Wet chemical synthesis of metal oxide nanoparticles: a review

A. V. Nikam ^a, B. L. V. Prasad ^b and A. A. Kulkarni *^a
aChem. Eng. Proc. Dev. Division, CSIR-National Chemical Laboratory, Dr. Homi Bhabha Road, Pashan, Pune-411008, India. E-mail: aa.kulkarni@ncl.res.in
^bPhysical and Material Chemistry Division, CSIR-National Chemical Laboratory, Dr. Homi Bhabha Road, Pashan, Pune-411008, India

First published on 16th July 2018

---

---

Arun Nikam

Arun Nikam is a Ph. D student at CSIR-National Chemical Laboratory. He received his Master's degree in Chemistry from Shivaji University, Kolhapur in 2010. He studies the flow synthesis of inorganic and organic nanomaterials. He is interested in controlling the physicochemical properties of polymeric drug nanoparticles and bimetallic nanoparticles.

B. L. V. Prasad

B. L. V. Prasad is a scientist in the Physical and Material Chemistry Division at CSIR-National Chemical Laboratory. He studies the synthesis of nano/biomaterials and their applications in the health care and energy sectors. He is an expert in self-assembled monolayers of nanoparticles and digestive ripening.

Amol Kulkarni

Amol Kulkarni is a scientist in the Chemical Engineering and Process Development division at the CSIR-National Chemical Laboratory, Pune (India). He leads a program on continuous flow synthesis of API, agrochemicals, intermediates, and nanomaterials. His group focuses on the design of flow reactors and the optimization and scaleup of continuous flow reactors.

---

1. Introduction

Nanomaterials show different optical, electronic, magnetic and catalytic properties compared to their bulk counterparts due to their high surface to volume ratios, surface defects, and quantum effects. Metal oxides have an intrinsic charge separation capacity that differentiates them from metals. The synthesis of metal oxide nanoparticles can be further explored to obtain desired or targeted properties. Progress in synthesizing metal oxide nanoparticles has been phenomenal due to their applications in electronics,¹ optics,² energy storage³ and catalysis.⁴ Since 1980, efforts have been made to invent methods to synthesize nanoparticles with desired properties. Syntheses of nanomaterials are categorized in two types: the top-down approach and the bottom-up approach. In the top-down approach, a large piece of material is broken into nanosized entities. This requires complex, expensive, highly energy intensive and specialized setups to maintain specific conditions, such as pressure, temperature, and environment (inert and non-flammable). It has been observed that despite the use of expensive setups/instrumentation, the top-down approach produces nanomaterials with surface defects and non-uniform shapes, which hinders their applicability. In the bottom-up approach, atomic or molecular species are integrated to form nanostructures. The bottom-up approach is largely based on wet chemical synthesis, which is relatively simple, modular and scalable. Wet chemical processes are used to obtain better control over the nanostructures. Wet chemical synthesis approach has achieved great success because it enables to tune kinetic and thermodynamic parameters to achieve great control over the sizes, shapes, and compositions of nanoparticles, which is reflected by changes in their optical, electronic and surface properties. While a few interesting examples have been reported in the literature, it is necessary to explore different wet chemical synthesis techniques to obtain an in-depth analysis of the control of size, shape, and crystallinity. Controlled synthesis of nanoparticles is a central prerequisite for achieving desired properties. The reaction parameters (viz. type of precursor, heating method, heating/cooling rates, temperature, concentration, mixing, types of ligands, solvent properties, addition of the sequence, and addition rate) play vital roles in deciding the sizes and shapes of nanoparticles. Therefore, it is necessary to compile literature data on the controlled wet chemical synthesis of metal oxide nanoparticles.

Wet chemical synthesis approaches have been realized to produce desirable sizes and shapes of metal oxide nanoparticles in a reproducible manner. Control over size and shape is achieved by a better understanding of elementary events, the mechanism of conversion of the precursor, surface stabilizing agents, and reagents in the system as well as their correlations with growth and nucleation rate. To realize the industrial application of colloidal metal oxides, the synthesis procedure should be scalable to meet industrial requirements in terms of the quality and quantity of the product. A typical conventional batch protocol used for scaleup can yield nanoparticles at an average production capacity of 10 mg mL⁻¹.⁵ To scale up the synthesis of metal oxide nanoparticles, it is desirable to increase the batch volume or reactor volume. Increasing the batch volume suffers from limitations of mixing, mass and heat transfer, which alters the nucleation and growth kinetics. The nucleation and growth kinetics of particle formation are deciding factors to control the size of nanomaterials in a solution. However, one may observe batch-to-batch variations in size, shape, and composition when using the same procedure. An alternative approach to avoid batch-to-batch variation of nanomaterials and to achieve high quality product with high throughput is to transform batch synthesis to flow synthesis and optimize the reaction parameters in a reproducible manner. In flow synthesis, smaller dimensions help to achieve rapid mixing and a higher interfacial area helps to enhance the mass transfer rates and high heat transfer rates throughout the reactor due to the high surface to volume ratio^6,7 and the ability to reduce the extent of axial dispersion or back mixing. Flow synthesis can afford reproducible products, and it can be automated. Due to the modular nature of the flow setup, one can control nucleation and growth by adjusting parameters such as the flow rate, dimensions of the reactor, temperature, concentration, and mixing.^8–10 By obtaining a detailed understanding of flow synthesis, it is possible to achieve large-scale production without compromising product quality.

This review gives in-depth insight into the diverse facets of methodologies for the synthesis of metal oxide nanoparticles using the wet chemical approach, which is relevant to meet industrial requirements. The different methods are discussed critically to elucidate their advantages and disadvantages. Various methods have been used for the synthesis of metal oxide nanoparticles. Some of the most commonly used methods include precipitation, sol–gel, reverse micelle, thermal decomposition, solvothermal, microwave-assisted and flow synthesis. We will ignore the sol–gel, precipitation and reverse micelle methods because they have already been deeply explored. In this review, we will focus on the microwave-assisted, thermal decomposition and hydrothermal synthesis methods, which can be scaled up without loss of quality of the metal oxide nanoparticles in flow synthesis.

These synthesis methods usually produce monocrystalline metal oxides due to the high temperatures used during synthesis. Various aspects of these synthesis methods with reference to the properties of the materials are elaborated in Table 1. The scope of this review is restricted to only metal oxides which are synthesized through wet chemical synthesis, which helps achieve greater control over the sizes and shapes of the nanoparticles. These include iron oxide, ferrites, zinc oxide, cobalt oxide, titanium oxide, copper oxides, silicon oxides, manganese oxide, zirconium oxide, stannous oxide, lanthanide oxide, and magnesium oxide.

---

2. Microwave-assisted synthesis

The microwave-assisted synthesis of nanomaterials has been under development for the last three decades. With technological advancements in the design of microwave generators, microwave synthesis of nanomaterials is becoming increasingly attractive. Rapid heating in the reaction system can be achieved using microwave radiation, which enhances the reaction rate and decreases the reaction time. Hence, microwave radiation has been used to reduce reaction time by enhancing the reaction kinetics. The first publication in this area appeared in 1985 from Komarneni and Roy, who reported the liquid phase synthesis of TiO₂ microspheres by a sol–gel process using microwave heating in kerosene.¹¹ Around this period, Gedye and Giguere et al. explored microwave heating to carry out organic reactions.^12,13 Over the years, the number of publications on the microwave-assisted synthesis of nanomaterials and organic compounds has rapidly increased. Earlier, the synthesis of nanomaterials was carried out in a domestic microwave oven with no control over temperature or pressure, leading to uncertainty in the product quality. Recently, microwave ovens or reactors have enabled in situ measurement of temperature and pressure with IR sensors. Online monitoring of reaction parameters, such as temperature and pressure, enables improvements in the quality of nanoparticles by manipulating the reaction parameters. Microwave radiation can be absorbed by solvents with high dielectric constants and produces heat during irradiation; this causes uniform heating in a reaction mixture through dipole–dipole interactions with the alternating electric field generated by the microwaves. Due to this property, microwave energy is more efficient for selective heating. This method requires less energy and time to obtain nanoparticles compared to conventional heating methods.^13–15 Conventional heating generates a thermal gradient in the reaction solution in the absence of efficient mixing; this can alter the local nucleation and growth kinetics of nanomaterials and may result in poor product quality, with significant variations in the shape and size distributions. Microwave heating induces homogeneous heating, which leads to uniform nucleation and growth kinetics in the entire solution and affords better quality products, preferably under efficient mixing. Details on the effects of the power density, frequency, microwave penetration depth and type of reactor material on the product quality will be elaborated separately. Various metal oxides that have been synthesized using microwave-assisted techniques include ZnO,¹⁶ α-Fe₂O₃, β-Fe₂O₃, Fe₃O₄,¹⁷ CuO,¹⁸ Cu₂O,¹⁹ Mn₃O₄, MnO₂,²⁰ TiO₂,²¹ and Co₃O₄.²² Iron oxides occur in different forms, such as FeO, α-Fe₂O₃, β-Fe₂O₃, γ-Fe₂O₃, ε-Fe₂O₃, and Fe₃O₄. Particularly, the magnetic properties and colors of iron oxides vary with their sizes, shapes, crystal phases and compositions at nanoscale dimensions. Hematite (α-Fe₂O₃), maghemite (β-Fe₂O₃) and magnetite (Fe₃O₄) have been extensively investigated due to their applications in the magnetic memory storage, catalysis, biomedical and pigment industries. Iron oxide nanoparticles show well-defined properties when they are monodisperse in size and shape. Most synthesis processes are driven by properties of the solvent, viz. boiling point and polarity. Hu et al. reported the synthesis of monodisperse spherical Fe₃O₄ nanoparticles with a polydispersity index of 3% in ionic solvent using a microwave-assisted method.²³

