Journal of Thermal Spray Technology

, Volume 25, Issue 3, pp 431–440 | Cite as

Thermoelectric Device Fabrication Using Thermal Spray and Laser Micromachining

  • Mahder Tewolde
  • Gaosheng Fu
  • David J. Hwang
  • Lei Zuo
  • Sanjay Sampath
  • Jon P. Longtin
Peer Reviewed


Thermoelectric generators (TEGs) are solid-state devices that convert heat directly into electricity. They are used in many engineering applications such as vehicle and industrial waste-heat recovery systems to provide electrical power, improve operating efficiency and reduce costs. State-of-art TEG manufacturing is based on prefabricated materials and a labor-intensive process involving soldering, epoxy bonding, and mechanical clamping for assembly. This reduces their durability and raises costs. Additive manufacturing technologies, such as thermal spray, present opportunities to overcome these challenges. In this work, TEGs have been fabricated for the first time using thermal spray technology and laser micromachining. The TEGs are fabricated directly onto engineering component surfaces. First, current fabrication techniques of TEGs are presented. Next, the steps required to fabricate a thermal spray-based TEG module, including the formation of the metallic interconnect layers and the thermoelectric legs are presented. A technique for bridging the air gap between two adjacent thermoelectric elements for the top layer using a sacrificial filler material is also demonstrated. A flat 50.8 mm × 50.8 mm TEG module is fabricated using this method and its performance is experimentally characterized and found to be in agreement with expected values of open-circuit voltage based on the materials used.


additive manufacturing (AM) laser micromachining thermal spray thermoelectric generators (TEGs) thermoelectric power generation waste-heat energy harvesting 


Thermoelectric devices offer unique power generation solutions for waste-heat energy harvesting. They are small, quiet, and have no moving parts, which makes them well suited for a variety of engineering applications, including electricity production from vehicle waste-heat to improve fuel economy in vehicles (Ref 1-3), and industrial waste-heat recovery to power wireless sensors and networks to communicate with and collect data from engineering systems (Ref 4-6). Commercially available TEGs, however, are flat, inflexible and are available in only limited range of sizes. Efforts to incorporate them onto components with curved surfaces such as exhaust pipes, pump housings, steam lines, mixing containers, reaction chambers, etc., require custom-built heat exchangers. This adds cost and is labor-intensive, in addition to presenting challenges in terms of space, thermal coupling, added weight, and long-term reliability. For example, recent attempts to incorporate TEGs into vehicles required extensive modification of the exhaust system, increasing total cost and complexity (Ref 7-9).

A very different approach is needed to fabricate the thermoelectric device directly onto the waste-heat component surfaces. In this article, the goal is to develop a fabrication process whereby TEGs are fabricated using thermal spray and laser micromachining. Thermal spray is an additive manufacturing technique used to fabricate a wide range of materials (metals, complex multicomponent alloys, ceramics, etc.) for different engineering components (Ref 10, 11). Interest has also developed in recent years in using thermal spray for non-traditional applications, such as sensors, electronic components, antennas, etc. (Ref 12-14). Adding to this body of work, this paper presents an approach to fabricating thermoelectric devices for electricity production using thermal spray and laser micromachining, with no adhesive or mechanical bonding required. The complete thermoelectric device is fabricated using a series of thermal spray layers applied sequentially, including the dielectric layers, metallic interconnects, and both p-type and n-type thermoelectric legs. Laser micromachining, as a rapid and scalable solution for material processing, is used to pattern the electrically isolated features on the thermally sprayed coatings. In this way, the traditional benefits of thermal spray technology, including rapid material synthesis and transfer, processing under ambient conditions, good coating adhesion, and durability, are made available to thermoelectric device fabrication (Ref 15). Both thermal spraying and laser micromachining are also well suited for conformal (non-flat) surfaces such as round exhaust pipes.

Thermoelectric Materials

Thermoelectric materials make use of the Seebeck effect to produce a voltage in response to a temperature gradient across the material. They are characterized as n-type (free electrons as major carriers) and p-type (free holes as major carriers), depending on whether their Seebeck coefficient is negative or positive, respectively. Heat flow across the material produces electrical current which is used to drive electrical loads (Ref 16).

