Error in running MEPGS calculation

From NWChem

Viewed 1882 times, With a total of 4 Posts
Jump to: navigation, search

Clicked A Few Times
Threads 8
Posts 20
Hello,

I am trying to run MEP calculation on a simple ethylene oxide molecule. 
I tried running it in Forward, Backward as well as default (both) directions
So i start the job from coordinates of optimized TS structure with negative imaginary frequencies and then run MEPGS
For Forward as well default calculations, the GS OPT starts (step0) but after 200 iterations fails to converge...which should be the starting structure (TS already optimized)
For the backward calculation the energy increases and gives me an energy convergence failure...
Since I have ran the same molecule with another software (so I know the IRC works)....I am not sure what part is not working ....please explain and suggest changes


Here is the input
start TS_Eth_h2o2_B3LYP_IRCb

  1. geo from M06-L

geometry units angstroms
C     0.70240430     1.02854192     0.01900758
C -0.42716827 1.85861995 0.04102502
H 1.29500754 0.97410043 -0.89356705
H 1.30159049 0.91820617 0.92428135
H -0.92398310 2.14543825 -0.88229354
H -0.91822796 2.11267884 0.97925056
O -0.00444306 -2.26509742 0.06769202
H -0.23814281 -2.41577937 -0.85742490
H -0.58224747 -0.47936882 0.76067322
O -0.16938925 -0.35319650 -0.10605228
end

basis
  • library aug-cc-pvdz
end

Driver
Maxiter 100
end

dft
XC B3LYP
Maxiter 100
grid xfine
end

task dft Freq

freq
reuse TS_Eth_h2o2_B3LYP_IRCb.hess
end

mepgs
maxmep 300
maxiter 200
inhess 2
noxyz
evib 0.0005
stride 0.01
Backward
mswg
end

task dft mepgs


error from forward and default calculations:
---------------------------------------------------------------
Failed to converge in maximum number of steps or available time
---------------------------------------------------------------


 Step       Energy      Delta E   Gmax     Grms     Xrms     Xmax   Walltime
---- ---------------- -------- -------- -------- -------- -------- --------
@ 200 -230.12944427 3.8D-12 0.00053 0.00018 0.00000 0.00000 20803.0
                                             ok       ok       ok

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



------------------------------------------------------------------------
mepgs: gsopt failed 0
------------------------------------------------------------------------
There is an error related to the specified geometry
------------------------------------------------------------------------

Error from backward calculation:

         Expected decrease in energy              0.000500

         Obtained decrease in energy             -0.000533

-------------------------
The energy has increased
-------------------------
------------------------------------------------------------------------
dft mepgs failed 0
------------------------------------------------------------------------
This type of error is most commonly associated with calculations not reaching convergence criteria


All the help is greatly appreciated

Thanks,
Pansy

  • Niri Forum:Admin, Forum:Mod, NWChemDeveloper, bureaucrat, sysop
    Profile
    Send PM
Forum Regular
Threads 4
Posts 207
Hi Pansy,

There are some issues with the MEPGS code. My suggestion would be to use the NEB or String approach. Both these approaches give you the minimum energy pathway and transition state. They are also efficient.

Best,
-Niri

Gets Around
Threads 24
Posts 86
How do we know that the initial configuration is a saddle point ? Only in this case it is possible to use the MEPGS code and Forward, Backward displacements lead to a decrease in energy.
So firstly optimize saddle more carefully, and run the calculation again with
Quote:username

mepgs
maxmep 300
maxiter 200
inhess 2
xyz
evib 0.0005
stride 0.01
Backward
mswg
end

if the 200 steps will not be enough, look *.bxyz file, and increase the stride or maxiter, so total displacement (stride*maxiter) will be enough to achieve a product or reactant geometry, because stride is a measure of the distance between adjacent geometries in * bxyz file, measured in sqrt(amu)*Bohr.
Edited On 3:23:31 AM PST - Sat, Nov 29th 2014 by Vladimir

Clicked A Few Times
Threads 8
Posts 20
Quote:Niri Nov 26th 5:23 pm
Hi Pansy,

There are some issues with the MEPGS code. My suggestion would be to use the NEB or String approach. Both these approaches give you the minimum energy pathway and transition state. They are also efficient.

Best,
-Niri


Thanks .... I will try the string method

Pansy

Clicked A Few Times
Threads 8
Posts 20
String method for three geoms
Hello
I am running string calculation to get mep with three input geom
I was able to run it successfully with freezing the first and the end geoms

Now I am trying to run the same calculation by first freezing those geoms and then relaxing it
also I added the HASMIDDLE which allows the code to use the the "midgeom" geometry to be used as the midpoint (in this case the midgeom is a Transition state)
part of the input with string module
string
 nhist 10
nbeads 10
maxiter 20
stepsize 0.10
print_shift 1
HASMIDDLE

 # don't allow the end points of the path to move
freeze1 .true.
freezeN .true.
end
task dft string ignore

string
 nbeads 20
maxiter 20

 # allow the end points of the path to move
freeze1 .false.
freezeN .false.
end
task dft string ignore


the job runs for two beads and then ends with an unspecified error:
"


zts_meps: gradient failed        0
------------------------------------------------------------------------
This error has not yet been assigned to a category
------------------------------------------------------------------------
For more information see the NWChem manual at http://www.nwchem-sw.org/index.php/NWChem_Documentation"

Please suggest suitable changes

Pansy


Forum >> NWChem's corner >> Running NWChem



Who's here now Members 0 Guests 1 Bots/Crawler 0


AWC's: 2.5.10 MediaWiki - Stand Alone Forum Extension
Forum theme style by: AWC