From owner-chemistry@ccl.net Wed Apr 18 17:28:01 2007 From: "Rachel Crespo rachel++fq.uh.cu" To: CCL Subject: CCL:G: PCM calculations Message-Id: <-34069-070418172430-22878-9BGHuwkY1fveVVdk5+YRjQ+*+server.ccl.net> X-Original-From: "Rachel Crespo" Date: Wed, 18 Apr 2007 17:24:25 -0400 Sent to CCL by: "Rachel Crespo" [rachel.:.fq.uh.cu] Dear CCL members, Which are the units of density and dielectric constant in the input of PCM calculations in gaussian? Best regards Rachel From owner-chemistry@ccl.net Wed Apr 18 20:53:01 2007 From: "Sengen Sun sengensun*yahoo.com" To: CCL Subject: CCL: molecular orbital analysis Message-Id: <-34070-070418203042-30943-48W9KCwLwo8ihTM1VWrEAQ[*]server.ccl.net> X-Original-From: Sengen Sun Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=iso-8859-1 Date: Wed, 18 Apr 2007 16:30:29 -0700 (PDT) MIME-Version: 1.0 Sent to CCL by: Sengen Sun [sengensun * yahoo.com] I admire Dr. Hultin's realistic view on this issue. In fact, Dr. Hultin made very insightful contributions to this discussion in the last 2 years. > But if we just treat this "big discussion" (or a bloody war?) as an isolated case, we are short-sighted. This engagement would have been essentially pointless to me. > I am really encouraged by someone (I cannot remember who) who asked a very important question several months ago on CCL: What are challenging problems in computational chemistry similar to challenging problems in organic synthesis? To me, this person has a truly inquiry mind. > This is the issue concerning the whole picture of comp. and theo. chemistry. In addition to singing pretty songs and writing beautiful poems to praise what we already achieved, our comp and theo chemists must ask at least the following two questions: > What are our ultimate goals and tasks in the centre of fundamental science? > What are current challenges in order to meet the goals and to fulfill the tasks? > As every one knows and as Popper pointed out, the majority of new organic chemical reactions do not proceed as expected. Theoretical guidelines in syntheses of new compounds remain very scarce in the modern era of advanced computational technologies. As far as to April, 2007, organic and medicinal chemists at bench rarely take computational approaches to rationalize their synthetic protocols, and fail frequently in the executions of new chemical reactions as THEY DID A CENTURY AGO! The reality is that the current computational tools to design and to predict chemical reactions are far from satisfaction and lag far behind synthetic advancements. > Facing such tremendous challenges, our computational and theoretical chemists have no reason to be indulged in any spiritual comfort without deep understanding of the physical behaviors of electrons and nuclei in the covalent bond-formation and -breaking. If we keep the challenges in mind, it is meaningless to argue in what sense "orbitals" mean "explanation" or "understanding". > In order to meet our big goals, we have no other choices than understanding physical or kinetic interactions between electrons and nuclei in the bond formation and breaking, so that our future generations can develop better tools to design and predict. Our younger generations must not be confined to the frames and thinking we are now creating for them. > It is true that even hundreds of years later, our offspring cannot predict, design, or explain EVERYTHING. But a sure thing is that they will do much better based on a better physical understanding of the nature. Sengen --- Phil Hultin wrote: > There is actually a kind of ¡°chicken-and-egg¡± > issue > here: is the bond longer/shorter because of orbital > influences, or do we observe different orbital > populations because changing nuclear separations > alter > electronic energies? > > We had a big discussion on CCL a while back about > the ¡°reality¡± of orbitals, and I DON¡¯T want to > re-ignite that. However, whenever we are talking > about molecular structure we should think about what > it means to ¡°explain¡± something. My view is that > molecular orbital theory can ¡°rationalize¡± > experimental observations and that this is in one > sense an ¡°explanation¡± but I would hesitate before > implying causation. > > > > There is also the question of whether ¡°bonds¡± > should > be regarded as physical entities at all, but ¡­ > > > > Dr. Philip G. Hultin > > Professor of Chemistry, > > University of Manitoba > > Winnipeg, MB > > R3T 2N2 > > hultin : cc.umanitoba.ca > > http://umanitoba.ca/chemistry/people/hultin > > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam > protection around > http://mail.yahoo.com > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From owner-chemistry@ccl.net Wed Apr 18 21:31:00 2007 From: "Wai-To Chan chan::curl.gkcl.yorku.ca" To: CCL Subject: CCL: Molecular orbital discrepancy in transition metal DFT calculations Message-Id: <-34071-070418212615-15638-Pq5mXeo240Axvs+i3eHBCg]![server.ccl.net> X-Original-From: Wai-To Chan Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii Date: Wed, 18 Apr 2007 21:28:42 -0400 (EDT) MIME-Version: 1.0 Sent to CCL by: Wai-To Chan [chan[A]curl.gkcl.yorku.ca] <<<<<<<<<<<<<< Dear All: I have been calculating geometry optimizations with DFT for some Molybdenum III complexes in a doublet ground state and I have found that the energy of the SOMO orbitals is always lower than the energy of the HOMO and some times even lower in energy than HOMO 1. When the calculations are carried out with the symmetry off the results are the same. The complexes with multiplicity of quartet also present some inconsistencies in the molecular orbital's energy ordainment. I have been made the optimization calculations in Jaguar with BP86 and BHandHLYP functionals. Does anyone have encountered this problem? It would be very helpful if someone could provide me a reference about this. Thank you >>>>>>>>>>>>>>>>>>>>>> I can think of two explanations of your problem. Incorrect relative energy levels of the SOMO and HOMO could be a sign of SCF convergence difficulties. In such situation I will experiment with different initial guess orbitals. There is no rule on how one can arrive at the right initial guess. For open-shell pi-electron systems for instance I may look into swapping the virtual orbitals with the SOMO or HOMO to change the number of pi electrons in the initial guess. With Guassian I also have the option of invoking the STABLE=OPT keyword in the hope forcing convergence to the ground electronic state. DFT is known to be less than reliable for the characterization of high-valent transition metal complexes. For instance see "Iron(IV) Porphyrin Difluoride Does Not Exist: Implications for DFT Calculations on Heme Protein Reaction Pathways" Ghosh & Taylor J. Chem. Theory Comput. vol 1 pg 597 (2005) According to this study DFT predicts Fe(IV) as the ground state which is incorrect--the correct spin state is Fe(III). A number of other studies have also reported similar limitations of DFT but I don't know if high spin Mo have been covered. I also doubt the reliability of your choice of BHandHLYP for organometallic systems. From my experience it seems to show better SCF convergence capability than other functionals but is less accurate in structural predictions. Wai-To Chan