1 Introduction

Metal oxide nanoparticles (MONPs) exhibit unique physical, chemical, optical and electronic properties with respect to those of their bulk counterparts due to the quantum confinement and more availability of the surface atoms than interior atoms for participating in any reaction. Applications of MONPs depend upon their properties such as surface area, shape, size, stability, crystallinity, anti-corrosiveness, conductivity, and photocatalytic activity etc. [1, 2]. It is highly desirable for the commercialization of MONPs to have good controllable qualities during synthesis. In general, the synthesis methods of MONPs can be divided into two groups: (i) physical methods such as ball milling, sputtering, laser ablation, electrospraying, electron beam evaporation etc. and (ii) chemical methods such as sol–gel method, polyol method, hydrothermal method, co-precipitation method, microemulsion technique, chemical vapor deposition etc. (Fig. 1). Physical methods are based on top to down strategy i.e., the synthesis is started from the bulk counterpart of any material which gets depleted systematically for the generation of the fine nanoparticles (NPs) while the chemical methods are mainly based on bottom to up approach i.e., the assembly of atoms or molecules to form distribution of different sizes of NPs. The main advantage of the chemical methods is that it allows the production of particles with a defined size, dimension, composition and structure, which could be useful for various applications such as catalysis, sensing and in electronic devices. In addition, the synthesis via some chemical methods particularly sol–gel route requires low processing temperature and less energy for carrying out the synthesis which makes this method more economical compared to physical methods [3]. The shape and size of the NPs directly affects their properties [4]. For instance, a study by Haruta et al. [5] suggested the catalytic activity of gold NPs depend on the NP-support contact; they found hemispherical NPs to perform better than the spherical NPs. In a similar study, Tian et al. [6] reported the enhanced catalytic activity of platinum NPs. Xu et al. [7] investigated the oxidation of styrene over cubic, truncated triangular nanoplates, and near-spherical Ag NPs and found the rate of reaction over the nanocube particles to be 14 times higher than over the nanoplates, and four times higher than over the nearly-spherical particles. These studies show the importance of the shape of the synthesized nanostructures and why the shape of the nanostructures during synthesis must be controlled. Thus, a method that can provide better control over shape and size of the NPs is the most suitable for wide range of applications. Different studies in the literature indicate that sol–gel method is one of the promising method to control the shape and size of the NPs [8,9,10].

Fig. 1
figure 1

Different physical and chemical methods used for the synthesis of MONPs

In this paper, we reviewed one of the most widely used chemical method i.e., sol–gel method for the synthesis of various MONPs such as TiO2, ZnO, SnO2 and WO3. The processing steps involved in the sol–gel process are discussed in detail. In addition, we have provided a brief discussion about the material characterization to study the crystal structure, optical properties, bandgap, elemental composition, shape, and size. The various applications of MONPs such as in photocatalysis, photovoltaics, sensors, hydrogen production and biomedical areas are discussed in detail.

2 Sol–gel method

Sol–gel method is one of the well-established synthetic approaches to prepare high-quality MONPs as well as mixed oxide composites. This method has excellent control over the texture and surface properties of the materials. In general, sol–gel method can be described in five key steps; hydrolysis, polycondensation, aging, drying and thermal decomposition [11] as described in Fig. 2.

Step 1:

Hydrolysis of the precursors such as metal alkoxides takes place in water or alcohols as shown in Fig. 2. For the synthesis of MONPs, oxygen is necessary for the formation of metal oxide which is supplied by water or organic solvents (e.g., alcohols). If water is used as reaction medium, it is known as aqueous sol–gel method; and use of organic solvent as reaction medium for sol–gel process is termed as nonaqueous sol–gel route. In addition to water and alcohol, an acid or a base also helps in the hydrolysis of the precursors. The general chemical reaction for the hydrolysis process is given below:

$$ {\text{M-OR}} + {\text{H}}_{ 2} {\text{O}} \to {\text{MOH}} + {\text{ROH}}\quad \left( {\text{Hydrolysis}} \right) $$

where M = metal, R = alkyl group (CnH2n+1)

The amount of water strongly influences the gel formation; a higher water content facilitates the formation of a higher ratio of bridging to nonbridging oxygens, thus yielding a more polymerized and more branched structure during the condensation [12].

Step 2:

This step involves condensation of adjacent molecules, where water/alcohol are eliminated and metal oxide linkages are formed, and polymeric networks grow to colloidal dimensions in the liquid state. Condensation occurs via two processes olation and oxolation. Olation is a process in which hydroxyl (–OH–) bridge is formed between two metal centers (metal–hydroxy-metal bonds) and oxolation is a process in which an oxo (–O–) bridge is formed between two metal centers (metal–oxo-metal bonds). The general chemical reaction for the condensation process is given below:

$$ {\text{M-OH }} + {\text{XO-M}} \to {\text{M-O-M }} + {\text{ XOH}}\quad \left( {\text{Condensation}} \right) $$

where M = metal, X = H or alkyl Group (CnH2n+1)

Condensation or polycondensation ultimately results in increasing the viscosity of solvent forming a porous structure maintaining liquid phase called gel. The size and the cross-linking within the colloidal particles mainly depend upon the alkoxide precursor and pH of the solution [13,14,15,16].

