From: Chitrak Gupta (
Date: Tue Sep 15 2015 - 10:47:52 CDT

Hi Van and Christopher,

I recently faced a problem at this stage of FFTK (although I don't remember
whether it was the exact same problem).

What was happening is, FFTK was asking for the *.chk file of the
optimization step, but when generating the *.gau input file, it had
"hess.chk" as the starting point. I don't know if this is a bug in FFTK (I
was planning on asking this question later once I finished the entire
process). I think you will need to verify whether that %chk line refers to
a file that ***already exists* in the same folder. If not, you can manually
change it to the filename of the optimization checkpoint file.

Hope this helps,

On Tue, Sep 15, 2015 at 6:07 AM, Mayne, Christopher G <>

> The Gaussian job is terminating with a FileIO error, which means that it
> can't access a particular file that it is looking for. One thing to note
> about the way ffTK sets up the Hessian calculation is that it will pull
> geometry information from the checkpoint file output from the geometry
> optimization (this is why you have to specify it in the input). ffTK makes
> a copy of this checkpoint file and names it using the same user-provided
> filename for writing the Gaussian input (GAU) file (i.e. if hess.gau is the
> output file, the checkpoint file will be copied to hess.chk). From the
> data you've provided, this file is named hess.chk. If you are running the
> Gaussian job elsewhere, you will need to also copy this CHK file along with
> the GAU file. Further, I have also run across instances where job
> submission scripts on a cluster will copy the input file to a compute node
> to perform the calculation. If the CHK file isn't copied to the compute
> node as well, you will generally get!
> a similar FileIO error message. This is solved either by modifying the
> submission script to copy the CHK file to the compute node as well, or
> modify the input GAU file to provide the full path to the checkpoint file--089e013cbc26dcdbd6051fcb1a48--