The performance of a TE material is quantified by its dimensionless thermoelectric figure of merit, \(\bar{Z}\bar{T}\) (Ref 16):
$$\bar{Z}\bar{T} = \frac{{S^{2} T\upsigma }}{k},$$
where S is the Seebeck coefficient (V/K), σ is the electrical conductivity (Ω−1/m), k is the thermal conductivity (W/m K), and \(\bar{T}\) is the average temperature (K) of the hot and cold side. The overall heat-to-electricity conversion efficiency η is given as
$$\upeta = \upeta_{\text{CR}} \frac{{\sqrt {1 + \bar{Z}\bar{T} } - 1}}{{\sqrt {1 + \bar{Z}\bar{T} } - T_{\text{C}} /T_{\text{H}} }},$$
where TH and TC are the device hot- and cold-side temperatures, respectively, and ηCR = 1 − TC/TH is the limiting Carnot efficiency (Ref 17). Most thermoelectric materials research focuses on achieving a value of \(\bar{Z}\bar{T} \ge 1,\) which is considered a benchmark for economically competitive applications. At low temperatures (<450 K), TEG efficiency is low (~6% even at high ZT). However, at medium (450-850 K) and high (>850 K) temperatures, the conversion efficiency increases, which makes their use for power generation more attractive. This temperature range includes a wide range of industrial waste-heat sources. For example, vehicle exhaust pipes and diesel generator exhaust pipes nominally experience hot-side temperatures of 770-870 K (Ref 2) and 750-780 K (Ref 18), respectively.

Thermoelectric Power Generators

Commercial TEGs are typically available as flat, rigid modules. They consist of many n- and p-type thermoelectric junctions configured electrically in series and thermally in parallel as shown in Fig. 1. Some flexible thin-film thermoelectric modules are also available (Ref 19-21), but are only suited for low power (~mW) applications.
Fig. 1

Schematic of a thermoelectric module

The current manufacturing process for thermoelectric devices varies based on the type of thermoelectric material employed (Ref 22). A large share of the manufacturing cost is in material preparation and assembly. Thermoelectric material synthesis begins with melting, high-energy milling, or mechanical alloying stoichiometric ratios of the constituent elements until a specific composition is achieved. Once the TE material is synthesized in powder or particle form, it is consolidated into ingots, either through hot pressing, hot isostatic pressing, or one of the several sintering techniques e.g., spark plasma sintering. Thermoelectric properties may be further optimized with post-sintering heat treatment of the materials. The ingots are then diced, polished, and cut to form the TE legs. The legs and metallic interconnects are assembled into a module using soldering, epoxy bonding, or mechanical pressing (Ref 23, 24).

Thermal Spray

The thermal spray processes consist of injecting 10-100 µm feedstock particles into a high-enthalpy jet that melts and accelerates the particles toward the surface to be sprayed. Splats of near-circular lamellae 10-100 µm in diameter and a few microns thick flatten and solidify rapidly upon impact with the substrate to form a dense coating (Ref 15). Thermal spray finds application in numerous industrial fields including aeronautical and terrestrial turbines (thermal barrier coatings, abradable seals, etc.), the biomedical industry (bio-integrable coatings onto orthoprostheses), and the paper industry (abrasion-, wear-, and corrosion-resistant coatings), among others (Ref 11). The thermal spray process affords versatility in the type of materials that can be deposited, requires no firing or curing of the deposited material, and can be used to fabricate multi-layer structures from different materials. In addition, the deposition of the key active layers can be optimized in such a way that they are nearly fully dense and do not require post-spray consolidation or post-heat treatment. Furthermore, the rapid cooling rates (~106 K/s) that the material experiences upon impact has been shown to result in improved thermoelectric performance compared to bulk materials, which has been demonstrated with melting spinning (Ref 25, 26) for example.

Recent studies have explored the use of thermal spray as an alternative means of fabricating high-quality thermoelectric materials including Mg2Si (Ref 27, 28), FeSi2 (Ref 29), and Si-Ge (Ref 30), and the optimal spray parameters for such materials were investigated. Although these studies were not able to achieve \(\bar{Z}\bar{T}\) values large enough for practical devices, the field remains largely unexplored. Research in this area is ongoing with the goal of further improvement in \(\bar{Z}\bar{T}\) values.

In this work, thermal spray technology will be used to fabricate the dielectric layers, conducting interconnects, and active materials of the TEG directly onto the component. Because thermal spray can be used on conformal (non-flat) geometries, the flat module design is no longer a restrictive requirement.

Laser Micromachining

Laser micromachining is a non-contact, single-step patterning process. It has several advantages over traditional micro-fabrication techniques such as chemical etching or mechanical milling (Ref 31, 32). Conventional material removal methods, such as machining and grinding, can be difficult to scale up in a cost-effective manner. Laser micromachining is also well suited for conformal geometries as well as for micromachining multi-layer coatings. The primary role of the laser in this application is for precision material removal to pattern electrically isolated pads on which the active TE legs will be placed. Additionally, laser micromachining will be used for cleaning material overspray in the isolated grooves.

Materials and Methods

Equipment Setup

All thermal spray work was done on-site at the Center for Thermal Spray Research (CSTR) at Stony Brook University. Atmospheric plasma spray (APS) was chosen for this work as it was found to provide the best feature quality. The powders used had a minimum of >99% purity and particle size of <60 μm. An F4-MB (Sulzer Metco, Switzerland) and SG100 plasma torch (Praxair Surface Technologies, USA) were used for spraying.

