Copper Sulfide–Zinc Sulfide Janus Nanoparticles and Their Seebeck

Feb 25, 2016 - Copper Sulfide–Zinc Sulfide Janus Nanoparticles and Their Seebeck Characteristics for Sustainable Thermoelectric Materials...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/JPCC

Copper Sulfide−Zinc Sulfide Janus Nanoparticles and Their Seebeck Characteristics for Sustainable Thermoelectric Materials Hiroyuki Shimose,† Maninder Singh,† Dipali Ahuja,† Wei Zhao,‡ Shiyao Shan,‡ Shunsuke Nishino,† Masanobu Miyata,† Koichi Higashimine,† Derrick Mott,*,† Mikio Koyano,† Jin Luo,‡ C. J. Zhong,‡ and Shinya Maenosono*,† †

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



S Supporting Information *

ABSTRACT: A heterostructured copper sulfide−zinc sulfide nanocomposite is explored as a new class of low temperature and sustainable thermoelectric materials. The nanoparticles are created through a wet chemical synthetic technique and display a remarkable Janus structure. These nanoparticles are processed as building blocks by molecular linking with short alkyl chain ligands to enhance their electrical conductivity. The nanomaterials are pressed into a pellet and subjected to subsequent thermal annealing to remove volatiles and enhance particle contacts through sintering. The resulting nanocomposite materials were characterized to assess the thermoelectric characteristics, revealing P-type conductivity.



INTRODUCTION Nanoparticles that have well-defined size, shape, composition and structure have found use in advanced applications such as for optical-based sensing,1,2 catalysis,3,4 bioimaging,1,5 and biomolecular diagnostics,1,2,5 among many others. The use of nanoparticles with highly uniform characteristics is important in these applications because the unique properties of the material are highly sensitive to the physical parameters. As a result, a lot of effort has been made to study the relationship between the individual nanoparticle characteristics (size, shape, composition, structure) and the resulting properties. However, in the field of thermoelectrics, although nanotechnology has proven invaluable to increasing the material efficiency, few studies have focused on true control of the nanoscale parameters. Thermoelectric materials encompass a field where the Seebeck, Peltier, and Thomson effect are harnessed to either create an electric current from a temperature gradient or use an electric current to generate a temperature gradient.6 The phenomenon efficiency is related by the thermoelectric figure of merit (ZT = σS2T/k), where σ is electrical conductivity, S is the Seebeck coefficient, T is temperature, and k is thermal conductivity of the material. Large scale devices made from thermoelectric materials have the potential to greatly enhance our energy production efficiency because much of the energy we produce is wasted as excess heat generation. In terms of electricity generation, thermoelectric materials can be used in automobiles7,8 or in any electricity production plant utilizing steam turbine technology (nuclear, coal, geothermal, etc.).8 In fact, even by using modest thermoelectric materials (ZT = 1 to 2) in © XXXX American Chemical Society

these applications, the energy production efficiency can be increased by as much as 15%,9 representing a monumental increase in energy generation efficiency, making the technology highly important in meeting energy needs while minimizing material consumption and waste production. Thermoelectric materials have been known of for quite a long time, along with their potential benefits, but have never found widespread use because of fundamental limitations in the energy conversion efficiency of even the best materials.9 The challenges associated with increasing the overall ZT value can be observed in the figure of merit equation where electrical conductivity must be increased while thermal conductivity suppressed. Physically, these two parameters are challenging to isolate individually and physical techniques that increase or decrease one of the parameters has the same effect on the other. Recently, however, it was found that nanotechnology offers one route to suppressing the thermal conductivity while essentially maintaining the electrical conductivity of a material.9,10 By imparting nanoscale crystal grain defects into the material, the thermal conductivity can be suppressed through scattering of the heat carrying phonon at these crystal grain interfaces, whereas the electrical conductivity is unimpeded because of the significantly smaller mean free path of the electron.10,11 The finding has generated a resurgence in interest in thermoelectric materials and has led to the development of many related techniques for Received: December 3, 2015 Revised: February 24, 2016

