Crystallization from melt

Discussion in 'General Electronics Chat' started by EngIntoHW, Jul 1, 2011.

  1. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi,

    I'm learning about CZ (Czochralski) crystal growing technique, which is a Liquid-Solid monocomponent growth technique by pulling a monocrystal seed from the melt.

    Say that you got a melt doped with Borun.
    The more you pull the monocrystal seed from the melt, the concentration of the Borun in the crystal is getting larger.

    So eventually, the crystal is not uniformly doped.

    Could anyone please explain why is this non-uniformity?

    Thank you.
     
  2. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    I know absolutely nothing about the Czochralski process, but a little Googling told me that the nonuniformity is due to segregation. I also know nothing about crystal formation, but I suspect that crystals abhor impurities, and your job in pulling the crystal is to adjust all the parameters at your disposal (rotation, temperature, magnetic field, pull rate, etc.) to coerce the crystal into incorporating the dopant into its structure at a rate that is as constant as you can make it.

    I could be totally wrong.
     
    EngIntoHW likes this.
  3. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi Ron,
    Thank you very much for the excellent post :)

    I was actually familiar with the Segregation coefficient, but I think I just now came to realize it.
    If the Seg Coeff. was constant during the entire pulling - not dependant on the growth rate - then we would have gotten a uniform doping across the crystal, right?
     
  4. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,392
    1,605
    The Segregation coefficient is constant, it's a constant. Its the melt that's changing.

    As Ron_H said "crystals abhor impurities" so as the crystal is pulled and it grows it is leaving behind impurities in the melt. Some impurities get pulled, but a much lower concentration in the pull then in the melt.

    Thus the concentration of impurities in the melt is rising as we pull more material from the melt.

    And if we pull from a higher concentration of impurities then we include more of them.
     
    EngIntoHW and Ron H like this.
  5. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Excellent clarification, Ernie.:)
     
  6. someonesdad

    Senior Member

    Jul 7, 2009
    1,585
    141
    The technique is also called zone refining. I remember going to a seminar to it in the late 60's put on by a guy I knew in the chemistry department. The chemists use it for a variety of things.
     
  7. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi,
    And thanks for your help :)


    As you can see below, the Segregation Coefficient is not constant, and it depents on the Growth Rate.
    As can be seen in the graph, the impurities concentration in the solid crystal increases as the Growth Rate increases.
    I thought that that was the reason for the non-uniformity of doping in the crystal formed.

    [​IMG]
     
  8. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi mates,

    I was wondering, how can we get a uniform concentration in the crystal? (when we complete its creation).

    Say that the Si melt is doped with Boron, which has a Ko of 0.8 (which is pretty close to 1).
    If we decrease the Boron's concentration in the melt (by adding to the melt more and more Si) during the pulling of the crystal, would we get a uniform Boron concentration in the crystal eventually?

    Thanks.
     
Loading...