Search

This article reports the design, fabrication, and demonstration of additively manufactured air jet impingement coolers for the thermal management of high-power gallium nitride (GaN) transistors. The polymer jet coolers impinge high-speed airflow with a velocity of 42–195 m/s (Reynolds number between $1.87 \times 10^{4}$ and $8.77 \times 10^{4})$ onto working GaN devices mounted on a printed circuit board (PCB). The air jet provides cooling heat fluxes of up to 58.4 W/cm2, cooling rates of up to 6.6 °C/s, and convective heat transfer coefficient ranging from 5.2 to 17.0 kW/( $\text{m}^{2}\cdot \text{K}$ ). The cooling performance is comparable to that of jet coolers made from other materials and manufacturing technologies. A key benefit of additive manufacturing (AM) is design freedom and geometric complexity, which we highlight by demonstrating three different packaging configurations, each enabled by a different jet cooler design that is customized for different types of packaging configurations: Cooler 1 directs two parallel impinging jets onto the top side of two devices; cooler 2 directs two air jets onto the front side and two air jets onto the back side of two devices; and cooler 3 directs air jets onto the front side of four devices mounted on parallel adjacent circuit boards. The second benefit of AM is the ability to consolidate multiple components into a single part, which we highlight by combining a nozzle, a fluidic delivery system, and a flow distributor within a volume of 80 mm $\times80$ mm $\times100$ mm. This work demonstrates the potential of AM to create complex, lightweight, fluidic delivery systems to achieve thermally and hydrodynamically optimized air jet cooling for high-power-density electronic devices.

Anisotropic polymer/nanoparticle composites display unique mechanical, thermal, electrical, and optical properties depending on confirmation and configuration control of the composing elements. Processes, such as vapor deposition, ice-templating, nanoparticle self-assembly, additive manufacturing, or layer-by-layer casting, are explored to design and control nanoparticle microstructures with desired anisotropy or isotropy. However, limited attempts are made toward nanoparticle patterning during continuous fiber spinning due to the thin-diameter cross section and 1D features. Thus, this research focuses on a new patterning technique to form ordered nanoparticle assembly in layered composite fibers. As a result, distinct layers can be retained with innovative tool design, unique material combinations, and precise rheology control during fiber spinning. The layer multiplying-enabled nanoparticle patterning is demonstrated in a few material systems, including polyvinyl alcohol (PVA)-boron nitride (BN)/PVA, polyacrylonitrile (PAN)-aluminum (Al)/PAN, and PVA-BN/graphene nanoplatelet (GNP)/PVA systems. This approach demonstrates an unprecedentedly reported fiber manufacturing platform for well-managed layer dimensions and nanoparticle manipulations with directional thermal and electrical properties that can be utilized in broad applications, including structural supports, heat exchangers, electrical conductors, sensors, actuators, and soft robotics. W.X. and R.F. contributed equally to this work. This work was funded by the Global Sports Institute (GSI) at Arizona State University and the U.S. National Science Foundation (NSF, EAGER 1902172). Scopus

Heat dissipation is a key obstacle to achieving reliable, high-power-density electronic systems. Thermal devices capable of actively managing heat transfer are desired to enable heat dissipation optimization and enhanced reliability through device isothermalization. Here, we develop a millimeter-scale liquid metal droplet thermal switch capable of controlling heat transfer spatially and temporally. We demonstrate the thermal switch by integrating it with gallium nitride (GaN) devices mounted on a printed circuit board (PCB) and measure heat transfer and temperature of each device for a variety of switch positions and heat dissipation levels. When integrated with a single GaN device (2.6 mm $\times4.6$ mm face area) dissipating 1.8 W, the thermal switch shows the ability to actively control heat transfer by conducting 1.3 W in the ON mode with the GaN device at 51 °C ± 1 °C, and 0.5 W in the OFF mode with the GaN device at 95 °C ± 1 °C. To elucidate the heat transfer physics, we developed a 1-D system thermal resistance model in conjunction with an independent 3-D finite-element method (FEM) simulation, showing excellent agreement with our experimental data. Finally, we demonstrated that when the switch is integrated with two GaN devices, the switch can balance the device heat transfer rate and enhance junction temperature uniformity and system reliability by lowering the device-to-device temperature difference from > 10 °C (no switch) to 0 °C.

We present the deep learning model for internal forced convection heat transfer problems. Conditional generative adversarial networks (cGAN) are trained to predict the solution based on a graphical input describing fluid channel geometries and initial flow conditions. Without interactively solving the physical governing equations, a trained cGAN model rapidly approximates the flow temperature, Nusselt number (Nu), and friction factor (f) of a flow in a heated channel over Reynolds number ranging from 100 to 27,750. For an effective training, we optimize the dataset size, training epoch, and a hyperparameter λ. The cGAN model exhibited an accuracy up to 97.6% when predicting the local distributions of Nu and f. We also show that the trained cGAN model can predict for unseen fluid channel geometries such as narrowed, widened, and rotated channels if the training dataset is properly augmented. A simple data augmentation technique improved the model accuracy up to 70%. This work demonstrates the potential of deep learning approach to enable cost-effective predictions for thermofluidic processes.