Step 3:

Aging process produces continuous changes in the structure and properties of the gel. During the aging process, polycondensation continues within the localized solution along with reprecipitation of the gel network, which ultimately decreases porosity and increases thickness between colloidal particles

Step 4:

The drying process is complicated because water and organic components are detached to form gel which disturbs its structure. There are different drying processes: atmospheric/thermal drying, supercritical drying, and freeze-drying, each one having different implications on the structure of the gel network as shown in Fig. 2. Heating the porous gel at high temperature causes densification [17] and the pores may be removed from gel to get xerogel, which has relatively a low surface area and pore volume and high shrinkage of the gel [18]. On the other hand, in supercritical drying aerogels are formed which have high pore volume and surface area and the original gel network remains almost intact. The third type of drying is done by freezing the solvents to form cryogel; the gel network shrinkage in cryogel is relatively lower than the xerogel. Another important parameter is the relative humidity (RH) during the drying that directly affects the stability and performance of nanomaterials. Particularly, nanofilms dried at lower RH are more stable than those which have been dried at higher RH [19]

Step 5:

Lastly, thermal treatment/calcination is performed to drive off the residues and water molecules from the desired sample, the calcination temperature is a very important parameter in controlling the pore size and the density of the material

In the next sections, we present a review on the sol–gel synthesis process for TiO2, ZnO, SnO2 and WO3 NPs and their characterizations.

Fig. 2
figure 2

Steps involved in sol–gel process to synthesize MONPs

2.1 Synthesis of TiO2 nanoparticles using sol–gel method

Sol–gel technique is one of the easiest and efficient methods to synthesize different TiO2 NPs [20, 21]. Several groups have reported the formation of TiO2 nanostructures via sol–gel method. Vijayalakshmi et al. [22] have prepared TiO2 NPs by mixing titanium IV Isopropoxide (TTIP) and ethanol using nitric acid to assist the hydrolysis process. The size of the NPs obtained was around 9 nm and their study showed that TiO2 NPs with sol–gel method were highly crystalline and had smaller crystallite size as compared to those prepared via hydrothermal method under similar reaction conditions. In another work, Jaroenworaluck et al. [23] prepared TiO2 NPs by mixing tetraisopropylorthotitanate (TIPT), methanol and ethanol in different molar ratios under stirring. The size of the synthesized anatase NPs was around 10 nm. Furthermore, it was observed that at lower calcination temperature anatase phase was formed and at higher temperatures (600–800 °C) anatase to rutile transformation occurs. In a similar study, Li et al. [20] produced TiO2 NPs to investigate the phase transformations of TiO2 NPs.

XRD pattern of multiphase TiO2 nanostructures is shown in Fig. 3i. The diffraction peaks of different TiO2 phases are well-matched with their corresponding standard data obtained from their respective JCPDS cards. The presence of sharp diffraction peaks at ~ 25.06°, 48.1°, 54.1°, 55.1° indicate the (101), (200), (105), (211) lattice planes, respectively. Hence, confirming pure and highly crystalline tetragonal anatase TiO2 phase with the lattice parameters a = b = 0.377 nm, and c = 0.948 nm (Fig. 3i-a). For brookite TiO2, the presence of main diffraction peaks at ~ 25.26°, 37.9°, 47.9°, 54.1°, 55.1° indicate the (210), (021), (321), (230), (421) lattice planes, respectively, confirming a pure and crystalline orthorhombic brookite TiO2 phase with lattice parameters a = 0.919 nm, b = 0.546 nm and c = 0.515 nm (Fig. 3i-b). Similarly, the presence of sharp diffraction peaks at the 27.5°, 36.2°, 39.3°, 41.4°, 44.0° indicate the (110), (101), (200), (111), (210) lattice planes, respectively, confirming a pure and highly crystalline tetragonal rutile TiO2 phase with the lattice parameters a = b = 0.458 nm, c = 0.295 nm (Fig. 3i-c).

Fig. 3
figure 3

i XRD pattern of multiphase TiO2: a anatase, b brookite, c rutile TiO2, ii SEM images of TiO2 nanostructures; a and b anatase nanostructures, c and d brookite nanostructures, e and f rutile nanostructures, iii Optical characteristics: a UV–Vis reflectance spectra, b and c optical bandgap evaluation of multiphase TiO2 NPs from Tauc plots, where the symbols A, B and R correspond to anatase, brookite and rutile-TiO2 [24,25,26,27,28,29,30,31,32]

Figure 3(ii-a, b) shows the SEM images of the TiO2 NPs prepared by Li et al. [26] depicting the structural evolution of anatase TiO2 NPs with the reaction time. After 1-h reaction time, smooth and solid spherical particles were observed and after 1-day reaction time it was observed that small surface platelets formed an urchin like prickly surface developing shell on the NPs. Figure 3(ii-c, d) shows the SEM images of brookite TiO2 prepared by Kandiel et al. [27] and Zhang et al. [28], depicting brookite rods with diameters up to 25 nm and lengths up to 150 nm. Figure 3(ii-e, f) shows the SEM images of rutile TiO2 prepared by Sarkar et al. [29] and Bai et al. [30], depicting microspheres of rutile TiO2 (diameter ~ 2 µm) consisting a cluster of nanorods with diameter ~ 20 nm. From the corresponding high magnified image (inset of Fig. 3ii-e), it can be seen that the nanorods are closely packed with square-shaped ends.

