Thermoelectric characterization of nickel-nanowires and nanoparticles embedded in silica aerogels

Embedded nickel nanowires (NiNWs) and nickel nanoparticles (NiNPs) in silica aerogels at three different concentrations are characterized by scanning thermal microscopy, a Hot disk method and four probe measurements to consider them as potential thermoelectric materials. NiNW samples exhibit 9 orders of magnitude improvement in thermoelectric figure of merit while the embedded NiNPs samples show a 6 orders of magnitude improvement when the concentrations are increased from 0 to 700 ppm. The electrical resistivity is highly sensitive to the concentration of NiNWs and NiNPs in the silica aerogels, while the thermal conductivity remains largely unchanged over temperature range 300 to 420 K. The electrical conductivity σ follows a percolation scaling law of the form σ ∝ (W − W c) with critical weight fraction (W c) to form a conductive network at range 0.04-0.06 Wt% and 0.08-0.1 Wt% for embedded NiNWs and NiNPs, respectively. The investigation suggest that further optimization of the concentration of nanomaterials in aerogels could yield promising thermoelectric properties. © 2018 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1063/1.5027889


I. INTRODUCTION
The challenge in creating high-performance thermoelectric materials lies in simultaneously achieving a high electrical conductivity, high Seebeck coefficients, and a low thermal conductivity.These parameters are temperature dependent and are closely associated in such a way that by enhancing one parameter the other parameter changes and most often deteriorates. 1hermal conductivity can be significantly reduced by increasing the number of phonon scattering centers within materials. 1,2Phonon scattering inside porous materials can be controlled by manipulating the materials porosity because phonons scatter from pore boundaries.Pore size, shape and orientation affect thermal conductivity. 3,4The incorporation of nanostructures also lowers the thermal conductivity and improves the thermoelectric properties by introducing phonon scattering centers. 2 Silica aerogel is of interest to thermoelectricity fundamentally and at applied level due to its extremely low thermal conductivity (0.017 Wm −1 K −1 ). 5 Potential applications include not only thermal insulation, but also infrared detectors, 6 microelectronic devices 7 and in space explorations. 8he electrical and thermal properties of silica aerogel can be modified by the incorporation of nanomaterials.Metal nanoparticles inserted into the aerogel improve its electrical conductivity [9][10][11] to values relevant to thermoelectric materials and enabling potential applications in carbon capture. 12nvestigations into the thermal properties of silica aerogels by the addition of nanostructures have been reported. 13The thermal conductivity of nanowires can be significantly reduced by embedding it in an aerogel medium.For instance, the thermal conductivity of ZnO nanowires has been reduced by a factor of ten within a temperature range of 150 to 300 K when they were embedded into silica aerogel. 13This reduction of thermal conductivity resulted from the scattering of ballistic phonons at the nanowires boundaries. 13Bi 2 Te 3 and Bi 2−x Sb x Te 3 aerogels prepared by the sol-gel method have also been synthesized and characterized. 14The thermal conductivity was decreased from 1.7 Wm −1 K −1 to 0.5 Wm −1 K −1 for Bi 2 Te 3, while it remained unchanged for Bi 2−x Sb x Te 3 aerogels. 14However, the room temperature thermoelectric power factor, which relates to electrical conductivity, was 3.3 × 10 −4 Wm −1 K −2 for Bi 2−x Sb x Te 3 and 5.0 × 10 −4 W m −1 K −2 for Bi 2 Te 3 aerogels. 14 highly versatile silica sol-gel process has increased the electrical conductivity of Pd embedded in silica aerogel by three orders of magnitude. 15This increase results from the formation of Pd metallic percolation networks. 15RuO 2 nanoparticles similarly generated a wired RuO 2 network at the surface of silica, with an electrical conductivity as high as 10 −3 S cm −1 , 16 while similarly a porous silica-graphene nanocomposite exhibits a conductivity of 0.5 S.cm −1 . 17Despite the promising results from the development of these materials, their characterization remains incomplete since several key thermoelectric properties have not yet been evaluated.In particular, there are no reports of silica-based aerogel materials that provide information on both the thermal and electrical conductivity.
This paper describes the characterization of nickel nanowires (NiNWs) and nickel nanoparticles (NiNPs) embedded in silica aerogels at various temperatures with the aim of evaluating key thermoelectric properties.The electrical conductivity was studied by four-point probe measurements 18 and by Hall Effect equipment, while thermal conductivity was measured by using the Hot Disk Transient Plane Source method (with a Hot Disk TPS 2500 S) and with scanning thermal microscopy (SThM).Here we report a full thermoelectric evaluation of silica-based aerogels with embedded nanowires and nanoparticles.The work shows that by optimizing the concentration of NiNWs or NiNPs, a silica aerogel composite may be potentially considered as a valuable thermoelectric material.

