Abstract
BaTiO3 powders doped with Ag at different Ag/Ba molar ratios were prepared by sol-gel method. The resistivity reached the lowest point of 5.644 Ω·m when Ag concentration was 0.10 at% and the powders were calcined for two times at 800°C and 500°C. XRD and FTIR investigations showed that no new substance was formed after the doping and calcining process, but the particle size and the strength of Ti-O bond in modified BaTiO3 crystal cell all changed. The conductivity of Ag-doped BaTiO3 powders with different Ag concentrations and through different preparing methods was discussed by using defect theory.
1. Introduction
Conductive powders are promising materials applied as film conductors [1–3]. It is also used as conductive fillers in polymers to make it electrically conductive, resistant to static electricity and to screen electromagnetic wave [4]. There are many series of conductive powders such as carbon, metals, and ceramics [5].
BaTiO3 with a perovskite structure is noteworthy for its exceptional dielectric, piezoelectric, electrostrictive, and electrooptic properties with corresponding electronic applications [6–8]. Various studies have been performed to obtain BaTiO3 of better electronic properties by doping different elements [9–12]. In order to obtain BaTiO3 of low resistivity, Zhao [13] doped Ag+ into BaTiO3 and conductive ceramics of Ω·m were prepared. Wu et al. [14] studied the influences on conductivity of doping different rare-earth elements into BaTiO3. Hao prepared BaTiO3 ceramics doped with Nb2O3 [15], Sm2O3 [16], Gd2O3 [17], and La3+ [18] (by gaseous penetration method) and achieved resistivity of = 2.364 × 108 Ω·m, 2.1 × 102 Ω·m, 1.27 × 105 Ω·m, and 8.2 × 104 Ω·m, respectively. Because the raw materials for synthesizing BaTiO3 are relatively cheap, to make the powders semiconductive by doping different ions into BaTiO3 seems to be not only a feasible but also a promising way to satisfy the needs for the application of conductive powders.
In this paper, BaTiO3 powders doped with Ag were synthesized and low resistivity of such powders is reported. Their electrical and compositional characteristics are measured and studied.
2. Experiments
2.1. Preparation of BATO Powders
The Ag-doped BaTiO3 powders (BATO) were prepared by sol-gel method [15, 19]. Ba(Ac)2 and Ti(OC4H9)4 were used as the starting materials. Dopant was chosen to be Ag in the form of AgNO3. The pH of the solution was adjusted to be 3.5 by adding glacial acetic acid. Gels were prepared through standard sol-gel procedure. The gel powders were calcined at 800°C for 2 h and then were calcined at 500°C for 2.5 h for the second time (SCalcination). Here the SCalcination method is an effective way to decrease the resistivity of BATO powders.
2.2. Analysis
The resistivity of the BATO powders was determined by using a standard four-point method (Keithley’s SourceMeter, model 2400, America). Structural analysis was performed by X-ray diffraction analysis (Seifert Debye Flex 2002). Fourier Transform Infrared spectroscopic measurements were performed by using an IR spectrophotometer (Nicolet AVATAR 320, America) ranging from 450 to 4000 cm−1.
3. Results and Discussion
3.1. Electric Properties of BATO Powders
The electric resistivity of BATO powders at different Ag concentrations is shown in Table 1. The influence of the second calcination at 500°C after the calcination at 800°C (SCalcination) on the resistivity can be seen in Figure 1.
The resistivity of undoped BaTiO3 powder is 4.0 × 109 Ω·m and is not shown in Table 1 or Figure 1.
It can be seen that as Ag concentration increased, the resistivity at first decreased, followed by an increase and another decrease. Such flat S-shaped trend was complied by the resistivity of both of pre- and post-SCalcination. The difference between the two processes is that SCalcination decreased the resistivity of BATO powders from the Ag/Ba molar ratio of 0.0005 to 0.0015, while it increased the resistivity at 0.0020 and 0.0025. The resistivity after SCalcination decreased the most at the Ag concentration of 0.15 at%, from 1821.04 Ω·m to 7.18 Ω·m.
3.2. XRD Analysis of BATO Powders
Figure 2 gives the XRD patterns. It can be concluded that no new substance was formed during the doping or calcination process other than BaTiO3 and BaCO3. It is also found that doping Ag and SCalcination both contribute to minimize the second phase and obtain purer BATO samples. The peaks in XRD patterns slightly changed their positions after doping and after SCalcination. The strongest diffraction of the sample 1# (undoped BaTiO3) is at 31.44°, while in 2# at 31.46° and 3# at 31.52°. The shift of BaTiO3 peaks showed that Ag entered BaTiO3 crystal lattice, leading to the increase of the conductivity of the powders.
The particle size of BATO powders was calculated using Scherrer Formula and FWHM of (200) reflection observed via the X-ray data. The particle size of sample 1# is calculated to be 11.81 nm, while the particle sizes of sample 2# and 3# are 20.31 nm and 18.29 nm, respectively. In other words, we can see that the particle size was increased after doping but reduced after the SCalcination.
