- Email: [email protected]
Effects of operation temperature in artificially aging of zinc oxide varistors by high current short impulses
Electric Power Systems Research 134 (2016) 145–151
Contents lists available at ScienceDirect
Electric Power Systems Research journal homepage: www.elsevier.com/locate/epsr
Effects of operation temperature in artificially aging of zinc oxide varistors by high current short impulses Credson de Salles a,∗ , Thiago Arantes Nogueira a , Estácio Tavares Wanderley Neto a , Manuel Luís Barreira Martinez a,1 , Alvaro Antonio Alencar de Queiroz b a b
High Voltage Laboratory of the Federal University of Itajubá, 37,500-903, Bairro Pinheirinho, Itajubá, Brazil Physics and Chemistry Institute of the Federal University of Itajubá, Av BPS n, 37,500-903, Bairro Pinheirinho, Itajubá, Brazil
a r t i c l e
i n f o
Article history: Received 5 November 2015 Accepted 18 January 2016 Keywords: Varistors ZnO Bi2 O3 vaporization High-current short impulses Varistor microstructure
a b s t r a c t Nowadays surge arresters combine a complex metal oxide varistor (MOV) technology inside a polymeric housing. MOV is a sintered polycrystalline ceramic based on zinc oxide (ZnO) and small amounts of other metallic oxides (additives) usually applied to the manufacturing of surge protective devices for overvoltage protection at all power system voltage classes. The effect of operation temperature on the chemistry and microstructure of commercial zinc oxide varistor (ZOV) submitted to accelerated electrical aging by high-current short impulses were investigated by scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), optical microscopy (OM), Vickers micro hardness and X-ray diffraction (XRD). The SEM analysis indicates the formation of micro voids in grain and some larger voids in grain boundary with increase of operation temperature. The EDS and XRD measurements suggests that the operation temperature in the interval of 333 K–353 K during the high-current short impulses promotes the volatilization of the Bi2 O3 phase and formation of the pyrochlore in the ZOV body. The results of this paper suggest the necessity of getting a better knowledge of the aging process of ZnO-based varistors submitted to high operational temperatures applied to the manufacturing of gapless metal oxide surge arresters (MOSA). © 2016 Elsevier B.V. All rights reserved.
1. Introduction The use of zinc oxide varistors (ZOV–MOV) have become widespread presenting a huge range of industrial applications in the manufacturing of surge protective devices for electric power lines and electronic systems [1–3]. The ZOV are polycrystalline ceramics manufactured by sintering approximately 90 mol% ZnO powder mixed with a great variety of oxides. The most common additives – dopants are Bi2 O3 , Sb2 O3 , MnO and CoO [4–10]. A number of other additives have also been reported, including, most recently, a vast literature about the use of the rare earths [11–17]. The ZnO interaction with the dopants introduces potential barriers at the ZnO grain boundary during the sintering and improves significantly the electrical properties of the ZOV, such as voltage gradient, nonlinear coefficient, leakage current and discharge current capability [18–21]. However, although a considerable progress
∗ Corresponding author. Tel.: +55 35 3622 3546; fax: +55 35 3622 3546. E-mail address: [email protected] (C. de Salles). 1 In memoriam. http://dx.doi.org/10.1016/j.epsr.2016.01.010 0378-7796/© 2016 Elsevier B.V. All rights reserved.
has been obtained in the manufacturing processes, many questions remain about the influence of the chemical nature in the electric charge transport and the interface conditions at the grain boundary of ZOV [22,23]. Aiming at finding a quantitative relationship between the microstructure and its electrical performance, a considerable amount of studies have been carried out on the effect of discharging high-current impulse amplitudes and the energy absorption capability of ZOV [24–28]. The results are in good agreement with the experimental observation that the aging of the ZOV is different for long and short duration current impulses [29,30]. Although the behavior of ZOVs under short high-current impulses (SHCI) has been extensively studied for a long time, there is no clear description about the exact mechanisms involved in their degradation [31–38]. At this sense, the most plausible explanation is based on the formation of a potential barrier in the surface layer of zinc–oxide grains, where Zn and O-vacancies act as acceptors and donors [39–42]. In spite of the large number of investigations about the influence of moisture on the ZOVs degradation, there are few studies on the chemical changes of ZOVs aged by SHCI [43–48]. However, these studies are still not completely satisfactory and the effect of the
146
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151
ambient, or better, operation, temperature on the accelerate aging of the ZOVs submitted to SHCI draws attention and some additional work. In this sense, as far as the authors are aware, no detailed physico-chemical investigations about the influence of the operation temperature in the microstructure of ZOVs during accelerate aging by SHCI were reported up to the moment. In order to obtain a better definition of the behavior of ZOVs under operational stresses, it is very important to know the acceptable ranges for the limits of the ZOVs electrical properties. Part of this knowledge, the principal goal of this paper, can be only achieved by carrying out a set of scientific experiments. This paper aims at evaluating the influence of the limits of the operation temperature on the microstructure of ZOVs submitted to aging by SHCI. 2. Experimental 2.1. Study of the current impulse on ZnO varistor sample This study considered detailed construction requirements for the manufacturing of metal oxide surge arresters for surge protection of medium and high voltage lines. The lightning discharge probability [49] and the operational conditions were also verified and considered. The ZOVs were manufactured according to the standard formulation for metal oxide gapless surge arresters for medium and high voltage systems (12 kV to 144 kV), therefore, applied to the design of high duty distribution and high voltage sub transmission metal oxide surge arresters. Initially, the leakage current and power losses, at the reference voltage, of 45 ZOVs samples, from the same manufacturing batch, were determined. These samples were divided in 3 sets of 15 samples. These first recorded data were used for a subsequent electrical evaluation of the aging of the ZOVs samples [50]. Each sample was conveniently accommodated inside nylon cells, equipped with two contact electrodes. Next, when necessary, before each SHCI application, the internal cells temperature were automatically raised from room temperature (293 K), monitored, and regulated to the final pre-defined testing temperatures (333 K and 353 K), by means of thermocouples introduced in a small drilled hole on the lateral surface of the cells, a heating rate of 5 K/min was applied. After reaching the testing temperature, the heating process was finalized, and in less than 1 min interval, that considers the heat losses of the cells, a 20 kA 8/20 s SHCI was applied to each cell-sample. Finally, the samples were left to cool down to room temperature (293 K) and a leakage current and power losses test was carried out again [50]. The test sequence considers 20 of such cycles of heating, SHCI application, cooling and leakage current and power losses evaluation [50]. The 20 kA level considers that microstructural aging effects at present a higher definition. This can be also verified for discharge current levels of 10 kA, 15 kA and 30 kA [50]. However, in these cases, the microstructural degradation is not well defined, once the microstructural stresses appear randomly throughout the samples. The choice also considers the possibility of evaluating deeply aged ZOVs, close to end of their technical life period. One sample of each of the 3 SHCI tested set (293 K—25 ◦ C, 333 K—60 ◦ C and 353 K—80 ◦ C) were randomly selected to be submitted to a microstructural evaluation. 2.2. Microstructural ZOV characterization In order to evaluate the effect of the operation temperature on the microstructure of commercial ZOVs aged by SHCI, four types of samples were evaluated. A standard and non aged one; a sample
degraded by SHCI at 293 K; a sample degraded by SHCI at 333 K and a sample degraded by SHCI at 353 K. The aged and unaged ZOV samples were analyzed through X-ray diffraction with phase quantification by Rietveld method. The phase transitions were monitored by X-ray diffractometer, Shimadzu XRD 6000, by using a monochromatized CuK␣ ˚ 40.0 kV, 30.0 mA) and diverradiation (graphite crystal, 1.5418 A, gence/reception slits of 2 mm/0.6 mm were used for collecting the XRD data at 10◦ < 2 < 70◦ , with 2 of 0.02◦ and a step time of 2 s. The X-ray diffraction data was refined by the RIETAN-2.000 Rietveld refinement program [51]. A Phillips XL30 scanning electron microscope coupled to energy dispersive X-ray spectrometer (SEM/EDS) was employed for the morphological and microstructural evaluation of the ZOVs before and after SHCI aging. Prior to investigation, the ZOVs samples were embedded into epoxy resin and then successively polished using different size of abrasives and diamond paste grits. Microstructural analysis was carried out after etching the polished surfaces with 6 M NaOH solution for about 5 min to get an enhanced topography and then washed carefully with large amount of distilled and deionized water. The specimens were sputtered with a thin film of carbon to make them conductive and improve image resolution. The grain sizes may be estimated through the linear intercept method [52]. The Vicker micro hardness (VHN) of ZOV samples before and after SHCI aging was determined using an MH-6 digital micro hardness tester (0.098–9.8 N). For the measurement of hardness, at first, the top face of the samples was ground and polished by the polishing machine. The VHN was estimated according to the following equation: VHN =
P d2
.C
(1)
where P is the applied load, d is the diagonal length of indenter impression and C is a constant that takes the value of 0.1891. 3. Results and comments The general microstructure of ZOV before and after SHCI aging tests was examined by SEM as shown in Fig. 1. The samples consisted mainly of ZnO and spinel grains, the latter of which were approximately 1 m in size. EDS coupled to SEM, in Fig. 2, indicates that ZOV samples showed significant amounts of bismuth oxide (Bi2 O3 ) and antimony oxide (Sb2 O3 ) dissolved in spinel phases. The Bi2 O3 and Sb2 O3 combined with ZnO grains produce highly non-Ohmic properties improving the non-linear coefficient conferring a high stability of the double Schottky barriers in ZOV [53,54]. Bi2 O3 is the most essential component for producing non-Ohmic behavior of ZOV whilst addition of Sb2 O3 controls the ZnO grain growth [55–57]. Due to the higher atomic weight of bismuth (Bi), the intergranular Bi2 O3 phase appears in SEM micrographs as a white phase, whereas the ZnO grains and spinel Zn7 Sb2 O12 appear light and dark grey respectively as shown in Fig. 1. The distribution of the elements in the ZOV samples before and after SHCI accelerated aging were measured by EDS according to the color mapping on the distribution of zinc, bismuth, antimony and oxygen as shown in Fig. 2. The micrographs of EDS colorful maps shown that three different phases could be readily identified in the microstructure of the ZOV before and after SHCI aging. The aged ZOV clearly show different microstructures relatively to the unaged ZOV. A ZnO phase was the most abundant which agrees deeply with the fact that ZnO is the main component of ZOVs. Before SHCI aging the Bi, O and Sb elements were distributed in the vicinity of grain boundary of the ZOV as in Fig. 2a. After SHCI aging the EDS mapping analysis – in Fig. 2b and d – indicated a Bi, Sb and oxygen enrichment of ZnO grains, as consequence of current localization, and the associated joule heating of the ZOVs
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151
147
Fig. 1. Scanning electron micrographs of commercial ZOV before (A) and after SHCI aging test at 293 K (B), 333 K (C) and 353 K (D). The samples were aged by SHCI of 20 kA 8/20 s. The Sp represents the spinel grains (Zn7 Sb2 O12 ). * pore generated by SHCI.