Referring to Fig. 2, all laser micromachining work was performed with a 527 nm arc-lamp pumped GM-30 Nd:YLF pulsed laser (Photonics Industries, USA). This laser was found to be most suitable for the micromachining conditions required. The laser output pulse has an average pulse duration of 200 ns, a maximum energy of 25 mJ, maximum power of 30 W, and a variable repetition rate up to 4 kHz. An external signal from the motion controller is used to control the laser output and synchronize beam delivery with sample position. The beam diameter was set using plano-convex lenses of 50, 100, and 150 mm focal length.
Fig. 2

Laser machining setup

Linear motion control for the sample is achieved using three linear translation stages (Coherent LabMotion, USA) mounted on a rigid post. Each stage has a resolution of 0.5 µm and a range of 50 mm. The stages are oriented mutually perpendicular to one other along x, y, and z axes. The motion control system can be programmed to control the laser beam delivery and execute a pre-programmed motion profile.

Material Selection

The ZT of a material depends on the crystal structure, phase, microstructure and, for semiconductors, the doping level. Bismuth telluride, which is the most commonly used TE material, is not suited for thermal spray because the APS process results in the oxidation of the powders used, significantly changing its phase and degrading its thermoelectric properties. The cost of bismuth telluride is also prohibitive for testing and presents health and environmental hazards. Other powders considered suitable for thermal spraying, e.g., Mg2Si, require extensive testing to optimize spraying parameters and may require post-heat treatment (Ref 27, 28).

Because of these factors, metal alloy powders were chosen to demonstrate the feasibility of this manufacturing method in this work. Although the resulting \(\bar{Z}\bar{T}\) is lower than that for other TE materials, metal alloys are inexpensive, straightforward to spray, and have been studied and characterized extensively. The powders consisted of 38% Ni and 62% Cu for the n-type material and 80% Ni and 20% Cr for the p-type material supplied from Sulzer Metco (Westbury, NY, USA). These alloys were chosen because they were readily available and are also similar to those used in a commercial E-type thermocouple (chromel-constantan), which has one of the largest Seebeck coefficients (~68 µV/°C), thus maximizing power output (Ref 33).

Device Design and Fabrication Process

One of the main requirements in the design of a thermoelectric module is to determine the optimum module geometry, based upon available thermoelectric material and manufacturing methods (Ref 34). The design also must consider choice of substrate and conducting material, which must conduct electricity and heat with as little impedance as possible. As such, thin layers for the ceramic and metal interconnects and thick layers for the active TE materials are preferred. The conducting layers will have a large area and much smaller thickness than the TE legs, making the requirement for the bulk electrical conductivity less severe in the choice of conductor material. Most metals or alloy with high electrical conductance, e.g., Cu, Ag, etc., meet this requirement.

From an adhesion perspective, thermal gradients, thermal cycling, and vibration can impose significant stress that can cause delamination between the different layers. Hence, adhesion is important for long-term reliable TEG performance. The ability to withstand thermal shock and temperature cycling is associated with variation in the coefficient of thermal expansion (CTE) of the different materials in the module. CTE mismatch can lead to residual stresses that develop at the joints, reducing interfacial bonding (Ref 35). The diffusion layer between two adjacent material layers should thus be at least a few microns thick to ensure good contact and thermal stability when exposed to hot-side temperatures over time (Ref 2). Good contact adhesion also improves heat transfer during laser micromachining. Poor adhesion between the conducting copper layer and the substrate can cause film peeling and delamination upon heating.

The thermally sprayed materials used for the module including their properties are shown in Table 1. The insulating and conducting layers are made from yttria-stabilized zirconia (YSZ) and copper, respectively. Both materials are straightforward to spray and have been characterized extensively.
Table 1

Properties of Materials Used (Ref 12, 37-42)

Material property



Ni80Cr20 (p-type)

Cu62Ni38 (n-type)

Electrical conductivity, Ω−1/m


9.8 × 106

2.5 × 105

6.0 × 105

Thermal conductivity, W/m K





Seebeck coefficient, V/K



CTE (× 10−6 K−1)





The open-circuit voltage, VOC, generated by a TEG is given by
$$V_{\text{OC}} = NS(T_{\text{H}} - T_{\text{C}} ),$$
where N is the number of thermoelectric junctions and S is the Seebeck coefficient of the TE junction. It can be shown that the optimal leg thickness is greater than 1 mm (Ref 34). This thickness target is well suited for thermal spraying, where coatings of thickness up to 2 mm are routinely sprayed. The number of junctions used in this work was N = 40.