A. Experimental details
Silica aerogel materials containing embedded NiNPs and NiNWs were fabricated via the sol-gel process and ambient pressure drying. 12Nickel nanowires were synthesized following the solvothermal process previously reported by Chen, et al. 19 The concentrations of NiNPs embedded in the silica aerogels were 500, 700, 900, and 0 ppm whereas the concentration of embedded NiNWs were 500 and 700 ppm.The sample sizes were prepared into 0.5 mm × 0.6 mm × 0.1 mm for the length, width and thickness, respectively.The thermoelectric properties of the silica aerogel samples at various temperatures have been examined and reported in this paper, except for the silica aerogel embedded with NiNPs at 900 ppm concentration.This is because during the silica gel preparation with 900 ppm of NiNPs, nickel nanoparticles have sediment during synthesis and the sample did not show any further increase in the NiNPs concentration in the final aerogel product. 12he electrical resistivity was measured by the van der Pauw method using two techniques.Fourpoint probe measurements were made with a CASCADE 4200SCS probe station from Keithley, and a Hall Effect system was also used.The benefit of using van der Pauw's technique is that it avoids inaccuracies arising from the sample geometry. 18Four small gold contacts were deposited using a thin film deposition system (Kurt Lesker PVD 75) on the edges of the samples.The electrical resistivity was measured from ambient temperature (300 K) to 423 K under atmospheric conditions.The uniformity of the material was assessed in terms of electrical resistivity and thermal conductivity by making multiple measurements on each sample.
Temperature dependent thermal conductivity and Seebeck coefficients in the range of 300 K to 423 K were measured using an atomic force microscope (AFM XE150 Park Systems) in SThM mode with a heating stage in passive mode.The SThM is based on Joule dissipation within the probe-sample system.The SThM experiments were conducted in conjunction with a Peltier heating stage controller, which enables the measurement of the Seebeck coefficient.The thermal conductivity was measured using the Hot Disk Transient Plane Source method (with a Hot Disk TPS 2500 S).

