Patent Application
20060169969
The invention relates to bandgap cascade cold cathodes, their use and methods of making them.
Manipulating electrons in an evacuated enclosure for electron microscopes, X-ray tubes, traveling wave tubes, free electron lasers, etc. requires a bright, steady, and well defined e-beam source. Thermionic emissions, typically from a tungsten filament, could deliver up to 10 Amp/cm2, but will incur a reduced life when the emission increases much beyond one A/cm2. Cold cathodes, particularly of a planar structure, could theoretically deliver a much higher current density if the electron affinity or work function of the emitter could be greatly reduced. In this disclosure, we will make use of “hot electrons” generated by the cascade of a breakdown avalanche multiplication of a semiconductor material such as the hot electrons in a Schottky diode reaching the solid/vacuum interface, to effectively reduce the work function for enhanced surface emissions.
Schottky Diode Under Avalanche Multiplication As an Example
Among semiconductor device configurations, the simplest is the Schottky diode. It is composed of a slice of semiconductor with one smooth surface coated with a metal electrode serving as the Schottky surface and the other not so smooth surface also made conductive serving as the source or sink. Electrons can flow relatively freely with small resistance from the rough sink electrode to the smooth Schottky electrode. But when the Schottky electrode becomes negatively charged, the smooth uniform field drives the free electrons away from the surface electrode, forming a carrier void and therefore a depleted insulating region. The depletion thickness extends as the field increases, and could reach the sink electrode. More bias delivers a higher field over the material.
Under an applied field, some electrons from the Fermi sea will always tunnel through the tilted band edge to form a leakage current. When the applied field is increased to a sufficient level so that carriers under the field are accelerated within a mean-free-path to gain sufficient energy so that the scattering would create an electron-hole (e-h) pair, and this nascent e-h pair in turn could also accelerate within a mean-free-path to gain sufficient energy to scatter and create subsequent e-h pairs, a cascade of avalanching multiplication is thus commenced. Depending on the e-h scattering and annihilation efficiencies, or the ionization efficiency of the semiconductor material, the multiplication factor of the avalanching carriers for each mean-free-path distance can be as low as one—no effective avalanche over the material, to as high as two—the carriers would double in number over each and every mean-free-path field length. This avalanche will terminate at the sink electrode whose in-situ equi-potential surface would provide no acceleration beyond. But for carriers reaching the sink electrode with kinetic energy of the bandgap, or fractions of the bandgap, their energy can readily be utilized to balance against the work function of the material for cold cathode emissions.
Thus, during avalanche multiplication under reverse bias over an Si/C layer, the bandgap energy cascades from the Schottky electrode to the sink electrode and is used to balance against the work function of the sink electrode, thereby allowing the sink electrode to function as a cold cathode emitter at a reduced applied external field.
Bandgap Energy Cascade
Avalanching carriers gaining sufficient energy in the semiconductor within one mean-free-path can scatter and create electron-hole pairs. The newly created carriers are the “hot carriers”, with each hot carrier having at least the bandgap energy in order to maintain the process of avalanche multiplication. These hot electrons cascading to the sink electrode become hot surface electrons. The kinetic energy of these electrons in a wide bandgap material is typically ˜3 eV, which can be used to balance against the work function of the material at ˜4.5 eV for cold cathode functions. Therefore, a construction of the Schottky cold cathode under bandgap cascade can be outlined as follows:
For example, a bandgap cascade cold cathode is obtained by treating surfaces of an effectively thin silicon wafer substrate with HF acid to remove the surface oxides and other impurities while leaving at least an epitaxial surface thereof with dangling hydrogen atoms; depositing on the epitaxial surface an effectively thin layer of Si either by SiH4 under CVD (chemical vapor deposition) or silicon beam under ALD (atomic layer deposition) at 650° C. to obtain an epitaxial surface layer thereon for superlattice construction; constructing a few periods of Si/C at about one nm/period on the epitaxial surface layer to form an Si/C surface; depositing on the Si/C surface a first metal effective to form a metal-silicide electrode a few 10s of nm thick; etching away the silicon substrate to termination chemically when it reaches the Si/C superlattice boundary to form an effectively smooth Si/C surface thereon; coating the effectively smooth Si/C surface with an effectively thick second effective metal such as nickel to form a Schottky electrode surface on which a layer of about 200 nm Pt or Au is coated for chemical stability while masking off an edge of the Schottky surface from the metal coatings for edge-termination of the field as a leakage protection; welding the Pt or Au coating onto a Cu electrode disc having the same diameter as the Schottky electrode surface, the Cu electrode being sufficiently massive for a heat sink of thermal management; and placing a ring-shaped contact on the Si/C surface for delivering a positive pulse that initiates avalanche multiplication over the Si/C as a cathode emitter.