Fig. 2. Elemental mapping by EDS of commercial ZOV samples before (A) and after SHCI aging at 293 K (B), 333 K (C) and 353 K (D). The samples were aged by SHCI of 20 kA 8/20 S.
148
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151 160 140
VHN (GPa)
120 100 80 60 40 20
ZnO
ZnO20
ZnO60
ZnO 80
Samples Fig. 3. Hardness versus temperature for ZOV before (ZnO) and after SHCI at operation temperatures of 293 K (ZnO20), 333 K (ZnO60) and 353 K (ZnO80).
at the interfaces. These findings appear to be consistent with the electro migration of ions within the depletion layer in the direction of current and oxygen desorption mechanisms [58,59]. However, the previous mechanisms appears to be accompanied by the Bi and Sb vaporization followed by crystallizing, on cooling, leaving with the formation of a monatomic layer of absorbed Bi and Sb at the ZnO grains [60]. To determine the element content an EDS analysis it was applied to measure quantitatively the chemical composition of ZOV before and after SHCI, whose results are shown in Fig. 2. As it can be clearly observed before SHCI aging, ZOVs have three different phases formed by a ZnO matrix doped with Co, the spinel (Zn7 Sb2 O12 ) and bismuth oxide (Bi2 O3 ). Fig. 2a shows a large concentration of bismuth detected from the EDS data at the ZOV grain boundary within the intergranular layers before the aging by SHCI, which explains the large rise in the Schottky barrier of this device [61–63]. At same time, the presence of localized oxygen at grain boundary observed in EDS spectra before SHCI aging at Fig. 2 may be justified by the enrichment of the grain boundaries with oxygen during sintering and their presence can determine the height of the potential barriers to a large extent in microstructure of ZOV [64]. Although ZOVs present a good resistance to degradation caused by continuous operating voltage their aging/degradation or destruction caused by SHCI remains a serious problem for electrical engineering [65–67]. As it was clearly shown by SEM and EDS analysis in Figs. 1 and 2, the microstructure of the ZOVs are greatly influenced by the sample operation temperature during the SHCI aging. According to these results more micro voids into the grains and some larger voids in the grain boundaries were formed with increasing of the operation temperature during SHCI aging. The presence of micro voids and larger voids contributes to the densification decreasing of ZOV microstructure owing to operation temperature during SHCI aging tests. Fig. 3 demonstrates the VHN plotted as a function of operation temperature during the SHCI aging of ZOV. The present measurements seem to indicate that the operation temperature of the samples of 333 K and 353 K appears to be an important factor at SHCI ageing of ZOV. An approximately monotonically linear decrease in VHN with increasing operation temperature was observed for ZOV. This means that the increase of operation temperature can substantially suppress micro crack of ZOV and
Fig. 4. Illustration of a hole structure in ZOV caused by the SHCI at 353 K.
consequently VHN is decreased. This result suggests a weakness of the coupling between the varistor grains with the increase of operation temperature and pores formation induced by the SHCI. Then, one can say that the mechanical resistance of ZOV becomes lower as result of pores discussed above. The discharge of a high-current pulse through the ZOV may lead to an electrical current filament formation with a high temperature gradient within the ceramic body and therefore inducing a higher thermal stress, causing the melting and evaporation of the dopants. The dopants volatilization induces the formation of larger porosity in the grain boundary and the vanish of the potential barriers [68,69]. It has been proposed that the raise in temperature T at the time t, caused by the Joule heating, is given by the following equation [70]: 1 T (t) = To+ mC v
t P(t)dt
(2)
o
where To is the uniform initial temperature of the ZOV (298 K),m ( = 5420 kg m−3 ) is the mass, Cv is the specific heat capacity (550 J kg−1 K−1 ) at a constant volume, and P is the electric power developed during a SHCI discharge. According to Eq. (2), a large average gradient temperature should be instantaneously (s) generated in ZOV by SHCI and, in case of current concentration on flow through preferential paths, the rapid Joule heating is accompanied by rapid fusion followed by vaporization and the formation of holes in the ceramics body. An example of a hole in the ZOV microstructure generated by the SHCI is shown in Fig. 4. The hole was approximately 136 m in diameter and there is evidence of once-molten flow on the ZOV surface. This observed hole structure in ZOV strongly indicates that a current filament had formed with temperatures high enough to melt the ZnO locally. The quantitative EDS analysis shown in Fig. 5 indicates that the operation temperature of the samples (333–353 K) appears to be an important variable to be considered in accelerated aging tests of ZOV by SHCI as large as 20 kA amplitude. It is clearly visible in Fig. 5(B) and (D) that the Bi2 O3 phase seems to have been lost during high-current impulses at 333 K and 353 K. The EDS analysis suggests that the SHCI behavior of the ZOVs is thermally activated. Therefore, for a high energy short time impulse, the amount of heat in the device can locally raise the temperature quickly. Therefore, the operation temperature increases as
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151
149
Fig. 5. EDS chemical mapping of commercial ZOV samples before (A) and after electrical aging by 20 kA SHCI at 293 K (B), 333 K (C) and 353 K (D). The red arrows indicate a possible volatilization of Bi2 O3 phase from varistor microstructure.
the direct solar exposure of the ZOV in tropical or desert countries should be sufficient to activate the thermal avalanche to start chemical changes in varistor device [71]. The temperature increase can be also result of a set of switching operations, during a storm, a common event in tropical regions. In spite of the large number of
investigations on the electrical properties of commercial ZOVs, as far as the authors are aware, the aging at high operational temperatures, due to, for example, excessive solar exposure or previous switching surge operations, appears not be sufficiently studied [71,72].
150
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151
Fig. 6. XRD patterns of ZOV commercial samples before (C) and after 20 kA SHCI at 293 K (A), 333 K (B) and 353 K (D). The crystalline phases identified in the ZOV microstructure are: ZnO (•), Bi2 O3 (), Sb2 O3 (), Zn7 Sb2 O12 (), Zn2 Bi3 Sb3 O14 (䊏), Co2 SbO4 ( ).
