• Korean Journal of Materials Research
• /
• v.5 no.6
• /
• pp.650-656
• /
• 1995

High frequency SAW filters for cellular communications were fabricated by metallizing 36$^{\circ}$Y-X LiTaO$_3$piezoelectric substrate with IIDT type electrodes. It was found that the center frequency of the filter was lowered than as designed. In order to overcome such a drawback and enable a fine tuning of its center frequency, dielectric SiN$_{x}$ thin films were deposited on LiTaO$_3$substrate by PECVD as passivation layer and then frequency responses were also characterized. As a result, the center frequency of the filter could be shifted to a higher frequency with increasing the thickness of SiN$_{x}$ film, because SAW velocity increased with increasing the ratio of the thickness of dielectric thin film to wavelength. The insertion loss of the filter, however, became larger with increasing the thickness of SiN$_{x}$ film.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2010.02a
• /
• pp.356-356
• /
• 2010

Tribological behaviors of the hard film on soft substrate system were explored using the hard thin film of diamond-like carbon (DLC) coated the soft polymer of polydimethysiloxane (PDMS). A DLC film with the Young's modulus of 100 GPa was coated on PDMS substrate with Young's modulus of 10 MPa using plasma enhanced chemical vapor deposition (PECVD) technique. The deposition time was varied from 10 sec to 10 min, resulting in nanoscale roughness of wrinkle patterns with the thickness of 20 nm to 510 nm, respectively, at a bias voltage of $400\;V_b$, working pressure 10 mTorr. Nanoscale wrinkle patterns with 20-100 nm in width and 10-30 nm height were formed on DLC coating due to the residual stress in compression and difference in Young's modulus. Nanoscale roughness effect on tribological behaviors was observed by performing a tribo-experiment using the ball-on-disk type tribometer with a steel ball of 6 mm in diameter at the sliding speed of 220 rpm, normal load of 1N and 25% humidity at ambient temperature of $25^{\circ}C$. Friction force were measured with respect to thickness change of coated DLC thin film on PDMS. It was found that with increases the thickness of DLC coating on PDMS, the coefficient of friction decreased by comparison to that of the uncoated PDMS. The wear tracks before and after tribo-test were analyzed using SEM and AFM.

• Korean Journal of Materials Research
• /
• v.21 no.12
• /
• pp.666-672
• /
• 2011

The a-Si:H/c-Si hetero-junction (HJ) solar cells have a variety of advantages in efficiency and fabrication processes. It has already demonstrated about 23% in R&D scale and more than 20% in commercial production. In order to further reduce the fabrication cost of HJ solar cells, fabrication processes should be simplified more than conventional methods which accompany separate processes of front and rear sides of the cells. In this study, we propose a simultaneous deposition of intrinsic thin a-Si:H layers on both sides of a wafer by dual hot wire CVD (HWVCD). In this system, wafers are located between tantalum wires, and a-Si:H layers are simultaneously deposited on both sides of the wafer. By using this scheme, we can reduce the process steps and time and improve the efficiency of HJ solar cells by removing surface contamination of the wafers. We achieved about 16% efficiency in HJ solar cells incorporating intrinsic a-Si:H buffers by dual HWCVD and p/n layers by PECVD.

• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.16 no.3
• /
• pp.116-120
• /
• 2006

We have investigated plasma spray coated $Al_2O_3$ layers on Al-60 series substrates for development of wafer electrostatic chuck in semiconductor dry etching system. Samples were prepared without/with cooling bar on backside of samples, at various distances, and with different powder feed rates. There were many cracks and pores in the $Al_2O_3$ layers coated on Al-60 series substrates without cooling bar on the backside of samples. But the cracks and pores were almost disappeared in the $Al_2O_3$ layers on Al-60 series substrates coated with cooling bar on the back side of samples, 15 g/min. powder feed rate and various 60, 70, 80 mm working distances. Then the surface morphology was not changed with various working distances of 60, 70, 80 mm. When the powder feed rate was changed from 15 g/min to 20 g/min, the crack did not appear, but few pores appeared. Also the $Al_2O_3$ layer was coated with many small splats compared with $Al_2O_3$ layer coated with 15 g/min powder feed rate. The deposited rate of $Al_2O_3$ layer was higher when the process was done without cooling bar on the back side of sample than that with cooling bar on the back side of sample.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2011.08a
• /
• pp.310-311
• /
• 2011

