Method achieving higher inversion layer mobility in novel...

Semiconductor device manufacturing: process – Silicon carbide semiconductor

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C438S705000, C438S766000, C438S966000, C438S969000

Reexamination Certificate

active

06407014

ABSTRACT:

FIELD OF THE INVENTION
This invention relates to a novel method of achieving low interface states between the interface of silicon carbide and thermally grown oxide and to novel silicon carbide semiconductor devices derived therefrom.
BACKGROUND OF THE INVENTION
Silicon carbide is a superior semiconductor material for the production of power MOSFET and Schottky barrier rectifiers. Indeed, silicon carbide MOSFETs have been reported with blocking voltages as high as 1400V. However, the on-resistance of these MOSFETs are orders-of-magnitude higher than is theoretically predicted for similarly rated ideal silicon carbide unipolar devices. The reason for this large on-resistance has primarily been the low inversion layer mobility, which is believed to be less than 10 cm
2
/Vs for these MOSFETs. This value of inversion layer mobility is more than two orders-of-magnitude lower than the bulk silicon carbide mobility. This low inversion layer mobility for silicon carbide MOSFETS is primarily due to the large interface state density between silicon carbide and thermally grown oxide. The lowest interface state density on p-type silicon carbide of which we are aware is 1×10
11
cm
−2
eV
−1
. This interface state density is much higher than the interface state density observed for MOS type devices in silicon technology, where this value is routinely observed to be about 1×10
10
cm
−2
eV
−1
.
Additionally, silicon carbide is chemically inert in nature and is not attacked by most of the common etchants at room temperature due to the strong bond between carbon and silicon in monocrystalline silicon carbide. At the same time, the bonds between silicon and carbon in amorphous silicon carbide are weak. In my work with B. J. Baliga, it was reported that monocrystalline silicon carbide is not attacked by most of the common laboratory etchants, such as HF, HNO
3
, KOH, HCl, etc.; that silicon is etched by using a hot mixture of HF and HNO
3
, that carbon can be etched using hot HNO
3
, and hence we suggested that monocrystalline silicon carbide be converted to amorphous silicon carbide and be etched by treating it as a mixture of silicon and carbon. See Alok et al, Journal of Electronic Materials, Vol. 24, No. 4, pp. 311-314, and the similar disclosure of U.S. Pat. No. 5,436,174, wherein this work is used to form a trench in a monocrystalline silicon substrate by directing first electrically inactive ions using ion implantation into a first portion of the monocrystalline silicon carbide substrate to create an amorphous silicon carbide region followed by removal of the first amorphous silicon carbide region to form a trench in the monocrystalline silicon carbide using an etchant which selectively etches amorphous silicon carbide at a higher rate than monocrystalline silicon carbide.
U.S. Pat. Nos. 5,318,915, 5,322,802, 5,436,174, and 5,449,925 use amorphization to create deep PN junctions or deep trenches in SiC wafers. However, these references do not produce integrated circuits, do not convert a SiC wafer part into silicon, and do not provide for improvement in speed and performance of integrated circuits. Other workers in the art (JPA 55024482 and JPA 07082098) have attempted to create SiC areas in a Si wafer by converting a thin layer of Si into SiC using ion implantation. Such thin layers can not be used to create high voltage (>1000V) vertical power devices. Moreover, attempts in our laboratory to convert part of a Si wafer to SiC using high temperature ion implantation were unsuccessful.
There is a continued need in the art for methods for achieving low interface states between the interface of silicon carbide regions and oxide regions, in particular thermally grown oxide regions, and thereby increasing the inversion layer mobility in silicon carbide MOS devices, and for novel silicon carbide semiconductor devices derived therefrom. There is also a continued need in the art for methods for creating silicon-rich regions in silicon carbide substrates such as silicon carbide wafers and for novel silicon carbide semiconductor devices derived therefrom.
SUMMARY OF THE INVENTION
An object of the invention is to increase the inversion layer mobility in silicon carbide MOS devices.
Another object of the invention is to increase the inversion layer mobility in silicon carbide MOS devices by reducing the interface states between selected silicon carbide regions and thermal oxide regions in said devices.
Another object of the invention is to provide method for creating silicon-rich regions in silicon carbide substrates such as silicon carbide wafers and to provide novel silicon carbide semiconductor devices derived therefrom.
These and other objects of the invention will be apparent from the description of the invention which follows.
It has been found that (i) the unwanted interface states in silicon carbide derived devices are attributed to carbon; (ii) carbon may be selectively removed to reduce the interface states between the silicon carbide region and oxide regions, either grown as a result of thermal oxidation, or by oxide deposition followed by thermal oxidation; and (iii) this reduction of interface states leads to increased inversion layer mobility and improved performance of silicon carbide MOS devices derived by the method of this invention and having such reduced interface states.
Thus in my present work, it is desired to selectively remove carbon from silicon carbide, using an etchant effective to selectively remove carbon that is present in silicon carbide, preferably an etchant that only attacks carbon and not silicon, or that etches carbon at a much faster rate than it etches silicon and/or silicon carbide. However, before any effective etching and removal of carbon from silicon carbide can be realized, it is necessary to break the bond between silicon and carbon that is present in the monocrystalline silicon carbide.
The invention relates to a method which comprises steps to remove at least an effective amount of carbon from a region, preferably a surface region, of silicon carbide prior to oxide formation on said region by: (a) performing at least one amorphizing step, preferably by ion implantation to a desired depth to convert at least an effective amount, preferably a substantial amount and most preferably all of selected regions of monocrystalline silicon carbide to regions of amorphous silicon carbide; (b) selectively removing or dissolving at least an effective amount of carbon from said amorphous silicon carbide region(s) through use of an etchant selective to carbon such as hot HNO
3
to form an amorphous silicon-rich region; and (c) forming an oxide on said amorphous silicon-rich region, preferably by subjecting the etched amorphous region to thermal oxidation or to an oxide deposition procedure followed by thermal oxidation.
The invention provides a method for the production of silicon carbide devices which have an oxide region on
(a) either an amorphous silicon-rich region which is (i) predominantly or entirely amorphous silicon or (ii) a mixture of predominantly amorphous silicon in combination with amorphous silicon carbide and /or silicon dioxide or
(b) a monocrystalline silicon region;
wherein (a) or (b) is present on a region of a silicon carbide substrate, or
(c) a region of a silicon carbide substrate,
and to novel silicon carbide devices derived therefrom.
An outline of exemplary methods and regions created thereby is given in FIG.
1
.
In specific embodiments, the method is used to produce high quality oxide on silicon carbide and/or silicon carbide devices having an oxide region on silicon carbide and includes the steps of:
(a) amorphizing silicon carbide in at least one region of a monocrystalline silicon carbide substrate to convert the silicon carbide in said region to amorphous silicon carbide on a monocrystalline silicon carbide substrate;
(b) removing at least an effective amount of the carbon from the resulting amorphous silicon carbide region with an etchant effective to selectively remove said effective amount o

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Method achieving higher inversion layer mobility in novel... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Method achieving higher inversion layer mobility in novel..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method achieving higher inversion layer mobility in novel... will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-2920682

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.