Thermal conductivity and structuring of multiwalled carbon nanotubes based nanofluids

Highlights

• Carbon nanotubes – based functionalized and metal nanoparticle decorated.

• Nanofluids for thermal applications with thermal conductivity measurements.

• Artificial neural networks for successful prediction of thermal conductivity.

• Solvation and confinement by molecular dynamics.

• Hydrogen bonding in nanofluids with functionalized nanotubes.

Abstract

the thermal conductivity of OH functionalized multiwalled carbon nanotube and its composites with Ag, Au and Pd in water, ethylene glycol and ethylene glycol-water (60:40 vol%) mixtures is studied using a combined experimental and theoretical approach. The experimental study was carried out in the 0.01 to 0.20 solid mass fraction range from 10 °C to 60 °C. The results show a large effect of the considered solvent on the thermal conductivity as well as increasing values with solid mass fraction and temperature. Artificial neural network (ANN) methods were successfully applied for the prediction of thermal conductivity of the considered nanofluids. The nanoscopic structuring of the studied nanofluids was analyzed by using molecular dynamics simulations, with particular attention to the nanotubes' solvation as well as the changes in the base fluids by the presence of the nanomaterials.

Introduction

The recent developments of nanotechnology have led to a novel class of fluids, which are named nanofluids, being defined as a suspension of nanoparticles (1–100 nm size) in a base fluid [1]. Particular attention has been paid for the application of nanofluids in thermal technologies because of the enhancement of thermal conductivity in comparison with base fluids, thus improving heat transfer operations [[2], [3], [4], [5]]. Likewise, the dispersion of nanoparticles leads to viscosity modification, which is also of great relevance for heat transfer and fluid mechanics applications [[6], [7], [8]]. These suitable thermal and rheological properties of nanofluids allows their application in energy related technologies such as cooling of electronics, solar energy systems, heating of buildings and automotive engine systems [[9], [10], [11], [12], [13], [14]].