Stochastic foam with hierarchy order pore structure possesses distinguished physical properties such as high strength to weight ratio, super lightweight, and extremely large specific area. These exceptional properties make stochastic foam as a competitive material for versatile applications e.g., heat exchangers, battery electrodes, automotive components, magnetic shielding, catalyst devices and etc. Recently, the more advanced hollow cellular (shellular) architectures with well-developed structure connections are studied and expected to surpass the solid micro/nanolattices. However, in terms of theoretical predicting and studying of the cellular foam architecture, currently no systematic model can be utilized to accurately capture both of its mechanical and thermal properties especially with hollow struts due to complexity induced by its stochastic and highly reticulate nature. Herein, for the first time, a novel packing three-dimensional (3D) hollow dodecahedron (HPD) model is proposed to simulate the cellular architecture. An electrochemical deposition process is utilized to manufacture the metallic foam with hollow struts. Mechanical and thermal testing of the as-manufactured foams are carried out to compare with the HPD model. HPD model is proved to accurately capture both the topology and the physical properties of stochastic foam at the similar relative density. Particularly, the proposed model makes it possible to readily access and track the physical behavior of stochastic foam architecture. Accordingly, this work will also offer inspiration for designing an efficient foam for specific applications.

Thermoelectric power generation offers a promising way to recover waste heat. The geometrical design of thermoelectric legs in modules is important to ensure sustainable power generation but cannot be easily achieved by traditional fabrication processes. Herein, we propose the design of cellular thermoelectric architectures for efficient and durable power generation, realized by the extrusion-based 3D printing process of Cu2Se thermoelectric materials. We design the optimum aspect ratio of a cuboid thermoelectric leg to maximize the power output and extend this design to the mechanically stiff cellular architectures of hollow hexagonal column- and honeycomb-based thermoelectric legs. Moreover, we develop organic binder-free Cu2Se-based 3D-printing inks with desirable viscoelasticity, tailored with an additive of inorganic Se82− polyanion, fabricating the designed topologies. The computational simulation and experimental measurement demonstrate the superior power output and mechanical stiffness of the proposed cellular thermoelectric architectures to other designs, unveiling the importance of topological designs of thermoelectric legs toward higher power and longer durability. The geometrical design of thermoelectric legs in modules is key for sustainable power generation but can be hardly achieved by traditional fabrication process. Here, the authors develop an extrusion-based 3D printing process of Cu2Se thermoelectric materials for efficient power generation.

This paper investigates heat transfer enhancement by means of additively manufactured static mixers during liquid water cooling of a horizontal, heated flat plate. The static mixers disrupt the thermal boundary layer and induce mixing, resulting in an increased heat transfer rate of about 2X larger than flows without mixers. Simulations of the flows provided insights into the flows near the mixers, and guided selection of specific mixer geometries. The mixers were fabricated directly into the flow channels using additive manufacturing and then assembled onto the heated plate. Two types of mixing structures were analyzed: twisted tape structures that are similar to conventional static mixers; and novel chevron-shaped offset wing structures. Heat transfer performance was measured for liquid water (510 ≤ Re ≤ 1366) cooling the heated section with convective heat flux ranging between 0.1 and 0.8 W/cm2. This work demonstrates the potential of additive manufacturing to enable novel flow geometries that can enhance convection heat transfer whilst minimizing pressure drop penalties and volumes.

This paper reports two-phase cooling in compact cross-flow microchannel heat exchangers with high power density up to 180 W/cm3. The performance is enabled by high-speed air flow through microchannels and two-phase condensation of refrigerant R245fa. The heat exchangers were realized in 1 cm3 blocks of copper alloy, using micro-electrical-discharging machining. Two heat exchanger designs were analyzed, fabricated, and tested. The first device has 150 air-side channels of diameter 520 μm, and the second device has 300 air-side channels of diameter 355 μm. In both cases the refrigerant channels are 2.0 × 0.5 mm2. The heat exchangers were operated with Reynolds number between 7500 and 20,500 for the air flow and with mass flux between 330 and 750 kg/m2 s for the refrigerant flow. The refrigerant temperature at the channel entrance was 80 °C, which is near the maximum operating temperature for some electronic devices. For comparison purposes, the devices were also tested with single-phase refrigerant flows. This work demonstrates the potential of high power density heat exchangers that leverage advanced manufacturing technologies to fabricate miniature channels.

