From owner-chemistry@ccl.net Fri Aug 30 08:03:01 2013 From: "Willy Offermans Willy%a%Offermans.Rompen.nl" To: CCL Subject: CCL: More than one imaginary frequencies Message-Id: <-49122-130830075552-23983-o5moCMx3dJUWM+94kI7o/g(!)server.ccl.net> X-Original-From: Willy Offermans Content-Disposition: inline Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=utf-8 Date: Fri, 30 Aug 2013 13:55:37 +0200 MIME-Version: 1.0 Sent to CCL by: Willy Offermans [Willy|,|Offermans.Rompen.nl] Dear colleagues, I gave it a second thought. Your TS is determined by a dE/dx = 0, and d2E/dxdy <0 for only one ``direction''. This is the direction along the RC. Your images along the RC are determined by dE/dx >< 0 for the direction along the RC and dE/dx = 0 for all others and d2E/dxdy <0 along the RC, if they lie above the inflection point of the RC and d2E/dxdy > 0 below the inflection point and d2E/dxdy = 0 at the inflection point of the RC. I hope this makes sense to you. So a force analysis and frequency analysis of the images along the RC can give you some information about the path. However this info is usually of less importance. On Thu, Aug 29, 2013 at 08:50:50AM -0400, Billy McCann thebillywayne[a]gmail.com wrote: > Ichraf, > > That you found a transition structure with one and only one negative > eigenvalue means you found a true TS. Running an IRC calculation will > connect the TS to the reactants and products. The reactants and products > themselves should have only positive eigenvalues. It is necessary to > validate these critical points using vibrational calculations in order to > verify their point on the potential energy surface (PES). > > However, the intermediate points along the PES which connect the critical > points will not yield those characteristic of critical points, because they > are not on a place on the PES where the derivative of the energy with > respect to geometry is zero. Therefore, these points will not have only > positive eigenvalues. Unless you use a more sophisticated method, the > vibrational calculations will yield no insight. > > The IRC is something of a constrained geometry optimization. Do not bother > yourself with frequency calculations on the intermediate points, as they > are not critical points. Typically, the IRC calculation is used to verify > that the transition structure you found is indeed the one which connects > your (expected) products and reactants. > > If you have verified your reactants, TS, and products as critical points, > and have verified that you've indeed found that TS that connects your > reactants and products, you've successfully solved your problem. > > Kind regards, > > Billy Wayne McCann, Ph.D. > Oak Ridge National Lab > Chemical Sciences Division > Chemical Separations Group > > > On Wed, Aug 28, 2013 at 10:32 AM, Close, David M. CLOSED=mail.etsu.edu < > owner-chemistry.^_^.ccl.net> wrote: > > > > > Sent to CCL by: "Close, David M." [CLOSED[*]mail.etsu.edu] > > Ichraf: > > You are right. But since you have more than on negative eigenvalue, you > > are not on a true TS. You are on a saddle point. You are however close to > > the answer. You have to cut out the coordinates and displace the geometry > > in the direction of the TS. Redo the calculation and you should arrive at > > the TS. There is a simple fortran program to perform this procedure. I > > can send it to you if you like. > > Regards, Dave Close. > > > > -----Original Message----- > > > From: owner-chemistry+closed==etsu.edu..ccl.net [mailto: > > owner-chemistry+closed==etsu.edu..ccl.net] On Behalf Of Ichraf Oueslati > > ichraf.oueslati(_)obspm.fr > > Sent: Wednesday, August 28, 2013 8:57 AM > > To: Close, David M. > > Subject: CCL: More than one imaginary frequencies > > > > > > Sent to CCL by: "Ichraf Oueslati" [ichraf.oueslati : obspm.fr] Hi all, > > > > In order to find a Minimum Energy Path, I run an IRC calculation with G09. > > When the frequencies of the points along the path were computed, I found > > more than one imaginary frequencies for each one. > > > > I already started IRC calculation with an optimized TS which has one and > > only imaginary frequency. For optimisation, I started by using B3LYP method > > then MP2 method with ts and noeigen keyword. > > This optimized TS connects to the reactant and product (I checked the > > geometries). I thought that the points along the MEP must have only > > positive frequencies. Am I wrong ?? > > > > Did the reaction path founded by IRC algorithm is correct? Why did I get > > those negative frequencies for the points along the MEP? what should I do > > to fix this problem?http://www.ccl.net/chemistry/sub_unsub.shtmlhttp://www.ccl.net/spammers.txt> > > > > -- Met vriendelijke groeten, With kind regards, Mit freundlichen Gruessen, Willy ************************************* Dr. W.K. Offermans CAT Fellow CAT Catalytic Center Institut für Technische und Makromolekulare Chemie RWTH Aachen Worringerweg 1, Raum 38C-150 D-52074 Aachen, Germany Phone: +49 241 80 28592 Fax: +49 241 80 22593 Home: +31 45 544 49 44 Mobile: +31 681 15 87 68 e-mail: Willy^_^Offermans.Rompen.nl e-mail: Willy.Offermans^_^CatalyticCenter.RWTH-Aachen.de From owner-chemistry@ccl.net Fri Aug 30 12:50:00 2013 From: "Victor Rosas Garcia rosas.victor]![gmail.com" To: CCL Subject: CCL: derivation of RESP charges? Message-Id: <-49123-130830124528-10298-J5h47sUmnVnqTVhA4U01hg . server.ccl.net> X-Original-From: Victor Rosas Garcia Content-Type: multipart/alternative; boundary=047d7b343884cc06a404e52cf211 Date: Fri, 30 Aug 2013 11:45:21 -0500 MIME-Version: 1.0 Sent to CCL by: Victor Rosas Garcia [rosas.victor[#]gmail.com] --047d7b343884cc06a404e52cf211 Content-Type: text/plain; charset=ISO-8859-1 Hello everybody, I am currently developing some molecular mechanics parameters for some large molecules. I have reason to believe that most of the atoms are well-parametrized except for a cyclopropane group that is part of a long chain. Regarding derivation of charges, I have calculated charges by RESP/B3LYP//cc-pVTZ and obtained the Bolzmann-averaged values using the RED program, but these calculations had to be done on a small substituted cyclopropane, because the original molecule is too big. Now I have a nice set of charges for my small molecule. Here comes the conundrum: If I just change the values by default to my new values for the charges, the total charge for the large molecule is about -0.044, not zero. I am thinking about applying a scaling factor to the new values, making negative charges slightly less negative and the positive values slightly more positive, but I am wondering if there are any recommendations about how to fix this. thanks for your attention Victor --047d7b343884cc06a404e52cf211 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
Hello everybody,

