There are many problems where we want to calculate the binding energy between two systems.

Normally, we have a system A, adsorbent (a surface, a nanotube, a protein, etc.), that interact with a system B, adsorbate (atoms, ions, molecules, etc.). The procedure is to calculate the energy of both systems separately $E_\text{A}$, and $E_\text{B}$ and to calculate the energy of the complex $E_{\text{AB}}$, then the binding energy can be calculated as:

$$ E_{\text{bind}} = E_{\text{AB}} - E_\text{A} + E_\text{B} \tag{1} $$

In case DFT calculations using basis set, we also need to apply the basis set superposition error (BSSE) correction (Molecular Physics, 1970, VOL. 19, NO. 4, 553-566).

The question is: what are the fundamentals of equation (1)?

  • $\begingroup$ You don't apply the basis set superposition error, but rather try to remove it. $\endgroup$ – Susi Lehtola Jun 18 '20 at 9:06
  • $\begingroup$ @SusiLehtola of course! I added "correction". It is ok now? $\endgroup$ – Camps Jun 18 '20 at 10:25
  • $\begingroup$ Yes, but it's also not specific to DFT but applies to any basis set calculation. $\endgroup$ – Susi Lehtola Jun 18 '20 at 11:14

The justification is simple and comes from a very fundamental law of thermodynamics: Internal energy is a complete differential form and is independent of intermediate states and only depends on start and final states: $$\Delta U = U_{2} - U_{1} $$

In your case: $U_{2} = E_{AB}$ and $U_{1} = E_{A} + E_{B}$ and if $\Delta U$ or binding energy $E_{\text{binding}}$ is positive, it means the start state is thermodynamically preferred but if $\Delta U$ is negative, it means the binding state is thermodynamically preferred. You can generalize it to more complex systems with more than two reacting systems or components.


In general, a system composed of $K$ interacting subsystems have a potential energy at a specific configuration of its parts. For instance, a system of $M$ nuclei and $N$ electrons can be separated into interacting subsystems with internal geometries, having $\{\mathbf{R}_A\}$ as nuclear positions for subsystem $A$ with $N_A$ electrons, $\{\mathbf{R}_B\}$ and $N_B$ for subsystem $B$, and so on. With all the remainder coordinates accounting for the distance between these subsystems, we can write:

$$ \mathbf{R}=\mathbf{R}_{\text{int}}+\mathbf{R}_{\text{ext}}=\sum_{A}^{K}\left[\sum_{a\in A}\left(\mathbf{R}_a+\sum_{b\in B}\mathbf{R}_{ba}\right)\right] $$

where $\mathbf{R}_a$ is the position of a nucleus $a$ within the subsystem $A$ (internal coordinates) and $\mathbf{R}_{ba}$ is the distance vector from a nucleus $b$ in the subsystem $B\neq A$ to the nucleus $a$.

According to this, we can write the total energy of the complex as one-system terms, two-system terms, and so on.

$$ E(\{\mathbf{R}\})=\sum_{A}^{K}E_A(\{\mathbf{R}_A\})+\frac{1}{2}\sum_{A}^{K}\sum_{B\neq A}^{K}E_{AB}(\{\mathbf{R}_B-\mathbf{R}_A\})+\dots $$

where the energy is adjusted to zero (subtracting the energy of the isolated subsystems from $E$) at infinite separation of the subsystems. Then, we can define a potential energy of interaction between the subsystems as

$$ \Delta E(\{\mathbf{R}\})=E(\{\mathbf{R}\})-\sum_{A}^{K}E_A(\mathbf{R}_A)=\frac{1}{2}\sum_{A}^{K}\sum_{B\neq A}^{K}E_{AB}(\{\mathbf{R}_B-\mathbf{R}_A\})+\dots $$

Now, this interaction energy can be calculated at any $\{\mathbf{R}\}$, however, most of the time we want to calculate the interaction energy at optimized geometries $\{\mathbf{R}^{\text{(opt)}}\}$:

$$ \Delta E=E(\{\mathbf{R}^{\text{(opt)}}\})-\sum_{A}^{K}E_A(\mathbf{R}^{\text{(opt)}}_A) $$

and this happen to be the binding energy with respect to a specific dissociation route from equilibrium.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.