What are the main reasons for rod inversion in polysilicon production?

The most likely reason is too high a temperature, which causes the silicon core to melt and spill over； Either it isn’t installed symmetrically and stably, and as the amount of deposition increases, it eventually collapses. Of course, there might be other reasons as well!

However, the temperature we set was not high, and there was a bias current at the silicon core and graphite collet (one side was longer while the other was not as long, and it was irregular).

It’s Jiangxi LDK, I heard that LDK is pretty bad

Is it caused by uneven air intake distribution, with different airflow speeds at various points across the cross-section? I’m just guessing too; I want to know if that’s the reason.

It is not clear under what circumstances the rod should be replaced: 1. The quality of the silicon core? 2. At what thickness should the rod be cut down? 3. Technical parameters, etc.? Each case needs to be analyzed on its own merits.

The master rod wasn’t chosen properly, and it wasn’t installed correctly as well.

The problem of polycrystalline silicon rod failure is a major constraint on polycrystalline silicon production volume; those who have not been involved in actual production do not have much say on this matter. There are many possible causes for such failures, and various factors need to be considered: firstly, the quality of the silicon core is important – especially ensuring that the doping is uniform. Secondly, it is essential that there is good ohmic contact between the graphite fixtures and the silicon core. The quality of the furnace installation also plays a role. Additionally, it is advisable to subject the silicon core to acid treatment. During production, the airflow must be even, current fluctuations should be kept to a minimum, and the temperature should be maintained within an appropriate range

1. The temperature we set is too high; 1130 is too high indeed. It has been proven that this temperature is excessive. In the past, the process lasted 24 hours, but now there are cases where it only lasts 8–14 hours. Therefore, it would be appropriate to lower the temperature a bit. Personally, I think 1080 would be sufficient. Based on measurements of the external temperature, the actual temperature is about 40 degrees higher than that of the chassis; thus, the temperature at the base of the silicon rod is around 1040. This temperature is more suitable for the growth of the root portion. In the early stages of polysilicon production, I believe it’s important to first stabilize the root area, while also paying attention to the crossbeams. We need to be aware of how the temperature is distributed throughout the chamber. You might ask why Furnace No. 7 uses these same parameters yet does not experience any issues. I analyzed this issue using infrared temperature readings. Every infrared thermometer has some degree of error, and if the thermometer in Furnace No. 7 happens to show a difference of several dozen degrees, it confirms that my previous judgment was correct. So, the set values aren’t fixed for every furnace. 2. The level at which the silicon core is installed is also a factor that can cause the rod to tip over; even a small deviation may lead to a larger deviation under pressure, resulting in the rod collapsing due to the increased weight once deposition begins. 3. It is necessary to reduce the initial rate of current increase, with the daily increase not exceeding 450A; in other words, the feeding speed of the TCS system needs to be adjusted downward in the initial stage. 4. Attention must be paid to fluctuations in pressure related to the amount of material fed. There are fluctuations in pressure in the TCS high-pressure tank when it reaches 7004, so it is important to activate the adjustment system for this tank as soon as possible. 5. Care must be taken to ensure proper contact between the root part and the graphite components, in order to avoid the formation of defects. 6. When drawing silicon cores using 5009, focus should not be solely on quantity; it is better to produce fewer batches while ensuring quality, paying attention to aspects such as uniform mixing and acid washing at each step of the process. 7 During these frequent periods of disruption, the commander must remain calm and composed, seeking answers amidst all the complex factors! :handshake

It can be said that there are many reasons for rod reversal in reduction furnaces; aside from current fluctuations, what are the other main causes? Please share your opinions! Thank you

To be specific, the reasons for reversal in the early stage are definitely different from those in the later stage