How to prevent wrong valencies during reaction?

Hi,


I have a simple condensation reaction (some benzene rings react to anthracene etc). However also a lot of molecules with wrong valencies (5-bonding C) are created. Is there a way to prevent that?





The reaction is:

Code:

C1=CC=CC=C1.C1=CC=CC=C1.C1=CC=CC=C1>>C1=CC2=C(C=C1)C3=CC=CC=C3C=C2.C1=CC2=CC3=CC=CC=C3C=C2C=C1.C1=CC2=C(C=C1)C3=C(C=CC=C3)C4=CC=CC=C24

My reaction library consists of 5 example starters.


Some of the wrong smiles are attached as example.


I am not sure if the whole setup is very clever, but it would be nice


to get rid of the 5 binding carbons and many other molecules are OK.





Thank you


Tobias

You can avoid those problems by writing reaction scheme properly. Anthracene (14 atoms) cannot be not formed from three benzene rings (18 atoms).





Please consider, that in reactions, bonds break or form. I think, that it helps if you:





- think about the mechanism of your reaction





- check, that your reaction schema has the same number of atoms on both sides of the arrow (not always required, but useful)





- map the reaction scheme, it marks the corresonding atoms an helps to recognize problems in the scheme





- include and map the required hydrogens as well explicitly in the scheme





(If you cannot solve this reaction, we can discuss it in webex)

Hello Gyuri,


there is so much potential in Reactor and so many things possible its sometimes just overwhelming.





You are also right it helps to think about the reaction and the whole setup.


I will try my best. I think with the common reactions I am totally fine. I also like to tinker a little bit


hoping that ignorance / mild imbecility leads to new exciting results :-).





Cheers


Tobias

Hi,


just some other topic, I noticed that in the old Reactor, I am running version 3.2.11 there is a noticeable lag or "thinking process" during the compound creation, where the CPU is not running fullspeed. This is under WINXP and JAVA 1.5.


This is not disk bound (its running on a RAMDISK and I checked


with FileMon there is no read/write activity).





So the timer counts (estimated time left), but the CPU is not


working fullspeed its up and down, which is really weird and usually comes from some "internal problems" like wrong scheduling or wrong threading or overflow errors. Reaction


is the reaction above.





This is usually not observed in other reaction schemata.





Tobias

Hi Tobias,





I can not reproduce the error. Anyway, I suggest you upgrade to a newer version of JChem (5.0.7 or 5.1).





Zsolt