Solvothermal synthesis of Zn_1−xMn_xO nanoparticles using oxalate precursor route: Optical and magnetic properties

1 Introduction

Incorporating impurity ions into a semiconducting host to extend its properties has been one of the most important techniques that paved the way for the modern technology based on spintronic devices. Compared with conventional semiconductor devices, the potential advantages of spintronic devices are to increase data processing speeds, decrease electric power consumption and increase integration densities. Experimentally, room temperature ferromagnetism has been reported in V (Hong et al., 2005), Co (Jayakumar et al., 2006), Ni (Schwartz et al., 2004), Fe (Karmarkar et al., 2007), Cu (Herng et al., 2007) and Mn (Guo et al., 2008) doped ZnO systems. The key need for achieving practical applications of spintronic devices is to increase the Curie temperature T_c of dilute magnetic semiconducting material to above room temperature.

Among all the transition metal doped ZnO, Mn-doped ZnO has its own advantage because of its intrinsic physical properties of a large magnetic moment due to the half-filled 3d band and relatively small ionic radius difference between Mn²⁺ (0.66 Å) and Zn²⁺ (0.60 Å) host cation (Sato and Yoshida, 2002; Mandal et al., 2006). Several experimental reports are available in the literature for the appearance of ferromagnetism above room temperature in Mn-doped ZnO synthesized by different methods such as solid state reaction, chemical and citrate method (Theodoropoulou et al., 2003; Norberg et al., 2004; Radovanovic and Gamelin, 2003; Sluiter et al., 2005) and also on thin films made under non-equilibrium conditions (Neal et al., 2006; Cho et al., 2002; Kittilstved et al., 2005; Heo et al., 2004). However, controversy between the research groups was reported on the magnetic behaviour of these materials. Sol–gel synthesis of Mn doped ZnO (1, 2, 3, 4, 5 at.%) showed different magnetic behaviour with different Mn concentration. 1% Mn exhibit diamagnetism, however 5% Mn show both paramagnetic and ferromagnetic behaviour (Omri et al., 2013). Recently, Mn-doped ZnO of a similar composition has been reported using reverse micelles (Khatoon et al., 2012) in which the size of nanoparticles was comparatively high (20–50 nm). Very recently, Zn_0.69Mn_0.31O (2–5 nm) nanoparticles exhibit paramagnetic behaviour with antiferromagnetic coupling (Khatoon et al., 2013).

In this paper, we report the optical and magnetic properties of Zn_1−xMn_xO nanoparticles synthesized by the solvothermal method through the oxalate precursor route. Recently, this method has been used for the preparation of nanocrystalline Zn_1−xNi_xO (Khatoon and Ahmad, 2012), Cd_1−xM_xO (M = Mn and Ni) (Ahmad et al., 2013; Ahmad et al., 2013) and In_2−xM_xO₃ (M = Mn and Ni) (Khatoon et al., 2012; Ahmad et al., 2013) with controlled morphology. Using this method, we have achieved a very high surface area for these nanoparticles for the first time.

2 Experimental

Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles were synthesized by the solvothermal method through the oxalate precursor route. All metal salt solutions were prepared in double distilled water of 0.1 M concentration. Stoichiometric amounts of zinc acetate dihydrate (CDH, 98.5%) and manganese acetate tetrahydrate (CDH, 99%) were well mixed in a 500 mL round bottom flask and stirred for about 30 min. The mixture was precipitated with 75 mL of aqueous solution of ammonium oxalate. A white suspension immediately formed. 75 mL of ethanol was also added to the reaction mixture to form the low boiling azeotrope with water. The reaction mixture was refluxed at about 80 °C for 12 h in the closed environment so that the volume of reaction mixture remains constant. The precipitate was recovered by centrifugation, washed several times with water and finally with acetone. The precipitate was dried in an oven at 55 °C and finally ground to powder. Zn_1−xMn_xO nanoparticles were obtained by thermal decomposition of the precursor at 450 °C for 6 h in air.

