Introduction

ZnO-based materials have found great attraction in the last years because of their useful properties in several applications, such as solar cells, luminescence (Choi et al. 2008), catalysis, sunscreen protection (Zak et al. 2011), cosmetic products, UV detectors (Becheri et al. 2008), paints (Abyaneh et al. 2011) and due to their antibacterial activity (Vigneshwaran et al. 2006). The nanostructured ZnO particles are one of the most used inorganic materials due to the flexibility of preparation in different morphologies and properties (Wang 2004), with various applications. Recently, several synthesis routes have been developed for nanostructured ZnO materials, such as sol–gel method (Spanhel 2006; Golic et al. 2011; Zhang et al. 2007b), sol–gel combustion, template-assisted growth, hydrothermal (Hu and Chen 2008), precipitation (Aimable et al. 2010), gas-phase reaction, spray pyrolysis (Milosevic and Uskokovic 1993; Andrés-Vergés et al. 1989; Turner et al. 2010) and pyrosol method. The nanostructured ZnO materials could be obtained in various shapes such as particles, belts, wires, rods, depending on synthesis routes (Van den Rul et al. 2006; Krishnakumar et al. 2011; Clavel et al. 2007; Iwasaki et al. 1997; Zhang et al. 2007a).

The particle size distributions, the degree of agglomeration, as well as chemical stability are other important characteristics of the obtained ZnO materials.

Among the presented preparation methods, the pyrosol method could be one of the simplest, effective, low cost and fast ways to synthesize ZnO nanomaterials with controlled morphology and high purity. The principle of the method is based on the formation of an aerosol of the diluted solution of precursors, by using a high-frequency ultrasound generator and carrying through a furnace, in a quartz tube, by a carrier gas. The reactions that occur during the passage of the aerosol through the furnace are forming a complete calcinated powder with no necessary post-production treatment (Vasile et al. 2010; Dinu et al. 2008).

The synthesis methods are very influent on the properties of the material, such as morphology, shape, stability, etc. All this parameters, as well as very small dimension and corresponding high reactivity, are influencing the biological effects of the nanomaterials (Lin et al. 2009).

The literature is abundant in reports of ZnO photoluminescence (PL). The PL spectra of ZnO powders usually presents two emission peaks in the UV and visible ranges, their relative intensity strongly depending on the O2 pressure used during the process (Bae et al. 2006).

The UV emission corresponds to the near band-edge emission (NBE) and the visible emission is commonly referred to as a deep-level or trap-state emission (Huang et al. 2001).

The relative strength of NBE to deep level defect emissions exhibits a dramatic threshold dependence on surface roughness: surface optical emission efficiency increases over tenfold as roughness decreases to unit cell dimensions, highlighting the coupled role of surface morphology and near-surface defects for high efficiency ZnO emitters (Doutt et al. 2009).

High intensity of green and blue-green luminescence vs. NBE are reported in literature whenever the luminescence spectra is recorded on powders as obtained, without further treatments (simple annealing or annealing in the presence of reducing atmosphere) (Yuldashev et al. 2003).

It has been reported that the sub-band-gap emission in ZnO depends on the morphology of the nanostructures (Kwok et al. 2005; Li et al. 2010; Zhang et al. 2003). A blue-green emission at 453 nm and a green emission at around 557 nm were found in literature for nest like nanoparticles (Li et al. 2010) and emissions at 421, 484.6, and 532 nm for nanocones (Zhang et al. 2003).

In this study, ZnO nanoparticles have been synthesized and a detailed analysis of the particle size and morphology as a function of the pyrosol method parameters is considered.

The effect of the temperature and concentration of precursor solutions on PL and UV–Vis properties of the ZnO material have been also investigated. The quantum confinement of nanocrystalline ZnO particles it is presented (Prasad 2004; Efros and Efros 1982).

