Yes you can identify the coordination bond of M-L by IR spectra in the range 200-600 cm-1 and by the calculation of Nephlauxetic paramater(beta) via UV-Visible spectra of complex solution

Hi, Good question. Are you using NBO5? In the second order perturbation interaction table you will have interactions with the Pt center. You should sort them by energy. From that list you need to identify donating interaction from Platinum to the bonds of interest. You will have pi type and sigma type, depending on what your ligands are.

I wouldn't just sum these energies, but rather look at each interaction and make a rational choice of which interaction is important to you.

Finally compare the donation from Platinum to a donation to platinum.

Please remember that all interaction in NBO analysis are stabilizing, so you are asking which is more stabilizing.

Also, make sure to check the results. I have encountered weird energies doing a very similar analysis with platinum in the past. To my memory these involved very large interaction energies with rydberg states. If this happens I would guess there is some convergence problem with the NBO analysis and I do not know how to solve it.

Dear Jarowski

Thanks for the reply

I'm using NBO 3.1 in gaussian 09 package.

Yes,  I understood. But my interaction is splitted into more contributions, example:

...

                                                                  E(2)      

61. LP ( 3)Pt 1     /248. BD*( 1) P 3 - C 7 2.73 0.45 0.031

61. LP ( 3)Pt 1     /249. BD*( 1) P 3 - C 8 3.26 0.45 0.034

62. LP ( 4)Pt 1     /248. BD*( 1) P 3 - C 7 0.87 0.48 0.018

62. LP ( 4)Pt 1     /250. BD*( 1) P 3 - C 9 4.01 0.49 0.040

...

I have to make the sum of these energies or I use the more representative?

Because I want the stabilization energy between Pt and P3-Cn*

I didn't understand the part of the Rydberg states. If I have a high stabilization energy between Pt and Rydberg states this means that there is a convergence problem with the NBO analysis? Example:

....

                                                           E(2)

59. LP ( 1)Pt 1     / 76. RY*( 7)Pt 1 32.74 2.69 0.266

59. LP ( 1)Pt 1    / 78. RY*( 9)Pt 1 43.15 10.29 0.597

59. LP ( 1)Pt 1    / 79. RY*( 10)Pt 1 47.10 15.23 0.759

59. LP ( 1)Pt 1    / 80. RY*( 11)Pt 1 12.41 0.81 0.090

59. LP ( 1)Pt 1    / 81. RY*( 12)Pt 1 49.16 27.81 1.048

59. LP ( 1)Pt 1    / 82. RY*( 13)Pt 1 26.45 1.53 0.180

59. LP ( 1)Pt 1    / 83. RY*( 14)Pt 1 39.55 5.20 0.406

59. LP ( 1)Pt 1 /    84. RY*( 15)Pt 1 13.91 1.61 0.134

...

Thank you very much for your help.

D.V.

By high I mean unreasonably high! Like hundreds of kjs. This data looks ok.

If you want to sum because no one interaction stands out, that is fine, but I would confine your summation to sigma, like you have shown. For any others group them in terms of chemistry (sigma, pi) then consider the type of interaction. Sometimes its a bond interacting with a bond, other times a LP interacting with a bond. Try to discover what matters. A lot of these numbers are just noise, so you will need to be statistical about it but also use your intuition. For example, you might have 10 interaction with a sub kcal/mol energy. If you sum those are you really getting something meaningful out of it? Its a hard game.

Just an aside, but I have never understood how a lone pair on Pt1 can interact with a Rydberg state on Pt1!! Unless someone else has an explanation, I would ignore these.

Dear Jarowski

 I understood. You have made clear my doubts.

The analysis of this data is not easy and requires a good experience.

I appreciate the time and effort you have spent sorting out that problem.

D.V.