According to the LaMer theory, to obtain uniformly sized nanoparticles, instant nucleation is necessary throughout the reaction solution; to achieve this, the temperature gradient should be eliminated throughout the reaction mass. Thus, the microwave heating mechanism under efficient mixing eliminates thermal gradients in the reaction system and allows nuclei to grow homogeneously. Iron oxide nanoparticles were prepared using ionic solvent [BMIM][BF4] as a co-solvent; it acts as a microwave absorber, helps to induce uniform heating of the solution and maintains thermal homogeneity by an ionic conduction mechanism that leads to the formation of monodisperse nanoparticles (see Fig. 1A). Guru et al. studied the influence of different glycols on iron oxide nanoparticles prepared by microwave-assisted synthesis.²⁴ They found that using different glycols in the reaction system affects the final phase of the iron oxide nanoparticles; also, the size or length of the glycol is sensitive to the particle size. Particles prepared with ethylene glycol, poly ethylene glycol and poly propylene glycol in water had crystal sizes of 35, 29.9 and 28.2 nm for Fe₃O₄, α-Fe₂O₃ and γ-Fe₂O₃, respectively. Ai et al. demonstrated the synthesis of rose-like nanocrystalline Fe₃O₄ superstructures using ethylene glycol as a solvent by a rapid microwave-assisted approach.²⁵ A three-step mechanism was proposed for the nano-rose formation. In the first step, Fe₃O₄ nuclei formed due to the absorption of P123. In the second step, the free P123 molecules could selectively adsorb onto certain surfaces of neighboring Fe₃O₄ nano-crystals to suppress their natural anisotropic growth. These small crystalline primary particles aggregated and formed interparticle pores, resulting in porous Fe₃O₄ nanoroses.

The effects of the metal precursor decide the morphology, size and final phase of nanoparticles because the anionic part of a metal complex can act as a surface-stabilizing agent or surface etchant. Bilecka et al. studied the effects of the precursor on the size of iron oxide nanoparticles.²⁰ Fe(acac)₂, Fe(ac)₂, and Fe(acac)₃ precursors were used for the preparation of 11 nm, 7 nm and 5 nm-sized Fe₃O₄ nanoparticles, respectively, in benzyl alcohol (see Fig. 1B). Osborne et al. also synthesized Fe₃O₄/Fe₂O₃ nanoparticles in water using FeCl₃ and dextran at 100 °C.²⁶ Dextran serves as a surface stabilizing agent and confines nucleation and growth of nanoparticles in a limited space, which affords smaller Fe₃O₄ particles in the range of 6.5 nm (see Fig. 1C). Bhavesh et al. reported the synthesis of extremely small fluorescein-labeled iron oxide nanoparticles 2.5 nm in size using a microwave in a single step as a viable tool for cell labelling and T1-MRI agents. In their method, they functionalized the surface of iron oxide nanoparticles with a carboxymethyl derivative of fluorescein isothiocyanate (FITC)-labelled dextran as an anchoring site for further biomodification and fluorescence signals, which act as T1 contrast agents. Lamellar nanostructures of sodium/potassium iron oxide nanosheets were obtained by heating a suspension of iron(II) sulphate, sodium thiosulfate, and sodium/potassium hydroxide in a microwave for 5 min (ref. 27) (see Fig. 1D). Nanosheets were observed after 1 min microwave radiation, and increasing the reaction time to 5 min afforded larger sodium iron oxide nanosheets. These results showed a rapid increase in the growth rate upon microwave irradiation. Wang et al. reported the synthesis of Fe₃O₄ and α-Fe₂O₃ nanoparticles by a fast microwave-assisted solution method.¹⁷ The elliptical α-Fe₂O₃ nanoparticles were formed by an oriented attachment mechanism (see ref. 17, Fig. 3C). Zhou et al. reported hexagonal Fe₃O₄ nanoplates with an average edge length of 80 nm synthesized by microwave irradiation.²⁸ The reaction time, microwave power, concentration of NaOH and additives such as citric acid, sodium acetate and sodium hypophosphite were adjusted to tune the morphologies and sizes of iron oxide nanoparticles. Citric acid plays a vital role in the formation of plate-like Fe₃O₄ nanoparticles (see ref. 28, Fig. 1C); it acts as a capping agent and shape-directing agent because it can chelate metals with its carboxylate groups. Meanwhile, without surfactant, the iron oxide formed aggregated nanoparticles with undefined shapes.

ZnO nanoparticles are attractive due to their antibacterial, antifungal, anti-corrosive and UV-filtering properties. Various structures of ZnO have been prepared using microwave synthesis and tested for multiple applications due to their size- and shape-dependent properties. Hu et al. reported size tuning of monodisperse ZnO colloidal nanocrystal clusters (CNCs) by a microwave polyol synthesis.¹⁶ ZnO nanocrystal clusters were prepared from Zn(OOCCH₃)₂·2H₂O in ethylene glycol by a seed-mediated mechanism, where the secondary structure of the ZnO CNCs was obtained from smaller ZnO nuclei (7 to 8 nm) attached to each other to form well-assembled spherical colloidal nanocrystals. Jalal et al. prepared ZnO nanocrystals in the presence of an ionic liquid [BMIM][NTf₂] by a microwave-assisted method. The average size of the ZnO nanoparticles was 41.3 nm, and they were used in nanofluids for antibacterial material.²⁹ Cho et al. reported shape-controlled growth of ZnO nano- and microstructures by microwave-assisted synthesis.³⁰ Different basic structures of ZnO, viz. nanorods, nanocandles, nanoneedles, nanodisks, nanonuts, microstars, micro-UFOs, and microballs, were reported at a low temperature (90 °C) with low power microwave-assisted heating and a subsequent aging process (see Fig. 2). These nanostructures were obtained by changing the metal precursors, the capping agents, and the aging times. Even more complicated ZnO structures, including ZnO bulky stars, cakes, and jellyfishes, were also reported to be synthesized by microwave irradiation of a mixture of the as-prepared basic ZnO structures. The evolution of the variable morphology can be explained by preferential growth of a particular plane and suppression of a specific plane. ZnO nanorods are obtained by preferential growth along the (0001) plane, which grows faster compared to other planes. As the reaction proceeds, dissolution becomes dominant due to the decreased concentration of the growth units. Due to this limited concentration growth, the ends of the nanorods are flat (see Fig. 2A). Capping agents have excellent binding capabilities to Zn²⁺ ions, which helps to form nanoneedles of ZnO. The small ethylenediamine (EDA) capping agent binds to six symmetric surfaces of the (1010) and (0001) faces. However, ZnO crystal growth rate along the (0001) plane is faster than in the normal six symmetric directions; hence, ZnO nanoneedles were formed (see Fig. 2B). ZnO nanodisc formation is facilitated by citrate ions, in which the COO⁻ groups have stronger binding ability to the (0001) plane than the (1010) plane and the capped citrate prohibits contact between the growth units and the (0001) plane. Under this circumstance, ZnO grows along six symmetrical planes (0001), which leads to hexagonal nanodisc formation (see Fig. 2C). Sharma et al. reported the synthesis of spherical ZnO nanoparticles in a microwave using zinc sulphate heptahydrate with sizes ranging from 10 to 15 nm.³¹ Bilecka et al. synthesized ZnO spherical nanoparticles in a microwave using a benzyl alcohol route.³² The growth rate of ZnO nanoparticles was accelerated by microwave heating (13.36 nm3 min⁻¹) in comparison to conventional heating (3.9 nm³ min⁻¹). Pimentel et al.³³ prepared uniform ZnO nanorods using microwave irradiation for 3 min with an aspect ratio of about 13. Higher microwave power (600 W) resulted in smaller ZnO nanorods compared to lower power (300 W). The aspect ratio of ZnO is higher at high power, and with increasing reaction time, the lateral growth of nanorods becomes facile; hence, the aspect ratio decreases as the reaction proceeds. On the other hand, the aspect ratio was smaller at low microwave power due to lower growth rates and growth at a specific plane. The authors have reported rapid nucleation at higher microwave power, which leaves very little precursor for growth; hence, the length of the ZnO nanorods is smaller in comparison to those obtained at lower microwave power. High power initiated faster nucleation in contrast with low power, where large numbers of small nuclei were formed that acted as seeds for nanorod growth. Due to the higher power, most of the precursors were converted into nuclei, and a small amount of precursor remained in the solution to facilitate the growth of nanorods. Therefore, at higher power, the nucleation rate is faster and the growth rate is slower, resulting in small ZnO nanorods.

