Nanofluids for Heat Transfer: Advances in Thermo-Physical Properties, Theoretical Insights, and Engineering Applications

Abstract

Nanofluids, fluids with different suspended nanoparticles, have shown improved thermo-physical properties in recent research outputs, and they have emerged as promising alternatives for the industrial fluids used in numerous heat exchange applications. Much research has been conducted around the world to develop heat transfer fluids with optimum thermo-physical properties with the help of nanotechnology, especially in the 21st century. Following the latest research outcomes, nanofluids with base fluids of industrially used coolants, such as water, engine oil, transformer oil, electronic coolants, etc., have shown imposing thermo-physical properties compared to their base fluid. Identifying the nanofluids with high performances and lesser practical obstacles to be used as heat transfer fluids is vital. This paper reviews the thermo-physical property improvements of nanofluid properties, such as thermal conductivity, viscosity, density, specific heat capacity, and flash point, along with their theoretical models. Recent studies on using surfactants to improve the stability of nanofluids are also included in this review. The next part of the study reviews the latest research outputs on the thermo-physical properties of nanofluids in applications in engineering disciplines. Later, research on molecular dynamics simulations of nanofluids are discussed. As the final section, this paper presents Nanofluid research related to neural network modeling. Cumulatively, this paper presents a comprehensive review of recent nanofluids research, along with theoretical developments. This review is a cumulative study of the recent studies of nanofluid research in different disciplines. Most of the recent reviews focused on specific applications of nanofluids and do not cover the field from the basics of the nanofluids to their applications. However, this review covers all the aspects of the nanofluid field, along with several important engineering applications.

1. Introduction

1.1. The Field of Nanofluids

Nanofluids is a field related to nanotechnology that represents a two-phase system (liquid phase and solid phase), which can be synthesized by dispersing nanoparticles in a base fluid by using different preparation methods. With the introduction of the “nanofluid” concept in 1995 [1,2] the concept of enhancing the thermophysical properties of a base fluid by adding nano-sized particles has found its way into different fields of engineering and science. The effort to enhance the thermal properties of liquids by adding micro-sized particles has been a developing topic for around 100 years. However, there were several technical barriers, such as clogging and the sedimentation of micro-sized particles. However, the development of nanotechnology during the past decades has helped to develop nano-sized particles (metallic, metal oxide, and carbon-based) with unique characteristics that could synthesize nanofluids, overcoming the traditional technical barriers.

Figure 1 shows the number of research articles with ‘‘nanofluids” in their title according to the data from several esteemed publishers, such as MDPI, Elsevier, Taylor & Francis, Wiley, and Nature, and shows the recent development of nanofluid research in different fields. According to previous investigations, nanofluids have enhanced thermophysical properties, such as thermal conductivity, kinematic viscosity, thermal diffusivity, density, specific heat capacity, and even flash point [3,4]. The great potential of nanofluid applications has been demonstrated in many fields, such as biotechnology [5], renewable energy [6], high-voltage engineering [7], mechanical engineering [8], HVAC [9], automobile engineering [10], and even medicine [11].

As mentioned earlier, depending on the engineering application, different base oils and nanoparticles were involved in this research area. Heat transfer fluids, such as water, engine oil, transformer oil, and ethylene glycol are among the most common base fluids. Metallic particles, such as Cu, Ag, and Au, and metal oxide particles, such as CuO, ZnO, Al₂O₃, and TiO₂, are mainly used for enhancing the thermal properties of base fluids. Also, single and multi-walled carbon nanotubes, reduced graphene oxide, and carbon nanodots are the most commonly used carbon-based nanoparticles. In addition to single nanoparticle types, there are also some efforts to apply hybrid nanoparticles to enhance the thermal properties [12].

The preparation of nanofluids basically uses the single-step method and the two-step method [4]. In the single-step method, nanofluid is synthesized, and dispersion in the base fluid happens in a single step. Following this, the stability of the nanofluid is high, with minimum particle agglomeration. However, in the two-step method, the nanomaterial is prepared separately, and these particles are dispersed into the base fluid. The dispersion process can be improved by using magnetic stirring and ultrasonication, as shown in Figure 2. Surfactants can also be added to improve stability.

Research on nanofluids has mainly investigated thermal conductivity and kinematic viscosity as the main thermophysical properties concerning the different weight or volume concentrations of nanoparticles or with temperature. In addition to the above properties, researchers have considered specific heat capacity, density, flash point, pour point, and thermal diffusivity. These properties are investigated by keeping the particle concentration and temperature as the leading independent variables. Some researchers have used different surfactants, different surfactant concentrations, and different sonication times as other parameters to investigate thermal properties [13,14,15]. All these different preparation methods, nanoparticles, and their applications in various engineering fields have made nanofluids a broader area for scientists and engineers.

1.2. Effects and Importance of Thermophysical Properties of Nanofluids

There are different kinds of nanofluid applications in various disciplines, as mentioned in the previous section. However, selective nanoparticles have the ability to significantly improve the thermal properties of selective liquids. Also, the importance of using nanoparticles is that they have the ability to improve such properties without harming the base fluids’ natural properties, which is difficult to achieve by using other bulk materials or liquids [16,17,18]. The incorporation of nanoparticles can improve the performance not only in terms of heat transfer but also in terms of energy storage, eco-friendliness, and system compactness as well. Also, a wide range of nanoparticles can be selected, considering the specific applications.

This review is a comprehensive study of the recent research on the thermophysical properties of nanofluids and various heat transfer applications. First, important thermophysical properties, such as thermal conductivity, viscosity, specific heat capacity, and flash point, are discussed along with the theoretical and experimental studies. Then, the effect of surfactants are discussed in several research studies that consider the stability and other thermal properties. The next chapter discusses the heat transfer applications of nanofluids from transformer oil, solar PV systems, engine oil, electronic cooling, machining fluids, and refrigerator systems. Then, molecular dynamics studies of nanofluids are presented with several examples. As the final specific area, the use of artificial intelligence in nanofluid research is discussed.

2. Thermal Conductivity

2.1. Introduction

According to the enhancement of the thermophysical properties of the fluids using nanoparticles, one of the most promising areas of research in the field of nanoparticles is their use to enhance the thermal conductivity of fluids. Thermal conductivity measures a material’s ability to conduct heat. These nanoparticles are suspended in a fluid; they create a large surface area that can interact with the fluid molecules. This interaction results in an enhancement of the thermal conductivity of the fluid. The enhancement in thermal conductivity can be several times higher than the base fluid alone. This characteristic of nanofluids makes them suitable for use in a wide range of applications, as mentioned in the introduction. It has been reported that different factors affect the thermal conductivity of nanofluids, as shown in Figure 3 [19,20]. The importance and the effect of the factors given in Figure 3 are discussed in detail in the next chapters. However, the effect of temperature and particle concentration on thermal conductivity is inevitable, according to most of the research articles referred to in the review paper. The effect of the other parameters needs to be discussed in depth due to their complex relation with thermal conductivity.

2.2. Nanofluid’s Thermal Conductivity Models/Theories

2.2.1. Effective Medium Theory

The effective medium theory is a theoretical framework that is often used to predict the properties of nanofluids. The basic idea behind the effective medium theory is that the properties of a composite material can be approximated by averaging the properties of the individual components. In nanofluids, the individual components are the fluid and the nanoparticles suspended within it. The effective medium theory is typically used to estimate the thermal conductivity and viscosity of nanofluids based on the properties of the fluid and nanoparticles. The Maxwell model is one of the most widely used models for effective medium theory [21]. The Maxwell model assumes that the nanoparticles are spherical in shape, neglects the dynamic state of the particles, and does not consider the interaction between the nanomaterial and the fluid. The parameters considered for this model are the thermal conductivities of the nanomaterial and fluid and the volume fraction of the nanomaterial of the nanofluid. However, in most cases, Maxwell’s model underestimates the nanofluid’s thermal conductivity. The drawback of Maxwell’s model is that it does not consider the basic properties of nanoparticles, such as the interfacial layer and nanoparticle shape. However, there are several theories that expanded the traditional Maxwell theory by solving some of its drawbacks. Bruggeman’s theory is one of the effective medium theories that is used to model the effective thermal conductivity of nanofluids [22]. The Bruggeman theory is based on the assumption that the thermal conductivity of the nanofluid can be described as a volume-weighted average of the thermal conductivities of the base fluid and the nanoparticles. This theory has been mainly used for spherical particles with higher concentrations. The Hamilton–Crosser model is another version of the traditional Maxwell basic model that considers the effect of the shape of nanoparticles in the thermal conductivity calculation [23].

2.2.2. Interfacial Layering Theory

The formation of an interfacial layer around the nanoparticles that are dispersed in a fluid has been investigated. In Figure 4, it can be seen that liquid molecules’ formation around the nanoparticle creates an interfacial layer. This interfacial layer can be identified as one of the significant mechanisms to increase the thermal conductivity of nanofluids. This layer around the particles exhibits a higher thermal conductivity than the base fluid, and it acts as a thermal bridge between the nanoparticles and the base fluid, according to the theoretical developments of Yu and Choi [24].