Optical characteristics and related bandgap properties studied by Coronado et al. [31] using UV–Vis spectroscopy are shown in Fig. 3(iii-a–c). The corresponding optical bandgaps of the TiO2 nanostructures are calculated from the UV–Vis reflectance spectra using the Tauc equation, \( \left( {\alpha h\nu } \right)^{n} = \, B\left( {h\nu-E_{g} } \right) \), where α is the absorption coefficient, is the energy of the incident photon, B is a constant (characteristic of the material used), Eg is the optical bandgap of the material used and n can have two different values, either 2 for direct or 0.5 for indirect bandgap transitions. The intercept on the x-axis of the Tauc plot \( \left( {\alpha h\nu } \right)^{n} \) versus gives the optical bandgap energy of the material (Fig. 3iii-b, c) The extrapolation of the reflectance versus wavelength curves determines the optical bandgap energy of TiO2 NPs giving the values of Eg to be 3.2 eV, 3.13 eV and 3.01 eV (taking n = 0.5) and 3.53 eV, 3.56 eV and 3.37 eV (taking n = 2) for anatase, brookite and rutile TiO2, respectively.

XPS spectra of TiO2 NPs synthesized by Singh et al. [33] showing elemental composition and binding energy of the elements present in the sample (Fig. 4). The curve is fitted using voigt function; curve for O 1s is fitted according to two peaks including Ti–O link at 529.9 eV and O–H link at 532 eV (due to absorbed surface hydroxyl group) [34, 35]. The double peaks of Ti 2p after deconvolution corresponds to Ti 2p1/2 at 464.4 eV and Ti3/2 at 458.7 eV.

Fig. 4
figure 4

XPS spectra of a O 1s and b Ti 2p of the TiO2 sample prepared by Singh et al. [33]

Apart from pure TiO2 NPs, several groups synthesized doped TiO2 NPs for enhancing their properties for particular application e.g., for degradation of various organic dyes such as rhodamine B, methylene blue (MB), methyl orange (MO) etc. present in waste water for water treatment [36, 37]. Kerkez et al. [38] prepared pure and doped samples of TiO2 using Titanium(IV) Butoxide, propanol, acetic acid and sulphuric acid as precursors for pure TiO2 and copper nitrate (Cu(NO3)2·3H2O), nickel nitrate (Ni(NO3)2·6H2O), cobalt nitrate (Co(NO3)2·6H2O), chromium nitrate (Cr(NO3)3·9H2O), manganese nitrate (Mn(NO3)2·5H2O) and iron(III) nitrate (Fe (NO3)3·9H2O) were used as Cu, Ni, Co, Cr, Mn and Fe sources, respectively to prepare the doped samples of TiO2. They studied the photocatalytic activity of the samples prepared and photocatalytic degradation rates of the MO and MB colorants in presence of visible light and found that the Cu doped TiO2 sample has highest photocatalytic activity under visible light for both of the dyes. Similarly, Yadav et al. [39] prepared Cu, Fe, Ni, Cr, and Co-doped TiO2 and undoped TiO2 nanostructures by sol–gel method and demonstrated photocatalytic dye degradation and antibacterial application.

Peerakiatkhajohn et al. [40] prepared silver doped TiO2 thin film on polyethylene terephthalate (PET) substrate using Titanium(IV) n-butoxide and silver nitrate (AgNO3) as precursors for the preparation of thin films. The Ag/TiO2 gel obtained from this was directly used for coating on the plastic substrate having potential applications in indoor air pollution treatment and photodegradation of gaseous pollutants such as benzene, toluene, ethylbenzene and xylene. Pan et al. [41] synthesized the surface modified NPs of TiO2 by octadecyltrimethoxysilane (WD-11) showing potential to be used as nanocomposites in the optical field and as lubricating oil additive. Although NPs already have high surface to volume ratio than their bulk counterparts, but still there is some room for increasing this ratio by having porous NPs. Few studies have showed the simple template-assisted sol–gel process for preparing porous TiO2 NPs one such study was done by Lakshmi et al. [42]. Their group for the first-time used sol–gel chemistry to prepare semiconductor nanofibrils and tubules within the pores of an alumina template membrane. Wang et al. [43] synthesized porous TiO2 NPs using tetraethyl orthosilicate (TEOS) as a template forming TiO2–SiO2 composite and then selective removal of SiO2 by dilute hydrofluoric acid (HF) solution. TiO2 prepared by this method exhibited remarkably improved performance due to its porous structure and large surface area for sensing wide range of RH. A summary of reported synthesis, chemical used, their characteristics and properties is presented in Table 1.

Table 1 A summary of reported synthesis of TiO2 nanoparticles via sol–gel method by different research groups

2.2 Synthesis of ZnO nanoparticles using sol–gel method

Solution-based approach (e.g., sol–gel) for synthesis of ZnO NPs is simple and less energy consuming. Due to the simplicity of the sol–gel method, many research groups have synthesized ZnO NPs using it. For instance, Mahato et al. [47] synthesized nanocrystalline ZnO using zinc acetate dehydrate, oxalic acid and ethanol as the precursors by following the method as reported by Hariharan [48] for using it as a decontaminant for the neurotoxic agent called sarin, Ristic et al. [49] prepared nanocrystalline ZnO powder by using Zinc 2-ethylhexanoate containing 1% of ethylene glycol monomethylether, tetramethylammonium((CH3)4NOH) aqueous solution and isopropanol as precursors. TEM results of the NPs showed that the particles produced by this study had size in the range of 20–50 nm.