XRD data were taken from the standard ZOV (before aging) and the degraded samples by SHCI in air atmosphere and operation temperatures of 293 K, 333 K and 353 K. The main phases and their modification after the degradation process are evaluated in Fig. 6. The XRD pattern confirmed the presence of ZnO phases with hexagonal wurtzite structure and secondary phases. The secondary ZOV phases showed a complex microstructure consisting of several spinel-type crystallites at the grain boundaries like as Co2 SbO4 , ZnSb2 O4 and Bi3 SbO7 . The zinc-antimony spinel phase, Zn7 Sb2 O12 was also identified by XRD. The Bi2 O3 phases were not found after the ZOV aging by SHCIs at 333 K and 353 K. However, a pyrochlore phase (Zn2 Bi3 Sb3 O14 ) was observed in Fig. 6. This result suggests that the ZOV submitted to a SHCI stress leading to a rapid Joule heating is accompanied by a solid-state chemical reaction of Bi and Sb forming a pyrochlore phase in the ceramic body.
4. Conclusions The effect of operation temperature in the microstructure of ZOV after accelerated aging by SHCI was investigated in the present study. The SEM and EDS results indicate that the operation temperature seems to have a significant influence in the topology and chemical composition of ZOV, respectively. SEM analysis indicates a significant modification in the microstructure of the SHP aged ZOVs showing that more voids can be observed at the grains and grain boundaries. The elemental EDS mapping indicates that environmental temperature promotes the electro migration of oxygen out at grain boundary with segregation of Sb and Bi at ZnO grain surface during SHCI aging tests. XRD investigation suggests that the increase of operation temperature induces the formation of a pyrochlore phase during the accelerated SHP ZOV aging through a state-solid chemical reaction between spinel phase and Bi2 O3 present at the multiple ZnO grain junctions. However, more detailed characterizations using transmission electron microscopy (TEM) and ac impedance techniques are being conducted in order to achieve an understanding of all the phenomena involved during the state-solid chemical changes in
commercial ZOV during SHCI aging tests and the influence of environmental temperature.
References [1] Relva C. Buchanan, Ceramic Materials for Electronics, Marcel Dekker, New York (EUA), 2004, pp. 431–464. [2] A. Janotti, Chris G. Van de Walle, Fundamentals of zinc oxide as a semiconductor, Rep. Prog. Phys. 72 (2009) 126501 (29pp). [3] P.E.J. Flewitt, R.K. Wild, Grain Boundaries: Their Microstructure and Chemistry, John Wiley & Sons, New York (EUA), 2001. [4] S. Bernik, S. Macek, A. Bui, The characteristics of ZnO–Bi2 O3 -based varistor ceramics doped with Y2 O3 and varying amounts of Sb2 O3 , J. Eur. Ceram. Soc. 24 (2004) 1195–1198. [5] M. Peiteado, J.F. Fernández, A.C. Caballero, Varistors based in the ZnO–Bi2 O3 system: microstructure control and properties, J. Eur. Ceram. Soc. 27 (2007) 3867–3872. [6] Y. Abdollahi, A. Zakaria, R.S. Aziz, S.N. Tamili, K.A. Matori, N.M. Shahrani, N.M. Sideki, M. Dorraj, S. Moosavi, Optimizing Bi2 O3 and TiO2 to achieve the maximum non-linear electrical property of ZnO low voltage varistor, Chem Cent J 7 (2013) 137. [7] M. Houabes, R. Metz, Rare earth oxides effects on both the threshold voltage and energy absorption capability of ZnO varistors, Ceram. Int. 33 (2007) 1191–1197. [8] M. Dorraj, Y. Abdollahi, S.B.M. Said, M.F.B.M. Sabri, N.A. Sairi, W.P. Meng, E. Abouzari-lotfe, Enhance protection of electronic appliances through multivariate modelling and optimization of ceramic core materials in varistor devices, R. Soc. Chem. Adv. 5 (2015) 21384–21395. [9] R. Metz, H. Delalu, J.R. Vignalou, N. Achard, M. Elkhatib, Electrical properties of varistors in relation to their true bismuth composition after sintering, Mater. Chem. Phys. 63 (2000) 157–216. [10] P. Duran, F. Capel, J. Tartaj, C. Moure, A strategic two-stage low temperature thermal processing leading to fully dense and fine-grained doped-ZnO varistors, Adv. Mater. 14 (2002) 137–141. [11] F. Jiang, Z. Peng, Y. Zang, X. Fu, Progress on rare-earth doped ZnO-based varistor materials, J. Adv. Ceram. 2 (2013) 201–212. [12] M. Houabes, R. Metz, Rare earth oxides effects on both the threshold voltage and energy absorption capability of ZnO varistors, Ceram. Int. 33 (2007) 1191–1197. [13] N.T. Hung, N.D. Quang, S. Bernik, Electrical and microstructural characteristics of ZnO–Bi2 O3 -based varistors doped with rare-earth oxides, Int. J. Mater. Res. 16 (2001) 2817–2823. [14] C.-W. Nahm, C.-H. Park, H.-S. Yoon, Microstructure and varistor properties of ZnO–Pr6 O11 –CoO–Nd2 O3 based ceramics, J. Mater. Sci. Lett. 19 (2000) 271–274. [15] H. Liu, H. Kong, D. Jiang, W. Shi, X. Ma, H. Zhang, Effects of cooling rate on the microstructure and electrical properties of Dy2 O3 -doped ZnO-based varistor ceramics, Rare Met. 26 (2007) 39–44. [16] C.-W. Nahm, Microstructure, electrical properties, and dc aging characteristics of Tb4 O7 -doped ZnO-based varistors, J. Mater. Sci. 43 (2008) 2857–2864.
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151 [17] J. Cai, Y.-H. Lin, M. Li, et al., Sintering temperature dependence of grain boundary resistivity in a rare-earth-doped ZnO varistor, J. Am. Ceram. Soc. 90 (2007) 291–294. [18] N. Daneu, S. Bernik, A. Reˇcnik, Inversion boundary induced grain growth in ZnO ceramics: from atomic-scale investigations to microstructural engineering, J. Phys. Conf. Ser. 326 (2011) 012003, http://dx.doi.org/10.1088/1742-6596/326/ 1/012003. [19] M. Peiteado, J.F. Fernández, A.C. Caballero, Processing strategies to control grain growth in ZnO based varistors, J. Eur. Ceram. Soc. 25 (12) (2005) 2999–3003. [20] A. Sawalha, M. Abu-Abdeen, A. Sedky, Electrical conductivity study in pure and doped ZnO ceramic system, Physica B 404 (2009) 1316–1320. [21] J.-L. He, J. Hu, Discussions on nonuniformity of energy absorption capabilities of ZnO varistors, IEEE Trans. Power Delivery 22 (2007) 1523–1532. [22] H.L. Mosbacker, Y.M. Strzhemechny, B.D. White, P.E. Smith, D.C. Look, D.C. Reynolds, C.W. Litton, L.J. Brillson, Role of near-surface states in ohmic-Schottky conversion of Au contacts to ZnO, Appl. Phys. Lett. 87 (2005) 012102-2–0121023. [23] M.A. Alim, S. Li, F. Liu, P. Cheng, Electrical barriers in the ZnO varistor grain boundaries, Phys. Status Solid 203 (2006) 410–427. [24] A.N.M. Karim, S. Begum, M.S.J. Hashmi, Performance and failure during energy testing of zinc oxide varistor processed from different powder size fraction and passivation thickness, Int. J. Mech. Mater. Eng. 5 (2010) 175–181. [25] C. Zhang, S. Zhang, The study on the pulse-degradation and destruction mechanisms of zinc oxide nonlinear varistors, J. High Voltage Eng. 26 (2000) 48–52. [26] Z. Zheng, H. Cui, Q. Wang, Y. Tu, Space charge characteristics of ZnO varistors after pulse degradation, IEEE Power Energy Eng. Conf. (2012) 1–4, http://dx. doi.org/10.1109/APPEEC.2012.6307387. [27] M. Bartkowiak, M.G. Comber, G.D. Mahan, Influence of nonuniformity of ZnO varistors on their energy absorption capability, IEEE Trans. Power Delivery 16 (2001) 591–598. [28] K.P. Mardira, T.K. Saha, R.A. Sutton, The Effects of Electrical Degradation on the Microstructure of Metal Oxide Varistor, IEEE, 2001, doi: IEEE 0-7803-72859/01. [29] M. Lengauer, D. RubesÏa, R. Danzer, Finite element modelling of the electrical impulse induced fracture of a high voltage varistor, J. Eur. Ceram. Soc. 20 (2000) 1017–1021. [30] K.J. Brown, Metal oxide varistor degradation, International Association of Electrical Inspectors (IAEI) News March–April (2004) 2–8. [31] K.L. Chrzan, Z. Wróblewski, Degradation and destruction of ZnO varistors caused by current pulses, in: International Conference on Advances in Processing, Testing and Application of Dielectric Materials, 2001. [32] M. Jaroszewski, J. Pospieszna, An assessment of ageing of oxide varistors exposed to pulse hazards using dielectric spectroscopy, in: International Conference on Solid Dielectrics, July 5–9, 2004, Toulouse, France, 2004. [33] C. Zhang, S. Zhang, The study on the pulse-degradation and destruction mechanisms of zinc oxide nonlinear varistors, J. High Voltage Eng. 26 (2000) 48–52. [34] A.S. Tonkoshkur, A.B. Glot, A.V. Ivanchenko, Percolation effects in dc degradation of ZnO varistors, J. Adv. Dielectr. 