To attain power generation with body heat, the thermal resistance matched design of the thermoelectric generator was the principal factor which was not critical in the case of thermoelectric generator for the waste heat generation. The dimension of thermoelectric legs and the number of thermoelectric leg-pairs dependent output power performances of the thermoelectric generator on the human wrist condition was simulated using 1-dimensional approximated heat flow equations with the temperature dependent material coefficients of the constituent materials and the dimension of the substrate. With the optimum thermoelectric generator design, thermoelectric generator modules were fabricated by using newly developed fabrication processes, which is mass production possible. The electrical properties and the output power characteristics of the fabricated thermoelectric modules were characterized by using a home-made test set-up. The output voltage of the designed thermoelectric generator were a few tens of millivolts and its output power was several hundreds of microwatts under the conditions at the human wrist. The measured output voltage and power of the fabricated thermoelectric generator were slightly lower than those of the designed thermoelectric generator due to several reasons.

Thermoelectric power generation offers a promising way to recover waste heat. The geometrical design of thermoelectric legs in modules is important to ensure sustainable power generation but cannot be easily achieved by traditional fabrication processes. Herein, we propose the design of cellular thermoelectric architectures for efficient and durable power generation, realized by the extrusion-based 3D printing process of Cu2Se thermoelectric materials. We design the optimum aspect ratio of a cuboid thermoelectric leg to maximize the power output and extend this design to the mechanically stiff cellular architectures of hollow hexagonal column- and honeycomb-based thermoelectric legs. Moreover, we develop organic binder-free Cu2Se-based 3D-printing inks with desirable viscoelasticity, tailored with an additive of inorganic Se82− polyanion, fabricating the designed topologies. The computational simulation and experimental measurement demonstrate the superior power output and mechanical stiffness of the proposed cellular thermoelectric architectures to other designs, unveiling the importance of topological designs of thermoelectric legs toward higher power and longer durability., The geometrical design of thermoelectric legs in modules is key for sustainable power generation but can be hardly achieved by traditional fabrication process. Here, the authors develop an extrusion-based 3D printing process of Cu2Se thermoelectric materials for efficient power generation.

A critical challenge in using thermoelectric generators (TEGs) for charging portable or wearable electronics has been limited outputs generated by wearable TEGs, because available temperature difference between human body and environment (ΔText) is typically a few degrees of Kelvins. The power generation from wearable TEGs strongly depends on the characteristics of integrated heat sinks and airflow configurations. Thus, when designing wearable TEGs, the optimization for the internal thermal resistance is necessary considering the large external thermal resistance at TEG–air interface. This chapter presents a theoretical model for designing a TEG when the heat transfer characteristics of heat sink and airflow configuration are given. Experimental data verify the strong dependence of TEG performance on the types of heat sink and airflow configuration.

This paper presents a heuristic optimization method for cooling channels with internal repeated-rib roughness. The method rapidly explores a design space to simultaneously optimize two geometric parameters, channel length, and rib roughness ratio. For a rapid and accurate optimization, the method combines a heuristic optimization technique, simulated annealing (SA), and numerically derived closed-form models of heat transfer and pressure drop. It is shown that approximately 1 million designs are evaluated within 6 s, resulting in optimal designs having minimal thermal resistance for given pressure thresholds. Closed-form correlations for developing and fully developed flow are derived by evaluating discrete design points using a finite volume model (FVM). The derived correlations predict the channel properties with acceptable ranges of mean absolute error (

A random forest algorithm is trained to recognize different transitions to turbulence in wavy-wall tubes, and then can classify a flow within 1/100th of a second.

Thermoelectric energy conversion offers a unique solution for generating electricity from waste heat. However, despite recent improvements in the efficiency of thermoelectric materials, the widespread application of thermoelectric generators has been hampered by challenges in fabricating thermoelectric materials with appropriate dimensions to perfectly fit heat sources. Herein, we report an extrusion-based three-dimensional printing method to produce thermoelectric materials with geometries suitable for heat sources. All-inorganic viscoelastic inks were synthesized using Sb2Te3 chalcogenidometallate ions as inorganic binders for Bi2Te3-based particles. Three-dimensional printed materials with various geometries showed homogenous thermoelectric properties, and their dimensionless figure-of-merit values of 0.9 (p-type) and 0.6 (n-type) were comparable to the bulk values. Conformal cylindrical thermoelectric generators made of 3D-printed half rings mounted on an alumina pipe were studied both experimentally and computationally. Simulations show that the power output of the conformal, shape-optimized generator is higher than that of conventional planar generators. Three-dimensional printing is changing the way we manufacture objects. Here, Kim et al. develop inks of inorganic thermoelectric materials to 3D print the legs of conformable thermoelectric generators, allowing waste heat recovery from hot water pipes.

