conversion efficiency of a thermoelectric device is strongly dependent on device geometry and the interface property. Therefore, novel designs of device geometry, construction of flexible thermoelectric modules, and development of solution processing (roll-to-roll and ink-jet printing) techniques can boost the power generation, Peltier cooling, and many other applications.

Until now, only a few power generators based on organic materials have been demonstrated. As an example, Crispin et al. reported an organic power generator consisting of 55 legs in a vertical architecture.²¹ For this device, a leg with a dimension of 1 × 1.5 × 0.03 mm³ is fabricated by filling cavities with precursor solutions with micropipettes. It should be noted that p-legs are composed of PEDOT-Tos deposited by chemical polymerization of EDOT, while the n-legs are composed of a blend of TTF-TCNQ with PVC. The thermoelectric module exhibits a maximum power output of 0.128 μW at DT of 10 K. Despite this interesting result, further improvement of the power is limited by the low ZT value of the n-type materials and small DT because of limited thickness. Another organic thermoelectric power generator in vertical geometry was reported by Zhu’s group.²² By using metal-organic conducting polymers such as poly-(metal 1,1,2,2-ethenetetrathiolate) as the n-type and p-type materials, large legs with dimensions of 5 × 2 × 0.9 mm³ were fabricated by pressing the polymer powder into pellets. In this way the thermoelectric module serves as a robust device for high power generation. Benefiting from legs with more balanced thermoelectric performance and increased thickness, a thermoelectric module consisting of 35 legs generated a maximum power and VOC of 750 μW and 0.26 V, respectively, when the DT is maintained at 82 K. The power output density reached 1.2 μW cm^–2 at 30 K.

Construction of flexible thermoelectric prototype modules is of vital importance for practical applications of organic thermoelectric materials. Recently, Zhu et al. demonstrated a highly integrated flexible module with 220 legs.²³ The prototype device showed an open voltage of 1.51 V and a short current of 2.71 mA. The maximum power output exceeded 1 mW, which is the best performance for organic thermoelectric module published to date. The relative high output voltage and large power output of the constructed module were even sufficient to independently drive a calculator with liquid crystal display. Interestingly, Fujifilm Corporation recently demonstrated a flexible thermoelectric converter²⁴ with a power generation capacity of few milliwatts capable of generating electricity with a temperature difference of only 1 K. The temperature difference resulting from placing a hand on the device creates enough electricity to power a toy car. It is expected the module will be used as a wearable power supply for a health monitoring device and will incorporate other integrated elements that harvest energy.

In addition to the conventional power generators, organic thermoelectric devices for multifunctional applications have recently attracted increased attention. For instance, organic thermoelectric devices can also be utilized to harvest light energy directly to serve as photo-thermo-electric (PTE) converters. In two recent works done by Kim et al. and Huang et al., PTE devices based on photoselenophene derivatives, (hexyl- 3,4-ethyl-enedioxyselenophene, EDOS-C6) and poly[Cux(Cuett)], have been constructed independently.^25,26 For these PTE devices, both photothermal (PT) and heat driven thermoelectric conversion can occur in a single element to allow PTE conversion. As for the EDOS-C6 based device, a high Seebeck voltage of up to 900 μV under moderate NIR light exposure (808 nm @ 2.33 W cm^–2) can be obtained. In addition to this PTE conversion, photoinduced excitation of the organic active materials can also occur, with an obvious influence on the thermoelectric properties. In Huang’s work, the NIR light irradiation on the poly[Cux(Cu-ett)]:PVDF film induced enhancement of the Seebeck coefficient from 52 ± 1.5 to 79 ±5.0 μV K^–1. Benefitting from prominent PTE and PTE effects of poly[Cux(Cu-ett)]:PVDF, allows for a PTE voltage of 12 mV to be obtained. This enables promising applications of organic TE materials in electricity generation from solar energy and NIR detection.