Despite of the large number of possible substances for base fluids, water (W) and ethylene glycol (EG) are largely used because of economy. Nevertheless, both W and EG have limitations. Water has high thermal conductivity (0.606 W m⁻¹ K⁻¹ at 20 °C [15]), but their freezing and boiling points are a clear limitation for its usability. EG has a wide temperature range for operation (210 K for liquid range), but its thermal conductivity is low (0.253 W m⁻¹ K⁻¹ at 20 °C [16]). Therefore, the combination of W and EG may lead to a base fluid with suitable properties, i.e. high thermal conductivity and large operation (liquid) range, even considering the formation of eutectic mixtures with low melting point [17]. The modification of thermal properties of base fluids by the addition of suitable nanoparticles, i.e. nanofluids formation, has been the subject of many studies considering nanomaterials such as carbon-based nanoparticles, including single walled (SWCNTs) and multiwalled carbon nanotubes (MWCNTs) [18], graphite [19], graphene [20,21], and graphene oxide [22,23]. The suitability of SWCNTs and MWCNTs has been probed for different applications in areas such as lithium ion batteries, polymer based composite materials, nanoelectronics, as diodes and transistors, and in super-capacitors like electromechanical actuators and sensors [24]. Likewise, SWCNTs and MWCNTs are very attractive considering that they are high aspect ratio nanoparticles (HARNs), which is known to enhance thermal, mechanical and chemical properties of the base fluids [[25], [26], [27], [28]]. The use of SWCNTs and MWCNTs in thermal – related applications stands on their large thermal conductivity [29], which leads to enhanced thermal performances with regard to base fluids. Thermal conductivity enhancement of SWCNTs/MWCNTs-nanofluid has been experimentally probed and increases with concentration and temperature [30]. Sabiha et al. [31] investigated the stability and thermal conductivity of SWCNT nanofluids showing improvement from 2.84% to 36.39% compared to water (base fluid) for 0.05–0.25 vol% and 20–60 °C. Kumaresan et al. [32] measured the thermophysical properties of water–ethylene glycol mixture based MWCNT nanofluids reporting an enhancement of thermal conductivity to a maximum of 19.73% at a MWCNT mass fraction of 0.45%, at the temperature of 40 °C, with a justification on the improvement based on the kinetics of MWCNTs aggregation and liquid layering. An additional improvement of thermal properties of nanofluids may be produced by the consideration of nanocomposites [33], which are produced by the combination of at least two different nanoparticles into one. It should be remarked that the thermal conductivity of HARNs such as SWCNTs/MWCNTs is in the range of 2000–6000 W K⁻¹ m⁻¹ [29] whereas for metallic or oxide spherical nanoparticles the thermal conductivity is remarkably lower (e.g. values in the 40 W m⁻¹ K⁻¹ range for oxides) [34,35]. Nevertheless, more efficient nanofluids are obtained by the consideration of nanocomposites involving SWNTs or MWCNTs combined with low thermal conductivity nanoparticles [36]. For this purpose, recent studies have developed hybrid carbon nanotube/metal or metal oxide nanofluids. Chen et al. [37] examined the thermal conductivity of aqueous MWCNT and nanocomposites of MWCNT decorated with silver nanoparticles (MWCNT-Ag), thus showing thermal conductivity improvements for MWCNT-Ag nanofluids in comparison with those for MWCNT with functionalized surfaces. In another study, Chen et al. [38] examined the water based nanofluid containing hybrid additives of MWCNTs and Fe₂O₃ nanoparticles. Their results displayed that low percentage hybrid fillers loading improve the thermal conductivity of water based nanofluid, because of the good dispersion and interfacial adhesion. Moreover, they reported that the thermal conductivity enhancement of water based nanofluid containing 0.05 wt% MWCNT and 0.02 wt% Fe₂O₃ nanoparticles is 27.75%, which is higher than that of nanofluid containing 0.2 wt% single MWCNT or Fe₂O₃ nanoparticles. Nine et al. [39] prepared and studied water based MWCNT-Al₂O₃ hybrid nanofluids in weight proportion of 97.5:2.5 to 90:10 over 1% to 6% weight concentration. They concluded that hybrid nanofluids with spherical particles show a smaller increase in thermal conductivity in contrast to cylindrical shape particles.

Considering the large number of possible nanofluids as well as the inherent difficulties for nanofluids synthesis and characterization, there is a large interest on the development of different approaches of modeling to predict thermal and rheological behavior of nanofluids accurately [[40], [41], [42], [43], [44], [45], [46], [47], [48]]. The artificial neural networks (ANN) – based methods are one of the suitable methods for predicting and also modeling complicated phenomena such as properties of complex nanofluids [49,50]. Likewise, the thermal properties of nanofluids are largely dependent on the nanoscopic behavior of these systems, for which a suitable approach stands on the use of Molecular Dynamics (MD) simulations [51,52]. MD studies have probed to be suitable for inferring the most relevant information on the structural arrangements, solvation and nanoparticles – base fluid interactions, thus providing a connection between the nanoscopic features and macroscopic relevant thermophysical properties.

In this work a study of hydroxyl functionalized MWCNTs (MWCNT-OH) and their nanocomposites with silver (MWCNT-OH-Ag), gold (MWCNT-OH-Au) and palladium (MWCNT-OH-Pd) nanoparticles in W, EG and EG + W mixture (60:40 vol%) as base solvents (BS) is reported. Thermal conductivity of the considered nanofluids is experimentally studied as a function of nanocomposites concentration and temperature for the three considered base fluids. Likewise, ANN modeling was considered for the prediction of thermal conductivity in the same composition – temperature ranges. Additionally, MD studies on the behavior of the considered nanocomposites particles in the studied base fluids was carried out. The objective of the work is to characterize new types of nanofluids both from experimental and theoretical approaches, thus probing their suitability for thermal operations as well as the nanoscopic roots of their behavior.