Another template-assisted sol–gel method was also used for synthesis of ZnO nanofibres [42]. Yue et al. [50] reported a simple method for synthesis of ZnO nanotubes with porous anodic aluminum oxide (AAO) membranes from two-step anodization in oxalic acid solution. The resulted ZnO nanotubes had size of around 70 nm and thickness of ~ 12 nm. Mayekar et al. [51] studied the role of temperature and sodium hydroxide in the preparation of zinc oxide NPs, their study showed that the particle size increases with increase in the temperature and also with the increase in the sodium hydroxide concentration.

Figure 5i is showing the FESEM and SEM images for the 0 Dimensional, 1 Dimensional, 2 Dimensional and 3 Dimensional ZnO structures of different shapes and sizes. Figure 5(i-a) shows the spherical ZnO NPs having a size of around 80 nm prepared by Zavar [52] after calcination for 2 h at 350 °C. SEM image of 1D ZnO nanostructures having length and diameter in the range of about 3 μm and 200–500 nm, respectively are shown in Fig. 5(i-b) [53]. SEM micrograph of ZnO nanosheets (Fig. 5(i-c)) prepared by Ju et al. [54] shows the network of uniform and homogenous nanosheets with a large size and density. Highly uniform flower-like particles consisted of hexagonal nanorods prepared by Yue et al. [55] have been shown in Fig. 5(i-d). The typical size of these individual nanorods ranged from 300 to 800 nm in diameter and about 2–3 μm in length.

Fig. 5
figure 5

i FESEM and SEM images for a 0D, b 1D, c 2D, d 3D zinc oxides [52,53,54,55], ii XRD pattern of Synthetic ZnO NPs and the standard ZnO wurtzite structure from JCPDS card No. 36-1451 [56] and iii-a UV–Vis spectrum of ZnO NPs and b Tauc plot for determination of bandgap of ZnO NPs [57]

XRD pattern of the synthesized NPs prepared as reported in a particular study [56] has been shown in Fig. 5ii. The observed experimental peaks are fitted to the standard JCPDS card values corresponding to the ZnO reflections from (100), (002), (101), (102), (110) and (103) planes, confirming the ZnO hexagonal wurtzite structure. The typical UV–Vis absorption spectrum of ZnO NPs has been shown in Fig. 5(iii-a) where it can be seen that the NPs exhibit strong absorption in UV region (around 279 nm). The determination of energy bandgap (3.76 eV) is done by plotting Tauc curve shown in Fig. 5(iii-b). This value of bandgap is slightly higher than the actual value of bandgap for ZnO. This enhancement in bandgap occurs due to the quantum confinement of carrier within the nano dimensions [57].

Figure 6a shows the XPS spectra of ZnO NPs prepared via sol–gel method by Zhang et al. [58]. The survey scan is showing the presence of zinc, oxygen and carbon (impurity). Figure 6b, c is showing the high-resolution XPS spectra of Zn 2p and O 1s, respectively of ZnO nanorods samples prepared by Navaneethan et al. [59]. The Zn 2p1/2 and Zn 2p3/2 peaks were observed at 1046.8 eV and 1023.8 eV, indicating the presence of bivalent oxidation state in the sample [60]. The O 1s peak positioned at the binding energy of 530.8 eV was assigned to O2− ions in the Zn–O bonding of the hexagonal wurtzite structure of ZnO and the peak located at 523.3 eV was associated with O and O2− ions in oxygen deficient regions in the sample matrix [61].

Fig. 6
figure 6

a XPS spectra of ZnO NPs annealed in air at 800 °C prepared by sol–gel method [58] b high-resolution XPS spectra of Zn 2p1/2 and Zn 2p3/2 and c high-resolution XPS spectra of O 1s [59]

Further, ZnO nanostructures doped by Al, Ga, Sn, Ni, In, N and Li have been widely studied by some research groups for their structural, optoelectronics, luminescence, photocatalytic, impedance and thermal properties [62,63,64,65,66,67,68,69,70,71]. Similarly, doped ZnO nanostructures have been reported by Ullah et al. and Hossain et al. [72, 73] to study the photocatalytic degradation of organic dyes with the use of manganese-doped zinc oxide NPs which were prepared by sol–gel technique. ZnO doped and undoped colloids were made using zinc acetate, sodium hydroxide, ethanol and manganese acetate (dopant).

Also, ZnO thin films have been reported by some research groups. Natsume et al. [74] prepared ZnO thin films on Pyrex glass substrate by spin coating various layers of the solution of zinc acetate dihydrate and methanol followed by annealing in the temperature range of 500–575 °C to study the resistivity of the films. They found out a minimum film resistivity of 28.2 Ω cm was at an annealing temperature of 525 °C. Nagase et al. [75] reported a method to prepare ZnO films by spin coating on quartz substrate by using KrF excimer laser irradiation of sol–gel derived precursors prepared by zinc acetate and monoethanolamine dissolved in 2-methoxymethanol. Laser irradiation resulted in two types of effects on crystalline ZnO film (i) Irradiation at low energy fluence produces low crystallinity with weak orientation and (ii) irradiation at high energy fluence produces high crystallinity with strong orientation. A summary of reported synthesis, chemical used, their characteristics and properties is presented in Table 2.

