Detailed Description
Referring to fig. 1, the double-trench SiC power MOS device of the present invention is characterized by including a source 1, a drain 6, a gate 3, an N + substrate layer 7, and an N-drift region 8, where the drain 6 and the N-drift region 8 are respectively disposed on the bottom surface and the top surface of the N + substrate layer 7; two sides of the upper surface of the N-drift region 8 are respectively provided with L-shaped P well regions 10 which are symmetrically arranged, and the upper surface of the outer end of the transverse section of the P well region 10 is provided with a P + region 11; a P shielding layer 5 is arranged in the middle of the P well region 10 on the two sides, and a super junction 9 is arranged below the P shielding layer 5; a grid electrode groove is arranged on the P shielding layer 5, and a grid electrode 3 is arranged in the grid electrode groove; a P-body region 4 is arranged between the grid electrode groove and the P well regions 10 at two sides, and an N + region 2 is arranged above the P-body region 4; a source electrode 1 is arranged above the P well region 10, the P + region 11 and the N + region 2; an interconnection electrode 12 is provided between the source electrode 1 and the gate electrode 3 on both sides. Where N-drift region 8 consists of an upper portion of N-drift region 8 and a lower portion of N-drift region 8 (seen as one region in fig. 1), located above and below superjunction 9, respectively. During manufacturing, the lower part of the N-drift region 8 and the upper part of the N-drift region 8 are formed by epitaxy twice.
Referring to fig. 2, the process flow of the method for manufacturing the device of the present invention is as follows (the specific operations of the following steps of the process all belong to the conventional technology):
step 1, forming the lower part of an N-drift region (8): epitaxially forming the lower part of an N-drift region (8) on one side of the N + substrate layer (7), and carrying out standard wet process cleaning on the surface; see diagram a in fig. 2.
Step 2, formation of the superjunction (9) (see diagram b in fig. 2):
(1) coating photoresist on the surface of the N-drift region (8) of the cleaned epitaxial wafer, etching a high-temperature ion implantation region of an N-type region in the super junction (9), and then performing high-temperature nitrogen ion implantation to form an N region of the super junction (9), wherein the thickness of the N region is 0.2 mu m, and the concentration of the N region is 3e +17cm-3。
(2) Coating photoresist on the surface of the epitaxial wafer on which the super junction (9) N region is formed, etching a P-type high-temperature ion implantation region in the super junction 9, and then performing high-temperature Al ion implantation to form a film with the thickness of 0.2 mu m and the concentration of 3e +17cm-3The P-type region of superjunction 9.
Step 3, forming the upper part of the N-drift region (8): epitaxial growth is carried out on the upper surface of the super junction (9) to obtain 11 mum thickness and 7.5e +15cm concentration at the upper part of the N-type drift region (8)-3(ii) a See diagram c in fig. 2.
Step 4, forming a source groove (13) and a gate groove (14): etching two ends of the upper surface of the upper part of the N-type drift region (8) to form a source electrode groove (13); etching and forming a gate trench (14) in the middle of the upper surface of the upper part of the N-type drift region (8); see diagram d in fig. 2.
And step 5, forming a P well region (10): the source trench is coated with photoresist, and a P well region (10) is etched to form a P well region with a concentration of 5e +17cm-3The depth was 0.3. mu.m. Formation of P shield layer (5): coating photoresist on the grid groove, etching a P shielding layer (5) injection region, and performing P-type ion injection to form a P shielding layer (5) with the concentration of 5e +18cm-3. See e diagram in fig. 2.
And 6, forming a P-body region (4): coating photoresist on the chip finished in the previous step, etching P-body (4) ion implantation regions at two sides of the gate trench, and performing P-type ion implantation to form P-body (4) regions with concentration of 1e +17cm-30.5 μm in depth; see diagram f in fig. 2.
And 7, forming an N + region (2): coating photoresist on the upper surfaces of the P-body region (4) and the P-well region (10), etching an implantation region of the N + region (2), and performing N-type ion implantation to form an N + (2) region with the concentration of 1e +20cm-3Depth of 0.2 μm; see graph g in fig. 2.
And 8, forming a carbon protective film on the surface of the N-drift region (8): after the P-type ion implantation of the step 5 is carried out, a carbon protective film is formed on the surface of the N-drift region (8);
step 9, high-temperature ion implantation activation: and (3) carrying out 1600 ℃ high-temperature ion implantation annealing on the chip completed in the step (9):
step 10, removing the carbon protective film: removing the carbon protective film on the surface of the chip:
step 11, preparing a grid oxide layer (15): cleaning the surface of the chip with the carbon protective film removed by HF acid, and then carrying out SiO2Growing a gate dielectric layer to form a gate oxide layer (15) with the thickness of 50 nm; see graph h in fig. 2.
Step 12, formation of a drain electrode (6): after the formation of the grid oxide layer (15), drain region metal deposition is carried out, two metals of Ni/Au are 20nm/240nm, and then a drain electrode (6) forming drain region ohmic contact is stripped.
Step 13, forming a source (1): and coating photoresist, etching a source region ohmic contact pattern, performing source region metal deposition, sequentially depositing 20nm/240nm Ni/Au metals, then performing ohmic contact annealing, and stripping to form a source electrode (1). See graph i in fig. 2.
Step 14, forming the gate 3: forming a gate electrode on the SiC sample subjected to the ohmic contact electrode annealing: Ni/Au of 20nm/240 nm; see graph j in fig. 2.
Step 15, formation of the interconnection electrode (12): coating a stripping glue and a photoresist on the surface of a chip on which a grid (6) is formed, etching contact holes of the grid (6), a source electrode (1) and an N + region (2), then depositing a metal layer, and stripping to form a pattern of an interconnection electrode (12) among the grid (6), the source electrode (1) and the N + region (2), wherein the interconnection electrode (12) is made of Ti/Au two-layer metal and has the thickness of 50nm/200nm respectively. See graph k in fig. 2.
The device of the invention can improve the breakdown voltage of the device, reduce the on-resistance and effectively prevent the advanced breakdown of the gate oxide, and the principle of the device is described as follows:
the existing single-groove UMOSFET has the advantages that the electric field of the device is mainly concentrated below the gate oxide layer, so that the reliability of the oxide is seriously tested and the oxide is easy to break down in advance, and after the double-groove design of source grooving is introduced, the electric field concentration effect at the bottom of the gate oxide layer can be relieved by the groove at the source position, so that the device is prevented from breaking down in advance, and the breakdown voltage of the device is improved. The doping concentration of the introduced P shielding layer is larger than that of Pbase, when the device works in a reverse state, the drain electrode is applied with high voltage, and the electric field of the device is dispersed to the P shielding layer and the N-drift region, so that the electric field concentration effect at the bottom of the gate oxide is relieved, and the advanced breakdown of the gate oxide is avoided. The voltage increases, the depletion region increases, and the electric field decreases. Therefore, the higher the source-drain voltage is, the higher the resistance of the drift layer is, and the higher the on-resistance thereof becomes; the super junction structure is characterized in that an N layer and a P layer in a drift layer are arranged in a longitudinal groove structure, after voltage is applied, depletion layers are connected in a transverse extending mode, finally the depletion layers under the gates are expanded, and the super junction structure breaks through the dependence of the doping concentration of a device and VB, so that on the basis of keeping high breakdown voltage, higher doping concentration can be kept, and the on-resistance of the super junction structure can be reduced.