This is one of my input files:
start
title "name"
geometry
load ironoxide.xyz
end
basis
Fe library ahlrichs_vtz
H library 6-31G*
C library 6-31G*
N library 6-31G*
O library 6-31G*
F library 6-31G*
S library 6-31G*
Cl library 6-31G*
end
driver
xyz geo
maxiter 1000
end
geometry adjust
zcoord
end
end
charge -3
dft
xc b3lyp
iterations 1000
mult 2
vectors atomic output fe.mo
end
task dft optimize
unset dft
charge -1
dft
xc b3lyp
mult 2
iterations 1000
vectors fe.mo
end
task dft optimize
ironoxide.xyz
2
geometry
Fe 0.00000000 0.00000000 0.38403467
O 0.00000000 0.00000000 -1.24308515
I get the least spin contamination after optimizing with charge -3 and changing to charge -1. However, the spin contamination increases during optimization and slight distortions in the input geometry (~0.01 Å) changes increases the spin contamination even more.
Expectation value of S2:
<S2> = 1.7552 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.7562 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.7568 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.7572 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.7573 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.7573 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.2067 (Exact = 0.7500)
Expectation value of S2:
<S2> = 0.7767 (Exact = 0.7500)
Expectation value of S2:
<S2> = 0.8908 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.1086 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.2432 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.2975 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.3333 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.3400 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.3407 (Exact = 0.7500)
Expectation value of S2:
<S2> = 1.3408 (Exact = 0.7500)
Are there any suggestions on how I can decrease the spin contamination?
|