Estimating Species Thermochemistry

Species thermochemistry can be estimated based on 2 approaches: 1. Benson group additivity (GA) 2. On-the-fly quantum-chemical methods based on the explicit 3D structure of the molecule.

For the on-the-fly methods, a number of QM packages [G03, OpenMopac, MM4] and a number of levels of theory [PM3, MM4, MM4HR] are supported. Since the accuracy of Benson GA for acyclic molecules is generally superior than for the QM methods, acyclic molecules are always estimated through GA.

For cyclic molecules 3 different strategies can be chosen: 1. BensonOnly 2. QMforCyclics 3. Hybrid

The BensonOnly strategy only uses the Benson Group Additivity framework to estimate thermochemistry of species. For compounds containing one or more rings, it identifies the smallest set of smallest rings in the molecules and perceives whether fused ring atoms are present, that indicate that a polycyclic ring system is present. Then, the databases of ring corrections and polycyclic ring corrections are used to search an adequate (polycyclic) ring correction. If an apt ring correction is not found, a warning is printed, but the thermochemistry without the ring correction is kept.

The QMforCyclics strategy preferentially choses the on-the-fly semi-empirical methods to estimate thermochemistry. If repeated failures occur for the QM methods, this strategy falls back to the BensonOnly strategy.

In the hybrid strategy, first, the Benson databases are searched for an adequate (polycyclic) ring correction. If such a correction is not found, this strategy falls back to the QMforCyclics strategy.

The thermochemistry calculated from on-the-fly quantum methods are now saved to $RMG_QM_LIBRARY, which defaults as QMThermoLibrary in the working directory. These results can be inputted into subsequent RMG runs by copying QMThermoLibrary from a previous run into the working directory of the new run.

The ThermoDataEstimator utility produces thermochemistry estimates without running RMG.

Instructions for Use

1. Create a text file named input.txt in any directory that you like. The text file should begin with a block to specify the desired Database.

   The next section allows to use on-the-fly generation QM techniques to estimate TD properties, rather than group additivity methods.

   The first flag defines the strategy to estimate thermochemistry: e.g.:

   //thermo strategy? [BensonOnly/QMforCyclics/Hybrid]
   Hybrid

   The following lines are not required if the BensonOnly flag is used:

   The QM package to be used, is specified: either choose “both”, “gaussian03”, “mopac”, “mm4”, or “mm4hr” e.g.:

   //QM method: [both/gaussian03/mopac/mm4/mm4hr]
   gaussian03

   Limit the use of on-the-fly methods to only species with less than the specified number of unpaired electrons. E.g. “0”, implies that radical species will not computed using on-the-fly QM methods. e.g.:

   //maxradnumforQM?
   0

   A check of the connectivity of the passed-in 3D structure can be performed ensuring that the quantumchemical data returned by the QM package is based on a 3D structure with correct connectivity. e.g.:

   //CheckConnectivity? [off/check/confirm]
   check

   Specify whether to keep all output files from Gaussian03 or Mopac. If no enviornment variable for qmCalculationsDir is set, the output files will be kept in $WorkDir/QMFiles.

   //KeepQMFiles? [yes/no] yes

   The next section allows the user to set atom constraints on the species to be processed by the ThermoDataEstimator, beyond the default values hard-coded in TDE. Only species that satisfy the set atom constraints, will be processed by TDE. Uncomment (//) the desired atom constraint, to set a particular constraint.

   E.g.:

   MaxCarbonNumberPerSpecies:     40
   //MaxOxygenNumberPerSpecies:     10
   //MaxRadicalNumberPerSpecies:    10
   //MaxSulfurNumberPerSpecies:     10
   //MaxSiliconNumberPerSpecies:    10
   //MaxHeavyAtomNumberPerSpecies: 100
   //MaxCycleNumberPerSpecies:      10
   END

   The next field should be the PrimaryThermoLibrary, which may be left empty if desired:

   Database: RMG_database
   
   PrimaryThermoLibrary:
   END

   Next, in the text file, create the adjacency list(s) corresponding to the molecule(s) of interest using the same syntax used to define a species in the condition.txt input file (see RMG manual). Hydrogens can be omitted for simplicity. For example, the adjacency list for 2,2,4,4-tetramethylpentane could be written as:

   1 C 0 {2,S}
   2 C 0 {1,S} {3,S} {4,S} {5,S}
   3 C 0 {2,S}
   4 C 0 {2,S}
   5 C 0 {2,S} {6,S}
   6 C 0 {5,S} {7,S} {8,S} {9,S}
   7 C 0 {6,S}
   8 C 0 {6,S}
   9 C 0 {6,S}

   An complete example can be found in the examples/ThermoDataEstimator directory.

2. At the command prompt, change directory to the location of the input.txt file and then, if you are using a unix-based operating system, execute the command

   $ java -classpath $RMG/bin/RMG.jar ThermoDataEstimator input.txt > RMG.log

   Or if you are using a Windows operating system you can instead run the batch file ThermoDataEstimator.bat. Either double-click it or at the command propmpt type:

   > ThermoDataEstimator.bat

The output of the program will automatically be written to RMG.log

3. The program will read RMG’s thermodynamics database, count groups, and output the NASA polynomials in CHEMKIN format, as well as another set of values labeled ThermoData. The format for ThermoData is

   ΔH_f,298 S₂₉₈ C_p,300 C_p,400 C_p,500 C_p,600 C_p,800 C_p,1000 C_p,1500

   Units are kcal/mol for ΔH_f,298 and cal/mol*K for the other quantities (entropy and heat capacity). The symmetry number (including contributions from internal rotations) will also be displayed.