PC GAMESS/Firefly-related discussion club

Learn how to ask questions correctly

Igor Polyakov
polyakoviv@gmail.com

Frankly speaking it was kind of hard to understand your message, that is why i will post some general comments:
1) Yes, you can determine the heat of formation of LiH from the reaction:
Li+H=LiH and yes E(LiH)-(E(Li)+E(H))=HeatofFormation(LiH)
For the E(LiH) you should use the LiH equilibrium geometry, for the E(Li)+E(H) you can do 2 separate calculations(i think in this case this is a preffered option) or make 1 calculation with a large distance between Li and H.
2) You can compare accuracy of different methods but you must use size consistent methods, for example Hartree-Fock, coupled cluster, many-body perturbation theory (to any order), and full configuration interaction (CI)
3)In order to compute the heat of formation more accurately you will need to compute hessian and do the thermochemical analysis for the given temperature.
4)In order to save computer time and get the very accurate result for the energy calculation people designed several computational schemes such as G1, G2 and etc, for example:

the G2(MP2) method involves the following steps: (copypasted from one of the QC books)
(1) The geometry is optimized at the HF/6-31G(d) level and the vibrational frequencies
are calculated.To correct for the known deficiencies at the HF level, these are
scaled by 0.893 to produce zero point energies.
(2) The geometry is re-optimized at the MP2/6-31G(d) level, which is used as the reference
geometry.
(3) An MP2/6-311+G(3df,2p) calculation is carried out, which automatically yields the
corresponding HF energy.
(4) The energy is calculated at the QCISD(T)/6-311G(d,p) level. This automatically
generates the MP2 value as an intermediate result, and the difference between
the QCISD(T) and MP2 energies is taken as an estimate of the higher order correlation
energy. The G2 method (not G2(MP2)) performs additional MP4 calculations
with larger basis sets to get a better estimate of the higher order correlation
energy.
(5) To correct for electron correlation beyond QCISD(T) and basis set limitations, an
empirical correction is added to the total energy, ΔEemp = −0.00481Na − 0.00019Nb,
where it is assumed that the number of b electrons is larger than or equal to the
number of b electrons. The numerical constants are determined by fitting to the
reference data. It should be noted that this correction makes the G2 methods nonsize
extensive.

I suppose, u get the general idea.

Glad if my post helps,

Best regards, Igor.


On Wed Oct 28 '09 3:31pm, Dominic P. Guaņa wrote
------------------------------------------------
>i am computing the heat of formation by first runtyp=energy for the energy of individual atoms.... adding them up then subtract with the energy computed from the their most stable geometry. Optimization, hessian at B3LYP/TVZ (pd) level... i want to test the accuracy of the theory level i am using. I tried this with LiH,

>This is my Formula, calculation at B3LYP/TZV (pd) level

>Heat of Formation = E(LiH)-(E(Li)+E(H))

>gives me -58.17 kcal/mol.

>HELP with this. THanxs in advance, i dont know if my formula is correct or will give an accurate result for the accuracy of the Theory Level.

>Thanxs in advance.