A

DOI: 10.1021/acs.jpcc.5b11857 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C enhancing a materials ZT value.9,12 Despite this progress, however, there is still little understanding of how exactly the nanoscale structuring can be optimized for obtaining the best ZT value for a given material because there are few studies focusing on thermoelectric materials composed of nanoparticles with well-defined size, shape, composition, and structure. Many studies of thermoelectric materials deal with top-down techniques where a bulk material is broken down, oftentimes with ball-milling, then is compressed back into a large scale material. Enhanced techniques utilize spark plasma sintering to maximize electrical conductivity and many specialized methods of material preparation revolve around the approach.9 In contrast, very few studies utilize bottom-up wet chemical nanoparticle synthesis as a production technique for the particles. Although such synthetic approaches can provide high quality particles with uniform parameters, there are several drawbacks. When nanoparticles are used as building blocks for the construction of thermoelectric materials, many challenges are introduced which must be overcome in order for the material to exhibit good thermoelectric activity. First, any organic ligands must be removed from the nanoparticle surfaces or replaced by something that is sufficiently electrically conductive.9,13 Next, the natural high surface area of the particles makes the creation of a dense pellet, which is required for a thermoelectric material, a fundamental challenge (high pressure, high heat, or other techniques are required that would destroy the nanoparticle characteristics).9 Finally, most nanoparticle synthetic techniques only successfully produce a minute amount of sample (typically 100 mg or less), which is much less than is required to reliably measure characteristics such as thermal conductivity. In order for the role of nanotechnology to be more fully understood and utilized in thermoelectric materials, these challenges must be researched and overcome. The study presented here attempts to answer some of the fundamental questions about nanoparticle size, shape, composition, and structure for a sustainable set of nanoparticle materials. Much information has been generated for both bulk materials and the nanostructured counterparts of many theoretically ideal materials such as bismuth telluride,14,15 bismuth antimony telluride,10,16 lead telluride,7,14 and so forth. However, these materials can never find widespread use because of the fundamental nonsustainability of such systems. Tellurium is highly valuable as a component in thermoelectric materials because of its favorable electronic band structure;8,12 however, tellurium is extremely rare in the earth, necessitating the development and discovery of new and sustainable thermoelectric materials. The chalcogenide class of materials has proven highly interesting (of which tellurium is a member) as an alternative, particularly selenium- or sulfur-based materials.8,17 The sulfur-based chalcogenides offer potentially high thermoelectric efficiency and are inherently sustainable because of the high elemental abundance. In this work, we have focused on copper- and zinc-based sulfides (copper sulfide− zinc sulfide) because of these elements sustainable nature, the finding that copper-sulfide-based materials possess a unique “phonon glass, electron crystal” effect,8,9,11 and the fact that zinc has traditionally been used as a dopant to control electronic structure in many related materials, which would be useful for tailoring the thermoelectric efficiency. The study encompasses the bottom-up wet chemical synthesis of unique Janus structure Cu2S−ZnS nanoparticles with varying compositions, and the processing of the particles into large-scale thermoelectric materials. The physical properties of the

nanomaterials are studied and the Seebeck coefficient is analyzed. We comment on the formation mechanism of these particles and the relationship between particle composition, structure, and the resulting Seebeck properties. Although it is still a technical hurdle to assess the thermal conductivity of the nanoparticle material produced in this study (which is required to truly appraise the thermoelectric efficiency), the results provide significant insight into how nanoparticles synthesized using a bottom-up synthetic strategy can be processed into an effective thermoelectric material and, more importantly, opens the door to using sustainable materials with heterostructure as a pathway to control the resulting thermoelectric properties.



EXPERIMENTAL SECTION Chemicals. Copper nitrate trihydrate (Cu(NO3)2·3H2O), zinc nitrate nonahydrate (Zn(NO3)3·9H2O), 1-octadecene, 1dodecanethiol, ethylene diamine, and common solvents were obtained from Sigma-Aldrich, used as received. Instrumentation. Transmission electron microscopy (TEM) images were collected on a Hitachi-7650 instrument operated at 100 kV. Scanning TEM with a high-angle annular dark field (STEM-HAADF) detector and EDS elemental mapping were performed on a JEOL JEM-ARM200F instrument operated at 200 kV, the nominal resolution is 0.8 Å. Samples for TEM, EDS, and STEM-HAADF were prepared by dropping suspended nanoparticles onto a carbon-coated copper grid and drying in air overnight. X-ray diffraction (XRD) patterns were collected in reflection geometry using a Rigaku SmartLab X-ray diffractometer at room temperature with Cu Kα radiation (wavelength 1.542 Å, 40 kV, 30 mA). X-ray photoelectron spectroscopy (XPS) was performed on a Shimadzu Kratos AXIS-ULTRA high performance system. Photoelectrons were excited by monochromated Al Kα radiation. Detection was done with a delay-line detector (DLD) and a concentric hemispherical analyzer (CHA). The X-ray tube was operated at 150 W and the pass energy of the CHA was 20 eV for narrow-scan spectra. The analyzed area on the specimen surface was 300 × 700 μm2 and was located in the center of the irradiated region. The instrument was operated at a vacuum level of 1 × 10−8 Torr. Nanoparticle Synthesis. Nanoparticles were synthesized using a wet chemical synthetic approach driven by thermolysis. A total of 0.374 mmol of Cu(NO3)2·3H2O and Zn(NO3)3· 9H2O precursors were used in the synthesis. First, 50 mL of solvent (1-octadecene) and 6.042 × 10−3 moles (1.5 mL) of dodecanethiol (DDT) was added into a three neck roundbottom flask. One of the flask necks was used for monitoring the reaction temperature via thermocouple probe, another neck was fitted with a gas trap and condenser to catch volatile materials during the synthesis, and the final neck was used for injecting reactants and for bubbling argon through the reaction solution. The reaction mixture was stirred at 600 rpm using a stirring bar with argon bubbling and was kept at room temperature for 5 min to remove oxygen. After that, the flask was heated to 240 °C, when the temperature of the reaction flask reached 225 °C a stock solution containing the copper and zinc precursors dissolved in 5 mL of methanol was injected into it. This high temperature injection ensures that the nanoparticle formation occurs quickly, helping to regulate the nanoparticle monodispersity. After the reaction temperature reached 240 °C the reaction temperature stabilized and was maintained for 2 h. The solution color changed from light gray to black, indicating the formation of nanoparticles. After the reaction, the heating B

