1 Introduction

Metal oxide semiconducting nanostructures doped with trivalent lanthanides ions (Ln3+) are one of the most promising materials having most different applications in the field of light-emitting displays, magnetic recording media, photocatalysis and optical storage [1,2,3,4]. The optical emissions arising from the surface defect states of these semiconductors are effectively controlled by incorporating rare earth (RE) ions [1]. RE ions are considered to be the better luminescent centers because of their intra-shell transitions which can produce narrow and intense emission lines [1, 5]. The RE ions are incorporated into metal oxide semiconductors for the modification of electronic structure to improve optical, magnetic and catalytic properties [6,7,8,9].

One of the difficulties related with the synthesis of semiconductor nanoparticles is the control of growth by variation reaction parameters [10,11,12,13], such as concentration, temperature, pH, surfactant agent and synthesis time. These parameters directly influence the size and the morphology of nanoparticles. Furthermore, changing a synthetic strategy route can also introduce promising results. Microwave-hydrothermal (MH) method has attracted great attention as a heating process for semiconductor materials due to their kinetic advantages in comparison with the conventional hydrothermal method. As a modification of conventional hydrothermal, MH method is rapid volumetric heating without the heat conduction process, which can achieve uniform heating in a short period of time, and it can clearly decrease the reaction temperature of synthesis, obtain highly crystalline products with high purity and low aggregation morphologies, can also screen a wide range of experimental conditions in order to optimize the material properties [14,15,16,17,18,19,20]. Due to these advantages the microwave-hydrothermal method becomes a promising choice in the synthesis of nanostructured materials.

Among many oxides, zinc oxide (ZnO) is an important functional semiconducting material with a direct band gap of 3.37 eV, physical and chemical stability, weak cytotoxicity, high photosensitivity and piezoelectric properties [21, 22]. This has made ZnO a good candidate for technological application such as energy conversion, photocatalysis, sensors, LED’s and antibacterial properties [23,24,25,26,27,28]. Furthermore, ZnO has attracted attention for environmental remediation such as industrial wastewater treatment [29,30,31]. Under UV–Vis light irradiation, pollutant could be decomposed into non-toxic substances on the surface of a ZnO based photocatalyst [32,33,34]. However, the fast recombination of photogenerated electron–hole pairs is the main obstacle for increasing the photocatalytic efficiency of ZnO [35,36,37]. To enrich the photocatalytic, optical and magnetic properties of pure ZnO, many efforts have been made, as morphologies tuning, homogenous particles obtention and modification by doping with non-metals, alkali metals, transition metals (TM) as well as rare-earth (RE) metals [38,39,40,41,42]. Rare-earth metals doped ZnO nanostructures it’s a good alternative, which produced impurity energy levels in band gap and traps for photogenerated charges carriers accelerating the interfacial charge transfer and decreasing the recombination of electron–hole pairs [43, 44]. Gadolinium with half-filled f orbital has a mainly positive effect on the optical, electronic and magnetic properties when is used as a dopant for ZnO [42, 45]. Some researches for the beneficial effects of Gd on photocatalytic and optical properties of ZnO nanostructures but also, other semiconductors have been reported [46,47,48,49].

In this work, ZnO doped with 1.0 and 2.0 mol% Gd3+ were prepared by microwave-hydrothermal method. Crystalline and morphological homogeneous structures, optical and photocatalytic activity of synthesized nanoparticles were studied using different techniques.

2 Materials and methods

2.1 Preparation of Gd3+-doped ZnO nanoplates

1.22 × 10−2 mol of Zn(CH3COO)2∙2 H2O was added to 40 mL of distilled water. This solution was placed under constant stirring until complete dissolution of the salt. To this solution was added 1.0 mL of polyethylene glycol 400 (PEG 400). The pH value was raised to approximately 12 with the addition of a 2 mol L−1 KOH solution. Gd3+-doped samples were obtained with the stoichiometric addition of the Gd(NO3)3 0.28 mol L−1 (1.0 and 2.0 mol%) solution under constant stirring. The solutions were transferred to a polytetrafluoroethylene cup which was inserted into the reactor. The reaction system was heated to 90 °C for 16 min, with a heating rate of 5 °C min−1. The pressure inside the autoclave was stabilized at 1.5 atm. The final products were centrifuged, washed repeatedly with distilled water and ethanol, and finally dried in an oven at 80 °C.