Overall Device Fabrication Steps

The fabrication of the device involves three steps: (i) deposition and patterning of the bottom layers, (ii) deposition of both p- and n-type thermoelectric legs, and (iii) deposition and patterning of the top conducting layer. The steps are shown schematically in Fig. 3. In steps b and d, blanket layers of insulating and conducting materials are sprayed, after which laser machining is performed. The n- and p-materials are deposited sequentially using masks in step c.
Fig. 3

Process steps to fabricate a TEG module (a) metal substrate, (b) apply bottom electrode and pattern, (c) deposit n- and p-type TE legs, (d) deposit top electrode layer

Fabrication of Bottom Layer

The laser machining setup shown in Fig. 2 is used to test different machining parameters on the YSZ-copper layer sprayed on the bottom substrate. The substrate is mounted on a motion stage and patterned. The laser removes a thin strip of the copper layer to form electrically isolated regions. The position of the sample with respect to the beam spot and rate of ablation is controlled by adjusting the stage position and speed. The depth of the grooves is determined by the sample translation speed, laser power density, and the number of passes. All of the copper between the adjacent regions must be removed in order to avoid shorting. Four different assist gases—helium, nitrogen, argon, and air were used to remove debris and improve the laser-machined feature quality. The gas pressure is regulated at 2.8 bar (40 psi) through a 0.76-mm-(0.030 in) diameter nozzle directed at the beam spot during processing. At such low delivery pressures, the gas has no detrimental effects on the ablation dynamics or feature structure. The optimal laser parameters that maximize material removal are empirically obtained using a trial-and-error approach.

Deposition of p-type and n-type Thermoelectric Legs

The next step is to spray the p-type and n-type TE legs on the target zone patterned on the bottom layer as shown in Fig. 3(c). Various masking techniques have been developed to protect component areas next to the target zone from impact by overspray particles, including metal shadow or contact masks, high-temperature tapes, and paint-on masking compounds (Ref 15). Metal masks can be placed on top of the substrate in direct contact with it or the mask can be offset from the substrate. The use of a mask works equally well for flat and curved surfaces. A slight outward taper was introduced in the mask cutouts to provide a narrow-feathered edge rather than a sharply defined edge in the final deposited TE legs. This reduced the stresses at the edge of the coating, which minimizes opportunities for delamination. The order of the TE material sprayed is not important, however, the mask for the second spray step needs to be modified so that it can accommodate the previously sprayed legs. Any excess TE material that is sprayed into the patterned grooves is removed with the laser as described above.

Fabrication of Top Layer

To complete the circuit, the top layer must be sprayed and laser processed to form the TE couples connected in series as shown in Fig. 3(d). This top layer provides metallization for the two adjacent p- and n-type legs. The top conducting layer is fabricated by spraying a blanket layer of copper over the TE element legs. Depositing this layer is challenging because it must bridge the air gap between two adjacent TE pillars. One option is to use a permanent filler material inserted into the gaps to act as a bridge when the top conducting layer is sprayed. However, this would reduce the performance of the TE device, as the filler material would provide a path for heat conduction and reduce the device efficiency. Since air is an excellent thermal insulator compared to a filler material, an alternative is to use a temporary filler material that can be removed after the top conducting layer has been deposited. The filler material is preferred to be in paste form with a low melting point and be chemically inert. The process is shown in Fig. 4.
Fig. 4

Gap-bridging technique to avoid electrical shorting on the top conducting layer, (left) Air gap filled with filler material, (center) top conducting layer sprayed, (right) filler removed leaving air gap

After the top copper layer has been deposited and processed, a heat sink can be integrated on the top layer. A dielectric is required in between the heat sink and the top layer to prevent shorting the TE legs. This layer can be optionally thermally sprayed as well.

Results and Discussion

Fabrication of Bottom Layer

Flat coupons of aluminum 50.8 mm × 50.8 mm × 3.2 mm thick were prepared for spraying by grit blasting the surfaces. The substrates were then sprayed with a ~0.3 mm of YSZ followed by ~0.5 mm layer of copper. The copper-YSZ layer was rough, exhibited good adhesion, and was uniformly spread across the substrate, as verified by thickness measurements made at several points across the sample. The top copper layer roughness is also advantageous for laser processing and subsequent spraying of the TE legs as no additional grit blasting is needed. The top and an SEM cross section of the spayed layers are shown in Fig. 5. As shown in the right image, the inter-diffusion length between the substrate and the insulator is ~10 μm, while that between the insulator and the copper is ~50 μm.
Fig. 5

(a) Results from APS spraying of bottom electrode layer on flat substrate and (b) SEM Images of the side-view