B. Electrical resistivity measurement
The temperature dependence of the electrical resistivity of silica aerogels with NiNPs concentrations of 0, 500, and 700 ppm were investigated (see figure 1a).Median values of electrical resistivity were found to be in the range of 10 13 , 10 10 and 10 7 Ωm at room temperature (300K) for undoped silica aerogel, silica aerogel with 500 ppm NiNPs and silica aerogel with 700 ppm NiNPs, respectively.The electrical resistivity was in the range of 10 6 and 10 5 Ωm for silica aerogel having 500 ppm NiNWs and 700 ppm NiNWs, respectively (see figure 1b).The electrical resistivity of the silica aerogel is reduced by six orders of magnitude when embedded with 700 ppm NiNPs compared with undoped silica aerogel, whereas the 700 ppm NiNW sample exhibits eight orders of magnitude improvement compared with undoped silica aerogel.The silica aerogel embedded with 500 ppm NiNPs exhibits three orders of magnitude resistivity reduction compared with the undoped silica aerogel and the silica 500 ppm NiNW sample reduces resistivity by seven orders of magnitude compared with undoped silica aerogel.
The variation in electrical conductivity with weight fractions (Wt%) of NiNPs and NiNWs in silica aerogels is shown in figures 1c and 1d, respectively.There is a six orders of magnitude increase in the electrical conductivity when the weight fraction of NiNPs in silica aerogels is increased from 0 to 4.9 Wt%.The electrical conductivity improves by eight orders of magnitude when the weight fraction of NiNWs in silica aerogels is increased from 0 to 1.2 Wt%.The electrical conductivities of composite silica samples have been analyzed in relation to the threshold weight concentration of NiNPs and NiNWs fractions.The scaling law of percolation theory was used, σ ∝ (W − W c ) t , where W is the weight fraction of the filler, W c is the critical weight fraction of the filler and σ is the electrical conductivity. 20,21The conductivity exponent, t, generally reflects the dimensionality of the system, and for example in carbon systems it has typical value of 1-1.3 and 1.6-2.0 for 2D and 3D, respectively. 22In this work, to obtain the best fit we have used values for the conductivity exponent of t∼ 2.2 and t∼2.7 for NiNPs and NiNWs in aerogels, respectively.The insets in figures 1c and 1d represent the best fit to the experimentally measured electrical conductivity data as a function of W − W c , expressed as weight fraction.This analysis reveals a percolation threshold of about 0.08-0.1 Wt% and 0.04-0.06Wt% for NiNPs and NiNWs in silica aerogels, respectively.The calculated percolation thresholds for silica aerogels are in good agreement with the reported percolation threshold for silica aerogel composites. 23Electrical measurements showed that the additional NiNWs and NiNPs in silica aerogels make an effective interaction between fillers and form a conductive percolation network, thereby enhancing the electrical conductivity.
At room temperature, the standard deviations of the median values of nine resistivity measurements for the NiNPs aerogel samples were within 5-6% (using four-point probe techniques) and 5-7% (using Hall Effect equipment).For NiNWs aerogel samples, the standard deviations of the median values of nine resistivity measurements were 4-5% (using four-point probe techniques) and 5-7% (using Hall Effect equipment).The insignificant variations in electrical measurements for all the samples demonstrates that incorporation of NiNPs and NiNWs does not degrade material uniformity, which is essential for a device-grade material.
Furthermore, there is a small decrease in the resistivity of pure silica aerogels as the temperature increases, whereas the aerogels containing NiNPs and NiNWs exhibit a metallic behavior, 24 showing an increasing resistivity with increasing temperature (see figure 1a).The electrical resistivity of silica aerogel embedded with 700 ppm NiNPs increases by approximately 250% at 420 K compared with its resistivity at 300 K, whereas the increase is approximately 200% for silica aerogel embedded with 500 ppm NiNPs.For silica aerogel embedded with NiNWs (see figure 1b), the increase in the electrical resistivity at 420 K compared with 300 K is approximately 240% for 700ppm NiNWs and 210% for 500ppm NiNWs.The electrical resistivity of pure silica aerogels decreases by 16% at 420 K compared with that at 300 K.
SThM was used to measure the Seebeck coefficient (s) for all silica aerogel samples.The Seebeck coefficient was used to extract the sample thermoelectric power factor, s 2 σ.The Seebeck coefficient of a material is the measure of the magnitude of the induced voltage in response to a temperature change.For all, pure silica aerogel, silica aerogel embedded with 500 ppm NiNPs and 700 ppm NiNPs, the median values of the Seebeck coefficient at room temperature (300 K) was ∼ 1.824 × 10 −3 VK −1 .The median values of room temperature Seebeck coefficient are 1.814 × 10 −3 VK −1 and 1.824 × 10 −3 VK −1 for silica aerogel with 500 ppm NiNWs and silica aerogel with 700 ppm NiNWs, respectively.The standard deviation of the median values of the Seebeck coefficient from nine measurement locations of all NiNP and NiNW samples are within 8-9% and 7-9%, respectively, showing good material uniformity.
The thermoelectric power factor of the samples is calculated for different temperatures and the data is shown in figure 2a.The NiNW samples exhibit a large improvement in power factor compared with NiNP samples.The silica with 700 ppm NiNWs shows a 10 10 improvement in power factor compared with the undoped silica aerogel sample while this improvement is 10 9 for silica with 500 ppm NiNWs.Figure 2a also shows there is a 10 6 improvement in power factor due to the increase in NiNP concentration from 0 to 700 ppm, while a 10 3 improvement in power factor is observed as the density of the NiNPs increases from 0 to 500 ppm.These results confirm previous reports that nanostructuring is a promising method to improve thermoelectric properties of materials. 25anostructures provide a chance to disconnect the linkage between thermal and electrical transport by introducing new scattering mechanisms. 25

