The README File

Analytical potential energy surfaces for O(³P) + C₂H₂ and O(¹P) + C₂H₂ reactive systems.

Keywords: Analytical potential energy surface, O+C2H2, triplet surface, singlet surface, quasi-classical trajectories

Kindly submitted to CCL by: Sophya V. Garashchuk from The University of South Carolina

Codes with data, libraries and instructions for compilation of analytical potential energy surfaces for O(³P)+C₂H₂ and O(¹P)+C₂H₂ reactive systems. They may be used as examples for quasiclassical trajectory dynamics. The surfaces are constructed in full dimensionality and are permutationally invariant as described in Ref. [1]

References:

Permutationally invariant potential energy surfaces in high dimensionality. B. J. Braams and J. M. Bowman. Int. Rev. Phys. Chem. 28 (2009), pp 577-606.
Analytical potential energy surface for O+C2H2 system. Sophya Garashchuk, Vitaly A. Rassolov, Bastiaan J. Braams, Chem. Phys. Lett.
Electronic population inversion in HCCO/DCCO products from hyperthermal collisions of O(³P) with HCCH/DCCD. S. Lahankar and J. Zhang and S. Garashchuk and G. C. Schatz and T. Minton. J. Phys. Chem. Lett. 4 (2013), pp 1315-1321.

Instructions

Download the pes.tar.gz (a compressed tar archive) to your $HOME directory. The archive contains a top mod subdirectory that contains the Fortran90 sources and example data. The Makefiles for Intel Fortran Compiler ifort (see, for example: http://software.intel.com/en-us/non-commercial-software-development and http://software.intel.com/en-us/articles/using-intel-compilers-for-linux-under-redhat-enterprise-linux-or-centos) are also included. These Makfiles assume that your files reside under $HOME/mod but of course, you can edit them if you want to place the distribution in some other directory. You will need the ifort compiler to run these examples. Then uncompress the archive as:
```
gunzip pes.tar.gz
tar xvf pes.tar
```
This will create a mod subdirectory. You should see the following files there:
```
cd mod
ls -l
...
-rw-rw-r--. 1 2404     1002 20071424 2013-07-02 12:33 inv-pes-xyz-090129.pax
-rw-rw-r--. 1 2404     1002 94034944 2013-07-02 12:33 pes-gr-090210.pax
-rw-r--r--. 1 2404     1002    13603 2013-07-02 15:12 pes_h2c2o1s.f90
-rw-r--r--. 1 2404     1002    13603 2013-07-02 15:12 pes_h2c2o1t.f90
-rw-rw-r--. 1 2404     1002     1029 2013-07-02 15:49 README
-rw-r--r--. 1 2404     1002     6220 2013-07-02 15:05 t0s.xyz
-rw-r--r--. 1 2404     1002     6200 2013-07-02 15:05 t0t.xyz
-rw-r--r--. 1 2404     1002     1060 2013-07-02 15:05 test_h2c2o1_st.f90
```

Untar both .pax files (these are uncompressed tar archives with sources and Makefiles):

cd $HOME/mod
tar xf inv-pes-xyz-090129.pax  # this creates inv and pes-xyz subdirectories
tar xf pes-gr-090210.pax       # this creates pes-gr-090210 subdirectory

The compilation requires lapac libraries. You may need to modify the Makefile to adjust location of your libraries, include files, and compiler location. Assuming your system knows how to compile with lapack do something like:
```
cd $HOME/mod/inv
make realclean
make libinv.a
```

Repeat compilation in the other directory:

cd $HOME/mod/pes-xyz
 make realclean
 make lib

The steps above should produce three libraries: inv/libinv.a, pes-xyz/libpes.a and pes-xyz/libpx.a. Only the libpes.a needs to be linked for the compilation below.
The libraries can now be used to compile and run the test code. Compile the file test_h2c2o1_st.f90 (it takes long time 10 min to hours), with command similar to:
```
cd $HOME/mod
ifort test_h2c2o1_st.f90 -Lpes-xyz -Linv -I. -I./inv -I./pes-xyz -lpes -o test.x
```

Run energies of points in t0s.xyz or t0t.xyz as:

cd $HOME/mod
./test.x pes-gr-090210/h2c2o1s-avtz-rccpt-fit pes-gr-090210/h2c2o1t-avtz-rccpt-fit t0s.xyz

that uses the t0s.xyz as an example.

File List for potential_energy_surface_O_C2H2 Directory

README [2kB] : README file with instructions
pes.tar.gz [31287kB] : code with libraries

The stuff below is not a part of the original distribution!!!

The CCL admin also tried to compile and run the sources under the CentOS 6.4 (64 bit) on his laptop (oh well, it is an i7 laptop with gobs of memory and ain't cheap) but without the Intel compiler, using the open source GNU stuff. These are good sources, since they compiled without a glitch. Since the admin does not do chemistry anymore, he even did not look at the results (they may be TOTALLY wrong since the compilation was very untidy), but this is what he did after unpacking the pes.tar.gz :

# as root
yum install gcc-gfortran makedepf90 lapack-devel lapack

# as a humble Joe Schmoe user:
cd $HOME/mod/inv
cp Makefile Makefile.orig
emacs Makefile # see these very quick and dirty edits
export TOPDIR=$HOME/mod
make realclean
make libinv.a
cd ../pes-xyz
emacs Makefile # see these very quick and dirty edits
cp ../inv/inv_wp.mod ../inv/inv.mod  .  # copied some modules
make realclean
make lib
cp libpes.a ..  # copied the library to the top directory
cd ..
f95 test_h2c2o1_st.f90 -Lpes-xyz -Linv -I. -I./inv -I./pes-xyz -lpes -o test.x
./test.x pes-gr-090210/h2c2o1s-avtz-rccpt-fit pes-gr-090210/h2c2o1t-avtz-rccpt-fit t0s.xyz > the_stdout.txt

and some results were collected in the the_stdout.txt file.