The sprayed copper layer is laser patterned to finalize the bottom electrode. The optimal processing parameters used are listed in the Table 2. Details of the selection of these parameters are described elsewhere (Ref 36). The grooves are ~0.7 mm wide and made with multiple passes of the optimized laser parameters. The result after micromachining the top copper layer is shown in Fig. 6.
Table 2

Optimal laser cutting parameters

Focal length

50 mm

Laser power

13 W

Scan speed

8 mm/s

Scan spacing

65 μm

Spot diameter

~50 μm

Scan length

50 μm

Number of passes


Assist gas

Compressed air

Fig. 6

Patterned bottom copper electrode on flat substrate

Deposition of p-Type and n-Type Thermoelectric Materials

The mask are manufactured from 3.2 mm thick Aluminum using a DSLS 3000 CNC milling machine (Taig Tools, Chandler, AZ, USA). Taper angles from 1° to 10° in 1° increments were explored for the holes on the masks. A 4° taper angle was found to provide high-aspect ratio features while still allowing easy removal of the masks. The average TE leg thickness after spraying was 1.6 mm for the p-type legs and 1.4 mm for the n-type legs, with height being limited by the mask thickness. The masks used with the corresponding sprayed TE legs sprayed are shown in Fig. 7.
Fig. 7

(a) p-type mask (b) thermal spray of p-type legs (c) n-type mask and (d) thermal spray of legs

Some TE material overspray was deposited into the grooves that electrically separate each TE junction. The laser was used to remove this material to avoid electrical shorting. A long depth-of-focus lens was used and the laser power was set to just below the ablation threshold to heat the debris until it had sufficient energy to detach from the surface (Ref 36). The cleaning operation was repeated until all of the over-sprayed material was removed. Electrical isolation between the processed regions was verified by measuring the resistance with a digital multimeter and confirming the meter reads an open circuit (>106 Ω).

Fabrication of Top Conducting Layer

After spraying, the top surface of the TE legs was polished to achieve a uniform thickness of 1.4 mm. For the temporary filler, an 80-20 wt.% mixture of fine sand and polyvinyl alcohol (white glue) was found to work well. The sand provides support and high-temperature tolerance for the thermal spray process, while the PVA acts as a binder that can be removed by moderate heating. Just enough glue is used to hold the sand into place when dry. Referring to Fig. 8, the mixture is squeegeed into the grooves and allowed to dry for 24 h. After drying, a ~0.5-mm blanket layer of copper is sprayed on top.
Fig. 8

(a) Temporary filler material applied to the legs and (b) blanket coating of copper sprayed and (c) final device after micromachining top layer

The sample is then placed in an oven at 350 °C for 15 min to burn out the glue, after which the sand is removed by gently tapping the sample on its side. The final step is to laser process the top coating to electrically separate the TE junctions and complete the device. The optimal parameters for patterning the top coating are listed in Table 3. The right image in Fig. 8 shows the final TEG with isolated electrical regions on the top copper layer.
Table 3

Laser cutting parameters for top coating

Focal length

50 mm

Laser power

8 W

Scan speed

6.5 mm/s

Scan spacing

65 μm

Spot diameter

~50 μm

Number of passes


Total width

0.7 mm

Assist gas used


Testing and Characterization

The measured internal resistance of the device was 90.1 Ω. Only 26 of the 40 TE couples were functioning due to delamination from the surface of the bottom conductor during the fabrication process. This delamination from the bottom substrate may have occurred because of cracks that developed in the ceramic layer due to thermally induced stresses during the spraying of the TE legs. Close inspection of the samples after spraying revealed that cracks had developed on the interface between the insulating layer and the bottom metal substrate, which resulted in the separation of several TE junctions from the metal substrate. Ceramic paste (Aremco Products, Cottage, NY, USA) was used to re-attach these junctions to the substrate.

The completed thermoelectric device was tested for its performance on a TE test rig. The heater section of the test rig consists of a copper block with four embedded 100 W cartridge heaters. The TE module under test sits on top of a temperature-controlled copper block as shown in Fig. 9. The temperature of the copper block is controlled using a thermocouple inserted near the top of the block for feedback and a fan is used to cool the top of the TEG. The entire assembly is housed in a high-temperature box made from 2.5 mm thick ceramic board insulation (Cotronics Corp.). The probes are connected to an electronic load (BK Precision 8500) to measure the output voltage when a temperature difference is applied across the TEG device.
Fig. 9

(a) Test stand to characterize TEG and (b) measurement of top-side temperature with an infrared camera

The top-side temperature is measured using both a thermocouple and an infrared camera. In order to determine the temperature difference across the TE materials, a thermal resistance model is used. Referring to thermal resistance network in Fig. 10, the temperature difference across the TE element (ΔTTE) can be calculated with Eq 4 from the measured values TH and TC.
$$\Delta T_{\text{TE}} = \frac{{0.5R_{\text{TE}} }}{{2R_{\text{Cu}} + R_{\text{YSZ}} + R_{\text{Al}} + 0.5R_{\text{TE}} }}(T_{\text{H}} - T_{\text{C}} ).$$
Fig. 10