You may also check the "charge displacement" method in http://pubs.acs.org/doi/abs/10.1021/ja0772647

I confirm that the Charge Displacement method works very well with coordination bonds (please see link). This method will be implemented in ADF soon.

http://pubs.rsc.org/en/content/articlelanding/2016/sc/c5sc02971f#!divAbstract

Dear Daniele,

Your best bet is vibrational spectroscopy as far as the experimental methods are concerned. Theoretical approach these days can predict/show almost anything, but the ultimate truth lies in the experiment. Typically we need to have a sample in which you would not expect to see much of the pi-bonding. These cases are abundant in the organometallic chemistry of Pt-ethylene [and its substituted derivatives]. Here we see true pi-bonding with a significant component of back-donation was clearly juxtaposed to the metallo-cyclopropane way of binding. If interested in experimental details - ask for more and I will provide you with a stack of references about both types of complexes. In this literature pile you will be able to see data of vibrational spectroscopy from which you can make a valuable assessment of the pi-bonding/back donation quite similar to those done for the phosphines binding (Tolman's cone angle).

Nick

Dear Daniele, 

Besides to the cited methods in this topic, another very interesting one is the CDA (Charge Decomposition Analysis) developed by Gernot Frenking. Ref: S. Dapprich, G. Frenking J. Phys. Chem., 1995, 99, 9352-9362

This method has been used in several computational investigation involving transition metal compounds (including some systems containing Pt).

Julio 

Mr. Veclani,

Please pay attention to attached equation showing the stabilization energy (E(2)). This energy term, in fact, provides information about population of donating orbital and energy difference between two orbitals. As it is given in shown (attachment) an example of p-p* transition. So that when you would like to evaluate M-to-L CT or L-to-M CT effects that you should evaluate only the E(2) of orbitals of M and L, involved in M-L bond formation (not all energies). But if you would like to perform a quantitative evaluation how much is this contribution comparing with other transitions, for example, studying an electronic absorption spectrum of a complex where along with CT bandy you have other set of bands associated with intra-ligand (between ligands, ligand to solvent molecules) and intra metal ion transitions, than you should take into consideration all energies.

Please pay as well as attention to a computation model of an NO molecule without metal ion (NBO analysis), showing E(2), Fij, Ei, Ej and more (level B3PW91/SDD), and accounting for only s, p and n-contributions. Or those are orbital contributions which in terms of metal-organics, particularly PtII complexes in your case, describing only a set of orbitals, are assciated with MLCT or LMCT effects.

In other words E(2) gives information about interacting orbitals, generally, without to be a specific characteristic for MLCT or LMCT effects.

Theory can provide details that experiment can not. A properly designed experiment is of course always a good idea, but reference molecules introduce all kinds of confounds. If you are interested in your specific molecule and don't have the time or background or equipment to become a spectroscopist and to do more synthesis, theory will enliven and elucidate your question in a way that is transferable to others. As an experimental/computational hybrid myself I have deep respect for the two approaches, but one must mix them in the right proportions. In this case, more experiment may simply lead to more questions and more theory when the question you currently have is fairly straightforward. 

By the way, if you have access to NBO5 (I have not used it) I believe you can turn on and off certain interaction and re-optimize the structure. This may reveal what interactions are most important for the bond you are interested in.

Dear all

Thank you for time.

I've already done the CDA analysis, which has given good results.

I'd like to support the CDA analysis also with NBO analysis.

I have studied these systems only from a computational point of view.

 I thank you again.

D.V.

If you don't want to be too much sophisticated for the complex you mention (I guess you wish to estimate the backdonation into the phosphine), you can perform a single point of the already optimized species, after its reorientation. You must have the metal exactly lying at the origin and the Pt-P bond aligned with the x axis  From the fraction of missing electron pair in d-xy (but also in xz because the phosphine is a double face acceptor), you can guess the amount of pi backdonation (provided that there is no other competing acceptor ligand in the plane). If you make a comparison with PF3, you may be surprised of the difference.

You can use single crystal X-ray to get the bond length of  Pd-M  and use it  as indication of double .