Reactor and process for endothermic gas phase reaction in a reactor

Abstract

The invention provides a process for endothermic gas phase reaction in a reactor, in which reactant gases are introduced into the reactor via a gas inlet apparatus and distributed homogeneously into a heating zone by means of a gas distribution apparatus, wherein the reactant gases are heated in the heating zone to a mean temperature of 500-1500 C. by means of heating elements and then conducted into a reaction zone, the reactant gases reacting in the reaction zone to give a product gas which is conducted out of the reactor via a gas outlet apparatus. Further subject matter of the invention relates to a process for endothermic gas phase reaction in a reactor, wherein the heating of the heating elements is controlled by temperature measurements in the reaction zone, at least two temperature sensors being present in the reaction zone for this purpose, and reactor for performance of the process.

Claims

1. A process for converting silicon tetrachloride to trichlorosilane in a reactor, said method comprising the following steps: (a) providing the reactor comprising: (i) a gas inlet apparatus; (ii) a gas distribution apparatus selected from the group consisting of a gas distributor plate and a gas distributor screen; (iii) a heating zone comprising heating elements; (iv) a reaction zone; (v) at least two temperature sensors provided in the reaction zone and not in the heating zone, wherein the at least two temperature sensors are spatially resolved from each other; and (vi) a gas outlet apparatus; (b) introducing reactant gases into the reactor via the gas inlet apparatus; (c) homogeneously distributing with the gas distribution apparatus the reactant gases into the heating zone; (d) heating the reactant gases with the heating elements in the heating zone to a mean temperature of 500-1500 C.; (e) conducting the reactant gases from the heating zone into the reaction zone; (f) reacting the reactant gases in the reaction zone, wherein the reactant gases are silicon tetrachloride and hydrogen, which react at a temperature of at least 600 C. to give a product gas comprising trichlorosilane and HCl; (g) acquiring temperature measurements from the at least two temperature sensors in the reaction zone; (h) controlling the heating of the heating elements based on temperature measurements acquired in the reaction zone; and (i) conducting the product gas out of the reactor via the gas outlet apparatus, wherein a mean deviation of temperatures determined at the temperature sensors from a mean thereof is not more than 50 K.

2. The process as claimed in claim 1, wherein at least one heat exchanger is provided, which heats reactant gas by a countercurrent principle by use of product gas produced in the reaction, the gas distribution apparatus being arranged between the at least one heat exchanger and the heating zone.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention will be described in conjunction with the following drawings, wherein:

(2) FIG. 1 shows a cross-sectional schematic view of an apparatus suitable for performance of the inventive process;

(3) FIG. 2 is a bar graph of damage probabilities without gas distribution;

(4) FIG. 3 is a bar graph of damage probabilities with gas distribution;

(5) FIG. 4 is a bar graph of damage probabilities with optimized gas distribution; and

(6) FIG. 5 is a bar graph showing the data of FIGS. 2-4 on a single graph.

LIST OF REFERENCE NUMERALS

(7) 1 gas feed including gas supply 1a and distributor apparatus 1b 2 heating elements 3 heating zone 4 deflecting apparatus 5 gas conduit 6 two temperature measurement devices 7 reaction zone 8 gas outlet

(8) The invention relates in general terms to endothermic gas phase reactions.

(9) The examples which follow relate to the conversion of silicon tetrachloride to trichlorosilane.

EXAMPLES

Example 1 (Comparative Example)

(10) For the comparative example, a reactor according to U.S. Pat. No. 4,536,642 was used.

(11) A gas mixture in the reactant stream consisting of 33 mol % of silicon tetrachloride and 67 mol % of hydrogen was used. The inlet temperature of the reactant gas stream was about 175 C.

(12) The pressure was set to 6 bar and the temperature of the gas in the reaction zone to 1000 C.

(13) After the reaction, the product gas was analyzed in a gas chromatograph, and the proportions of trichlorosilane and silicon tetrachloride were determined. The temperature of the exiting product gas stream was about 350 C.