Thermal resistor network model of TE pair

The contact resistances across the interfaces of the different materials are neglected. The thermal resistances of each element are calculated based on the material properties listed in Table 4. Here, RTE, RCu, RYSZ, and RAl represent the thermal resistances of the thermoelectric materials (based on n-type), copper layer, YSZ insulating layer, and the aluminum substrate, respectively. It can be seen that the thermal resistance of the YSZ layer is the largest resistance due to its low thermal conductivity.
Table 4

Thermal resistances of various elements across the TEG






L, mm





k, W/m K





R, K/W

1.80 × 10−5

2.50 × 10−4

1.0 × 10−5

2.8 × 10−4

Figure 11 shows the measured open-circuit voltage as a function of temperature difference ΔT applied across the TE device (data points). The line represents the estimated open voltage using Eq 3 and 4 and the data in Table 4. The measured output shows the same trend as the prediction, although the actual values are 10-40% lower than the predicted values. The reason for the discrepancy may include the fact that the contact resistance between the elements was not included in the resistance network shown in Fig. 10, which would tend to reduce the overall heat flow through the device for a given applied temperature difference.
Fig. 11

Measured open-circuit voltage as function of temperature difference across the TE legs (ΔTTE)

Conclusions and Future Work

This work presents the feasibility of using thermal spray and laser micromachining to fabricate thermoelectric generators (TEGs) directly onto engineering components. Target applications include automotive exhaust systems and high-volume waste-heat sources. The steps required to fabricate a thermoelectric device are presented, including the formation of the bottom and top metallic layers and the thermoelectric legs using thermal spray and laser micromachining. A technique for bridging the air gap between adjacent thermoelectric elements for the top layer based on a sacrificial filler material was also demonstrated. The results from the characterization of a flat 50.8 mm × 50.8 mm TEG module fabricated with this method are presented. The output voltage was found to be within the estimated values based on the material properties used in the device. Opportunities for future work include improving the thermoelectric performance of thermally sprayed materials and exploring the use of additional thermoelectric materials that are well suited for thermal spray.



The authors gratefully acknowledge support for this work from the New York State Energy Research and Development Authority (NYSERDA) under Agreement # 25222 and the National Science Foundation under Grant CBET #1048744.