3 Characterization techniques

The crystalline phases of Zn_1−xMn_xO were characterized by powder X-ray diffraction using a Bruker D8 Advance X-ray diffractometer with Ni-filtered Cu-Kα radiation. Normal scans were recorded with a step size of 0.050° and step time of 1 s. Thermal decomposition of the samples were studied by thermogravimetry (TG) and DTA with a EXSTAR 6000 instrument in nitrogen atmosphere at a heating rate of 15 °C min⁻¹ with alumina as a reference sample. The order of reaction and activation energy can be calculated from the thermogram using Freeman and Carroll relation (Freeman and Carroll, 1958): $$\frac{(-E/2.3R)\mathrm{\Delta }(1/\text{T})}{\mathrm{\Delta }\text{log}{W}_{\text{r}}}=-\text{x}+\frac{\mathrm{\Delta }\text{log}(\text{dw}/\text{dt})}{\mathrm{\Delta }\text{log}{W}_{\text{r}}}$$ where E is the activation energy, R is the gas constant, dw, dt and ΔT are the change in mass, time and temperature (in kelvin) respectively and W_r is the change in dw. The activation energy was determined from the slope (−E/2.3 R) by plotting a graph of [Δlog(dw/dt)/ΔlogW_r] Vs [Δ(T)⁻¹/ΔlogW_r].

The composition of solid solutions was estimated by ICP-MS analysis using Agilent 7500a Inductively Coupled Plasma (ICP) with MS Detector. The crystal size and morphology have been studied by transmission electron microscopy (TEM) using FEI Technai G² 20 transmission electron microscope with an accelerating voltage of 200 kV. The TEM specimens were prepared by dispersing the samples in ethanol and placing a drop of the dispersed sample in a copper grid. Room temperature optical properties were recorded using the ocean optics lambda-25 UV–visible spectrophotometer in the reflectance mode. The powder samples were compacted to pellets. The energy band gap of these nanoparticles can be further calculated by the reflection spectra using the Kubelka–Munk remission function F(R) (Kortum, 1969): $$\text{F}(\text{R})={(1-\text{R})}^2/2\text{R}$$ where R is the diffuse reflectance

Specific surface areas of the samples were determined by Quantachrome instruments (Model NOVA 2000e surface area and pore size analyser). Samples were degassed at 250 °C for 3 h in vacuum because of the removal of any adsorbed gases. After degassing, specific surface areas were obtained from the nitrogen adsorption experiments measured at 77.35 K. Magnetic properties of as-prepared Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles were investigated using a Quantum design physical properties measurement system.

4 Results and discussion

Fig. 1 shows TG and DTA curves recorded at 15 °C min⁻¹ up to 500 °C for the thermal decomposition of Mn-doped zinc oxalate precursors in nitrogen atmosphere. The TG curves display two well-defined weight loss steps. The first one is attributed to the loss of water of hydration of Mn-doped zinc oxalate in the temperature range 95–200 °C. The measured weight loss accompanying this step is about 18.5% of the total weight resulting in the elimination of two water molecules. It is characterized by a broad endothermic DTA peak in the temperature range of 105–220 °C. This dehydration step was found to be completed at about 200 °C. The anhydrous oxalate was thermally stable up to 330 °C after which it starts to decompose in the second TG step. Heating from 330–440 °C results in a complete collapse of the material to its corresponding oxides with the simultaneous evolution of CO and CO₂. The decarbonylation and decarboxylation reactions are characterized by the two exothermic DTA peaks in the temperature range 330–410 °C and 410–450 °C respectively, attributed to the decomposition of anhydrous oxalate with the formation of carbonate followed by the formation of its corresponding oxides. The residual weight corresponds to Mn-doped ZnO, which is confirmed by powder X-ray diffraction studies. On the basis of TG and DTA results, brown precipitate of Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles were obtained by the decomposition of precursors at 450 °C for 6 h.

Close

Figure 1 TG/DTA curves of Zn_1−xMn_xC₂O₄ nanoparticles for x = (a) 0.021, (b) 0.059, and (c) 0.097.