2.2 Characterization of the as-prepared samples

The X-ray diffraction (XRD) analyses were used to obtain information about the crystalline structure of the ZnO samples. The measurements were performed on Shimadzu X-ray diffraction (XRD) 6000 (Japan) equipped with CuK radiation (λ = 1.5406 Å) in the 2θ range from 10° to 100° with 0.02°/min scan increment and steps with a fixed-time of 2 s. The refinement of XRD diffraction results was performed using the Rietveld’s profile analysis method [50, 51], with the General Structure Analysis System (GSAS) program suite [52], with EXPGUI interface [53]. Raman spectra were performed at room temperature with an Ocean Optics portable spectrometer equipped with λ = 785 nm laser, operating at 499 mW. The morphological characterization was performed using a scanning electron microscope model EVO MA 10 of Zeiss. X-ray (EDX) spectroscopy was performed on an Oxford Instruments, operating at 200 kV. UV–visible spectra of ZnO and Gd3+-doped ZnO samples were obtained on a Cary5G spectrophotometer in the 200–800 nm regions. Photoluminescence (PL) spectra were recorded at room temperature by a thermal Jarrel-Ash Monospec27 monochromator and a Hamamatsu R446 photomultiplier (λexc = 350.7 nm). The adsorption of N2 and the desorption isotherms at 77.35 K were measured on a Quantachrome TouchWin automatic surface analyzer version 1.1. Before analysis, the samples were degassed under vacuum at 90° C for 3 h. The surface area was calculated according to the BET method (Brunauer, Emmett and Teller) [54].

2.3 Photocatalytic activity

The photocatalytic performance of the as-prepared undoped ZnO and Gd3+-doped samples was evaluated by the degradation of methylene blue (MB) under UV–Vis light irradiation. A xenon lamp (300 W) was employed light source. The light beam was passed through a KG1 filter to remove UVB and UVC photons, thus the photoreactor was illuminated with wavelengths between 320 and 800 nm. In a typical procedure, 0.16 mg catalyst was dispersed into 16 mL aqueous solution of MB (0.02 mmol L−1) in a double wallet borosilicate reactor. After stirring in the dark for 60 min, the suspensions were placed under UV–Vis light irradiation. The reactor was kept at 25 °C by a water bath, the samples were collect at regular intervals, centrifuged to remove the excess of powders and then subjected to spectrophotometric analysis in the UV–Vis region (SHIMADZU 1650PC with 0.1 cm optical path quartz cuvette). The discoloration curves were plotted from the absorbance values at the maximum wavelength of 663 nm for aqueous solutions of methylene blue at pH close to neutrality. The percentage of discoloration was calculated by following the formula (Eq. 1):

$$ \% \;{\text{Decolorization}} = \left[ {1 - \left( {\frac{{A_{f} }}{{A_{i} }}} \right)} \right] \times 100 $$
(1)

where Af is the absorbance of the solution at time t > 0, and Ai is the initial absorbance of the solution. Photonic efficiency (ξ), which is defined as the ratio of the degradation rate of methylene blue (MB), r, and I the incident photon flux was also calculated as described elsewhere by using (Eq. 2) [55, 56]. The UV–Vis incident photon flux was determined by ferrioxalate actinometry [57] (9.20 × 10−8 Einstein s−1).

$$ {\text{Photonic}}\;{\text{efficiency}}\;(\xi ) = \frac{r \times 100}{I} $$
(2)