Table 2 A summary of reported synthesis of ZnO nanoparticles via sol–gel method by different research groups

2.3 Synthesis of SnO2 nanoparticles using sol–gel method

Different research groups performed synthesis of SnO2 NPs using sol–gel method. Gu et al. [76] prepared nanoscale SnO2 by simple sol–gel method using tin chloride pentahydrate (SnCl4·5H2O) and ammonia solution (NH4OH). The reported diameter of the prepared NPs was about 2.6 nm. Adnan et al. [77] also prepared the SnO2 NPs using tin chloride pentahydrate (SnCl4.5H2O) and ammonia solution (NH4OH 25%). The high amount of ammonia leads to decrease in surface area and agglomeration of particles. Their study showed that SnO2 prepared at high alkaline condition (pH 10.2) exhibited higher catalytic activity in the hydrogenation reaction of styrene. Zhang et al. [78] used Nitric acid, granulated tin and aqueous ammonia (25%) as precursors and used citric acid as complexing agent. They showed that the use of citric acid slows down the hydrolysis and condensation of the precursor and thus inducing nucleation at early stage of the process resulting in monodispersed particles ranging from 2.8 to 5.1 nm in size. De Monredon et al. [79] prepared monodispersed crystalline NPs of SnO2 by hydrolysis of tin isopropoxide with acetylacetone and p-toluene sulfonic acid. Ristic et al. [80] studied the size dependence of nanocrystalline SnO2 on the synthesis route. They used Sn(IV) chloride (SnCl4), Sn(IV)-isopropoxide (Sn(OCH(CH3)2)4) and isopropanol as precursors. The average particle size reported in their study was around 2.05 nm found by the low-frequency Raman Scattering method. Aziz et al. [81] synthesized tetragonal phase SnO2 NPs using tin chloride (SnCl2·2H2O) and polyethylene glycol (PEG), ethanol and acetylacetone. Results showed that the SnO2 calcined at lower temperature with the incorporation of high molecular weight of PEG results in non-uniform tiny crystallites compared to the others. Gnanam et al. [82] prepared tin oxide NPs using different solvents such as methanol, ethanol, and water to study their effect on the optical properties of SnO2 using tin tetrachloride pentahydrate (SnCl4·5H2O) as a precursor to form sols. The SnO2 nanocrystallites prepared in methanol, ethanol, and water formed with a particle size of 3.9, 4.5 and 5 nm, respectively.

From the reported work by Gnanam et al. [82], TEM and HR-TEM images of the SnO2 NPs to study the morphologies and particle size have been shown in Fig. 7i. Figure 7(i-a) shows methanol mediated SnO2 NPs with spherical morphology of size about 3 nm. Figure 7(i-b, c) shows TEM pattern of sample prepared in ethanol and water and the size of the NPs was 4.5 nm and 5 nm, respectively. Figure 7(i-d–f) shows the high-resolution TEM (HR-TEM) image of the SnO2 NPs calcined at 400 °C, for 2 h. It shows clear lattice fringes, indicating the established crystallinity of SnO2 powders. The XRD pattern of the SnO2 NPs prepared with different solvents namely methanol, ethanol and water calcined at 400 °C for 2 h have been shown in Fig. 7ii, which corresponds to the tetragonal rutile structure of SnO2 having lattice parameters a = b = 0.4739 nm and c = 0.316 nm. The XRD pattern of the organic mediated sample (Fig. 7ii-a, b) shows slight broadening of peaks compared to that of the aqueous mediated sample (Fig. 7ii-c). The room temperature UV–Visible spectrum is shown in Fig. 7(iii). The methanol mediated SnO2 NPs has the absorption edge at 312 nm (Fig. 7iii-a), which is smaller than the band edges observed at 323 and 337 nm for ethanol and water as shown in Fig. 7(iii-b, c) and the corresponding bandgap energies calculated to be 3.97, 3.83 and 3.68 eV for methanol, ethanol and water-mediated SnO2 NPs, respectively which is larger than the bulk SnO2.

Fig. 7
figure 7

i TEM of the as-prepared SnO2 NPs mediated in a methanol, b ethanol, and c water. HR-TEM images of the as-prepared SnO2 NPs mediated in d methanol, e ethanol, and f water, ii XRD pattern of SnO2 NPs mediated in methanol (a), ethanol (b), and water (c), iii UV–Vis absorption spectra of SnO2 NPs mediated in methanol (a), ethanol (b), and water (c) [82]

The XPS spectra of SnO2 NPs prepared by Al-Hada et al. [83] are shown in Fig. 8a. The presence of Sn and O elements is confirmed by the XPS curve. The deconvoluted Sn peaks i.e., Sn 3d3/2 and Sn 3d5/2 are shown in Fig. 8b with binding energies of 494.9 eV and 487.8 eV, respectively which are also in good agreement with the research led by other groups [84,85,86]. The deconvoluted O 1s spectra (Fig. 8c) demonstrate binding energies at 529.7 eV and 531.1 eV which is correlated with SnO2 [87, 88]. A summary of reported synthesis, the chemical used, their characteristics and properties are presented in Table 3.

Fig. 8
figure 8

a XPS spectra of SnO2 NPs, b Sn 3d5/2 and Sn 3d3/2 XPS spectra, c O 1s XPS spectra [83]

Table 3 A summary of reported synthesis of SnO2 nanoparticles via sol–gel method by different research groups

2.4 Synthesis of WO3 nanoparticles using sol–gel method

Different structures of WO3 like NPs, nanofilms and nanocomposites have been prepared via sol–gel method by various groups [89,90,91]. Shi et al. [92] prepared WO3 NPs in the range of 25–100 nm for detection of acetone rendering its potential application in the diagnosis of diabetes. For the synthesis of NPs, a solution of peroxopolytungstic acid (PTA) was prepared from tungsten powder, hydrogen peroxide and a block copolymer EO20PO70EO20. The developed sensors possess a detection limit down to ~ 0.05 ppm.