DOI: 10.1021/acs.jpcc.5b11857 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C mantle was removed and the reaction mixture was cooled to room temperature. Ethanol (99.5%, 300 mL) was added to the reaction mixture to precipitate the nanoparticles by centrifuging at 3500 rpm for 5 min, then the clear supernatant was removed and discarded. After this first washing the particles were further purified. Particles were dispersed in 10 mL of hexane with the help of sonication and then 20 mL ethanol-methanol (1:1) mixture was added to the solution. Ethanol-methanol mixture acts as a poor solvent and allows precipitation of the nanoparticles. After the addition of poor solvent the reaction mixture is centrifuged at 4000 rpm for 5 min. The clear supernatant is removed and discarded. This washing process was repeated twice. The resulting nanoparticles were dried and used for subsequent characterization. Sample Preparation for Seebeck Measurement. The dried particles were dispersed in a 5 volume percent solution of ethylene diamine (EDA) in toluene (a sufficient amount to suspend the dried particles) via sonication for 2 h. Then, the dispersion was centrifuged for 3 min at 3500 rpm. This process was repeated twice to ensure that a majority of the long chain DDT was replaced with the short chain linker EDA. The aggregated particles were dried in vacuum and then were pelletized using a hydraulic pellet press. The sample was pressed with a pressure of 25 MPa for 3 min and a pellet of approximately 2 mm thickness and 10 mm diameter was obtained. Next, the pellet was placed in a quartz tube that was purged with flowing nitrogen gas and was annealed at a temperature of 400 °C for 2 h. The pellet maintained a black color throughout the treatment and XRD analysis showed that the material retained the independent copper and zinc sulfide phases with no oxide formation. After this procedure, the sample was ready for Seebeck coefficient measurement using a constant heat-flow method.18 The pellet was broken into several pieces and a single piece was attached to a glass slide. Gold paste was used to create connections on opposite sides of the pellet piece. Copper and constantan thermocouple wires were attached to both gold contacts on the pellet piece. These wires were then connected to a digital volt meter, which was used to measure voltage fluctuations during heating of the pellet. One side of the connecting wire was kept in an ice bath as a reference for the thermoelectric temperature−voltage measurement (constant 0 °C). Then, one side of the pellet was gently heated using a contact heating probe and the resulting voltage change and corresponding temperatures at hot and cold sides of the pellet were recorded. This data was used to calculate the Seebeck coefficient of the materials. After the synthesis of the nanoparticle samples, a dispersion of each in hexane was drop-cast on a carbon coated copper microgrid and was analyzed using transmission electron microscopy (TEM).

Figure 1. TEM images of as-synthesized nanoparticles synthesized with Cu:Zn molar feeding ratios of (A) 4:0, (B) 3.5:0.5, (C) 3:1, (D) 2.5:1.5, (E) 2:2, and (F) 1.5:2.5.

is more evident in the magnified image insets. For nanoparticles synthesized using a molar feeding ratio of 3.5:2.5 Cu:Zn (Figure 1B), some particles appear to have a slightly lighter section at their periphery. These features become more pronounced for particles synthesized with a molar feeding ratio of 3:1 Cu:Zn (Figure 1C), where a sharpened point appears to extend from the periphery of the disk shaped particle. This feature becomes more evident and rounded for 2.5:1.5 and 2:2 Cu:Zn molar feeding ratio samples (Figure 1D and E). However, for the highest feeding ratio of zinc (1.5:2.5 Cu:Zn, Figure 1F), the particles are suddenly much larger, are irregular in size, and appear to be spherical. Table 1 shows the calculated size distributions of all nanoparticle samples where about 200 particles were counted for each. Table 1. TEM Determined Size and Size Distributions of Nanoparticle Samples sample Cu4Zn0 Cu3.5Zn0.5 Cu3Zn1 Cu2.5Zn1.5 Cu2Zn2 Cu1.5Zn2.5