5 (2015) 1–8. [35] J. Liu, J. He, J. Hu, W. Long, F. Luo, Statistics on the AC ageing characteristics of single grain boundaries of ZnO varistor, Mater. Chem. Phys. 125 (2011) 9–11. [36] T. Tada, Degradation of ZnO varistors as estimated by aging tests, Electr. Eng. Japan 170 (2010) 661–674. [37] A.S. Tonkoshkur, A.Y. Lyashkov, I.V. Gomilko, A.V. Ivanchenko, Effect of longterm electrical degradation on the distribution of donor impurities in ZnO varistor ceramics, Inorg. Mater. 36 (2000) 745–748. [38] A.S. Tonkoshkur, I.V. Gomilko, A.Y. Lyashkov, Percolation effects in the capacitive properties of metal-oxide varistors in the range of high voltage, J. Adv. Dielectr. 4 (2014) 1–5. [39] V. Hinrichsen, Metal-Oxide Surge Arresters in High-Voltage Power Systems: Fundamentals, third edition, Siemens AG, Erlangen (DE), 2012. [40] R.C. Buchanan, Ceramic materials for electronics, Marcel Dekker, New York (EUA), 2004, pp. p441. [41] J. Ott, A. Lorenz, M. Harrer, E.A. Preissner, C. Hesse, A. Feltz, A.H. Whitehead, M. Schreiber, The influence of Bi2 O3 and Sb2 O3 on the electrical properties of ZnO-based varistors, J. Electroceram. 6 (2001) 135–146. [42] A. Sawalha, M. Abu-Abdeen, A. Sedky, Electrical conductivity study in pure and doped ZnO ceramic system, Physica B 404 (2009) 1316–1320. [43] K. Chrzan, J. Andino, Electrical strength of air containing by-products of intensive partial discharges, IEEE Trans. Dielectr. Electr. Insul. 8 (2001) 607–611. [44] K.L. Chrzan, Influence of moisture and partial discharges on the degradation of high-voltage surge arresters, Eur. Trans. Electr. Power 14 (2004) 175–184. [45] L. Kaczmarek, G. Lentka, Problems of varistor quality assessment during exploitation, Metrol. Meas. Syst. XIX (2012) 395–404.
151
[46] M.-H. Wang, K.A. Hu, B.Y. Zhao, N.F. Zhang, Degradation phenomena due to humidity in low voltage ZnO varistors, Ceram. Int. 33 (2007) 151–154. [47] W.P. Chen, H.L.W. Chan, Water-induced degradation in Zinc oxide-based ceramic varistors, J. Am. Ceram. Soc. 85 (2003) 1625–1627. [48] W.P. Chen, H.L.W. Chan, Degradation of ZnO ceramic varistors induced by water and AC voltages, Japan J. Appl. Phys. 43 (2004) 701–702. [49] C. de Salles, A.F. Picanc¸o, M.L.B. Martinez, H.R.P.M. de Oliveira, Determination of the Discharge Current on Distribution Network Surge Arresters, in: Powertech 2009, 28 June–2 July, 2009, Bucharest, Romania, 2009. [50] C. de Salles, M.L.B. Martinez, A.A.A. Queiroz, Surge Ageing of Metal Oxide Varistors, in: 2015 Electrical Insulation Conference (EIC), 7–10 June 2015, Seattle, Washington, USA, 2015. [51] F. Izumi, T. Ikeda, A Rietveld-analysis program RIETAN-98 and its applications to zeolites, Mater. Sci. Forum 321–324 (2000) 198–203. [52] J. Luan, G. Liu, H. Wang, H. Ullah, On the sampling of three-dimensional polycrystalline microstructures for distribution determination, J. Microsc. 244 (2011) 214–222. [53] L. Madler, S.E. Pratsinis, Bismuth oxide nanoparticles by flame spray pyrolysis, J. Am. Ceram. Soc. 85 (7) (2002) 1713–1718. [54] S.C. Pillai, J.M. Kelly, D.E. McCormack, R. Rameshd, High performance ZnO varistors prepared from nanocrystalline precursors for miniaturised electronic devices, J. Mater. Chem. 18 (2008) 3926–3932. [55] D. Xu, L. Shi, Z. Wu, Q. Zhong, X. Wu, Microstructure and electrical of ZnO–Bi2 O3 based varistor ceramics by different sintering processes, J. Eur. Ceram. Soc. 29 (2009) 1789–1794. [56] M. Peiteado, J.F. Fernández, A.C. Caballero, Varistors based in the ZnO–Bi2 O3 system: microstructure control and properties, J. Eur. Ceram. Soc. 27 (2007) 3867–3872. [57] Choon-W. Nahm, Effect of sintering temperature on varistor properties and aging characteristics of ZnO–V2 O5 –MnO2 ceramics, Ceram Int 35 (2009) 2679–2685. [58] A.V. Ivanchenko, A.S. Tonkoshkur, Modeling of the degradation electromigrational process in structures with intercrystallite potential barriers, Multi Model Mater. Struct. 3 (2007) 477–490. [59] J. He, J. Liu, J. Hu, R. Zeng, W. Long, Non-uniform ageing behavior of individual grain boundaries in ZnO varistor ceramics, J. Eur. Ceram. Soc. 31 (2011) 1451–1456. [60] W. Onreabroy, N. Sirikulrat, C. Hammond, S.J. Milne, Properties and intergranular phase analysis of a ZnO–CoO–Bi2 O3 varistor, Solid State Ionics 177 (2006) 411–420. [61] A. Janotti, C.G. Van de Walle, Fundamentals of zinc oxide as a semiconductor, Rep. Prog. Phys. 72 (2009) 1–29. [62] D. Fernández-Heviaa, M. Peiteado, J. de Frutosa, A.C. Caballero, J.F. Fernández, Wide range dielectric spectroscopy of ZnO-based varistors as a function of sintering time, J. Eur. Ceram. Soc. 24 (2004) 1205–1208. [63] D. Fernández-Hevia, J. de Frutos, A.C. Caballero, J.F. Fernández, Mott–Schottky behavior of strongly pinned double Schottky barriers and characterization of ceramic varistors, J. Appl. Phys. 92 (2002) 2890–2898. [64] C.F. Klingshirn, A. Waag, A. Hoffmann, J. Geurts, Zinc Oxide: From Fundamental Properties Towards Novel Applications, Springer-Verlag, Berlim (De), 2010, pp. p114. [65] J. Liu, J.L. He, J. Hu, W.C. Long, R. Zeng, Statistical pulse degradation characteristics of grain boundaries in a ZnO varistor based on microcontact measurement, J. Am. Ceram. Soc. 93 (2010) 2473–2475. [66] G. Yin, J. Li, G. Yao, S. Li, A new view on impulse degradation of ZnO-based ceramics, in: Proceedings of the Ninth International Conference on Properties and Applications of Dielectric Materials, July 19–23 2009, Harbin, China, 2009, pp. 1137–1140. [67] H. Andoh, S. Nishiwaki, H. Suzuki, S. Boggs, J. Kuang, Failure mechanisms and recent improvements in ZnO arrester elements, IEEE Electr. Insul. Mag. 16 (2000) 25–31. [68] K. Bavelis, E. Gjonaj, T. Weiland, Modeling of electrical transport in zinc oxide varistors, Adv. Radio Sci. 12 (2014) 29–34. [69] T. Tada, Degradation of ZnO varistors as estimated by aging tests, Electr. Eng. Jpn. 170 (2010) 1–18. [70] M. Lengauer, D. RubesÏa, R. Danzer, Finite element modelling of the electrical impulse induced fracture of a high voltage varistor, J. Eur. Ceram. Soc. 20 (2000) 1017–1021. [71] F. Munteanu, F.M. Frigura-Iliasa, E. Cazacu, About establishing the functional limits of a ZnO varistor based surge-arrester, Revue Roumaine des Sciences Techniques - Serie Électrotechnique et Énergétique 52 (2007) 443–452. [72] A. Gubaski, W. Mielcarek, K. Prociów, J. Warycha, J.M. Wróbel, The effect of aluminium additive on the electrical properties of ZnO varistors, Mater SciPoland 27 (4/2) (2009) 1207–1218.
