Process for making low-resistivity CVC

Abstract

A process for making low resistivity CVC silicon carbide. Applicants have developed a better process for adding nitrogen to silicon carbide which has the safety economic advantages of doping with N.sub.2 with the ease of N.sub.2 release advantages of using NH.sub.3. Preferred embodiments of the present invention include a NH.sub.3 generator with a source of H.sub.2 and a source of N.sub.2 and an arc discharge apparatus adapted to produce NH.sub.3 gas from a combination of the H.sub.2 and N.sub.2 sources.

Claims

1. A process for making low-resistivity CVC SiC and low resistivity CVD SiC comprising the following steps: A) provide a NH.sub.3 generator comprising: 1) a source of H.sub.2 and a source of N.sub.2 2) an arc discharge apparatus adapted to produce NH.sub.3 gas from a combination of the H.sub.2 and N.sub.2 sources, B) install in a CVD reactor a substrate compatible with a thermally activatable reactant gas to produce chemical vapor deposition vapors and other reaction products, C) introduce the reactant gas into the reactor along with a gas stream from the NH3 generator, D) thermally activate the reactant gas and the gas stream from the NH.sub.3 generator such that the reactant gas reacts to produce CVD vapors and the gas stream from the NH.sub.3 generator produces atomic nitrogen, E) deposit materials from the CVD vapors and atomic nitrogen on the substrate with the atomic nitrogen dispersed within the materials from the CVD vapors.

2. The process as in claim 1 wherein the CVD reactor comprises a source of solid particles or fibers and the reactor is a CVC reactor.

3. The process as in claim 2 wherein the solid particles or fibers is introduced into the reactor along with the gas stream from the NH3 generator.

4. The process as in claim 1 wherein the NH3 generator comprises a spark plug, an ignition coil a MSD ignition control element and an ignition tester.

5. The process as in claim 4 wherein the NH3 generator is powered by an automotive battery.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0020] FIG. 1 is a prior art drawing depicting a CVC SiC system.

[0021] FIG. 2 is a drawing describing an arc discharge apparatus.

[0022] FIG. 3 shows an apparatus for making low resistivity CVC SiC.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

[0023] Applicants preferred process for adding nitrogen to silicon carbide can be describe by reference to FIGS. 2 and 3. FIG. 2 shows a technique for utilizing nitrogen (N.sub.2) and hydrogen (H.sub.2) source gasses in molecular form to produce NH.sub.4 which more readily gives up atomic nitrogen the course of chemical deposition processes. The atomic nitrogen feed equipment was constructed from commercial-off the-shelf-components. The premise was to flow N.sub.2 and H.sub.2 into the apparatus and utilize arc discharge to crack the strongly covalently bonded N.sub.2 and H.sub.2 molecules, which then react to form toxic NH.sub.3. The NH.sub.3 is substantially more efficient source of atomic nitrogen than N.sub.2. The nitrogen feed equipment includes spark plug 40, N.sub.2 source 42, H.sub.2 source 44, SS tube with conflat ends 46, CF electrical feed-through (rated to 40 KV) 48, Ignition coil 50, 12-volt auto battery 52, MSD-ga Ignition control 54 and MSD ignition tester 56.

[0024] FIG. 3 is a combination of FIG. 2 and prior art FIG. 1 where the output of the equipment shown in FIG. 2 replaces apportion of the reactant gas supply shown in FIG. 1.

Experimental Results

[0025] To achieve improvements in the nitrogen doping process for low resistivity CVC SiC®, Trex designed and built an arc discharge system, shown schematically in the FIG. 1 below. The premise was to flow N.sub.2 and H.sub.2 into the apparatus and utilize the arc discharge to crack the strongly covalently bonded N.sub.2 and H.sub.2 molecules, which then react to form ammonia. Ammonia produces a more efficient nitrogen dopant source to incorporate an n-type dopant into the CVC lattice.

[0026] The arc discharge apparatus was constructed using commercial-off-the-shelf automotive spark plug and ignition coils, shown in FIG. 2. The apparatus was assembled and bench tested prior to installation into Applicants' reactor. Trex installed the arc discharge apparatus into its 0.4 m (16″) reactor and defined an experimental test plan (Table 1) for low resistivity CVC SiC® development. Trex machined the nitrogen-doped material that was made during these runs into 101 mm×3 mm×3 mm bars and utilized an in-house four point measurement to determine resistivity.

[0027] The lowest resistivity achieved in the test plan was 0.3 ohm-cm (by the four point probe method described above), which is significantly lower than the approximately 0.5-1.0 ohm-cm achieved in the past with nitrogen doping CVC SiC® without the arc discharge. Adjustments to the hydrogen flow along with modest increases in nitrogen and spark frequency could enable the 0.1 ohm-cm target for semiconductor applications. This avenue was considered during the next phase of experimentation.