Cobalt oxide is another important metal oxide which has played significant roles in memory storage, electronics and catalysis due to its superior magnetic properties and surface properties. Li et al.³⁴ reported controlled Co₃O₄ synthesis achieved by microwave hydrothermal synthesis using Co-MPA (3-mercaptopropionic acid) within 10 min. The shape of Co₃O₄ depends upon the temperature. While low temperature favors spherical nanoparticles, high temperature favors cubic 12 to 20 nm-sized particles. However, Bhatt et al. also prepared spherical Co₃O₄ nanoparticles with sizes in the range of 3 to 12 nm by a fast microwave-assisted synthetic route using a surfactant, tri-octyl phosphine octane.³⁵ Mesoporous Co₃O₄ nanoflakes prepared by a microwave-assisted method and a low temperature conversion method were reported by Chen et al.³⁶ It was observed that the formation of Co₃O₄ contained layered intermediate cobalt carbonate hydroxide hydrate Co(CO₃)_0.5(OH)_0.11H₂O with interconnected architectures. Bilecka et al. reported the synthesis of CoO using cobalt acetate in benzyl alcohol at 200 °C in a size range of 45 to 75 nm within a few seconds. TiO₂ nanoparticles have attracted great attention because they have been used as pigments in cosmetics, paints, food additives and self-cleaning glass because they break down dust in sunlight by photocatalysis. Dar et al. demonstrated TiO₂ synthesis in a mixture of alcohols (ethanol + benzyl alcohol) and obtained 5 nm and 7 nm nanocrystals by a microwave-assisted route.²¹ The average size of TiO₂ is 295 nm when ethyl alcohol is used as the solvent. Here, randomly oriented aggregation results in a decrease of the overall surface energy. Oriented aggregation was observed, which is actually enabled by the rapid formation of nuclei and the absence of a growth controlling agent. In another report, Wang et al. synthesized hierarchical TiO₂ nanocrystallite aggregates composed of 10 nm nanocrystallites with a size of 500 nm by a microwave-assisted method at 150 °C in a short time.³⁷ Ethanol and TiCl₄ were selected as the solvent and titanium precursor, respectively. The rapid heating rate and superheating/“hot spots” of the reaction system under microwave irradiation resulted in the instantaneous formation of larger numbers of nuclei, which led to the formation of numerous clusters. This implies that it is possible to use a microwave to generate a large number of nuclei which can be used subsequently to grow larger particles. Details about the reaction parameters of metal oxides synthesized by microwave-assisted methods are summarized in Table 2.

---

Here, we have summarized the effects of solvents, with or without surfactant, and the effects of precursors on the shapes and sizes of iron oxide nanoparticles using a microwave as a heating tool. One must carefully choose a solvent with a high dielectric loss that can accelerate the reaction by maximum absorption of microwaves. The hydrophobic and hydrophilic parts of the surfactant and its length influences stability, shape, and size, as mentioned above. In water, without surfactant, in the absence of surface passivation, metal oxide tends to form nano-agglomerates. In some cases, metal oxides form agglomerates despite the use of surfactant due to the formation of nanoparticles at a rate far greater than the rate at which the surfactant binds to the active sites. The active sites on the nanoparticles have high surface energy, and the nanoparticles attempt to minimize this energy by agglomeration. The surfactant must have a higher binding ability that is more suitable to bind the active sites of nanoparticles; this helps to control the sizes and shapes of nanoparticles during synthesis (viz. amines, thiols, and high molecular weight polymers).

3. Thermal decomposition

The chemical process in which a metal precursor is heated above its decomposition temperature in a high boiling point solvent is known as thermal decomposition. Usually, the use of aqueous and volatile solvents for metal oxide preparation suffers from polydispersity, agglomeration and poor crystallinity. Metal oxide nanoparticles obtained by thermal decomposition do not require post-synthesis thermal treatment, and these nanoparticles are highly monocrystalline. In addition to their monocrystalline nature, the particles are monodisperse in size and shape. The nucleation and growth rate of metal oxide formation by thermal decomposition is very well understood, and good control over particle size and shape has been demonstrated by slight changes in the reaction parameters. Most of the metal oxide nanoparticles synthesized by this method use organometallic precursors dissolved in an organic solvent and a surface stabilizing agent at a high temperature in an inert environment. During decomposition, the presence of several reagents and their ionic forms actually leads to complex reactions. Park et al. have shown separation between nucleation and growth phases by adjusting the reaction parameters and optimizing the synthesis to obtain tuneable-sized nanoparticles.⁴¹ Son et al. have shown that the shape of iron oxide nanoparticles can be controlled by the addition of small amounts of a gadolinium oleate (Gd-oleate) complex as a surface passivation agent at particular planes. Interestingly, the addition of Gd-oleate introduces shape changes in iron oxide nanoparticles from spheres to cubes. This shape change has been explained by DFT calculations, which proved that the Gd-oleate complex strongly binds to the 100 plane compared to the 111 plane; this reduces energy by surface passivation, which results in slow growth at the 100 plane. On the other hand, the iron oxide nanoparticles grow along the 111 plane, resulting in cube shapes.⁴² Park et al. reported the ultra large scale synthesis of monodisperse metal oxide nanocrystals. Fe₂O₃, CoO, MnO, FeO@Fe, and MnFe₂O₄ were synthesized using a metal oleate precursor in a high boiling point solvent.⁴¹ The effects of the boiling point of the solvent on the size of Fe₂O₃ nanoparticles have been studied; 5 nm-sized iron oxide nanoparticles were synthesized using 1-hexadecene (boiling point 274 °C) and trioctylamine (boiling point 365 °C), yielding 22 nm iron oxide nanoparticles. Sun et al. synthesized monodisperse MFe₂O₄ nanoparticles (M = Fe, Co, Mn) in phenyl ether (boiling point 258 °C, 4 nm) and benzyl ether (boiling point 298 °C, 6 nm).⁴³ With increasing reaction temperature, the particle size increased, resulting in much higher growth rates.

Comparatively, mixed metal oxides are quite difficult to synthesize, and it is more difficult to control their sizes, shapes, compositions, and crystallinity than those of monometallic oxides. Mixed metal oxide synthesis has its own intricacies because it involves more than one metal precursor with different decomposition temperatures. Bau et al. have reported the thermal decomposition of Ni and Fe precursors to produce mixed metal oxides with tuneable compositions of Fe and Ni.⁴⁴ The poor solubility of iron in the NiO lattice is known from its equilibrium phase diagram. The decomposition time of oleate and the initial ratio of nickel to iron afford different shapes of NiFe₂O₄ nanoparticles, viz. stars, cubes and spheres (see Fig. 3A). The shape of the nanoparticles changes with increasing reaction time as well as by changing the composition. As the iron content in the reaction mixture increases, the concentration of the strong absorbing ligand oleic acid (OA) decreases, leading to stable NiFe₂O₄ spherical nanoparticles. Star-shaped NiFe₂O₄ is obtained because OA binds the nuclei immediately and arrests growth when the nuclei concentration is high (see Fig. 3A). Muscas et al. showed the synthesis of CoFeO₄ oxide nanoparticles with different residual oxygen contents and studied its effect on their morphology.⁴⁵ The formation of CoFe₂O₄ nanoparticles under ambient conditions and in vacuum can be explained using LaMer theory (see Fig. 3B). Under ambient conditions, the nucleation and growth of nanoparticles can be separated by controlling the amount of oxygen available for the nucleation and growth phases (Fig. 3C). In vacuum or in restricted residual oxygen, partial nucleation occurs. During the growth process, monomer consumption by the nanoparticles overlaps with the formation of small nuclei. Ripening was seen to facilitate the formation of self-focusing monodispersed nanoparticles (Fig. 3D).