size (nm) 7.2 6.5 9.4 8.7 9.5 18.0

± ± ± ± ± ±

0.4 0.4 0.7 0.6 0.8 5.7

The nanoparticles were also analyzed using X-ray diffraction crystallography (XRD) to study the crystalline structure of the particles. Figure 2 shows the resulting XRD patterns produced for each sample in the analysis. The peaks are indexed to the hexagonal crystal structure of copper sulfide and the wurtzite structure of zinc sulfide. The patterns do not show any significant material impurity such as oxides.21,22 The relatively broad peaks present in the patterns reflect the corresponding small crystal grain sizes observed in the TEM analysis. For nanoparticles synthesized with a molar feeding ratio of 4:0 Cu:Zn, the pattern is consistent with pure copper sulfide. As zinc is introduced, however, a subtle broadening of the peaks in the range of 25−30° and 46−50° begins to occur. This is most evident for the nanoparticles synthesized with a molar feeding ratio of 2.5:1.5 Cu:Zn where the peaks from zinc sulfide overlap with those of copper sulfide. However, as the feeding ratio of



RESULTS AND DISCUSSION Nanoparticle Characterization. Figure 1 shows a representative set of TEM images of the nanoparticle samples synthesized. The nanoparticles appear to have highly uniform size and shape, but a close inspection reveals a subtle evolution in the particle morphologies. For particles synthesized using only copper precursor (Figure 1A), uniformly sized particles with a platelet or disk morphology are observed, where the particles lay on the TEM grid face down (resulting in the apparent spherical morphology), which has been observed in previous studies.19,20 However, as zinc is added in the particle synthesis, a new feature begins to appear on the particles, which C

DOI: 10.1021/acs.jpcc.5b11857 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 2. XRD patterns of the nanoparticle samples for Cu:Zn molar feeding ratios of (A) 4:0, (B) 3.5:0.5, (C) 3:1, (D) 2.5:1.5, (E) 2:2, and (F) 1.5:2.5. The peaks are indexed to the hexagonal crystal structure of copper sulfide and the wurtzite structure for zinc sulfide as indicated by the circle and square symbols in the graph.21,22

zinc is increased beyond this point, the peaks originating for zinc sulfide quickly decrease in intensity. For a molar feeding ratio of 1.5:2.5 Cu:Zn, no zinc sulfide peaks can be observed and only copper sulfide appears present. The peaks from this sample are also markedly narrower, which is a result of the larger size of the nanoparticles in this case. This data suggests that these nanoparticles possess a heterostructure or Janus structure, where individual particles are phase segregated into copper sulfide and zinc sulfide, which is further supported by the morphologies revealed in the TEM images. The structure appears to evolve as the metallic feeding ratio is varied, where the zinc sulfide phase first emerges as the zinc feeding ratio is increased until a critical maximum at 2.5:1.5 Cu:Zn and then quickly decreases as more zinc is added. With the observation from the TEM and XRD data that some samples appear to contain both Cu2S and ZnS phases, we used scanning transmission electron microscopy with a highangle annular dark field (STEM-HAADF) detector to gain more direct information on individual particle composition. Figure 3 shows STEM-HAADF images and a corresponding elemental map for the nanoparticle sample synthesized with a Cu:Zn molar feeding ratio of 2.5:1.5. The STEM-HAADF images show that individual particles contain two distinct phases, as evidenced by the different brightness in single particles. The magnified view of a single particle in Figure 3F clearly shows that the particle contains two distinct halves with different crystalline structure or orientation. By referencing the corresponding positions in the Cu and Zn elemental maps, we can observe that half the particle contains copper, whereas the neighboring area contains zinc. The overlay image shows this particularly well where individual particles are observed to contain both a copper phase and a zinc phase, whereas sulfur can be observed throughout all of the particles. The data provides a good indication that these particles have a Janus structure, with roughly half the particle composed of Cu2S and the other half composed of ZnS. The occurrence of this structure is highly exciting in the field of thermoelectric materials because of the implications to heightened and controlled suppression of thermal conductivity, which could be achieved by greater scattering of the heat carrying phonon at the interface of Cu2S and ZnS. In other systems, this could be enhanced by using materials with very a different crystal

Figure 3. STEM-HAADF image of Cu2S−ZnS NPs synthesized with a metallic feeding ratio of 2.5:1.5 (A). Elemental maps were collected corresponding to the initial image for copper K line (B), zinc K line (C), and sulfur K line (D). An overlay of all three mapped elements is shown in (E), and a magnified STEM-HAADF image is shown in (F). The scale bar shown in (A) is the same as that for B−E.

