The new version of MultiWell is now available on the website:

http://aoss-research.engin.umich.edu/multiwell/

This version includes several enhancements and corrects several minor bugs. Always check for updates and use the latest version of the software! 

(My apologies if you receive multiple versions of this message!)

NEW IN VERSION 2014 (January 2014):

1. MULTIWELL: Bug Fix: when densities of states are sparse (e.g. for small molecules at low energies), tunneling from one Well to another was allowed to take place, even when tunneling into an energy grain that did not contain a state (an "empty grain"). In this event, the stochastic trial did not terminate correctly. This problem has now been addressed by forbidding tunneling into an empty grain and issuing a warning in the rate constant tables in the output files. However, note that this pragmatic "fix" does not address the more fundamental problem that wells connected by tunneling should possess states in matching energy grains. Fortunately, the effects of this problem are small, except at very low energies in very small molecules: it does not occur at all in models that do not contain empty grains. Probably, very few users have been affected. If more information is needed, contact [log in to unmask].

2. MULTIWELL: Improvement: the format of output file "multiwell.dist" was revised to make it more convenient for users, although the file is now larger than before.

3. MULTIWELL: Improvement: For convenience, the temperature dependence of the collisional energy transfer parameter ALPHA has been revised to the following form:

Alpha = [C(1) + E*C(2) + E*E*C(3)]*(T/300)**C(8)

Previously, the constant term, C(1), was not multiplied by the temperature-dependent factor. Furthermore, the temperature-dependent factor was changed by replacing T with the ratio T/300. All of the temperature dependent factors have been changed analogously. The revised form is expected to be more convenient for users.

4. DENSUM: Bug-fix: A minor bug prevented calculation of sums and densities of states using the KRO degree of freedom. This bug has been fixed. Since this degree of freedom is not employed routinely, this bug probably did not affect any users.

5. THERMO, MOMINERT: improvement: in both of these codes, it is now an option to use user-defined filenames with extension ".dat" for the data file (e.g. "C2H5.dat") in place of the default names (defaults: "densum.dat" and "mominert.dat"). The output files take the user-defined file name with extension ".out." (e.g. "C2H5.out") in place of the default names (defaults: "densum.out" and "mominert.out").

6. ADENSUM: Bug Fix: ADENSUM can be used with THERMO to compute thermodynamic data and canonical TST rate constants for fully vibrationally-coupled anharmonic molecules and transition states. It does this by passing a data file (with the ".qvib" suffix), which contains partition functions, zero point energy, etc. When separable degrees of freedom (e.g. hindered rotors) are included in the ADENSUM calculation, their zero point energies should have been added to the anharmonic zero point energy, but were not. This omission has now been fixed.

7. gauss2multi: revision: When using gauss2multi to create data files for the MultiWell codes, Users must specify pressure units, which are used only for creating "multiwell.dat". The thermo.dat data file requires specifying the standard state, but TOR pressure units are not an allowed standard state and previous versions would give error messages when creating thermo.dat. The gauss2multi code has been revised so that thermo.dat is created with "BAR" as default when TOR units are specified by Users (thus avoiding unnecessary error messages). Other data files created by gauss2multi are not affected; in particular, multiwell.dat can still be created with TOR pressure units, when they are specified.

8. In the User Manual and in the output files from the various codes, the hindered rotor degree of freedom that was previously described as an "Unsymmetrical hindered rotor" (hindered rotor type "hrd"), is now described more properly as a "General hindered rotation". This change was made because the hrd type could always be used for any 1-D hindered internal rotation, symmetrical, or unsymmetrical, although it is particularly useful for the latter.