Hutchison et al. prepared In₂O₃, ITO/In₂O₃ core/shell, and In₂O₃/ITO/In₂O₃ core/shell/shell nanoparticles by sequential addition of metal precursor.⁴⁶ A Sn:In ∼ 1:9 mole ratio was used for the synthesis of ITO, followed by immediate addition of a solution of indium(III) acetate in oleylalcohol by maintaining a fixed addition rate (0.35 mL min⁻¹). The amount of oleylalcohol plays an important role in determining the particle size because the oxide formation is initiated through an esterification process that involves the alcohol. Increasing the amount of oleylalcohol increases the size of the Fe₂O₃ nanoparticles. Ming et al. have reported the synthesis of oxides of Mn, Nb, Zn, Cr, Ga, Fe, Cd, In, Co, Ni, Mo, Sn, Pb, and Sb using metal stearate precursors without any solvent and surfactant. They were able to control the sizes of the nanoparticles in the range of 7 nm to 50 nm.⁴⁷ Here, stearate was used as a surface stabilizing agent to control the sizes and shapes of the metal oxides. Jeong et al. reported a digestive ripening method for the synthesis of monodisperse sub-10 nm lanthanide oxide. In the first step, non-uniform Ln₂O₃ nanodiscs were obtained through decomposition of Ln–Ac in oleic acid at 280 °C. Non-uniform seeds were ripened with oleic acid, which acts as a digestive ripening agent, to obtain monodisperse nanodiscs in the range of 10 nm, such as Ho₂O₃ (7 nm), La₂O₃ (7 nm), Eu₂O₃ (7.5 nm), Gd₂O₃ (6 nm), Tb₂O₃ (6.4 nm), Dy₂O₃ (6.8), Ho₂O₃ (6.4 nm), Er₂O₃ (8.5 nm), Tm₂O₃ (4.2 nm) and Y₂O₃ (7.9 nm); the approximate thickness was 1.1 nm.⁴⁸ Details of the reaction parameters for the synthesis of various metal oxides synthesized by thermal decomposition are summarized in Table 3.

---

Analysis of the above reports and the vast body of literature shows that while the thermal decomposition method is useful for the synthesis of monodisperse, monocrystalline and shape-directed metal oxide nanoparticles, no quantitative information is available on the conversion, reaction kinetics and various by-products that are generated during decomposition. Inline or in situ techniques to measure growth and nucleation rates will greatly enable reliable use of that information to fabricate these materials in large quantities.

4. Solvothermal synthesis

Chemical processes performed in closed vessels above ambient pressure and temperature are known as solvothermal processes. During solvothermal methods, reaction vessels or autoclaves are operated in a temperature range of 100 to 1000 °C and a pressure range of 1 to 10000 bar.⁵⁷ Due to the high pressure and temperature, interactions between reactants are facilitated and single crystalline products are obtained. Products obtained with this method are highly selective and reproducible in terms of purity, crystallinity and morphology. Large scale synthesis using this method is intricate due to the high temperature, high pressure and prolonged reaction time. However, depending upon the value of the material, it can be adapted for scale-up. Abdul Rashid et al. have shown the transformation of iron oxide into 0D to 3D structures using a hydrothermal approach.⁵⁸ Time-dependent studies revealed the formation of single crystalline α-Fe₂O₃ nanocubes after 8 hours; these were transformed into smaller γ-Fe₂O₃ nanoparticles due to deformation of the nanocubes after 10 hours. Time-dependent TEM images showed that the nanocubes were obtained from quantum dots of α-Fe₂O₃ capped with oleate and coordinated through oleylamine, which formed aggregates and transformed into nanocubes due to the equivalent growth rates of the (012), (112) and (102) planes. Yang et al. synthesized monodisperse polyhedrons of iron oxide nanoparticles without surfactants.⁵⁹ In the course of formation, the iron oxide edges and corners were etched selectively due to the higher surface energy at the edges and corners than the 102 plane of the octahedra under assistance of Cl⁻ and H⁺ ions. After attaining equilibrium, unchanged truncated octahedra (700 nm) were observed. Kim et al. demonstrated the synthesis of monocrystalline TiO₂ in toluene in the presence of oleic acid as a surfactant at 250 °C for 20 h.⁶⁰ The shape of anatase TiO₂ nanoparticles can be tuned from spherical to elongated dumbell shapes, such as nanorods, either by increasing the concentration of oleic acid with respect to titanium isopropoxide (TIP) or increasing the concentration of TIP in toluene. Yan et al. demonstrated the hydrothermal synthesis of rutile TiO₂ nanotubes in NaOH water ethanol solution with diameters of 20 nm and lengths up to several microns.⁶¹ The formation of rutile phase requires higher temperatures; however, with a hydrothermal process, it can be achieved at a low temperature (130 °C). Li et al. synthesized near-monodisperse TiO₂ nanoparticles and metal-ion-doped (such as Sn⁴⁺, Fe³⁺, Co²⁺, and Ni²⁺) TiO₂ nanocomposites by solvothermal reactions using lauryl alcohol (LA) as the solvent as well as a surfactant.⁶² The sizes, shapes, dispersibilities, and components of the TiO₂ nanoparticles were controlled through fine tuning of the reaction parameters. When 6 mmol of LA was used, small aggregated particles were obtained. When the amount of LA was increased to 21 mmol and 75 mmol, nanorods 25 nm and 50 nm in length were formed, respectively. When NH₄CO₃ was introduced into the reaction mixture, the morphology of the TiO₂ nanoparticles was seen to change. When 4 mmol NH₄CO₃ and 21 mmol LA were used, elongated particles were obtained. On the other hand, when the amount of NH₄CO₃ was increased (25 mmol), aggregated spherical particles were observed. The decomposition of NH₄CO₃ provides H₂O for a hydrolyzation reaction, which aids rapid nucleation and produces smaller TiO₂ particles. Additionally, LA serves as a hydrolysing agent and also as a coordinating agent to promote anisotropic growth. Solvothermal and hydrothermal methods are being used for the preparation of monocrystalline metal oxide nanopowders.

A variety of sizes and shapes of metal oxide nanopowders can be obtained by adjusting the reaction parameters (ligand to metal precursor ratio, temperature, pressure, reaction time, etc.). The reaction parameters for a range of metal oxide nanoparticles synthesized by solvothermal methods are summarized in Table 4. In most literature studies, the data obtained for the synthesis of particular (specific) metal oxides cannot be correlated due to a lack of sufficient information on the nucleation and growth kinetics and their relationship with the % conversion of the metal precursors. In order to develop scalable synthesis methods, one needs to study the kinetics of the formation of metal oxide nanoparticles, which accounts for the effects of individual components (types of solvent, ligand and precursor and their concentrations at constant temperature and pressure) and the quantitative yield of nanoparticles. Although the solvothermal synthesis route produces monocrystalline materials and has potential to synthesize monodisperse and uniformly shaped metal nanoparticles, scale-up of solvothermal methods is quite complex due to their high temperature and pressure conditions; very little information is available on the transient transformations of the structures of materials under these conditions.