Experimental

Powders preparation

The pyrosol method was employed for preparing ZnO powders in order to achieve fine zinc oxide nanostructured materials and homogeneous dispersion at the nanometer level of powders. The synthesis of ZnO particles was carried out by using, as precursor, a diluted solution of zinc nitrate hexahydrate Zn(NO3)2∙6H2O (Sigma Aldrich) of purity 98 %, prepared under vigorous stir, with concentration of 0.1, 0.05 or 0.025 M.

The pyrosol method consists in the formation of an aerosol from the diluted solution of precursor, by using a high-frequency ultrasound generator, which is carried through the furnace, in a quartz tube, by a carrier gas. The reactions that occur during the passage of the aerosol through the furnace are evaporation, decomposition, and consolidation of the powder (Dinu et al. 2008). The experimental pyrosol production setup is presented in Fig. 1.

Fig. 1
figure 1

Experimental nanoparticles pyrosol production setup

Full size image

The preparation of ZnO was done by controlling parameters such as synthesis temperature, frequency of the piezoelectric ceramics and concentration of precursor solution as well as gas flow. The gas flow during all synthesis was maintained at ~6 L/min using a flow controller and the vibration frequency of piezoelectric ceramics at a value of 3.8 MHz. In this case, at the end of the quartz tube, the powder collection is done in deionised water, in one powder collection bottles. Afterwards, the powders were dried using an electric oven at 140 °C, for 24 h.

The temperature of synthesis used was 400, 500, 600 and 700 °C in order to establish the best temperature for which zinc oxide as single phase is obtained. The next study was done at the same synthesis temperatures and the concentrations of the starting solutions were modified from 0.1 to 0.025 M in order to decrease the particle size.

Powders characterization

The obtained powders were characterized using analysis such as X-ray diffraction (XRD), scanning electron microscopy (SEM), HR/TEM with selected area electron diffraction (SAED) and PL and UV–VIS spectrophotometry.

XRD analysis was carried out on Panalytical X’Pert Pro MPD equipment, with a Cu K α radiation, over a scan range of 2θ = 10° to 2θ = 90° angle. The SEM analysis was carried out by using a Quanta Inspect F microscope from FEI Company with field emission gun (FEG) and a 1.2 nm resolution, equipped with an energy-dispersive X-ray spectrometer (EDXS) with a resolution at Mn K α of 133 eV.

The bright field and high resolution images coupled with SAED were obtained using a Tecnai™ G2 F30 S-TWIN transmission electron microscope (FEI, the Nederland’s), equipped with a STEM/HAADF detector, EDAX spectrometer (for energy dispersive X-ray analysis) and Gatan EELS spectrometer (for electron energy loss spectroscopy). The microscope operates at an acceleration voltage of 300 kV (Shottky field emitter) with a transmission electron microscopy (TEM) point resolution of 2 Å and a TEM line resolution of 1.02 Å. The sample preparation for TEM measurements was done as follows: a small amount of powder was diluted into pure ethanol and left into an ultrasonic bath for approximately 15 min. After that, a small drop of the diluted solution was put onto a 400 mesh holey carbon coated film copper grid and left to dry for approximately 30 min, prior to be analysed.

Fluorescence measurements were made with a Perkin-Elmer LS 55 spectrofluorimeter using a Xe lamp as a UV light source, in the range 200–800 nm, with all the samples in solid state. The measurements were made with scan speed of 200 nm min−1, slit of 10 nm, and cut-off filter of 1 %. An excitation wavelength of 320 nm was used. UV–Vis spectra measurements were made with a JASCO V560 spectrophotometer with solid sample accessory, in the domain 200–800 nm, with a speed of 200 nm min−1. All measurements were made at ambient temperature.

Results and discussions

X-ray diffraction characterization

The qualitative phase composition of the powders was investigated by XRD, the XRD patterns being presented in Fig. 2.

