Reaction barrier of Pd-catalyzed reaction

moiumuio

Hi everyone,
I am calculating the reaction barrier of a Pd-catalyzed reaction. The process I would follow on Gaussian software is:
1. Optimize+frequency calculate with PBE0-D3(BJ) function and basis sets of SDD for Pd and Br and 6-31G** for other atoms. The temperature was set at 383 K.
2. SP energy is calculated with PBE0-D3(BJ) function and basis sets of SDD for Pd and Br and 6-311+G** for other atoms. The temperature was set at 383 K, and toluene was used as a solvent.
Is my procedure that logically good, or are there any matters you would change?
Thank you

sobereva

1 SDD for Pd is OK, but SDD for Br is too poor, because there is no d-polarization function. For convenience, you can simply use def2-SVP for all atoms, which is usually adequate for opt freq purpose.

2 Using def2-TZVP is much more recommended. Note that single point energy is independent of temperature, this point was discussed in http://sobereva.com/423

moiumuio

sobereva 发表于 2024-4-16 04:04
1 SDD for Pd is OK, but SDD for Br is too poor, because there is no d-polarization function. For con ...

I have therefore updated my path to:
1. Optimize and frequency calculate with the PBE0-D3(BJ) function and basis sets of def2-SVP for all atoms. The temperature was set at 383 K.
2. SP energy is calculated with the PBE0-D3(BJ) function and basis sets of def2-TZVP for all atoms. Toluene was used as a solvent.
Do I need diffuse function for SP energy calculation?
And if my computer is not fast enough for the DEF2 basis set, should I reduce it to Pople and those Lanl/SDD basis sets?

sobereva

moiumuio 发表于 2024-4-16 10:28
I have therefore updated my path to:
1. Optimize and frequency calculate with the PBE0-D3(BJ) fun ...

Diffuse function is not needed.
def2-TZVP is never expensive. However if you really find you cannot afford it, you can use def-TZVP, which is much cheaper. Pople series of basis set of 3-zeta level was not well constructed, its price/performance ratio is poorer than def/def2 series.

There is no basis set named Lanl, it is a series, including such as lanl2MB, lanl2DZ, lanl08, lan2TZ, lanl08(d), lanl2TZ+ and so on.

Note that it is not "function", but "functional"

moiumuio

sobereva 发表于 2024-4-17 04:18
Diffuse function is not needed.
def2-TZVP is never expensive. However if you really find you cann ...

Yes, sir. I will take care of it.
Def2-ECP will only be used for Pd and not for Br if I use def2-SVP for all atoms. Am I right?

sobereva

moiumuio 发表于 2024-4-17 12:23
Yes, sir. I will take care of it.
Def2-ECP will only be used for Pd and not for Br if I use def2- ...

If you simply write keyword such as # PBE1PBE/def2SVP opt freq, then you do not need to specifically set def2-ECP, because in this case Gaussian will automatically employ def2-ECP for all elements beyond the fourth row.

moiumuio

sobereva 发表于 2024-4-18 07:06
If you simply write keyword such as # PBE1PBE/def2SVP opt freq, then you do not need to specifical ...

Greeting Prof.,
Should I include a solvent model (Toluene in my case) in the optimization step if pivalic acid is used as a catalyst? Toluene is not a highly polar solvent; moreover, my experiments have shown that adding water reduces the reaction's efficiency. Am I okay to omit the solvent model from the optimization step?

sobereva

moiumuio 发表于 2024-5-3 00:40
Greeting Prof.,
Should I include a solvent model (Toluene in my case) in the optimization step if ...

You can omit it since toluene is nonpolar