Susanti et al. [93] prepared nano-WO3 based gas sensor via sol–gel method to detect carbon monoxide (CO) gas using a mixture of tungsten hexachloride and ethanol to prepare the tungsten alkoxide precursor W(OC2H5)6 and a surfactant (Triton X-100) to prepare the WO3 gels. Their result showed that the sensitivity of the material toward CO gas was dependent on the calcination temperature. Santato et al. [94] prepared mesoporous semiconducting WO3 films using colloidal solution from tungstic acid, sodium tungstate, stabilized by an organic additive such as polyethyleneglycol (PEG) 300. Their study showed that thin (less than 1 µm thick) nanocrystalline WO3 films have several interesting features such as excellent adherence and mechanical stability, open mesoporous structure and good transparency, making them suitable candidates for electrochromic device applications.

Figure 9i shows the XRD patterns of the different forms of WO3 matched with their corresponding JCPDS card: (a) hexagonal WO3 (b) monoclinic WO3 and (c) orthorhombic WO3 [95, 96]. UV–Vis spectrum of the WO3 nanocubes showing the absorption at about 473 nm and bandgap determined by Tauc plot with value of 2.58 eV is shown in Fig. 10ii [97]. TEM and SEM image of WO3 NPs prepared by Shi. et al. [92] having spherical shape with size ranging from 25 to 100 nm is shown in Fig. 10iii.

Fig. 9
figure 9

i XRD patterns of WO3: hexagonal tungsten trioxide (a), monoclinic WO3 (b), orthorhombic WO3 (c), ii UV–Vis spectra and inset is plot of \( \left( {\alpha h\nu } \right)^{n} \) versus photon energy (Tauc plot), iii Bright-field TEM image of WO3 nanocrystals (a), SEM micrograph of the WO3 nanocrystals [92, 95,96,97]

Fig. 10
figure 10

a XPS survey spectra of tungsten oxide samples, XPS of tungsten oxide hydrate and annealed samples b O 1s and c W 4f7 [90]

XPS spectra of nanostructured WO3 thin films prepared by Breedon et al. [90] have been discussed in Fig. 10a which reveal the presence of tungsten, oxygen, sodium and carbon. The Spectra of O 1s and W 4f7 is shown in Fig. 10b and c, respectively. The Na 1s peak at 1072 eV was found due to the presence of the by-products of the sodium tungstate precursor. The O 1s spectra of the prepared sample shown in Fig. 10b, showing a weak oxygen peak at 533.2 eV which is due to the oxygen atoms bound with or in WO3 and the main peak found at 530.6 eV. As the annealing temperature increased the smaller peak diminishes due to the expulsion of the water molecules and the main oxygen peak intensifies. The W 4f orbital is resolved in Fig. 10c into W 4f5/2 and W4f7/2 at 37.5 eV and 35.4 eV, no evidence of the formation of substoichiometric WO3−x was found as no peak mainly occurring at 34.8 eV as seen by other groups [98] can be seen here.

Nanocomposites prepared by Yao et al. [99] showed the formation of colloids to improve photochromism of WO3 by the combination of TiO2, most electrons will accumulate in WO3 resulting in the efficient suppression of the recombination process of electrons and holes after the combination. They used Sodium tungstate (Na2WO4·2H2O), oxalic acid (H2C2O4·2H2O) and titanium butoxide Ti(O-Bu)4 to form TiO2 and WO3 colloidal sols. Benoit et al. [100] reported a simple approach to decorate ordered TiO2 nanotube layers with WO3 to form WO3–TiO2 nanocomposite by using WCl6, ethanol and annealed TiO2 nanotubes as precursors. Their results showed that TiO2 decorated with WO3 led to remarkable enhancement in electrochromic properties. A summary of reported synthesis, chemical used, their characteristics and properties is presented in Table 4.

Table 4 A summary of reported synthesis of WO3 nanoparticles via sol–gel method by different research groups

3 Applications of metal oxide nanoparticles

MONPs can be useful in different areas ranging from cosmetics, color pigments, screen coatings, photovoltaics, photocatalysis, drug delivery, etc. (Fig. 11). Different applications of MONPs are discussed below:

Fig. 11
figure 11

Different applications of TiO2, ZnO, SnO2 and WO3 NPs

3.1 Photocatalysis applications

MONPs can produce photogenerated charge carriers upon excitation with required amount of light energy and having technological significance in environmental remediation and electronics. The abundant MONPs (TiO2, ZnO, SnO2 and WO3) have been used as photocatalyst due to their controllable optical and electronic properties. Most of the MONPs are investigated for use as antibacterial, self-cleaning, and deodorization system [102, 103].