Fig. 2
figure 2

X-ray diffraction pattern for ZnO powders obtained at different temperatures for different concentration of precursor solution A 600 °C, B 700 °C, and concentration of precursor solution of 0.1, 0.05 and 0.025 M C

Full size image

The XRD analyses indicate that, at the same concentration of precursor solution, the formation of wurtzite structure of zinc oxide powders started at 600 °C (Fig. 2A). At a synthesis temperature of 400 and 500 °C, the XRD diffraction patterns are showing peaks which indicate that the powder is composed from a mixture of zinc nitrates, zinc hydroxide and only a small amount of zinc oxide.

In order to analyse the influence of precursor solution concentration on the dimension of zinc oxide particles, at 600 and 700 °C, concentrations of precursor solution of 0.1, 0.05 and 0.025 M have been used. The corresponding XRD patterns are shown in Fig. 2B and C.

The XRD patterns for ZnO powders obtained at 600 and 700 °C and concentration of precursor solution of 0.1, 0.05 and 0.025 M reveal no significant variation of the intensity of the diffraction peaks due to the decrease of concentration of precursor solution and the increase of synthesis temperature.

All diffraction peaks for samples obtained at temperatures of 600 and 700 °C can be assigned to the wurtzite crystal structure of the zinc oxide powder (Fig. 2A–C), in accordance with International centre for diffraction data (ICDD) file number 04-004-4531 (Kihara and Donnay 1985).

In order to estimate the average crystallite size, the obtained peaks were analysed and the Scherrer’s equation has been used (Ada et al. 2008; Fernandes et al. 2009).

In all cases, at the concentrations and synthesis temperatures considered, through pyrosol method, nanocrystalline particles have been obtained.

XRD crystallite size variation against synthesis temperature and concentration of precursor solution is shown in Table 1. As it can be seen, zinc oxide powders with the average nanocrystalline dimensions in the range between 7.4 and 15.3 nm for synthesis temperature of 600 °C, and in the range 5.7–18.4 nm for synthesis temperature of 700 °C, have been obtained. The crystallites sizes of zinc oxide powders obtained at 600 and 700 °C and 0.1 M precursor solution concentration are smaller than those obtained at the same synthesis temperature and 0.5 and 0.025 M concentration of precursor solution.

Table 1 Crystallites size variation against synthesis temperature and concentration of precursor solution from XRD analysis
Full size table

SEM analysis

The SEM images of the ZnO sample obtained at synthesis temperature of 600 and 700 °C starting from 0.1 M concentration of precursor solution are presented in Fig. 3.

Fig. 3
figure 3

The SEM images for ZnO powders obtained at 600 °C a, b and 700 °C c, d synthesis temperatures at a concentration of precursor solution of 0.1 M

Full size image

The SEM images of ZnO sample obtained through pyrosol method at 600 and 700 °C and 0.1 M precursor solution concentration are showing spherical-like and nanostructured particles. The particles sizes are varying between 0.2 and 1 μm.

From the SEM image detail presented in Fig. 3b and d, we can clearly see that the samples are evidencing nanostructured morphology of the ZnO particles. The particles are homogeneous for both samples, with some broken particles. In both cases, the nanoparticles are forming a very well-defined structure. By comparing the powders obtained at 600 and 700 °C synthesis temperatures, it can be seen that at 700 °C the powder presents a much uniform morphology. The data are in accordance with XRD analysis obtained for these samples.

The 600 °C heat treated powder morphology is evidenced in secondary electrons image (Fig. 3), and still some broken particles could be observed, with different shapes and dimensions.

TEM–HRTEM analysis

The TEM, high resolution transmission electron microscopy (HRTEM) and SAED patterns of the ZnO sample obtained at synthesis temperatures of 600 and 700 °C starting from 0.1 M concentration of precursor solution are presented in Figs. 4, 5 and 6.

