The Rochow Direct Method Process for making methylchlorosilanes by effecting reaction between powdered silicon and methyl chloride in the presence of a copper catalyst is shown by U. S. Pat. 2,380,995. As taught by Rochow in Chemistry of the Silicones, Second Edition, (1951), John Wiley and Sons, New York, pg. 79-80, dimethyidichlorosilane is the source of dimethylsiloxy, or"D"units when hydrolyzed, and is basic for the manufacture of high molecular weight polymers. It has been long recognized, as shown by Rochow et al., J. Am. Chem. Soc., 63 798 (1941) that in addition to copper, key metallic promoters, such as zinc and tin, also can enhance the formation of dimethyldichlorosilane during direct method methylchlorosilane crude production.

T. Margaria et al., WO/95/01303, (1994) have further described the use of phosphorus as a promoter for dimethyldichlorosilane formation in the direct method. Copper phosphide has been citedby xHalm fet al., in U. S. patent 4,762,940 (1990), as useful source of phosphorus for improving dimethyldichlorosilane selectivity in direct method procedures. As shown in U. S. patent 5,059,343, Haim et al., also investigated the effects of using phosphorus to improve dimethyidichlorosilane selectivity in direct method operations in combination with copper, tin and zinc. Even though significant advances have been made with respect to procedures for improving the yields of dimethyldichlorosilane in direct method methylchlorosilane manufacture, further techniques are constantly being sought.

BRIEF SUMMARY OF THE INVENTION The present invention is based on the discovery that while phosphorus is recognized as a dimethyldichlorosilane promoter, its effectiveness for dimethyidichlorosilane formation during direct method operation can be substantially enhanced by introducing phosphorus into the reactor, while maintaining a high weight ratio of copper to zinc. Support for such conclusions involving phosphorus usage, is shown in a direct method study operating over a Cu/Zn ratio range of 25 to 250.

SUMMARY OF THE INVENTION A method for promoting the formation of dialkyldihalosilane during direct method alkylhalosilane production, comprises effecting reaction between alkylhalide and powdered silicon in the presence of a catalyst comprising copper, silicon, zinc, and phosphorus, where there is maintained during alkylhalosilane formation, a reaction mixture comprising by weight, an average of from about 1% to about 5% of copper based on the weight of silicon, a sufficient amount of zinc to provide a Cu/Zn weight ratio having a value of from about 25 to about 250, and a proportion of from about 100 ppm to about 1000 ppm of phosphorus, based on the weight of reaction mixture.

DETAILED DESCRIPTION OF THE INVENTION Among the reactors which can be used in the practice of the invention, there are included a fixed bed, a stirred bed and a fluid bed. The invention can be practiced in a continuous, or semi-continuous manner. Although methyl chloride is the alkyl halide of choice, other alkyl halides such as C14 alkyl chlorides, for example, ethyl chloride, propyl chloride, etc., also can be used.

In instances where-a fluid bed reactor is used, methyl chloride, an inert gas such as argon, or mixture thereof can be used to fluidize the bed of silicon particles. The silicon particles can have an average size of below 700 microns, with an average size of greater than 20 microns and less than 300 microns. Preferably, the mean diameter is in the range of 100 to 150 microns.

Among the copper compounds which can be used as copper sources in the practice of the invention are carboxylic salts of copper, and partially oxidized copper. Additional copper sources are particulated cupric chloride and cuprous chloride, copper flake, brass and bronze.

Zinc metal powder, halides of zinc, such as zinc chloride, zinc oxide, have been found effective as sources of zinc. Relative to copper and tin weight ratios, tin can be present in the reaction mixture during dialkyldihalosilane formation, at from 200-3000 ppm of tin, per part of copper. Sources of tin include, tin metal dust, tin halides, tin oxide, tetramethyl tin, alkyl tin halides, brass and bronze.

Figure 1 shows a cylindrical, heat resistant glass fixed bed reactor at 1, having a porous glass frit at 2 and a charge of silicon powder at 3. The fixed bed reactor is enclose in a tin oxide-coated glass heating tube at 4 having electrical terminals not shown. A series of thermocouples inserted in stainless steel tubing are shown at 5 and 6 which respectively monitor temperatures of the powdered silicon mass and the preheat zone. An outer glass insulation and safety tube is shown at 7. A methyl chloride inlet is shown at 8 and an outlet for silane vapor and unreacted methyl chloride is shown at 9.