  1. 1.
    J. Yang and T. Caillat, Thermoelectric Materials for Space and Automotive Power Generation, MRS Bull., 2006, 31(03), p 224-229CrossRefGoogle Scholar
  2. 2.
    J. Yang and F. Stabler, Automotive Applications of Thermoelectric Materials, J. Electron. Mater., 2009, 38(7), p 1245-1251CrossRefGoogle Scholar
  3. 3.
    M.A. Karri, E.F. Thacher, and B.T. Helenbrook, Exhaust Energy Conversion by Thermoelectric Generator: Two Case Studies, Energy Convers. Manag., 2011, 52(3), p 1596-1611CrossRefGoogle Scholar
  4. 4.
    W.S. Wang, W. Magnin, N. Wang, M. Hayes, B. O’Flynn, and C. O’Mathuna, Bulk Material Based Thermoelectric Energy Harvesting for Wireless Sensor Applications, J. Phys: Conf. Ser., 2011, 307(1), p 012030Google Scholar
  5. 5.
    J. Dilhac, R. Monthéard, M. Bafleur, V. Boitier, P. Durand-Estèbe, and P. Tounsi, Implementation of Thermoelectric Generators in Airliners for Powering Battery-Free Wireless Sensor Networks, J. Electron. Mater., 2014, 43(6), p 2444-2451CrossRefGoogle Scholar
  6. 6.
    M. Tewolde, C.C. Lin, H. Tao, H. Chen, G. Fu, D. Liu, T. Zhang, D. Benjamin, L. Zuo, D. Hwang, and J. Longtin, Sensors for Small Modular Reactors Powered by Thermoelectric Generators, ASME 2014 Small Modular Reactors Symposium, American Society of Mechanical Engineers, 2014Google Scholar
  7. 7.
    A. Schmitz, C. Stiewe, and E. Muller, Preparation of Ring-Shaped Thermoelectric Legs from PbTe Powders for Tubular Thermoelectric Modules, J. Electron. Mater., 2013, 42(7), p 1702-1706CrossRefGoogle Scholar
  8. 8.
    S. Kumar, S.D. Heister, X.F. Xu, J.R. Salvador, and G.P. Meisner, Thermoelectric Generators for Automotive Waste Heat Recovery Systems Part II: Parametric Evaluation and Topological Studies, J. Electron. Mater., 2013, 42(6), p 944-955CrossRefGoogle Scholar
  9. 9.
    D. Crane, J. LaGrandeur, V. Jovovic, M. Ranalli, M. Adldinger, E. Poliquin, J. Dean, D. Kossakovski, B. Mazar, and C. Maranville, TEG On-Vehicle Performance and Model Validation and What It Means for Further TEG Development, J. Electron. Mater., 2013, 42(7), p 1582-1591CrossRefGoogle Scholar
  10. 10.
    L. Pawlowski, The Science and Engineering of Thermal Spray Coatings, Wiley, Chichester, 2008CrossRefGoogle Scholar
  11. 11.
    H. Herman, S. Sampath, and R. McCune, Thermal Spray: Current Status and Future Trends, MRS Bull., 2000, 25(07), p 17-25CrossRefGoogle Scholar
  12. 12.
    S. Sampath, Thermal Spray Applications in Electronics and Sensors: Past, Present, and Future, Journal of Thermal Spray Technology, 2010, 19(5), p 921-949CrossRefGoogle Scholar
  13. 13.
    M. Gardon, O. Monereo, S. Dosta, G. Vescio, A. Cirera, and J.M. Guilemany, New Procedures for Building-Up the Active Layer of Gas Sensors on Flexible Polymers, Surface & Coatings Technology, 2013, 235, p 848-852Google Scholar
  14. 14.
    H. Ronkainen, U. Kanerva, T. Varis, K. Ruusuvuori, E. Turunen, J. Perantie, J. Putaala, J. Juuti, and H. Jantunen, Materials for Electronics by Thermal Spraying, Physical and Numerical Simulation of Materials Processing Vii, L.P. Karjalainen, D.A. Porter and S.A. Jarvenpaa, Eds., 2013, p 451-456Google Scholar
  15. 15.
    J.R. Davis, Handbook of Thermal Spray Technology, ASM International, Materials Park, 2004Google Scholar
  16. 16.
    D.M. Rowe, Thermoelectrics Handbook: Macro to Nano, CRC Press, Boca Raton, 2005CrossRefGoogle Scholar
  17. 17.
    B. Sherman, R.R. Heikes, and J.R.W. Ure, Calculation of Efficiency of Thermoelectric Devices, J. Appl. Phys., 1960, 31(1), p 1-16CrossRefGoogle Scholar
  18. 18.
    T. Hendricks, N. Karri, T. Hogan, and C. Cauchy, New Perspectives in Thermoelectric Energy Recovery System Design Optimization, J. Electron. Mater., 2013, 42(7), p 1725-1736 (in English)CrossRefGoogle Scholar
  19. 19.
    K. Kato, Y. Hatasako, M. Kashiwagi, H. Hagino, C. Adachi, and K. Miyazaki, Fabrication of a Flexible Bismuth Telluride Power Generation Module Using Microporous Polyimide Films as Substrates, J. Electron. Mater., 2014, 43(6), p 1733-1739CrossRefGoogle Scholar
  20. 20.
    L. Bell, Flexible thermoelectric circuit, Google Patents, 2003Google Scholar
  21. 21.
    R. Venkatasubramanian, E. Siivola, T. Colpitts, and B. O’Quinn, Thin-Film Thermoelectric Devices with High Room-Temperature Figures of Merit, Nature, 2001, 413(6856), p 597-602CrossRefGoogle Scholar
  22. 22.
    S. LeBlanc, Thermoelectric Generators: Linking Material Properties and Systems Engineering for Waste Heat Recovery Applications, Sustainable Materials and Technologies, 2014, 1-2, p 26-35CrossRefGoogle Scholar