Foams offer extremely large surface area per unit mass, making them competitive material for heat exchangers and energy storage systems. Understanding the influence of foam characteristics, i.e., size, distribution and concentration of pores, ligament defects as well as foam architecture, on thermal transport is important when designing the foam-based devices. In this article, we present the effective thermal conductivity of open-cell polyurethane (PU) foam (20 PPI) with ~10 μm thick nickel coating measured by transient plane source (TPS) method. A calibration methodology for TPS method is developed to obtain accurate measurements. A finite element model and thermal resistance model are developed for the heat transfer in metal coated foams occurring near room temperature. For precisely modeling the foam architecture topology, an X-Ray tomography is employed. The developed models are used to investigate how the Ni coating thickness affects the effective thermal conductivity. Lastly, we discuss how the model assumptions are related to the discrepancy between the model predictions and measurements for the polymer-metal foams.

We report an air jet cooler made with additive manufacturing. The air jet cooler directs an impinging air jet directly onto electronic devices. The jet system was fabricated by a single manufacturing process using a resin-based three-dimensional printer, and monolithically integrates two nozzles, air delivery channel, flow distributor and mechanical fixtures within a volume of 80×80×80 mm3, To demonstrate the viability of the jet cooler, high power gallium nitride (GaN) transistors were cooled using air jets at up to 195 m/sec. With the air jet, a GaN transistor could dissipate heat flux up to 60 W/cm2, which was 7X larger than the maximum allowable heat flux under natural convection cooling. The air jet cooler is also capable of rapid switching of the cooling air. Direct impingement of air jet could reduce the GaN transistor temperature by ~70°C within 11 seconds. This work demonstrates the potential of additively manufactured air jet coolers as a compact thermal management scheme for high-power electronics. Since the geometry of the air jet coolers can be easily tailored to various shapes, the demonstrated concept can be applied for cooling a variety of electronics with different topologies and different layouts of hot spots.

Segmented thermoelectric generators (TEGs) comprising multiple TE elements can operate over a large thermal gradient without inherent conversion efficiency (ZT) losses of materials. However, despite excellent theoretical efficiencies, the performance of actual segmented TEGs are critically affected by several challenges related to material incompatibility and limited design flexibility in conventional fabrication processes. Herein, we report the multi-material 3D printing of composition-segmented BiSbTe materials by the sequential deposition of all-inorganic viscoelastic TE inks containing BixSb2-xTe3 particles, tailored with Sb2Te42− chalcogenidometallate binders. The peak ZTs of the 3D-printed materials controllably shifted from room temperature to 250 °C by composition engineering of BixSb2-xTe3 particles. We fabricated the optimally designed TEG comprising the 3D-printed, composition-segmented tri-block Bi0.55Sb1.45Te3/Bi0.5Sb1.5Te3/Bi0.35Sb1.65Te3 TE leg, which extends the peak ZTs and satisfies full compatibility across the entire temperature range, realizing a record-high efficiency of 8.7% under the temperature difference of 236 °C. Our approach offers a promising strategy to optimize segmented TEGs.

Flexible thermoelectric modules (TEMs) are often built with a polymer filler as structural support. However, the filler may serve as a thermal bypass, leading to reduced temperature difference across a TEM and degradation of the output. Herein, we present flexible TEMs made of porous PDMS fillers and Bi2Te3-based thermoelectric legs. The heat conduction through the filler is suppressed by leveraging the low thermal conductivity of porous PDMS (0.08 W/m∙K), leading to an increase in the temperature difference across the TEMs. The porous PDMS TEM exhibits 23% enhanced power density and a figure of merit (ZT) of 0.75 around the room temperature. Lastly, the efficacy of the porous PDMS TEM is demonstrated by testing the TEMs with body heat. Our approach would offer a promising insight into the flexible TEM designs.

Controlling crystalline phases in polymorphic materials is critical not only for the fundamental understanding of the physics of phase formation but also for the technological application of forbidden, but potentially useful physical properties of the nominally unstable phases. Here, using tin oxide (SnO2) as a model system, we demonstrate a new way to enhance the mechanical hardness of an oxide by stabilizing a high-pressure dense phase through nitrogen integration in the oxide. Pristine SnO2 has a tetragonal structure at the ambient pressure, and undergoes phase transitions to orthorhombic and cubic phases with increasing pressure. Leveraging the enhanced reactivity of nitrogen in plasma, we are able to synthesize tin oxynitride (SnON) thin films with a cubic phase same as the high-pressure phase of SnO2. Such nitrogen-stabilized cubic SnON films exhibit a mechanical hardness of ∼23 ± 4 GPa, significantly higher than even the nitride counterpart (Sn3N4) as the result of the shortened atomic distance of ...