Contents lists available at ScienceDirect
Electric Power Systems Research journal homepage: www.elsevier.com/locate/epsr
Effects of operation temperature in artificially aging of zinc oxide varistors by high current short impulses Credson de Salles a,∗ , Thiago Arantes Nogueira a , Estácio Tavares Wanderley Neto a , Manuel Luís Barreira Martinez a,1 , Alvaro Antonio Alencar de Queiroz b a b
High Voltage Laboratory of the Federal University of Itajubá, 37,500-903, Bairro Pinheirinho, Itajubá, Brazil Physics and Chemistry Institute of the Federal University of Itajubá, Av BPS n, 37,500-903, Bairro Pinheirinho, Itajubá, Brazil
a r t i c l e
i n f o
Article history: Received 5 November 2015 Accepted 18 January 2016 Keywords: Varistors ZnO Bi2 O3 vaporization High-current short impulses Varistor microstructure
a b s t r a c t Nowadays surge arresters combine a complex metal oxide varistor (MOV) technology inside a polymeric housing. MOV is a sintered polycrystalline ceramic based on zinc oxide (ZnO) and small amounts of other metallic oxides (additives) usually applied to the manufacturing of surge protective devices for overvoltage protection at all power system voltage classes. The effect of operation temperature on the chemistry and microstructure of commercial zinc oxide varistor (ZOV) submitted to accelerated electrical aging by high-current short impulses were investigated by scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), optical microscopy (OM), Vickers micro hardness and X-ray diffraction (XRD). The SEM analysis indicates the formation of micro voids in grain and some larger voids in grain boundary with increase of operation temperature. The EDS and XRD measurements suggests that the operation temperature in the interval of 333 K–353 K during the high-current short impulses promotes the volatilization of the Bi2 O3 phase and formation of the pyrochlore in the ZOV body. The results of this paper suggest the necessity of getting a better knowledge of the aging process of ZnO-based varistors submitted to high operational temperatures applied to the manufacturing of gapless metal oxide surge arresters (MOSA). © 2016 Elsevier B.V. All rights reserved.
1. Introduction The use of zinc oxide varistors (ZOV–MOV) have become widespread presenting a huge range of industrial applications in the manufacturing of surge protective devices for electric power lines and electronic systems [1–3]. The ZOV are polycrystalline ceramics manufactured by sintering approximately 90 mol% ZnO powder mixed with a great variety of oxides. The most common additives – dopants are Bi2 O3 , Sb2 O3 , MnO and CoO [4–10]. A number of other additives have also been reported, including, most recently, a vast literature about the use of the rare earths [11–17]. The ZnO interaction with the dopants introduces potential barriers at the ZnO grain boundary during the sintering and improves significantly the electrical properties of the ZOV, such as voltage gradient, nonlinear coefficient, leakage current and discharge current capability [18–21]. However, although a considerable progress
∗ Corresponding author. Tel.: +55 35 3622 3546; fax: +55 35 3622 3546. E-mail address: [email protected] (C. de Salles). 1 In memoriam. http://dx.doi.org/10.1016/j.epsr.2016.01.010 0378-7796/© 2016 Elsevier B.V. All rights reserved.
has been obtained in the manufacturing processes, many questions remain about the influence of the chemical nature in the electric charge transport and the interface conditions at the grain boundary of ZOV [22,23]. Aiming at finding a quantitative relationship between the microstructure and its electrical performance, a considerable amount of studies have been carried out on the effect of discharging high-current impulse amplitudes and the energy absorption capability of ZOV [24–28]. The results are in good agreement with the experimental observation that the aging of the ZOV is different for long and short duration current impulses [29,30]. Although the behavior of ZOVs under short high-current impulses (SHCI) has been extensively studied for a long time, there is no clear description about the exact mechanisms involved in their degradation [31–38]. At this sense, the most plausible explanation is based on the formation of a potential barrier in the surface layer of zinc–oxide grains, where Zn and O-vacancies act as acceptors and donors [39–42]. In spite of the large number of investigations about the influence of moisture on the ZOVs degradation, there are few studies on the chemical changes of ZOVs aged by SHCI [43–48]. However, these studies are still not completely satisfactory and the effect of the
146
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151
ambient, or better, operation, temperature on the accelerate aging of the ZOVs submitted to SHCI draws attention and some additional work. In this sense, as far as the authors are aware, no detailed physico-chemical investigations about the influence of the operation temperature in the microstructure of ZOVs during accelerate aging by SHCI were reported up to the moment. In order to obtain a better definition of the behavior of ZOVs under operational stresses, it is very important to know the acceptable ranges for the limits of the ZOVs electrical properties. Part of this knowledge, the principal goal of this paper, can be only achieved by carrying out a set of scientific experiments. This paper aims at evaluating the influence of the limits of the operation temperature on the microstructure of ZOVs submitted to aging by SHCI. 2. Experimental 2.1. Study of the current impulse on ZnO varistor sample This study considered detailed construction requirements for the manufacturing of metal oxide surge arresters for surge protection of medium and high voltage lines. The lightning discharge probability [49] and the operational conditions were also verified and considered. The ZOVs were manufactured according to the standard formulation for metal oxide gapless surge arresters for medium and high voltage systems (12 kV to 144 kV), therefore, applied to the design of high duty distribution and high voltage sub transmission metal oxide surge arresters. Initially, the leakage current and power losses, at the reference voltage, of 45 ZOVs samples, from the same manufacturing batch, were determined. These samples were divided in 3 sets of 15 samples. These first recorded data were used for a subsequent electrical evaluation of the aging of the ZOVs samples [50]. Each sample was conveniently accommodated inside nylon cells, equipped with two contact electrodes. Next, when necessary, before each SHCI application, the internal cells temperature were automatically raised from room temperature (293 K), monitored, and regulated to the final pre-defined testing temperatures (333 K and 353 K), by means of thermocouples introduced in a small drilled hole on the lateral surface of the cells, a heating rate of 5 K/min was applied. After reaching the testing temperature, the heating process was finalized, and in less than 1 min interval, that considers the heat losses of the cells, a 20 kA 8/20 s SHCI was applied to each cell-sample. Finally, the samples were left to cool down to room temperature (293 K) and a leakage current and power losses test was carried out again [50]. The test sequence considers 20 of such cycles of heating, SHCI application, cooling and leakage current and power losses evaluation [50]. The 20 kA level considers that microstructural aging effects at present a higher definition. This can be also verified for discharge current levels of 10 kA, 15 kA and 30 kA [50]. However, in these cases, the microstructural degradation is not well defined, once the microstructural stresses appear randomly throughout the samples. The choice also considers the possibility of evaluating deeply aged ZOVs, close to end of their technical life period. One sample of each of the 3 SHCI tested set (293 K—25 ◦ C, 333 K—60 ◦ C and 353 K—80 ◦ C) were randomly selected to be submitted to a microstructural evaluation. 2.2. Microstructural ZOV characterization In order to evaluate the effect of the operation temperature on the microstructure of commercial ZOVs aged by SHCI, four types of samples were evaluated. A standard and non aged one; a sample
degraded by SHCI at 293 K; a sample degraded by SHCI at 333 K and a sample degraded by SHCI at 353 K. The aged and unaged ZOV samples were analyzed through X-ray diffraction with phase quantification by Rietveld method. The phase transitions were monitored by X-ray diffractometer, Shimadzu XRD 6000, by using a monochromatized CuK␣ ˚ 40.0 kV, 30.0 mA) and diverradiation (graphite crystal, 1.5418 A, gence/reception slits of 2 mm/0.6 mm were used for collecting the XRD data at 10◦ < 2 < 70◦ , with 2 of 0.02◦ and a step time of 2 s. The X-ray diffraction data was refined by the RIETAN-2.000 Rietveld refinement program [51]. A Phillips XL30 scanning electron microscope coupled to energy dispersive X-ray spectrometer (SEM/EDS) was employed for the morphological and microstructural evaluation of the ZOVs before and after SHCI aging. Prior to investigation, the ZOVs samples were embedded into epoxy resin and then successively polished using different size of abrasives and diamond paste grits. Microstructural analysis was carried out after etching the polished surfaces with 6 M NaOH solution for about 5 min to get an enhanced topography and then washed carefully with large amount of distilled and deionized water. The specimens were sputtered with a thin film of carbon to make them conductive and improve image resolution. The grain sizes may be estimated through the linear intercept method [52]. The Vicker micro hardness (VHN) of ZOV samples before and after SHCI aging was determined using an MH-6 digital micro hardness tester (0.098–9.8 N). For the measurement of hardness, at first, the top face of the samples was ground and polished by the polishing machine. The VHN was estimated according to the following equation: VHN =
P d2
.C
(1)
where P is the applied load, d is the diagonal length of indenter impression and C is a constant that takes the value of 0.1891. 3. Results and comments The general microstructure of ZOV before and after SHCI aging tests was examined by SEM as shown in Fig. 1. The samples consisted mainly of ZnO and spinel grains, the latter of which were approximately 1 m in size. EDS coupled to SEM, in Fig. 2, indicates that ZOV samples showed significant amounts of bismuth oxide (Bi2 O3 ) and antimony oxide (Sb2 O3 ) dissolved in spinel phases. The Bi2 O3 and Sb2 O3 combined with ZnO grains produce highly non-Ohmic properties improving the non-linear coefficient conferring a high stability of the double Schottky barriers in ZOV [53,54]. Bi2 O3 is the most essential component for producing non-Ohmic behavior of ZOV whilst addition of Sb2 O3 controls the ZnO grain growth [55–57]. Due to the higher atomic weight of bismuth (Bi), the intergranular Bi2 O3 phase appears in SEM micrographs as a white phase, whereas the ZnO grains and spinel Zn7 Sb2 O12 appear light and dark grey respectively as shown in Fig. 1. The distribution of the elements in the ZOV samples before and after SHCI accelerated aging were measured by EDS according to the color mapping on the distribution of zinc, bismuth, antimony and oxygen as shown in Fig. 2. The micrographs of EDS colorful maps shown that three different phases could be readily identified in the microstructure of the ZOV before and after SHCI aging. The aged ZOV clearly show different microstructures relatively to the unaged ZOV. A ZnO phase was the most abundant which agrees deeply with the fact that ZnO is the main component of ZOVs. Before SHCI aging the Bi, O and Sb elements were distributed in the vicinity of grain boundary of the ZOV as in Fig. 2a. After SHCI aging the EDS mapping analysis – in Fig. 2b and d – indicated a Bi, Sb and oxygen enrichment of ZnO grains, as consequence of current localization, and the associated joule heating of the ZOVs
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151
147
Fig. 1. Scanning electron micrographs of commercial ZOV before (A) and after SHCI aging test at 293 K (B), 333 K (C) and 353 K (D). The samples were aged by SHCI of 20 kA 8/20 s. The Sp represents the spinel grains (Zn7 Sb2 O12 ). * pore generated by SHCI.