  23. 23.
    M. Yahatz and J. Harper, Fabrication of Thermoelectric Modules and Solder for Such Fabrication, Google Patents, 1998Google Scholar
  24. 24.
    R. Joachim and W. Heinz, Thermoelectric Couple with Soft Solder Electrically Connecting Semi-conductors and Method of Making Same, Google Patents, 1969Google Scholar
  25. 25.
    A.G. Gillen and B. Cantor, Photocalorimetric Cooling Rate Measurements on a Ni-5 wt% A1 Alloy Rapidly Solidified by Melt Spinning, Acta Metall., 1985, 33(10), p 1813-1825CrossRefGoogle Scholar
  26. 26.
    Q. Li, Z. Lin, and J. Zhou, Thermoelectric Materials with Potential High Power Factors for Electricity Generation, J. Electron. Mater., 2009, 38(7), p 1268-1272CrossRefGoogle Scholar
  27. 27.
    G. Fu, L. Zuo, J. Longtin, C. Nie, and R. Gambino, Thermoelectric Properties of Magnesium Silicide Fabricated Using Vacuum Plasma Thermal Spray, J. Appl. Phys., 2013, 114(14), p 6CrossRefGoogle Scholar
  28. 28.
    G. Fu, L. Zuo, J. Longtin, C. Nie, Y. Chen, M. Tewolde, and S. Sampath, Thermoelectric Properties of Magnesium Silicide Deposited by Use of an Atmospheric Plasma Thermal Spray, J. Electron. Mater., 2014, 43(7), p 2723-2730CrossRefGoogle Scholar
  29. 29.
    M. Fukumoto, M. Itoh, Y. Tanaka, H. Yakabe, and K. Kikuchi, Preparation of Beta-Fesi2 Thermoelectric Coatings by Plasma Spraying of Mechanically-Alloyed Powders, J. Jpn. Inst. Metals, 1998, 62(5), p 449-456 (in Japanese)Google Scholar
  30. 30.
    Y. Tanaka, Y. Tokimoto, and M. Fukumoto, Fabrication and Improvement of Plasma Sprayed Si-Ge Thermoelectric Coating, J. Jpn. Inst. Metals, 1999, 63(8), p 1029-1035 (in Japanese)Google Scholar
  31. 31.
    Narendra B. Dahotre and S.P. Harimkar, Laser Fabrication and Machining of Materials, Springer, Berlin, 2008Google Scholar
  32. 32.
    D. Bäuerle, Laser Processing and Chemistry, Springer, Berlin, 2011CrossRefGoogle Scholar
  33. 33.
    R.S. Figliola and D.E. Beasley, Theory and Design for Mechanical Measurements, 5th ed., Wiley, Hoboken, 2015Google Scholar
  34. 34.
    D.M. Rowe and G. Min, Design Theory of Thermoelectric Modules for Electrical Power Generation, IEE ProceedingsScience, Measurement and Technology, 1996, p. 351-356Google Scholar
  35. 35.
    T. Clyne and S. Gill, Residual Stresses in Thermal Spray Coatings and Their Effect on Interfacial Adhesion: A Review of Recent Work, J. Therm. Spray Technol., 1996, 5(4), p 401-418CrossRefGoogle Scholar
  36. 36.
    M. Tewolde, D. Liu, D. Hwang, and J. Longtin, Laser Processing of Thermal Sprayed Coatings for Thermoelectric Generators, ASME 2013 Heat Transfer Summer Conference, ASME, 2013Google Scholar
  37. 37.
    J.R. Davis, ASM Specialty Handbook—Copper and Copper Alloys, ASM International, Materials Park, 2001Google Scholar
  38. 38.
    Y. Itoh, S. Suyama, and H. Fukanuma, Thermal and Electrical Properties of Copper Coatings Produced by Cold Spraying, J Soc Mater Sci Jpn, 2010, 59(2), p 143-148CrossRefGoogle Scholar
  39. 39.
    T. Tong, J. Li, Q. Chen, J.P. Longtin, S. Tankiewicz, and S. Sampath, Ultrafast Laser Micromachining of Thermal Sprayed Coatings for Microheaters: Design, Fabrication and Characterization, Sensors and Actuators A, 2004, 114(1), p 102-111CrossRefGoogle Scholar
  40. 40.
    M. Scagliotti, F. Parmigiani, G. Chiodelli, A. Magistris, G. Samoggia, and G. Lanzi, Plasma-Sprayed Zirconia Electrolytes, Solid State Ionics, 1988, 28-30(Part 2), p 1766-1769CrossRefGoogle Scholar
  41. 41.
    H. Guo, S. Kuroda, and H. Murakami, Microstructures and Properties of Plasma-Sprayed Segmented Thermal Barrier Coatings, J. Am. Ceram. Soc., 2006, 89(4), p 1432-1439CrossRefGoogle Scholar
  42. 42.
    R. Brandt, L. Pawlowski, G. Neuer, and P. Fauchais, Specific heat and thermal conductivity of plasma stabilized yttria-stabilized zirconia and NiAl, NiCrAl, NiCrAlY, NiCoCrAlY coatings, High Temp. High Press, 1986, 18, p 65-67Google Scholar

Copyright information

© ASM International 2015

Authors and Affiliations

  • Mahder Tewolde
    • 1
  • Gaosheng Fu
    • 1
  • David J. Hwang
    • 1
  • Lei Zuo
    • 3
  • Sanjay Sampath
    • 2
  • Jon P. Longtin
    • 1
  1. 1.Department of Mechanical Engineering, Center for Thermal Spray ResearchStony Brook UniversityStony BrookUSA
  2. 2.Department of Materials Science and Engineering, Center for Thermal Spray ResearchStony Brook UniversityStony BrookUSA
  3. 3.Department of Mechanical EngineeringVirginia TechBlacksburgUSA

Personalised recommendations