2.2.3. Percolation Theory

This theory suggests that the thermal conductivity enhancement is due to the formation of a “percolation” network of nanoparticles within the fluid. When the concentration of nanoparticles is low, they are dispersed randomly in the fluid and do not contribute to thermal conductivity. However, as the concentration of nanoparticles increases, they begin to aggregate to form a network that enhances thermal conductivity. However, these aggregated clusters lead to reduced Brownian particle velocity within the nanofluid. Wang et al. [25] developed a model based on the models of effective medium theory to predict the thermal conductivity of nanofluids considering the clustering of nanoparticles within the nanofluid.

2.2.4. Brownian Motion

This theory suggests that the enhancement of thermal conductivity is due to the random motion of the nanoparticles in the fluid, which helps to transfer heat more efficiently. Figure 5 shows the nanoparticles’ random motion within the fluid to transfer heat. However, in the discussion of the contribution of the Brownian motion, there is a significant impact due to the increase in temperature. The temperature increment decreases the fluid’s viscosity and reduces the agglomeration while increasing the Brownian motion.

2.2.5. Phonon Theory

This theory suggests that thermal conductivity enhancement is due to the increased electron–phonon interactions between the nanoparticles and the fluid. Phonon energy is energy that passes through atoms as vibrational energy. Interfacial layers also enhance the heat transfer through phonons, as well as contributing to Brownian motion [26,27,28].

2.3. Nanofluids’ Thermal Conductivity Theories/Models

Table 1. Nanofluid thermal conductivity models.

Refs.	Model Name	Formula	Remarks
[21,29]	Maxwell	${\displaystyle \frac{k_{nf}}{k_{bf}}}={\displaystyle \frac{k_{np}+{2k}_{bf}+{\varphi (k}_{np}-k_{bf})}{k_{np}+2k_{bf}-\varphi (k_{np}-k_{bf})}}$	Basic thermal conductivity formula for most of the thermal conductivity models. This model is based on the effective medium theory.The Maxwell model is most suitable for low-concentration (<5%) nanofluids with spherical nanoparticles.
[22]	Bruggeman	${\displaystyle \frac{k_{nf}}{k_{bf}}}=\frac{(3\varphi -1)\frac{k_{nf}}{k_{bf}}+[3(1-\varphi )-1]+\surd ∆}{4}$ Δ = [$(3\varphi -1)\frac{k_{nf}}{k_{bf}}+{[3(1-\varphi )-1]}^2+8\frac{k_{nf}}{k_{bf}}$	This model is also based on the effective medium theory of nanofluids. Compared to the other models, this one specifically focuses on higher concentrations of nanoparticles. The Bruggeman model does not depend on the nanoparticle concentration; however, at low concentrations, the Maxwell model and Bruggeman model display identical results.
[23]	Hamilton–Crosser	${\displaystyle \frac{k_{nf}}{k_{bf}}}={\displaystyle \frac{k_{np}+\left(n-1\right)k_{bf}-(n-1){\varphi (k}_{bf}-k_{np})}{k_{np}+{(n-1)k}_{bf}+\varphi (k_{nf}-k_{np})}}$	Developed by considering the shape factor of the nanoparticles. n = 3 for spheres, n = 6 for cylinders.
[28,30]	Wasp	${\displaystyle \frac{k_{nf}}{k_{bf}}}={\displaystyle \frac{k_{np}+2k_{bf}-2{\varphi (k}_{bf}-k_{np})}{k_{np}+{2k}_{bf}+\varphi (k_{bf}-k_{np})}}$	This is the derived form of the Hamilton–Crosser model by using n = 3.
[31]	Pak and Cho	${\displaystyle \frac{k_{nf}}{k_{bf}}}=1+7.47\varphi$	Developed under the assumption that all the nanoparticles contribute to enhancing the thermal conductivity of the nanofluid.
[32]	Lu-Lin	${\displaystyle \frac{k_{nf}}{k_{bf}}}=$ 1 + aϕ + bϕ2	Here, a and b are empirical coefficients that depend on the type of nanoparticles and the base fluid.
[21]	Maxwell–Garnett	${\displaystyle \frac{k_{nf}}{k_{bf}}}={\displaystyle \frac{{(1-\varphi )(k}_{np}+2k_{bf})+3{\varphi k}_{np}}{{(1-\varphi )(k}_{np}+2k_{bf})+3\varphi k_{bf}}}$	This model is also based on the effective medium theory of nanofluids.
[24]	Yu and Choi	${\displaystyle \frac{k_{nf}}{k_{bf}}}={\displaystyle \frac{k_{np}+{2k}_{bf}+2{\varphi (k}_{np}-k_{bf}){(1+\beta )}^3}{k_{np}+2k_{bf}-\varphi (k_{np}-k_{bf}){(1+\beta )}^3}}{k}_{bf}$	This model is a modified version of the Maxwell model. This one considers the effect of the nanolayer on the thermal conductivity.

shows some of the important thermal conductivity models for the nanofluids’ thermal conductivity predictions.

2.4. Nanofluids’ Thermal Conductivity Measurements

Thermal conductivity enhancement is the most commonly investigated physical property of nanofluids by researchers.

Table 2. Recent research results on the thermal conductivity of nanofluids.

Ref.	Nanomaterial and Particle Size	Base Fluid/Surfactant	Concentration	Temp. Range	Maximum Thermal Conductivity Improvement
[33]	MWCNT (multi-walled carbon nanotubes): 5–15 nm (outer diameter) 3–5 nm (inner diameter) CuO: 30 nm–50 nm	Water	0.05% to 0.6 vol.%	25–50 °C	30.38% at 50 °C and 0.6% volume concentration.
[34]	Graphene: 2–18 nm TiO2: 10–25 nm Graphene/TiO2—Hybrid	Water	0.005% to 0.5 vol.%	25–75 °C	27.84% of enhancement for TiO2–graphene/water hybrid nanofluid was observed at a volume fraction of 0.5% and a temperature of 75 °C.
[35]	Graphene oxide	Water	0.01 to 0.5 wt.%	25–60 °C	19.9% enhancement at 25 °C for 0.5 wt.%.
[36]	Al2O3—45 nm Al2O3—500 nm	Water Ice Diathermic oil	0.01% to 0.1 vol. % 0.01% to 0.1 vol. % 0.01% to 0.04 vol. %	293 K (Water) 253 K (Ice) 293 K (Oil)	The highest thermal conductivity for micro-sized and nano-sized alumina particles was discovered at the highest volume ratio in addition to a 0.01% vol. ratio for nanosized Alumina dispersed in ice.
[37]	Fe3O4	Water	0.1% to 3 vol.%	20–55 °C	90% enhancement at a solid volume fraction of 3% and at the temperature of 55 °C.
[38]	Graphene oxide	Water	1.0–4.5 mg/mL	25–50 °C	25.27% enhancement in 4.5 mg/mL mass fraction at 50 °C.
[39]	Fly ash	Water	0–0.5 vol.%	30–60 °C	An enhancement of 11.9% was observed at 60 °C.
[40]	Al2O3—20 nm CeO2—50 nm Al2O3: CeO2—50:50 Hybrid	Deionized water	0.01–0.5% vol. 0.01–0.5% vol. 0.01–0.5% vol.	35–50 °C	Al2O3, CeO2, and their hybrid nanofluids showed 5.3%, 3.3%, and 8.8% maximum enhancements, respectively, at the 0.5% volume concentration at 50 °C.
[41]	TiO2—20 nm	Distilled water (CTAB—cetyltrime thylammonium bromide, SDS—sodium dodecyl sulfate)	0.025–1.25 vol.% 0.025–1.25 vol.%	20 °C	Maximum enhancement of 10% and 8% at a 1.25% vol. ratio for CTAB-treated and SDS-treated nanofluid samples.
[42]	CuO—23 nm CuO—31 nm	Deionized water (SDS, PVP-Polyvinylpyrrolidone)	0.1–0.5 vol. %	Room temperature	Maximum enhancements of 38% and 34% at 0.4 wt.% of SDS and PVP, while CuO volume concentration was 0.5% vol for 23 nm-sized CuO particles.
[43]	Cu	Engine oil	0.2%, 0.5%, and 1 wt. %	40–100 °C	The highest enhancement of 49% was observed for a 1% weight fraction.
[44]	Al2O3—8 to 12 nm TiO2—10 nm	Engine oil: 5W30 (oleic acid)	0.05% Al2O3 + 0.05% TiO2	18–132 °C	The highest enhancement of 8.6% was observed at 100 °C.
[45]	CuO—10 nm TiO2—25 nm	SAE 15W40 engine oil	0.1–1 wt. %	25–50 °C	There was a 21.84% enhancement for CuO in 1% wt. and a 20.2% enhancement for TiO2 in 1% wt.
[46]	ZnO	SAE 50 engine oil	0.125–1.5 vol. %	25–55 °C	The maximum enhancement of 8.74% was obtained at the volume fraction and temperature of 1.5% and 55 °C, respectively, compared to the base fluid at the same temperature.
[47]	MWCNT Graphene nanosheets	15W50 engine oil	1.5 wt.% 0.5 wt.%	40 °C	The highest enhancement, 77%, was obtained at 2 wt.%.
[48]	Al2O3–MWCNT hybrid	10W40 engine oil	0.05–1 vol. %	25–65 °C	Maximum enhancement of 30.35% was obtained at 1 vol% at 65 °C.
[49]	Graphene nanoplate: 10–20 nm	5W30 engine oil	0.15 wt. %	-	29.9% highest enhancement was observed at 0.15 wt.%.
[50]	WO3—23 to 65 nm MWCNT—external: 20 to 30 nm Internal: 5 to 10 nm hybrid	40W10 engine oil	0.05–0.6 vol. %	20–60 °C	The maximum enhancement of 19.85% was recorded at 60 °C temperature and 0.6% vol. ratio for the hybrid nanofluid.
[51]	Ag–WO3 hybrid	Transformer oil	1–4 wt.%	40–100 °C	The maximum enhancement of 41% was reported at 4 wt.% and 100 °C./
[52]	SiO2–graphene hybrid	Transformer oil	0.01–0.08 wt.%	20–100 °C	The maximum enhancement of 80% was reported at 0.04 wt.%, pH 9, and 100 °C.
[53]	Amorphous graphene sheets	Transformer oil	0.0012–0.01 wt.%	35, 45, 55 °C	30% maximum enhancement was observed at 0.01 wt.%, 55 °C.
[54]	TiO2 SiC	Natural ester oil	0.004 wt.% 0.004 wt.%	25–90 °C	25% maximum enhancement at 40 °C; 58% maximum enhancement at 40 °C.
[55]	Hexagonal boron nitride (h-Bn)	FR3 Insulating oil	0–0.1 vol.%	25 and 90 °C	Enhancement of 14% at 0.1 vol.% at 90 °C.
[56]	Graphene oxide (GO) nanosheets	Cottonseed oil (SDS)	0.01, 0.02, 0.03 and 0.05 wt. %	45, 60, 75, 90 °C	The highest enhancement is 36.4% at 0.05 wt.% at 65 °C.
[57]	Graphene–carbon nanotube (Gr–CNT) hybrid	Ethylene glycol	0.0175, 0.035, 0.0525, and 0.07 vol.% (1:1 ratio for Gr and MWCNT)	30 and 50 °C	0.07 vol.% Gr–CNT hybrid material showed the maximum enhancement of 18 and 50% at 30 and 50 °C.
[58]	Functionalized multi-walled carbon nanotubes together with Fe3O4	Ethylene glycol	0 to 2.3 vol.%	25–50 °C	The highest enhancement of 29.7% was reported at 2.3 vol.% at 50 °C.
[59]	SiO2	Liquid paraffin (Oleic acid)	0.005 to 5 wt.%	25–70 °C	The highest enhancement was 38% at 5 wt.%, 70 °C.
[60]	Fe3O4	Liquid paraffin (Oleic acid)	0.005–0.03 vol.%	20–90 °C	The greatest thermal conductivity enhancement (28.49%) was obtained at 90 °C and 0.03 vol%.
[61]	MWCNTs	Kapok seed oil	0.1 wt.%	30–90 °C	6.15% enhancement was observed at 90 °C.
[62]	MXene (Ti3C2)	Silicone oil	0.05, 0.08, and 0.1 wt.%	25–125 °C	64% improvement was found for the 0.1 wt.% concentration at 150 °C.

