ZnO based ceramic semiconductors are widely used as gas sensors [1], piezoelectrics, electrodes for solar cells, phosphors, transparent conducting films and varistors [2]. Varistors possess high energy absorption capability against various surges. They are extensively used as protective devices to regulate transient voltage surges of unwanted magnitudes due to their fast response to over voltage transients. They sense and clamp transient voltage pulses in nanosecond speed [3]. The exact role of many additives in the electronic structure of ZnO based varistors is uncertain. Improving the electrical properties of ZnO based varistors is under research. The formulation of varistors that have high non-linear characteristics is the most important parameter to consider. Varistors are formed with small amounts of other metal oxides such as Al₂O₃, Bi₂O₃, Co₃O₄, Cr₂O₃, MnO, Sb₂O₃, TiO₂, etc. Such additives are the main tools that are used to improve the non-linear response and stability of the varistor [4]. This nonlinear response can be explained by the mechanism concerning the grain boundaries and associated defect concentration gradients [5]. Electrical properties of the ceramic ZnO depend on the distribution of vacancies, impurities and their behavior. Much has been done in I–V characterization of ZnO based varistors in previous studies [4]. It is essential to obtain information on the optical absorption behavior of the ceramic ZnO doped with metal oxides for the examination of electronic states. The optical absorption behavior of ZnO doped with MnO₂, Co₃O₄ is discussed for the different doping levels of V₂O₅ at different processing conditions.

ZnO (99.9% pure, Alfa Aesar) was doped with 0.4MnO₂ (99.999% pure, Alfa Aesar), 0.4Co₂O₃ (99.7% pure, Alfa Aesar) and xV₂O₅ (99.6% pure, Alfa Aesar) where x = 0–1.5 mol%. The detail of the composition is given in Table 1. Powder of all ingredients (24 hour ball milled) of each mole percent was pre-sintered at 700 °C for 90 minutes in open atmosphere at a heating and cooling rate of 5 °C/min. Samples were ground and polyvinyl alcohol (1.1 wt %) was added as a binder. The dried powder was pressed under a force of 800 kg cm⁻² to form a disk of 10 mm diameter. Finally the disks were sintered at 950 and 1025 °C for 2 hours in air at a heating and cooling rate of 4 °C min⁻¹. The disk from each sample was ground for 1 hour and granulated by sieving through a 75-mesh screen for the photopyroelectric (PPE) spectroscopy and XRD analysis. Density was calculated using geometrical method. Polished samples were thermally etched for microstructure analysis. Grain size was determined by the grain boundary crossing method. Cu K_α radiation with PANAalytical (Philips) X’Pert Pro PW1830 was used for X-ray analysis. XRD data were analyzed by X’Pert High Score software.

The measurement of PPE signal amplitude has been described elsewhere [6]. A light beam (300 to 800 nm) from one kW Xenon arc lamp, mechanically chopped at 10 Hz was used for PPE measurements. Optical absorption coefficient (β) varies with the excitation photon energy (hυ) [7]. It is given by the expression, (βhυ)² = (hυ − E_g), where hυ is the photon energy, C is constant and E_g is the optical energy band gap. PPE signal intensity (ρ) is directly proportional to β, hence (ρhυ)² is related to hυ linearly. From the plot of (ρhυ)² versus hυ, E_g is obtained by extrapolating the linear fitted region to zero.

Optical absorption edge has been observed in a variety of crystalline and amorphous materials. The optical-absorption edge has an important role in electron or exciton-phonon interactions [8]. It is found that PPE signal intensities plotted semi logarithmically vary linearly with the photon energy just lower than the fundamental absorption edge [9]. Therefore, an empirical relation for absolute measuring temperature (T) and photon energy (hυ) is given by the equation: (1) P = P 0 e ( σ ( h υ − h υ 0 k T )where k is the Boltzmann constant and P₀, σ, υ₀ are fitting parameters [10,11]. The value σ/kT determines the exponential slope, where σ is the steepness factor and is characterized in optical absorption edge. The steepness factor is found (σ_A in region-A and σ_B in region-B) from the PPE spectrum.

The XRD pattern of the V₂O₅ doped ZnO for the sintering temperature of 1025 °C can be seen in Figure 1. The samples at 0 mol% of V₂O₅ contain small peaks related to Co₃O₄ (reference code 00-042-1467) at both sintering temperatures but peaks are clearer at 1025 °C. Very small peaks related to ZnMn₂O₄ (reference code 01-077-0470) were also found at both sintering temperatures. Samples doped at 0.3 mol% V₂O₅ contain the secondary phases Zn₃(VO₄)₂ (reference code 00-034-0378), Zn₄V₂O₉ (reference code 01-077-1757). The same phases are also found at all higher doping levels of V₂O₅.

