0
$\begingroup$

I am trying to rigid scan two bond distances simultaneously in Gaussian 09/16. For example, I have taken a water molecule and started a rigid scan. Below is my input.

# b3lyp/6-31+g(d,p) scan

Title Card Required

0 1
 O              
 H                  1            B1
 H                  1            B2    2            A1

   A1           109.50000006

B1 0.96 S 20 0.1
B2 0.96 S 20 0.1

The output scan plot is this:

enter image description here

It is a 3D plot of 441 (21 steps for H1 atom * 21 steps for H3 atom) grid points. I want both the H atoms to move simultaneously, so I get a 2D plot of only 21 points. Below is the type of plot I want for scanning 2 H atoms. Any help with the input?

enter image description here

$\endgroup$
4
  • $\begingroup$ I'm not sure, but it's certainly easy enough to try: use a single variable for both bond lengths. $\endgroup$ Mar 25, 2023 at 1:12
  • $\begingroup$ That seems fine, though. However, I would be thankful to you if you please share any sample input. $\endgroup$
    – Pro
    Mar 27, 2023 at 9:29
  • 1
    $\begingroup$ In your z-matrix substitute B2 with B1 and delete the last line before the mandatory blank line. $\endgroup$ Mar 28, 2023 at 16:56
  • $\begingroup$ it works, thanks $\endgroup$
    – Pro
    Mar 30, 2023 at 14:32

1 Answer 1

0
$\begingroup$

From the obvious hints @Martin provided, it became easy. We can simultaneously scan only one coordinate. So, for this particular problem the input should be as follows,

# hf/3-21g scan

Title Card Required

0 1
 O              
 H                  1            B1
 H                  1            B1    2            A1

   A1           109.50000006

B1 0.96 S 20 0.1

Explanation: Since bond the bond lengths of OH are same, one has to put only B1 parameter in the z-matrix and vary that parameter. But, for a special case, if $OH_1\neq OH_2$ then this will not work. I will look for more general case and post that as update here.

$\endgroup$
2
  • 1
    $\begingroup$ In my opinion don't bother looking for a more general case. Gaussian interface is not flexible enough for that. There is possibly an easier way to generate a series of xyz coordinates programmatically with python of so and run single points on these. For future questions like these, I recommend looking at Matter Modeling. $\endgroup$ Mar 30, 2023 at 16:46
  • 1
    $\begingroup$ Oh... I forgot, you might have luck with GIC. But in all honesty, learning python is probably easier... $\endgroup$ Mar 30, 2023 at 16:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.