(14) The relative selectivity is given by the molar proportion of trichlorosilane relative to silicon tetrachloride.

(15) For the sake of simplicity, the relative selectivity attained in the comparative example is defined as 100% if all heating elements are working.

(16) FIG. 2 shows the relative probability of damage to the individual heating elements (A-N) as a function of the heating element number for the comparative example.

(17) It is clear that the spatial distribution of the occurrence of damage to the heating elements apparently does not obey any recognizable laws.

(18) This constitutes the prior art.

(19) If at least one heating element has failed, the power of the remaining functioning heating elements is regulated such that the target temperature which is measured in the middle of the reaction zone with a temperature sensor is maintained.

(20) However, it was found that, even in the event of one failed heating element, the relative selectivity falls to about 97%.

(21) The occurrence of by-products has risen by 3%.

Example 2

(22) In Example 2, essentially the same boundary conditions are used as in Example 1.

(23) However, the gas supply is distributed better into the heating zone using a gas distribution apparatus.

(24) The gas distribution apparatus homogenizes the gas stream fed in over different gas passage dimensions along the gas feed into the cylindrical heating zone.

(25) FIG. 3 shows a distinct improvement in the probability of damage resulting from the incorporation of a gas distribution apparatus.

(26) The random cases of damage are altered to become a systematic distribution.

(27) FIG. 3 additionally shows an inhomogeneous but systematic distribution of the gas phase supplied.

(28) The relative probability of damage to the heating elements is reduced and the reactor can be operated for longer.

(29) By virtue of the now systematic gas distribution, it is possible by further optimization steps to adjust and further improve the apparatus for better gas distribution according to the geometry of each individual reactor.

(30) This can be accomplished by further adjusting the dimensions of the distribution apparatus.

(31) FIG. 4 shows the optimized probability of damage to the heating elements, which is distributed homogeneously over each element.

(32) In contrast to FIG. 3, however, no reduction in the number of instances of damage is discernible.

(33) This optimized damage distribution has to be established individually for every reactor and shows the systematic representation here.

(34) FIG. 5 shows the direct comparison of the different probabilities of damage.

(35) The relative probability of damage to the heating elements is reduced and the reactor can be operated for longer.

Example 3

(36) In Example 3, essentially the same boundary conditions are used as in Example 1.

(37) However, the reaction zone is equipped with an additional four temperature measurement devices, such that it is possible to additionally measure the temperature in the reaction zone in a spatially resolved manner.

(38) The temperature measurement devices are arranged radially around the centre of the base plate within the reaction zone.

(39) FIG. 1 shows, by way of example, the position of two of these additional temperature measurement devices 6.

(40) If the determination of the temperature is carried out not as described in Example 1 with only one temperature measurement device but with the mean of all the values from the available temperature measurement devices, it is found that, in the event of failure of a heating element, the relative selectivity falls only to 99.5%, since the direct influence of temperature on the regulation temperature is reduced.

(41) The unwanted by-products from an increased reaction temperature occur only up to 0.5%.

Example 4

(42) In Example 4, in addition to Example 2, the heating elements are regulated such that the deviation of the temperatures measured at the heating elements from the mean T is at a minimum.

(43) This is done at every time point.

(44) It has been found that, even when all heating elements are working, there can be significant differences in the temperatures.

(45) The reason for this is probably the geometry of the heating zone (and hence the gas flow) and/or the geometry of the heating elements.

(46) If a T of less than 50 K is set when all the heating elements are working, a relative selectivity of 110% compared to Example 1 can be achieved.

(47) Even if one heating element is no longer working, a relative selectivity of 107% compared to Example 1 is still achieved.

(48) Here too, the relative probability of damage to the heating elements is reduced.

(49) The service life, the conversion and the reliability of the reactor can be distinctly prolonged or increased as a result.