Harvesting Nanocatalytic Heat Localized in Nanoalloy Catalyst as a

Letter pubs.acs.org/Langmuir

Harvesting Nanocatalytic Heat Localized in Nanoalloy Catalyst as a Heat Source in a Nanocomposite Thin Film Thermoelectric Device Wei Zhao,† Shiyao Shan,† Jin Luo,† Derrick M. Mott,‡ Shinya Maenosono,*,‡ and Chuan-Jian Zhong*,† †

Department of Chemistry, State University of New York at Binghamton, Binghamton, New York 13902, United States School of Materials Science, Japan Advanced Institute of Science and Technology, 1-1 Asahidai, Nomi, Ishikawa 923-1292, Japan

‡

S Supporting Information *

ABSTRACT: This report describes findings of an investigation of harvesting nanocatalytic heat localized in a nanoalloy catalyst layer as a heat source in a nanocomposite thin film thermoelectric device for thermoelectric energy conversion. This device couples a heterostructured copper−zinc sulfide nanocomposite for thermoelectrics and low-temperature combustion of methanol fuels over a platinum−cobalt nanoalloy catalyst for producing heat localized in the nanocatalyst layer. The possibility of tuning nanocatalytic heat in the nanocatalyst and thin film thermoelectric properties by compositions points to a promising pathway in thermoelectric energy conversion.

■

INTRODUCTION The exploration of thermoelectric energy conversion, which has attracted increasing interest in recent years, has to address major challenges in terms of enhancing thermoelectrics and harvesting heat sources. Nanotechnology has had largest impact on the field of thermoelectrics, leading to a significant improvement in the dimensionless figure of merit (ZT) for many traditional thermoelectric materials such as bismuth telluride.1−3 The ability to tune the thermoelectrics at nanoscale shows significant improvement in the dimensionless figure of merit (ZT) for many traditional thermoelectric materials (e.g., bismuth telluride). For the very best thermoelectric materials available today (i.e., Bi2Te3, BiSbTe3, PdTe, etc), either rare or toxic elements are used, which limit their practical applications.4−7 One emerging approach involves sulfur-based nanomaterials such as chalcopyrite (CuFeS2) and Cu2S-related materials,8−11 which are attractive because of the abundant nature of the constituent elements. On the other hand, the ability to harvest heat from hydrocarbon combustion effectively on thermoelectric devices would enable broader applications of thermoelectric energy conversion.12,13 This is especially attractive considering the possibility of abundant and renewable biofuels. A key problem is the lack of effective catalysts that are low cost and sustainable.14−18 Herein, we demonstrate that the combination of the thermoelectrics from earth-abundant Cu2S-based nanomaterials and the highly localized heat from catalytic combustion of hydrocarbons over noble metal-reduced nanoalloy catalysts is potentially viable for constructing a thermoelectric energy conversion device. Scheme 1 illustrates a thin film thermoelectric conversion device on which a thin film of nanocatalyst is introduced for the proof-ofconcept demonstration. The heat from room-temperature flameless catalytic combustion of hydrocarbon vapor over a nanocatalyst, i.e., © XXXX American Chemical Society

CH3OH + 3/2O2 = CO2 + 2H 2O (ΔH = −764kJ/mol)

is highly localized in the nanocatalyst layer, in contrast to conventional catalytic combustion to release heat simply to the environment. The localized heat is harvested by the thin film thermoelectric nanomaterials for thermoelectric conversion. In addition, use of Pt-alloy catalysts for the thermoelectric conversion promises to increase the catalytic activity and stability while reducing the use of noble metals.

■

EXPERIMENTAL SECTION

Synthesis of CuS and CuZnS Nanoparticles. Thermoelectric nanoparticles were synthesized using a wet chemical synthetic approach driven by thermolysis. A total of 0.374 mmol of Cu(NO3)2 H2O and Zn(NO3)2·9H2O precursors were used in the synthesis. 50 mL solvent (1-octadecene) and 6.042 × 10−3 moles (1.5 mL) of dodecanethiol (DDT) was added into a round-bottom flask. The reaction mixture was stirred under argon bubbling and was kept at room temperature for 5 min to remove the air. When the temperature of the reaction flask reached 225 °C, a stock solution containing the copper and zinc precursors dissolved in 10 mL of octadecene was injected into it. After the reaction temperature reached 240 °C, the reaction temperature stabilized and was maintained for 2 h.19 This reaction technique is based on our own modification of the traditional thermolysis technique9 for creating copper sulfide nanoparticles. Ethanol (99.5%, 300 mL) was added to the reaction mixture to precipitate the nanoparticles. The resulting nanoparticles were washed and dried. Both n- and p-type Bi2Te3 nanoparticles were also synthesized, which were detailed in a previous report.20 Received: August 26, 2015 Revised: October 2, 2015