represents some of the recent studies of nanofluids’ thermal conductivity. In these studies, the most focused variables are the base fluid type, nanoparticle type, surfactant type, concentration, and temperature.

3. Viscosity

3.1. Introduction

The fluid viscosity is closely associated with its ability to reduce friction on the surface of solid body contacts. Generally, fluid with the least viscosity value is desired, which forces the two moving surfaces apart to achieve “fluid bearing” conditions in contrast to too viscous a fluid, which will require a large amount of energy to move [63]. Viscosity strongly depends on temperature, pressure, and shear rate. This phenomenon’s occurrence is due to the molecular distances among molecules in fluids. When changing the temperature of the liquid, the molecules vibrate more rapidly at random and, in doing so, establish a space around them that is proportional to their kinetic energy. Furthermore, the viscosity must be high enough to maintain a lubricating film at different temperatures while remaining low enough to flow around machinery parts under all conditions.

3.2. Nanofluid’s Viscosity Models/Theories

Several models have been developed to explain the viscosity of nanofluids, such as theoretical models for the thermal conductivity of nanofluids as given in the

Table 3. Nanofluid viscosity models.

Ref.	Model Name	Formula	Remarks
[64]	Einstein	${\displaystyle \frac{{\mu }_{nf}}{{\mu }_{bf}}}=1+2.5\varphi$	Basic viscosity model for nanofluid viscosity. This model assumes the solid shape of the nanoparticles and the low volume fraction of the nanofluid.
[65]	Brinkman	${\displaystyle \frac{{\mu }_{nf}}{{\mu }_{bf}}}={(1-\varphi )}^{-2.5}$	This is an extended version of the famous Einstein model. This model is also based on the assumption of the spherical shape of the nanoparticles.
[66]	Krieger–Dougherty	${\displaystyle \frac{{\mu }_{nf}}{{\mu }_{bf}}}={\left(1-{\displaystyle \frac{\varphi }{{\varphi }_m}}\right)}^{-\eta {\varphi }_m}$	This model covers several characteristics of nanoparticles. η is the intrinsic viscosity of nanoparticles, and ϕm is the maximum concentration at which the flow can occur.
[67]	Batchelor	${\displaystyle \frac{{\mu }_{nf}}{{\mu }_{bf}}}=(1+2.5\varphi +{6.5\varphi }^2)$	The effect of the interactions among the nanoparticles has been encountered for this model.
[68]	Lundgren	${\displaystyle \frac{{\mu }_{nf}}{{\mu }_{bf}}}=[1+2.5\varphi +(25/4){\varphi }^2+f({\varphi }^3)]$	This model considers one of the most important properties of nanofluids, namely the Brownian motion. The bulk stress created by the particles has also been taken into account.

.

Einstein’s model is the most popular model that explains the viscosity that is mainly used for low-concentration nanofluids with spherical particles [64]. This model predicts that the viscosity of a nanofluid only depends on the volume fraction of the nanoparticles in the fluid. Most of the other viscosity models are expanded models of Eintein’s initial models, and most of them use the nanoparticle volume ratio as the only variable in the equations.

3.3. Viscosity Measurements

Table 4. Recent research studies on the viscosity of nanofluids.

Ref.	Nanomaterial and Particle Size	Base Fluid (Surfactant)	Concentration	Temp. Range	Maximum Viscosity Improvement
[4]	Fly ash— 11.5 nm	Water	0.1–0.5 vol.%	30 to 50 °C	Maximum improvement is 13% at a temperature of 30 °C and a 0.5% volume concentration.
[7]	Amorphous graphene quantum dots	Transformer oil	0.001 wt.%	20 to 80 °C	Maximum viscosity is observed at 20 °C and decreases with temperature. Nanoparticles do not cause an increase or decrease in viscosity.
[13]	Graphene nanoplatelets TiO2	Distilled water and ethylene glycol (CTAB)	0.1–0.025 wt.%	30 to 70 °C	Mono nanofluid (graphene nanoplates nanofluid) with 0.1 wt% showed the highest viscosity, 32.54% at 40 °C.
[14]	Al2O3 CuO	Water	0.05–0.15 wt.%	-	Both Al2O3 and CuO increase the viscosity. The concentration of SDBS (sodium dodecylbenzene sulfonate) also improves the viscosity of the nanofluid.
[15]	Graphene nanoplatelets	Deionized water (SDS)	0.01–0.1 vol.%	10 to 70 °C	The highest enhancement of 4.9% was observed at 0.1 vol.% and 10 °C suspension.
[27]	Ag–Cu alloy	Hydrocarbon rotary pump oil	0.003–0.015 vo.%	-	Viscosity reached 100 mPas at 0.003 and 0.015 vol.% compared to the 98.5 mPas viscosity of the pure oil.
[35]	Graphene oxide	Distilled water	0.01–0.5 wt.%	25–60 °C	A slight increment of the viscosity from 0.01 wt.% to 0.1 wt.%. A severe increment from 0.1 wt.% to 0.5 wt.%
[43]	Cu	Engine oil	0.2–1 wt.%	40 to 100 °C	A 37% improvement was observed for a 1% weight fraction at 40 °C.
[45]	CuO, TiO2	SAE 15W40 engine oil	0.1–1 wt.%	40 to 100 °C	The highest enhancements are 10.88% and 8.8% for CuO and TiO2, respectively, at 1 vol.%, 40 °C.
[47]	MWCNTs/GNs hybrid MWCNTs	15W50 engine oil	0.5–2 wt.% 0.5 wt.%	40 and 100 °C	The highest enhancement in kinematic viscosity was recorded at 2 wt.% as 73.4% and 76.8% at 40 °C and 100 °C, respectively.
[49]	Graphene nano-plate Average lateral dimension (x and y) length ≤ 5 µm	SAE 5W30 engine oil	0.03–0.15 wt.%	40 and 100 °C	Maximum improvement was 10.5% at 0.15 wt.% of GNs at 40 °C.
[52]	Hybrid SiO2–graphene	Transformer oil	0.01–0.08 wt.%	40 and 100 °C	The highest enhancement of hybrid SiO2–graphene-based nanofluids was 29.7% at 40 °C..
[59]	SiO2: 20 nm diameter (approximately)	Liquid paraffin (oleic acid)	0.005–5 wt.%	25 to 70 °C	The highest enhancement of 495% was reported at 5 wt.% and 70 °C.
[61]	MWCNT	Kapok seed oil	0.1 wt.%	30 to 90 °C	-
[62]	MXene (Ti3C2)	Silicone oil	0.05–0.1 wt.%	25 to 125 °C	No noticeable change with respect to nanoparticle weight concentration.

shows several recent studies related to nanofluid viscosity measurement. The main variables that have been used for these measurements are the base fluid type, nanoparticle type, concentration, and temperature. Viscosity, along with thermal conductivity, makes a crucial contribution to determining the heat transfer capability of a fluid.