Dual-frequency (DF) capacitively coupled plasmas (CCP) are used to separately control the mean ion energy and flux at the electrodes [1]. This separate control in capacitively coupled radio frequency discharges is one of the most important issues for various applications of plasma processing. For instance, in the Plasma Enhanced Chemical Vapor Deposition processes such as used for solar cell manufacturing, this separate control is most relevant. It principally allows to increase the ion flux for high deposition rates, while the mean ion energy is kept constant at low values to prevent highly energetic ion bombardment of the substrate to avoid unwanted damage of the surface structure. DF CCP can be analyzed in a fashion similar to single-frequency (SF) driven with effective parameters [2]. It means that DF CCP can be converted into SF CCP with effective parameters such as effective frequency and effective current density. In this study, comparison of DF CCP and its converted effective SF CCP is carried out through particle-in-cell/Monte Carlo (PIC-MCC) simulations. The PIC-MCC simulation shows that DF CCP and its converted effective SF CCP have almost the same plasma characteristics. In DF CCP, the negative resistance arises from the competition of the effective current and the effective frequency [2]. As the high-frequency current increases, the square of the effective frequency increases more than the effective current does. As a result, the effective voltage decreases with the effective current and it leads to an increase of the ion flux and a decrease of the mean ion energy. Because of that, the negative resistance regime can be called the preferable regime for solar cell manufacturing. In this preferable regime, comparison of DF (13.56+100 or 200 MHz) CCP and SF (60 MHz) CCP with the same effective current density is carried out. At the lower effective current density (or at the lower plasma density), the mean ion energy of SF CCP is lower than that of DF CCP. At the higher effective current density (or at the higher plasma density), however, the mean ion energy is lower than that of SF CCP. In this case, using DF CCP is better than SF CCP for solar cell manufacturing processes.

• Korean Journal of Materials Research
• /
• v.9 no.12
• /
• pp.1155-1159
• /
• 1999

Thin film solar cells on a glass/$SnO_2$ substrate with p-SiC/i-Si/n-Si heterojunction structures were fabricated using a plasma-enhanced chemical-vapor deposition system. The photovoltaic properties of the solar cells were examined with varying the gas phase composition, x=$CH_4/\;(SiH_4+CH_4)$, during the deposition of the p-SiC layer. In the range of x=0~0.4, the efficiency of solar cell increased because of the increased band gap of the p-SiC window layer. Further increase in the gas phase composition, however, led to a decrease in the cell efficiency probably due to in the increased composition mismatch at the p-SiC/i-Si layers. As a result, the efficiency of a glass/$SnO_2$/p-SiC/i-SiC/i-Si/n-Si/Ag thin film solar cell with $1cm^2$ area was 8.6% ($V_{oc}$=0.85V, $J_{sc}$=16.42mA/$cm^2$, FF=0.615) under 100mW/$cm^2$ light intensity.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.34 no.4
• /
• pp.251-255
• /
• 2021

a-Si is commonly considered as a primary candidate for the formation of passivation layer in heterojunction (HIT) solar cells. However, there are some problems when using this material such as significant losses due to recombination and parasitic absorption. To reduce these problems, a wide bandgap material is needed. A wide bandgap has a positive influence on effective transmittance, reduction of the parasitic absorption, and prevention of unnecessary epitaxial growth. In this paper, the adoption of a-SiO_x:H as the intrinsic layer was discussed. To increase lifetime and conductivity, oxygen concentration control is crucial because it is correlated with the thickness, bonding defect, interface density (D_it), and band offset. A thick oxygen-rich layer causes the lifetime and the implied open-circuit voltage to drop. Furthermore the thicker the layer gets, the more free hydrogen atoms are etched in thin films, which worsens the passivation quality and the efficiency of solar cells. Previous studies revealed that the lifetime and the implied voltage decreased when the a-SiOx thickness went beyond around 9 nm. In addition to this, oxygen acted as a defect in the intrinsic layer. The D_it increased up to an oxygen rate on the order of 8%. Beyond 8%, the D_it was constant. By controlling the oxygen concentration properly and achieving a thin layer, high-efficiency HIT solar cells can be fabricated.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.08a
• /
• pp.154-155
• /
• 2012

