From owner-chemistry@ccl.net Tue Oct 21 07:37:01 2014 From: "Sebastian Kozuch seb.kozuch ~~ gmail.com" To: CCL Subject: CCL: Computer cluster Message-Id: <-50620-141021073447-4446-tcLXl69FSuu0sw9l+6dd8A..server.ccl.net> X-Original-From: Sebastian Kozuch Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=utf-8; format=flowed Date: Tue, 21 Oct 2014 14:34:35 +0300 MIME-Version: 1.0 Sent to CCL by: Sebastian Kozuch [seb.kozuch__gmail.com] Dear colleagues, As many of you have done several times, I need to buy a computer cluster, and I was hoping to get some tips from experienced people. The idea is to run mostly DFT jobs on rather big molecules, with some heavy wave function calculations for benchmark analysis. I am probably going for 2x12 core nodes, with 128GB RAM and SSD hard disks (~800 GB), which I think is good for DFT jobs, but the memory and scratch will probably be too small for couple cluster calculations. What would you do, obviously considering that the budget is far from infinite? Any advice will be appreciated. Sincerely, Sebastian From owner-chemistry@ccl.net Tue Oct 21 11:38:00 2014 From: "Simon Cross simon.{=}.moldiscovery.com" To: CCL Subject: CCL: SSC14 - Scandinavian Symposium on Chemometrics Message-Id: <-50621-141021093316-7794-aNFWrOSxLbTqfvTpwaXihg{=}server.ccl.net> X-Original-From: Simon Cross Content-Type: multipart/alternative; boundary="------------070907040306070300080302" Date: Tue, 21 Oct 2014 15:33:06 +0200 MIME-Version: 1.0 Sent to CCL by: Simon Cross [simon/./moldiscovery.com] This is a multi-part message in MIME format. --------------070907040306070300080302 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Dear colleagues, The SSC14 Organizing Committee cordially invites you to attend the 14th Scandinavian Symposium on Chemometrics that will be held in Sardinia (Italy) on *June **14th-17th, 2015*. The SSC series are major scientific events that have been organized in different cities in Northern Europe since 1988 (Lappeenranta, Finland). They have a fascinating scope covering rational approaches in chemometrics, cheminformatics, design, omics and related sciences and worldwide attendance from both academia and industry. In the SSC14 various practical applications of chemometric methods in chemistry, biochemistry and chemical technology fields will be presented, using classical and newly developed algorithms. Therefore the symposium will touch upon the development and applications of chemometrics methods in Pharmaceutical, Cheminformatics, Omics, Process control, Analysis of chemical processes, Optimization and Food technology. Besides the scientific aspects of the program you will have the opportunity to visit some of the most charming places of the Mediterranean in a wonderful resort close to the sea. It is our intention to make this Symposium a memorable event for the attendees, both scientifically and socially. Please visit _http://www.ssc14.org/_ for further information about the venue, the main topics and the important deadlines. We look forward to welcoming you in Sardinia! With best regards, the organizing committee _ _ _http://www.ssc14.org _ --------------070907040306070300080302 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit

Dear colleagues,

The SSC14 Organizing Committee cordially invites you to attend the 14th Scandinavian Symposium on Chemometrics that will be held in Sardinia (Italy) on June 14th-17th, 2015.  

The SSC series are major scientific events that have been organized in different cities in Northern Europe since 1988 (Lappeenranta, Finland). They have a fascinating scope covering rational approaches in chemometrics, cheminformatics, design, omics and related sciences  and worldwide attendance from both academia and industry.

In the SSC14 various practical applications of chemometric methods in chemistry, biochemistry and chemical technology fields will be presented, using classical and newly developed algorithms. Therefore the symposium will touch upon the development and applications of chemometrics methods in Pharmaceutical, Cheminformatics, Omics, Process control, Analysis of chemical processes, Optimization and Food technology.

Besides the scientific aspects of the program you will have the opportunity to visit some of the most charming places of the Mediterranean in a wonderful resort close to the sea. It is our intention to make this Symposium a memorable event  for the attendees, both scientifically and socially.

Please visit http://www.ssc14.org/ for further information about the venue, the main topics and the important deadlines. We look forward to welcoming you in Sardinia!