4. Specific Heat Capacity

4.1. Introduction

Specific heat capacity is a measure of the amount of heat energy required to raise the temperature of a substance by a given amount. Specific heat capacity depends on the type of material, which makes some materials better at storing heat.

In most cases, the specific heat capacity of nanofluids is typically higher than that of their base fluid due to the presence of nanoparticles in the fluid. The higher specific heat capacity of nanofluids is due to the enhanced heat transfer resulting from the increased thermal conductivity and the enhanced convection caused by the suspended nanoparticles. However, these enhancements depend on the type of material, base fluid, and the nanomaterial concentration.

4.2. Specific Heat Capacity Measurements

Several research studies have been mentioned related to specific heat capacity, as given in

Table 5. Recent research works on the specific heat capacity of nanofluids.

Ref.	Nanomaterial and Particle Size	Base Fluid/Surfactant	Concentrations	Specific Heat Capacity Improvement
[15]	Graphene nanoplates have a thickness from 2 to 8 nm and a diameter from 4 to 12 µm	Water (SDS)	0.01, 0.05, and 0.10 vol.% with graphene: SDS ratios are 0.5:1, 1:1, and 1.5:1	0.1 vol.% of graphene nanoplatelets with a 1.5:1 surfactant ratio caused the thermal property of the base fluid to be reduced to values between ~28.12% (70 °C).
[69]	CeO2	Water	0.01, 0.05, 0.1, 0.2, and 0.3 vol.%	0.3% volume fraction had specific heat of about 5% lesser than the base fluids at the temperature of 35 °C.
[70]	SiO2—10 nm, 20 nm, 30 nm	Base salt (NaNO3 and KNO3)	0.5, 1.0, 1.5, and 2.0 wt.%	Highest enhancement of 26.7% for 20 nm SiO2.
[71]	Coal fly ash nanoparticles—14 nm	Water (Triton—X 100)	0.1, 0.3, and 0.5 vol.%	A 21.19% decrease was observed for 0.5 vol.% at 30 °C.
[72]	ZnO CuO	Deionized water (EBT—eriochrome black T, OA—olylamine)	0.1 wt.%	No comparison has been performed with the base fluids. The parameters considered are temperature and sonication time. However, a recognizable change was not observed with respect to sonication time.
[73]	MgO—diameter: 25–45 nm TiO2—diameter: 18–23 nm	Water (SDS)	0.1–0.5 vol. % with MgO: TiO2 = 50:50, 80:20, 20:80, 60:40, and 40:60	80 wt.% MgO—20 wt.% TiO2 decreased by 1.08% with a solid volume concentration of 0.5%.
[74]	CNTs (carbon nanotubes)—diameter: 10 to 20 nm. Length ranges from 10 to 30 μm.	Water (SDS, PVP)	0.1, 0.3, 0.5 and 1 wt.% with CNT: Surfactant = 1:0.5 and 1:1	At 60 °C, enhancements of 57%, 61%, 63%, and 65% at 0.1, 0.3%, 0.5%, and 1 wt.% were discovered, respectively. Notably, the type of surfactant did not affect the specific heat capacity enhancement.
[75]	Al2O3—ZnO Al2O3—29 nm ZnO—70 nm	Water	0.33, 0.67, 1%, 1.33% and 1.67% with ratio of Al2O3-ZnO = 1:2, 1:1, 2:1	Nanofluids at a 2:1 mixture ratio have a maximum viscosity increase of 96.37% and maximum specific heat decrease of 30.12% at a temperature of 25 °C and a volume concentration of 1.67%
[76]	CuO + MWCNT: average diameter = 20, 30, 40, 50 nm MgO + MWCNT: average diameter = 20, 30, 40, 50 nm SnO2 + MWCNT: average diameter = 20, 30, 40, 50 nm	Deionized water (CTAB)	0.25, 0.50, 0.75, 1.00, 1.25, 1.50, 80:20 each (by weight ratio) with the addition of the CTAB surfactant at 3:2	The maximum decrease in specific heat capacity is about 12.84% at φ = 1.50%, 25 °C at an average particle size of 50 nm and 20 nm for CuO + MWCNT. The maximum decrease in specific heat capacity compared to the base fluid has been reported at 25 °C, φ=1.50%, and average size 20 nm h is about 15.09%. The highest decrease of about 13.23% was observed at 1.50 vol%, 25 °C, and 20 nm for SnO2 + MWCNT.
[77]	GO: Al2O3 diameter = 30 ± 5 nm	Water (SDS)	0.05, 0.07, 0.01, 0.12, and 0.15 wt.%	The maximum reduction ratio was almost 7% at 0.15 wt.% at 20 °C
[78]	CNT—diameter: 10 to 20 nm. Length ranges from 10 to 30 μm	Water (SDS, PVP)	0.1, 0.3, 0.5, and 1 wt.% with CNT: surfactant = 1:0.5 and 1:1	At 60 °C, enhancements of 57%, 61%, 63%, and 65% at 0.1, 0.3%, 0.5%, and 1 wt.% were discovered, respectively. Notably, the type of surfactant did not affect the specific heat capacity enhancement.

.

5. Flash Point

5.1. Introduction

The flash point is the minimum temperature at which a liquid gives off enough vapor to ignite. It is an important concept because liquids with low flash points are considered to be highly flammable and pose a significant fire risk. It is also important to identify that the flash point of a liquid can be influenced by a number of factors, such as the presence of impurities, the temperature and pressure of the surroundings, and the method used to measure the flash point. The presence of nanoparticles can enhance the thermal conductivity of the nanofluid, which, in turn, can affect its flash point. However, the exact effect on the flash point depends on several factors, such as the size, concentration, type of nanoparticles, and base fluid. The flash point is measured mainly using the ASTM D93 [79]. It has been reported that the increment of the flash point and the decrement of the flash point change with the addition of the nanoparticles. The decrement of the flash point can be explained by an increment of the thermal conductivity, which causes an increase in the heat transfer from the liquid to the vapor phase, which leads to rapid vaporization and ignition. Then, the increment of the flash point can be explained by the increment of the viscosity of the nanofluid, which slows down the vaporization.

5.2. Flash Point Measurements

Table 6. Recent research works on the flash point of nanofluids.

Ref.	Nanomaterial and Particle Size	Base Fluid	Concentration	Flash Point Improvement
[78]	CNT diameter: 10–40 nm; length: 20 μm	Mobil gear 627 paraffinic oils	0.1, 0.5, 1, and 2 wt.%	The flash point for Mobil gear 627 and paraffinic oils was increased by about 13 and 25%, respectively.
[80]	SiO2, CQD (carbon quantum dots)	Mineral oil	0.01 wt.%	3.33% improvement.
[81]	ZnO—diameter: 30 nm	SAE50 engine oil	0.1 to 1.5 vol.%	7.2% improvement at 1.5 vol%.
[82]	CQD: SiO2 hybrid, CQD 2.5 nm, SiO2: 15–40 nm	Transformer oil	0.01, 0.05, and 0.1 wt.%	6.67% for CQD–SiO2 hybrid nanofluid.
[83]	TiO2, ZnO, Al2O3	Virgin mineral oil	-	TiO2—14.7%, ZnO—12.3%, Al2O3—2.3%.
[84]	Exfoliated hexagonal boron nitride (Eh-BN)	Mineral oil Pongamia pinnata oil	0.01 wt.%	30% for mineral oil-based nanofluid and 3.6% for Pongamia pinnata oil.
[85]	TiO2, Al2O3, MoS2	Transformer oil	0.025 wt.%	TiO2—10.56, Al2O3%—7.0%, MoS2—4.2%.
[86]	MWCNT	Turbine meter oil	0.05, 0.1, 0.2, 0.3, and 0.4 wt.%	4.44% increment at both 0.3 and 0.4 wt.%.
[87]	Graphene oxide: 2 μm diameter	Engine oil—SAE-50	0.01, 0.25, 0.50, and 1.00 wt.%	8% improvement at 1 wt.%.
[88]	ZnO, diameter: 0.064 nm	Polyol ester oil	0.1, 0.3, and 0.5 wt.%	8% improvement at 0.5 wt.%.
[89]	SiO2, diameter: 30–50 nm	SAE20W40 engine oil	0.3, 0.6, 0.9, 1.2, and 1.5 wt.%	Maximum reduction of 6.97% in flash point at 0.6 wt.%.
[90]	Al2O3	Lube oil stock—60	0.25, 0.65, 1.05, 1.45, and 1.85 wt.%	9.73 increment at 1.85 wt.%.
[91]	FeO3	Kernel palm oil methyl ester	0.10, 0.15, and 0.20 wt.%	The highest decrease of 9.13% observed in the 0.10 wt.% sample.

shows several recent studies of flash points of nanofluids.