structure or lattice parameter, which is very similar to the wellknown cases of superlattice materials and their remarkably low thermal conductivity.23 To further gain an understanding of the overall material composition, X-ray photoelectron spectroscopy (XPS) was performed. High resolution XPS spectra were collected in the copper 2p, zinc 2p, and sulfur 2p regions and are shown in Figure 4. Each sample shows well-defined single peaks in the

Figure 4. XPS spectra in the Cu 2p, Zn 2p, and S 2p areas for nanoparticles synthesized with 4:0, 3.5:0.5, 3:1, 2.5:1.5, 2:2, and 1.5:2.5 molar feeding ratio of Cu:Zn from bottom to top.

copper and zinc regions that correspond to chalcogenides. Although it is difficult to distinguish the peak positions for oxides or sulfides, an analysis of the oxygen area (O 1s peak) reveals only a low intensity peak that likely arises as a result of adsorbed species on the material surfaces. By using the calculated peak area (and the corresponding sensitivity factors) for the copper 2p3/2 and zinc 2p3/2 spin state peaks, the metallic atomic composition was calculated for each sample and is D

DOI: 10.1021/acs.jpcc.5b11857 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

(one of the two faces) of the Cu2S discs.25 The nucleation of ZnS takes place on only one face because the Cu2S disc is anisotropic with one of the 001 faces being cationic and the other anionic.26 The cationic face of the disk is very strongly protected with alkanethiol molecules, so the ZnS only has access to the weakly protected anionic face of the Cu2S crystal. This accounts for the observation of Janus nanoparticles using this synthetic technique. Once ZnS begins to grow on the 001 face of the Cu2S, it continues to do so until the zinc precursor is used up. Simultaneously, however, the reaction temperature is sufficiently high for ZnS to nucleate and form independently of the Cu2S. The ZnS particles are ultrafine, with a very small size (∼1 nm) and are effectively transparent in the nanoparticle solution.24 The synthesis then becomes a kinetic race between the formation of Janus nanoparticles and the ultrafine ZnS particles. As the feeding ratio of zinc is increased, the formation of ultrafine ZnS becomes increasingly dominant. The reason these ultrafine ZnS particles are not detected in subsequent composition analysis is because they are very challenging to isolate with the nanoparticle purification process used in this study, which causes them to be washed away during the purification procedure. This explanation also accounts for the fact that elemental analysis shows less zinc in the sample as the zinc feeding ratio is increased in our reaction, or the more ultrafine ZnS particles that are formed, the lower the ratio of zinc that is detected in the isolated Janus nanoparticles. Finally, for a very high zinc feeding ratio, Cu2S nanoparticles are observed with a larger size and very low zinc content (∼2%). This occurs because the low feeding ratio of copper inhibits the early formation of Cu2S particles, allowing primarily ultrafine ZnS particles to form independently. The relatively small amount of Cu2S in the reaction solution then undergoes a very slow growth process to large particles. The Cu2S does not become incorporated to the ultrafine zinc particles because when ZnS forms independently, it possesses the sphalerite (cubic) structure, which is incompatible with the preferred crystal structure of Cu2S. This is interesting given the fact that when the ZnS forms on the Cu2S surface, it adopts the wurtzite structure (hexagonal), which is not observed spontaneously at low temperature.25 Scheme 1 shows the general nanoparticle formation pathway as well as the types of nanoparticles formed under different metallic feeding ratio regimes. Thermoelectric Characterization. The Seebeck coefficient S of each sample was measured by a constant heat-flow method as described in the Experimental Section. The sample with the lowest Cu:Zn metallic feeding ratio was not measured because the yield was too low to collect enough sample to form a pellet. The resulting Seebeck analysis is shown in Figure 5, where the thermoelectric voltage response is plotted versus the corresponding temperature change. A linear regression of the plots is used to calculate the Seebeck value, represented by the slope of the line. The analysis reveals P-type conductivity for all of the samples (positive slope) with the calculated Seebeck coefficients at room temperature shown in Table 3. Contrary to our expectations, the pure Cu2S has the highest Seebeck coefficient of 146 μV/K, which is significantly higher than the reference value for comparable Cu2S materials (80 μV/K at 700 K maximum).27 The observation might arise because of the unique surface states of this material (very high surface area and small crystalline size) or it could be because the material adopts a copper deficient phase such as djurlite (Cu1.8S) or other similar phase, which is challenging to detect for our material (because of the very small crystalline grain size). Whichever the

shown in Table 2. Sulfur content can also be calculated using this technique; however, because the nanoparticles are coated Table 2. Calculated Particle Metallic Composition from XPS Data Cu4Zn0 Cu3.5Zn0.5 Cu3Zn1 Cu2.5Zn1.5 Cu2Zn2 Cu1.5Zn2.5

