exercises:2014_ethz_mmm:basis_sets

**This is an old revision of the document!**

In this exericse you will compare different basis sets and use them for computing the binding energy of an H2 molecule.

The cp2k basis set format is the following:

Nsets n1 lmin lmax Nexp Ncontr exp1 contr1 contr2 contr3 ... exp2 contr4 contr5 contr6 ... Nexp.... n2 lmin lmax Nexp Ncontr exp1 contr1 contr2 contr3 ... exp2 contr4 contr5 contr6 ... Nexp.... ...

Nsets = number of sets.

n1,n2… = main quantum number (but it gets ignored by the program!)

lmin = min *l* (angular quantum number) of the set ( s = 0 ; p = 1 ; d = 2 …)

lmax = max *l* (angular quantum number) of the set ( s = 0 ; p = 1 ; d = 2 …)

Nexp = number of exponents of the set

Ncontr = number of contaction coefficients per angular quantum number

As an example:

2 2 Sets 1 0 0 1 1 Set 1: lmin=0 ; lmax=0 (→ s functions!) ; 1 exponent ; 1 contraction 0.35 1 exponent1 of set 1 ; contraction 1 1 1 1 1 Set 2: lmin=1 ; lmax=1 (→ p functions!) ; 1 exponenet ; 1 contraction 0.6 1 exponent1 of set 2 ; contraction

Save the following input file. Comment lines are marked with !

- mybasis.inp
&GLOBAL PROJECT H-mybasis RUN_TYPE ENERGY &END GLOBAL &FORCE_EVAL METHOD Quickstep ! Electronic structure method (DFT,...) &DFT LSD ! Requests a spin-polarized calculation for an odd number of electrons MULTIPLICITY 2 ! Multiplicity = 2S+1 (S= total spin momentum) &POISSON ! Solver requested for non periodic calculations PERIODIC NONE PSOLVER MT ! Type of solver &END POISSON &QS ! Parameters needed to set up the Quickstep framework METHOD GAPW ! Method: gaussian and augmented plane waves &END QS &XC ! Parametes needed to compute the electronic exchange potential &XC_FUNCTIONAL NONE ! No xc_functional &END XC_FUNCTIONAL &HF ! Hartree Fock exchange. In this case is 100% (no fraction specified). &SCREENING ! Screening of the electronic repulstion up to the given threshold. This section is needed EPS_SCHWARZ 1.0E-10 ! Threshold specification &END SCREENING &END HF &END XC &END DFT &SUBSYS &CELL ABC 10.0 10.0 10.0 PERIODIC NONE ! Non periodic calculations. That's why the POISSON scetion is needed &END CELL &COORD H 0.0 0.0 0.0 &END COORD &KIND H &BASIS ! Basis set for H 2 1 0 0 1 1 0.35 1 1 0 0 1 1 0.6 1 &END POTENTIAL ALL ! Specifes that the potential is for all electron calculations. &POTENTIAL ! Usual all eletcron potential for H 1 0 0 0.20000000 0 &END POTENTIAL &END KIND &END SUBSYS &END FORCE_EVAL

Try to change the basis set, and report the obtained energy values for H. After a couple of tries on your own, try to use some of the literature basis sets (reported at the end of this section). At the end, you should get a table like this :

Basis set | Energy (Eh) |
---|---|

mybasis (from given input) | …. |

basis try 1 | …. |

basis try 2 | …. |

…. | …. |

pc-0 | …. |

pc-1 | …. |

pc-2 | …. |

…. | …. |

Is always good to keep record of self-created basis sets, to track the effect of a change in value and number of exponents, contractions….etc..

Repeat the procedure for H2.

H2 coordinates:

H 0 0 0 H 0.8 0 0

The H2 molecule has an even number of electrons. Remember to take out the LSD and MULTIPLICITY keywords.

Binding energy:

What is the effect of:

- increasing/decreasing the value of the exponents for the given basis?
- adding sets with p,d symmetry to the basis? You have the same effect in H and H2?

exercises/2014_ethz_mmm/basis_sets.1395762214.txt.gz · Last modified: 2014/03/25 15:43 by sclelia

Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-ShareAlike 4.0 International