---

5. Continuous flow synthesis of metal oxide nanoparticles

Recently, flow synthesis has emerged as a mature technique to synthesize nanocrystalline materials due to its faster heat and mass transfer rates, rapid mixing, precise control of reaction parameters, and superior physical stability.^73,74 A typical setup for continuous flow synthesis is presented in Fig. 4(A). The high surface area-to-volume ratios and decreased diffusional dimensions of the tube (milli) and micro-reactor allows reactions to be carried out in a rapid and controllable manner. Control over the nucleation and growth kinetics can be achieved by minimizing local variations in temperature and concentration to obtain monodisperse particles in an efficient way. Zhang et al. employed microfluidic technology for the controlled synthesis of TiO₂ nanoparticles and showed great control over the sizes and phases of the TiO₂ nanoparticles.³ Once the TiO₂ nanoparticles are formed, they can be enwrapped by organic solvents, which prevents further growth of the nanoparticles in the microfluidic environment. Bagbhanzadeh et al. showed shape tuning of TiO₂ nanocrystals from rods to spherical nanoparticles using a continuous flow solvothermal method.⁷⁵ For the synthesis of TiO₂, titanium tetraisopropoxide (TTIP) was taken in oleic acid and toluene and injected in a stainless steel tube coiled reactor maintained at 250 °C with a residence time of 10 min. Nanorods and spherical TiO₂ nanoparticles were obtained at 250 °C and 300 °C, respectively; this is ascribed to the relative rates of the stages, governed by thermodynamics and reaction kinetics. Face-directed growth observed at low temperature as well as interparticle ripening and inter-particle Ostwald ripening leads to spherical particle growth at higher temperatures. Hakuta et al. synthesized BaTiO₃ nanoparticles by a continuous flow hydrothermal method. The temperature of the system alters the size as well as the Ti conversion.⁷⁶ At subcritical temperatures (300 °C to 380 °C), the size of the BaTiO₃ nanoparticles decreased with increasing temperature up to 380 °C due to rapid nucleation. Meanwhile, at supercritical temperatures, a wide particle size range from 10 to 150 nm was observed due to undesired mixing around the mixing point at 400 °C. Sue et al. reported a continuous hydrothermal synthesis of Pr-doped CaTiO₃ nanoparticles from Pr(NO₃)₃, Ca(NO₃)₂, and TiO₂ sol at 673 K and 30 MPa using a T-type micro-mixer. Increasing the KOH molality, residence time, and Ca/Ti ratio led to the formation of CaTiO₃.⁷⁷ Pr-doped CaTiO₃ nanoparticles were produced through three steps: dissolution of the TiO₂ sol, formation of a hydroxide precursor containing Ca²⁺ and Ti⁴⁺ through hydrolysis, and nucleation-growth of CaTiO₃ through dehydration condensation. Giesche et al. have reported the synthesis of monodisperse SiO₂ nanoparticles by a controlled growth process in a continuous flow tube coiled reactor. The effects of the concentrations of seed, TEOS and ammonia and there different addition rates were studied to optimize the reaction conditions for the preparation of monodisperse SiO₂ particles. Silica particles up to 3.6 μm in size and mass fractions of up to 10 volume% were prepared using a continuous flow approach. In a typical seed-mediated synthesis, the size is dependent on the seed to TEOS ratio. As the seed to seed to TEOS ratio decreased from 1:1 to 1:100, the particle sizes increased from 135 ± 3 to 469 ± 9 nm, respectively. Khan et al. have demonstrated the use of segmented flow reactors for the synthesis of SiO₂ nanoparticles.⁷⁸ TEOS was used as a silica source and NH₃ as a precipitating agent. SiO₂ nanoparticles obtained by laminar flow synthesis have broad size distributions due to axial distribution in the flow pattern, commonly referred to as axial dispersion. To eliminate axial dispersion, a gas phase was injected to generate segments in the microfluidic reactor. Internal recirculation of fluid in the plugs allowed homogeneous mixing, which minimizes local variation during the nucleation and growth process and produces a narrow size distribution of silica nanoparticles. Corradi et al. prepared monodisperse spherical SiO₂ nanopowder using a continuous microwave hydrothermal method.⁷⁹ The microwave decreases the reaction time due to its rapid heating rate, which enhances the rates of the nucleation, growth and crystallization processes. The obtained silica nanoparticles were 100 nm in size without surfactant only at the lowest flow rate of 43 mL min⁻¹. This shows that when scaling up SiO₂ synthesis, it is necessary to have precise control over the flow rate, which ultimately influences the size and aggregation.

Wacker et al. employed microfluidic technology to produce functionalised silica nanoparticles.⁸⁰ Silica nanoparticles with different sizes were tuned by altering the reaction time and the concentrations of the reactants. Flow focusing devices were made from poly dimethyl siloxane (PDMS) so that the two streams of TEOS and hydrolysing solution mixed before droplet formation and the droplets were obtained by pinch-off due to the side-by-side flowing streams of alkoxide solution and hydrolysing mixture on a microfluidic chip using a Fluorinert oil as the continuous phase. The growth rate of SiO₂ particles was higher in the microfluidic device; this enabled precise tuning of the size of the silica particles by changing the reaction time. Bondoli et al. prepared spherical ZrO₂ nanoparticles using a continuous flow microwave hydrothermal reactor. The hydrothermal microwave process decreases the reaction time because the homogeneous heating enhances the reaction kinetics.⁸² ZrO₂ nanoparticles prepared by a continuous flow microwave hydrothermal method (10 ± 5 nm) are smaller than those obtained by a traditional sol gel method (220 ± 20 nm). Residence time is seen to play an important role in scaling up the synthesis of size-controlled ZrO₂ nanoparticles. The continuous flow method is more efficient than interrupted injection (also known as the stop flow method) because the steady state synthesis retains the quality and morphology of the product and increases the production rate without clogging the channels. Salazar-Alvarez et al. prepared variable-sized Fe₃O₄ nanoparticles using flow synthesis.⁸³ Ferric chloride, ferrous chloride and NaOH were used to precipitate iron oxide at 80 °C with 10 s residence time. The obtained magnetite nanoparticles had a narrow size distribution in the range of 2 to 7 nm. The effects of the concentration of NaOH were studied to change the size of the iron oxide nanoparticles. At high concentration, sodium hydroxide provides a large number of nucleating sites that lead to the formation of a large number of nuclei. The total iron ion concentration influences the reaction kinetics and alters the size of the nanoparticles. As the iron ion concentration increases, a chemical gradient is generated in the system, where iron ions diffuse to the areas of lower density surrounding the Fe₃O₄ nanoparticles which are present as nucleating sites which contributes to growth of Fe₃O₄ nanoparticles. Kumar et al. prepared iron oxide nanoparticles using a stable passively-driven capillary-based droplet reactor⁸⁴ in which dextran acted as a capping agent. Ferric chloride, ferrous chloride and ammonia were injected through an auxiliary capillary in the main stream of inert phase octadecene to generate droplets in the capillary at 60 °C. The obtained iron oxide nanoparticles were 3.6 nm in size and showed a superparamagnetic nature, with a saturation magnetization of 58 emu g⁻¹. Frenz et al. have reported an electro-coalescence-based technique in a microfluidic reactor to synthesize iron oxide nanoparticles.⁸⁵ Two droplets of iron precursor and ammonia were fused inline in perflurocarbon oil by applying a voltage (U = 200 V) across the capillary. The droplet-based system is suitable for slow reactions because mixing is limited to each droplet hemisphere. Inline pairing of the droplets by hydrodynamic coupling mixing is limited to each droplet sphere. Meanwhile, by electro-coalescence, the two reagents from the droplets mix homogenously in each hemisphere, which produces iron oxide nanoparticles with reproducibility. The monocrystalline iron oxide is 4 nm in size and shows a superparamagnetic nature. Hassan et al. also studied iron oxide synthesis in a single nozzle co-axial flow operating under a laminar flow regime.⁸⁶ Trimethyl ammonium hydroxide solution was injected through the outer capillary, whereas the iron precursor was infused through the inner capillary by maintaining a Q_out:Q_in ratio of 4. Clogging was avoided by slower diffusion of the iron solution towards the wall of the reactor with the aid of trimethyl ammonium hydroxide. Darr et al. prepared CoO nanoparticles in a continuous hydrothermal reactor under laminar and turbulent flows.⁸⁸ CoO nanoparticles were prepared by mixing a high temperature (450 °C) solution of H₂O₂ with cobalt acetate solution at a constant flow rate and 24 mPa pressure at a production rate of 20 g h⁻¹, with 91% yield by mass. The size of the nanoparticles is reported to change with variation in the flow pattern with respect to the concentration of cobalt precursor. The nucleation and growth kinetics were drastically altered due to transition of the flow from a laminar to a turbulence flow. Fig. (4B) indicates the effects of the flow pattern on the size of the nanoparticles; this can be explained by understanding the nucleation kinetics and growth kinetics in the flow, which is quite different from a batch process. In the case of laminar mixing at higher concentrations, an increase in super saturation upon mixing affords smaller particles. The growth of nanoparticles in a laminar flow is diffusion limited; hence, the collision of smaller nanoparticles becomes a rare event in a laminar flow, which produces small particles. In turbulent mixing at higher concentrations, super saturation is achieved very rapidly due to back mixing, and the growth of nanoparticles proceeds more favourably through coalescence due to collisions between nuclei (see Fig. 4B). Hussain et al. applied a continuous microwave flow synthesis process to obtain phase pure nanosized SnO₂.⁸⁹ This technique was shown to be very reliable for large scale rapid synthesis of SnO₂ at relatively low temperatures and pressures. As the concentration of Sn precursor was increased from 0.25 M to 0.75 M, the particle size increased from 3.74 to 7.13 nm and the particles agglomerated; there was no change in morphology with changing concentration of the Sn precursor. An overview of the observations based on continuous flow synthesis of metal oxide nanomaterials is provided in Table 5.