As a field-assisted technique, spark plasma sintering (SPS) enables densification of specimens in a very short period of time compared to other sintering techniques. For high performance thermoelectric material synthesis, SPS is widely used to fabricate nanograin-structured thermoelectric materials by rapidly densifying the nanopowders suppressing grain growth. However, the microstructural evolution behavior of thermoelectric materials by SPS, another important process during sintering, has been rarely studied. Here, we explore SPS as a tool to control the microstructure by long-time SPS. Using p-type (Bi 0.25 Sb 0.75 ) 2 Te 3 thermoelectric materials as a model system, we systematically vary SPS temperature and time to understand the correlations between SPS conditions, microstructural evolution, and the thermoelectric properties. Our results show that the relatively low eutectic temperature (∼420 °C) and the existence of volatile tellurium (Te) are critical factors to determine both microstructure and thermoelectric property. In the liquid-phase sintering regime, rapid evaporation of Te leads to a strong dependence of thermoelectric property on SPS time. On the other hand, in the solid-phase sintering regime, there is a weak dependence on SPS time. The optimum thermoelectric figure-of-merit (Z) of 2.93 × 10 −3 /K is achieved by SPS at 500 °C for 30 min. Our results will provide an insight on the optimization of SPS conditions for materials containing volatile elements with low eutectic temperature.

We report the facile synthesis and the chemoresistive performances of villi-like WO 3 nanostructures (VLWNs) which were fabricated by RF sputter with glancing angle deposition (GAD) mode. A fast deposition of GAD effectively forms self-assembled anisotropic nanostructures with high porosity and surface-to-volume ratio. The sensing tests were examined for NO 2 detection in dry air and a mixture of reducing gases. As a result, these sensors at 200 °C exhibit an highly selective and sensitive NO 2 detection down to 800 parts per trillion (ppt) level, and could also respond well to NO 2 in the concentration range of 0.2–1 parts per million (ppm) without the interference of gas mixture. These results show that the enhanced sensing properties to NO 2 are attributed to the highly efficient surface modulation by double potential barriers at nano-necks of WO 3 nanostructures.

Unlike bulk materials, many physical properties of thin-film materials depend on film thickness due to extrinsic effects such as residual stress and the dead layer. In this work, the effect of thickness on the electrocaloric properties of Pb0.9La0.1(Zr0.65Ti0.35)O3(10/65/35) (PLZT) films grown by the sol–gel method was studied experimentally. In the sol–gel synthesis, the annealing process results in residual stress and the metal–dielectric contact generates a dead layer. These extrinsic effects influence the dielectric and ferroelectric properties of the thin films, and their roles are film thickness dependent. For PLZT of nanometer thickness (from 420 nm to 1080 nm), the permittivity and polarization, and their temperature dependences, showed strong dependence on film thickness. In particular, in the temperature range from 70°C to 350°C, the electrocaloric temperature change showed threefold improvement (from 0.2°C to 0.6°C) as the thickness was increased from 420 nm to 840 nm. This work will aid development of polycrystalline thin films including PLZT for electrocaloric applications.

This paper explores machine learning approach as a heat transfer correlation. Machine learning significantly reduces the effort to develop multi-variable heat transfer correlations, and is capable of readily expanding the parameter domain. Random forests algorithm is used to predict the convection heat transfer coefficients for a high-order nonlinear heat transfer problem, i.e., convection in a cooling channel integrated with variable rib roughness. For 243 different rib array geometries, numerical simulations are performed to train and test ML model based on six input features. Machine learning model predicts closely to numerical simulation data with high determination of coefficient (R2), e.g., R2 > 0.966 for the testing dataset. The capability and limitation of random forests algorithm are discussed with validation dataset.

This article reports a computationally efficient optimization for wavy surface roughness in cooling channels based on simulated annealing. Our approach rapidly calculates the performance metrics of cooling channels using closed-form models and approximates an optimum design via a random optimization technique, simulated annealing. The proposed method optimizes the wall roughness of cooling channels for maximizing the heat transfer rate while satisfying the pressure drop constraint. The channel wall smoothly converges and diverges throughout to locally modulate the heat transfer rate and the pressure drop. For a given pressure constraint and heat load distribution, the optimizer is able to compare more than 10,000 possible channel designs within a minute and generate a distinct channel design that achieves the optimization objective. The channel temperature predicted by the optimizer differs up to 10.9% to the estimations by a finite volume model. Lastly, optimized wavy channel designs are introduced that exhibit up to 2.7 times greater performance factor than a conventional channel geometry. This work demonstrates the potential of algorithm-based optimization techniques for designing efficient thermodynamic systems.