For example, the wastewater can be purified by the removal of harmful bacteria and other organic pollutants with semiconducting photocatalysis process [104]. The mechanism of photocatalysis involving MONPs for photodegradation of organic dyes is described in Fig. 12 and the reactions taking place under this process can be explained by Eqs. (1)–(7) [105]

$$ {\text{MONPs }} + h\nu \to {\text{Me}}_{\text{CB}}^{-} + {\text{ Mh}}_{\text{VB}}^{ + } $$
(1)
$$ {\text{Mh}}_{\text{VB}}^{ + } + {\text{ H}}_{ 2} {\text{O}} \to {\text{H}}^{ + } + {\text{ OH}}^{ \cdot } $$
(2)
$$ {\text{Mh}}_{\text{VB}}^{ + } + {\text{ OH}}^{ - } \to {\text{OH}}^{ \cdot } $$
(3)
$$ {\text{Me}}_{\text{CB}}^{ - } + {\text{ O}}_{2} \to {\text{O}}_{ 2}^{ \cdot - } $$
(4)
$$ {\text{Organic pollutant}} + {\text{OH}}^{ \cdot } \to {\text{degradation products}} $$
(5)
$$ {\text{Organic pollutant }} + {\text{ Mh}}_{\text{VB}}^{ + } \to {\text{oxidation products}} $$
(6)
$$ {\text{Organic pollutant }} + {\text{ Me}}_{\text{CB}}^{ - } \to {\text{reduction products}} $$
(7)

where  = energy of photon, \( {\text{Me}}_{\text{CB}}^{ - } \) = conduction band electron in MONPs, \( {\text{Mh}}_{\text{VB}}^{ + } \) = valence band hole in MONPs.

Fig. 12
figure 12

Mechanism of photodegradation of organic dyes under illumination for different MONPs [102]

The photocatalytic antibacterial applications of TiO2 based nanostructures were summarized by Yadav et al. [106]. Similarly ZnO, SnO2 and WO3 NPs have also shown antibacterial and high photocatalytic activities as reported in different works [107,108,109].

3.2 Photovoltaics applications

Since several decades, photovoltaic technologies utilize semiconducting metal oxides. The outstanding flexibility of their optoelectronic properties and simple preparation method offers metal oxides an exclusive place to next-generation photovoltaics. Moreover, their outstanding capability to preserve or improve the device characteristics allows their application in semi-transparent and flexible solar cell devices, and printable electronics [110,111,112].

MONPs are used for different purposes in solar cells e.g., as an electron transport layer/hole blocking layer and mesoporous scaffolds in perovskite solar cells (PSCs) and dye sensitized solar cells (DSSCs) (Fig. 13b, c). Significant light scattering effect has been shown by MONPs when embedded in the organic materials [112, 117,118,119,120]. In PSCs, it is seen that the LUMO level of the perovskite light absorber layer is little higher than the conduction band edge of the MONPs (Fig. 13a), thus it is possible for perovskite material to transfer its electron to the MONPs layer for facilitating efficient charge transfer and collection.

Fig. 13
figure 13

Conduction band minimum and valence band maximum of TiO2, ZnO, SnO2 and WO3 along with the typical HOMO–LUMO levels of methylammonium lead iodide (MAPbI3) perovskite (a) [113,114,115], MONPs as electron transport layer and mesoporous scaffolds in perovskite solar cell (b) and in dye sensitized solar cell (c) [116]

TiO2 NPs are one of the most used materials for photovoltaic applications; it’s been mainly used in PSC and DSSC devices [121, 122]. For devices on flexible substrate, ZnO is preferred candidate owing to its higher electron mobility in comparison to TiO2 and low-temperature synthesis which requires no heating/sintering [123]. Mesoporous SnO2 thin films can also be used to replace mesoporous TiO2 scaffold layers in PSCs to achieve good power conversion efficiencies [124]. WO3 nanosheets coated with thin layer of TiO2 also resulted in good electron transport layer in PSC [125].

3.3 Gas sensing applications

MONPs gas sensors are the most investigated group of gas sensors owing to their large surface to volume ratio and high surface reactivity. One of the challenges for MONPs gas sensors is to achieve high selectivity [126]. Several MONPs have been used for sensing combustible, reducing, or oxidizing gases by observing electrical conductivity [127].

A schematic representation of gas sensing device is shown in Fig. 14. The interaction with the surrounding atmosphere changes the conductivity of metal oxide and because of its transducer function electrical signal is transmitted to the electrode. The indirect heating of the device is used to change its electrical conductivity. A promising research has been done for metal oxide-based gas sensors [128]. SnO2 and WOx have a long history for sensing applications, especially in gas sensing [129].

Fig. 14
figure 14

Schematic representation of gas sensing device and sensing principle [116]

3.4 Hydrogen fuel production

The conversion of solar energy into hydrogen represents an attractive and challenging alternative for photovoltaic solar cells. In 1972, a crucial discovery in the field of photocatalysis was done by Honda and Fujishima explaining the Honda–Fujishima effect which shows the electrolysis of water and the evolution of the hydrogen gas [103]. Since then, many nanomaterials have been seen which act as a photocatalyst to produce hydrogen fuel by inducing photocatalysis in water. Process for H2 evolution by a photocatalyst in presence of the sacrificial reagents is shown in Fig. 15, photogenerated holes irreversibly oxidize the sacrificial reagents instead of water making the photocatalyst electron-rich thus increasing the H2 evolution process.

Fig. 15
figure 15

Photocatalytic H2 evolution in the presence of a reducing agent [130]

Several metal oxide semiconductors were used to split water into hydrogen and oxygen upon illumination. The most straightforward method to produce hydrogen from water and sunlight is by coupling an electrolyzer to a solar cell array. TiO2 and SnO2 possess high chemical stability over a broad range of pH and applied potentials; however under light illumination the decay of ZnO was observed [131]. WO3 and other metal oxide semiconductors such as iron oxide (Fe2O3) have also been studied extensively for the use of photoanodes [132, 133].

