From chemistry-request #at# server.ccl.net Sat Jan 29 17:00:55 2000 Received: from ivory.trentu.ca (trentu.ca [192.75.12.103]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id RAA15181 for ; Sat, 29 Jan 2000 17:00:54 -0500 Received: from el1 ([204.225.13.50]) by trentu.ca (PMDF V5.2-32 #29543) with SMTP id <01JLA6D8Q7H0003C7D /at\trentu.ca> for chemistry /at\ccl.net; Sat, 29 Jan 2000 15:53:47 EST Date: Sat, 29 Jan 2000 15:57:27 -0500 From: Errol Lewars Subject: KCl CLUSTERS ANSWERS SUMMARY To: chemistry -8 at 8- ccl.net Message-id: <38935437.4F46 $#at#$ trentu.ca> Organization: Trent University MIME-version: 1.0 X-Mailer: Mozilla 3.01Gold (WinNT; I) Content-type: text/plain; charset=iso-8859-1 Content-transfer-encoding: 8BIT 2000 Jan 29 Hello, here is the summary of answers to my query about KCl clusters. Thanks very much to all who replied. I hope I included all responses. [#1--#9, below] E. Lewars ============== THE QUESTION 2000 Jan 18 Hello, A colleague in my Department has looked at the mass spectra of clusters consisting of one K+ surrounded by several KCl molecules (or ion-pairs, if you prefer), and clusters of one Cl- surounded by several KCl molecules (this latter with _negative_-ion mass spec). I think these can be thought of as K+ solvated with KCl, and Cl- solvated with KCl. He's found that some clusters are particularly prominent, as if there were certain magic numbers of stable KCl "solvent" molecules (a bit reminiscent of prominent clusters in Cn mixtures, where a big C60 peak led to the hypothesis of buckminsterfullerene). QUESTION: Is there a program that will handle from a few up to about 100 KCl molecules plus a K+ or Cl- ion? Please note that these clusters may not be simple periodic systems like a KCl crystal. I don't know if they can even be viewed as a doped KCl crystal. We are looking for a rationalization of the fact that clusters of certain sizes seem to be favored. Thanks E. Lewars ========== ========= THE ANSWERS #1 Dear E. Lewars, Regarding your question about clusters made up from ionic salts: I have been personally involved in the theoretical modelling of such clusters with an ab initio yet quite simple methodology. Rather than explaining myself in this mail, I think it is better if you have a look at our recent papers: "Theoretical Study of Small (NaI)n Clusters" Andres Aguado et al. J. Phys. Chem. B 101, 5944-5950 (1997). "Structural and Electronic Properties of Smalll Neutral (MgO)n Clusters" Eduardo de la Puente et al. Phys. Rev. B 56, 7606-7614 (1997). "Structure and Bonding in Small Neutral Alcali Halide Clusters" Andres Aguado et al. Phys. Rev. B 56, 15353-15360 (1997). "Ab initio Calculations of Structures and Stabilities of (NaI)nNa+ and (CsI)nCs+ Cluster Ions" Andres Aguado et al. Phys. Rev. B 58, 9972-9979 (1998). "Structures and Stabilities of Doubly-Charged (MgO)nMg2+ (n=1-29) Cluster Ions" Andres Aguado et al. J. Chem. Phys. 110, 4788-4796 (1999). The two last ones are perhaps the most interesting to you as the clusters considered have the composition with an extra ion you mention in your mail. In brief, the main aproximation of the model (that makes it faster than others electronic-structure methods) is the neglect of the polarization forces. This can be a quite reasonable assumption for alkali halides if you don't look for very thin details. However, it is not a good assumption at all for oxides, so in the last paper we included such terms semiempirically. The method we employ is called the PI method, and is useful for studying ionic materials, both crystals (pure or doped) and clusters. You'll find references to the original works with the PI model in our works. If the field of doped crystals is interesting to you, you may have a look to the following work: "Lattice Distortions around a Tl+ Impurity in NaI:Tl+ and CsI:Tl+ Scintillators: An ab initio study involving Large Active Clusters" Andres Aguado et al. Phys. Rev. B 58, 11964-11969 (1998). "Calculations of the Band Gap Energy and Study of Cross Luminescence in Alkaline-Earth Dihalide Crystals" Andres Aguado et al. J. Phys. Soc. Jpn. 68, 2829 (1999). Hope this helps. The method is quite able to manage with clusters of about 100 atoms if you do not pretend to perform molecular dynamics simulations, but just static calculations. To perform dynamical calculations, tou should turn to parameterized pair