6. Comparative Studies of Nanofluids with the Theoretical Models

In addition to the theoretical modeling of nanofluids’ properties and experimental studies of their thermophysical properties, there is a considerable amount of scientific literature that compares the experimental results with theoretical results [29]. Eric C. Okonkwo et al. [92] conducted research using Al₂O_3-–water–and Al₂O₃–Fe–water nanofluids to compare the thermal properties with the theoretical results. In this research, specific heat capacity results were compared with the mixture formula and the Pak and Cho model. In this comparison, mono nanofluid showed results that were similar to the Pak and Cho model, while the hybrid nanofluid displayed results that were similar to the mixture formula. The dynamic viscosity values were compared with Einstein, Brickman, and Batchelor models; the least deviation was recorded with the Einstein model. The thermal conductivity experimental values were compared with the Maxwell, Bruggeman, Yu and Choi, and Maiga models, and all the models displayed considerable deviation from the experimental values. J.A. Esfahani et al. [93] compared thermal conductivity values with the Hamilton and Crosser and Yu and Choi models of Ag-, Cu-, and TiO₂-based insulation oil nanofluids. Viscosity values were also compared with those of the Singh, Brinkman, and Maiga models. Shriram S. Sonawane et al. [94] conducted a comparative TiO₂ nanoparticle-based study with different base fluids, namely water, ethylene glycol, and paraffin oil. This research compared the thermal conductivity experimental values with the Maxwell and Bruggeman models. However, ethylene glycol- and paraffin oil-based nanofluids agreed with the Bruggeman model. In the study conducted by Gianluigi Bovesecchi et al. [95], thermal conductivity results were compared with the Nan, Feng, and Prasher models at varying particle concentrations and temperatures. The Feng model showed a minimum deviation for lower temperatures.

7. Effect of Surfactants on Thermal Properties of Nanofluids

7.1. Introduction

Stability is the main problem that arises during the preparation of the nanofluids. Sometimes, it might be very difficult to create a stable nanofluid sample with an even distribution of nanoparticles. This stability is mainly caused by the type of base fluid combined with the nanoparticle type. Hydrophilic nanoparticles, such as metal oxides, are easily dispersed in polar-type base fluids, such as water. However, hydrophobic nanoparticles, such as graphene-type nanoparticles, are easily dispersed in non-polar base fluids, such as oils. However, in cases of instability in nanofluids, researchers have to apply surfactants to increase the stability, in addition to using magnetic stirring and ultrasonication. The surfactant reduces the surface tension between the base fluids and nano-additives. It also helps to uniformly disperse the nanoparticles in the base fluid [96]. Furthermore, surfactants with various chemical structures perform differently due to their hydrophilic head charge.

7.2. Recent Studies of Surfactants on Nanofluids

Table 7. Recent research work on the surfactants of nanofluids.

Ref.	Nanomaterial	Base Fluid	Surfactant	Thermal Effects and Stability by Surfactants
[14]	Al2O3, CuO	Distilled water	Sodium dodecylbenzene sulfonate (SDBS)	The thermal conductivity increased with surfactant concentration but decreased significantly at greater concentrations. In this study, the higher concentration of nanoparticles decreased the stability (zeta potential), but stability increased with the SDBS concentration.
[15]	Graphene nanoplates (thickness from 2 to 8 nm, diameter from 4 to 12 µm)	Water	Sodium dodecyl Sulfate (SDS)	Even though SDS improves the stability of nanofluids, it lowers their thermal conductivity and specific heat capacity. The lower concentration of SDS stabilized the nanofluid for 24 h while the higher concentration maintained the stability for up to 45 days.
[73]	MgO–TiO2	Distilled water	SDS	No identical change was observed since all the samples were prepared with the same surfactant ratio. In this study, the SDS concentration was maintained consistently while varying the MgO and TiO2 ratio. The highest stability was obtained for the MgO–TiO2 = 8:2 sample.
[74]	CNT	Water	Gum Arabic (GA), polyvinyl pyrrolidone (PVP), sodium dodecyl sulfate (SDS)	Surfactants were used at the 1:0.2, 1:0.5, and 1:1 ratios. It was discovered that the highest stability existed at the 1:0.5 and 1:1 ratios. Among all the surfactants, SDS displayed the highest stability compared to the other two. These surfactants did not affect the thermal properties.
[97]	FeC	Water	Low-viscosity carboxymethyl cellulose sodium salt	There was no effect on thermal conductivity, and the stability of the nanofluid was improved by the surfactant.
[98]	Al2O3	Water	Cetyl trimethyl ammonium bromide (CTAB), SDBS, SDS	Only SDBS produced a stable and greater distribution of nanoparticles in nanofluid at 2:1 ratio. DBS-containing nanofluid showed a slight decrement in thermal conductivity compared to nanofluid containing no surfactant.
[99]	TiO2	Water	CTAB, acetic acid (AA), oleic acid (OA), SDS	Only CTAB and AA produced stable nanofluids for more than 500 h, and the stability was improved with surfactant concentration according to surface tension data. The TEM images also displayed the availability of stable clusters of 147 nm and 207 nm.
[100]	TiO2	Water	CTAB, OA	CTAB produced a more stable and homogeneous nanofluid than OA. It prevented nanoparticle clustering.
[101]	Ag	Water, hexane ethylene glycol	GA GA, OA	Gum Arabic lowered the thermal conductivity of water. The stability of the nanofluids were not focused on in this study.
[102]	Al2O3	Therminol	OA	The nanofluid stability was analyzed using Turbiscan LabExpert equipment and FTIR measurements. It was found that the nanofluid samples prepared at 120 °C showed the highest stability. Surfactants did not affect the thermal conductivity of the nanofluid.
[103]	Mg(OH)2	Water	CTAB, SDS, OA	All the nanofluids samples had recorded zeta potential exceeding 45 mV on the 30th day after the preparation, showing that all the surfactants were suitable for the preparation. Among them, CTAB produced the most stable nanofluid.
[104]	Cu Al2O3	Water	SDBS	The effect of the surfactant was investigated along with the pH value for both nanofluids. The highest stability was obtained for Cu at pH = 9 and for Al2O3 at pH = 8. Surfactants have increased thermal conductivity with a surfactant concentration of 0.1 wt.%.
[105]	TiO2	10W30 engine oil	Triton X	The stability study was performed using a UV–Vis spectrometer, demonstrating an absorbance decrement after 2 h of preparation and electrostatic stability after 168 h. The surfacant improvedthe load-carrying capacity, friction-reducing, and anti-wear abilities of the nanofluid.
[106]	h-BN Al2O3	15W40 diesel engine oil	OA	The visual recordings of the stability of the nanofluids were obtained after 24 h, 72 h, 168 h, and 720 h. However, both nanofluids displayed better stability at 168 h but not at 720 h. The dispersion of h-BN nanoparticles was better than that of Al2O3 nanoparticles. The total acid number of the nanofluid with added h-BN added showed a slight increment.

shows several studies of the effect of the surfactants on nanofluids.

8. Heat Transfer Applications of Nanofluids

8.1. Transformer Oil-Based Nanofluids

There are two main types of transformers based on their insulation type, namely the dry type and the oil-filled type. Oil-type transformers are transformers which use oil as their insulation material as well as coolant. Most oil-filled-type transformers use petroleum-based mineral oil, which is typically obtained from crude petroleum. This material mainly consists of a complex hydrocarbon with a large number of carbon and hydrogen atoms [107]. The failure of a transformer can result in a high economic loss due to power supply interruption. It also takes a long time to fix and costs a lot in terms of repairs. Transformer oil-based nanofluids are suspensions of nanoparticles in transformer oil. These nanoparticles can enhance the thermal conductivity of the transformer oil, allowing for more efficient cooling of the transformer. Additionally, the nanoparticles may also improve the dielectric properties of the transformer oil, making it more resistant to electrical breakdown. The use of transformer oil-based nanofluids is a relatively new technology and is still being researched for its potential benefits and drawbacks. As with other nanofluid applications, it can be identified that researchers are applying metal, metal oxide, carbon, and hybrid nanoparticles to improve the thermal properties of nanofluids. Sadegh et al. [51] have used WO₃–Ag hybrid nanoparticles to enhance thermal conductivity and dielectric strength with different weight ratios. Even though this experiment improves thermal conductivity, adding nanoparticles decreases the dielectric strength of pure transformer oil. TiO₂ is also a common nanoparticle type used with transformer oils, which shows some promising results in enhancing the dielectric and thermal properties [108,109,110]. Mojtaba Parvar et al. [108] investigated the effect of ZnO nanoparticles in the case of thermal properties, such as the thermal conductivity and viscosity in transformer oil for different volume fractions at different temperatures. Another study that was conducted in 2019 [83] compared the effect of several metal oxide nanoparticles (ZnO, TiO_2, and Al₂O₃) by measuring both thermal and electrical properties. Ahmad et al. [7] developed a graphene quantum dots–transformer oil nanofluid that displayed a significant enhancement in thermo-physical properties, such as breakdown voltage, thermal conductivity, viscosity, and flash point. The use of amorphous graphene oxide sheets can also develop the thermal and electrical insulation properties of transformer oils by a superior percentage [53]. Compared to other materials, carbon can be identified as a different material that is represented in different structures with unique properties in the material science world. However, compared to metal oxide-based research, it can be identified that the use of carbon-based materials can create more stable nanofluids while exhibiting superior thermal and electrical properties [7,53,111]. Along with the enhancement of the thermophysical properties of the existing transformer oils, several other research experiments have been carried out to replace mineral oil-based insulation oils with vegetable oil-based nanofluids as a more environmentally friendly solution [54,55,56,84,112]. In some of these studies, vegetable oils have exhibited higher thermal and electrical properties than mineral oils, making them a more suitable solution

8.2. Solar PV System Cooling and Energy Storage of Nanofluids

The photovoltaic system is the main type of different solar power generation systems. In this method, electricity is generated through solar PV panels. Solar radiation is the heat source of solar power generation. The absorption of solar radiation makes solar panels a type of heat exchanger. In addition to electricity generation, solar panels produce heat energy. However, researchers are investigating using this produced energy for various other applications by using solar collectors. These solar collectors absorb the produced heat, which can be used for different applications. This phenomenon can help increase solar PV systems’ efficiency, since their efficiency increases at lower temperatures. There are different types of PVT systems based on the cooling/working material type (air or water), number of fluids, solar collector material type, and the fluid flow method [113,114]. Two methods are employed to achieve efficient heat transfer, which are active and passive methods. The active method has to be used with an external energy source. However, passive methods are recommended because of their lower maintenance and operating costs. In addition to water and air, several other solar collectors use various phase-changing materials (PCMs), such as paraffin, salt hydrates, salt hydrides, and eutectics. These materials can absorb the heat produced by the PV systems, and this stored energy can be used for other applications.