3.5 Biomedical applications

Metal oxides owning special electrical, optical, and magnetic properties are promising for healthcare and medical applications like cosmetics, self-cleaning coatings, biosensors, bio-separation, antibacterial, and drug delivery [134]. MONPs are widely used in tissue engineering for creating biological alternatives of tissues and organs in mammals (Fig. 16). MONPs porous scaffolds can guide the cell growth and tissue regeneration in scaffold by the formation of reactive oxygen species such as O.−2 and H2O2 which can be helpful in cell function, adhesion, proliferation, angiogenesis and wound healing [135].

Fig. 16
figure 16

Scheme showing the role of MONPs in tissue engineering scaffolds [135]

In few studies, different MONPs have been studied like TiO2 nanocoral architecture that provides a highly controllable drug release system under UV light for cancer chemotherapy [136]. ZnO nanostructures show superior antibacterial, anticancer, biosensor, bioimaging, antimicrobial, and excellent UV-blocking properties [137, 138]. SnO2 NPs can be used for its antibacterial properties [83]. WO3–SnO2 sensor is used in non-invasive detection of human breath analysis [139]. Tungsten oxide NPs can be a good alternative material for electrochemical non-mediated biosensors [140]. Hybridized tungsten oxide nanostructures are also used for food quality assessment as a dye detector [141].

3.6 Energy storage applications

Supercapacitors are center of attention for their use as an energy storage device. The charge storage mechanism of a supercapacitor strongly depends on the surface properties of the electrode materials. These nanomaterials possess high surface area along with high surface energy, but the aggregation of nanoparticles is the most challenging problem. To overcome this problem mesoporous nanomaterials are widely used for electrode fabrication of supercapacitors. Mesoporous nanomaterials such as composites of metal oxides of TiO2 and V2O5 (vanadium pentaoxide) are employed as electrode material for the supercapacitor due to their excellent physical properties [142]. Various studies are reported for MONPs based electrodes in supercapacitors. For example, Sasirekha et al. [143] synthesized ZnO NPs via sol–gel method, coated with carbon (ZnO/C) to prepare an electrode for supercapacitor, exhibiting a specific capacitance of 92 F/g. Similarly, Bonu et al. [144] and He et al. [145] studied the role of SnO2 quantum dots and SnO2 NPs embedded in carbon microspheres which showed the specific capacitance as high as 420 F/g.

3.7 Artificial photosynthesis

Artificial photosynthesis is a chemical process that mimics the natural process of photosynthesis, a process that converts sunlight energy, CO2 and water to fuel. An artificial photosynthetic system that directly produces fuels from sunlight could provide an approach to scalable energy storage and a technology for the carbon–neutral production of high-energy–density transportation fuels [146]. Various doped and undoped photoelectrodes have been explored by researchers to create such system using metal oxides nanostructures such as TiO2, WO3, Fe2O3 and Cu2O [147,148,149,150]. Although these metal oxides make stable photoanodes, but their large bandgaps can only absorb very limited portion of the sunlight. However, there are some other inorganic semiconductors such as Si, CdTe, InP etc. which can also do the same task with an ultrathin protective layer of metal oxide which is nonporous, catalytic and have appropriate interfacial energetics. TiO2 and NiOx have shown ideal protection for the photoanodes as reported in some studies [151,152,153].

4 Future perspectives

The focus of research for creating novel materials is shifting nowadays towards size reduction due to which MONPs are paving their way in different industrial applications like electronics, drug delivery tools, sensors, solar energy etc. Currently, it is a global demand to change the approach for electrical energy production from burning fossil fuels to some form of a renewable energy source to fix serious environmental concerns like climate change and global warming. Utilizing the energy received from sun is by far one of the most explored form of renewable energy which ultimately gave rise to the field of photovoltaics. The synthesis of different nanostructures like nanorods, nanospheres, nanotubes, nanofibers, nanosheets etc. having different dimensionality and porosity has demonstrated their role in evolution of electronic devices especially in solution processed solar cells for achieving breakthrough efficiencies. A hunt for creating novel materials by a simple and cost-effective sol–gel processing can be helpful to execute powerful ideas to advance the nanoscale photovoltaic technology to the next level which can help in solving the energy crisis globally. For the production of different nanostructures with desired specific properties, the reaction parameters such as pH, reaction time, temperature, pressure during synthesis can be tuned to achieve novel nanostructures. Also, beside their useful applications it is important to keep an eye on minimizing the environmental hazards which can arise due to by-products of the reactions and by MONPs itself.

5 Conclusion

In conclusion, we have summarized the reported sol–gel synthesis route for different metal oxide nanoparticles (MONPs). The various processing steps involved in sol–gel synthesis are systematically elaborated. Sol–gel synthesis route has shown various advantages over other physical and chemical synthesis procedures due to ambient temperature processing, good control over texture of the produced NPs, faster process and economically less expensive. Sol–gel procedures to synthesize particularly TiO2, ZnO, SnO2 and WO3 MONPs are discussed along with a brief summary of the materials characterization to study the crystal structure, morphology, optical properties, bandgap, elemental composition, shape, size etc. Also, the role of MONPs involved in the different applications such as photocatalysis, photovoltaics, hydrogen production, gas sensors, biomedical, artificial photosynthesis, and energy storage is discussed. We think, scientific discussion and summarized data provided in this review will be helpful for the researchers around the world working on MONPs and their applications.