---

6. Properties and applications

Metal oxides are important for optoelectronics, catalysis, energy storage, memory storage, pigments, gas sensing and biomedical applications. Iron oxides, ZnO, MgO, SnO₂, TiO₂, SiO₂ and rare earth metal oxides have been explored exhaustively due to their size and shape-dependent properties. Similarly, iron oxide exhibits size-dependent properties and is freely available in nature in the forms of α-Fe₂O₃ (hematite), γ-Fe₂O₃ (maghemite), α-FeOOH (geothite), γ-FeOOH (lepidocrocite), and Fe₃O₄ (magnetite). Hematite and goethite were used in the paintings in the Altamira cave in northern Spain. It is evident that the sizes, compositions and morphologies of iron oxide nanoparticles alter their final properties and performance. For promising applications in cancer therapy and MRI, size-dependent magnetic properties require study. Demortiere et al. showed that size and composition have significant impacts on saturation magnetization (M_s)⁹⁰ (see Fig. 5A). The M_s value increases as the size of iron oxide nanoparticles increases. At 2.5 nm, M_s is 29 emu g⁻¹ and at 14 nm, M_s is 77 emu g⁻¹; these values are close to those of bulk γ-Fe₂O₃ (74 emu g⁻¹) and Fe₃O₄ (90 emu g⁻¹). This change in M_s is attributed to the fractions of atoms on the nanoparticle surfaces and the surface environments of the atoms, which induce changes in the number and symmetry of co-ordination. Due to the magnetic properties of iron oxides, they are interesting for use in targeted drug delivery applications, cell separation and magnetic resonance imaging (MRI) agents for diagnosis. The superparamagnetic nature of iron oxide nanoparticles is used to generate heat energy by applying an alternating external magnetic field for thermo-responsive drug delivery, where iron oxide nanoparticles with embedded drugs in a thermo-responsive polymer matrix act as carriers. For MRI, low cost, biocompatible and magnetic iron oxide nanoparticles have been extensively investigated as T2 contrast agents to replace conventionally used paramagnetic metal chelates, which act by shortening T1 relaxation times.⁹¹ Iron oxide nanoparticles functionalised with dimercaptosuccinic acid act by shortening the T2 relaxation time (the time constant describes the relaxation constant, usually of protons interfering with each other). In the field of an external magnet, the iron oxide enhances the local field of neighboring protons when the protons resume their original state after a radiofrequency pulse. Due to the transition from high energy to low energy, photons are released in the form of energy, which is imparted into MRI images (see ref. 91, Fig. 3).

ZnO has a wide band gap (3.4 eV) and appears to be a less expensive, less toxic material with size dependent fluorescence properties; it has been used in sunscreen lotions, anti-reflectant layers and field emission transistors. A nano ZnO oxide prepared by Nanophase is used under the brand name Z-Cote by BASF for skin protection from sunlight. Self-cleaning products have been sold in the market, such as glass, textiles and plastics. Zhang et al. reported a combination of anti-reflecting and self-cleaning properties of a coating of SiO₂ nanoparticles and TiO₂ nanoparticles obtained by a simple electrostatic attraction method.⁹² A SiO₂ nanoparticle single layer was used to provide a porous structure with a low refractive index, which acts as an anti-reflecting material. Coating nanosized TiO₂ on the SiO₂ surface provided a self-cleaning and hydrophilic layer. Even with the high refractive index of the TiO₂ nanoparticles, the SiO₂–TiO₂ coating showed 99% transmittance and can be used as self-cleaning glass for houses and cars. Xu et al. demonstrated the synthesis of ZnO quantum dots with tuneable sizes in the range of 2.0 to 7.8 nm which showed a broad range of fluorescence emissions from violet to orange.⁹³ The size-dependent fluorescence properties were associated with an increase in the band gap between the conduction band and valence band as the size decreased. Hence, this ZnO composite material has potential applications in optoelectronics for security purposes. Diep et al. prepared ZnO nanotetrapods which emit green light upon UV irradiation.⁹⁴ These ZnO nanotetrapods embedded in polydimethylsiloxane (PDMS) sheets can be used for optoelectronics applications (see Fig. 5B). MgO is a white crystalline material with low electrical conductivity and high thermal conductivity. It is often used as an insulator, catalyst support, laxative and paper de-acidifier. MgO nanomaterials are being extensively studied as bactericidal agents, catalyst supports and gas separators. MgO nanocrystals exhibit fluorescence emission in the visible light region with high temperature stability. Xie et al. reported tunable fluorescence of MgO nanoparticles by surface functionalization.⁹⁵ The MgO spherical nanoparticles prepared at variable temperatures, including 320 °C, 400 °C and 480 °C, were 150 nm in size and had grain sizes of 27.6, 19.3, and 14.6 nm, respectively; these nanoparticles showed blue to violet, green and yellow photoluminescence (see Fig. 5C). Due to its high photoluminescence stability, MgO has potential applications in biomedical and biology fields as fluorescent labels. Scale-up of the synthesis of metal oxide nanoparticles suffers from a lack of stability of the particles in powder and dispersed forms; hence, the use of metal oxide nanoparticles is limited. For practical applications, the metal oxide powders should be dispersible in solution; also, the metal ions should preserve their oxidation states, and excellent control of the sizes and shapes of the nanoparticles is necessary. These many challenges can be conquered by understanding the surface chemistry, formation mechanisms and effects of post-synthetic treatments on metal oxide nanoparticles.

7. Conclusions

Microwave-assisted synthesis, thermal decomposition and solvothermal methods provide monocrystalline nanomaterials as well as modular routes to scaleup. These wet chemical routes enable size and shape control of metal oxide nanoparticles. However, the growth and nucleation kinetics of metal oxides cannot be generalised to all metal oxides due to the high sensitivity of the material to slight changes in the reaction parameters (including influences on the formation kinetics of the nanomaterials).

It is necessary to study the effects of each reaction parameter on the sizes, shapes, and compositions of nanomaterials. This is because the growth and nucleation kinetics are highly sensitive to reaction parameters, which can be correlated with changes in the sizes, shapes, and compositions of nanoparticles to predict optimum conditions to attain better control over the properties of metal oxide nanoparticles. For controlled and large-scale production, more attention must be paid to understanding the role of each (reagent) component. Scaleup of batch synthesis to large volumes is associated with limitations, such as mixing, mass and heat transfer, and batch-to-batch variations in products. Continuous flow synthesis has potential to afford nanoparticles in large quantity without compromising the quality of the nanomaterials, which can meet industrial requirements in terms of the quality and quantity of the nanomaterials. Increasing production at low cost is likely to remove the main impediment to realize industrial scale application.

In situ monitoring of particle growth at the nanoscale in solution is a challenging task which will provide avenues for understanding mass transfer and diffusion of surface atoms. Separation of nanomaterials is tedious and energy intensive; thus, special attention should be paid to resolving the separation issue. Production of 2D and 3D nanostructures of metal oxides using flow synthesis is expected to show interesting potential. Automated continuous flow synthesis using wet chemical synthesis will be a future technology for the production of metal oxide nanoparticles with great control over the reaction conditions. Metal oxide nanoparticles with uniform sizes are being realized for applications in the areas of biomedicine, energy storage, optoelectronics, and catalysis.

Recommendations for further work and a few challenges

Scale-up of microwave-assisted synthesis in batches is usually limited up to 2 to 3 L due to the limitation of the penetration depth of microwave radiation. It is always advised to use polar solvents in microwave-assisted synthesis because the dipolar heating mechanism helps to achieve rapid heating of the reaction solution, leading to spontaneous nucleation and fast synthesis. The process of translation of lab scale microwave-assisted synthesis to industrial scale production faces many challenges in terms of continuous microwave generator design, reactor design, reactor material, and fittings. Microwave reactors made of quartz are often difficult to connect with tubes and mixers. Teflon and metallic fittings are not feasible with microwave radiation because the former obstructs the path of the microwaves and the latter is known to induce sparks in the reactors. Hence, the development of microwave-assisted synthesis for scalable production of metal oxide nanoparticles can be realized after resolving the abovementioned issues. Mass production of metal oxides using solvothermal methods can be envisaged after testing sustainability processes for longer times under high pressure and temperature conditions without any failure which is important for human safety. Better understanding of microwave-assisted and thermal decomposition methods will enable us to produce uniform anisotropic metal oxide nanoparticles with great control over the growth kinetics.

The abovementioned observations from the continuous flow synthesis of nanomaterials must be further explored for in situ determination of the nucleation and growth of nanoparticles. In situ monitoring and automatic data analysis systems will solve the issues of the optimization reaction parameters and reactor selection; this requires various synergistic areas of expertise (chemists, chemical/electronic/mechanical/instrumental engineers, mathematicians, and software developers).

