Covalent/Ionic Character of Polar HF bond

[zhangx...@gmail.com]

Hey, guys.

I am a beginner to VB theory, and now need for your kind comments and help.

Questions came when I obtained the dissociation curve of the polar HF molecule, and both the covalent H-F and the major ionic H+F- structure. The VB calculations were perfomed at the L-VBCISD/6-31+G** level, and the wavefunction of the single VB structure was seperately optimized. The minumum of the covalent and ionic curve are -56.2 and -58.8 kcal/mol, respectively. From these results, it seems that the H-F bond has more ionic character, less covalent character.

But, from the hamiltonian matrix and weight of the full VB calculation, the weight of covalent VB structure is larger than 0.5, and I thought that the covalent VB structure should contribute more to the bonding than the ionic VB. But now I was puzzled at the weights of covalent and ionic component, since the above seperate wavefunction optimization suggests a more ionic character.

Any advices and comments are appreciated. A sample of input file with HF at a distance of 0.91 Angstrom is provided.


XMVB Input xmi file
HF molecule,3 structures
$ctrl
frgtyp=sao orbtyp=hao str=full
iscf=2 vbcisd ncor=1 guess=read
nae=2 nao=2
itmax=2000
$end
$Frag
1 1 1 1
spz 1
spzdxxyyzz 2
pxdxz 2
pydyz 2
$end
$orb
1 1 1 1 1 1
2
2
3
4
2
1
$end

Xiaoyong Zhang
KU Leuven

[应富鸣]

Dear Xiaoyong,

Very interesting.

I think you should have noticed that the orbitals of your calculations with different structures are different from each other.

In VBSCF and VBCISD calculations including all structures, covalent and ionic structures share the same set of orbitals which are different from the calculations they are optimized separately.

It means that the orbitals for all structures are not the best for a single structure, which puzzled you.

I tried a BOVB  calculation in which the orbitals in different VB structures are treated as different orbitals and optimized separately. Following are the weights for different calculations:

VBSCF:

1     0.58681  ******  1:4    5   6

2     0.39832  ******  1:4    5   5

3     0.01487  ******  1:4    6   6

VBCISD:

1     0.56297  ******  1:4    5   6 

2     0.41345  ******  1:4    5   5 

3     0.02358  ******  1:4    6   6 

BOVB:

1     0.55572  ******     1   1   2   2   3   3   4   4   5   6

2     0.35836  ******     7   7   8   8   9   9  10  10  11  11

3     0.08592  ******    12  12  13  13  14  14  15  15  16  16

You may find that the weight of covalent structure is even smaller than in VBCISD. Considering that BOVB can be treated as VBSCF + part of VBCIS, the weight of covalent structure could be even smaller than BOVB IF a VBCISD with different orbitals in different structures is proceeded.

I think it may help you.

Best regards,

在 2018年4月3日星期二 UTC+8下午7:45:30，zhangx...@gmail.com写道：

[zhangx...@gmail.com]

Dear fuming,

Very thanks for the helpful comments!  

Regards