Cu atom %

Zn atom %

100 89 83 79 80 98

0 11 17 21 20 2

in alkanethiol molecules, the contribution from this source cannot be separated from the sulfur content in the chalcogenide structure, so we did not include it in this calculation. These compositions confirm that the zinc content in the nanoparticle samples does not share a linear relationship with the metallic feeding ratio. In addition to revealing the nanomaterial metallic content, inspection of the S 2p region also appears to contain a visual trend in the peak shape. These nanoparticles are expected to contain two different types of sulfur including sulfide, which is part of the chalcogenide component (the copper sulfide or zinc sulfide), and thiolate, which is the sulfur contained in the dodecanethiol adsorbed to the particle surfaces. Deconvolution of these chemical species was performed using a Gaussian function to fit S 2p3/2 and S 2p1/2 peaks for sulfide and thiolate; however, the results did not show a significant trend in the relative amounts of each sulfur species as a function of metallic feeding ratio or elemental composition. Instead, the peak shape is likely changing as a result of differences in the ratio of copper sulfide and zinc sulfide contained in each sample, which we could not isolate by deconvolution. The peak fitting, fitting parameters, and calculated ratio of sulfide and thiolate is shown in the Supporting Information. Proposed Nanoparticle Formation Pathway. The observation of the Janus nanoparticle structure and the fact that there is a maximum amount of zinc that can be incorporated into the individual particles presents a fundamental question on the particle formation pathway. In classical terms, if all of the metal precursor is converted to nanoparticles, then the measured particle composition should always be similar to the metallic feeding ratio. In our study, the amount of zinc that appears in each nanoparticle sample is lower than the corresponding metallic feeding ratio, reaching a maximum content around 50% molar feeding ratio, and dropping off sharply as the ratio of zinc is further increased. Here, we discuss how the Janus nanoparticles form and why the measured particle compositions deviate from the expected metallic feeding ratio based on our experimental observations. In this reaction, the particle formation is driven by thermolysis, which is essentially a decomposition reaction where the metal assists in catalytically cleaving the sulfur−carbon bond in the alkanethiol molecule, causing sulfur to be incorporated and leading to the chalcogenide material. For pure copper, this begins to occur at a temperature of 135 °C19 and at 190 °C for pure zinc .24 As temperature is increased, the rate of thermolysis increases as well. Given these two reference values for thermolysis, it is likely that Cu2S quickly forms first, followed by formation of ZnS. After the Cu2S nanoparticles form, the ZnS undergoes nucleation and growth on one of the 001 planes E

DOI: 10.1021/acs.jpcc.5b11857 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C Scheme 1. Proposed Formation Pathway to Janus Cu2S−ZnS Janus Nanoparticles Synthesized with Various Metallic Feeding Ratios

Table 3. Seebeck Coefficients at Room Temperature S (μV/K) Cu4Zn0 Cu3.5Zn0.5 Cu3Zn1 Cu2.5Zn1.5 Cu2Zn2

146 113 47 44 37

With this in mind, it may be that the heterostructure will display an overall enhanced ZT value when compared to the pure copper sulfide. The determination of the full range of thermoelectric properties for this material is part of our continued studies in this area.



CONCLUSIONS Cu2S−ZnS nanoparticles have been synthesized and studied in terms of their Seebeck coefficient, an important parameter for thermoelectric materials. The demonstration of the thermolysis-based synthetic approach has provided important fundamental information on how the Cu2S and ZnS materials interact to create Janus-type particles during the synthetic formation, which could be harnessed to create related materials with complex structures. The material’s Seebeck coefficient has also revealed a particularly high value for the pure Cu2S, where further studies may lead to important insight into the benefits that nanomaterials carry for explorations of the thermoelectrics.29 As the nanoparticles ZnS content begins to increase, the Seebeck coefficient decreases as a result of dilution of the overall material because the Cu2S provides the major contribution to the semiconducting activity, whereas the ZnS is relatively inactive. Further study of the thermal conductivity of these materials will be key in fully exploring their suitability for thermoelectrics. The results related here provide a new way of looking at nanomaterials for thermoelectrics where control of the nanoparticle structure can lead to new and unexpected semiconducting properties.