[0028] Several improvements were made to arc discharge apparatus for the next phase of experimentation: an oil-cooled coil was installed which maintained a lower operating temperature, stainless steel (SS) wool was added as a catalyst to promote gas ionization and thereby encouraging ammonia production, and a larger arc discharge chamber was construction to allow a higher volume of N.sub.2 and H.sub.2 to be ionized, thereby increasing the volume of ammonia generated.

[0029] Samples from each run were sent to a certified lab for volume resistivity measurements along with an un-doped CVC SiC control sample (TK18474). Test methods ASTM D4496 (AC measurement) and D257 (DC measurement) were used. Results were as follows:

[0030] TK18474 (control): 428 ohm-cm (DC), 882 ohm-cm (AC)

[0031] TK16422: 62 ohm-cm (DC), 85 ohm-cm (AC)

[0032] TK16423 (doped powder): 99 ohm-cm (DC), 142 ohm-cm (AC)

[0033] In order to further reduce volume resistivity the CVC SiC® doping process was moved to the 0.46 m (18″) reactor, which allowed for higher N.sub.2/H.sub.2 gas volumes, thereby theoretically permitting higher ammonia production.

[0034] Trex also opted to experiment with a modified CVC® manufacturing process. Trex retrofitted the 0.46 m (18″) reactor to flow NH.sub.3 (ammonia) directly into the chamber in the form of a 1% NH.sub.3 in Ar gas mixture. The rationale was to determine the relative effectiveness of 1% NH.sub.3 as the dopant gas versus generating ammonia in situ with the arc discharge apparatus. Run TK18620 was conducted using baseline CVC SiC® run parameters plus 5 slm 1% NH.sub.3-Ar. Preliminary run analysis suggests that the density of this material is lower than Trex's routine CVC SiC®, 3.0 g/cm.sup.3 vs. 3.21 g/cm.sup.3 respectively. The cause of this is still under evaluation. Samples from this run were sent to the same certified lab for volume resistivity analysis, along with another conductive CVD SiC sample with a resistivity advertised as <1 ohm-cm. ASTMs D4496 and D257 were used. Results are as follows:

[0035] Trex (TK18620): 27 ohm-cm

[0036] Third party sample: 98-140 ohm-cm

[0037] In parallel material from TK18620 was tested at the third party source's lab (certification unknown). Results indicate a volume resistivity value of 0.006 ohm-cm.

[0038] In the interim since the provisional application was filed Applicant submitted samples from its most recent low resistivity CVC SiC run to Orton Ceramic (a certified materials testing lab) along with commercially available low resistivity CVD SiC from a third party source (this third party source supplies the semiconductor industry with most of their CVD SiC material). Orton Ceramic used ASTM D4496 and ASTM D257 test methodologies to determined the volume resistivity of Applicants low resistivity CVC SiC and the commercially available low resistivity CVD SiC. Results are shown in the table below:

[0039] Results illustrate two important points: [0040] 1. Trex's low resistivity CVC SiC material has up to 100× lower volume resistivity than credible competition based on two different test methods. [0041] 2. The third party low res CVD SiC has a published resistivity value of less than 1 ohm-cm using method ASTM D4496, which is inconsistent with the certified lab results Trex obtained on their material.

[0042] To further qualify the material, a sample from TK18620 was sent to an EDM (electrical discharge machining) shop for wire EDM testing. EDM is a standard machining method used on conductive (low resistivity) materials and is significantly less expensive than diamond grinding, which is the standard machining method for non-conductive, hard ceramics like silicon carbide. The EDM shop indicated that TK18620 material cut beautifully.

[0043] What can be deduced from these results is that Trex's low resistivity CVC SiC material is superior to credible competition and is more suitable for semiconductor low resistivity and ultralow resistivity applications than said third party source.

[0044] One other point of note: the aforementioned third party source has stopped making their CVD SiC altogether. The semiconductor industry will soon find itself in a material source crisis. Trex's low resistivity CVC SiC is poised to become the semiconductor industry's material of choice.

[0045] Applicants' conclusion is Trex's material is clearly of lower resistivity than credible competition and these results validate our methodology for low resistivity CVC SiC material.

Variations

[0046] Persons skilled in the chemical vapor deposition art will recognize that many variation to the specific embodiments described above are possible. For example, many changes in the parameters disclosed can be made to increase the amount of nitrogen incorporated into the CVC SiC which will have a direct effect on the electrical resistance. The processes describe herein can also be applied to standard chemical vapor deposition. Therefore, the scope of the present invention should be determined by the appended claims.