3 Results and discussion

XRD patterns refined by the Rietveld method are presented in Fig. 1a–c. Undoped ZnO and Gd3+-doped samples present a single phase of wurtzite structure (JCPDS 36-1451), without the formation of secondary phases. Table 1 shows the Rietveld refinement indexes, the lattice parameters and volume, obtained by Rietveld refinement for undoped ZnO and Gd3+-doped samples. The results were in good agreement with those observed and calculated XRD patterns. These results suggest that little fraction of Gd3+ are inserted into lattice or located as clusters on the surface of ZnO particles. It is known that the ionic radius of the Gd3+ and Zn2+ is 0.94 Å and 0.74 Å, respectively [45]. The wurtzite presents an open structure with interstices that can accommodate the Gd3+, suggesting that rare-earth 3 + ions are replacing or displacing Zn2+ ions in the crystalline structure. Although differences of ionic radius, the replacing by Gd3+ ions do not cause changes in the crystalline wurtzite structure at long-range. In addition, the diffraction peak (101) of wurtzite phase get shifted towards the lower angle (Fig. 1d) which leads to an expansion of the unit cell and is due to the large ionic radius of rare earth Gd3+ than Zn2+ ions into ZnO sites [6, 45, 58]. This behavior is attributed to the substitution of Zn2+ by Gd3+ ions and therefore confirms the doping procedure. It means that ZnO lattice is expanded along the c-axis for Gd3+-doped samples due to the mismatch between Gd3+ and ZnO of lattice in which ionic radii of Gd3+, Zn2+ and O2– is 0.94 Å, 0.74 Å and 1.32 Å, respectively [4, 59]. A slight expansion along a-axis was also observed. The increase for dopant promotes an increase in the number of defects in the crystalline lattice of the ZnO, thus leading to an increase in the lattice parameters.

Fig. 1
figure 1

Rietveld refinement plots for ZnO (a), 1.0 mol% of Gd3+ (b), 2.0 mol% of Gd3+ (c) and expanded region of the diffraction peak (101) (d)

Full size image
Table 1 Parameters obtained from Rietveld refinement for ZnO, 1.0 mol% Gd3+ and 2.0 mol% of Gd3+ samples
Full size table

The cell parameters (Table 1) indicate that doping with rare earth ions generates distortions in the crystalline lattice of ZnO and increases the amount of structural defects, such as oxygen and interstitial oxygen vacancies [43, 58, 60], since a slight increase was observed in the values of the cell parameters and consequently in an expansion of the unit cell when compared with the pure sample. This effect is consistent because the ions Gd3+ coordinate at octahedral positions preferentially in the crystalline lattice, occupying the interstitial sites [6].

Typical SEM images and EDX spectra of undoped ZnO and Gd3+-doped ZnO nanostructures prepared by microwave-hydrothermal method are shown in Fig. 2. The SEM images of ZnO Gd3+-doped ZnO samples reflect that have nearly similar particles morphology with varying agglomeration degree. Plate like morphology was observed for all samples. In addition, the SEM image of undoped ZnO sample shows irregular, slightly agglomerated plates with an average thickness of 45 nm, Fig. 2a. The particles display slightly larger particles size than Gd3+-doped samples. Both Gd3+-doped samples (Fig. 2b, c) showed formation of agglomerated plates with a smaller thickness (35 nm). EDX analysis confirmed the presence of Gd3+ in the samples (Fig. 2d). The silicon peak (Si) observed in the spectra is due of to substrate signal. No other peak related to impurities was detected in the spectra which further confirm that Gd3+-doped ZnO samples were successfully synthesized.

Fig. 2
figure 2

SEM images of samples obtained by microwave hydrothermal method of ZnO (a), 1.0 mol% of Gd3+ (b) and 2.0 mol% of Gd3+ (c). The EDX spectra are on the right side of the micrographs

Full size image

According to the literature, under conditions of thermodynamic equilibrium, the faster growth rate of ZnO is along the direction [0001] because of its higher surface energy and under certain physical chemical conditions, growth along this direction can be partially suppressed [61]. The formation of ZnO nanoplates or nanosheets may be associated with excess OH− in the reaction medium allied to a prolongation of synthesis time (16 min). These external factors possibly retard the growth in the direction [0001], favoring the direction \( [01\bar{1}0] \) in the formation and growth of the nanoplates [23].

The Raman spectra of ZnO and Gd3+-doped ZnO powders are shown in Fig. 3. Wurtzite-type ZnO hexagonal symmetry belongs to the space group P63mc, with two formula units per primitive cell. Zinc oxide presents eight optical Raman mode sets (phonon modes) at the point of the Brillouin zone in the ZnO monocrystal, being described as follows by the representation: Г = A1 + A2 + B1 + B2 + 2 E1 + 2 E2, where modes B1 are silent, A1 and E1 are polar modes, active in both the Raman and the infrared, while E2 modes are nonpolar active only in the Raman [62, 63].