I am currently dev= eloping some molecular mechanics parameters for some large molecules.=A0 I = have reason to believe that most of the atoms are well-parametrized except = for a cyclopropane group that is part of a long chain.

Regarding derivation of charges, I have calculated charges by RESP/B3LY= P//cc-pVTZ and obtained the Bolzmann-averaged values using the RED program,= but these calculations had to be done on a small substituted cyclopropane,= because the original molecule is too big.=A0 Now I have a nice set of char= ges for my small molecule.=A0

Here comes the conundrum: If I just change the values by default to my = new values for the charges, the total charge for the large molecule is abou= t -0.044, not zero.=A0 I am thinking about applying a scaling factor to the= new values, making negative charges slightly less negative and the positiv= e values slightly more positive, but I am wondering if there are any recomm= endations about how to fix this.

thanks for your attention

Victor
=
--047d7b343884cc06a404e52cf211-- From owner-chemistry@ccl.net Fri Aug 30 14:01:00 2013 From: "Prof. Curt Breneman brenec^_^rpi.edu" To: CCL Subject: CCL: derivation of RESP charges? Message-Id: <-49124-130830135908-3083-RCJJW6YIH2T2Zco3kzQUeA[#]server.ccl.net> X-Original-From: "Prof. Curt Breneman" Content-Language: en-us Content-Type: multipart/alternative; boundary="----=_NextPart_000_05B6_01CEA589.1C579ED0" Date: Fri, 30 Aug 2013 13:59:14 -0400 MIME-Version: 1.0 Sent to CCL by: "Prof. Curt Breneman" [brenec{:}rpi.edu] This is a multipart message in MIME format. ------=_NextPart_000_05B6_01CEA589.1C579ED0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Don't do it that way! There is something a bit fishy when you get those sorts of results, so you should examine your goals, and what you'd like to use the charges to accomplish. Generally, RESP is designed to constrain charges to integers (or zero) for each residue in a protein or other biopolymer. Cheers, Curt Breneman Professor and Head RPI Chemistry & Chemical Biology > From: owner-chemistry+brenec==rpi.edu{=}ccl.net [mailto:owner-chemistry+brenec==rpi.edu{=}ccl.net] On Behalf Of Victor Rosas Garcia rosas.victor]![gmail.com Sent: Friday, August 30, 2013 12:45 PM To: Breneman, Curt Subject: CCL: derivation of RESP charges? Hello everybody, I am currently developing some molecular mechanics parameters for some large molecules. I have reason to believe that most of the atoms are well-parametrized except for a cyclopropane group that is part of a long chain. Regarding derivation of charges, I have calculated charges by RESP/B3LYP//cc-pVTZ and obtained the Bolzmann-averaged values using the RED program, but these calculations had to be done on a small substituted cyclopropane, because the original molecule is too big. Now I have a nice set of charges for my small molecule. Here comes the conundrum: If I just change the values by default to my new values for the charges, the total charge for the large molecule is about -0.044, not zero. I am thinking about applying a scaling factor to the new values, making negative charges slightly less negative and the positive values slightly more positive, but I am wondering if there are any recommendations about how to fix this. thanks for your attention Victor ------=_NextPart_000_05B6_01CEA589.1C579ED0 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Don’t do it that way!  There is something a bit fishy when = you get those sorts of results, so you should examine your goals, and = what you’d like to use the charges to accomplish.  Generally, = RESP is designed to constrain charges to integers (or zero) for each = residue in a protein or other biopolymer.

 

Cheers,

Curt Breneman

Professor and Head

RPI Chemistry & Chemical Biology

 

 

From:= = owner-chemistry+brenec=3D=3Drpi.edu{=}ccl.net = [mailto:owner-chemistry+brenec=3D=3Drpi.edu{=}ccl.net] On Behalf Of = Victor Rosas Garcia rosas.victor]![gmail.com
Sent: Friday, = August 30, 2013 12:45 PM
To: Breneman, Curt =
Subject: CCL: derivation of RESP = charges?

 

Hello = everybody,

I am currently developing some molecular = mechanics parameters for some large molecules.  I have reason to = believe that most of the atoms are well-parametrized except for a = cyclopropane group that is part of a long chain.

Regarding = derivation of charges, I have calculated charges by RESP/B3LYP//cc-pVTZ = and obtained the Bolzmann-averaged values using the RED program, but = these calculations had to be done on a small substituted cyclopropane, = because the original molecule is too big.  Now I have a nice set of = charges for my small molecule. 

Here comes the conundrum: = If I just change the values by default to my new values for the charges, = the total charge for the large molecule is about -0.044, not zero.  = I am thinking about applying a scaling factor to the new values, making = negative charges slightly less negative and the positive values slightly = more positive, but I am wondering if there are any recommendations about = how to fix this.

thanks for your = attention

Victor

------=_NextPart_000_05B6_01CEA589.1C579ED0--