8. Abbreviations

TOP	Tetradecyl-hosphonic acid
OA	Oleic acid
OAm	Oleyl amine
TMAO	Trimethyl N-oxide
TOA	Trioctyl amine
PA	Palmitic acid
HAD	Hexadecyl amine
L	Length
W	Width
D	Diameter
ID	Inner diameter
OD	Inner diameter
FITC	Fluorescein isothiocynate
PEG	Polyethylene glycol
EG	Ethylene glycol
HMT	Hexamethylene tetraamine
PPG	Polypropylene glycol
MPA	Mercaptopropionic acid
CTAB	Cetyl tetraammonium bromide

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

A. N. acknowledges CSIR New Delhi for a Senior Research Fellowship. AAK and BLVP acknowledge the support from DST Advanced Manufacturing Technology.

References

1. M. E. Franke, T. J. Koplin and U. Simon, Small, 2006, 2, 36–50 CrossRef PubMed .
2. S. Brovelli, N. Chiodini, R. Lorenzi, A. Lauria, M. Romagnoli and A. Paleari, Nat. Commun., 2012, 3, 690 CrossRef PubMed .
3. L. Zhang, Y. Li, Q. Zhang, G. Shi and H. Wang, Chem. Lett., 2011, 40, 1371–1373 CrossRef .
4. Z. Zhang, C. Dong, C. Yang, D. Hu, J. Long, L. Wang, H. Li, Y. Chen and D. Kong, Adv. Synth. Catal., 2010, 352, 1600–1604 CrossRef .
5. S. E. Skrabalak and R. L. Brutchey, Chem. Mater., 2016, 28, 1003–1005 CrossRef .
6. J. M. Köhler, S. Li and A. Knauer, Chem. Eng. Technol., 2013, 36, 887–899 CrossRef .
7. R. L. Hartman, J. P. McMullen and K. F. Jensen, Angew. Chem., Int. Ed., 2011, 50, 7502–7519 CrossRef PubMed .
8. L. Zhang and Y. Xia, Adv. Mater., 2014, 26, 2600–2606 CrossRef PubMed .
9. S. Duraiswamy and S. A. Khan, Small, 2009, 5, 2828–2834 CrossRef PubMed .
10. V. Sebastian, M. Arruebo and J. Santamaria, Small, 2014, 10, 835–853 CrossRef PubMed .
11. S. Komarneni and R. Roy, Mater. Lett., 1985, 3, 165–167 CrossRef .
12. R. Gedye, F. Smith and K. Westaway, Tetrahedron Lett., 1986, 26, 279 CrossRef .
13. R. J. Giguere, T. L. Bray, S. M. Duncan and G. Majetich, Tetrahedron Lett., 1986, 27, 4945–4948 CrossRef .
14. D. E. Clark and W. H. Sutton, Annu. Rev. Mater. Sci., 1996, 26, 299–331 CrossRef .
15. Y. J. Zhu, W. W. Wang, R. J. Qi and X. L. Hu, Angew. Chem., 2004, 116, 1434–1438 CrossRef .
16. X. Hu, J. Gong, L. Zhang and J. C. Yu, Adv. Mater., 2008, 20, 4845–4850 CrossRef .
17. W.-W. Wang, Y.-J. Zhu and M.-L. Ruan, J. Nanopart. Res., 2007, 9, 419–426 CrossRef .
18. W.-W. Wang, Y.-J. Zhu, G.-F. Cheng and Y.-H. Huang, Mater. Lett., 2006, 60, 609–612 CrossRef .
19. A. V. Nikam, A. Arulkashmir, K. Krishnamoorthy, A. A. Kulkarni and B. Prasad, Cryst. Growth Des., 2014, 14, 4329–4334 CrossRef .
20. I. Bilecka, I. Djerdj and M. Niederberger, Chem. Commun., 2008, 886–888 RSC .
21. M. I. Dar, A. K. Chandiran, M. Gratzel, M. K. Nazeeruddin and S. A. Shivashankar, J. Mater. Chem. A, 2014, 2, 1662–1667 RSC .
22. T. Hyeon, S. S. Lee, J. Park, Y. Chung and H. B. Na, J. Am. Chem. Soc., 2001, 123, 12798–12801 CrossRef PubMed .
23. H. Hu, H. Yang, P. Huang, D. Cui, Y. Peng, J. Zhang, F. Lu, J. Lian and D. Shi, Chem. Commun., 2010, 46, 3866–3868 RSC .
24. S. Guru, D. Mishra, S. S. Amritphale and S. Joshi, Colloid Polym. Sci., 2016, 294, 207–213 CrossRef .
25. Z. Ai, K. Deng, Q. Wan, L. Zhang and S. Lee, J. Phys. Chem. C, 2010, 114, 6237–6242 CrossRef .
26. E. A. Osborne, T. M. Atkins, D. A. Gilbert, S. M. Kauzlarich, K. Liu and A. Y. Louie, Nanotechnology, 2011, 23, 215602 CrossRef PubMed .
27. L. Wu, H. Yao, B. Hu and S.-H. Yu, Chem. Mater., 2011, 23, 3946–3952 CrossRef .
28. H. Zhou, R. Yi, J. Li, Y. Su and X. Liu, Solid State Sci., 2010, 12, 99–104 CrossRef .
29. R. Jalal, E. K. Goharshadi, M. Abareshi, M. Moosavi, A. Yousefi and P. Nancarrow, Mater. Chem. Phys., 2010, 121, 198–201 CrossRef .
30. S. Cho, S.-H. Jung and K.-H. Lee, J. Phys. Chem. C, 2008, 112, 12769–12776 CrossRef .
31. D. Sharma, S. Sharma, B. S. Kaith, J. Rajput and M. Kaur, Appl. Surf. Sci., 2011, 257, 9661–9672 CrossRef .
32. I. Bilecka, P. Elser and M. Niederberger, ACS Nano, 2009, 3, 467–477 CrossRef PubMed .
33. A. Pimentel, D. Nunes, P. Duarte, J. Rodrigues, F. M. Costa, T. Monteiro, R. Martins and E. Fortunato, J. Phys. Chem. C, 2014, 118, 14629–14639 CrossRef .
34. L. Li and J. Ren, Mater. Res. Bull., 2006, 41, 2286–2290 CrossRef .
35. A. S. Bhatt, D. K. Bhat, C.-w. Tai and M. S. Santosh, Mater. Chem. Phys., 2011, 125, 347–350 CrossRef .
36. S. Chen, Y. Zhao, B. Sun, Z. Ao, X. Xie, Y. Wei and G. Wang, ACS Appl. Mater. Interfaces, 2015, 7, 3306–3313 CrossRef PubMed .
37. X. Wang, J. Tian, C. Fei, L. Lv, Y. Wang and G. Cao, RSC Adv., 2015, 5, 8622–8629 RSC .
38. R. Bhavesh, A. V. Lechuga-Vieco, J. s. Ruiz-Cabello and F. Herranz, Nanomaterials, 2015, 5, 1880–1890 CrossRef PubMed .
39. S. C. Padmanabhan, D. Ledwith, S. C. Pillai, D. E. McCormack and J. M. Kelly, J. Mater. Chem., 2009, 19, 9250–9259 RSC .
40. S. Vijayakumar, A. K. Ponnalagi, S. Nagamuthu and G. Muralidharan, Electrochim. Acta, 2013, 106, 500–505 CrossRef .
41. J. Park, K. An, Y. Hwang, J.-G. Park, H.-J. Noh, J.-Y. Kim, J.-H. Park, N.-M. Hwang and T. Hyeon, Nat. Mater., 2004, 3, 891–895 CrossRef PubMed .
42. H.-W. Song, N.-Y. Kim, J.-e. Park, J.-H. Ko, R. J. Hickey, Y.-H. Kim and S.-J. Park, Nanoscale, 2017, 9, 2732–2738 RSC .
43. S. Sun, H. Zeng, D. B. Robinson, S. Raoux, P. M. Rice, S. X. Wang and G. Li, J. Am. Chem. Soc., 2004, 126, 273–279 CrossRef PubMed .
44. J. A. Bau, P. Li, A. J. Marenco, S. Trudel, B. C. Olsen, E. J. Luber and J. M. Buriak, Chem. Mater., 2014, 26, 4796–4804 CrossRef .
45. G. Muscas, G. Singh, W. Glomm, R. Mathieu, P. A. Kumar, G. Concas, E. Agostinelli and D. Peddis, Chem. Mater., 2015, 27, 1982–1990 CrossRef .
46. A. W. Jansons and J. E. Hutchison, ACS Nano, 2016, 10, 6942–6951 CrossRef PubMed .
