A semiconductor bridge (SCB) is a two-terminal silicon based electronic device which consists of a heavily doped polysilicon. Its application can be extended to the automobile airbag, actuator, and igniter ordnance system including missile etc. Among all the electro-explosive detonators (EED’s), SCB has been proven to be one of the most cost-effective, reproducible, and reliable ignitors. Above all things, in that SCB is relatively simple, small, and fully compatible with the microelectronics IC industry, it is obvious that SCB will be a promising initiator that should take the place of the conventional metal hot-wire.
In this work, the SCB chips of 1Ω, 2Ω with various bridge sizes were fabricated with the conventional CMOS process including thermal oxidation, LPCVD, dopant drive-in, and lithography. In drive-in process, POCL3 was diffused 4∼5 times at 1000℃ for 120min so that the bridge may have a desired resistance value. These chips were mounted on the TO5 ceramic package and wire-bonded so as to make electrical connection between Al pad and the outer lead. Capacitive discharge circuit was adopted to dump the stored electrical energies into SCB swiftly. At this mode, the applied voltage and the capacitance were chosen to be as 25V, 25μF, respectively, considering the RC-time and stored energy of the circuit. And oscilloscope and photo-diode were used to measure the electrical, optical characteristics of SCB from which the plasma generation mechanism could be explained. Theoretical energies needed in phase transition of poly-Si (melting, vaporization, and ionization) were calculated. By analyzing the voltage-time curve and comparing it with energies calculated above, the time consumed in phase transformation (melting, vaporization, and ionization) and the plasma duration time were estimated.
Post-firing analysis of the SCB loaded with explosives was performed by Bruceton test, a kind of statistical confidence test, which could provide us with All-fire voltage(AF) & No-fire current(NF) data. All-fire voltage is related with efficiency of SCB, and No-fire current with safety of SCB. Of these two important factors that affect the operation of SCB, no-fire current condition was studied in detail by comparing the theoretical values with the experimental results.
SCB chip was packaged in TO5 can. Au wire was often disconnected due to the weakness of the strength of the wire upon loading explosive powder in TO5 can installed with SCB. Therefore, a soldering was applied to the package of SCB as the replacement of the wire-bonding. A solder alloy was selected as Bi-Pb(58wt%-42wt%) considering melting point of the package, wettability on Al metal pad. In prior to soldering, precleaning of Al pad ,Ni-electroless plating and Au coating were performed in sequence. Through this experiment, economical and reliable packaging method of SCB was studied.
During the SCB operation , the voltage change across the bridge was measured by oscillioscope where the 2 obvious peaks were found to relate with the operation of SCB. While the n-type device showed an apparent 1st, 2nd peaks ,the p-type device showed a flat 1st peak where bridge melted. As the bridge length increases, the duration time(the time between 1st peak and 2nd peak) increased due to the decrease in electric field (E=V/d). As the filling pressure of explosives increases, the ignition time became shorter because the heat transferring rate transferred to the explosives become larger.
We have examined the impact of structure(wedge) and material(TiSix) on electrical characteristics of such as AF & NF. We found that wedge type and TiSix bridge affect the AF/NF due to local stress and higher heating energy of refractory metal, respectively. The simulation results(ABAQUS) with modified structure (wedge type) exhibit that the stress(70~100Mpa) on the wedge type bridge is larger than that(60~70Mpa) of the normal type due to tensile stress across the wedge type bridge. Due to this local stress, the electrical characteristics of wedge type are improved significantly.
From the bridge volume, the mass$(8.4\times10^{-9}g)$ and moles$(3.0\times10^{-10}moles)$ can be calculated. Theoretical energy needed to be liquified and vaporized are calculated as 0.0248mJ, 0.1052mJ ,respectively. The total energy consumed from solid to vapor becomes 0.13mJ. The power dissipation can be obtained by multiplying the applying current by induced volatage across the SCB. The total energy consumed is obtained by integrating the power dissipation with time. The 1st and 2nd peak in case of 20×90(LxW)μ㎡ device occurred at 250ns, 600ns, respectively. Therefore SCB is vaporized within 550ns and the plasma generated at the 2nd peak (600ns).
The position of photo-diode signal is equal to that of the 2nd peak which indicates the generation of the plasma from SCB. Photo-diode output voltage and FWHM decreased when the applied voltage decreased from 25V to 5V because of the increase in plasma intensity with increasing the electric field to SCB.
All-fire energy tends to increase with the bridge volume. Likewise, no-fire current shows that the current increases with the bridge size. In an effort to explain this phenomena, Hot spot model is adopted. The maximum or critical current which should not evaporate the bridge is proportional to the one fourth power of the bridge area.
