ISYMGEN - Basis setsΒΆ

The basis set information is in the ISYMGEN file which is created when you run NRLMOL. This file contains the basis set for each identity atom. For each orbital of a given atom, same set of primitive Gaussians is used. For example, the default basis set for carbon is written as below :

The following is the Pederson-Porezag (NRLMOL default) basis in the ISYMGEN . The newer version allows use of few other basis sets such as : 6-31G*, 6-311G**, STO-3G, TZVP, DGDZVP etc. The Pederson-Porezag (NRLMOL default) is specially optimized for the PBE functional.

   6   6                       ELECTRONIC AND NUCLEAR CHARGE        
ALL                            ALL-ELECTRON ATOM TYPE  
   3                           NUMBER OF ATOMS OF TYPE CAR  
ALL-CAR001    
ALL-CAR002    
ALL-CAR003    
EXTRABASIS         CONTROLS USAGE OF SUPPLEMENTARY BASIS FUNCTIONS   
 12                NUMBER OF BARE GAUSSIANS   
   5   4   3       NUMBER OF S,P,D FUNCTIONS   
   0   0   1       SUPPLEMENTARY S,P,D FUNCTIONS  

       0.22213361D+05       0.33317370D+04       0.75790135D+03  
       0.21454372D+03       0.69924889D+02       0.25086135D+02  
       0.95910418D+01       0.38024557D+01       0.14891854D+01  
       0.57487653D+00       0.21494732D+00       0.77209650D-01  
  
       0.19792249D+00       0.36998977D+00       0.63644615D+00  
       0.10124931D+01       0.14480787D+01       0.17173689D+01  
       0.14931932D+01       0.68987161D+00       0.86072247D-01  
      -0.16566695D-02       0.37766033D-03      -0.47105343D-04  
 
      -0.45005260D-01      -0.84621052D-01      -0.14496564D+00  
      -0.23535601D+00      -0.34215368D+00      -0.44595124D+00  
      -0.45263971D+00      -0.32216414D+00      -0.12988420D-01  
       0.20135471D+00       0.12769913D+00       0.14135467D-01  

       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.10000000D+01       0.00000000D+00       0.00000000D+00  

       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.10000000D+01       0.00000000D+00  

       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.10000000D+01  

       0.23138630D-01       0.42649133D-01       0.74658851D-01  
       0.12024115D+00       0.18351176D+00       0.24706804D+00  
       0.30714219D+00       0.31372706D+00       0.26726340D+00  
       0.14756585D+00       0.47585576D-01       0.72796459D-02  

       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.00000000D+00       0.00000000D+00       0.00000000D+00  
       0.10000000D+01       0.00000000D+00       0.00000000D+00  


       ......   
       ......   
       ......

Here, the first line specifies the nuclear and electronic charges in the atom. The nuclear and electronic charges specifies the actual atom (e.g. Nuc charge = 6 for carbon) but the electronic charge depends on whether all electron or pseudo potential calculations are used (e.g. for carbon it would be 6 for all electron and 4 for a pseudopotential calculation)

The second line specifies the type of calculation :

ALL for all-electron BHS for BHS pseudopotential TAB for tabulated user-supplied pseudopotential

The third and fourth lines specify the number of total such atoms in the geometry and their symbols in the SYMBOL file. EXTRABASIS =1 in SYMBOL file will signal the program to use the supplementary basis functions. Then comes the number of primitive Gaussians followed by the number of the s, p and d -type contracted Gaussians. The number of supplementary functions of s, p and d type are written next. These informations are followed by some blocks of numbers. The first block lists the exponents of the primitive Gaussians. This is then followed by Ns blocks where Ns is the number of contracted s-type Gaussians. The first block is the coefficients multiplying the primitive Gaussians for the 1s contracted Gaussian, the second block is for 2s Gaussian and so on. After the Ns number of such blocks, comes the Np blocks corresponding to the p-type contracted Gaussians followed by similar Nd number of d-type Gaussians. In the example for carbon atom above, the 1s and 2s contracted Gaussian is a linear combination of the all the primitive Gaussians whereas the higher unoccupied s orbitals are taken as single long-range Gaussians. Similarly for p orbitals where only the 2p orbital is occupied. These are then followed by similar blocks corresponding to the supplementary functions. The supplementary functions are used only when

EXTRABASIS = 1 in the SYMBOL file. These are generally polarization functions and are usually only used for the calculation of IR and Raman intensities and possibly for dipole moments and polarizabilities.

The carbon atom basis set in the example might be called a “12-1211+++G(3d)” basis in Pople/Gaussian parlance if we consider the last s,p,d orbitals as diffuse functions. This means 12 primitive gaussians contracted to the core 1s orbital, then a triple-zeta valence set (three sets of s and three sets of p orbitals) represented by 12, 1 and 1 primitive gaussians, three sets of d polarization functions and three sets of diffuse functions (s,p,d) each represented by a single primitive gaussian. This is an example of a “generally contracted” basis in which the same set of primitives contributes to all the orbitals. If the diffuse functions are considered as valence, then this basis has a quadruple-zeta functions.