The issue of material composition in the polysilicon reduction production process
Thread Content
In the production of polysilicon using the Siemens process, two main reactions occur to trichlorosilane inside the reduction furnace: one is a reduction reaction with hydrogen, and the other is its own decomposition reaction. Based on what I know at the moment, the proportion of self-decomposition reactions is relatively high. So in actual production, what effects and disadvantages would arise if the mixture were replaced with 100% trichlorosilane gas fed into the reduction furnace? Meanwhile, regarding the ratio of hydrogen to trichlorosilane in the mixture, is it true that the higher the proportion of hydrogen, the better? I hope everyone can exchange ideas :)
When the amount of hydrogen is high, the deposition rate is slow
There are very strict ratios to follow; if the proportion of hydrogen is high, the popcorning rate is low and the surface remains smooth, while an excessive amount of hydrogen reduces the deposition rate
Thank you to the two people above. Well, assuming an extreme scenario: what impact would it have on production if 100% trichlorosilane gas were used in place of the mixture and fed into the reduction furnace?
The furnace will be filled with smoke, making it impossible to see the silicon rods at all; it’s also not possible to measure the surface temperature of these rods using an infrared thermometer. It’s impossible to grow polycrystalline silicon; haha, a low hydrogen content poses various problems – the view through the lenses becomes blurred, let alone when the hydrogen content is 100%. This gentleman might as well give it a try by turning off the hydrogen control valve and see what happens
The values cited in the literature can vary, such as 1:5, 1:10, or even 1:20. However, in actual production, the exact value is likely to differ little among different manufacturers, or it will remain within a certain range. But does this value stay constant throughout the entire cycle, or is it necessary to make adjustments from time to time?
The highest value reported in the literature seems to be 1:10; any higher ratio would only result in a slower reaction rate, as hydrogen dilutes trichlorosilane significantly.
The ratio is generally around 10:1. We use TELEDYNE’s online analysis instruments to monitor the hydrogen content and implement interlock control
Actually, the ratio of SiHCl3 to H varies from manufacturer to manufacturer! Some are 4:1, some are 8:1 or 10:1, depending on the actual conditions of each factory! It can basically be controlled within a range of 4:1 to 10:1!
Ours is around 5:1, but it often blooms
In that case, if you use temperature to control the current, it will end in a disastrous outcome: the current rises rapidly in a short time and the rod melts. If current control is used, I know only that the temperature of the rod will drop sharply, and the reaction cannot be stopped; adding hydrogen again will restore things to normal
A higher hydrogen ratio results in a higher conversion rate of trichlorosilane, but the deposition speed is slower
The general ratio is 5:1 to 8:1. From a growth perspective, a lower ratio of trichlorosilane results in higher consumption, a lower conversion rate, and a faster deposition rate; decomposition due to heat is the main mechanism in this case. A higher ratio is generally used for melting materials in the initial stages of production, as a smooth and dense surface is required, but the deposition rate is slow, meaning growth is slower. Now, zone-melted single crystals are rare; they’re all Czochralski-grown, with no particular requirements regarding the surface, as it’s going to be broken anyway
It seems that the improved Siemens model will soon be fully domestically produced, and the investment required for production lines with a capacity of 1,500 tons per year is likely to decrease significantly!