Unit of polarizability and meaning of the value

[timber]

I checked polarizability of H2O

   h2O_xyz = psi4.geometry('''
    0 1
    O 0 0 -0.11
    H 0 -1.4 1.2
    H 0 1.2 1.2
    ''')
_, wfn_stylene = psi4.properties(
    #'cc2/3-21g', molecule=h2O_xyz , return_wfn=True,
    #'ccsd/3-21g', molecule=h2O_xyz , return_wfn=True,
    'ccsd/6-31G(d)', molecule=h2O_xyz , return_wfn=True,
    properties=['polarizability'])

and information I got is

                 CC2 Dipole Polarizability (Length Gauge) [(e^2 a0^2)/E_h]:
                  0                     1                     2        
    0      7.349326778459673    -0.068101265084527     0.000000000000000
    1     -0.068101265084527     4.271260031707426     0.000000000000000
    2      0.000000000000000     0.000000000000000     0.782088084846100
	alpha_(0.000) =       4.134224965004 a.u.
 
                 CCSD Dipole Polarizability (Length Gauge) [(e^2 a0^2)/E_h]:
                   0                     1                     2        
    0      7.208682091476423     0.000000000000000     0.000000000000000
    1      0.000000000000000     0.775821679244794     0.000000000000000
    2      0.000000000000000     0.000000000000000     4.210846362700781
	alpha_(0.000) =       4.065116711141 a.u.

I do not undestant the unit of the values. On the other hand, according to the reference, The dipole polarizability of a water molecule (H2O) is typically represented by the symbol α0. The dipole polarizability of H2O is approximately 1.47~1.48 ų ( Untitled Document ). and Other paper say ~1.2 ų ( https://www.researchgate.net/figure/Dipole-Moments-Polarizabilities-and-Sizes-of-the-Molecules-Considered-in-This-Study_tbl1_259278978 )

My result is proper?

Another question is that Dipole Polarizability, here, means electronic Polarizability? ionic Polarizability? or orientation Polarizability ?

[ajaymk]

These numbers are in atomic units and need to be converted to ų. See Table 2 in this paper for conversion factors. To compare against experimental results, bigger basis sets are needed (see https://doi.org/10.1063/1.5051458). Using aug-cc-pvdz will give ~1.37 ų, which is close to the values you mentioned.

[timber]

Super thanks! I’ll test it!

[crawdad]

In case it’s helpful, here are some notes that I wrote on this topic and updated yesterday. It explains the conversion between atomic units and cgs for polarizabilities.