Abstract: Post-implant annealing of Al implanted 4H-SiC has been performed in the temperature
range from 1600°C to 1750°C. Annealing was conducted in a hot-wall CVD reactor using a silanerich
ambient. Ar was used as the carrier gas to deliver the silane to the annealing zone where the
sample was heated via RF induction. The resulting annealed surfaces exhibited a step-bunch free,
smooth morphology when viewed on SEM and AFM. The maximum surface roughness as
measured via AFM was 0.65 nm RMS for the sample annealed at 1750°C.
Abstract: It is investigated whether the homogeneous depth profiles of epitaxially doped B or Al
are changed or preserved by implantation of various implanted species and annealing processes. We
have found a strong decrease in the atomic B concentration in epitaxially B-doped layers after
implantation of N+, Al+, and P+ and subsequent annealing at 1700 °C. On the other hand, the Al
profiles in epitaxially Al-doped layers are preserved after the same processes.
Abstract: 4H-SiC samples implanted at 600°C with 1020 cm-3 of B or B and C to a
depth of ~0.5 μm, capped with (BN/AlN), and annealed at temperatures ranging from
1400°C – 1700°C were studied using variable temperature cathodoluminescence. New
emission lines, which may be associated with stacking faults, were observed in the
samples co-implanted with B and C, but not in the samples implanted only with B.
For both the B and B and C co-implanted samples, the intensity of the line near 3.0 eV
decreases with increasing annealing temperature, TA, and this line is not observed
after annealing at 1700°C. The D1 defect related emission lines are observed in the
luminescence spectra of all samples and their relative intensities seem to vary with the
implantation-annealing schedule and excitation conditions.
Abstract: A box like Ge distribution was formed by ion implantation at 600°C. The Ge concentration
was varied from 1 to 20 %. The TEM investigations revealed an increasing damage formation
with increasing implantation dose. No polytype inclusions were observed in the implanted regions.
A detailed analysis showed different types of lattice distortion identified as insertion stacking faults.
The lattice site location analysis by “atomic location by channelling enhanced microanalysis” revealed
that the implanted Ge is mainly located at interstitial positions.
Abstract: Hydrogen-exfoliation has become a viable approach to transfer SiC thin layers onto
different substrate materials. However, little attention has been paid to the exfoliation-inducing
annealing conditions. To investigate the annealing conditions, 4H SiC wafers were implanted with
either 2.5×1016 H2
+ cm-2 or 5.0×1016 cm-2at 37 KeV. Post-implant, multi-step annealing sequences
were examined in order to promote more efficient blistering, and it was found that a low
temperature initial annealing step (T ≈ 500°C) can decrease the annealing time necessary in the
high temperature regime; this was attributed to a nucleation of hydrogen induced platelet defects
during the low temperature annealing regime and efficient splitting during a higher temperature
(900 °C) anneal. This process is similar to what is observed for InP and Si exfoliation, except that
the annealing processes occur at higher temperature.
Abstract: While nickel ohmic contacts to n-type silicon carbide have good electrical properties, the
physical contact, and therefore the reliability, can be poor. An approach is described for using the
good electrical properties of Ni ohmic contacts while using another metal for its desired
mechanical, thermal and/or chemical properties. In the present work, once the Ni contacts have
been annealed forming nickel silicides and achieving low contact resistance, they are etched off.
Removing the primary Ni contacts also eliminates the poor morphology, voids, and at least some of
the excess carbon produced by the Ni/SiC reaction. The Ni contacts are then replaced by a second
contact metal. This second metal displays low contact resistance as-deposited, indicating that the
critical feature responsible for the ohmic contact has not been removed by the primary contact etch.
Not only does this approach provide more flexibility for optimizing the contact for a given
application, it also provides some insight into the ohmic contact formation mechanism.
Abstract: Ohmic contacts on SiC have been investigated extensively in the past decade. However,
the mechanism for ohmic contact formation has been a troublesome issue. The interfacial structures
at the atomic scale responsible for forming ohmic contacts have not been revealed. Our previous
results have shown that carbon can form ohmic contacts on SiC after thermal annealing, and that an
interfacial carbon layer between Ni and the SiC improves the contacts significantly. In this study,
we have investigated the interactions between Ni and carbon, and ohmic contact formation on SiC
using x-ray diffraction (XRD) and Raman spectroscopy. After annealing, ohmic behavior was
observed and Ni graphite intercalated compounds (GICs) were found on Ni/C/SiC structures.
Unlike conventional graphite intercalated compounds, the Ni atoms substitute for carbon atoms in
the graphitic networks in these Ni-GICs. XRD peaks at 21.6° due to the Ni graphitic intercalation
compound (Ni-GIC) and at 26.3° due to graphite have been observed. The distance between
graphitic sheets is 0.403nm in the Ni graphite intercalated compounds, whereas it is ~20% larger in
the graphite. The thickness of the interfacial carbon layer does not affect the formation of Ni-GIC.
Abstract: The characteristics of Ni, Monel (Ni-Cu alloy, Ni55mol%-Cu45mol%), Monel/Si,
Ni/Ti/Ni and Mo electrodes were studied for ohmic contact to C-face N-type 4H-SiC. Low contact
resistivity (ρC) was not compatible with reduction of graphite precipitation in the case of Ni, Monel,
Ni/Ti/Ni, and Mo electrodes. Monel/Si achieved less graphite precipitation and low ρC, which is
enough to apply for actual rectifier, because a Monel/Si electrode forms a silicide without reaction
between the deposits and the substrate.
Abstract: Structural properties of Ni/Ti films deposited on 4H-SiC and annealed at temperatures
from 800 to 1040°C have been studied. Films with three different metal deposition sequences were
investigated by X-ray diffraction and Auger electron spectroscopy: (A) Ti(100 nm)/Ni(50 nm); (B)
Ti(4 nm)/Ni(50 nm)/Ti(100 nm); and (C) Ti(4 nm)/Ni(150 nm). A distinct spatial separation of
nickel silicide and titanium carbide layers was observed in all samples. It was discovered that the
distribution of the products of the solid state chemical reaction in samples (A) and (B) was
independent on the deposition sequence of Ti and Ni layers. The titanium carbide layer located on
the interface and covered by the clearly separated nickel silicide layer was detected in both samples
after heat treatments.
Abstract: We report on the die bonding processes and how the surface roughness and metallization
schemes affect the processes of die bonding in 4H-SiC device fabrication using a soldering test and
die shear test (DST) with differently prepared 4H-SiC samples. The first set of samples (FZ#1 and
FZ#2) was capped with sequentially evaporated Ti and Au on an annealed Ni layer. The second set of
samples (FZ#3 and FZ#4) and the third set of samples (FZ#5 and FZ#6) were prepared by 4μm-thick
Au electroplating on an annealed Ni layer and an un-annealed Ni layer, respectively. The quality of
the soldering, such as the solder coverage, void, and adhesion, was characterized by optical
microscope, X-ray microprobe, and DST. We found that the samples (FZ#4 and FZ#6) deposited by
Au electroplating on C-face (bottom-side) 4H-SiC provided a satisfactory result for the tests of solder
coverage, void, and DST and also realized the cleaning process prior to the electroplating and
soldering was the most crucial in the die packaging processes of vertical structure devices. The void
fraction measured by X-ray microprobe for the samples, FZ#4 and FZ#6 was 2.2% (average for 5
samples) and 0.8% (average for 3 samples), respectively.