A

DOI: 10.1021/acs.langmuir.5b03193 Langmuir XXXX, XXX, XXX−XXX

Letter

Langmuir

Scheme 1. Thin-Film Thermoelectric Device Harvesting Nanocatalytic Heat in the Nanocatalyst As a Heat Source for Energy Conversiona

a

This device couples the thermoelectrics of nanocomposite thermoelectric materials and the localized heat of catalytic combustion of hydrocarbon fuel over a nanocatalyst layer. HT: high temperature; RT: room temperature.

Figure 1. Morphologies of the nanocomposite thermoelectric materials and the nanoalloy catalysts. (A) TEM images of Cu2S nanoparticles (left, synthesized with Cu:Zn molar feeding ratio of 4:0 (7.2 ± 0.4 nm)) and Cu2.5Zn1.5S nanoparticles (right, synthesized with Cu:Zn molar feeding ratio of 2.5:1.5 (8.7 ± 0.6 nm)). (B) Plots of thermoelectric voltage versus temperature for Cu2S (black) and CuZnS (red), showing p-type conductivity with Seebeck coefficients (μV/K) of 146 and 44, respectively. (C) TEM image of PtCo nanoparticles (Pt70Co30, 8.0 ± 1.5 nm). (D) XRD pattern of the Pt55Co45 alloy nanoparticles (a, 2.0 ± 0.6 nm) and Pt55Co45/C catalyst after the thermal treatment (b, 6.6 ± 1.0 nm). Preparation of PtCo Nanocatalysts. The general synthesis of PtCo nanoparticles involved the use of three metal precursors, Pt(acac)2 and Co2(CO)8 (or Co(acac)2), in controlled molar ratios. These metal precursors were dissolved in an octyl ether solvent. A mixture of

oleylamine and oleic acid was also dissolved in the solution and used as capping agents. 1,2-Hexadecanediol was used as a reducing agent for the reduction of the Pt precursor, and elevated temperature was used to initiate the thermal decomposition of the Co precursor. The loading of B

DOI: 10.1021/acs.langmuir.5b03193 Langmuir XXXX, XXX, XXX−XXX

Letter

Langmuir the metal nanoparticles on the carbon support was controlled by their feeding ratios, which were heated at 260 °C in 20% O2/80% N2 for 90 min for removing the organic shells and then heated at 400 °C in 15% H2/85% N2 for 120 min for calcination. The loading was determined using TGA.21−23

under optimal placement of thermocouple the temperature as high as 500−600 °C can be observed.25,26 The thermocouple placement was not optimized, serving mainly as an indication of the temperature change, from which the detected small change in temperature outside the reactor (20−30 °C) supported the viability of a high gradient of temperature for the nanocatalytic reaction. The temperature change measured inside the reactor detected by a thermocouple probe at the center of the catalyst was indeed much greater. For methanol-air fuel over the nanoalloy catalyst, the measured temperature change can be tuned from room temperature to ∼300 °C easily by the fuel-air flow rate, reflecting the increase of heat in the nanocatalyst. Experimentally, the temperature change of catalyst layer and the thermoelectric (TE) voltage change under the nanocatalytic heat were measured using a combination of thermocouple and ptype nanomaterials. For example, p-type copper sulfide nanocomposites in combination with n-type Bi2Te3 were used in fabricating a thermoelectric device (Scheme 1, and Figure S3). The pretreated nanomaterials were first suspended in a solvent (e.g., toluene for p-type Cu2S, and polyvinylidene fluoride and NMethyl-2-pyrrolidone for n-type Bi2Te3) forming the desired inks. The inks were then printed onto a flat substrate printed with gold electrodes, forming two thermoelectric electrodes. A catalyst layer was then applied onto the hot end of the two thermoelectric electrodes. After drying, the thermoelectric properties were measured under controlled flow of fuel-air over the catalyst layer. For example, for a Pt55Co45/C layer (36 wt %, 60 °C drying for 1 h before testing) on the thermoelectric device, two thermocouples were attached to both the hot and cold sides of the thermoelectric devices to monitor the temperature change. Upon blowing controlled methanol/air fuel over the catalyst layer, the temperature for the hot end was monitored, which went as high as 160 °C (Figure 2). The temperature for the cold end also showed a small temperature