Fig. 4
figure 4

The TEM bright field images of ZnO powders obtained at 600 °C a, and 700 °C c synthesis temperatures at a concentration of precursor solution of 0.1 M. Insets b and d, represents the SAED for the obtained ZnO powders at 600 and 700 °C synthesis temperature

Full size image
Fig. 5
figure 5

The TEM bright field images of ZnO powders (detail for a single particle) obtained at 600 °C a and 700 °C b and concentration of starting precursor solution of 0.1 M

Full size image
Fig. 6
figure 6

The HRTEM images of ZnO powders obtained at 600 °C a and 700 °C b and concentration of starting precursor solution of 0.1 M

Full size image

The TEM bright field image obtained on ZnO particles heat treated at 600 and 700 °C, shown in Fig. 4a and c, reveals that the powder is composed from spherical shaped particles, with an average particle size of 0.5–2 μm for both samples. By comparing the bright field images, we can see that at 600 °C the morphology of the sample is slightly different than the one treated at 700 °C in what regards the shape. We can observe that at 600 °C the particles are not uniform in size distribution. The powder also presents the tendency to break down into smaller pieces with various shapes and sizes. The inset presented in both bright field images from Fig. 4b and d, represents the SAED taken from one single sphere. The only phase identified from SAED is the polycrystalline hexagonal form of ZnO for both samples.

By analysing the bright field images presented in Fig. 5a and b, we can state that the spherical-shaped particles are actually agglomerates, formed from polyhedral-shaped grains, with an average size of 14.75 nm for the 600 °C temperature of synthesis and 13.95 nm for 700 °C. The average grain size was calculated by measuring approximately 500 grains from various regions of the samples. The powders synthesized at 600 °C have a more pronounced tendency to form agglomerates; this is probably due to the fact that at this synthesis temperature the crystalline phases are not fully formed. The histogram of size distribution for 600 and 700 °C synthesis temperature for ZnO powders are presented in Fig. 8, where it can be seen that the grain size distribution of 600 °C heat treated powder is monomodal (Fig. 7a), while in the case of 700 °C treated powder which is bimodal (Fig. 7b).

Fig. 7
figure 7

Grain size distribution of ZnO powders obtained at 600 °C a and 700 °C b and concentration of precursor solution of 0.1 M

Full size image

The variation of nanocrystalites size with synthesis temperature and concentration of precursor solution from TEM measurements is presented in Table 2.

Table 2 Nanoparticles size variation against synthesis temperature and concentration of precursor solution from TEM analysis
Full size table

By analysing the nanoparticles sizes for the 0.025 and 0.05 M concentration of precursor solution and for both temperature of synthesis we can state that the nanoparticles size is almost the same size with the crystallites determined from XRD. In the case of 0.1 M concentration of precursor solution at 600 and 700 °C, the mean nanoparticles size measured from TEM data is not in concordance with XRD nanocrystallite sizes calculated from XRD data, being 2–3 times bigger in size (see Fig. 6a), which means that the particles are formed from 2 to 3 crystallites. We can see that the temperature of synthesis has no so obvious influence on grain size, but, by decreasing the concentration of precursor solution, the average grain sizes increases, for both synthesis temperatures. The HRTEM images obtained on nanocrystalline ZnO powders are shown in Fig. 6a and b. The image shows clear lattice fringes of polycrystalline nanopowder of d = 2.47 and 2.81 Å corresponding to the (101) and (100) crystallographic planes of hexagonal ZnO for the powder synthesized at 600 °C. The regular succession of atomic planes indicates that the powder is structurally uniform. The HRTEM image presented in Fig. 6b represents one single crystal of hexagonal ZnO. The image also shows clear lattice fringes of d = 2.47 Å corresponding to the (101) crystallographic planes of hexagonal ZnO.

Photoluminescence and UV–Vis properties

Figure 8a–c is presenting the PL spectra of ZnO powders obtained at 600 and 700 °C and all concentrations of precursor solution. The samples have been measured with an excitation of the 320 nm.