The applied field of the Schottky surface diode can be modulated at high frequencies, and for each pulse, a breakdown voltage is applied in order to commence an avalanche multiplication, to generate hot electrons cascading to the sink electrode. This sink electrode surface under the influence of a separate external field becomes the enhanced cathode emitter. While a Schottky diode is used for illustration, any p-n junction diode material could be applied with the same avalanche multiplication rationale under a breakdown bias.
Preferred embodiments will now be described with reference to an illustrative but not limitative drawing wherein:
Combining the Static and Dynamic Processes
Four recent static schemes of surface structuring and material selections described by the Fowler-Nordheim model can be summarized as seeking field emission with the lowest applied field. A dynamic model on the other hand, may be compared to the activities of a Faraday cup, where carriers would enter the cup, or as considered below travel to the vacuum, by passing through a gate potential barrier. This barrier usually consists of a chopper to be used for energy discrimination of the charged particles. All the static schemes of surface structuring and barrier lowering with space charges etc. can be considered as lowering the effective gate potential which increases the probability for the charged particles to passing through.
In our dynamic model, the kinetic energy of the electron cascade may be substantially more energetic than the gate potential or the work function of the material, therefore the emission will be limited mostly by the avalanching current and not by the detailed structures of the surface barrier as governed by the Fowler-Nordheim model. The static and dynamic energy profiles using Si/C superlattice are outlined in
Avalanche Cascade Under a Schottky Electrode
Schottky diodes are usually constructed with a heavily doped substrate with a nearly intrinsic epitaxial (epi)-layer coated with the Schottky electrode. The heavily doped substrate is necessary to reduce the material resistance to the forward current, while the near intrinsic epi-layer with low carrier concentration is necessary to sustain the reverse bias with a small leakage current. When the Schottky electrode is applied with a negative bias, the applied field pushes the electrons away from the electrode, causing a void of free carriers and thus a depletion region with high negative resistance. This negative bias from the Schottky electrode can be exerted equally well by a positive bias placed at the source electrode opposite to the Schottky electrode. As the bias increases, the depleted region extends, and as the depletion thickness reaches to the sink electrode, the entire Schottky layer becomes depleted.
Increasing the bias further, the level of the applied field over the dielectric becomes so high that the leakage carriers in the region become accelerated within a mean-free-path to scatter and create nascent electron-hole (e-h) pairs, which in turn, would initiate additional e-h pairs etc. and thus initiate the dynamics of avalanche cascade. The cascade current is limited by the number density of the e-h pairs as their polarized presence would neutralize the applied field in the dielectric material.
Cascade Carriers with Bandgap Energy
For our cold cathode application, the material responsible for forward resistance is of no concern because there is no current flow from the vacuum to the electrode. We need to focus on the energetic avalanching electrons to balance against the work function, implying that the larger the bandgap to create the e-h pairs, the higher the hot electron energies and the more useful for them to fly across the work function barrier. This combination of high current density with large bandgap energy is rather unique in our requirement of the material characteristics. Designs of Zener diode, for example, call for mainly a precise breakdown voltage, which is not sufficient to satisfy the functions of the proposed bandgap cascade emitter. The bandgap, the dielectric constant, the breakdown voltage, and the thermal conductivity of several well known potential materials are listed in Table 1 below:
From Table 1 above, cold cathode with a very high emitter current may be obtained by using large bandgap for highly energetic hot electrons with material having high carrier concentration. Insulators such as SiO2 have high bandgap but low carrier concentration, while silicon can have high carrier concentration but low bandgap value. We have some in-house expertise for Si/O and Si/C superlattice fabrications, and their bandgaps are 2.3-3.1 eV as measured by photo-luminescence. These bandgap values may be engineered to an even higher level by reducing the silicon fraction in the superlattice construction.