At present, researchers are trying to enhance the efficiency of the PVT systems by applying nanofluids instead of pure water, which have better thermophysical properties. Sekar et al. [115] have used ZnO and CuO nanofluids with and without PCM systems containing paraffin wax as a phase-changing material. They achieved an improved electrical energy output improvement of around 15% with CuO and the PCM system and around 12% with ZnO and the PCM system. Munzer et al. [116] investigated the different thermal properties of TiO₂- and Al₂O₃-based nanofluid in a cooling system of three PV panels, such as surface temperature, power production, and efficiency, with different flow rates and weight ratios. MWCNT is also a promising nanoparticle type used in many thermal applications in various engineering fields. Tong et al. [117] applied MWCNT/Fe₃O₄ to enhance the efficiency of a PVT system with an active cooling system. Also, they compared numerical results generated with finite element analysis software and experimental results. Ali H.A. Al-Waeli et al. [118] used paraffin wax as a nano-PCM and nano-SiC with water and nanofluid in a PVT system. This study’s electrical and thermal efficiencies were 13.7% and 72.0%, respectively. Ahmad Fudholi et al. [119] have achieved around 85–89% energy efficiency by using TiO₂/water nanofluid as a coolant in PVT systems at different weight percentages, solar radiations, and flow rates. Using SiO₂–Al₂O₃ nanofluid is another example of applying hybrid nanofluids in PVT systems [120]. This research was carried out for indoor and outdoor conditions and achieved an overall efficiency of 68.09% and 75.26%, respectively. In a PVT system, Irem Karaaslan et al. [115] CuO/Fe water nanofluid achieved rates of 2.14% and 5.4%, respectively, for electrical and thermal efficiencies. They also compared the efficiency with mono nanofluids and reported that these values were much higher than for mono nanofluids. Yijie Tong [116] conducted a comparative study of MWCNT/Fe₃O₄ hybrid nanofluid compared to Fe₃O₄ nanofluid. MWCNT/Fe₃O₄ hybrid nanofluid achieved around two times the thermal energy efficiency of a mono nanofluid. Irem Karaaslan et al. [121] conducted research to analyze the thermo-physical properties of water-based CuO and Fe nanofluids in VPT systems. The research was conducted for water, CuO–water, and hybrid nanofluids using ANSYS Fluent 18.2 software.

Carbon-based nanofluids have also gained significant attention in solar thermal applications. It has been shown that ultra-stable carbon quantum dot nanofluids can be used as good spectral beam splitters (SBSs) in PVT applications [122]. This research showed that carbon quantum dot (CQDs)-based nanofluids, with polyethylene glycol and glycerol, can improve solar energy absorption, especially in the solar irradiance wavelengths expected from 650 nm to 1050 nm, which is the ideal performance that can be expected from an SBS in PVT systems. The use of carbon quantum dots (CQDs), along with antimony tin oxide (ATO), can also significantly enhance the performance of SBS in PVT applications [123]. Yanqiong Bao et al. [124] used MWCNT/SiC aqueous nanofluid in the solar collectors, achieving 64.7% solar energy absorption efficiency.

In addition to improving the performance of PVT applications, nanoparticles can be used to desalinate water with solar energy. Xiaopeng Bai et al. [125] developed TiN nanospheres for solar desalination processes, considering the high cost of using noble materials. In this research, TiN nanospheres were embedded in highly porous poly (vinyl alcohol) films to function as a photo-thermal material for solar seawater desalination. They reported that the highest evaporation rate was 3.8 km⁻²h⁻¹. Alejandro Espejo Sanchez et al. [126] developed a novel hybrid solar nanophotonic distillation membrane that can simultaneously perform desalination (with a 15–32% improvement compared to the undoped membrane) and electrical power generation of 0.36 W. The membrane was doped with carbon-coated Cu nanoparticles. Dmitrii M. Kuzmenkov et al. [127] analyzed the effect of three different nanoparticles (MWCNT of 49 and 72 nm, Fe₃O_4, and a commercial paste based on carbon nanotubes). All the particles were used at 10% concentration, and the best results were obtained for the 5% wt. concentration of carbon nanotubes.

In addition to conducting performance analysis, several studies have focused on total cost analysis. Shek Rahman et al. [128] evaluated the performance of a tube solar collector (ETSC) using Al₂O₃/water nanofluids. Several thermophysical properties, such as thermal conductivity, specific heat capacity, density, and zeta potential, were evaluated in this research. In this research, the results were compared with previous TiO₂/water-based research, and a significant reduction in cost was reported (68.15%). Muhammad Amar et al. [129] conducted an energy, exergy, and economic (3E) analysis of a flat-plate solar collector using a gallic acid-based multi-walled carbon nanotube in water. The research reported a considerable improvement in the payback period compared to the water-based system. The payback period was 6.228% shorter than using water, with a saving of 321.72 MJ of embodied energy.

8.3. Engine Oil-Based Nanofluids

Motor oil, engine oil, or engine lubricants are any substances used to lubricate engines while they are running. The three types of engine oils are represented in Figure 6.

Furthermore, there are two grades of engine oils, i.e., single-grade and multi-grade, in the field. Multi-grade engine oils are popular because of their ability to change their viscosity depending on different temperatures. Several types of engine oils are prepared for diesel and petrol engines, and a standard grading system grades them according to their viscosity and specific temperature values. When considering a multi-grade engine oil (10W30, for example), it has two viscosity profiles, namely 10 and 30 at cold (40 °C) and hot (100 °C) temperatures, respectively. Reducing friction, cooling, and cleaning are the three major functions of engine oils, and a typical engine has a complex collection of additives that increase its properties. On average, engine oil consists of 15 to 25% additives by weight and 75 to 85% by weight of base oil. Research conducted by Sadegh Aberoumand et al. [43] observed that Cu nanoparticles could enhance the thermal conductivity and viscosity by a significant degree; additionally, it was observed that this nanofluid displayed a non-Newtonian behavior at lower temperatures. Another study showed that the use of Al₂O₃/TiO₂ hybrid nanoparticles can enhance thermal efficiency by 3.9–8.6% [44]. Another important breakthrough that was made in this research was the use of thermogravimetric analysis. It showed that the oxidation of the base oil can be delayed by 54.9 °C by these nanoparticles. SiO₂ nanoparticles can also be a perfect candidate for improving the thermophysical properties of engine oils at different concentrations [89]. The viscosity and thermal conductivity enhancements are especially high compared to the base oil. In addition to metal and metal oxide nanoparticles, carbon-based nanoparticles can also be a perfect candidate for improving the thermophysical properties of engine oils. Research conducted by Bahaa M. Kamel et al. [47] showed that the use of MWCNT nanoparticles along with graphene nanoparticles can also improve many thermal properties. Another interesting research shows that the graphene nanoparticles alone can also increase the thermophysical properties of engine oils [49]. This research has also proved that graphene nanoparticles can decrease the friction of engine oil. Farid Soltani et al. [50] took a different approach by comparing WO₃ and MWCNTs-based engine oils at several volume fractions. The researchers have observed that the MWCNT’s effect is much higher compared to WO₃ nanoparticles, improving the thermal conductivity of engine oils. Chanaka Galpaya et al. [130] conducted an interesting study by applying several nanomaterials, such as fullerene, TiO₂, and Fe₂O₃, in engine oils. The results showed that the effect of fullerene on the thermophysical properties of engine oil is negligible compared to that of the other two nanomaterials. This is an important observation of the neutral behavior of fullerene nanoparticles compared to other carbon nanomaterials.

