Thermal conductivity of conventional fluids can be improved using two approaches i.e. usual suspensions and Nanofluids. Nanofluids generally have higher thermal conductivity than usual micro-sized suspensions. However, when toluene, transformer oil, ethylene glycol and water are used as base fluids, the improvement in thermal conductivity achieved by using Nanofluids is better than that could have been achieved through suspensions. Nano-particles can be categorized into three different types i.e. Carbon nanotubes (CNTs), ceramic particles and pure metallic particles. These particles can be used in different combinations to form different Nanofluids.
ANL group investigated the ceramic Nanofluids for the first time. The pioneer work in this field was carried out to determine the thermal conductivity of Cu-O and Al2O3 Nano-particles in the base fluid of ethylene glycol and water. Conventional transient hot-wire (THW) method was used to measure thermal conductivity. Different volumetric fractions i.e. 1-5% were used to analyze thermal conductivity enhancement. It was concluded that thermal conductivity was greatly enhanced by using Al2O3 and Cu-O Nano-particles. For the 4% volumetric fraction of CuO, the increase in thermal conductivity was 20%. Though the increase in thermal conductivity in the case of water as a base fluid was substantial but the rise in thermal conductivity was comparatively higher in the case of ethylene glycol. The increase of 12% and 10% for 3.5 vol. % of Cu-O and 4 vol. % of Al2O3 was achieved in the case of water as a base fluid. These values of thermal conductivity were higher than the values obtained for suspensions in the study carried out by Maxwell in 1962 . The model he used employ point source method to predict the values of effective thermal conductivity as a weighted average of liquid and solid conductivity. Previous work also predicted the thermal conductivity of Al2O3 and Cu-O water Nanofluids. They used smaller Nano-particles of size 23 nm and 28 nm for Cu-O and Al2O3 respectively. They used engine oil (Pennzoil 10W-30) and ethylene glycol as the base fluids and results showed that thermal conductivity is strongly dependent on the size of particle and method of dispersion. Xie et al.  used particles of Al2O3 of size 1.2-3.02 nm to measure the thermal conductivity with water as a base fluid. They studied the effect of base solution as well as the size of particle on thermal conductivity. Though the ceramic particles are not known for their high thermal conductivity enhanced the thermal conductivity of Nano-fluid. Many researches have been conducted on oxide particles-based Nanofluids due to their easy availability. Murshed et al.  predicted the thermal conductivity of TiO2-Water Nano-fluid. He used cylindrical and spherical shaped particles of size 10 × 40 nm and 15 nm respectively. They measured an increase of 33% in thermal conductivity for a volumetric fraction of 5%. The obtained values of thermal conductivity were significantly higher than the values predicted by Hamilton-Crosser model. They also found that, at lower volume fraction, particle concentration and thermal conductivity are non-linearly related to each other.
Though the importance of Nano-fluid was obvious from ceramic-based Nanofluids but the arrival of metallic based Nanofluids was a one-step forward. A study was done to first measure the thermal conductivity of copper-transformer oil based Nano-fluid. Large sized particles of size approximately equal to 100 nm were used in this study and 55% increase in thermal conductivity was obtained at a volumetric fraction of 5%. ANL group used copper Nano-particles of 10 nm in size in ethylene glycol and narrated an increase of 40% with the volumetric fraction of only 0.3 % which was a major breakthrough. This study proved that thermal conductivity is greatly affected by the size of Nano-particle. They used thioglycolic acid to stabilize the Nano-fluid. Predicted values of thermal conductivity are shown in Fig. 3. Patel et al.  used silver and gold particles for the first time in preparing Nanofluids and employed transient hot wire method to measure the thermal conductivity. They reported a noticeable increase in the thermal conductivity for very small concentrations of Nano-particles. The increase in the thermal conductivity for toluene-gold Nano-fluid was 3-7% for a very low volume fraction of only 0.005–0.011% and the increase was around 3.2-5% for 0.0013–0.0026% volumetric fraction in the case of water-gold Nano-fluid. The key reason of increase was the particle size which was around 10-20 nm. Water based Nano-fluid showed more improvement due to use of bare particles whereas thiolate coated Nano-particles were used in toluene-gold Nano-fluid to avoid agglomeration. Though the thermal conductivity of silver is higher but the silver-water Nano-fluid showed relatively lower thermal conductivity due to the large sized Nano-particles (∼60–80 nm). This is important to note that size of particle can dominate the concentration effects or thermal conductivity. Xie et al.  studied the effect of base fluid on thermal conductivity. Glycerol-water mixture, glycol-water mixture, glycerol, pump oil, ethylene glycol, and deionized with α-Al2O3 were used in the study. Base fluids with higher thermal conductivity showed a decrease in thermal conductivity. Study Iron-Ethylene Glycol Nano-fluid and obtained an increase of 18% in the thermal conductivity for the volumetric fraction of only 0.55%. It was observed that thermal conductivity is greatly affected by the sonication of Nano-fluid which indirectly proves the dependence of thermal conductivity on particle size.