Effect of the ratio of distillation products B and P on the resistivity of polycrystalline silicon of type N

When the levels of boron and phosphorus in the product are similar (high compensation), the resistivity is very high. However, to look back at the ratio of boron to phosphorus in trichlorosilane, it is necessary to know the respective \"conversion efficiencies\" of B and P during deposition, that is, the relationship between their contents in trichlorosilane and their contents in the final product.

Reply to 2# missyou4ever: When the levels of B and P in the product are equal, pore formation and electron compensation occur, which results in variations in the resistivity measured for phosphorus. Even though both the concentration of donor impurities and that of acceptor impurities is low, the resistivity of the P-type material turns out to be particularly high, and the material becomes PN-type instead. At present, most entities that use polysilicon still rely on N-type and P-type resistivity as criteria for assessing quality; even if the impurity concentration in PN-type polysilicon is low, they consider it unqualified, and explaining this is rather complicated. It is indeed complex to achieve the desired product quality through control of raw materials and manufacturing processes; I believe this is something that every person working in the polysilicon industry strives for. I hope everyone will work together in practice, and I also hope that we can share our achievements and improve together.

People in silicon are experts. Since no post-treatment of polysilicon has been carried out, the final quality of the polysilicon is not well understood. I have been looking for factors that affect the quality of polysilicon in order to consider options related to the distillation of trichlorosilane; I was wondering if you could offer some guidance. My email address is qingyanqi@126.com. Thank you.

Reply to 3# Silicon Expert: It seems you’re using the phosphorus-based or boron-based method. I don’t think it’s necessary to tell them that it’s of the PN type; after crystal growth, if the subsequent region is P, then report it as P, and if it’s N, report it as N. The grading criteria only specify the contents of donors and acceptors. When they grow single crystals, they still just check the resistivity at the shoulder area, and this approach doesn’t work for materials with high compensation. A high shoulder resistance rate is a necessary condition for a good product, but not a sufficient one. Do those who come to buy materials not bring along people who understand the technology? With silicon supplies so tight these days, the upstream silicon manufacturers hold all the power; those who buy the material from them don’t dare to be picky. If they get too insistent, the manufacturers just ignore them, haha. Also, I think that at present, a production volume of not two or three thousand tons cannot be considered satisfactory. If you have low-temperature infrared equipment, you’ll know right away after processing a single crystal B/P sample; otherwise, boron testing is extremely troublesome. How many batches can you test for boron in a day? When we are busy, we run 5-6 furnaces per day, and results are required by the next day. Without low-temperature infrared technology, how many people would be needed to produce single crystals? . I think low-temperature infrared is still necessary. As for the relationship between P/B in trichlorosilane and P/B in the final product, it is not something that ordinary people can figure out, as not everyone has access to all the data from the raw materials to the finished product. If it’s possible to gain access and conduct extensive statistics, it’s not an impossible task. Dude, if possible, add me on QQ so we can chat when we get the chance. I’ll send it to you via QQ internal messages.

There are sensible people upstairs. This is indeed difficult for ordinary chemical workers.