Figure 2 shows a cylindrical, heat resistant glass fluidized bed reactor at 20, having a fluidized bed of silicon at 21 and a stirrer at 22. A porous frit for containing silicon powder, which allows methyl chloride flow into the reactor, is shown at 23. The reactor is enclose in a glass heating tube at 24, which is wrapped in Nichrome wire at 25. Thermocouples inserted in stainless steel tubing are shown at 26 and 27 to monitor the temperature of the bed and the reactor. An outer glass safety tube is shown at 28. A methyl chloride inlet port is shown at 29 and a silane vapor and unreacted methyl chloride exit port is shown at 30. A vibrator, not shown, can be used in particular situations to stabilize the fluidized bed.

The graph at Figure 3 shows the dramatic impact phosphorus can have, at a 500 ppm level, on promoting dimethydichlorosilane formation during direct method methylchlorosilane production. The data is based on operating a fixed bed, such as shown in Figure 1, at 20% silicon utilization, over a broad Cu/Zn weight ratio range, employing copper phosphide as a phosphorus source. As shown by the graph, Cu/Zn ratio values of about 30 to about 100 are particularly significant.

In order that those skilled in the art will be better able to practice the present invention, the following description is given by way of illustration, and not by way of limitation. All parts are by weight unless otherwise indicated.

EXAMPLE 1.

A fixed bed reactor, as shown by Figure 1, is used in a dimethyldichlorosilane promotion study. The fixed bed reactor is a glass tube 20 cm long having a 1.3 cm outside diameter. A glass frit is located 6 cm from the end to support a bed of powdered silicon. The bed size is 6 grams. The highest anticipated reaction rate is 1 gram crude/hr-gm silicon.

Analysis is conducted to determine crude composition. Silicon utilization is determined by weighing crude samples at various times. Typically data is recorded at 20% by weight silicon utilization. Selectivity is determined by gas chromatography,"GC", with thermal conductivity detectors. Individual silane component analysis is performed by analysis of reagent grade silanes obtained from Aldrich Co. GC calibrations are performed by analysis of silane mixtures of known composition.

Silicon powder used in the study, is furnished by Elkem Co. of Alloy, West Virginia. The silicon powder is ground to achieve an average surface area of 0.38 meters2/gm. The major elemental components in ppm present in the silicon are AI (1800), Ca (20), Fe (5000), P (40). Several of the catalyst sources, such as cuprous chloride, powdered tin, and powdered zinc used in the study are obtained from the Aldrich Co. of Milwaukee, Wisconsin.

Among the sources of phosphorus used, there are included, (Cu) 3P from Green Back Industries of Greenback, Tennessee, Zn3P2, PCt3, P (CH3) 3, and P (C2H5) 3 from the Aldrich Co.

A preferred copper source used in the study, is cuprous chloride which is preferably pre-reacted at 350°C with silicon powder to form a contact mass as shown by the following equation: Si + CuCI--> Contact Mass (Cu3Si) + 1/4 SiC14 The contact mass is added to the reactor with other catalyst components.

The contact mass is prepared as follows: A mixture of 40 g of silicon powder (0.38 micron particle size), and 5.43 g of hexane is combined with a hexane slurry of 12.67 g of solids consisting of CuCI and tin dust having a Cu/Sn ratio of 1000/1. The mixture is initially dried with a stream of nitrogen under ambient conditions, and then fully dried under reduced pressure at ambient temperatures.

The pre-contact mass is placed in a crucible and then into a furnace and heated to 300°C under a flow of argon until SiC14 evolution ceases. An NH40H indicator is used to monitor the effluent. The weight change due to loss of volatile products accompanying the reaction to prepare contact mass is within experimental error equal to the calculated weight loss.

The fixed bed reaction is run over a period of 8 hrs. at a temperature of 300°C to 310°C. There is used a contact mass based on an initial charge of 6 g of powdered silicon, and 5% by weight copper of the contact mass in the form of CuCI. Sufficient powdered tin and powdered zinc are also blended into the contact mass to provide a Sn/Cu ratio having a value of about 1000 ppm, and a Cu/Zn ratio having a value of 10/1. A methyl chloride flow rate is maintained at 35 mL/min. After about 20% silicon utilization, based on weight of crude formed, the selectivity for dimethyldichlorosilane is found by gas chromatography to be 86.6%.

Following substantially the same procedure, a series of fixed bed reaction runs are made involving the use of contact mass carried to at least 20% silicon utilization, with and without phosphorus at initial Cu/Zn ratios at 10/1 and100/1.

Prior to charging the reactor, appropriate levels of zinc dust, and sources of phosphorus, such as Cu3P are blended into the contact mass. Volatile phosphorus compounds, such as Pots, and (CH3) 3P, are injected upstream with methyl chloride flow into the contact mass.