47. Q. Dou and K. M. Ng, Powder Technol., 2016, 301, 949–958 CrossRef .
48. J. Jeong, N. Kim, M.-G. Kim and W. Kim, Chem. Mater., 2016, 28, 172–179 CrossRef .
49. H. Zhang, X. Wang, C. Chen, C. An, Y. Xu, Y. Dong, Q. Zhang, Y. Wang, L. Jiao and H. Yuan, Inorg. Chem. Front., 2016, 3, 1048–1057 RSC .
50. S. Mehra, A. Bergerud, D. J. Milliron, E. M. Chan and A. Salleo, Chem. Mater., 2016, 28, 3454–3461 CrossRef .
51. J. Buha, I. Djerdj and M. Niederberger, Cryst. Growth Des., 2007, 7, 113–116 CrossRef .
52. C. Cavelius, K. Moh and S. Mathur, Cryst. Growth Des., 2012, 12, 5948–5955 CrossRef .
53. Z. C. Orel, A. Anzlovar, G. Drazic and M. Zigon, Cryst. Growth Des., 2007, 7, 453–458 CrossRef .
54. J. Paek, C. H. Lee, J. Choi, S.-Y. Choi, A. Kim, J. W. Lee and K. Lee, Cryst. Growth Des., 2007, 7, 1378–1380 CrossRef .
55. G. Leem, S. Sarangi, S. Zhang, I. Rusakova, A. Brazdeikis, D. Litvinov and T. R. Lee, Cryst. Growth Des., 2008, 9, 32–34 CrossRef .
56. Q. Li, J. Wang, Y. He, W. Liu and X. Qiu, Cryst. Growth Des., 2009, 9, 3100–3103 CrossRef .
57. R. I. Walton, Chem. Soc. Rev., 2002, 31, 230–238 RSC .
58. N. M. Abdul Rashid, C. Haw, W. Chiu, N. H. Khanis, A. Rohaizad, P. Khiew and S. Abdul Rahman, CrystEngComm, 2016, 18, 4720–4732 RSC .
59. S. Yang, Y. Xu, Y. Sun, G. Zhang and D. Gao, CrystEngComm, 2012, 14, 7915–7921 RSC .
60. C.-S. Kim, B. K. Moon, J.-H. Park, B.-C. Choi and H.-J. Seo, J. Cryst. Growth, 2003, 257, 309–315 CrossRef .
61. J. Yan, S. Feng, H. Lu, J. Wang, J. Zheng, J. Zhao, L. Li and Z. Zhu, Mater. Sci. Eng., B, 2010, 172, 114–120 CrossRef .
62. X. L. Li, Q. Peng, J. X. Yi, X. Wang and Y. Li, Chem. – Eur. J., 2006, 12, 2383–2391 CrossRef PubMed .
63. H. Zhou, M. Kang, D. Wu and B. Lv, CrystEngComm, 2016, 18, 5456–5462 RSC .
64. N. M. A. Rashid, C. Haw, W. Chiu, N. H. Khanis, A. Rohaizad, P. Khiew and S. A. Rahman, CrystEngComm, 2016, 18, 4720–4732 RSC .
65. A. T. Raj, K. Ramanujan, S. Thangavel, S. Gopalakrishan, N. Raghavan and G. Venugopal, J. Nanosci. Nanotechnol., 2015, 15, 3802–3808 CrossRef PubMed .
66. G. Clavel, M. G. Willinger, D. Zitoun and N. Pinna, Adv. Funct. Mater., 2007, 17, 3159–3169 CrossRef .
67. Y. Tan, X. Xue, Q. Peng, H. Zhao, T. Wang and Y. Li, Nano Lett., 2007, 7, 3723–3728 CrossRef .
68. A. L. Tiano, G. C. Papaefthymiou, C. S. Lewis, J. Han, C. Zhang, Q. Li, C. Shi, A. M. Abeykoon, S. J. Billinge and E. Stach, Chem. Mater., 2015, 27, 3572–3592 CrossRef .
69. S. Chaianansutcharit, O. Mekasuwandumrong and P. Praserthdam, Cryst. Growth Des., 2006, 6, 40–45 CrossRef .
70. X. L. Zhang, R. Qiao, J. C. Kim and Y. S. Kang, Cryst. Growth Des., 2008, 8, 2609–2613 CrossRef .
71. L. Kumari, W. Li, J. Xu, R. Leblanc, D. Wang, Y. Li, H. Guo and J. Zhang, Cryst. Growth Des., 2009, 9, 3874–3880 CrossRef .
72. Q. Tang, W. Zhou, W. Zhang, S. Ou, K. Jiang, W. Yu and Y. Qian, Cryst. Growth Des., 2005, 5, 147–150 CrossRef .
73. A. Shavel, D. Cadavid, M. Ibanez, A. Carrete and A. Cabot, J. Am. Chem. Soc., 2012, 134, 1438–1441 CrossRef PubMed .
74. T. W. Phillips, I. G. Lignos and R. M. Maceiczyk, Lab Chip, 2014, 14, 3172–3180 RSC .
75. M. Baghbanzadeh, T. N. Glasnov and C. O. Kappe, J. Flow Chem., 2013, 3, 109–113 CrossRef .
76. Y. Hakuta, H. Ura, H. Hayashi and K. Arai, Ind. Eng. Chem. Res., 2005, 44, 840–846 CrossRef .
77. K. Sue, S.-i. Kawasaki, T. Sato, D. Nishio-Hamane, Y. Hakuta and T. Furuya, Ind. Eng. Chem. Res., 2016, 55, 7628–7634 CrossRef .
78. S. A. Khan, A. Günther, M. A. Schmidt and K. F. Jensen, Langmuir, 2004, 20, 8604–8611 CrossRef PubMed .
79. A. B. Corradi, F. Bondioli, A. M. Ferrari, B. Focher and C. Leonelli, Powder Technol., 2006, 167, 45–48 CrossRef .
80. J. B. Wacker, I. Lignos, V. K. Parashar and M. A. Gijs, Lab Chip, 2012, 12, 3111–3116 RSC .
81. B. F. Cottam, S. Krishnadasan and M. S. Shaffer, Lab Chip, 2007, 7, 167–169 RSC .
82. F. Bondioli, A. Bonamartini Corradi, A. M. Ferrari and C. Leonelli, J. Am. Ceram. Soc., 2008, 91, 3746–3748 CrossRef .
83. G. Salazar-Alvarez, M. Muhammed and A. A. Zagorodni, Chem. Eng. Sci., 2006, 61, 4625–4633 CrossRef .
84. K. Kumar, A. M. Nightingale, S. H. Krishnadasan, N. Kamaly, M. Wylenzinska-Arridge, K. Zeissler, W. R. Branford and E. Ware, J. Mater. Chem., 2012, 22, 4704–4708 RSC .
85. L. Frenz, A. El Harrak, M. Pauly, S. Bégin-Colin, A. D. Griffiths and J. C. Baret, Angew. Chem., Int. Ed., 2008, 47, 6817–6820 CrossRef PubMed .
86. A. A. Hassan, O. Sandre, V. Cabuil and P. Tabeling, Chem. Commun., 2008, 1783–1785 RSC .
87. H. Tang, Y. He, B. Li, J. Jung, C. Zhang, X. Liu and Z. Lin, Nanoscale, 2015, 7, 9731–9737 RSC .
88. C. J. Denis, C. J. Tighe, R. I. Gruar, N. M. Makwana and J. A. Darr, Cryst. Growth Des., 2015, 15, 4256–4265 CrossRef .
89. M. Akram, A. T. Saleh, W. A. W. Ibrahim, A. S. Awan and R. Hussain, Ceram. Int., 2016, 42, 8613–8619 CrossRef .
90. A. Demortiere, P. Panissod, B. Pichon, G. Pourroy, D. Guillon, B. Donnio and S. Begin-Colin, Nanoscale, 2011, 3, 225–232 RSC .
91. P. Tartaj, M. P. Morales, T. Gonzalez-Carreño, S. Veintemillas-Verdaguer and C. J. Serna, Adv. Mater., 2011, 23, 5243–5249 CrossRef PubMed .
92. X.-T. Zhang, O. Sato, M. Taguchi, Y. Einaga, T. Murakami and A. Fujishima, Chem. Mater., 2005, 17, 696–700 CrossRef .
93. X. Xu, C. Xu, X. Wang, Y. Lin, J. Dai and J. Hu, CrystEngComm, 2013, 15, 977–981 RSC .
94. V. M. Diep and A. M. Armani, Nano Lett., 2016, 16, 7389–7393 CrossRef PubMed .
95. S. Xie, X. Han, Q. Kuang, Y. Zhao, Z. Xie and L. Zheng, J. Mater. Chem., 2011, 21, 7263–7268 RSC .

This journal is © The Royal Society of Chemistry 2018