Fig. 2. Elemental mapping by EDS of commercial ZOV samples before (A) and after SHCI aging at 293 K (B), 333 K (C) and 353 K (D). The samples were aged by SHCI of 20 kA 8/20 S.
148
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151 160 140
VHN (GPa)
120 100 80 60 40 20
ZnO
ZnO20
ZnO60
ZnO 80
Samples Fig. 3. Hardness versus temperature for ZOV before (ZnO) and after SHCI at operation temperatures of 293 K (ZnO20), 333 K (ZnO60) and 353 K (ZnO80).
at the interfaces. These findings appear to be consistent with the electro migration of ions within the depletion layer in the direction of current and oxygen desorption mechanisms [58,59]. However, the previous mechanisms appears to be accompanied by the Bi and Sb vaporization followed by crystallizing, on cooling, leaving with the formation of a monatomic layer of absorbed Bi and Sb at the ZnO grains [60]. To determine the element content an EDS analysis it was applied to measure quantitatively the chemical composition of ZOV before and after SHCI, whose results are shown in Fig. 2. As it can be clearly observed before SHCI aging, ZOVs have three different phases formed by a ZnO matrix doped with Co, the spinel (Zn7 Sb2 O12 ) and bismuth oxide (Bi2 O3 ). Fig. 2a shows a large concentration of bismuth detected from the EDS data at the ZOV grain boundary within the intergranular layers before the aging by SHCI, which explains the large rise in the Schottky barrier of this device [61–63]. At same time, the presence of localized oxygen at grain boundary observed in EDS spectra before SHCI aging at Fig. 2 may be justified by the enrichment of the grain boundaries with oxygen during sintering and their presence can determine the height of the potential barriers to a large extent in microstructure of ZOV [64]. Although ZOVs present a good resistance to degradation caused by continuous operating voltage their aging/degradation or destruction caused by SHCI remains a serious problem for electrical engineering [65–67]. As it was clearly shown by SEM and EDS analysis in Figs. 1 and 2, the microstructure of the ZOVs are greatly influenced by the sample operation temperature during the SHCI aging. According to these results more micro voids into the grains and some larger voids in the grain boundaries were formed with increasing of the operation temperature during SHCI aging. The presence of micro voids and larger voids contributes to the densification decreasing of ZOV microstructure owing to operation temperature during SHCI aging tests. Fig. 3 demonstrates the VHN plotted as a function of operation temperature during the SHCI aging of ZOV. The present measurements seem to indicate that the operation temperature of the samples of 333 K and 353 K appears to be an important factor at SHCI ageing of ZOV. An approximately monotonically linear decrease in VHN with increasing operation temperature was observed for ZOV. This means that the increase of operation temperature can substantially suppress micro crack of ZOV and
Fig. 4. Illustration of a hole structure in ZOV caused by the SHCI at 353 K.
consequently VHN is decreased. This result suggests a weakness of the coupling between the varistor grains with the increase of operation temperature and pores formation induced by the SHCI. Then, one can say that the mechanical resistance of ZOV becomes lower as result of pores discussed above. The discharge of a high-current pulse through the ZOV may lead to an electrical current filament formation with a high temperature gradient within the ceramic body and therefore inducing a higher thermal stress, causing the melting and evaporation of the dopants. The dopants volatilization induces the formation of larger porosity in the grain boundary and the vanish of the potential barriers [68,69]. It has been proposed that the raise in temperature T at the time t, caused by the Joule heating, is given by the following equation [70]: 1 T (t) = To+ mC v
t P(t)dt
(2)
o
where To is the uniform initial temperature of the ZOV (298 K),m ( = 5420 kg m−3 ) is the mass, Cv is the specific heat capacity (550 J kg−1 K−1 ) at a constant volume, and P is the electric power developed during a SHCI discharge. According to Eq. (2), a large average gradient temperature should be instantaneously (s) generated in ZOV by SHCI and, in case of current concentration on flow through preferential paths, the rapid Joule heating is accompanied by rapid fusion followed by vaporization and the formation of holes in the ceramics body. An example of a hole in the ZOV microstructure generated by the SHCI is shown in Fig. 4. The hole was approximately 136 m in diameter and there is evidence of once-molten flow on the ZOV surface. This observed hole structure in ZOV strongly indicates that a current filament had formed with temperatures high enough to melt the ZnO locally. The quantitative EDS analysis shown in Fig. 5 indicates that the operation temperature of the samples (333–353 K) appears to be an important variable to be considered in accelerated aging tests of ZOV by SHCI as large as 20 kA amplitude. It is clearly visible in Fig. 5(B) and (D) that the Bi2 O3 phase seems to have been lost during high-current impulses at 333 K and 353 K. The EDS analysis suggests that the SHCI behavior of the ZOVs is thermally activated. Therefore, for a high energy short time impulse, the amount of heat in the device can locally raise the temperature quickly. Therefore, the operation temperature increases as
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151
149
Fig. 5. EDS chemical mapping of commercial ZOV samples before (A) and after electrical aging by 20 kA SHCI at 293 K (B), 333 K (C) and 353 K (D). The red arrows indicate a possible volatilization of Bi2 O3 phase from varistor microstructure.
the direct solar exposure of the ZOV in tropical or desert countries should be sufficient to activate the thermal avalanche to start chemical changes in varistor device [71]. The temperature increase can be also result of a set of switching operations, during a storm, a common event in tropical regions. In spite of the large number of
investigations on the electrical properties of commercial ZOVs, as far as the authors are aware, the aging at high operational temperatures, due to, for example, excessive solar exposure or previous switching surge operations, appears not be sufficiently studied [71,72].
150
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151
Fig. 6. XRD patterns of ZOV commercial samples before (C) and after 20 kA SHCI at 293 K (A), 333 K (B) and 353 K (D). The crystalline phases identified in the ZOV microstructure are: ZnO (•), Bi2 O3 (), Sb2 O3 (), Zn7 Sb2 O12 (), Zn2 Bi3 Sb3 O14 (䊏), Co2 SbO4 ( ).