Composite phase change materials consisting of a high-latent-heat phase change material (PCM) embedded in a high-thermal-conductivity matrix are desirable for thermally buffering pulsed heat loads via rapid absorption and release of thermal energy at a constant temperature. This paper reports a composite PCM thermal buffer consisting of a Field's metal PCM having high volumetric latent heat (315 MJ/m3) embedded in a copper (Cu) matrix having high intrinsic thermal conductivity [384 W/(m·K)]. We demonstrate thermal buffer samples fabricated with Cu volume fractions from 0.05 to 0.2 and sample thicknesses ranging between 1 mm and 4 mm. Experiments coupled with finite element method simulations were used to determine the figures of merit (FOMs), cooling capacity ηeff, energy density Eeff, effective thermal conductivity keff, and the buffering time constant τ. The cooling capacity was measured to be as high as ηeff = 72 ± 4 kJ/(m2·K1/2·s1/2) for the 1.45 mm thick thermal buffer sample having a Cu volume fraction of 0.13, significantly higher than theoretical values for aluminum–paraffin composites [45 kJ/(m2·K1/2·s1/2)] or pure paraffin wax [8 kJ/(m2·K1/2·s1/2)]. Our work develops design guidelines for high-FOM thermal buffer devices for pulsed heat load thermal management.

High-performance computing systems, especially 3D ICs, are yet facing thermal exacerbation. Inter-tier liquid cooling microchannel layers have been introduced into 3D ICs as an integrated cooling mechanism to tackle thermal degradation. Many research works optimize microchannel designs based on runtime-expensive numerical simulations or inaccurate thermofluid models. In this work, we propose accurate closed-form models on tapered microchannel to capture the relationship between channel geometry and heat transfer performance. To improve the accuracy, our correlation is based on developing flow model and derived from numerical simulation using a subset of multiple channel parameters. Our models reduce error by 57 % in Nusselt number and 45 % in pressure drop for channels with inlet width 100-400µm compared to commonly used fully developed flow based models in optimization. Obtained correlations show potential as solid foundation to achieve close to optimal design through runtime-efficient microchannel design optimization.

Thermoelectric devices have attracted a great attention for renewable energy harvesters and solid-state coolers. For practical applications, the mechanical properties of thermoelectric materials become critical for the device reliability, a persistent performance with a long time and high operation cycles. Bi–Te based single-crystals, mostly used in commercial thermoelectric devices, are intrinsically brittle with weak van der Waals bonding, often leading to device failures such as crack and debonding during fabrication and operation. Thus, it is highly desirable to enhance the mechanical property of Bi–Te based alloys as well as the thermoelectric property. Here, we investigate the effect of B 4 C nanoparticles (less than 0.5 wt%) dispersed in p -type Bi 0.4 Sb 1.6 Te 3 matrix on the mechanical properties. X-ray diffraction (XRD) result confirms that B 4 C-dispersed Bi 0.4 Sb 1.6 Te 3 has a single phase. We observe that the grain size of Bi 0.4 Sb 1.6 Te 3 becomes decreased with the B 4 C nanoparticle concentration by electron backscatter diffraction (EBSD) technique. Hardness, Young’s modulus, and flexural strength of B 4 C-dispersed Bi 0.4 Sb 1.6 Te 3 are enhanced, compared to the B 4 C-free Bi 0.4 Sb 1.6 Te 3 polycrystals. On the other hand, the thermoelectric figure-of-merit of B 4 C-dispersed Bi 0.4 Sb 1.6 Te 3 is almost identical to that of the pure Bi 0.4 Sb 1.6 Te 3 . Such enhancements of the mechanical properties of the B 4 C-dispersed Bi 0.4 Sb 1.6 Te 3 are attributed to the grain boundary hardening and second-phase hardening. Beyond thermoelectric materials, our result implies that the grain refinement by nanoparticle dispersion is a simple and promising way to strengthen the mechanical properties of other brittle materials with layered structure.

An accurate and rapid characterization of a thermoelectric module (TEM) is critical to understand the problems in module design and fabrication. We describe an apparatus and a method that directly measure the cooling performance of a TEM such as current for maximum cooling (Imax), maximum cooling power (Qc,max), and maximum temperature difference (ΔTmax). The apparatus is designed based on a finite element model to ensure a simple heat flow measurement. We evaluate the module performance metrics based on differential measurement between the cooling powers with temperature difference across the module under transient conditions. The use of transient data reduces measurement time, and the use of a differential technique enables compensation of the thermal losses. The measured data fit well with conventional theoretical relations for the TEM performance metrics. We test a commercial TEM and validate the results using the Harman method.