8.4. Radiator Cooling

In thermal management studies, it is crucial to improve the cooling system performance while reducing the size of systems. In addition to using engine oil-based nanofluids, several research studies have been conducted on vehicle radiator coolants. These research studies basically incorporated water, ethylene glycol, and water–ethylene glycol mixtures [131,132,133]. Aditi et al. [134] tried to improve the coolant performance by incorporating Al₂O₃ and Ag nanoparticles in water and ethylene glycol in CFD simulation using the ANSYS Fluent software. The study evaluated various factors, such as temperature gradient, outlet temperature, and outlet pressure. All the evaluated factors gave positive results for nanofluids. In the research conducted by Tugba Tetik et al. [135], a water–ethylene glycol mixture was used in a 7:3 ratio along with an Al₂O₃–SiO₂–TiO₂ nanoparticle composite in different ratios. It was reported that the heat transfer ratio improved from 14% to 30%. Amr M. Hassaan [136] conducted a similar kind of research by incorporating MWCNT–Al₂O₃ nanoparticles in water using a car radiator. Several parameters, such as heat transfer rate, Nusselt number, and pressure drop, were evaluated in the study, and all of them reported positive results except the pressure drop and the friction factor. Another research study conducted in 2022 used graphene nanoplatelets/cellulose nanocrystals in automobile cooling systems [137]. This research was conducted as an experimental study as well as a simulation. Both samples gave positive results with the increment of the convective heat transfer coefficient (CHTC), overall heat transfer coefficient (OHTC), pressure drop, and Nusselt number. At the end of the research, the researchers proposed the reduced dimensions of the radiator for the given nanofluid. Another study conducted by Waqar Ali et al. [138] studied the potential of using SiO₂–MWCNT nanoparticles in a louvered fin flat-tube radiator while using water as the base fluid. Heat transfer rate, heat transfer coefficient, and Nusselt number were studied in this research. Positive results were achieved when using hybrid nanoparticles, with the maximum enhancement being for SiO₂–MWCNT at an 8:2 ratio. Mahdi Hajiakbari et al. [139] focused on improving the efficiency of a four-stroke diesel engine using TiO₂, SiO_2, and Al₂O₃ nanoparticles in the cooling system. This study has reported a heat transfer rate improvement of 75.1% and 128.1% at 85 °C and 95 °C, respectively. Also, the economic analysis showed an approximate cost saving of USD 2700.

8.5. Electronic Cooling

The usage and development of the electronic industry have been extensive in the recent past. In high-performance applications, the main challenge faced by such kinds of systems is maintaining a healthy temperature. Such electronic cooling methods are essential in data centers, battery cooling, and different kinds of computers, from workstations to personal computers. There are different kinds of cooling methods available for electronic devices, such as microchannels, spray cooling, and Peltier cooling [140,141]. It has been shown that most of the investigations/studies in this field are related to nanofluid circulation in microchannels. Emad E. Mahmoud et al. [142] conducted research to investigate the thermal efficiency and the pressure drop of Al₂O₃–water nanofluids in a novel micro-heat sink to cool a CPU. This study investigated the heat transfer coefficient and the pressure drop at different Reynolds numbers. It has been reported that the pressure drop was 15% when the Reynolds number was 1000, and the heat transfer coefficient values were recorded at 24.5% for the same Reynolds number. Another study used the same kind of nanofluid in a microchannel experimental setup to investigate the efficiency of the cooling at different volume flow rates, from 2 mL/s to 8 mL/s, and also at different volume concentrations, from 0.1% to 0.5%. The highest heat transfer coefficient, namely 16,524.01 mW/m²K, was recorded as 8 mL/s, with a 0.5% volume ratio [143]. Al₂O₃ nanofluids have become a popular option in heat transfer applications due to their higher stability in water. Another study proved this statement by using Al₂O₃ nanofluid in flattened heat pipes for electronic cooling [144]. This research considered several variables, such as the tilting angle of the pipe, and compared the Al₂O₃ nanofluid with pure water and TiO₂ nanofluid. The results showed that Al₂O₃ nanofluids perform better at the 45° inclined angle, with a heat transfer coefficient of 4.6 W/m²K. Another similar kind of comparative study was conducted by Imène Saad et al. based on a theoretical model of a nanofluid-filled copper cylindrical heat pipe for electronics cooling applications for CuO/water and Al₂O₃/water nanofluids [145]. However, in this study, it has been reported that the efficiency of CuO/water nanofluid is much higher in terms of the thermal resistance decrease and the capillary limit increase. The study conducted by HassanWaqas et al. [146] focused on the heat transfer rate and thermal resistance using TiO₂/water nanofluid in circular microchannels. It is crucial to investigate the behavior of the ionic nanofluids within magnetic fields. One such study investigated temperature and heat dissipation characteristics in a ferrofluid under the influence of no magnet and I-, L-, and T- magnetic field patterns using a lab-scale experimental setup [147]. According to the results of the research, the heat transfer rate was much higher with the magnetic field than when field was disabled. Basma Souayeh et al. [148] also investigated the magnetic effect of using magnetic baffles on the Fe₃O₄ magnetic nanofluid in mini-channel heat transfer applications. ANYS Fluent 18.1 software was used for this study, and the Nusselt number and pressure drop values were measured without a magnetic field, with one magnetic baffle, and with two magnetic baffles. According to the results, the Nusselt number increased with the magnetic field strength, while the pressure drop showed the opposite behavior.

8.6. Machining Fluids

Machining or cutting fluids are essential liquids in the machining processes that help to reduce the temperature and friction of the machining while removing the metal chips. Extensive research has been carried out to develop different kinds of nanofluids dedicated to machining fluid applications. The main objectives of this research were to increase the efficiency of the fluid while following minimum quantity lubricant (MQL) standards [8,149,150,151]. Research of this kind has been conducted by varying several factors, like metal or alloy type, nanofluid type, and different concentrations of nanofluids. Aysegul Yucel et al. [152] investigated the effect of mineral oil-based MoS₂ nanofluid in machining AA 2024 T3 Al alloy. Several parameters, such as surface roughness, surface topology, and maximum temperature, were investigated in this study. The use of MoS₂ resulted in a greater enhancement compared to dry machining and base fluid machining. In the study conducted by Abdelkrem Eltaggaz et al. [153], Al₂O₃-based nanofluids with different concentrations (0, 2, 4 wt.%) were used for the machining of Ti-6Al-4V alloy, resulting in a positive effect on the average surface roughness. The use of hybrid nanofluids in such applications gives several advantages over mono nanofluids. Aqib Mashood Khan et al. [154] carried out a comprehensive comparative analysis of using Al–GnP (graphene nanoparticles) hybrid nanofluids vs. liquid LN₂ in the machining of Ti-6Al-4V alloy. This research focused not only on the surface quality but also on the cutting power, specific energy consumption, and CO₂ generation. Even though the nanofluid displayed a significant improvement in some parameters, LN₂ showed the highest improvement. Hongfei Wang et al. [155] analyzed the wear of micro-machining the Ti6Al4V alloy with graphene nanofluid and a diamond tool. The results were evaluated using both molecular dynamic simulations and SEM analysis.

Some researchers have tried to replace traditional machining fluids with nanoparticle-incorporated vegetable oils in several studies, considering the environmental impact of some commercial cutting fluids [156,157,158,159].

Enes Usluer et al. [160] used MWCNT-enforced cutting nanofluid in machining and conducted a cost analysis. It was reported that the total machining costs were lowered by 78%, while the performance of the machining application was improved.

8.7. Refrigerator Systems

A considerable amount of research has been conducted to improve the efficiency of refrigerant fluids, such as R32, R134a, R410a, R600, and HFC32 [161,162,163]. Seboka Gobane et al. [164] conducted research to investigate the possibility of replacing traditional refrigerants with CuO-based Jatropha oil. However, in this study, only two parameters were evaluated, namely the thermal conductivity and dynamic viscosity at 0.5, 1.0, 1.5, and 2% volume concentrations. Both the parameters show an increment in the nanoparticle concentration. In the study conducted by Sunday A. Afolalu et al. [165] used rice husk-based nanoparticles with different concentrations (0.2, 0.3, 0.5, and 0.6 wt.%) in R600a refrigerant, the use of rice husk nanoparticles greatly increased the temperature control of the system. Another study examined the effect of Al₂O₃ nanoparticles in the HFC-32 and R-410a refrigerants at different volume concentrations from 0.06 to 0.14% in a MATLAB simulation model [166]. The results showed that the net refrigerant effect increased from 77% and 79% to 81.2% and 83.5% in HFC-32 and R-410a, respectively.

8.8. Industrial Manufacturing of Nanofluids

In addition to the lab-scale research on nanofluids, there are several industrial manufacturers of nanofluids for various applications, such as cooling, heat transferring, and lubricants. Meliorum Technologies is one such industrial nanofluids manufacturer for automotive coolant additives and several nanoparticles [167]. Tool-x is another industrial nanofluid manufacturer that basically produces cutting fluids incorporated with inert carbon nanoparticles [168]. Such products help to reduce the wear and tear of mechanical and cutting tools. HYDROMX is another company that provides nanofluid solutions for various applications, such as green buildings, HVAC systems, and solar thermal applications [169]. Synano is another nanotechnology-based company that provides nanofluid cooling solutions for electronic devices, data centers, and electrical vehicles [170].

In the applications of nanofluids, it is very difficult to determine the most suitable type of nanofluid since the properties of nanofluids are highly dependent on the nanoparticle type, concentration, and base fluid type. However, according to the obtained results, the stability and the thermal performance of the carbon nanoparticles are much higher in oil-based applications compared to aqueous fluids. On the other hand, metal and metal oxide nanoparticles have also displayed higher performance in aqueous and ethylene glycol-based fluids. However, suitable surfactants can be used to improve the stability and thermal performance of selected nanofluids. Figure 7 represents the number of research that has been published from 2010 to 2024 relative to different engineering applications of nanofluids.