potential models, from which those of Paul Madden are the most accurate known to me up to date. Best regards, Andres Aguado. --Date: Tue, 18 Jan 2000 08:19:11 +0100 From: Andres Aguado ========= #2 This has already been done, at least for (NaCl)_n: cf. the Cambridge Cluster Database; in this case: http://brian.ch.cam.ac.uk/~jon/structures/NaCl.html >From a suitable plot of the cohesive energies given there, you should be able to get a first idea about "magic numbers", up to and including n=35. For programs that can handle this type of problem, given a suitable model potential, cf. the work of the CCD authors (Doye, Wales, et al.), or alternatively my work, cf. http://www.theochem.uni-stuttgart.de/~hartke If the above mentioned NaCl results are not yet sufficient for you, and if you want to go beyond n=35, I would be interested in a collaboration on this project. Just contact me at hartke \\at// theochem.uni-stuttgart.de In contrast to the CCD authors, I have succeeded in treating the global cluster geometry optimization problem also on the DFT and ab-initio levels, see the refs. given on my home page above (although, obviously, this can be done only for considerably smaller systems...). I know of only one other group attempting a similar approach, albeit not for geometry optimization but for MD: Novaro et al., J.Chem.Phys. 109 (1998) 2176. Bernd Hartke -- PD Dr. Bernd Hartke e-mail: hartke:~at~:theochem.uni-stuttgart.de Dep. of Theoretical Chemistry e-mail: bernd.hartke#* at *#rus.uni-stuttgart.de University of Stuttgart http://www.theochem.uni-stuttgart.de/~hartke Pfaffenwaldring 55 Phone: +49-711-685-4409 70569 Stuttgart FAX: +49-711-685-4442 GERMANY ============= #3 Just look at my web page, the address is below. On Jan 20, 12:07pm, Errol Lewars wrote: > Subject: Re: Clusters > Donald E. Williams wrote: > > > > Mpa/mpg can handle cluster energy minimization without assuming any symmetry. > > You have to provide the nonbonded energy functions. > > > > -- > > Dr. Donald E. Williams email:dew01:~at~:xray5.chem.louisville.edu > > Department of Chemistry http://www.louisville.edu/~dewill01 > > University of Louisville phone:502-852-5975 > > Louisville, KY 40292 fax: 502-852-8149 > > ==== > Hello, > > Thank you for your reply. I'll see if I can track down mpa/mpg, which i > guess are molecular mechanics-type (force field) programs. > > EL > ==== >-- End of excerpt from Errol Lewars -- Dr. Donald E. Williams email:dew01 \\at// xray5.chem.louisville.edu Department of Chemistry http://www.louisville.edu/~dewill01 University of Louisville phone:502-852-5975 Louisville, KY 40292 fax: 502-852-8149 ====== #4 Before spending a lot of time on this calculation, which would be interesting, and may I think already be in the literature, note that these magic clusters appear to be a kinetic rather than thermodynamic effect in the ms of alkali halide cluster ions, both +ve and -ve mass spectra. They are instrument geometry and size and accelerating voltage dependant in magnetic instruments --- i.e. deppend on time available for equilibration. Prof. Jack M. Miller, Associate Vice-President, Research and Dean of Graduate Studies, Professor of Chemistry, Brock University, St. Catharines, Ont., Canada, L2S 3A1. Phone (905) 688 5550, ext 3789 FAX (905) 682 2277 e-mail jmiller ^%at%^ brocku.ca http://chemiris.labs.brocku.ca/~chemweb/faculty/miller/ ===== #5 I cannot provide immediate information regarding calc-programs, but these magic numbers effects are well-known for metallic clusters (of comparable size). You might look up what these people are using to predict/explain them (e.g. for Na_n or K_n). -- Jochen Heinrich-Heine-Universität, Institut für Physikalische Chemie I Universitätsstr. 