Fig. 8
figure 8

The photoluminescence spectra of powders obtained by pyrosol process at 600 and 700 °C

Full size image

The results show that the as-synthesized samples at temperatures of 600 and 700 °C, have a wurtzite structure, with a weak UV emission at about 395–400 nm, two blue-green emissions, one centred at 450–456 nm and a second at 483 nm, and a weak green emission at around 516 nm.

The weak UV emission at 395–400 nm is assigned to the overlap of the free exciton emission from the wide band gap (3.16 eV as calculated from electronic spectra) of ZnO with the violet emission from 407 nm assigned to the radiative defects related to the interface traps existing at the grain boundaries (Jin et al. 2000).

The three other emission bands at 450, 483, and 516 nm in the blue-green range are defect-related emissions. The luminescence bands at 450 and 483 nm are caused by the transition from the level of the ionized oxygen vacancies to the valence band (Prasad 2004). The green band is assigned to many different kinds of defects, e.g. oxygen vacancies (Vanheusden et al. 1996), donor acceptor recombination (Lin et al. 2001), or antisite oxygen (Reynolds et al. 2001). It was shown that the intensity of the green defect emission band from ZnO depends on the oxygen concentration during growth process (Tseng et al. 2003). However, the correlation of oxygen vacancies and the green band is still under discussion (Schneider et al. 2006; Shan et al. 2005).

We attributed the green emission peak from 516 nm to the transition from conduction band to the acceptor level corresponding to antisite oxygen (Ratheesh Kumar et al. 2007). Annealing and size studies suggest that the yellow-orange luminescence of ZnO is most likely caused by bulk defects associated with excess oxygen (Hsu et al. 2006). The absence of the yellow-orange luminescence in our samples and the presence of the blue-green and green luminescence can be a clear indicative that only defects associated with oxygen deficiency are presented in our ZnO samples (Reynolds et al. 2001; Shin et al. 2008).

The PL spectra for the ZnO nanoparticles obtained at different concentrations presented also an increase of intensity for all the emission peaks in the series 0.025, 0.05 and 0.1 M. The analysis of the PL spectra for the samples obtained from various concentrations at both temperatures reveals that although for C M = 0.025 M the emission intensity is higher for the sample obtained at 600 °C, as concentration of precursor solution increase the emission intensity of the sample obtained at 700 °C becomes higher. This can be correlated with the crystallite size as it was determined from the XRD spectra. For the samples obtained from 0.025 M solutions at 600 °C we have a smaller crystallite size than for those obtained at 700 °C. While at 0.05 M the crystallite size is about the same for the samples obtained at 600 and 700 °C, at 0.1 M the crystallite size for the 700 °C sample is smaller than those from 600 °C. The PL spectra intensity presents a similar variation pattern, the samples with smaller crystallite size having a higher PL. As the blue-green PL is assigned to surface defects of the particles, a smaller size can translate into a higher surface, which means also a higher concentration of the surface defects.

The UV visible spectra for ZnO nanoparticles synthesized by pyrosol method displayed exciton absorption peak at 363 nm which implies a low ZnO crystallite size. The optical transmission spectra of the ZnO nanostructured particles taken in the wavelength region 200–800 nm are shown in Fig. 9a and b. The maximum transmittance in the visible region was found to be 80 %.

Fig. 9
figure 9

The optical transmission spectra of the ZnO particles synthesized at 600 °C a and 700 °C b

Full size image

The electronic spectra data were used to evaluate absorption coefficients of the ZnO particles at different wavelengths. The optical band gap (E g) for samples was calculated from the Tauc plot, by following equation:

$$ (\alpha h\upsilon ) = A{(h\upsilon - {E_{\text{g}}})^n}, $$
(1)

where n is 0.5, 1.5, 2 and 3 for direct, indirect forbidden, direct forbidden and indirect transitions (Goswami 1996; Pankove 1971; Rani et al. 2006; Lethy et al. 2008), α is absorption coefficient, A is the band edge sharpness constant, h is Planck’s constant, and hυ is the photon energy.