case, as zinc is introduced to the material there is a continual decrease in the Seebeck coefficient. On the basis of the observed morphology, this likely arises as a result of a dilution effect, where the primary contribution to the semiconductor activity comes from the Cu2S component of the particles and the ZnS is contributing little activity. This makes sense because pure ZnS has a low carrier density without being doped by a carrier element.28 As the particle composition becomes dominated by ZnS, the Seebeck value decreases toward that of ZnS. Although these results suggest that the copper sulfide is the best thermoelectric material candidate from the perspective of Seebeck coefficient, a full characterization of the materials thermal and electrical conductivity is required to assess the true thermoelectric efficiency (i.e., ZT value). For these materials, these measurements are still a great technical hurdle, especially for determining the thermal conductivity, which requires a very large amount of sample, typically at least 1 g. In the case of our own materials, we are dealing with about 20 mg of sample, which allows the reliable determination of Seebeck coefficient, but restricts us from determining other physical properties such as thermal conductivity. Scaling up the synthesis is an attractive option, but we found the kinetics of the reaction were greatly affected by a simple scale up scheme, resulting in particles with significantly different size, shape, composition, and structure.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.5b11857. XPS spectra of S 2p area and fitting parameters. (PDF)

Figure 5. Thermoelectric voltage measurement of the Cu2S−ZnS nanoparticle materials showing P-type conductivity for Cu:Zn molar feeding ratios of (A) 4:0, (B) 3.5:0.5, (C) 3:1, (D) 2.5:1.5, and (E) 2:2. F

DOI: 10.1021/acs.jpcc.5b11857 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C



(16) Zhao, Y.; Dyck, J. S.; Hernandez, B. M.; Burda, C. Enhancing Thermoelectric Performance of Ternary Nanocrystals Through Adjusting Carrier Concentration. J. Am. Chem. Soc. 2010, 132, 4982−4983. (17) Maignan, A.; Guilmeau, E.; Gascoin, F.; Breard, Y.; Hardy, V. Revisiting Some Chalcogenides for Thermoelectricity. Sci. Technol. Adv. Mater. 2012, 13, 053003. (18) Suekuni, K.; Tsuruta, K.; Kunii, M.; Nishiate, H.; Nishibori, E.; Maki, S.; Ohta, M.; Yamamoto, A.; Koyano, M. High-performance Thermoelectric Mineral Cu12‑xNixSb4S13 Tetrahedrite. J. Appl. Phys. 2013, 113, 043712. (19) Sigman, M. B.; Ghezelbash, A.; Hanrath, T.; Saunders, A. E.; Lee, F.; Korgel, B. A. Solventless Synthesis of Monodisperse Cu2S Nanorods, Nanodiscs, and Nanoplatelets. J. Am. Chem. Soc. 2003, 125, 16050−16057. (20) Zhuang, Z.; Peng, Q.; Zhang, B.; Li, Y. Controllable Synthesis of Cu2S Nanocrystals and Their Assembly into a Superlattice. J. Am. Chem. Soc. 2008, 130, 10482−10483. (21) Reference patterns accessed from the International Centre for Diffraction Data database 2015, card number 00-046-1195 for hexagonal copper sulfide and 01-079-2204 for wurtzite zinc sulfide. (22) Lyubutin, I. S.; Lin, C.-R.; Starchikov, S. S.; Siao, Y.-J.; Shaikh, M. O.; Funtov, K. O.; Wang, S.-C. Synthesis, Structural and Magnetic Properties of Self-organized Single-crystalline Nanobricks of Chalcopyrite CuFeS2. Acta Mater. 2013, 61, 3956−3962. (23) Venkatasubramanian, R.; Siivola, E.; Colpitts, T.; O'Quinn, B. Thin-film Thermoelectric Devices with High Room-temperature Figures of Merit. Nature 2001, 413, 597−602. (24) Kuzuya, T.; Tai, Y.; Yamamuro, S.; Sumiyama, K. Synthesis of Copper and Zinc Sulfide Nanocrystals via Thermolysis of the Polymetallic Thiolate Cage. Sci. Technol. Adv. Mater. 2005, 6, 84−90. (25) Kolny-Olesiak, J. Synthesis of Copper Sulphide-based Hybrid Nanostructures and their Application in Shape Control of Colloidal Semiconductor Nanocrystals. CrystEngComm 2014, 16, 9381−9390. (26) Connor, S. T.; Hsu, C.-M.; Weil, B. D.; Aloni, S.; Cui, Y. Phase Transformation of Biphasic Cu2S-CuInS2 to Monophasic CuInS2 Nanorods. J. Am. Chem. Soc. 2009, 131, 4962−4966. (27) Ge, Z.-H.; Zhang, B.-P.; Chen, Y.-X.; Yu, Z.-X.; Liu, Y.; Li, J.-F. Synthesis and Transport Property of Cu(1.8)S as a Promising Thermoelectric Compound. Chem. Commun. 2011, 47, 12697−12699. (28) Diamond, A. M.; Corbellini, L.; Balasubramaniam, K. R.; Chen, S.; Wang, S.; Matthews, T. S.; Wang, L.-W.; Ramesh, R.; Ager, J. W. Copper-alloyed ZnS as a p-type Transparent Conducting Material. Phys. Status Solidi A 2012, 209, 2101−2107. (29) Zhao, W.; Shan, S.; Luo, J.; Mott, D. M.; Maenosono, S.; Zhong, C. J. Harvesting Nanocatalytic Heat Localized in Nanoalloy Catalyst as a Heat Source in a Nanocomposite Thin Film Thermoelectric Device. Langmuir 2015, 31, 11158−11163.