■

RESULTS AND DISCUSSION For nanoparticles synthesized using only a copper precursor, uniformly sized particles (Figure 1A left image and Figure S1A) were obtained. Based on detailed analysis of the image and previous work on similar systems, some of the particles appear to exhibit platelet or disk like morphology. The detailed synthesis and structural characterization of the thermoelectric copper/zinc sulfide nanocomposites have been recently reported.9 When zinc is incorporated to the nanoparticles (a metallic feeding ratio of 2.5:1.5 Cu:Zn), the particles are slightly larger with a slightly elongated shape (Figure 1A, right image and Figure S1B). The resulting Seebeck analysis is shown in Figure 1B. Both Cu2S and CuZnS exhibit p-type conductivity. The semiconducting and optical properties for this type of materials are known.24 From the slope of the thermoelectric voltage for the CuZnS nanoparticle materials, Seebeck coefficients (μV/K) are 146 (Cu4Zn0) and 44 (Cu2.5Zn1.5). After the thermal treatment, the nanomaterials in a tube furnace under N2 flow was heated at 380 °C for 40 min, the nanostructures remained largely the same as that before the treatment. However, there were subtle changes in nanocrystal phases (Figure S2), which was largely due to sintering of the nanoparticles after removing the capping molecules. For the demonstration of the nanocatalytic heat production for the thermoelectric conversion, PtCo nanoparticles of different compositions were synthesized, with the size and composition being controlled by the concentration ratio of Pt(acac)2 and Co(acac)2 precursors in a chemical reduction process in an octyl ether solvent in the presence of oleylamine and oleic acid.21−23 Figure 1C shows an example of the assynthesized PtCo nanoparticles (Figure S1C). PtCo nanoparticles of other compositions were also obtained (e.g., Pt55Co45 nanoparticles). For the preparation of nanocatalysts, the nanoparticles were further subjected to thermal treatment, as described previously,22 producing activated, carbon-supported PtCo nanoparticles (PtCo/C). As revealed by X-ray diffraction (XRD) pattern (Figure 1D), the particles exhibit an alloy characteristic with an face-centered cubic (fcc)-type lattice constant of 0.390 ± 0.002 nm for the carbon-supported Pt55Co45 nanoparticles after the thermal treatment. Note that the particles were somewhat increased in size after thermal treatment due to sintering.22,23 There are different ways to measure the approximate temperature change associated with the nanocatalytic reaction that differ in terms of their closeness to the true temperature in the nanoparticles. For example, in a simple nonoptimized configuration where the reaction was carried out in a thermal microreactor, the measurement of nanocatalytic heat used a thinlayer thermal reactor channel constructed with a 2 mm thick Teflon gasket sandwiched between two stainless steel plates (1 mm) with nanoalloy catalyst adhering to the inner side of the plate. It was sealed around the perimeter of the channel with an inlet and an outlet at the opposite ends of the channel. A controlled amount of commercial Pt/C catalyst (40% Pt loading) was placed on the inner side of the stainless steel plate. The temperature change was measured in the outer side of the stain steel plate upon fuel vapor passing over the catalyst. Note that

Figure 2. Thermoelectric voltage produced by a custom-made thin film thermoelectric device. The temperature was measured at the hot end (T1) and cold end (T2), and the temperature change (ΔT = T1 − T2) for a custom-made thin film TE device (p-type: Cu2S; n-type: Bi2Te3) coated with a layer of Pt55Co45/C catalyst under controlled methanol-air flow (“on/off” indicates flowing air through a methanol bubbler at 200 mL/min). Inset in the right panel: IR image of the device from the top side of the catalyst layer on the device. (The device in the inserted photo has a dimension of 2.0 cm × 3.3 cm; two thermocouples were attached to the hot side (upper front) and the cold side (bottom front) of the thermoelectric device.) C