With best regards, the organizing committee


http://www.ssc14.org

--------------070907040306070300080302-- From owner-chemistry@ccl.net Tue Oct 21 15:22:01 2014 From: "Zork Zou zorkzou.]*[.gmail.com" To: CCL Subject: CCL:G: Isotope effects: is your thermochemistry calculation correct? Message-Id: <-50622-141021151743-18601-zQuMaoQ4r5zjGyl+xG/jRA]*[server.ccl.net> X-Original-From: Zork Zou Content-Type: multipart/alternative; boundary=089e015369243023cf0505f3afea Date: Tue, 21 Oct 2014 14:17:37 -0500 MIME-Version: 1.0 Sent to CCL by: Zork Zou [zorkzou^^^gmail.com] --089e015369243023cf0505f3afea Content-Type: text/plain; charset=UTF-8 In a quantum chemistry calculation, there may be two related symmetries: the symmetry of nuclear charges (Sz) and the symmetry of atomic nuclei (Sn). The former doesn't depend on atomic masses whereas the latter does. Usually the two symmetries are the same, but in the case of isotope substitution they can be different. For example, in H2O, both of them are C2v, but in HDO they are C2v and Cs, respectively. Symmetry may affect thermochemistry results significantly through the rotational entropy (S_rot): S_rot = 177.671 eu/mole + R * ln(T) + R * ln(I/s) ... (linear molecule) or S_rot = 267.643 eu/mole + (3/2) * R * ln(T) + R * ln[sqrt(I1 * I2 * I3)/s] ... (nonlinear molecule) where "I" is the moment of inertia, and "s" is the symmetry number, which "is defined as the number of indistinguishable positions into which the molecule can be turned by simple rigid rotations". (see M. A. Paul, Principles of chemical thermodynamics, 1951, p. 663) Obviously, not only "I", but also "s" depends on the isotopes. However, by checking the thermochemistry results of most quantum chemistry programs including Gaussian, I found a wrong symmetry (Sz) being used for "s" by all of them, so "s" is independent of isotopes by mistake. In the case of isotopic substitution in high-symmetric molecule, this may lead to big errors in the Gibbs free energy. For example, s=60 for fullerene C60 (Sz = Sn = Ih) and s=1 if one carbon atom is replaced by C14 (Sz = Ih and Sn = C1). At the PM6 level of theory and at 1 atm and 298.15 K, the thermal corrections to Gibbs free energy are 0.351192 Hartree (s=60) and 0.347326 Hartree (s=1), that is, the error by using Ih symmetry is about 2.4 kcal/mol !!! In the example above HDO, the error in Gibbs free energy is about 0.4 kcal/mol. So, be careful with your thermochemistry calculations with isotopic substitution. --089e015369243023cf0505f3afea Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
In a quantum chemistry calculation, there may be two = related symmetries: the symmetry of nuclear charges (Sz) and the symmetry o= f atomic nuclei (Sn). The former doesn't depend on atomic masses wherea= s the latter does. Usually the two symmetries are the same, but in the case= of isotope substitution they can be different. For example, in H2O, both o= f them are C2v, but in HDO they are C2v and Cs, respectively.
Symmetry may affect thermochemistry results significantly throu= gh the rotational entropy (S_rot):

S_rot =3D 177.6= 71 eu/mole + R * ln(T) + R * ln(I/s) ... (linear molecule)

or

S_rot =3D 267.643 eu/mole + (3/2) * R = * ln(T) + R * ln[sqrt(I1 * I2 * I3)/s] ... (nonlinear molecule)
<= br>
where "I" is the moment of inertia, and "s&quo= t; is the symmetry number, which "is defined as the number of indistin= guishable positions into which the molecule can be turned by simple rigid r= otations". (see M. A. Paul, Principles of chemical thermodynamics, 195= 1, p. 663)

Obviously, not only "I", but = also "s" depends on the isotopes. However, by checking the thermo= chemistry results of most quantum chemistry programs including Gaussian, I = found a wrong symmetry (Sz) being used for "s" by all of them, so= "s" is independent of isotopes by mistake. In the case of isotop= ic substitution in high-symmetric molecule, this may lead to big errors in = the=C2=A0Gibbs=C2=A0free energy. For example, s=3D60 for fullerene C60 (Sz = =3D Sn =3D Ih) and s=3D1 if one carbon atom is replaced by C14 (Sz =3D Ih a= nd Sn =3D C1). At the PM6 level of theory and at 1 atm and 298.15 K, the th= ermal corrections to Gibbs free energy are 0.351192 Hartree (s=3D60) and 0.= 347326 Hartree (s=3D1), that is, the error by using Ih symmetry is about 2.= 4 kcal/mol !!!

In the example above HDO, the error= in Gibbs free energy is about 0.4 kcal/mol.

So, b= e careful with your thermochemistry calculations with isotopic substitution= .
--089e015369243023cf0505f3afea--