keto/enolate TS structure in Solvation model

Dear CCL users:
 I am playing around with a keto/enolate formation reaction catalyzed by
 gernal acid and base(methyl amine/ammonium ion). I would like to locate
 the TS structure in organic solvent such as THF or cyclohexane based on
 the TS structure previously located in gas phase. Despite semi-empirical
 method AM1 in spartan was able to locate the initial strucuture in
 solvation, but further optimization in Gaussian was unsuccesful and
 many times.(Gaussian program complain that there is no negative engien
 while desired one)I am not sure if the initial structure located in
 spartan was
 not good enough or there was some trap in refinement in Gaussian.Can
 with experiences of this kind of reaction in solvation give me any
 Your answer will be highly appreciated! Have a good one.
 The keyword I am using in SCRF caculation
 #HF/6-31G(d) OPT=(Tight,TS) SCRF=(PCM,Read,Solvent=THF) Freq Test
 I also attached the PDB file of the TS in gas phase (located in
 and the TS in hexadacane and water.( located in AM1 ) if you are
 Yong. L. Y.
 Department of Chem.
 SUNY-Stony Brook

Attachment: 12_01_SE_hexa_output.pdb
Description: Protein Databank data

Attachment: 12_01_SE_water_output.pdb
Description: Protein Databank data

Attachment: 11_27_enolate_output_01.PDB
Description: Protein Databank data