1, Geb. 26.43.02.29, 40225 Düsseldorf, Germany phone 02118113681 fax 02118115195 -- www-public.rz.uni-duesseldorf.de/~jochen Jochen#* at *#Uni-Duesseldorf.de -- Jochen.Kuepper#* at *#FernUni-Hagen.de -- Kuepper %-% at %-% ACM.org ========== #6 Errol, What kind of simulation did you have in mind? Classical with empirical interactions? Gas-phase? I simulated gas-phase KCl clusters classically using a Born-Mayer potential, and I imagine simply adding one more ion would be trivial. However, I don't know if this empirical potential would work with a net charge or not. I unfortunately cannot get my hands on my old papers about KCl clusters (both neutral and charged). If memory of the literature serves, there will be several types of "magic" clusters with rocksalt structure, such as 3x3x3 or 5x5x5 arrangements of K and Cl in a cubic shape (and many other possible rectangular geometries) similar to the 2x2x2 and 4x4x4 magic structures for neutral clusters. For anything over about 10 ions, rocksalt structures should always give you the most stable energy, even for singly-charged clusters. These experiments remind me of similar, but more complicated, simulations of (KCl)_n + e- using quantum path-integral methods to simulate the thermodynamic behavior of the free electron. Once again, rocksalt structures were often observed, with the electron occupying a "negative ion vacancy" (like an F-color center). Sometimes the electron was distributed on the surface, however. Let me know if these ideas are of interest to you! -- Keith --------------------------------------------------------------- Keith Ball, Dill Group Department of Pharmaceutical Chemistry University of California at San Francisco 3333 California Street, Suite 415 San Francisco, CA 94118-1944 phone: (415) 476-8910 fax: (415) 502-4222 e-mail: kdb-: at :-maxwell.ucsf.edu http://rainbow.uchicago.edu/~kdb1 --------------------------------------------------------------- ================ #7 Maybe I am saying something stupid, but Is it a solid cluster? If it was solid, and maybe even if not, it would be interesting if there is a structural model that could account for this, more than the number of molecules (like in the C60) the possible structures accompanying the number. At this respect, I would initially, advice, classical mechanic codes since this is are ionic ions parametrized in literature and the gain in speed would allow the calculation of many geometrical configurations in a reasonable time, i.e. 200 atoms it is not an issue for these codes. As well, as the use of some structure determination experimental techniques, neutron Diffraction, for instance. I hope it helps, regards Manuel ps Sorry if I say some nonsense, but you don't give much information in the mail, and I am still under the flu influence, :-) [All suggestions are very welcome--EL] ___________________________________________________________________ Manuel Melle Franco -Ph.D. Student- Dipartimento di Scienza dei Materiali Universita' degli Studi di Milano-Bicocca via Cozzi, 53 - 20125 Milano ITALY Tel. Office (+39)-02-6448 5232 Tel. Home (+39)-02-6610 2542 Fax. (+39)-02-6448 5403 ___________________________________________________________________ ======== #8 Dear Dr. Lewars, I think that you need to do statistical mechanical simulation for your system since there would be enormous number of isomers for such a large system. Depending upon the temperature of the experiment several isomers might have contributed to the specrtra ( for a similar problem in the case of Si-C clusters ( even with a very small system compared to yours) see, JPC A, 103, 1999, 6442). Pradipta -- ***************************************** * Dr. Pradipta Bandyopadhyay * * AMES LAB * * Department of Chemistry * * Iowa State Unievrsity * * Ames, IA 50011 * * USA * * e-mail: pradipta ":at:" si.fi.ameslab.gov * * Phone : 515-294-4604 (Lab) * * : 515-232-8067 (Residence) * * Fax : 515-294-0105 * * URL: http://www.msg.ameslab.gov/ * * Group/pradipta/index.html * ***************************************** ------------------------------------------------------------------------------ You may my glories and my state dispose, But not my griefs; still am I king of those. -- William Shakespeare, "Richard II" ------------------------------------------------------------------------------ #9 Dear Dr. Lewars, the mono-anionic Clusters are known. Have a look into T.P. Martin, Physics Reports 95, 167 (1983) However, there is currently a search for small di-anions, and so far the "smallest" detected species are BeF4(2-) and MgF4(2-). I would be very interested to know if your colleague sees any evidence for KCl3(2-) or K2Cl4(2-). Sincerely yours, Thomas Sommerfeld ------------------------------------------------------------------------------ Dr. Thomas Sommerfeld Tel: +49 6221 545264 Theoretische Chemie, Universitaet Heidelberg Fax: +49 6221 545221 Im Neuenheimer Feld 229, 69120 Heidelberg, Germany Email: Thomas.Sommerfeld ^%at%^ urz.uni-heidelberg.de ========= ==============