I haven't experimented with Geom=connectivity before but I believe it needs a special input. It looks like you had a cycle of  SCF optimization done, the convergence criteria was not met but since there was no Opt keyword it stopped there.   Once it gets close to it's convergence criteria (1e-8 default) the structure will likely look much more plausible.

I hope this helps.

Do an optimization instead of single point energy. More over the level of theory which u r using is not sufficient to give the equilibrium geometry of ur molecule. Just choose a higher level methods with improved basis set which give reliable results. don't worry abt others things. Finally i think u r downloaded this file from some database and performed a calculation, u did nt take much care about valency of any atom. for example each C contains four bonds, if it is not satisfied add hydrogen to satisfy them as shown in the attached file, use no of processors and mem according to u........GOOD LUCK    

I examined the output file of Gaussian. I found that many atoms are missing. The correct formula is C7H15Cl2N2O2P, the molecule has MW=261.09 and the CAS number is 3778-73-2. I will send you more details about the calculation later.

1. I found the SMILES format of the molecule:O=P1(OCCCN1CCCl)NCCCl

2. I opened the program Avogadro for Linux. I use the version 1.0.3 for OpenSuse 12.2. The last version is 1.1.0 for OpenSuse 13.1.

3. I select Build, Insert, SMILES. I put from 1 the formula. The structure of isophosphamide appeared in three dimensions.

4. Then I selected Extensions, Optimize Geometry.

5. Molecular Mechanics, Setup Force field. I selected MMFF94.

6. Conformer search. I found 972 conformers. From View I selected Conformer properties and a Table appeared with all conformers. I put the conformers in order of increasing energy. I send you the conformer which I consider with minimum energy. (-697.44 kJ/mol).

I will continue later with ab initio calculation with Gaussian 09 for Linux. I will send you all the output files. I suggest you to install Avogadro and Gabedit. With these programs you can visualize very well the results.

[email protected]

Dear Dr.Charis Semidalas

I am trying to build a molecule in Gaussian 03 windows version. But the bonds are broken in the ring. So i collect mol file from the websites and give the same for calculation. But as you said, it contains only 14 atoms. all the 15 hydrogen atoms are missing in the mol file. 

Thank you for your answer and help sir.

First, build the molecule correctly. Use the complete molecular structure (available many places) and build it in GaussView. For example, in your structure you have P-N distances of ~0.8 A! This is not correct and suggests that the problem is as much with your general knowledge of chemistry than with GaussView! Second, submit the correct input file to Gaussian - when it is optimized, visualize it again in GaussView. Third, realize that GaussView is just an art program, it doesn't know any chemistry - it draws lines based on interatomic distances, not on any chemistry. So if a bond is shorter than a certain threshold, it may draw a "double bond" - likewise, if a bond is longer than a given value, it may draw no bond! YOU must interpret (based on chemical principles) whether this is really a double bond or a broken bond or just an artifact of the art program.

I selected the conformer with the minimum energy (iso_697.cml) for ab initio calculation  in Gaussian 09 for Linux. I present some distances I found: P2-O3=1.6253 A, P2-O1=1.4828 A, P2-N7=1.6833 A, P2-N11=1.6935 A. All the rest calculations are written in the file iso_697.out. Also, I send the figure of optimized molecule by the method B3LYP and basis set 6-31G(d). The input file for Gaussian is iso_697.txt. In order to run the job change the extension to com.

I choose the key Freq=Raman in order to find the frequencies. All the vibrational frequencies are real, meaning that none are imaginary, the point that Gaussian converged to is a local minimum on the potential energy surface.

The only way to feel comfortable that you have a global minimum is to do enough calculations to convince yourself that you have explored the potential energy surface.

Dear Dr. Charis Semidalas

Thank you very much for sending input and output files.