In the present work, 0.01–0.05wt.% Sn-doped Bi2(Te0.9Se0.1)3 alloys were prepared by mechanical deformation followed by hot pressing, and their thermoelectric properties were studied. We observed that the Sn element is a very effective dopant as an acceptor to control the carrier concentration in the n-type Bi2(Te0.9Se0.1)3 alloys to optimize their thermoelectric property. The n-type carrier concentration can be controlled from 4.2 × 1019/cm3 to 2.4 × 1019/cm3 by 0.05wt.% Sn-doping. While the Seebeck coefficient and the electrical resistivity are both increased with doping, the power factor remains the same. Therefore, we found that the thermoelectric figure-of-merit becomes maximized at 0.75 when the thermal conductivity has a minimum value for the 0.03wt.% Sn-doped sample.

This paper reports improvements of atomic force microscopy (AFM) mass spectrometry (MS), in which ∼1 attoliter of analyte is desorbed by a heated AFM cantilever tip and analyzed with a mass spectrometer. Decoupling the AFM sampling apparatus from the MS system enabled analysis of the microscale transport physics independent of analyte ionization efficiency. Using this approach, we find that the transport efficiency is governed by the air velocity during sampling, and not mass flow rate as reported in the literature. We also find that an unheated sampling tube results in higher efficiency compared to a heated tube. Optimization of the transport parameters improved the system efficiency by 2.5-fold over the state of the art.

SnO 2 thin films were grown by atomic layer deposition (ALD) with dimethylamino-2-methyl-2-propoxy-tin(II) (Sn(dmamp) 2 ) and O 3 in a temperature range of 100–230 °C. The ALD window was found to be in the range of 100–200 °C. The growth per cycle of the films in the ALD window increased with temperature in the range from 0.018 to 0.042 nm/cycle. Above 230 °C, the self-limiting behavior which is a unique characteristic of ALD, was not observed in the growth because of the thermal decomposition of the Sn(dmamp) 2 precursor. The SnO 2 films were amorphous in the ALD window and exhibited quite a smooth surface. Sn ions in all films had a single binding state corresponding to Sn 4+ in SnO 2 . The concentration of carbon and nitrogen in the all SnO 2 films was below the detection limit of the auger electron spectroscopy technique and a very small amount of carbon, nitrogen, and hydrogen was detected by secondary ions mass spectroscopy only. The impurity contents decreased with increasing the growth temperature. This is consistent with the increase in the density of the SnO 2 films with respect to the growth temperature. The ALD process with Sn(dmamp) 2 and O 3 shows excellent conformality on a hole structure with an aspect ratio of ∼9. This demonstrates that the ALD process with Sn(dmamp) 2 and O 3 is promising for growth of robust and highly pure SnO 2 films.

The effect of Sn doping on the thermoelectric properties of p-type Bi0.4Sb1.6Te3 (BST) thin films was studied. Sn-doped BST films were grown on 4° tilted GaAs (001) substrates by metal–organic chemical vapor deposition. To control the Sn ion concentration in the films, we systematically controlled the dose of the Sn precursor by varying the H2 flow rate from 0 sccm to 100 sccm. The hole carrier concentration increased as the H2 flow rate was increased. Interestingly, the Seebeck coefficient of the films simultaneously increased with the carrier concentration when the H2 flow rate was increased up to 60 sccm. This might be attributed to the formation of virtual bound states in the valence band by Sn doping. Consequently, the Sn ion doping contributed to the thermopower enhancement of the BST films.

Accuracy of the Harman measurement largely depends on the heat transfer between the sample and its surroundings, so-called parasitic thermal effects (PTEs). Similar to the material evaluations, measuring thermoelectric modules (TEMs) is also affected by the PTEs especially when measuring under atmospheric condition. Here, we study the correction methods for the Harman measurements with systematically varied samples (both bulk materials and TEMs) at various conditions. Among several PTEs, the heat transfer via electric wires is critical. Thus, we estimate the thermal conductance of the electric wires, and correct the measured properties for a certain sample shape and measuring temperature. The PTEs are responsible for the underestimation of the TEM properties especially under atmospheric conditions (10–35%). This study will be useful to accurately characterize the thermoelectric properties of materials and modules.

The performance of infrared (IR) sensing bimaterial cantilevers depends upon the thermal, mechanical and optical properties of the cantilever materials. This paper presents bimaterial cantilevers that have a layer of black silicon nanocone arrays, which has larger optical absorbance and mechanical compliance than single crystal silicon. The black silicon consists of nanometer-scale silicon cones of height 104–336 nm, fabricated using a three-step O 2 –CHF 3 –Ar + Cl 2 plasma process. The average cantilever absorbance was 0.16 over the 3–10 μm wavelength region, measured using a Fourier transform infrared (FTIR) microspectrometer. The measured cantilever responsivity to incident IR light compares well to a model of cantilever behavior that relate the spectral absorbance, heat transfer, and thermal expansion. The model also provides further insights into the influence of the nanocone height on the absorbance and responsivity of the cantilever. Compared to a cantilever with smooth single crystal silicon, the cantilever with black silicon has about 2× increased responsivity. The nanocone array fabrication technique for silicon bimaterial cantilevers presented here could be applied to other IR sensors.