Fig. 3
figure 3

The Raman spectra of ZnO (a), 1.0 mol% of Gd3+ and 2.0 mol% of Gd3+ (b) samples obtained by microwave hydrothermal method

Full size image

The most intense mode in the Raman spectra is the E2H at 438 cm−1, characteristic of the wurtzite phase of ZnO, attributed to the zinc-oxygen vibration in the crystalline lattice [63, 64]. The slight asymmetries observed between samples are attributed to short-range structure disorder as well as to non-harmonic phonon–phonon interactions. As the concentration of the Gd3+ ions increases, a decrease in the intensity of the E2H mode is observed, due to the distortions generated and consequently, a short-range structural disorder (Fig. 3b). The band located in the 210 cm−1 region refers to the 2E2L mode of the second-order phonons. The low frequency bands around 334 cm−1 can be attributed to the second order Raman spectrum resulting from the E2H–E2L phonons. The polar phonons A1 and E1, as opposed to the E2 phonons, are both divided into phonons TO and LO. The vibrational mode E1 (LO) located in the region around 580 cm−1 rises from background, which originate from second-order Raman scattering, is associated with structural defects formed by oxygen vacancies [64, 65]. The low intensity band attributed to the order–disorder degree of ZnO A1 (TO) around 378 cm−1 was observed to ZnO pure. The band at 409 cm−1 for the E1 (TO) mode is observed as a “shoulder” of the band at 438 cm−1 (E2H) and with the insertion of the dopant into ZnO crystal lattice it is covered by the E2H mode (Fig. 3b). A narrow band referring to the intrinsic TA + LO mode of ZnO was observed around of 660 cm−1 [65, 66].

In the literature, a similar behavior was observed, confirming that the influence of RE ion on the crystal structure arises from the strong dependence on the size and polarizability of the dopant ion [67]. The incorporation of Gd3+ ions promotes a tendency to electronic cloud distortion due to the increased polarization intensity. The Gd3+ presents a higher ionic radius and has the capacity to promote a distortion of the electronic cloud.

To investigate the specific area and the porosity of the ZnO and Gd3+-doped ZnO nanoplates, Brunauer–Emmett–Teller (BET) adsorption and desorption of N2(g) measurements were performed. The nitrogen adsorption and desorption isotherms are shown in Fig. 4. All samples presented the type IV curve accompanied by a H3 type hysteresis, which is attributed to the predominance of mesopores according to the IUPAC classification [68, 69]. In addition, the BET surface area and pore distribution of all the samples are summarized in Table 2. The BET tests show that surface areas are 11.89 m2 g−1, 13.94 m2 g−1 and 7.84 m2 g−1 for undoped ZnO, 1.0 mol% of Gd3+ and 2.0 mol% of Gd3+, respectively. 1.0 mol% of Gd3+-doped ZnO with a relatively higher specific surface possibly provides more active sites and is more beneficial for the photocatalytic reaction. Small particle sizes and minimized agglomeration increase the specific surface area and therefore also increase the number of active surface sites where the photogenerated charge carriers can react efficiently with absorbed molecules to form hydroxyl radicalar species [70, 71].

Fig. 4
figure 4

N2 adsorption/desorption isotherms of ZnO (a), 1.0 mol% of Gd3+ (b) and 2.0 mol% of Gd3+ (c) samples obtained by microwave hydrothermal method

Full size image
Table 2 BET surface area of plate-like ZnO and Gd3+-doped samples with different Gd3+ concentrations
Full size table

Variation in the specific surface area has been observed in the literature from insertion of dopants in the ZnO lattice. Liang et al. [32] synthesized Ce-doped ZnO microflowers at different concentrations (0.25–5.0% of Ce) and observed a decrease in the specific when the Ce4+ concentration was increased to 3.0%. Li et al. [72] obtained Er-doped ZnO samples via sol–gel method at concentration of 0.2%, 0.5% and 1.0% of Er. The authors reported that Er3+ concentrations were increased of 0 to 0.5%, the specific surface area increased, while the specific surface area decreased when the concentration increased to 1.0%.