The density increases from 57.8 to 91.8% of the theoretical density at a sintering temperature of 925 °C for a 2 hour sintering time, Figure 2. The density increases with the increase of V₂O₅ mol% and is in accordance with the literature [12]. The density increase at 925 °C above 0.7 mol% of V₂O₅ indicates that the densification process is essentially completed at the sintering temperature above 900 °C [12]. It is expected that the vanadium-rich liquid phase Zn₃(VO₄)₂ enhances the densification by a solution and re-precipitation of ZnO [13]. The density of the ceramic is increased from 62.8 to 82.4% for the sintering temperature of 1025 °C. Density increases slowly compared to at the lower sintering temperature. The density has a lower value above 0.7 mol% of V₂O₅ at a sintering temperature of 1025 °C than 925 °C. This lower density may be due to the volatility of V₂O₅ [13].

Examination of the microstructure, Figure 3, shows that the grain size of the ceramic at 0 mol% of V₂O₅ is 2.8 and 3.1 μm and is increased to 8.1 and 9.48 μm at the sintering temperature of 925 and 1025 °C, respectively. The grain size is increased with the increase of sintering temperature at all mol% of V₂O₅. Large grains have oblong shape and small grains are also found in the ceramic. Exaggerated ZnO grain growth is found in the samples, Figure 4. This is due to the high reactivity of the V-rich liquid phase during sintering, which causes abnormal grain growth [14]. The addition of V₂O₅ can enhance the densification and grain growth behavior. This fact can be attributed to the formation of Zn₃(VO₄)₂, which acts as a liquid phase sintering aid [12]. EDX analysis shows that the vanadium is distributed at the grain boundaries as well as triple point junction [15]. Co and Mn are distributed in the grain boundaries and in the grain interiors [16]. The value of E_g is reduced from 3.2 eV (pure ZnO) to 2.28 and 2.54 eV at 0 mol% of V₂O₅ for the sintering temperatures of 1025 and 925 °C, respectively (Figure 5). This is due to 0.4 mol% of MnO₂ and Co₃O₄ because the reduction of E_g is due to the introduction of interface states by Mn and Co ions as the ionic radius of Co and Mn is smaller than that of Zn²⁺. With the addition of 0.3 mol% of V₂O₅, the E_g decreases from 2.28 and 2.54 eV to 2.17 and 2.11 eV at 1025 and 925 °C, respectively. It is due to the introduction of the interface states.

The ionic radius of Zn²⁺ is 0.74 Å and the ionic radius of Vanadium is 0.59 Å, so the reduction in the value of E_g at 0.3 mol% of V₂O₅ is due to the limited substitution of Vanadium ions in the ZnO lattice. The value of E_g is increased with the doping level of V₂O₅ beyond 0.3 mol%. It is expected that this may be due to the segregation of the V₂O₅ forming secondary phases Zn₃(VO₄)₂ and Zn₄V₂O₉ and reduces the interface states. The further increase in the value of E_g may be due to the high volatility of V₂O₅ at the high sintering temperature of 1025 °C. The steepness factor σ_A, Figure 6, increased with the increase of V₂O₅ doping level for the both sintering temperatures 925 and 1025 °C for the 2 hour sintering time. The increase in the value of σ_A with the doping level indicates the decrease in the PPE signal intensity. The decrease in the PPE signal intensity corresponds to the decrease in structural disordering. This indicates the decrease in the interface states with the doping level of V₂O₅. Resultantly, the value of E_g increases slightly as shown in Figure 5.

Generally an exponential tail (in region-B) for crystalline semiconductors can be characterized by: (2) ( σ B / K T ) − 1 = A < U 2 > T / C oC_o is the exponential tail parameter of the order of unity and <U²>_T is the thermal average of the square of the displacement of the atoms from their equilibrium positions. The term <U²>_T expresses the energy of displacement of atoms [17,18].

The value of the steepness factor (σ_B) decreases with the increase of doping of V₂O₅ for sintering temperatures of 925 and 1025 °C, Figure 7. This indicates that the average thermal displacement energy of atoms is increasing. This increase indicates an increase in structural disordering. Thus, the value of E_g decreases. Above 0.3 mol%, the value of σ_B increases with the increase of V₂O₅ for both sintering temperatures of 925 and 1025 °C. This indicates that the average thermal displacement energy of atoms is decreasing, which indicates a decrease in structural disordering. Correspondingly, the value of E_g increases. This may be due to the high volatility of the V₂O₅ and the secondary phases developed in the ceramics.

XRD results confirm the hexagonal phase of ZnO and few peaks of the secondary phases ZnMn₂O₄, Zn₃(VO₄)₂ and Zn₄V₂O₉. The EDX analysis shows that the V, Co and Mn are distributed at the grain boundaries and grain interiors. The maximum and minimum grain size is found to be 9.48 μm for 1025 °C and 2.8 μm for the 925 °C sintering temperature, respectively. The maximum and minimum relative density is found to be 82.4 for 1025 °C and 57.9 for 925 °C sintering temperature, respectively. The optical energy band-gap is reduced to 2.11 eV for 0.3 mol% V₂O₅ at the sintering temperature of 1025 °C. PPE spectrometry proved to be a useful tool to study the optical absorption behavior along with the other electrical measurements of ZnO based varistors.