The CDA algorithm itself may also lead to this seemingly problematic data. Essentially, the CDA analysis employs Mulliken method to compute the compositions, a well known drawback of this method is that it may result in compositions larger than 100% and negative compositions. If the composition is slightly larger than 100%, you can view it as 100%; while if the value is slightly negative, you can simply view it as 0%.

Hi, I'm trying to decompose the orbital contribution of a molecule into its fragment.
If I assume that AB is the complex, while A and B is its moiety, What I did is

1. Get each molden file of AB, A, and B
2. use [16 Charge decomposition analysis (CDA) and plot orbital interaction diagram] to get CDA info, and then,
3. get contribution info from [2 Show fragment orbital contributions to specific complex orbital]

When do this, for some AB orbital, the contribution from A or B moiety orbital is even higher than 100% while its counter moiety's orbital gives a very large negative contribution.
Is there any physical meaning for this large contribution? or telling me the basis set (or ecp) is not good enough to proceed with the CDA process? (I'm using def2- double or triple zeta basis set with def2-ecp and the system has more than 50 heavy metals)

Thank you for the help!