The following Table shows the results obtained, where"D"is dimethyldichlorosilane: % D at 20% Si Utilization Cu Source Cu/Zn Ratio P (ppm as (CH3) 3P) CuCI 10 : 1 0 89. 7 ""500 88.4 100: 1 0 85.6 500 94.2 Consistent with the results of Figure 3, a significant dimethyidichlorosilane promotional effect is shown as a result of the use of 500 ppm of phosphorus at Cu/Zn ratios of 100/1. For example, at 0% phosphorus an 85.6% D level is provided, while at 500 ppm of phosphorus, a 94.2% D level is shown.

Similar results are obtained using (CH3) 3P, (C2H5) 3P, or PC13 as phosphorus sources. For example, at 500 ppm, (CH3) 3P provides a 91.3% yield at 10/1 Cu/Zn weight ratio, and a 93.2% yield at 100/1 Cu/Zn weight ratio.

EXAMPLE 2.

The fixed bed reaction procedure of example 1 is substantially repeated, except in place of contact mass, the source of copper is copper flake and powdered brass obtained from OMG Americas, Research Triangle Park, North Carolina. The brass comprises by weight, 80% copper, 19.5% zinc and 0.5% tin.

The fixed bed reactions are conducted at 5% by weight copper levels, based on bed weight, comprising blends of powdered silicon, copper flake and appropriate amounts of brass. Master batch slurries in hexane or toluene of silicon, copper flake, and brass are made. A Cu/Zn weight ratio of about 11: 1 is made using 135 g of silicon, 4.29 g of copper flake, and 3 g of brass.

Another slurry is prepared to provide a Cu/Zn weight ratio of 35: 1. The solvent is separated under reduced pressure, prior to charging the resulting blend into the fixed bed reactor.

The following results are obtained, employing (CH3) 3P at 0 and 500 ppm, and at Cu/Zn weight ratios of 11: 1 and 35: 1, Cu Source Cu/Zn Ratio P (ppm as (CH3) 3P) % D at 20% Si Utilization Cu flake 11: 1- 0 88.3 ""500 90.8 "35: 1 0 83.3 ""500 91 EXAMPLE 3.

A fluid bed reactor, similar to figure 2, is used in a direct method study. There is employed a copper/zinc catalyst to determine whether the effectiveness of phosphorus as a dimethyldichlorosilane promoter is influence by the particular Cu/Zn weight ratio range used during the reaction. The fluid bed reactor consists of a 3.8 cm ID glass tube with a glass frit at the center to support the silicon bed. The reactor is enclose in a glass heating tube which is wrapped in Nichrome wire. In order to fluidize the bed, a stirrer is used in combination with a vibrator not shown.

The reactor is initially charged with 20 g of contact mass having 5% copper and a starting Cu/Zn weight ratio of 10: 1. Methyl chloride is fed into the reactor over a 24 to 28 hr period at a bed temperature of 300-310°C. After 35% silicon utilization, the crude is analyzed for % dimethyldichlorosilane by gas chromatography. The same procedure is repeated, except that the bed includes 10 mg of Cu3P, which provides 500 ppm of phosphorus.

A similar fluid bed procedure is repeated with and without 10 mg of Cu3P, except a Cu/Zn weight ratio of 100: 1 is used. The Cu/Zn weight ratios of 10: 1 and 100: 1 are respectively repeated several times. The following are the results obtained from the fluid bed reactor using 10: 1 and 100: 1 Cu/Zn weight ratios: Fluid Bed Reactor Cu/Zn ratio Phosphorus (ppm) % D 10/1 0 87 "500 90 100/1 0 83.2 "500 92.9 The above results are average values obtained from several repeat runs which are made to determine whether a statistically significant change occurs with respect to % D in methylchlorosilane crude, based on employing phosphorus at different Cu/Zn ratios. No significant statistical increase is found when comparing the 87% D yield (standard deviation=2) using a Cu/Zn ratio of 10: 1 without phosphorus, to the 90% D yield (standard deviation=3.7) with phosphorus. However, at a Cu/Zn ratio of 100/1, a statistically significant increase occurs when the value 83.2% (standard deviation=2.3) is obtained without phosphorus, as compared to the 92.9% (standard deviation=2.9) obtained with phosphorus.

Although the above examples show only a few of the very many variations of conditions and materials which can be used in illustrating the practice of the present invention, it should be understood that the present invention is directed to a much broader scope as set forth in the description preceding these examples and the appended claims.