Export to PPT

Close

Figure 2 [Δlog (dw/dt)/ΔlogW_r] Vs [Δ(T)⁻¹/ΔlogW_r] plot for the determination of order of reaction and activation energy of Zn_1−xMn_xC₂O₄ nanoparticles for x = (a) 0.021, (b) 0.059 and (c) 0.097.

Export to PPT

Close

Export to PPT

Close

Figure 3 Powder X-ray diffraction pattern of Zn_1−xMn_xO nanoparticles for x = (a) 0.022, (b) 0.061, and (c) 0.098.

Export to PPT

The actual percentage of manganese in the doped Zn-oxalate and zinc oxide nanoparticles was obtained by ICP-MS analysis. Result showed that the amount of manganese incorporated into the Zn host lattice is less than the amount of Mn used during the synthesis. This may be due to the less solubility of Mn than Zn. The ICP data and estimated molecular formula of Zn_1-xMn_xC₂O₄ and Zn_1−xMn_xO using ICP is shown in Tables 1 and 2 respectively.

The morphology and particle size have been estimated from transmission electron microscopic (TEM) studies. The TEM micrograph of Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles, shown in Fig. 4, indicates that the particles are uniform in size and hexagonal in shape. The average grain size estimated from these micrographs was found to be 18, 15 and 9 nm respectively which decreases on increasing Mn concentration in Zn lattice. The decrease in average particle size with increasing dopant ion concentration may be due to the lattice distortion and the internal stress which arises from the radius difference between Mn²⁺ and Zn²⁺ (Wei et al., 2008; Lee et al., 1998, Ahmad et al., 2013).

Close

Figure 4 TEM micrographs of Zn_1−xMn_xO nanoparticles for x = (a) 0.022, (b) 0.061, and (c) 0.098.

Export to PPT

Room temperature UV–visible spectroscopic measurements were carried out in the range 200–900 nm to study the effect of manganese doping on the band gap of ZnO. Fig. 5 shows the reflectance spectra of Mn-doped ZnO nanoparticles. The first absorption band appeared at around 360 nm, which is the characteristic absorption edge of ZnO. Apart from the band gap transition, other absorption edges were also appeared. The second strong absorption in the range of 410–510 nm corresponds to the ⁶A₁(S) → ⁴T₂(G) transition at Mn²⁺ sites. Similar spin forbidden transition is reported earlier for Mn-doped ZnO system (Norberg et al., 2004; Fukumura et al., 1999; Bates et al., 1966; Jin et al., 2000; Deka and Joy, 2007). This may be due to the lattice distortion which arises because of the incorporation of larger Mn²⁺ ions in ZnO matrix (Norberg et al., 2004; Fukumura et al., 1999; Bates et al., 1966; Jin et al., 2000; Deka and Joy, 2007). A weak absorption peak appearing at around 680 nm may be due to spin forbidden ⁶A₁ (S) → ⁴T₁ (G) transition of Mn²⁺ (Bates et al., 1966; Deka and Joy, 2007). Thus, optical studies confirmed the presence of Mn²⁺ ions in tetrahedral sites. The band gap of these nanoparticles is calculated from the Kubelka–Munk plot (Fig. 6) and comes out to be 3.32, 3.30 and 3.28 nm respectively which decreases on increasing Mn concentration from 2.19% to 9.8% respectively. The decrease in the band gap may be due to the sp-d exchange interactions between the band electrons and the localized d electrons of the substituted Mn²⁺ ions (Ogale et al., 2003; Park and Kim, 2003; Bouloudenine et al., 2004; Colis et al., 2006; Bouaine et al., 2007; Khatoon and Ahmad, 2012; Khatoon et al., 2013).

Close

Figure 5 UV–visible reflectance spectra of Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles.

Export to PPT

Close

Figure 6 [F(R) × Eg]² versus energy plot for direct band gap determination of Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles.