C. Thermal conductivity measurement
The thermal conductivity of the samples was measured using SThM and the results were validated with a HotDisk analyzer.Data for temperature dependent thermal conductivity measurements are presented in the figures 2b, 2c and figures 2d, 2e for the NiNPs and NiNWs samples using SThM and HotDisk analyser, respectively.There is an excellent agreement between the two techniques.At 300 K the thermal conductivity remains constant as the density of NiNPs increases from 500 ppm to 700 ppm and there is only a slight increase in thermal conductivity (0.005 W.m -1 .K -1 ) compared with pure silica aerogel.The median values of thermal conductivity from nine SThM measurements around the samples are 25 × 10 −3 Wm −1 K −1 , 30 × 10 −3 Wm −1 K −1 and 30 × 10 −3 Wm −1 K −1 for the pure silica, silica with 500 ppm NiNPs and silica with 700 ppm NiNPs, respectively.The median value of thermal conductivity from nine HotDisk analyzer measurements are 25 × 10 −3 Wm −1 K −1 , 29 × 10 −3 Wm −1 K −1 and 30 × 10 −3 Wm −1 K −1 for the pure silica, silica with 500 ppm NiNPs and silica with 700 ppm NiNPs, respectively.
The thermal conductivity of the silica with 500 ppm and 700 ppm NiNWs is the same at 300 K.The median value of thermal conductivity from nine SThM measurements is 41 × 10 −3 Wm −1 K −1 for both samples.Using the HotDisk analyser, the median value of nine measurements is 43 × 10 −3 Wm −1 K −1 for the NiNW samples.The thermal conductivity of NiNWs embedded in silica aerogel is 10 × 10 −3 Wm −1 K −1 higher than NiNPs embedded in silica aerogel.The lower thermal conductivity in the NiNPs sample could be due to grain boundary scattering compared with the NiNWs sample, since the grain boundary scattering plays a major role in conductivity decrease as the size approaches the electron mean free path. 26he thermal conductivity of the pure silica aerogel is in good agreement with values reported in literature. 27The improvement in electrical resistivity by incorporating NiNWs, while simultaneously maintaining a low thermal conductivity, demonstrates the huge potential of these NiNWs embedded in aerogel as thermoelectric materials.
The total thermal conductivities of the NiNWs and NiNPs silica composites are a combination of lattice and electronic thermal conductivity. 28It is essential to distinguish between the lattice and electronic thermal conductivity to understand the role of electrons and phonons in thermal conduction.The electronic component of the thermal conductivity k e was calculated by the Wiedemann-Franz equation, k e = LT /ρ, where T is temperature and L is the Lorentz number, L = 2.44 × 10 −8 V 2 K −2 . 27he electronic component of the thermal conductivity has a negligible effect on the total thermal conductivity, because it is very small of the order of 10 −19 Wm −1 K −1 in the undoped silica aerogel at 300 K. Therefore, it is assumed that the measured thermal conductivity of undoped silica aerogel is a lattice (effective) thermal conductivity.
As temperature increases from 300 to 420 K in the silica samples, the phonon mean free path increases and the lattice vibrations dominate the thermal conductivity.Consequently, the thermal conductivity increases and causes a reduction in the Seebeck coefficient of both the pure silica aerogel and the silica aerogel embedded with NiNPs and NiNWs.The increase in thermal conductivity at elevated temperatures is substantially larger for the silica aerogel embedded with NiNPs and NiNWs than for the pure silica aerogel.This could be due to radiative heat transfer. 29Heat transfer through radiation takes place in the form of electromagnetic waves, which is a consequence of the thermal agitation of the material's molecules or charged particles.For the pure silica aerogel at 420 K the thermal conductivity is 20% higher than the thermal conductivity at 300 K.However, for the silica aerogel embedded with 500 ppm NiNPs it is 43% larger than at 300 K, and for the silica aerogel embedded with 700 ppm NiNPs it is 50% larger that at 300 K.At 400 K the thermal conductivity of silica aerogel embedded with 700 ppm NiNPs is increased by 7% compared with the 500 ppm NiNPs silica.For the silica aerogel embedded with NiNWs, the thermal conductivity is 9% larger at 420 K compared with thermal conductivity at 300 K for both the 700 ppm and 500 pm samples.This increase in thermal conductivity with increase in temperature is due to the domination of lattice vibrations that are responsible for heat conduction.
Water presence inside the aerogel pores can have effect on thermal conductivity, so the thermal conductivity of the silica (700 ppm NiNPs) was measured using SThM, firstly as the temperature was increased from 300 to 420 K and then when the temperature was decreased from 420 to 300 K.There is a 2 × 10 −3 Wm −1 K −1 difference in measured thermal conductivity at 300 K and a difference of 3 × 10 −3 Wm −1 K −1 at 410 K.The silica aerogel sample was placed in contact with the heating stage for 30 minutes prior to performing the measurement to obtain the thermal equilibrium condition.However, the thermal conductivity sweeps at 423 K was performed at the same time.The figure contains the thermal conductivity sweep measurement is presented in the figure 2f.The change in thermal conductivity measurements is likely to be due to the evaporation of water inside the pores when the temperature starts to increase.The increased thermal conductivity at higher temperature explains the reduced power factor observed at higher temperature in figure 2a.
All measured thermal and electrical parameters have been used to calculate the overall thermoelectric figure of merit ZT (see figure 3).There is an improvement in ZT of 3 and 6 orders of magnitude for silica with 500 ppm and 700 ppm NiNPs, respectively, compared with pure silica at room temperature (300 K).However, the corresponding ZT improvement for silica NiNW samples are 8 and 9 orders of magnitude for silica with 500 and 700 ppm NiNWs, respectively.Room temperature ZT values reached 5.27 × 10 −15 , 4.8 × 10 −12 , and 4.8 × 10 −9 , for pure silica, silica with 500 ppm NiNPs and silica with 700 ppm NiNPs, respectively.ZT reached 6.07 × 10 −7 and 6.49 × 10 −6 for the silica 500 ppm NiNWs and 700 ppm NiNWs samples, respectively.The results show that the thermoelectric properties of silica aerogel material can be improved further by utilizing silica aerogel with an increased NiNW concentration.This may decrease electrical resistivity further, thereby enabling this material to become a potential thermoelectric material in the future.