An understanding of the thermal behavior of heated microcantilevers is essential to their use. In this article, we investigated the infrared (IR) emission of two silicon cantilevers with integrated solid-state heaters over the 2500–3000 cm−1 spectral range. Two cantilevers were examined: the first had a heater surface area of 17 × 20 μm and the second had a heater surface area of 170 × 100 μm. We calculated the spectral power emitted by the cantilever based on the Planck function, dielectric function of the doped silicon at elevated temperatures, and cantilever spectral emissivity. Measurements of the cantilever spectral power compared well with predictions. The first cantilever had a radiative power of 4.2 μW at a heating power of 15.7 mW and maximum temperature of 1,150 K, and the second cantilever had a radiative power of 70.1 μW at a heating power of 54.9 mW and maximum temperature of 850 K. The cantilever emissive power depended on the spatial variation in the cantilever temperature and cantilever em...

Interfaces, such as grain boundaries in a solid material, are excellent regions to explore novel properties that emerge as the result of local symmetry-breaking. For instance, at the interface of a layered-chalcogenide material, the potential reconfiguration of the atoms at the boundaries can lead to a significant modification of the electronic properties because of their complex atomic bonding structure. Here, we report the experimental observation of an electron source at 60° twin boundaries in Bi2Te3, a representative layered-chalcogenide material. First-principles calculations reveal that the modification of the interatomic distance at the 60° twin boundary to accommodate structural misfits can alter the electronic structure of Bi2Te3. The change in the electronic structure generates occupied states within the original bandgap in a favourable condition to create carriers and enlarges the density-of-states near the conduction band minimum. The present work provides insight into the various transport behaviours of thermoelectrics and topological insulators., Grain boundaries in polycrystalline materials may offer the opportunity to explore physical phenomena that do not normally occur within the crystal grains. Here, the authors show that twin boundaries in Bi2Te3 works as an electron supply for the whole bulk material.

Although the Harman method evaluates the thermoelectric figure-of-merit in a rapid and simple fashion, the accuracy of this method is affected by several electrical and thermal extrinsic factors that have not been thoroughly investigated. Here, we study the relevant extrinsic effects and a correction scheme for them. A finite element model simulates the electrical potential and temperature fields of a sample, and enables the detailed analysis of electrical and thermal transport. The model predicts that the measurement strongly depends on the materials, sample geometries, and contact resistance of the electrodes. To verify the model, we measure the thermoelectric properties of Bi2-Te3 based alloys with systematically varied sample geometries and either with a point or a surface current source. By comparing the model and experimental data, we understand how the measurement conditions determine the extrinsic effects, and, furthermore, able to extract the intrinsic thermoelectric properties. A correction scheme is proposed to eliminate the associated extrinsic effects for an accurate evaluation. This work will help the Harman method be more consistent and accurate and contribute to the development of thermoelectric materials.

This paper investigates how silicon nitride thickness impacts the performance of silicon nitride–aluminum bimaterial cantilever infrared sensors. A model predicts cantilever behavior by considering heat transfer within and from the cantilever, cantilever optical properties, cantilever bending mechanics, and thermomechanical noise. Silicon nitride–aluminum bimaterial cantilevers of different thicknesses were designed and fabricated. Cantilever sensitivity and noise were measured when exposed to infrared laser radiation. For cantilever thickness up to 1200 nm, thicker silicon nitride results in improved signal to noise ratio due to increased absorptivity and decreased noise. The best cantilever had an incident flux sensitivity of 2.1 × 10 −3 V W −1 m 2 and an incident flux signal to noise ratio of 406 Hz 1/2 W −1 m 2 , which is more than an order of magnitude improvement compared to the best commercial cantilever.

This article reports the thermomechanical sensitivity of bimaterial cantilevers over a mid-infrared (IR) spectral range (5–10 μm) that is critical both for chemical analyses via vibrational spectroscopy and for direct thermal detection in the 300–700 K range. A physics-based model of cantilever bending was developed by including heat transfer to and within the cantilever, temperature-dependent cantilever bending, and cantilever and optical system IR characteristics. Detailed measurements of the optical system IR characteristics were used as inputs to the model, including Fourier transform infrared (FT-IR) spectral characterization of cantilever absorbance as well as characterization of the light source and monochromator. Mechanical bending sensitivity and noise were modeled and measured for six commercially available microcantilevers, which consist of either an aluminum film on a silicon cantilever or a gold film on a silicon nitride cantilever. The spectral sensitivity of each cantilever was measured by ...