Export to PPT

Fig. 7 shows the N₂ adsorption–desorption isotherm of Mn-doped ZnO nanoparticles. Isotherm displays the type IV curve with H3 type hysteresis loop. This type of hysteresis loop usually appeared for mesoporous materials (Sing et al., 1985). The specific surface area of these nanoparticles is estimated from BET plots in the P/P₀ ranging from 0.05 to 0.30 and comes out to be 255.4, 520.4 and 582.9 m²g⁻¹ respectively. This is the highest value of the surface area as reported earlier to the best of our knowledge. It has been observed that the surface area increases on increasing the dopant ion concentration. This may be due to the decrease in particle size. The pore size has been determined from the BJH method (Fig. 8) and found to be 18.38, 16.42 and 16.39 Å respectively. The increase in surface area and decrease in pore radius on increasing the Mn concentration are attributed to the decrease in average grain size due to an increase in disorder with the incorporation of Mn ions in Zn matrix (Barick et al., 2008). Variations of surface area, pore radius and particle size of Mn-doped ZnO have been tabulated in Table 3.

Close

Figure 7 Nitrogen adsorption–desorption isotherms of Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles.

Export to PPT

Close

Figure 8 BJH plot of Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles.

Export to PPT

The temperature dependence (5 K < T < 300 K) of the molar magnetic susceptibility and inverse molar magnetic susceptibility of Mn-doped ZnO solid solutions is plotted in Fig. 9. The molar magnetic susceptibility decreases continuously with increasing temperature, indicating the paramagnetic behaviour of the samples. It can be clearly seen from the inverse molar susceptibility versus temperature plot that ${\chi }_{\text{M}}^{-1}$ vary linearly with temperature. The result is consistent with the Curie–Weiss equation $$\chi =C/(T+\theta )$$ where, χ is the magnetic susceptibility, C is the Curie constant and θ is the Curie–Weiss temperature. The Curie–Weiss temperature for Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) is close to zero. The exact temperature ranges of Curie–Weiss behaviour depend on the composition of the solid solutions. Using the least square fitting procedure, Curie constants and the effective magnetic moment have been determined and come out to be −0.4, −7 and −12 K respectively. This indicates the antiferromagnetic exchange between Mn ions is very weak. The calculated effective magnetic moment, μ_eff comes out to be 4.91, 4.67 and 4.18 μ_B/Mn²⁺ respectively. It is found that μ_eff for the sample with low doping concentration (x = 0.022) is consistent with the theoretical value (5 μ_B/Mn²⁺), indicating the presence of Mn²⁺. However, the effective magnetic moment decreases on increasing the Mn concentration. The lower value of the effective magnetic moment may be due to some mixed valence state at manganese site.

Close

Figure 9 Temperature dependence of molar magnetic susceptibility and inverse susceptibility plot of Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles measured at magnetic field of 1 kOe.

Export to PPT

Additional information of magnetic behaviour can be achieved from magnetic field versus magnetic moment curve (Fig. 10). All curves passing to zero, confirmed the paramagnetic nature of Mn-doped ZnO nanoparticles. Hence all the studies provide strong evidence that the 2.2% Mn ions were completely substituted in ZnO matrix.

Close

Figure 10 Magnetic field versus magnetic moment plot of Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles measured at temperature of 5 K.

Export to PPT

5 Conclusions

Zn_1−xMn_xO (x = 0.022, 0.061 and 0.098) nanoparticles have been synthesized by the solvothermal method through the oxalate precursor route. The kinetic analysis according to the Freeman and Carroll method showed an increase in activation energy with increasing manganese concentration. XRD results revealed that the solubility limit of Mn²⁺ ions in ZnO is less than 2.2%. The appearance of ⁶A₁(S) → ⁴T₁(G) and ⁴T₁→ ⁴A₂ transitions confirmed the incorporation of Mn²⁺ in ZnO lattice. The energy band gap of these nanoparticles decreases on increasing Mn content. TEM studies showed that the particle size decreases (18 to 9 nm) on increasing Mn concentration. The very high surface area of 255.4, 520.4 and 582.9 m²g⁻¹ respectively has been obtained for the first time. All the samples showed paramagnetism with very weak antiferromagnetic interactions. Collective evidence from X-ray diffraction, reflectance and magnetic measurements suggests that Mn²⁺ is incorporated in ZnO lattice site.