DOI: 10.1021/acs.langmuir.5b03193 Langmuir XXXX, XXX, XXX−XXX

Letter

Langmuir

catalyst, a thicker layer of catalyst showed a temperature increase of about 0.9 °C. However, the corresponding voltage increase was very small (0.03 mV). While the temperature changes detected are similar, the voltage difference indicates that an increased thickness could prevent an effective heat transfer from the catalyst layer to the thermoelectric materials. This observation demonstrates the importance of controlling the thickness of the catalyst layer. Note that a current flow was also detected under multiple pulses of methanol-air fuel over the catalyst layer, showing an increase of current upon flowing. To demonstrate the generality of the finding, the thermoelectric conversion was also tested with a commercial TE device (p-type (Bi2Te3)−n-type (Bi2Te3) (Hi-Z Technology, Inc.)) coated with a layer of the same PtCo/C catalyst used for the thinfilm TE device (see Figure S3C). Upon flowing methanol-air over the catalyst layer, an increase of temperature as high as 180 °C and an increase of voltage as high as 18 mV were detected (Figure 3). The results have also demonstrated the viability of

increase. In the custom-made thermoelectric device, a corresponding voltage change was detected, which was shown to be as high as 15 mV (Figure 2). The repetition of the measurements indicated that the measured temperature and voltage were very reproducible. The temperature increase of the TE device was also mapped by infrared imaging of the device during the fuel-air flow over the catalyst layer (see right panel in Figure 2), revealing that the hot zone is largely confined in the catalyst layer, though there is some spread-out around the catalyst layer and the device. The temperature difference between the hot side and cold side was found to be around 120 °C. Based on the temperature change, the thermoelectric Seebeck coefficient was approximately estimated to be about 127 μV/K. Considering the Seebeck value of 146 μV/K for Cu2S, and 200 μV/K for n-type Bi2Te3 measured in this work, the theoretical Seebeck value for the entire device is estimated to be about 346 μV/K. Based on the Seebeck value, it translates to an apparent efficiency of ∼38%. Different methanol/air ratios in the fuel-air flow were also tested, showing a relatively small effect on the result. In another example with similar device components, the highest temperature detected was about 250 °C, at which a voltage increase of about 25 mV was detected (Figure S4A). However, a certain degree of carbon burning was also found to be operative under the catalytic heat, causing a decrease in the temperature. In this case, the temperature change detected was about 10 °C, and the voltage increase was about 1 mV (Figure S4B) The composition of the nanocatalyst plays an important role in determining the thermoelectric performance of the above TE devices. In comparison with the PtCo/C catalyst, a significant decrease in the thermoelectric voltage was observed with commercial Pt/C (40 wt %) catalyst. This can be explained by the smaller increase in temperature as a result of the lower catalytic activity of the Pt/C for the catalytic combustion. For a device with Cu2S as p-type material and Bi2Te3 as n-type material in the TE device coated with a layer of Pt/C catalyst, the temperature measured on the surface of the catalyst was shown to increase about 1 °C upon flowing methanol-air over the catalyst layer. However, the voltage shows an increase around 0.6 mV in response to the temperature increase (Figure S5A). However, the temperature showed here is the temperature measured from the catalyst surface by the thermocouple, which is much lower than the temperature of the localized heat produced in the nanocatalyst. As such, the temperature transferred to the thermoelectric materials could be much higher than 1 °C. With the methanol ratio changing from 80 to 99%, the temperature difference at the hot end changes from 0.5 to 0.8 °C. While the corresponding voltage change showed a linear relationship between the voltage measured and the methanol-air percentage flowing over the catalyst layer (Figure S5B), the magnitude of the change was much smaller than that for the case of PtCo alloy catalyst. Several different p-type nanomaterials were also studied to assess their thermoelectric characteristics. One example involved a TE device with CuZnS (thermally treated in N2 at 380 °C for 40 min) in combination with n-type Bi2Te3 onto which commercial Pt/C (40 wt %) catalyst on the hot end of the two thermoelectric electrodes was added. Upon blowing methanol/air fuel over the catalyst layer, the temperature showed a small increase of 0.7 °C. The voltage measured (