State Key Laboratory: The Effect of Barium Titanate Powder Particle Size on MLCC Performance

Preface: As new electronic components continue to develop toward chip-scale, miniaturization, high-frequency, broadband, high-precision, integration, and environmental friendliness, MLCC products—as one of these components—are also moving toward miniaturization, high volumetric efficiency, high temperature resistance, and high reliability. To achieve these performance requirements, one of the key areas for research is barium titanate (BaTiO₃) material. Barium titanate dielectric materials possess excellent dielectric properties, including high dielectric constant, low dielectric loss, and good dielectric tunability. By incorporating trace amounts of modifying compounds, the material's dielectric constant and Curie temperature can be adjusted over a wide range. Furthermore, by controlling the particle size of ultrafine barium titanate powder, ultra-thin ceramic dielectric layers for capacitors can be produced. This paper focuses on studying the influence of barium titanate particle size on the performance of MLCC products based on the same proportion of modifying additives.

BaTiO₃ powders of different particle sizes were prepared using the hydrothermal method (purity > 99.9%, Ba to Ti molar ratio 0.998–1.000, unit cell parameter c/a > 1.002). The powders were each mixed with modifying compounds at the same ratio to obtain ceramic powders with different particle sizes. The composition of the ceramic powders with different BaTiO₃ particle sizes is shown in Table 1.

SEM images of BaTiO₃ powders with particle sizes of 200 nm and 400 nm are shown in Figure 1.

For each particle size group listed in Table 1, 5 kg of ceramic powder was mixed with organic solvent (toluene:anhydrous ethanol = 1:1), binder (PVB resin:ceramic powder = 7:100), and other modifying oxides. The mixture was dispersed at high speed in a bead mill to form a ceramic slurry. An 8 μm-thick dielectric film was formed using an ultra-flat, high-precision tape-casting machine. Inner electrodes were printed on the dielectric film using nickel electrode paste. Two hundred and fifty layers of dielectric film were alternately stacked using a laminator, then densified by isostatic pressing and cut into ceramic green chips. The green chips were heated to 450 °C in a nitrogen atmosphere, debindered for 40 hours, and then sintered in a bell furnace to form ceramic chips. After chamfering, polishing, termination, termination firing, and electroplating, MLCC samples of specification 1210 (3.2 mm × 2.5 mm × 2.5 mm) were produced with a nominal capacitance of 4.7 μF and a rated voltage of 100 V.

MLCC products were fabricated according to the requirements of each group in Table 1. Since the particle sizes differ, the sintering temperatures required to form the ceramics also differ to some extent, while the other processes are essentially the same. Generally, the smaller the powder particle size, the higher the surface activity, the easier the sintering process, and the lower the sintering temperature. Using the same modifying additives (dopants), appropriate sintering temperatures were determined for BaTiO₃ powders of different particle sizes in Table 1 to ensure the ceramics were dense and the grain growth in the ceramics was uniform (as shown in Figure 2).

From the curves in Figure 3, it can be seen that both the dielectric constant and dielectric loss of the product increase as the BaTiO₃ particle size increases. During the hydrothermal synthesis of BaTiO₃ powder, grain growth entails a transition from the cubic phase to the tetragonal phase, and the tetragonal phase content increases with increasing ceramic grain size. Since the tetragonal phase has a higher dielectric constant, MLCCs exhibit a higher dielectric constant when the powder particle size is larger.

On the other hand, as the grain size decreases, the proportion of grain boundaries (which have a low dielectric constant) per unit volume increases significantly, while the proportion of grain cores (which have a high dielectric constant) decreases. Additionally, BaTiO₃ powders with smaller grain sizes have a larger specific surface area, allowing more thorough and uniform contact with the modifying agent. After sintering, the penetration of the modifying agent further increases the proportion of grain boundaries. The increased quantity of grain boundaries with low dielectric constants has a "dilution" effect on the dielectric performance of the product.

In summary, in the particle size range of 200 nm to 500 nm, the smaller the BaTiO₃ powder particle size, the lower the dielectric constant of the resulting MLCC product, and accordingly the lower the dielectric loss.