XRD data were taken from the standard ZOV (before aging) and the degraded samples by SHCI in air atmosphere and operation temperatures of 293 K, 333 K and 353 K. The main phases and their modification after the degradation process are evaluated in Fig. 6. The XRD pattern confirmed the presence of ZnO phases with hexagonal wurtzite structure and secondary phases. The secondary ZOV phases showed a complex microstructure consisting of several spinel-type crystallites at the grain boundaries like as Co2 SbO4 , ZnSb2 O4 and Bi3 SbO7 . The zinc-antimony spinel phase, Zn7 Sb2 O12 was also identified by XRD. The Bi2 O3 phases were not found after the ZOV aging by SHCIs at 333 K and 353 K. However, a pyrochlore phase (Zn2 Bi3 Sb3 O14 ) was observed in Fig. 6. This result suggests that the ZOV submitted to a SHCI stress leading to a rapid Joule heating is accompanied by a solid-state chemical reaction of Bi and Sb forming a pyrochlore phase in the ceramic body.
4. Conclusions The effect of operation temperature in the microstructure of ZOV after accelerated aging by SHCI was investigated in the present study. The SEM and EDS results indicate that the operation temperature seems to have a significant influence in the topology and chemical composition of ZOV, respectively. SEM analysis indicates a significant modification in the microstructure of the SHP aged ZOVs showing that more voids can be observed at the grains and grain boundaries. The elemental EDS mapping indicates that environmental temperature promotes the electro migration of oxygen out at grain boundary with segregation of Sb and Bi at ZnO grain surface during SHCI aging tests. XRD investigation suggests that the increase of operation temperature induces the formation of a pyrochlore phase during the accelerated SHP ZOV aging through a state-solid chemical reaction between spinel phase and Bi2 O3 present at the multiple ZnO grain junctions. However, more detailed characterizations using transmission electron microscopy (TEM) and ac impedance techniques are being conducted in order to achieve an understanding of all the phenomena involved during the state-solid chemical changes in
commercial ZOV during SHCI aging tests and the influence of environmental temperature.
References [1] Relva C. Buchanan, Ceramic Materials for Electronics, Marcel Dekker, New York (EUA), 2004, pp. 431–464. [2] A. Janotti, Chris G. Van de Walle, Fundamentals of zinc oxide as a semiconductor, Rep. Prog. Phys. 72 (2009) 126501 (29pp). [3] P.E.J. Flewitt, R.K. Wild, Grain Boundaries: Their Microstructure and Chemistry, John Wiley & Sons, New York (EUA), 2001. [4] S. Bernik, S. Macek, A. Bui, The characteristics of ZnO–Bi2 O3 -based varistor ceramics doped with Y2 O3 and varying amounts of Sb2 O3 , J. Eur. Ceram. Soc. 24 (2004) 1195–1198. [5] M. Peiteado, J.F. Fernández, A.C. Caballero, Varistors based in the ZnO–Bi2 O3 system: microstructure control and properties, J. Eur. Ceram. Soc. 27 (2007) 3867–3872. [6] Y. Abdollahi, A. Zakaria, R.S. Aziz, S.N. Tamili, K.A. Matori, N.M. Shahrani, N.M. Sideki, M. Dorraj, S. Moosavi, Optimizing Bi2 O3 and TiO2 to achieve the maximum non-linear electrical property of ZnO low voltage varistor, Chem Cent J 7 (2013) 137. [7] M. Houabes, R. Metz, Rare earth oxides effects on both the threshold voltage and energy absorption capability of ZnO varistors, Ceram. Int. 33 (2007) 1191–1197. [8] M. Dorraj, Y. Abdollahi, S.B.M. Said, M.F.B.M. Sabri, N.A. Sairi, W.P. Meng, E. Abouzari-lotfe, Enhance protection of electronic appliances through multivariate modelling and optimization of ceramic core materials in varistor devices, R. Soc. Chem. Adv. 5 (2015) 21384–21395. [9] R. Metz, H. Delalu, J.R. Vignalou, N. Achard, M. Elkhatib, Electrical properties of varistors in relation to their true bismuth composition after sintering, Mater. Chem. Phys. 63 (2000) 157–216. [10] P. Duran, F. Capel, J. Tartaj, C. Moure, A strategic two-stage low temperature thermal processing leading to fully dense and fine-grained doped-ZnO varistors, Adv. Mater. 14 (2002) 137–141. [11] F. Jiang, Z. Peng, Y. Zang, X. Fu, Progress on rare-earth doped ZnO-based varistor materials, J. Adv. Ceram. 2 (2013) 201–212. [12] M. Houabes, R. Metz, Rare earth oxides effects on both the threshold voltage and energy absorption capability of ZnO varistors, Ceram. Int. 33 (2007) 1191–1197. [13] N.T. Hung, N.D. Quang, S. Bernik, Electrical and microstructural characteristics of ZnO–Bi2 O3 -based varistors doped with rare-earth oxides, Int. J. Mater. Res. 16 (2001) 2817–2823. [14] C.-W. Nahm, C.-H. Park, H.-S. Yoon, Microstructure and varistor properties of ZnO–Pr6 O11 –CoO–Nd2 O3 based ceramics, J. Mater. Sci. Lett. 19 (2000) 271–274. [15] H. Liu, H. Kong, D. Jiang, W. Shi, X. Ma, H. Zhang, Effects of cooling rate on the microstructure and electrical properties of Dy2 O3 -doped ZnO-based varistor ceramics, Rare Met. 26 (2007) 39–44. [16] C.-W. Nahm, Microstructure, electrical properties, and dc aging characteristics of Tb4 O7 -doped ZnO-based varistors, J. Mater. Sci. 43 (2008) 2857–2864.
C. de Salles et al. / Electric Power Systems Research 134 (2016) 145–151 [17] J. Cai, Y.-H. Lin, M. Li, et al., Sintering temperature dependence of grain boundary resistivity in a rare-earth-doped ZnO varistor, J. Am. Ceram. Soc. 90 (2007) 291–294. [18] N. Daneu, S. Bernik, A. Reˇcnik, Inversion boundary induced grain growth in ZnO ceramics: from atomic-scale investigations to microstructural engineering, J. Phys. Conf. Ser. 326 (2011) 012003, http://dx.doi.org/10.1088/1742-6596/326/ 1/012003. [19] M. Peiteado, J.F. Fernández, A.C. Caballero, Processing strategies to control grain growth in ZnO based varistors, J. Eur. Ceram. Soc. 25 (12) (2005) 2999–3003. [20] A. Sawalha, M. Abu-Abdeen, A. Sedky, Electrical conductivity study in pure and doped ZnO ceramic system, Physica B 404 (2009) 1316–1320. [21] J.-L. He, J. Hu, Discussions on nonuniformity of energy absorption capabilities of ZnO varistors, IEEE Trans. Power Delivery 22 (2007) 1523–1532. [22] H.L. Mosbacker, Y.M. Strzhemechny, B.D. White, P.E. Smith, D.C. Look, D.C. Reynolds, C.W. Litton, L.J. Brillson, Role of near-surface states in ohmic-Schottky conversion of Au contacts to ZnO, Appl. Phys. Lett. 87 (2005) 012102-2–0121023. [23] M.A. Alim, S. Li, F. Liu, P. Cheng, Electrical barriers in the ZnO varistor grain boundaries, Phys. Status Solid 203 (2006) 410–427. [24] A.N.M. Karim, S. Begum, M.S.J. Hashmi, Performance and failure during energy testing of zinc oxide varistor processed from different powder size fraction and passivation thickness, Int. J. Mech. Mater. Eng. 5 (2010) 175–181. [25] C. Zhang, S. Zhang, The study on the pulse-degradation and destruction mechanisms of zinc oxide nonlinear varistors, J. High Voltage Eng. 26 (2000) 48–52. [26] Z. Zheng, H. Cui, Q. Wang, Y. Tu, Space charge characteristics of ZnO varistors after pulse degradation, IEEE Power Energy Eng. Conf. (2012) 1–4, http://dx. doi.org/10.1109/APPEEC.2012.6307387. [27] M. Bartkowiak, M.G. Comber, G.D. Mahan, Influence of nonuniformity of ZnO varistors on their energy absorption capability, IEEE Trans. Power Delivery 16 (2001) 591–598. [28] K.P. Mardira, T.K. Saha, R.A. Sutton, The Effects of Electrical Degradation on the Microstructure of Metal Oxide Varistor, IEEE, 2001, doi: IEEE 0-7803-72859/01. [29] M. Lengauer, D. RubesÏa, R. Danzer, Finite element modelling of the electrical impulse induced fracture of a high voltage varistor, J. Eur. Ceram. Soc. 20 (2000) 1017–1021. [30] K.J. Brown, Metal oxide varistor degradation, International Association of Electrical Inspectors (IAEI) News March–April (2004) 2–8. [31] K.L. Chrzan, Z. Wróblewski, Degradation and destruction of ZnO varistors caused by current pulses, in: International Conference on Advances in Processing, Testing and Application of Dielectric Materials, 2001. [32] M. Jaroszewski, J. Pospieszna, An assessment of ageing of oxide varistors exposed to pulse hazards using dielectric spectroscopy, in: International Conference on Solid Dielectrics, July 5–9, 2004, Toulouse, France, 2004. [33] C. Zhang, S. Zhang, The study on the pulse-degradation and destruction mechanisms of zinc oxide nonlinear varistors, J. High Voltage Eng. 26 (2000) 48–52. [34] A.S. Tonkoshkur, A.B. Glot, A.V. Ivanchenko, Percolation effects in dc degradation of ZnO varistors, J. Adv. Dielectr. 