III. CONCLUSION
The thermoelectric properties of silica aerogels with different concentrations of NiNPs and NiNWs were examined.Very low lattice thermal conductivity was achieved which appeared relatively insensitive to NiNP and NiNW concentration (up to 700 ppm).However, the electrical resistivity was reduced by 6 orders of magnitude as the concentration of the NiNPs increased from 0 ppm to 700 ppm and it was reduced by 8 orders of magnitude as the concentration of the NiNWs increased from 0 ppm to 700 ppm.Analyzing the embedded silica aerogel samples in relation to the NiNW and NiNP weight fraction to silica aerogel revealed the formation of a conductive network during the material preparation, which causes the enhanced electrical conductivity.The thermoelectric power factor of the materials consequently increased by 10 orders of magnitude as the NiNW concentration increased from 0 to 700 ppm and by 6 orders of magnitude for the NiNP samples up to 700 ppm.The study also confirmed that incorporating NiNPs and NiNWs does not degrade the uniformity of aerogels.Additional studies to increase the concentration of NiNWs beyond 700 ppm are required to improve electrical resistivity further and determine whether silica aerogel can become a viable thermoelectric material.

FIG. 1 .
FIG. 1.(a) The electrical resistivity temperature dependence for silica aerogel at two different concentrations (500 and 700 ppm) of NiNPs and pure silica aerogel (0 ppm) using two different techniques; (b) The electrical resistivity temperature dependence for silica aerogel at two different concentrations (500 and 700 ppm) of NiNWs using two different techniques; Variation in electrical conductivity with weight fractions of embedded silica at a fixed weight fraction.Inset: variation in electrical conductivity with (W − W c ), NiNPs (c); NiNWs, (d).

FIG. 2 .
FIG. 2. (a) Thermal properties showing the enhancement in power factor compared with pure silica aerogel (0 ppm) for the NiNP and NiNW samples; (b) thermal conductivity of the silica aerogel embedded with NiNPs samples by SThM; (c) thermal conductivity of the silica aerogel embedded with NiNWs samples by SThM (d) thermal conductivity of the silica aerogel embedded with NiNPs samples by HotDisk analyzer (e) thermal conductivity of the silica aerogel embedded with NiNWs samples by HotDisk analyzer (f) effect of water on thermal conductivity using SThM for silica aerogel embedded with 700 ppm NiNP.

FIG. 3 .
FIG. 3. Thermoelectric figure of merit (ZT) of the samples at different temperatures.