The UV–visible diffuse reflectance spectra of the samples of pure zinc oxide and doped with the Gd3+ ions synthesized at 16 min times, are shown in Fig. 5. The Gd3+ ion with 4f7 configuration is very stable, f–f electronic transitions are difficult to detect, commonly is observed in the vacuum ultraviolet region (10–200 nm) [73]. To determine the gap energy (Egap) values of the synthesized materials, the Kubelka–Munk method [74] was used. The Egap values obtained were 3.28 eV; 3.25 eV and 3.26 eV, for samples of undoped ZnO, doped with 1 mol% Gd3+ and doped 2 mol% Gd3+, respectively. The insertion of a dopant in the structure of the zinc oxide did not promote significant changes in the Egap values of the materials, which is possibly due to the low concentrations of the doped ions Gd3+ inserted in the crystalline lattice of ZnO. But also, the estimated Egap experimental values were close to the theoretical Egap value of ZnO (3.37 eV) [75, 76].

Fig. 5
figure 5

UV–Vis diffuse reflectance spectra of nanoplates ZnO and doped with 1.0 and 2.0 mol% of Gd3+

Full size image

Photoluminescence spectra obtained at room temperature for samples of pure ZnO and doped with the Gd3+ ions synthesized at 16 min are shown in Fig. 6a. All samples presented a maximum broad band around 613 nm, which corresponds to the green and orange emission contributions. The photoluminescence intensity of 2.0 mol% of Gd3+ sample increased, suggesting that the recombination of the electrons and holes increases. This increase can be attributed to the surface defects of vacancies created by an increase in dopant concentration in the crystalline lattice, which act as centers of radiative recombination [32, 77]. The similar behavior was observed by Sowik et al. [78] that reported the modification of ZnO quantum dots with rare earth metals. The modification of ZnO with small amount of rare earth metal influenced lower PL intensity, however modification of ZnO by 0.18 or more mmol of La and Er caused increased in PL emission intensity.

Fig. 6
figure 6

a Photoluminescent emission spectra (λexc = 350.7 nm) for ZnO and Gd3+-doped ZnO, b decomposition emission bands for the samples of undoped ZnO, c doped with 1.0 mol% of Gd3+ and d doped with 2.0 mol% of Gd3+

Full size image

It is widely accepted that the deep-level-emission is closely related to the structural defects such as oxygen vacancies or Zn interstitials in zinc oxide [4]. By using the Gaussian analysis, Fig. 6b–d shows the decomposition curves for the samples of undoped ZnO and doped with Gd3+ ion. From the results it was possible to attribute how much each color contributed to the photoluminescent emission. The bands were decomposed into five components, two components in the region of green (515 and 562 nm), one component in the region of orange (613 nm) and two components in the red region (665 and 736 nm). Table 3 shows the maximum peak values for each emission component and its percentage of area.

Table 3 Relative percentage of the Gaussian deconvolution of the photoluminescence bands for the samples
Full size table

Doping can generate defects, which result in different recombination processes influencing emission in the visible region. As the dopant concentration increased in the ZnO crystal lattice, a higher percentage of green emission-related area can be observed by the contributions of the two peaks P1 and P2. The greater percentage of area in the green region correlate with the increase in the amount of dopant in the structure is attributed to the increase in the density of structural defects created, such as interstitial oxygen and zinc vacancies, as well as electronic defects [79,80,81,82]. Red emission bands can be attributed to free electron recombination and may be associated with excess oxygen occupying interstitial sites Oi [23, 83, 84]. The slight increase in emission in the orange-red regions from doping can be directly associated with the transitions of impurity levels induced by the presence of rare earth ion and also for changes in levels of impurities associated with native defects in ZnO.

The photocatalytic activity of the samples was evaluated against the degradation of the MB dye under UV–Vis irradiation. The degradation curve of MB without any catalyst added is also presented. As shown in Fig. 7a, there is a small decrease in the absorbance of MB in the blank experiment after 180 min of irradiation, indicating that the photolysis of MB solution resulted in degradation of 12%.