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors gratefully acknowledge financial support by the Mitani Foundation Grant and the Nippon Sheet Glass Foundation for Materials Science and Engineering in this research.



REFERENCES

(1) Jain, P. K.; Huang, X.; El-Sayed, I. H.; El-Sayed, M. A. Noble Metals on the Nanoscale: Optical and Photothermal Properties and some Applications in Imaging, Sensing, Biology, and Medicine. Acc. Chem. Res. 2008, 41, 1578−1586. (2) Saha, K.; Agasti, S. S.; Kim, C.; Li, X.; Rotello, V. M. Gold Nanoparticles in Chemical and Biological Sensing. Chem. Rev. 2012, 112, 2739−2779. (3) Narayanan, R.; El-Sayed, M. A. Catalysis with Transition Metal Nanoparticles in Colloidal Solution: Nanoparticle Shape Dependence and Stability. J. Phys. Chem. B 2005, 109, 12663−12676. (4) Wang, D.; Li, Y. Bimetallic Nanocrystals: Liquid-phase Synthesis and Catalytic Applications. Adv. Mater. 2011, 23, 1044−1060. (5) Alivisatos, P. The Use of Nanoparticles in Biological Detection. Nat. Biotechnol. 2004, 22, 47−52. (6) Rowe, D. M. Thermoelectrics Handbook: Macro to Nano; CRC Press: Boca Raton, FL, 2006. (7) Heremans, J. P.; Jovovic, V.; Toberer, E. S.; Saramat, A.; Kurosaki, K.; Charoenphakdee, A.; Yamanaka, S.; Snyder, G. J. Enhancement of Thermoelectric Efficiency in PdTe by Distortion of the Electronic Density of States. Science 2008, 321, 554−557. (8) Gonçalves, A. P.; Godart, C. New Promising Bulk Thermoelectrics: Intermetallics, Pnictides and Chalcogenides. Eur. Phys. J. B 2014, 87, 42−71. (9) Kanatzidis, M. G. Nanostructured Thermoelectrics: The New Paradigm. Chem. Mater. 2010, 22, 648−659. (10) Poudel, B.; Hao, Q.; Ma, Y.; Lan, Y.; Minnich, A.; Yu, B.; Yan, X.; Wang, D.; Muto, A.; Vashaee, D.; et al. High-thermoelectric Performance of Nanostructured Bismuth Antimony Telluride Bulk Alloys. Science 2008, 320, 634−638. (11) Liu, H.; Shi, X.; Xu, F.; Zhang, L.; Zhang, W.; Chen, L.; Li, Q.; Uher, C.; Day, T.; Snyder, G. J. Copper Ion Liquid-like Thermoelectrics. Nat. Mater. 2012, 11, 422−425. (12) Zhao, L.-D.; Hao, S.; Lo, S.-H.; Wu, C.-I.; Zhou, X.; Lee, Y.; Li, H.; Biswas, K.; Hogan, T. P.; Uher, C.; et al. High Thermoelectric Performance via Hierarchical Compositionally Alloyed Nanostructures. J. Am. Chem. Soc. 2013, 135, 7364−7370. (13) Ibanez, M.; Korkosz, R. J.; Luo, Z.; Riba, P.; Cadavid, D.; Ortega, S.; Cabot, A.; Kanatzidis, M. G. Electron Doping in Bottom-up Engineered Thermoelectric Nanomaterials through HCl-Mediated Ligand Displacement. J. Am. Chem. Soc. 2015, 137, 4046−4049. (14) Kovalenko, M. V.; Spokoyny, B.; Lee, J.-S.; Scheele, M.; Weber, A.; Perera, S.; Landry, D.; Talapin, D. V. Semiconductor Nanocrystals Functionalized with Antimony Telluride Zintl Ions for Nanostructured Thermoelectrics. J. Am. Chem. Soc. 2010, 132, 6686−6695. (15) Scheele, M.; Oeschler, N.; Meier, K.; Kornowski, A.; Klinke, C.; Weller, H. Synthesis and Thermoelectric Characterization of Bi2Te3 Nanoparticles. Adv. Funct. Mater. 2009, 19, 3476−3483. G

DOI: 10.1021/acs.jpcc.5b11857 J. Phys. Chem. C XXXX, XXX, XXX−XXX