3.3. FTIR Spectra Analysis
The FTIR spectra of samples 1#, 2#, and 3# are shown in Figure 3.
The characteristic absorption at 3410 cm−1 is assigned to–OH stretching vibration, due to the water brought by KBr or absorbed on the powder surface. The characteristic absorption at 1440 cm−1 is assigned to the stretching vibrations of carboxylate. This is because there is a small amount of BaCO3 in the samples, which is in agreement with the XRD patterns. All three samples exhibit strong absorptions around 550 cm−1 and 450 cm−1, which can be assigned to the stretching and bending vibrations of the Ti-O bond in [TiO6]2− octahedron. It is also a characteristic absorption of BaTiO3. But the wave number of the strongest absorption around 550 cm−1 varied slightly for the three samples: 547.8 cm−1 for sample 1#, 551.8 cm−1 for sample 2#, and 565.7 cm−1 for 3#. Since the wave number increases when infrared light of higher frequency and thus stronger energy is absorbed, it can be concluded that the Ti-O bond was strengthened after the doping and further strengthened after SCalcination.
3.4. Discussions on the Resistivity of BATO
During the calcination process AgNO3 decomposed into Ag2O, which entered BaTiO3 lattice. Since Ag+ possesses only one positive charge and Ba2+ possesses two, the substitution must be charge-compensated to maintain charge neutrality. According to the defect theory established by Kroger and Vink forty years ago, the incorporation of Ag2O as an acceptor dopant can be written as (1) to (3):
Since the perovskite structure is relatively a close packed one, the formation of interstitial ions is not as easy as the formation of vacancy. So the incorporation of Ag2O would prefer the way presented in (1).
The formation of oxygen vacancy can be proved by the FTIR result, in which the wave number of the characteristic absorption of Ti-O bond changed among the three samples. As stated previously, the strength of Ti-O bond in the [TiO6]2− octahedron was stronger in the sample doped with Ag than the undoped BaTiO3 sample. It can be explained as that, since the formation of oxygen vacancies means that there is O2− leaving the cell, the [TiO6]2− octahedron was distorted and Ti4+ gained a stronger attraction of the O2− left, rendering the Ti-O bond stronger.
The oxygen vacancy formed was in equilibrium with the formation of hole:
So that is the ultimate reason why doping AgNO3 would render BaTiO3 semiconductive.
As is shown in Figure 1(a), increasing AgNO3 will decrease resistivity because the conductivity is proportional to the concentration of oxygen vacancy. However we should also be aware that a high concentration of defect would distort the crystal structure thus undermining conductivity. Such two factors together influence the conductivity and resulted in the trend of conductivity change before Ag concentration of 0.15 at% in Figure 1(a).
However the following drop of resistivity at Ag concentration of 0.20 at% and 0.25 at% may result from the formation of silver during the drop of temperature in the cooling process after calcination of the BATO powders. The silver would remain in the grain boundary and serve as a good conductor, reducing the grain boundary resistivity. Since the total resistivity of BaTiO3 is composed of grain resistivity and grain boundary resistivity, the overall resistivity is reduced. This explains the drop of resistivity after Ag concentration of 0.15 at% in Figure 1(a).
To explain the conductivity trend presented in Figure 1(b), we must first understand that not all the AgNO3 doped in BaTiO3 proceeded by (1); part of Ag+ followed the path in (2) and (3). It is the part of Ag that occupied Ba site that formed the oxygen vacancy and thus contributed to the conductivity. So the more Ag entering Ba sites, the lower the resistivity is. Since heating provides ions with more energy to satisfy defect formation energy, SCalcination facilitated more Ag to move from interstice to Ba site. This can be proved by the calculation result of grain size that the cell size is smaller after SCalcination than before, because interstitial Ag+ destabilized the crystal and increased its size.
However at higher Ag concentrations (0.20 at% and above) the resistivity is higher after SCalcination than before. This may be due to two reasons. Firstly, as discussed previously, when the concentration of defect mounts to certain level, it will damage the conductivity. So SCalcination would worsen such process. Secondly, SCalcination would oxidize the silver formed during the cooling process after the first calcination.
4. Conclusions
BATO conductive powders were prepared by sol-gel method. The lowest resistivity of ρ = 5.644 Ω·m was obtained when (Ag) : (Ba) = 0.0010 and after the powder was processed by calcination at 800°C for 2 h for the first time and 500°C for 2.5 h for the second time.
XRD and FTIR investigations and measurements of electric conductivity combined proved that Ag+ entered the lattice of BaTiO3 and substituted for Ba2+, forming oxygen vacancies, which are the main reasons for the semiconductivity.
Acknowledgment
This work was supported by the National Science Foundation in China (no. 20571020).