<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;"><DIV>Dear Prof. Blaha,</DIV>
<DIV>Thank you very much for your message. Now I can see what the problem was. For a typical implantation experiment, the ratio of the implanted ions to host atoms is about 1:1000000 (one is to a million or more). I cannot run the code with such a big supercell. So far we have done calculations with supercells consisting of 4 Cu atoms and one Be atom and also with 32 Cu atoms and one Be atom. In the case of larger supercell (consisting of 32 Cu atoms and one Be atom), we have seen that the electron density at beryllium nucleus is decreasing after force minimization compared to the smaller supercell, but it is still more than what we are getting for similar calculation with aluminum supercell. So you are suggesting that if we can continue with bigger and bigger supercell, then we shall see the effect of the higher electron affinity of copper and then the electron density at the Be nucleus would be larger when Be is implanted in aluminum compared to
copper. </DIV>
<DIV> </DIV>
<DIV>Since these would be very big calculations and may not be doable for us, so is there any simpler way to do such calculations for such very dilute impurity implantation? Can I qualitatively say that the host atoms with higher electron affinity will take away more s-valence electrons from 7Be compared to another host of lower electron affinity? Can I say that if the Cu is placed closer to 7Be, then it will take away more s-valence electrons than if it is placed further away? </DIV>
<DIV> </DIV>
<DIV>Earlier we were doing similar calculations using TB-LMTO(Stuttgart code) without making any supercell. But we were always finding that for higher electron affinity host atoms, the s-valence electrons in 7Be muffin-tin radius was decreasing and interpreting this result as a loss of electron density at the Be nucleus. If we would make the lattice size smaller, then the loss of s-valence electrons would increase further. Assuming that the 2s electron density at Be nucleus is proportional to the number of s-valence electrons in Be muffin-tin sphere, we could understand a lot of implantation data in a rather quantitative way. </DIV>
<DIV> </DIV>
<DIV>I was wondering if there might be any such simpler method for WIEN2K calculations? </DIV>
<DIV> </DIV>
<DIV>With best regards</DIV>
<DIV> Amlan Ray</DIV>
<DIV>Address</DIV>
<DIV>Amlan Ray</DIV>
<DIV>Variable Energy Cyclotron Center</DIV>
<DIV>1/AF, Bidhan Nagar</DIV>
<DIV>Kolkata - 700064</DIV>
<DIV>India</DIV>
<DIV> </DIV>
<DIV> </DIV></td></tr></table><br>