5 (2015) 1–8. [35] J. Liu, J. He, J. Hu, W. Long, F. Luo, Statistics on the AC ageing characteristics of single grain boundaries of ZnO varistor, Mater. Chem. Phys. 125 (2011) 9–11. [36] T. Tada, Degradation of ZnO varistors as estimated by aging tests, Electr. Eng. Japan 170 (2010) 661–674. [37] A.S. Tonkoshkur, A.Y. Lyashkov, I.V. Gomilko, A.V. Ivanchenko, Effect of longterm electrical degradation on the distribution of donor impurities in ZnO varistor ceramics, Inorg. Mater. 36 (2000) 745–748. [38] A.S. Tonkoshkur, I.V. Gomilko, A.Y. Lyashkov, Percolation effects in the capacitive properties of metal-oxide varistors in the range of high voltage, J. Adv. Dielectr. 4 (2014) 1–5. [39] V. Hinrichsen, Metal-Oxide Surge Arresters in High-Voltage Power Systems: Fundamentals, third edition, Siemens AG, Erlangen (DE), 2012. [40] R.C. Buchanan, Ceramic materials for electronics, Marcel Dekker, New York (EUA), 2004, pp. p441. [41] J. Ott, A. Lorenz, M. Harrer, E.A. Preissner, C. Hesse, A. Feltz, A.H. Whitehead, M. Schreiber, The influence of Bi2 O3 and Sb2 O3 on the electrical properties of ZnO-based varistors, J. Electroceram. 6 (2001) 135–146. [42] A. Sawalha, M. Abu-Abdeen, A. Sedky, Electrical conductivity study in pure and doped ZnO ceramic system, Physica B 404 (2009) 1316–1320. [43] K. Chrzan, J. Andino, Electrical strength of air containing by-products of intensive partial discharges, IEEE Trans. Dielectr. Electr. Insul. 8 (2001) 607–611. [44] K.L. Chrzan, Influence of moisture and partial discharges on the degradation of high-voltage surge arresters, Eur. Trans. Electr. Power 14 (2004) 175–184. [45] L. Kaczmarek, G. Lentka, Problems of varistor quality assessment during exploitation, Metrol. Meas. Syst. XIX (2012) 395–404.
151
[46] M.-H. Wang, K.A. Hu, B.Y. Zhao, N.F. Zhang, Degradation phenomena due to humidity in low voltage ZnO varistors, Ceram. Int. 33 (2007) 151–154. [47] W.P. Chen, H.L.W. Chan, Water-induced degradation in Zinc oxide-based ceramic varistors, J. Am. Ceram. Soc. 85 (2003) 1625–1627. [48] W.P. Chen, H.L.W. Chan, Degradation of ZnO ceramic varistors induced by water and AC voltages, Japan J. Appl. Phys. 43 (2004) 701–702. [49] C. de Salles, A.F. Picanc¸o, M.L.B. Martinez, H.R.P.M. de Oliveira, Determination of the Discharge Current on Distribution Network Surge Arresters, in: Powertech 2009, 28 June–2 July, 2009, Bucharest, Romania, 2009. [50] C. de Salles, M.L.B. Martinez, A.A.A. Queiroz, Surge Ageing of Metal Oxide Varistors, in: 2015 Electrical Insulation Conference (EIC), 7–10 June 2015, Seattle, Washington, USA, 2015. [51] F. Izumi, T. Ikeda, A Rietveld-analysis program RIETAN-98 and its applications to zeolites, Mater. Sci. Forum 321–324 (2000) 198–203. [52] J. Luan, G. Liu, H. Wang, H. Ullah, On the sampling of three-dimensional polycrystalline microstructures for distribution determination, J. Microsc. 244 (2011) 214–222. [53] L. Madler, S.E. Pratsinis, Bismuth oxide nanoparticles by flame spray pyrolysis, J. Am. Ceram. Soc. 85 (7) (2002) 1713–1718. [54] S.C. Pillai, J.M. Kelly, D.E. McCormack, R. Rameshd, High performance ZnO varistors prepared from nanocrystalline precursors for miniaturised electronic devices, J. Mater. Chem. 18 (2008) 3926–3932. [55] D. Xu, L. Shi, Z. Wu, Q. Zhong, X. Wu, Microstructure and electrical of ZnO–Bi2 O3 based varistor ceramics by different sintering processes, J. Eur. Ceram. Soc. 29 (2009) 1789–1794. [56] M. Peiteado, J.F. Fernández, A.C. Caballero, Varistors based in the ZnO–Bi2 O3 system: microstructure control and properties, J. Eur. Ceram. Soc. 27 (2007) 3867–3872. [57] Choon-W. Nahm, Effect of sintering temperature on varistor properties and aging characteristics of ZnO–V2 O5 –MnO2 ceramics, Ceram Int 35 (2009) 2679–2685. [58] A.V. Ivanchenko, A.S. Tonkoshkur, Modeling of the degradation electromigrational process in structures with intercrystallite potential barriers, Multi Model Mater. Struct. 3 (2007) 477–490. [59] J. He, J. Liu, J. Hu, R. Zeng, W. Long, Non-uniform ageing behavior of individual grain boundaries in ZnO varistor ceramics, J. Eur. Ceram. Soc. 31 (2011) 1451–1456. [60] W. Onreabroy, N. Sirikulrat, C. Hammond, S.J. Milne, Properties and intergranular phase analysis of a ZnO–CoO–Bi2 O3 varistor, Solid State Ionics 177 (2006) 411–420. [61] A. Janotti, C.G. Van de Walle, Fundamentals of zinc oxide as a semiconductor, Rep. Prog. Phys. 72 (2009) 1–29. [62] D. Fernández-Heviaa, M. Peiteado, J. de Frutosa, A.C. Caballero, J.F. Fernández, Wide range dielectric spectroscopy of ZnO-based varistors as a function of sintering time, J. Eur. Ceram. Soc. 24 (2004) 1205–1208. [63] D. Fernández-Hevia, J. de Frutos, A.C. Caballero, J.F. Fernández, Mott–Schottky behavior of strongly pinned double Schottky barriers and characterization of ceramic varistors, J. Appl. Phys. 92 (2002) 2890–2898. [64] C.F. Klingshirn, A. Waag, A. Hoffmann, J. Geurts, Zinc Oxide: From Fundamental Properties Towards Novel Applications, Springer-Verlag, Berlim (De), 2010, pp. p114. [65] J. Liu, J.L. He, J. Hu, W.C. Long, R. Zeng, Statistical pulse degradation characteristics of grain boundaries in a ZnO varistor based on microcontact measurement, J. Am. Ceram. Soc. 93 (2010) 2473–2475. [66] G. Yin, J. Li, G. Yao, S. Li, A new view on impulse degradation of ZnO-based ceramics, in: Proceedings of the Ninth International Conference on Properties and Applications of Dielectric Materials, July 19–23 2009, Harbin, China, 2009, pp. 1137–1140. [67] H. Andoh, S. Nishiwaki, H. Suzuki, S. Boggs, J. Kuang, Failure mechanisms and recent improvements in ZnO arrester elements, IEEE Electr. Insul. Mag. 16 (2000) 25–31. [68] K. Bavelis, E. Gjonaj, T. Weiland, Modeling of electrical transport in zinc oxide varistors, Adv. Radio Sci. 12 (2014) 29–34. [69] T. Tada, Degradation of ZnO varistors as estimated by aging tests, Electr. Eng. Jpn. 170 (2010) 1–18. [70] M. Lengauer, D. RubesÏa, R. Danzer, Finite element modelling of the electrical impulse induced fracture of a high voltage varistor, J. Eur. Ceram. Soc. 20 (2000) 1017–1021. [71] F. Munteanu, F.M. Frigura-Iliasa, E. Cazacu, About establishing the functional limits of a ZnO varistor based surge-arrester, Revue Roumaine des Sciences Techniques - Serie Électrotechnique et Énergétique 52 (2007) 443–452. [72] A. Gubaski, W. Mielcarek, K. Prociów, J. Warycha, J.M. Wróbel, The effect of aluminium additive on the electrical properties of ZnO varistors, Mater SciPoland 27 (4/2) (2009) 1207–1218.