The breakdown voltage of the products was tested at a voltage ramp rate of 200 V/s; the results are shown in Figure 4.

The insulation resistance was measured under the rated voltage; the results are shown in Figure 5.

As the grain size decreases, both insulation resistance and breakdown voltage increase. To prevent oxidation of the nickel inner electrodes, a reducing atmosphere containing H₂ is required during product sintering. The H₂ concentration is one of the most critical factors affecting the product's insulation performance. Since all four groups of products (Table 1) were sintered under the same atmosphere, their insulation resistance values are within the same order of magnitude. However, as previously described, the proportion of grain boundaries in the dielectric layer increases as the grain size used decreases. The high insulation characteristics of these grain boundaries give products made with smaller grains better insulation properties and voltage resistance. Consequently, the four groups in Table 1 still exhibited significant differences in insulation resistance.

Figure 6 shows the temperature-dependent capacitance variation curves for MLCCs prepared with BaTiO₃ of different particle sizes (Table 1).

It can be observed that the smaller the grain size, the flatter the capacitance-change-versus-temperature curve of the product. It is generally believed that due to the presence of modifying agents, the grains in the dielectric layer of the sintered product exist as a "core–shell" structure. BaTiO₃ with a core–shell structure exhibits a flat dielectric-temperature curve. Studies indicate that the high-temperature dielectric constant of core–shell-structured BaTiO₃ is determined by the volume fraction of the grain cores, while the intensity of the low-temperature dielectric peak is determined by the volume fraction of the grain shells. The initial BaTiO₃ powder particle size affects the volume fraction of the grain shells. In the range of 200–500 nm, the smaller the grain size, the larger the volume fraction of the grain shells and the smaller the volume fraction of the grain cores, resulting in a smaller rate of capacitance change at both low and high temperatures and thus better temperature characteristics.

Since the actual service life of MLCCs is relatively long, the lifespan can be predicted by accelerating the test using voltage and temperature factors. Using parameters such as temperature and voltage determined in the experiments, the Arrhenius equation can be applied to estimate the service life of the product under market application conditions:

Where:

( L_X ) = estimated service life under market application

( L_H ) = censoring time of the accelerated test

( V_X ) = applied voltage under market conditions

( V_H ) = applied voltage during the accelerated test

( T_X ) = application temperature under market conditions

( T_H ) = temperature during the accelerated test

( K ) = Boltzmann constant

( E_a ) = activation energy

( n ) = voltage acceleration factor

Based on experience, the ( E_a ) for MLCCs generally lies between 1.0 and 1.5, and ( n ) generally lies between 3 and 5. In this experiment, ( E_a = 1.2,\text{eV} ) and ( n = 3.5 ) are considered approximately correct.

Practical and mathematical theory indicates that the failure distribution of MLCCs can be approximately described by the Weibull distribution. Figure 7 shows the Weibull distribution fitting curves for the accelerated life tests of the four sample groups.

The censoring time of the accelerated test for each sample group was obtained through calculations based on the fitted data, thereby allowing the actual service life of the samples to be estimated, as shown in Table 2.

As the BaTiO₃ grain size used decreased, the actual service life of the product increased significantly.

The particle size of BaTiO₃ powder has a decisive influence on the performance of MLCC products. As the powder particle size used decreases, the dielectric constant of the product decreases, and the dielectric loss also decreases accordingly. Products made with smaller particle sizes exhibit better insulation and voltage resistance characteristics, and their temperature characteristics also show certain improvements. Notably, the BaTiO₃ powder particle size has a significant effect on the service life of the product: products made with smaller BaTiO₃ particle sizes exhibit a considerably extended service life.

Therefore, in the particle size range of 200–500 nm, using BaTiO₃ powder with a smaller particle size can significantly enhance the electrical performance and reliability of MLCC products.

Source: Electronic Process Technology, September 2020, Volume 41, Issue 5

Authors: An Kerong, Huang Changrong, Chen Weijian