Sensing is another interesting area of application for organic thermoelectric materials. In that regard, thermoelectric devices represent good candidates for temperature sensors. Interestingly, recent work by Zhang et al. demonstrated the microstructure-frame-supported organic thermoelectric material (MFOTE) can also be applied as pressure-temperature dualparameter sensors for self-powered e-skin applications.²⁷ In this device, the temperature difference between the device and object is detected by the Seebeck effect, while the biased pressure can be probed by the change in device resistance. The incorporation of piezoresistive and thermoelectric mechanisms enables the simultaneous detection of temperature and pressure stimuli without an additional decoupling process, and even features a self-powered pressure sensitivity and temperaturesensing accuracy of >20 kPa–1 and <0.1 K, respectively. It should be noted the flexible dual-parameter sensor can be self-powered and integrated into array, enabling their intelligent application in a wide range of robotics and health-monitoring products.

Conclusion and Perspectives

After rapid development in the past few years, the field of organic thermoelectrics has made remarkable advancements. High ZT values of >0.2 at room temperature have been successfully demonstrated using both p-type (PEDOT:PSS) and n-type (Poly[Kx(Ni-ett)]) materials. This performance is comparable to those of many inorganic counterparts at low temperatures. Moreover, several organic thermoelectric devices have been fabricated using printing methods, indicating rapid progress of this cutting-edge field.

Despite these achievements, organic thermoelectrics are still in the initial development stage and face a number of challenges. One challenge is the need to identify more organic thermoelectric materials with high ZT value. Until now, only limited organic candidates have been explored in thermoelectric studies, leaving the investigation of structure-property relationship an open question. An in-depth understanding of the critical roles of a conjugated backbone, side chains, polymer molecular weight, and energy levels will enable development of state-of-the-art organic thermoelectric materials. Many high-mobility organic semiconductors are good candidates for thermoelectric applications by principle. In general, their low carrier density limits the improvement of electrical conductivity, impeding their use in thermoelectrics. Controlling charge carrier concentration in organic semiconductors emerges as a key challenge; therefore development of high efficient doping methods will accelerate the achievement of high performance thermoelectronics from organic semiconductors.

The second challenge is lack of understanding of the fundamental mechanisms of organic thermoelectrics. The combined effect of many physical processes, including charge transport, phonon transport, and phonon scattering are involved in thermoelectric conversion and lead to a complicated operating mechanism. There are a number of important questions to be answered. For example, what kind of charges dominate the charge transport organic materials from the standpoint of charge transport? What is the best model for intrinsic charge transport in various organic semiconductors and their doped systems? How can we better characterize the important role of carriers and phonons in determining the thermal transport property of organic semiconductors? Furthermore, the S-σ trade-off relationship, σ-k relationship, and the relationship between energy level and S constitute several questions for organic thermoelectric materials, which require further investigation. Understanding these mechanisms and reconciliation of the trade-off relationships remain challenging tasks.

The development of flexible organic thermoelectric devices with high power generation density constitutes the third challenge. The output power of organic thermoelectric devices is not only governed by the properties of materials, but also strongly related to the condensed structure of the material and the properties of their interfaces with electrodes. Precisely modulated micro/nanostructure and functional interfaces deserve focused attention to maximize the output power of the device. Another opportunity relies on the construction of flexible devices for low-cost applications. To meet this requirement, flexible integrated modules consisting of a huge number of thermoelectric legs should be fabricated utilizing solution-processing techniques. However, both device geometry and integration of organic thermoelectric devices are not well developed.

Precise measurement of key thermoelectric parameters including S, σ, and k constitutes the fourth challenge in organic thermoelectrics. While characterization techniques are established for inorganic thermoelectrics, a standard protocol for organic thermoelectric materials does not exist. To ensure thermoelectric performance is accurately evaluated, the size and shape of organic materials and electrodes, as well as the properties of the substrate must be clearly reported. It should be noted the power factor is usually utilized to evaluate the performance of organic materials, since in-plane thermal conductivity cannot be obtained in a straightforward way. Although the 3ω method has been adopted for organic thermoelectric materials in the past few years, the difficulties in device fabrication and measurement setup makes wide acceptance of this method difficult.