9. Molecular Dynamics of Nanofluids

In addition to the experimental studies and theoretical studies of nanofluids, molecular dynamics studies play a significant role in the field of nanofluids. In molecular dynamics studies, simulations are conducted to investigate the fundamental interactions and mechanisms of the nanofluids. Molecular dynamics studies help to understand the micro-level behavior of particles, which paves the way for the prediction and understanding of the thermophysical properties of nanofluids [171].

Amirhosein Mosavi et al. [172] conducted a computational fluid dynamics simulation to analyze the effect of the different sizes of spherical barriers on Ar/Cu nanofluids’ thermal properties. This team used LAMMPS (large-scale atomic/molecular massively parallel simulator fluid simulation) fluid simulation software to analyze the physical parameters, such as the atomic temperature, total energy, number of nanofluid atoms at the gas phase, radial distribution function, and thermal conductivity of the nanofluid. The research of Yuhuan Chen et al. [173] modeled the surface tension of nanofluids containing N, N-dimethyl cyclohexylamine (DMCHA), and several alcohol mixtures, including 1-pentanol, 1-hexanol, 1-heptanol, 1-octanol, 1-nonanol, and 1-decanol. Fahim Mahtab Abir et al. [174] compared the experimental and simulated specific heat capacity of salt-based 1 wt.% SiO₂ nanofluid using a differential scanning calorimeter for the thermogravimetric analysis as well as the MedeA simulation software. Mohammad Bagheri Motlagh et al. [175] analyzed the convection heat transfer and Poiseuille flow in a Cu nanochannel of a nanofluid using the LAMMPS fluid simulation software. In this study, the Nusselt number was calculated and compared with the base liquids to investigate the convection heat transfer. This study showed that the increment of the diameter of nanoparticles negatively affects the Nusselt number due to the decrement of the surface-to-volume ratio of the nanoparticles. T. Khamliche et al. [176] also used LAMMPS software to analyze the molecular dynamics of self-synthesized Cu nanoparticle behavior within ethylene glycol. Kamal Ghani Dehkordi et al. [177] also used the same software to conduct molecular dynamics simulations of Fe₂O₃ nanoparticle-based water nanofluid. This study specifically analyzed the boiling phenomenon of the nanofluid in different microchannels made with Pt, Fe, and Au materials in the presence of an external electrical field. In another study, the cooling performance of Al₂O₃/water nanofluid in a novel multi-walled heat sink was investigated [178]. The maximum temperatures were evaluated using the Reynolds number, and the nanoparticle volume fraction was determined using CATIA and CAMSOL software. The use of surfactants improved the stability of the nanofluids; however, it is crucial to understand the nano-level mechanism of these surfactants to determine the most suitable surfactant for the relevant nanofluids using molecular dynamics simulations, and there is currently little research investigating this micro-mechanism [179,180,181].

10. Nanofluids and Artificial Intelligence

Recent advances in artificial neural networks (ANNs) have significantly enhanced the design, analysis, and deployment of nanofluids in thermal management, renewable energy, and industrial systems. Artificial neural networks (ANNs) are computational models inspired by biological neural systems with interconnected nodes and layers, which learn patterns from training data in order to perform various tasks, like classification, prediction, and decision-making. In ANNs, there are basically three types of layers, namely the input layer, output layers, and several hidden layers in between them, as shown in Figure 8. In a complex field, such as nanofluids, it is very important to develop a data-driven approach to predict the properties of different kinds of nanofluids [182,183,184]. This data-driven approach gives higher accuracy than empirical models. Also, the empirical models are adapted for use with certain types of nanofluids, as mentioned earlier (e.g., the Maxwell model is suitable for spherical nanoparticles with less than 5% v/v concentration.), but the use of ANNs can overcome this issue and ensure higher accuracy.

Artificial neural networks (ANNs) have emerged as a powerful tool for predicting the thermal properties of nanofluids. Na Wang et al. [185] modeled the thermal conductivity of MgO-based nanofluids using mathematical correlation, multivariate adaptive regression spline (MARS), and group method of data handling (GMDH) methods. Several past research studies were considered for modeling, including those using different base fluids, such as water, engine oil, and ethylene glycol. Mohammad Hemmat Esfe et al. [186] developed an ANN using MATLAB 2021a software to predict the rheological behavior of MWCNT–Al₂O₃ (3:7)-based SAE40 hybrid nanofluid. The researchers used 203 different samples with different temperatures and concentrations to train the model, and 70% of the data were dedicated to the training, while 15% were used for the validation, and the remaining 15% were also used for training. Another similar research study conducted by Saleem Nasir et al. [187] developed a BRT (Saleem Nasir) ANN to predict the Nusselt number and heat transfer coefficient of Ag+MgO/H₂O hybrid nanofluids at different volume concentrations, Hartmann numbers, pressure distributions, and Rayleigh numbers in a porous cavity. Tarikul Islam et al. [188] used an ANN to optimize the thermal performance of Cu–Al₂O₃/water hybrid nanofluid flow in a wavy enclosure with inclined periodic magnetohydrodynamic (MHD) effects. This study used an ANN and response surface methodology (RSM) to maximize the Nusselt number and optimize the thermal performance of the nanofluid, achieving a mean square error of 0.00018. The research by Saeed Alqaed et al. [189] took a different approach to modeling the exergy and energy of parabolic trough solar collectors, which is not a widely discussed topic in the nanofluid field. The study was conducted for a Cu–Al₂O₃∕water hybrid nanofluid in a turbulator, which was used as a parabolic trough solar collector system. The study varied the volume fraction from 1% to 5% and the Reynolds number from 6000 to 18,000. The results showed a higher accuracy of ANNs at low and high turbulences of 0.51% and 1.46%, respectively.

11. Discussion

In this study, nearly 200 articles have been analyzed, including journals, reviews, book chapters, conferences, and online sources, providing a broad base of knowledge about the nanofluid discipline. However, according to the research articles that have been reviewed, there are several points that can be considered for future studies and clarifications.

• Many research articles have focused on the thermal properties of nanofluids, such as their thermal conductivity, viscosity, and specific heat capacity. However, there is lack of studies in the research articles focused on other properties, such as flash point, which is a crucial property in many applications. The number of theoretical studies is also not at a satisfactory level.
• Most of the theoretical models are based solely on the nanoparticle volume fraction. However, the extensions of these models have considered the effect of other important properties, such as interfacial layer and particle shape. However, there is lack of models which consider the effect of temperature, which has a significant effect of thermal properties of nanofluids.
• In the nanofluid theoretical models, there are several limitations, such as concentration and particle size and shape. It can be seen that not enough research has been conducted to investigate suitable nanofluid types with which to use these models without significant deviations.
• Many research studies have been conducted considering the applications of nanofluids; however, there has been a lack of financial analysis concerning the use of nanofluids in such applications.
• The use of ANNs in thermal property predicting can be identified as a positive trend in the recent past, especially due to their higher accuracy compared to theoretical models.

12. Conclusions

In this study, the thermo-physical improvements in nanofluid properties, such as thermal conductivity, viscosity, specific heat capacity, and flash point, were reviewed and summarized. Several conclusions can be obtained based on the recent research outputs. Several experimental analyses have been conducted to obtain thermal conductivity, viscosity, and flash point measurements of different nanofluids, and we can draw conclusions based on the results obtained. A clear improvement in thermal conductivity, viscosity, specific heat capacity, and flash point can be observed in nanofluids compared to the relevant base fluid, and this improvement depends on numerous parameters, such as the base fluid type, nanomaterial type, nanoparticle shape, nanoparticle concentration, nanoparticle size, surfactant type, temperature, nanofluid preparation method, etc. The following conclusions can be drawn from the summarized research outcomes:

• The thermal conductivity, viscosity, and flash point of nanofluids increase with the nanoparticle concentration.
• A clear relationship between the specific heat capacity of nanofluids and the nanoparticle concentration or nanoparticle type was not observed. Both increments and decrements in the specific heat capacity of nanoparticles can be seen with the increase in nanoparticle concentration.
• The concentration and type of surfactants negatively and positively affected the increased thermal conductivity depending on the type of nanofluid.
• The applications of nanofluids have made a significant impact on several engineering and scientific disciplines in improving the performance of heat transfer applications.
• Molecular dynamics simulations demonstrated the need for a new field in the nanofluid research area in order to understand the nano-level impact of the nanoparticles in nanofluids.
• The collaboration of artificial intelligence can make a significant impact on nanofluid studies by providing a data-driven approach to predict important thermo-physical properties.

According to a number of research articles considered in this review, it can be seen that nanofluids’ physical properties are highly dependent on the nanoparticle type, the base fluid type, and the concentration and temperature. Furthermore, surfactants can improve these properties to some extent. However, according to these studies, it is very difficult to select a specific nanoparticle type for a certain application. However, the higher performance of carbon nanoparticles in oil-related applications and metal oxide-based nanoparticles’ higher performance in aqueous fluids are apparent.

It is very difficult to declare a clear common relationship between some of the thermo-physical properties and different parameters of nanofluids, as there is no common procedure or protocol followed by researchers worldwide. The most impactful method to develop relationships is to conduct these studies separately for different kinds of nanofluids since the thermal properties are dependent on the type of nanoparticle and the fluid type, especially with the contribution of technologies like ANN.13.