Fig. 7
figure 7

a Photolysis of the methylene blue in the absence of the catalyst at a total time of 180 min and b photodegradation curves of MB by different catalysts under UV–Vis irradiation

Full size image

The 1.0 mol% of Gd3+-doped ZnO nanoplates catalyst exhibit higher photocatalytic activity than those observed for the other samples (Fig. 7b), which corroborates with the PL behavior. The photocatalytic activity is affected by photo-generated charge transfer and separation, so the PL emission spectra can be used to estimate the separation capability of the electron–hole pairs. The PL emission is the result of the radiative recombination of electrons and holes. So, lower PL intensity reflects lower recombination yields of electron–hole pairs [7, 48]. Therefore, photocatalyst with weaker PL intensity tends to show higher catalytic ability.

The 1.0 mol% Gd3+-doped ZnO shows lower emission intensity and consequently, its photocatalytic efficiency is higher when compared to the undoped ZnO. The introduction of Gd3+ ions can generates an increase in the number of structural defects in the crystalline ZnO lattice, as oxygen vacancies. The oxygen vacancies act as electron acceptors during photocatalytic process, which trap the photogenerated electrons temporally to reduce the recombination of electrons and holes, providing an alternative path [43, 85]. If the amount of rare earth ions exceeds the dopant acts as recombination center for electrons and holes, so the photocatalytic activity decreases, also observed in literature [48, 86].

Yi et al. [4] reported that the degradation rate of MB on Gd3+-doped ZnO films with various Gd3+ contents under UV illumination. With increasing Gd3+ content, the photocatalytic activity of the Gd3+-doped ZnO films decreases. This similar tendency is also observed for Divya et al. [87]. They have prepared Gd3+-doped ZnO varying the Gd3+ concentration between 0.05 and 0.2 wt%. The sample with 0.05 wt% Gd3+ has shown better photoactivity as compared to the other samples. Yayapao et al. [88] reported the degradation of MB under UV–Vis illumination in presence of Gd3+-doped ZnO and also observed increased photoactivities for doped samples.

The photocatalytic degradation of methylene blue under UV–Vis irradiation is in accordance with the Langmuir–Hinschelwood model, which is well established in the literature [89,90,91]. Pseudo-first order rate constants (k) could be obtained by linear fittings of ln (A/A0) versus the irradiation time, Fig. 8. A is the absorbance of the solution at time t > 0, and A0 is the initial absorbance of the solution after the equilibration in the dark. The rate constants were calculated along with the percentage of methylene blue discoloration and the photonic efficiency (ξ), Table 4. The results confirmed the improvement of the photocatalytic activity of ZnO due to the doping with 1.0 mol% of Gd3+.

Fig. 8
figure 8

Degradation of MB plotted as—ln (A/A0) as a function of irradiation time using different catalysts, adjusted according to a pseudo-first order kinetics

Full size image
Table 4 Observed rate constants (k), percentages of MB discoloration and photonic efficiencies (ξ) for different photocatalysts, under the irradiation conditions employed in this work (9.20 × 10−8 Einstein s−1)
Full size table

The possible mechanism for describing the methylene blue photodegradation is illustrated in Fig. 9. The oxygen vacancies (V +o ) and Gd3+ ion will act as electron acceptor, interstitial oxygen (Oi) will capture the holes, retarding the recombination of the charge carriers. New intermediate energy levels located between the valence band and the conduction band of ZnO are generated by the Gd3+ doping. The photoinduced holes react with adsorbed H2O generating the oxygen radical species, responsible for the degradation of organic contaminants [7, 32, 43].

Fig. 9
figure 9

Proposed scheme for the photodegradation of methylene blue under UV–Vis light irradiation

Full size image

4 Conclusions

In summary, nanoplates of undoped ZnO an doped with Gd3+ ions were successfully synthetized using microwave hydrothermal method at only 90 °C for 16 min. XRD, Raman and photoluminescence confirmed the insertion of Gd3+ ions into ZnO crystal lattice. Are also investigated the influence of native defects of ZnO and Gd3+ ions on the photoluminescence and photocatalytic efficiency of the samples undoped ZnO an doped with Gd3+. The BET surface measurements showed that 1 mol% of Gd3+ exhibited the higher surface area. All samples presented photocatalytic activity for the degradation of MB. Moreover, the sample doped with 1.0 mol% of Gd3+ presented highest photocatalytic activity. The results showed that the insertion of dopant Gd3+ increased the number of structural defects of the ZnO crystal lattice, therefore, the oxygen vacancies played a relevant role in the enhanced photocatalytic efficiency.