I'm assuming that you mean that there is no S-S bond shown in GaussView. In that case, yes, you are correct - the distance is simply larger than GV is expecting for a single bond line to be drawn. You can use the "bond distance" menu to add the bond back for display purposes. There is no way to avoid this - you, however, know more than GV does, so put the bonds where you know chemically that they are. By the way, this has nothing to do with zero imaginary frequencies, it's just the way GV draws molecules.

in my opinion your suggestion about the bond broken is correct. check the S-S bond length from the literature and compare them with one you achieved from Gaussian output. removing the bond between sulfur atoms is attributed to the default graphic of software. in few cases you can add bond in your final output. I suggest after adding bond, indicate the bond length in your article or state it in other places which you want to discuses about your work. check other answer to ensure about my suggestion. 

I'm assuming that you mean that there is no S-S bond shown in GaussView. In that case, yes, you are correct - the distance is simply larger than GV is expecting for a single bond line to be drawn. You can use the "bond distance" menu to add the bond back for display purposes. There is no way to avoid this - you, however, know more than GV does, so put the bonds where you know chemically that they are. By the way, this has nothing to do with zero imaginary frequencies, it's just the way GV draws molecules.

Great, thank you both!!

Dear Maria,

The geometry optimization and, in particular bond lengths, is dependent on the used method/basis set. No imaginary frequency is sufficient to ensure that your structure is in the minimum (local or even global minimum) on the potential energy surface (PES) without considering this fact that some bonds are shown or not by a graphical visualizer program such as Gauss view. Note that if optimized bond length is longer than that of defined in the visualizer program as default,  corresponding bond is not shown. You can try with other method/ basis set, e.g, B3LYP using larger basis set of 6-311++G**. This try may meet your expectation. As mentioned, focus just for the absence of any imaginary frequency for your optimized structure and never worry  for the bonds which may not shown.

Best,   

Dear Maria,

You can add that bond using viewerlite from Materials Studio suite. Regarding S-S bond, Raman spectrum is the best way to characterize it. S-S bond will not be active in IR spectrum.

Best,

Dear Maria,

The bond order S-S is broken in the Gaussian output due to the lowest bond energy. The S-S bond energy is 265 kJ/mol. We can confirm its virtual presence in the FT-Raman spectrum. you can find the vibrational band at 715 cm−1 corresponds to S-S bond. 

Best wishes

by

T.karthick

Dear T. Karthick,

I was expecting the S-S band located approx. at 512 1/cm. One report seems to support this value.

Szafranski, C. A.; Tanner, W.; Laibinis, P. E.; Garrel, R. L. Surface-Enhanced

Raman Spectroscopy of Aromatic Thiols and Disulfides on Gold Electrodes.

Langmuir, 1998, 14, 3570-3579.

Dear maria,

Open your .out file using chemcraft, if you are able to see the bond between S-S it means bond is present between them and in gauss view alone is not visible. But if the bond is absent it means there is bond breaking occurs in ur molecule and you have to recompute them and reoptimize them. I have came accross same problem and chemcraft is the best one to identify us the problem.

Best wishes....

You can check that S-S single bond really exists.

If you already submit input file with frequency calculation, you can check vibrational mode

of S-S,

The bond  may be broken  if you cannnot find S-S translational mode in output file

Also, it is nesseary to check the moelcular hybridization number in sulfur atom using NBO

calculation. 

Hope this help.

Thank you very much all for your answers. I added polarization functions on my basis set and the bond lengths (for S-S and S-C bonds) decreased significantly and  visualized by gaussview. We now have an X-Ray structure for the molecule am studying computationally, so I can compare the experimental with the theoretical bond lengths, and indeed when I used simply the 6-31G or 6-311G basis sets (the ones that  gaussview didn't visualize the S-S bonds)  we had a huge error on the S-S bond lengths (up to 10% error), while when we add polarization functions we had an %error of 1 to 2. Thanks for the idea for NBO calculation, I didn't use it before, but once there is free time I will have a go (just to see)!

Dear Maria,

Nbo analysis will help you to find the charge transfer occur within the molecule it may help you to choose or change your polarization function. Kindly not the natural population analysis and hybrids obtained from the nbo table, it may help you a lot.

Thanks.......