By plotting \( {(\alpha h\upsilon )^2} \) the optical band-gap energy was determined by extrapolating the linear portion of the graph to hυ = 0. It is observed that for all the particles, the best straight line is obtained for n = 1/2, which is expected for direct allowed transition.

The transition for all six samples appears to occur in a single step, indicative of phase uniformity of the samples.

As shown in Fig. 10a, the optical band gap values for the products obtained at 600 °C are 3.18, 3.19 and 3.16 eV, respectively, when the ZnO powder is synthesized from solutions with C M = 0.025, 0.05 and 0.1 M. For the ZnO powder obtained at 700 °C (Fig. 10b), the band gap values are 3.16, 3.17 and 3.15 eV, respectively. In both cases, we obtained smallest band gap for the highest concentration, C M = 0.1 M. The calculated band gap values are slightly smaller than that of bulk ZnO (3.37 eV). This difference may originate from the introduction of defects during the decomposition reaction. The energy band gap of a semiconductor is affected by the residual strain, defects, charged impurities, disorder at the grain boundaries and also particle size confinement (Li et al. 2010; Zhang et al. 2003). Also, crystallite with smaller size has higher surface area and can cause more band bending effect (Vinodkumar et al. 2010).

Fig. 10
figure 10

The optical band gap of the ZnO nanoparticles synthesized at 600 °C a and 700 °C b

Full size image

The lowest band gap shown by samples obtained from highest concentration and highest temperature (700 °C and 0.1 M concentration of precursor solution) can be due to the enhanced band bending effect at the particle boundaries (Srikant and Clark 1997; Dutta et al. 2007) and because that sample also presents the smallest crystallite size as seen from XRD spectra.

The absorption coefficient near the band edge shows an exponential dependence on photon energy (Urbach 1953).

$$ \alpha = {\alpha_0}\exp [(h\upsilon - {E_{\text{i}}})/{E_{\text{u}}}], $$
(2)

where E u is the Urbach energy which corresponds to the width of the band tail, E i and α0 are constant.

Figure 11 shows the variation of ln(α) vs. hυ for the ZnO nanoparticles. Urbach energy was calculated from the reciprocal gradient of the linear portion of these curves, the calculated values being in the interval 141–168 meV as can be seen in Table 3.

Fig. 11
figure 11

Variation of lnα vs. hυ for the ZnO particles obtained at 600 and 700 °C

Full size image
Table 3 The optical band gap and Urbach energy of the ZnO particles synthesized at 600 and 700 °C
Full size table

Such a change in the band tail feature points out the role of the solution concentration as the defect generator within the ZnO lattice.

The most evident differences are done first by concentration of precursor solution and afterwards by temperature of synthesis. The effect of the temperature is decreasing with the raising of the concentration used, so that samples of 0.1 M concentration present only minor influence of temperature.

Conclusions

We report the synthesis of zinc oxide nanostructured particles through pyrosol method starting from diluted solution of Zn(NO3)2∙6H2O at final concentrations of 0.1, 0.05 and 0.025 M. The temperature of synthesis used was 400, 500, 600 and 700 °C in order to establish the best temperature for which zinc oxide single phase is obtained, as well as the dimensions and properties of the obtained particles. The ZnO powders were characterized using XRD, SEM, HR/TEM, PL and UV–Vis spectrophotometry.

The grain size increase with respect of decrease of concentration of solution is strongly correlated with the synthesis process presented in Fig. 2. The UV visible spectra for ZnO nanoparticles synthesized by pyrosol method displayed exciton absorption peak at 363 nm which implies a low ZnO crystallite size of 13.95 nm, as determined from TEM data.

In this study, we confirmed that it is possible to prepare ZnO nanoparticles without the aid of special organic additives, or face-specific organic molecules, starting from a simple precursor, such as a diluted solution of zinc nitrate, by using pyrosol method.