The promise of nano-crystalites (nc) as a technological material, for applications including display backplane, and solar cells, may ultimately depend on tailoring their behavior through doping and crystallinity. Impurities can strongly modify electronic and optical properties of bulk and nc semiconductors. Highly doped dopant also effect structural properties (both grain size, crystal fraction) of nc-Si thin film. As discussed in several literatures, P atoms or radicals have the tendency to reside on the surface of nc. The P-radical segregation on the nano-grain surfaces that called self-purification may reduce the possibility of new nucleation because of the five-coordination of P. In addition, the P doping levels of ${\sim}2{\times}10^{21}\;at/cm^3$ is the solubility limitation of P in Si; the solubility of nc thin film should be smaller. Therefore, the non-activated P tends to segregate on the grain boundaries and the surface of nc. These mechanisms could prevent new nucleation on the existing grain surface. Therefore, most researches shown that highly doped nc-thin film by using conventional PECVD deposition system tended to have low crystallinity, where the formation energy of nucleation should be higher than the nc surface in the intrinsic materials. If the deposition technology that can make highly doped and simultaneously highly crystallized nc at low temperature, it can lead processes of next generation flexible devices. Recently, we are developing a novel CVD technology with a neutral particle beam (NPB) source, named as neutral beam assisted CVD (NBaCVD), which controls the energy of incident neutral particles in the range of 1~300eV in order to enhance the atomic activation and crystalline of thin films at low temperatures. During the formation of the nc-/pm-Si thin films by the NBaCVD with various process conditions, NPB energy directly controlled by the reflector bias and effectively increased crystal fraction (~80%) by uniformly distributed nc grains with 3~10 nm size. In the case of phosphorous doped Si thin films, the doping efficiency also increased as increasing the reflector bias (i.e. increasing NPB energy). At 330V of reflector bias, activation energy of the doped nc-Si thin film reduced as low as 0.001 eV. This means dopants are fully occupied as substitutional site, even though the Si thin film has nano-sized grain structure. And activated dopant concentration is recorded as high as up to 1020 #/$cm^3$ at very low process temperature (＜ $80^{\circ}C$) process without any post annealing. Theoretical solubility for the higher dopant concentration in Si thin film for order of 1020 #/$cm^3$ can be done only high temperature process or post annealing over $650^{\circ}C$. In general, as decreasing the grain size, the dopant binding energy increases as ratio of 1 of diameter of grain and the dopant hardly be activated. The highly doped nc-Si thin film by low-temperature NBaCVD process had smaller average grain size under 10 nm (measured by GIWAXS, GISAXS and TEM analysis), but achieved very higher activation of phosphorous dopant; NB energy sufficiently transports its energy to doping and crystallization even though without supplying additional thermal energy. TEM image shows that incubation layer does not formed between nc-Si film and SiO2 under later and highly crystallized nc-Si film is constructed with uniformly distributed nano-grains in polymorphous tissues. The nucleation should be start at the first layer on the SiO2 later, but it hardly growth to be cone-shaped micro-size grains. The nc-grain evenly embedded pm-Si thin film can be formatted by competition of the nucleation and the crystal growing, which depend on the NPB energies. In the evaluation of the light soaking degradation of photoconductivity, while conventional intrinsic and n-type doped a-Si thin films appeared typical degradation of photoconductivity, all of the nc-Si thin films processed by the NBaCVD show only a few % of degradation of it. From FTIR and RAMAN spectra, the energetic hydrogen NB atoms passivate nano-grain boundaries during the NBaCVD process because of the high diffusivity and chemical potential of hydrogen atoms.