Outline of the Bandgap Cascade Cathode
Avalanching carriers gaining sufficient energy in the semiconductor within one mean-free-path to scatter and create e-h pairs, are the “hot carriers”, and when they arrive at the sink electrode, they can be the desired hot surface electrons. The kinetic energy of these electrons is typically ≧3 eV, which will be used to balance against the work function of the cold cathode emitter at about 4 eV, or effectively below 3 eV when certain potential lowering structures are incorporated. The applied field of the Schottky electrode can be modulated with high frequencies, and for each pulse, a breakdown voltage would be exerted to induce the avalanching cascade. This avalanche generates hot electrons all the way to the sink electrode. The sink electrode, under a separate applied field, becomes the enhanced cold cathode emitter. The desired characteristics of the Schottky bandgap cascade cold cathode are outlined as follows:
For a Schottky diode, the forward current can typically be orders of magnitude higher than the limit of thernionic emissions under a few A/cm2. The avalanching cascade driven by a reverse bias, or the breakdown current, is governed by the carrier concentration and the mobility of the material. Power dissipation during the avalanche, however, can be much higher than the forward current because of the high potential drop across the layer under the reverse bias. Consider for example, the silicon Schottky under a breakdown field of 0.4 MV/cm over a thickness of 25 μm. To drive the breakdown, it would require an applied voltage ˜1,000 Volts, and if an avalanching cascade of 10 A/cm2 is maintained during the breakdown, the thermal load of the depletion layer can be 10 kW/cm2, a thermal budget well beyond any semiconductor material to conduct and dissipate. The key here is therefore to reduce the semiconductor layer thickness from tens of μm to sub-μm, while allowing the cold cathode to be mounted on a massive electrode as heat sink.
The disclosed depleted Schottky layer at sub-μm is really a thin and uniform device constructed epitaxially with two terminals; one for thermal management as well as for the delivery of electrons, and the other as emitter pulsed by a sufficiently high applied voltage to drive the bandgap cascade. To construct the thin, uniform and precise semiconductor layer, we will use a silicon-based superlattice Si/C with high bandgap (˜3 eV), high thermal conductivity (˜4 W/° C.-cm) and high dielectric constant (˜9). The construction steps are outlined as follows:
For a Schottky diode, the forward current can typically be several hundred A/cm2 as compared to thermionic emissions of at most a few A/cm2. The avalanching current under a reverse bias, or the breakdown current is governed by the same scattering-limited carriers movements of the material. Power dissipation during the avalanche multiplication, however, can be much higher than the forward current because of the high reverse field. Consider for example, a silicon Schottky under a breakdown field of 3 Volt/mfp (mean-free-path in silicon is ˜3 nm), a thickness of 10 μm, during breakdown would require an applied voltage of up to 10,000 Volts. If an avalanching current of 100 A/cm2 is obtained during the breakdown, the power density experienced by the layer can be 106 Watt/cm2, a thermal budget much too high for most semiconductor materials to dissipate. A pulsed emission, with a duty cycle under 0.1% may therefore be necessary. In addition, a reduction of the semiconductor layer thickness for a reduced bias would reduce the power as well as enhance the heat dissipation. The layer thickness can be reduced from tens of μm to under 0.1 μm while allowing the cathode emitter to be constructed on a massive electrode as heat sink in order to maintain a robust thermal management.
The proposed bandgap cascade under avalanche multiplication in a Schottky diode at 0.1 μm is really an ultra-thin and ultra-uniform film constructed epitaxially with two terminals; one Schottky electrode for thermal management as well as for the delivery of electrons, and one sink electrode under a sufficiently high applied field to deliver the carriers under avalanche multiplication with cascading bandgap energy. The pulse is also sufficiently short to reduce the effective duty cycle for a limited thermal load.
To construct the ultra-thin, uniform and precise wide bandgap semiconductor structure, we will use a silicon-based superlattice Si/C whose construction steps are disclosed as follows:
The other wide bandgap thin film materials for the bandgap cascade design, GaN, AlN, or both, have wider bandgap than SiC now considered preferred in the disclosure.
Variations, combinations and permutations of the above described method and device invention as will occur to those of ordinary skill are contemplated as within the scope of the following claims.
| Number | Date | Country | |
|---|---|---|---|
| 60649216 | Feb 2005 | US |
