NMR predictor couldn't calculate charged species?

Hello, I really like MarvinSketch (compared to ChemBioDraw), mostly because NMR predictor is so useful to me. But sometimes NMR predictor failed for charged species, like the one below. When I clicked CNMR or HNMR prediction for this structure, MarvinSketch just automatically closed itself without producing any error message or asking if you want to save the previous structures. Is it a bug or have I done anything wrong here? Thanks!

Hi,

we have tried to predict 1H and 13C NMR spectra for your test molecule with a few Marvin versions but we were unable to reproduce this error. Please find attached the prediction reports in pdf. Which version have you used for NMR prediction?

If you dowload Marvin 5.11.3 (or use Marvin applet, see http://www.chemaxon.com/marvin/sketch/index.php), you should be able to run NMR for your molecule.

Best regards,

Csaba

Hi Csaba,

Thank you very much for your quick response! I tried the molecule in other computers (32-bit, windows XP) and it worked well, perhaps my computer (64-bit, windows 7) has some conflicts between java and marvin. I'll reinstall java and marvin.

It seems like the NMR predictor doesn't give very good prediction for conjugate resonance systems: if the charge center was switched, then the result changed a lot. I saw there's a discussion between you and Hans-Ulrich, so hope the problem can be sovled in future. Again thank you very much for your help!

Sincerely,

Ying C.

Hi,

the reinstall might help. If not, it is worth checking your Java verison number. The problem related to conjugate systems will be solved in the future.

Best regards,

Csaba

Dear Csaba,

I must comment here once more about mesomeric resonance structures:

As already shown in the topic NMR-prediction problems for mesomeric resonance systems

MarvinSketch 5.11.3 gives different results for two mesomeres, here for this imidazolium ion.

I think Marvin should block those systems before giving wrong results.

Best regards, Hans-Ulrich

Dear Hans-Ulrich,

thank you for the hint, we will consider it in the future. Now we have tasks having higher priority than this issue.

Best regards,

Csaba

Dear Csaba,

do you have really tasks with higher priority ?

It's simply wrong to get different results for a molecule or ion drawn in mesomeric resonance formulas. Both resonance formulas represent "virtuality", to use this modern phrase. And in the example discussed here there is only one cation having singular properties.

Best regards, Hans-Ulrich

Hi Csaba and Hans-Ulrich

I reinstalled both Marvin and Java (the version is right) and MarvinSketch works great for most tasks, but it still automatically closed itself whenever I tried NMR predictor for charged conjugate systems, so it really block itself before giving wrong results on my computer:) Anyway, thanks for your help~

Best regards,

Ying

Hi Ying,

I found the "not opening window effect", too. And it's not really reproducable. And it's on different machines with different Linux versions and different Java versions. Sometimes the CNMR calculation runs some time and then there is no new window showing the results, it simply stops. But sometimes, doing first a HNMR calculation there is a window  showing HNMR and then doing a CNMR calculation there is the CNMR window, too. How to understand ?

There must be a problem in the handling of opening new windows.

Best regards, Hans-Ulrich

Dear All,

we are examiming this issue. I will let you know as soon as we have found the error.

Best regards,

Csaba

Dear Csaba,

thank you for your answer. And in the next days I will try to find the conditions for opening the CNMR-windows and the conditions for not opening. May be I will be able to reproduce it.

Best regards, Hans-Ulrich

Dear Hans-Ulrich,

thank you for your help. We have been unable to reproduce this error so far.

Best regards,

Csaba

Dear Csaba,

one sequence was reproducable: Drawing a larger molecule f.e. Chrysen C18H12 in MarvinSketch 5.11.4 and then using CNMR, the calculation starts and the little window shows the continuation. At the end the little window closes and nothing more. The window showing the spectra is not opened.

But starting now the calculation once more, it runs and now the new window opens and shows the spectrum and the CNMR data.

It seems to me, it's a problem of timing inside the program. Is it possible the window object is called before the calculation has finished ? As a JAVA programmer I had sometime such timing problems for parallel objects. Is there a check if the calculation has finished before opening the new window ?

So I use the workaround: Starting the calculation simply once more.

Best regards, Hans-Ulrich

PS: It seems, the bug is not associated with charges.

Dear Hans-Ulrich,

thank you again for your help. I have tried to reproduce this bug for the chrysene molecule with Marvin 5.11.4, but no problems arose. However, you gave us a good hint, and my colleague, who implemented the NMR GUI, will have a look at the GUI code.

Best regards,

Csaba